The Typology of Detective Fiction

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Reading 22

Tzvetan Todorov

THE TYPOLOGY OF DETECTIVE


FICTION (1966)

Introduction and context

T Z V E T A N T O D O R O V is a literary theorist, originally born in Bulgaria, who


played a key role in the development of structuralism as a forceful current in
literary studies in France during the 1960s. ‘The Typology of Detective Fiction’
(1966), reproduced here in full, encapsulates the structuralist approach to classify-
ing and making sense of large quantities of literary data. Todorov begins his discus-
sion of detective fiction by locating the analysis within the context of genre and
sub-genre (see also Readings 11 and 26 on film genres). Just like different genres of
film have particular ‘rules’ of narrative development, so too do different genres of
popular literature, for example romance, Science fiction and in this case ‘detective
fiction’. Furthermore, the broad genre of ‘detective fiction’ contains within it further
‘types’ – or sub-genres – of story, which develop in accordance with their own, par-
ticular rules. The sub-genres of literary detective fiction that form the basis of
Todorov’s analysis are the ‘whodunit’, typified by Agatha Christie, the ‘thriller’,
typified by Dashiell Hammett and Raymond Chandler, and the ‘suspense novel’, typi-
fied by William Irish and Patrick Quentin. Todorov thus compares and contrasts
the (sometimes overlapping) rules governing the narrative development of these
sub-genres of detective fiction in order to establish a structuralist framework within
which the crime novel might be read critically as a literary device with its own
internal logics and conventions, as well as being read for pleasure and distraction.

Detective fiction cannot be subdivided into kinds.


It merely offers historically different forms.
– Boileau and Narcejac, Le Roman palicier, 1964
294 TZVETAN TODOROV

If I use this observation as the epigraph to an article dealing precisely with ‘kinds’ of
‘detective fiction,’ it is not to emphasize my disagreement with the authors in ques-
tion, but because their attitude is very widespread; hence it is the first thing we must
confront. Detective fiction has nothing to do with this question: for nearly two
centuries, there has been a powerful reaction in literary studies against the very notion
of genre. We write either about literature in general or about a single work, and it is
a tacit convention that to classify several works in a genre is to devalue them. There is
a good historical explanation for this attitude: literary reflection of the classical period,
which concerned genres more than works, also manifested a penalizing tendency –
a work was judged poor if it did not sufficiently obey the rules of its genre. Hence
such criticism sought not only to describe genres but also to prescribe them; the grid
of genre preceded literary creation instead of following it.
The reaction was radical: the romantics and their present-day descendants have
refused not only to conform to the rules of the genres (which was indeed their privilege)
but also to recognize the very existence of such a notion. Hence the theory of genres
has remained singularly undeveloped until very recently.Yet now there is a tendency
to seek an intermediary between the too-general notion of literature and those
individual objects which are works. The delay doubtless comes from the fact that
typology is implied by the description of these individual works; yet this task of
description is still far from having received satisfactory solutions. So long as we cannot
describe the structure of works, we must be content to compare certain measurable
elements, such as meter. Despite the immediate interest in an investigation of genres
(as Albert Thibaudet remarked, such an investigation concerns the problem of universals),
we cannot undertake it without first elaborating structural description: only the
criticism of the classical period could permit itself to deduce genres from abstract
logical schemas.
An additional difficulty besets the study of genres, one which has to do with the
specific character of every esthetic norm.The major work creates, in a sense, a new genre
and at the same time transgresses the previously valid rules of the genre. The genre of
The Charterhouse of Parma, that is, the norm to which this novel refers, is not the French
novel of the early nineteenth century; it is the genre ‘Stendhalian novel’ which is created
by precisely this work and a few others. One might say that every great book establishes
the existence of two genres, the reality of two norms: that of the genre it transgresses,
which dominated the preceding literature, and that of the genre it creates.
Yet there is a happy realm where this dialectical contradiction between the work
and its genre does not exist: that of popular literature. As a rule, the literary master-
piece does not enter any genre save perhaps its own; but the masterpiece of popular
literature is precisely the book which best fits its genre. Detective fiction has its
norms; to ‘develop’ them is also to disappoint them: to ‘improve upon’ detective
fiction is to write ‘literature,’ not detective fiction.The whodunit par excellence is not
the one which transgresses the rules of the genre, but the one which conforms to
them: No Orchids for Miss Blandish1 is an incarnation of its genre, not a transcendence.
If we had properly described the genres of popular literature, there would no longer
be an occasion to speak of its masterpieces. They are one and the same thing; the best
1
Thriller by James Hadley Chase, first published in 1939. It is the subject of a famous essay
by George Orwell, ‘Raffles and Miss Blandish’ (Collected Essays, Journalism and Letters, Vol. 3.)
THE TYPOLOGY OF DETECTIVE FICTION 295

