Lewin Theory Phenomenology Modes of Perception

Music Theory, Phenomenology, and Modes of Perception
Author(s): David Lewin

Source: Music Perception: An Interdisciplinary Journal, Vol. 3, No. 4 (Summer, 1986), pp.
327-392
Published by: University of California Press
Stable URL: https://www.jstor.org/stable/40285344
Accessed: 01-12-2019 20:22 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
University of California Press is collaborating with JSTOR to digitize, preserve and extend
access to Music Perception: An Interdisciplinary Journal
This content downloaded from 128.195.70.55 on Sun, 01 Dec 2019 20:22:31 UTC
All use subject to https://about.jstor.org/terms
Music Perception ©1986 BY THE REGENTS OF THE
Summer 1986, Vol. 3, No. 4, 327-392 UNIVERSITY OF CALIFORNIA
Music Theory, Phenomenology, and Modes of Perce
DAVID LEWIN
Harvard University
Recent years have seen an increasing influence on music t

perceptual investigations that can be called phenomenolog
sense of Husserl, either explicitly or implicitly. The trend is p
particularly in what one might call its sociology, but it is
promising. Potential or at least metaphorical links with Art
telligence are especially suggestive. A formal model for
perceptions," incorporating some of the promising feature
interesting things in connection with Schubert's song Morgen
model helps to circumvent some traditional difficulti
methodology of music analysis. But the model must be used w
since, like other perceptual theories, it appears to make "li
paradigmatic musical activity. Composer/performer/play
actor/director/poet can be contrasted here to listener/reade
genera can be compared in the usual ways, but also in some no
ways. The former genus may be held to be perceiving in the c
and some influential contemporary literary theories actual
members of this genus to those of the other as perceivers. Th
can be modified, I believe, to allow a more universal stance
regards acts of analytic reading/listening as poetry.
Part I: Phenomenological Preface
Overtly phenomenological study of music in Husserl's

the man himself, who made central to his theories of per
analysis for perceiving a sustained tone (Husserl, 1964,
sis is highlighted by Izchak Miller in a recent philosop
which the interested reader will find especially clear (
puts the heart of the matter as follows:
. . . whereas it does seem true that I am hearing that tone

certain interval of time, it does not seem it can be true that
all of it (or an extended part of it) at any given instant of
Yet . . . throughout that interval I continuously experience
ance, or the continuity, of that tone, and this requires (con
Requests for reprints may be sent to David Lewin, Department of

versity, Cambridge, Massachusetts 02138.
327
328 David Lewin
previous hypothesis) that I e

than a mere instantaneous pha
neous perceptual experience of
passage, of a tone possible?
Answering this and other r
awareness is of crucial import
yond the mere desire to provi
perception. The subject mat
conscious experience, and Huss
our various experiences and di
scious experiences, or - as H
ness, are themselves process
then, succeed in being reflec
continuity, or the passage, of
words, succeed in reflecting
than the corresponding mome
According to Husserl, the str
makes the continuous percepti
sible is the very same structur
the temporal passage of our m
bility of the first is, thus, acc
(Miller, 1984, pp. 2-3).
Miller also devotes much att

Melody" (Miller, 1984, pp. 1
account of listening to the
certo, does not itself invoke
any theorist interested in
Narmour - or serialism for th
the context of Miller's book,
ences.
Among explicitly phenomenological writers who do invo

cated music-theoretical concepts, Judith Lochhead is especi
thy. Her dissertation in particular projects an avowedly phe
view of Western art music from many periods; it comment
tively on temporal issues that have to do with our finding mu
sic recalcitrant to received analytic approaches, a problem
phenomenological in nature (Lochhead, 1982). 1
Thomas Clifton also proclaimed a phenomenological appro
theory, although of a quite different sort. The title of his recen
1. Lochhead worked extensively with the philosopher Don Ihde, who a

portant work on the phenomenology of hearing (Ihde, 1976).
Music Theory 329
book (Clifton, 1983) reflects his stance

of Clifton's earlier work does not ex
worth reading for anyone interested i
applying phenomenology to music t
has recently been reviewed concisel
(1983) and by James Tenney (1985).
Few professional music theorists ha
enological a program or approach as
phenomenological thinking is implicitl
well. Jonathan Kramer's temporal st
modes of thought.3 So do Christoph
less obvious way, does a recent study
article builds a numerical model whi
number of time-spans I recall from th
duration d.4 In this way I construct a
folding durational-interval vector"
concept underlying my construction
model of perceptual time, a model t
impressions," impressions that follow
serl's "retentions," projections of re
tions) into my present consciousness. L
volved with something much like H
future expectations into present consc
have found the idea of an "unfolding r
gestive in connection with a great var
2. Unfortunately, further development of Clif

death.
3. Kramer is a consulting editor for Music

devoted entirely to time and rhythm in musi
vance to this article. Its editor is Lewis Row
temporality in music. Of phenomenological in
"The Subconscious Language of Musical Tim
that "the terms for temporality in music . .
abstract thing that is measured by the gest
ties in suggestively with remarks I shall make
of perception.
4. For the model, it is perfectly workable to

"more-or-less-d, as distinguishable from other
5. HusserPs terms and pertinent diagrams ar

lowing).
6. That will be manifest in my forthcoming book, Generalized musical intervals and

transformations (New Haven: Yale University Press, expected 1987).
330 David Lewin
Marvin Minsky - like my

phenomenologically orient
would find itself very mu
to really understand how m
simply have to use much mo
that can describe how pro
makes his statement in co
'transformational' meth
Schenkerian methods. The
Narmour when, protesting
patterns" of Schenkerian th
in musical process is to be
prospect ... in relation to
1977, p. 40). This sort of d
primal impressions are pa
implying-and-realizing; ret
present perception; proten
present perception.
The works of the non-phen
mulate and examine very a
of Husserl's perception-stru
the way in which such str
loops with other perception
ments. The other perception
tionships to the given struc
realization, denial), and tho
tions between perceptions,
objects or arguments of per
Let me illustrate the sort o
lish text. Consider a thi
Siegmund's watching S
Sieglinde's watching . . . (etc
of p by using a pair of fini
ally recursive relationsh
SGL = Sieglinde's watching
ate another infinite perce
watching . . . (etc). A comp
evaluation routine (let us c
7. Compare this excerpt from M

scious experiences, or - as Husse
processes, albeit mental processes
any given moment of the continu
Music Theory 331
EVALuating SGL. There is of course a

EVALuation would go on forever, trapp
two ways to avoid this difficulty that
tual musical implications and in ligh
computer science. One would be to ha
more global part of the system interru
ers. Another would be to have some so
applied to the environment before a
The parser could spot the endless loop;
tual EVALuation to be terminated (by
or what you will) after a certain numbe
all, is what we ourselves do in writin
special symbols ". . ." and/or "(etc.)"
loop structure has been made clear
avoiding the infinite loop could be c
put like "Siegmund's watching Siegli
. . . Sieglinde's suddenly noticing Hu
excellent external interrupt.)
Having explored the abstract textua
feeling for the kinds of recursive syst
now examine an abstract musical exa
I shall be especially interested first
musical perceptions; this correspo
SGM-as-an-object-of-SGL, in the En
ested in the specifically recursive aspe
Fig. 1.
8. If the parser applies itself only to a restricted family of formal strings called "percep-
tions," and the perceptions do not engage the parsing language itself, then certain technical
"Church-Turing" problems should not arise. Computer buffs will know what I mean (al-
though they may not agree). For other readers, one might put the matter this way in intuitive
discourse: if parsing is to be applied, then musical perceptions should not form a "lan-
guage," and/or the parsing itself should be "imperceptible."
332 David Lewin
tures; this corresponds to th

sidered together as a percepti
Imagine a string ensemble
ducing an acoustic signal wh
am I "perceiving" as I listen m
ponding to cursor-time X o
what I am perceiving - let us
work of things, things inclu
themselves, and their relatio
perceptions (al), "V7 harmon
bass over the last beat," and
beat." I perceive how the p
among themselves. I perceiv
Perception(a) at cursor-time X
(a2), and (a3) each relate to y
gree in the bass over the last
jects retained perceptions of
last beat" at every perceptual
of clock time preceding X;
"Dominant-seventh harmony
an analogous family of objec
time well behind the G7 chor
built musical contexts of th
acts of "perceiving a domina
the extent that "dominant"
ideas and language, their cont
the entire musical performan
Particularly interesting as an
responding to the score of F
Perception(a) does not notic
ception^) in certain relation
least "protension" (if not "im
of affairs and the traditional
tional view, Perception(b) "ha
"expect" it, perhaps with a ce
serlian view, Perception(b) d
ceive at time X the structure
perception - along with cer
ception^) at cursor-time X
("coming up"), mensurated
likelihood ("very likely in th
"The C eight-three chord" is
least not directly. The chord
Music Theory 333
ception(b), which is as yet perceived on

not "the C chord" which is "very li
rather it is "the confirmation-time fo
coming up in one beat's time," as I p
time to Signal(a), I do not form the
coming up over the next beat as a co
mental construct of a C major chord c
in the context of a broader mental con
We are now in a position to explore
pects of musical perception-structur
quiring after the objects of Perceptio
ception^) itself, in a particular re
terminology, we could describe the r
perceive in Perception(b) includ
implication-realized. Here we encounte
inquire what we perceive that is implie
tion^) in a relation of realization-imp
pect of the situation by formulating e
manner of SGM and SGL: IMP = (a
realization of IMP.
For a more general model of percep
isolate recursive relationships in this
our primary focus must be on the per
is how the model I shall soon propose w
Perceptions (a) and (b) will each be def
in the list for Perception(a), we might
ing of the pair [Perception (b), impl
tion^) we should then place a formal
[Perception(a),realization]. One can im
the situation to lie within two sym
Perception(a) ...(... (Perception(b),i
tion^) ...(...(... (Perception(a),realiz
course into symbolic computer languag
possible utility of Artificial Intellige
studying these matters. Thereby I me
tact with Minsky, and with certain
9. Within the DEFINE list for Perception(

suitably qualified by a formal probability valu
of expectation or predictability associated w
something-(b)-like in some well-stipulated sen
like these into a formal definition of "implic
334 David Lewin
well.10 Minsky (1982) devote

Miller (1984, pp. 93-97) uses
that suggests an AI enviro
more detail.
Before I get to that, thoug
slipped the reader's attentio
Perception(b) as part of wh
tion^) as part of what-(b)-
supposed that our imaginary
producing an acoustic stim
some exploration.
First let us suppose that th
of Figure l(a) to produce a
stead perform Figure l(c),
appropriate Perception(c).
about Perception(b) and its
main exactly as we have a
Signal(c) at now-time Y ch
ready discussed continues t
all the functions it had at t
in connection with Percep
ceives" is precisely that (b) is
imagine a computer stat
tion(b),denial) ...)...). In ord
must be at hand at that tim
that of (c). One must not
placing" (b).
Let us consider next the tr
ure l(b), producing acoustic
now have at hand a new Per
ent from our old acquainta
defined at cursor-time X,
10. My stance here is not partic

serPs phenomenology and the wo
instance, of a recent publication
The editors' Introduction contain
telligence" (pp. 17-19) and "Huss
tion characterizes Minsky's fram
similar to HusserPs for represen
work in this area has been done
addresses recursive aspects of perc
hefty body of other work by L
Report Music and Mind, An Art
Music Association, 1981).
Music Theory 335
it additionally becomes at cursor-time

yes). We might say that (b-yes) "conf
previous case "denied" (b). Symb
Perception(b-yes) . . . (Perception(
Part II: A General Model
If one were to sequester the notion of "good" continuation as a descrip-

tor ... in tonal music, one would have to introduce . . . powerful con-
cepts of relation - including those of contradiction, opposition, and
paradox - as natural to the process, even necessary to it. (Browne,
1985, p. 6).11
To help us entertain the ideas discussed in Part I, and others of their ilk, I
propose as a provisional model for "a musical perception" this basic for-
mula:
p = (EV,CXT,P-R-LIST,ST-LIST).
Here the musical perception p is defined as a formal list containing four
arguments. The argument EV specifies a sonic event or family of events be-
ing "perceived." The argument CXT specifies a musical context in which
the perception occurs. The argument P-R-LIST is a list of pairs (pi,ri); each
pair specifies a perception pA and a relation tx which p bears to pl. The argu-
ment ST-LIST is a list of statements si, . . ., sK made in some stipulated lan-
guage L.
As an example we can construct one formal musical perception pertinent
to our intuition of "what we hear" when a quartet plays the last quarter-
note of Figure l(c) to finish a performance of Figure l(c). For the formal
perception, EV is "this thing that happens on the last beat." CXT is all-of-
Figure l(c), and also a culturally conditioned theoretical component that
makes us responsive to categories we call beats, keys, tonics, dominants, et
al. The P-R-LIST includes a pair (Perception(b),denial). The ST-LIST might
include, in a suitable language L, a statement, "deceptive cadence."
One might wonder why we need an argument EV at all, in the specific
example or in the general model. In the example, we describe EV as "this
thing that happens on the last beat." Now "on the last beat" is a perceptual
statement that might very easily be added to our ST-LIST. Generalizing that
observation, we can plausibly wonder what words we could possibly use, in
11. I must express very heartfelt gratitude to Fred Lerdahl and Diana Deutsch who, by
inviting me to give a lecture about musical perception, started me thinking along the lines of
the present paper, and in particular along the lines of the model here proposed. The lecture,
"Changing Perceptions over a Passage in Schubert," was given at the Fourth Workshop on
Physical and Neuropsychological Foundations of Music, Ossiach, Austria, in August of
1983.
336 David Lewin
pointing to an EV, that coul

