Seppo Heikkinen

The Christianisation of Latin Metre

A Study of Bedes De arte metrica















Academic dissertation to be publicly discussed, by due permission of the Faculty of Arts at
the University of Helsinki in auditorium M1, on the 21
st
of March, 2012, at 12 oclock.










































ISBN 978-952-10-7808-8 (paperback)
ISBN 978-952-10-7807-1 (PDF)

Unigrafia
Helsinki 2012

Contents

1. Introduction 1
1.1. General observations 1
1.2. Grammar and metre in Anglo-Saxon England 2
1.3. The dating of Bedes De arte metrica 5
1.4. The role of metrics in Bedes curriculum 9
1.5. The structure and aims of De arte metrica 11
1.6. Bedes Christian agenda and its implementation in his discussion of
metrics 13



2. Hexameter verse and general prosody 17
2.1. The dactylic hexameter in Anglo-Saxon England 17
2.2. Classical and post-classical prosody: common syllables 24
2.2.1. Plosives with liquids 26
2.2.2. S groups 29
2.2.3. Productio ob caesuram and consonantal h 34
2.2.4. Hiatus and correption 40
2.2.5. Hic and hoc 44
2.2.6. Summary 47
2.3. Other observations on prosody 49
2.4. The structure of the dactylic metres 58
2.4.1. The dactylic hexameter 58
2.4.2. The elegiac couplet 72
2.5. The aesthetics of verse 75
2.5.1. Enjambment 77
2.5.2. Bede on word order 81
2.5.2.1. The golden line 81
2.5.2.2. Other observations on word order 92
2.6. Word division and caesurae 97
2.7. Elision and hiatus 108
2.8. Bede on prosodic licences 118
2.9. The differences between pre-Christian and Christian poets 126
2.10. Conclusion 134

3. The lyric metres 137
3.1. The phalaecean hendecasyllable 141
3.2. The sapphic stanza 147
3.3. The terentianean metre 152
3.4. The anacreontic metre 158
3.5. The iambic trimeter 164
3.6. The iambic dimeter 168
3.7. The trochaic septenarius 175
3.8. Conclusion 185

4. Rhythmic verse 187
4.1. Introduction 187
4.2. The evolution of prosodic terminology in late antiquity 187

4.2.1. From mechanical to vocal ictus 187
4.2.2. The redefinition of rhythm 190
4.3. The origins of rhythmic poetry 192
4.4. Bedes definition of rhythm 195
4.5. Conclusion 204

5. Summary 207

Bibliography 211

Acknowledgements

I am deeply indebted to several people for their support in the completion of this thesis.
Firstly, I must thank my supervisors Heikki Solin and Matti Kilpi for their insights,
technical and cultural alike. They are both humanists in the true sense of the word and I
enjoyed our sessions more than I could ever have anticipated. Many of my other colleagues
have given me their time, help and inspiration. They include Anneli Luhtala, Olli Salomies,
Paavo Castrn, Leena Pietil-Castrn, Martti Leiwo, Maarit Kaimio and Reijo Pitkranta; this
list, of course, is far from exhaustive. I would also like to thank my pre-examiners Gerd
Haverling and J yri Vaahtera whose suggestions improved my thesis immeasurably. Special
thanks are due to my sister Kanerva Blair-Heikkinen and her husband John Blair for their
never-flagging belief in my work as well as their insights into the Anglo-Saxon world. John
also kindly read my whole draft and corrected my English. All the remaining mistakes are, of
course, mine. I could not have accomplished what I have without the help and support of my
parents Olavi and Katri Heikkinen and my aunt Kirsti Hmlinen. I am grateful for their
having been so endlessly supportive and gently pushing me at times when I was not inclined
to push myself. My research was financially supported by the Reinhold Ekholm Foundation,
the Research Foundation of the University of Helsinki and the Emil Aaltonen Foundation.
Lastly, I must thank my fellow musicians, dead and alive, professional and amateur alike, for
their inspiration. The art and scholarship of Bede may seem only tenuously connected to
those of Bach or, say, Shostakovich, but they were all dedicated to the coordination of sound
with time, to our pleasure and, ultimately, improvement.

Helsinki, 24 February 2012
Seppo Heikkinen






1

1. Introduction

1. 1. General observations

Bedes De arte metrica is, in many respects, a revolutionary work: it is one of the first
metrical treatises composed for an audience who were not native speakers of Latin. At the
same time it is the first practical presentation of metrics intended for the purposes of the
medieval monastic curriculum. Its role in the transmission of the classical poetic tradition
cannot be underestimated; however, its departures from what we consider the classical norm
of quantitative metrics also exerted a strong influence on medieval poetry. Despite the
background of the work, it turned out to be surprisingly long-lived, and its popularity can be
said to have outlasted the curriculum for which it was composed. De arte metrica remained in
use throughout the Middle Ages and, although its influence began to wane somewhat after
the cultural revolution of the twelfth century, it was not really abandoned: it is telling that its
first printed edition appeared shortly after the invention of the printing press, and the work
was circulated, even in vernacular translations, as late as the sixteenth century.
1

The modern age, however, has been less kind to Bedes writings on grammar.
Without doubt, this is largely due to Bedes prominent role as exegete and historian, which
has often eclipsed his minor grammatical works, but also to their hitherto inaccurate dating.
Until quite recently, scholars viewed Bedes works on orthography (De orthographia), metre
(De arte metrica) and schemes and tropes (De schematibus et tropis) as early works; the
idea seems to have been that Bede honed his skills in what is considered a preliminary field
of scholarship before moving on to such grander things as biblical exegesis. Ostensibly, the
early dating of Bedes grammatical writings corroborated their place as a kind of juvenilia
within his oeuvre. Yet another factor which has contributed to the underrated status of Bedes
grammatical writings has been the role of grammar itself, as it was perceived in antiquity:
grammar, which formed a part of the trivium, was literally a trivial subject; that Bede,
representing an entirely different educational system with a curriculum of its own, did not
share this view has not always been recognised or appreciated, although it has a direct
bearing on the role and prominence of these works within his literary production. The
scrutiny to which these works have been subjected has, inevitably, suffered from the
disdainful view which scholars sometimes take of Roman grammarians. Admittedly, Late

1
Avalle 1992, 400-401.
2

Latin grammarians were not always the most original of thinkers: in their work we often
encounter much that is recycled from earlier sources with little criticism or imagination. A
reader acquainted with the works of Bedes predecessors would be perfectly justified in
failing to be impressed by his first glance at Bedes De arte metrica, as it, superficially,
seems to fit into this mould: much in it is derived from earlier sources, and even verbatim
quotations from other grammarians are conspicuously frequent. It is only when one looks
closer that one sees the true individuality of Bedes views on metre. Bede rarely goes so far
as to refute openly anything in his sources, but through subtle manipulation of wording, as
well as careful elimination and introduction of material, he has often managed to alter the
traditional presentations of metrical rules in ways that often amount to virtual redefinitions. In
addition to the individual revisions of his material, Bede can, with some justice, be regarded
as the creator of a literary genre: his De arte metrica presents syllable prosody and the poetic
metres as a unified system rather than as separate subjects, and his presentation served as the
primary model for the subsequent artes metricae of the Carolingian era and the high middle
ages.
2



1. 2. Grammar and metre in Anglo-Saxon England

The relationship between Christianity and grammatical scholarship had been an
uncomfortable one in late antiquity, as the study of grammar was seen as inextricably linked
to the study of pagan literature. Vivien Law (1997, 74) cites several historical anecdotes
which illustrate this sad state of things, the best-known of which is undoubtedly J eromes
vision where he was accused of being Ciceronianus, non Christianus.
3
As Christianity
gained the upper hand, it nevertheless required an ever-growing number of educated young
men for its offices. The school system of the Late Empire, however, was still largely based on
the old, pagan, tradition of the artes liberales, to which the church had to adjust, and as the
socio-economic structure of the empire collapsed, the Christians ultimately had to take over
education. In 529, the Council of Vaison instructed parish priests to care for the education of
youths in schools that were founded as adjuncts of the diocesan system. Although nominally
Christianised, the episcopal and diocesan schools of late antiquity did not differ markedly

2
Leonhardt 1989, 75. However, Leonhardt, as Manitius (1911, 74) before him, underestimates the actual
content of Bedes treatise (p. 75: Dabei bringt Beda, wie bereits Manitius zu Recht bemerkt hat, nichts, das
nicht auch bei den antiken Grammatikern zu finden wre.).
3
Hier. epist. 22, 30.
3

from their predecessors. The education they provided was still loosely modelled on that of the
traditional schools, albeit cosmetically adapted to the needs of a Christian society.
4

Another approach to the question of education and Christianity was taken in the
convents of late antiquity. The monastic schools were dedicated to askesis, or conversatio,
the Christian life, and their teaching was fully geared to its implementation. They renounced
many, if not most aspects of the traditional school system, being, in C. W. Joness words,
apprenticed, vocational and democratic;
5
even the concept of liberal education was
rejected, implying, as it did, a society of masters and servants. The curriculum was revised to
contain only that which advanced the vocational needs of the monks. The study of Scripture
was an integral element of the spiritual life, and the teaching of letters was moulded in its
service. This is the tradition that ultimately became the model for the insular monastic
schools, which, in turn served as the foundation for the Carolingian system of education.
Nevertheless, the relationship between letters and Christianity was somewhat
different on the British Isles, where the Latin language was introduced together with
Christianity; subsequently, grammar was, from its arrival, permeated with the Christian
religion.
6
As Latin was the language of the church, instruction in basic grammar was of the
essence. This was, in itself, a subject of no little difficulty, as the existing grammatical
literature had been composed for native speakers of the Latin language. Trying to learn Latin
from the writings of Donatus or Priscian would have been an impossible task, as they provide
no paradigms or any other such features as we associate with books of elementary grammar.
7

Furthermore, although many of the Late Roman grammarians had been Christian, this was
hardly reflected in their works, as they had been composed in the spirit of the earlier
grammatical tradition. Christian authors had made some effort to Christianise the subject of
grammar, mainly through the occasional introduction of J udaeo-Christian material as
substitutes for the usual classical quotations, but this was still largely superficial. These
features of the continental tradition necessitated the alteration of grammatical teaching in
such a way that it would better serve a) the educational needs of a non-native user of Latin
and b) the vocational needs of the monastic system. The latter meant a genuine need for a
thorough Christianisation of the subject, to which the insular grammarians generally reacted
with prefaces excerpted from the Early Fathers, word-lists revised to contain more
ecclesiastical vocabulary, and biblical quotations inserted in place of quotations from the

4
J ones 1975, v.
5
J ones 1975, vi.
6
Law 1984, 82.
7
Law 1997, 75.
4

classics;
8
all of these are features which we can encounter in Bedes grammatical works and
which his contemporaries would have taken as a matter of course.
Although the present thesis deals not with a work of elementary grammar but
with a treatise on metre, the peculiar linguistic conditions of Anglo-Saxon England
nevertheless played a role which cannot be ignored. The rules of quantitative syllable prosody
were a cause of considerable bewilderment for the Anglo-Saxons, as the phenomenon had
disappeared from the spoken Latin of their day and could only laboriously be garnered from
books. Admittedly, this is a problem they had partly inherited from their predecessors: we
may call to mind Augustines confession that he knew nothing of syllable quantity,
9
as well
as the number of grammatical works devoted to the final syllables of words, where the
discrepancy between the spoken Latin of late antiquity and previous poetic practice was the
most blatant. The main drawback of the late antique sources on prosody and the poetic metres
was that they discussed them as separate subjects. The Anglo-Saxon poet-scholars Aldhelm
and Bede transformed the genre of metrical treatises by incorporating syllable prosody into
their discussion of verse technique. Even the title of Bedes treatise on metre, De arte
metrica, is the first of its kind, and sets it apart from its predecessors (generally with titles
along the lines of De metris):
10
it sets out to portray the art of verse composition as a whole,
ranging from its smallest components, sounds and syllables, to the broader issues of poetic
style. Bedes examples of the basic elements of quantitative verse are almost invariably
drawn from hexameter poetry, preferably that of the Christian poets, and this, conversely,
means that the Christian authors who set the norm for good poetic style also became
authorities on questions of general prosody. Hence, we can see that in Bedes treatise, the
traditional presentations of the grammarians often come second to the authors own
observations on Christian verse.

8
Law 1984, 82.
9
Aug. mus. 3, 2.
10
Leonhardt 1989, 77 n. 26.
5

1. 3. The dating of Bedes De arte metrica

The traditional consensus has, until recently, been that Bedes De arte metrica and its
companion work De schematibus et tropis are early works.
11
This misconception has been
founded not only on the idea that they somehow show less maturity and sophistication than
his major works, but also on an ambiguous turn of phrase which appears in the epilogue of
De arte metrica: Bede addresses the work to one Cuthbert, whom, as he has previously been
interpreted, he calls his beloved son and fellow deacon. The passage forms a preface to the
following De schematibus et tropis, functioning as a bridge between the two libelli. Bede
writes is as follows:

Haec tibi, dulcissime fili et conlevita Cuthberte, diligenter ex antiquorum opusculis
scriptorum excerpere curavi, et quae sparsim reperta ipse diuturno labore collegeram
tibi collecta obtuli, ut quemadmodum in divinis litteris statutisque ecclesiasticis
imbuere studui, ita et in metrica arte, quae divinis non est incognita libris, te solerter
instruerem. Cui etiam de figuris vel modis locutionum, quae a Graecis schemata vel
tropi dicuntur, parvum subicere libellum non incongruum duxi, tuamque dilectionem
sedulus exoro ut lectioni operam inpendas illarum maxime litterarum, in quibus nos
vitam habere credimus sempiternam.
12

[I have taken pains to make these extracts from the handbooks of the ancient scholars
for your benefit, beloved son and fellow deacon, Cuthbert, and I have offered to you
this collection of poetic excerpts, which, as I came upon them here and there, I
laboriously gathered over a long period of time in order that I might instruct you
intelligently in the art of metrics, which is not unknown in the Bible, just as I
endeavoured to give you your first training in divine letters and ecclesiastical law. To
complement it I have also drawn up a little work on the figures and mannerisms of
speech which the Greeks call schemes and tropes, and I earnestly beg you to devote
your efforts especially to the informed reading of that Book in which we believe that
we have eternal life.]
13


The passage is interesting and illuminating in many respects: firstly, Bede explains the extent
to which the works are based on his own research, and the statement that they have been
compiled over a long period of time would make a very early date improbable.
Furthermore, the closing sentence clearly demonstrates the extent to which Bede regarded the

11
Laistner 1957, 95: Bedes earliest treatises De arte metrica, De schematibus et tropis, De orthographia
were intended for school use. They prove that he was brought up on, and, when he became himself a teacher,
adapted and excerpted such writers as Donatus, Charisius, Audax, Caper, and other grammarians of the later
Roman imperial age. Also Palmer (1959, 573), who categorically refutes the immaturity of De arte metrica
and De schematibus et tropis, nevertheless refers to them as Bedes earliest works. Also Blair 1970, 5; 249-
250; Brunhlzl 1975, 201; Leonhardt 1989, 75.
12
Bede, De arte metrica, ed. Kendall 1975 (hereafter DAM) 25, 36.
13
Kendall 1991, 167.
6

study of grammar as inseparable from the study of scripture: by his own admission, Bede did
not view it as a subject that was merely ancillary to more serious scholarly pursuits.
However, what particularly intrigues us is the term conlevita (which Kendall has, in accord
with previous tradition, translated as fellow deacon). The origins of the term, apparently of
Bedes coinage, are in the practice, widespread in the churches of the East and the West alike,
of equating the Christian ecclesiastical hierarchy of deacon, presbyter and bishop with the
Jewish one of Levite, priest and high priest.
14
Thus, levita was commonly used as a synonym
for diaconus. Whether Bede here means that both he and his addressee were deacons at the
time of the works composition is, however, another matter.
The Cuthbert whom Bede addresses is generally held to be his long-time pupil,
who later, in 735, became the abbot of Wearmouth and Jarrow and to whom the famous
description of Bedes death (De obitu Bedae) is also attributed. The double expression
beloved son and fellow deacon is problematic, as the former term seems to imply that the
recipient was considerably Bedes junior, whereas the latter one would mean that they were
both roughly of the same age. On the assumption that Bede was a deacon at the time of the
works completion, C. Plummer placed the date of the works composition between 691 or
692, the time of Bedes ordination as deacon, and 701 or 702, his ordination as priest.
15
To
render the expression beloved son explicable, M. L. W. Laistner further refined this dating
by estimating that it was composed immediately before Bedes priesthood.
16
This dating was
still accepted at face value by C. B. Kendall in his 1975 editions of De arte metrica and De
schematibus et tropis: in his introduction he simply states: DAM and DST were among the
earliest of Bedes numerous works. He was still a deacon (levita) when he composed them,
very probably in 701 or 702.
17

The problem with these dates is that they focus solely on the supposition that
Bede was a deacon at the time of the works completion while ignoring the probable age of
the recipient who certainly was one.
18
Curiously enough, the first person to challenge this
dating was Charles W. J ones in his preface to the volume of Bedes didactic works which
contained Kendalls edition in other words, in the same volume we encounter two opposite

14
Dassmann 1970, 198-214.
15
Plummer 1896, I, cxlv.
16
Laistner & King 1943, 131-132.
17
Kendall 1975, 74.
18
C. V. Franklin has expressed this problem amusingly: If, as it is thought, the Cuthbert of the dedication is the
future abbot of Wearmouth-J arrow who died in 777, a dating at the beginning of the century would make him a
very precocious student indeed. Franklin 2000, 200.
7

views on the date of the work.
19
Thereafter, the new dating for Bedes De arte metrica won
an increasing number of followers, the most prominent contributors being Martin Irvine
20
and
George H. Brown.
21
In his 1991 edition of De arte metrica, even Kendall admitted that the
final version of the book may have been completed at a later date, and attributes the term
conlevita to Bedes use of his old lecture notes.
22
Arthur Holder, however, has pointed out
in his 1999 article that Bedes use of conlevita should in no way be taken to imply that he
was himself a deacon at the time of the works composition, referring to the use of the
similarly prefixed words condiaconus, conpresbyter and conepiscopus in Augustine.
23

Conlevita would, in his opinion, simply mean deacon who is my colleague in the ministry of
Christ.
24

New light on the probable dating of De arte metrica has been shed in C. V.
Franklins 2000 article which points out an exegetical interpretation which occurs in De
schematibus et tropis. Based on what we know about the dating of Bedes exegetical works,
this interpretation would, indeed, seem to give Bedes twin work a late date. The passage is in
Bedes exposition of syllempsis in sensu, a device where a plural is used for a singular, or
vice versa. Bede interprets the expression reges et principes (kings and princes) in Psalm
2:2 as an allusion to Herod and Pilate:

Item pro uno multi, ut: adstiterunt reges terrae, et principes convenerunt in unum.
Reges enim pro Herode, principes pro Pilato positos apostoli intellexerunt.
25

[Likewise, many things take the place of one in: The kings of the earth stood up and
the princes met together. For the apostles understood kings to refer to Herod and
princes to refer to Pontius Pilate.]
26


This interpretation does not yet appear in Bedes first commentary on the Acts of the
Apostles (Expositio Actuum apostolorum), which we know was certainly not written before
709, although in his revised discussion of the subject (Retractatio in Actus apostolorum),

19
J ones 1975, x-xi.
20
Irvine 1986, 43.
21
Brown 1987, 35-36.
22
Kendall 1991, 28-29.
23
All of these terms appear in Augustines letter 149, with which Bede was thoroughly acquainted. We may
safely assume similar knowledge in his recipient.
24
Holder 1999, 395. It must be added that already the Carolingians found conlevita confusing: Remigius of
Auxerre seems to have shared Plummers and Laistners mistake in his 9
th
-centure gloss to De arte metrica.
Kendall 1975, 141: CONLEVITA id est simul levita quia et beatus Beda tunc erat diaconus. (Conlevita, that
is, Levite at the same time, as the blessed Bede was also a deacon at that time.)
25
Bede, De schematibus et tropis, ed. Kendall 1975, 1, 61-63.
26
Trans. Kendall 1991, 173.
8

written not before 716 and possibly as late as 725-732, it does.
27
As Franklin states, it is not
clear whether De schematibus et tropis or Retractatio is the earlier work, but suggests that
Bedes treatise on schemes and tropes may have been composed between the two treatises on
the Acts. Even this, however, would not rule out even a considerably later date for Bedes
twin work on grammar.
28
It is worthy of note that Bedes even briefer treatise on
orthography, De orthographia, was likewise re-dated as early as 1982 on similar grounds,
29

and it is apparent that it, too, was composed no earlier than 709. In other words, none of
Bedes grammatical works seem to belong to the early part of his career, and despite their
perhaps deceptive conciseness must be regarded as works of considerable maturity.
The re-dating of Bedes grammatical works has several implications: firstly, it
is impossible to regard them as immature or derivative, as they obviously represent the result
of years of research. Secondly, when it comes to De arte metrica, it is apparent that Bede was
already an accomplished poet at the time of its composition. This means that the concept of
Bedes poetry as a scholars half-hearted attempt at the practical application of metrics must
be refuted (such views have led to Bedes verse being, if possible, even more underrated than
his works on grammar).
30
Rather, this provides us with another way of looking at Bedes
views on metre: De arte metrica is more than an exposition of metrical rules in the abstract,
and it is fair to assume that it strongly reflects Bedes own verse technique.
This, indeed, is the starting point of Neil Wrights 2005 article, where he
discusses several prosodic and stylistic features of Bedes metrical Vita Sancti Cuthberti,
finding several correspondences with his rulings in his De arte metrica. Judging by an early
draft of the work (the so-called B redaction), possibly composed around 705,
31
Bede appears
to have undergone some changes of mind regarding prosodic issues: the final version of the
hagiography was reworked in some crucial respects, and corresponds more closely with
Bedes later rulings on prosody.
32
It is also apparent that Bede, far from regarding metre and
grammar as merely preliminary subjects, found them an indispensable companion to all
scholarly activity, and retained an active interest in them throughout his career. Nothing
speaks more strongly for this than the very fact that the dating of Bedes grammatical
writings is based on the evolution of his views on biblical exegesis.

27
Franklin 2000, 202; see also Laistner 1939, xii-xvii.
28
Franklin 2000, 203.
29
Dionisotti 1982, 125.
30
e.g. Turner 1836, 376.
31
Lapidge 1995, 346-347.
32
Wright 2005, 150-170.
9

1. 4. The role of metrics in Bedes curriculum

If indeed, as Bede himself put it, the study of grammar was necessary for the informed
reading of that Book in which we have eternal life, we cannot afford to take lightly either the
tone or content of Bedes writings on grammar. The starting point of Bedes treatise on metre
was the belief that metrics were an inalienable part of J udaeo-Christian heritage. This belief
was based on the writings of such Christian apologists as Cassiodorus, Jerome and Isidore,
33

who taught that considerable portions of the Old Testament had been composed in verse, and,
even more astonishingly, in hexameters and pentameters.
Such claims, which to us may seem outlandish, mainly served to refute the
claims that Christians were barbarians or cultural upstarts: Moses was recast as the Christian
Homer, and the origins of poetry were transplanted into a biblical sphere. Of course,
Cassiodorus and J erome found hexameters and pentameters in the Bible because they wanted
to, but their assertions went unquestioned by subsequent generations of Christian scholars.
With their negligible knowledge of Hebrew, they were forced to trust Cassiodorus and
J erome implicitly.
34

Late antiquity, starting with J uvencuss third-century Evangeliorum libri, had
seen the birth of Christian Latin literature in classical metres, and the works of Christian epic
in hexameter verse by J uvencus and his followers Arator, Prudentius and Sedulius formed an
important model for the hexameter poets of Anglo-Saxon England. Nevertheless, the
influence of Vergil persisted both in the classroom and in didactic literature on prosody and
metre. Although such authors as J ulian of Toledo and Aldhelm had already introduced more
Christian material into their treatises, they still, to a high degree, relied on examples drawn
from the classics. The cited material in Bedes De arte metrica, on the other hand, has been
thoroughly overhauled: Bedes presentations of prosody in general and the dactylic
hexameter in particular are dominated by the example of the Christian epic poets, most of all
Sedulius. When Bede cites from Vergil, it is often as a last resort for want of an appropriate
Christian example, and strikingly often Vergil is used as a specific example of what not to do:
Bede emphasises, and frequently exaggerates, the prosodic differences between Vergil and
later (Christian) poets. He seems to have believed that there, indeed, existed two distinctly

33
Cassiod. in psalm. 118, 23-26; Hier. praef. Vulg. Iob; Arator ad Vigil. 80-81; Isid. orig. 1, 39, 11.
34
The concept that barbarians were in possession of a cultural heritage more ancient than that of the Hellenes
was, admittedly, not a novelty; such thoughts had been expressed by the Greeks themselves already in pre-
Christian times. Understandably, however, the Christians were more than willing to make full use of them. See
e.g. Ridings 1995, 24- 27.
10

different poetic practices, pagan and Christian. This tendency has been duly noted by
scholars, but the connection between Bedes treatment of his sources and the then-prevalent
view of the Hebraic origins of verse has not received the attention it deserves: Bede did not
merely think that the hexameter was a pagan invention which had been adopted and
possibly improved by Christian poets. Rather, believing in the biblical origins of all verse,
he saw that Christian poets were reclaiming from the pagans what was rightly theirs and
restoring it to its former glory. Bedes sincere efforts to help this project are apparent
throughout his twin works De arte metrica and De schematibus et tropis. The tone of these
works is already set by the opening chapter De littera, where Bede discusses the Latin
alphabet in a surprisingly extended form: he has added the Greek letters p, ,, and o to the
alphabet because of the common abbreviations for Iesus (IHS) and Christus (XPS), and o and
c because of their being mentioned in the Apocalypse (21:8).

Qui etiam post perceptionem Dominicae fidei H et X et P et A Graecas litteras, etsi
non in alphabeti ordinem recipiunt, divinis tamen paginis inditas continent, H
videlicet...intromittentes propter auctoritatem nominis Ihu, X et P propter nomen Xpi,
O propter auctoritatem Dominici sermonis, Ego sum A et O.
35

[Since their conversion to Christianity, they have also taken over the Greek letters
which are found in the Bible, eta, chi, rho, and omega and alpha, although they have
not admitted them into the order of the alphabet. To be specific, they have introduced
eta...on the strength of the name of Jesus, chi and rho on the authority of the name of
Christ, and omega on the authority of the Lords words, I am the Alpha and the
Omega.]
36


As we can see, Bede has here attempted to lend a particularly Christian tone to even the
smallest elements of verse. Not all aspects of metre, of course, allowed for a similar
treatment, but if we observe the quotations Bede has employed by way of illustration, the
very same tendency comes to the fore. Besides substituting, wherever possible, Christian
quotations for classical ones, Bede also, as I see it, tries to manipulate the content of the
examples he employs. The Christian citations are, for the most part, appropriately uplifting
even when fragmentary, and even in his classical quotations Bede seems to avoid actual
pagan content. It is telling that chapter eleven of his treatise, which deals with the aesthetics
of verse (Quae sit optima carminis forma), contains a longish quotation from the opening of
Lucans Pharsalia.
37
This may seem like a surprising choice in a chapter where all the other

35
DAM 1, 11-18.
36
Trans. Kendall 1991, 37. Bedes chapter on the alphabet has been discussed thoroughly in Robert B. Palmers
1959 article Bede as a Textbook Writer: A Study of his De Arte metrica, 573-584.
37
DAM 11, 64-69.
11

examples of good verse are Christian, but the motivation behind this may be the fact of
Lucans secularity: as a work of pre-Christian epic, the Pharsalia has the advantage of being
almost free of any mention of the pagan gods whose presence pervades the Aeneid in a way
that must have seemed repellent to Bedes sensibilities.


1. 5. The structure and aims of De arte metrica

Bedes presentation of the elements of verse generally follows the structure of Donatuss Ars
maior. He first describes letters (chapter 1), then syllables and their lengths (chapters 2-9),
then metrical feet (chapter 9), devoting the rest of his treatise (chapters 10-24) to actual
metrics. The role of the dactylic metres (the hexameter and the elegiac couplet) in his treatise
is pre-eminent: not only does Bede devote the most space to the discussion of these metres,
but his examples of syllable prosody are mostly derived from hexameter lines. Didactically
speaking, one of Bedes main accomplishments is the integration of syllable prosody with the
structures of metrical verse: this is an approach obviously necessary in an age where the
classical system of syllable quantity had died out. Bede was right in observing that metre
could only be taught by simultaneously keeping an eye on the issues of elementary prosody.
38

Only towards the end of his treatise does Bede present what amounts to a slim compendium
of lyric metres (chapters 17-23). This section mainly contains such metres as were employed
in ecclesiastical music, the most notable being the iambic dimeter and the trochaic
septenarius. Other lyric metres have made it to Bedes selection because they have been used
by Christian authors: the anacreontic metre (used in the proemium of Prospers Poema
coniugis ad uxorem), the phalaecean hendecasyllable (used in the introduction of Cyprianus
Galluss Exodus) and the sapphic strophe (used by Paulinus of Nola). Conspicuously, all
examples of lyric verse are Christian, and the wealth of verse-forms in Horace, the primary
model for lyric verse in earlier grammars, is absent. This is probably due not only to Bedes
lack of direct contact with Horaces verse but also Horaces secularity. Bede himself goes so
far as to say that many other metres can be found in earlier literature, but he has not deemed
them worthy of discussion, as they are pagan, or, from his point of view, irrelevant.
39
As
Bedes treatises were intended for the purposes of the monastic curriculum, learning for its
own sake could not be encouraged: rather, grammar was to be taught in such a form that it

38
Leonhardt 1989, 75-76.
39
DAM 24, 8-9.
12

complemented the study of scripture and the vocational life of the cloisters. Inevitably, this,
too, is reflected in Bedes presentation of metre: poetic forms worthy of study, alongside
Christian epic, were those forms of poetry which were of immediate importance in liturgy.
This explains the scarcity of lyric metres in Bedes De arte metrica and the way in which
they are primarily limited to those employed in hymnody. This also explains why Bedes De
arte metrica is the first metrical treatise to give a satisfactory presentation of non-
quantitative, or rhythmic, poetry. Bedes chapter De rithmo (chapter 24) is generally
acknowledged to constitute the most revolutionary portion of De arte metrica. Discussions of
rhythmic verse in earlier grammar are few; their emphasis is generally on its lack of syllable
quantity, and, as in the case of Julian of Toledo, they are limited to the admonition to avoid it.
Bede, on the other hand, recognised that the traditional terminology of prosody and metre
was not sufficient for the description of all liturgical texts. Remarkably, he also appears to
have realised that the rhythmic verse of early medieval hymns was based on quantitative
iambo-trochaic models and suggested that metre could find an equivalent in a system without
syllable quantity.
The final chapter of Bedes treatise (Quod tria sint genera poematos, DAM 25)
contains a very brief description of the principal types of narrative, drawn from Diomedes,
but supplemented by the writings of the Church Fathers and illustrated with references to
Scripture. The chapter serves both as an epilogue to De arte metrica and as a bridge to its
companion De schematibus et tropis, which describes different figures of speech. Although
this subject is usually seen to belong in the realm of rhetoric, something unsuitable for the
vocational studies of Christians, it had effectively been incorporated into grammar in the
monastic schools, largely through the influence of Donatuss discussion of schemes and
tropes in his Ars maior.
40
The Christian tone of De schematibus et tropis is even stronger than
that of its companion, where Bede was limited to verse in his choice of examples; in De
schematibus et tropis, on the other hand, he was able to draw freely on the Bible, and, indeed,
nearly all rhetorical figures are illustrated with biblical quotations, together with a handful of
citations from the Early Fathers.

40
For his 1991 edition, commentary and translation of the two works, Kendall has used the title The Art of
Grammar and Rhetoric, something which Bede would certainly have considered inappropriate. see Brown
2009, 23.
13

1. 6. Bedes Christian agenda and its implementation in his discussion of metrics

The purpose of the present thesis is to explore the ways in which Bede, in the composition of
his De arte metrica, had utilised his sources, grammarians and poets alike: what he chose to
leave out, what he paraphrased and what he boldly redefined; we can find abundant examples
of all this in his De arte metrica. Bedes primary sources on prosody and metre were the
Latin grammarians of late antiquity whose works have all been edited in Heinrich Keils
Grammatici Latini: they include, alongside Donatus, his commentators Sergius and
Pompeius, Audax, Marius Victorinus, Maximus Victorinus, Mallius Theodorus, Diomedes
and Charisius.
41
This, conversely, means that Bede probably did not have access to the works
of such classics as Cicero, Quintilian, the Rhetorica ad Herennium or Varro, let alone Greek
sources. When it comes to poets, the situation is fairly similar: Bedes examples of verse are
primarily taken from the Christian poets of late antiquity and Vergil, with, as we have already
stated, a strong preference for the former, although Bede also cites the opening of Lucans
Pharsalia, as well as one line from Lucretius and another from Horace; it is likely, though,
that he did not know the latter two authors at first hand. Although De arte metrica contains
many verbatim and almost-verbatim quotations from the grammarians, Bede has surprisingly
often altered their phrasing in ways which, taken at a glance, may seem insignificant but
which result in a fundamental change of meaning. Take, for example the description of the
dactylic hexameter as presented by Mallius Theodorus:

Constat autem metrum dactylicum hexametrum heroicum ex dactylo et spondio vel
trochaeo, ita ut recipiat spondium locis omnibus, dactylum locis omnibus praeter
ultimum, trochaeum vero loco tantum ultimo.
42

[The heroic dactylic hexameter consists of the dactyl, the spondee and the trochee in
such a way that it takes the spondee in every foot, the dactyl in every foot except the
last one, and the trochee only in the final foot.]

Bede follows this definition almost to the letter, but not quite:

Constat autem ex dactylo et spondeo vel trocheo, ita ut recipiat spondeum locis
omnibus praeter quintum, dactylum praeter ultimum, trocheum vero loco tantum
ultimo.
43


41
Keil in gramm. VI, 220-221; Irvine 1986, 32. On Bedes library, see Laistner 1957, 117-149; Lapidge 2006,
34-17; 191-228. The Roman grammarians have been given a clear and concise presentation in Law 1982, 11-29.
42
gramm. VI, 589, 230-233.
43
DAM 10, 2-13; my italics.
14

[It is formed from the dactyl, the spondee, and the trochee in such a way that it takes
the spondee in every foot except the fifth, the dactyl in every foot except the last, and
the trochee only in the final foot.]
44


As we can see, Bede has added the words praeter quintum to Malliuss definition, thereby
effectively ruling out spondaic lines (towards which, as we shall observe, Bede harboured a
vehement dislike). A similar addition appears in Bedes description of rhythmic poetry, this
time as a vindication of a new, Christian literary genre. Maximus Victorinus gives the
following portrayal of the similarities between rhythm and metre:

Plerumque tamen casu quodam invenies rationem metricam in rhythmo, non artificii
obseruatione servata, sed sono et ipsa modulatione ducente.
45

[However, you can often by chance find measured quantities even in rhythm, not
because the regular artistic arrangement has been preserved, but from the influence of
the sound and the beat itself.]

Bede, although otherwise in agreement with Victorinus, takes a stand for the literary merits of
rhythmic verse by the simple addition of one clause:


Plerumque tamen casu quodam invenies rationem metricam in rhythmo, non artificii
observatione servata, sed sono et ipsa modulatione ducente, quem vulgares poetae
necesse est rustice, docti faciant docte.
46

[However, you can often by chance find measured quantities even in rhythm, not
because the regular artistic arrangement has been preserved, but from the influence of
the sound and the beat itself. The common poets inevitably do this awkwardly, and
the learned poets skilfully.]
47


Similar minor alterations appear throughout the work. It must be noted that Bede rarely goes
so far as to contradict his predecessors openly.
Alongside with its Christianising element, an important feature of Bedes De
arte metrica is its practical approach to verse technique. Bede often simplifies the
presentations of his predecessors, especially where they are cluttered with unhelpful jargon.
We encounter this characteristic in Bedes discussion of metaplasms, or metrical licences,
where he has pared the dozen or so types mentioned by earlier grammarians down to merely
four ones which play an actual role in verse composition, focusing particularly on the

44
Trans. Kendall 1991, 97; my italics.
45
gramm. VI, 206, 7 207, 3.
46
DAM 24, 16-19; my italics.
47
Trans. Kendall 1991, 201; my italics.
15

technique of elision, a source of some bewilderment for his contemporaries. Especially in
Bedes treatment of the lyric metres, his definitions are simplified to such a degree that they
mainly seem to be descriptions of the poems which Bede uses as their illustration. As Bedes
examples of the lyric metres are drawn mainly from Late Latin hymnody, where syllable
resolution is scarce, he barely touches on the phenomenon at all: Bede presents the trochaic
septenarius and the iambic trimeter as isosyllabic metres (with a fixed number of syllables),
and syllable resolution in the iambic dimeter is only mentioned as an afterthought.
As Bedes views on prosody and hexameter technique are based heavily on
what the Christian poets of late antiquity, above all Sedulius, did or did not do, they strongly
reflect the prosodic features of Late Latin, as well as exhibiting some more contrived
techniques that had become popular in post-classical verse. Bede presents the prosody of final
vowels in an essentially post-classical form: above all, the final os of first-person verb forms
and third-declension nouns are, by his definition, short, and he condones the strange post-
classical practice of treating the letter h as a consonant, a common feature in the verse of the
Christian epic poets. On the other hand, he condemns outright several other poetic licences
(apparently because they were not used by Sedulius), labelling them as essentially pagan
practices. These include, above all, spondaic lines (Bede appears virtually incredulous of
their existence) and hiatus.
Bedes notion that these features are pagan practices is not wholly justified, as
they do appear also in Christian poetry while not being particularly common even in classical
verse, but it gives him the possibility of using Vergilian lines with these features as a caveat,
thereby underlining his view of Christian verse technique being an improvement on pagan
practices in every respect. Possibly disingenuously, he also devotes one whole chapter to
explaining away Christian lapses from the prosodic norm which he elsewhere seeks to
delineate.
Bedes observations on poetic style, although not numerous, are considerably
original and equally telling in their commendation of techniques which Bede had noted in
Sedulius and other Christian authors. Bede is apparently the first author to describe the kind
of double hyperbaton which in modern scholarship is known as the golden line. The wide-
spread use of the term is probably the indirect result of Bedes observations, and although
usually applied to classical hexameter verse, the feature is really a characteristic of Seduliuss
verse style. Similarly, Bede commends a kind of line that consists of an asyndetic list of
words, a common feature of much later medieval poetry.
16

This thesis explores the variety of ways in which Bede sought to create a
compendium of rules for what he viewed as an ideal way of composing Christian verse. I will
look at the historical backgrounds of the phenomena which he struggled to cast in a clear-cut
and unambiguous way, drawing on his sometimes conflicting sources and his own
observations as a scholar and poet. I will also venture to shed light on the ways in which
Bedes reforms of the nomenclature of metre and prosody, as well as his personal likes and
dislikes, foreshadowed the practices of medieval poetry. Many of the features which I discuss
have been noted in previous articles, but hitherto they have not been observed as an entity,
which they deserve to be. I have decided to forego a closer examination of Bedes chapters 1
(De littera) and 25 (Quod sint tria genera poematos), which, although in their own way
illuminating, have no actual bearing on questions of prosody or the scansion and composition
of verse. The former has been studied in detail by Palmer (1959, pp. 573-584); a discussion
of the latter would more properly belong together with a study of Bedes De schematibus et
tropis.
Although Bedes discussion of grammar and metre seems highly pragmatic on
the surface, his belief in the importance of these subjects was in essence idealistic, based as it
was on the concept that metre was ultimately created by God. In this thesis, I will venture to
study, and, as far as possible, explain, this subtle interplay of the practical and the idealistic,
as it is manifested in Bedes writings on metre and grammar.
17

2. Hexameter verse and general prosody


2.1. The dactylic hexameter in Anglo-Saxon England

The Anglo-Saxons are generally acknowledged as the first non-Romance nation to create
quantitative Latin poetry of any consequence, and therefore they enjoy a unique position in
the cultural history of the Middle Ages. The native language of the last representatives of
Roman hexameter poets was still Latin: Arator (fl. 550) was a Latin-speaking Ligurian,
Corippus (fl. 550) was a North African speaker of Latin, and even Venantius Fortunatus (ca.
540-600), although he made his career in Merovingian Gaul, was originally from the Latin-
speaking north of Italy.
1

The first attempts at hexameter verse and other forms of quantitative poetry in
post-Roman Europe were prosodically shaky, few in number and meagre in scope. This is
obvious in the poetic creations of Visigothic Spain: King Sisebuth (d. 620) composed a poem
of about fifty lines on eclipses of the sun and the moon, whereas Bishop Eugene of Toledo (d.
647), possibly a Visigoth himself, wrote a handful of occasional poems, none of which
exceed the length of twenty lines.
2
The early verse composed by the learned clerics of sixth-
and seventh-century Ireland, on the other hand, seems to have been exclusively non-
quantitative,
3
and it is highly questionable whether the early Irish scholars had any grasp of
quantitative prosody or the poetry based on it.
4
The earliest Irish hexameter poetry was
composed no earlier than the eighth century, and is demonstrably based either on Anglo-Latin
or Carolingian models.
5
Even in sixth-to-eighth-century Lombardy, closer to the heartland of
the Latin sphere, the general decline in classical learning had an inevitable effect on poetry as
well: this is demonstrated by numerous poems composed in what must be a form of the
dactylic hexameter, albeit without any sense of syllable quantity.
6

St. Aldhelm, abbot of Malmesbury and bishop of Sherborne, is without doubt
the first Latin hexameter poet of any consequence to emerge after Venantius Fortunatus.

1
Lapidge 1979, 210.
2
ibid.
3
The sole exception may be a handful of poems in quantitative adonic verse, which have been attributed,
probably spuriously, to St. Columban (540-615). Bolton 1967, 42-43.
4
Roger 1905, 267-268.
5
Lapidge 1999, 373.
6
Meyer 1905, 230-234; Norberg 1958, 101-104. Both Meyer and Norberg have questioned whether these verses
have any intention of passing for metrical poetry; rather, they simply seem to imitate the word accents of
classical verse without taking syllable quantity into account, being in essence rhythmic hexameters.
18

Although, in time, the interval between the composition of Venantiuss hexameter
hagiography of St. Martin and Aldhelms main poetic work, Carmen de virginitate, is not
huge - the former was written around 600 and the latter at the end of the seventh century
their cultural background was fundamentally different: Venantiuss native language was still
Latin and his schooling classical.
7
Although the poetry of Venantius Fortunatus can be seen
as medieval in spirit and his verse, at times, reflects the collapse of classical syllable
quantity, he is still a representative of an unbroken classical tradition. Aldhelm, on the other
hand, was an Anglo-Saxon, his native language Old English and his training and education
monastic. The quantitative prosody of classical Latin was, for him, not only alien but also
difficult to grasp: it no longer played a role in the spoken Latin of his day, and there was no
equivalent feature in his native language. This meant that Aldhelm had to resort to a number
of techniques and solutions that were radically different from those of his late Roman
predecessors. These methods were passed on to his followers and can be said to have
influenced Latin hexameter poetry for centuries to come.
Aldhelm presumably became acquainted with the rudiments of quantitative
versification at the monastic school of Canterbury, where he had the opportunity of studying
under two pre-eminent southern scholars, Theodore of Tarsus and the North African Hadrian.
It is probable that Aldhelms verse technique is based on the teachings of these two men.
8

Aldhelms approach to versification is well illustrated by his didactic/poetic Epistola ad
Acircium, an epistle to the learned king Aldfrith of Northumbria. The work, known also by its
full name De metris et enigmatibus ac pedum regulis, consists of four parts: a rambling
preface on the symbolic value of the number seven is followed by two treatises on different
aspects of verse, known respectively as De metris and De pedum regulis. Sandwiched
between them is a collection of hexameter riddles, mainly modelled after the Origins of
Isidore, the ostensible purpose of which is to illustrate the metrical rules discussed in the rest
of the book. De metris, which mainly relies on Audaxs Excerpta,
9
discusses the structure of
the hexameter in detail, outlining the use of dactylic and spondaic feet, the principal types of
caesurae and the use of elision, which was apparently an issue of considerable difficulty for
Anglo-Saxon poets. Aldhelm follows the example of the Late Latin grammarians in citing

7
Manitius 1911, 170.
8
According to Bede, the subjects taught by Theodore and Hadrian included astronomy, computus and metrics:
ita ut etiam metricae artis, astronomiae et arithmeticae ecclesiasticae disciplinam inter sacrorum apicum
volumina suis auditoribus contraderent. (So that they passed to their pupils, among the study of sacred books,
also the disciplines of metre, astronomy and ecclesiastical arithmetic.) Bede, Hist. eccl. 4.2.
9
Although Aldhelm is supposed to have been acquainted with Priscians Institutiones grammaticae , the work
does not reflect Priscians theories in any perceptible way. Ruff 2005, 150.
19

several classical and late antique poets to illustrate the metrical rules he discusses, but he
also, on occasion, uses verses of his own, especially when he wants to demonstrate the
possible ways of combining dactyls and spondees. Throughout his work Aldhelm displays a
tendency to view metre as an exercise in arithmetic: for him, poetry consists simply in finding
a way of combining units with the right syllable lengths to produce larger prosodic entities.
This propensity of Aldhelms is reflected both in his discussion of metrical rules and in his
own verse technique, as demonstrated by his poetic works. Aldhelms adoption of Audaxs
nomenclature of different types, or schemata, of hexameter lines, in itself of dubious use to
students of poetry, betrays the authors obsession with patterns: he first classifies hexameters
by the number of dactyls and spondees in their first five feet, and then further categorises
these line-types according to the number of permutations they allow. Hence, for instance, he
refers to lines with one dactyl and lines with four dactyls collectively as pentaschemi, as they
both allow five permutations (dssss, sdsss, ssdss, sssds, ssssd and dddds, dddsd, ddsdd,
dsddd, sdddd, respectively). Aldhelm demonstrates each of these schemata in wearisome
detail with hexameter lines on the theme Christ on the cross saved the world. The lines are
virtually identical both in content and wording, and illustrate Aldhelms technique of
substitution and permutation admirably. As examples of schemata with four dactyls Aldhelm
presents the following lines:

Iam veneranda Dei soboles cruce mundum salvat; (dddds)
iam veneranda patris soboles salvat cruce mundum; (dddsd)
en veneranda Dei proles cruce saecla coruscat; (ddsdd)
iam pietas immensa Dei cruce cuncta beavit; (dsddd)
mundum iam veneranda Dei soboles cruce salvat.
10
(sdddd)

As we can see, most of the words in the lines allow liberal permutation. Furthermore,
Aldhelm has at his disposal a number of synonyms or near-synonyms, which he uses to create
some additional variation. Such words can be prosodically similar (like the iambic Dei and
patris) or dissimilar (like the spondaic proles and the dactylic soboles); in the former case,
they can be employed to avoid tautology,
11
in the latter, they can be used as substitutes for
each other in metrically different positions. In addition, Aldhelm often resorts to short and
semantically predictable words (like cruce) to fill in gaps in his lines. The fourth part of

10
Ehwald 1919, 88.
11
Admittedly, Aldhelms use of synonyms is often gratuitous and pleonastic; for instance, both escarum saginis
and alimenta ciborum for food occur in successive lines (10 and 11) of his riddle on the date palm (Enigmata
XCI, Ehwald 1919, 139).
20

Aldhelms Epistola, De pedum regulis, is little more than a list of prosodically different
words (pyrrhic, iambic, trochaic, spondaic, dactylic, anapaestic etc.), which he apparently
considers useful mainly as providing a wide selection of such metrical gap-fillers. Aldhelms
view of the hexameter line is straightforward: a poet must first learn the metrical
framework of the hexameter line and then absorb a sufficient number of prosodically
familiar words and phrases to fill it.
12
This approach is practically the diametrical opposite of
the method employed by Donatus and other earlier grammarians, who first discuss the
smallest elements of language (letters, then syllables and words) and only then combine them
into longer units such as metrical feet.
13
It is well worth asking whether a fledgling poet could
actually have learnt to compose verse with the aid of Aldhelms treatise; in reality, he would
have had to rely more on the example of previous poets and his own ear.
14
It is also worthy of
note that the mathematically minded Aldhelm presents metrics as something entirely
divorced from style and meaning:
15
for him verse composition, both in theory and practice,
appears to be equivalent to the completion of a crossword puzzle.
Aldhelms own poetic technique shows notable parallels with this mechanistic
approach to the structure of the hexameter. His hexameter verse is, to quote Andy Orchard,
almost wholly cobbled together from a combination of repeated phrases, both borrowed and
newly coined.
16
Most of his lines are formed from three distinct building blocks: the first
two and a half feet, followed by a strong penthemimeral caesura, form one. This, in turn, is
usually followed by a molossus of three long syllables and a final cadence formed by a dactyl
and a spondee (or trochee).
17
Aldhelm achieves variety mainly by altering his repeated
phrases: for instance, in his Carmen de virginitate, the final cadences regna polorum,
claustra polorum, sceptra polorum and astra polorum all signify heavenly realms.
18


12
Lapidge 1976, 213.
13
One possible reason for Aldhelms ostensibly impractical approach may be that De metris was not really
intended to be a propaedeutic work and that, according to Aldhelms assumption, the reader was already
acquainted with Donatus. See Ruff 2005, 155.
14
Wright 1985, 188.
15
Ruff 2005, 153.
16
Orchard 1994, 111.
17
The following passage from his Carmen de virginitate (lines 44-48, Ehwald 1919, 354-355) may serve as an
example:
Omnia regnando / dispensat / saecula simplex
en promissa novo / scribantur / carmina versu
garrula virgineas / depromat / pagina laudes,
colaque cum pedibus / pergant et / commata ternis.
[God guides in rule all generations as one. Behold, let these promised songs be composed in new
verse! Let the fluent page issue praise of virgins, and let the clauses and caesuras of the verse proceed
with three types of foot.]
- Trans. Rosier 1985, 104.
18
Lapidge 1975, 226.
21

Otherwise, his hexameter poetry is exceedingly monotonous and foursquare: the ends of
clauses almost invariably coincide with line-endings, and variation in the placement of
dactyls and spondees, for all of Aldhelms talk about schemata, is minimal. This type of verse
structure had its undoubted advantages: a line that consisted of three prosodically predictable
and transferable blocks was fairly easy to grasp even for inexperienced poets and could be
used in a cut-and-paste manner to produce new verse. This is probably one of the chief
reasons for Aldhelms otherwise perhaps surprising persistence as a model for Anglo-Saxon
hexameter poets.
Bedes approach to the problems of the dactylic hexameter differed from
Aldhelms both in theory and practice, and the differences between these two poet-
grammarians reflect their respective artistic temperaments. The most apparent difference
between Aldhelm and Bede is the latters emphasis on variety that is only equalled by the
formers complete disregard for it. Bede also differs from Aldhelm in seeing the
interconnectedness of metre, syntax and rhetoric (which, of course, was incorporated into
grammar): whereas Aldhelm appended hexameter riddles and a numerological treatise to his
De metris and De pedum regulis, Bedes De arte metrica is accompanied by De schematibus
et tropis, a handbook on figures of speech, and, throughout his treatise, it is evident that Bede
keeps a keen eye not only on metrical structures but also on syntax and literary expression. It
has been suggested that Bedes treatise as well as his own hexameter technique are, at least
partly, a reaction against Aldhelm, and even that when Bede warns against excessive
repetition, Aldhelm is his main target.
19

Bedes own poetry testifies to his ability to look beyond the inner metric of the
hexameter line. Unlike Aldhelm, whose lines are invariably end-stopped, Bede frequently
practices enjambment; his lines are less heavily spondaic than Aldhelms, and there is
considerably less repetition of favourite schemata. Bede is also noticeably more
sophisticated in his use of elision, the stumbling-block of many Anglo-Saxon poets. Bedes
verse hagiography of St. Cuthbert is valuable evidence for Bedes views on metre, as it
frequently shows striking parallels with his own teaching in De arte metrica: Bede made full
use of the various metrical devices at his disposal, rather than merely discussing them as
problems that needed to be solved.
Both of these quite different approaches to hexameter verse found their
followers, and it has been argued that Anglo-Saxon learning brought forth two different

19
Wright 2005, 166.
22

schools of hexameter verse. The more rigid Aldhelmian hexameter was emulated by poets
of what Andy Orchard has termed the Southumbrian school of hexameter poets, whereas
Bedes lighter and more varied hexameter style won the following of the Northumbrian
school.
20
Possibly the most prominent representative of the latter was Alcuin, through whom
the influence of Bedes views and the poetic diction that reflected them spread to the
Carolingian mainland.
Anglo-Latin hexameter poetry exhibits several common prosodic and lexical
features that, at least in some cases, betray its formulaic nature. When it comes to prosody, it
must be remembered that Anglo-Latin poets relied on the example, and often meticulous
study, of both the classics and the Christian poets of late antiquity. In this context, their more
striking prosodic liberties cannot really be seen to constitute metrical flaws. Some of the
typically late-antique prosodic features adopted by Anglo-Latin poets are the following:
1) The shortening of long final vowels also in words not subject to the brevis
brevians law, a feature attested in much of post-classical Latin poetry (although, in Silver-
Age poets, never in dative or ablative forms),
21
was commonly applied to the final o of first-
person verb forms, the nominative forms of third-declension nouns and even the dative and
ablative of second-declension words; similar abbreviation also took place in the final e of
adverbs, the final vowels of imperative forms and even the final a of first-declension
ablatives.
22

2) The consonantal use of h: word- initial h can either create a position after a
closed syllable or cause a hiatus (as in Sedul. carm. pasch. 3, 296, vr humilis maesto
deiectus lumine terram, also cited by Bede).
23

3) The so-called s impurum, or a word-initial s group (sc, sp or st) that creates a
position (as in Ven. Fort. Mart. 3, 1. hactenus in bibulis fix stetit anchora terris). An
opposite case of word-internal s groups that do not create a position is a possibly hypercorrect
feature typical of Aldhelm (e.g. gstat and rstat).
24


20
Orchard 1994, 239-283.
21
Raven 1965, 23.
22
Campbell 1953 passim; Norberg 1958, 6.
23
Norberg 1958, 6-7: this practice is apparently based on Vergils line terga fatigams hasta (Aen. 9, 610),
probably a case of productio ob caesuram, which had become a standard textbook example.
24
ibid; Ehwald 1919, 755; Orchard 1994, 75-76.
23

4) A number of other more or less established wrong quantities typical of late
antique poets: ige, triduum, ecclsia, pter, sttim, quque, c, foris. In more arcane
vocabulary and Greek words, in particular, false quantities are extremely common.
25

It is typical that even the more learned poets sometimes deviated from the
quantities prescribed by the grammarians. Clearly the reason is that, as writers of Christian
poetry, they were overly dependent on the example of their Late Latin predecessors. In some
cases, however, it appears to have been a matter of conscious choice: Bedes own poetry
shows some metrical liberties of peculiarly Late Latin nature that do not correspond with his
own description in his De arte metrica;
26
at least in some cases they can probably be
attributed to his admiration of Sedulius.
The lexical and stylistic features typical of Anglo-Latin verse are sometimes
caused by gratuitous ostentation, especially when it comes to the obscure vocabulary
favoured by the so-called hermeneutic school of Anglo-Latin writing whose founding
father was Aldhelm itself.
27
The hermeneutic school was fond of archaisms, neologisms and
foreign words mainly extracted from glosses. The roots of this school can be seen in the
Late Latin prose of such authors as Apuleius, Ammianus Marcellinus, Martianus Capella,
Ennodius and Sidonius Apollinaris, and the tradition was carried on by such insular figures as
Gildas, Columban, Virgilius Maro, the composers of Hisperica famina and Aldhelm. In
verse, the hermeneutic aesthetic often manifested itself in the form of some rather contrived-
sounding metrical devices, which, on the other hand, also served as quick fixes to the
prosodic problems frequently encountered by Anglo-Latin authors. Some typical lexical
features of Anglo-Latin verse (abundant in, but by no means confined to, the works of the
hermeneutic school) are the following, as presented by Michael Lapidge:
1) The use of distributive numerals and numeral adverbs and the tendency to
express numbers in the form of multiplication. This was, of course, necessitated by the
prosodic impracticability of many cardinal numbers (e.g. undecim, duodecim), but sometimes
became a mannerism.
2) The use of ast for at.
3) The frequent use of compounds with -dicus, -loquus and -loquax.
4) The use of neuters with -amen in the ablative singular (-amine) or in the
plural (-amina), especially to fill in the fifth foot of the hexameter line.

25
Campbell 1953, 14: Greek words not regularly found in earlier verse are scanned wrong as often as not, and
so are rare Latin words.
26
ibid.
27
Lapidge 1972, 86; Lapidge 1975, 67-90.
24

5) The use of archaic passive infinitives with -ier to fill in the fifth foot before a
word with an initial vowel.
6) The ablative of the gerund (with a short o) used as a substitute for the present
participle, especially to fill in the beginning of the fifth foot (as in meditand rediscens).
7) The free use of pentasyllabic words to fill in the fifth and sixth feet of the
hexameter line.
28

Some of these features, such as passive infinitives with -ier and pentasyllabic
words in line-endings,
29
are typical of archaic Latin verse, some are post-classical, but most
of them were necessitated by the Anglo-Saxon authors need for dactyls in the appropriate
parts of the line, and nowhere more so than in the fifth foot, which in Anglo-Latin poetry is
invariably a dactyl, spondaic fifth feet having been effectively proscribed both by Aldhelm
and Bede. This is a feature which we shall discuss later in detail.
To recapitulate: the Anglo-Latin hexameter poets were compelled to make do
with what had been handed down to them from previous generations of poets. They resorted
to techniques from various ages of Roman letters, archaic, classical and post-classical alike,
and often they did so indiscriminately: their choices were more often than not dictated by the
necessity of keeping the basic fabric of the hexameter line together, and actual stylistic issues
did not always enter the picture. Bede, both as a poet and as a grammarian, stood apart from
many of his contemporaries in exhibiting an unusually keen and analytical sense of style.
This is evident in his verse as well as his guidelines for other poets in De arte metrica, where
he exhibits not only an interest in stylistic questions in general, but also an ability to
distinguish between different poetic styles and their metrical manifestations. Bedes typology
of old, or pagan, and new, or Christian, hexameter styles is a case in point, and, above
all, remarkable in the way it pervades his whole presentation of the dactylic hexameter.


2.2. Classical and post-classical prosody: common syllables

Bedes discussion of prosody follows the model of previous grammarians (e.g. Donatus) in
starting out from the simplest elements of language: firstly, the letter, and secondly, the
syllable. In this respect, Bedes approach must be considered at the same time more practical
than Aldhelms, as well as more successful in its incorporation of prosody into the metrica

28
Lapidge 1999, 372.
29
Raven 1965, 32; Wilkinson 1963, 90-91.
25

ars. The second chapter of Bedes treatise, titled De syllaba, presents the elementary rules of
syllable quantity and the differences between long and short syllables. The chapter is
markedly traditional, if not hidebound, in its adherence to the terminology of classical
grammarians, exemplified by its application of the terms acute and circumflex to the
Latin word-accent. Of greater interest for a study of Bedes views on prosody is the third
chapter, De communibus syllabis, which deals with common syllables, or cases where a
syllable may be interpreted as either short or long. From a stylistic point of view, the
questions pertaining to common syllables are important above all because, as Bede himself
attests,
30
they often belong to the realm of poetic licence, where the strict rules of prosody
can be bent to suit the requirements of an individual poetic style, but also because classical
and post-classical poets dealt with these questions differently. In all their ambiguity, common
syllables must have been especially difficult for Anglo-Saxon students of Latin prosody, and
Bede correspondingly discusses this problematic subject in detail, drawing on nearly all of his
major sources: Diomedes, Donatus, Mallius Theodorus, Maximus Victorinus, Marius
Victorinus, Pompeius, Sergius and Servius.
Bedes typology of common syllables is mainly taken from Sergius
31
and
Maximus Victorinus
32
, who recognise nine different types of common syllable. Bedes
presentation of these nine types cannot be said to be particularly systematic, as he explains
the cases in roughly the same order as his predecessors without any attempt at further
classification. The types of common syllable, as listed by Bede, are the following:
1) Syllables where a short vowel is followed by a combination called muta cum
liquida, i.e. a plosive (or, in some cases, an f) and a liquid.
2) Final syllables followed by a word-initial h.
3) Final syllables ending in a short vowel that are followed by an s group, i.e.
the combination of initial s and another consonant; Bede also discusses initial ps in this
context.
4) Final short syllables that are lengthened before a caesura (productio ob
caesuram).
5) Cases of hiatus where an unelided final diphthong can be shortened before
the initial vowel of the following word.

30
DAM 3, 1-3: Communes autem syllabae modis fiunt novem, quibus aut naturaliter longae poetica licentia in
breves aut naturaliter breves transferuntur in longas. (There are nine ways in which common syllables are
formed. In these ways syllables which are naturally long may be changed by poetic licence into short syllables
or syllables which are naturally short may be changed into long syllables.) - Trans. Kendall 1991, 47.
31
gramm. IV, 478, 29 480, 3.
32
gramm. IV, 230, 7-24.
26

6) Cases of hiatus where an unelided long vowel can be shortened before the
initial vowel of the following word.
7) The pronouns hic and hoc, which can be interpreted as long, even before an
initial vowel.
8) Short vowels followed by z.
9) The final syllables of lines.
Almost as an afterthought, at the end of his presentation Bede also discusses x and word-
initial gn. As we can see, the order in which Bede presents the different types of common
syllable is remarkably haphazard. In a more analytical presentation, the different types of
consonant groups (mutae cum liquidis and s groups) should more properly be discussed in the
same context. It is also notable that, although the double consonants z and x could be
classified as s groups (as, indeed, Bede has done with ps), Bede discusses them separately.
The consonantal use of h, a Late Latin oddity, is more properly a special case of productio ob
caesuram, but Bede has chosen to discuss these types of common syllables separately, even
going to the length of warning his reader not to confuse them. To facilitate my presentation of
Bedes views on prosody, I have chosen to discuss the various consonant groups together,
and to analyse Bedes views on consonantal use h in conjunction with productio ob
caesuram, as these types are historically connected, although Bede has either failed or been
unwilling to see their connection.