novel will be the one about which there is nothing to say. This is a generally unnoticed
phenomenon, whose consequences affect every esthetic category. We are today in the
presence of a discrepancy between two essential manifestations; no longer is there
one single esthetic norm in our society, but two; the same measurements do not apply
to ‘high’ art and ‘popular’ art.
The articulation of genres within detective fiction therefore promises to be rela-
tively easy. But we must begin with the description of ‘kinds,’ which also means with
their delimitation. We shall take as our point of departure the classic detective fiction
which reached its peak between the two world wars and is often called the whodunit.
Several attempts have already been made to specify the rules of this genre (we shall
return below to S. S. Van Dine’s twenty rules); but the best general characterization
I know is the one Butor gives in his own novel Passing Time (L’Emploi du temps). George
Burton, the author of many murder mysteries, explains to the narrator that ‘all detec-
tive fiction is based on two murders of which the first, committed by the murderer,
is merely the occasion for the second, in which he is the victim of the pure and
unpunishable murderer, the detective,’ and that ‘the narrative . . . superimposes two
temporal series: the days of the investigation which begin with the crime, and the days
of the drama which lead up to it.’
At the base of the whodunit we find a duality, and it is this duality which will
guide our description. This novel contains not one but two stories: the story of the
crime and the story of the investigation. In their purest form, these two stories have
no point in common. Here are the first lines of a ‘pure’ whodunit:

a small green index-card on which is typed:


Odel, Margaret.
184 W. Seventy-first Street. Murder: Strangled about 11 P.M. Apartment
robbed. Jewels stolen. Body found by Amy Gibson, maid. [S. S. Van Dine,
The ‘Canary’ Murder Case]

The first story, that of the crime, ends before the second begins. But what
happens in the second? Not much. The characters of this second story, the story of the
investigation, do not act, they learn. Nothing can happen to them: a rule of the genre
postulates the detective’s immunity. We cannot imagine Hercule Poirot or Philo
Vance2 threatened by some danger, attacked, wounded, even killed. The hundred and
fifty pages which separate the discovery of the crime from the revelation of the killer
are devoted to a slow apprenticeship: we examine clue after clue, lead after lead. The
whodunit thus tends toward a purely geometric architecture: Agatha Christie’s Murder
on the Orient Express, for example, offers twelve suspects; the book consists of twelve
chapters, and again twelve interrogations, a prologue, and an epilogue (that is, the
discovery of the crime and the discovery of the killer).
This second story, the story of the investigation, thereby enjoys a particular status.
It is no accident that it is often told by a friend of the detective, who explicitly
acknowledges that he is writing a book; the second story consists, in fact, in explaining
how this very book came to be written. The first story ignores the book completely,
2
Hercule Poirot is the detective in many of Agatha Christie’s novels, and Philo Vance is the
detective in many of S. S. Van Dine’s novels.
296 TZVETAN TODOROV