..." and placed among the s
be expanded as necessary. C
specific example, without
meld L and L' into one super
only a syntactic dummy, a
three arguments: a context
and a ST-LIST of statements
ing certain statements that
"event" in the given Conte
I go over this possibility so
and rejected it, even though
Occam's Razor, and yet I w
plicitly denied the existen
statements about it that co
The social and political histo
my discomfort, and I freely
ceptibility to the Will-to-Tr
position. I prefer to believe
a perception are about som
thing EV will have at the ve
collect, and compare a cer
perceptions-about-EV. The
spect with Miller's analysi
that there must be ". . .a fea
the (purported) object of the
ties, a feature which prov
ported) object through a co
meanings of our act may s
perceptual noematic Sinn is
It seems that what Husserl
perceptual act is a 'purely re
meaning of an indexical, pro
'this.' " (Miller, 1984, pp. 7
demonstrable-X, my EV, al
sity for perceptual discourse
The necessity for a mus
clearer. For example, when p
ure 1 (c), I have one set of im
isolated harmonic structure
12. We shall explore just such shi

EVents, in the Schubert analysis t
Music Theory 337
ing it in the context of Figure l(c) as a

not make statements involving a "d
any sort, or indeed a "key" or a "beat.
involve further problems, in that I
within the pertinent music: there migh
music.
To illustrate the problem of locatability more thoroughly, let me suppose
I have before me, poised and ready, a classical orchestra. I bring down my
hand, cuing them in, and they produce a chord, forte and staccato, that lasts
about one-third of a second; then they rest for about two-thirds of a second;
then I cut them off. This "this," this EVent, this determinable-X of the situ-
ation, is produced by these instruments on these notes: flutes on Et 6 and
Bl>5; oboes on G5 and El>5; clarinets on Et5 and G4; bassoons on EH and
E!>3; horns on EH and G3; trumpets on Et5 and EH; kettledrum on Et3;
first violins on G5, BH, EH, and G3; second violins on Et5, EH, and G3;
violas on EH and G3; cellibass on Et 2 and EH. I turn to you and ask,
"What was that?" You reply, "It must be the opening of the Eroica Sym-
phony." "No," I respond, "it was actually measure 2 of the symphony."
"Unfair!" you exclaim. But why is it unfair? I had indeed instructed the
players to play measure 2 when I cued them; they were in fact all looking at
measure 2 in their parts as they played. In any conceivable sense you might
imagine, they did play measure 2. Only you did not perceive measure 2!
Your sense of unfairness arises here precisely because there is a crucial phe-
nomenological sense in which measure 2 is not a well-formed ConteXT.
Measures l-and-2-together are a well-formed ConteXT; you would be able
to locate measure 2 in that context. Measure 1 by itself, or rather measure 1
preceded by a certain amount of sound typical of "orchestra-not-playing,"
is also a well-formed musical ConteXT. That is why you immediately per-
ceived measure 1. According to my model, you were quite correct in that
perception; indeed it would have been impossible, in the formal sense of the
model, for you to have perceived anything else in the context at hand.
Let us study another example. Suppose I refer to the place in the Wald-
stein Sonata "where it goes like this" playing or pointing to Figure 2. Al-
though the event is perfectly well defined as an acoustic stimulus, even as an
auditory perception, there is no musical perception at hand, since you have
Fig. 2.
338 David Lewin
only the vaguest idea of w

under discussion is concer
"goes like this," has not be
a CXT, you can not have a
fectly clear auditory percep
third eighth of measure 2
teXT. Event-and-context a
If, on the other hand, I ref
note of the reprise measur
one that is suggested by Fig
I am claiming, in as radical
the reprise measure" does N
focused-on-the-pertinent-ev
ception. We are usually so
event "sounds." In contrast,
by the specific event give
like Figure 2. But that is a v
The problem of locatability
Here, I shall indicate only
Suppose that my orchestr
played anything for you. I
produce a chord, fortissim
second; then they rest for
The chord is as follows: f
El>5 and G4; bassoons on E
Et5 and Et4; kettledrum o
second violins on G5, Bl>4
Fig.3.
Fig. 4.
Music Theory 339
bass on Et 3 and Et 2. 1 turn to you

Beethoven, or a responsible conducto
orchestral 'cellist, and so on, you m
measure from the first movement o
More likely, though, you will answe
"No," I respond, "it was actually th
movement, measure 690." "Unfair!"
is of a very different type. For the ch
nically speaking." You are protesting o
making on your ear, and on your know
sonable to you. You will nevertheless
lent ear and a thorough knowledge of
chord. You will further admit that
class, or an advanced orchestration clas
such subtle differences between sound
that a student in an advanced analys
ponder why Beethoven comes so close
690, but does not reproduce it exactly
the propriety, if not the sufficiency, o
and of itself. That is, you will admit "in
now expounding, the propriety of
ConteXT for itself-as-EVent.
The question remains, however, to wh
should be considered "practically" lo
would agree that the demands I migh
are not to be made of "the listener," a
any phenomenological theory of mu
"phenomenology," will primarily ad
whose role vis-a-vis the Eroica differs f
conductor playing it, or Beethoven
perceptions of "the listener" have so
things a composer does, and to things
want to equate the roles of composer, p
our culture as it is today for better or
in this connection seems crucial to me.
of this article.
Meanwhile I should recall that locatab
volved when we stipulate a ConteXT
tion. Even when all the events at issue
p of our basic formula - depends on th
earlier point, the a-minor chord of
its own context, and a quite differen
l(c) as a whole. In the analysis of a
340 David Lewin
Part III of this paper, we sh

cal EVents whose perceptual
expand and/or contract in
Now let us return to th
LIST), and devote some att
recalled, is a list of pairs (pi
relation ^ which p bears to
p2, "imply" p3, "support"
then include the pairs (p
(p4,support), and (p5,succe
ceive pi-being-denied, p2-be
perception.
The P-R-LIST enables us to model recursive aspects of perception-
structuring; as we saw in earlier discussion, that is a powerful and charac-
teristic feature of the model. Earlier, for example, we could speak of Percep-
tion^) as perceiving Perception(b)-being-implied, while Perception(b)
perceived Perception(a)-being-realized. The P-R-LIST for Perception(a)
thus contained the pair (Perception(b),implication), while the P-R-LIST for
Perception(b) contained the pair (Perception(a),realization).
Eventually it may be necessary to formulate rules that determine when
certain recursive P-R configurations are malformed. A few such rules may
already appear obvious, but I would urge extreme caution in the matter.
After studying Parts III and IV of this paper, the reader will see why I want
to proceed so carefully. There we shall see that the geometry and logic of
musical perception are not easily inferable from the geometry of Euclid or
Descartes and the logic of Zermelo/Fraenkel or Gôdel/Bernays.
In any event, we must not declare to be "malformed" loops that are sim-
ply infinite, like the implication/realization loop for Perceptions (a) and (b),
or the trysting loop for Siegmund and Sieglinde. While exploring the Wag-
nerian loop, we investigated two ways to prevent an EVALuator from get-
ting trapped in the loop; the same expedients are available for the
implication/realization loop, and for a large class of similarly structured
loops. The first expedient is to apply higher-level parsing to the environ-
ment before attempting to EVALuate the perception-strings. The parser
would spot the loop and supply the EVALuator with a symbol like". . ."or
"(etc.)" to finish off with, once a certain number of trips around the loop
had made the recursive structure clear.
The second expedient is to break off EVALuation upon a trigger signal
from an external interrupt, thus: Siegmund's watching Sieglinde's watching
13. The reader who wants to explore the abstract theory of ConteXTs farther will be
interested in an extended study by Raphael Eric Atlas (1983). This work explicitly and sys-
tematically investigates the roles of varying musical contexts in building perceptions involv-
ing enharmonic relationships of all sorts within tonal compositions.
Music Theory 341
Siegmund's watching Sieglinde's wat

Hunding. Just like Hunding's tubas, ou
Figure l(c) could function as this sor
triad as a signal from outside the liste
implication of (b)'s realization of (a)
(a)'s implication denied by (c). If the
of Signal(c), the C-major eight-three
function, stopping EVALuation of th
and introducing the new Perception
implication of (b)'s realization of (a)'s
confirmed by (b-yes). The interrupt
chord of Signal(c), or for the C-majo
attractive theoretical conceit. The ha
acoustic signals external to perception-
of function from any they might c
within formal p-structures.
The species of function is different b
nal interrupt necessarily presupposes,
aspect of musical time that is not a me
temporal exigencies impinge upon th
like the metaphor of that model ver
keeping the EVALuator out of infini
not necessarily presuppose any musical
tener; the parser, along with the EVAL
the apparatus through which a listen
of space and time for the music percei
We return once more to the Basic F
LIST), and focus now upon ST-LIST,
made in some stipulated language L. De
as a "list" is only a formatting conven
example, be abstracted to represent
More generally, the language L might
notational systems with a symbolic t
nacular discourse like everyday Eng
stead or as well poetic sayings or writ
associations. It might involve gestura
cative systems not usually brought un
like writing down original compositi
passages. In Part V, I shall devote qu
composing and performing as means
shall more or less withhold that atten
Imagining our utterances or gestures
ming convenience, as I said before; i
342 David Lewin
More of-the-essence, and m

as modeled by the basic fo
tures of some kind. With th
that formal musical percep
tions," since each one emb
newly observed qualities of
model goes even farther in
broad sense, for the way in
on observation. Our sense
thereby necessarily involved
guage L, and our acquisitio
that the language L involves
we must be ready to consid
theoretical component, alo
may possess. To illustrate th
duced by a piano playing th
Calling that signal "Signal
upon hearing its last chord.
may say, "I hear a fourth-d
ST-LIST for a pertinent p, i
the bass F is four steps up a
cal context can be symboliz
Figure 5(b) is not projected
CXT for the perception un
"degrees" and to hear the
carries a long historical/cul
of the major scale, the Rule
listener may well be "unaw
Fig. 5.
14. This is one of the meanings for "apperception" given in The American Heritage Dic-
tionary of the English Language (New York: American Heritage Publishing Co., Inc.,
1969), p. 63.
Music Theory 343
Another listener may want to hear

context as a "subdominant." To use tha
able formal p, this listener will invoke
part of the CXT; Figure 5(c) would se
bass F of the EVent in question is
dominant below its theoretical toni
generated the G that lies the-interval-
the term "subdominant" means, when
cal shadow involving Continental har
nineteenth centuries, along with their
cal tonic of Figure 5 (c) is middle C, no
who invokes the "subdominant" cont
invoke another theoretical context,
relates the middle C of Figure 5(c) in
lower, the viola C that figures in the a
Figure 5(c) itself already presupposes
it assumes that the-interval-of-a-domi
tio 2:3; that interval is an octave sm
harmonic interval given by the ratio 1
fundamental-frequency-to-partial-f
octave-equivalence in this sort of co
work by Descartes, Rameau, and D'A
A third listener might perceive the f
preparation," thereby invoking the
with its Schenkerian shadows. The G in
not eventuate in the acoustic continuat
since the G in the theoretical context
at the end of Signal 5 by the listener
As a linguistic resource, the theoretica
less to do with acoustic signals than do
D of Figure 5(b); those are equally lin
listeners to make other kinds of perce
languages.
Part HI: A Passage from Schubert
To illustrate what the model of Part II can bring out in analysis, I shall
discuss some aspects of Schubert's song Morgengrufi that are characteristi-
cally addressed by that model. Figure 6 transcribes aspects of the strophe,
and gives the concomitant text for the first stanza. I shall assume that the
344 David Lewin
Fig. 6.
reader knows the piece well enough not to need more reminder of the com-
plete music and text.15
Figure 7 tabulates aspects of the formal perceptions I propose to discuss.
The perceptions are listed as pi through p9 in the left-hand column of the
figure. Each perception, following the model, involves a family of EVents, a
ConteXT for that family, a Perception-Relation-LIST, and a STatement-
LIST. EVents are located by entries in the second column of the figure; Con-
teXTs are located by entries in the third column. "Tonal theory" in some
15. The impetus for my discussion comes from a long unpublished essay I wrote on this
piece, and on the methodology of analysis, in 1974. Fred Lerdahl and Ray Jackendoff (1983)
generously credit the essay during their interesting analysis of the strophe in A Generative
Theory of Tonal Music (pp. 264-269). Their analysis illustrates excellently the resources
and powers of their theory. Since it uses extensively a different language L from mine, it
"perceives" things differently; otherwise I do not sense any major incompatibilities between
their readings and mine. Their methodological approach to ambiguous readings definitely
does differ from mine, both as expressed in the unpublished essay and as I shall develop it
over Parts III and IV of this paper.
The 1974 essay devoted a good deal of attention to the four-strophe form of the song.
Thereby it found a large-scale sense of balance about the temporal extents of tonic and dom-
inant in the song, a balance that resolves on a very high rhythmic level some of the discom-
fort Lerdahl and Jackendoff feel about those extents in the context of the-strophe-once-
around. I too feel that discomfort in that context. The discussion in Part III here will not
engage any context as extensive as even one strophe.
Music Theory 345
Fig. 7.
heuristic sense is understood as a component of each ConteXT. Selected

pairs from the P-R-LISTs are entered in the fourth column of the figure, and
selected STatements from the ST-LISTs are entered in the fifth column, by
reference to graphic examples that will presently be forthcoming.
The perception px in Figure 7, for example, addresses the EVents of mea-
sure 12 in the ConteXT of measure 12 (and tonal theory). So, in the row of
Figure 7 headed by "pi" on the left, "m.12" is entered in the second
column, the column of EVents, and "m.12" is also entered in the third
column, the column of ConteXTs. Nothing is entered in the fourth column,
the column of salient P-R pairs for pl. This inferentially asserts that it is not
crucial to hear pi in relation to other perceptions hereabouts, in order to
346 David Lewin
perceive "what we are hea