2.2.1. Plosives with liquids

Bede starts his presentation of common syllables with the most common case, where a short-
vowelled syllable remains short by position before a consonant group known as muta cum
liquida, or mute with liquid. Consonant groups that behave in this way are bl, br, cl, chl, cr,
chr, dr, fl, fr, gl, gr, pl, phl, pr, phr, tr, and thr, or, in other words, all combinations of either a
plosive or an f followed by a liquid (l or r). In early comedy, such syllables are almost always
interpreted as short,
33
as, indeed, they appear to have been in the prosody of spoken classical
Latin. In dactylic verse, however, they were variable already in the works of Ennius,
34
and in
late antiquity they exhibit a growing tendency to be treated as long; to some degree, this

33
Raven 1965, 25.
34
Allen 1965, 90.
27

phenomenon may reflect the evolution of spoken Latin.
35
Bedes presentation of this type of
common syllable is somewhat simplified, as he does not specify the type of consonant that
can be followed by a liquid, as Donatus,
36
or Diomedes,
37
for instance, have done:

Brevis quippe transfertur in longam, cum correpta vocalis in eodem verbo a duabus
explicitur consonantibus, quarum posterior est liquida. Est enim natura brevis in hoc,
mens tenebris obscura suis (Sedul. carm. pasch. 2, 209). Est positione longa in hoc,
mortisque tenebras (Iuvenc. 1, 128). In quo Sergius modo iniusto utitur exemplo,
neve flagella (Verg. georg. 2, 299). Flagellum enim in capite verbi habet liquidam
litteram consonanti subiectam, quae positio numquam brevem natura syllabam verbi
praecedentis potest facere longam.
38

[A short syllable may be lengthened when a short vowel is followed in the same word
by two consonants, the second of which is a liquid. It is short by nature in this
example: mens tenbris obscura suis (the mind is clouded by darkness). But it is
long by position in this one: mortisque tenbras (and the shadows of death).
Sergius illustrates this type with the inappropriate example, nev flagella (nor the
topmost shoots of the vine). For in flagellum, the liquid is next to a consonant at the
beginning of the word, which is a combination that can never make the final syllable
of a preceding word long when it is short by nature.]
39


Even in this very brief presentation both Bedes originality and independence come to light,
possibly together with his pro-Christian bias. He cites the Christian poets Sedulius and
J uvencus as his examples but renounces the Vergilian example (neve flagella) quoted by
Sergius as irrelevant. Bedes critical assessment of Sergiuss Vergilian citation shows
remarkable perspicacity: word-initial mutes with liquids hardly ever lengthen the final
syllable of the previous word in Latin poetry. The few exceptions to this rule in the classics
are probably due to the emulation of Greek models,
40
and it is interesting to note that this is
one of the examples of pre-Christian prosodic negligence that came under Bedes attack in
his chapter on the prosodic differences between pagan and Christian poets.
41

Especially in Greek metre, the nasals m and n are usually classified together
with the liquids proper, because after plosives they have similar prosodic features: the

35
e.g. classical Latin ntegrum but Italian intro (from late Latin intgrum). Norberg 1958, 13.
36
Holtz 1981, 606, 7-9.
37
gramm. I, 428, 30 429, 8.
38
DAM 3, 3-12.
39
Trans. Kendall 1991, 47, except where Kendalls 1991 edition gives the reading neve flagello for Verg.
georg. 2, 299.
40
Raven 1965, 25; e.g. lappaequ tribolique (Verg. georg. 1, 153); per impotenti freta (Catull. 4, 18).
41
DAM 16; Orchard 1994, 75-76. Curiously, Aldhelm, with characteristic individuality, made extensive use of
this liberty in his poetry. This feature may have been caused by his imitation of Venantius Fortunatus.
Aldhelms use of plosives and liquids is idiosyncratic also in other ways: he frequently shortens a notionally
long final vowel before an initial plosive and liquid. This is particularly striking in ablative forms, as in Carmen
de Virginitate 526: Quem Deus aethrali ditavit grati gratis, where gratia is an ablative that should have a
long final a. This feature is apparently without parallel in any other poets. Orchard 1994, 77-78.
28

combination of plosive and nasal either may or may not create a position after a short vowel.
In Greek verse, the rules, as in the case of proper liquids, are considerably more flexible than
in Latin, but usually a word-initial combination of plosive and nasal (as in tmesis, pneuma
etc.) does not create a position after a short vowel.
42
In Latin verse, the rules governing Greek
loans with plosives with nasals are essentially the same as those governing plosives with
liquids. The only native consonant-nasal compound of any relevance in this context is gn,
which creates a position word-internally but never word-initially; in reality, the g may have,
already in classical Latin, been pronounced as a velar nasal (as in English hangnail), except at
the beginning of a word, where it was possibly silent.
43

Bede discusses both word-internal and word-initial gn at the end of his chapter
on common syllables. He mentions the compound in conjunction with the letter x, their
common feature being the fact that they only create a position word-internally but not word-
initially. Bede accepts the classification of n as a liquid,
44
because initial gn is analogous to
initial plosive and liquid, but emphasises that, unlike plosives with other liquids, it does not
make a true common syllable. Bede implies that this observation is his own, and
demonstrates this with an appropriately uplifting quotation from Prosper:

N quoque littera pari ratione, ni fallor, cum in medio verbo consonanti alteri fuerit
subiecta, praecedentem syllabam sive natura seu positione semper longam habet, ut
regna, calumnia. Cum vero in primordio verbi fuerit alii subiecta consonanti, ut
Gneus, gnarus, profecto ultimam syllabam verbi prioris, si in brevem desierit
vocalem, brevem hanc, ut fuerat, remanere permittit, neque ullam producendi habet
potestatem, Prospero teste, qui ait:
Nec tamen hos toto depellit foedere gnarus,
naturam errantum dividere a vitiis (Prosp. epigr. 67, 3-4).
45

[The letter n also, if I am not mistaken, in like fashion always makes a preceding
syllable long either by nature or by position when it is placed after another consonant
in the middle of a word, as regna, calumnia. But when it is placed after another
consonant at the beginning of a word, as Gneus, gnarus, there is no question but that
it permits the final syllable of the previous word, if the latter ends in a short vowel, to

42
Raven 1968
2
, 23.
43
Allen 1978
2
, 23-25; the evidence given by the grammarians is inconclusive: for the proper name Gnaeus
(abbreviated Cn.), Terentianus Maurus suggests the pronunciation Gnaeus with an ordinary g (Ter. Maur. 894);
Varro, according to J ulius Pariss De praenominibus epitome (10, 5), on the other hand, has noted the spelling
Naeus: Quod unum praenomen varia scriptura notatur: alii enim Naeum, alii Gnaeum alii Cnaeum scribunt.
(One first name is written with a variety of spellings: some, namely, write Naeus, others Gnaeus and yet others
Cnaeus.) It is also worthy of note that the well-documented Greek transcription is for Gnaeus is Noio. See
also Zirin 1970, 27-29.
44
In accordance with traditional terminology, Bede classifies nasals as liquids already in his chapter on the letter
(De littera), see DAM 1, 56-60.
45
DAM 3, 127-136.
29

remain short as it was before. Nor does it have any power of making length, as
Prosper can witness, who says:
Nec tamen hos toto depellit foeder gnarus,
naturam errantum dividere a vitiis
(However, he does not drive away these from the whole law, being skillful in
separating the nature of the sinners from their vices.)]
46



2.2.2. S groups

The third type of common syllable exposed by Bede involves what can be termed s groups,
or, in traditional nomenclature, s impura, namely, initial consonant groups beginning with s,
primarily the combinations of s and plosive (sp, st and sc). The behaviour of s groups is in
some ways analogous to that of mutes with liquids, as they on occasion result in similar
variability of quantity after short vowels, and in Late Latin poetry and grammar this tendency
seems to have been even more pronounced.
In Latin poetry, unlike Greek verse, initial s groups normally do not lengthen
the previous syllable, should it end in a short vowel. In pre-classical and classical poetry,
deviations from this are occasionally found (e.g. Catull. 64, 186 null spes), especially in
what D. S. Raven terms studied imitation of Greek lines (e.g. Verg. Aen. 8, 425
Brontesqu Steropesque).
47
After the Augustan period, this device seemed to lose
popularity, and most poets avoid short final vowels in such positions altogether. There are,
however, notable exceptions among the Late Latin poets, and Dag Norberg notes that certain
authors, above all Sedulius, uniformly lengthen short-vowelled final syllables in these
positions.
48
Later Christian authors who shared Seduliuss taste for this licence include
Venantius Fortunatus, Aldhelm and Alcuin.
49
Bedes description of this metrical liberty is
terse and based entirely on Christian examples (Venantius Fortunatus and Sedulius). Bede
passes no judgement on this prosodic liberty, but implies that it is a normal part of metrical
practice, once again showing his indebtedness to the example of Sedulius:

Tertius modus est communis syllabae, cum verbum aliquod in vocalem desinens
correptam excipitur a duabus consonantibus, quarum prior sit s. Est enim natura
brevis in hoc Fortunati:
Ordinibus variis alba smaragdus inest (Ven. Fort. carm. 8, 3, 264).

46
Trans. Kendall 1991, 59-61.
47
Raven 1965, 24-25.
48
Norberg 1958, 77.
49
Orchard 1994, 77.
30

Est positione longa in hoc Sedulii:
Adveniat regnum iam iamque scilicet illud (Sedul. carm. pasch. 2, 249).
50

[The third kind of common syllable occurs when a word ending in a short vowel is
followed by a word beginning with two consonants, the first of which is s. The final
syllable of alba is short by nature in this example from Fortunatus:
Ordinibus variis alb smaragdus inest
(A bright emerald with varying patterns is on the diadem).
The final syllable of iamque is long by position in this example from Sedulius:
Adveniat regnum iam iamqu scilicet illud
(May that kingdom, to be sure, come even now).]
51


Bedes exposition of s groups shows curious parallels with a prosodic feature peculiar to the
verse of his predecessor Aldhelm. Aldhelm shows an idiosyncratic tendency to treat word-
internal s groups like individual consonants: in Aldhelms opinion, the rules of common
syllables apparently work both ways, and, consequently, we find in Aldhelm numerous cases
where, contrary to normal metrical practice, a word-internal s group does not create a position
after a short vowel (e.g. r-stat, g-stat, cri-stalli):
52
Bede appears to reflect this specious
logic in his uniquely faulty analysis of a Vergilian line:

Hanc Virgilius et in medio verbo alteri consonanti praepositam, ubi commodum
duxit, liquentium more transiluit, ut est:
Hortatur Mnestheus, nunc nunc insurgite remis (Verg. Aen. 5, 189);
nisi forte versum ita scandendum putamus ut sit horta spondees, turmne spondees,
stheusnunc spondeus, coniunctis scilicet vocalibus, quod diptongon vocant.
53

[Even in the middle of the word, when it suited his convenience, Vergil passed over
an s standing before another consonant, as though it were a liquid, as is the case in:
Hortatur Mnestheus, nunc nunc insurgite remis
(Mnestheus exhorts them, Now, now, pull on the oars);
unless indeed we were to argue that this verse should be scanned taking horta as a
spondee, turmne as a spondee, and stheusnunc as a spondee: that is, with the vowels
e and u joined in the combination which is called a diphthong.]
54


Bede, in other words, supposes that Mnestheus is to be scanned short-short-long (Mn-sth-
us), which would prosodically not be dissimilar to such Aldhelmian quantities as t-stan-tur
and g-stat. Only as a second alternative does Bede present the true scansion of the word
(Mnes-theus, with two long syllables). Bedes muddled analysis of the line reflects the
difficulties inherent in the syllable quantities of Greek words and names and his dependence

50
DAM 3, 28-34.
51
Trans. Kendall 1991, 49.
52
Norberg 1958, 6-7; Ehwald 1919, 755; Orchard 1994, 75-76.
53
DAM 3, 42-48.
54
Trans. Kendall 1991, 51.
31

on Vergil and other auctoritates in unscrambling them. Simultaneously, it demonstrates the
fundamental logic behind Bedes analysis of common syllables: in his (as, apparently, in
Aldhelms) eyes, s groups are prosodically analogous to the combination of plosive and
liquid. Bedes and Aldhelms idea that an s group may not necessarily create a position
within a word is probably a back-formation from the rules governing the word-internal use of
plosives and liquids.
In conjunction with the more usual types of s group Bede also touches on word-
initial ps, which he apparently classifies as a kind of inverted s group. Initial ps and x only
appear in Greek loans: in Christian Latin verse, the former had attained unprecented
popularity, mainly owing to such quintessentially Christian words as psallo and psalmus. As
Bedes intention was to write a guidebook for Christian poets, he could not overlook the
prosody of these words (where the initial ps does not create a position). The examples are
from Sedulius and Venantius Fortunatus; Bede also uses the opportunity gently to reprimand
Pompeius, who had neglected this issue:
55


Cum vero s in capite verbi alii fuerit consonanti subiecta, nequaquam potest ultimam
verbi prioris syllabam producere, quae in brevem desierat vocalem, ut Sedulius:
Stare choro et placidis caelestia psallere verbis (Sedul. carm. pasch. 1, 25);
et Fortunatus:
Vocibus alternis divina poemata psallunt (Ven. Fort. carm. 8, 3, 7).
Falsoque definivit Pompeius s non posse liquescere, nisi ipsa antecedat, ut ponite
spes sibi quisque (Verg. Aen. 11, 309).
56

[But an s placed after another consonant at the beginning of a word can never
lengthen the final syllable of a preceding word which ends in a short vowel. For
example, from Sedulius:
Stare choro et placidis caelesti psallere verbis
(To stand in a choir and chant the divine psalms in agreeable words);
and, from Fortunatus:
Vocibus alternis divina poemat psallunt
(They chant the divine hymns with reciprocal voices).
Pompeius therefore incorrectly claims that s cannot blend with another consonant
unless it comes first, as in ponit spes sibi quisque (put aside that hope, each man
has his own hope).]
57


Bede does not discuss initial x or z in this context, sharing as he does his predecessors literal-
minded approach to orthography and prosody: Bede deals with letters rather than sounds, and
the letters x and z, are, in his nomenclature, not analogous to other s compounds that are

55
gramm. V, 109, 9-10
56
DAM 1, 35-42.
57
Trans. Kendall 1991, 51.
32

actually spelt with an s (although, adhering to traditional terminology, he calls x and z by the
terms duplex littera, double letter, and consonans duplex, double consonant). He does,
however, cover both letters at the end of his chapter on common syllables (DAM 3, 109-114;
119-126), immediately before word-initial gn, which he considers analogous to x. In this he
follows the example of Sergius, who also discusses z at the end of his presentation of
common syllables (gramm. IV, 479, 25).
The case of the letter z is problematic, as it is a Greek loan in the Latin alphabet,
and its prosodic nature in verse was the result of not only its actual sound value but also its
historical background. Although the original sound value of Greek was originally either
[dz] or [zd], by the time late republican Romans adopted it, it had presumably evolved into a
voiced s sound [z], which was geminated word-internally [zz].
58
Word-initial z would,
therefore, not need to make a position after a final short vowel, but probably from adherence
to its earlier sound-value, Roman poets generally avoided placing an initial z after a short
vowel altogether (except such cases as Zacynthos in Verg. Aen. 3, 270).
59
Although the late
antique grammarians, as well as Bede, persistently referred to z as a double letter or a
double consonant, several Christian poets (including Juvencus and Fortunatus) frequently
treated it as a single consonant even word-internally. Not surprisingly, Bede condones this
post-classical practice:

Octavus modus est, cum correptam vocalem in eadem parte orationis sequitur z
consonans Graeca duplex. Est enim longa in hoc Iuvenci:
Difficile est terris adfixos divite gaza (Iuvenc. 3, 522).
Est brevis in hoc eiusdem.
Et gaza distabat, rerum possessio fulgens (Iuvenc, 3, 499).
60

[The eighth kind of common syllable occurs when the Greek double consonant z
follows a short vowel in the same word. For example, the first syllable of gaza is
long in this verse of Juvencus:
Difficile est terris adfixos divite gza
(It is difficult for those attached to the earth to be separated from their rich
treasure).
It is short in this verse of the same poet:
Et gza distabat, rerum possessio fulgens
(And treasure, the glittering possession of things, was at a distance)].
61


58
Allen 1978
2
, 45-46.
59
In early Greek hexameter verse (e.g. Homer), is still clearly a double consonant and makes a position, even
word-initially. Exceptions to this rule are words with an initial short syllable followed by a long one (Zukuv0o,
Zcciu), which in hexameter verse would be prosodically impossible after a long syllable. Vergils nemorosa
Zacynthos (Aen. 3, 270), an often-quoted example, reflects this practice. Zacynthos with a preceding short
syllable occurs also in Ovid (her. 1, 87). Post-classical examples of initial z not creating a position are
prosodically more varied; e.g. zelotypus in J uvenal (5, 45 and 6, 278) and zonae in Manilius (3, 319).
60
DAM 3, 108-114.
33


Word-initial z, however, must in Bedes opinion never make a position, and he is very
explicit on this point on several occasions elsewhere in his treatise. In his second chapter,
where he presents his classification of short, long and common syllables, he states
unequivocally:

Cum vero parte aliqua orationis in brevem vocalem terminata sequens sermo a littera
z incipit, nullam producendi habet potestatem, unde est nemorosa Zacinctos (Verg.
Aen. 3, 270).
62

[but when a word ending in a short vowel is followed by a word which begins with
the letter z, the z cannot lengthen the short vowel, which accounts for the phrase
nemorosa Zacinctos.]
63


Although Bede here relies on the example of Vergil, Vergil is the very poet whom he later
rebukes for neglecting this rule. In his chapter on the differences of pagan and Christian
poets, Bede cites Vergils Euriqu Zephyrique (Verg. Georg. 1, 352) as an example of
outdated prosody.
64
On the question of the prosodic value of z, Bede, in other words, departs
once again from classical practice: whereas in classical prosody, z is almost invariably treated
as a double consonant (with certain words like Zacynthos a notable exception), Bede, relying
on the example of Vergils Zacynthos and Iuvencuss gaza, views it essentially as a single
consonant that can make a position word-internally but never word-initially. This is a further
restriction that does not occur in the presentation of Sergius who treats a short vowel before z
as an ordinary case of common syllable.
65

Initial x, which Bede discusses at the end of his presentation of common
syllables, only appears in Greek names (mainly Xerxes) and is of less interest for Christian
poetry than ps and z, and, correspondingly, Bedes discussion of the feature is less original
than that of other s groups and more indebted to previous grammarians. Notably, the
hexameter line Bede uses as his illustration is from Ennius (ann. 13, 4), also cited by
Terentianus Maurus
66
(neither of whom he presumably knew at first hand), probably by way
of Sergius.
67
Bede concurs with normal prosodic practice in declaring that an initial x cannot
lengthen a final syllable ending with a short vowel, although word-internally, it is always a

61
Trans. Kendall 1991, 57-59.
62
DAM 3, 37-39.
63
Trans. Kendall 1991, 45.
64
DAM 16, 28-29.
65
gramm. 4, 479, 23-24.
66
Ter. Maur. 1160.
67
gramm. IV, 479, 29.
34

double consonant (unlike Sergius, who treats both z and x as cases where the syllable can be
common, not discussing their position in the word at all)
68
:

Cum vero in primordio fuerit verbi, non potest producere finem prioris verbi quod in
brevem desierat vocalem, ut:
Pontibus instratis conduxit litora Xerxes.
69

[But when it is at the beginning of the word, it cannot lengthen the final syllable of a
previous word which ends in a short syllable, as in:
Pontibus instratis conduxit litora Xerxes
(Xerxes connected the shores with a series of bridges).]
70



2.2.3. Productio ob caesuram and consonantal h

The fourth type of common syllable in Bedes list is productio ob caesuram, or productio in
arsi, a prosodic feature which the author also touches on in his presentation of the
consonantal h, which, as Dag Norberg notes, is ultimately its derivative.
71
Productio ob
caesuram involves the lengthening of a short final syllable in the arsis of the foot, or before a
word-break in the middle of a foot. Generally this takes place before the strong central
caesura of a hexameter line. In Vergils verse, there are fifty-four cases of productio ob
caesuram,
72
the most famous of which is omnia vincit amr / et nos cedamus amori (ecl.
10, 69), which became a textbook commonplace.
73
Surprisingly, many post-classical poets
eschewed this licence altogether, being in this matter firmer than Vergil and Ovid, even when
composing in an age where the prosodies of final syllables were becoming increasingly
shaky.
74
Consequently, Bede, as Mallius Theodorus before him, associates this technique
strongly with Vergil, although his example is taken from Sedulius:

Est enim modus quartus syllabae communis, cum post pedem quemlibet una syllaba
brevis remanserit de verbo, quae vel in vocalem desinens excipiatur a consonante

68
gramm. IV, 479, 25-26: Sciendum est etiam quod et x communem syllabam facit, sicut et z Graecum.
69
DAM 3, 123-126.
70
Trans. Kendall 1991, 59.
71
Norberg 1958, 7.
72
Allen 1973, 117.
73
The case of amor may be a simple archaism, as third-declension nouns in -or (gen. -oris) generally have a
long final syllable in the nominative in Plautus and Ennius; similarly, in Vergil, we also encounter dolr (Aen.
12, 422), domitr (Aen. 12, 550), labr (georg. 3, 118), melir (georg. 4, 92) and Numitr (Aen. 5, 768).
Archaic vowel lengths may also have been intended in such verb forms as ert (Aen. 7, 174) and dabt (Aen.
10, 383), but deliberate archaisms by no means account for all cases of lengthened final syllables in Vergil.
see J ohnston 1897, 19-24.
74
Klopsch 1973, 73-74; the situation is different in high medieval poetry, especially rhyming hexameters, where
the feature is ubiquitous.
35

verbi sequentis vel in consonantem desinens excipiatur a littera vocali. Est enim
natura brevis in hoc:
Cuius onus leve est, cuius iuga ferre suave est (Sedul. carm. pasch. 1, 290).
Est longa permissu poetico in hoc:
Frondea ficus erat, cuius in robore nullum (Sedul. carm. pasch. 4, 46).
75

[The fourth kind of common syllable occurs when, after any metrical foot, one short
syllable of the word remains and is followed by a consonant in the next word or ends
in a consonant and is followed by a vowel. For example, the final syllable of cuius is
short in this verse:
Cuis onus leve est, cuius iuga ferre suave est.
(Whose burden is light, whose yoke is pleasant to bear).
It is long by poetic licence in this:
Frondea ficus erat, cuis in robore nullum.
(Leafy was the fig tree, in whose strength there was nothing).]
76


Bede does not bring the term caesura into the discussion, and his formulation of this metrical
rule, drawn from several equally tangled sources,
77
is undeniably complicated. It is also
slightly misleading: productio ob caesuram, at least in Vergil, almost invariably occurs in
final syllables that end with a consonant, and the latter case (words with a final short vowel
followed by an initial consonant) is unusual in classical verse,
78
although an opposite case
can be made, as I believe Bede has done: it has been noted that Vergil generally placed words
that, in the spoken Latin of the day, would have been subject to the brevis brevians law
before the central caesura, where the vacillating length of the final vowel would have made
no difference, hence Arma virumque cano / Troiae qui primus ab oris. Although we scan
can with a long o, Bede would most certainly have regarded can as the regular scansion, as
his chapter on the final syllables of verbs attests.
79
In Bedes own hexameter verse, where
many word-final vowels (mainly o in third-declension nouns and first-person verb forms) are
generally short, they can still be long in the arsis.
80


75
DAM 3, 53-59.
76
Trans. Kendall 1991, 53.
77
Maximus Victorinus (gramm. VI, 220, 1); Marius Victorinus (gramm. VI, 27, 12-13).
78
The most important exceptions in Vergils verse are liminaqu laurusque (Aen. 3,91) and Chloreaqu
sybarimque (Aen. 12, 363), both of which represent a formula where the first -que is routinely lengthened (as
in Brontesqu Steropesque, lappaequ tribolique etc.). Possibly spurious cases in Vergil include Dona
dehinc auro gravi sectoque elephanto (Aen. 3, 464), for which, however, the alternative reading gravi(a) ac
(with elision) has been suggested. The line appears in several grammarians (e.g. Servius, gramm. IV, 424, 27;
Sergius, gramm. IV, 479, 14) and has probably influenced Bedes presentation, although he does not mention it.
See Johnston 1897, 19-24; Housman 1927, 10-12.
79
DAM 7, 5-6.
80
J aager 1935, 17-18. For example, line 657 of Bedes Vita metrica Sancti Cuthberti (J aager 1935, 108),
Emisit, cred flammis ut mundior illis, has credo with o in the arsis. The classical scansion would be cred,
but, in reality, cred with productio ob caesuram may have been what Bede intended.
36

It is not surprising that Bede has, after his wont, Christianised the licence of
productio ob caesuram by substituting quotations from Sedulius for the usual Vergilian lines,
but then he unexpectedly goes on to condemn this practice, quoting Mallius Theodorus:
81


Quod genus syllabae inter longas vel omnino refugiendum vel parcissime
usurpandum est. Unde et in recentioribus poetis non facile eius invenies exemplum,
quamvis et apud Virgilium non rarissimum, apud Homerum vero frequentissimum
reperiatur.
82

[This licence should either be avoided entirely or it should be used very sparingly.
Hence you will not easily find an example of it among the more modern poets,
although it is found not infrequently in Vergil and very frequently indeed in
Homer.]
83


Bede has expanded Malliuss mild criticism of this metrical licence with his own dichotomy
of pagan and Christian techniques: although his first example is taken from Sedulius, whose
prosody Bede elsewhere seems to consider exemplary, he considers productio ob caesuram
generally typical of Vergil and even more characteristic of Homer (whom he certainly did not
know at first hand), and appears to argue that one of the metrical advances made in Christian
poetry was the avoidance of such constructions. We must remember that Bede may have
overestimated Vergils use of the device because, as we noted, Bede regarded several word-
final vowels as short, although they are long in classical verse. This might have led him to see
productio ob caesuram where none existed. All the same, Bedes views on this metrical
feature appear unusually inconclusive.
Recent studies of Bedes own verse indicate that Bede may, indeed, have been
of two minds regarding productio ob caesuram. In a manuscript of Bedes Vita metrica
Sancti Cuthberti known as the B redaction, which apparently is an early draft of the poem,
84

the licence is conspicuously common, appearing no less than thirteen times, which is a quite
impressive figure in a poem of about 1,000 lines. In the final, or vulgate version, all but one
of these have been suppressed.
85
This change may be attributed to Malliuss admonition that

81
gramm. VI, 587, 15-17.
82
DAM 3, 60-64.
83
Trans. Kendall 1991, 53.
84
See Lapidge 1995, 346-347. Lapidge conjectures that the B redaction was composed around 705 and the
vulgate version in the second decade of the eighth century.
85
Wright 2005, 154. Wright does not take into account that final vowels that in Bedes prosody although not
in classical prosody are normally short can appear as long syllables in the arsis. Jaager 1935, at pp. 19-20
concludes that the true frequency of productio ob caesuram in Bedes verse is difficult to establish: Wieweit
die Stellung in der Arsis Lngung bewirkt, ist nicht deutlich ersichtlich, da in den fraglichen Fllen die Quantitt
der Silben schwankt. The only clearcut case of productio ob caesuram occurs in line 892 (Jaager 1935, 127):
Quo sacer astra petens corps / exsangue reliquit (Where, reaching for the stars, the holy man left his
37

this licence should either be avoided entirely or used very sparingly. Apparently, in the B
redaction, Bede chose the latter course, but in the vulgate of the Vita Cuthberti the former.
Possibly this change was brought about by Bedes closer study of Christian poets and his
growing conviction that productio ob caesuram was one of the pagan licences he could
well do without.
As the second type of common syllable, Bede discusses one of the most
peculiar features of late antique prosody, the consonantal use of h, which probably came
about through a misunderstanding. As Dag Norberg sees it, this practice is based on the
model of Vergils line terga fatigams hasta, nec tarda senectus (Verg. Aen. 9, 618), where
the final syllable of fatigamus is scanned as long.
86
Generally this scansion is supposed to be
a case of productio ob caesuram, but several late antique grammarians attributed this to the
initial h of hasta, which correspondingly came to be treated as an ordinary consonant. The
line became a standard textbook example, and the consonantal use of h, ostensibly sanctioned
by its authority, spread to the technique of several Late Latin poets. It is remarkable that in
Late Latin usage, most markedly in the poetry of Venantius Fortunatus, a consonantal h not
only appears after a caesura, but also in the beginning of a foot (e.g. Ven. Fort. Mart. 3, 43:
angit in ancipiti, qud hoc aenigma beati). In some cases, it may even prevent elision (e.g.
Ven. Fort. Mart. 1, 86: servantur simul, ille fide, hic corpore vivens). Bedes examples of
consonantal h, however, taken from Sedulius, belong to the conservative type that is not too
different from productio ob caesuram: in both cases, the lengthened syllable is in the arsis of
the foot, and in the second, the h coincides with the trithemimeral caesura. Curiously, though,
both times the h creates a position after a monosyllable.
87


Item natura brevis syllaba ad votum poetarum transferri potest in longam, cum
correpta vocalis in consonantem desinit et excipitur ab h littera. Est enim natura
brevis in hoc:

bloodless body behind). The line is an allusion to Verg. Aen. 2, 542, which, in the usual reading, has
corpusque exsangue (with elision and without productio ob caesuram).
86
Norberg 1958, 7: Partant dun exemple comme Terga fatigamus hasta, nec tarda senectus, on avait formul
la rgle quune consonne suivie de h pouvait former une syllabe longue, rgle que les auteurs de la grammaire
apprenaient aussi bien la fin de lAntiquit quau Moyen Age, et que beaucoup de potes appliquaient -
Misguided as the grammarians were, there may have been a grain of truth in their mistake. Of the fifty-four
cases of productio ob caesuram in Vergils poetry, fourteen occur before a word-initial h, although initial h in
Latin is about ten times less frequent than initial vowels. Arguably, the articulation of initial h may have played
some role in Vergils prosody, albeit a marginal one. In all classical verse, hiatus is more common before an
initial h than before initial vowels, and vowels placed before hic and its derivatives are generally avoided by all
poets. - See Allen 1973, 148.
87
Ceccarelli 2006 (p. 206) has plausibly suggested that the monosyllable has probably contributed toBedes
analysis : Per quanto riguarda gli allungamenti, Bede (o la sua fonte) riconosce dunque lazione dellh iniziale
solo nei casi non giustificabili in base alla pratica virgiliana.
38

Porcinum tenuere gregem, niger, hispidus, horrens (Sedul. carm. pasch. 3, 84).
Est voluntate poetarum longa in hoc:
Vir humilis maesto deiectus lumine terram (Sedul. carm. pasch. 3, 296);
et item:
Mors fera per hominem miserum sibi subdidit orbem (Sedul. hymn. 1,69).
88

[Similarly, poets can choose to change a syllable which is short by nature into a long
syllable, when a short vowel ending in a consonant is followed by the letter h. It is
short by nature in this example:
Porcinum tenuere gregem, nigr, hispidus, horrens.
(They possessed a porcine herd, dark, shaggy, bristly).
But in this one the poet has chosen to make it long:
Vr humilis maesto deiectus lumine terram
(The humble man, with his sorrowful eyes cast on the ground):
And likewise:
Mors fera pr hominem miserum sibi subdidit orbem
(Savage death subjects the world to itself through wretched man).]
89


It is, in this context, unclear whether Bede sanctioned the consonantal use of h in all
positions, but it must be noted that in his first example, where the word-initial h does not
create a position, it is at the beginning of the fifth foot of the line, whereas the two other
examples, where such lengthening occurs, have an h in the middle of the foot. What is
surprising, however, is Bedes renunciation of the Vergilian example which, at least in
Norbergs view, was the indirect cause of the consonantal h in the first place:

Ubi item quidam grammaticorum dubium ponunt exemplum, terga fatigamus hasta
(Verg. Aen. 9, 618). Nam etsi h non sequeretur, mus tamen esse posset longa poetica
licentia, quia plenis pedibus superfuit, sicut hoc quod item ponunt:
Omnia vincit amor, et nos cedamus amori, (Verg. ecl. 10, 69)
ubi mor ideo potuit produci, quia post emensos pedes integros partem terminat
orationis, tametsi vocalis sequatur.
90

[Some grammarians put terga fatigams hasta (we worry the backs of the oxen
with the reversed spear) here. This is a dubious example because, even if h did not
follow it, the syllable mus could be long by poetic licence, since it is a final syllable
and follows two full metrical feet, just as in another example, which they also give:
Omnia vincit amr, et nos cedamus amori
(Love conquers all, let us also yield to love),
the syllable mor was able to be lengthened even though a vowel follows it, because it
comes at the end of a word and follows two full metrical feet.]
91



88
DAM 3, 13-20.
89
Trans. Kendall 1991, 47.
90
DAM 3, 21-27.
91
Trans. Kendall 1991, 47.
39

Although Bedes terminology is here more than usually convoluted (quia plenis pedibus
superfuit, literally because it is left over after complete feet), it is clear that he is referring
to (Vergilian) productio ob caesuram, where a word-final short syllable is interpreted as long
before a caesura and that he sees this as distinct from the (Christian) use of consonantal h,
which by his time had become an established metrical feature of Christian poetry. This
demonstrates both his analytical mind but also his unwillingness, also perceptible elsewhere,
to use Vergil as an illustration when he had Christian examples at his disposal.
Bede also broaches the consonantal use of h in his extensive chapter on elision
(De synalipha),
92
a technique that apparently was very hard to grasp for Anglo-Saxons.
Discussing word-final m, a particularly thorny issue, he suggests that initial h may prevent
elision after it, although this is up to the discretion of the poet. Characteristically, this seems
to be the only case where Bede actually approves of hiatus, which he otherwise deems an
outdated device, and this, too, is due to the authority of the Christian poets Juvencus and
Sedulius, whom he also quotes on the subject:

Quaecumque ergo verba in m terminantur, nisi adpositione consonantis alicuius
defendantur, synalipha inrumpente syllabam ultimam aut perdunt semper aut
minuunt, excepto cum ab h littera sequens sermo inchoaverit. Tunc et enim in arbitrio
poetarum est, utrum haec instar fortium consonantium synalipham arceat, an pro
modo suae fragilitatis nihil valeat. Valuit namque in hoc, quia voluit poeta:
Nomine Iohannem hunc tu vocitare memento (Iuvenc. 1, 26);
et:
Progenitum fulsisse ducem, hoc caelitus astra (Sedul. carm. pasch. 2, 77).
Item nihil iuvit ad propellendam synalipham, quia poeta neglexit:
Qui pereuntem hominem vetiti dulcedine pomi (Sedul. carm. pasch. 1, 70).
93

[Every word which ends in m will always either lose its final syllable or have it
reduced by the action of elision, unless elision is blocked, which happens when the
immediately following word begins with any consonant except h. In that case, it is up
to the judgement of the poets as to whether the h should prevent elision after the
fashion of the stronger consonants, or should have no force on account of its
weakness. It has the force of a consonant, since the poet willed it, in this example:
Nomine Iohannem / hunc tu vocitare memento
(Remember to call his name J ohn);
and in
Progenitum fulsisse ducem, / hoc caelitus astra
(The Son and king has appeared, the star from Heaven witnesses it).
But it has no power to block elision, because the poet disregarded it, in this example:
Qui pereunt(em) hominem vetiti dulcedine pomi
(You who restore mankind, damned by the sweetness of the forbidden fruit).]
94


92
DAM 13.
93
DAM 13, 31-43.
94
Trans. Kendall 1991, 117.
40


In Bedes metrical Vita Sancti Cuthberti, h generally does not create a position. As Werner
J aager has noted, most of the cases where a short syllable preceding an h is treated as a long
one, the short syllable coincides with the arsis of the foot, and lengthening can be attributed
to the following caesura.
95
Given Bedes ambiguous views on productio ob caesuram, we
cannot really say what he had in mind. There are, however, two cases of consonantal h which
cannot be attributed to productio ob caesuram: in both of them, the h is at the beginning of
the foot. Conspicuously, in both of the cases h is preceded by a monosyllable (sed), as in the
Sedulian examples he cites in his treatise:

Servatur sd haec puero victoria lecto
96

[But this victory is saved for the chosen youth].
Acta rotant! nullo noceor sd hostis ab ictu
97

[However often they roll moving rocks at me, I am unharmed by any blow of the
enemy].

Characteristically, Bede has made full use of the poetic licence which he endorses in his
presentation of the letter h: his Life of Cuthbert also has sed hostis without lengthening of
sed:

Propinasque tuis praetristia musta, sd hostis
98

[You serve the sorrowful wine to your people, but the enemy...].


2.2.4. Hiatus and correption

Hiatus in metrics can best be defined as absence of elision. One of Bedes most extensive
chapters (DAM 13, De synalipha) deals with the intricacies of elision and when to do it;
hiatus, on the other hand, is also touched on in his chapters on common syllables
99
and the
differences between old (pagan) and new (Christian) poets.
100

Hiatus in Latin poetry is not a very common feature. In the classics it is
generally considered a stylistic Graecism, and it often appears in conjunction with Greek

95
J aager 1935, 19.
96
J aager 1935, 66, line 107.
97
J aager 1935, 93, line 473.
98
J aager 1935, 119, line 792.
99
DAM 3, 65-96.
100
DAM 16.
41

names or other quasi-Greek metrical devices, such as unusual caesurae or spondaic lines (e.g.
Verg. georg. 1, 437: Glauco / et Panopaee / et Inoo melicertae). Licences of this nature
usually echo the verse of Homer (where hiatuses are sometimes residues of word-initial
digamma) or the Alexandrians, and, in some cases, can be their direct paraphrases.
101
In
Vergil, unelided final long vowels or diphthongs are sometimes shortened by the process
known as correption, which is a further Graecism (e.g. Verg. Aen. 3, 211: insulae Ionio in
magno, where the ae of insulae is scanned as short).
Considering that hiatus by itself is an unusual feature and correption arguably a
mere curiosity, Bedes thorough exposition of these features may seem unexpected. Bede,
however, had reasons for his interest in these phenomena: firstly, as hiatus is the opposite of
elision, which apparently was one of the most difficult verse techniques for the Anglo-
Saxons, Bede had to make plain how final vowels should not be handled. Secondly, hiatus is
one of those metrical licences which (together with spondaic verses) Bede associates most
strongly with pre-Christian literature, and Bede has expanded his presentation of hiatus and
correption with a very partisan discussion of the differences between old poets (veteres),
essentially Vergil, and our poets (nostrates), the Christians. Apart from his excursus on
pre-Christian and Christian practices, Bedes presentation is primarily drawn from Servius
102

and Sergius,
103
but supplemented with Donatus.
104

The fifth and sixth types of common syllable in Bedes treatise are cases where
correption either does or does not occur in a final syllable before hiatus, the fifth type being
final diphthongs and the sixth, long vowels, a distinction which grammarians, Bede included,
tended to give undue prominence. Bede makes it clear from the start that hiatus, with or
without correption, is an outdated technique best avoided. Bede shows some originality in
extending the rule of correption to word-internal cases: having denounced both hiatus and
correption of final vowels as obsolete, he goes on to purport that modern (i.e. Christian)
poets only use correption inside a word. This claim may seem surprising, as, in Latin, vowels
followed by another vowel are almost invariably short, word-internal correption having taken
place already in archaic Latin. Apparently, this is an attempt to rationalise the variable length
of some word-internal vowels: in his chapter on elision, Bede uses what he perceives as
word-internal correption as the explanation for such prosodic doublets as unius/unius and

101
Raven 1965, 24.
102
gramm. IV, 424, 29-34.
103
gramm. IV, 479, 15-17.
104
Klotz 606, 12-13.
42

fidei/fdi.
105
Here Bede uses the example of Maria, always an important name for Christian
poets, although he also discusses the e of Eous (of the dawn, eastern, or, as a masculine
noun, morning star), again using Sedulius as an example. Even in classical verse, the length
of e in Eous fluctuates, being generally short in the end of the line but long elsewhere,
106
and
Sedulius has followed suit:

Item modus est quintus, cum pars orationis desinit in diptongon sequente statim altera
vocali. Est enim per naturam longa in hoc, Musae Aonides (Cornelius Severus?).
Est per licentiam breuis in hoc, insulae Ionio in magno (Verg. Aen. 3, 211), quod
posteriores poetae magis in una parte orationis fieri voluerunt. Unde nec huius
exemplum facile in nostratibus poematibus invenies. At cum diptongus a vocali
alterius verbi excipitur, tum hanc per synalipham transiliendam esse dicebant, ut
Prosper in praefatione epigrammatum:
Nec nostrae hoc opis est, sed ab illo sumitur hic ros,
qui siccam rupem fundere iussit aquas (Prosp. epigr. praef. 7-8).
Sextus modus est, ut Donatus ait, cum producta vocalis est vocali altera consequente.
Est enim longa in hoc:
O utinam in thalamos invisi Caesaris issem (Lucan. 8, 88).
Brevis in hoc:
Te Coridon, o Alexi, trahit sua quemque voluptas (Verg. ecl. 2, 65).
Quod moderni versificatores in eadem potius parte orationis facere consuerunt, ut
Eoi uenere magi (Sedul. carm. pasch. 2, 74) et:
Splendidus auctoris de vertice fulget Eous (Sedul. carm. pasch. 5, 191).
Et rursus longa est per naturam ita:
Angelus intactae cecinit properata Mariae (Sedul. carm. pasch. 2, 36).
Brevis per licentiam ita:
Exultat, Mariae cum prima adfamina sensit (Iuvenc. 1, 91).
At cum longa vocalis vel etiam brevis, quae partem terminat orationis, excipitur a
vocali alterius uerbi, priorem per synalipham absumunt, ut Prosper:
Nam si te virtute tua ad caelestia credas
scandere, de superis pulsus in ima cadis (Prosp. epigr. 31, 3-4).
Quamvis et Arator imitatus veteres dixerit:
O utinam nostris voluisses fida iuventus
consiliis parere prius, ne litora Cretae
linqueres insani rabiem passura profundi (Arator 2, 1107-1109).
107

[Similarly, the fifth kind of common syllable occurs when a word ends in a diphthong
and is immediately followed by a word beginning with another vowel. For example,
the final syllable of Musae is long by nature in Musae Aeonides (the Heliconian
Muses). But the final syllable of insulae is short by poetic licence in insulae Ionio
in magno (islands in the great Ionian sea). Later poets have preferred to transfer
this licence to the interior syllables of a word, so that you will not easily find an
example of this in Christian Latin poetry. Indeed, the grammarians have generally

105
DAM 13, 45-57.
106
cf. dies primo surgebat ous (Verg. Aen. 1, 288); terras irrorat ous (Verg. georg. 1, 288), but oas
acies et nigri Memnonis arma (Verg. Aen. 1, 489).
107
DAM 3, 65-96.
43

stated that a diphthong followed by another word beginning with a vowel should be
elided, as Prosper does in the Preface to his Epigrams:
Nec nostr(ae) hoc opis est, sed ab illo sumitur hic ros,
qui siccam rupem fundere iussit aquas
(This is not of our power, but this dew is consumed by him,
who commanded the dry rock to issue forth the waters).
The sixth kind of common vowel occurs, as Donatus says, when a long vowel is
followed by another vowel. For instance, the interjection o is long in this verse:
utinam thalamos invisi Caesaris issem
(Oh, would that I had married hateful Caesar);
and short in this:
Te Coridon, Alexi, trahit sua quemque voluptas
(Coridon pursues you, oh Alexis, his own pleasure leads each one on).
Modern poets prefer to exercise this licence within the same word. So the first
syllable of Eous is long by nature in oi venere magi (the Magi came from the
east) and short by poetic licence in
Splendidus auctoris de vertice fulget ous
(The brilliant morning star shines from the head of the Maker).
And again the penultimate syllable of Maria is long by nature in this case:
Angelus intactae cecinit properata Marae
(The Angel prophesied to the Virgin Mary about things to come),
and short by poetic licence in this one:
Exultat, Marae cum prima adfamina sensit
(The offspring of Elizabeth leaped in her womb after he heard the first
salutation of Mary).
When either a long or short vowel which comes at the end of a word and followed by
a vowel at the beginning of the next word, poets elide the former, as Prosper does in
Nam si te virtute tu(a) ad caelestia credas
scandere, de superis pulsus in ima cadis
(For if you think you are climbing to heaven by your own power, driven from
above you fall into Hell).
However, Arator, imitating the ancients, said:
O utinam nostris voluisses fida iuventus
consiliis parere prius, ne litora Cretae
linqueres insani rabiem passura profundi
(Oh, would that you, faithful youth, would obey our counsel rather than leave
the shores of Crete to suffer the rage of the stormy deep).]
108


The end of the passage is striking: Bede presents a quotation from Arator, where the poet
(quite in keeping with standard metrical practice) has a hiatus following the interjection o.
Leaving aside the question of how exactly Bede expects o to be elided, his interpretation
shows his firm belief in the existence of two distinct poetic techniques: pagan and Christian.
In Bedes opinion, Arators hiatus is simply attributable to his imitation of pre-Christian poets
and, as such, does not truly correspond with the requirements of up-to-date Christian verse,

108
Trans. Kendall 1991, 53-57. Kendalls translation has Christian Latin poetry where Bede simply has in
nostratibus poematibus, but Christianity, of course, is implied.
44

where all final vowels are elided and correption only takes place within the confines of a
single word. In other words, Bede recognises the possibility of using archaising devices in
poetry, but only in a Christian context as an allusion to pre-Christian poetry. That hiatus was,
even in classical verse, an archaising or Graecistic device never seems to have entered Bedes
mind, no doubt owing to his negligible knowledge of Greek literature. Bedes bias against
pre-Christian verse technique remains, to all appearances, remarkably unfair: Christian poets
like Arator seem to have remained in his good graces, whatever they did, whereas there was
no excuse for abnormal prosody in pagan verse.


2.2.5. Hic and hoc

As the seventh type of common syllable, Bede discusses the quantitatively problematic
demonstrative pronoun hic (nominative singular) and its neuter form hoc, both of which are
generally scanned with a geminated final c (hicc, hocc) before vowels, making them, in
classical prosody, almost always long by position. This classical practice is the result of
analogy: hic and hoc evolved from the combinations of the pronouns *ho (masculine) and
*hod (neuter) and the deictic particle -ce, through the intermediate forms hice (with a
weakened vowel) and hocce (with assimilation of d with c), both of which have been attested
in Plautus and Terence.
109
By analogy, in the syncopated forms hic and hocc, the geminated c
of the neuter form also spread to the masculine form hic, and classical prosody treats both
forms as heavy syllables ending in a double consonant. Historically, the pronouns hic and hoc
have proved puzzling to many generations of scholars. As W. Sidney Allen has noted, their
scansion as long has occasionally led to confusion as to their vowel quantity, and older
dictionaries have often given them the historically incorrect pronunciations hic and hc.
110

Apparently, hic and hoc were considered difficult already in late antiquity: Marius Victorinus
emphasises that they are truly long syllables and warns his readers not to confuse this with
productio ob caesuram:

Ergo illae non ideo longae fiunt quod pars orationis finitur, ut putant plurimi (nam
idem et in hac effici posset), sed, ut dixi, in pronominibus c littera sonum efficit

109
Allen 1978
2
, 76-77.
110
Allen 1978
2
,75-76.
45

crassiorem et naturam i litterae inter vocales positae ac per hoc sonum geminantis
imitatur.
111

[Therefore, they are not long because they terminate a part of speech, as most believe
(for this could also be possible), but, as I said, in pronouns the letter c is pronounced
more fully and resembles the quality of the geminate i which occurs between vowels.]

Occasional departures from this norm in classical poetry are rare and usually deliberate
archaisms that reflect pre-classical pronunciation (e.g. Verg. Aen. 4, 22: solus hc inflexit
sensus animumque labantem). Textbook examples of this kind inevitably led later-day
grammarians astray, including Bede, who, quoting Sergius and Servius,
112
classifies
pronouns ending with the letter c as true common syllables:

Septimus modus est, cum pronomen c littera terminatum vocalis statim sequitur. Est
enim longa in hoc:
Non quia summus Pater est, et Filius hic est,
sed quia quod summus Pater est, et Filius hoc est. (Sedul. carm. pasch. 1, 319-
320)
Brevis in hoc:
Hic vir, hic est, tibi quem promitti saepius audis. (Verg. Aen. 6, 791)
113

[The seventh kind of common syllable occurs when a word beginning with a vowel
immediately follows a pronoun ending in the letter c. For example, the pronoun hic is
long in this verse.
Non quia summus Pater est, et Filius hic est,
sed quia quod summus Pater est, et Filius hoc est.
(Not that the one who is the supreme Father is also the Son, but that that which
is the supreme Father is also the Son).
It is short in this one:
Hic vir, hc est, tibi quem promitti saepius audis
(This, this is the man whom you have heard promised so often).]
114


It is not surprising that Bede has, once again, Christianised his material by choosing a
quotation from Sedulius, his favourite, to illustrate the usual case of hic as a long syllable,
although, in demonstrating the less usual case of hic as a short syllable, he has been
compelled to rely on Vergils example. In this case, Bede has, perhaps surprisingly, not
recognised the latter for what it really is, an archaism, and, for this reason, this peculiarly old-
fashioned feature has escaped his condemnation: for Bede, as for Servius, hic and hoc are
truly common (rather than long) syllables. The implication is that, at least in theory, also hoc

111
Mariotti 1967, 95, 23-27.
112
Sergius, gramm. IV, 479, 20-22; Servius, gramm. IV, 424, 34-36.
113
DAM 3, 95-102.
114
Trans. Kendall 1991, 57.
46

could be a scanned as a short syllable, although Bede, understandably, has provided no
example of such a case.
115

Bede follows up his slightly flawed exposition of hic and hoc with another
analogy, which is based on a possibly faulty reading of Paulinus: Bede contends that also
adverbs ending in a c can also have a common final syllable:

Sed et adverbium c littera terminatum communem syllabam facit. Est enim longa in
hoc Paulini:
Donec adspirante Deo conatibus aegris (Paulin. 15, 299).
Brevis in hoc Prosperi:
Ut morbo obsessis praestanda est cura medendi,
donec in aegroto corpore vita manet (Prosp. epigr. 112, 1-2).
116

[But an adverb ending in the letter c also makes a common syllable. The final syllable
of donec is long, for example, in this verse of Paulinus:
Donc adspirante Deo conatibus aegris
(Until with the help of God for the struggling sick),
It is short in this verse of Prosper:
Ut morbo obsessis praestanda est cura medendi,
donc in aegroto corpora vita manet
(So those afflicted with this disease should be cared for as long as life remains
in their sick body).]
117


On the strength of his reading of Paulinus 15, 299, Bede has assumed that the final c of
adverbs and conjunctions such as donec has qualities similar to those of hic and hoc.
Although the usual reading of Paulinuss line is Donec et adspirante, rather than Donec
adspirante, Bede is here touching on an existing phenomenon that played a genuine role in
the prosody of Late Latin verse; donec (with long final syllable) has been documented at least
in Venantius Fortunatus (carm. 9, 2, 65 donc adventum). This, however, is not attributable
to any special qualities of the final c, but, rather, to the fact that donec, together with some
other particles and adverbs, may have been pronounced with a oxytone, or word-final, accent
(donc), which, in turn, led to the lengthening of the final syllable.
118
These, however, are

115
Not all scholars shared Bedes over-generalisation. Marius Victorinus notes dismissively: Pro brevi autem
hic vir, hic est. Sed pro brevi bis fere tantum. (It is short in hic vir, hic est, but it is short in hardly more
than two cases.) Mariotti 1967, 95, 13-14.
116
DAM 3, 103-108.
117
Trans. Kendall 1991, 57.
118
Townend 1950, 33. Priscian (gramm. III, 74, 11-16) asserts a penultimate accent for donec but not for other
adverbs ending in c:
In c primitiva inveniuntur sc, donc (quod etiam donicum antiqui dicebant...) et derivativa
pronominum illc, istc, hc, illc, istc, hnc, illnc, istnc, hc, istc, illc.
[Original adverbs with c include sc and donc (which the ancients also called donicum), and the
pronominal derivatives illc, istc, hc, illc, istc, hinc, illinc, istinc, hc, istc, illc.]
47

highly contested issues of which Bede was not aware, and it seemed logical to him to
attribute the quantity donec to the fact that it, like hic and hoc, ended with a c. The role of
such forms as hic and illic, where the vowel is long, may also have affected Bedes ruling by
analogy.


2.2.6. Summary

As we have witnessed, Bede regarded the Christian writers of late antiquity with a very
thinly-veiled favouritism. Bede has demonstrated nearly all of the rules pertaining to common
syllables with quotations from Christian authors, with Sedulius a clear favourite; one could
almost surmise that, as the poet to be emulated, Sedulius had, for Bede, become something
like a Christian Vergil. Bede has not eliminated Vergilian examples altogether, but it is
clear that he often resorts to them for want of a better, or, that is to say, a Christian example,
as in the cases of word-initial z and x. There are some phenomena which he presents with
exclusively pre-Christian material, but often he portrays them as outdated, pagan
techniques, the employment of hiatus and correption being a case in point.
Bedes Christian bias is also reflected in the manner in which he evaluates the
various ways in which poets have treated common syllables and the respective prosodic
liberties they have taken. Although he does not express this explicitly, there appear to be two
different prosodic practices for him: pagan and Christian. Prosodic liberties that occur in the
classics but were later given up are, in his eyes, examples of archaic unruliness and, at least
implicitly, he appears to attribute the subsequent improvements in poetic technique to the
onset of Christianity. Conversely, he regards many of the unusual or unclassical prosodies of
the late antique Christian poets, above all Sedulius, most favourably. Bedes list of common
syllables is more than merely an expanded and Christianised version of Sergius and
Pompeius: it can be read as a veiled attempt to codify the prosodic practices of Christian
poets. This tendency, which is only implicitly expressed in Bedes chapter on common
syllables, is later formulated unequivocally in his chapter on the differences of pagan and
Christian poets.
119


The pronunciation donc is apparently based on analogy: the oxytone accent of such words as illc and istc
evolved from the penultimate accent of the unabbreviated forms illce and istce, whereas donec (probably from
dnicum) is prosodically different. See also Allen 1978
2
, 87.
119
DAM 16.
48

A summary of common syllables and their proper Christian usage, as
understood by Bede, can be presented as follows:
1) The combination of plosive and liquid (or f and liquid) can create a position
word-internally but never in an initial position. This corresponds with both classical and
Christian practice, although some deviations from the latter rule appear in Vergil, a feature
which Bede condemns as outdated elsewhere in his treatise. Bede ignores evidence of the
same usage in the Christian poets Venantius Fortunatus and Aldhelm.
2) The combination of initial s and another consonant can create a position
word-initially, a patently post-classical feature that occurs in Sedulius. Conversely, Bede
suggests that similar s groups may not necessarily make a position word-internally, a feature
that is virtually non-existent in all Latin poetry (apart from Aldhelm, who makes copious use
of it). Bedes assumption, which must be considered erroneous, is based on the faulty
analysis of one single line in Vergil, although it is not impossible that his analysis has been
influenced by Aldhelms prosody.
3) Bede regards z fundamentally as a single consonant (although, out of
adherence to traditional terminology, he calls it a double letter). It can, however, create a
position word-internally, if the poet so chooses. Bedes presentation is almost the opposite of
classical practice, where z always makes a position word-internally and is only treated as a
single consonant initially in prosodically difficult words.
4) Bedes views on productio ob caesuram, where a short final syllable can be
interpreted as long before a caesura are ambiguous and inconclusive: the phenomenon
appears in Sedulius, so Bede cannot wholly discredit it. On the other hand, Bede seems to
regard it as more typical of Vergil and therefore potentially old-fashioned and not altogether
to be recommended.
5) Bede condones the consonantal use of initial h, which, if the poet so chooses,
can create a position after a short final syllable. As he states elsewhere in his treatise, an
initial h can also prevent elision from taking place, although apparently only after a final m.
This ruling reflects a Christian practice that probably evolved through the grammarians
faulty analysis of Vergils terga fatigams hasta (more properly a case of productio ob
caesuram).
6) All other forms of hiatus are archaic and best avoided. Hiatus, if encountered
in Christian poetry, must be understood as an allusion to pre-Christian verse.
7) Hic and hoc are, in Bedes nomenclature, common syllables, although in
classical verse they are almost always long. This generalisation, which Bede shares with most
49

of his predecessors, is based on Vergils archaising use of short hic (Aen. 6, 791), which had
become a textbook commonplace. Relying on what appears to be false analogy, Bede also
makes the suggestion that the final syllable of donec (and, by implication, other adverbs
ending with a c) can be common.

As we can infer from this presentation, Bede had studied both Vergil and the Christian poets
known to him very thoroughly indeed, obviously in the belief that there existed an ideal kind
of Christian prosody distinct from pagan practice. Bedes prime model for contemporary
verse appears to be Sedulius, whose example he generally follows even in its idiosyncrasies.
Sometimes his willingness to generalise and synthesise has led him astray, and he seems to
give undue prominence to the example of individual lines (some of his Vergilian examples
are genuine rarities even in Vergils verse). But, arguably, even this fault may have served to
make his delineation of Christian verse technique all the more persuasive.
The final type of common syllable, the final syllable of a line, is something
which Bede discusses only cursorily, as this was hardly an issue that was expected to confuse
readers, or where stylistic issues played a role.


2.3. Other observations on prosody

Bedes chapters on syllables and their respective lengths are lengthy and detailed, and very
demonstrative of the techniques that were employed in the teaching of classical prosody in
the Middle Ages. They are, however, also highly derivative in their dependence on previous
authors, and generally concur with classical prosodic practice, and, as they mainly consist of
catalogues of words or their endings, they make wearisome reading. There are, nonetheless,
some instances where Bede departs from classical norm and recommends scansions that are
more in keeping with the prosody of the Late Latin Christian authors, mainly Sedulius, and it
is these cases that we must discuss briefly.
Apart from the common syllables, Bedes departures from classical syllable
quantity are generally restricted to the final syllables of words, usually final vowels that are
long in the classics but often shortened in post-classical verse. Such variation also often
occurs in words with a final s. Most typically, the final os of third-declension nouns and
first-person verb forms are shortened, a feature that occurs already in the Silver Age poets,
50

even in words that do not adhere to the brevis brevians law.
120
Bede is content to mention
such prosodic features in passing; apparently, by his time they had become established to
such an extent that Bede took them for granted, and did not deem them worthy of further
analysis.
Bede discusses the final syllables of words in his sixth, seventh and eighth
chapters, titled De ultimis syllabis nominum, pronominum, participiorum (On the final
syllables of nouns, pronouns and participles), De ultimis syllabis verborum et adverbiorum
(On the final syllables of verbs and adverbs) and De regulis syllabarum coniunctionum,
prepositionum, interiectionum (On the rules governing the syllables of conjunctions,
prepositions and interjections), respectively. These chapters are based mainly on Serviuss
De finalibus and Maximus Victorinuss De finalibus metrorum, and Bede largely quotes them
verbatim. Needless to say, they are also fairly consistent with classical prosody. Bedes
unusual lack of independence in these chapters is demonstrated by his inclusion of a Horatian
quotation (from Ars poetica 65); elsewhere in his treatise, Horace, even more than Vergil, has
been thoroughly supplanted by Christian lines.
121

At the very beginning of his sixth chapter Bede expresses his approval for a
number of short final vowels that are typical of post-classical poetry, starting with the Silver
Age poets. The chapter starts with a long catalogue of syllables, taken almost word for word
from Servius, where Bede states that the final o of third-declension nouns and the final u of
fourth-declension neuters are short: Nominativus singularis has habet breveso, ut ordo,
virgo; u, ut cornu
122
Bede, as Servius, seems to consider the short forms virg and corn
to be the normal ones (Servius only gives virgo and cornu,
123
and ordo appears to be Bedes
own addition). Remarkably, none of these words are subject to the brevis brevians law, and
therefore represent demonstrably post-classical prosody. Bede does not even mention the
possibility of a long scansion for the final syllables of these word-types.

120
Raven 1965, 23.
121
DAM 6, 35-38:
US, cum in genetivo crescente longa permanserit, producitur, ut virtus virtutis, tellus telluris,
excepto uno palus, quod in genetivo dis terminatur, palus paludis, unde est haec: sterilisque diu
palus aptaque remis.
[Final us is long in imparisyllabic nouns when the vowel remains long in the genitive, as, virtus
virtutis and tellus telluris, with the one exception of palus , which takes the ending dis in the genitive:
as for example, sterilisque diu pals aptaque remis (and the sterile marsh once fit only for boats).]
Trans. Kendall 1991, 77. The quotation from Horace is one of only four in Bedes entire output
identified in Michael Lapidges The Anglo-Saxon Library (2006), 212.
122
DAM 6, 1-5.
123
gramm. IV, 451, 28.
51

In his chapter on the final syllables of verbs and adverbs, De ultimis syllabis
verborum et adverbiorum, Bede presents some prosodic rules that deviate similarly from
what we consider to be the classical standard. Most noticeably, Bede presents the final o of
first-person verb forms as short, although the rule is presented somewhat more ambiguously
than the rule governing the final os of nouns: In omnibus o corripitur, ut amo, sedeo,
cerno, nutrio, tametsi auctoritas variet.
124
(In all conjugations final o is short, as amo,
sedeo, nutrio, cerno, although learned opinion differs on this point.)
125
The ruling, together
with this minor reservation, is based on Maximus Victorinuss De finalibus metrorum,
126
and
shows that something that had been a prosodic liberty in Silver Latin had, by Bedes time,
become the norm. Bedes own metrical practice is generally consistent with his unclassical,
or postclassical, rulings on final vowels. In his Vita metrica Sancti Cuthberti, short final o
seems to be the rule for both third-declension nouns and first-person verb forms as in:

Regia virg venit, regis quae sponsa perennis
127

[The royal maiden arrived, who had long been betrothed to the king].
Tu, rog, summe, iuva, donorum spiritus auctor
128

[I beseech Thee, o highest, help me, the creating spirit of gifts].

Deviation from this practice only occurs in the arsis of the foot (before a strong caesura) and
could equally well be attributed to productio ob caesuram, as in:

Audet adhuc supplex virg / pulsare propheten
129

[Still the kneeling virgin dared to touch the prophet].
Dic ergo, adiur / summi per regna tonantis
130

[Say, therefore, I vow by the realms of the highest God].

Such variation is by no means unusual in post-classical poetry, and these examples may seem
highly trivial. There is, however, one aspect that we must take into consideration: Bede
employs this variation very logically, and the logic is fully understandable if we bear in mind
his own teachings on metre: if, indeed, virg, ord, cern, corn etc. were what Bede, by his
own admission, considered the normal pronunciation, nearly all cases to the contrary, with a
final long vowel in the arsis of the foot, may constitute cases of productio ob caesuram. It is,

124
DAM 7, 5-6.
125
Trans. Kendall 83.
126
gramm. VI, 231, 20-21.
127
J aager 1935, 95, line 495.
128
J aager 1935, 61, line 35.
129
J aager 1935, line 518.
130
J aager 1935, line 504.
52

therefore, apparent that Bedes definition of productio ob caesuram, along with his
exhortation to use it sparingly, needs to be re-examined with this feature of his prosody in
mind. As Jaager has suggested, we may already see traces of high medieval caesural practice
in Bedes verse and its treatment of word-final syllables.
131
As we know, in some later
medieval grammars penthemimeres , the term for the strong caesura of the third foot of a
hexameter line, often came to be understood as synonymous with productio ob caesuram,
132

and, especially in Leonine hexameter verse with penthemimeral rhyme, the rhyming syllable
that precedes the caesura is commonly indifferent.
133
Although Bedes unclassical but
systematic treatment of word-final syllables is still more classical than that of the high
medieval poets (Bede does not shorten the long final vowels of datives or ablatives, nor are
the final vowels of first-declension nouns scanned as long in his verse), his definitions,
together with contemporary verse, probably contributed to the evolution of medieval prosodic
practice.
It must be noted that, in Bedes Vita metrica Sancti Cuthberti, similarly
consistent variation in the final syllables (long in the arsis, short elsewhere) occurs also in
other words, some of which normally have a short final vowel even in classical verse: the
most important of these are the pronouns ego, mihi, tibi, sibi, cui.
134
In his De arte metrica,
Bede does not mention ego but implies, albeit very obliquely, that it is analogous to nouns
ending with o; the final is of mihi, tibi and sibi are by Bedes own admission variable:

Dativus vero, sicut in nomine, semper longus est, excepto mihi, tibi, sibi, quae
indifferenter poni possunt. Sic reliqui quoque casus regulam sumunt ex nomine.
135

[The final syllable of the dative, just as in the noun, is always long, with the
exception of mihi, tibi and sibi, which can be treated as common. And in the same
way the remaining cases take their rule from the noun.]
136


Bedes presentation is taken verbatim from Servius (gramm. IV, 453, 25 454, 2). As we see,
Bede does not discuss ego, but the suggestion that the remaining cases of pronouns take
their quantities from noun cases would suggest an analogy with nouns with final o. Bede
seems to rely frequently on such analogy, especially where it seems corroborated by verse
technique.

131
J aager 1935, 20.
132
Klopsch 1972, 75.
133
e.g. Carmina Burana 5.8: Ordo, pudicitia, / pietas, doctrina, sophia (Order, modesty, piety, learning and
wisdom). Also note the short o in ordo, as well as the Italianate prosody in sopha.
134
J aager 1937, 17.
135
DAM 6, 67-69.
136
Trans. Kendall 1991, 79.
53

There are two phenomena which we must finally discuss, possibly of lesser
consequence than Bedes rulings on final os but nevertheless illustrative of his stubborn and
sometimes misguided attempts at regularisation: firstly, in his chapter on the final syllables of
nouns, adjectives and pronouns, Bede presents an idiosyncratic analysis of fifth-declension
nouns. Bede states that their genitive and dative endings should have both a long e and a long
i:

Ubi notandum est quod nomina quintae declinationis, quae in ei litteras genetivo et
dative casu terminantur, et has divisas et utramque longam habent, ut faciei, diei,
fidei.
137

[It should be noted here that in nouns of the fifth declension, which end in the letters
ei in the genitive and dative case, the e and i should be pronounced separately and
both are long, as faciei, diei and fidei.]
138


Bedes ruling on this matter does not correspond with usual classical prosody, where fifth-
declension nouns only have a long e in the genitive and dative forms when the preceding
letter is an i: strictly speaking, faciei and diei are the normal classical forms, but fidei, where
e is preceded by a consonant, is not. This confusion is probably attributable to the shaky
prosody of fifth-declension nouns in Late Latin. The form fidei (with long e) is archaic, and it
appears in the poetry of Ennius, Plautus and Lucretius, but it also reappeared in Late Latin
and was common in the early Middle Ages, as several hymns testify.
139
There apparently was
a deep-set tendency for the two types of genitive (diei and fidei) to affect each other by
analogy, as in Late Latin we also encounter the opposite case dii. Bedes presentation of
fifth-declension nouns as always having a long e in the genitive is a hypercorrect
generalisation, certainly affected by Late Latin verse but possibly also a reaction against such
forms as dii. Remarkably, Bede does not observe his rule in his own verse, where fdi is the
norm (e.g. line nine of Vita metrica Sancti Cuthberti: ut cunctum nova lux fidi face fusa
sub axem).
140
Bede does, however, give a thorough explanation for this variation in his
chapter on elision (De synalipha): he states explicitly that although, as he sees it, fidei is the
normal form, fidi is also possible by way of word-internal correption of the type he has
previously described in his chapter on common syllables.
141


137
DAM 6, 46-48.
138
Trans. Kendall 1991, 77.
139
Norberg 1958, 12. The prosodies are easily identifiable even in rhythmic verse, as they are accentuated
differently: fdei and fidi.
140
J aager 1935, 58.
141
DAM 13, 44-57
54

At the very end of his chapter on the final syllables of verbs and adverbs, Bede discusses the
numeral adverbs and departs from classical norm in ruling that the e in the suffix -ies is short:
es in numeris corripitur, ut toties, quoties, septies, decies.
142
Although such
shortened final syllables are by no means uncommon in Late Latin verse, the ruling is
curious, as we know not only that the e is long in classical prosody,
143
but that numeral
adverbs also have the older alternative spelling -iens (totiens, quotiens, septiens, deciens etc.),
where the final syllable is long even by position. This was by no means unknown to Bede,
who contests the alternative spelling in his normative guide to orthography, De orthographia:
Quoties, toties, septies sine n.
144
The treatment of such words does not seem to have
been entirely consistent even in classical antiquity: in inscriptional evidence, the suffix -ies
seems to have coexisted with the older -iens at a fairly early date (this is analogous to such
forms as mesibus/mensibus, cosul/consul etc.), although this seems to have applied mainly to
the definitive numerals like quinquies, sexies and septies.
145
On the subject of quotiens,
totiens and other indefinitives, the grammarians generally seem to have been adamant in their
defence of -iens.
146
Bede, however, appears to have gone out of his way to rule out all
scansions where the final syllable of numeral adverbs is long; furthermore, he does not
distinguish between the definitive numerals (septies, decies) and the indefinitives quoties and
toties.
The question of numeral adverbs and their use in Anglo-Latin verse is
somewhat more important than it may seem on the surface: one of the most recognisable
mannerisms of Anglo-Latin poetry is the habit of expressing numbers in the form of
multiplication, that is to say, with a numeral adverb and a distributive numeral (as in bis seni,
or two times six, for twelve). This is understandable if we remember that some Latin
numerals are prosodically unusable in dactylic verse (for instance, duodecim, with four short
syllables, does not scan, but bis seni does), but there are good examples of gratuitous use of
multiplication, especially in the verse of Aldhelm and his followers.
147
However, even

142
DAM 7, 46.
143
See Probus at gramm. IV, 247, 8-9.
144
J ones 1975, 45.
145
Long 1901, 12. In late antiquity, the definitive numeral adverbs often became even further contracted (sexis,
septis etc.).
146
e.g. Priscian (Passalacqua 1987, 10, 23 - 11, 4): Reliqua omnia in es productam desinuntinfinitis tamen
numerorum adverbiis etiam n interponitur, quotiens, totiens, similiter multotiens. (All the others end in a long
es, however, in the indefinite numeral adverbs an n is also placed in the middle: quotiens, totiens, and, similarly,
multotiens.) Caper concurs (gramm. VII, 95, 8-9); Marius Victorinus (Mariotti 1967, 88, 20) advocates that all
numeral adverbs should be spelt with an n.
147
e.g. Aldhelms riddle on the woman pregnant with twins (Enigmata XC):
Sunt mihi sex oculi, totidem simul auribus exsto;
55

numeral adverbs have their prosodic limitations: if they are scanned with a long final syllable,
several of them are cretic in structure (long-short-long) and therefore useless in dactylic verse
(e.g. quinquies, sexies, septies, octies, centies), whereas their shorter forms (quinquis, sexis,
octis, centis) are not. This may certainly have contributed to the proliferation of the shorter
forms in Late Latin poetry, and as poetic metre was one of Bedes primary sources on all
prosody, he simply codified what he had encountered in verse. The short final syllables of
numeral adverbs are a phenomenon that is parallel to the shortened final vowels of nouns and
verbs, but in this case, more than in many others, prosody was probably based on necessity.
Bedes ruling that the final syllables of numeral adverbs are always short shows a consistency
with contemporary metrical practice, and I find it possible that he also ruled out the
alternative spelling with -iens for the very reason that it did not scan.
It must be added that Bede does not always follow his own prosodic rules
consistently. In such cases, his grammatical writings are invariably and predictably more
conservative than his verse, where departures from the official prosodic rules are generally
due to the influence of late antique poetry. For one thing, in his De ultimis syllabis nominum,
pronominum, participiorum, Bede declares the final syllables of genitives and ablatives are
long with the exception of the third declension: Genetivus, dativus et ablativus producuntur.
Sed genetivus, cum tertiae fuerit declinationis, cum ablativo suo e tantum littera terminato
breviatur, ut a fonte fontis. (The final syllables of the genitive, dative and ablative cases
are long. But the genitive of the third declension together with its ablative in e are short, as, a
fonte, fontis.)
148
In other respects, Bede follows his own ruling on the matter but,
remarkably, treats the final o in the ablative of the gerund as short, which, of course, is
consistent with much of Late Latin and nearly all of medieval Latin verse, where the ablative
of the gerund was regularly used as a substitute for the present participle.
149
It is commonly

sed digitos decies senos in corpore gesto:
ex quibus ecce quater denis de carne revulsis;
quinquies at tantum video remanere quaternos.
[I have six eyes, and take things in from as many ears;
but I have sixty toes and fingers on my body.
Look, when forty or these have been torn from my flesh,
I see that only twenty will remain.]
Ehwald 1919, 136; trans. Lapidge 1985, 90. Note the short e in quinquies at line four.
148
DAM 6, 41-43; trans. Kendall 1991, 77.
149
The first documented uses of the ablative of the gerund with a short final o are in Seneca; see J. Fitch 1987,
154.
56

placed in the end of the line so that the two final syllables take up the beginning of the fifth
foot.
150
Bedes use of the gerund conforms to this practice:

Num nostrum e speculis dixit temptand latenter
lustrabas itiner?
151

[He said: Did you observe my journey testing me secretly?]