that is, it never confesses its literary nature (no author of detective fiction can permit
himself to indicate directly the imaginary character of the story, as it happens in
‘literature’). On the other hand, the second story is not only supposed to take the
reality of the book into account, but it is precisely the story of that very book.
We might further characterize these two stories by saying that the first – the
story of the crime – tells ‘what really happened,’ whereas the second – the story of
the investigation – explains ‘how the reader (or the narrator) has come to know about
it.’ But these definitions concern not only the two stories in detective fiction, but also
two aspects of every literary work which the Russian Formalists isolated forty years
ago. They distinguished, in fact, the fable (story) from the subject (plot)3 of a narrative:
the story is what has happened in life, the plot is the way the author presents it to us.
The first notion corresponds to the reality evoked, to events similar to those which
take place in our lives; the second, to the book itself, to the narrative, to the literary
devices the author employs. In the story, there is no inversion in time, actions follow
their natural order; in the plot, the author can present results before their causes,
the end before the beginning. These two notions do not characterize two parts of the
story or two different works, but two aspects of one and the same work; they are two
points of view about the same thing. How does it happen then that detective fiction
manages to make both of them present, to put them side by side?
To explain this paradox, we must first recall the special status of the two stories.
The first, that of the crime, is in fact the story of an absence: its most accurate
characteristic is that it cannot be immediately present in the book. In other words, the
narrator cannot transmit directly the conversations of the characters who are implicated,
nor describe their actions: to do so, he must necessarily employ the intermediary of
another (or the same) character who will report, in the second story, the words heard
or the actions observed. The status of the second story is, as we have seen, just as
excessive; it is a story which has no importance in itself, which serves only as a mediator
between the reader and the story of the crime. Theoreticians of detective fiction
have always agreed that style, in this type of literature, must be perfectly transparent,
imperceptible; the only requirement it obeys is to be simple, clear, direct. It has even
been attempted – significantly – to suppress this second story altogether. One pub-
lisher put out real dossiers, consisting of police reports, interrogations, photographs,
fingerprints, even locks of hair; these ‘authentic’ documents were to lead the reader
to the discovery of the criminal (in case of failure, a sealed envelope, pasted on the last
page, gave the answer to the puzzle: for example, the judge’s verdict).
We are concerned then in the whodunit with two stories of which one is absent
but real, the other present but insignificant.This presence and this absence explain the
existence of the two in the continuity of the narrative. The first involves so many
conventions and literary devices (which are in fact the ‘plot’ aspects of the narrative)
that the author cannot leave them unexplained. These devices are, we may note, of
essentially two types, temporal inversions and individual ‘points of view’: the tenor of
each piece of information is determined by the person who transmits it, no observation

3
This translation of the Russian Formalists’ terms, fabula and sjuzet, is not entirely
satisfactory, since ‘story’ and ‘plot’ are used loosely and sometimes interchangeably in
much criticism of prose fiction. ‘Discourse’ is perhaps a more satisfactory rendering of
sjuzet. [...]
THE TYPOLOGY OF DETECTIVE FICTION 297

exists without an observer; the author cannot, by definition, be omniscient as he was