text."
What STatements can we make about "what we are hearing when we
hear measure 12 in its own context," understanding also a context of tonal
theory? Here is one: the measure elaborates g6 harmony, with a D in the
principal upper voice. That statement is entered in the fifth column of Fig-
ure 7, the column of salient STatements about plB To save space, the English
sentence is represented on the table by a reference to Figure 8.1, an example
which projects the sense of the statement in a compact graphic format.
Fig. 8.
Music Theory 347
Parts 8.1, 8.2, and so on of Figure 8 c

and so on of Figure 7. Our exegesis o
complete: pi perceives measure 12 in it
tonal theory; therein the events elabor
bass and D5 in a principal melodic vo
We hear quite different and various o
we hear the events of that measure i
contexts. The whole point of the prese
some precision the variety of formal p
a variety of formal CXTs for the E Ve
lated families of EVents. It is meaning
arbitrary - to invoke C major and its d
inant, when we are talking about perc
Whether or not we wish to utter perc
in-its-own-context is another matter.
tion pi a useful entity to have at hand
perceptions. To speak roughly in tradi
be able to refer to "the g6," when I w
degree-function or other function in a
footnote later on will develop the m
There are other kinds of perception
events of measure 12. For example, I m
in the accompaniment: only one pitch
come one-per-written-eighth-note. Bu
features of measure 12 in the context
comparing it to other events, especi
larger contexts. I would certainly not
sure 12 lies "in a high register" whe
context only. (By the italicized word,
logical context which makes me awa
poor bass that comfortably reaches a
tending to notice above, about the at
bass "in measure 12," are not featu
rather matters that involve how wh
Perception-Relations with what-I-
broader ConteXTs. Our model enable
these matters.
The perception p2, in Figure 7, en
EVents of measure 12, that is the con
ones attention is drawn to the attack-
paniment in measure 12, and it wou
statements about those matters on th
p2, one might then refer to perceptio
348 David Lewin
tures for measures 9-10, f

those perceptions relate to p
ments and relations for cov
On that example, the P-R-L
inclusion); this pair perceiv
terminal segment within th
LIST for p2 on the example
this pair notices a perceptio
ceives measure 12 as a challe
sures 9-12. The specific co
sures 12-13, or measures 1
ask why the context of mea
formed, considering the s
quarter of measure 11 and
tion. To address it in my ow
G root over all four measur
the principal melodic voice,
the entire vocal part of meas
(V-percept, questioning) t
root-percept, prolongati
(leading- tone-percept, de
state of affairs are covered s
On that example, the V ro
leading into their symbolic
function is depicted by the f
question mark and exclamat
sion about the denial of lea
clearly prolongs "dominan
Some critical readers may b
a confusing role for the even
phrase articulation very awk
harmony as a confusing min
own phrase boundary?" The
hope, making it clear why I
perception here. For the tim
modate very well the hos
fronted by p2. The model an
perceive in the music. We p
as the beginning of a new
mally" and, also, retains o
attacked).
Perception p3a, in Figure 7, hears measure 12 "as the beginning of a new
phrase . . . which . . . continues "normally'," and Perception p3b extends
Music Theory 349
the ConteXT for P3a backwards so a

The pair (p2,denial) appears on the P-
p3b's denial.
p3a hears the E Vents of measures 12-
sical phrase of measures 12-13 coinc
text; that is part of the EVent and p
and the harmony implied by the Con
STatement of Figure 8.3. In this Conte
of d minor, d minor, in this context (
there is no hint of C major tonality in
selves. Figure 8.3 suggests the way in w
"notVi\"
Beyond Figure 8.3, we will also want to image on the STatement-LISTs
for p3a and p3b various other STatements involving various Perception-
Relations on their P-R-LISTs. For instance we might STate, upon hearing
measures 12-13 in their own context, "Oh, now I hear where that g minor
six-chord is going." That Statement involves inter alia the pair (piincipital
inclusion), a pair on the P-R-LIST of p3a. The time-span of p3a continues the
time-span of pi, the span in which one perceives "that g minor six-chord."
We might also STate, upon hearing measures 12-13 in their own context,
"d minor is being tonicized." This STatement involves a mentally con-
structed d minor tonic, at measure 14 or thereabouts, upon which the
dominant-of-d in measure 13 will discharge. The mental construction is
symbolized in Figure 8.4, a sketch which pertains to a perception p4, a per-
ception of the tonicization satisfied in protension. The pair (p4,implication)
appears on the P-R-LIST for p3a, and of course the pair (p3a, realization)
appears on the P-R-LIST for p4.
Upon listening to measures 12-13 in the context of measures 9-13, we
might also STate, "Aha! So the g minor six chord is not a confusing minor
dominant of C major; it is rather iv-of-ii in a C-major progression that toni-
cizes ii." This STatement, the Statement of the Critical Readers, can be
imagined on the ST-LIST for p3b. The pairs (p2,denial) and (p3a,reinforce-
ment) accordingly appear on the P-R-LIST for p3b.16 In connection with the
Critical Readers, one would put on the ST-LIST of p3b additional state-
ments, e.g.: "There is a big phrase boundary between measure 11 and the
pickup to measure 12."
Perception p4 hears the d minor tonic that we expect to continue from the
EVents of measures 12-13. We discussed earlier the pairs (p4,implication)
and (p3a,realization) on the P-R-LISTS for p3a and p4 respectively. The re-
16. Figure 8.3 should technically be annotated some more to show how p3b, perceiving C
major tonality in its larger context, would analyze the harmony in the key of "C:ii," rather
than in the key of d.
350 David Lewin
cursive structure is by now

appears with diamond-shap
in protension only, so far as
8.4 STates inter alia that the events of measures 12-13 are about to dis-
charge upon a constructed d minor tonic event.
The mentally constructed d minor tonic here interrelates with a d minor
harmony we heard earlier. That harmony was tonicized via a fleeting C# in
the vocal line of measure 8, the first chromatic note of the song. p4 thus
expands and elaborates upon an earlier perception of tonicized-d. The pair
(earlier d tonicization, elaboration) appears on the P-R-LIST for p4. It
would be more exact to introduce in this connection a new perception p4a
whose ConteXT includes measure-8-in-retention as well as measures 12-13
and 14-in-protension.
Perception p5 models our effort to make sense of tonicized-d-minor (p4)
following directly upon a prolonged dominant-of-C perceived over mea-
sures 9-11. We hear that the melodic D5 of measure 13, the D5 which also
figures as the diamond-shaped goal of melodic tonicization in Figure 8.4,
prolongs the D5 where the voice signed off in measure 10, a D5 introduced
and cadenced upon as the fifth of the dominant harmony there. And the
other diamond-shaped note of Figure 8.4, the D4 which is the fundamental
bass for the tonicization there, can be heard as part of a bass arpeggiation
within the dominant harmony. Hearing these things, we expect that the
tonicized d minor of p4 (of Figure 8.4), having arisen in a larger context as
an elaboration of dominant harmony, will again return to dominant har-
mony. The statement made by the preceding sentence is elaborated and
symbolically sketched in Figure 8.5, to which reference is made on the ST-
LIST for p5. The diamond-shaped notes on the example portray contextual
elements we construct protensively upon hearing the EVents of measures
9-13. A slur extends to the right of the melodic D5 within the ii harmony;
that symbol suggests that we mentally prolong the melodic D5 through the
protensive dominant-of-C which follows; the slurred D5 thereby resumes
its earlier role as fifth-of-a-dominant-harmony.
Figure 8.5 embeds Figure 8.4 within its middle, and the larger progres-
sion "makes good sense" of the smaller. Our model reflects these observa-
tions by putting the pairs (p4,medial inclusion) and (p4,reinforcement) on
the P-R-LIST for p5. That P-R-LIST also contains the pairs (p3b,reinforce-
ment) and (p2, virtual annihilation). That is to say, p5 (Figure 8.5) continues
and mightily intensifies the denial of p2 that began with the construction of
P3b. On Figure 8.5, the bracket, the parentheses, and the filled-in noteheads
suggest how the g minor six-chord is here perceived as completely
/brairds-looking, inflecting a subsequent (protensive) d minor harmony;
in this perception, the g minor chord has no direct prolongational relation
Music Theory 351
to the dominant harmony that precede

the g minor six-chord, perceives that-
Perception p6a addresses our hearing
the ConteXT of measures 12-14, conf
tion p4. (p4, confirmation and elaborat
Figure 8.6 confirms and elaborates Figu
and (c). Stage (a) shows the protensive
the inner voices of the mentally const
flects the top voice of stage (a) by a (p
voices of stage (b), leaving the passing
replaces the A natural in the harmony
variant A flat. The bass, alto, and so
by the actual acoustic signals of me
shaped D4 of stage (c) is a mental con
the d minor tonicization upon the harm
representative for a d minor root fun
perceived (by perception p6a!) as an i
chord, the d being mentally constru
14.
Perceiving the C5 of Figure 8.6 as
structed d harmony in a context of
perception protensively, a perceptio
plished. Perception p6b puts 6a togethe
BH to come in the melody and G3 to c
sure 15. (p6b,implication) appears on
specifically implied by p6a as follows.
14 is perceived by p6a as dissonant, a
should therefore resolve, we expect, to
in the melody of acoustic measure 14
nant, a passing seventh. It should ther
melody of measure 15. We expect Bl> r
17. When I refer to "the g minor chord," I d

structed pi to reference such a mental object
measure-12-by-itself, enables me to refer to a "
inant of C major" nor a "subdominant of d
tively, to say that ps hears the g minor chord
dominant of C. If I substitute "the subdominant
ing sentence, I make p5 hear tautologously in m
nant of C" for "the g minor chord" in the sa
erroneous fashion.
Classical theories of consonance and/or triadi
claim, that I can perceive "the g minor chord"
ration" and/or "resolution" in any larger cont
352 David Lewin
ConteXT for p6b includes a

a theoretical-psychological
plicitly noted in Figure 7. W
include Figure 8.6 follo
hypothetical-measure- 1 5 . 18
Perception p6b has on its P
on its P-R-LIST a pair (p7a, m
same EVents in the same t
measure 14 in the context
But p7a expects quite differe
ized by Figure 8.7. As the ex
to step down from measure 1
however, does not expect t
awaits B natural, not B flat,
tions p7a disagrees with p6b
modify each other. The pair
P7a-
p7a expects B natural, not B flat, and p7a is not concerned with maintain-
ing a d minor context, because the perception is listening for sequential pat-
terns in its ConteXT. This feature of the perception is symbolized by the
annotation "(seq.)" in the CXT column of Figure 7. Perceiving the acoustic
signal of measure 14 in the context of measures 12-14, p7a recognizes that
the text of measure 14 is analogous to the text of measure 12 in a context
where a complete sentence underlay measures 12-13; p7a also recognizes
that the acoustic harmony of measure 14 has the same intervallic structure
as the acoustic harmony of measure 12. Accordingly, p7a expects that mea-
sure 15 will continue from measure 14 by analogy with the way measure 13
continued from measure 12: measure 15 will finish the sentence begun in
measure 14, and measures 14-15 will project the progression iv6 V in "c
minor" just as measures 12-13 projected the same progression in d minor.
p7a in its context (Figure 8.7) thus perceives p3a (Figure 8.3) becoming
expanded sequentially. The pair (p3a,sequential expansion) appears on the
P-R-LISt for p7a. Figure 8.7 asserts a local tonic of c minor, not C major, for
measures 14-and-expected-15. That is because c minor, not C major, is the
18. Idomatic harmony in d minor for the Bt-over-G, consistent with the level of
complexity introduced into that key by stage (c) of Figure 8.6, is quite conceivable. For
instance, the Bt-over-G could be harmonized by iv7, giving rise to an elaboration of p6b
through the following progression: iv6(ml2), V(ml3), iji (ml4 as stage(c) of Figure 8.6), iv7
(harmonizing the protensive Bt-over-G at hypothetical-measure- 15), V^, i6, and so on.
The harmonic exercise is not sheer pedantry. Our "language L" includes the discourse of
traditional tonal theory, and the urge to work out a reasonable harmonization is the urge to
show that the STatement of pa> involving Bt-over-G in d minor is in fact a grammatical (i.e.,
possible) construction in the language L.
Music Theory 353
literal sequential analog for the d mi

knows nothing of any larger context i
has a question mark on the progression
13 to its "c minor" of measures 14-15
that makes the progression "logical," e
dential dominant of d in measure 13 to
would not help the "logic" much to inv
problem is that p7a perceives the harm
of c (or C), and that function has no c
the dominant of d which precedes it.
p7a thereby modifies p6b in yet anoth
lematic about the progression from m
V of d progressing very logically to a s
as indicated in stages (a)-(b)-(c) of F
tions of p7a and p6b in this connection
ble emploi brought into our presen
acoustic signal of measure 14 signifies
p6b - actually already in p6a - the sam
perceptions p7a and p6b that carry t
contexts that are coextensive in curs
measure 14 in the ConteXT of meas
15.
To say these things about the two dis
about p7a and p6b, is very different f
acoustic object, "the chord of measur
a d chord "at the same time." I put "is
marks to draw special attention to t
parlance here, which speaks as if th
extends were the only temporal frame
essing, and interrelating the two m
much more to say on such methodolog
Perception p7b notices that the acou
12-14 through acoustic measure 15 a
p7a - to a dominant of c (or C) with
Thus (p7a,confirmation) appears on
measure 15 denies the protensive B fla
pected) measure 15. Accordingly, the
LIST for p7b. p7b does not, however, de
to confirm the expectations of p5: p
notes of Figure 8.5 are in fact eventua
sures 14-15 just as p5 had imagined
represented by the acoustic signal of
354 David Lewin
tion of p6a (Figure 8.6). This