Bede also deviates from his prescribed prosody in the genitive singular of the fourth-
declension spiritus, where the u is short on two occasions, e.g. Spirits extinxit sacri quae
praescia virtus (Which the virtue that knew the Holy Spirit defeated) in line 333 of his
metrical Life of Cuthbert.
152
This is understandable from a strictly technical point of view: the
normal genitive formspirits is cretic and therefore unusable in dactylic verse, whereas the
unclassical form spirits is not. Bedes unusual licence is probably based on the emulation of
Sedulius, in whose Carmen paschale it appears at least three times.
153
Bede does not discuss
this particular licence in detail, but he does touch on the problems inherent with cretic words
in the fifteenth chapter of his treatise.
154
Although similar shortening of the final us in fourth-
declension genitives is by no means unusual in Late Latin and medieval verse, in Bedes
poetry it seems restricted to spiritus, without doubt a necessary word for a Christian poet, and
the phenomenon can almost certainly be attributed both to prosodic necessity and the
authority of Sedulius.
A final word must be said about the degree to which Bede has Christianised the
examples which he uses in his presentation of final syllables: on the whole, apart from
endorsing some post-classical syllable lengths, Bedes chapters on final syllables are fairly
conventional, and less dominated by the authority of Sedulius or other Christian poets than
the other chapters of his treatise. This is understandable when we bear in mind that Serviuss
and Victorinuss word-lists hardly gave Bede the scope he would have needed fully to
implement his Christian agenda, as the poetic quotations they employed rarely exceed the
length of one word. There are, however, some minute but nevertheless telling alterations that
Bede has undertaken, and these are apparent in his presentation of Greek nouns. Where

150
e.g. Ven. Fort. Mart. 3, 153: Carnutis hinc etiam dum praetereund veniret (When [the saint] was passing
by the Carnutes). The ablative of the gerund can have a final long o in the arsis, as testified by Aldhelms
Carmen de virginitate, line 44 (Ehwald 1919, 354): Omnia regnand dispensat saecula simplex ([God]
guides in rule all generations as one). Trans. Rosier 1985, 104.
151
J aager 1935, 76, lines 241-242.
152
J aager 1935, 82.
153
Huemer 1885, 395; Wright 2005, 158. Sedulius also shortens the final us in the genitive forms of vagitus,
domus and sexus.
154
DAM 15, 1-11.
57

Servius gives Musa and Euterpe as examples of Greek nouns with final a and e,
155
Bede has
eliminated these mythic figures and replaced them with the innocuous and Christian ecclesia,
baptisma, schole and synagoge:

In Graecis vero nominativus singularis has habet breves: a, ut ecclesia,
baptisma...longas vero has: e, ut scole, sinagoge, quae Latina consuetudo in a
terminat...
156

[In Greek words the nominative singular is short in these short final syllables: a, as
ecclesia, baptisma...the nominative singular has these long final syllables: e, as in
schole, synagoge, which in Latin usage ends in a...]
157


Bedes presentation is certainly not accurate, as far as Greek prosody is concerned: in Greek,
first-declension words usually have a long final a, except words with the suffix ju (as in
Mqociu, auiociu or tpuacu), and ecclesia does not fall into this category; rather, Bede is
describing its Latinised pronunciation. It is also apparent that Bedes choice of examples is
not entirely successful: the words schole and synagoge are, in Latin, generally spelt with a
(schola, synagoga), something which Bede himself readily admits. Bedes efforts at
presenting Greek prosody in a more Christian form do not end here: in his lists of Greek
vocatives, he predictably uses Peter (Petros, vocative Petre) instead of Maximus Victorinuss
Phoebos, Phoebe.
158
But where Bede really clutches at straws is his list of Greek first-
declension vocatives: Bedes presentation includes the vocatives of schole, synagoge,
pentecoste and parasceve, all identical with the nominative and all of them arguably things
one is unlikely to address in the second person:

Vocativus, cum a terminatur, in masculinis tantum longus est, ut Aenea: nam in
femininis corripitur, ut cathedra. E terminatus producitur, ut scole, sinagoge,
pentecoste, parasceve, exceptis his quorum nominativus os terminatur, ut Petros
Petre.
159

[Final a in the vocative is only long in the masculine, as Aenea. In feminine words it
is short, as cathedra. Final e is long, as schole, synagoge, pentecoste, parasceve, with
the exception of those words whose nominative ends in os, as Petros Petre.]
160


The passage, in all its superfluity, can only be attributed to Bedes attempts to make his
presentation of Greek prosody as Christian as possible: as in his list of Latin letters, which he

155
gramm. IV, 452, 2.
156
DAM 6, 70-74.
157
Trans. Kendall 1991, 79.
158
gramm. VI, 234, 15
159
DAM 6, 80-84.
160
Trans. Kendall 1991, 81.
58

had elaborated with his own Greek additions, practical considerations seem to have come
second to piety.


2.4. The structure of the dactylic metres


2.4.1. The dactylic hexameter

As we have witnessed, Bedes presentation of verse prosody shows certain well-recognisable
departures from the practice of classical poets. Generally these differences are small and can
almost always be attributed to the verse technique of the Christian hexameter poets of the
Late Empire, although, as we have observed, Bedes own generalisations, and, at times, false
analogy also played a part. In the main, Bede is very consistent about his reformed rules of
Christian prosody, especially in those cases where his Christian models have shown similar
consistency. In general, though, prosody as presented by Bede is still remarkably close to
classical practice: this is shown by the way in which he revised his own poetry, eliminating
incorrect vowel lengths where his closer study of earlier poetry proved them wrong.
161
His
opposition to some unusual features of classical verse, such as the use of hiatus and productio
ob caesuram, cannot be regarded as particularly radical; what is a new element in his treatise
is their blanket denunciation as old, or, in other words, potentially un-Christian practices:
for Bede, purely prosodic considerations appear to have become inextricably enmeshed with
ideological ones. Nevertheless, these outdated practices were, by and large, uncommon
even in classical verse, and also many pagan authors had avoided them strenuously. Bedes
presentation of classical and post-classical prosody tends to exaggerate the differences
between classical and Christian poets, and although it illuminates his agenda very well, and
ultimately affected the poetic diction of later poets, it still mostly corresponds with classical
practice.
From a purely theoretical point, Bedes description of the hexameter line must
be considered an even bolder departure from the traditions of Latin verse and verse theory, as
it essentially constitutes a redefinition of the dactylic hexameter itself. If we take a typical

161
Lapidge 1996, 343; Wright 2005, 153-155.
59

definition of the dactylic hexameter line, as presented by virtually all grammars and treatises,
ancient and modern alike, it is more or less the following:

The hexameter is the metre of epic and much other non-dramatic poetry, and is
occasionally found as a stichic length in the drama. As the name implies, it consists
of six dactylic metra. Of these, the first five may take the dactylic or the spondaic
form, although the spondee is rare in the fifth place; the last metron must be a
spondee.
162


Contrast this with what Bede writes in the introductory part of his chapter De metro dactylico
exametro vel pentametro:

Metrum dactylicum exametrum, quod et heroicum vocatur, eo quod maxime heroum,
hoc est, virorum fortium facta canerentur, ceteris omnibus pulchrius celsiusque est.
Unde opusculis tam prolixis quam succinctis, tam vilibus quam nobilibus aptum esse
consuevit. Constat autem ex dactylo et spondeo vel trocheo, ita ut recipiat spondeum
locis omnibus praeter quintum, dactylum praeter ultimum, trocheum vero loco
tantum ultimo; vel, ut quidam definiunt, spondeum ultimo loco semper et omnibus
praeter quintum, trocheum vero nusquam, quia, etsi ultima brevis est natura, tamen
spondeum facit ad votum poetarum, qui, ut praediximus, ultimam versus omnis
syllabam indifferenter accipiunt.
163

[The dactylic hexameter, which is called heroic, because the deeds of heroes, i.e. of
brave men, used to be sung in it, is more beautiful and loftier than all the rest. Hence
it is usually as suited to extended works as short ones, to common works as to
dignified ones. It is formed from the dactyl, the spondee, and the trochee in such a
way that it takes the spondee in every foot except the fifth, the dactyl in every foot
except the last, and the trochee only in the final foot. Or, as some prosodists explain
it, it takes the spondee in the last foot and in all feet except the fifth, but it never takes
the trochee, because, even if the final syllable is short by nature, the metre makes a
spondee at the will of the poets, who, as I have stated above, treat the final syllable of
every verse as common.]
164


Bedes description of the hexameter line is, for the greater part, largely borrowed from
Mallius Theodorus, but Bede has elaborated it with additions of his own to make it conform
to his severe approach to metrical structures. One may note the rather lengthy excursus on the
nature of the final foot of the line (whether it should be a spondee or a trochee), where Bede
reminds his reader that the final feet of lines are syllabae ancipites and that every final foot
can ultimately be interpreted as a spondee. This discussion is possibly designed to serve as a
warning to readers still struggling with Latin quantitative prosody: the trochee is not one of

162
Raven 1965, 44.
163
DAM 10, 2-13.
164
Trans. Kendall 1991, 97.
60

the building blocks of the hexameter line and must never be used, except as the final foot of
the line.
More conspicuously, Bede has reformulated the make-up of the hexameter line
itself. According to most definitions, the fifth foot of the hexameter is usually a dactyl, but
Bede states explicitly that it is always a dactyl (spondeum locis omnibus praeter quintum),
and later reminds his reader that the last foot must always be a spondee but the fifth foot,
never (spondeum ultimo loco semper et omnibus praeter quintum). We can see how subtly
radical Bedes tinkering with the definition of the hexameter is if we contrast it with his
source, Mallius:

Constat autem metrum dactylicum hexametrum heroicum ex dactylo et spondio vel
trochaeo, ita ut recipiat spondium locis omnibus, dactylum locis omnibus praeter
ultimum, trochaeum vero loco tantum ultimo.
165

[The heroic dactylic hexameter consists of the dactyl, the spondee and the trochee in
such a way that it takes the spondee in every foot, the dactyl in every foot except the
last one, and the trochee only in the final foot.]

Bedes exposition of the dactylic hexameter is almost a verbatim quotation from Mallius, but
where Mallius has spondium locis omnibus, Bede has wedged in praeter quintum. Bede has
effectively ruled out the fifth-foot spondee from his very definition of the hexameter line, or,
having recognised that the fifth foot is usually a dactyl, he has ruled that it must never be a
spondee.
As in Bedes revision of classical prosody, this break with grammatical tradition
must be viewed in its proper historical context. It is obvious that what Bede perceived as the
essential differences between pagan and Christian prosody has once again played a role, as
well as his perhaps over-zealous habit of presenting as rules what had previously been mere
tendencies. For Bede, spondaic fifth feet in hexameter lines were, together with hiatus, a relic
from a bygone age, a claim that can be validated by the study of classical and Late Latin
verse, although arguably not to the extent that Bede implies.
The spondaic fifth foot was, already in antiquity, considered an unusual feature,
indeed so much so that lines with this feature were given a name of their own: they are called
spondaic verses (Greek oaovocioov:c, Latin versus spondiaci). Greek hexameter poets used
spondaic verses from time to time, although sparingly; the individual styles of the poets also
played a role in the way in which this feature was employed. Hesiod, in particular, was

165
gramm. VI, 589, 20-23
61

apparently fond of this device, as the opening line of his Works and Days attests.
166
On the
average, spondaic verses appear to constitute approximately five per cent of all Greek
hexameter lines, although individual variation seems to be considerable (ranging from one
every 6.9 lines in the Hellenistic Aratus to none at all in the late antique Nonnus and
Musaeus). The appendix of George E. Duckworths Vergil and Classical Hexameter Poetry
gives the following figures:
167


Spondaic verses, one every x lines

Homer, Iliad 18.3
Homer, Odyssey 20.6
Homer, total 19.4
Hesiod 15.7
Aratus 6.9
Theocritus, I-XIII 63.1
Callimachus, Hymns 63.1
Apollinaris 11.9
Quintus 14.3
Nonnus -
Musaeus -

Spondaic verses are a feature that was radically limited already in the earliest Roman
attempts at hexameter poetry. One may question the reason for this development, but it is
probable that as the Latin hexameter, for several historical reasons, was from the start less
dactylic than its Greek counterpart, it was considered necessary for it to have a dactylic
cadence to render its metrical structure more recognisable. Already in Ennius, spondaic
verses are considerably less common than in most Greek hexameter poetry (one every 51.5
lines according to Duckworth, although any statistics on Enniuss verse are, of course,
inconclusive), and, over time, they show a slow but steady decline. Excepting Catullus, who
in his Peleus and Thetis employed spondaic verses abundantly, apparently in emulation of the
Alexandrians, the classical hexameter poets were very cautious in their use of spondaic
verses, as we can observe from the following selection of statistics - again courtesy of
Duckworth:

Spondaic verses, one every x lines

Cicero 745.0

166
Moooi Hicop0cv, oiooi kcocooi (Pierian Muses, praising with your song).
167
Duckworth 1969, table III.
62

Lucretius 230.2
Catullus, LX 13.6
Vergil, Eclogues 276.0
Vergil, Georgics 437.0
Vergil, Aeneid 409.5
Horace, Satires -
Horace, Epistles 1,967.0
Horace, Ars poetica 476.0
Ovid, Metamorphoses 323.5

As we can see, in the classics, spondaic verses are not only an unusual verse-type but a
downright marginal one. It is also interesting to note that Horaces Satires have no spondaic
verses at all, although, in other respects, their quasi-colloquial style is generally viewed as
looser than that of epic poetry. Apart from Catullus, the highest instance of spondaic verses
is in Lucretius, where they are probably attributable to his more archaic style, still influenced
by Ennius. In Vergils Eclogues, which show a higher number of spondaic verses than his
other work, they are obviously due to the emulation of Theocritus, as in Catullus. All this
would seem to indicate that by the age of the Augustans, spondaic verses had come to be
viewed as a somewhat contrived artistic device (this view is corroborated by the admittedly
few, and arguably weak, attempts at parody they excited).
168
Usually they were used in direct
paraphrases of Greek lines, in conjunction with Greek names, or otherwise in Greek-related
contexts, where they were often paired with other Graecistic devices such as hiatus or
unusual caesurae.
169
Owing to their perceived heaviness, they could also be used to express
great size. In some cases, they alluded to archaic Latin poetry (Ennius or Lucretius).
170
The
following is a random sampling of spondaic lines as employed by Vergil and Ovid with brief
analyses of the probable motives for their use:
- Cara deum suboles, magnum Iovis incrementum (Verg. ecl. 4, 49) (Beloved progeny of
the gods, Jupiters great offspring). Spondaic verse suggesting size.
- Ante tibi Eoae Atlantides abscondantur (Verg. georg. 1, 221) (First let Atlass daughters
vanish from your sight at dawn). Spondaic verse in conjunction with Greek names; also
note the Greek hiatus before Atlantides.

168
The most famous example of a spondaic verse parodied is Ciceros (Att. 7, 2): Flavit ab Epiro lenissimus
Oncesmites (The gentle Onchesmites blew from Epirus); the line mocks the style of Catullus and the
Neoterics, with whom the spondaic verse was a stock Greek affectation. - The only spondaic verse in Horaces
Ars poetica (line 467) is in his admonition not to save mad poets, should their vanity drive them to self-
destructive acts: Sit ius liceatque perire poetis; invitum qui servat, idem facit occidenti (Poets should have
the right to kill themselves; who saves them against their will as good as murders them). The association of
spondaic verses with overly pompous poets is evident.
169
Raven 1965, 101.
170
Eden 1975, 32.
63

- Tune ille Aeneas, quem Dardanio Anchisae / alma Venus Phrygii genuit Simoentis ad
undam? (Verg. Aen. 1, 617-618) (Are you the Aeneas whom kindly Venus bore to the
Dardanian Anchises on the shores of the Phrygian Simois?) Profusion of Greek names;
hiatus before Anchisae.
- Cum patribus populoque, penatibus et magnis dis (Verg. Aen. 8, 675.) (With the senate
and the people, the Penates and the great gods). Allusion to Ennius;
171
spondaic verse
suggesting size.
- Penelopaeque socer cum Parrhasio Ancaeo (Ov. met. 8, 315) (Penelopes father-in-law
with the Parrhasian Ancaeus). Greek names with Greek hiatus.
The marginalisation of the spondaic verse in classical hexameter poetry appears
to be the by-product of the general standardisation the hexameter line, above all its final
cadence: in classical hexameter verse, the final two feet of a normal hexameter correspond
either with the type cndere gntem or cnde seplcro: in other words, the line ends with a
dactyl and a spondee, and the final word has either two or three syllables, thereby ensuring
that the accent coincides with the ictus in the final two feet of the line. These two types were
overwhelmingly popular already in Ennius (around 80 per cent of the lines), and other types
gradually fell into disuse.
172
This standard cadence made the hexameter line more uniform,
and, for most Latin speakers, it was probably its most recognisable part. Unusual line-endings
were reserved for special effect, and they often have a ring of the literary about them,
alluding to faraway places or earlier poets, either in homage or parody.
173

The practice of the classics was generally followed by the poets of the Silver
Age and late Antiquity: their use of spondaic verses shows frequencies similar to their
Augustan predecessors, although, conspicuously enough, several poets seem to have given up
spondaic verses altogether. This applies, above all, to a group of late Christian authors, whom

171
ibid; Lloyd 1956, 38-46. The phrase cum magnis dis appears in Pyrrhuss speech, as portrayed in Enniuss
Annals 190: Dono ducite doque volentibus cum magnis dis (This I grant you with the favour of the
great gods). Vergil has also used the phrase in Aen. 3, 12: Cum sociis natoque penatibus et magnis dis
(With my companions, son, the Penates and the great gods).
172
Raven 1965, 100-101. Unorthodox line-endings include all those in which the word-accent clashes with
ictus, as in most lines ending in a quadrisyllabic word or word-combination or a (stressed) monosyllable (e.g.
Verg. Aen. 10, 442: soli mihi Pallas; Aen. 5, 481: procumbit humi bos. Quinquasyllabic words or word-
compounds such as frugiferentis or omn(e) animantum, freely used by Lucretius, also fell into general disuse,
although they appear to have made a comeback in late antiquity, and later were a favourite technique of the
Anglo-Latin poets. The above restrictions do not, however, apply to the conversational hexameter lines of
Roman satire.
173
Nougaret 1948, 46, suggests that already Ennius and Lucretius used spondaic lines as a vehicle of parody.
The notorious Ennian line (dubia 9 Skutsch) Introducuntur legati Minturnenses (The Minturnese legates are
brought in), which consists entirely of spondees, can be seen to reflect the pomposity of the dignitaries.
64

Duckworth has dubbed post-Ovidian.
174
Of these, such poets as Paulinus of Prigueux,
Dracontius, Cyprianus and Arator did not use spondaic lines at all, and Sedulius, Bedes
apparent favourite, only once.
175
Here, again, we may perhaps look to Sedulius as a possible
model for Bedes redefinition of the dactylic hexameter: Bede was aware of Seduliuss single
spondaic line in Carmen paschale (5, 196), but he offered it an alternative scansion, as if to
exculpate his model of what in his eyes was a grave metrical fault.
176
But we shall discuss
this later in detail.
Meanwhile, spondaic verses met with increasing disapprobation in the works of
grammarians. According to Gelliuss Attic Nights, Seneca notoriously sneered on Vergils
allusions to Ennian verse as a gratuitous archaism (although, admittedly, he did not specify
spondaic verses);
177
Quintilian, on the other hand, labelled spondaic verses, together with
similar prose clausulae and hexameter lines ending with five-syllable words as weak
(praemolle).
178
Of the late antique grammarians, the most hostile stand was taken by Audax
in his Excerpta de Scauro et Palladio (gramm. VI, 337, 17-20):

Minusque lenis est versus, qui quinto loco spondeum magis quam dactylum habuerit,
et vocabitur spondiazon, veluti est ille: Aut leves ocreas lento ducunt argento.
(Aen. 7, 634)
[A verse that has a spondee, rather than a dactyl, in the fifth foot is uncouth, and it is
called a spondeiazon, as the following one: Aut leves ocreas lento ducunt argento
(Or they make smooth greaves of malleable silver).]


174
Duckworth 1969, 132-133; table I.
175
ibid; Ceccarelli 2008 vol. 2, 88.
176
DAM 15, 55-60..
177
Gell. 12, 2, 10:
De Vergilio quoque eodem in loco verba haec ponit: Vergilius quoque noster non ex alia causa duros
quosdam versus et enormes et aliquid supra mensuram trahentis interposuit, quam ut Ennianus
populus adgnosceret in novo carmine aliquid antiquitatis.
[In the same place, he (i.e. Seneca) says these words about Vergil: Even our Vergil used, from time
to time, some harsh, unusual and somewhat overlong lines for no other reason than that friends of
Ennius might recognise something of the old in a new poem.]
178
Quint. inst. 9, 4, 65:
Est in eo quoque nonnihil, quod hic singulis verbis bini pedes continentur, quod etiam in carminibus
est praemolle, nec solum ubi quinae, ut in his, syllabae nectuntur, fortissima Tyndaridarum (Hor.
sat. 1, 1, 100), sed etiam quaternae, cum versus cluditur Apennino et armamento et Orione.
[There is also something in the fact that here two feet are included in one word, something that is
excessively weak even in verse, not only when a word of five syllables ends a verse, as fortissima
Tyndaridarum, but even when the concluding word consists of but four syllables, as Apennino,
armamentis, Orione.]
Quintilians objection is probably based on the accentuation of verse-endings of this type: although Latin
presumably had a secondary accent, it was not considered strong enough for the beginning of the fifth foot (see
Wilkinson 1963, 90-91; Raven 1965, 32; Allen 1973, 190).
65

Aldhelm, Bedes immediate precursor, seized Audaxs condemnation of spondaic verses and
more or less formulated it into a rule. In his De metris, Aldhelm quotes Audax in his aesthetic
objection to spondaic verses: minus lenis est versus, qui quinto loco spondeum habuerit (a
line with a spondee in the fifth foot is uncouth).
179
Somewhat later, he reiterates the warning,
this time in stricter terms: nec in fine dactilus poni debet nec spondeus in quinto loco (one
must never place a dactyl at the end of a line or a spondee in the fifth foot).
180
But Aldhelm
does not stop here: in the invocatory prooemium of Carmen de virginitate, his most extensive
hexameter work, he goes yet one step further and actually calls divine powers to his aid in
order to avoid spondaic verses (line 49): Spondaei quintam contemnat sillaba partem (Let
the syllable of the spondee shun the fifth foot).
181

Aldhelms patent hostility towards spondaic verses is reflected in his own work,
which, as Anglo-Latin poetry in general, is uniformly devoid of them. The only exception
may be Aldhelms own spondaic lines in his De metris, where he demonstrates his
classification of the schemata of dactyls and spondees in hexameter lines.
182
But these
spondaic verses are intended as purely theoretical models, and it is plain that when it came to
actual poetry, Aldhelm regarded them as inapplicable. Aldhelms take on spondaic verses
largely corresponds with his remarkably formulaic approach to metre and versification: as we
stated previously, the final two feet of the hexameter line were for Aldhelm and his followers
an immutable block that invariably consists of a dactyl and a spondee. As Carin Ruff has
formulated this spurious logic: That the dactyl is characteristic of the fifth foot explains (in
Aldhelms view) why the spondee is characteristic of the sixth foot.
183
Aldhelms views
were probably influenced by Audaxs purely hypothetical example of what a hexameter line
that consists of six dactyls would sound like. As an illustration, Audax uses two such lines,
one apparently of his own invention and the other a highly eccentric paraphrase of Vergils
Aen. 9, 503-504: At tuba terribilem sonitum procul aere canoro / increpuit (But the brazen
trumpet let out a terrible sound from afar):

Posset itaque esse etiam versus ex dactylis sex, si modo metrici admitterent, veluti
dictus est ille, interea tenero mihi bucula pascere gramine, item at tuba terribilem
sonitum procul excitat horrida.
184


179
Ehwald 1919, 83.
180
Ehwald 1919, 83.
181
Ehwald 1919, 355.
182
Ehwald 1919, 84-89.
183
Ruff 2005, 158.
184
gramm. VII, 340, 1-5.
66

[In other words, there could also be a line of six dactyls, if the prosodists only
allowed it, as the following one: Interea tenero mihi bucula pascere gramine
(Meanwhile, my herd grazes on the soft grass); similarly At tuba terribilem
sonitum procul excitat horrida (But the frightful trumpet raises up its terrible sound
from afar).]

Bede appears to have followed Aldhelms reasoning, as to the interconnectedness of the
compulsory fifth-foot dactyl and the sixth-foot spondee. However, he also seems to have
been led astray by Audaxs purely hypothetical discussion of (non-existent) sixth-foot
dactyls: in his chapter on the outdated practices of pagan authors and the subsequent
improvements made by the Christians, he suggests that the earlier poets did not always follow
the ideal type of verse-ending, but sometimes used spondees in the fifth foot of the line and
dactyls in the sixth. This misunderstanding is obviously due to Bedes attribution of Audaxs
hexameter paraphrase of Vergil to Vergil himself:

Nam et in exemplis antiquorum inveniuntur aliquoties duo spondei in fine versus,
sicut et duo dactyli nonnumquam, ut sunt illa Maronis:
At tuba terribilem sonitum procul excitat horrida
et:
Aut leves ocreas lento ducunt argento (Aen. 7, 634).
185

[In the model passages of the ancient poets two spondees are found several times at
the end of the line, and similarly two dactyls, as in these verses of Vergil:
At tuba terribilem sonitum procul excitat horrida
(But the frightful trumpet raises up its terrible sound from afar);
and:
Aut leves ocreas lento ducunt argento
(Or they fashion smooth greaves from malleable silver).]
186


This must be considered a highly atypical oversight on Bedes part, but as we have witnessed,
he was sometimes prone to misquotation, especially when the misquotation appeared to
confirm a theory.
187
Nevertheless, Bede appears to have accepted implicitly Aldhelms view
that the fifth and sixth foot of the hexameter line are interconnected, that deviation from the
norm mainly occurs in pre-Christian poetry, and that the veteres, who were notoriously prone
to metrical irregularities, may have gone beyond the employment of spondaic verses in their
transgressions against the norm. Bedes assumption that pagan authors used lines with a
dactyl in the sixth foot also appears in his chapter on syllable fusion and resolution (De

185
DAM 16, 3-7.
186
Trans. Kendall 1991, 141.
187
It must be noted that Audaxs quasi-Vergilian prevarication is also quoted by Aldhelm (Ehwald 1919, 89-
90), who also gives the original line as Vergil wrote it, attributing only the latter to Vergil.
67

episynalipha vel dieresi), where he discusses the genitives of second-declension nouns with -
ius and -ium. He presents a line from Paulinus of Nola (18, 280), where the final -ii is in the
last element of the line and presumably to be scanned as one syllable: Oblectans inopem
sensu fructuque peculii. Bede defends this interpretation of the line but makes an allusion to
the ancients with their habit of ending a hexameter line with a dactyl:

Quod si quis dixerit hic eum more antiquorum dactylum in fine posuisse versiculi,
legat quod idem alibi dicit:
Excoluit, biiugis laquearii et marmore fabri (Paulin. 27, 385).
188

[For if anyone should claim that in this instance Paulinus placed a dactyl at the end of
the line in accordance with the practice of the ancients, let him read what the same
poet says elsewhere:
Excoluit, biiugis laquearii et marmore fabri
(It was adorned, with the chariots of the panelled-ceiling maker and with the
marble of the stonecutter).]
189


Apparently Bede took it for a fact that pre-Christian poets used not only fifth-foot spondees,
but also sixth-foot dactyls. Moreover, he seems to assume that his reader shares this view:
otherwise he would not have gone to such lengths as to refute an analysis based on this
alleged feature of pagan poetry.
Bedes opposition to spondaic verses went far beyond their mere condemnation.
That the structure must have seemed utterly unpalatable to him is manifest in the way in
which he virtually tied himself in knots in order to exonerate his metrical authorities of what
he perceived as an unpardonable flaw: in essence, Bede tried to prove that when Christian
authors appear to use spondaic lines, they are, in reality, doing nothing of the sort. He
contrived to accomplish this by extending the rules of syllable resolution to suggest
alternative scansions where the fifth-foot spondee would de facto be scanned as a dactyl. The
most problematic case, from Bedes point of view, was indubitably the solitary spondaic
verse in Sedulius, who, in most stylistic and prosodic matters, was Bedes champion. The line
appears in Seduliuss portrayal of the crucifixion (carm. pasch. 5, 196), and undoubtedly the
poet had aimed for an air of unusual heaviness, quite compatibly with the classical use of
the device: Scribitur et titulus: Hic est rex Iudaeorum (And the inscription is written: This
is the king of the Jews). This, however, was plainly not good enough for Bede, who, in his
chapter on prosodic liberties, suggests an alternative scansion, turning the fifth-foot spondee
into a dactyl:

188
DAM 14, 88-90.
189
Trans. Kendall 1991, 131,
68


Quod quomodo scandendum iudicaverit, videat qui potest, utrum Iudaeorum duos
spondees quinta et sexta regione contra morem, an solutis syllabis, iuxta quod supra
monstravimus, dactylum fieri voluerit et spondeum.
190

[Anyone who can may decide for himself how the poet determined that this line
should be scanned: whether he wanted Iudaeorum to make two spondees in the fifth
and sixth feet contrary to custom, or to become a dactyl and a spondee with the
syllables resolved according to the principles which I have discussed before.]
191


It is unclear what Bedes suggested scansion for Iudaeorum would be. He makes a rather
oblique reference to his previous chapter on the fusion and resolution of vowels (DAM 14),
but does not specify how this should be applied to the line in question. Kendall has suggested
the scansion -u-de--rum,
192
although I find it equally possible that Bede had I-da-e--rum
in mind. In either case, Bedes suggestion is remarkably far-fetched compared to the normal
scansion of the line. A further and even bolder elaboration of the rules of fusion and
resolution is a device, apparently Bedes own invention, which consists in the insertion of a
short prosthetic vowel (i or e) between two consonants when one of them is an r. This, as we
may deduce from Bedes own description, is a feature common to medieval ecclesiastical
singing, but Bede only discusses the phenomenon in conjunction with the hexameter line with
the apparent purpose of giving an alternative scansion to spondaic verses. All the examples
Bede presents are by Christian authors (Juvencus, Paulinus of Nola and Prudentius), and their
use of spondaic verses must have seemed particularly jarring to Bedes religious sensibilities,
but, rather than sidestepping the issue, Bede has boldly attempted to explain why these lines
are only ostensibly spondaic:

Recipit et r littera solutionem quamvis ordine dissimili. Ibi enim discissis sive
conglutinatis vocalibus syllaba contra naturam aut adcrescit aut interit; hic autem ea
vocali, quae nequaquam adscripta est, in sono vocis adsumpta superadcrescere
tantum syllaba consuevit, ut:
Illi continuo statuunt ter dena argenti (J uvenc, 4, 426);
et Paulinus:
Et spatii coepere et culminis incrementa (Paulin. 28, 202);
et rursus:
Sic prope, sic longe sita culmina respergebat (Paulin. 28, 91);
et Prudens in Psychomachia:
Dixerat haec et laeta libidinis interfectae (Prud. psychom. 98);
et idem in eadem:

190
DAM 15, 57-60.
191
Trans. Kendall 1991, 139.
192
Kendall 1991, 139: I.e. IUDE, dactyl, with resolution of IU, and the shortening of a long vowel or
diphthong before a second long vowel.
69

Palpitat atque aditu spiraminis intercepto (Prud. psychom. 594):
Neque enim in quinta regione versus heroici spondeum ponere moris erat. Sed ita
tamen versus huiusmodi voluisse reor, ut addita in sono vocali, quam non scribebant,
dactylus potius quam existeret spondeus, verbi gratia, intericepto, incerementa,
interifectae, resperigebat, et per synalipham denarigenti. Quod ideo magis r littera
quam ceterae consonantes patitur, quia, quae dure excipitur; atque ideo sonus ei
vocalis adponitur, cuius temperamento eius levigetur asperitas. Quod etiam in
cantilenis ecclesiasticis saepe in eadem r littera facere consuerunt qui antiphonas vel
responsoria vel cetera huiusmodi, quae cum melodia dicuntur, rite dicere norunt.
193

[The letter r can also be resolved, although not in the same way. For in the examples
we have been discussing, a syllable contrary to its nature was either split into two or
disappeared by means of vowels which were resolved or fused. In this case however
the practice was simply to augment the syllable by adding in pronunciation an
unwritten vowel, as, for example:
Illi continuo statuunt ter dena argenti
(They immediately agree upon thirty pieces of silver for him);
and Paulinus:
Et spatii coepere et culminis incrementa
(They began to have increases both of width and height);
and again:
Sic prope, sic longe sita culmina respergebat
(So the storm soaked the buildings located near and far);
and Prudentius in the Psychomachia:
Dixerat haec et laeta libidinis interfectae
(She had spoken these things, and rejoicing in the death of lust);
and the same poet in this line:
Palpitat atque aditu spiraminis intercepto
(She shudders, and with the entrance of the breath cut off).
Since it was not the custom to place a spondee in the fifth foot of a heroic verse, I
believe that these poets intended such verses to be scanned so that, with the audible
addition of a vowel, which they did not write, there should be a dactyl rather than a
spondee in that foot; for instance, intericepto, incerementa, interifectae, resperigebat,
and by elision, denarigenti. This occurs more readily with the letter r than with other
consonants because r which is naturally voiced harshly is made harsher when it is
followed by other consonants; and therefore a vowel sound is added to it to smooth
its harshness by blending of the vowel. Those who have been trained to chant
antiphons and responses and other such pieces properly are in the habit of treating the
letter r very frequently in this way even in the songs of the Church.]
194


Bedes attempt seems desperate but I see no reason to doubt its sincerity. Seen as an
explanation for the use of spondaic verses in Christian poetry, it is certainly not satisfactory:
J uvencus, Paulinus and Prudentius, unlike Sedulius, belong to a group of late Latin poets

193
DAM 14, 53-79.
194
Trans. Kendall 1991, 127-129.
70

whose use of spondaic verses parallels or even exceeds the classics,
195
and this phenomenon
is scarcely explainable by the postulation of an extra vowel in the fifth foot of the line. Bedes
reference to antiphons and canticles reveals the background of his argument: the insertion of
prosthetic vowels is a practice that is generally postulated for the ecclesiastical music of the
early Middle Ages; such interpolated vowels are presumably what early neumic notation
implied with grace notes known as liquescent neumes.
196
It seems natural that Bede,
himself a teacher of ecclesiastical music, must have been attracted by such a solution to an
apparent metrical flaw, although applying it to hexameter verse intended for reading rather
than singing seems a desperate measure and betrays both his favouritism when it came to
Christian poets and the surprising rigidity of his views on prosody and metre.
Bedes description of the hexameter and its placement of feet shows us the
curiously dual aspect of his scholarly temperament: Bede was undoubtedly more original and
empirical in his approach to metre than most grammarians. He is seldom content with merely
quoting from his predecessors, and everywhere in his treatise we come across observations
that are clearly based on his own observation and meticulous study. At the same time, he
shows a pigheaded predilection for creating systems and rules, and his observations often
appear to have been influenced by a pre-existing theoretical superstructure of his own device,
even to the extent of the suppression of evidence and the creation of inapplicable prosodic
rules. Bedes presentation of the hexameter, in particular, seems to have been impaired by his
visions of an ideal type of verse, a feature that we must now observe more closely.
We may call to mind how Bede discusses the sixth foot of the hexameter line.
Bede is not explicit as to whether it should be interpreted as a trochee or a spondee, but
suggests that it would make more sense always to view it as a spondee. Bede argues in favour
of this interpretation by reminding the reader that the final syllable of every line is
indifferent, i.e. a syllaba anceps. However, Bede presents a further argument that strikes a
post-medieval reader as fanciful, if not downright metaphysical:

Alioquin legitimum numerum XXIIII temporum versus exameter non habebit, quia
tot illum pro sui perfectione habere decebat, quot habet libra plena semiuncias.
197


195
See Duckworth 1969, table I: J uvencus has one spondaic verse per 526.7 lines, a figure comparable with
Vergils Aeneid, but Prudentius has one spondaic verse per 134.3 lines and Paulinus one per 256.5 lines in the
context of Roman hexameter verse, very high frequencies indeed.
196
Treitler 2003, 391.
197
DAM 10, 13-16.
71

[Otherwise the hexameter will not have the correct number of twenty-four morae,
since it was considered fitting that it have for its perfected state as many morae as a
full pound has half-ounces.]
198


Bede clearly views the hexameter line as something that has an ideal, perfect, form that
harmonises with the rest of creation; consequently, he regards the concept of an odd number
of morae in a hexameter line as aesthetically objectionable. His exposition of this principle is
taken from Audax, who is even more adamant in his opposition to the concept of trochaic
sixth feet but presents no arguments in favour of his view except that a hexameter line with
twenty-three morae would, by definition, be catalectic (colobos).
199
Bedes parallel with
weights and measures is apparently his own addition and probably reflects his general world-
view of symmetry and harmony, to which the hexameter, too, is subject.
Bedes discussion of the hexameter shows, once again, that his aim is to write a
guide to the proper composition of Christian poetry. The quotation which he gives as an
illustration of the dactylic hexameter is the first line of Venantius Fortunatuss Christian De
virginitate, an extensive work in elegiac couplets that became a model for much of the
Carolingian poetry in the same metre. To facilitate scansion, Bede has divided the line into
feet: Culmina. multa po.los radi. anti. lumine. complent (Many mansions fill Heaven with
a radiant light). In other words, Bedes prime example of a hexameter line is not taken from
Vergil but from a Christian work, and one in elegiac couplets (rather than pure hexameters) at
that. But there is logic behind this choice: Bede discusses the hexameter and the pentameter
under the same heading, and goes to great lengths to demonstrate their connectedness as well
as their prosodic parallels, illustrating both line-types with the same couplet. Arguably, the
hexameters of Venantius are also easy to scan for beginners: they are almost invariably

198
Trans. Kendall 1991, 97.
199
gramm. VII, 336, 14-23. Audaxs discussion of this purely theoretical point may seem gratuitous:
Quid est dactylicum metrum? Quod constat dactylo et spondeo. Cur non addis ultimo interdum
trochaeo? Quia bono iudicio metrici complures hunc pedem de versu hexametro excludendum
censuerunt. Quippe omnis syllaba in ultimo versu adiaforos est, id est indifferenter accipitur, nec
interest utrum producta sit an correpta, siquidem positione longa fiat, cum partem orationis in exitu
finit. Eo accedit quod vitiosus eius modi versus est, qui trochaeum admittit. Nam cum ita ratio
exposcat, ut in pleno versu viginti quattuor tempora sint, admisso utique trochaeo minuitur temporum
numerus, et erunt tempora viginti tria, qui est versus colobos.
[Why is it a dactylic metre? Because it consists of the dactyl and the spondee. Why do you not add
that it sometimes has a trochee in the last foot? Because several prosodists have, with good judgement,
opined that this foot should be excluded from the hexameter line. Namely, every syllable at the end of
a line is adiaphoros, that is indifferent, and it does not matter whether it is long or short, as it becomes
long by position when it ends a part of speech at the line-ending. In addition, a line that admits a
trochee is faulty. For since the rules require that a full line should have twenty-four morae, the
admission of a trochee, so to speak, diminishes the number of morae, and the line would have twenty-
four morae, making it a catalectic line.]
72

heavily dactylic, the caesurae are regular almost to the point of tedium and the clash of accent
and ictus is generally avoided, except in the very middle of the line.
200



2.4.2. The elegiac couplet

In conjunction with the dactylic hexameter, Bede also discusses the elegiac couplet, which
consists of a hexameter line followed by a pentameter. In his introduction to the subject, Bede
has seen it fit to warn his reader that the pentameter only appears together with the hexameter
line,
201
an observation which, as he cautiously implies, he has made on his own. This is
something that previous generations presumably took for granted, but Bede, ever wary of
overestimating his reader, has stated it explicitly. His definition of the dactylic pentameter
line and its structure are taken from Mallius Theodorus, who is also his main source on the
lyric metres, probably because his definitions are simpler and more practical than those of
many other grammarians, and unencumbered by excessive theoretical speculation. His
example of the pentameter line is taken from the second line of the couplet by Venantius
Fortunatus which also provided him with the model for the hexameter. Once again, Bede has
facilitated the scansion of the line by dividing it into feet that correspond with Malliuss
description of the metre:

Huic cognatum est et quasi familiariter adhaerens, ita ut sine ipsius praesidio
numquam id positum viderim, metrum dactylicum pentametrum, quod recipit
spondeum loco primo et secundo, dactylum locis omnibus, catalecton in medio et in
fine. Huius exemplum:
Laetan.turque pi.is. agmina. sancta cho.ris (Ven. Fort. carm. 8, 3, 2).
202

[The dactylic pentameter is related to the hexameter and, as it were, clings intimately
to it, so that I have never seen the pentameter used without the other being in
attendance. It takes a spondee in the first and second foot, a dactyl in all feet, and a
catalectic foot in the middle and at the end. An example of this meter is:
Laetan.turque pi.is. agmina. sancta cho.ris.
(And the holy host rejoice in sacred choirs).]
203



200
Heikkinen 2004, 27.
201
This observation is, indeed, true, as far as extant literary works are considered. The best-known exception is
the anonymous Pompeian inscriptional poem Nihil durare potest tempore perpetuo (CIL IV 9123), famously
used by Carl Orff in his Catulli carmina. All four lines of the poem are in the dactylic pentameter.
202
DAM 10, 20-25.
203
Trans. Kendall 1991, 98-99.
73

To avoid confusion, Bede also presents an alternative way of dividing the pentameter line,
which, as he is quick to point out, is impractical as it ignores the central caesura of the line, as
well as obscuring its dactylic nature. According to this model, the pentameter line is cut up
into metrical feet of similar lengths so that the beginning of the line has two dactyls, and the
end, two anapaests.
204
Bede does not condemn this theory outright, but cleverly, and with
characteristically wry modesty, demonstrates its impracticality with a line of his own,
constructed from two identical half-lines.
205
Bedes analysis shows that the hexameter is, for
him, the primary metre and the pentameter its derivative:

Huius metri versus quidam ita scandendos astruunt, ut quinque absolutos pedes eis
inesse doceant, spondeum sive dactylum loco primo et secundo, spondeum tertio
semper, quarto et quinto anapestum, veluti si dicas:
Quaerite regna poli, quaerite regna poli,
quaerite dactylus, regnapo dactylus, liquae spondeus, ritere anapestus, gnapoli
anapestus. Quod rationi eiusdem metri, ni fallor, minus videtur esse conveniens, cum
universi qui huic metro usi sunt versum omnem in medio diviserint, quem duabus
pentimemeris constare voluerint, quarum prior dactylum sive spondeum licenter in
utraque regione recipit, posterior solum dactylum in utraque.
206

[Some prosodists claim that verses of this metre should be scanned this way - they
teach that they have five complete feet: a spondee or a dactyl in the first and second
foot, always a spondee in the third foot, and an anapaest in the fourth and fifth foot,
as though you should pronounce the verse
Quaerite regna poli, quaerite regna poli
(Seek the kingdom of Heaven, seek the kingdom of Heaven),
quaerite (dactyl), regnapo (dactyl), liquae (spondee), ritere (anapest), gnapoli
(anapest). This seems less suitable for the scheme of the metre, if I am not mistaken,
since all who have used it have divided the whole line in the middle and have
intended it to consist of two two-and-a-half-foot segments, the first of which takes
the dactyl or the spondee freely in either full foot, and the second, the dactyl
exclusively in both full feet.]
207


Bede shows a firm sense of pedagogics: he has opted for the most lucid means of explaining
the structure of the pentameter. As in his later chapters on the structure of the lyric metres, he
has jettisoned the excess baggage of idle theoretical speculation and simultaneously
demonstrated the simplest building blocks of the metre.

204
This cumbersome analysis of the dactylic pentameter is given by Marius Victorinus (gramm. VI, 109, 29
110, 8).
205
Bede is, of course, not the first author to resort to this device; it occurs already in Terentianus Maurus
(Desine Maenalios / desine Maenalios Ter. Maur. 1730) and is virtually ubiquitous in Late Latin
grammarians.
206
DAM 10, 26-37.
207
Trans. Kendall 1991, 99.
74

Bede also briefly discusses the origins of the hexameter and the elegiac couplet,
as well as the etymology of their names. This part of his presentation cannot be considered
particularly original, as it is borrowed almost verbatim from Isidore. Bede describes the
origin of the hexameter as follows. Hoc metrum post Homerum heroicum nomen accepit,
Pithium antea dictum, eo quod Apollinis oracula illo sint metro edita.
208
(This metre
received the name of heroic verse after Homers time, having previously been called
Pythian, because the oracles of Apollo were uttered in it.)
209
The etymology of the word
elegiac comes from the same source: Hoc autem et superius metrum ubi iuncta fuerint,
elegiacum carmen vocatur. Elegios namque miseros appellant philosophi, et huius modulatio
carminis miserorum querimoniae congruit, ubi prior versus exameter, sequens est
pentameter.
210
(When the pentameter is joined with the hexameter, the verse is called
elegiac. For scholars speak of elegiac poetry as sad, and the modulation of this verse, where
the first line is a hexameter and the next a pentameter, is suited to the lamentations of the
miserable.)
211
Bede, however, departs from this traditional material with a brief excursus
into what he and his contemporaries considered the biblical use of the hexameter and the
elegiac metre:

Quo genere metri ferunt canticum Deuteronomii apud Hebreos, sed et psalmos
CXVII et CXLIII esse scriptos. Namque librum beati Iob simplici exametro scriptum
esse asseverant.
212

[It is said that the song of Moses in Deuteronomy (Deut. 32) and Psalms 118 and
144 were written in this metre in Hebrew, while the book of the blessed Job was
written in plain hexameters.]
213


Bede has drawn this presentation from Cassiodoruss and Jeromes commentaries,
214
which
constitute the most influential attempts to impute a biblical origin on poetic metres. By
Bedes time, their assertions were no longer really questioned, and it is no wonder that Bede
takes the questions of prosody in hexameter verse so seriously. Far from viewing the dactylic
metres as pagan metres that had been adopted by Christian writers, he saw them as essentially
J udeo-Christian, although for a period of time they had also been in the used by Greek and

208
DAM 10, 18-19.
209
Trans. Kendall 1991, 97.
210
DAM 10, 37-41.
211
Trans. Kendall 1991, 99.
212
DAM 10, 41-44..
213
Trans. Kendall 1991, 99.
214
Cassiod. in psalm. 118, 23-26; Hier. praef. Vulg. Iob.
75

Roman pagans. This, apparently, is the real starting point of all his observations on metre and
poetic diction.


2.5. The aesthetics of verse

In pronounced contrast to many of his predecessors, and Aldhelm in particular, Bede does not
limit his treatise of the poetic metres to their inner metric: he goes beyond mere prosodic
issues and the placement of feet in attempting to explain how individual lines could best be
combined into larger entities. He discusses these questions already at the end of his
presentation of the structures of the hexameter and the pentameter.
The elegiac couplet, in Bedes opinion, seems to have required particular
scrutiny. In part this is undoubtedly due to the fact that it was less familiar to his readers than
pure hexameter verse: for one thing, it was less prominently represented in the works of the
authors who belonged to the monastic curriculum. Presumably, the classic elegiac poets
Ovid, Tibullus and Propertius were still largely unknown in England during Bedes lifetime
(apart from the lines quoted by grammarians). Although we know that, even in his discussion
of hexameter poetry, Bede preferred Christian examples to classical ones, in the case of the
elegiac couplet he really had little choice: all his quotations are drawn from the Christian
poets Sedulius, Prosper and Venantius Fortunatus.
215

The one aspect of elegiac verse which in Bedes eyes needed to be spelt out
explicitly is the fact that the hexameter and the pentameter form a sense-unit. This may not
have been self-evident to readers only acquainted with pure hexameters: although the lines
in Anglo-Latin hexameter verse of the Aldhelmian type are more often than not end-stopped,
hexameter verse also permits free enjambment. In elegiac verse, on the other hand, an
individual couplet may not be concatenated with the following one. Bedes discussion of the
structure of the couplet and the emphasis on the close connection of its two lines is probably
intended as a caveat to prospective poets with some grasp of hexameter verse. Drawing on
the example of Sedulius and Prosper, he presents two alternative ways of constructing a
couplet: that where the individual lines make sense on their own, and that where enjambment
takes place between the hexameter and the pentameter:

215
Late Latin elegiac verse is, in some respects, looser than its classical predecessors, and above all, the
Augustan refinement of always ending the pentameter line with a two-syllable word is not consequently
observed. Bede was probably ignorant of such restrictions, and in any case, they would have been meaningless
to him, and consequently are not discussed in his treatise.
76


Observandum est autem in carmine elegiaco nequid umquam de sensu versus
pentametri remaneat inexplicatum, quod in sequente versu exametro reddatur, sed vel
uterque sensibus suis terminetur versus, ut Sedulius:
Cantemus, socii, Domino, cantemus honorem,
dulcis amor Cristi personet ore pio (Sedul. hymn. 1, 1-2);
vel sibi mutuo prior exameter ac pentameter subsequens, prout poetae placuerit,
conserantur iuxta illud Prosperi:
Solus peccator servit male, qui licet amplo
utatur regno, sat miser est famulus (Prosp. epigr. 8, 5-6).
Nam sequentes versiculi, etsi his sunt subiuncti, sibimet sunt tamen invicem
coniuncti et secundus primo dat supplementum. Sequitur enim:
Cum mens carnali nimium dominante tyranno
tot servit sceptris dedita quot vitiis (Prosp. epigr. 8, 7-8).
216

[Furthermore, it should be observed that in elegiac poetry the sense of the pentameter
line is never left to be completed by the following hexameter. Either each line makes
complete sense separately, as in Seduliuss couplet:
Let us sing, companions, let us sing glory to the Lord,
let the sweet glory of Christ sound out from our pious mouths,
or, if the poet prefers, there may be enjambment between the hexameter and the
following pentameter, in the manner of this couplet of Prosper:
The sinner serves evil alone; although he may possess
a great kingdom, he is a wretched enough slave.
The following lines, although they are subjoined to these, are nevertheless linked to
each other in their turn, and the second completes the sense of the first. This is the
next couplet:
When the tyrant sin excessively dominates the flesh,
the mind serves as many masters as the number of vices to which it is given.]
217


Although Bede here discusses something that is a mandatory feature of elegiac verse, he has
extrapolated his discussion of outer metric to hexameter verse. Far from limiting himself to
the bare necessities of metrical rules, he tries to give guidelines for its architectonics on a
larger scope. The eleventh chapter of his treatise, titled Quae sit optima carminis forma (On
the best kind of verse) constitutes a slim compendium of what Bede apparently regarded as
particularly elegant poetic devices. The chapter is too modest to make an actual treatise on
poetic style, but it allows us a glimpse at Bedes aesthetic mindset.
For Bede, the key elements of the best kind of verse are variety and scope,
and their worst enemies are excessive repetition and short-windedness, both flaws that mar
the verse of less accomplished poets and are prominent in Aldhelms poetry. It has been
suggested that Bedes aesthetic is in many respects a reaction to Aldhelms poetic diction and
even that when Bede speaks against excessive repetition he is actually taking a sideswipe at

216
DAM 10, 44-59.
217
Trans. Kendall 1991, 101.
77

Aldhelm.
218
The four devices which Bede recommends as hallmarks of sophisticated verse
are the following:
1) Enjambment of consecutive hexameter lines.
2) The employment of what has traditionally been called a golden line.
3) Asyndetic lists of nouns or verbs.
4) The principle that adjectives should preferably be placed before their nouns.
This may seem like a random choice of poetic devices, and Bedes motives in choosing them
may, on the surface, seem hard to fathom, but apparently his main intention has been the
promotion of a better and more varied hexameter style. The chapter also displays Bedes
obvious admiration of Seduliuss poetry, and his discussion of the aesthetics of verse shows
that this admiration went far beyond questions of prosody. Bede by no means intended these
four techniques to be quick fixes for unimaginative poets; on the contrary, he repeatedly
warns his readers not to overuse these techniques, lest they cause tedium.


2.5.1. Enjambment

Bede starts his discussion with an indirect reference to the structure of the elegiac couplet
which he discussed in the previous chapter: Bede says that, in hexameter verse (as opposed to
elegiac poetry) the concatenation of several lines often causes a pleasing effect, calling forth
the very Christian example of Arator and Sedulius. Bede first cites six enjambed lines from
Seduliuss Paschale carmens portrayal of Lots wife, following it up with a five-line
description of the Temple in Arators De actibus apostolorum.

At vero in exametro carmine concatenatio versuum plurimorum solet esse gratissima,
quod in Aratore et Sedulio frequenter invenies, modo duobus, modo tribus, modo
quattuor, aut quinque versibus, nonnumquam sex vel septem vel etiam pluribus ad
invicem conexis, quale est illud:
Loth Sodomae fugiente chaos, dum respicit uxor,
in statuam mutata salis stupefacta remansit,
ad poenam conversa suam: quia nemo retrorsum,
noxia contempti vitans discrimina mundi,
aspiciens salvandus erit, nec debet arator
dignum opus exercens vultum in sua terga referre (Sedul. carm. pasch. 1, 121-
126);
et Arator:

218
Wright 2005, 167.
78

Iura ministerii sacris altaribus apti
in septem statuere viris, quos undique lectos
Levitas vocitare placet. Quam splendida coepit
ecclesiae fulgere manus, quae pocula vitae
misceat et latices cum sanguine porrigat agni (Arator act. 1, 552-556).
219

[In hexameter verse, on the other hand, the concatenation of many lines is usually
very pleasing. You will frequently find this kind of thing in Arator and Sedulius,
where enjambment links two, three, four or five lines, such as this passage: As Lot
was fleeing from the chaos of Sodom, his wife looked back and was rooted to the
spot, stunned and changed into a statue of salt, converted into her own punishment:
because no one who looks back while trying to avoid the injurious hazards of the
despised world will be saved, nor should the ploughman exercising his worthy labour
look behind him;
or these lines of Arator: They decree that the duties of the office of deacon,
which has to do with the sacred altar, should fall upon seven men, whom they chose
from among all the disciples and agreed to call deacons. How splendidly did the band
of deacons of the church begin to shine, the band which administers the cup of life
and serves the water with the blood of the lamb.]
220


Bede, however, warns his reader not to overdo it:

Verum huiusmodi conexio si ultra modum procedat, fastigium gignit ac taedium.
221

[But if a sequence of run-on lines goes on too long, it is distasteful and
wearisome.]
222


Curiously, he also warns not to do this in antiphonal hymns:

Hymnos vero, quos choris alternantibus canere oportet, necesse est singulis versibus
ad purum esse distinctos, ut sunt omnes Ambrosiani.
223

[And indeed those hymns which are properly sung by antiphonal choirs must be
carefully arranged line by line according to sense, as all the hymns of Ambrose
are.]
224


Bedes emphasis on the importance of enjambment is clearly a reaction to the Aldhelmian
form of hexameter versification, where the individual lines were composed piecemeal, and
the clauses rarely extend beyond line-endings. From the evidence of his own Vita metrica
Sancti Cuthberti we know that Bede was a master in the use of enjambment, and in this his

219
DAM 11, 2-18.
220
Trans. Kendall 1991, 103.
221
DAM 11, 19-20.
222
Trans. Kendall 1991, 103.
223
DAM 11, 20-22.
224
Trans. Kendall 1991, 103.
79

primary models were, indeed, Sedulius and Arator. As a prime example we may observe the
opening of Bedes verse hagiography:

Multa suis Dominus fulgescere lumina saeclis
donavit, tetricas humanae noctis ut umbras
lustraret divina poli de culmine flamma.
Et licet ipse Deo natus de lumine Christus
lux sit summa, Deus sanctos quoque iure lucernae
ecclesiae rutilare dedit, quibus igne magistro
sensibus instet amor, sermonibus aestuet ardor;
multifidos varium lichinos qui sparsit in orbem,
ut cunctum nova lux fidei face fusa sub axem
omnia sidereis virtutibus arva repleret.
225

[The Lord gave many lights to shine in their times, so that the divine light from the
summit of heaven would illumine the gloomy shades of human night. Although
Christ himself, born from God who is the light, is the greatest light, God also gave
saints rightly to redden as lamps of the church, through whom, with fire as their
teacher, love might press on our senses, ardour might burn in our speeches. He
scattered their manifold lamps over the diverse parts of the world, so that the new
light of faith, poured from the torch under the whole heaven, might fill all the land
with heavenly powers.]
226


As we see from Bedes long periods, what he has aimed at is a sort of uninterrupted stream of
verse; if anything, even more so than in the examples he has chosen from Sedulius and
Arator. The effect is enhanced by Bedes subtle placement of clause-breaks: rather than
making them coincide either with the line-ending or the penthemimeral caesura, Bede uses
the less usual trithemimeral caesura in line two (donavit, / tetricas humanae noctis ut
umbras), the even less usual trochaic caesura of the second foot in line five (lux sit summa,
/ Deus sanctos quoque iure lucernae) and the hepthemimeral caesura in line six (ecclesiae
rutilare dedit, / quibus igne magistro). That Arator, also quoted at length by Bede in his
presentation of enjambment, was an influence on Bedes verse can be seen not only from the
overall structure of the passage but from the direct allusions to Arators verse which it
contains: in line six, igne magistro plays on Arators Ecclesiae nascentis erat quibus igne
magistro (act. 1, 126); line seven, on the other hand, is an almost direct quotation of his
Mentibus instat amor, sermonibus aestuat ardor (act. 1, 147), and line ten is a paraphrase of
Arators Omnia qui fidei virtutibus arva serenant (act. 2, 1222): Bede has merely
substituted qui fidei for sidereis and repleret for serenat.

225
J aager 1935, 58-59, lines 1-10.
226
Trans. Steen 2007, 24.
80

If we compare this passage from Bedes Vita with the opening lines of, say,
Aldhelms Carmen de virginitate, the contrast could not be greater:

Omnipotens genitor, mundum ditione gubernans,
lucida stelligeris qui condis culmina coeli,
nec non telluris formas fundamina verbo,
pallida purpureo pingis qui flore vireta,
sic quoque fluctivagi refrenas caerula ponti,
mergere ne valeant terrarum litora lymphis,
sed tumidos frangant fluctus obstacula rupis...
227

[Almighty Progenitor, guiding the world by Your rule, Who are the Creator of the
shining heights of the star-filled heaven, (Who) also formed the foundations of the
earth by Your Word; You Who paint the pale greensward with purple blossom, and
restrain the azure surface of the wave-wandering sea so that the shores of the lands
are not submerged by water, but rather that obstacles of rocks may break the swollen
waves...]
228


The image Aldhelm conjures is one of intense breathlessness, but it has been achieved by a
string of parallel clauses, none of which extend beyond the line-ending. Even the order of the
individual lines could be changed without any fundamental alteration to the content. In other
words, the whole aesthetic of the poem is different and reflects Aldhelms short-spanned
approach to versification, where the main focus was on keeping the inner structure of the
individual line together, rather than any attempt to craft the lines into a congruous whole. All
in all, the respective poetic styles of Bede and Aldhelm betray their personal attitudes
towards poetry, one seeing it as adjacent to rhetoric, the other to arithmetic. However,
Aldhelms verse is certainly easier to read to Bedes, and its structural straightforwardness
does lend it a kind of vigour and panache, which some critics have preferred over Bedes
broader but, at the same time, subtler style.
229

Characteristically, Bede, in his presentation of enjambment, has bypassed any
mention of Vergil or other ancients altogether and, through his choice of poetic examples,
managed to present the technique of enjambment as something typically Christian; in other
words, his exposition of enjambment, once again, exhibits his relentless effort to substitute
Sedulius for Vergil. The sudden introduction of antiphonal hymns into the discussion may

227
Ehwald 1919, 352.
228
Trans. Rosier 1985, 102.
229
e.g. Turner 1836, 376: Our Venerable Bede attempted Latin poetry, but the Muses did not smile on his
efforts. On Bedes Vita metrica Sancti Cuthberti, Turner is particularly harsh: It has little other merit than that
of an Anglo-Saxon labouring at Latin prosody in the dark period of the seventh century. It has not the vigour or
the fancy which occasionally appear in Aldhelms versification. And further, quoting the opening of the work:
His invocation is much inferior to Aldhelms. Nowhere does Turner specify what he means by inferior, nor
does he take into account that Bedes laboured prosody is actually much superior to Aldhelms.
81

strike the modern reader as strange, but we must bear in mind that Bede was himself a teacher
of ecclesiastical music
230
and that antiphonal hymns were the poetic form with which his
readers were most acquainted. Bede has tried to be as clear as possible about the metrical
phenomenon that he is discussing: as in the case of the elegiac line, he expounds that certain
features of hexameter verse are not applicable to other forms of poetry. Bedes observations
do not amount to an actual discussion of the differences between stichic and strophic poetry,
and even those chapters of his treatise which deal with the lyric metres are virtually devoid of
any guidance to poetic style: Bedes brief chapter on the iambic dimeter, as used by
Ambrose,
231
discusses only its metrical structure, and the outer metric of the line or its
stylistic features are not mentioned at all. This underlines the overall impression of
enjambment, and hexameter technique in general, as a novelty that was probably largely
alien, and certainly difficult, for Bedes readers.