in the classical novel. The second story then appears as a place where all these devices
are justified and ‘naturalized’: to give them a ‘natural’ quality, the author must explain
that he is writing a book! And to keep this second story from becoming opaque, from
casting a useless shadow on the first, the style is to be kept neutral and plain, to the
point where it is rendered imperceptible.
Now let us examine another genre within detective fiction, the genre created
in the United States just before and particularly after World War II, and which is
published in France under the rubric ‘série noire’ (the thriller); this kind of detective
fiction fuses the two stories or, in other words, suppresses the first and vitalizes the
second. We are no longer told about a crime anterior to the moment of the narrative;
the narrative coincides with the action. No thriller is presented in the form of memoirs:
there is no point reached where the narrator comprehends all past events, we do not
even know if he will reach the end of the story alive. Prospection takes the place of
retrospection.
There is no story to be guessed; and there is no mystery, in the sense that it was
present in the whodunit. But the reader’s interest is not thereby diminished; we realize
here that two entirely different forms of interest exist. The first can be called curiosity;
it proceeds from effect to cause: starting from a certain effect (a corpse and certain
clues) we must find its cause (the culprit and his motive). The second form is suspense,
and here the movement is from cause to effect: we are first shown the causes, the
initial données (gangsters preparing a heist), and our interest is sustained by the expec-
tation of what will happen, that is, certain effects (corpses, crimes, fights). This type
of interest was inconceivable in the whodunit, for its chief characters (the detective
and his friend the narrator) were, by definition, immunized: nothing could happen to
them. The situation is reversed in the thriller: everything is possible, and the detective
risks his health, if not his life.
I have presented the opposition between the whodunit and the thriller as an
opposition between two stories and a single one; but this is a logical, not a historical
classification. The thriller did not need to perform this specific transformation in
order to appear on the scene. Unfortunately for logic, genres are not constituted in
conformity with structural descriptions; a new genre is created around an element
which was not obligatory in the old one: the two encode different elements. For this
reason the poetics of classicism was wasting its time seeking a logical classification of
genres. The contemporary thriller has been constituted not around a method of pre-
sentation but around the milieu represented, around specific characters and behavior;
in other words, its constitutive character is in its themes. This is how it was described,
in 1945, by Marcel Duhamel, its promoter in France: in it we find ‘violence – in all its
forms, and especially the most shameful – beatings, killings. . . . Immorality is as
much at home here as noble feelings. . . . There is also love – preferably vile – violent
passion, implacable hatred.’ Indeed it is around these few constants that the thriller is
constituted: violence, generally sordid crime, the amorality of the characters.
Necessarily, too, the ‘second story,’ the one taking place in the present, occupies a
central place. But the suppression of the first story is not an obligatory feature:
the early authors of the thriller, Dashiell Hammett and Raymond Chandler, preserve
the element of mystery; the important thing is that it now has a secondary function,
subordinate and no longer central as in the whodunit.
298 TZVETAN TODOROV

This restriction in the milieu described also distinguishes the thriller from the
adventure story, though this limit is not very distinct. We can see that the properties
listed up to now – danger, pursuit, combat – are also to be found in an adventure
story; yet the thriller keeps its autonomy.We must distinguish several reasons for this:
the relative effacement of the adventure story and its replacement by the spy novel;
then the thriller’s tendency toward the marvelous and the exotic, which brings it
closer on the one hand to the travel narrative, and on the other to contemporary
science fiction; last, a tendency to description which remains entirely alien to the
detective novel. The difference in the milieu and behavior described must be added to
these other distinctions, and precisely this difference has permitted the thriller to be
constituted as a genre.
One particularly dogmatic author of detective fiction, S. S. Van Dine, laid down,
in 1928, twenty rules to which any self-respecting author of detective fiction must
conform. These rules have been frequently reproduced since then (see for instance
the book, already quoted from, by Boileau and Narcejac) and frequently contested.
Since we are not concerned with prescribing procedures for the writer but with
describing the genres of detective fiction, we may profitably consider these rules a
moment. In their original form, they are quite prolix and may be readily summarized
by the eight following points:

1. The novel must have at most one detective and one criminal, and at least one
victim (a corpse).
2. The culprit must not be a professional criminal, must not be the detective, must
kill for personal reasons.
3. Love has no place in detective fiction.
4. The culprit must have a certain importance:
(a) in life: not be a butler or a chambermaid.
(b) in the book: must be one of the main characters.
5. Everything must be explained rationally; the fantastic is not admitted.
6. There is no place for descriptions nor for psychological analyses.
7. With regard to information about the story, the following homology must be
observed: ‘author: reader = criminal: detective.’
8. Banal situations and solutions must be avoided (Van Dine lists ten).

If we compare this list with the description of the thriller, we will discover an
interesting phenomenon. A portion of Van Dine’s rules apparently refers to all
detective fiction, another portion to the whodunit. This distribution coincides,
curiously, with the field of application of the rules: those which concern the themes,
the life represented (the ‘first story’), are limited to the whodunit (rules 1–4a); those
which refer to discourse, to the book (to the ‘second story’), are equally valid for the
thriller (rules 4b–7; rule 8 is of a much broader generality). Indeed in the thriller
there is often more than one detective (Chester Himes’s For Love of Imabelle) and more
than one criminal (James Hadley Chase’s The Fast Buck).The criminal is almost obliged
to be a professional and does not kill for personal reasons (‘the hired killer’); further,
he is often a policeman. Love – ‘preferably vile’ – also has its place here. On the
other hand, fantastic explanations, descriptions, and psychological analyses remain
banished; the criminal must still be one of the main characters. As for rule 7, it has lost
THE TYPOLOGY OF DETECTIVE FICTION 299