pair (p5, confirmation (via p
In confirming both p7a and p
the f chord in measure 14 of
p6a. Our model has no proble
measure 14 of p7a" and "the
objects in our model; "the d
14 of p5" is yet another objec
acoustic signal, does complet
sures 12-13 according to the
subdominant of the local to
sure 12 is a subdominant of
p7b also perceives that the pr
fact come to pass over meas
(Figure 8.6) that allows the si
mony. There is no logical con
p7b perceives one object as "b
Perception p8 puts the At-
enough context so that the
tion of the At-G in the bass
relationship. Figure 8.8 mak
the two approaches to G-in
measures 9-15 we identify
labels the two dominants, p
suitable Perception-Relation
8.5) heard the span of mea
mony in C major, and p8 (F
ported by it.
Figure 7 does not include o
measure 9. Intuitively, we can
these intuitions could be r
STatements if we wanted to
notice, for instance, that t
notes from the voice entranc
bass and/or the harmony sta
ment texture). We notice th
chordal chromaticism in the
context of measures 5-9 (o
bass into the At of measure
motion of the bass during m
measure 9 is the first large-s
it is the melodic and rhythm
sures 5-11. We notice that t
Music Theory 355
prolonged above the At-G gesture in th

the vocal line in the context of mea
notice that this vocal F is dissonant ov
9. We notice that the vocal F sets a sub
ter, we shall notice curious thing
resolutions of that F in the context
about the dramatic irony of the tex
makes the contrary-to-factness of the
if something were the matter" - as if
Such observations, and others like t
LIST of the earlier At-G perception t
thereby p8 is itself indirectly enriche
tions, pursuant to these matters, cou
we shall now certainly notice in the
"dominant" EVents of measures 9-1
measure 9 by a series of questions: each
sures 9-15 ends with a question mark
high F of the voice in measure 9 essen
an understood "questioning" domina
dominant elaboration at hand. This per
signal of measures 16-17, where the hig
matching rhythmic motive, as the la
to a large-scale tonic and the subjuncti
bring out the last-mentioned relatio
text stresses in setting the rhyming v
war' dir (was geschehen)" and "so mu
dir was geschehtn" and "so mutë ich w
19. N.B.: the youth does not "go." Rather h

remainder of the strophe. Even then he does no
My unpublished typescript analyzes and inter
tant E-D-C round between voice and piano,
analysis, the round expands into recurrent lar
in one strophe, through the vocal D of mlO-12
in that next strophe. The E of ml6-in-that-next
E in the foreground round reasserts itself, in
foreground round. The music for the strophe
round-motive within each strophe. The youth
mufi, has been attached to the act of going,
ground E of mufi, m 16, is linearly connect
gehen, m 1 7. The background connection is n
one strophe; the background line emerges onl
wieder through the D of measures 10-12. The m
strophes go around, the less accent we hear th
of measure 9; the more structural weight, corr
12.
356 David Lewin
Figure 8.8 indicates succin

phe, measures 9-15, is also
large-scale dominant is the
much as it is with contrary
measure 9. Consequently,
that projects a more chrom
than we have hitherto exam
LIST for p8.
Figure 8.9 sketches a symbo
measures 9-15 in their ow
tensive resolution to follo
things, the harmonic functi
iv-of-ii after all, but rather
According to the figure at
sixth degree of a local tonic
the A natural in the bass of
Rather, the Bt in the bass
altered member of a large
the A natural in the bass a
chromatically away from it
behave as scale degrees borr
idiomatically to 5, and fla
flat-6 at measure 14 (via th
In Figure 8.9, a slur binds t
sion of the "Mullerin" motive from the voice in measure 6. In the fore-
ground, the music moves from tonic harmony at the vocal entrance (mea-
sure 5) to dominant harmony (measure 6); the foreground dominant is then
prolonged by the Mûllerin motive. Just so, the music moves on a larger scale
from tonic harmony at the beginning of each strophe to the big dominant
harmony at measure 9; p9 then hears that big dominant prolonged by the
(minor) Mullerin motive as depicted under the slur of Figure 8.9.
The pair (p2,confirmation) is a characteristic member of the P-R-LIST for
p9: what-p9-perceives includes the perception that p2 does (did) in fact make
sense, even though it was (is) "denied" by p3b and "virtually annihilated"
by p5. We do not have to have recourse to "posthumous rehabilitation"
heje. p2 is not necessarily "really" dead, just because p3b and p5 honestly
perceived it as dying and dead. We are now somewhere else, perceiving
something else along with p9. To put the matter more elegantly: p2, P3b, P5,
and p9 are not all cohabiting the same phenomenological place at the same
phenomenological time. They are different objects (or acts) in different
parts of phenomenological space-time, exercising a variety of interrelation-
ships as reflected in our model by a variety of P-R pairs. I shall discuss the
methodological point at greater length during Part IV. Meanwhile we can
Music Theory 357
note that p9, in confirming p2, denies

the P-R-LIST for p9. p3b, it will be re
"Aha! That g minor chord is not a con
of-ii in a progression tonicizing ii."
E Vents of measures 12-13 in the C
denies p3b by STating: "Doch, doch! Th
dominant of C, a questioning, doubt
giating the G root which set in at meas
the E Vents of measures 9-15 in th
protensive measure 16. In thus deny
contrary-to-fact contrary-to-fact con
ined earlier: "as if something were wr
The pair (pg,support) appears on the
tions mutually reinforce each other
(p5,qualification). The qualification-r
be analyzed into a number of compone
that both perceive measures 9-15 as
in C. (Compare Figure 8.9 with Figur
to the manner of the elaboration; thei
fering symbols that appear in Figur
symbols which have been amply discus
review in this connection our earlier e
lated" p2. p9 takes a longer view than p
not happen" or "was wrong," any m
wrong" when p5 took a longer view an
deny or otherwise bad-mouth some o
experiences in this sort of situation w
Part IV: Methodology
We have already started to note and

enables us to bypass certain false dich
omies that arise when we implicitly
discussing one phenomenon at one
time, when in fact we are discussing m
locations. We can review the point b
shown in Figure 9.
When we contemplate such politica
spectively or in debate with other a
symptoms of a deficiency in traditi
forts arise whenever we make, about
of syntactic form, "The X is. . . ." T
358 David Lewin
Fig. 9.
Figure 9(a) by saving, "The harmony of measure 12 is . . .," we are already

falsely constraining our musical perceptions by implicitly asserting that
there is one phenomenological object called "the harmony of measure 12,"
and we are also constraining our perceptions by saying of this object that it
"is," putting it at one location in one present-tense system that renders
falsely coextensive a number of different times: the historical time in which
the piece continues to exist for its listeners and performers, every time in
which an acoustic signal projects the score of measure 12, the time during
which a listener may be former and processing perception pl5 ditto percep-
tion p2, ditto perceptions p3a and p3b, ditto perception p4, or p5, or p8, or p9,
and so on, and the time in which I am now writing this sentence, and the
Music Theory 359
time in which you are now reading it.

to the differences among what-happ
part-of-pi, ditto p2, ditto p3a, and the
within the model, not one object called
wise, our model makes us sensitiv
structures pi can occupy different me
and pj impinge upon us at the same me
them to do so separately; indeed, pi m
R-LIST while pj simultaneously had
cussed earlier how EVALuation of the
nation during this time: a higher-le
loop prior to EVALuation, arranging f
ternal to mental processing could in
cording to some pre-structured config
Any phenomenological theory should
sity for conceptually distinguishing
like those mentioned above: the tim
think about it while writing this artic
as you now think about it while readin
the-clock, the time in which measur
create a pertinent acoustic signal by ce
time in which measure 12 "is" when a
acoustic signal via certain psycho-phys
ing to a recording at home - for the fi
and so on. Each of these occupational c
mental constructs for perceiving the
gist will not assume a priori that the
morphic. Indeed the transformations
tem into the others should be presum
reflect both the autonomy and the int
I have called my false dichotomies
into the position of voting for a slate
in adversary judicial proceedings, as w
wrong but fantastically wrong. My
rules for analysis: mistrust anything
20. The metaphor of an obligatory interrupt

we sense the psychological immanence of the
tells us "it is time to be moving on" from the
measures 9-15, an impasse portrayed nicely in
The downbeat moves us on by its high-level r
vocal E, and its new verb mu/i.
360 David Lewin
impression you have once for

listen any more to music th
one thing going on (or two
things). The false dichotomi
find the phenomenology of t
omies without abandoning r
The dichotomies illustrate we
wont to lay for the ear, not
straints of this or that theor
more in connection with ou
sloppiness in using the word
is our too facile recourse to
sentational modeling, a recou
the useful metaphors of the p
To illustrate the treacherou
plane, one need only glance
within a portion of a Euclid
barline of measure 12 in the l
reference a unique "point" of
dinate and a unique horizon
of the plane. The geometric m
cious idea that there is one un
an object which impinges up
the time in which the B fla
unique spatial location is pro
the B flat notehead-point on
lacious sense that only one m
time-system, is prompted by
notehead-point in the sam
dimensional representation
understanding, a certain cr
harmony of measure 14" is
adjacent points in the plane
bounded by the vertical lines
as a connected region in Euc
is the projection of that regio
is also connected; it suggest
ciously suppose our harmon
metaphors, we conclude "lo
that "it" [sic] cannot be both
time." And so we begin tryin
tual phenomena [sic], not re
for the analytic task at hand.
Music Theory 361
Our model helps us to abandon, along

leading expressions of the species "m
These utterances help us to dismiss i
quential. The linguistic mannerism cre
and - worse - into our mental habits
music. When we fall into such discours
measure 12, apparently an inconvenien
be "naught but" the beginning of a
then that event, on a yet larger level, t
elaboration of v-of-C after all. In this
all a malformation: "The ... g minor
now also notice the way in which
"merely" sneak in, so that we are enab
tions at the expense of others, again as
expressions tell us not to explore furth
gripped us; the parlance violates my
True, we will modify our perceptio
tending their P-R-LISTS, creating ne
"deny" old ones, and so forth. Perhap
of mental surgery upon them. We ca
too, during the time in which we com
is perfectly reasonable. Indeed, our
the process at work, for example, in
modifying p2, both as we listen to the
more deeply. What is not reasonable
bad-mouth perceptions we are comin
that underlies such an urge is a good
logical business at hand, that the atten
away from something which the mi
unexplored. One thinks of Freud's Zu
21. Freud (1955, pp. 64-65). Joan Riviere (1

as "forced back" and Unterdruckung as "supp
to Freud as Verdràngung = "repression." Freu
ysis of everyday errors (Fehlleistungen). On t
that ". . . a suppression (Unterdruckung) of a
dispensable condition for the occurrence of a
has told us that the speaker may or may not be
case ". . . it has been forced back (zuruckgedr
vert the idea into speech and then . . the tend
itself against his will and gains utterance . . .
(Versprechen)."
Earlier, Freud classifies many fehlleistung
(1955, p. 18; Riviere, 1952, p. 29). One of the
event. The one who verspricht sich typically
to say . . ." The one who verhorst sich typic
that what I really heard was. . . ."
362 David Lewin
When we are using words li

ceptions, we are likely to c
portant." Our perception-m
They are suspect because the
"unimportant" or "less imp
tance" is too imprecise a w
casually suggests unspecified
been stated explicitly and
same time - the word can be
syntactic system" or "rank i
compounding each other, c
priority with aesthetic val
when one is using Schenke
point is worth two example
The first example is litera
clearly appreciate in a literar
tion and aesthetic signific
stares at his bloody hand, w
Will all great Neptune's ocea

Clean from my hand? No. Th
The multitudinous seas inca
Making the green one red.
- Act II, Scen
About the last sentence of

legal dichotomy. On the on
the sentence, is thereby "mo
which is ("merely") an adje
other hand, "multitudino
"This." "Multitudinous" is a f
the exception of "Neptune
syllables; it is also a bomba
with the exception of "Nep
words; it is thereby the fir
surge up at "rather," clim
dine," and subsides through
of Saxon monosyllables, "t
peated Saxon monosyllabic
polysyllabic Latinate surge
echoing off into "Making
So should we then vote f
"This"? No. We are not vo
called "the most important w
Music Theory 363
of such a mental object the idea that it

tion; finally, "importance" is a useless te
to make it reference two very different
usage, "importance" refers to height on
other usage, the same term refers to c
plex poetic phrase.
Naturally we are more interested in S
dures, than we are in the fact that his te
English syntax. That is not at issue he
our possibly confusing the fact that Sh
manner in which he did so. The fact doe
of English-users whose texts interest us
guish him, and cannot be separated fr
So, while it is not interesting that the
abstract English sentence, it is intensely
"This," used a a substantive noun, the
whose compositional structure we have
the multitudinous seas. The grandiose cli
whose significant ConteXTs and Perce
excised from contact with the sentence a
lar with the opening and the subject of t
dramatic poetry by impatiently twiddlin
through the less impressive but unfortu
the more magnificent and heaven-storm
An actor who behaves as if we did res
For no audience can possibly miss the
while an untrained or insensitive actor ca
sage by not sufficiently exploring and
word "This" works for the poetry. Speci
if the ocean might wash "this blood" c
"No." Then he begins a new thought w
that "This" is an adjective, and that the
Or, if the actor makes us feel that "Th
pose that it stands for "This blood." Bu
do not get the word "blood"; instead we
the blood never wash off the hand, e
hand" have fused into a compact and i
union for which the appositional form i
Macbeth's question concerned three dis
and the hand; his answer condenses th
and the multitudinous seas. Hand and
guilt fuse into one for the character. T
blood into the blood-hand creates a tight
364 David Lewin
released by the expansion of

dinous" and "incarnadine."
by noticing that I was insti
unclenching it, gradually sp
possible extent, as I intone
shall say more about such p
So, "This" is indeed the d
etic subject of Macbeth's d
sentence, a subject that beco
this hand, and this act all in
ble "This," the very antithe
dinous." Which word sha
reader will by now have tak
cal expression here, and it p
vote between English sent
My second example involv
a musical analysis. In that
signs to a certain event a sy
ignoring how much "more
is somewhat analogous to
the Schenkerian reading a
protests about our calling
"multitudinous" is clearly
test the pertinence of Schen
ally protest English syntax,
on.
The music is Handel's familiar setting of the carol Joy to the Wor
the Schenkerian reading is by Allen Forte and Steven E. Gilbert
182-183). In connection with this reading the authors bring up a
Schenkerian syntax: it is not possible to assert a well-formed Ur
starts on â, e.g. at the word "Joy"; a well-formed Urlinie can, h
start on 5, for example, at "world." This Urlinie can descend fr
with appropriate support from an Ursatz while Heaven and angel
contrast, there is no syntactic support from any well-formed Ursat
putative descent from è (Joy) to $ (world) within an Urlinie that m
on 8. As the authors put it, "the steps between & and $ are . . . over
harmony; this contrasts with the full support given the slow descen
to Î over the last seven measures." The melodic gesture of è -to-
tonic-pedal is described by the Schenkerian term, "Leerlauf."
The authors' analysis might at first seem utterly inconsistent with
mendous accentual impact of the musical attack on "Joy." Is not
liant impetus the most striking thing about the piece? And in that
can one presume to assert that "world" is "more important"? Th
Music Theory 365
recognizes, I hope, the analogy with

view, the Schenkerian reading does not
tant" than "J°y"; rather it asserts th
formed Schenkerian Ursatz, much as "
English sentence in Macbeth's speech
The critic will go on to demand of an
nence of such grammatical observat
works at hand. I have tried to produ
peare passage in this regard, and I shal
Handel piece.
There, the Schenkerian syntax sugg
age. Handel's joy is cosmic. It fills the
harmony. It does not move from on
does not leave its heavenly orb and t
ductive medium, for example, throu
human time, like an Urlinie. Rather
suffuses all things, the world in par
nance. The Leerlauf transmits the radi
heater, through empty space. No con
the world, here on 5, can we set about
that conducts one event to another th
and now can a Schenkerian Urline ge
voice of an Ursatz.
These metaphors belong squarely wi
tinguishes the (harmonic) Music of
Music, a tradition extending back t
Zarlino carried the tradition into a
would have been sensitive to it at least
son. And Schenker's mature theory re
other form: he presents his Ursatz as
idealized human voices, of a categor
by Nature.22
Whatever the relevance or irrelevanc
metaphor gives us a poetic reading t
syntax presented by Forte and Gilbert
ral urge to sing the word "J°y>" in
22. Relevant material from Boethius is tra

Gioseffo Zarlino's ideas are succinctly discus
1558/1968, pp. xviii-xxiv). An eighteenth-ce
Mattheson (1739, p.6); Mattheson's prose par
The article on Mattheson in the New Grov
friendship and mutual influence of Handel
about the Ursatz appear in Free Composition
366 David Lewin
radiantly as possible. This re

and conductors, as a way of
to the vocal problem posed b
steam at this point in the m
first four notes. No self-r
something special, just as
"multitudinous" is someth
some coaching with the deli
actors can use some coaching
it will surely help a chorus t
lishing a harmonic resonanc
continuing world, rather th
that begins at joy and end w
A point should be taken u
much freer to reject Schenk
in which to make perceptu
reject traditional English gr
is indeed the case. But, wh
this or that music theory
language L in which to ma
free to accept or reject the n
which we can discuss thin
"strong beats," "beats," etc.
ority of any kind to such t
been discussing must come u
fuse the assertion of system
tural downbeat, with the
event is "more important" t
I have suggested that our
suppressing certain "less i
that, can be a psychologic
priate or disturbing, a kin
own experience, I have alway
this assumption, whenever I
Zuruckdràngung is its only
dency to confuse arguments
sitions in the language L, or
osition over another in th
attributing relative value or
lies at hand from our discus
we can consider Sentences
(a) "The Urlinie for a pertin
(b) "The Urlinie for a pert
Music Theory 367
(c) "The Urlinie for a pertinent Urs

Sentence (a) is false; it can be demons
ventions of Schenkerian language - c
"Ursatz" - along with an appeal to
score of the piece - at the noteheads, a
listening; they require perception on
read English or German text and mu
arguments. A hypothetical "percepti
pa = (8 of "J°y"> whole piece, . . , S
dismissed as "malformed." Criticism of
at some vague and wrong-headed no
opening "Joy" seem "too important."
verifiable fact that Sentence (a) is not
guage L.
Neither Sentence (b) nor Sentence (c) is false, in the sense that Sentence
(a) is. However, the logical conjunction of (b) and (c) is false in that sense:
within the language L, specific rules tell us that the sentence "(b) and (c)" is
false. The truth of (b) logically entails the falsity of (c), and vice versa.
So much for the logic of sentences (b) and (c) within L. When we con-
struct corresponding perceptions pb and pc, however, we are not within L.
We cannot call either perception "true" or "false," even conditional upon
the other. Both perceptions are well-formed since, inter alia, neither Sen-
tence (b) nor Sentence (c) is in itself false (or malformed in L). The matter
can stand some elaboration. Let us define the perceptions as pb = ($ of
"world," whole piece, . . . (pc,denial) . . ., Sentence(b)) and pc = (3 of
"and," whole piece, . . . (pb,denial) . . ., Sentence (c)). Since Sentence (b) and
Sentence (c) are mutually exclusive within L, it is impossible to perceive a
well-formed thing called "pb-and-pc" at one-and-the-same-time in one-
and-the-same-place. But our model does not propose that we consider pb
and pc to be in the same phenomenological place at the same phenomeno-
logical time. Quite the contrary: the model enables us and indeed urges us
to articulate different locations for pb and pc in phenomenological space-
time. Thus a political/legal dispute over "pb? or pc?" is out of place.
We can rationally argue in a political/legal way over grounds for prefer-
ring Sentence (b) to Sentence (c), or vice versa. We can point to aspects of
Schenkerian theory, and/or aspects of the score, that make one or the other
sentence preferable. Thus, to support a preference for Sentence (b), we
could point to the mini-descent from 5 to Î, with full Ursatz-type support
from the bass and the harmony, that shapes the first cadence of the music
(". . . world, the Lord is come"). Or we could point to the lack of bass sup-
port under the acoustic attack of the 3 at "And heav'n"; the entrance of the
bass voice is delayed so as to support the agogically accented % that follows
shortly after, on "sing." And so forth. We can also carry out such an argu-
368 David Lewin
ment by invoking the text

reader's enthusiasm for Sent
volving the Leerlauf. An op
asm for Sentence (c) by al
though, could argue for pref
One either has the perceptio
(b) and do not experience per
who already experiences p
both perceptions, at differen
tening process, would find t
P-R-LISTS involved.
Another example will help us distinguish the logic of sentences in L from
the logic of perceptions in our model. Suppose any common theory of tonal
harmony as a component of a language L. Consider two sentences within
that language. Sentence XDY reads, "Event X functions harmonically as a
dominant of Event Y," and sentence YDX reads, "Event Y functions har-
monically as a dominant of Event X." Clearly each sentence is well formed.
And, just as clearly, the sentence that is the logical conjunction of XDY with
YDX must be false. If XDY is true under a certain allowable substitution
for X and Y, then YDX must be false under that substitution, and vice
versa. We are assured of this without even considering any musical score,
let alone doing any listening. Now let us turn out attention to Figure 10,
which sketches a cadence by Siegmund just before the last passage sung by
Sieglinde in Act I of Die Walkiire.
It seems at first that we have at hand here a perception-structure that in-
volves exactly the sentence just branded as false, that is, the logical conjunc-
tion of XDY with YDX. For the X event is evidently perceived to resolve as
a dominant seventh into the Y event at the moment of the cadence. (The
textual alliteration on the vocal Gs amplifies the effect.) And, apparently at
the same time and in the same place, we perceive the Y event as a dominant
Fig. 10.
Music Theory 369
to the X event that immediately prece

returns, at the end of Sieglinde's pas
quite like the one that contains the X
one perception or the other, in order
But our difficulty is only apparent. T
improperly reified one percept (as
one percept called YDX; the confusi
speaking that makes us believe we h
time," so that we try to imagine one c
ception of both-XDY-and-YDX." Our
confusions, by articulating a variety o
phenomenological space and time. T
ceptive cadence (Figure 1) will serve
with Figure 10, we can formulate the
among others.
q! = (Event X,
Figure 10 up the pause,
. . (q2,implication) . . .,
V-of-an-expected-I)
q2 = (Event X,
Figure 10 ending with G\ instead of e#6,
. . . (q^realization) . . .,
cadential dominant)
q3 = (Event X,
Figure 10 without the bass and figure for the event at the end,
. . . (q2,confirmation) . . ,
cadential dominant)
q4 = (Event Y,
Events X and Y,
. . . (q5,implication) . . .,
dominant of X)
q5 = (Event Y,
Events X and Y plus a protensive X' that projects D7 harmony,
. . . (q4,realization) . . .,
dominant in transit from X to X')
q6 = (Event Y,
Figure 10 and on through Sieglinde's passage,
. . (q5,confirmation and elaboration) . . .,
structural dominant in transit from Siegmund's cadential G:V to
Sieglinde's)
370 David Lewin
AU these percepts are well

formed:
q7 = (Event X-and- Y,
Figure 10,
. .'XDY,YDX,. . .)•
The STatement-LIST for q7 is malformed within the language L, so q7 is

malformed as a perception. There can not be any phenomenological place
and time where q7 "is." In contrast to that, one observes how Siegmund's
protensive G^ event, involved implicitly in qu q2, and q3 as that-of-which-
X-is-the-dominant, is a different phenomenological object from Sieglinde's
Y event, an E raised-sixth, involved in q4, q5, and q$ as that-which-is-a-
dominant-of-retained-X.
One more example will focus our attention even more sharply on the dis-
congruity between the logic of sentences-in-the-language-L and the logic of
perceptions in our model. A well-known drawing outlines a Gestalt that
can be seen as either a rabbit or a duck. In this connection we can construct
a visual percept r, perception-of-rabbit, and a visual percept d, perception-
of-duck; evidently both r an d are well-formed and relevant. One can make
verifiable statements on a STatement-LIST for r: these are ears; here is the eye;
and so on. One can make verifiable statements in the same language about d:
this is the bill; here is the eye; and so on. Present-day computer pro-
grams (at least in theory) could recognize such features of the drawing, find
them well-formed, and tell us both "Here is a rabbit" and "Here is a duck"
according to stipulated L-criteria for uttering those remarks. However,
though "I see a rabbit" and "I see a duck" are both valid perception-
utterances, "I see a-rabbit-and-a-duck" is not; at least to my knowledge no-
body ever sees both animals at the same time (in the same phenomenologi-
cal place). We would not want our computer to tell us "Here is a
both-rabbit-and-duck." We would want the machine to know there is no
such animal as a both-rabbit-and-duck.
Thus, even though "I perceive rabbit" and "I perceive duck" are both
valid perceptions, we cannot infer the validity of "I perceive rabbit-and-
duck." We can infer "(I perceive rabbit) and (I perceive duck)," but only
under a very special logical interpretation of the conjunctive "and": the
meaning of the conjunction here does not imply "at the same time in the
same place." That is, we must understand: "Somewhere I perceive rabbit
and somewhere I perceive duck." In this linguistic form, the operator
"Somewhere I perceive" does not distribute over conjunction of its argu-
ments: "((Somewhere I perceive) (thing 1)) and ((Somewhere I per-
ceive)(thing 2))" does not mean the same as "(Somewhere I perceive) ((thing
1) and (thing 2))." So, in particular, "((Somewhere I perceive) (rabbit)) and
Music Theory 371
((Somewhere I perceive) (duck))" i

ceive) (rabbit and duck)" is not only
formed, since rabbit-and-duck is no
language.
In just the same way, "((Somewhere I perceive) (XDY) and ((Somewhere
I perceive)(YDX))" is loosely speaking valid, if we mean by Y here "some-
thing I infer from the acoustic signal during the indicated clock-time." But
"(Somewhere I perceive) (XDY-and-YDX)" is not valid: there is no such
thing as XDY-and-YDX in the language of harmonic theory.
And in just the same way, "((Somewhere I perceive) (a S Urlinie)) and
((Somewhere I perceive) (a 3 Urlinie))" is not malformed, though I do not
myself assert it of the Handel composition; however, "(Somewhere I per-
ceive) (both-a-S-Urlinie-and-a-3-Urline)" is malformed, since there is no
such thing as both-a-S-Urlinie-and-a-S -Urlinie in the language of Schenker-
ian theory. If we wanted to, we could develop a post-Schenkerian theory in
which a piece could logically have more than one "Urlinie." Using that new
theory as a component within a new language L, we could then render the
conjunctive perception well formed. Presumably we would change the vo-
cabulary of our neo-theory and our STatements, since "Ur" no longer
seems appropriate.
We should certainly be willing to alter our theoretical discourse in this
way, whenever a certain mass of perceptual experience leads us to believe
that the alterations might enable us to articulate valuable analytic insights.
But we should think long and hard before subjecting a received theoretical
discourse to fundamental modification. In changing the language, we risk
losing our ability to express some of the features that characterize what is
problematic about a tricky perceptual situation. For instance we could cre-
ate a new word "dubbit," defined as the Gestaltist drawing recently dis-
cussed; by changing my language in this way I could say "I see a dubbit"
and thereby "solve the problems" involved in saying both "I see a rabbit"
and "I see a duck." But it is just the "problems" in the perceptual situation
that we find characteristic and interesting, worthy of extended analysis; our
linguistic expedient has turned the interesting phenomenon into a hum-
drum affair. So you see a dubbit. Who cares if you see a dubbit?
We should generally take the same methodological tack when some of
our perceptions about a piece of music involve STatements that are logi-
cally incompatible-in-L with other STatements that we articulate in con-
nection with other perceptions. In such a situation, we should generally
want our analysis to convey the characteristic multiplicity of the percep-
tions involved and the characteristic incompatibility of their assertion in-
the-same-place at-the-same-time. The rhythm of the dialectic thus engaged
will be a significant aspect of our rhythmic response to the music.
Indeed, one of the most interesting features of our model is the way in
372 David Lewin
which it implicitly engages

tated. The model suggests, f
sage from Morgengrufi invo
considered as "rhythmic,"
Pi) P2) p3a> and so on come i
P-R situations, recede from
dues, all the while the aco
model is at present not wor
cision compared to tradition
the Renaissance. To provid
reasons as well, it needs to h
various sorts of time, som
perceptions pi are formed, i
sorts of time might well in
Euclidean time within whi
sense of classical physics
marked by events that are
(e.g., neuron firings or patt
phenomenological time m
"measures" or "breves" or
certain pieces that posit su
model within which some
structures takes place, possi
nipulates configurations of
time within which EVALuat
the processing system fro
listed.
The project, when sketched this way, may strike the reader as hopelessly
extensive. In fact it strikes me that way. I think that our model, no matter
how much development it may undergo, will always remain incomplete
and informal in some of its most compelling rhythmic aspects. That is
surely a defect in the model regarded as a component within a potential for-
mal theory of music-perception. But it does not damage the model irrepara-
bly as a linguistic tool for making analytic statements about pre-existing
pieces of music. In the discussion of Morgengrufi, I hope to have exem-
plified some ways in which I feel the model can in fact convey new and char-
acteristic ideas about aspects of a piece that are undeniably "rhythmic." I
used English prose and a few graphics for the purpose. I can imagine using
other media as well: poetry, other languages, other kinds of graphic art,
theater arts, musical performance of the piece, or of excerpts therefrom, or
of a series of examples (with or without commentary), composed Lehr-
stiicke of various sorts, and so on, the various media alone or in combina-
tion. The graphic conventions of Schenkerian or post-Schenkerian theories,
Music Theory 373
for example, might enable one to repr

ConteXTs by "windows" framing
graphs. Figure 8.1 through 8.9 sug
amount of rhythmic theory could be f
interrelationships of such windows-
derlying such formalization, that is the
monitor, would have to be taken in
fluence on the theory be underestimat
Earlier on, I suggested that the p-mo
tween the undefined "importance" of
ity of elements within a language L th
subjects of sentences, tonic harmoni
nien. The model can also distinguish ot
in avoiding fruitless political/legal con
a category called "finality": pi is mor
includes that of p2 in all respects and a
of the piece. We can also define "P-
emblematic than p2 if the P-R list
inclusive in some other defined way, t
"ST-emblematicity" in the same spirit.
aesthetic values to these categories if
believe and claim that more-final pe
tant" (of greater aesthetic value) than
can as legitimately believe and claim th
are the "more important" ones; and
I argue that discriminations of this
not because I believe that value judgm
context but - on the contrary - precis
important. We ought to be correspo
are, to ourselves and - where the oc
why we should not mistakenly confus
rationalist systems. The confusion ca
systematics and our valuations. To pu
ever the individual critics of the prece
know what they are talking about.
23. The tree-structures of Lerdahl and Jac

to elaboration through such "windows." Ext
modifying their methodology, particularly on
tual priority. As the reader will have gathered
that modification in this regard would in fac
engaging, powerful, and significant. Recent
munications, lead me to believe that the theo
like my "windows," although of course on its
374 David Le win
Part V: Perception and th
At the very beginning of th

phenomenological studies of
larly in what one might call
that thought.
The problems I want to cons
ception, to suppose that there
Y that is perceived. Typicall
mind that might be God or
serted, explicitly or implicitl
ity" or "existence" or "bein
agrees that a tree Y does alw
ing it. (See Russell, 1945, p.
sion of Berkeley's argument.)
Classical European philosop
suggest to us that we call X a
ing a verb that describes a re
that-is-not-X; X is "observin
sort.
Husserl proceeds quite differently in these matters, as do other modern

philosophers among his precursors, contemporaries, and followers (Mille
1984, pp. 7-32). But they still recognize a distinction of X and Y in som
form. Y is crucially not X-itself but rather some thing(s) demonstrabl
"other" - this tree here now (that is not me), this acoustic signal here a
listen to it over this time span, that is impinging upon me (but is not me),
this artwork as I perceive it or understand it, perhaps as Z made it, or even
as Z is making it, but not (NB) as-it-is-emerging-now-from-me, let alone as
it-is-being-me-and-I-am-being-it.25
24. Or a plant? I once saw a fast-action film of a vine that reversed its direction of growth
along the ground 180 degrees, and crawled back for some distance in that direction to reac
a stake that had been put in the ground there; the vine then proceeded to climb the stake. Did
the vine perceive the stake? If not, why not?
25 . I phrase aspects of the sentence to recall Hegel, for it might appear at first that Hegel
phenomenology does precisely obliterate, or attempt to obliterate, the X/Y distinction. In
sense that is true. But the picture it gives of Hegel's procedure is not complete enough. T
Phenomenology of Mind does not deny subject-perceiving-object and substitu
Understanding-Understanding-Understanding. Rather the book portrays a process of e
lightenment, a journey that begins at subject-perceiving-object and ends at Understandin
Understanding-Understanding. The journey is a very different thing from the destination
trip from Des Moines to Chicago to New York to Paris to Damascus is not the same thing
Damascus, nor does it deny Des Moines. Damascus is not a substitute for Des Moines in th
connection. For "Des Moines" read "Consciousness" or "Perception"; for "Damascus"
read "Self-consciousness" or "Understanding." The air carriers and intermediate airpor
are the dialectic process and the stages of dialectic transition. According to Gadamer, H
Music Theory 375
The habit of distinguishing X from Y

not in itself pose a danger for music t
though, when we add an assumption
fundamentally perceptual in nature o
take as a point of departure for mus
musical perception) a paradigm in w
some "music" Y that is demonstrably
"the music" Y is profoundly and funda
prior to any activity of X-now, even
has Gegebenheit and Dasein, not ju
speaking, X finds Y given and there^ n
even if, in some situations, Z might be
The X/Y paradigm can accommoda
ratus of Husserl's phenomenology.26 Bu
tense activities of composers and perfo
is composing right now is not somethi
poser; on the contrary, we say precise
poser." Nor is the music-as-it-is-bei
prior to the activity of the-compose
composition involves producing someth
This is as true for the symphonetes, t
of 41 33'\ as it is for the phonascus, th
The Ring. To be sure, a traditional com
noematic exchanges, even into subjec
or portions of the piece already drafte
ries will then be pertinent to the wor
indispensable. But the music-tfs-it-is-b
the composer's activity, nor is it somet
poser.
gel's The Phenomenology of Mind demonstrates "the necessary transition [emphasis mine]
from consciousness to self-consciousness. . . . R. Wiehl ... has shown that in looking back
from the chapter on 'Force and Understanding,' one must view 'Sense Certainty' as the point
of departure: namely, . . . consciousness as yet entirely unconscious of its essential self-
consciousness [X thinking 'I perceive Y' and taking it for granted that Y is something not-
me] . . . Hegel's claim that the dialectical transitions are necessary [emphasis mine] is made
good . . . again and again if one reads carefully." (Gadamer, 1976, p. 36).
26. Chapter 1 of Miller (1984) also addresses this issue. The differentiability of Y from X
is clear in Husserl's insisting that "the 'direct' objects of our perceptual acts are ordinary
physical objects, and not anything else in their stead" (Miller, 1984, p. 14). Miller continues
by citing Husserl's own text: "... I perceive the thing, the object of nature, the tree there
[emphasis mine] in the garden; that and nothing else is the real object of the perceiving 'in-
tention.' ... an 'inner image' of the real tree that stands out there [emphasis mine] before me
[emphasis mine] is nowise given . . ." (ibid.).
376 David Lewin
Once the music has been c

nomenon for the composer.
The record has special value
critics, but for the compose
cisely what the sight of sk
skier who has navigated a tr
self on the Eiffel Tower mea
first trip to Paris. Not just t
the same in all three cases: "
In contrast, the composer
cwm-is." And the listener-pe
thing else: "I am here-now
perceiving is involved in t
not, nor is the composer-h
does not "perceive" the ar
sense); the composer either
this well:
Composition is a deed, an action . . . The climber in the high mountains

is intent upon the steps he is taking, on the practical realization of those
steps . . . [The composer's] psychology is not dissimilar . . . extremely
often the completed work is incomprehensible to him immediately after
it is finished.
Why? Because his experience in creating the work is incalculably
more intense than any later experience he can have from it; because the
finished product is, so to speak, the goal of that experience and not in
any sense a repetition of it. He cannot relive the compositional experi-
ence . . . And yet he is too close to it to detach himself to the extent nec-
essary to see his work objectively, and to allow it to exert its inherent
power over him. (Cone, 1979, pp. 25-26).27
The X/Y paradigm fits poorly in the same ways with the performer in the
act of performing. "The music" that this person is playing now is not "over
there" for the player; it is not something other-than-me, prior to any activ-
ity on my part. As with composing, the gestus of performing involves pro-
ducing something that is not "there" prior to the activity, something "o/'the
artist" at the time of creation. To be sure, a traditional performer at work
can enter into noetic/noematic exchanges, even subject/object relation-
ships, with parts of the acoustic signal already produced; to that extent per-
27. The passage is taken from "The Composer and His Message," a lecture delivered at
Princeton University in the Fall of 1939.
I have modified the sense of the passage by one of my omissions. Sessions writes: "he
cannot relive the compositional experience without effort which seems quite irrelevant." I
do not see how the experience can be relived at all.
Music Theory 377
ceptually oriented music theories are

as what-is-being-played-right-now is f
tivity. Here, even more than in the ca
recognize "the music," after it become
sician, as a trace or record of that per
word "record" in precisely that conn
There is not space here to explore t
may be useful to working composers
extent to which useful theories in t
those explicitly bound to ways of perc
performances (rather than those boun
ence, logic, dialectics, et al.). Personally
kinds can be useful beyond analysis
action, ways of suggesting what migh
what has been done. But I shall leave
and proceed instead to sum up my p
thing you do, and not just something
of music can not be developed fully
(with or without an ancillary dialectic
Actually, I am not very sure what a "
"theory of modern Western art-music
either) that includes a theory of musi
the broader study of what we call p
that includes competent listening to b
and performance. Here I understand
in the sense of high art but also as
"noodling," and in a whole spectrum
rubric of noodling I include rhythm
like patterns of walking, finger-drum
clude singing, whistling, or humming
variations on familiar tunes; I also inc
ing metal objects, knocking on wood
28.1 differ explicitly here with the stance of L

their book reads: "We take the goal of a theo
musical intuitions of a listener who is experi
doff, 1983, p. 1.) I am impressed but not pers
just brought up, arguments which can be fo
implicitly throughout it.
It is true that the musical intuition of their li
for that person. But the musical intuition is no
this, like the "real tree" of footnote 26. When
instance of) Morgengrufi. That is what I am l
ears and my brain, are things I am listening w
there.
378 David Lewin
sounds without pitched fu

blowing on conch-shells, th
so on. The range of activitie
bad-and-incompetent perfo
ones (where the performer r
some sense - cognitive or
band or orchestra, or singi
dancing in more or less str
chamber music or jazz or
formally, writing passages o
semi-formal groups to play,
or "shows," improvising so
score up on the piano rac
promptu transcriptions fir
sound of an ensemble piece
piano or synthesizer keybo
The p-model we have been s
of musical behavior, and it
various activities as formal
guages L" of some kind. I hin
the "language L" of the p-m
possibilities recurrently thr
not sure that "language" is
though there are preceden
guage of Love). And even if
will not very likely to able t
The activities as listed abo
sociology of the matter at
time around contemporary
be aware of many ways in
academic - separate competen
isting music, and in unders
this separation whether we
feelings. The reader has gath
conveniences, mainly in that
much more serious to my wa
posers, performers, and sc
ing "effective" sounds, upo
ing art as something "giv
young people instead to conc
in all cases as ways of mak
more on the latter subject f
Music Theory 379
Speaking in particular as a profess

about the many examinations I have at
are certified as competent musical "per
way in which they run critical analyse
languages L that are not music. Someti
go to the piano, play something, and a
student will freeze upon being asked t
by "fooling around" of this sort at the
ing that I would encounter such bloc
course with an examinee. Remarkably,
ther way, between the keyboar
susceptibility to this paralysis.29
I am not concerned here about ad
proper, who are being examined in the
technique of analysis, or in their acqua
cal literature. I am concerned rather w
are being encouraged by our educati
standing of music from its production
cal "understanding" with an ability t
and/or in certain symbolic language
"there," art works whose species are
amination. When we certify "under
behaving like the authorities who ce
basis of questions asked in English,
existing French texts. (I suppose that o
(c), (d), or (e) "in French" as well as "
student certified as "competent in F
Causons musique," or, "II me faut sort
result may be substantial or total para
"I don't speak or write French, but
the student will be justly indignant si
propagated the myth that such a m
29. In an upper-division analysis course I o

play the first piano sonata of Boulez very wel
piece by Debussy for two weeks, when she ca
hear at all. I said, "I can't hear what you mean
replied, "I don't play the piece." "Well," said
mean as you did when you heard (such-and-su
piece at all; I thought this was a course in an
at the piano, but still paralyzed. Some student
give it their best try. On the other hand, som
best, while some who play badly freeze up (a
ability).
380 David Lewin
French" exists - that it is

speaking or writing it, and
English (or baby "French"
knowledge of Trench in som
Our conceiving (and enco
speak nor write French is
teners" to music who do n
(and encounter) "fans" wh
ences" for political debates
activity but rather wa
eventually - perhaps - vote.
ered "musical" if its inhab
other, to the best of their
sic "lovers" boast of the
probably commute), their c
proved attendance at conce
their superb hi-fis, their st
And our academies are righ
enjoy, in their erotic-ac
Canon in D can be yours for
Poor Pachelbel - he thought
"ability to read" French, t
ciation" and to a significant
a great social swell, a movem
consumers, rather than en
movement is wrong. The L
serve, understand, and appr
Naturally one cannot sim
from perceptive understand
or at least without some exp
to dispute the value of studi
claim of Lerdahl and Jacken
active listeners as well." (
however, qualify it so as to
have done a great deal of ex
of concentrated readiness in
spire." Schoenberg puts t
could imagine it: "Theory
that all was well done.' " (
Schoenberg is speaking of
ation," so the stress in his c
30. The quotation is from the
Music Theory 381
nation follows creation. In the context

the word "saw" as much or more: first
the Lord perceive what He had (already
berg's metaphor suggests a powerful
by what He sees, He responds to Hi
more, or something new. Thus:
And God said, Let the waters ... be gath
appear: and it was so.
. . . and God saw that it was good.
And God said, Let the earth bring forth
and it was so.
Genesis, 9-1 1
Here one could say that the Lord u

fresh creation, but it is equally impor
tion as a mode of response to his latest
of perception-STatement: "LOOKS
trees." (And it was so.) Many compos
rhythm familiar: one recognizes that
way it ought to be when - and sometim
composition begins to take shape as
Making fresh music as a mode of m
chain of perception-and-creation is
have so far considered, including m
worked out as yet. Perhaps the link ca
text of received conceptual systems.
mental act, and describes it as somethi
see as yet, though, how he might dist
listening, acts of performing, and acts
tual response in various musical contex
The link might be supplied by som
Harold Bloom, who asserts that "th
poem, but another poem, a poem not
as it stands does not transfer easily to
problems attaching themselves to t
Suppose we modify the notion and,
any crafted artwork, claim that "a poe
ing of another poem, a poem not itsel
In that case, when we play excitedl
after a stimulating concert we are not
3 1 . Avenues of phenomenological investigat

and Fisher (1968, pp. 23-39).
382 David Lewin
the memory of perception; r

other poem being our impr
play fascinatedly again and a
passionata; we are not mat
hands to a preconceived "per
of perceiving the theme as w
performances that we essay,
process that is our act of p
minor Piano Concerto was
zart's c minor Concerto. Th
ingly a species of perception
many other things too; the
in phenomenological spac
Beethoven about Mozart's p
his various perceptions. Our
bal, but because they are i
ems."
That feature of Beethoven's verbal remarks highlights an important dif-
ference of the post-Bloomian view from the Bloomian one. The post-
Bloomian view does not exclude critical utterances as poetry. No more does
it exclude acts of analysis. The making of an analysis can be an act of per-
ception, in this view, to the extent - and only to the extent - that the ana-
lytic report which traces the deed of perception is itself "another poem."
The broad interpretation of "poem" allows us to admit traditional vari-
eties of interpretative studies into the canon of critical perception, thereby
weakening the force of Bloom's original assertion while expanding its do-
main.32 The broad interpretation specifically admits under the post-
Bloomian rubric not only the score of Beethoven's c minor Concerto, and
not only my playing the theme from the finale of the Appassionata this way
and that, but also analyses like those of Lerdahl and Jackendoff , or like my
discourse involving the syntax of Macbeth's sentence. In doing so, the criti-
cal approach brings sharply to our attention the need for studies in the po-
etics of analysis. To the degree that analytic records of musical perceptions
are poems, ski tracks tracing the poetic deeds that were the perceptions
themselves, then critics - if not analysts - must concern themselves with the
poetic resources at hand, that is, the sorts of poetic spaces analysts inhabit
and the varieties of poetic media through which they move in executing
their deeds.
I take this search for poetics to be the core of the critical position pro-
jected by James Randall, Elaine Barkin, and Benjamin Boretz in recent
32. I phrase my text here so as to connect with Jonathan Culler's (1981, pp. 14, 107-
111) critique of Bloom, interpretation, and the dissonances between them.
Music Theory 383
years; their writings "about" music m

with critical theory, analysis, more-or
musical composition that has close c
"compositional" activities of Kenneth
others.33 Also concerned with poetic
of "other poems," are the Functio
whose ideas considerably antedate Bl
called 'the death of the hired man,' pe
Cal Tech in 1982 on the occasion of
etry lounge (a version of a New Eng
wooden benches and desks, whose t
'Mending Wall' stencilled across them
To characterize the cited writings as
ysis is not to succumb to a superficial
ner in the vulgar sense. A casual re
might easily be misled by such an imp
positions they occupy in a dialectic tha
historical, a dialectic that involves the
Babbitt and the history of PNM. In a s
Babbitt is "scientific" and "objectiv
etic" and "subjective." The superficia
very far from adequate to engage the
hopelessly trivializes. The writings o
broad interpretation of the post-Bloom
dall. In that view, the issue is not whe
what sorts of poems there shall be, an
ued.
33. Except for the works of Ashley, the recent writings and the compositions are repre-
sented by contributions to Perspectives of New Music, starting with the Spring/Summer
1972 issue, which contains Randall (1972). Among other things, the article projects an at-
tempt to build a very new sort of perceptual ConteXT in which to hear Alberich's opening
passage within Gotterdàmmerung, Act II, Scene 1. Barkin is represented by a number of
substantial pieces in the subsequent issues of PNM. Of special interest in the present connec-
tion is " 'play it AS it lays'," which records a perception of Arnold Schoenberg's piano piece,
opus 19, number VI (vol. 17, no. 2, Spring-Summer, 1979, pp. 17-24). The enormous labors
of love through which Benjamin Boretz influenced the journal over many years are only
hinted at in his modest editorial apologia, "Afterward(: a foreword)," (vol. 22, Fall- Winter,
1983 and Spring-Summer, 1984, pp. 557-559). Kenneth Gaburo is celebrated by a large
number of contributions to volume 18 (Fall- Winter, 1979 and Spring-Summer, 1980, pp. 7-
256). The contribution by Garburo himself is a lecture/composition/performance/talk poem
("Brain: . . . Half A Whole," pp. 215-256). The reader may want to approach it, or to re-
view it, after perusing the discussion of David Antin and Marjorie Perloff later in this article.
Pieces by and about Robert Ashley appear in Formations, vol. 2, no. 1 (Spring, 1985), pp.
14-63. Musicians may not be familiar with this journal; it is published in Madison by the
University of Wisconsin Press.
384 David Lewin
Marjorie Perloff focuses th

say (Perloff, 1984), she qu
topher Clausen, who has the
Few doubt that the rise of sc

ing [poetry] as a publically im
ple accept as being centrall
reading public that figurativ
and discourse retain their val
been notably successful. . .
believe that anything of real
Perloff examines brilliantly

not untypical of discussio
sumptions are: "First, that
modes of discourse. Second
that poetry once served an
fourth, that poetry is inher
such, stands opposed to 'c
figurative, straightforward
'truth' of poetry is one of s
(Perloff, 1984, p. 43).
These observations lauch a
cluding perceptive analyse
Stein, that culminates in t
Frost. Toward the end of he
By this time, the audience ha

the connection between fro
tween the status of Armaja
design the poetry lounge, a
Antin's casual talk has been
writing poetry with referen
etry must be based on actua
text puts forward that poeti
through a series of narrative
are finally not quite sure w
poetry? Lecture or story? Ph
. . Antin does not regard "c
as the enemy; . . . and althou
importance, he is more inter
does it mean to do x [emphas
sistencies [sc. rabbits and d
truths of moments, situati
61)
Music Theory 385
My post-Bloomian proposition, that th

sides in the (active) making of another p
"performance" in traditional terms, is n
rage of scholarship over the last few pag
convey the point, I will copy out a wond
- Accori accori accori, uom, a la strada!

- Che ha', fi' de la putta? - I son rubato.
- Chi t'ha rubato? - Una, che par che ra
come rasoi', si m'ha netto lasciato.
- Or come non le davi de la spada?
- I dare' anzi a me. - Or se' 'mpazzato?
- Non so; che'l dà? - Cosi mi par che va
or t'avess'ella cieco, sciagurato!
- E vedi che ne pare a que' che'l sanno?
- Di' quel, che tu mi rubi. - Or va' con D
- Ma ando pian, ch'i vo' pianger lo dann
- Che ti diparti? - Con animo rio.
- Tu abbi'l danno con tutto'l malanno!
- Or chi m'ha morto? - E che diavol sacc'io?
- Cecco Angiolieri (1250-1319)34
In trying to "perceive" the poem so that it makes sense to you, are you
not taken by an urge to perform it - to read it aloud and act the roles of the
three characters, with appropriate vocal modifications? I am. So far as I ki-
netically sense the vigorous movements of the characters while they
converse - which I do to a considerable degree as I am reading their parts -
I am also trying to direct the scene for a theatrical production, as part of my
mode of perception. This is not to say that I would consider irrelevant to my
perceptions closely reasoned studies of the syntactic structure, the historical
contexts of thirteenth-century Italy (including the rise of the vernacular in
literature and the development of the sonnet), the intrinsic sound-structure
of the text, the rhythms in the changes of speakers, the ways in which those
rhythms counterpoint the regularities of the sonnet "form," contributing
thereby to the fantastic modulation and theatrical coup when the woman
34. The text is taken from Kay (1958, p. 68). Kay provides a "plain prose translation":
"Run, run, run, man, along that street!" "What's wrong, whoreson?" "I've been robbed."
"Who robbed you?" "A woman, who shears like a razor, she's left me so bare." "Well, why
didn't you have at her with your sword?" "I'd sooner turn it on myself." "Are you mad?" "I
don't know; what makes you think so?" "The way you are going on: it's as good as if she had
blinded you, you wretch!"
"See how it appears to people who understand?" "Let them know that you rob me." "O
go away!" "I'm going, but slowly, for I must weep my loss." "How do you leave me?" "In
bad heart." "Well, you can suffer your loss' and every illness with it, for all I care!" "Who is
killing me now?" "How the devil should I know?"
386 David Lewin
herself appears on the scene

these studies would help cla
tions. But they would not s
least I do not think they wo
urge to recite, act, and di
ficient" reflective knowledg
that contains them. I only
denser, more compelling,
my earlier analysis of Mac
sponse there: to the exten
response to the text, to tha
the same mode; fresh ana
nous") will not wean one a
prove the performance, or
ances.
A skeptic could point out that I am discussing a play (b

and an unusually theatrical sonnet (by Cecco); it is only na
to these works in a theatrical mode. Fair enough, and I do no
mote a priori any one mode of perception as universally "
other. Only I believe we are in some danger, these days,
more productive modes of perception; I think we underes
the extent to which those modes are alive and active even in situations
where their pertinence is not so immediately apparent as it is with Shakes-
peare and Cecco, situations where we think of ourselves as "readers," not
as speakers, writers, actors, and directors; as "listeners," not as players and
composers.35
To illustrate my point, I shall ask you to imagine the following scenario.
You are a young warrior of ancient Rome, taking flight from an armed
mass of pursuing enemies. Desperately seeking refuge, you burst unwit-
tingly into the Temple of the Vestal Virgins, a shrine forbidden to males
under penalty of death. Amazed and irate priestesses surround you. Col-
lecting yourself as best you can, you turn to them and say - what?
Well you certainly do not say "Pardon ME!" presumably tipping your
helmet to the ladies and looking about surreptitiously for a convenient ex-
it. At least you do not say that unless you are the person who composed the
College Board Examination in Latin that I took some thirty-five years ago.
Of course the question on that person's examination was not "What did the
young man say to the Virgins?" but rather "What is the correct translation
of 'Ignoscite' in the above passage?"
35. The cultural-historical bias behind our underestimation is explored by Barish

(1981).
Music Theory 387
We were offered five translations from

"Pardon me" as the first of the options, I
nation hall, drawing some indignant atten
ual. How nice, I thought, an examiner w
read the other four "answers," the awf
in fact the "correct" answer. Indeed, accor
to support "Pardon me" not only as a pl
in the context, but as a correct translatio
tion, but the correct translation.
The examiner, of course, had thought
not been thinking (perceiving) anything
scenario tries to suggest, the modes in wh
as best I could under the examination con
because I was myself a young man in a
by older authorities. The examiner had ce
the connections of the expression "pard
and particularly in conversational usage
the context in which "Ignoscite" appea
taking in (perceiving) the Latin passage
have written "Forgive my blasphemy,
something of the sort. These translation
civic service and civic virtue that is impli
and the events of the drama. "Pardon me,
geois British colloquialism. When read
worse, the world of slapstick comedy, a m
exCUse ME") or Charlie Chaplin (who w
tipping bit, not to mention the escape sce
don me" is just as wrong a translation
know, do one?" and it is wrong in exactly
The examiner, however, was not testing
imaginatively, using a Latin text, into th
the ability to bring into such an imaginat
conceptual matrices of ones own culture.
do with an examination "in Latin"? As Per
tions. First, an examination "in Latin"
Latin, which is separable from conver
drama, or writing original Latin tex
"grammar" and "translation"; as a resu
comes an examination in English about La
an effective command of English is no
consists of selecting from among five g
Subassumptions: five choices are plent
some implicit question; each answer is
388 David Lewin
"correct" and it is "the" c

can take place in a hushed,
read aloud nor move about
and gestures have nothing t
these assumptions, we then
schooling perceive somethi
classics"!
Let me put the matter this way: the gesture and English utterance that
you make when you act the young Roman in his predicament are not phe-
nomena that are separable from your understanding of what "ignoscite"
can mean in Latin. Just so, the vocal and bodily gestures that you make
when you act Macbeth saying "This" are not phenomena that are separable
from what you perceive in the scene as a playgoer or reader.37 Just so, the
way you sing or conduct the first four notes of "Joy to the world" is not
something that is separable from the way you perceive structural functions
for the notes on which you sing "Joy" and "world." Likewise, your percep-
tions of Morgengrufi are not separable from how long you wait on the fer-
mata at measure 15 before it feels right to go on, when you sing or accom-
pany the song, or when you transcribe it for piano solo. Your perceptions of
the song are likewise not separable from how long you want to dwell on the
lonely B flat in the piano at measure 12, before allowing the next note of the
accompaniment to enter. (Our formal perceptions pi through p9 intermesh
with just such performance activities.) And your perceptions of the "XDY-
and-YDX" cadence in Die Walkiire are not separable from the way you
conduct the fp dynamic and the change of tempo, nor is either of these sepa-
rable from the way in which you act Sieglinde's discovery at this moment
that her adulterous-lover-to-be, the savior promised her by her father, is in
fact her own long-lost brother.38
36. How many people does one find in modern Italy who do not use gestures as part of
their language? Should we assume that earlier inhabitants of the region were more con-
strained? Did their fascination with rhetoric, when they discovered it, reflect a desire to keep
still while they spoke? Did Cicero deliver his speeches without moving a muscle? In court? In
the Senate? Or did he just mail Xerox copies of the written texts to the jurors and the Sena-
tors, so that they could "read" the speeches as our high-school students do?
37. I make contact here again with the sorts of ideas expressed by Lochhead and Fisher
(1982).
38. Wagner's stage directions say that she tears herself loose from Sigmund's embrace in
the most extreme intoxication, and confronts him as a model for comparison ("reifit sich in
hôchster Trunkenheit von ihm und stellt sich ihm gegenuber"). Each twin has been ordained
by Wotan to be the mirror and (dominant) support of the other; Sieglinde comes to realize
that at just this moment.
All of Sieglinde's deceptive cadences in G are laden with this dramatic import, as are all
the G cadences through Act I. Most of them are deceptive. The deceptive ones typically in-
volve harmonies including an E and/or a Ctf and/or a Bl>, as well as a G. The dramatic upres-
Music Theory 389
The musical examples just above inv

me make them vivid for non-musician
ent way. Still, the reader who reviews
bert, Handel, and Wagner passages w
their purely musical analysis that is
performance issues raised in the pre
literary and theatrical contexts I ha
naturally also appropriate, and enric
Indeed it is quite possible to approac
atrically" as well. To illustrate the po
role of "F# IGV\ within the drama tha
Fifth Symphony. In one of your dual
or subdominant of C. C is the tonic of
the clock-face of the chromatic scale or
You enter magnificently, surrou
diminished-seventh chord that is th
since the first theme got underway. Yo
piece and the first tutti of the piece.
to maintain the tonic C against your
serious chromatic excursion of the pie
iv were local affairs. You enter here w
to G; but you abruptly hurl the cloa
underneath as Gt, upper neighbor to F
solves not as V-of-(V-of-C) but as V-
enharmony, you single-handedly ach
Et major, from the first thematic gro
matic group. A new theme enters dire
horn call.
In the reprise you replay this whol
enter on your climactic diminished
waiting for you to throw off your F#
throw off your Fjt cloak all right, bu
neath it! You resolve as leading-tone to
ence" of Wotan throughout Act I is often mis

bass events of the music beyond the Valhal
wonder why the lovers can't get down to busi
Sieglinde a good G cadence, as she predicts th
come out with the Redemption theme in G -
But the Redemption cadence is spoiled and tur
Bt-CI in the trombones that undermine the ca
ing them ominous. Sieglinde must flee from
father, who is now clearly identified as the s
bass, and the diminished-seventh harmony.
390 David Lewin
(V-of-C) after all. The horn

manage the horn-call the
handedly warded off the mo
inC.39
During the development y
After the first theme has
quence whose local tonics mo
C, to G minor. F and G are
capacities as Gt and Ffl re
gets to G minor, it starts to
ens, the bass moves by step
Fjt ! Thereupon a new motiv
ond thematic group. This s
of the circle of fifths, via
music gets to F minor, it s
Entwicklung tightens, the b
in F minor; then, pivoting t
finally arrive at a tonicized
your leading tone.
And then, after all, you thr
monically as Ftf all the tim
"becalmed" accordion-type
winds and strings shift to
triad in first inversion. The
finally pianissimo, are uniq
and fortissimo bluster of C
ing, riding-gently-up-and-
frenzied Sturm and Drang
say that of a phenomenon so
the way back to C minor
country" as a member of V-
refused to assume upon yo
reprise is still to come.)
Now that I have coached y
gency of the theatrical mod
duct the pertinent music,
about it? Is not the way you
separable from things-tha
"playing" the role of Et, or
things-that-you-perceive in
to different characters.)
39. Atlas (1983, pp. 26-27, 32-
Music Theory 391
My skeptic will point out that this

matic" one, and ask how my contentio
dramatic music. Here, finally, I must
proclaiming the virtues of any one mo
only concerned that our society enc
modes. To the skeptic above I say, "F
are convinced is neither poetic nor dra
ter further."
References
Atlas, R. E. The diachronic recognition of enharmonic equivalence: A theory and its appli-
cation to five instrumental movements by W. A. Mozart. Ph.D. dissertation, Yale Un-
iveristy, New Haven, CT, 1983.
Barish,J. The anti-theatrical prejudice. Berkeley: University of California Press, 1981.
Barkin, E. Play it as it lays. Perspectives of New Music, 1979, 17(2), 17-24.
Bloom, H. The anxiety of influence. New York: Oxford University Press, 1973.
Boretz, B. Afterward (: a foreword). Perspectives of New Music, 1983/84, 22, 557-559.
Browne, R. The dialectic of good continuation in tonal music. Music Analysis, 1985, 4(1/2),
5-13.
Clifton, T. Music as heard: A study in applied phenomenology. New Haven: Yale University
Press, 1983.
Cone, E. T. (Ed.). Roger Sessions on music. Princeton: Princeton University Press, 1979.
Cook, N. [Review of Clifton, Music as heard\. Music Analysis, 1983, 2(3), 291-294.
Culler, J. The pursuit of signs. Ithaca, NY: Cornell University Press, 1981.
Dreyfus, H. L., & Hall, H. Husserl, intentionality, and cognitive science. Cambridge, MA:
The MIT Press, 1982.
Forte, A., & Gilbert, S. E. Introduction to Schenkerian analysis. New York: W. W. Norton
& Co., 1982.
Freud, S. Vorlesung zur Einfiihrung in die Psychoanalyse. Berlin: Gustave Kiepenheuer
Verlag, 1955.
Gaburo, K. Brain: . . . Half a whole. Perspectives of New Music, 1979/80, 18, 215-256.
Gadamer, H. G. Hegel's dialectic (P. C. Smith, Trans.). New Haven: Yale University Press,
1976.
Gréer, T. A. Listening as intuiting: A critique of Clifton's theory of intuitive description. In
Theory Only, 1984, 7(7,8), 3-21.
Hasty, C. Rhythm in post-tonal music: Preliminary questions of duration and motion, jour-
nal of Music Theory, 1981, 25(2), 183-216.
Husserl, E. The phenomenology of internal time consciousness (J. S. Churchill, Trans.).
Bloomington: Indiana University Press, 1964.
Ihde, D. Listening and voice: A phenomenology of sound. Athens, OH: Ohio University
Press, 1976.
Kay, G. R. (Ed.). the Penguin book of Italian verse. Harmondsworth, Middlesex: Penguin
Books Ltd., 1958.
Keller, H. FA No. 1: Mozart, K. 421. The Score, 1958, 22, 56-64.
Kramer, J. New temporalities in music. Critical inquiry, 1981, 7(3), $5y-5$b.
Laske, O. Toward an explicit cognitive theory or musical listening. Computer Music jour-
nal 1980, 4(2).
Lerdahl, F., & Jackendof f, R. A generative theory of tonal music. Cambridge, MA: The MIT
Press, 1983.
Lewin, D. Some investigations into foreground rhythmic and metric patterning. In R.
Browne (Ed.), Music theory: Special topics. New York: Academic Press, 1981, pp. 101-
136.
392 David Lewin
Lochhead, J. The temporal stru

Ph.D. dissertation, State Univers
Lochhead, J., & Fisher, G. The p
Semegen's Three Pieces for Cla
39.
Mattheson, J. Der vollkommene
Miller, I. Husserl, perception, a
"A Bradford Book," 1984.
Minsky, M. Music, mind, and m
York: Plenum, Press, 1982, pp.
Narmour, E. Beyond Schenkeris
Perloff, M. Postmodernism and
Randall, J. Compose yourself - a
10(2), 1-12.
Riviere, J. A general introductio
1952.
Rowell, L. The subsconscious language of musical time. Music Theory Spectrum, 1979, 1,
96-106.
Russell, B. A history of Western philosophy. New York: Simon & Schuster, 1945.
Schenker, H. [Free composition (E. Oster, Trans. & Ed.)]. New York: Longman, 1979.
Schoenberg, A. Structural functions of harmony. New York: W. W. Norton & Co., 1954.
Strunk, O. Source readings in music history. New York: W. W. Norton & Co., 1950.
Tenney J. [Review of Clifton, Music as heard.} journal of Music Theory, 1985, 29(1), 291
294.
Zarlino, G. [The art of counterpoint, part three of Le istitutioni harmoniche, 1558 (G. A.
Marco & C. V. Palisca, Trans.)]. New Haven: Yale University Press, 1968.

Lewin Theory Phenomenology Modes of Perception

Uploaded by

Copyright:

Available Formats

Lewin Theory Phenomenology Modes of Perception

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lewin Theory Phenomenology Modes of Perception

Uploaded by

Copyright:

Available Formats

Music Theory, Phenomenology, and Modes of Perception

Author(s): David Lewin

Music Theory, Phenomenology, and Modes of Perce

Recent years have seen an increasing influence on music t

Part I: Phenomenological Preface

Overtly phenomenological study of music in Husserl's

. . . whereas it does seem true that I am hearing that tone

Requests for reprints may be sent to David Lewin, Department of

previous hypothesis) that I e

Miller also devotes much att

Among explicitly phenomenological writers who do invo

1. Lochhead worked extensively with the philosopher Don Ihde, who a

book (Clifton, 1983) reflects his stance

2. Unfortunately, further development of Clif

3. Kramer is a consulting editor for Music

4. For the model, it is perfectly workable to

5. HusserPs terms and pertinent diagrams ar

6. That will be manifest in my forthcoming book, Generalized musical intervals and

Marvin Minsky - like my

7. Compare this excerpt from M

EVALuating SGL. There is of course a

tures; this corresponds to th

ception(b), which is as yet perceived on

9. Within the DEFINE list for Perception(

well.10 Minsky (1982) devote

10. My stance here is not partic

it additionally becomes at cursor-time

Part II: A General Model

If one were to sequester the notion of "good" continuation as a descrip-

pointing to an EV, that coul

12. We shall explore just such shi

ing it in the context of Figure l(c) as a

only the vaguest idea of w

bass on Et 3 and Et 2. 1 turn to you

Part III of this paper, we sh

Siegmund's watching Sieglinde's wat

More of-the-essence, and m

Another listener may want to hear

Part HI: A Passage from Schubert

heuristic sense is understood as a component of each ConteXT. Selected

perceive "what we are hea

Parts 8.1, 8.2, and so on of Figure 8 c

tures for measures 9-10, f

the ConteXT for P3a backwards so a

cursive structure is by now

to the dominant harmony that precede

17. When I refer to "the g minor chord," I d

ConteXT for p6b includes a

literal sequential analog for the d mi

tion of p6a (Figure 8.6). This

prolonged above the At-G gesture in th

19. N.B.: the youth does not "go." Rather h

Figure 8.8 indicates succin

note that p9, in confirming p2, denies

Part IV: Methodology

We have already started to note and

Figure 9(a) by saving, "The harmony of measure 12 is . . .," we are already

time in which you are now reading it.

20. The metaphor of an obligatory interrupt

impression you have once for