2.5.2. Bede on word order


2.5.2.1. The golden line

Bede is often credited as the first grammarian to recognise the construction that traditionally,
in the classrooms and grammars of the English-speaking sphere, is termed the golden line.
The construction can best be defined as a double hyperbaton with the predicate verb placed in
the centre, although usage of the term tends to vary considerably. What often adds to the
confusion is that this, apparently, was the state of things already for the early modern
grammarians, and that the term has, until fairly recently, been virtually unknown outside the
English-speaking world.
232
The best-known use of the term is in the preface to Drydens
Sylvae (1685), which was famously quoted in L.P. Wilkinsons 1963 Golden Latin Artistry:
that verse which they call golden, of two substantives and two adjectives with a verb betwixt
them to keep the peace.
233
Wilkinson has reworked this definition by first noting that
conjunctions, prepositions, etc. can be ignored, then by limiting the term in a way that
probably corresponds with how it was traditionally taught in the English classroom:

230
Hist. eccl. 5, 24.
231
DAM 21.
232
Mayer 2002, 139.
233
Wilkinson 1963, 215.
82


Let us restrict the term, as is generally done, to lines in which the epithets and nouns
appear in the corresponding order, that is, a b C A B: as in
grandia per multos tenuantur flumina rivos.
234


In other terms, the first word of the line agrees with the penultimate one and the second word
with the final one. Wilkinson further refined this definition by creating other, mildly jocular
terms for other types of line-spanning hyperbaton which fall short of his definition of
golden:

The chiastic form a b C B A (shall we call it the Silver Line, since it is not quite so
absolute?) can also be used in this way, as at the end of Georgic II (540):
impositos duris crepitare incudibus ensis.
235


And further:

Another form of hexameter (shall we call it the Bronze Line?) is that framed by
epithet and noun, as in (Aen. VI, 137)
aureus et foliis et lento vimine ramus.
236


Wilkinsons nomenclature of golden, silver and bronze lines was obviously not intended to be
taken too seriously, but it was adopted at least in the 1998 Latin grammar of Dirk Panhuis
(translated in 2006), which presents golden, silver and bronze lines as a veritable grammatical
rule.
237

An alternative, and much stricter, definition of the golden line, which also has
modern-day followers, was presented already in the 1652 Latin grammar of Edward Burles,
the first known source to use the term golden for lines of this type. Burless definition of
the golden line is essentially the same as Wilkinsons, but with the further restriction that the
adjectives should come at the beginning of the line and the nouns at the end:
238


Epithets are elegantly set before their Substantives, and if the Verse doe consist of
two Adjectives, two Substantives and a Verb only, the first Adjective agreeing with

234
ibid; Ov. rem. 445 (Great streams are weakened by many tributaries).
235
Wilkinson 1963, 216; Verg. georg. 2, 540 (Nor had the people heard swords clattering on hard anvils).
236
Wilkinson 1963, 217; Verg. Aen. 6, 137 (A bough, golden both in leaves and in flexible twig).
237
Panhuis 2006, 206. Frustratingly enough, Panhuiss definition of the bronze line differs from Wilkinsons
line framed by epithet and noun, being simply a verse which contains one or two hyperbata with the scheme
abBA. Panhuis also appears to specify that adjectives should precede nouns, being stricter than Wilkinson or
Dryden, and more in line with Burles.
238
Burles 1652, repr. 1973, 357.
83

the first Substantive, the second with the second, and the Verb placed in the midst, it
is called a Golden Verse; as, Lurida terribiles miscent aconita novercae.
239

Pendula flaventem pingebat bractea crinem.
240


Burless little-known and amusingly concise grammar gives several tips on how to write
proper (classical) hexameter verse. Interestingly, he reiterates the ban on fifth-foot spondees
we have witnessed in both Bede and Aldhelm;
241
however, he also warns his pupils not to use
rhyme (in allusion to the medieval rhyming hexameters of the Leonine type, by his time
totally exploded).
242
Burless stricter definition of the golden line is apparently the
combination of the double hyperbaton, as recommended by Dryden, with the preference for a
word-order where adjectives precede their nouns. Burless definition of the golden line can be
represented with the following formula:

adj A adj B V noun A noun B

If this stricter definition of the golden line is followed, not all of the lines presented by, say,
Wilkinson as golden really qualify as such.
243
Apparently, similar inconsistency in the
nomenclature of golden lines and similar structures seems to prevail even in modern
academia, sometimes resulting in widely differing estimates as to their frequency: to name
but one example, Young, in his 1932 article, found thirty-nine golden lines in Vergils
Eclogues,
244
but Kenneth Mayer, in 2002, using stricter criteria, only fifteen (together with
seven silver lines).
245
The differences in the definitions of the golden line are due to diverging
views as to whether chiastic silver lines are included, whether pronouns, adverbs or extra
nouns and verbs are allowed, whether participles in periphrastic verb forms qualify as
adjectives and so forth. Understandably, the whole feature and its relevance have been
questioned: Mayer, himself probably the most industrious compiler of statistics on the

239
Ov. met. 1, 147 (Terrible stepmothers mix deadly aconites).
240
The hanging gold leaf dyed his/her hair yellow. The origins of the line are unknown, but its earliest
known use in a textbook is in J ohn Clarkes 1633 Manuductio ad artem carmificam seu poeticus (p. 345).
Mayer 2002, 166.
241
Burles 1652, repr. 1973, 356: The fifth foot must be a Dactyle, the sixth a Spondey, the other four Dactyls or
Spondeys at our pleasure: and this Verse is also called Hexameter.
242
Burles 1652, repr. 1973, 357: In making Verses avoid Rhimes: as,
Singula spectando ventos superare volando.
Nutrix dum rotat, canit & cunabula motat.
243
e.g. Horaces portrayal of the town mouses abode in sat. 6, 103: tincta super lectos canderet vestis eburnos
(the dyed cloth shone on the ivory couches); the second noun (lectos) comes second and its adjective
(eburnos) only at the end of the line.
244
Young 1932, 517.
245
Mayer 2002, 161.
84

phenomenon, is extremely dismissive of the very concept of the golden line, and views it as a
medieval, or even post-medieval invention.
246

The golden line as such is never discussed in antique sources. What is possibly
the first attempt at its description appears in Diomedess De pedibus metricis sive
significationum industria and its fanciful and highly impractical list of good and bad
verse types.
247
Among these, one type that is specially commended is what Diomedes calls a
teres versus, or rounded verse. Diomedess list has been thoroughly discussed in Mayers
2002 article on the golden line.
248
Most of Diomedess categories are useless for the actual
composition and study of poetry; his list is highly unanalytical in confusing purely metrical
features with rhetorical figures, and some of the good line-types commended by him are
patently faulty and virtually nonexistent in Latin poetry.
249
Similar lists abound in Late Latin
grammarians, and this tradition was perpetuated by medieval and renaissance scholars.
Special hexameter lines seem to have been an academic hobby already in the Silver Age:
both Martial and Quintilian mention hexameter lines that, when read backwards, are in
another metre;
250
Gellius, in his Noctes Atticae talks reprovingly of a book on metrical trivia
such as the name for the verse that grows in each word by one syllable.
251
Impossibly
convoluted nomenclatures were imposed on these verse types by the grammarians, who often
differ widely in their application and also their judgement on whether the respective line-
types are to be recommended or not.
Diomedess list of good and bad hexameter types is, by and large,
unenlightening, but his example of a rounded verse is certainly what we would call a
golden line. Diomedes does not discuss word order, his actual definition of his rounded

246
Mayer 2002, 166.
247
gramm. I, 498-500. The good verses include types which Diomedes terms as inlibati, iniuges, aequiformes,
quinquipartes, partipedes, fistulares, aequidici, teretes, sonores and vocales, or, in Mayers translation, intact,
detached, equal-shaped, five-part, foot-divided, pipe-like, even-worded, rounded, resounding or vocalic. The
bad verses, in turn, are mutili, exiles, ecaudes, fragosi and fluxi (truncated, scanty, tailless, rough and flabby).
Mayer 2002, 144-145.
248
Mayer 2002, 144-158.
249
Diomedes, for instance, recommends a verse-type which he calls partipes (foot-divided), where the word-
breaks coincide with the ends of feet: Partipedes sunt qui in singulis pedibus singulas orationis partes
adsignant, ut miscent fida flumina candida sanguine sparso. (Foot-divided verses are those that mark off
individual words in individual feet, as in: miscent fida flumina candida sanguine sparso.) gramm. I, 499, 10-
12, trans. Mayer 2002, 149. As Mayer notes, Marius Victorinus at gramm. VI, 71,24-27 listed this line not as
an optimus versus, but as one of the worst. Moreover, the line-type is in crass violation of Roman usage of
caesurae, and therefore virtually non-existent in Latin poetry, apart from such curiosities as Enniuss infamous
Sparsis hastis longis campus splendet et horret (Enn. varia 14V), which, in Wilkinsons words, was a joke
already to Lucilius. Wilkinson 1963, 96.
250
Mayer 2002, 140-141; Mart. 2, 86; Quint. inst. 9, 6, 90.
251
Mayer 2002; Gell. 14, 6, 4: et quis adeo versus sit, qui per singula vocabula singulis syllabis increscat.
85

verse is well-nigh unintelligible, and it is not altogether impossible that he is in fact
discussing some other prosodic or lexical feature:

Teretes sunt, qui volubilem et cohaerentem continuant dictionem, ut
Torva Mimalloneis inflatur tibia bombis.
252

[Rounded verses are those that conjoin a fluent and contiguous phrase, such as
Torva Mimalloneis inflatur tibia bombis.
(The stern flute swells with Mimallonean blasts.)]
253


And indeed, it is only in Bedes De arte metrica that we encounter the first reliable and
comprehensible description of a verse that has the structure of Drydens golden line:

Optima autem versus dactylici ac pulcherrima positio est, cum primis penultima ac
mediis respondent extrema, qua Sedulius frequenter uti consuevit, ut:
Pervia divisi patuerunt caerula ponti (Sedul. carm. pasch. 1, 136);
et
Sicca peregrinas stupuerunt marmora plantas (Sedul. carm. pasch. 1, 140);
et:
Edidit humanas animal pecuale loquelas (Sedul. carm. pasch. 1, 162);
item in pentametro:
Dignatus nostris accubitare thoris (Sedul. carm. pasch. 1, 2);
et:
Rubra quod adpositum testa ministrat holus (Sedul. carm. pasch. 1, 16).
254

[The best and most beautiful arrangement of a dactylic hexameter verse is when the
next to the last word agrees with the first word and the final word agrees with a word
in the middle, an arrangement which Sedulius was accustomed to use frequently, as
in:
The unobstructed waters of the divided sea lay open;
and
The dry marble surface astonished the soles of the strangers feet;
and
The asinine animal uttered human speech.
And likewise in the pentameter:
Having deigned to recline on my couch;
and
The herb, which, having been popped in, the red pot serves.]
255


The reader may note the differences between Diomedess and Bedes approaches: rather than
presenting the various types of good verse as an elaborate system, Bede gives a few examples

252
gramm. I, 499, 20-22.
253
Trans. Mayer 2002, 152, except that of the verse, which is mine. The line is a paraphrase of Persiuss torva
Mimalloneis inplerunt cornua bombis (1, 99), itself a golden line.
254
DAM 11, 24-34.
255
Trans. Kendall 1991, 103-105.
86

of the phenomenon, which, moreover, are taken from his own reading of Sedulius. Bedes
presentation of the golden line is therefore probably not influenced by Diomedes (who does
not specify what he means by fluent and contiguous), but is, rather, Bedes own analysis of
a stylistic feature with which he has been thoroughly acquainted through experience. One
must note, furthermore, that Bede does not give the verse-type a name: apparently, he was not
in the least attracted by the Late Latin practice of composing lists of good and bad verse-
types.
Bedes description of the verse-type corresponds with the Wilkinsonian
refinements of Drydens golden line, while at the same time being more lax than Burless
1652 model: what Bede describes is a double hyperbaton of the abCAB type rather than the
chiastic, or silver, model (abCBA). Unlike either Burles or Dryden, Bede does not discuss
verbs or their placement at all.
256
Bede also does not, at least in this context, specify whether
nouns should or should not be restricted to the latter half of the line, although in the first two
cited examples they seem to be. The third example, edidit humanas animal pecuale
loquelas, however, does not follow Bedes own presentation of the poetic device: it simply
consists of a noun and its attribute framed by another such word-pair with the verb at the
beginning.
257
This may seem like an uncharacteristic slip on Bedes part, but, above all, it
shows that he was not dogmatic about his definition. Neither of Bedes examples of good
pentameter lines are golden lines in the strictest sense: the first line dignatus nostris
accubitare thoris has only one pair of noun and attribute, whereas in the latter, rubra quod
adpositum testa ministrat holus, the nouns and attributes are in their prescribed order but the
verb is not, as it intrudes between the final nouns (abACB).
Bedes ostensible lapses from his own description of a good hexameter line
must, first and foremost, be seen as an indication that, unlike some of his predecessors, he
was not trying to create a typology of verse.
258
He starts with what he has found a particularly
pleasing line-type, giving an analysis of its word-order, and goes on to give examples of
similarly pleasing, although structurally not quite identical lines. In essence, Bede could be
said to recommend all lines with hyperbatons, preferably (but not necessarily) with double
ones.

256
Wright 2005, 162-163.
257
As Neil Wright has noted, this construction is very common in both classical and post-classical verse, being
also a favourite structure of Sedulius: Indeed, so common is it that it will be helpful to term lines with this
structure initial verb golden lines. Wright 2005, 162. In Panhuiss (though not Wilkinsons) terminology this
would be a bronze line.
258
Mayer 2002, 142.
87

Bedes attribution of the golden line to Sedulius may seem like yet another case
of his favouritism: in much of modern scholarship, the golden line is seen as a typically
classical feature. L. P. Wilkinsons discussion of the feature relies, inevitably, on exclusively
classical examples,
259
whereas Michael Lapidge views Bedes use of the golden line as an
example of his mastery of Vergilian techniques.
260
It appears, however, that the role of the
golden line in classical poetry has been overestimated and that Sedulius was indeed a more
enthusiastic user of the technique than any of his classical predecessors.
According to the statistics which Kenneth Mayer has compiled on the frequency
of golden and silver lines in classical and post-classical poetry, the feature is actually
comparatively uncommon in most of the classics proper: golden lines rarely exceed two
percent of the lines, and the frequencies for silver lines are lower still. The highest figures are
in Senecas Apocolocyntosis Divi Claudii (six out of forty-nine hexameter lines), Laus
Pisonis (6.13 per cent), Catullus 64 (4.41 per cent) and Ciris and Culex from the Appendix
Vergiliana (4.35 and 4.99 per cent, respectively). In Vergils Aeneid, the figure is as low as
0.34 per cent, and in Ovids Metamorphoses 1.05 per cent.
261
It would appear that, of the
major classical poets, only Catullus went out of his way to employ the golden line and that,
for the others, it was simply one of many possible word orders, enjoying no special
prominence. Mayer has also questioned the claim, made, among others by Wilkinson, that
classical poets used golden lines especially to round off periods.
262

The situation is completely different in late antique and early medieval poetry.
Noticeably high figures occur in Ausoniuss Mosella (3.73 per cent), and, inevitably,
Sedulius (3.93 per cent). At least Mayers statistics seem to bear out Bedes view that golden
word order is typical of Seduliuss verse technique. Seduliuss example was followed by
Aldhelm, whose use of golden lines in his Carmen de virginitate exceeds even the figures
encountered in Sedulius.
263
It appears that, for Aldhelm, the golden line became something of
a mannerism, which, together with his disregard for enjambment, often creates a tedious
effect. As Neil Wright has noted, Aldhelm sometimes allowed himself as many as three

259
Wilkinson 1963, 215-217.
260
Lapidge 2005, 747-748.
261
Mayer 2002, 161.
262
Wilkinson 1963, 216: Virgil also used it finely to round off periods. Here are two splendid examples from
the great finale to Georgics, 1, which also constitute that overarching superflux of rhythm at the end of a
period. Mayer 2002, 161: While several scholars have claimed that the golden line is mainly and artfully
used to close periods and descriptions, the texts I have investigated do not seem to bear this out.
263
According to Mayers statistics, Aldhelm has 188 golden and twenty-three silver lines in his 2,904-line
Carmen de virginitate (6.47 and 0.79 per cent respectively). Mayer 2002, 162.
88

consecutive golden lines, citing the following example (Carmen de virginitate 813-815) as
an illustration (the first two lines are genuine golden lines and the third a silver one):
.
Aequora per vitreos bullirent turbida campos
Priscaque turgescens non nosset litora Pontus,
Caerula fluctivagis sed regnant molibus alta.
[(As) the turbid seas raged throughout their glassy fields,
and the swelling ocean did not know its old shores
but the high seas reigned with their flooding masses.]
264


Bede apparently saw the risks inherent in the repetitive use of golden lines, and, as before in
his discussion of enjambment, he cautions his reader against their overuse, possibly with
Aldhelms verse in mind: Nec tamen hoc continuatim agendum, verum post aliquot
interpositos versus. Si enim simper uno modo pedes ordinabis et versus, tametsi optimus sit,
status statim vilescit.
265
(However, this should not be done constantly, but only after
intervals of several lines. For if you always arrange your feet and verses in the same way,
even if it is the best way, your composition is at once cheapened.)
266
Possibly Bedes
flexible presentation of golden verse and similar, but not quite identical, constructions was
also tailored to preclude the mechanical reiteration of completely identical word patterns.
Several scholars of Bedes own verse have noted the exceptional care with
which he applied Seduliuss golden line and its near-equivalents in his own verse Vita Sancti
Cuthberti. Although Michael Lapidge, in his 2005 article, sees Bedes cultivation of golden
symmetry as a result of his meticulous study of Vergil,
267
Bedes main model was probably
Sedulius (it is, of course, probable that through his study of Sedulius Bede had become
acutely aware of the structure and certainly recognised it when encountering it in Vergil).
Bedes use of hyperbaton is varied, ranging from the strict golden word-order of line 24

Aurato nitidae lustrat fulgore loquelae
[(John Chrysostom) illuminates (Constantinople) with the golden radiance of his
speech]

to the looser

Qui facit humanas asinam reboare loquelas

264
Ehwald 1919, 387-388; trans. Wright 2005, 165-166.
265
DAM 11, 35-37.
266
Kendall 1991, 105.
267
Lapidge 2005, 196-197.
89

[Who made the ass bray human words]

at line 74. As we can see, the latter line is a reworking of Seduliuss edidit humanas animal
pecuale loquelas, which Bede specially recommended in his treatise.
268
More notably,
however, Bedes use of various types of hyperbaton differs markedly from Aldhelms and,
to a lesser extent, Seduliuss occasional reiteration of structurally identical golden lines. In
this respect, Bedes loose and arguably inconsistent description of the phenomenon, as well
as his admonition to avoid excessive repetition of identical structures, seem to correspond
with his own metrical practice.
269
The ultimate disadvantage of golden lines in their strictest
form is that their structure is syntactically and semantically self-contained and that their
continuous use impedes with the enjambment of lines (for which Bede shows a high
appreciation) and results in a monotonously end-stopped style. If we examine the first three
lines of Bedes Vita metrica Sancti Cuthberti, we can see how Bede has solved this problem.
I have used different fonts to indicate the respective word-pairs:

Multa suis Dominus fulgescere lumina saeclis
donavit, tetricas humanae noctis ut umbras
lustraret divina poli de culmine flamma.

As we can see, none of the lines is golden in the strictest sense. The first line comes the
closest: it has the a-b-A-B word order which we associate the golden line, but the predicate
verb has been transposed to the beginning of the following line. The next line has the chiastic,
or silver, word-order a-b-B-A, but here, again, the predicate verb has undergone the same
transposition. The third line has a simple hyperbaton without a central verb. As we can see,
enjambment has been achieved through the manipulation of the placement of verbs, which,
perhaps wisely, Bede leaves undiscussed in his presentation of golden word order.
An important aspect of Bedes presentation of Seduliuss hyperbaton, golden or
otherwise, is its connection with end-rhyme, which is considered a typical feature of
Seduliuss verse, and is probably inseparable from his penchant for certain types of word-
order.
270
It has been suggested that golden word-order ultimately led to the emergence of

268
Lapidge 2005, 195. As Lapidge notes, Bede has further elaborated on his model by adding cacemphaton, or
ugly-sounding syllable reiteration (humanas asinam), to emphasise the braying voice of Balaams donkey.
269
Wright 2005, 163-166.
270
Rhyme is also prominent in Seduliuss abecedary hymn A solis ortus cardine, where it is consistently
present though not in a consistent pattern (Hiley 1993, 284):
A solis ortus cardine
ad usque terrae limitem
90

leonine rhyme in hexameter poetry: if the second word of a hexameter line agrees with the
final one, and they both belong to the same declension, the result is inevitably rhyme-like. In
classical Latin poetry, it appears that some poets actually went out of their way to eliminate
rhymes, usually by tampering with word-order, although long sequences of lines with internal
rhyme have also been documented.
271
Kenneth Mayer has suggested that the rising popularity
of the golden line in late antique poetry was connected with fondness for end-rhyme: that
chiastic, or silver, hyperbatons did not enjoy a similar surge is, in his view, a result of their
inability to produce rhyme.
272
Whether rhyme was a by-product of hyperbaton or golden lines
resulted from the pursuit of rhyme is, of course, an academic chicken-or-egg question of a
rather fruitless kind; at best, we can say that in late antique verse antique word order and
rhyme are strongly related. Of the five lines of Sedulius quoted by Bede, four have pure
monosyllabic rhyme (divisi-ponti; peregrinas-plantas; humanas-loquelas; nostris-thoris) and
one a milder form of word-final assonance (adpositum-holus). Remarkably, even Bedes
examples of not-quite-golden lines have rhyme, and it seems possible that this feature played
a part in Bedes presentation of the feature, although he does not say it in so many words.
Although in his De arte metrica he does not discuss the phenomenon,
Seduliuss use of rhyme certainly did not go unnoticed by Bede: his De schematibus et tropis,
the companion work of De arte metrica, contains a presentation of homoeoteleuton (in
practice, monosyllabic rhyme), where one of Bedes examples is a line by Sedulius (carm.
pasch. 1, 136). Strikingly, it is the very same line with which he illustrates the optimal type of
word order, and it is plain that Bede was quite aware of their connection:


Christum canamus principem
natum Maria virgine.
Beatus auctor saeculi
servile corpus induit,
ut carne carnem liberans
non perderet quod condidit. (Sedul. hymn. 2, 1-8)
[From the rising of the sun in the east
to the furthest reaches of the West
let us sing the glory of Christ, our Lord
born from the Virgin Mary.
The blessed creator of the world
put on a servants body,
that, liberating flesh with his flesh,
He might not lose what He had founded.]
271
Marouzeau 1949, 99-100; Wilkinson 1963, 33.
272
Mayer 2002, 163. If we take Vergil and Sedulius as examples, the phenomenon is most striking: although in
Seduliuss Carmen paschale golden lines are over ten times more common than in Vergils Aeneid (3.93 per
cent in Sedulius, 0.34 per cent in Vergil), the frequencies of silver lines are almost equal (0.23 per cent in
Sedulius and 0.26 per cent in Vergil). Mayer 2002, 161-162.
91

Omoeoteleuton est similis terminatio, quoties media et postrema versus sive
sententiae simili syllaba finiuntur, ut Ecclesiastes (Eccl. 6:9): Melius est videre quod
cupias, quam desiderare quod nescias. Et iterum (Eccl. 7:6): Melius est a sapiente
corrigi, quam stultorum adulatione decipi. Hac figura poetae et oratores saepe
utuntur. Poetae hoc modo:
Pervia divisi patuerunt caerula ponti (carm. pasch. 1, 136).
[Homoeoteleuton or like ending, is the name for the figure when the middle and
final sections of a verse or clause end in the same syllable, as in Ecclesiastes (6:9):
Better it is to see what you may desire (cupias), than to desire that which you cannot
know (nescias). Or in (Eccl. 7:6): It is better to be rebuked (corripi) by a wise man
than to be deceived (decipi) by the flattery of fools. This figure is frequently
employed by poets and by writers of polished prose. Poets use it like this.
The unobstructed waters of the divided (divisi) sea (ponti) lay open.]
273


This passage is notable not only for its endorsement of end-rhyme in poetry: in Bedes De
schematibus et tropis, his Christianising tendency is even stronger than in De arte metrica:
Bedes examples of rhetorical figures are almost exclusively drawn from the Bible, with
some quotations from the Early Fathers being the most prominent exception. That Bede chose
to cite Sedulius in his strictly Christian presentation of schemes and tropes shows that he
viewed him, at least in this context, as being as good as Scripture.
Bedes chapter on optimal word order also contains a gentle admonition to
place adjectives before nouns. Although this is not the typical word order of Latin prose, it is
a tendency that seems to prevail even in classical poetry, the idea being that the weightier
nouns should preferably come at the end of the line.
274
Bede cites examples of this feature by
quoting Sedulius, again a line with roughly golden word-order (abACB) and internal rhyme:

Studendum est praeterea metricis, quantum artis decori non obstitit, ut mobilia
nomina fixis nominibus praeponant, sed nec concinentia nomina coniunctim ponant,
verum interposita qualibet alia parte orationis, ut:
Mitis in inmitem virga est animata draconem (Sedul. carm. pasch. 1, 132).
275

[Poets should also strive, so long as it does not interfere with the grace of their art, to
place adjectives before their nouns, but not to put nouns and adjectives that are in
agreement with each other side by side, but rather to interpose some other word, as:
The harmless rod was changed to a fierce dragon.]
276


In other words, Bede not only recommends that adjectives should precede their nouns but
reiterates his endorsement of hyperbata and, by implication, internal rhyme this time by
cautioning against the placement of agreeing nouns and adjectives next to each other. If we

273
Trans. Kendall 1991, 177-179.
274
See e.g. Winbolt 1903, 153; Norden 1916, 400-402; Harrison 1991, 138.
275
DAM 11, 48-52.
276
Trans. Kendall 1991, 105.
92

observe this rule together with Bedes previous definition of the ideal type of verse, it is
easy to see where the early modern concept of the golden line in its strictest sense originated:
it is simply a fusion of the double hyperbaton as described by Bede, and the principle, also
recognised in De arte metrica, that adjectives should precede rather than follow their noun
heads. Kenneth Mayers studies of renaissance grammar show how Bedes two rules were, in
succeeding grammatical works, collated into one, until we arrive at golden verse as
understood by Burles (adj A - adj B C noun A noun B). That this was merely one
tendency in medieval and post-medieval grammar is demonstrated by the considerable lack of
consensus as to what exactly constitutes a golden line, and the correspondingly varying
estimates as to how common it truly was in classical and post-classical poetry.
That the golden line was essentially a Late Latin feature needs to be more
widely acknowledged. Although the argument can be made, and, indeed, has been made, that
it is a classical technique that came to be overused by post-classical and medieval poets, the
very definition of the structure as we know it is ultimately derived from Bede, whose purpose
was to define something that he perceived as a typically Christian verse technique. Although
Sedulius was certainly not the first user of golden lines, they were, more than in classical
poetry, a predominating feature of his poetic style. And it is probably due to Sedulius, Bede
and Bedes latter-day followers that this technique and its varying analyses came to enjoy
such widespread popularity even in modern classical scholarship.


2.5.2.2. Other observations on word order

Bede is not entirely dogmatic in his recommendation of preferred word-orders. Having stated
that adjectives should ideally precede their main nouns, and that words in agreement with
each other should be apart, he gives some examples to the contrary, again drawn from his
own reading of Christian verse, in this case Prosper. Bede states quite explicitly that what he
has presented as ideal word-order is not a fixed rule but nevertheless contributes to the
overall aesthetic of verse: ...non quod haec simper observari necesse sit, sed quia, cum fiunt,
decori sint.
277
He further adds that beautiful verse can be composed even without observing
this principle:


277
DAM 11, 56-61.
93

Nam et Prosper mutato hoc ordine fecit versum decentissimum:
Moribus in sanctis pulchra est concordia pacis (Prosp. epigr. 29, 1);
et item:
Lex aeterna Dei stabili regit omnia nutu
nec mutat vario tempore consilium (Prosp. epigr. 41, 1-2).
278

[For Prosper has composed a very fitting verse, with this order reversed:
The peaceful harmony in holy conduct is very lovely;
and again:
The eternal law of God rules all things with a stable will,
nor does Gods plan change with the changing of the times.]
279


In the first line, the adjective sanctis follows the noun moribus. In the hexameter line of the
second quotation, the adjective aeterna follows its noun lex, and, in the pentameter, there is
no word interposed between vario and tempore, although the adjective does precede the
noun. These are hardly crass transgressions of good poetic style, as the lines also contain
word orders of the type recommended by Bede (pulchra concordia; stabili nutu). The
examples are certainly well chosen, if their aim is to demonstrate the flexibility and
variability of word order in verse.
Bedes use of the Christian poets Sedulius and Prosper as examples of elegant
word order is hardly surprising, given the overall tone of his treatise, but in the very same
chapter Bede gives his presentation a surprise twist by referring to the authority of the
classical, and pagan, Lucan, quoting nothing less than the opening of his Pharsalia (1, 1-3
and 10-12):

Et Lucanus, poeta veteranus, Caesaris et Pompei proelia descripturus, ita incipit:
Bella per Emathios plus quam civilia campos
iusque datum sceleri canimus populumque potentem
in sua victrici conversum viscera dextra.
Cumque superba foret Babilon spolianda trophaeis
Ausoniis umbraque erraret Crassus inulta,
bella geri placuit nullos habitura triumphos.
280

[And when the ancient poet Lucan sets about to describe the battles of Caesar and
Pompey, he begins in this way: Wars more civil upon the Emathian plains, and
license conceded to lawlessness, I sing; and a powerful people turning with
victorious right-hand against its own vitals...And, while proud Babylon was to be
spoiled of the Ausonian trophies, and the shade of Crassus was wandering
unavenged, has it pleased you that wars, doomed to produce no triumphs, should be
waged?]
281



278
ibid.
279
Trans. Kendall 1991, 107.
280
DAM 11, 62-69.
281
Trans, Kendall 1991, 107.
94

The passage is supposed to provide yet another example of how a poet may depart from the
best kind of word order, as defined previously. The departures, on the whole, are not striking:
there are four instances of an adjective (or participle) following its noun head: ius datum and
populum potentem in line two, umbra inulta in line four and bella habitura in line six.
It is remarkable that Bede, who is usually scrupulous in using only Christian
examples to illustrate his prosodic rules and stylistic rulings, has opted for a relatively
lengthy quotation from a pagan author. He obviously realises this himself, underlining that
Lucan is one of the ancient poets, using the term veteranus here in the same sense that he
uses priscus and vetus elsewhere in his book. We may recall that old, or ancient, in
Bedes vocabulary, simply means pre-Christian, all Christian poets starting with J uvencus
being, by his reckoning, modern. We have seen Bedes reluctance to use Vergil as an
illustration, so why Lucan? My suggestion is that this is yet another subtle manoeuvre on
Bedes part to dispense with pagan learning: the most prominent feature of Lucans Pharsalia
is that, unlike Vergils Aeneid, it is a secular epic, or an epic without pagan gods. This feature
had certainly been widely recognised already in antiquity, often with no little puzzlement and
sometimes with displeasure; Isidore, for one, states that Lucan seemed to have written history
rather than poetry.
282
This lack of the supernatural, or fictional, element, which some had
viewed as a drawback, was, for Bede, a veritable virtue: although Lucan certainly was one of
the old poets, not being Christian, he was not blatantly pagan, and his verse, having no
references to pagan gods, was more compatible with Christian learning. In essence, although
he does not say this explicitly, Bede seems to recommend Lucan over Vergil as the old
poet to be studied.
Bedes treatise is remarkable for the way it attempts to sidestep everything that
is related to pagan religion. His approach, in this respect, is certainly subtler than that of some
of his predecessors. Earlier Christian poets were often very blatant about their denunciation
of pagan beliefs, often in such a way that they could simultaneously exhibit both their
classical learning and their Christian piety. We can encounter this tendency in Sedulius, who
compares benighted pagans to Theseus wandering in the Labyrinth:

Quid labyrintheo, Thesidae, erratis in antro
caecaque Daedalei lustratis limina tecti?
283

[Why, children of Theseus, do you wander in the labyrinthine cave

282
orig. 8, 7, 10: Unde et Lucanus ideo in numero poetarum non ponitur, quia videtur historias conposuisse,
non poema. See also Klopsch 1980, 9-12.
283
Sedul. carm. pasch. 1, 43-44.
95

And roam about the blind thresholds of that Daedalean hall?]
284


Aldhelms renunciation of paganism is considerably more heavy-handed in its combination
of the scholarly and the sanctimonious. Consider, for example, the first lines of his riddle
on the Sun and the Moon (Enigmata LXXIX):

Non nos Saturni genuit spurcissima proles
Iupiter, immensum fingunt quem carmina vatum,
nec fuit in Delo mater Latona creatrix;
Cynthia non dicor nec frater Apollo vocatur,
sed potius summi genuit regnator Olimpi,
qui nunc in caelis excelsae praesidet arci.
285

[The foul offspring of Saturn, namely Jupiter whom the songs of poets picture as
mighty did not produce us, nor was Latona our mother on Delos; I am not called
Cynthia and my brother is not Apollo. Rather, the ruler of high Olympus, Who now
resides in His heavenly citadel on high, produced us.]
286


Bede, obviously, saw no need to go so far; rather, as an example of acceptable pre-Christian
verse, he simply chose Lucan over Vergil, thus making any allusions to pagan beliefs
unnecessary. This is another variant of the technique of omission which Bede also employs
elsewhere in his treatise: we may recall how the names of there Greek deities Phoebus and
Euterpe were simply omitted from Bedes presentation of final syllables, which otherwise
was an almost verbatim reproduction of Servius and Victorinus. Even there, Bede had left in
such ancient but non-godly and therefore innocuous figures as Tanaquil and Dido;
287
here, the
reference to Crassuss ghost was apparently something that even Bedes Christian
sensibilities could manage.

Yet another stylistic device which, in Bedes opinion, is typical of the best kind of verse and
which merits closer observation here is the typically medieval feature that Curtius has termed
verse-filling asyndeton (versefllendes Asyndeton).
288
This device consists in the use of
asyndetic lists of words belonging to the same word class and is a hallmark of the poetic style
of Venantius Fortunatus.
289
Bede was certainly aware of this, and, indeed, all his examples of
this device are from Venantius:

284
Trans. P. McBrine, http://www.pmcbrine.com/translations/ 2010.
285
Ehwald 1919, 134.
286
Trans Lapidge 1985, 87.
287
DAM 6, 8; 6, 85.
288
Curtius 1948, 289.
289
Tardi 1927, 261.
96


Aliquando versum nominibus tantum perficere gratum est, ut Fortunatus:
Lilia, narcissus, violae, rosa, nardus, amomum,
oblectant animos germina nulla meos (Ven. Fort. carm. 8, 3, 237-238).
Quod idem et in propriis fecit nominibus:
Sarra, Rebecca, Rachel, Hester, Iudith, Anna, Noemi.
quamvis praecipue culmen ad astra levent (Ven. Fort. carm. 8, 3, 99-100).
Fecit et in verbis:
Blanditur, refovet, veneratur, honorat, obumbrat,
et locat in thalamo membra pudica suo (Ven. Fort. carm. 8, 3, 127-128).
290

[Sometimes, it is pleasing to compose a line with nouns alone, as Fortunatus did in:
The lily, the narcissus, violets, the rose, the nard, the spiceplant,
no such plants please my mind.
The same poet also did this with proper nouns:
Sarah, Rebecca, Rachel, Esther, Judith, Anna, Naomi,
although they especially raise their eminence to Heaven.
And he did it with verbs:
He cherishes, restores, respects, honours, protects,
and places her chaste limbs in his marriage bed.]
291


This time, Bede does not warn his reader against the overuse of such word-lists, although he
well could have, given some of the more outr examples in Venantiuss poetry.
292
Although
Bede here commends this construction, he himself appears to have made little use of it in his
own verse, based as it is on the model of Sedulius, and generally not very exuberant in tone.
It seems that Bedes endorsement of Venantiuss idiosyncratic use of asyndeta may have
been little more than a whim. Nevertheless, Bedes enthusiastic support for the device
presentation may certainly have contributed to its popularity in later Anglo-Latin poetry,
293

and medieval verse in general.

290
DAM 11, 38-47.
291
Trans. Kendall 1991, 104.
292
e.g. the over-the top description of a leper in Ven. Fort. Mart. 1, 490-492:
Vir maculis varius, cute nudus, vulnere tectus,
tabe fluens, gressu aeger, inops visu, asper amictu,
mente hebes, ore putris, lacerus pede, voce refractus.
[A man mottled with spots, with naked skin, covered with wounds,
dripping with pus, with infirm step, failing sight and rough clothing,
dim mind, putrescent mouth, lacerated feet and broken voice.]
293
The device seems to have been particularly popular among the representatives of the so-called hermeneutic
school of Insular Latin poets. See Lapidge 1975, 107-111.
97

2.6. Word division and caesurae

The twelfth chapter of Bedes De arte metrica, titled De scansionibus vel caesuris heroici
versus, is a highly condensed presentation of the types of word division and caesurae
employed in hexameter verse. Bedes analysis follows Audaxs presentation of word division
(in Audax caesurae, in Bede scansiones) and supplements it with what he considered the
principal types of caesurae. Relying as he does on previous grammarians, Bede is far from
exhaustive in his presentation, recognising the penthemimeral and hepthemimeral caesura,
the trochaic caesura of the third foot and the bucolic diaeresis; unlike some of his
predecessors he does not discuss the respective importance of these types of word-break. The
simplicity of Bedes presentation is, first and foremost, the result of his reliance on Late Latin
grammarians: as Bedes own use of sense-pauses, judiciously placed in widely varying parts
of the hexameter line, testifies, his metrical sophistication on a practical level exceeded what
he presented in his grammatical writings. The chapter ends with a brief but vague explanation
of cola and commata, illustrated with a quotation from Scripture, which demonstrates both
Bedes unwillingness to impose too much theory on his reader and his Christianising
tendency.
The role of word division is central in Latin hexameter poetry; arguably more
so than in Greek verse. The main reason for this is the nature of the Latin system of
accentuation. Latin words almost invariably have an accent on the penultimate or
antepenultimate syllable, and as a result, in hexameter verse, strong caesurae, or word-breaks
in the middle of a foot, create a clash between accent and ictus, whereas other types of word-
break (diaereses and trochaic, or weak, caesurae) have the opposite effect. As in classical
hexameter verse the general tendency is to make the accent clash with the ictus in the middle
of the line and coincide with it towards the end, the preponderance of strong caesurae in Latin
hexameter verse is marked.
294
Most Latin hexameter lines have a strong caesura in the third
foot of the line, known as the penthemimeral or semiquinarian caesura (e.g. Verg. Aen. 1, 1
Arma virumque c|no / Tr|iae qui primus ab oris), which more or less ensures that, at least
in the third foot, accent and ictus do not coincide. An alternative solution is to have a strong
caesura in the fourth foot of the line (known as hephthemimeral or semiseptinarian); in
developed hexameter verse this is always supported by another strong caesura (trithemimeral)

294
e.g. Wilkinson 1963, 120-121.
98

in the second foot (e.g. Aen. 3, 490 Sic oculos / sic ille manus / sic ora ferebat).
295

Diaereses (word-breaks coinciding with the end of a foot) and trochaic caesurae, on the other
hand, produce feet where accent and ictus coincide. An extreme example of the former is
Enniuss notorious Sprsis | hstis | lngis | cmpus | splndet et horret (Varia 14V),
whereas even in Vergil we occasionally find lines where the strong caesurae are neglected. In
Aen. 4, 486 Sprgens | mida | mlla / so|prife|rmque pa|pver (Sprinkling wet honey
and sleep-inducing poppy), all the word breaks are either trochaic or coincide with the ends
of feet, resulting in a perfect coincidence of accent and ictus, and a line that is, literally,
soporifer. Apparently Roman grammarians of the classical period were already aware of the
importance of word division in Latin hexameter verse, and realised that words should
preferably not end with feet in the middle of a line, although it does not seem that they were
able to present their findings in the form of an actual system, and they appear to have been
largely unaware of their prosodic implications. Aulus Gellius testifies that Varro had already
pondered these issues and emphasised the role of the penthemimeral caesura:

In longis versibus, qui hexametri vocantur, item in senariis, animadverterunt metrici
primos duos pedes, item extremos duo, habere singulos posse integras partes
orationis, medios haut umquam posse, sed constare eos semper ex verbis aut divisis
aut mixtis atque confusis. M. etiam Varro in libris disciplinarum scripsit observasse
sese in versu hexametro, quod omnimodo quintus semipes verbum finiret et quod
priores quinque semipedes aeque magnam vim haberent in efficiendo versu atque alii
posteriores septem, idque ipsum ratione quadam geometrica fieri disserit.
296

[In long verses, known as hexameters, and similarly in senarii, prosodists noticed that
the first two feet, and similarly the two final ones, could each contain complete
words, but the middle ones hardly ever, consisting always of words that are either
divided, mixed or blended. Also Marcus Varro, in his books on the disciplines, wrote
that he had observed that in hexameter verse the fifth half-foot invariably ended a
word and that the first five half-feet had the same importance in making up the verse
as the seven following ones, and he asserted that this took place through a
geometrical law of some kind.]

Varros (or Gelliuss) observation seems to overplay the role of the penthemimeral caesura,
undeniably predominant in Latin hexameter verse, while ignoring the hepthemimeral caesura.
Subsequent generations of grammarians seem to have done little to refine this presentation,
apart from the observation that in hexameter verse words should generally not end with feet.
Diomedes, sharing many other grammarians penchant for descriptions of good and bad

295
Raven 1965, 94-96.
296
Gell. 18, 15.
99

hexameters, inevitably took this even further and presented as the best kind of all hexameter
lines a type where as many words as possible cut across feet:

Versus heroicus is dignitate primus est et plenae rationis perfectione firmatus ac
totius gravitatis honore sublimis, multaque pulchritudinis venustate praeclarus, qui
sine ulla coniunctione quascumque alias orationis partes ita mutuis inter se
conexionibus colligat, ut in scansione propria nullus pes nisi novissimus tantum
[interdum] integram partem orationis includat.
297

[The kind of hexameter verse that is foremost in dignity, strong in the full perfection
of its art, sublime in the honour of all its weight and brilliant in the charm of its
plentiful beauty is the one where, without any conjunction, all words are assembled
and connected with each other in such a way that in its scansion no foot except
sometimes the last one encloses an entire part of speech.]

Diomedess example of such a remarkable hexameter line is Vergils (Aen. 4, 129)
Ocea|num intere|a sur|gens Au|rora re|linquit (Meanwhile the rising dawn leaves behind
the ocean). This aesthetic judgement was shared by Audax, who in his Excerpta de Scauro
et Palladio (gramm. VII, 340) attempted to create a typology of hexameter verse based on its
word division. Although flawed and arguably impractical, Audaxs presentation was adopted
in slightly altered form by both Aldhelm and Bede in their respective presentations of
hexameter caesurae. Audaxs analysis of hexameter word-division is fundamentally sound, as
it grasps the importance of whether word-endings coincide with the ends of feet or not;
however, it relies heavily on hexameter types that are unusual or virtually non-existent, and
must therefore be considered ill-presented and misleading. Audax lists four categories of
word division which also appear in other grammarians, sometimes under other names, and
sometimes even in the lists of special hexameter lines compiled by such authors as
Diomedes and Marius Victorinus.
298
Audax refers to these types of word division with the
term caesurae, which may strike the modern reader as an unusual practice, but the meaning
of the word in antiquity was not quite fixed. Early Greek writers on metrics were, as far as we
know, not preoccupied with caesurae and it is uncertain to what extent they had either a
terminology or a theoretical framework for them. The Greek :onp (for which caesura is a
calque) was, to our knowledge, first employed in a cogent manner by the third-century
Aristides Quintilianus, and it originally meant a segment of verse rather than an end of a

297
gramm. I, 495, 27-34.
298
e.g. Audaxs versus districtus, where words end with feet, is identical with versus partipes (foot-divided
verse), praised by Diomedes as a versus optimus (gramm. I, 499, 10-12) and castigated by Marius Victorinus as
a bad one (gramm. VI, 71, 24-27).
100

segment.
299
Although, already in late antiquity, several grammarians used the term caesura in
the modern sense to denote a word-break, Audax appears to use it for a system of word-
breaks. Audaxs caesurae include four types which he calls coniunctus, districtus, mixtus
and divisus:

De caesuris. Quot sunt species in caesura hexametri versus? Quattuor. Quae sunt?
Coniunctus, districtus, mixtus, divisus. Coniunctus qui est? Qui in scandendo ita
concatenatus sibi est, ut nusquam finito sensu divisa inter se verba ponantur, quod
genus versificationis laudabile habetur, veluti est:
Infandum, regina, iubes renouare dolorem (Verg. Aen. 2, 3).
Qui districtus? Qui in scandendo sensum seu partes orationis separatas habet, ut puta
veluti:
Dic mihi, Clio, quisnam primus fingere versus.
Qui mixtus? Qui utrumque in se habet, ut in quibusdam coniunctus, in quibusdam
separatus sit, ut puta:
Hic currus fuit, hoc regnum dea gentibus esse (Verg. Aen. 1, 17).
Qui divisus? Qui in priapeo deprehenditur metro. Quale est metrum priapeum? Cum
in hexametro versu primi tres pedes concatenati inter se, a reliquis tribus sequentibus
divisi separatique sunt, ut in bucolicis:
Aut Ararim Parthus bibet aut Germania Tigrim (Verg. ecl. 1, 62),
Utque viro Phoebi chorus adsurrexerit omnis (Verg. ecl. 6, 66),
Est mihi namque domi pater, est iniusta noverca (Verg. ecl. 3, 33).
300

[On caesurae. How many ways of dividing a hexameter verse are there? Four. What
are they? The conjunctive, the districtive, the mixed and the divided. Which line is
the conjunctive? That which, when scanned, is bound together in such a way that
parts of speech are never placed divided among themselves when their meaning is
completed; this type of line is considered worthy of praise, as in:
Infan|dum, re|gina, iu|bes reno|vare do|lorem
(O queen, you force me to relive an unspeakable pain).
Which line is districtive? That, where the meaning or the parts of speech are
separated in scansion, as for example:
Dic mihi, | Clio, | quisnam | primus | fingere | versus
(Tell me, o Clio, who was the first to compose verse).
Which line is mixed? That which has both types so that words are sometimes
conjoined, sometimes separated, as in:
Hic cur|rus fuit, | hoc reg|num dea | gentibus | esse
(Here was her chariot, this the goddess wanted to be the ruler of peoples).
Which line is divided? That which is seen to be in the priapeian metre. What is the
priapeian metre? When the first three feet of the hexameter are bound together and
divided and separated from the three following ones, as in the Bucolics:
Aut Ara|rim Par|thus bibet | aut Ger|mania |Tigrim
(Or the Parthian will drink from the Sane, or Germany from the Tigris),
Utque vi|ro Phoe|bi chorus | adsur|rexerit | omnis
(And how all of Phoebuss choir rose to the man),
Est mihi | namque do|mi pater, | est in|iusta noverca

299
Bassett 1917, 86-87; Bassett 1919, 348-349. Aristides is also the first author to differentiate between a
caesura and a diaeresis, a distinction not generally made in antiquity.
300
gramm. VII, 340, 6-23.
101

(For I have a father at home, and an unjust stepmother).]

We can see the logic behind Audaxs presentation, although it is not well commented and
potentially misleading: Audaxs versus districtus is a line where all the word-endings
coincide with the foot-division; it has no central caesura, and as such, it is a genuine rarity in
Latin verse (it is telling that Audax has not been able to illustrate it with a quotation but has
instead used a wholly artificial line, presumably of his own making). Its opposite is the versus
coniunctus, where none of the words end with the feet, and this type of verse Audax actually
ventures to commend. What Audax calls the mixtus is a line with both types of word-division.
Audax does not make it plain that what really counts is the word division in the middle of the
line, or what we call the main caesura, and therefore his presentation is plainly useless as a
means of teaching verse composition. This is underlined by his inclusion of what he terms
divisus, or a divided line, a verse-type that is arguably little more than a curiosity: it is simply
a line where a penthemimeral caesura is followed by a pyrrhic word (in the cited examples
bibet, chorus and pater, respectively); the following diaeresis lends it an air of being cut
neatly in half, an unusual but by no means abnormal effect. Audaxs presentation also leaves
much to be desired as far as clarity is concerned: he does not discuss caesurae, and barely
mentions metrical feet, and his discourse on divided or separated words would border on
the impenetrable if it were not for his cited examples.
Both Aldhelm and Bede undertook to improve Audaxs presentation by
incorporating it with a presentation of the principal caesurae of hexameter verse. In
developed Latin verse, a hexameter invariably has either a penthemimeral or a hepthemimeral
caesura, which both Aldhelm and Bede are quick to point out in their discussion of Audaxs
versus districtus.
301
With typical perversity, Aldhelm has completely reshuffled Audaxs
terminology: in his presentation coniunctus has become districtus, districtus had become
divisus, and divisus is known simply as priapeius.
302
Remarkably, Aldhelm has also
undertaken to Christianise Audaxs cited material, providing several quotations from widely

301
Ehwald 1919, 93. Aldhelm implies that lines without caesurae are an archaic feature, bringing up what he
sees as major advances in modern (Christian?) versification. This comment seems to foreshadow Bedes
dichotomy of pre-Christian and Christian techniques: Sed huiuscemodi versum pentimemere vel eptimemere
carentem modernus usus in exametero dactilico non libenter admittit. (But this kind of line, lacking a
penthemimeral or hephthemimeral caesura, is not freely admitted in the dactylic hexameter by modern
practice.) Trans. Wright 1985, 208.
302
Ehwald (1919, xix) has suggested that Aldhelms presentation may be based on an alternative, unknown
source rather than Audax. However, his wording is largely identical with Audaxs, and the change in
nomenclature may simply be Aldhelms attempt to improve Audaxs presentation. R. Leotta 1980, 248.
However unaccountable this may seem, it is parallel to Aldhelms confusing manner of using a dialogue form
with the Greek letters A and M, not for oioookoo (teacher) and no0p:p (pupil) as one might expect, but the
other way around, apparently for discipulus and magister.
102

varying Christian sources,
303
although, in the main, Audaxs Vergilian examples are still
there, as is his artificial example of the versus districtus. The wealth of quoted lines is
striking; with what appears to be intentional humour on the authors part (or perhaps a dig at
previous grammarians) the disciple of Aldhelms dialogue-form presentation snaps at his
teacher: Melius exemplis quam nuda verbositate informor (I would learn better from
examples than unsupported eloquence).
304
Aldhelms improvements on Audaxs
presentation are, however, limited, and he confuses his exposition further by adding a
discussion of different types of pathe or passiones of hexameter verse, metrical liberties that
occur in Homeric verse but are totally irrelevant for the Latin hexameter.
305

In his presentation of word division in hexameter verse, Bede essentially
paraphrases Audax, but his alterations to both word and content are conspicuous:

Scansionum autem in versibus sunt species quattuor, coniuncta, districta, mixta,
divisa. Coniuncta, quae ceteris laudabilior habetur, illa est, ubi nusquam pes cum
verbo finitur, ut:
Inmortale nihil mundi conpage tenetur (Iuvenc. praef. 1).
Districta, ubi verba cum pedibus terminantur, ut:
Haec tua sunt, bona sunt, quia tu bonus omnia condis (In laude Cerei 1,1),
quam versificationis speciem rarissime invenies. Nam et si non post duos vel tres
pedes syllaba superfuerit, quod pentimemerim et eptimemerim uocant, ratus haberi
versus nequit, sicut hic post duos pedes sunt, post tres tu superest.
Mixta est scansio quae utrumque in se habet, ut in quibusdam coniunctus, in
quibusdam vero separatus sit versus:
Nobis certa fides aeternae in saecula laudis (Iuvenc. praef. 17);
et:
Pacificos Deus in numerum sibi prolis adoptat (Iuvenc. 1, 465).
Divisa est, ubi primi tres pedes concatenati inter se a reliquis pedibus separati sunt,
ut:
Inde Dei genetrix pia virgo Maria coruscat (Ven. Fort. carm. 8, 3, 25);
et Prosper:
Corde patris genitum creat, et regit omnia Verbum (Prosp. epigr. 104, 5).
306

[There are four types of scansion in lines of heroic verse; the conjunctive, the
districtive, the mixed and the divided. The conjunctive scansion, which is considered
more praiseworthy than the others, is that in which the end of a word never coincides
with the foot division, as in:
Inmor|tale ni|hil mun|di con|page te|netur
(Nothing is imperishably preserved by the structure of the world).

303
Aldhelm cites from Arators De actibus apostolorum, Prospers Epigrams, an anonymous Latin translation of
the Oracula Sibyllina, Symphosiuss Aenigmata and one line of his own, attributing it, with probably
disingenuous modesty, to a poet. Wright 1985, 207-209.
304
Ehwald 1919, 92; trans. Wright 1985, 207.
305
Wright 1985, 187.
306
DAM 12, 2-22.
103

The districtive scansion is that in which the ends of words coincide with the foot
endings, as in.
Haec tua | sunt, bona | sunt, quia | tu bonus | omnia | condis
(These things are yours, they are good because You who make all things are
good.)
You very rarely find this kind of scansion, because, unless there is a word break after
the first syllable in the third or the fourth foot, which is called a penthemimeral or a
hepthemimeral caesura, the verse cannot be considered acceptable. In this case the
word sunt follows the second foot and the word tu follows the third foot. The mixed
scansion is that which combines the two, so that the line is conjunctive in some feet,
but districtive in others, as in:
Nobis | certa fi|des ae|ternae in | saecula | laudis
(Firm faith will confer the immortal glory of eternal praise through the ages on
me);
and:
Pacifi|cos Deus | in nume|rum sibi | prolis a|doptat
(God adopts the peacemakers into the ranks of his elect).
The divided scansion is that in which the first three feet of the verse are joined by
consecutive scansion and separated by diaeresis from the remaining feet, as in:
Inde De|i gene|trix pia | virgo Ma|ria co|ruscat
(Then the holy mother of God, the virgin Mary, trembles);
and in this line of Prosper:
Corde patris geni|tum creat, | et regit | omnia | Verbum
God creates the son in the heart of the Father, and rules all things through the
Word).]
307



As we can see, Bede has retained Audaxs typology of word-division, but subjected its
presentation to some minute and apparently well thought-out alterations. Firstly, Bede has
discarded Audaxs broader use of the term caesura for scansio, obviously because he
elsewhere uses the word caesura in the modern sense to denote a word-break and has wanted
to avoid unnecessary ambiguity. Secondly, the Christianisation of the quoted examples has
been carried out more thoroughly than in Aldhelms De metris: Bede has eliminated Audaxs
Vergilian citations altogether in favour of J uvencus, Venantius Fortunatus and Prosper.
Bedes discussion of the districta scansio is also a clear improvement on Audaxs utterly
artificial presentation, having, as it does, an authentic quotation for its model, the anonymous
poem known as De laude cerei quoted by Augustine in his City of God.
308
The cited verse is

307
Trans. Kendall 1991, 109-111.
308
Aug. civ. 15, 22:
Quod in laude quadam cerei breviter versibus dixi:
Haec tua sunt, bona sunt, quia tu bonus ista creasti.
Nil nostrum est in eis, nisi quod peccamus amantes
ordine neglecto pro te, quod conditur abs te.
[Which I shortly said with verses in a certain praise of the paschal candle: These things are yours,
they are good because you who are good created these things. There is nothing of us in them except
the sin we commit in ignoring the order of things and loving your creations instead of you.]
104

metrically not quite as implausible as Audaxs example, as the central caesura is not totally
neglected, which Bede takes care to point out: in Bedes words, sunt follows the second foot
and tu the third one (i.e. the lin has both a penthemimeral and a hephthemimeral caesura).
The presentations of mixed and divided scansion follow Audax more closely than those
of the conjunctive and districtive ones, but Bedes examples of these types are taken
from Christian poetry. All in all, Bedes presentation is more lucid than Audaxs; one may
contrast Bedes clear-cut definition of the coniuncta scansio (ubi nusquam pes cum verbo
finitur) with Audaxs obscure ut nusquam finito sensu divisa inter se verba ponantur.
Bedes description of the actual caesurae is terse and simple. In one respect it
can be regarded as simpler than those in his Late Latin predecessors: Bede does not
differentiate, as several of his forerunners did, between primary and secondary caesurae.
This dichotomy is probably based on practical observation: a Latin hexameter line invariably
has either a penthemimeral or a hephthemimeral caesura, and in most treatises these are
accorded the first place; other types of caesura that are mentioned as secondary are the
trochaic caesura of the third foot, the even less usual trochaic caesura of the fourth foot
309
and
(although in our terminology not caesura but usually discussed as such by the grammarians)
the bucolic diaeresis between the fourth and fifth feet.
310

That the trochaic caesurae are discussed at all reflects the conservative nature of
Latin grammar, as does the fact that the trithemimeral caesura (a strong word-break in the
second foot), all but compulsory in hexameter lines with a hepthemimeral caesura, never
made it to the terminology of the Late Latin grammarians.
311
Bedes presentation of caesurae
is taken from Maximus Victorinus (gramm. VI, 240). Obviously Bede did not expect much
prosodic sophistication from his prospective readers, as he has added a warning, taken from
Diomedes (gramm. I, 497) not to be led astray by the concept of trochaic caesurae: Bede,
as Diomedes, emphasises that trochees as feet do not occur in the middle of a hexameter line.
All cited examples are taken from Seduliuss Carmen paschale:


It is probable that the hexameter poem is not by Augustine himself, who even by his own admission knew
nothing of syllable quantity (Aug. mus. 3, 2). Although Bedes quotation does not follow the usual reading in
Augustine it is possible that Bede attributed the poem to him.
309
Most grammarians do not mention this caesura, but Marius Victorinus, Priscian and, curiously, Aldhelm do.
310
See Klopsch 1972, 66. The dichotomy of primary and secondary caesurae appears in Marius Victorinus
(gramm. VI, 64, 31 65), the Fragmentum Sangallense (gramm. VI, 638, 5-11), Iulius Severus (gramm. VI,
645) and Priscian (gramm. III, 460). Aldhelm (Ehwald 1919, 97) Bede and Cruindmel (Huemer 1883, 35), who
generally follows Bede, do not differentiate between the roles of the caesurae.
311
Klopsch 1972, 65-66: Einerseits werden die weiblichen Zsuren mitgeschleppt, andererseits wird die lngst
eingebrgerte Trithemimeres nicht in den Katalog aufgenommen.
105

Nec minus caesurarum intuendus est status, quae et ipsae sunt quattuor,
pentimemeris, eptimemeris, cata triton trocheon, bucolice ptomen. Pentimemeris est,
ubi post duos pedes invenitur semipes, qui versum dividat et partem terminet
orationis, ut cum sua gentiles. Eptimemeris, ubi post tres pedes invenitur syllaba,
ut cum sua gentiles studeant (Sedul. carm. pasch. 1, 17). Dicta autem pentimemeris
et eptimemeris Graece quasi semiquinaria et semiseptinaria, quia cum per spondeos
fiunt, haec quinque syllabis, illa constat septem, et in hac quinta syllaba semipedem,
in illa tenet septima. Cata triton trocheon, ubi tertio loco invenitur trocheus, non quod
in medio versu esse possit trocheus, sed sublata una de dactylo syllaba remanet
trocheus, ut grandisonis pompare modis (Sedul. carm. pasch. 1, 18). Bucolice
ptomen, ubi post quattuor pedes non aliquid remanet, ut:
Semper principium, sceptrum iuge, gloria consors (Sedul. carm. pasch. 1, 34);
et:
Cristus erat panis, Cristus petra, Cristus in undis (Sedul. carm. pasch. 1, 159).
Quae caesura inde nomen habet, quod in bucolicis saepius inveniatur.
312

[Attention should be given as well to the essential features of the caesuras of heroic
verse, of which there are also four: the penthemimeral, the hephthemimeral, the
trochaic, and the bucolic diaeresis. In the penthemimeral caesura a half-foot which
coincides with the end of a word and divides the line is found after the first three feet,
as in cum sua | genti|les ||. In the hepthemimeral caesura the final syllable of a word
is found after the first three feet, as in cum sua | genti|les stude|ant || (since the
pagans study their own fictions). These are called penthemimeris and
hephthemimeris in Greek; and similarly semiquinaria and semiseptinaria in Latin,
because when they are fashioned with spondees the former consists of five and the
latter of seven syllables and in the former the fifth syllable makes up a half-foot and
in the latter the seventh syllable. The trochaic caesura is that in which a trochee is
found in the third foot not that there can be a trochee in the middle of a line, but
when you take the last syllable from a dactyl a trochee remains, as in grandiso|nis
pom|pare || mo|dis (to make a show with swelling measures). In the bucolic
diaeresis there is a word break after the first three feet, as in:
Semper principium, sceptrum iuge, gloria consors
(Always power, dominion forever, shared glory):
and:
Cristus erat panis, Cristus petra, Cristus in undis
(Ghrist was the bread, Christ was the rock, Christ was in the waters).
This caesura takes its name from the fact that it is quite often found in bucolic
verse.]
313


As we can see, Bede, as his predecessors, overemphasises the role of the trochaic caesura.
Although it was popular with Greek poets, it cannot exist on its own in Latin verse: it is
normally always supported by trithemimeral and hephthemimeral caesurae, and usually the
hephthemimeral caesura constitutes the primary break of such lines. This is certainly the case

312
DAM 12, 23-39.
313
Trans. Kendall 1991, 111-113.
106

in the line which Bede cites as an example of the trochaic caesura, as we can see if we (unlike
Bede) observe the line in its entirety:

Grandisonis pompare modis, / tragicoque boatu.

As a matter of fact, in a hexameter line with trithemimeral and hephthemimeral caesurae
there need be no third-foot caesura at all:

Inde to|ro / pater | Aene|as / sic orsus ab alto (Verg. Aen. 2, 2)

The retention of the bucolic diaeresis in the terminology of the grammarians is arguably quite
as artificial: in the examples cited by Bede, the penthemimeral caesura is at least equally
important (e.g. Semper principium, / sceptrum iuge, gloria consors). Bedes presentation of
caesurae might strike one as unusually hidebound, derivative and largely unhelpful from a
didactic point of view: it, unhappily, presents Audaxs artificial typology of word-division,
which is largely based on non-existent or unusual types of hexameter verse and ignores
normal metrical practice. In his presentation of the actual caesurae, Bede follows the example
of previous grammarians. He emphasises the role of the penthemimeral and hephthemimeral
caesurae, certainly the most important ones, but, probably out of adherence to traditional
terminology, also discusses the trochaic caesura of the third foot and the bucolic diaeresis,
which in reality are little more than side issues. In this respect Bede is neither better nor
worse than his predecessors, although probably the most useful part of his presentation is the
almost parenthetical statement that a hexameter line should have either a penthemimeral or a
hephthemimeral caesura. Bede acknowledges neither the existence nor the importance of the
trithemimeral caesura, but neither did any of his predecessors: apparently even Varro and
Gellius limited their observations on hexameter caesurae to the third and fourth feet of the
line. It must be noted that in his metrical Vita Cuthberti Bede followed classical practice
meticulously in having a trithemimeral caesura in all lines with a hephthemimeral one.
It would, indeed, seem that Bede primarily intended his chapter on caesurae to
be a reference work of well-established prosodic terminology rather than a guide to the use of
caesurae. This would explain his discussion of largely irrelevant caesura types as well as his
neglect of more well-established types of word division. Bedes own work testifies that he
was well aware of the good use to which word-breaks in various parts of the line could be
put. Although even in Bedes verse the penthemimeral caesura dominates, his sense-pauses
107

often appear in the most unusual places, as we noted in the opening of his Vita Cuthberti: in
line two (donavit, / tetricas humanae noctis ut umbras) the first clause ends with the
trithemimeral caesura (for which Bede did not have a name) and in line five with the even
less usual trochaic caesura of the second foot (lux sit summa, / Deus sanctos quoque iure
lucernae). If we observe Bedes presentation of Audaxs divisa scansio, we can also see that
he was certainly aware of the role of the diaeresis after the third foot of the line, but, alas, he
did not have a name for such a diaeresis and did not venture to coin one. It is also noteworthy
that chapter twelve is the only portion of Bedes treatise where he actually uses the term
caesura: elsewhere, as in his discussion of common syllables in chapter three, he relies on
circumlocutions (e.g. DAM 3, 23-24: plenis pedibus superfuit) which may seem tortuous to
us but reflect earlier usage.
The end of Bedes chapter contains a brief discussion of cola and commata.
Here again, the probable purpose is to acquaint the reader with some central terms of
grammatical theory. The precise difference of colon and comma is something that does not
always come across unambiguously in earlier literature, and Bede has chosen to conclude that
in prose the terms are virtually interchangeable. When it comes to verse, however, Bede
seems to have adopted Pompeiuss idiosyncratic idea that the difference is one of word-
division: a segment of verse ended by a strong caesura is a comma, whereas one terminated
by a diaeresis after the second foot is a colon.
314
In practice, Bedes definition of comma is
simply another way of describing the penthemimeral caesura and probably reflects the
chapters purpose of being a compendium of metrical terminology:

Item ubi post duos pedes superest syllaba, comma dicitur; ubi post duos pedes nihil
remanet, colon vocatur. Quae tamen nomina apud oratores indifferenter ponuntur, qui
integram sententiam periodon appellant; partes autem eius cola et commata
dicuntur.
315

[A phrase which is two and a half feet long is said to be a comma; a phrase which
is only two feet long is called a colon. But these names are employed without
regard to distinctions by professors of rhetoric, who call the whole sentence a
period and its parts colons and commas.]
316


Bedes definitions of commata and cola in verse are historically imprecise: the concept of the
colon was originally derived from poetry, and originally corresponded with a sense-unit

314
gramm. V, 133, 23 134, 2.
315
DAM 12, 42-46.
316
Trans. Kendall 1991, 113.
108

roughly equivalent to a hexameter line,
317
whereas a comma is generally described as
something shorter. As we can see, the Pompeian definition adopted by Bede turns things
the other way around, as both terms are applied to segments of hexameter verse, and a colon
has actually become the shorter one of the two. Bede ends his chapter with some samples of
cola and commata in prose, where, as he contends, the two terms are wholly undifferentiated.
The Biblical quotation gives Bedes simple presentation the required Christian veneer; the
passage is remarkably similar to De arte metricas companion work De schematibus et tropis
with its discussion of rhetorical figures and their Biblical illustrations:

Ut puta, sustinetis enim, siquis vos in servitutem redigit (2 Cor. 11:20), colon est;
si quis devorat, colon est; si quis accipit, colon est; si quis extollitur, et cetera
usque ad plenam sententiam, cola sunt et commata.
318

[For example: For you suffer if a man bring you into bondage is a colon; if a man
devour you is a colon; if a man take from you is a colon; and if a man be lifted
up, and so on to the end of the sentence are colons and commas.]
319



To sum up: Bedes chapter on word division and caesurae must be primarily viewed as an
introduction to the traditional terminology of hexameter caesurae rather than a guide to their
actual employment. Although Bede has attempted to elucidate his presentation, especially
when it comes to Audaxs typology of caesurae, it is evident that he feels uncomfortable with
the terminology which he discusses, and the chapter seems lacking in the practical approach
otherwise typical of Bedes treatise.