its pertinence with the disappearance of the double story. This proves that the devel-
opment has chiefly affected the thematic part, and not the structure of the discourse
itself (Van Dine does not note the necessity of mystery and consequently of the double
story, doubtless considering this self-evident).
Certain apparently insignificant features can be codified in either type of detective
fiction: a genre unites particularities located on different levels of generality. Hence
the thriller, to which any accent on literary devices is alien, does not reserve its
surprises for the last lines of the chapter; whereas the whodunit, which legalizes the
literary convention by making it explicit in its ‘second story,’ will often terminate the
chapter by a particular revelation (‘You are the murderer,’ Poirot says to the narrator
of The Murder of Roger Ackroyd). Further, certain stylistic features in the thriller belong
to it specifically. Descriptions are made without rhetoric, coldly, even if dreadful
things are being described; one might say ‘cynically’ (‘Joe was bleeding like a pig.
Incredible that an old man could bleed so much,’ Horace McCoy, Kiss Tomorrow
Goodbye).The comparisons suggest a certain brutality (description of hands: ‘I felt that
if ever his hands got around my throat, they would make the blood gush out of my
ears,’ Chase, You Never Know with Women). It is enough to read such a passage to be sure
one has a thriller in hand.
It is not surprising that between two such different forms there has developed a
third, which combines their properties: the suspense novel. It keeps the mystery of
the whodunit and also the two stories, that of the past and that of the present; but it
refuses to reduce the second to a simple detection of the truth. As in the thriller, it is
this second story which here occupies the central place. The reader is interested not
only by what has happened but also by what will happen next; he wonders as much
about the future as about the past. The two types of interest are thus united here –
there is the curiosity to learn how past events are to be explained; and there is also the
suspense: what will happen to the main characters? These characters enjoyed an
immunity, it will be recalled, in the whodunit; here they constantly risk their lives.
Mystery has a function different from the one it had in the whodunit; it is actually a
point of departure, the main interest deriving from the second story, the one taking
place in the present.
Historically, this form of detective fiction appeared at two moments: it served as
transition between the whodunit and the thriller and it existed at the same time as the
latter. To these two periods correspond two subtypes of the suspense novel. The first,
which might be called ‘the story of the vulnerable detective,’ is mainly illustrated by
the novels of Hammett and Chandler. Its chief feature is that the detective loses his
immunity, gets beaten up, badly hurt, constantly risks his life, in short, he is inte-
grated into the universe of the other characters, instead of being an independent
observer as the reader is (we recall Van Dine’s detective-as-reader analogy). These
novels are habitually classified as thrillers because of the milieu they describe, but we
see that their composition brings them closer to suspense novels.
The second type of suspense novel has in fact sought to get rid of the conven-
tional milieu of professional crime and to return to the personal crime of the
whodunit, though conforming to the new structure. From it has resulted a novel we
might call ‘the story of the suspect-as-detective.’ In this case, a crime is committed in
the first pages and all the evidence in the hands of the police points to a certain person
(who is the main character). In order to prove his innocence, this person must himself
300 TZVETAN TODOROV