2.7. Elision and hiatus

The prosodic feature known variously as elision or synaloephe has traditionally been one of
the major stumbling-blocks for students trying to learn the proper scansion of Latin verse. In
practice, elision means that if a word that ends in a vowel (or an m) is followed by a word

317
Thus, for instance, in Demetrius (1,1). Cicero has the analogy to hexameter verse in his Orator 221-222,
where he uses the Latin membra for cola: Constat enim ille ambitus et plena comprehensio e quattuor fere
partibus, quae membra dicimus, ut et auris impleat et neque brevior sit quam satis sit neque longiorE quattuor
igitur quasi hexametrorum instar versuum quod sit constat fere plena comprehensio. (For the full
comprehensive period consists of roughly four parts, which we call members, that it may satisfy the ear and be
neither shorter nor longer than necessary... Therefore, the full period consists of four parts equivalent to a
hexameter line.).
318
DAM 12, 46-49.
319
Trans. Kendall 1991, 113.

109

with an initial vowel (or an h), the final vowel (or vowel and m) of the previous word is left
unpronounced (e.g. multum ille >mult ille) or this, at least, is the view generally taken by
the grammarians, and one that has been considered good enough for the classroom.
Nevertheless, the precise nature of elision (or synaloephe) has, in modern literature, been a
matter of some debate: it is unclear whether an elided final vowel was truly dropped, whether
it became fused with the following initial vowel, or whether it was glossed over in some other
manner;
320
nevertheless, the feature is assumed to have existed, at least to some extent, even
in regular speech.
There are several things which further complicate the matter of elision. Firstly,
elision is normally not indicated in writing, and it can be broken against, resulting in what is
known as hiatus. Generally, the only way to verify whether elision takes place or not, is to see
whether the line scans. Secondly, the verb form est acts differently from other words with an
initial vowel. If est follows a word-final vowel (or m), the initial e is eliminated by a process
known as prodelision or aphaeresis (e.g. multum est >multumst).
321
Unlike regular elision,
prodelision is often spelt out in manuscripts of archaic drama and inscriptional verse, but
usually not in manuscripts of literary classical poetry, and Late Latin grammarians generally
seem to have been ignorant of its existence.
Bedes treatise on metre is unusually advanced in its presentation of elision as a
form of verse technique, and, in this respect, he follows in the footsteps of Aldhelm.
322

Earlier grammarians (e.g. Donatus) followed the less practical course of presenting elision
among what they termed metaplasms, or metrical licences. In Donatuss Ars maior, for
instance, elision, together with other metaplasms, is placed at the end of the treatise, together
with schemes and tropes, and presented separately from actual metre. Donatuss list of
metaplasms reads as follows; it is worthy of note that he in no way differentiates between
structures that are mere prosodic curiosities and those that are an integral part of verse
technique:

Metaplasmus est transformatio quaedam recti solutique sermonis in alteram speciem
metri ornatusve causa. Huius species sunt quattuordecim: prosthesis, epenthesis,

320
A particularly intricate system of synaloephe, which combines complete omission of vowels with synizesis
(vowel fusion) and final u:s and i:s turned into consonants is offered by Allen 1978
2
, 78-82, but even he admits:
However, if the English reader chooses to apply elision in all cases of vowel junction, and thereby avoid the
uncertainties inherent in other solutions, he will at any rate be no further removed from classical practice than
some of the Latin grammarians were; and only very rarely will such reading lead to any ambiguity.
321
Allen 1978
2
, 123. In archaic verse, prodelision also takes place after word-final s.
322
Wright 1985, 185.
110

paragoge, aphaeresis, syncope, apocope, ectasis, systole, diaeresis, episynaliphe,
synaliphe, ecthlipsis, antithesis, metathesis.
323

[Metaplasm is a type of transformation of direct and unbound speech into another
kind for the sake of metre or embellishment. There are fourteen kinds of metaplasm:
prosthesis, epenthesis, paragoge, aphaeresis, syncope, apocope, ectasis, systole,
diaeresis, episynaliphe (=episynaloephe), synaliphe (=synaloephe), ecthlipsis,
antithesis and metathesis.]

An even more extensive presentation of metaplasms is given by Consentius in his De
barbarismis et metaplasmis.
324
Bede, on the other hand, has limited his discussion of
metaplasms to those that have a practical effect on the scansion of hexameter verse (i.e.
synaloephe, synaeresis, diaeresis and syncope) and ignores the rest. His terminology is also
simplified even within the scope of the phenomena which he discusses: unlike many of his
predecessors, he uses one term (synalipha) for all forms of elision, whereas, e.g., Isidore,
Donatus, Consentius and even Aldhelm use two different terms, synalipha and ecthlipsis, for
different (and variously defined) types of elision.
325
Bede also discusses other metaplasms in
his chapter on the fusion and resolution of syllables, which, in accordance with older
terminology, he calls episynalipha and diaeresis.
326
The fifteenth chapter of his treatise,
which deals with the prosodic licences of poets, is also fundamentally based on earlier
presentations of metaplasms, although, here too, Bedes approach to the matter is much more
practical, as he has jettisoned the earlier nomenclature with its elaborate names for the
lengthening (ectasis) and shortening (systole) of syllables altogether.
327

The focus of Bedes presentation of metaplasms is on elision. That such an
approach was necessary for the Anglo-Saxon scholars is reflected in the obvious difficulties
which the use of elision presented them with. Andy Orchard, among others, has demonstrated
the different approaches to the technique of elision taken by the first major Anglo-Latin
poets: in Aldhelm, elision is scarce and generally limited to the first two feet of the line,
328

whereas Bede seems to have gone out of his way to introduce elision wherever possible.
Although, even in Bede, elision is not nearly as common as in, say, Vergil, his eagerness

323
Holtz 1981, 660, 8-10.
324
Niedermann 1937, 22-32.
325
Lausberg 1998, 230-231; Holtz 1981, 663; Niedermann 1937, 27-28; Isid. orig. 1, 35, 5. The general
consensus appears to be that, in some way, synaloephe is the smooth transition from one vowel to another,
whereas ecthlipis is something harsher. Isidore and Donatus seem to understand ecthlipsis to mean the elision
of final m (synaloephe would be the elision of final vowel), and Aldhelm follows their example. Consentius
suggests that the difference is that in synaloephe, a vowel is skipped rather than omitted (transilitur, non
exclusa est). Niedermann 1937, 28, 16-17.
326
DAM 14.
327
e.g. Donatus in Holtz 1981, 661-662.
328
Wright 1985, 185-186; Orchard 1994, 79-80.
111

sometimes has unsatisfactory results: from time to time, Bedes use of elision appears either
gratuitous or awkward, especially in his elision of monosyllables,
329
in itself not atypical in
Anglo-Latin verse.
330
All of this would indicate that Bede viewed elision as one of the
hallmarks of sophisticated poetic style, something also reflected in his disdain of its opposite,
the hiatus. Indeed, together with spondaic lines, hiatus was apparently one of Bedes main
annoyances and one of those features which, in his opinion, most marred the verse of the pre-
Christian poets. Bedes severity regarding the matter may be biased, but contemporary
evidence shows that Bedes concern was justified: the occasional poetry of less accomplished
Anglo-Latin poets exhibits strikingly liberal use of hiatus, usually without any regard for
poetic style. Hiatus is particularly conspicuous in Insular Latin Hymns, a feature unique in
early medieval rhythmic verse where elision and hiatus were usually equally strenuously
avoided.
331

Bedes presentation of elision (in Bedes terminology, synalipha, in accordance
with Donatuss nomenclature) starts with a short explanation of the phenomenon, together
with a concise etymology of the term, taken from Consentius:
332


Synalipharum quoque commemoranda ratio est, quia nonnumquam ultima verbi
syllaba vel particula syllabae videtur absumi. Unde synalipha Graece dicitur quasi
quodam saltu transmittens. Fit autem duobus modis quia, cum aliqua pars orationis
aut in vocalem litteram aut in m consonantem desinit, incipiente a vocali sequente
parte orationis, illa quae sequitur pars orationis praecedentem vel litteram vocalem
vel syllabam quae in m desierat sua vocali absumit.
333

[The principle of elisions should also be discussed, because sometimes the final
syllable or part of the final syllable of a word seems to be taken away. Hence in
Greek elision is called synaloepha, a kind of leaping across, as it were. There are
two kinds of elision, since, when a word ends in either a vowel or in the consonant m
and the following word begins with a vowel, the following word absorbs with its
own vowel the preceding vowel or the syllable ending in m.]
334


Shortly thereafter Bede ventures to give a few examples of this practice, quoting from his
favourite Christian authors. Bede shares with some of his Late Latin predecessors a total
ignorance of the practice known as aphaeresis or prodelision, where the initial e of the verb

329
Jaager 1935, 20. J aager specifically mentions the inelegant elision of the first syllable of the line in verses
457 qu(i) obliti, 685 (qu(am) imperiti) and 762 (qu(i) obsequio) of Vita metrica Sancti Cuthberti.
330
Orchard 1994, 80.
331
Norberg 1958, 32-33; Orchard 1994, 38.
332
Niedermann 1937, 27, 1-8.
333
DAM 14, 2-9.
334
Trans. Kendall 1991, 115.
112

form est is elided after a final vowel or m.
335
Bede assumes normal elision to take place in
such cases, as his first example makes evident:

Quod dico huiusmodi est:
Arta via est verae quae ducit ad atria vitae (Prosp. epigr. 19,1).
Scanditur enim ita, artavi dactylus, estve spondeus, intercepta a syllaba per
synalipham.
336

[This, in my view, is how elision works:
Arta vi(a) est verae quae ducit ad atria vitae
(The path is narrow which leads to the temple of true life).
The line scans like this: artavi (dactyl), estve (spondee), with the syllable a being
absorbed by elision.]
337


The correct reading would be arta viast, rather than arta viest. The same misunderstanding
occurs several more times in Bedes chapter on elision: he cites the following examples, all
from Prosper and all more properly cases of prodelision:

Nullus enim est insons (Prosper epigr. 43, 3);
Magnum praesidium est (Prosper epigr. 72, 1);
Non cognitus ordo est (Prosper epigr. 40, 3).

Bedes suggested scansions are nullus enest insons (rather than enimst), magnum
praesidiest (rather than praesidiumst), and we may suppose that he implies the scansion
cognitus ordest (rather than ordost) for the last quotation. Of course, we know better, but,
ultimately the issue is a fairly irrelevant one, having, as it does, no actual bearing on the
metrical structure of any line, and therefore being untraceable unless spelt out.
Bede reiterates the fact, which Anglo-Saxons probably found perplexing, that
also word-final ms are elided.
338
This practice reflects the phonetics of classical Latin, where

335
On aphaeresis or prodelision, see e.g. Raven 1965, 28. Of the grammarians, especially Consentius is in an
utter muddle: he recognises the possibility of eliding the first vowel of a word (he refers to this as a form of
ecthlipsis), but appears to suggest that it is freely interchangeable with ordinary synaloephe (where the final
vowel is skipped):
Ergo sic dubium erit, utrum hoc an illo modo scandas, utrum synalipham an ecthlipsin facias,
coniugio Anchise (Verg. Aen. 3, 475). Nam si sic scandas: coniugi Anchi, synalipham fecisti,
transiluisti enim o syllabam; quae transilitur, non exclusa est. Si sic scandas: coniugi Onchise,
ecthlipsis erit; exclusa est enim a et quasi expressa. Niedermann 1937, 28, 13-18.
[Therefore, if there is any doubt whether you should scan in this or that particular way, making either
a synaloephe or an ecthlipsis, (consider) coniugio Anchise. For, if you scan thus: coniugi Anchi, you
have made a synaloephe, skipping the syllable o, which is skipped, not omitted. If you scan: coniugi
Onchise, it will be an ecthlipsis; namely, the a is omitted and, so to speak, squeezed out.]
336
DAM 14, 9-13.
337
Trans. Kendall 1991, 115.
338
That the issue was taken seriously by Anglo-Saxon scholars is proven by the singular use of elision in the
works of both Aldhelms works: he is presumably the only hexameter poet in whose verse the elision of m is
113

final ms were very weakly articulated and may simply have nasalised and lengthened the
previous vowel.
339
Bede has, strangely enough, added that initial h has the power to block
elision after an m. This is in line with his exposition of the consonantal value of the letter h in
his chapter on common syllables and reflects the metrical practice of the Christian epic poets
J uvencus and Sedulius.
340
Bedes formulation of the rule in such a form that hiatus may take
place when the previous word ends in an m and the following word begins with an h appears
to be his own attempt to codify this practice:

Quaecumque ergo verba in m terminantur, nisi adpositione consonantis alicuius
defendantur, synalipha inrumpente syllabam ultimam aut perdunt semper aut
minuunt, excepto cum ab h littera sequens sermo inchoaverit. Tunc et enim in arbitrio
poetarum est, utrum haec instar fortium consonantium synalipham arceat, an pro
modo suae fragilitatis nihil valeat. Valuit namque in hoc, quia voluit poeta:
Nomine Iohannem hunc tu vocitare memento (Iuvenc. 1, 26);
et:
Progenitum fulsisse ducem, hoc caelitus astra (Sedul. carm. pasch. 2, 77).
Item nihil iuvit ad propellendam synalipham, quia poeta neglexit:
Qui pereuntem hominem vetiti dulcedine pomi (Sedul. carm. pasch. 1, 70).
341

[Every word which ends in m will always either lose its final syllable or have it
reduced by the action of elision, unless elision is blocked, which happens when the
immediately following word begins with any consonant except h. In that case, it is up
to the judgement of the poets as to whether the h should prevent elision after the
fashion of the stronger consonants, or should have no force on account of its
weakness. It has the force of a consonant, since the poet willed it, in this example:
Nomine Iohannem / hunc tu vocitare memento
(Remember to call his name J ohn);
and in
Progenitum fulsisse ducem, / hoc caelitus astra
(The Son and king has appeared, the star from Heaven witnesses it).
But it has no power to block elision, because the poet disregarded it, in this example:
Qui pereunt(em) hominem vetiti dulcedine pomi
(You who restore mankind, damned by the sweetness of the forbidden fruit).]
342



actually more common than that of final vowels. This feature of Aldhelms verse seems inexplicable, but is
possibly due to hypercorrection. Even in Bede the elision of final m is conspicuously common: in his Vita
metrica Sancti Cuthberti: together with elision of final que, it comprises the majority of elisions. Orchard 1994,
81-82; Jaager 1935, 20.
339
e.g. Allen 1978
2
, 31.
340
Curiously enough, Bede does not appear to have made use of this licence in his own verse. In Aldhelm, on
the other hand, it is copious enough, although Aldhelm in his De metris does not recognise this possibility.
Orchard 1994 (p. 84) states: One might note that of some twenty-four instances in Aldhelm where a word
beginning in h- is in a position which may potentially cause elision, only seven in fact do so (=29.17%). It
seems probable that cases of elision blocked by initial h in Aldhelm are due to the influence of either Venantius
Fortunatus or Sedulius.
341
DAM 13, 31-43.
342
Trans. Kendall 1991, 117.
114

Bede here also touches on a prosodic feature which he had discussed in his chapter on
common syllables, where he renounced the use of hiatus as an old-fashioned practice typical
of the pagans, together with the correption of a long vowel in hiatus. However, as what
appeared to be his own extrapolation of the subject, he alleged that such correption could take
place within a word, using the examples of Maria/Maria and ous/ous.
343
Here Bede again
takes up the subject, this time emphasising that elision never takes place within a word, but
that a word-internal vowel may be shortened before another vowel:

Sciendum autem quod numquam in eadem parte orationis media fieri potest
synalipha. Verum si in medio verbo duae vocales conveniunt, quarum prior sit longa,
potest illa quae sequitur priorem facere brevem de longa, si sic poeta voluerit.
Auferendi autem funditus potestatem non habet. Est enim longa naturaliter in illo
Paulini:
Ut citharis modulans unius uerbere plectri (Paulin. carm. 27, 72).
Est brevis licenter in illo Sedulii:
Unius ob meritum cuncti periere minores (Sedul. hymn. 1,5).
Item natura longa est in hoc Paulini:
Discutiebat ovans galea scutoque fidei (Paulin. carm. 16, 125);
licentia brevis in hoc Prosperi:
Delicias iam nunc promissi concipe regni
virtute et fidei quod cupis esse tene (Prosp. epigr. 102, 17-18).
Et hoc, ut supra rettulimus, inter communes syllabas computatur.
344

[It should be understood that elision can never occur in the middle of a word. It is
true that if two vowels the first of which is long, come together in the middle of a
word, the second vowel can shorten the first, if the poet so wishes; but it does not
have the power of eliding it completely. For instance, the i in unius is long by nature
in this line of Paulinus:
Ut citharis modulans unus uerbere plectri
(As one singing to the accompaniment of the lyre, with the strike of the
plectrum alone).
But it is short in this line of Sedulius:
Unus ob meritum cuncti periere minores
(On account of the fault of one man all his descendants perished).
Similarly, the e in fidei is long by nature in this line of Paulinus:
Discutiebat ovans galea scutoque fidi
(Triumphing with the helmet and the shield of faith, he scattered their
weapons);
but it is short by poetic licence in this couplet of Prosper:
Delicias iam nunc promissi concipe regni
virtute et fidi quod cupis esse tene
(Take now the wealth of the promised kingdom,
maintain that you desire to be by virtue and faith).

343
DAM 3, 81-87.
344
DAM 13, 44-59.
115

A vowel which may be shortened this way is reckoned among the common syllables,
as I have mentioned above.
345


Bedes idea that a long vowel may be shortened inside a word if another vowel follows it has
hit the nail on the head, albeit accidentally: this is precisely what happened at the junctions of
vowels in archaic Latin.
346
As a consequence, classical Latin has precious few words where a
long vowel is followed by another vowel: the main exceptions to the rule are most forms of
fio, the pronominal genitive ending -ius, Greek loans such as er and a few others.
Therefore, Bedes effort to present this as a poetic technique seems misguided,
as classical Latin has few words where a long vowel remains long before another vowel, and
the applications of the rule, as presented by Bede, are extremely limited. The examples which
Bede here discusses are the pronominal genitive illius and the genitive/dative form fidei. In
classical Latin, the normal form of illius has a long i in the hiatus, but the short form illius is
also common in archaic verse, and coexisted with illius even in Vergil (e.g. Aen. 1, 16: hic
illus arma), and, probably through his influence, in later verse. Fidei (with long e) on the
other hand, is an archaic pronunciation the classical form is fidi which, probably through
analogy to diei, had gained a new popularity in Late Latin.
347
As we noted in Bedes chapter
on final syllables, he considers the archaic fidei to be the normal form,
348
although he himself
uses fidi in his verse. The reasons for this are clarified by Bedes theory of word-internal
correption: illius is the normal form and illius the result of correption; similarly, fidei, which
Bede erroneously considers the normal form, can be shortened to fidi.
The segment on word-internal correption, and Bedes emphasis on elision never
taking place word-internally, is probably a reaction agains Maximus Victorinus, who in his
De ratione metrorum suggested the opposite. Victorinus proposes an utterly nonsensical
scansion, with a verbose and scarcely intelligible theoretical explanation, for Verg. Aen. 5,
186 Nec tota tamen ille prior praeeunte carina (Yet not first by a whole ships length).
Victorinuss suggested scansion is prae-ente (with omission of the u),
349
which Bede attacks

345
Trans. Kendall 1991, 118-119.
346
See e.g. Baldi 1999, 265-266.
347
Norberg 1958, 12.
348
DAM 6, 47-48.
349
Scanditur enim sic, nec to spondeus, ta tamen dactylus, ille pri dactylus, or prae spondeus, hic vides posse
alios dicere dactylum esse ratione coniunctarum syllabarum, quoniam prae syllaba vocali finitur et excipitur a
vocali, quae est e. Sed cum synalipha fieri possit faciatque spondeum, cur non eam viam sequamur, quam nobis
consuetudo frequentior tritam reliquit? Antenovissimus pes est huius versus ente ca, quem ideo praepono. Cum
in synalipha haec sit consuetudo, ut anterior vocalis excludatur, tunc tamen non excludi scias, cum sola syllabam
facere potest anterior, quae excludi non potest. Quod dico, huiusce modi est: praeeunte per duas e scribitur; est
ergo spondeus or prae, sequunturque circa se duae uocales, eunte carina. Sed quoniam e littera eadem et syllaba
116

vehemently, proposing instead his theory of word-internal correption. In Bedes opinion the
ae in praeeunte is shortened so that both prae and e are short syllables:
350


Ita scandendum esse ratio probat, ut primo sit necto spondeus, deinde tatamen
dactylus, illepri dactylus, orpraee dactylus, adbreuiata diptongo propter vocalem
quae sequitur, unteca rina dactylus et spondeus, qui terminent. Quis enim audiat
Victorinum docentem ut scandamus enteca rina, facientes synalipham in media parte
orationis, quod numquam fecere priores?
351

[Hence metrical theory proves that Vergils line...should be scanned so that first there
is necto, a spondee, then tatamen, a dactyl, illepri, a dactyl, and orpraee, a dactyl,
with the diphthong ae shortened on account of the following vowel. The line ends
with unte carina, a dactyl and a spondee. For who will agree with Victorinus, who
teaches that we should scan enteca rina, eliding in the middle of the word, which
earlier writers never did?]
352


If, in his use of priores for earlier writers Bede is referring to the pre-Christian authors, this
may be the highest compliment he pays them in his treatise.
Bedes presentation of elision ends with the statement that elision can be used
in every foot of the line, even the last one, something which Aldhelm had conspicuously
neglected,
353
and this may be yet another effort on Bedes part to encourage a richer and more
varied hexameter style. Bede also presents an example of elision in the final syllable of a line,
a fairly unusual technique which does not appear in his own verse. Although this is a
Vergilian feature, Bede, true to his habits, does not present it as such but uses the example of
Paulinus:

Fit autem synalipha in omni parte versus, etiam in extrema, ut Prosper:
Sed rerum auctori nullus non cognitus ordo est (Prosper epigr. 40, 3).
Fit et post versum synalipha, quae ad sequentis uersus caput intendat, ut Paulinus:
Quae decus omne operum perimebant inproba foedaque

est, u autem, syllaba ut sit, adhibet sibi consonantes n et t, idcirco rectius ente carina quam unte carina per
synalipham scandimus. gramm. VI, 218-219.
[It is namely scanned thus: nec to is a spondee, ta tamen, a dactyl, ille pri, a dactyl, and or prae a spondee. Here
you see that some may say that it is a dactyl by way of joined syllables, because the syllable prae ends in a
vowel and is followed by a vowel, which is e. But if elision may take place and make a spondee, why not follow
the road which has been made familiar to us by custom? The penultimate foot of this verse is ente ca, which I
prefer for this reason: although in elision it is customary that the previous vowel is omitted, you should know
that it may not be elided when it comprises the entire previous syllable, which may not be excluded. Therefore I
say it is as follows: praeeunte is written with two es, or prae is therefore a spondee, and followed by two
vowels. But as the letter e is, at the same time, also a syllable but u, in order to make a syllable, also has the
consonants n and t, we are more correct in scanning ente carina (by way of elision) than unte carina.]
Victorinuss theory is probably based on the fact that one-letter words like the interjection o are not elided.
350
More probably, this is a case of vowel fusion that Bede fails to recognise.
351
DAM 66-72.
352
Trans. Kendall 1991, 121.
353
Orchard 1994, 82.
117

obice prospectum caecantia (Paulin. carm. 28, 65-66).
Sunt namque ultimi versus illius pedes, inproba dactylus, foeda spondeus, at primi
sequentis per synalipham, quobice dactylus, prospec spondeus.
354

[Elision may occur in every foot of the line, even the last, as in Prosper:
Sed rerum auctori nullus non cognitus ordo est.
(There is no sequence of events which is not already known to the Creator.)
There even occurs elision which links the end of one verse to the beginning of the
next, as in Paulinus:
Quae decus omne operum perimebant inproba foedaqu(e )
obice prospectum caecantia.
(Ungainly and unsightly things which ruined every grace of the building, and
ruined the view with a barrier.)
For the last feet of the first line are inproba (dactyl), foeda (spondee), and the first
feet of the second line as as result of elision are quobice (dactyl), prospec
(spondee).]
355


Bedes presentation of elision must be considered a vast improvement on both the
grammarians of late antiquity and his immediate predecessor Aldhelm: Bede treats elision as
an integral part of the scansion and composition of verse, rather than as a side issue.
Furthermore, he has practically discarded the traditional terminology of metaplasms, calling
all forms of elision by the name of synalipha and disregarding the cumbersome and useless
distinction between synalipha and ecthlipsis, as employed by Isidore and Aldhelm. Bede also
presents the rules of elision in a logical and straightforward manner, erring only in his
ignorance of prodelision, which makes several of his suggested scansions historically
incorrect. He also refutes the idea of elision taking place in the middle of the word,
suggesting instead a technique where correption of long vowels in word-internal hiatus takes
place; Bedes suggestion may be dubious and of limited use, but arguably it testifies to the
same willingness to standardise, regularise and simplify which we can encounter everywhere
in his treatise. Bedes presentation of elisions taking place in all the feet of the verse, on the
other hand, is probably intended to promote a more varied kind of hexameter verse. We know
that elision was difficult for the early Anglo-Latin poets, and especially in Aldhelm its use
was both limited and inconsistent. In view of this, we can read Bedes presentation of elision
as a supplement to the stylistic guidelines offered in the earlier part of his treatise.




354
DAM 13, 73-82.
355
Trans. Kendall 1991, 121.
118

2.8. Bede on prosodic licences

Bede gives us a further glimpse on his respective views on Christian and pre-Christian
prosody in two successive chapters of his treatise. These chapters, numbered fifteen and
sixteen, are placed towards the end of his presentation of the dactylic hexameter and have the
respective titles Quod et auctoritas saepe et necessitas metricorum decreta violet and Ut
prisci poetae quaedam aliter quam moderni posuerunt. Chapter fifteen (in Kendalls
translation, Concerning the fact that the rules of the prosodists are often broken both by
authority and from necessity
356
) discusses the various deviations from strict prosody
necessitated either by the structure of the dactylic metres or other stylistic considerations.
Although some of Bedes examples of such liberties are taken from Vergil, the chapter deals
primarily with the metrical liberties of Christian poets and can also be viewed as an attempt
to justify some of their transgressions. The sixteenth chapter (in Kendall, Prosodic
differences between ancient and modern poets
357
), on the other hand, deals with the metrical
practices of pre-Christian poets, most notably Vergil, and condemns a number of their
prosodic licences, most notably hiatuses and spondaic lines, which, as we have seen, were
Bedes bte noire. The ultimate purpose of the chapter seems to be to demonstrate the extent
to which pre-Christian poets (whom Bede here, tellingly, calls prisci poetae) are outdated,
contrasted with the prosodic regularity of the modern, or Christian, poets.
In chapter fifteen, Bede primarily discusses the alterations of syllable lengths,
which, in hexameter poetry, are sometimes necessary to make the words prosodically usable.
Such devices are age-old and were well established already in the classical age; usually they
involve the lengthening of one or more syllables to make a prosodically cumbersome word fit
the metre. This practice, known as epic lengthening, occurs already in Greek epic poetry,
and in Latin it was used primarily to alter the prosodic structure of words with a surfeit of
short syllables. To mention but one example, the word religio, with three consecutive short
syllables would be inapplicable in dactylic verse, but its poetic forms religio and relligio
(with a long first syllable) are not. Similarly difficult are words with cretic structures, or short
syllables sandwiched between long ones, as in imperator (-u-x), for which already Ennius
substituted the rather contrived induperator (-uu-x). The post-classical evolution of Latin
prosody provided poets with other ways of circumventing metrically difficult word-forms:
the shortened final vowels of first-person verb forms and third-declension nouns proved

356
Kendall 1991, 133.
357
Kendall 1991, 141.
119

highly useful in providing poets with additional short syllables, and, as we have already
stated, the short final es of numeral adverbs such as quinquies made previously useless cretic
words employable in dactylic verse. For Bede, however, these were the regular forms of the
words in question (we have his own word on the subject in his chapters on the final syllables
of words), and therefore required no further explanation.
The lengthening and shortening of syllables belong to the features which in the
earlier grammarians are, impractically enough, lumped together with other metaplasms and
discussed together with schemes and tropes, or figures of speech. As in the case of elision,
Bede has chosen a more practical course: he understands that this prosodic feature has a
direct effect on the scansion of verse and discusses it accordingly. Remarkably, he has also
abandoned the traditional terms for lengthened, shortened, eliminated or interpolated
syllables
358
as being simply confusing for the point he is trying to make.
The examples which Bede gives of such prosodic tinkering generally fall into
the religio/religio category, and cannot be said to represent a radical departure from classical
norm, although in some cases the analyses which Bede gives his examples are, yet again,
idiosyncratic. Also noteworthy is Bedes the inclusion of the Christian trinitas as a
prosodically difficult cretic word:

At tamen intuendum est nobis quia et auctoritas nonnumquam et necessitas metricae
disciplinae regulas licite contemnit. Necessitas quidem in his verbis quae non aliter in
versu poni possunt, ut sunt ea quae quattuor syllabas breves habent, ita, basilica,

358
e.g. Donatus in his Ars maior:
Prosthesis est appositio quaedam ad principium dictionis litterae aut syllabae, ut gnato pro nato et
tetulit pro tulit. Epenthesis est appositio ad mediam dictionem litterae aut syllabae, ut relliquias pro
reliquias, induperator pro imperator. Hanc alii epenthesin, alii parenthesin dicunt. Paragoge est
appositio ad finem dictionis litterae aut syllabae, ut magis pro mage et potestur pro potest. Hanc alii
prosparalempsin appellant. Aphaeresis est ablatio de principio dictionis contraria prosthesi, ut mitte
pro omitte et temno pro contemno. Syncope est ablatio de media dictione contraria epenthesi, ut
audacter pro audaciter, commorat pro commoverat. Apocope est ablatio de fine dictionis paragoge
contraria, ut Achilli pro Achillis et pote pro potest. Ectasis est extensio syllabae contra naturam verbi,
ut Italiam fato profugus, cum Italia correpte dici debeat. Systole est correptio contraria ectasi, ut
aquosus Orion cum Orion producte dici debeat. Holtz 1981, 660, 7 662, 2.
[Prosthesis is the placement of a letter or a syllable in the beginning of a word, as in gnato for nato
and tetulit for tulit. Epenthesis is the placement of a letter or a syllable in the middle of a word, as in
relliquias for reliquias or induperator for imperator. This is called epenthesis by some and
parenthesis by others. Paragoge is the placement of a letter or a syllable in the end of a word, as in
magis for mage and potestur for potest. Some call this prospalalempsis. Aphaeresis is the removal (of
a letter or a syllable) from the beginning of a word, contrary to prosthesis, as in mitte for omitte and
temno for contemno. Syncope is the removal (of a letter or a syllable) from the middle of the word,
contrary to epenthesis, as in audacter for audaciter or commorat for commoverat. Apocope is the
removal (of a letter or a syllable) from the end of the word, contrary to paragoge, as in Achilli for
Achillis and pote for potest. Ectasis is the extension of a syllable against the nature of the word, as in
'taliam fato profugus, since Italia should be pronounced short (i.e. with a short initial i). Systole is
the correption of a syllable, contrary to ectasis, as in aquosus rion (Verg. Aen. 4, 52), since Orion
should be pronounced long (i.e. with a long initial o).]
120

Italia, religio, vel tres primas syllabas breves, ut reliquiae, vel unam in medio
brevem, ut veritas, trinitas, quae neque dactylum consuetum neque spondeum
facere possunt, quod propriis nominibus maxime solet evenire.
359

[Nevertheless we should keep in mind that the rules of prosody are sometimes
properly disregarded both by those poets whom we consider authoritative and from
necessity. They must be disregarded from necessity in those words which otherwise
have four short syllables, like Italia, basilica, and religio, or those which have three
initial short syllables, like reliquiae, or those which have one short syllable between
two long ones, like veritas and trinitas. Words like these can make neither an
ordinary dactyl nor a spondee. This problem is especially common with certain
specific nouns.]
360


Bede, for once, cites an example from Vergil (Aen. 5, 629) with the name Italia, where
lengthening takes place in the initial i. Strangely enough, he does not take it for granted that
this is what the poet intended but suggests an alternative scansion where the first foot of the
line is a tribrach (uuu) rather than a dactyl:
Huius exemplum:
Italiam sequimur fugientem et mergimur undis.
I contra naturam pro longa posuit, quia non aliter Italiam, quam saepius erat
nominaturus, appellare valebat, nisi aut syllabam quae natura brevis erat produceret
aut tribrachum loco dactyli poneret.
361

[For example:
Italiam sequimur fugientem et mergimur undis
(We seek Italy which keeps receding and we are overwhelmed in the seas).
Here the poet has put thei, contrary to its nature, in the position of a long syllable,
because he could not otherwise have referred to Italy, which he was going to have to
name rather frequently, unless he were either to lengthen a syllable that was short by
nature or put a tribrach in place of a dactyl.]
362


Bedes quotation from Vergil is not the usual reading (volvimur rather than mergimur), but
this has no bearing on its scansion. Bedes suggestion that a short syllable could, at least
theoretically, be used in the place of a long one is, however, a misconception possibly
influenced by earlier grammar, such as Aldhelms discussion of the pathe or passiones
(prosodic liberties) of hexameter verse.
363
That Bede thought such tampering with the basic
fabric of the hexameter possible is demonstrated by other examples where poets have altered

359
DAM 15, 2-11.
360
Trans. Kendall 1991, 133.
361
DAM 15, 11-16.
362
Trans. Kendall 1991, 133.
363
Aldhelm, citing Aen. 1, 2 (Italiam fato profugus) suggests that Vergil permitted the barbarism of the
substitution of a tribrach for a dactyl (Etenim barbarismo tribrachum pro dactylo admisit) Ehwald 1919, 94;
trans. Wright 1985, 209. It is telling that both Aldhelm and Bede use Italia as an example, even if the actual
quotations are different.
121

syllable lengths but Bede suspects them of substituting other metrical feet for dactyls and
spondees. A telling example is another quotation from Vergil (Georg. 4, 34), where,
apparently baffled by the word alvearia, Bede suggests that the poet used an antibacchius (--
u) in the place of a spondee:

Sic cum de apibus loquens alvearia nominare uellet, necessitate posuit antibachium
in versu dactylico:
Seu lento fuerint alvearia vimine texta.
364

[So, when speaking of bees, he wanted to refer to beehives by name, he had to
substitute an antibacchius for a dactyl out of necessity:
Seu lento fuerint alvearia vimine texta
(Or let the beehives be woven of pliant twigs).]
365


Apparently Bede suggests a reading where the third foot of the line is rint alve (--u); in
reality, the line is not quite as exotic as all that: more plausibly, vea-ri-a could be scanned as
a dactyl, with synizesis of e and a.
366
That this solution had not entered Bedes mind is
probably attributable to the same mulishness that we encounter in his chapter on common
syllables, where he steadfastly refuses to recognise the eu in Mnestheus for a diphthong.
367
It
is nevertheless surprising that although Bede devotes one extensive chapter (number
fourteen) to the subject of synizesis, he has serious difficulties recognising it in his own
reading.
The other examples of prosodic licence in Bedes presentation or, rather, their
analyses are similarly bizarre. Bede cites the examples religio and basilica (with lengthened
first syllable),
368
quoting from Paulinus and Venantius Fortunatus, and gives them an

364
DAM 15, 16-19.
365
Trans. Kendall 1991, 133-135.
366
E in hiatus may have been pronounced as something resembling a semivowel. See Grandgent 1907, 94; Allen
1978
2
, 51. The authenticity of alvearia in Vergils Georgics has been debated, and most modern editions have
the synonymous (and prosodically more plausible) alvaria. See Johnston 1897, 15.
367
DAM 3, 42-51.
368
DAM 15, 22-28:
Et Paulinus:
Qui simul huc sancta pro religione coistis (Paulin. carm. 27, 637);
re contra naturam pro longa posuit, quia non aliter hoc nomen versus exameter recipere valebat. Tale
est et illud eiusdem:
Basilicis haec iuncta tribus patet area cunctis (Paulin. carm. 27, 637).
Namque alibi quia potuit pro brevi ponitur eadem syllaba, dicente Fortunato:
Hic Paulina, Agnes, Basilissa, Eugenia regnant (Ven. Fort. carm. 8, 3, 35).
[And in:
Qui simul ac sancta pro religione coistis
(You who assemble together for the sake of holy religion),
Paulinus has put re, contrary to its nature, in the position of a long syllable, because this noun could
not otherwise go into a hexameter verse. Of such nature is also this line of the same poet:
122

adequate explanation but then goes on to cite further examples of poets introducing, as it
seems to him, unusual feet into the hexameter line. The remaining examples fall, in Bedes
terms, into the category of liberties sanctioned by the auctoritates without actual prosodic
necessity, and Bedes presentation of these licences amounts to little less than a carte blanche
for Christian poets. The substance of Bedes presentation is simply this: at times, if the
content of a verse is of sufficient consequence, prosodic rules can be broken so as to
emphasise the superiority of the divine truth over human learning. All the examples which
Bede uses are from Sedulius, Bedes champion, and all of them consist in the manipulation of
syllable lengths in ways not entirely consistent with usual prosody. Awkward as these lines
may seem, the prosodic analyses Bede gives them are stranger still, as he persistently
proposes the insertion of unusual feet (trochees and antibacchii) into the metrical framework
of the hexameter line. His main motives seem to be the defence of Seduliuss prosody, even
in its irregularities, and the unwillingness to accept such metrical features as hiatus or
spondaic lines in Christian poetry, which in turn has resulted in some very elaborate
alternative interpretations for the lines in question. His first example of Christian content
overriding grammatical rules is a line from Seduliuss Hymns (1, 110), where Sedulius has
scanned the ablative form spiritu with a short first syllable:

Auctoritate autem contemnitur regula grammaticorum, ut Sedulius in clausula
carminis, cuius supra memini, cum dixisset:
Gloria magna Patri, semper tibi gloria, Nate (Sedul. hymn. 1, 109),
subdidit:
Cum sancto Spiritu gloria magna Patri (Sedul. hymn. 1, 110).
Spiritus enim primam syllabam habet longam; unde vera scansio versus istius haec
est: cumsanc spondeus, tospiri antibachius, non dactylus. Sed poeta, ut gloriam
sanctae et individuae trinitatis clara voce decantaret, neglexit regulam grammaticae
dispositionis.
369

[The rules of the grammarians, moreover, are sometimes disregarded by poets whom
we consider authoritative. So, for example, at the end of the poem which I mentioned
above, Sedulius, after saying:
Gloria magna Patri, semper tibi gloria, Nate
(Great glory be to the Father , glory be always to you, Son),
added:

Basilicis haec iuncta tribus patet area cunctis
(The courtyard adjoining the three churches is accessible to all).
For elsewhere, because of its natural quantity, ba is found in the position of a short syllable, as when
Fortunatus says:
Hic Paulina, Agnes, Basilissa, Eugenia regnant
(Here Paulina, Agnes, Basilissa and Eugenia reign).]
-Trans. Kendall 1991, 135.
369
DAM 15, 35-45.
123

Cum sancto Spiritu gloria magna Patri
(Great glory be to the Father with the holy Spirit).
The first syllable of spiritus is long, and therefore the correct scansion of this last line
is: cumsanc (spondee), tospiri (antibacchius, instead of dactyl). But the poet, in order
to celebrate clearly the glory of the holy and undivided Trinity, has neglected the rule
forbidding the placement of an antibacchius in elegiac verse.]
370


The genitive and ablative forms of spiritus (spirits and spirit) are prosodically impossible
in dactylic verse, being, as they are, cretic in form. What is apparent here is that Sedulius has
solved the problem by shortening the first syllable of spiritu. This, however, is not how Bede
sees it; rather, he supposes that the poet has used an antibacchius (--u) in the place of a dactyl.
The importance of the passage, of course, is obvious, as cum sancto Spiritu is a direct
quotation from the Latin Gloria, and, as Dag Norberg has demonstrated, direct quotations
from the Bible and liturgical texts often have prosodically unorthodox syllable
combinations.
371
Bede himself took some licences regarding the prosody of spiritus: in his
Vita metrica Sancti Cuthberti, the genitive singular of spiritus appears twice, both times with
an unclassical short u, which, as Neil Wright has suggested, may have been inspired by
Seduliuss usage, although, of course, the unshortened form would inevitably be useless in
dactylic poetry.
372

The next quotation from Sedulius receives a similarly idiosyncratic treatment
from Bede. The line (carm. pasch. 1, 321) presumably has a hiatus, something of a rarity in
Seduliuss verse and apparently something that Bede refused to stomach, as in Bedes
nomenclature of ancient and modern verse techniques hiatus was something that he
considered typical of pre-Christian poets. Here the hiatus appears in the phrase ego in patre,
which is a direct quotation from the Gospel of J ohn (10:38),
373
but instead of allowing
Sedulius his hiatus, Bede assumes that the o of ego is elided and the line has a trochee in
place of a spondee:

Idem ipse in carmine paschali:
Sic ait ipse docens, ego in patre et pater in me.
Sic enim scanditur, sicait dactylus, ipsedo dactylus, cense trocheus, ginpa spondeus,
ablata o per synalipham, aut si candere vis censego et facere dactylum, contra morem

370
Trans. Kendall 1991, 135-137.
371
Norberg 1988, 17-18.
372
Wright 2005, 158-159.
373
One alternative scansion, suggested in Heiric of Auxerres gloss (Kendall 1975, 129), but not by Bede
himself, is to scan the e in ego as long and elide the final o (egin patre).
124

ipsius Sedulii, quem per omnia seruauit, agis ut inmunis stet vocalis altera
superveniente vocali de foris.
374

[Likewise the same poet in the Paschale carmen writes:
Sic ait ipse docens, ego in patre et pater in me.
(Thus he himself teaches when he says: I am in the Father and the Father is in
me).
The verse is scanned in this way: sicait (dactyl), ipsedo (dactyl), cense (trochee),
ginpa (spondee, with elision of the o). Or, if you prefer to scan censego, making a
dactyl, you block the elision of a vowel, which is followed by another vowel at the
beginning of a word, contrary to Sedulius own custom, which he always
observed.]
375


Bedes hostility towards hiatus is evident, and in this case it has probably warped his
judgement. This peculiar approach to Seduliuss prosodic liberties shows that Bedes starting
point in poetry was prosody, and the composition of poetry consisted in uniting appropriate
syllable combinations to create larger metrical structures (rather than filling a metrical
structure with the appropriate syllables, as Aldhelm seems to have viewed it). Undeniably,
Bedes approach to prosody seems too unyielding: he seems more willing to tamper with the
metre itself rather than with syllable lengths or the rules of elision.
Yet another quotation from Sedulius has been given a similar treatment in
Bedes exposition: in this case, Sedulius, quoting from J ohn 12:28, uses the phrase clarifica
nomen tuum (illuminate your name), which presents difficulties similar to the previous
examples: unless the final syllable is elided, the phrase would normally end in a cretic
construction. Seduliuss probable intention is to scan tuum as a monosyllable (by way of
synizesis), a possibly inelegant but nevertheless plausible course. This time, Bede is not clear
about what he assumes the intended scansion to be, but by analogy to the previous examples,
we must suppose that, in his opinion, Sedulius has used a long syllable in the place of a short
one. Bedes defence of this licence is, in all but wording, identical with the previous ones:

Idem in eodem opere:
Clarifica, dixit, nomen tuum. Magnaque caelo (Sedul. carm. pasch. 5, 8).
In quo, ut veritatem Dominici sermonis apertius commendaret, postposuit ordinem
disclipinae saecularis.
376

[In the same work, Sedulius says:
Clarifica, dixit, nomen tuum. Magnaque caelo
(Glorify your name, he said. And a great voice resounding from Heaven...).
In this verse, in order to commend more clearly the truth of the Lords Word, he set
aside the order of worldly learning.]
377


374
DAM 15, 45-51.
375
Trans. Kendall 1991, 137.
376
DAM 15, 51-55.
125


Bedes final example of prosodic licence in Seduliuss verse is an awkward attempt to recast
a spondaic verse from his Paschale carmen 5, 196 (Scribitur et titulus: Hic est rex
Iudaeorum) as an ordinary hexameter line. We already discussed this in conjunction with
Bedes incredulity at, and condemnation of, spondaic verses (p. 68). Here Bedes approach
seems the very opposite of the previous examples: rather than admitting that Sedulius could
have used a spondee in the fifth foot of a hexameter line, he suggests that the final word
Iudaeorum be scanned as a dactyl and a spondee (possibly I-da-e--rum, although Kendall
suggests -u-de--rum
378
). It is worthy of note that here, too, the reason for Seduliuss licence
is that the phrase hic est rex Iudaeorum is quoted verbatim from the Bible (Luke 23:28).
The middle of Bedes chapter on prosodic licences further contains a curious
allusion to J eromes preface to the book of Job, where he alludes to the metrical practices of
the ancient J ews to lend credibility to his exposition of prosodic licences. Jerome was one of
the foremost Christian apologists who sought to portray Scripture as the well-spring of all
human learning, and Bede certainly took him at his word when he declared that the book of
Job had been written in hexameters. Jeromes preface, in Bedes interpretation, also seems to
lend legitimacy to the prosodic liberties of the type he describes, although J erome and Bede
are equally oblique on the subject:

His et aliis huiusmodi necessitatibus credo factum, quod de libro beati Iob loquens
Hieronimus, cum dixisset eum maxima ex parte versibus apud Hebreos esse
descriptum, addidit: qui dactylo spondeoque currentes propter idioma linguae crebro
recipiunt et alios pedes eorundem quidem temporum, sed non earundem
syllabarum.
379

[I believe that this is what was done in these cases and in other cases of this kind
involving necessity, because Jerome, speaking of the book of the blessed Job, after
saying that the Hebrews had written it for the most part in hexameter verses which
were composed of dactylic and spondaic feet, on account of the idiom of the
language frequently incorporated also other feet of the same number of morae but of
a different number of syllables.]
380


Kendall suggests, with what appears to be candid and well-justified bafflement, that Bede is
trying to justify the substituting of feet of the same number of syllables but of a different
number of morae by analogy to the very opposite practice (the substitution of feet of the

377
Trans. Kendall 1991, 137.
378
DAM 15, 55-60.
379
DAM 15, 29-34.
380
Trans. Kendall 1991, 135.
126

same number of morae but a different number of syllables) in Hebrew verse.
381
In other
words, Bedes quotation from J erome sheds no light on why spondees or antibacchii should
be employed in a metre that consists of dactyls and spondees. It would seem that, although
J erome is describing a quantitative phenomenon, Bede has interpreted it as something
completely different, although it is hard to say whether Bede had any actual theory of what
J erome meant by his statement.
382
As an argument in defence of prosodic licences in
Christian poetry, Bedes use of Jerome seems a desperate attempt.
Bedes chapter on the prosodic licences of poets shows how meticulous his
study of Seduliuss verse had been and how relentless his efforts are to present him as the
epitome of good poetic style. The unusual prosodic features of the lines which Bede quotes
result from Sedulius citing verbatim either from Scripture, or, on one instance, the Latin
mass. Bede was certainly aware of this, and certainly his emphasis on the superiority of
divine over secular learning is based on this observation. At the same time, Bede has clearly
misunderstood the nature of Seduliuss poetic liberties, and the alternative scansions which
he suggests for his lines are unduly fanciful. Bedes stubbornness regarding such features as
hiatus and spondaic verses, which he saw as characteristic of pre-Christian poetry, has led
him astray, as has his inability fully to grasp the fusion of vowels. Bedes message is
nevertheless plain: Christian poets, or, at any rate Sedulius, can do no wrong, as any or all of
their metrical liberties can be attributed either to prosodic necessity or the superiority of
divine truth. The latter excuse is obviously something which the pagan authors did not have
recourse to, and, as we shall shortly observe, Bede was fairly ruthless in his treatment of their
prosodic licences.


2.9. The differences between pre-Christian and Christian poets

Bedes final notes on dactylic verses in his sixteenth chapter deal with stylistic and prosodic
issues which he has already touched on in the earlier part of his treatise. The chapter may,
superficially, appear to be a mere curiosity: here Bede, with exaggerated emphasis, discusses
the prosodic irregularities of pre-Christian poets, mainly Vergil, implicitly arguing that the

381
Kendall 1991, 135.
382
Bede may have, in his muddled way, seen Jeromes portrayal of Hebrew prosody as something akin to the
rhythmic poetry of the early Middle Ages; it is telling that Bede, as one of the first authors, uses the word
rhythmus (rhythm), also employed by Jerome in his preface to Job, as the designation for poetry without
quantitative structure.
127

hexameter had undergone vast improvements over the subsequent centuries, mainly owing to
the diligence of the Christian poets. Nowhere else in his treatise is Bede quite as explicit
about his belief that there are two distinct styles of hexameter verse, pagan and Christian.
Bedes dichotomy is reflected in his choice of terminology: he refers to pre-Christian poets
with the terms veteres, prisci and antiqui, so as to emphasise their obsolescence, whereas the
Christian poets, regardless of their true age, are called either moderni poetae, modern poets,
posteriores poetae, later poets, or nostrates poetae, our poets. The term veteres already
appears in earlier Latin grammars, where it usually alludes to archaic or pre-classical authors;
Bede, on the other hand, uses it specifically for the authors whom we regard as classical.
Bedes terminology shows us that, as he saw it, history could be divided into two distinct
ages, and that their differences were manifest everywhere, even in such minute matters as
syllable lengths.
Bedes examples of the prosodic unruliness of earlier poets are mainly based on
a handful of well-circulated quotations from the works of Vergil. As examples of Vergils
verse technique they are hardly illuminating, as the very reason they had become so well-
studied was their unusual prosody: such exceptional features as hiatuses, productio ob
caesuram and spondaic verses were invariably demonstrated with quotations from Vergil,
and consequently Vergilian lines which exhibited these features came to be overrepresented
in the grammarians. That they were unusual even in Vergils verse is something which Bede
maybe deliberately ignored.
Bedes case for the superiority of Christian prosody mainly revolves on Vergils
use of two licences which Bede particularly abhorred: hiatuses and spondaic verses. We have
witnessed the variety of excuses and alternative scansions Bede made for Christian authors
who had used these devices, most notably his hypothesis of prosthetic vowels in spondaic
verses, which would make the fifth foot dactylic (e.g. intercepto >intericepto, dena argenti >
dena arigenti).
383
However, spondaic verses appear to have been so repellent that Bede could
not attribute such a monstrosity even to Vergil. This is demonstrated by the beginning of his
sixteenth chapter, where he cites a Vergilian spondaic verse (Aen. 7, 634) but even here
tentatively suggests that the insertion of a parasite vowel might solve the problem
(presumably ducunt argento >ducunt arigento), something we already discussed in
conjunction with Bedes exposition of the hexameter (p. 69). We may also recall how Bede
quotes a hexameter line with six dactyls from Audax (at tuba terribilem sonitum procul

383
DAM 14, 53-79
128

excitat horrida
384
) in the mistaken belief that its author was Vergil. Apparently, in Bedes
eyes, the ancients cavalier way of dealing with the ending of the hexameter line was their
most serious metrical transgression, although his presentation of it is ultimately faulty:
relying on a badly-understood passage in Audax, he believes that pagan authors used dactyls
in the sixth foot of the hexameter line, but is, at the same time, strangely incredulous of their
well-attested use of spondaic verses.
The rest of Bedes chapter deals primarily with what Bede appears to have
considered less serious faults of pagan prosody, but even here the examples are sometimes
misquoted and frequently misinterpreted. The larger part of the chapter discusses the rules
regarding common syllables, something which Bede himself discussed at length in the third
chapter of his treatise, sometimes drawing his own conclusions where his reading of Christian
verse seemed to demand it. In the sixteenth chapter of his treatise, he focuses on some minor
deviations from standard prosodic practice in the pre-Christian authors; all the examples are
arguably both trivial and easily explained, and Bedes main motive in using them has
probably been to demonstrate the irregularity of pagan prosody.
Remarkably, Bede has pounced on Audaxs presentation of qu creating a
position in Lucretius, a poet whom Bede certainly did not know at first hand and who was
definitely irrelevant for his contemporaries. Audax mentions this feature in his list of long
syllables:

Aut cum correpta uocalis excipitur a littera q, quam necesse est ut consequatur littera
u, quae cum altera uocali iuncta loco consonantis accipitur, ut apud Lucretium:
Quae calidum faciunt aquae tactum atque uaporem (Lucr. 6, 868).
385

[Or when a short vowel is followed by the letter q, which must be followed in turn by
the letter u, which in conjunction with another vowel is treated as a consonant; as in
Lucretius:
Quae calidum faciunt quae tactum atque vaporem.
(Which give a warm feel and steam to the water).]

The quotation from Lucretius in Audax is not our usual reading (the accepted reading has
laticis instead of aquae), but, arguably, the feature genuinely exists in Lucretiuss verse: there
are several instances in his De rerum natura, where the words aqua and liquidus appear to be
scanned with a long first syllable.
386
It is uncertain whether this is due to some colloquial

384
gramm. VII, 340, 5.
385
gramm. VII, 328, 18 329, 3.
386
Lucr. 6, 552; 6, 868; 6, 1072 (for aqua); 1, 349; 3, 427; 4, 1259 (for liquidus).
129

pronunciation of qu (cf. acqua in modern Italian),
387
but this phenomenon seems largely
restricted to Lucretius (with such possible exceptions as Paulin. carm. 21, 785 quo totiens
quae),
388
and Bedes unusual attention to it is probably a reaction against Audaxs
inaccurate presentation of long syllables:

Qui et aliis in metrico opere regulis multum libere utebantur, quas moderni poetae
distinctius ad certae normam definitionis observare maluerunt. Nam et vocalem
brevem, quae q et u et vocali qualibet exciperetur, voluerunt esse communem, ut
Lucretius:
Quae calidum faciunt aquae tactum atque vaporem (Lucr. 6, 868).
389

[The ancients also used to take a very casual attitude toward other rules of
versification, which modern poets have preferred to observe more carefully with the
standards of established principles. Not only did they claim that a short vowel
followed by qu and any vowel was common, as Lucretius did in:
Quae calidum faciunt aquae tactum atque vaporem
(Which give a warm feel and steam to the water)...]
390


This is followed by other supposed examples of pagan latitude regarding common syllables:
Bede cites several lines where Vergil has allowed the short final vowel to be lengthened by
an initial plosive-liquid group or the letter z, both practices on which he touches in his chapter
on common syllables (Bede only allows such lengthening word-internally):
391

Et vocalem in fine verbi brevem, quae exciperetur a consonante et liquida, inter
communes syllabas deputarunt, ut Virgilius:
Aestusque pluviasque et agentes frigora ventos (Verg. georg. 1, 352),
et:
Si tibi lanitium curae, primum aspera silva
lappaeque tribulique absint, fuge pabula laeta (Verg. georg. 3, 384-385).
Quod nunc poetae in eadem parte orationis, ut supra docuimus, magis fieri oportere
decernunt. Idem vocalem in fine verbi correptam, quae excipitur a littera z, inter
communes syllabas deputavit, ut: Eurique Zephirique tonat domus.
392

[...but they also classed among the common syllables a short vowel at the end of a
word which was followed in the next word by a consonant plus a liquid, as Vergil did
in:
Aestusqu pluviasque et agentes frigora uentos
(Summer heat and rains and winds bringing frost),
and:
Si tibi lanitium curae, primum aspera silva
lappaequ tribulique absint, fuge pabula laeta

387
Allen 1978
2
, 18.
388
Holford-Stevens 1979, 393.
389
DAM 16, 15-21.
390
Trans. Kendall 1991, 141.
391
DAM 3, 8-12; 3, 111-114.
392
DAM 16, 22-32.
130

(If your concern is wool, first see to it that prickly bushes, burrs and thorns are
absent, avoid lush pastures).
Poets have now decided that it is more fitting to do this within the same word, as I
have taught above. The same poet classified among the common syllables a short
vowel at the end of a word followed in the next word by the letter z, as in Euriqu
Zephirique tonat domus (when the house of Eurus and of Zephyr thunders).]
393


Bede ignores the fact that lines of this type are a genuine rarity even in Vergils verse, and
that they mostly follow the same formula: a pair of nouns, both with the enclitic que, the first
of which is lengthened.
394
The phrase, with lengthened que, is almost invariably placed in the
beginning of the line, and the particle is generally lengthened before a consonant group which
normally does not create a position word-initially: in Vergil, long que can appear before
plosives and liquids or, on one occasion, fl:

terrasqu tractusque (ecl. 4, 51; georg, 4, 222)
lappaequ tribolique (georg. 1, 153; georg. 3, 385)
tribulaqu traheaeque (georg. 1, 164)
aestusqu pluviasque (georg. 1, 352)
Cretesqu Dryopesque (Aen. 4, 146)
spiculaqu clipeique (Aen. 7, 186)
Noemonaqu Prytanimque (Aen. 9, 767; end of line)
ensemqu clipeumque (Aen. 12, 89)
fontisqu fluviosque (Aen. 12, 189)

The same can take place before a word-initial z or x:

Euriqu Zephyrique (georg. 1, 371)
Drumoqu Xanthoque (georg. 4, 336)

Lengthened -que can also be followed by a word-initial s group:

Brontesqu Steropesque (Aen. 8, 425)

There are also cases where que is lengthened before a single consonant:

liminaqu laurusque (Aen. 3, 91)

393
Trans. Kendall 1991, 143.
394
J ohnston 1897, 19-20. Housman (1927, p. 12) states the rules regarding the formula in the following terms:
All of them are surrounded by severe restrictions: the preceding word must fill a whole foot, the following
word must be a spondee or an anapaest, and a second que must be subjoined; and furthermore the lengthening is
confined to a few authors.
131

Chloreaqu Sybarimque (Aen. 12, 363)

As we can see, all the cases of such lengthening of -que are highly formulaic; Vergil has, in
fact, sometimes used the same phrase twice, or with minute alteration of wording. It is
generally assumed that this is a Graecistic device based on the use of the enclitic :c in Greek
epic. The phenomenon, which Bede here condemns is a highly unusual one even in Vergil;
on the other hand, instances of a position being created by initial plosive-liquid groups have
been attested even in Christian literature, most notably in the verse of Venantius Fortunatus
and Aldhelm,
395
and they should more properly be considered a post-Classical feature.
396

Bede does not discuss positions created by word-initial s groups (as in Brontesque
Steropesque), because, although highly unusual in the classics, they had become a favourite
technique in Seduliuss verse, which Bede regarded as exemplary. As all of these lines have
the lengthened que in the arsis of the foot, they could arguably be interpreted as cases of
productio ob caesuram, a course which Bede has not chosen.
The remaining examples of pre-Christian irregularity are lines where Vergil has
used a hiatus. Hiatus in Vergil is often a Graecistic device (these cases are easily recognisable
as they occur in conjunction with Greek names or in paraphrases of Greek hexameter lines),
and, consequently, its use is more common in the Ecloques than in the Aeneid. On some
occasions, a sense-pause, or the end of a clause, usually coinciding with a caesura, may
account for the phenomenon; in other words, Vergil seems, on rare occasions, to have used
hiatus as a form of punctuation, a feature that he probably inherited from the archaic Latin
poets.
397
Hiatus is a feature which Bede discusses both in his chapter on common syllables
and his chapter on elision; logically enough, as hiatus simply means absence of elision.
Although Bedes presentation of hiatus and its variations is very thorough, the substance of it
is the exhortation not to use it, as it is, in his view, archaic. The end of the sixteenth chapter is
mainly a recapitulation of what Bede has previously said on the subject, and it covers all the
different types of elision one is likely to encounter (after m, after vowels and after
diphthongs, with or without correption of the final long vowel or diphthong):


395
Orchard 1994, 76.
396
A. E. Housman 1927, 3. Housman exemplifies this by citing several examples from Manilius (whom he
himself had edited).
397
Lindsey 1922, 239-241; Shipley 1924, 142. In Plautus, hiatus often occurs to indicate the change of speaker.
Of Bedes examples, georg. 1, 4 and ecl. 3, 79 would certainly seem to fall into this category. On the other hand,
the phrase qui amant (with shortened but unelided i) in 8, 818 is a frequent enough phrase in archaic verse.
132

Qui eadem libertate synalipha utebatur, siquidem et m, ubi voluit, in fine uerbi
positam a supervenientis vocalis absumptione reservavit, ut iterum iterumque
monebo (Aen. 3, 436); et longam vocalem longam remanere permisit, ut:
Sit pecori, apibus quanta experientia parcis (georg. 1, 4);
et longam, cum voluit, breviavit, ut:
Et multum formose vale vale, inquit, Iolla (ecl. 3, 79);
et:
Credimus, an qui amant ipsi sibi somnia fingunt (ecl. 8, 108);
et diptongon reservavit, ut:
Ulla moram fecere, neque Aoniae Aganippe (ecl. 10, 12);
et eandem breviavit, ut insulae Ionio in magno (Aen. 3, 211).
398

[He (Vergil) used elision with the same freedom, since, when he wished, he not only
kept an m at the end of a word from being absorbed by a following vowel, as in
iterum iterumque monebo (I will warn again and again); but he also permitted a
long vowel to remain long, as in:
Sit pecor, apibus quanta experientia parcis
(What care for the flock, how much experience for the thrifty bees);
he shortened a long vowel, when he wished, as in:
Et multum formose vale, val, inquit Iolla
(And she kept saying, farewell, farewell, beautiful Iolla);
and in:
Credimus, an qu amant ipsi sibi somnia fingant
(Do we believe? or do lovers invent their own dreams?);
he kept a diphthong, as in:
Ulla moram fecere, neque Aoniae Aganippe
( For the ridges of Aonian Aganippe did not impede you).
and shortened it, as in insulae Ionio in magno (islands in the great Ionian
sea).]
399


The first example of Vergils use of hiatus is either a corrupt reading or a lapse of memory on
Bedes part: the received reading of Aen. 3, 246 has iterumque iterumque, without hiatus.
This lapse seems to indicate Bedes eagerness to find hiatuses in Vergils verse, possibly to
make a stronger case for its lack of discipline in prosody.
Bedes precise motives for this parade of metrical licences are unclear. That he
is hostile to all forms of hiatus (except before the letter h),
400
is abundantly obvious, judging
by the other chapters of his treatise. It is possible that he has found it necessary to show all
the possible forms in which hiatus may appear in Vergils verse so as to facilitate scansion;
on the other hand, he seems to have deliberately gone out of his way to show how disorderly
pre-Christian prosody was. As we know how common hiatus is in early Anglo-Saxon verse,
the idea that hiatus should be presented as a predominantly pagan feature seems grossly

398
DAM 16, 34-44.
399
Trans. Kendall 1991, 143, except the last four lines, which are mine; through some editorial lapse, the
translation in Kendall 1991 is incomplete.
400
DAM 13, 31-43
133

unfair (it is, of course, possible that Bede aimed to wean his readers from the use of hiatus by
casting it as pagan). Bede counters his presentation of undisciplined prosody in pagan
literature with a strangely prettied-up image of Christian poetry, most of all in the fourteenth
and fifteenth chapters of his treatise: it is remarkable how Bede elsewhere manages to come
up with various pretexts for the occasional deviations from regular prosody which he
encountered in the Christian authors, explaining at length that what appear to be hiatuses or
spondaic lines are really nothing of the sort, and finally concluding that secular learning
like prosody can be subjugated to divine truth, which would seem to cover all the
remaining prosodic licences in Christian poets.
No quarter, on the other hand, is given to Vergils pagan licences: Bede does
not discuss the stylistic or historical reasons for Vergils metrical irregularities, and, of
course, this would have been beyond the scope of his learning. Once, in his third chapter,
Bede makes a reference to Homer in conjunction with Vergils use of hiatus,
401
but this is a
connection which he certainly only knew at second hand. Bedes quotations from Vergil are,
especially in the sixteenth chapter of his treatise, partly imprecise and misanalysed, and he
gives overdue prominence to the evidence of individual lines.
In most of the chapters of Bedes treatise, citations from Vergil and other
classics have been thoroughly substituted with Christian quotations, and, for example, in
Bedes brief chapter on poetic style, the material is almost exclusively Christian.
402
Of the
remaining thirty-six Vergilian quotations which I have counted in Bedes De arte metrica,
nine are in his sixteenth chapter as specific examples of what not to do.
403
Nothing is more
telling than the words with which he concludes his sixteenth chapter, and his whole
exposition of dactylic verse: Quae cuncta posteriores poetas, ut dixi, distinctius observare
repperies.
404
(You will find, as I have said, that the later poets observe these things more
conscientiously.)
Bedes views on pre-Christian and Christian verse are understandable if we
keep in mind his implicit trust in J eromes and Cassiodoruss claims that the hexameter and
pentameter, together with other poetic metres, were already used in the Old Testament; that
they, in other words, belonged to the sphere of the very earliest and least corrupt learning that
mankind had. Many Christian authors of the late antiquity were largely motivated by a need
to create an independent corpus of Christian learning, and, furthermore, a whole Christian

401
DAM 3, 62-64.
402
DAM 11
403
Heikkinen 2007, 106.
404
DAM 16, 44-45.
134

culture that would no longer be dependent on pagan thought: this notion is explicitly
formulated in Augustines De doctrina Christiana.
405
Bede, on the other hand, lived in a
world where Christian culture was dominant, and where there was an extensive body of
Christian literature in existence. There was no need for him to try to overthrow the classical
tradition. However, as he viewed the hexameter, together with other poetic forms, as
something which the ancient Hebrews had first, it was natural for him to see the Christian
poets of the late antiquity as representatives of an ageless J udaeo-Christian tradition. For
Bede, hexameter poetry was not a pagan medium which the Christians had appropriated;
rather, it was for him a biblical vehicle of expression that had temporarily been borrowed by
the pagans. If the Christian apologists had viewed Moses as the J ewish Homer, Sedulius
was for Bede the Christian Vergil. Seduliuss poetic diction is reflected in Bedes own verse,
and Bedes efforts to codify Latin prosody are, to high degree, based on what Sedulius did
and what he did not do. Sometimes Bede overgeneralises and lapses into false analogy, and
certainly exaggerates the contrast between Christian and pagan poets by overplaying the
poetic licences in Vergil and explaining away those in the Christian poets. In reality, there is
no noticeable rift between pagan and Christian prosody, but Bedes world-view dictated
that there must be, and that it must be discovered and thoroughly explained. Bedes
presentation of pre-Christian verse and its outdated licences plays a central role in his efforts
to delineate Christian poetic technique in its ideal form.


2. 10. Conclusion

The larger part of Bedes De arte metrica is taken up by its composite presentation of
prosody and the structure of the dactylic hexameter. Such a combination may seem odd, but
Bede had rightly realised that a discussion of metrical structures that excluded syllable
lengths, or vice versa, would have been useless to his readers: the knowledge of syllable
lengths was no longer a feature of the spoken Latin of his day, nor had it been for some
centuries. The only practical use of syllable quantity was its central role in the traditional
forms of metrical poetry which the Christians had adopted from the pagan cultural heritage.
The hexameter was an obvious vehicle for the presentation of prosodic rules, as it is more
complex than the shorter lyric lengths which Bede discusses briefly in the latter part of his

405
Bonner 1973, 86-87.
135

treatise. Bede had also inherited from his predecessors the view that the hexameter was
nobler than the other poetic metres.
406
Although such views in antiquity were due to its role
as the metre of epic poetry, Bede ultimately based his views on the biblical use of the
hexameter purported by Cassiodorus and J erome. Bede was committed to portraying the
noblest poetic metre in its noblest possible form, and this task had involved the close study of
the Christian epic poets of late antiquity, above all Sedulius, whom Bede appeared to hold on
a par with the Church Fathers; remarkably, Sedulius is the only poet whom Bede cites in his
even more exclusively Christian De schematibus et tropis.
As Bede trusted his own practical observation more than the grammatical
auctoritates, and as he largely had to derive the rules of syllable prosody from pre-existing
verse, his choice of reading matter ultimately affected his views on prosody, down to the
lengths of individual syllables. This comes to the fore, above all, in his chapter on common
syllables (DAM 3). Although the chapter is ostensibly derived from the presentations of
Sergius and Maximus Victorinus, nearly all of Bedes poetic samples are taken from
Christian hexameter poetry. Consequently, he presents several features of post-classical
prosody as the norm: these include the consonantal use of h and the so-called s impurum, or a
word-initial s group that creates a position. At the same time, he condemns several prosodic
features more strongly than his predecessors did, again mainly on the strength of Seduliuss
usage: these features include hiatus and the practice known as productio ob caesuram, where
a short syllable preceding a caesura is scanned as long (admittedly his views on the latter
appear inconclusive). His rather purist take on metre extends to his presentation of the very
structure of the hexameter. As the first author on poetic metre, Bede, for all practical
purposes, denies the possibility of spondaic verses rather than merely condemning them.
Subsequently, he has to explain away the cases of spondaic lines in Christian poets by
postulating elaborate forms of syllable resolution and prosthetic vowels; Bedes suggested
alternative scansions seem to have been influenced by contemporary vocal technique, but,
applied to hexameter verse, they seem like little more than a ruse.
The final chapter of Bedes presentation of hexameter poetry is the most
illuminating, as it shows his ideological bent more explicitly than do his chapters on purely
prosodic issues. There his dichotomy of ancient, or pagan, and modern, or Christian,
poets is presented in a nutshell. Mainly, Bede presents a list of cases where Vergil (or, in one
case, Lucretius), had broken against the rules of metre and prosody which Bede had

406
DAM 10, 2-4.
136

meticulously presented in his earlier chapters. Remarkably, there are no examples of bad
prosody by a Christian poet in Bedes treatise, and no examples of metrical flaws that are not
attributed to the fact that they are from the pen of a pagan author. Bede presents one sole case
of a hiatus in a Christian author, but contends that the line has been composed imitatus
veteres. The overall picture one gets is incredibly one-sided: according to Bede, all Christian
verse is good and deviations from standard prosody can be explained away; on the other fact,
metrical faults in pagan verse are due to the very fact that they are pagan.
Apart from purely prosodic and metrical considerations, Bede presents some
guidelines for good verse. They mainly include the use of enjambment to preclude a
foursquare, end-stopped poetic style, and some instructions on word-order, which, again owe
a great deal to Sedulius. The most interesting of these is Bedes presentation of the so-called
golden line, a type of double hyperbaton (abCAB), which, in modern literature, is presented
as a Vergilian technique, although it is really far more common in the verse of Late Latin
poets, and conspicuously so in Sedulius. Paradoxically enough, Bedes description of this
feature affected later scholarship to such an extent that it became embedded in the
terminology of modern philology and, in effect, far outlived the popularity of Seduliuss
verse.
The influence of Bedes views on hexameter verse and general prosody was
threefold. Bedes presentation which successfully assimilated the presentation of syllable
structure into the metrica ars served as a model for subsequent generations of authors on
verse composition. His codification of prosody, which was based on the practices of Christian
hexameter verse, exerted a similar influence on medieval poets. But, apparently, his influence
does not end there: indirectly, Bedes views on good poetic style had an impact on the
modern study of the self-same pre-Christian classics that he had so zealously tried to
substitute with Christian authors.
137

3. The lyric metres

In comparison to Bedes thorough study of the hexameter and the elegiac couplet, their rules
and metrical structure, his presentation of other Graeco-Roman metres is surprisingly slender.
Of the whole array of non-dactylic metres covered by the previous generations of
grammarians, Bede has deemed only the following seven worthy of discussion:

1. The phalaecian hendecasyllable (in Bedes title, metrum Falleucium)
2. The sapphic stanza (metrum Sapphicum)
3. The terentianean metre (metrum dactylicum tetrametrum catalecticum)
4. The iambic senarius (metrum iambicum hexametrum)
5. The iambic dimeter (metrum iambicum tetrametrum)
6. The anacreontic metre (metrum Anacreontium)
7. The trochaic septenarius (metrum trochaicum tetrametrum)

This selection is conspicuously sparse and ostensibly random especially when it comes to
aeolic verse: of the aeolic metres known to Roman poets of the classical era, only the
phalaecian hendecasyllable and the sapphic stanza are represented in Bedes treatise, and,
with the exception of the sapphic stanza, the whole panoply of lyric metres employed by
Horace is ignored. On the other hand, a curious post-classical lyrical length termed, for lack
of a better word, the terentianean has been discussed in a chapter of its own. Bedes
presentation of the iambo-trochaic metres seems more logical even in its scantiness: it is
reduced to the iambic trimeter and the trochaic tetrameter, historically the most important of
the iambo-trochaic metres, together with the iambic dimeter, which had won the favour of the
early Christian hymnodists Sedulius and Ambrose.
This drastically reduced system of Roman metrics can be viewed as the indirect
result of the decline of lyric metres in Christian literature of late antiquity and the early
middle ages. We may consider it peculiar that the dactylic metres, and the hexameter in
particular, retained their popularity throughout the transition from pagan antiquity to
medieval Christianity despite the prosodic difficulties they presented to medieval Latin
speakers, while many of the simpler quantitative metres fell into obsolescence. This
development appears to be the result of several historical coincidences. Although early
Christian scholars embraced the dactylic hexameter enthusiastically for their epic works on
Biblical themes, no such development took place in lyric poetry. The churches of the Greek
East produced a wealth of Christian hymns in classical lyric metres, but despite the efforts of
such figures as Hilary to follow suit, the more classical of the lyric metres never fully took
138

root in western hymnody, which preferred to adopt the form of the simplest of the iambo-
trochaic metres, mainly the iambic dimeter, as employed by Caelius Sedulius and Ambrose of
Milan, and the trochaic tetrameter or septenarius, as used by Prudentius and Hilary. These
metres, together with their later non-quantitative counterparts, came to form the most
important models for subsequent Christian hymns. The choice appears to have been relatively
simple to make: these metres are, for the most part, largely isosyllabic and show a high
degree of ictus-accent coincidence, making them ideal for singing. As, at the same time, the
preferred metres of Christian epic and didactic poetry were the hexameter and the elegiac
couplet, the simple iambo-trochaic metres of early Christian hymnody, and subsequently their
rhythmic variants, pushed the majority of lyric metres into the very margin of literature. Paul
Klopsch gives Venantius Fortunatus as an example:
1
the bulk of his poetic output is
composed in elegiac couplets, the hexameter hagiography Vita Sancti Martini being the most
notable exception. The only non-dactylic poems in his oeuvre are the hymn Pange lingua
gloriosi (carm. 2, 2) in the trochaic septenarius, the Ambrosian hymn Vexilla regis prodeunt
in the iambic dimeter and the abecedary De Leontio episcopo, in the same metre. Only once,
in his Carmina 10, 7, did Venantius venture to employ a classical lyric metre (the sapphic
stanza), and then it was at the behest of his friend Gregory of Tours. Even then, Venantius
apparently had to model his poem after Terentianus Mauruss tortuous verse presentation
rather than the earlier lyric poets.
2

Bede apparently made his narrow selection of lyric metres by a process of
elimination: the seven metres that made it to his final choice have all been used by Christian
poets on Christian themes. This seems to have been a very important criterion for Bede: in
the beginning of his twenty-fourth chapter (De rithmo), he professes knowledge of Serviuss
exhaustive De centum metris, but declares that the metres he does not discuss are pagan.
3

Bedes verdict has presumably not been directed against the person of Servius, but rather his
material. It is notable that the only metre that Bede has illustrated with pre-Christian
quotations is the dactylic hexameter, obviously because the subject of Vergil and his example
could hardly be left undiscussed; after all, Vergil still belonged to the set books of the
Anglo-Saxon monastic curriculum. When it comes to the other Graeco-Roman metres, Bede
could afford to be much more radical in his endorsement of Christian poetry and excise pagan

1
Klopsch 1972, 93.
2
ibid.
3
DAM 24, 3-9:Praeterea sunt metra alia perplura, que in libris Centimetrorum simplicibus monstrata exemplis
quisque cupit repperietQuae, quia pagana erant, tangere non libuit. (There are, besides, a great many other
metres, which anyone who desires will find set forth with plain examples in the Book of a Hundred MetresI
have preferred not to deal with them because of their pagan nature.) Trans. Kendall 1991, 161.
139

writers completely from his selection of auctoritates. Another factor that may have
contributed to Bedes simplified presentation of the non-dactylic poetic metres has probably
been his lack of first-hand acquaintance with such prominent users of lyric metres as Horace,
Martianus Capella and Boethius. When it comes to Horace, however, Bedes apparent
ignorance may have been at least partly deliberate: if Bede had not decided to leave Horace
and his verse undiscussed, the grammarians at his disposal would certainly have provided
him with all the quotations necessary for a thorough presentation. As we have witnessed, in
his presentation of the hexameter Bede did not shy from quoting Lucretius, a poet whom he
most certainly did not know at first hand.
The lyric metres discussed by Bede can be divided into three categories: firstly,
Bede gives us three metres which he defines as dactylic: the phalaecian hendecasyllable
(DAM 17), the minor sapphic stanza (DAM 18) and an enigmatic metre found for the first
time in Terentianus Maurus and hence termed the terentianean (DAM 19). All of these
metres, with the possible exception of the terentianean, are in reality aeolic metres, but in the
Late Latin grammars they had generally come to be classified as dactylic. The second
category consists of three metres that Bede terms iambic: the iambic trimeter (DAM 20),
the iambic dimeter (DAM 21) and the anacreontic metre (DAM 22), a Greek lyric metre
which Bede interprets as a catalectic form of the iambic dimeter. The third category consists
of the trochaic metres, represented in De arte metrica only by the trochaic septenarius (DAM
23). Bedes selection is not only rather eclectic but also somewhat unanalytic: of the three
iambic metres which he presents, one is not a true iambic metre at all, and Bede does not
clearly differentiate between the classical and archaic, or Plautine, versions of the iambo-
trochaic metres. This is evident in his chapter on the iambic trimeter, where the definition is
that of the classical iambic trimeter, although Bede insists on calling it metrum iambicum
senarium,
4
a term more properly used for its archaic form, the senarius. This confusion, of
course, is something that Bede had inherited from his late antique predecessors, who often
exhibit similar inability to distinguish the two types of iambo-trochaic verse. Alongside the
classical iambic trimeter and dimeter, Bede presents the trochaic septenarius as the only
trochaic metre in his treatise.
5
Bedes only trochaic metre is, unlike the iambic metres, not
described in its classical form: the metre is in essence the archaic form of the trochaic
tetrameter catalectic, as used by Plautus and Terence (although Bede has managed to impose
some additional, and obviously erroneous, rules on its structure). This apparent lack of logic

4
DAM 20, 2. The title (De metro iambico exametro), however, refers to the metre as the iambic hexameter.
5
DAM 21.
140

is largely due to Bedes narrow selection of poems with which he illustrates his metres: in
most cases, the description of a metre appears to be little more than an analysis of the poem
he presents as its illustration. One could even argue that, rather than illustrating his
presentation of a metre with a poetic citation, Bede derives his very presentation from the
poem in question. Bede rarely ventures to discuss poetic metres in the abstract, and he shows
little interest in the history of metrics or the evolution of the individual metres (it is also very
typical of him that his poetic quotations are often much longer than their metrical analyses).
Faithful to this very pragmatic principle, Bede has excluded most of the iambo-trochaic
metres from his treatise. The remaining metres are those which his monastic readers were
likely to encounter in the course of their studies, and their description corresponds with the
structure of those poems which they would read.
Bede was not the sole representative of this trend in early medieval grammar.
The insular 9
th
-century Cruindmel, whose treatise on metre was strongly influenced by
Bedes De arte metrica, presents an even slimmer selection of lyric metres, only discussing
the iambic dimeter, as employed by Ambrose, and the so-called terentianean metre. For him,
there are plainly but two kinds of poetry: that composed in dactylic hexameters, and Christian
hymns, and the latter barely merit discussion in a treatise on metre:

Sed hos hymnos idcirco scandere neglegimus, quia per hexametrum dactylicum
heroicum, quo maxime metrici utuntur, non sunt compositi.
6

[But we refrain from citing these hymns, because they are not composed in the
dactylic hexameter, the metre most employed by metric poets.]

In this respect, Bede can be said to represent an intermediate position between the late
antique grammarians and Cruindmel. His exposition of the lyric metres is indubitably narrow,
indeed too much so for the later generations of poets who had better access to the works of
such polymetrists as Horace, Prudentius, Martianus Capella and Boethius and sought to
emulate them. Bearing this in mind, it is all the more significant that several of Bedes
definitions still exerted a notable influence on the Carolingian poets, even to the extent of
leading them astray.
7
This must be considered as evidence of the immense authority his
treatise enjoyed.

6
Huemer 1883, 48, 29. Klopsch has pointed out that, for Cruindmel, the only conceivable form of lyric poetry
are hymns. Klopsch 1972, 94.
7
The most significant example of this is Bedes erroneous assertion that a trochaic septenarius should always
have a trochee in its third foot (DAM 23, 4), which was slavishly followed at least by Hrabanus Maurus,
Walahfrid Strabo and Hincmar of Reims. Norberg 1958, 76-77.
141

3.1. The phalaecian hendecasyllable

The first lyric metres discussed by Bede are the phalaecian hendecasyllable and the minor
sapphic strophe, which are described in the seventeenth and eighteenth chapters of his
treatise, titled De metro Falleucio and De metro Sapphico, respectively. Both of these metres
belong to the type of verse generally called aeolic, because in Greek antiquity they first
appeared in the poetry of the Aeolian poets Sappho and Alcaeus.
8
Originally, the metres
termed aeolic did not consist of a sequence of regular metres; rather, the common feature of
all aeolic metres is that they are structured around what is termed a nucleus formed by a
metron known as a choriamb. The choriamb consists of the sequence long-short-short-long (-
uu-), and each nucleus can have one or several choriambs. The choriambic nucleus, in turn, is
generally expanded in each direction by a sequence of short and long syllables.
In some of the shorter lyric lines, the nucleus consists of a single choriamb. To
this group belong many of the most frequently used aeolic verses, including the glyconic:

Collis,| O Heli|conii (Catull. 61, 1)
-u | -uu- | u-

the phalaecian hendecasyllable:

Passer, | deliciae | meae puellae (Catull. 2, 1)
-- | -uu- | u-u--

and the sapphic verse:

Integer vi|tae, sceleris|que purus (Hor. carm. 1, 11, 1)
-u-- | -uu- | u--

In the longer aeolic verses, the nucleus can consist of two choriambs, as in the minor
asclepiad:

Maece|nas, atavis | edite re|gibus (Hor. carm. 1, 1, 1)
-- | -uu- | -uu- | u-

or even three, as in the major asclepiad:

8
Raven 1968
2
, 71; Raven 1965, 133.
142


Quae nunc | oppositis | debilitat | pumicibus | mare (Hor. carm. 1, 2, 5)
-- | -uu- | -uu- | -uu- | u-

In general, the Greek aeolic metres are much more varied and complex than their Roman
equivalents, which, from the start, tended towards greater regularisation. Above all, in the
shorter aeolic verses, the beginning of the line, which in Greek poetry tended to be very
variable, evolved in a much more regular direction. Similarly, the placement of the caesurae
in aeolic verse underwent considerable standardisation during the Augustan period.
The phalaecian hendecasyllable (u | -uu- | u-u-x) is an aeolic verse form that
was used only occasionally in early Greek poetry. Although already employed by Sappho, it
never attracted a wide following among the Greek poets, appearing only sporadically in tragic
choruses and as an ingredient in Attic scolia or drinking songs, where it formed stanzas
together with other lyric lengths. It was only in the Hellenistic age that poets like Theocritus
and Callimachus began to employ the phalaecian hendecasyllable in stichic poetry.
9

Subsequently, it became one of the favourite metres of the early Roman lyric poets. One of
the metres features that the Roman poets found attractive may have been its simplicity: the
phalaecian line is relatively short and, in its Hellenistic form, suitable for stichic poetry and
therefore employable in sustained narrative. To students of classical literature, the phalaecian
hendecasyllable is best known through the poetry of Catullus (about two-thirds of his lyric
poetry are in this metre), but it was also used by Varro, Petronius, the author(s) of the
Priapea, and more importantly by Statius and Martial, and, at a later date, Ausonius.
10
It was
also embraced fairly early by the Christian poets, including Prudentius and Cyprianus
Gallus.
11

In its original form, the beginning of the line showed considerable variation.
The initial pair of syllables could manifest itself as either a spondee (--), a trochee (-u) or an
iamb (u-). Although from the start Roman poets showed a strong preference for a line that
began with a spondee, the other types still appear in the poetry of Catullus:
12


-- | -uu- | u-u-- Passer, deliciae meae puellae (Catull. 2, 1)
-u | -uu- | u-u-- Arida modo pumice expolitum (Catull. 1, 2)
u- | -uu- | u-u-- Et acris solet incitare morsus (Catull. 2, 4)

9
See e.g. Raven 1968
2
Raven 1965, 81.
10
Nougaret 1948, 103.
11
Norberg 1958, 77.
12
Raven 1965, 139.
143


Even in Catullus, however, the first type was overwhelmingly the most popular, and in the
poetry of Statius and Martial, the two other types are already nonexistent. Starting with
Catullus, the phalaecian line also had a regular caesura either after the fifth syllable (as in
arida modo / pumice expolitum) or, more commonly, the sixth one (as in lugete, o
Veneres / cupidinesque). The writings of the grammarians generally concur with the
imperial, standardised version of the metre, although some of them, including the second-
century Terentianus Maurus, still acknowledge the possibility of a trochee or an iamb at the
beginning of a phalaecian line. Typically, Terentianuss analysis itself has been composed in
phalaecian hendecasyllables:

Verum mobilis hic locus frequenter
non solum recipit pedem, ut loquebar,
spondeum, sed et aptus est trochaeo,
nec peccat pede natus ex iambo.
13

[But this variable position
does not only take the spondee, as I said,
but the trochee is apt as well,
and one that begins with an iamb is not wrong either.]

The aeolic metres were obviously considered difficult by late Latin authors, and the longer
line-units seem to have largely fallen into disuse in Christian poetry. The only aeolic metres
present in Bedes De arte metrica belong to the relatively simple type with a single choriamb
for its nucleus, with the possible exception of the Terentianean metre, if indeed it is an aeolic
metre. Another fact that illustrates the problematic nature of the aeolic metres for the scholars
of late antiquity is the tendency of grammarians to impose a regular metron-scheme on them.
Generally, the aeolic metres came to be viewed as a variety of dactylic poetry. Already the
silver-age Caesius Bassus in his De metris discusses the variety of ways in which the
phalaecian hendecasyllable can be analysed and reaches a total of six, some of them quite
fanciful. Bassus mentions the vulgar analyses where the beginning of the line is drawn
from the hexameter, the most obvious division being one where the two and a half opening
feet of the hexameter (-- | -uu | -) are combined with the beginning of an iambic trimeter (u-u-
| u):


13
Ter. Maur. 2564-2565, 2558-2559.
144

Sed prima vulgaris quidem illa divisio, quae docet eum partem habere ex heroo,
partem ex iambo, cuius exemplum: castae Pierides meae Camenae. Ex heroo sic
dividitur: castae Pierides sonitum dedit aere canoro; ex iambico sic: meae
Camenae caelitum testor genus.
14

[But the first one is that vulgar division, which teaches that the metre has one part
from the heroic metre and another from the iambic, as in castae Pierides meae
Camenae. It is divided from the heroic line thus: castae Pierides / sonitum dedit
aere canoro; from the iambic thus: meae Camenae / caelitum testor genus.]

This method of deriving difficult metres from more familiar ones, mainly the dactylic
hexameter and iambo-trochaic sequences, appears to have been a popular mnemonic device
in the late empire. Bassuss vulgar division which views the phalaecian metre as a
combination of the opening of the dactylic hexameter and that of an iambic line was also
favoured by Terentianus Maurus
15
and is still presented by Diomedes:

Hendecasyllabum phalaecium a Phalaeco inventum tale est: vidi credite per lacus
Lucrinos. Huius pars prior de hexametro est, quam supplebimus sic, vidi credite
per liquidos Nereida fluctus, posterior autem pars de principio iambici est, quam si
suppleamus, integrum iambicum faciemus sic, lacus Lucrinos inter alta navium.
16

[The phalaecian hendecasyllable, invented by Phalaecus, is as follows: vidi credite
per lacus Lucrinos. The first part of it is from the hexameter, and we can fill it out
thus: vidi credite per / liquidos Nereida fluctus, the latter part, however, is from the
beginning of an iambic line, and if we complete it, we shall make it into a whole
iambic verse in this way: lacus Lucrinos / inter alta navium.]

The main advantage of this interpretation of the metre is that, from the point of view of the
student, it is relatively simple and easy to grasp, as the division-line between the dactylic
sequence (-- | -uu | -) and the iambic sequence (u-u-u) coincides with the caesura that usually
comes after the sixth syllable of the line.
Bassus presents another simple, and apparently popular, way of dividing the
Phalaecian hendecasyllable. This consists in viewing it as the two first feet of a hexameter (--
/ -uu) followed by a trochaic sequence called the ithyphallic (-u-u--), which is best-known
from the archilochean stanzas of Horaces poetry, where it is combined with dactylic and
iambo-trochaic elements.
17
Bassus illustrates this with a line where the caesura comes after
the fifth syllable of the line:


14
Mazzarino 1955, 136, 10-16.
15
Ter. Maur. 2575-2578.
16
gramm. I, 509, 10-17.
17
e.g. Raven 1965, 87-88; 113.
145

Altera divisio est, cuius feceram mentionem, cum de epodo Callimachi dicerem hoc
exemplo: Siccas ducite machinae carinas. Nam primi duo pedes reliquis quattuor
adsumptis faciunt heroum sic: Siccas ducite remigio subeunte carinas; reliqua pars
ithyphallicum facit: machinae carinas.
18

[The other division is the one which I had mentioned when discussing the epode of
Callimachus with this example: Siccas ducite machinae carinas. For the first two
feet, when four more are added to them make a hexameter in this way: Siccas ducite
/ remigio subeunte carinas; the remaining part makes an ithyphallic: machinae
carinas.]

This way of explaining the structure of the phalaecian metre could be further simplified by
analysing the final ithyphallic as a sequence of three trochaic feet (-u | -u | -u), and this,
indeed is the way in which many of the later grammarians describe it, including Mallius
Theodorus, whose presentation of the metre was Bedes direct model:

Metrum dactylicum Phalaecium hendecasyllabum constat ex spondio et dactylo et
tribus trochaeis. Huius exemplum: fulgens divitiis et ore clarus.
19

[The dactylic Phalaecian hendecasyllable consists of a spondee, a dactyl, and three
trochees. An example of this: shining with wealth and with a resplendent face.]

As we can see, Mallius views the phalaecian metre as an unequivocally dactylic verse form,
even going so far as to append the word dactylic to its name.
Bedes description of the phalaecian hendecasyllable mainly relies on that of Mallius, with
some alteration of terminology. As Bedes treatise of the phalaecian hendecasyllable is his
first chapter on the lyric metres, he has added a short preface which gives us a rather
illuminating picture of the decline of the lyric metres in the early middle ages. In his
introduction, Bede makes it clear that the real focal point of his treatise is the dactylic
hexameter and that in the remaining chapters he is plainly dealing with leftovers:

Verum quia de metro heroico quae videbantur tractavimus, libet et aliorum genera
metrorum, ea dumtaxat quae magis usui haberi reperimus, parumper commemorare.
Est igitur metrum dactylicum Falleucium pentametrum, quod constat ex spondeo et
dactylo et tribus trocheis.
20

[Now that I have dealt with everything that seemed appropriate about heroic verse, I
should like to mention briefly the other types of metres, those at least which I
consider are considered to be of more use. There is therefore the phalaecian dactylic
pentameter, which consists of a spondee, a dactyl and three trochees.]
21



18
Mazzarino 1955, 136, 17-22.
19
gramm. VI, 590, 21-22.
20
DAM 17, 2-6.
21
Trans. Kendall 1991, 145.
146

Bedes portrayal of the phalaecian metre is very strict and simple, and it follows the late
antique tradition of dissecting the metre into feet and classifying it as dactylic. What is novel
about Bedes approach is that he calls the phalaecian line a pentameter. The logic behind this
innovation is, in its way, quite irrefutable: if a line can be broken up into distinct metrical
feet, it makes perfect sense to call it by the number of the feet, as this, after all, is the standard
practice in the nomenclature of the dactylic and iambo-trochaic metres. It is also conspicuous
that Bede has given up the term hendecasyllabus, still employed by Mallius Theodorus.
Obviously Bede has had trouble grasping its function, the term being, as it was, a remnant
from the times when the Graeco-Roman lyric metres were not yet subjected to a rigid scheme
of metrical feet. Bede has apparently sought to bring his analyses of the lyrical metres into
line with those of the other poetic metres, and, accordingly, prefers a nomenclature based on
the number of feet to one based on the number of syllables. He has implemented a similar
alteration of terminology in his chapters on the sapphic strophe and the terentianean metre,
calling them, respectively, the sapphic dactylic pentameter and the dactylic tetrameter
catalectic,
22
although his direct models Mallius Theodorus and J ulian of Toledo had applied
the term hendecasyllabus to both of them.
23

Bede has, in his customary manner, given a very generous sample of this
modestly-presented metre by quoting no fewer than fifteen lines from Cyprianus Galluss
Exodus, which forms a part of his Heptateuchos, an epic paraphrase of the Old Testament.
Alongside verse hagiography, poetic paraphrases of biblical texts were one of the favourite
genres of late antique and early medieval poetry. Although Cyprianus Gallus today is
generally considered an obscure figure, his poetry appears to have enjoyed considerable
popularity in Anglo-Saxon England, and it has even been surmised that Cyprianuss Exodus
may have served as an inspiration for the much-celebrated Old English poem of the same
name. Bedes lengthy quotation from Cyprianus paraphrases the Song of Moses (Ex. 15:1-
18).
Bedes short chapter on the phalaecian hendecasyllable illustrates well the three
main tendencies of his work: the regularisation, simplification and Christianisation of the
Roman metrical heritage. Previously, Bedes predecessors in late antiquity had begun to treat
the phalaecian hendecasyllable as a regular sequence of metrical feet and, on the strength of
its first half, defined it as a dactylic metre. Following the example of Mallius Theodorus,
Bede took this tendency one step further by classifying the phalaecian hendecasyllable as a

22
DAM 18, 2; 19, 2.
23
gramm. VI, 591, 8-9; 12; Maestre Yeyes 1973, 229, 9-10.
147

pentameter and thus giving it an analysis that was analogous to those metres which even
originally had a regular metron-scheme. The age-old term hendecasyllabus, a relic from the
days when Greek letters still flourished in the West and the structure of the lyric metres was
perhaps better understood, was for him useless ballast that he could well do without.


3.2. The sapphic stanza

The sapphic stanza is based on a common aeolic form where the choriambic nucleus (-uu-) is
preceded by an opening sequence of four variable syllables and followed by three: uu | -uu-
| u-x. A more regular form, where the opening sequence is more fixed (-u-x) is generally
known as the sapphic line, or the sapphic hendecasyllable, after the poet Sappho. Sappho,
together with Alcaeus, employed it in a strophe of originally three lines, where two ordinary
sapphic lines are followed by a longer line that has been extended with a sequence of five
syllables (-uu-x). In later times, the extension of the third line came to be understood as a
separate line known as the adonic, and the sapphic stanza was consequently seen as
consisting of three sapphic lines and one adonic.
24
The adonic line can be analysed as a
simple choriamb followed by one syllable (-uu- | x), but it also resembles the final two feet of
the hexameter, and, analogously, it is often interpreted as a combination of a dactyl and a
spondee or a trochee (-uu | -x). In modern terminology, the sapphic stanza is sometimes
termed the minor sapphic to distinguish it from a related metre (correspondingly dubbed
the major or greater sapphic) encountered in Horaces Ode 1, 8).
25

The sapphic stanza was adopted by the Roman poets of the late republic, and its
most notable early user is Catullus, whose poems in this metre still show the same flexibility
of form as their Greek models in having either a long or a short syllable in the fourth position
of the line:

Sive in Hyrcanos Arabasve molles -u-- | -uu- | u-x
seu Sagas sagittiferosque Parthos -u-u | -uu- | u-x
sive quae septemgeminus colorant -u-- | -uu- | u-x
aequora Nilus.
26
-uu- | x


24
Nougaret 1948, 106; Raven 1968
2
, 79-80.
25
See e.g. Raven 1965, 145.
26
Catull. 11, 5-8.
148

The sapphic stanza was employed by Horace in both his Odes and in his Carmen saeculare.
True to his habits, Horace standardised the metre by making the fourth syllable of the longer
line always long and regularising its caesuras. The sapphic lines of Catullus do not always
have a caesura at all, and a similar feature appears even in Horace,
27
but in its more evolved
classical form, the line usually has a caesura after its fifth syllable (as in Hor. carm. 1, 22, 1:
integer vitae / scelerisque purus), or less commonly, after the sixth one (as in Hor. carm. 1,
10, 1: Mercuri, facunde / nepos Atlantis).
28
In later Latin poetry, starting with the Silver
Age poets, the caesura after the fifth foot became the rule. The original structure of the
sapphic stanza as a three-line strophe is still reflected in Catullus and Horace, as well as some
later-day poets who followed Horaces example: although the Romans viewed the adonic line
as a unit of its own, the third and fourth lines were often seen to be connected, which
frequently led to concatenation of the third line with the following adonic. In Catullus and
Horace, there is sometimes no word-division between the third and fourth lines, and there are
even some examples of hypermetric elision.
29

After Horace, the sapphic stanzas most prominent users include Statius (Silv.
4, 7) and Seneca, who used the stanza in a whole tragic chorus of his Medea (Med. 579-606).
Seneca also used the sapphic line in a stichic form without the adonic in his Phaedra (Phaedr.
274-324). Late antique users of the metre include Prudentius and Ausonius. After the onset of
Christianity, the sapphic stanza appears to have retained its popularity rather better than the
other lyric metres, and the metre continued to be popular throughout the middle ages.
30
The
Late Latin Christian poets who used the sapphic strophe include Paulinus of Nola and
Venantius Fortunatus, who, in a rare effort at composing lyric poetry, used the metre once in
his Carmina (10, 7). The Late Latin and medieval poets who composed verse in the sapphic
stanza generally followed the restrictions imposed on it in Silver Age: the fourth syllable is
always long, and the caesura generally comes after the fifth foot. Medieval departures from
these rules can generally be attributed to imitation of Horace.
The Roman grammarians tried, from the start, to give the sapphic stanza a
definition that was analogous to other, simpler metres. The Silver-Age Caesius Bassus,
ever inclined to derive all metres from other metres, presented an analysis of the longer line

27
Nougaret 1948, 103; e.g. Hor. carm. 1, 2, 30: quam Iocus circumvolat et Cupido.
28
Raven 1965, 144; the latter type of caesura, often called trochaic or weak, is common in the fourth book
of Horaces Odes and in his Carmen saeculare, but was shunned by later generations of poets.
29
See Nougaret 1948, 107; Raven 1965, 144. Famous examples include Catull. 11.11-12 (Gallicum Rhenum
horribile aequor ulti//mosque Britannos) and Hor. carm. 1.2.19-20 (labitur ripa Iove non probant(e) u//xorius
amnis). On similar features in medieval poetry, see Norberg 1958, 77-78.
30
Norberg 1958, 77.
149

as the combination of the beginning of the trochaic tetrameter (-u - - | -) and the iambic
trimeter (uu - u - | -) with resolution of the first element. He demonstrates this by drawing
Horaces Ode 1, 2, 1 from iambic and trochaic lines of unknown origin, possibly his own
invention (the iambo-trochaic sequence genibus haerebo tuis also appears in his presentations
of other lyric metres, including the alcaic stanza):

Cola huius carminis singula ex duobus commatibus, quorum quod antecedit ex
archilochio quadrato nascitur sic: iam satis terris magisque genibus haerebo tuis;
hoc enim par est huic: nunc ita est, magis magisque genibus haerebo tuis. Sequens
comma trimetri iambici primam habet partem; componitur sic: nivis atque dirae
genibus haerebo tuis, ut sit par huic: magis magisque genibus haerebo tuis.
31

[The individual cola of this verse consist of two commata, the first of which can be
derived from the trochaic tetrameter in this way: iam satis terris / magisque genibus
haerebo tuis; for this is similar to nunc ita est, magis magisque genibus haerebo
tuis. The following comma has the first part of an iambic trimeter, and it can be
composed thus: nivis atque dirae / genibus haerebo tuis, as it is similar to this:
magis magisque genibus haerebo tuis.]

The adonic line, on the other hand, could easily be taken from the end of the dactylic
hexameter. Bassus elucidates this, in turn, by presenting a hybrid of the opening line of
Vergils Aeneid and Horaces Ode 1, 2, 4:

Clausula strophes huius haec est terruit urbem, quae nascitur ab heroo hexametro
sic: Arma virumque cano Troiae qui terruit urbem. Nam primus ab oris, pro qua
hanc clausulam posui, par est huic non tantum numero, sed etiam pedibus: constat
enim ex dactylo et spondeo.
32

[The cadence of this strophe is the following: terruit urbem, which comes from the
hexameter in this way: Arma virumque cano, Troiae qui / terruit urbem. For
primus ab oris, for which I substituted this cadence, is similar to it not only
regarding rhythm, but the feet themselves: it consists of a dactyl and a spondee.]

Although several alternative ways of interpreting the longer line were possible, the adonic,
probably for the sake of simplicity, was invariably viewed as a hexameter fragment.
Diomedes gives the sapphic line a similarly iambo-trochaic interpretation, also using
Horaces Ode 1, 2 as his illustration:

Hendecasyllabum sapphicum Sappho poetria invenit. Exemplum huius tale est, iam
satis terris nivis atque dirae. Superior pars ex trochaico est. Nam si haec verba iam
satis terris suppleas, facies integrum trochaicum sic: Iam satis terris virente secta

31
Mazzarino 1955, 144, 7-14.
32
ibid.
150

pinus in Crago. Inferior autem, verba haec nivis atque dirae, de principio iambici
sunt. Denique additis quae desunt iambicus poterit inpleri sic: Nivis atque dirae
secta pinus in Crago.
33

[The poet Sappho invented the sapphic hendecasyllable. An example of this is iam
satis terris nivis atque dirae. The first part is from a trochaic line, for from these
words iam satis terris you can complete a whole trochaic line in this fashion: Iam
satis terris / virente secta pinus in Crago. The latter part, however, the words nivis
atque dirae, are from the beginning of an iambic line. Therefore it can be made into
a complete iambic line by adding the missing parts in this way: Nivis atque dirae /
secta pinus in Crago.]

The growing tendency to define all lyric metres as dactylic ultimately affected the sapphic
stanza as well. Rather than attempting to derive the longer line from iambic and trochaic
sequences, Mallius Theodorus in his De metris has given it a decidedly dactylic presentation,
possibly to make it harmonise with his analysis of the phalaecian hendecasyllable. The
phalaecian metre had already earlier come to be viewed as a dactylic metre, an analysis to
which it lends itself more readily than the sapphic line, as its beginning is similar to the first
feet of the dactylic hexameter. The process of the sapphic metres dactylisation involved
breaking the line up into feet of two or three syllables: -u | -- | -uu | -u | -x. This meant that the
first and fourth feet of the line would be trochees, arguably a highly atypical feature in a
dactylic metre:
Metrum dactylicum sapphicum hendecasyllabum constat ex trochaeo et spondio et
dactylo et duobus trochaeis sive trochaeo et spondio, ut est apud Horatium: iam satis
terris nivis atque dirae. Huic autem metro post tres versus additur finis heroici
versus, ut est apud eundem Horatium: terruit urbem.
34

[The dactylic sapphic hendecasyllable consists of a trochee, a spondee, a dactyl and
two trochees or a trochee and a spondee, as in Horaces iam satis terris nivis atque
dirae. After three verses, to this metre is added the end of a heroic verse, as in
Horaces terruit urbem.]

Bede based his definition of the Sapphic metre on that of Mallius, apparently wanting to
make his descriptions of the lyric metres as simple and uniform as possible. Bedes subtle
alterations of wording are similar to those in his chapter on the phalaecian hendecasyllable:

Metrum dactylicum sapphicum pentametrum constat ex trocheo, spondeo, dactylo,
duobus trocheis, cui metro post tres versus additur semis heroici versus.
35


33
gramm. I, 508, 21-29.
34
gramm. VI, 591, 8-9.
35
DAM 18, 2-4.
151

[The dactylic sapphic pentameter consists of a trochee, a spondee, a dactyl and two
trochees, to which, after three verses, is added one half of a heroic verse.]
36


Parts of Bedes definition are taken verbatim from Mallius Theodorus. Mainly, Bede has
retained Malliuss definition of the sapphic metre as dactylic and retained his foot-division.
Curiously, from Malliuss duobus trochaeis sive trochaeo et spondio (two trochees or a
trochee and a spondee) Bede has left out the words sive trochaeo et spondio (or a trochee
and a spondee), apparently finding the mention of this possibility uninformative and
redundant; as we may recall, Bede had discussed word-final syllables as one of the types of
common syllable in his third chapter.
37
Moreover, we may note that in analogy to his chapter
on the phalaecian hendecasyllable, Bede has substituted the term pentametrum for the word
hendecasyllabum, wishing to draw his readers attention to the foot-division presented in the
analysis of the metre rather than its number of syllables.
38
What may strike us as odd is the
description of the adonic as semis heroici versus, one half of a heroic verse, which seems
unusually imprecise for Bede: after all, the two feet of the adonic only make up one third of a
hexameter line. The word semis appears in all the manuscripts of De arte metrica,
39
and it
seems logical to suppose that it is originally a misreading or corruption of the word finis in
Mallius.
As in his presentations of the other lyric metres, Bede has not only followed a
policy of simplification but also one of Christianisation. Remaining true to his principles, he
has eliminated every reference to Horace and, instead, presents two quotations of no less than
four strophes altogether from Paulinus of Nolas Carmen 17 (lines 1-4 and 45-56). This is all
the more remarkable as Horaces Ode 1, 2 appears to have remained the grammarians staple
illustration of the sapphic stanza over a period of many centuries. It is obvious that Bedes
knowledge of Horaces poem probably did not extend beyond what he had found quoted by
the grammarians, and the substance of Horaces description of Jupiters vengeance on
impious Romans would certainly have struck him as highly unsuitable for an essentially
Christian guide to metre. Nevertheless, it is just possible for us to hear the same sentiment
echoed, in a subtly Christianised form, in his quotation from Paulinuss description of the
plagues of Egypt:

36
Trans. Kendall 1991, 147.
37
DAM 3, 115-118.
38
Coronati 1981-82, 57: Luno e laltro verso sono per definiti pentametrum, a differenza di Mallio Teodoro
che invece parla di hendecasyllabum, e forse anche in questo caso la scelta bedaica dovuta alla consueta
esigenza di non creare confusioni, inserendo una definizione terminologicamente incoerente con il resto di
manuale.
39
Kendall 1975, 132.
152


Sicut Egypto pereunte quondam
noctis et densae tenebris operta
qua Dei vivi sacra gens agebat,
Lux erat orbi.
40

[J ust as once when Egypt was perishing,
covered with the shades of thick night,
where the holy people of the living God were living,
He was a light to the world.]
41



3.3. The terentianean metre

One of the most enigmatic and, for most classical scholars, least familiar metres covered in
Bedes De arte metrica is discussed in the nineteenth chapter of his treatise, titled De metro
tetrametro catalectico. The metre, which Bede terms the dactylic tetrameter catalectic, is a
stichic verse-form of obscure origins that made its first documented appearance in
Terentianus Mauruss second-century didactic poem De litteris, de syllabis, de metris.
42

Consequently, the metre is in our time commonly called the terentianean metre, or versus
terentianeus, a term first employed as trentianen by Dag Norberg in his Introduction
ltude de la versification latine mdivale.
43

Even in late antiquity, the metre was never a common one, appearing mainly in the refrains of
Martianus Capellas De Nuptiis Philologiae et Mercurii and Ausonius, but most notably as a
27-line poem in Boethiuss De consolatione philosohiae (1, 4, 2).
44
Apart from being
unusual, the metre has been considered difficult to classify. Terentianus himself presents the
metre as a sub-species of dactylic poetry, where the beginning of a hexameter line (right up to
the penthemimeral caesura at 2.5 feet) is combined with its final two feet:

Si penthemimeris talis praemissa tome sit
quae primo spondeon habet, mox dactylon addit
tum post semipedem veniant duo fine revulsi
incolumi sermone pedes sine parte priorum
postquam res Asiae veluti, tunc primus ab oris.
45

[If a line is cut off at the penthemimeral caesura, having first a spondee, then a
dactyl, then after the half of a foot come two feet which have been seized, intact,

40
Paulin. 17, 45-48 (cit. DAM 18, 12-14).
41
Trans. Kendall 1991, 147.
42
Ter. Maur. 1939-1956.
43
Norberg 1958, 79-80.
44
Norberg 1998, 255.
45
Ter. Maur. 140-144.
153

from the end of the line without the preceding ones, as in postquam res Asiae,
followed by primus ab oris.]

Terentianus further demonstrates his analysis with four lines, which have mainly been
cobbled together from bits and pieces of Vergil and which essentially read like a radically
condensed version of the Aeneid:

Postquam res Asiae primus ab oris
at regina gravi saucia cura
sic fatur lacrimans, mittit habenas
et tandem Euboicis labitur oris.
46


The metrical scheme of each of the lines is ---uu- | -uu-x. In other words, the metre resembles
the first two and a half feet of a hexameter line followed by an adonic (or the two final feet of
the hexameter), the beginning and the end of the line being invariably separated by a strong
caesura. In Terentianuss presentation, the dactylic character of the verse is enhanced by his
Vergilian loans, which beyond doubt lend the metre an air of a hexameter line with the
middle cut out.
Terentianuss own analysis of the metre has often been accepted at face value,
although it is in some respects problematic. For one thing, the beginning of the line differs
from that of a typical Roman hexameter in one crucial respect: the terentianean line begins
invariably with a spondee, whereas the opening foot of a Roman hexameter is in most cases a
dactyl.
47
His analysis, however, corresponds with the tendency of the Late Latin grammarians
to view all the lyrical metres as ultimately dactylic, as we noted in the case of the phalaecian
hendecasyllable. Other ways of classifying the terentianean metre have at various times been
propounded by scholars of Late Roman poetry, and Dag Norberg has provided us with a
fairly exhaustive list of these in his 1995 article Le vers Trentieen.
48

Firstly, there are some parallels between the terentianean metre and the
phalaecian hendecasyllable, which has prompted some authors to view the former as an
anaclastic variant of the latter: the opening sequence of the syllables (-- | -uu-) is identical
in both metres, as is the number of syllables (eleven), to which some grammarians were
prone to attach great importance. Julian of Toledo was a supporter of this analysis of the

46
Ter. Maur. 1949-1952.
47
See Duckworth 1969, passim.
48
Norberg 1998, 257.
154

metre,
49
although Terentianus Maurus himself had been very explicit in distinguishing the
two metres.
50

There have been several attempts to derive the terentianean metre from other
aeolic metres, which makes sense if the line is analysed as having a double choriamb for its
nucleus: -- | -uu- | -uu- | x. In other words, the line could be read as a catalectic form of the
minor asclepiad (-- | -uu- | -uu- | ux), which is the analysis W. Meyer propounds in his
Gesammelte Abhandlungen.
51
Parallels with other forms of lyric poetry have also been
proposed over the centuries, although some of these seem barely intelligible: Huguccio of
Pisa, in his 13
th
-century treatise on pronunciation De dubio accentu called it
semisaphonicum tetrametrum, constans in prima parte ex pentametro saphonico, in postrema
adonico (a semisapphic tetrameter, consisting in its first part of the sapphic pentameter, in
its second of the adonic).
52
A. Lentini and F. Avagliano, in their 1974 edition of the poems
of Alfanus I, speak of gliconei catalectici in syllabam+adoni;
53
several medieval authors,
on the other hand, view the metre as a fusion of the minor asclepiad and the adonic, speaking
often of metrum Asclepiadeum Adonicum, or asclepiadic adonic metre, although in some
cases the analysis appears to have become confused with the strictly dactylic interpretation of
the line.
54
Norberg himself echoes this interpretation in his Introduction: il se compose
dun demi asclpiade +un adonique.
Somewhat more bizarrely, it has also been put forward that the terentianean
metre is, in fact, a form of ionic poetry. This would, in effect, make it a trimeter composed of
minor ionics (uu--), with a mandatory fusion of the two short elements in the first metron: ---
| uu-- | uu--. This theory was proposed in 1907 by K. Wagner, and supported by Paul Klopsch
in his Einfhrung in die mittellateinische Verslehre.
55
Structurally speaking, this analysis is
not impossible, but it seems unnatural for several reasons. Firstly, the regular fusion of the
two short elements in the first metron seems highly atypical; secondly, the placement of the

49
Maestre Yeyes 1973, 229, 11-13.
50
Ter. Maur. 1945-1947: Fiet hendecasyllabos, sed alter, / namque hic de genere est phalaeciorum, / cuius mox
tibi regulam loquemur. (It will become a hencecasyllable, but another one, for this one belongs to the
phalaecian type, the rules of which I will shortly expound.)
51
Meyer 1905, 225: Das quantitierende Vorbild ist nicht die phalaecische Zeile, sondern die kleine
Asklepiadeer: Saevus bella serit barbarus horrens.
52
Limone 1984, 382.
53
Lentini & Avagliano 1974, 143.
54
e.g. J ohn of Garland in his 13th-century De Parisiana poetria: Metrum Asclepiadeum Adonicumconstat
ex medietate versus exametri et fine suo. (The asclepiad adonic metreconsists of the half of a hexameter
verse and its ending.) Lawler 1974, 198. Although Johns description of the metre, with its reference to both
the asclepiad and the hexameter, seems exceedingly confusing, we must bear in mind that the aeolic metres, too,
were in the middle ages generally understood to be a form of dactylic poetry.
55
Wagner 1907, 28; Paul Klopsch appears to consider Wagners analysis more or less conclusive: Dass der
Vers als Trimeter aus Ionici a minore zu deuten ist, hat Wagner wahrscheinlich gemacht. Klopsch 1972, 98.
155

caesurae in the line does not correspond with the conventional use of the minor ionic in Latin
poetry. The minor ionic, admittedly, was never widely used by the Roman poets, the most
important example being Horaces Ode 3, 12 (1-3):

Miserarum est neque amori dare ludum neque dulci
mala vino lavere, aut exanimari metuentis
patruae verbera linguae

It is worthy of note that in Horaces ode, word-division invariably falls between the metrons:
miserarum (e)st | nequ(e) amori | dare ludum | neque dulci, etc. In the terentianean metre, if
it indeed were a form of Ionic verse, this would not be the case, because the central caesura of
the line would always intrude before the final element of the second metron: Postquam res |
Asiae / pri|mus ab oris. The overall aural impression created by Horatian minor ionics is
sufficiently different from that of the Terentianean metre to make the connection
imperceptible and Wagners hypothesis, at the very best, dubious.
The problem of the terentianean metres structural nature is complicated by the
fact that several poets have deviated from the previously presented formula of the
terentianean metre. Above all, Boethius and his medieval emulators, apparently viewing the
metre as a curtailed hexameter line, treated the beginning of the line as they would the
opening feet of a hexameter, with the corresponding variations. In principle, the first half of a
hexameter line (the part preceding a penthemimeral caesura) permits the following
combinations of feet:
a) two dactyls (-uu | -uu | -), as in arma virumque cano,
b) a dactyl and a spondee (-uu | -- | -), as in Italiam fato,
c) a spondee and a dactyl (-- | -uu | -), as in inferretque deos,
d) two spondees (-- | -- | -), as in quam Iuno fertur.
Type c is identical with the first half of the terentianean metre as presented by Terentianus,
and even in Boethius it is the most common one, appearing 17 times in the 27-line poem of
De consolatione philosophiae 1, 4, 2.
56
The other types, however, are present in Boethius as
well: there are four instances of type a,
57
five instances of type b,
58
and one sole instance of
type d.
59
This would indicate that, whatever the origins of the terentianean line, it was in
Boethiuss eyes simply another dactylic metre, permitting the same variations as the

56
e.g. line 1: heu quam praecipiti
57
e.g. line 2: mens hebet et propria
58
e.g. line 3: tendit in externas
59
line 6: hic quondam caelo
156

hexameter, albeit, in contrast to the hexameter, with a preference for a spondee in the opening
foot. Boethius was not without followers, and several Carolingian and Anglo-Saxon poets
allowed these very same variations in the first half of the terentianean line, as well as some of
their own invention.
60

The perhaps surprising popularity of the terentianean metre in the early Middle
Ages is, however, not mainly attributable to the example of Boethius, let alone Ausonius or
Martianus Cappella, none of whom were known to the earliest Anglo-Saxon authors, but
rather to its early adaptation by Christian hymnodists. The best-known examples of Christian
verse in this metre are the two hymns cited by Bede in his De arte metrica: Squalent arva soli
and Obduxere polum. Typically, Bedes analysis of the terentianean metre paraphrases Julian
of Toledo, who also has given the first line of Squalent arva soli as an example, but Bedes
quotations are much lengthier, and his treatise of the metre can be essentially be read as an
analysis of his cited examples. Bedes definition of the terentianean metre as a dactylic
tetrameter catalectic is apparently his own, and as a classification it is considerably more
practical thanJulians interpretation of the metre as a variety of the Phalaecian
hendecasyllable. Julian writes:

Dactylicum phalaecium hendecasyllabum. Quales pedes recipit? Spondeum,
dactylum et tres trochaeos. Da eius exemplum. Fulgens divitiis, et ore clarus. Fit
alio modo? Per spondeum, dactylum, catalecton et dactylum, ita ut in fine trochaeum
aut spondeum habeat. Da eius exemplum: Squalent arva soli pulvere multo.
61

[The dactylic phalaecian hendecasyllable. Which feet does it take? A spondee, a
dactyl and three trochees. Give an example of it. Resplendent with riches and
radiant of face. Can it be constructed otherwise? With a spondee, a dactyl, a
catalecton and a dactyl, so that in the end it has spondee or a trochee. Give an
example. The fields lie deep in dust.]

Bede, on the other hand, has an entirely different interpretation:

Metrum dactylicum tetrametrum catalecticum constat ex spondeo, dactylo, catalecto,
dactylo, spondeo. Quo usus est sanctus Ambrosius in precatione pluviae, cuius
exordium hoc est:
Squalent arva soli pulvere multo,
pallet siccus ager, terra fatescit,
nullus roris honos, nulla venustas,
quando nulla viret gratia florum.
Tellus dura sitit nescia roris,
fons iam nescit aquas, flumina cursus.
62


60
see Norberg 1998, 258-60.
61
Maestre Yeyes 1973, 229, 8-13.
62
DAM 19, 2-11.
157

[The dactylic tetrameter catalectic consists of a spondee, a dactyl, a catalecton, a
dactyl and a spondee. Saint Ambrose used this metre in a prayer for rain, of which
this is the beginning:
The fields lie deep in dust,
the dry land yellows, cracks appear in the earth,
no grace of blossoms shows green.
The hard earth thirsts, deprived of water,
the springs now are dried up, the rivers lack currents.]
63


Bedes interpretation of the terentianean metre as a dactylic tetrameter catalectic is much
simpler and more understandable than Julians presentation of the metre as a variant of the
phalaecian hendecasyllable. Always striving for the least cluttered ways of analysing things
and never reluctant to jettison unhelpful terminology, he makes no reference to the phalaecian
metre and none to hendecasyllables; we have noted that elsewhere, too, Bede shows a strong
reluctance to classify verse types according to the number of their syllables. Bedes choice of
the term tetrameter to describe the terentianean metre does seem somewhat more difficult
to explain, but it is possible that he perceived the metre as a dactylic pentameter lacking the
two final syllables.
64

Commentators have found it puzzling that Bede attributes Squalent arva soli
and Obduxere polum to Ambrose of Milan. That they have certainly not been composed by
Ambrose has been fairly well established on stylistic grounds; furthermore, none of
Ambroses authentic poems are in the terentianean metre. Efforts have, however, been made
to trace the origins of these hymns. They have both been published by Bulst in Analecta
hymnica XXVII because, together with several other texts in the same metre,
65
they have
formed a part of Visigothic hymnody. Wilhelm Meyer, in his Gesammelte Abhandlungen,
indicated that the hymns 203, 204 and 206 have probably been composed by the same poet,
66

and F. J. E. Raby placed their author in 5
th
-century Southern Italy.
67
More importantly,
however, the hymns in question had been used in the monastery of Monte Cassino, where
they were presumably referred to with the generic phrase Ambrosiani. It is known that the
hymns were, at least on one occasion, transmitted in the manuscript that also contained the
Benedictine rule, and it is presumably through the Benedictine usage of Ambrosianus as a
nonspecific expression that Bede came under the impression that the hymns were actually

63
Trans. Kendall 1991, 149.
64
Kendall 1991, 149.
65
i.e. AH XXVII 200 Rex aeterne deus, 201 Iram quam merito, 202 Huius supplicium and 205 Tristes nunc
populi.
66
Meyer 1905, 295.
67
Raby 1947, 1-3.
158

composed by Ambrose.
68
It is obvious that Bedes mistaken belief that the hymns were by
Ambrose must have invested the metre with exceptional authority and impelled Bede to
include it in his very sparse corpus of lyrical metres.
It is worth mentioning that the hymns quoted in Bedes chapter on the so-called
Terentianean metre all conform to the strict verse structure ---uu- | -uu-x, as presented by
both Terentianus Maurus and Julian of Toledo: they are conspicuously devoid of the
variations introduced by Boethius. Arguably, Bedes codification of the Terentianean metre
may have contributed to its popularity among medieval poets and hymnodists; however, we
also know that their efforts were often modelled on the looser, Boethian form than the strict
verse type represented by Bedes pseudo-Ambrosian models. But this may ultimately be due
not only to the example of Boethius but also to Bedes unambiguously dactylic interpretation
of the metres structure.


3.4. The anacreontic metre

The presence of the anacreontic metre among Bedes planed-down selection of lyric metres
may seem surprising to many readers, given the historical background of the metre: not only
is the anacreontic metre often associated with imagery of a somewhat profane nature, but the
Roman poets appear to have made hardly any use of it. As we shall shortly observe, Bedes
decision to include the metre in his treatise appears, once more, to have been made on the
strength of a single poem by a single, possibly falsely attributed poet; a decision that
ultimately had far-reaching repercussions on later Christian poetry.
The anacreontic metre (uu-u | -u--) has received its name from the Greek lyric
poet Anacreon, who employed the metre in his poetry on love and wine. It is generally
classified together with the ionic metres, which normally consist of a sequence of two short
syllables alternating with two long ones, the so-called minor ionic metron (uu--) having two
short syllables followed by two long ones, and the so-called major ionic, the other way
around (--uu). The traditional analysis of the anacreontic metre is that it is an anaclastic
form of the minor ionic dimeter (uu-- | uu--): in other words, that it is a minor ionic dimeter
where the final syllable of the first metron has changed places with the initial syllable of the

68
Norberg 1998, 256-57; Schaller 1984, 73-90.
159

second one.
69
This analysis of the anacreontic metre has been contested by C. M. J. Sicking,
70

among others, but the fact remains that in Greek poetry the anacreontic line frequently
alternates with the normal (non-anaclastic) minor ionic dimeter, as testified by the poetry of
Anacreon himself.
The anacreontic metre is surprisingly rare in Latin poetry: in the poetic works of
classical Roman antiquity it makes only a few isolated appearances and even then often in
conjunction with other metres. We know from a quotation by Terentianus Maurus that
Petronius used it,
71
and in late antiquity, it makes up the first part of the strophes in
Claudians Fescennina 2. Perhaps more importantly for scholars of Latin literature, it forms a
part of the Hellenistic verse-type known as the galliambic, which is best known for its use in
Catulluss so-called Attis poem (Catull. 63). The galliambic can best be analysed as a
combination of two anacreontic lines, the latter of which is catalectic (docked of its last
element): uu-u | -u-u | uu-u | -u-. The galliambic metre also allows frequent fusion of the short
elements into long ones and the resolution of long elements into double-shorts. The fusion of
short elements mainly takes place at the beginning of each anacreontic:

sectam meam exsecutae duce me mihi comites.
72
--u | -u-- | uu-u | uuu-

The anacreontic and and its derivative, the galliambic, were traditionally associated either
with light-hearted subjects or images of orgiastic revelry, which is still reflected by the
modern usage of the word anacreontic simply to mean a drinking song. In the Greek east, the
anacreontic metre was employed by later-day emulators of Anacreon in classicising texts
collected in a corpus called Carmina anacreontea, which greatly contributed to this trend.
But in the Greek-speaking world, the metre also survived the transition to Christianity
surprisingly well, as did many other lyric metres: it was adopted by several early Christian
poets, including the pre-eminent Gregory of Nazianzus, who used the metre in two of his
poems, as well as such figures as John of Nyssa and bishop Synesius of Cyrene. Christian
poetry in anacreontics appears to have been cultivated, in particular, in the schools of
Palestine and Syria.
73


69
Raven 1968
2
, 69; Raven 1965, 129: To some extent the outside syllables of the metron seem to be variable,
uu-- being replaced by uu-u and -u--; and these two alternatives are joined in the form known as anacreontic
uu-u -u--, which in Greek verse often occurs as a variant of the normal ionic dimeter.
70
Sicking 1993, 124.
71
Ter. Maur. 2852; 2862-2865; Petr. fr. 20.
72
Catull. 63,15.
73
Petit 1924, 1856-72; Lapidge 1996, 231.
160

In Late Latin grammar and poetry, on the other hand, the prosodic nature of the
anacreontic metre became obscured, and it frequently came to be treated as an iambic metre.
It was often viewed by the grammarians as a variant of the iambic dimeter catalectic (x-u- | x-
x) with a mandatory resolution of the first element. This analysis of the anacreontic is
understandable, as a fusion of the two short elements at the beginning of the line would result
in what is in essence a iambic line. This trend can also be interpreted as another instance of
the Late Latin tendency to view all lyric metres as forms of either dactylic or iambo-trochaic
verse. The iambic dimeter catalectic itself was never very popular among the Romans, one of
the best examples being another fragment of Petronius (fr. 19), and it does not appear to have
gained much popularity in late antiquity. Its most notable appearance in Christian Latin is the
hymn Hymnus ante somnum in Prudentiuss Cathemerinon 6. The poem consists of thirty-
eight strophes of seven lines each, and, according to my observation, in 33 cases the opening
foot of the line is an anapest, resulting in a verse that is identical with the anacreontic.
74
This
is all the more notable as resolution of a long element or syllaba anceps into a double-short
does not occur anywhere else in the poem and is, indeed something of a rarity in the iambic
poetry of late antiquity.
75
This feature of the hymn was also recognised by Dag Norberg, who
discusses the iambic dimeter catalectic in conjunction with the anacreontic metre,
76
and it is
analogous to a similar resolution in the (acatalectic) iambic dimeter of the Ambrosian
hymns.
77
In the grammarians, the analysis of the anacreontic line as a form of iambic verse
can be encountered, most notably, in De metris by Mallius Theodorus, which apparently
served as the primary model for Bedes own analysis of the anacreontic metre:

Metrum iambicum tetrametrum colobon, quod anacreonticum dicitur, recipit supra
dictos pedes omnes et pro quarto pede una tantum syllaba ponitur; quod maxime fit
sonorum, si primus pes anapaestus ponatur, post duo iambi, deinde syllaba. Huius
exemplum: Triviae rotetur ignis, / volucrique Phoebus axi / rutilum pererret
orbem.
78

[The iambic tetrameter catalectic, which is called the anacreontic, takes all the
aforementioned feet, but, in place of the fourth foot, only one syllable; this becomes
most sonorous if the first foot is an anapest, followed by two iambs and one syllable.
An example of this: (You see how) Trivias fire turns and Phoebus, on his flying
chariot, travels the golden orb.]


74
e.g. line 10 redit et quietis hora.
75
Norberg 1958, 72-73.
76
ibid.
77
Norberg 1988, 17-18.
78
gramm. VI, 593, 24-27; Petr. fr. 20.
161

Malliuss presentation is curious in many ways. His terminology is typical of the Late Latin
grammarians in having totally abandoned the classical concept of the iambic metron: for
Mallius, the iambo-trochaic metres consist of autonomous feet, which also results in his
applying the term tetrameter to what more properly is a dimeter. His definition of the iambic
dimeter catalectic, with the phrase the aforementioned feet, refers to his earlier definition of
the iambic trimeter, which is entirely classical.
79

Mallius appears to regard the anacreontic as merely another name for the iambic
dimeter catalectic. He does also present an analysis of the proper anacreontic metre,
although simply as a superior variant of the metre, which he illustrates with a three-line
quotation from a fragment of Petronius previously cited by Terentianus Maurus. Although no
more than four lines of the original poem have been preserved to us, it is apparent that it was
originally composed in anacreontics in the strictest sense of the word;
80
furthermore, we also
know Petronius to have employed the iambic dimeter catalectic as a metre in its own right.
81

Malliuss presentation of the proper anacreontic does not differ from the
ordinary iambic dimeter catalectic only in having an initial pair of short syllables: Mallius
specifically states that in the preferred form of the metre, the third foot is an iamb (not a
spondee), and he rules out all syllable resolution in the latter part of the line. The prosodic
structure of the most sonorous form of the anacreontic line is strictly determined and
without the prosodic variation that is typical of the genuine iambic metres.
Bedes chapter on the anacreontic, drawn from Mallius, also presents the
anacreontic metre under the name of the iambic dimeter catalectic, but, in stark contrast to
Mallius, Bede only presents the anacreontic metre in the strictest sense, not discussing other
possible types of the iambic dimeter catalectic:

Metrum iambicum tetrametrum colophon
82
, quod Anacreontium dicitur, recipit
anapestum, duos iambos et semipedem.
83

[The iambic tetrameter catalectic, which is called the anacreontic, takes an anapest,
two iambs and a half-foot.]
84


79
Recipit vero metrum iambicum hexametrum pedes hos: iambum locis omnibus, tribrachyn locis omnibus
praeter novissimum, spondium, dactylum et anapaestum locis tantum inparibus, pyrrichium loco tantum
ultimo. (The iambic hexameter (=trimeter) takes the following feet: the iamb in every foot, the tribrach in
every foot but the last one, the spondee, the dactyl and the anapest only in the odd feet, and the pyrrhic only in
the last foot.) gramm. VI, 593, 6-9.
80
The first line, triplici vides, ut ortu, not cited by Mallius, is also an anacreontic line in the strict sense.
81
e.g. Petr. fr. 19: Horatium videmus / versus tenoris huius.
82
Colophon is a corrupt form of colobon (Gr. kooov) given by manuscripts of De arte metrica. Keil, in his
Grammatici Latini, has emended it to colobon.
83
DAM 22, 2-3.
84
Trans. Kendall 1991, 157.
162


Bede, like Mallius Theodorus, apparently considers the anacreontic to be simply another
name for the iambic dimeter catalectic. What is curious is that Bede has not derived his
explanation of the metres composition from that of the other iambic metres: he only presents
the structure of the proper anacreontic, not discussing the true iambic dimeter catalectic at
all. All that is left of Malliuss presentation of the anacreontic as an iambic verse-form, apart
from the wording of the metrical analysis, is that Bede discusses the anacreontic immediately
after the iambic dimeter. Bede illustrates his analysis with a quotation from a poem known as
Poema coniugis ad uxorem or Ad coniugem suam, generally attributed to Prosper of
Aquitaine:
85


Quo usus est Prosper Tiro in principio exhortationis ad coniugem, ita dicens:
Age, iam precor mearum
comes inremota rerum,
trepidam brevemque vitam
Domino Deo dicemus.
Celeri vides rotatu
rapidos dies meare
fragilisque membra mundi
minui perire labi.
Fugit omne quod tenemus,
neque fluxa habent recursum,
cupidasque vana mentes
specie trahunt inani.
Ubi nunc imago rerum est?
Ubi sunt opes potentum,
quibus occupare captas
animas fuit voluptas?
86

[Tiro Prosper uses this metre at the beginning of his exhortation to his wife. He says:
Come now, I pray,
inseparable companion of my affairs,
let us dedicate an anxious, brief life
to the Lord God.
You see our days rapidly
whirling away
and the inhabitants of a fragile world
declining, stumbling, perishing.
Everything which we possess is fugitive,
nor do fleeting things ever return,
and vain things attract the avaricious heart
with insubstantial appearance.
Where now is the shadow of reality?

85
In manuscript tradition, the poem is variously attributed either to Prosper or Paulinus of Nola, although the
most recent opinion seems to be that it is by Prosper. Santelia 2009, 10-13.
86
DAM 22, 4-21.
163

Where is the wealth of the powerful,
with which pleasure was able to ensnare
captive minds.]
87


The bulk of Poema coniugis is composed in elegiac couplets; only the introduction is in the
anacreontic metre, in keeping with the Roman custom of combining the anacreontic metre
with other types of verse. It has, indeed, been suggested that the immediate model of the
poem is Claudians Fescennina 2, and that Prosper has sought to create a spiritual equivalent
of Claudians epithalamion:
88
the poem is an exhortation to a fundamental life change and the
adoption of a pious and chaste life-style in times of turbulence. It may be telling that both
poems begin with the exhortation age.
89

Bede cites the introduction in its entirety. Two things are evident from Bedes
treatise of the anacreontic metre: firstly, Bede, in presenting the metrical scheme uu-u | -u-- as
the only type of the Anacreontic metre, has followed his usual practice of trimming his
definitions down to the bare essentials. Secondly, it is obvious that this simplified analysis is
based on his quoted example, the introduction to Poema coniugis. In his presentation of the
anacreontic metre, Bede has been faithful to the method he applies throughout De arte
metrica: although the phrasing of his definitions may be borrowed from previous treatises,
their content is not, being more or less a description of his cited material. The wording may
resemble that of Mallius Theodorus, but the substance is based on the evidence of a single
poem.
The importance of Poema coniugis for Bede is surely attributable to his belief
that the poem was the work of Prosper of Aquitaine. Undoubtedly, the subject and ethos of
the poem must also have appealed to Bede greatly, impelling him to quote its introduction in
its entirety. The introduction of Poema coniugis served as a model for later generations of
Latin poets: we know that at least Walafrid Strabo and Godescalc of Orbais have imitated it
in their works.
90
In limiting his definition of the anacreontic metre to the form presented in
Poema coniugis, Bede also inadvertently restored the metre, bringing it closer to its classical
models as a metre manifestly different from the iambic verse forms with which it was often
confused in late antiquity.


87
Trans. Kendall 1991, 157.
88
Chippaniniello 2007, 115-138.
89
Claudian. fesc. 2: Age, cuncta nuptiali / redimita vere tellus / celebra toros eriles (Come now, all earth,
garlanded by the nuptial spring, celebrate the wedding of your master.)
90
Norberg 1958, 72-73; see PAC II, 412; III, 733; IV, 246.
164


3.5. The iambic trimeter

The iambic trimeter is the metre of dramatic dialogue par excellence of classical Greek
drama. In its classical form, the iambic trimeter consists of three iambic metra (x-u-), or, in
other words, metra of two iambs (u-) where the first element is a syllaba anceps and can
manifest itself as either long or short: x-u- | x-u- | x-ux. The metre can allow frequent syllable
resolution: both the long elements and the syllabae ancipites at the beginning of the metron
can be resolved into two short syllables. To put it in the terms of the Latin grammarians, this
means that the iambic feet can be substituted with tribrachs (uuu) in every foot (but the final
one), but in the odd feet, where the initial syllable can be either long or short, also with
spondees (--), dactyls (-uu) and anapests (uu-). If resolution occurs both in the syllaba anceps
and the following long element, even proceleusmatics (uuuu) are possible, although highly
unusual.
91
Several variants of the iambic trimeter were used in Roman antiquity. Catullus, in
his fourth poem, experimented with a form that allowed no syllabae ancipites and where
every foot is an iamb, and a similar type of verse alternates with dactylic hexameters in
Horaces sixteenth Epode. Generally, though, such extreme regularity was seldom practised
by the Roman poets, although the iambic trimeter, as other metres, underwent some
regularisation over the course of its history in Latin literature.
Already in the Golden Age, the more outr cases of syllable resolution were
going out of vogue in iambic verse. Resolution of the long element and the metron-initial
anceps are comparatively rare in the Epodes of Horace: such lines as alitibus atque canibus
homicidam Hectorem
92
(- uu u - | u uu u uu | - - u -) are genuine exceptions from this general
tendency. Some of the Silver Age poets swung rather in the opposite direction (mainly
Seneca in his tragedies),
93
but the Late Latin users of the iambic trimeter generally shunned
any form of syllable resolution; this general tendency seems to apply to iambo-trochaic verse
in general.
94

Dag Norberg has discussed the late antique evolution of the iambo-trochaic
metres extensively in his 1988 Les vers latins iambiques et trochaques au Moyen Age et

91
e.g. Sen. Med. 670: Pavet animus, horret, magna pernicies adest (uu uu u - | - - u - | uu - u -).
92
Hor. epod. 17, 12.
93
Raven 1965, 51-52.
94
Norberg 71-72. The most notable exceptions are the syllable resolution that usually occurs in the beginning of
the Ambrosian iambic dimeter (as in Ambros. hymn. 6, 19 Geminae gigas substantiae, also quoted by Bede in
DAM 21, 22) and a similar resolution in the iambic dimeter catalectic of Prud. cath. 6, which may result from
analogy to the anacreontic metre.
165

leurs rpliques rythmiques, where his starting hypothesis is that the simplification and
regularisation which these metres underwent in late antiquity ultimately led to the birth of the
rhythmic verse of the middle ages. This simplification and regularisation took the form of
several developments that guided the iambo-trochaic metres in a more pronouncedly
isosyllabic and accentually regular direction. The most important phenomena are the gradual
disappearance of elisions and syllable resolution and an increasing tendency to make the
word-accent coincide with the ictus, mainly towards the end of the line.
95

The classical iambic trimeter, as presented above, had its equivalent also in
the popular forms of iambo-trochaic verse employed in the comedies of Plautus and Terence.
This form, known as the iambic senarius, differed from the classical variety of the iambic
trimeter mainly in the respect that the syllaba anceps was not restricted to the odd feet of the
line but could appear anywhere but in the final foot. The resulting metre consists of a regular
sequence of feet (rather than the two-foot metra of the classical variety): x - x - | x - x - | x - u
-. In the dialogue of Plautus and Terence, syllable resolution is also very common, and, given
the higher frequency of syllabae ancipites, can make the metre at times almost
unrecognisable.
96
As a vehicle of popular literature, the iambic senarius retained its
popularity throughout classical antiquity, and was employed as an archaising metre by
several late antique authors as well, including Ausonius and, in Christian poetry, Hilary of
Poitiers.
97
Oddly enough, despite Hilarys example, the metre seems to have fallen into near-
complete disuse in the middle ages.
98
In late antiquity, the archaic iambic senarius seems to
have been subject to changes similar to those its classical equivalent, the iambic trimeter: it,
too, became increasingly isosyllabic, and syllable resolution, which in the plays of Plautus
and Terence was very profuse, all but disappeared.
99

In the earlier grammarians, the classical iambic trimeter was still clearly
distinguished from its popular form, the iambic senarius. Terentianus Maurus in his De

95
Norberg 1988, 66. As examples of the gradual disappearance of elisions and syllable resolution, Norberg
mentions verses 155-271 of Paulinus of Nolas Carmen 21, which contain 40 elisions and 50 cases of syllable
resolution, contrasting them with Prudentiuss hymn O Nazarene, lux Bethlem, verbum Patris (Prud. cath. 7),
the 220 lines of which only contain 27 elisions and 11 cases of syllable resolution. Still later, in seventh-century
Visigothic Spain, the poetry of Braulio of Zaragoza and Eugene of Toledo appears even more simplified: in a
80-line hymn of Braulio, there are only 6 cases of elision and no occurrences of syllable resolution, and in the
64 iambic trimeters of Eugene, both phenomena are completely absent.
96
e.g Ter. heaut. neque ut animum decuit aegrotum adulescentuli (uu uu uu | u - - uu | - - u -).
97
e.g. Ausoniuss Ludus septem sapientium and Hilarys hymn Fefellit saevam, which has been thoroughly
analysed in Meyer 1905, 164.
98
Norberg 1988, 69.
99
Norberg 1958, 72: A ce propos, nous voulons faire remarquer quil est relativement rare quune syllabe
longue se decompose en deux brevesaprs saint Ambrose et Prudence il est rare que lon rencontre des cas de
ce genre, ceci valant aussi bien pour le vers ambique classique que pour le vers ambique arcasant.
166

littera, de syllaba, de pedibus emphasises that those poets who use the latter do so not
through ignorance but deliberately (in metra peccant arte, non inscitia).
100
He is also
notable for still using the terms trimeter and senarius in distinctly different meanings (sic fit
trimetrus qui fuit senarius).
101
In later grammarians, the terminology became much more
confused, which was certainly compounded by the increasing tendency to view iambic metres
as consisting of individual feet rather than metra of two feet. In several Late Latin
grammarians we see the iambic trimeter referred to as the iambic senarius, but also as the
iambic hexameter, in this context an outrageously unclassical term. Even when the clear
distinction between the two types of line had become blurred, some grammarians retained
some idea of their differences. The fourth-century Maurus Servius, in his De centum metris,
writes in the following manner:

Metra iambica locis inparibus quinque recipere possunt pedes, iambum, tribrachum,
spondeum, dactylum, anapaestum, locis autem paribus tantum iambum vel
tribrachum, et apud comicos frequenter anapaestum, ita tamen ut multarum brevium
iunctura vitetur.
102

[The iambic metres can take five feet in the odd positions: the iamb, the tribrach, the
spondee, the dactyl, and the anapest, but in the even positions only the iamb or the
tribrach, and in the comic writers, often the anapest, but in such a way that a juncture
of many short syllables is avoided.]

Mariuss description of the comic versions of iambic metre is undeniably imprecise: of all
the feet made possible by syllabae ancipites in the even feet, he only recognises the anapest;
however, he still shows some grasp of the prosodic differences between these two verse
types. There is nothing left of this in Mallius Theodoruss De metris, which served as Bedes
model for his short chapter on the iambic trimeter:

Recipit vero metrum iambicum hexametrum pedes hos: iambum locis omnibus,
tribrachyn locis omnibus praeter novissimum, spondium, dactylum et anapaestum
locis tantum inparibus, pyrrichium loco tantum ultimo.
103

[The iambic hexameter (=trimeter) takes the following feet: the iamb in every foot,
the tribrach in every foot but the last one, the spondee, the dactyl and the anapest
only in the odd feet, and the pyrrhic only in the last foot.]

Bede has retained Malliuss wording to a high degree:

100
Ter. Maur. 2239.
101
Ter. Maur. 2265.
102
Soraci 1988, 26, 24 27, 2.
103
gramm. VI, 593, 6-9.
167


Metrum iambicum senarium recipit iambum locis omnibus, tribrachin locis omnibus
praeter novissimum, spondeum, dactylum, et anapestum locis tantum inparibus,
pyrrhichium loco tantum ultimo.
104

[The iambic senarius takes the iamb in every foot, the tribrach in every foot except
the last, the spondee, the dactyl, and the anapest in the odd feet only, and the pyrrhic
only in the final foot.]

As we can see, Bede has once more tinkered with Malliuss presentation: he has substituted
the word senarium for Malliuss hexametrum. Otherwise, the presentations are identical and
certainly portray the classical iambic trimeter, not the senarius. Malliuss choice of the word
hexametrum may reflect his effort to portray an analogy between the dactylic hexameter and
the iambic trimeter:

Metro autem dactylico heroico metri iambici hexametri proxima definitio visa est.
Altero enim producuntur acies, ubi proelia conseruntur, altero clarorum hominum
vita atque fortuna scenam personat; atque idcirco utriusque conexio in earum rerum,
quae sunt maximae, expressione ac declaratione versatur.
105

[Next to the heroic dactylic metre, it seems most fitting to define the iambic
hexameter. For the one is used to portray armies and battles, with the other the lives
and fates of famous men are resounded on-stage; therefore each of the two is used in
expressing and declaring those things which are of the greatest importance.]

To quote this would not have fitted well with Bedes Christian ethos. Instead of alluding to
war and the stage, Bede has presented the iambic trimeter with a quotation from Prudentiuss
Psychomachia. Bedes introduction to the citation can be read as a Christian paraphrase of
Malliuss presentation. It can even be understood to imply the connection of the iambic
trimeter and the dactylic hexameter as portrayers of great things, albeit in a Christian context:

Quo nobilissimus Hispaniarum scholasticus, Aurelius Prudens Clemens, scripsit
proemium Psychomachiae, id est, libri quem de virtutum vitiorumque pugna heroico
carmine conposuit. Ita enim inchoat:
Senex fidelis, prima credendi via
Abram, beati seminis serus pater,
adiecta cuius nomen auxit syllaba,
Abram parenti dictus, Abraham Deo (Prud. psychomach. praef. 1-4).
106

[Aurelius Prudentius Clemens, the celebrated scholar of Spain, wrote in this metre
the prologue of the Psychomachia, i.e., the book which he composed in heroic verse
on the battle of the virtues and vices. It begins this way:
107


104
DAM 20, 2-5.
105
gramm. VI, 593, 2-5.
106
DAM 20, 5-12.
168

The faithful patriarch who first showed the way of believing,
Abram, late in life the father of a blessed progeny,
whose name was lengthened by a syllable,
for he was called Abram by his father, but Abraham by God.]
108


In the light of Malliuss and Bedes thorough exposition of all the possible feet which can
appear in the iambic trimeter, the introduction to Psychomachia is very subdued in the matter
of syllable resolution: there are only four instances in its seventy lines, and all of them belong
to the common type where the initial syllaba anceps is a double-short. If we compare Bedes
exposition of the iambic trimeter with those of the other iambo-trochaic metres, the contrast
is quite remarkable: in his examination of the iambic dimeter, Bede only discusses syllable
resolution as an afterthought, and in his chapter on the trochaic septenarius it is not
mentioned at all; this would mainly be the result of his basing the portrayals of the metres in
question on his analyses of the individual poems he uses as their illustration. In view of this,
Bedes description of the iambic trimeter would certainly appear to be more thorough than his
material warrants, and, in this instance, Bede himself comes across as unusually derivative,
and rather less than usually empirical, in his reliance on previous authors.


3.6. The iambic dimeter

The prominence of the iambic dimeter in Late Latin poetry can be wholly attributed to
Christianity and Christian hymnody. Although this metre appears in Greek drama (and even
there mainly as a component of larger metrical units),
109
its use in classical Latin poetry was
infrequent, and even then it mainly appeared in conjunction with other types of usually
iambic verse.
The structure of the classical iambic dimeter is quite analogous to that of the
iambic trimeter: it consists of two iambic metrons (x - u-) of two iambs each. The first foot of
the metron is, as in the iambic trimeter, a syllaba anceps. Both the long syllables and the
syllabae ancipites can be substituted with a double-short, and therefore the metre, at least in
theory, allows the same panoply of possible metrical feet as the iambic trimeter: tribrachs
(uuu) in every foot but the last, spondees (--), dactyls (-uu), anapests (uu-) and
proceleusmatics (uuuu) in the odd feet.

107
Trans. Kendall 1991, 153.
108
Trans. Thomson 1949, 275.
109
Raven 1968
2
, 33.
169

The iambic dimeter very rarely appeared on its own in classical Latin verse (the
tragedies of Seneca being the most prominent example), and they are mainly known to
students of the Roman classics from the Epodes of Horace, where they appear in alternation
with other lengths.
110
Horaces emulators in their use include Martial and Ausonius. Analyses
of Horaces Epodes show that the iambic dimeter was from the very start treated as a more
isosyllabic verse form than the more common trimeter: there are only two instances of
syllable resolution in his dimeters.
111
Horaces Silver-Age emulators were freer in their use of
syllable resolution, but this was to become increasingly rare towards the end of antiquity,
although resolution still occurs sporadically in Bedes main authorities on the iambic dimeter,
Sedulius and Ambrose.
It is unclear why the iambic dimeter, for all its previous obscurity, became the
metre of choice of the early Latin hymnodists. Apparently its short length made it suitable for
the singing of hymns; Wilhelm Meyer, in his Gesammelte Abhandlungen, has, on the strength
of some statements by St Augustine, even ventured to derive the origin of Latin hymns in the
iambic dimeter from oriental models.
112
In Christian hymns in the iambic dimeter, four lines
normally form a stanza, and as the hymns were commonly sung antiphonally, the number of
the stanzas is usually even. The best-known representatives of Christian hymns in the iambic
dimeter are the hymns by Ambrose of Milan and Caelius Sedulius, but other contributors to
this genre include Ennodius of Pavia, Paulinus of Nola and Venantius Fortunatus. Although
this would not be strictly necessary in verse that is primarily intended for singing, the early
Latin hymns in the iambic dimeter by Sedulius, Ambrose and their near-contemporaries have
been composed in a strictly classical form of iambic metre: only the first syllable of each
metron (or the first syllables of odd feet) is a syllaba anceps, and syllable resolution is
generally restricted to the beginning of the metron, a feature common with the Late Latin
iambic trimeter.
113
Ambrose of Milan composed altogether 14 hymns in the iambic dimeter.
These hymns comprise 448 lines and show only 18 cases of syllable resolution; of these

110
Horaces first ten epodes have the iambic dimeter alternating with the iambic trimeter (e.g. epod. 1.1-2: Ibis
Liburnis inter alta navium / amice, propugnacula). In his epodes 11-13 and 14-15, the iambic dimeter is used in
conjunction with dactylic elements in what are termed archilochean stanzas: the so-called iambelegus, which
consists of an iambic dimeter and a hemiepes (x-u- x-u- | -uu -uu -) forms the second archilochean stanza
together with the dactylic hexameter, and the so-called elegiambus, where the same components are in an
inverted order (-uu -uu - | x-u- x-u-) forms the third archilochean stanza together with the iambic trimeter. See
Raven 1965, 112-13.
111
Raven 1965, 60.
112
Meyer 1905, 119; Augustine describes the nightly services of Milan in his Confessions, 9, 7: Tunc hymni et
psalmi ut canerentur secundum morem orientalium partium, ne populus maeroris taedio contabesceret
(Then it was first instituted that after the manner of the Eastern Churches, Hymns and Psalms should be sung,
lest the people should wax faint through the tediousness of sorrow)
113
Norberg 1958, 71-72.
170

fourteen occur in the first syllable of the metron.
114
Although also Prudentius applied syllable
resolution in his hymns, Sedulius refrained from it altogether in his hymn A solis ortus
cardine, and Ennodius only used it in proper names.
115
At the same time also elisions became
increasingly uncommon, and there was a pronounced tendency to eliminate the clash of
accent and ictus from the end of the line, thereby giving it a regular proparoxytone
(antepenultimate) accent. All of these developments served to make the iambic dimeter easier
to grasp for audiences without training in classical verse and in time also paved the way for
the emergence of rhythmic poetry.
Prior to Bedes De arte metrica, the growing importance of the iambic dimeter
was in no way noted by the grammarians who generally lump it together with the other
iambic metres, if they bother to discuss it at all. Diomedes does not mention the iambic
dimeter, and in Mallius Theodoruss De metris, the metre, which Mallius calls an iambic
tetrameter, is sandwiched between the iambic trimeter and the iambic dimeter catalectic. The
description of the metre itself is very sparse:

Metrum iambicum tetrametrum recipit supra dictos pedes omnes. Huius exemplum:
Merulae quod os vetustius.
116

[The iambic tetrameter takes all the aforementioned feet. An example of this:
Merulae quod os vetustius.]

With his expression supra dictos pedes, Mallius refers to his definition of the iambic trimeter,
as described above (p. 166), which is in essence perfectly classical (iambs and tribrachs in
every foot but the last, spondees, dactyls and anapests in the odd feet). Although Bede usually
follows Malliuss wording fairly closely in his descriptions of the lyric metres, the description
of the iambic dimeter in De arte metrica reads as a redefinition that is no longer based on the
general classical rules of iambic verse, but portrays metre as it had been adopted by the
Christian hymnodists. As a consequence, the definition is very much streamlined: Bede has
left out most of the metrical options which the classical form of the metre, at least in theory,
allows. Bede acknowledges the option of syllable resolution only grudgingly and then mainly

114
Norberg 1988, 17-18; the first foot is an anapest seven times (e.g. geminae gigas substantiae) and a dactyl
three times (martyribus inventis cano); the second foot is once a tribrach (carnis vitia mundans caro), and
the third foot is an anapest seven times (mortis sacrae meritum probat). Norberg also mentions that in a hymn
in honour of John the Evangelist that is attributed to Ambrose there are some unusual cases where, contrary to
classical practice, the second foot is an anapest (followed by a hiatus). These lines, however, are direct
quotations from the Gospel of John: in principio erat verbum, and in principio apud deum, from John 1:1.
115
Norberg 1988, 18. Some of Ennodiuss syllable resolutions run counter to the classical rules of iambic metre
(e.g. vatis Cypriani et martyris with an anapaest in the second foot) and almost seem closer to the archaic type
of Latin iambs.
116
gramm. VI, 593, 21-23.
171

in the position to which it was usually confined in Late Latin metre: the initial element of the
metron. Even this liberty is absent from the beginning of the chapter where Bede lays down
the rules of the metre. All that remains from Mallius Theodoruss presentation is the
definition of the iambic dimeter as a tetrameter. Bede has illustrated his redefinition of the
iambic dimeter with the opening lines of Seduliuss hymn A solis ortus cardine, Seduliuss
best-known contribution to the genre of Christian hymnody:

Metrum iambicum tetrametrum recipit iambum locis omnibus, spondeum locis
tantum inparibus. Quo scriptus est hymnus Sedulii:
A solis ortus cardine
adusque terrae limitem.
117

[The iambic tetrameter takes the iamb in every foot, and the spondee only in the odd
feet. A hymn of Sedulius is written in this metre:
From the rising of the sun in the east
to the ends of the earth in the west]
118


Bedes initial definition of the iambic dimeter is, in other words, entirely isosyllabic: for
Bede, the metre consists essentially only of iambs, with the option of spondees in the odd
feet. As Norbergs analysis of A solis ortus cardine shows, this definition corresponds with
the structure of Seduliuss hymn, which is structurally even simpler than the hymns of
Ambrose and does not contain a single case of syllable resolution. As Bede understandably
did not want to exclude Ambrosian verse from his treatise, he has found it necessary to
elucidate his discussion of the iambic dimeter with quotations of Ambrosian hymns, and
thereby also bring up the question of syllable resolution. Bedes first quotation from Ambrose
consists of a sampling of the opening lines of some of his best-known hymns (5, 4, 3, and 2,
respectively):

Sed et Ambrosiani eo maxime currunt:
Deus creator omnium;
Iam surgit hora tertia;
Splendor paternae gloriae;
Aeterne rerum conditor;
et ceteri perplures.
119

[But above all the hymns of Ambrose are composed in this metre; for example:
God the Creator of all things;
Now the third hour rises;
The splendour of the Fathers glory;

117
DAM 21, 2-6.
118
Trans. Kendall 1991, 155.
119
DAM 21, 7-11.
172

Eternal creator of things;
and a great many others.]
120


Somewhat surprisingly, Bede takes an aesthetic stand for a more regularised form of
iambic verse, where the odd feet of the line are regularly spondees, creating a repeated
spondee-iamb sequence (--u- | --u-). For his illustration of this type of iambic dimeter he has
chosen the opening lines of Ambroses fifteenth hymn, Aeterna Christi munera:

In quibus pulcherrimo est decore conpositus hymnus beatorum martyrum, cuius loca
cuncta inparia spondeum, iambum tenent paria, cuius principium est:
Aeterna Christi munera
et martyrum victorias,
laudes ferentes debitas,
laetis canamus mentibus.
121

[One of them is a hymn of great beauty and dignity on the blessed martyrs, of which
all the odd feet are spondaic and the even feet, iambic. This hymn begins:
Let us sing with joyful hearts
the eternal gifts of Christ
and the victories of the martyrs,
which bring about merited praises.]
122


Only at the very end of his exposition does Bede discuss the question of syllable resolution,
resorting once more to Mallius Theodoruss general description of the iambic metres. He
furnishes his discussion with a sole example drawn from Ambrosiuss sixth hymn:

Recipit hoc metrum aliquoties, ut scribit Mallius Theodorus, etiam tribrachin locis
omnibus praeter novissimum, dactylum et anapestum locis tantum inparibus. Unde
est:
Geminae gigas substantiae
alacris ut currat viam. (Ambr. hymn. 6, 19-20)
Ceterorum rara habes exempla.
123

[As Mallius Theodorus points out, this metre occasionally also takes the tribrachs in
every foot except the last, and the dactyl and the anapest in the odd feet only. This
explains:
Like a giant of two-fold nature,
he swiftly runs his course.
You will find few examples of the others.]
124



120
Trans. Kendall 1991, 155.
121
DAM 21, 12-18
122
Trans. Kendall 1991, 155.
123
DAM 21, 19-24
124
With the exception of the last sentence, the translation is from Kendall 1991,155.
173

The resolution presented by Bede for the benefit of his readers represents the type most
commonly encountered in the iambic verse of late Latin poetry. The line geminae gigas
substantiae (uu-u- | --u-) has syllable resolution of the initial syllaba anceps, or to put it in
the terms of the grammarians, has an anapest for its first foot. As Norbergs study shows, the
overwhelming majority of syllable resolutions in Ambroses hymns take place in the first
element of the metron, i.e. they have an anapest in an odd foot. Other types of syllable
resolution, of which there are but four instances in all of Ambroses hymns, are so rare that
Bede has apparently not bothered to illustrate them. We can see that here, again, Bedes
definition of a poetic metre is closely based on his exclusively Christian reading matter,
which also functions as its illustration. In Christian usage, the iambic dimeter was on its way
to becoming a totally isosyllabic metre: syllable resolution occurred seldom, and then mainly
in very restricted places. In Bedes eyes, this phenomenon did not merit discussion in the
actual definition of the metre, being for all intents and purposes a curiosity of marginal
interest. Yet Bedes pervasive tendency to simplify and regularise goes even further than this:
he specifically commends a type of iambic dimeter verse where even the use of spondees is
preordained, and all the odd feet are spondees. Bedes ideal form of the iambic dimeter can
be presented as an invariable chain of spondaic and iambic feet: -- u- | -- u-.
Bedes own hymns in the iambic dimeter corroborate his partiality for this
improved iambic dimeter. The iambic dimeter was Bedes metre of choice in his Liber
hymnorum, which has regrettably not survived to us in its entirety. According to Bedes own
account, he had composed hymns not only in the iambic dimeter but other metres as well, and
may even have composed rhythmic (non-quantitative) verse.
125
Although the authenticity of
some of the hymns that have circulated in Bedes name has been debated, it would seem that
the extant hymns that can with any certainty be attributed to him are all in the iambic dimeter.
As Bedes hymns were all intended for regular ecclesiastical use, it is sensible to presume
that their most important model was the so-called Canterbury Old Hymnal which had been
brought to Wearmouth-Jarrow in the late seventh century.
126
Although no extant manuscript
of the hymnal remains, its contents are known from secondary sources.
127
The Canterbury
Hymnal consisted of sixteen hymns altogether, all composed either in the iambic dimeter or
its rhythmic (non-quantitative) equivalent. Of the eleven metrical hymns, nine are almost
certainly by Ambrose, and it makes sense to see Ambrose, rather than Sedulius, as the

125
Bede mentions in his Historia ecclestiastica that he had composed a book of hymns in diverse metres and
rhythms (librum hymnorum diverso metro sive rythmo). Hist. eccl. 5, 24.
126
Lapidge 1996, 322. On the Old Hymnal itself, see Gneuss 1968, 10-40.
127
Lapidge 1996, 322; Gneuss 1968, 16-17 and 24-25.
174

primary model of Bedes own hymns in the iambic dimeter. One must bear in mind that
Bedes Liber hymnorum was, first and foremost, intended to be a supplement to the Old
Hymnal: Bede composed his hymns for those occasions which were not covered by the pre-
existing hymns at his disposal, and it obviously made sense to strive for a style that was, as
far as possible, uniform, consistent, and compatible with that of the older hymns.
The popular but arguably unfair view on Bedes hymns often appears to be have
been that they are somewhat bland and uninspired, which no doubt is based on the traditional
conception that good scholars make poor poets. They are, however, elegantly composed, and,
needless to say, metrically faultless. They are also quite in line with Bedes treatise of the
iambic dimeter in De arte metrica. The metre in Bedes hymns is, as one might expect, all but
isosyllabic: Dag Norberg has detected one sole case of syllable resolution, in Analecta
hymnica 82, 22, 2: dominus potens et fortis est (uu-u- | --u-), where the first foot is an
anapest,
128
making it prosodically similar to Ambroses geminae gigas substantiae, cited
above. It would also appear that Bedes admiration for the type of iambic verse where
spondees and iambs follow each other in strict alternation was not of a passive kind. As an
example we may cite the first four stanzas of Bedes perhaps most famous hymn, composed
for the day of the Holy Innocents:

Hymnum canentes martyrum --u- | --u-
dicamus innocentium, --u- | u-u-
quos terra flendo perdidit, --u- | --u-
gaudens sed aethra suscipit. --u- | u-u-

Vultum patris per saecula --u- | --u-
quorum tuentur angeli --u- | u-u-
eiusque laudant gratiam, --u- | --u-
hymnum canentes martyrum. --u- | --u-

Quos rex peremit impius, --u- | u-u-
pius sed auctor colligit u-u- | --u-
secum beatos collocans --u- | --u-
in luce regni perpetis. --u- | --u-

Qui mansiones singulis --u- | --u-
largitus in domo Patris, --u- | u-u-
donat supernis sedibus, --u- | --u-
quos rex peremit impius.
129
--u- | u-u-


128
Norberg 1988, 18.
129
Ed. Fraipoint 1955, 412.
175

[Chanting a hymn of joyous praise
we sing martyrs who are innocent
whom the earth relinquished in tears
whom the heavens receive rejoicing.

They whose face the angels of the Lord
gaze on in peace forever,
and praise His mercy
chanting a hymn of joyous praise.

Those whom the wicked king destroyed,
the merciful Creator now receives,
drawing to Himself the blessed ones,
in the light of the perpetual realm,

He Who grants mansions to all
inside His Fathers house,
now grants eternal dwelling-place
to those whom the wicked king destroyed.]
130


We can see that out of the sixteen lines, nine (in italics) consist of spondees and iambs in
strict alternation (--u- | --u-). Moreover, only one line out of sixteen begins with an iamb.
Bedes predilection for an iambic dimeter with as many spondees as possible, or at least a
spondaic opening, is made very apparent by this example. The hymn shows the strong
influence of Ambrose, and both its form and content have noticeable parallels with
Ambroses Aeterna Christi munera, quoted in De arte metrica, that go beyond the purely
metrical.
It is curious that apart from the dactylic hexameter, the only type of metrical
verse on which Bede makes an actual aesthetic judgement is the iambic dimeter. The obvious
reason is that these are the two metres which he himself mainly had employed as a poet. In
this instance, Bedes treatise on metrics is not only a description of existing verse and its
structures: it is a normative guide to what Bede perceived as the best possible form of verse,
as reflected by his own tastes and his own poetic technique.


3.7. The trochaic septenarius

The only trochaic metre even briefly discussed in Bedes De arte metrica is the trochaic
tetrameter catalectic, or, more properly, its archaising variety, the trochaic septenarius. That

130
Trans. Lapidge 1996, 329.
176

the iambic metres (the iambic dimeter and trimeter) are presented by Bede in their classical
form, whereas the works sole example of trochaic metre is given in its archaising variety, is
the direct result of certain metrical practices which were applied to the iambo-trochaic metres
in late antiquity, as well as Bedes choice of reading material: as we have stated previously,
Bedes slim corpus of lyric and iambo-trochaic metres is based on those metres that he had
encountered in the works of the foremost Christian poets of late antiquity. That a metre had
previously been discussed by the grammarians was for Bede not a strong enough criterion to
merit its presentation in his essentially Christian treatise on metre. Correspondingly, Bede
modelled his presentations of poetic metres as much on the form which they took in the
poems at his disposal as on how the metrical structures had been analysed by previous
grammarians. Bedes analysis of the trochaic septenarius is, in essence, based on the structure
of one single hymn, Hymnum dicat turba fratrum, a poem composed in archaising and highly
isosyllabic trochaic verse, and Bedes analysis, as usual, owes more to his own observation
than to the authority of any previous grammarian. Curiously, the analysis is also faulty: in
this case, Bedes do-it-yourself approach has resulted in an erroneous interpretation of the
metres prosodic structure. Bede has presented as rules features that are more properly mere
tendencies in the hymn at hand and ended up with a presentation that does not properly
correspond with any known form of Latin trochaic verse; archaic, classical, or archaising.
In Latin poetry, the trochaic metres were subject to variation similar to that
which had affected the iambic metres. In essence, both iambic and trochaic metres consist of
alternating short and long elements, but whereas in an iamb the short element precedes the
long one (u-), in a trochee the long element comes first (-u). The classical trochaic metron
consists of two trochaic feet, but, in certain positions, the short element can be a syllaba
anceps and thereby replaced with a long one. In this respect, the trochaic metron is the
reverse of the iambic one: in the trochaic metron, the last element is a syllaba anceps (-u-x),
in the iambic metron, the first one (x-u-). Syllable resolution, too, can take place in the long
elements as well as the syllabae ancipites, which can be substituted with a double-short (uu).
In effect, any trochaic foot can manifest itself as a tribrach (uuu), and the even ones, where
the latter syllable is an anceps, as spondees (--), dactyls (-uu), anapests (uu-),
131
and, at least
in theory, proceleusmatics (uuuu).
132


131
Raven 1965, 76.
132
Raven 1965, 43, 45 and 79: For prosodic reasons, proceleusmatics are much less common in trochaic metres
than in iambic ones: in iambo-trochaic poetry, a resolved anceps readily precedes but very seldom follows
resolved long, making trochaic feet (-x) with resolution of both elements considerably less common than iambs
(x-) with the same feature.
177

Quite like the iambic metres, the trochaic metres also have popular, or archaic,
variants, where the latter element of each foot can be a syllaba anceps, resulting in a regular
sequence of long syllables and syllabae ancipites: - x - x | - x - x Syllable resolution,
especially in the early comedy writers, can be profuse, although usually less extreme than in
iambic metres. This characteristic, compounded by the greater number of syllabae ancipites
makes a wide variety of different feet possible, often rendering archaic trochaic lines
challenging to read, as demonstrated by this line from Plautus (rud. 620):
133


Facit(e) hic lege potius liceat quam vi victo vivere.
uu - - u | uu - uu - | - - - - | - u x

The differences in the nomenclature of the classical and archaic varieties are parallel to those
of the iambic metres: the names of the classical types are derived from the number of metra
in the line (dimeter, tetrameter etc.), whereas those of the archaic or popular type have
names based on the number of feet in each line, although the terminology is in some cases
misleading. This applies, above all, to the trochaic tetrameter catalectic, the popular form of
which is known as the trochaic septenarius.
The trochaic tetrameter catalectic is the most common of the trochaic metres. It
is used in the dialogue of Greek drama, where it usually lends the lines a more agitated air
than does the more common iambic trimeter. It is assumed that the trochaic tetrameter
catalectic was the original metre of tragic dialogue and that it was only later supplanted by
the iambic dimeter. The trochaic tetrameter catalectic consists of three trochaic metra, the last
of which is catalectic, i.e. docked of its last syllable: - u - x | - u - x | - u - x | - u - . There is
commonly a diaeresis after the second metron. Apart from drama, the trochaic tetrameter
catalectic also appears in the Greek lyric poets.
134
In classical Latin poetry, this form of the
metre appears, above all, in the classicising tragedies of Seneca. The popular counterpart of
the trochaic tetrameter catalectic is known, rather perversely, as the trochaic septenarius. The
name of the metre is drawn from the number of complete feet in the line, as the trochaic
septenarius can be seen to consist of seven complete trochees and one extra syllable: - x | - x |
- x | - x | - x | - x | - u | -. It is worthy of note that even in the archaic septenarius, where most
of the short elements can manifest themselves as syllabae ancipites, the penultimate element
is always short, a feature that, as we shall see, may have contributed to Bedes erroneous

133
Raven 1965, 43.
134
Raven 1961, 34.
178

definition of the metre. The trochaic septenarius is a common metre in the dialogue of early
comedy. However, the use of the trochaic septenarius was not restricted to early drama, but
the metre appears to have coexisted with the classical trochaic tetrameter even in later poetry,
being particularly common in popular verse of all kinds. Recorded examples of its popular
usage include such lines as urbani servate uxores, moechum calvum adducimus
135
, which
the Roman soldiers chanted to Caesar in his triumphal procession, and childrens verse such
as habeat scabiem quisquis ad me venerit novissimus and rex erit qui recte faciet, qui non
faciet non erit.
136
We have adequate evidence to suggest that the trochaic septenarius, unlike
most archaic forms of iambo-trochaic verse, never completely fell into disuse but continued
its survival on the margins of poetry, having established itself as the Roman doggerel-writers
metre of choice.
137

Even the earliest examples of the trochaic septenarius commonly have a
diaeresis after the fourth foot, a feature that proved decisive for the later evolution of the
metre and its treatment by the grammarians.
138
Often the lines also have a secondary diaeresis
after the second foot.
139
The evident fact that trochaic septenarii tend to have word-breaks
coinciding with the endings of the feet serves to emphasise the lines prosodic regularity: in
the trochaic septenarius, the overwhelming tendency is for the accent to coincide with the
ictus, a feature that made it an obvious choice for marching songs.
140
That the trochaic
septenarius had a background of being used on parades also made it adaptable to Christian
processional hymns. Medieval texts in this genre often show a structural similarity to the
vernacular pagan examples cited above in having systematic accent-ictus coincidence, a
diaeresis after the fourth foot and often after the second foot as well.
141
It is curious to note
that in Late Latin poetry, including early Christian verse, the classical form of the trochaic
tetrameter catalectic (- u - x | - u - x | - u - x | - u -) appears to have coexisted with the
archaising trochaic septenarius (- x - x | - x - x | - x - x | - u -). Important examples of the

135
Suet. Iul. 51. On the other hand, Suetonius also cites some chants where the technique appears to be quite
classical: the odd feet are all trochees in e.g. Gallias Caesar subegit, Nicomedes Caesarem / ecce Caesar nunc
triumphat, qui subegit Galliam, / Nicomedes non triumphat, qui subegit Caesarem. Suet. Iul. 49.
136
Norberg 1988, 84.
137
ibid.
138
Raven 1965, 79.
139
Norberg 1958, 74; e.g. ecce Caesar || nunc triumphat || qui subegit Galliam, / Nicomedes || non triumphat ||
qui subegit Caesarem. Suet. Iul. 49.
140
ibid; Il faut remarquer que, dans tous ces vers qui taient destins a tre chants en marche ou en chur, les
accents concident avec les ictus. Whether the more literary poets took pains to avoid the clash of accent and
ictus appears to have been largely a matter of personal preference. That this was so is best demonstrated by the
anonymous Pervigilium veneris and the classicising Prudentius (e.g. Cath. 6), where there appears to be no
attempt to regularise the accentual rhythm. On the other hand, in Venantius Fortunatuss trochaic hymn Pange
lingua gloriosi, the accentual rhythm is regular in 90 per cent of the lines.
141
ibid; cf. Walahfrid Strabo in PAC II, 405: Imperator || magne vivas || semper et feliciter.
179

former include the anonymous Pervigilium Veneris and the sixth hymn of Prudentiuss
Cathemerinon, the latter was employed by Hilary of Poitiers in his hymn Adae carnis
gloriosae, and the anonymous composer of Hymnum dicat turba fratrum, previously also
attributed to Hilary and cited by Bede in his De arte metrica.
The descriptions of the trochaic tetrameter by the grammarians in general
correspond with the structure of the classical form of the metre. The huge popularity of the
trochaic tetrameter/septenarius and the relative obscurity of other trochaic metres is reflected
in the Ars of Diomedes, where the author does not specify that he is discussing the trochaic
tetrameter: for him the trochaic tetrameter is simply the trochaic metre:

Trochaicum metrum recipit pedes quinque: dactylum, spondeum, anapaestum
tribrachyn trochaeum, a quo nominatur. In triplicem autem feritur dipodian, et uni
cuique sine dubio pedes tam praeponuntur hi, trochaeus, dactylus, tribrachys, quam
subiungitur qui libet de supra memoratis quinque pedibus. Catalexin facit aut in
amphimacro aut in epitrito quarto.
142

[The trochaic metre takes the following five feet: the dactyl, the spondee, the anapest,
the tribrach, and the trochee, from which it receives its name. It is divided into three
dipodies (=metra), and at the beginning of each one are, without doubt, placed these
feet: the trochee, the dactyl, and the tribrach. This is followed by any one of the
aforementioned five feet. It makes a catalexis either in the amphimacer or the fourth
epitrite.]

Diomedes gives a general outline of the metres structure, which, however, is curiously
flawed. Diomedes is well aware that the odd feet of a trochaic line are unlike the even ones,
but he commits the error of allowing a dactyl in the odd feet of the metre, which is plainly
impossible, if we are discussing the metre in its classical form and do not permit a syllaba
anceps in an odd foot. It would appear that even to the grammarians, the distinction between
the classical trochaic tetrameter and the looser septenarius was not always perfectly clear-cut.
This is probably largely due to their excessive reliance on Greek authorities and the absence
of a theoretic framework for the archaic forms of the iambo-trochaic metres. Despite its
shortcomings, the analysis illustrates that even some of the late grammarians still perceived
the trochaic tetrameter as a line that consisted of two-foot metra, or dipodies, rather than
individual feet. This take on the trochaic metres was, however, on its way out, as Mallius
Theodoruss De metris, Bedes prime source on lyric and iambo-trochaic metres
demonstrates:


142
gramm. I, 504.
180

Iambico autem metro trochaicum metrum, quamvis et spondium et anapaestum et
dactylum recipiat, ut iambicum, eo tamen est contrarium, quod in iambico supra dicti
pedes locis tantum inparibus, in trochaico autem non nisi paribus conlocari licet.
Trochaeus vero in hoc metro, quod ex eo nomen accepit, locis omnibus, tribrachys
locis omnibus praeter septimum ponitur.
143

[The trochaic metre is converse to the iambic metre, although it, like the iambic,
takes the spondee, the anapest and the dactyl, because in the iambic metre, the
aforementioned feet may only be placed in the odd feet; in the trochaic metre, on the
other hand, only in the even ones. In this metre, the trochee, from which it receives
its name, is placed in all the feet, the tribrach in every foot but the seventh one.]

Mallius has abandoned the use of dipodies from his terminology. Although this is never
mentioned explicitly, the trochaic metre which he discusses is obviously the trochaic
tetrameter catalectic, judging by his mention of the seventh foot where the tribrach is not
allowed. Mallius spends some time discussing the ways in which the trochaic tetrameter
catalectic can be constructed, contrasting a type of line consisting entirely of trochees with
one where all the even feet are spondees, also pondering the advantages and drawbacks of
syllable resolution. Nowhere, however, does he give a description of the popular trochaic
septenarius.
Bedes description of the trochaic septenarius is, seen from this light, his most
radical redefinition of a Latin poetic metre and represents a complete break with previous
tradition. Although the wording of his treatise contains, as usual, some verbatim quotations
from Mallius, the content is all his own:

Metrum trochaicum tetrametrum, quod a poetis Graecis et Latinis frequentissime
ponitur, recipit locis omnibus trocheum, spondeum omnibus praeter tertium. Currit
autem alternis versiculis, ita ut prior habeat pedes quattuor, posterior pedes tres et
syllabam.
144

[The trochaic tetrameterwhich the Greek and Latin poets very frequently used,
takes the trochee in every foot, and the spondee in every foot but the third. It is
formed with alternating lines in such a way that the first has four feet and the second
three feet and an extra syllable.]
145


Bedes main deviations from previous tradition are the following:
1) The trochaic tetrameter is described as a couplet, where the first line consists
of four feet and the second one of three feet and one syllable. In other words, the central
diaeresis which almost invariably comes after the fourth foot has been recast as a line ending.

143
gramm. VI, 594, 28 595, 3.
144
DAM 23, 2-6.
145
Trans. Kendall 1991, 159.
181

That Bede calls it a tetrameter in spite of this alteration is due to Late Latin practice, where
the nomenclature of iambo-trochaic metres is based on the number of feet rather than metra.
In this context, the word tetrameter must be viewed as analogous to Bedes use of the word in
conjunction with the iambic dimeter: for Bede, the metre has two alternating lines of four feet
each, the first being acatalectic and the second catalectic (docked of its last syllable).
2) Bedes description of the trochaic tetrameter corresponds more closely with
the structure of the septenarius than with that of the classical trochaic tetrameter: Bede allows
spondees in the odd feet as well as the even ones. The exception, in Bedes presentation, is
the third foot of each line. This means that, quite in keeping with standard metrical practice,
the penultimate syllable is always short. That Bede thinks that this applies to the third foot of
the first half of the line as well is an obvious error that requires closer observation.
3) Bede precludes all types of syllable resolution from his analysis of the metre.
For him, the trochaic metre only consists of trochees and spondees and is perfectly
isosyllabic, having an invariable number of fifteen syllables.
The form of the trochaic septenarius, as Bede sees it, can be summed up as
follows: - x - x | - u - x || - x - x | - u -. As we can see, the structure matches neither that of the
classical trochaic tetrameter nor that of the septenarius. How did Bede arrive at such a
construction? As usual, we probably have to attribute the occurrence to the poem Bede uses
as an example in his treatise:

Huius exemplum totus est hymnus ille pulcherrimus:
Hymnum dicat turba fratrum,
hymnum cantus personet,
Christo regi concinentes
laudes demus debitas.
146

[An example of this is the whole of that very beautiful hymn:
Let the congregation sing a hymn,
let the music swell the hymn,
singing in chorus to Christ the King,
let us offer his merited praises.]
147


The hymn Hymnum dicat turba fratrum has been preserved in several early insular hymnals,
including the Irish Antiphonary of Bangor. In the earliest manuscripts, including those of the
Antiphonary, it has generally been attributed to St. Hilary of Poitiers, whom we know to have

146
DAM 24, 6-11.
147
Trans. Kendall 1991, 159.
182

been one of the pioneering figures of early Latin hymnody.
148
The hymn is exceptional in the
Antiphonary of Bangor because it is quite evidently of continental origin; furthermore, it is
the sole representative of quantitative verse in the Antiphonary. It has been suggested by
some critics that the hymn is, in fact, rhythmic, and several authors have noted the obvious
metrical flaws that appear in the hymn. These, however, generally belong to a type that is not
uncommon in Late Latin poetry: the prosodically faulty words, in most cases, have a stressed
syllable where the vowel has been lengthened contrary to classical practice, or where a final
syllable has been lengthened.
149
In Late Latin verse, licences of this kind are less unusual
than one might believe, and Bede, as we may recall, dedicated a whole chapter of his treatise
to the defence of such prosodic liberties, in particular in Christian hexameter poets. Similar
features also occur in the hymn Adae carnis gloriosae that we with certainty know to be by
Hilary.
150
The prosodic similarities between Hymnum dicat and Adae carnis, as Dag Norberg
suggests in his Les vers latins ambiques et trochaques, are probably due to the history of the
hymns dissemination: it was probably brought to Ireland in a continental hymnal, possibly
from Poitiers, where its anonymous composer may have been influenced by the verse of St
Hilary.
A closer examination of the hymn Hymnum dicat turba fratrum sheds some
light on Bedes ostensibly eccentric description of the metre. Firstly, the poem is completely
isosyllabic, showing hardly any signs of syllable resolution despite its sometimes eccentric
prosody; only line 30 (duodecim viros probavit, per quos vita discitur) can be interpreted as
having a double-short in its first element (uu u - u | - u - - | - - - u | - u -), but the initial
duodecim can equally well be read with fusion of the u and o, making the first element a long
syllable. In the light of what appears to be a complete absence of syllable resolution in the
poem, it must have seemed sensible to Bede to leave this prosodic option undiscussed: we
must recall that elsewhere, too, Bede limits his discussion of metrical rules to those features
that appear in the poem at hand.
As for Bedes seemingly absurd rule that the third foot of the trochaic
septenarius is always a trochee, this feature does, to some extent, seem to correspond with the
structure of Hymnum dicat turba fratrum as well, although by no means to the degree one
might suppose from Bedes presentation of the metre. A quick analysis of the poems

148
Walpole 1922, 1-3.
149
Kabell 1960, 64; Norberg 1988, 87.
150
Norberg 1988, 87: Il est tonnant quHilaire ait commis plusieurs erreurs prosodiques: nihil 6,3, 9,2, in 4,2
et en outre meus 8,2 et mihi 8,3 dans une citation de la Bible. De mme lauteur de Hymnum dicat : via (citation
de la Bible) et prolongement de la finale de discutit 18,1, de scandere 20,1, de nuntiat 28,2.
183

seventy-four lines (or, if we follow Bedes restructuring of the metre, seventy-four couplets)
shows that in the overwhelming majority of the lines the third foot is indeed a trochee. Only
thirteen lines in the whole text would appear to have a spondaic third foot, although the exact
metrical structure of some of the lines is open to interpretation, owing to the poets at times
idiosyncratic prosody. Bede acknowledges this feature and even gives two examples of third-
foot spondees from the poem but implies that they are exceptional. Following his recasting of
the trochaic septenarius as a couplet, he persistently refers to them as occurring in the third
foot of the first line:

In quo aliquando et tertio loco prioris versiculi spondeum reperies, ut:
Factor caeli, terrae factor;
et:
Verbis purgat leprae morbos.
151

[In it you will find here and there a spondee even in the third foot of the first line, as
in:
Maker of heaven, Maker of earth;
and:
He cures the disease of leprosy with a word.]
152


It remains unclear why Bede decided to formulate this tendency of the poem into a fixed
metrical rule. Admittedly, Bede had ventured similarly bold redefinitions of metrical
structures elsewhere in his treatise. It could simply be that, for obscure reasons of his own,
Bede had taken an aesthetic dislike to third-foot spondees in trochaic verse; if this is the case,
we can view it as analogous to his vehement opposition to fifth-foot spondees in hexameter
poetry.
There may also be yet another factor that contributed to Bedes presentation of
this previously nonexistent metrical rule, and this is his redefinition of the trochaic
septenarius as a couplet. Even in the archaic form of the trochaic septenarius, the penultimate
element is always short; in other words, the seventh foot of each line is invariably a trochee
regardless of any variation in the earlier portion of the line. In Bedes analysis of the trochaic
septenarius as a couplet, the seventh foot of the line has become the third foot of the second
line. Then, by analogy, Bede has extended the rule that had previously only applied to the end
of the line to the first half of the line as well. This might very well have brought about Bedes
idiosyncratic view that the third foot of each line is generally a trochee, although the third

151
DAM 23, 12-16.
152
Trans. Kendall 1991, 159.
184

foot of the first line of each couplet (i.e. the true third foot) may also be a spondee. This
observation was no doubt corroborated by Bedes own study of Hymnum dicat turba fratrum.
Bedes new definition of the trochaic septenarius seems to have exerted a
surprisingly strong influence on poets of the later generations, particularly the Carolingians.
According to Dag Norberg, several poets followed Bedes eccentric metrical rulings to the
letter. No lesser figures than Hrabanus Maurus, Walahfrid Strabo and Hincmar of Reims
composed poems in a metre that structurally resembles the archaic trochaic septenarius with
the exception that the third foot of the line is always a trochee.
153

It is also probable that Bedes reformed interpretation of the trochaic
septenarius played a role in the evolution of some forms of medieval rhythmic verse, and in
this respect it may have had even more durable results than in its erroneous ban on spondaic
third feet: Bedes perception of what had previously been a line as a couplet of alternating
line-types may ultimately have contributed to the increasing popularity of new strophes
where the halves of the original line were combined in new and creative ways. One of the
best-known examples of this is the versus tripertitus caudatus of the Stabat mater sequence,
which is based on the rhythmic, or non-quantitative, variant of the trochaic septenarius, where
the first half of the line is reduplicated (e.g. Stabat mater dolorosa / iuxta crucem lacrimosa /
dum pendebat filius). Similarly, there is a wealth of medieval poetry where only one half of
the trochaic couplet is employed, again in its non-quantitative form, as in the Dies irae
sequence where the first half of the verse (8p) is reiterated (Dies irae, dies illa / solvet
saeclum in favilla / teste David cum Sibylla). On the other hand, the latter half of the
trochaic septenarius, basically identical with the latter half of the iambic trimeter (7pp), was,
at times, used on its own, as in the following lines by Godescalc dOrbais:

Inest quibus caritas,
ipsis placet veritas.
154

[Those, in whom there is charity,
love the truth.]

Verse of this type became ever more popular during the Carolingian era. One must also bear
in mind that Bedes definition of the trochaic septenarius as an unambiguously isosyllabic
verse form made it more adaptable to rhythmic poetry.

153
Norberg 1958, 77; Meyer 1905 II, 348.
154
PAC VI, 106. See Norberg 1958, 117-118.
185

Bedes description of the trochaic septenarius represents one of the more radical
reinterpretations of metrical forms in De arte metrica. The line has become a couplet;
syllable resolution is banished from its metrical structure, and to crown it all, new restrictions
have been imposed on its construction. Although the last of these features had apparently
come about through the overzealous authors error, it too, had repercussions on later poetry.
The two first ones, on the other hand, paved the way for new poetic forms by codifying some
of the stylistic and prosodic changes to which the metre had been subjected in late antiquity.


3. 8. Conclusion

Bedes presentation of lyric metres is more modest in scope than his discussion of the
dactylic metres, which is largely due to the prosodic features of the respective metres as well
as their differing roles in the monastic curriculum. The hexameter is prosodically more
complex than the shorter lyric lengths which Bede had chosen to present; furthermore, the
influence of the pre-Christian classics was still prominent in dactylic verse, as Vergil and
Lucan belonged to the curriculum authors and could not be entirely bypassed. The lyric
metres which Bede had chosen to discuss, on the other hand, are metres which had been
employed by Christian authors and many of them played a central role in hymnody. As pre-
Christian lyric verse was largely unknown and definitely irrelevant to Bede and his
contemporaries, there was no need to discuss the respective features of the pagan and
Christian usage, when it came to lyric metres. Correspondingly, Bede presents these metres in
the form in which they occur in Christian poetry, relying more on his own observations than
those of the grammarians who largely drew on the example of Horace and other pre-Christian
authors.
This means that the lyric metres, as Bede presents them, are considerably
simpler than their pre-Christian forms. In the phalaecian metre and the sapphic strophe, the
variable syllable lengths of the opening of the line had disappeared. Therefore, Bede could
confidently present the former as beginning with a spondee and a dactyl, and the latter,
with a trochee and a spondee. In Christian iambo-trochaic verse, on the other hand, the
resolution of the long element had grown increasingly rare, and Bede only treats it as a side
issue in his presentations of the iambic metres but not at all in his discussion of the trochaic
septenarius. At the same time, Bede tries to impose additional restrictions on some of the
metres: in his chapter on the iambic dimeter, he specifically commends a type of verse where
186

all the odd feet are spondees, whereas in his chapter on the trochaic septenarius he presents
the erroneous rule that the third foot of the line must never be a spondee. This
misunderstanding is probably due to the limitations of Bedes sources, which only discuss
classical trochaic verse, ignoring the popular septenarius; in his presentation of the metre,
Bede was left to his own devices.
Bedes simplification of his metrical analyses extends to his choice of
terminology. He follows the growing trend of the grammarians to view all metres as either
dactylic or iambo-trochaic. Bede breaks all metres into metrical feet of two or three
syllables, which basically enables him to present all metres as sequences of dactyls, spondees,
trochees and iambs. The classical aeolic metres (the phalaecian metre and the sapphic verse)
have, in Bedes terminology, become dactylic pentameters, and, following the example of
Mallius Theodorus, he presents the iambic trimeter as a hexameter and the iambic dimeter
as a tetrameter, treating the foot, rather than the classical two-foot metron, as their basic
unit. The trochaic septenarius, on the other hand, remains a tetrameter but only because
Bede treats it as a couplet of two lines with four or three-and-a half feet each. Remarkably,
Bede gives hardly any guidance on style in his chapters on lyric metres. Obviously, he did not
regard them as likely to cause any real difficulty to his reader; furthermore, when it comes to
lyric verse, the pre-Christian classics did not enter the discussion, whereas Bede clearly
viewed them as problematic in his discussion of the hexameter. We must also bear in mind
Bedes background as a teacher of church music and Cuthberts as his pupil: the hymns were,
for both of them, the most familiar kind of verse, requiring far less instruction than the
hexameter, the mastery of which formed the pinnacle of poetic scholarship.
Bedes straightforward presentation of the lyric metres is, in many respects,
even more thoroughly Christianised than his discussion of the dactylic hexameter and the
elegiac couplet, and it is the first consistent effort to present the metres of Christian hymnody
in a practical form. Bedes efforts to standardise the metres he discusses played a role in the
verse of subsequent generations of poets: as Norberg has noted, Bedes misguided rulings on
the structure of the trochaic septenarius were followed by several Carolingian poets. Also, his
portrayal of most of these metres as isosyllabic, with variation of syllable lengths and syllable
resolution cut to the minimum, reflects the transition from quantitative metres to their
rhythmic counterparts, something to which Bede devoted unprecedented attention in the
twenty-fourth chapter of his treatise.
187

4. Rhythmic verse

4.1. Introduction

By almost universal consensus, De rithmo, the twenty-fourth chapter of Bedes De arte
metrica, is considered to be its most radical and historically significant portion. In his brief
chapter on rhythm, Bede, for the first time in grammatical literature, introduces rhythmic
poetry to his readers as a literary genre in its own right. The chapter is also noteworthy as it
constitutes the first arguably correct and unambiguous analysis of rhythmic poetry: in his
treatise on rhythm, Bede states explicitly that he is dealing with a form of poetry based not on
the length but the number of syllables. If Bedes work as a whole is taken to codify many of
the central features of early medieval metric poetry, in its chapter on rhythm we can see the
origins of those forms of medieval Latin poetry which are usually considered to be the most
characteristic of the Middle Ages, and on which most West European poetry in the modern
languages is also based. The importance of Bedes chapter on rhythm was fully grasped by
the works medieval readers, which is corroborated by its distribution in vernacular languages
as late as the sixteenth century, when one would have expected it to be superseded by more
advanced presentations.
1



4.2. The evolution of prosodic terminology in late antiquity

4.2.1. From mechanical to vocal ictus

In order fully to understand the context of Bedes presentation of rhythmic poetry, we must
first take a look at how the central terminology of prosody and metrics evolved during late

1
D Arco Silvio Avalle cites as an example a sixteenth-century French translation by Claude Fauchet: Il y a
apparence que les Rhythmes tiennent du Metre: pour ce que cest une harmonieuse composition de paroles, non
par mesure et certain ordre tel que celuy qui se garde en la composition des Metres au vers, ains par nombres de
syllables, selon quil plaist auz areilles. Et tells sont les Cantiques des Poetes vulgaires. De vray le Rhythme
peut ester faict par soy (parfois?) sans Metre: mais le Metre ne peut ester sans le Rhythme ou mesure. Ce que
lon peut dire plus clairement, Metre est un chant constraint par certaine raison; Rhythme est un chant libre et
non subject aucune loy. Vray est que bien souvent vous trouverez de la raison ou mesure certaine au Rhythme,
non pour ce que le compositeur sy soit assubjecti, mais pour ce que le son (ou ton , selon Victorinus) et
harmonie la paradventure conduit et men iusques caste raison. Laquelle il est de necessit que les Poetes
vulgaires suivent lourdement, et les savants sciemment. Comme lHymne qui sensuit, lequel est tresbien en
faon de vers ambiques: [Rex aeterne domine] et autres en assez bon nombre de saint Ambroise. Encore sen
chante il en faon de Trochaques, comme cestuy-ci du iour du iugement, compose par alphabet: [Apparebit
repentina]. Avalle 1992, 400-401.
188

antiquity. As the classical syllable lengths of spoken Latin began to disappear, the prosodic
nature of Latin poetry became increasingly blurred as well. This led to the gradual and
possibly inadvertent redefinition of such terms as arsis, thesis, ictus, and the very concept of
rhythm itself.
The term ictus has been a subject of some controversy among scholars, mainly
because, in its classical sense, it is an abstract and, according to some critics, inaudible
feature.
2
Quintilian describes the term unequivocally as the beating of time with the foot and
finger of the performer of a poem:

Tempora etiam animo metiuntur et pedum et digitorum ictu.
3

[People measure time even in their minds with a stroke of the feet and the fingers.]

Similarly, in his Odes Horace instructs his choir in the following manner:

Lesbium servate pedem meique pollicis ictum.
4

[Observe the Lesbian metre and the stroke of my thumb.]

In other words, classical ictus appears to be a phenomenon unrelated to dynamic stress, and
no classical author mentions the practice of stressing the metrical ictus by raising the pitch or
volume of the voice.
5
All this was to change with the decline of the quantitative prosody of
Latin. When the first syllables of such words as cano and eget, for example, were pronounced
long, as became customary in late antiquity, the metrical structure of dactylic or lyrical
metres was no longer perceptible. As the average reader could no longer hear the metrical
beat behind the metre, the role of word accent became increasingly important. Subsequently,
such metres as the dactylic hexameter could only be taught by the practice known as
scansion where the first syllables of all feet are stressed and ictus and accent became
inseparable. The term scandere versus itself seems to have its origins in the schools of the
late empire, and several authors from the third century onwards describe this practice. One of
the earliest and most explicit descriptions of this practice is by the late third-century
Sacerdos:


2
Wilkinson 1963, 92-94. Wilkinson himself prefers to speak of metrical pulse.
3
Quint. 9, 4, 51.
4
Hor. carm. 4, 6, 36. Wilkinson (1963, 92) curiously considers it unlikely that there was a conductor in
antiquity who beat time for a choir. Horace's thumb was picking at the strings of an imaginary lyre.
5
Norberg 1988, 13; Avalle 1992, 394-395.
189

Hoc tamen scire debemus, quod versus percutientes, id est scandentes, interdum
accentus alios pronuntiamus quam per singula verba ponentes: toro et pater
acutum accentum in to et pa, scandendo vero inde toro pater Aeneas in ro et ter.
6

[However, we must know this: when we are scanning verse, at times we pronounce
the accents differently than in individual words: toro and pater take the acute
accent on to and pa, but when scanning inde toro pater Aeneas, on ro and ter.]

It is clear that Sacerdoss presentation is either faulty or corrupt, as the accent can in no way
fall on the syllable ter in pater (ter is the final syllable of a dactylic foot and does not
coincide with the ictus). Nevertheless, the quotation gives us a general idea of the practice of
scanning verse. Scansion involved the segmentation of a line into units of one foot each,
which in turn were provided with an individual accent. The fourth-century Maximus
Victorinus, one of Bedes main sources, gives a description of this practice in his De ratione
metrorum, where Vergils omnia vincit amor et nos cedamus amori is segmented to omnia
(dactyl), vincita (dactyl), moret (spondee), and so forth.
7
This became a standard practice in
later grammatical literature and was used by Bede himself.
8
This practice appears to have
been applied primarily to the dactylic metres (the hexameter and the elegiac couplet) for the
obvious reason that in Roman poetry they have an extremely high rate of clash between
accent and ictus, and, indeed, some of the Roman innovations to the structure of the dactylic
metres are specifically intended to enhance this very feature.
9
Iambic and trochaic metres,
with their higher ratio of accent-ictus coincidence, could easily be recited without resorting to
this practice, as we shall soon observe.
As ictus became identified with accent, the terms arsis and thesis, originally
terms for the upbeat and downbeat of a metrical foot, became identified with stressed and
unstressed syllables.
10
We can find numerous examples of this change in the grammatical
literature of late antiquity. In the fourth century, Marius Victorinus writes:

Arsis igitur et thesis, quas Graeci dicunt, id est sublatio et positio, significat pedis
motum. Est enim arsis sublatio pedis sine sono, thesis positio pedis cum sono. Item
arsis elatio temporis, soni vocis, thesis depositio et quaedam contractio syllabarum.
11

[Therefore, the arsis and the thesis, as the Greeks call them, i. e. raising and lowering,
signify the movement of feet. Therefore arsis is the raising of the foot without sound,

6
gramm. VI, 448, 20-22.
7
gramm. VI, 219, 13-14.
8
DAM 10, 17.
9
e.g. Wilkinson 1963, 120-21; Raven 1965, 98-99.
10
Avalle 1992, 394.
11
gramm. VI, 40, 13-16; cf. Wilkinson 1963, 90: I shall avoid the terms thesis and arsis, because they
actually exchanged meanings in antiquity, and are only confusing.
190

thesis the lowering of the foot with sound. Similarly, arsis is the raising of the
syllable and the voice, the thesis a kind of lowering and contraction of the syllables.]

Marius Victorinuss rather tortuous definition of arsis and thesis is a curious melange of old
and new, and serious confusion between the concepts of rhythm, stress and quantity seems to
prevail. All this can be attributed to his reliance on earlier authors on prosody, and the
prosodic changes that Latin was undergoing at his time. All the same, we can see that such
terms as arsis, thesis and ictus, which were initially employed to describe phenomena of a
quantitative nature, had been given a radically new definition to suit the didactic purposes of
the Late Latin grammarians, although the clarity of the new terminology sometimes tends to
be undermined through contamination by older definitions.


4.2.2. The redefinition of rhythm

In antiquity, the word rhythmus had a double usage: it could have the more general meaning
of musical pulse; on the other hand, it was used as a more specific metrical term to describe
the technique and structure of lyric poetry.
12
This definition of rhythm can still be
encountered in some grammarians of late antiquity. Mallius Theodorus, as one of the last,
warns his readers:

Siqua autem apud poetas lyricos aut tragicos quispiam reppererit, in quibus certa
pedum conlocatione neglecta sola temporum ratio consideranda sit, meminerit ea,
sicut apud doctissimos quosque scriptum invenimus, non metra, sed rhythmos
appellari oportere.
13

[Should someone find something in the lyric or tragic poets, where the regular
placement of feet is neglected and only the order of the tempora is to be observed, let
him remember that, as we can find even in the writings of the most learned authors,
these are to be called rhythms, not metres.]

In this quotation, the terms metrum and rhythmus are contrasted in a way that may appear
misleading if we think in the modern terms of metric (i.e. quantitative) and rhythmic
(qualitative or non-quantitative) poetry. The author, however, makes it abundantly clear that
rhythmus is by his definition a decidedly quantitative phenomenon: sola temporum ratio
consideranda sit. In other words, the word rhythmus basically describes such quantitative

12
Meyer 1905, 140; Avalle 1992, 396.
13
gramm. VI, 585-586.
191

phenomena that cannot be analysed as consisting of regular metrical feet. This would mainly
apply to Greek lyric metres.
The contrast of metrum and rhythmus is a common enough subject in the works
of Late Latin grammarians. Marius Victorinus states:

Inter pedem autem et rhythmum hoc interest, quod pes sine rhythmo esse non potest,
rhythmus autem sine pede decurrit.
14

[Between a foot and a rhythm there is this difference: a foot cannot exist without
rhythm, but rhythm can flow even without feet.]

as does Augustine in his De musica:

Quocirca omne metrum rhythmus, non omnis rhythmus etiam metrum est.
15

[Therefore all metre is rhythm, but not all rhythm is metre.]

These late definitions of rhythm are, however, increasingly more ambiguous as to the precise
prosodic nature of the phenomenon. This, too, would appear to reflect the disappearance of
syllable quantity in spoken Latin. Like ictus, arsis and thesis, the concept of rhythm was
gradually given a new definition: it came to be used to denote the fluctuation of stressed and
unstressed syllables within lines that did not conform to the requirements of strict metrical
quantity. In his De musica, Augustine implicitly recognises the existence of non-quantitative
prosody by pleading ignorance of syllable quantity:

Iudicium aurium ad temporum momenta moderanda me posse habere non nego, quae
vero syllaba producenda vel corripienda sit, quod in auctoritate situm est, omnino
nescio.
16

[I do not deny that my ears have the power to judge the measurement of time, but I
do not in the least know which syllable is to be lengthened or shortened, which is
determined by the authors.]

We find the same contrast of metrica ratio with iudicium aurium in the Ars Palaemonis de
metrica institutione by Maximus Victorinus, one of Bedes sources:


14
gramm. VI, 44.
15
Aug. mus. 3, 2.
16
Aug. mus. 3, 5. Cicero uses the term iudicium aurium and its equivalents in conjunction with quantitative
prosody, cf. Cic. or. 173, iudicium ipsa natura in auribus nostris conlocavit [Nature herself has placed the
power of judgment in our ears] and 177, aures enim vel animus aurium nuntio naturalem in se continet vocum
omnium mensionem. (For the ears, or the mind through the medium of the ears, contain in themselves the
natural ability to measure all sounds.)
192

Metro quid videtur esse consimile? Rhythmus. Rhythmus quid est? Verborum
modulata compositio non metrica ratione, sed numerosa scansione ad iudicium
aurium examinata, ut puta veluti sunt cantica poetarum vulgarium.
17

[What is there that seems to resemble metre? Rhythm. What is rhythm? The
harmonious composition of words, not by the means of a metrical system, but by a
rhythmic scansion, judged in accordance with the way they sound to the ear, as are,
say, the songs of common poets.]

This definition of rhythm is already remarkably close to Bede's own. We shall presently
observe the subtle alterations that Bede made to this description in adapting it to his own
purposes.


4.3. The origins of rhythmic poetry

There have been three theories about the origin of medieval rhythmic poetry:
18

1) The romantic view that rhythmic poetry was of either of Celtic, Germanic or Romance
origin, propounded among others by K. Vossler in Die Dichtungsformen der Romanen
(Stuttgart 1951, 30) and P. Verrier in Le vers franais (Paris 1931, vol. 1, 19).
2) The view that rhythmic poetry is a phenomenon of primarily Hebraic origin, as expressed
by W. Meyer in Gesammelte Abhandlungen zur mittellateinischen Rythmik (Berlin 1905-
1936).
3) The modern view that medieval rhythmic poetry is a descendant of classical metres
adapted to a different prosodic system.
19

Although the prevailing opinion is that rhythmic poetry has evolved from classical metric
poetry, there is some evidence of Semitic influence in certain rhythmic poems of late
antiquity: the verse pamphlets Psalmus contra partem Donati by Augustine and Psalmus
contra Vandalos Arrianos by his follower Fulgentius. These are non-quantitative, isosyllabic
and roughly trochaic poems with a primitive rhyme and are presumably based on Semitic
models.
20
Norberg, however, does not consider them to have had any influence on subsequent
Latin poetry.
21


17
gramm. VI, 206.
18
Avalle 1992, 391.
19
Cf . Norberg 1958, passim.
20
Norberg 1988, 131.
21
Norberg 1988, 132: Pour l' volution du vers latin rythmique ces psaumes ne doivent avoir eu aucune
influence.
193

The current opinion is that most of the rhythmic poetry of the middle ages is
based on non-quantitative forms of classical iambic and trochaic metres. Neither the dactylic
nor the lyric metres were ever extensively used as models for rhythmic verse. There are a
number of reasons that precluded the extensive use of, say, the hexameter in rhythmic form.
First of all, the hexameter was considered too dignified for such use, and although
Commodian had composed a number of non-quantitative hexameters in the fourth century, he
had found no followers. As there was an existing corpus of strictly quantitative Christian
hexameter poetry based on classical models (e.g. the works of Juvencus, Arator, Sedulius,
Prudentius and others), Christian poets were not inclined to break away from this tradition.
The structure of the hexameter is also ill suited to a rhythmic reinterpretation.
Compared with the iambo-trochaic metres, the hexameter has a very low rate of ictus-accent
coincidence and the resolution of long elements plays an important role in the metre. The
iambic and trochaic metres, on the other hand, can be composed in a practically isosyllabic
form with no resolution of the long elements and, the trochaic tetrameter in particular, with a
very high rate of accent-ictus coincidence.
As Norberg states in his Les vers iambiques et trochaques au Moyen Age et
leurs rpliques rythmiques, the rhythmic variants of iambo-trochaic verse did not evolve
simply by substituting word-accent for ictus. Rather, the priests and monks who had not
received a classical training tried to imitate the rhythms they heard in verse when it was read
with the regular prose accents.
22
The process of transforming metric to rhythmic, or
quantitative to qualitative, was facilitated by the increasing prosodic regularisation of iambic
and trochaic verse in late antiquity. If we take as an example the iambic dimeter which was
one of the favourite metres of early hymnody, we can observe four possible accent patterns
when the lines are read with prose accents. Norberg demonstrated these four patterns with the
example of the following lines:
1. aetrne rrum cnditor
2. splndor patrnae glriae
3. slvit plum calgine
4. si rspicis lpsus cdunt

22
Norberg 1988, 2 and 13-16. The author claims that this proves that, even in late antiquity, the reading
technique known as scansion was not as prevalent as is generally supposed. On the other hand, it could simply
mean that scansion was primarily applied to dactylic verse and that iambo-trochaic verse was commonly read
with the usual prose accents.
194

The fourth pattern where the penultimate syllable is accented was well on its way out already
in antiquity, as it does not harmonise well with the others.
23
The elimination of the fourth
pattern led to an easily identifiable verse with a regular number of syllables and a regular
rising, or proparoxytone, cadence, with some accentual variation in the first half of the verse.
The iambic trimeter and the trochaic tetrameter catalectic underwent a similar evolution, and
as a result we have a system of verse that is identifiable by the number of syllables per line
and a regular cadence, the types of cadence being rising, or proparoxytone, and falling, or
paroxytone.
24
The accent patterns of such lines were subsequently copied by (initially
uneducated) writers without adherence to syllable length.
This type of non-quantitative poetry won particularly great popularity in the
British Isles and was presumably the only form of Latin verse created by the earliest Irish
scholars, who are generally presumed to have had no knowledge of syllable quantity,
25
and
even Aldhelm, the first major insular author of quantitative poetry, contributed to this genre
with his Carmen rhythmicum.
26
Aldhelms Carmen has been composed in a verse form
commonly termed the continuous octosyllable, a highly alliterative verse form with a regular
proparoxytone ending and primitive end-rhyme. This verse form, also cultivated by the
eighth-century Boniface, is presumably based on seventh-century Hiberno-Latin models but
is probably ultimately derived from the rhythmic forms of Classical metres, mainly a non-
strophic form of the iambic dimeter.
27

Some insular poets went so far as even to jettison the regularly accented verse
ending. This resulted in a form of verse that was solely based on the number of syllables per
line, a solution that may be practical, or understandable, in texts that are intended for singing.
The hymns attributed to the sixth-century Columba, although based on hymns in the iambic
dimeter by Ambrose and Sedulius, only resemble their models in the number of syllables, as
do most of the hymns in the late seventh-century Antiphonary of Bangor.
28





23
Norberg 1988, 2.
24
ibid.
25
Roger 1905, 267-268; Meyer III 1936, 303-346; Norberg 1988, 42-43; Lapidge 1979, 210. The earliest Irish
hexameters are presumably based on Anglo-Latin or Carolingian models; cf. Lapidge 1999, 273.
26
Orchard 1994, 19-72.
27
ibid.; Orchard 1999, 339.
28
Norberg 1988, 43-45.
195

4.4. Bede's definition of rhythm

Bede's main sources in his short chapter on rhythmic poetry are De Scauri et Palladii libris
excerpta per interrogationem et responsionem by Audax and Ars Palaemonis de metrica
institutione by Marius Victorinus, whose descriptions of rhythm are virtually identical.
29
At
first glance, this would seem to apply to Bede as well, whose description of rhythm is taken
almost verbatim from the Ars Palaemonis. Marius Victorinus writes:

Metro quid videtur esse consimile? Rhythmus. Rhythmus quid est? Verborum
modulata compositio non metrica ratione, sed numerosa scansione ad iudicium
aurium examinata, ut puta veluti sunt cantica poetarum vulgarium. Rhythmus ergo in
metro non est? Potest esse. Quid ergo distat a metro? Quod rhythmus per se sine
metro esse potest, metrum sine rhythmo non esse potest. Quod liquidius ita definitur,
metrum est ratio cum modulatione, rhythmus sine ratione metrica modulatio.
Plerumque tamen casu quodam invenies rationem metricam in rhythmo, non artificii
observatione servata, sed sono et ipsa modulatione ducente.
30

[What is there that seems to resemble metre? Rhythm. What is rhythm? The
harmonious composition of words, not by the means of a metrical system, but by a
rhythmic scansion, judged in accordance with the way they sound to the ear, as are,
say, the songs of common poets. Is there then no rhythm in metre? There can be.
How, then, does it differ from metre? The difference is that rhythm can exist on its
own without metre but metre cannot exist without rhythm. This can be defined more
clearly as follows: metrical verse is a (quantitative) system with a regular beat,
whereas rhythm is a regular beat without a metrical system. However, you can often,
by chance, find measured quantities even in rhythm, not because the regular artistic
arrangement has been preserved, but from the influence of the sound and the beat
itself.]
31



Bede paraphrases:

Videtur autem rithmus metris esse consimilis, quae est verborum modulata
conpositio, non metrica ratione, sed numero syllabarum ad iudicium aurium
examinata, ut sunt carmina vulgarium poetarum. Et quidem rithmus per se sine metro
esse potest, metrum vero sine rithmo esse non potest. Quod liquidius ita definitur:
metrum est ratio cum modulatione, rithmus modulatio sine ratione. Plerumque tamen
casu quodam invenies etiam rationem in rithmo, non artificii moderatione servata,

29
Palmer 1959, 574: It is virtually impossible to know whether Bede used Audax exclusively or turned at times
to the combination to be found in the Ars Victorini grammatici - Ars Palaemonis. Victorinus-Palaemon and
Audax are so similar in wording that it is obvious both rest on the same sourceSince the sections which
radically distinguish the two are not used by Bede, the problem will probably never be solved. Kendall in CC
123A, p.138, attributes the quotation in Bede' s definition of rhythm to both sources.
30
gramm. VI, 206-207.
31
My translation paraphrases Kendalls translation of Bedes De arte metrica, 1991, 161.
196

sed sono et ipsa modulatione ducente, quem vulgares poetae necesse est rustice, docti
faciant docte.
32

[Rhythmic verse appears similar to metrical verse. Rhythmic verse is the harmonious
composition of words, not by the means of a metrical system, but by a rhythmic
scansion, judged in accordance with the way they sound to the ear, as are, say, the
verses of common poets. Rhythm can certainly exist by itself without metre, but
metre cannot exist without rhythm. This can be defined more clearly as follows:
metrical verse is a quantitative system with a rhythmical beat, while rhythmic verse
has a rhythmical beat without a quantitative system. However, you can often by
chance find measured quantities even in rhythm, not because the regular artistic
arrangement has been preserved, but from the influence of the sound and the beat
itself. The common poets inevitably do this awkwardly, and the learned poets
skilfully.]
33


Superficially, the most noticeable difference is that Bede has dispensed with his predecessor's
stilted dialogue form. Most of Victorinuss turns of phrase are still there: modulata
compositio, iudicium aurium and the reference to common poets. Where Bede departs from
Marius Victorinuss description is in supplanting his ambiguous numerosa scansio
34

(rhythmic scansion or regular beat) with numerus syllabarum. In other words, Bede
unequivocally states the isosyllabic nature of rhythmic verse: it is a form of poetry based on
the number of syllables rather than merely poetry composed without regard to syllable
quantity. Apparently, Bede has taken advantage of the double meaning of numerus as either
rhythmical beat or number and, by substituting the latter for the former transformed the
older, vaguer, definition of rhythmical poetry into a newer and less ambiguous one.
35
The
observation of the number of syllables being a deciding factor in rhythmic verse already
occurs in the Ars grammatica of Diomedes:

Rhythmus est versus imago modulata servans numerum syllabarum.
36

[Rhythm is a rhythmic image of a verse, which preserves its number of syllables.]


32
DAM 24, 10-19.
33
My translation paraphrases Kendalls translation of Bedes De arte metrica, 1991, 161.
34
The term numerosus occurs in classical Latin as a prosodic term but, like rhythmus, without any contrast to
quantitative metrica ratio; cf. Cic. Brut. 8, 43 on prose clausulae: Ipsa enim natura circumscriptione quadam
verborum comprendit concluditque sententiam, quae, cum aptis constricta verbis est, cadit etiam plerumque
numerose. (Nature herself limits the number of words that can be employed in a sentence; when these words
are aptly chosen and well organised, the sentence usually falls rhythmically.)
35
J ulian of Toledo employs an expression that is ostensibly very similar: verborum modulata compositio non
metrica ratione sed numero ad iudicium aurium examinata. That Julian simply has numero, rather than numero
syllabarum, would imply that he is using the term numerus in the same sense as Victorinuss numerosa scansio:
to denote a regular beat. Amusingly enough, Maestre Yeness 1973 edition of J ulians Ars has numero
syllabarum with the explanation addidi ex Beda. - Maestre Yenes 1973, 222-223.
36
gramm. I, 474.
197

Bedes skillfully eclectic use of his sources is demonstrated by the way in which he has
managed to transplant this observation into Mariuss description of rhythm, which, in
content, is considerably different. Bedes insight may also owe something to the fact that, as
an Anglo-Saxon, he was exceptionally observant in some matters of prosody. As Cuthbert
testifies in his De obitu Bedae,
37
Bede appears to have been well versed in Anglo-Saxon
poetry, which, as we all know, has a practice of versification entirely different from Latin
rhythmic poetry. This, in turn, may have made him more aware that the principle of having a
fixed number of syllables per line constitutes a rule or restriction in much the same way as
the principle of having a fixed order of long and short syllables. All this is something that
may not have occurred to his predecessors who had a more exclusively Latin background and
a narrower frame of reference.
This, however, is not the only way in which Bede departs from his most
immediate source. What is striking is Bedes vindication of this poetic from in his phrase
quem vulgares poetae necesse est rustice, docti faciant docte. Bede, in other words, implies
that rhythmic poetry is a perfectly legitimate literary form capable of a high degree of
sophistication; that it is not simply a vehicle for vulgar or popular expression but can be
composed by educated people in an educated manner. We may comfortably say that Bede is
revolutionary not only in giving the first correct analysis of the structure of rhythmic verse
but also in recognising its legitimacy as a literary genre. This is particularly remarkable, as
some of his Christian predecessors, like Julian of Toledo, were openly hostile to all forms of
non-quantitative verse.
38

Bedes other sources include De metris by the fourth-century Mallius
Theodorus, whose definition of rhythm was quoted in chapter 4.2.2. Although Mallius is one
of Bedes favourite authors and Bede is heavily indebted to him for his chapters on the lyric
and iambo-trochaic metres, the two authors take a completely different stance on the question
of rhythm. We may recall that Mallius still adheres closely to the classical definition of

37
Published in the edition of Bede' s Historia ecclesiastica by Colgrave and Mynors (Oxford 1969), 579-587.
38
In a letter to his friend bishop Modoenus, J ulian of Toledo warns him against the use of this type of verse:
Tua aetas gravisaut fortes prosa exequatur sententias aut metricis dictis pandat iure camenas et rithmis uti,
quod plebegis est solitum, ex toto refugiat: (Being a man of venerable age, you should express your bold
thoughts in prose or compose poems in metrical form, as is right and proper, but totally shun the use of rhythms,
which is typical of commoners.) Bischoff 1966, 288-298. Norberg believes this corroborates the existence
of a living tradition of rhythmic versification: Mais Julien n' aurait pas eu raison d' avertir son collgue contre
la versification rythmique si celle-ci n' avait pas exist. Norberg 1988, 100. In his Ars grammatica, J ulian
gives an example of rhythmic poetry by quoting the unattributed line lupus dum ambularet viam, incontravit
asinum (Maestre Yenes 1973, 222-223). It is open to discussion whether the line is a prosodically faulty
trochaic septenarius (lpus, vam, sinum) or its rhythmic equivalent. In other respects, J ulians description of
rhythm is virtually identical with that of Audax and Maximus Victorinus, and, whatever the case may be, Bedes
attitude towards rhythmic poetry is the diametric opposite of Julian's.
198

rhythm as a quantitative element of prosody. It seems surprising that, in his chapter on
rhythm, Bede should have casually ignored this source and redefined rhythm in a way that is
no way reconcilable with Mallius description. It is all the more surprising if we bear in mind
that Mallius is one of Bedes main sources and that Bedes definitions of the iambo-trochaic
and lyric metres rely heavily on his. The answer to this apparent riddle can be found in the
seventeenth and eighteenth chapters of De arte metrica.
Rhythm, in its classical meaning, was frequently used to describe the structure
of those lyric metres that could not be analysed as consisting of regular metra or metrical feet
(note the words certa pedum conlocatione neglecta sola temporum ratio consideranda sit).
The description of such lyric metres in Bedes De arte metrica has been pruned down to the
bare minimum: he only discusses the phalaecian hendecasyllable and the minor sapphic
stanza. Furthermore, he analyses both the phalaecian and sapphic metres as consisting of
metrical feet: the phalaecian metre consists of a spondee, a dactyl and three trochees whereas
the sapphic stanza consists of a trochee, a spondee, a dactyl and two trochees, to which is
added a half of a hexameter line. For Bede, these metres are simply another form of
dactylic poetry, analogous to the hexameter and the elegiac couplet, and they can be analysed
accordingly by being segmented into feet. By his definition, all quantitative poetry consists of
metrical feet. Hence, the classical definition of rhythm as quoted by Mallius Theodorus
(certa pedum conlocatione neglecta) had, for Bede, become redundant and devoid of
meaning. Accordingly, the term rhythmus could be used to describe a phenomenon of greater
importance for Bede and his contemporaries.
Who, then, are the vulgares poetae, to whom both Victorinus and Bede allude?
It would be tempting to suppose that Victorinus was thinking of the verse pamphlets of
Augustine and Fulgentius, but as we have stated previously, these works are an isolated
phenomenon that inspired no followers. Furthermore, there is no evidence that Bede knew
either of these poems; Augustines anti-Donatist writings in general appear to have been
largely unknown in Anglo-Saxon England.
39
In his 1991 commentary to De arte metrica, C.
B. Kendall ignores the fact that the term vulgares poetae already appears in Victorinus and
has been merely borrowed by Bede as he states: Tempting as it is to assume that Bede is
here referring to poets writing in their native Anglo-Saxon tongue, it is more likely that he
has in mind composers of crudely isosyllabic, non-quantitative Latin poems like many of

39
Laistner 1957, 143.
199

those in the Antiphonary of Bangor and the Irish Liber Hymnorum.
40
Of course, Victorinuss
term may have reminded Bede of these very works.
We can interpret the term vulgaris in this context as something that refers to the
audience of the poems rather than their composers. If this is the case, we can take vulgaris to
mean popular, populist or accessible to the people, as in the vulgaris sermo, or common
language, for the use of which Jerome felt inclined to apologise. Of course, in a country
inhabited by speakers of a Germanic language, rhythmic poetry written in Latin was no
longer a form of peoples poetry any more than other forms of Latin literature were, and, in
this respect, we can regard Bedes use of Victorinus term vulgaris as rather outdated. Latin
rhythmic poetry was common in Bedes age only in the sense that it did not, as a rule,
belong to the official curriculum, or set books of the monastic schools and that it did not
require the long and tedious apprenticeship necessary to, say, an aspiring hexameter poet.
Bede illustrates his description of rhythmic verse with two poems:

Quomodo ad instar iambici metri pulcherrime factus est hymnus ille praeclarus:
Rex aeterne domine,
rerum creator omnium,
qui eras ante saecula
semper cum Patre Filius;
et alii Ambrosiani non pauci. Item ad formam metri trochaici canunt hymnum de die
iudicii per alfabetum:
Apparebit repentina
Dies magna Domini,
Fur obscura velut nocte
Improvisos occupans.
41

[In this way was very beautifully composed in the likeness of iambic metre that
celebrated hymn:
Eternal Lord King,
creator of all things,
you who were before the worlds
eternally the son with the father;
and quite a few other hymns by Ambrose. Similarly, they sing a hymn on Judgement
day an abecedary in the form or trochaic metre:
The great day of the Lord
will suddenly appear,
like a thief in the dark of night
seizing those not expecting it.]
42



40
Kendall 1991, 161.
41
DAM 24, 19-31.
42
Trans. Kendall 1991, 163.
200

The anonymous poems quoted by Bede had been well known in Anglo-Saxon England, and
they have been published in Walpoles Early Latin Hymns.
43
Bede has observed, quite
correctly, that iambic or trochaic verse need not be quantitative in order to be recognisable.
This idea may well have been inspired by Diomedess statement that rhythmic poetry is
versus imago modulata, a phenomenon that echoes metrical poetry without being quite the
same thing.
44
And indeed, the poems quoted by Bede have been written in the rhythmic
equivalents of two quantitative metres presented elsewhere in his De arte metrica: the iambic
dimeter and the trochaic septenarius.
Bedes first poem, Rex aeterne Domine, has been documented as early as the
sixth century. It is a hymn of indubitably continental origin and appears for the first time in
the Regula virginum of Caesarius and Aurelianus of Arles, where it is prescribed to be sung
at the first nocturns of Sunday (die dominica ad primos nocturnos) in lieu of the more
common Mediae noctis tempus est, which was used on the other nights of the week.
45
The
hymn originally consisted of sixteen stanzas of four lines each, quite like a quantitative
Ambrosian hymn. Originally the hymn, in its entirety, was a portrayal of the whole
redeeming work of Christ, but from the Xth century onwards the first seven stanzaswere
taken by themselves to form an Easter hymn, for which purpose they are not especially
suited.
46
The hymn was brought to England as part of the Old Hymnal which contained
hymns for different occasions of the ecclesiastical year.
47
Some of its hymns have been
composed in the quantitative iambic dimeter, but some are in their rhythmical counterpart.
The hymn Rex aeterne domine is the most notable representative of the latter group.
The prosodic structure of Rex aeterne Domine emulates the iambic dimeter,
which had become the favourite metre of Christian hymnody, mainly through the influence of
Ambrose and Sedulius. The metre, in its quantitative form, is described in the twenty-first
chapter of De arte metrica. We may recall that the description of the metre is simple and
sparse, but it is evident that we are here dealing with a strictly quantitative form of poetry.
Quoting Mallius Theodorus, Bede also acknowledged the possibility of resolution of the long
element into a double-short, which is regarded as one of the central features of metric
poetry.
48
Bede illustrated this phenomenon with a quotation from Ambrose (hymn. 6, 19:

43
Walpole 1922, 211-217; 381-384. The first hymn (Rex aeterne Domine) also appears in W. Bulst 1956, VI 2,
92; the second one (Apparebit repentina) in MGH, Poet. Lat. IV, pp. 507-510.
44
gramm. I, 474.
45
Walpole 1922, 212.
46
ibid.
47
Gneuss 1968, 24.
48
DAM 21, 19-24:
201

geminae gigans substantiae). This line has quite obviously not been written numero
syllabarum, as the very concept of syllable resolution means that the number of syllables in
each line is variable.


As Bede has defined a fixed number of syllables as a central feature of rhythmic
verse, the distinction between the metrical form of the iambic dimeter (with possible syllable
resolution) and its rhythmic equivalent is obvious. At the same time he is obviously aware of
the origin of the rhythmic poem he quotes: it is composed ad instar iambici metri, in the
likeness of iambic metre. We must also note his words ut sunt Ambrosiani non pauci.
Although it is quite possible that Bede is here falsely attributing the poem to Ambrose, as we
know that he has done with other poems (most notably the anonymous hymns Squalent arva
soli and Obduxere polum in the terentianean metre at DAM 19), it is plain that Bede has
understood the similarity of the metre to that of the Ambrosian iambic dimeter correctly,
while at the same time remaining aware of their prosodic differences. In other words, he
appears to have analysed the metrical pedigree of the verse correctly.
If we undertake a closer prosodic analysis of the hymn as quoted by Bede, we
shall be able to observe the differences between the rhythmic and metric forms of the iambic
dimeter. Superficially, the most striking feature of the hymn is that the first line, Rex aeterne
Domine, only has seven syllables. Some critics have analysed the line as corrupt and have
tried to emend it by changing it to O Rex aeterne Domine.
49
Norberg, however, points out
that this may not be necessary. His analyses of early rhythmical variants of the iambic
dimeter show several similarly faulty lines that appear to be one syllable short. These lines
are invariably the odd lines of the hymns. Norberg considers this a sign that the rhythmic
form of the iambic dimeter may have consisted of units of two lines and that the first syllable
of the first line of the couplet could be omitted at will. This makes perfect sense if we
remember that these hymns have been sung: the weak first syllable constituted a sort of
upbeat to the first accentuated note of the melody.
50
Similarly, especially Irish poets were
in the habit of adding such an upbeat to the rhythmic variant of the trochaic tetrameter
catalectic, thereby creating a hypercatalectic line.
51
These liberties were common in early

49
e.g. Avalle 1992, 418: Il primo ipometro (manca di una sillaba) e gli editori in genere suppliscono una O
prima di Rex.
50
Norberg 1988, 39: Le rythme accentuel tant rgulier dans environ 50% de cason pouvait omettre la
premire syllabe du vers aussi bien que la premire note de la mlodie comme une sorte de msure dattaque. It
is noteworthy that some sources give the third line of the poem qui eras ante saecula in the form qui es ante
saecula, where a similar omission seems to have taken place. Cf. Bulst 1956, VI 92.
51
Norberg 1988, 98.
202

medieval hymnody and form a rare exception to the rule of having a fixed number of
syllables in each line of rhythmic verse.
Let us ignore the first foot of the first line for a while and analyse the remaining
fifteen feet of the example. Bede maintains that in the metrical form of the iambic dimeter, all
feet can take an iamb and only the odd feet can take a spondee. A quick analysis of the four
lines quoted fromRex aeterne Domine shows us that five of the remaining fifteen feet are
flawed if seen in the context of metrical poetry: in the first line, Rex aeterne domine, the
second and third foot are unmetrical, the second being a spondee and the third a pyrrhic. In
the third line, qui eras ante saecula, the first foot is a trochee and the second one a
spondee,
52
and in the fourth line, semper cum patre filius, the second foot is a spondee.
These metrical flaws are, of course, compounded by the omission of the first syllable in the
first line, a practice alien to the iambic dimeter in its quantitative form. At the same time, the
use of hiatus is conspicuous: the cited portion of the poem has qui eras, but the rest of the
hymn at times virtually abounds in them (the second strophe, for example, has mundi in
primordio, plasmasti hominem and tuae imaginis in consecutive lines).
The other example of rhythmic poetry quoted by Bede, Apparebit repentina, is
an abecedary hymn on Judgement day. It consists of strophes of four lines which begin with
the consecutive letters of the alphabet. Each strophe is followed by the refrain in tremendo
die iudicii, which does not appear in Walpoles edition. Bedes citation of the hymns first
strophe in his De arte metrica is the first documented appearance of the hymn, which,
according to Walpole, is certainly as old as the VIIth cent., possibly much older.
53
The
author of the hymn is unknown: it has, at times, been ascribed to Hilary, but it may well be of
insular origin. Lapidge has tentatively put forward that it may have been composed by Bede
himself, as the poem shares many features with Bedes own (metrical) poem on Judgement
day,
54
as well as the vernacular Bedes death song which he is alleged to have sung in his
native tongue on his deathbed and which is quoted in Cuthberts Epistula de obitu Bedae.
55

We must also consider the way in which Bede presents the poem: although he characterises
Rex aeterne domine as pulcherrime factus, he refrains from any such praise in his
presentation of Apparebit repentina: if the poem were, indeed, his own, this would be
understandable modesty.

52
If we, however, choose the reading qui es ante saecula, where the first syllable has been left out in the same
way as in the first line of the poem, all the remaining feet are iambs.
53
Walpole 1922, 380.
54
Ed. Fraipoint 1955, 439-44.
55
Lapidge 1996, 328; Bedes death song in Colgrave and Mynors 1969, 580.
203

Structurally, Apparebit repentina has been composed in what is basically the
rhythmic equivalent of the trochaic tetrameter or septenarius, where the central diaeresis of
the classical original has evolved into a break between two lines of alternating lengths.
56
The
quantitative version of the metre is described in Bedes twenty-fourth chapter (De metro
trochaico tetrametro), and we may recall that, even in its quantitative form, the metre is
rendered in a form that is considerably less classical than those of the other quantitative
metres. The most crucial departure from classical tradition is the splitting of the original line
into a couplet, but we also noted the apparently confused, or faulty, definition of the metres
prosodic structure (Bede precludes spondees in the third foot of the line, although he shortly
thereafter recants and admits them as an afterthought, probably after studying the hymn he
uses as his illustration). As we noted, Bedes description of the trochaic septenarius also
differs from his portrayal of the iambic metres in one crucial respect: the possibility of
syllable resolution in the long element is not mentioned at all in his definition. Bede portrays
the septenarius, even in its metrical form, as an essentially isosyllabic verse form that is
remarkably close to its rhythmic variant. This, however, is not to say that Bede views them as
identical.
If we perform a metrical analysis of Apparebit repentina based on Bedes
interpretation of the trochaic septenarius as a distich of two trochaic lines (one acatalectic and
one catalectic: -x -x -x -x / -x -x -u x) where all but the third foot of the second line can be
either trochees or spondees (for the present, we may forego Bedes additional rule that the
third foot of each line must be a trochee), we can conclude that from the viewpoint of
quantitative metre, the quoted four lines contain four metrically faulty feet. In the first line,
apparebit repentina, the third foot is an iamb. In the second line, dies magna Domini, the
first foot is an iamb and the third foot a pyrrhic. In the third line, fur obscura velut nocte,
the third foot is an iamb, although, admittedly, velut does occasionally appear with a long e
even in Late Latin metrical poetry.
Such metrical lapses are common enough even in the quantitative poetry of Late
Antiquity and the early Middle Ages, but out of respect for his anonymous authors, Bede has
preferred to present the poems as examples of good rhythmic poetry rather than bad metric
poetry. He does not claim that the first poem, Rex aeterne Domine, has been composed in an
iambic metre but rather ad instar iambici metri, in the likeness of iambic metre. Similarly, his
other example, Apparebit repentina, has been written ad formam trochaici metri, in the form

56
Coronati 1981-1982, 53-62.
204

of trochaic metre. Bede has quite correctly observed that the alternation of long and short in
metrical verse can find its equivalent in the alternation of stressed and unstressed in rhythmic
verse, making it possible to transform quantitative structures into rhythmic ones.
Bedes distinction between quantitative verse and its non-quantitative
counterpart appears, at least at times, to have gone unheeded. Under Urban VIII, the hymn
Rex aeterne domine underwent fundamental revision in the hands of humanist correctors who
changed a fundamentally non-quantitative hymn into a quasi-classical one in quantitative
iambic dimeters. The resulting hymn, known by its opening line Rex sempiterne caelitum,
became the Roman Breviary hymn for Matins of Sundays and weekdays during Paschal
Time.
57
This revision has come under heavy attack from later-day hymnologists who have
viewed the revision as needless and misguided, but, undoubtedly, the close affinity of the
hymn with its earlier quantitative counterparts must have appeared particularly inviting to
prosodic purists of the renaissance and the modern era. Walpole himself, in his Early Latin
Hymns, quotes Bedes description of the hymn in De arte metrica but appears completely
oblivious of what Bede has meant: The first thing that strikes us in this ancient hymn is the
number of lines defective in metre. In view of them Bedes words quomodo ad instar
iambici metri pulcherrime factus est hymnus ille praeclarus rex Aeterne Domine seem
somewhat strange.
58
Walpole also undertakes several desperate attempts to emend
individual lines to make them conform to the quantitative model of the iambic dimeter,
having failed to grasp Bedes meaning. Although Bede elsewhere in his treatise sometimes
comes across as pigheadedly dogmatic (we may recall his inflexible views on spondaic
hexameter lines), on the question of rhythmic poetry he appears to have been both more
insightful and more radical than many of his later-day colleagues.


4.5. Conclusion

Although superficially Bede seems indebted to previous authors in his description of
rhythmic poetry, he is in many respects revolutionary. First of all, he has, for the first time,
correctly analysed rhythmic poetry as a verse form based on a fixed number of syllables.
Secondly, he states that rhythmic poetry can be written docte and pulcherrime, in an educated
and aesthetically pleasing way; that it, in fact, is a legitimate form of poetic expression rather

57
Henry 1912.
58
Walpole 1922, 210-211.
205

than merely a vulgar version of real poetry. Thirdly, he has observed that quantitative
metrical structures can find their equivalents in non-quantitative verse. This is a realisation
that can be considered essential to the transmission of the Graeco-Roman poetic heritage to
the European poets of the middle ges and beyond.
How is Bedes apparently unprecedented interest in rhythmic poetry to be
explained? The most obvious reason, of course, is the extreme vitality of this literary idiom
and its widespread use among the scholars of Anglo-Saxon England. Rhythmic poetry was
also an exclusively Christian form of poetry, whereas hexameters and elegiac couplets had
been composed by such pagan personalities as Vergil and Lucan, and for all the efforts of the
Christian poets and Bede himself, the pre-Christian classics still largely constituted the
standard for dactylic verse.
The authority of the Early Fathers would also have played a role in Bedes
opinion on rhythmic poetry. As we noted previously, Augustine, in his De musica, had
implicitly sanctioned the idea of prosody without quantity in opposing iudicium aurium with
metrica ratio and the auctoritas of poets and grammarians. We must also consider the
ambiguous evidence of the phrase et alii Ambrosiani non pauci, which Bede uses in
conjunction with the poem Rex aeterne Domine. If Bede did not use the term Ambrosiani in
a generic sense but genuinely believed the hymn to have been composed by Ambrose, this
would definitely have increased his respect for rhythmic poetry.
Bede may also have been inspired by Jeromes preface to the book of Job. The
preface itself is one of the most famous examples of the Early Fathers efforts to present the
Bible as the fountainhead of all learning, metrics and rhetoric included.
59
In his preface,
Jerome claims that the book of Job has mainly been composed in hexameters but that it also
contains other prosodic structures, including a mysterious entity called rithmus ipse.
60


Exametri versus sunt, dactilo spondeoque currentes et propter linguae idioma crebro
recipientes et alios pedes non earundem syllabarum, sed eorundem temporum.
Interdum quoque rithmus ipse dulcis et tinnulus fertur numeris lege solutis, quod
metrici magis quam simplex lector intellegunt.
[There are hexameter verses running in dactyls and spondees, and, owing to the
idiom of the language, often other feet as well, not with the same syllables, but the
same quantities. Sometimes, also, the rhythm itself is carried, sweetly and
melodiously, by numbers untied to metre, which is more understandable to scholars
of prosody than to the ordinary reader.]


59
J ones 1975, ix.
60
Hier. praef. Vulg. Iob.
206


J erome is presumably using the word rithmus in its older, classical sense, so as to denote
quantities that are not arranged into a scheme of metrical feet. J eromes expression numeris
lege solutis is, in fact, a direct loan from Horaces ode in praise of Pindar (carm. 4, 2, 9-12):

Laurea donandus Apollinari,
seu per audacis nova dithyrambos
verba devolvit numerisque fertur
lege solutis.
[He merits the award of Apollos laurels,
whether he rolls new words through his bold dithyrambs,
or is carried by rhythms unbound by rules.]
61


Jeromes classical allusion was probably lost on Bede, deprived as he was of first-hand
access to Horace. However, it is not impossible to think that Bede may have construed
J eromes ambiguous passage as a description of something similar to the rhythmic poetry that
he knew, lending it a previously unthought-of legitimacy and a lineage that went far beyond
the bounds of classical antiquity.

61
What Jerome precisely meant by his description of the prosodic structure of the Book of J ob is unclear and
beyond the scope of this presentation, but the subject has been discussed by Baroway 1950, esp. at 119-120.
Confusingly, grammarians frequently used the expression lege solutus as a way of describing prose, cf. Isid.
orig. 1, 38, 1: Prosa est producta oratio et a lege metri soluta. (Prose is a long utterance free from metrical
rules.) It may also be noted that Horaces use of the same phrase was eagerly snatched by eighteenth- and
nineteenth-century champions of vers libre such as Goethe and Hlderlin as a vindication of their art form.
207

5. Summary

Bedes De arte metrica continues the trend of Christianisation of grammatical literature that
began in late antiquity. On the surface, the Christian features of the treatise closely resemble
earlier attempts in this direction: we can witness Bedes eagerness to supplement classical
quotations with Christian ones and to introduce ecclesiastical vocabulary into word-lists;
something which is already present, although to a relatively modest degree, in the works of
such figures as Priscian, Isidore and Julian of Toledo. There are, however, some fundamental
differences between De arte metrica and its predecessors, and this can be largely attributed to
Bedes hands-on approach to questions of prosody. Bede generally relies more on his own
extensive reading of classical and Christian verse than he does on the often abstruse
definitions and typologies suggested by previous generations of grammarians; as a
consequence, although the work is intended to be a prescriptive guide to verse, it largely
reflects the poetic practices of Bedes favourite authors, above all Sedulius, who appears
almost as a replacement for Vergil throughout Bedes treatise. This is also the partial result of
Bedes integration of syllable prosody into his presentation of dactylic verse. Rather than
presenting the rules of prosody in their age-old and well-circulated guise with Christian
quotations thrown in, Bede largely derived his rules from his poetic examples. This, in turn,
has resulted in an unprecedented overhaul of prosodic rules, which, in Bedes treatise, largely
reads as a codification of a Christian metrical standard. The departures from classical
practice are minute but consistent, reflecting several post-classical features of the Christian
epic poets, mainly in their use of common syllables such as word-initial s groups and hs
creating position, but condemning others, such as the use of hiatus and spondaic verses, often
casting them as pagan practices. It is evident throughout Bedes treatise that his main
model for good Christian verse is Sedulius, as he ignores the evidence of such pagan
practices in Venantius Fortunatus and several other Christian poets.
Bedes crude dichotomy of pagan and Christian verse techniques is surely
subjective, and at times the logic behind it is hard to follow. It is difficult to say whether Bede
condemns some metrical practices because he truly regards them as pagan, or whether he
simply justifies his subjective aesthetic objections to them by labelling them as pagan.
Evidently, there is a deeply ingrained streak of aesthetic purism behind Bedes generally
practical approach to verse: in his chapter on the structure of the hexameter, for example, he
argues that a hexameter line must have twenty-four morae, exactly like there are twenty-four
half-ounces in a Roman pound. It is probably on account of such purism that Bede rejects the
208

spondaic verse more vehemently than any of his predecessors, not merely condemning it but
excluding it from his very definition of the hexameter line. The hexameter is, for Bede, a part
of creation, a view borne out by the statements of the Church Fathers who found, or claimed
they had found, hexameters in the Bible. Consequently, Bede goes to great lengths to give
plausible alternative analyses to lines in Christian epic which fail to fit his idealised mould of
the perfect hexameter line. On the other hand, Bede generally attributes similar deviations
from what he sees as the norm in pre-Christian verse to the simple fact of their pagan nature.
When it comes to lyric verse, Bedes treatise exhibits no such dichotomy,
because its presentation in De arte metrica has been Christianised to the point of total
exclusion of pagan material. Bede mainly discusses such metres as had been used by the early
Christian hymnodists (in addition to the phalaecean and sapphic metres employed by
Cyprianus and Paulinus). The presentation of these metres is streamlined and, in places,
profoundly simplified, mainly when it comes to Bedes analysis of their structure: Bede
dissects all metres into feet of two to three syllables. He follows Mallius Theodorus in his
rejection of the classical two-foot metron in iambo-trochaic verse, and, like several of his
predecessors, presents the phalaecean and sapphic metres (together with the post-classical
aeolic length known as terentianean) as forms of dactylic verse. On the other hand, his
discussion of the iambic dimeter used in the hymns of Ambrose is, if anything, more
extensive than anything that can be found in Bedes secondary sources, which yet again
indicates that, apart from hexameter and elegiac verse, the most important form of poetry for
him was the Christian hymn. Similarly, his presentation of the trochaic septenarius is unlike
anything we encounter in the works of previous grammarians: Bede presents the line as a
couplet of two lines and imposes further metrical restrictions on its structure. Bedes
idiosyncrasies must at least partly be attributed to the fact that the two lyric lengths he
discusses in the greatest detail are, for him, above all metres of Christian hymnody, and both
of them had generally been either ignored or, in the case of the septenarius, misrepresented in
the grammars of late antiquity.
For Bede, the central role of these two metres also necessitated a closer and
more analytical discussion of the non-quantitative hymns which had evolved in late antiquity.
As studies have shown, these hymns were based on the accent-patterns of the iambo-trochaic
metres but without regard for syllable quantity, something which Bede certainly had
perceived. As the difference between, say, the metrical form of the iambic dimeter and its
rhythmic variant would have been largely imperceptible when sung, it was necessary for
Bede to emphasise the differences between these two poetic forms: that the reader should not
209

look for syllable quantity where none existed; that both metrical and rhythmic verse were
legitimate forms of poetic expression; and that they were mutually exclusive. Bedes implicit
views on the infallibility of the Christian auctoritates dictated that a well-established form of
ecclesiastical poetry could not be faulty: therefore, he had to emphasise that also verse
without a quantitative structure could be tastefully composed. This is all the more significant
if we bear in mind how meticulous Bede was in his approach to questions of prosody in
quantitative verse, where he showed little tolerance for vacillating syllable lengths, often
showing himself to be stricter than poetic practice necessitated. In Bedes view, the rules of
prosody should be observed to the letter or not at all, and, as a consequence, it was necessary
for him to recognise rhythmic verse as a literary form in its own right, a delineation that was
historically significant for many generations of medieval poets.
The few but characteristically Late Latin instructions on good poetic style
which Bede gives in his presentation of hexameter technique also foreshadow the practices of
medieval verse, metrical and rhythmic alike. Although Bede, at least in his De arte metrica,
does not discuss rhyme, a prominent feature of medieval poetry, he was certainly aware of
the its existence, and, in recommending a line-type with the kind of word order we term
golden, may indirectly have contributed to its proliferation, as internal rhyme is a common
feature of golden lines, and, indeed, of all hyperbata where the noun head and its attribute
belong to the same declension. Both golden lines and internal rhyme are, yet again, typical of
the poetic style of Sedulius, and appear frequently together in later rhyming hexameters of
the Leonine type. In other words, of the most immediately recognisable features of medieval
verse, Bede effectively defines non-quantitative prosody and at the very least hints at rhyme,
although not in the same context. Although the bulk of De arte metrica discusses the dactylic
hexameter and rarely wanders far from classical practice, it anticipates, on a more modest
scale, several poetic practices that only reached their full fruition in the high Middle Ages.
The importance of De arte metrica is manifest on several levels: firstly, as a didactic work, it
constitutes one of the first treatises on metre that was aimed specifically at a monastic
audience. At the same time, in incorporating the presentation of syllable prosody into its
presentation of quantitative metres, it presents a major advance over the various treatises on
metre composed by the Late Latin grammarians; this, of course was necessitated by the fact
that it was not aimed at an audience of native Latin-speakers. Consequently, it formed the
prime model for the artes metricae composed in the following centuries. As for its content,
De arte metrica effectively codifies several metrical practices of late antiquity, and seeks to
create a standard for good hexameter verse that is ultimately modelled on the example of one
210

man: Caelius Sedulius. This partisanship, taken together with Bedes purist approach to
verse, resulted in sometimes insightful, sometimes impractical definitions which nevertheless
had an impact on poets of later generations. Apart from hexameter and elegiac verse, Bedes
main preoccupation was with Christian hymnody, where, by his time, rhythmic verse already
coexisted with the older quantitative metres. In recognising these two as separate forms, he
laid the ground for the various verse forms that appeared in the Middle Ages, and, ultimately,
even the vernacular verse of our own times may owe more to Bedes observations than is
generally realised.
211

Bibliography


Allen, W. Sidney 1973: Accent and Rhythm. The Prosodic Features of Greek and Latin: A
Study in Theory and Reconstruction. Cambridge.
Allen, W. Sidney 1978
2
: Vox Latina. The Pronunciation of Classical Latin. Cambridge.
Avalle, DArco Silvio 1992: Dalla metrica alla ritmica. Lo spazio letterario del medioevo.
1. Il medioevo latino. Roma. 391-476.
Baldi, Philip 1999: The Foundations of Latin. Berlin (Trends in Linguistics, Studies and
Monographs 117).
Baroway, Israel 1950: The Accentual Theory of Hebrew Prosody: A Further Study in
Renaissance Interpretation of Biblical Form. ELH: English Literary History vol. 17.
115-135.
Bassett, Samuel E. 1917: The Hephthimimeral Caesura in Greek Hexameter Poetry.
Transactions and Proceedings of the American Philological Association vol. 48. 85-
110.
Bassett, Samuel E. 1919: The Theory of the Homeric Caesura According to the Extant
Remains of the Ancient Doctrine. The American Journal of Philology vol. 40. 343-
372.
Bischoff, Bernhard 1966-1967: Mittelalterliche Studien. Stuttgart.
Blair, Peter Hunter 1970: The World of Bede. New York.
Bolton, W. F. 1967: A History of Anglo-Latin Literature 597-1066. Volume 1: 597-740.
Princeton.
Bonner, Gerald 1973: Bede and Medieval Civilization. Anglo-Saxon England 2. 71-90.
Brown, George Hardin 1987: Bede the Venerable. Boston.
Brown, George Hardin 2009: A Companion to Bede. Woodbridge; Rochester.
Brunhlzl, F. 1975: Geschichte der lateinischen Literatur des Mittelalters, Bd. 1. Mnchen.
Bulst, Walther 1956: Hymni Latini antiquissimi LXXV, psalmi III. Heidelberg.
Burles, Edward 1652 (repr. 1972): Grammatica Burlesa. Menston (English Linguistics 1500-
1800. A Collection of Facsimile Reprints. No. 307).
Campbell, Alastair 1953: Some Linguistic Features of Early Anglo-Latin Verse and Its Use
of Classical Models. Transactions of the Philological Society 1953. 1-20.
212

Ceccarelli, Lucio 2006: Note sullh iniziale in Venanzio Fortunato. Latin vulgaire Latin
tardif VII. Actes du VII
me
Colloque internationas sur le latin vulgaire et tardif.
Sville, 2-6 septembre 2003 (ed. Carmen Arias Abelln). Sevilla. 203-219.
Ceccarelli, Lucio 2008: Contributi per la storia dell esametro latino I-II. Roma.
Chiappiniello, Roberto 2007: The Carmen ad uxorem and the Genre of the Epithalamium.
Poetry and Exegesis in Premodern Latin Christianity. The Encounter between
Classical and Christian Strategies of Interpretation (eds. Willemien Otten and Karla
Pollmann). Leiden / Boston 2007. 115-138.
Colgrave, Bernard & Mynors, R. A. B. (eds.) 1969: Bedes Ecclesiastical History of the
English People. Oxford.
Coronati, Lia 1981-1982: La dottrina del tetrametro trocaico in Beda. Romanobarbarica 6.
53-62.
Curtius, Ernst Robert 1948: Europische Literatur und lateinisches Mittelalter. Bern.
Dassmann, E. 1970: Die Bedeutung des Alten Testaments fr das Verstndnis des
kirchlichen Amtes in der frhpatristischen Theologie. Bibel und Leben 11. 198-214.
Dionisotti, Carlotta 1982: On Bede, Grammars and Greek. Revue Bndictine 92. 111-140.
Drexler, Hans 1967: Einfhrung in die rmische Metrik. Darmstadt.
Duckworth, Georce E. 1969: Vergil and Classical Hexameter Poetry: A Study in Metrical
Variety. Ann Arbor.
Eden, P. T. 1975: A Commentary on Vergil: Aeneid VIII. Leiden.
Ehwald, Rudolf 1919: Aldhelmi opera. Berolini (Monumenta Germaniae historica. Auctores
antiquissimi 15).
Fitch, J ohn 1987: Senecas Hercules Furens: A Critical Text with Edition and Commentary.
Ithaca, New York.
Fraipoint, J . (ed.) 1955: Opera rhythmica. Bedae Venerabilis opera homiletica. Opera
rhythmica (eds. D. Hurst and J. Fraipoint). Brepols (Corpus Christianorum. Series
Latina 122).
Franklin, Carmela Vircillo 2000: The Date of Composition of Bedes De schematibus et
tropis and De arte metrica. Revue Bndictine tome 110. 99-204.
Gneuss, Helmut 1968: Hymnar und Hymnen im englischen Mittelalter. Tbingen.
Grandgent, Charles W. 1907: An Introduction to Vulgar Latin. Boston.
Harrison, S. J. 1991: Discordia Taetra: The History of a Hexameter-Ending. The Classical
Quarterly, New Series vol. 41. 138-149.
213

Heikkinen, Seppo 2004: The Poetry of Venantius Fortunatus: The Twilight of Roman
Metre. The Eastern Mediterranean in the Late Antique and Early Byzantine Periods
(eds. Maria Gourdoba, Leena Pietil-Castrn and Esko Tikkala). Helsinki (Papers and
Monographs of the Finnish Institute at Athens vol. IX). 17-31.
Heikkinen, Seppo 2007: Vergilian Quotations in Bedes De arte metrica. The Journal of
Medieval Latin vol. 17. 101-109.
Heikkinen, Seppo 2010: Quae non habet intellectum: The Disappearance of Spondaic Fifth
Feet from Dactylic Hexameter verse. Interfaces between Language and Culture in
Medieval England. A Festschrift for Matti Kilpi (eds. Alaric Hall, Olga Timofeeva,
gnes Kiricsi and Bethany Fox). Leiden Boston (The Northern World vol. 48). 81-
98.
Henry, H. T. 1912: Rex sempiterne caelitum. The Catholic Encyclopedia, vol. XIII. New
York. Rep. http://www.newadvent.org/cathen/13017b.htm
Hiley, David 1993: Western Plainchant: A Handbook. Oxford/New York.
Holder, Arthur 1999. (Un)dating Bedes De Arte Metrica. Northumbrias Golden Age (eds.
Jane Hawkes and Susan Mills). Sutton. 390-395.
Holford-Strevens, L. A. 1979: Nola, Vergil and Paulinus. The Classical Quarterly vol. 29.
391-393.
Holtz, Louis 1981: Donat et la tradition de lenseignement grammatical: tude sur lArs de
Donat et sa diffusion (IVe IXe sicle) et dition critique. Paris (Documents, tudes et
rpertoires).
Housman, A. E. 1927: Prosody and Method. The Classical Quarterly vol. 21. 1-12.
Huemer, J ohann (ed.) 1883: Cruindmelis sive Fulcharii Ars metrica. Beitrag zur Geschichte
der karolingischen Gelehrsamkeit. Wien.
Huemer, J ohann (ed.) 1885: Sedulii opera omnia / recensuit et commentario critico instruxit
Iohannes Huemer; accedunt excerpta ex Remigii expositione in Sedulii Paschale
carmen. Wien (Corpus scriptorum ecclesiasticorum Latinorum, vol. 10).
Irvine, Martin 1986: Bede the Grammarian and the Scope of Grammarical Studies in Eighth-
Century Northumbria. Anglo-Saxon England 15. 15-44.
J aager, Werner (ed.) 1935: Bedas metrische Vita Sancti Cuthberti. Leipzig (Palaestra 198).
J ohnston, Harold Whetstone 1897: A Collection of Examples Illustrating the Metrical
Licenses of Vergil. Chicago.
J ones, Charles W. (ed.) 1975: Bedae Venerabilis opera. Opera didascalica. CCSL 123A.
Turnhout.
214

Kabell, Aage 1960: Metrische Studien 2. Uppsala (Uppsala universitets rsskrift 6).
Kendall, Calvin B. (ed.) 1975: De arte metrica et de schematibus et tropis. Bedae
Venerabilis Opera. Opera didascalica. (ed. Charles W. Jones). Turnhout (Corpus
Christianorum. Series Latina 123A). 59-171.
Kendall, Calvin B. (ed.) 1991: Bede. Libri II De arte metrica et de schematibus et tropis. The
art of Poetry and Rhetoric. Saarbrcken (Bibliotheca Germanica. Series Nova vol. 2).
Klopsch, Paul 1972: Einfhrung in die mittelalterliche Verslehre. Darmstadt.
Klopsch, Paul 1980: Einfhrung in die Dichtungslehren des lateinischen Mittelalters.
Darmstadt.
Laistner, M. L. W. (ed.) 1939: Bedae Venerabilis Expositio actuum apostolorum et
retractatio. Cambridge, MA (The Medieval Academy of America, no. 35).
Laistner, M. L. W. 1957: The Intellectual Heritage of the Early Middle Ages: Selected
Essays. Ithaca, N.Y.
Laistner, M. L. W. and King, H. H. 1943. A Hand-List of Bede Manuscripts. Ithaca, New
York.
Lapidge, Michael 1972: Three Latin Poems from thelwolds School at Winchester.
Anglo-Saxon England 1. 85-137.
Lapidge, Michael 1975: The Hermeneutic Style in Tenth-Century Anglo-Latin Literature.
Anglo-Saxon England 4. 67-112.
Lapidge, Michael 1979: Aldhelm' s Latin Poetry and Old English Verse. Comparative
Literature vol. 31. 209-231.
Lapidge, Michael (trans.) 1985: The Enigmata. Aldhelm: The Poetic Works. Translated by
Michael Lapidge and James L. Rosier with an Appendix by Neil Wright. Cambridge.
70-94.
Lapidge, Michael 1996: Anglo-Latin Literature. London and Rio Grande.
Lapidge, Michael 1999: Poetic Technique, Latin. The Blackwell Encyclopaedia of Anglo-
Saxon England (eds. Michael Lapidge, J ohn Blair, Simon Keynes and Donald
Scragg). Oxford. 372-373.
Lapidge, Michael 2005: Bede and the Poetic Diction of Vergil. Poesa latina medieval
(siglos V-XV). Actas del IV Congreso del Internationales Mittellateinerkomitee
Santiago de Compostela, 12-15 de septiembre de 2002 (eds. M. C. Daz y Daz and J.
M. Daz de Bustamente). Firenze. 739-748.
Lapidge, Michael 2006: The Anglo-Saxon Library. Oxford.
215

Lausberg, Heinrich 1998: Handbook of Literary Rhetoric: A Foundation for Literary Studies;
with a Foreword by George A. Kennedy (eds. David E. Orton and R. Dean Anderson).
Leiden.
Law, Viven 1982: The Insular Latin Grammarians. Woodbridge
Law, Vivien 1984: Irish Symptoms and the Provenance of Sixth- and Seventh-century Latin
Grammars. Matriaux pour une histoire des thories linguistiques (eds. Sylvain
Auroux, Michel Glatigny, Andr Joly, Anne Nicolas and Irne Rosier). Lille.
Law, Vivien 1997: Grammar and Grammarians in the Early Middle Ages. London.
Lawler, T. 1974: The Parisiana Poetria of John of Garland. New Haven (Yale Studies in
English 182).
Lentini, A. & Avagliano, F. 1974: I carmi di Alfano I. Salerno (Miscellanea cassinese 38).
Leonhardt, J rgen 1985: Dimensio syllabarum: Studien zur lateinischen Prosodie- und
Verslehre von der Sptantike bis zur frhen Renaissance. Gttingen (Hypomnemata
92).
Leotta, Rosario 1980: Una classificazione aldelmiana. Giornale italiano di filologia vol.
32. 245-250.
Limone, Oronzo 1984: Il liber de dubio accentu. Studi Medievali 25.
Lindsey, W. M. 1922: Early Latin Verse. Oxford.
Lloyd, Robert B. 1956: Penatibus et magnis dis. The American Journal of Philology vol.
77. 38-46.
Long, Omera Floyd 1901: On the Usage of Quotiens and Quotienscunque in Different
Periods of Latin. Baltimore.
Luiselli, Bruno 1976: Il De arte metrica di Beda di fronte alla tradizione metricologica
tardo-latina. Grammatici latini d' et imperiale. Miscellanea filologica. Genoa
(Pubblicazioni dell' istituto di filologia classica e medievale, 45). 169-180.
Maestre Yenes, M. A. H. 1973: Ars Iuliani Toletani Episcopi. Una gramtica latina de la
Espaa visigoda. Estudio y edicin critica. Toledo.
Manitius, Max 1911: Geschichte der lateinischen Literatur des Mittelalters. Erster Teil.
Mnchen (Handbuch der klassischen Altertumswissenschaft 9, 2, 1).
Mariotti, Italo 1967 (ed.): Marii Victorini ars grammatica / introduzione, testo e commento a
cura di Italo Mariotti. Firenze (Testi con commento filologico 6).
Mayer, Kenneth 2002: The Golden Line: Ancient and Medieval Lists of Special Hexameters
and Modern Scholarship. Latin Grammar and Rhetoric. From Classical Theory to
Medieval Practice (ed. Carol Dana Lanham). London New York. 139-179.
216

Mazzarino, Anatonio (ed.) 1955: Grammaticae Romanae fragmenta aetatis Caesareae. Vol.
1. Torino.
Meyer, Wilhelm 1905-1936: Gesammelte Abhandlungen zur mittellateinischen Rythmik I-III.
Berlin.
Niedermann, Max (ed.) 1937: Consentii Ars de barbarismis et metaplasmis. Victorini
Fragmentum De soloecismo et barbarismo. Neuchtel: Universit de Neuchtel
(Facult des lettres, Universit de Neuchtel. Recueil de travaux, fasc. 18).
Norberg, Dag 1958: Introduction l' tude de la versification latine mdivale. Stockholm
(Acta Universitatis Stockholmiensis. Studia Latina Stockholmiensia 5).
Norberg, Dag 1988: Les vers latins iambiques et trochaques au Moyen Age et leurs rpliques
rythmiques. Uppsala (Filologiskt arkiv 35).
Norberg, Dag 1998: Au seuil du Moyen Age. II. tudes linguistiques, mtriques et littraires
1975-95. Uppsala (Filologiskt arkiv 40).
Nougaret, Louis 1948: Trait de la mtrique latine classique. Paris.
Orchard, Andy 1994: The Poetic Art of Aldhelm. Cambridge. (CSASE 8).
Orchard, Andy 1999: Octosyllables, Latin. The Blackwell Encyclopaedia of Anglo-Saxon
England (eds. M. Lapidge, J. Blair, S. Keynes and D. Scragg). Oxford. 339.
Palmer, Robert B. 1959: Bede as a Textbook Writer: A Study of his De Arte metrica.
Speculum 34. 573-584.
Panhuis, Dirk 2006: Latin Grammar (English translation/adaptation by Dirk Panhuis and
Gertrud Champe). Ann Arbor.
Passalaqcua, Marina 1987 (ed.): Prisciani Caesariensis opuscula. Vol. 1, De figuris
numerorum; De metris Terentii; Praeexercitamina / edizione critica a cura di Marina
Passalacqua. Roma (Sussidi eruditi 40).
Petit, F. 1924: Anacrontiques. Dictionnaire darchologie chrtienne et de liturgie I.2.
(publ. par Fernand Cabrol et Henri Leclercq). 1863-72.
Plummer, C. (ed.) 1896: Venerabilis Baedae opera historica. Oxonii.
Raby, F. J. E. 1947: On the Date and Provenance of Some Early Latin Hymns. Medium
Aevum 16. 1-5.
Raven, D. S. 1965: Latin Metre. London.
Raven, D. S. 1968
2
: Greek Metre. London.
Ridings, Daniel 1995: The Attic Moses: The Dependency Theme in Some Early Christian
Writers. Gteborg (Studia Graeca et Latina Gothoburgensia LIX).
Roger, Maurice 1905: L' enseignement des lettres classiques d' Ausone Alcuin. Paris.
217

Rosier, James L. (trans.) 1985: The Carmen de virginitate. Aldhelm: The Poetic Works.
Translated by Michael Lapidge and James L. Rosier with an Appendix by Neil
Wright. Cambridge. 95-167.
Ruff, Carin 2005: The Place of Metrics in Anglo-Saxon Latin Education: Aldhelm and
Bede. The Journal of English and German Philology vol. 104. 149-170.
Santelia, Stefania 2009 (ed.): Prospero dAquitania: Ad coniugem suam. In appendice: Liber
epigrammatum. (Studi Latini 68). Napoli.
Schaller, Dieter 1984: Der alkaische Hendekasyllabus im frhen Mittelalter.
Mittellateinisches Jahrbuch XIX. 73-90.
Shipley, F. W. 1924. Hiatus, Elision, Caesura, in Virgils Hexameter. Transactions and
Proceedings of the American Philological Society vol. 55. 137-158.
Sicking, C. M. J. 1993: Griechische Verslehre. Mnchen (Handbuch der
Altertumswissenschaft: Abt. 2; Teil 4).
Steen, J anie 2008: Verse and Virtuosity: The Adaptation of Latin Rhetoric in Old English
Poetry. Toronto.
Tardi, D. 1927: Fortunat. tude sur un dernier reprsentant de la posie latine dans la Gaule
mrovingienne. Paris.
Thraede, Klaus 1978: Der Hexameter in Rom. Verstheorie und Statistik. Mnchen (Zetemata:
Monographien zur klassischen Altertumswissenschaft: Heft 71).
Townend, G. B. 1950: Oxytone Accentuation in Latin Elegiacs. The American Journal of
Philology vol. 71. 22-39.
Treitler, Leo 2003: With Voice and Pen. Coming to Know Medieval Song and How it was
Made. Oxford.
Turner, Sharon 1836: The History of the Anglo-Saxons from the Earliest Periods to the
Roman Conquest. London.
Verrier, Paul 1931: Le vers franais: formes primitives: dveloppement: diffusion, vol.1.
Paris.
Vossler, Karl 1951: Dichtungsformen der Romanen. Stuttgart.
Wagner, J . Kurt 1907: Quaestiones neotericae imprimis ad Ausonium pertinentes. Leipzig.
Walpole, A. S. 1922: Early Latin Hymns. Cambridge.
Wilkinson, L. P. 1963: Golden Latin Artistry. Cambridge.
Winbolt, Samuel Edward 1903: Latin Hexameter Verse: An Aid to Composition. London.
218

Witke, Charles 1971: Numen Litterarum. The Old and the New in Latin Poetry from
Constantine to Gregory the Great. Leiden (Mittellateinische Studien und Texte, Bd.
5).
Wright, Neil 1985: Appendix: Aldhelms Prose Writings on Metrics. Aldhelm: The Poetic
Works. Translated by Michael Lapidge and James L. Rosier with an Appendix by Neil
Wright. Cambridge. 181-219.
Wright, Neil 2005: The Metrical Art(s) of Bede. Latin Learning and English Lore. Studies
in Anglo-Saxon Literature for Michael Lapidge, vol. I (eds. Katherine OBrien
OKeeffe and Andy Orchard). Toronto. 150-170.
Young, Arthur M. 1935: Schematized Word-Order in Virgil. The Classical Journal 1932.
515-522.
Zirin, Ronald Andrew 1970: The Phonological Basis of Latin prosody. The Hague (J anua
linguarum, Series practica; 99).