find the real culprit, even if he risks his life in doing so. We might say that, in this case,
this character is at the same time the detective, the culprit (in the eyes of the police),
and the victim (potential victim of the real murderers). Many novels by William Irish,
Patrick Quentin, and Charles Williams are constructed on this model.
It is quite difficult to say whether the forms we have just described correspond to
the stages of an evolution or else can exist simultaneously.The fact that we can encoun-
ter several types by the same author, such as Arthur Conan Doyle or Maurice Leblanc,
preceding the great flowering of detective fiction, would make us tend to the second
solution, particularly since these three forms coexist today. But it is remarkable that the
evolution of detective fiction in its broad outlines has followed precisely the succession
of these forms. We might say that at a certain point detective fiction experiences as an
unjustified burden the constraints of this or that genre and gets rid of them in order
to constitute a new code. The rule of the genre is perceived as a constraint once it
becomes pure form and is no longer justified by the structure of the whole. Hence in
novels by Hammett and Chandler, mystery had become a pure pretext, and the thriller
which succeeded the whodunit got rid of it, in order to elaborate a new form of interest,
suspense, and to concentrate on the description of a milieu.The suspense novel, which
appeared after the great years of the thriller, experienced this milieu as a useless
attribute, and retained only the suspense itself. But it has been necessary at the same
time to reinforce the plot and to re-establish the former mystery. Novels which have
tried to do without both mystery and the milieu proper to the thriller – for example,
Francis Iles’s Premeditations and Patricia Highsmith’s The Talented Mr Ripley – are too
few to be considered a separate genre.
Here we reach a final question: what is to be done with the novels which do not
fit our classification? It is no accident, it seems to me, that the reader habitually con-
siders novels such as those I have just mentioned marginal to the genre, an intermedi-
ary form between detective fiction and the novel itself. Yet if this form (or some
other) becomes the germ of a new genre of detective fiction, this will not in itself
constitute an argument against the classification proposed; as I have already said, the
new genre is not necessarily constituted by the negation of the main feature of the
old, but from a different complex of properties, not by necessity logically harmoni-
ous with the first form.

Critical connections

There is a wealth of literature on the enduring popularity of the detective novel, and
the structure and appeal of crime fiction more generally (Knight 2004; Priestman,
2003; Clarke, 2001).Tzvetan Todorov’s ‘The Typology of Detective Fiction’ is included
here because, like many of the readings in this collection, it is a defining example
within its field. Crime fiction, and perhaps especially the detective novel, remains one
of the most popular forms of media entertainment available today. Mandel’s (1984)
estimate two decades ago that more than 10,000 million copies of crime stories had
been sold worldwide since 1945 can today be increased considerably. Agatha Christie
novels alone have sold more than one billion copies, and up to one third of the fiction
published in English belongs to the crime genre (Knight, 2004). Sutherland (2007),
writing in The Times newspaper, noted the remarkable increase in the popularity of
THE TYPOLOGY OF DETECTIVE FICTION 301

the detective novel in recent years, and in particular the new ‘multinationalism’ of
crime writing which reflects the global breaking down of national boundaries (Dan
Brown’s (2003) stellar-selling The Da Vinci Code is a case in point). Given its appar-
ently growing popularity, its clear prevalence in libraries and bookshops, and the fact
that many if not most of the crime films we watch today began life as crime novels,
this literary form clearly merits the critical attention of media criminology.
Considered alongside other readings in this collection, Todorov also illustrates the
significance of genre across different media forms and formats. The excerpts by Steve
Neale (Reading 11) and Barry Langford (Reading 26) make explicit use of genre analy-
sis in their discussions of film studies research methodologies and critical readings of the
gangster movie respectively. Both directly reference Todorov’s contribution to the
field. Brian Jarvis’ Marxist and psychoanalytic treatment of the serial killer movie
(Reading 27) is also grounded in a clear understanding of genre analysis.

Discussion questions

● What distinctive genre rules does Todorov identify for the ‘whodunit’, ‘the thriller’
and ‘the suspense novel’?
● How many other sub-genres of detective fiction can you think of? What are their
principal rules of narrative development and how do they differ structurally
from those ‘types’ of detective fiction identified by Todorov?
● Can Todorov’s method of analysis be used for any novel?

References

Clarke, J. (2001) ‘The pleasures of crime: interrogating the detective story’ in J.Muncie
and E. McLaughlin (eds) The Problem of Crime, London, Sage.
Knight, S. (2004) Crime Fiction 1800–2000, Basingstoke, Palgrave.
Mandel, E. (1984), Delightful Murder: A Social History of the Crime Story, London:
Pluto.
Priestman, M. (2003) Cambridge Companion to Crime Fiction, Cambridge: Cambridge
University Press.
Sutherland, J. (2007) ‘Hot on the Trail of Global Sleuths’, Times, April 29 2007.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy