The status of representation in Michel Butor’s
‘L’emploi du temps’ and ‘La modification’
Michel Butor has for a long time been regarded as one of the principal practitioners of the Nouveau roman. But, as the nouveau romanciers have become increasingly involved in theorizing their own practice and in defining themselves, as a group, more sharply in opposition to the prevailing literary ideology, so Butor’s position within the group has become rather ambiguous. This is essentially because his novels stress the conscious reintegration of the individual subject with ‘his’ cultural heritage, rather than the radical subversion of that culture by a body of new texts. In fairness to Butor, it is true to say that he sees reintegration only as a necessary first step towards change and liberation; but then it is also true that his novels are weighted far more heavily on the side of reintegration, the project of changing the world being not much more than the note of rather pious hope on which each of the novels ends.
Thus Ricardou complains that ‘la mise en jeu de l’ensemble culturel dans le texte (de Butor) . . . ne conteste pas du tout le domaine culturel. Au contraire, elle l’enrichit de relations nouvelles dans son cadre même’;1 and Robbe-Grillet’s rather more brutal formulation of Butor’s position is ‘il cherche toujours à retrouver un accord possible du moi et du monde, alors que pour nous ça n’a plus guère de sens.’2
The conflict hinges to a large extent on the concept of representation, since Butor defines culture as a system of collective representations of reality; and in his novels, the search for the ‘accord possible’ takes the form of exploring the different possible representations of the world which can be constructed by the subject, and which will in turn define for him his possibilities of action.
In other words, the idea of representation is rejected by Robbe-Grillet3 and Ricardou, whereas it is one of Butor’s main concerns. But this is not to say that his novels are simply representational; rather, they work on the problematics of representation—foregrounding it until it becomes a principal theme in all of the novels. The texts enact the gradual building up of a particular organization of the world around a particular individual. In fact, this process of construction and ‘modification’ is superficially almost indistinguishable from the textual production advocated by Ricardou and others in the place of representation. It is the modification whereby the given entity of ‘reality’—Bleston, the railway carriage, the geography class—is gradually invested with an accumulation of meanings which shift around as the text builds up parallels and contrasts, as it defines its lacunae and then fills them in, and as it operates a continual recontextualization of its elements.
The agent of the process is given in La modification as ‘une machine mentale.’ It seems reasonable to link this phrase with the ‘machines à écrire’ which the hero spends his time promoting; and from there we could go on to identify the text as a whole with a ‘writing machine,’ which would lead us to a notion of pure textual productivity to which Ricardou and Robbe-Grillet would quite happily subscribe. Also, it is obvious that Butor employs many of the textual devices most characteristic of the Nouveau roman: the play on identity and difference, the use of repetition, the treatment of chronology, even the use of a deliberately stereotyped representation of reality (the ‘Courrier du coeur’ aspect of La modification) as a starting point for the work of the text.
There is, however, an important difference, which is that the end-product is also a representation. That is, the work on the initial representation is assimilated completely to the mental explorations of the hero, who is in turn represented with perfect psychological realism as a character in the novel. If one could define La modification, for instance, as the textual production of a representation, there would be no conflict between Butor and Robbe-Grillet; but it is in fact a representation of a textual/psychological production of a representation—and the ‘textual/psychological’ ambiguity, which comes out precisely in the difference between ‘machine mentale’ and ‘machine à écrire’ is central to the larger ambiguity of Butor’s position vis-à-vis the rest of the nouveau romanciers.
One specific instance of this is raised in L’emploi du temps by the enigmatic and inexhaustible nature of Bleston. The city always ‘knows’ more than Revel or the reader, and this poses the problem of how to evoke the existence of things which have no positive presence in the text. That is, the representation must be seen to be inadequate to its object. Since the object has no real existence against which this could be judged, the representation must contain within itself clear signs of its incompleteness. It is this incompleteness which is ambiguous, because, on the one hand, it is a move towards the open-endedness of textual production, in so far as it prevents a definitive closure of the text: and indeed, the spectacularly arbitrary ending of L’emploi du temps suggests that in principle it could go on for ever. But at the same time it is an extremely effective representational device, evoking en creux the objective and inexhaustible existence of further things-to-be-known and situating them clearly in relation to what is already know. Thus the main lacuna in L’emploi du temps is 29 February; this is a ‘known unknown’ whose place is firmly fixed in the existing pattern, where it functions as a deliberate blank, signifying that there is something more to be discovered in the representation (rather than produced in the text).
On the whole, then, it would seem that the representation is paradoxically strengthened by its own incompleteness—at least in this case. But the novel then reveals a further twist on the same theme. In terms of the schema of ‘representation of a production of a representation’, the middle term—the represented (textual) production—is supplied here by Revel’s writing of his diary. This is structured by the series of dates on which he writes—as opposed to the series of dates he writes about, which belong to the third term—the narrator’s representation—and which includes the missing 29 February. In contrast to this heavily emphasized lacuna, there is also another far less obvious gap in the series of dates on which he writes; in other words, a gap in the textual production. The entry for 4 September begins: ‘Dans le texte daté du jeudi 3 juillet que je lisais hier, j’ai décelé une importante lacune . . .’ However, there is in fact no section dated 3 July: the text jumps from Wednesday 2 to Friday 4, so that the ‘lacune’ mentioned on 4 September, although it seems to refer to an omission in what has been recorded, turns out to be in reality a gap in the writing process itself. No explanation is given for this, except that the sentence immediately preceding the gap (i.e. the end of the entry for 2 July) is: ‘Je vais remettre le négatif entre les pages du roman.’ The fact that there are two missing dates—29 February and 3 July—and hence that both the representation and the textual production contain lacunae, but that these occur at different points, serves as an image of the non-coincidence of the two processes: textual production cannot be entirely assimilated to representation, but has at least some distinct weight of its own.
The definition given earlier—a representation of a production of a representation—implies that the productive element of the text is that which maintains the distance between the two levels of representation. It remains to be shown how that distance operates rather differently in La modification as compared with L'emploi du temps, and how as a result the status of the two novels as representations is also different.
Representation presupposes an object and a subject, both given in advance. Thus representational (as opposed to ‘abstract’) art is usually defined as being of something, as bearing some fairly simple relation to an object in the world. But a representation is also always for someone—it implies a predefined ‘representing’ subject. Therefore, corresponding to the two levels of representation, one can locate two representing subjects in Butor’s novels: the author, and the narrator, who is also a represented subject (i.e. part of the author’s representation). One index of the distance between the two levels will thus be the degree to which these two subjects overlap or, conversely, diverge.
In both L’emploi du temps and La modification, the production of meaning emerges from a process of interaction between the mind of the narrator and external reality, and both terms are relevant to both novels. There is, however, a difference in the degree to which the ‘input’ of external reality is already itself structured, and hence the degree to which the structuring operations of the narrator’s mind are constrained.
In La modification the input is fragmentary and more or less random: Delmont calls it ‘cette donne du jeu.’ All the metonymic connections whereby a meaningful articulation emerges are produced by the ‘machine mentale (qui) s’est constituée . . . faisant glisser l’une sur l’autre les régions de mon existence.’ This appears ambiguous at first because while ‘mentale’ suggests that it is the narrator’s mind which is doing the ‘programming,’ the reflexive verb seems to attribute it to some other, deliberately undefined agent. This use of the impersonal reflexive is almost obsessive throughout La modification: ‘ce grand labeur qui se poursuit en vous’; ‘le mouvement qui s’est produit dans votre esprit’; etc. Significantly, the reflexive verb always occurs together with a reference to the narrator—‘en vous’, ‘dans votre esprit’—thus making it clear that the process is not going on outside him; the fact that the psychological subject never coincides with the syntactic subject of the sentence simply means that the former can never be seen as the conscious agent of the process. The narrator is presented as a divided subject engaged in a kind of unconscious bricolage of whatever images happen to be available, and it is this bricolage which is responsible for the production of meanings. The fact that the patterns (e.g. railways /falling rain/telegraph wires/ musical score) can be accounted for entirely without recourse to anything other than the narrator’s unconscious mind means that La modification can be read quite coherently as a realist discourse, of a psychological kind: a sophisticated but non-problematic representation of a certain model of subjectivity.
The situation is different, and less clear, in L’emploi du temps. Here it is impossible to identify the source of the production of meaning completely with the unconscious mind of Revel, because although some of the links are subjective—his identifying himself with Theseus, for example—there is also a whole configuration of mysterious connections which are presented as existing objectively, outside Revel’s mind. They centre on the detective story, whose plot of fratricide refers to the Old Cathedral with its window dedicated to Cain, who is the patron saint of Bleston: and also, via the Bailey sisters to whom the narrator lends the book, it is connected with Richard Tenn whose house is identical to that described in the novel and whose brother died in a mysterious accident, which gives Richard Tenn a motive for wanting to murder the author of the detective story; similarly, the fact that Mrs Jenkins is the daughter of the architect of the New Cathedral so unflatteringly described in the story, and that the fly set in glass which she wears as a ring is a copy of the fly carved in the Cathedral, and that there is also a real fly tormenting Burton in hospital—none of these coincidences and causal connections are the product of Revel’s mind.
There are, obviously, objective coincidences in La modification as well, such as that between the Guichet du Louvre and the Piazza Navone, each of which has one Egyptian, one classical and one Christian monument. These, however, are all static; the development of the images is wholly subjective; whereas in L’emploi du temps some of the coincidences are events which evolve independently of Revel. The contrast between Revel trying to find out about other people’s real lives, and Delmont inventing stories about the other people in his compartment, also points to the difference between the two novels.
So far, there is nothing in this which infringes the conventions of realist representation: one can imagine a perfectly realistic novel about a young Frenchman spending a year in Manchester and getting involved in a lot of mysterious and possibly criminal goings-on. But L’emploi du temps is not presented like this. The patterns of coincidence and motivation are all seen as significant in the sense of being intentional; they are deliberately set up by the city of Bleston itself, which is thus constituted as another subject. In other words, confronting the realistic subject Revel and his attempts to construct a representation of Bleston-as-object, we find the magic subject Bleston, sometimes referred to as ‘Circé,’ whose ‘gigantesque sorcellerie insidieuse’ allows it to construct a far more powerful and sinister representation of itself as object, one which affects all the inhabitants: James Jenkins speaks of the ‘pouvoir’ of the streets resulting in ‘une sorte de peur permanente,’ thus corroborating what could otherwise be taken as simply a figment of Revel’s imagination.
The failure of his attempts is attributed to Bleston’s malevolence; he is reduced to passivity, and, like Theseus in the labyrinth, is obliged to follow ‘une piste tracée à mon intention . . . pour mieux me perdre.’ The fact that the central mystery of Burton’s accident is eventually seen to be just a red herring is taken as yet another sign of ‘une volonté mauvaise’: Bleston outwitting Revel by spinning a web of misleading clues. The borderline between illusion and reality is blurred, as Revel comes to realize that ‘toutes les illusions par lesquelles tu auras su m’égarer, finalement font aussi bien partie de ta réalité que les aspects de toi-même que tu t’avoues.’
That is, the reality of Bleston is not so much factual as mythical. Some of the particular arrangements of reality produced by Bleston are illusions; but the basic assumption that it is indeed Bleston and not Revel which is producing the pattern is never presented as an illusion. Throughout the novel Bleston is accorded the status of an enigmatic and mythical subject.
The relation between Revel and Bleston is thus not one of subject to object, but of intersubjectivity: he speaks to the city more than he does about it. The process of representation has become an inter-subjective contest.4 The original double level of representation (author/narrator) is thus now complicated by the presence within the novel itself of two subjects—a realistic and a magical or mythical one—fighting over the representation of an object which is identical with one of them (the realistic Bleston-object and the mythical Bleston-subject).
This treatment of myth is very different from that which we find in La modification, which moves steadily towards Delmont’s realization that ‘le mythe romain’ is nothing more than an impossible nostalgia for an empire, that is for a world unified around a central city: ‘le temps où le monde avait un centre.’ There is also the idea that this desire is in turn merely a projection in bad faith onto the geographical world of his desire to recreate himself as a unified subject. In this way the mythical status of Rome is challenged; it is shown as a problem which needs to be analyzed, not in its own terms but on the level of Delmont’s individual psychology: ‘ce qui'il vous faudrait maintenant examiner à loisir et de sang-froid, c’est l’assise et le volume réel de ce mythe que Rome est pour vous . . .’
In contrast, Bleston’s mythical status and quasi-magical powers are allowed to stand unqualified in any way by the text. The effect of this is to ‘block’ a realist reading of L'emploi du temps, since—despite the many elements in the text which are realistic in a completely conventional sense—any such reading comes up against the obstacle of the subjective autonomy attributed to Bleston.
As a result of these various factors, the structure of representations (the author’s, the narrator’s, and the gap between them) operates very differently in the two novels. In La modification, author and narrator are engaged in the same kind of process: constructing representations. Butor is writing a representation of Delmont’s representation, but since the text contains nothing extraneous to the latter’s perspective, there is no obvious difference between the two levels. Author and narrator coincide perfectly, and so the two levels appear to collapse into one. There results a kind of optical illusion, which the novel reproduces in miniature in the image of the ‘tableau de Pannini,’ which has ‘ceci de remarquable qu’il n’y a aucune différence de matière sensible entre les objets représentés comme réel et ceux représentés comme peints.’ The ‘double’ representation—in the picture, the painting of paintings, in the novel the representation of the narrator’s representation—is indistinguishable from the simple version.
Similarly, the congruence is reinforced by the circularity of the novel’s structure: the book which Delmont finally decides to write is to be understood as identical to the book which Butor has in fact written, so that the narrator’s project can be superimposed exactly on the author’s achievement.
But in L'emploi du temps the two levels come apart. In the first place, this is because the author is here doing something essentially different from the narrator: Revel is passively immersed in a pre-existing mass of obscure connections, whereas Butor is, obviously, constructing them. This in itself merely deprives L’emploi du temps of the particular type of reinforcement and streamlining which in La modification is secured by the homogeneity of the two levels of representation. However, the nature of the connections raises a further point. The accumulation of coincidences is simply so unlikely—so ‘unrealistic’—that it seems as though the author is deliberately building up a pattern of symbolic relations which will stretch the conventions of naturalistic fiction to breaking point. Each separate incident is perfectly naturalistic; but taken together, they are perceived by the narrator as evidence for the existence of some mysterious agent manipulating his environment. Were it not for Revel’s interpretation, which assumes the necessity of a subject logically prior to any production of patterns of meaning, one could argue that this proliferation of links was in fact an exercise in textual production, and that its invraisemblance was a kind of deliberate provocation and subversion of realist representation. This interpretation is somewhat forestalled by the narrator’s attribution of the patterns to a second representing subject, Bleston, thus ensuring that all elements in the text are recuperated into the process of representation. However, the ambiguous status of the novel is such that this second subject in fact poses more problems as to the validity of the representation. In much the same way as the above-mentioned lack of homogeneity between author’s and narrator’s modes of activity, the presence of two represented subjects further disrupts the congruence between the two levels, which in La modification is to a large extent responsible for the smoothness of the overall representation. But what is even more crucial is that one of these subjects is realistic and the other fantastic, and that neither surrenders completely to the other. If Revel entered completely into Bleston’s fantastic world, the fact of him being a psychologically realistic character would simply enhance, by contrast, the latter’s impact (in fact this is a standard device of fantastic fiction). But his continual struggle and partial success in imposing his own realistic representation means that the novel as a whole cannot function coherently as a representation of either a realistic or a fantastic kind, but hovers between the two. It is a ‘broken mirror.’5
Returning to the general issue of the representational versus the purely textual status of the two novels, the position would seem to be as follows: in La modification the process of constructing representations is treated as a psychological one, and is represented as taking place entirely within the narrator’s unconscious mind. The novel can, therefore, be read within the framework of psychological realism; and this obscures the aspect of textual work on representation. But L'emploi du temps, for the reasons given above, does not altogether succeed as representational discourse, and to the extent that its points of incoherence and relative lack of homogeneity as representation invite a different mode of reading, suggesting even that Revel and Bleston are perhaps merely devices whereby the text produces its own structures of meaning, to this extent at least, the space occupied by textual production here becomes visible.
1 From the discussion at the Colloque de Cerisy held on the Nouveau roman in 1971.
3 Robbe-Grillet’s novels use representational elements as starting points; but the work of the text consists in a process of deconstruction which gradually ‘de-realizes’ them.
4 A recent episode of the television serial, Dr Who, showed exactly the same situation: Dr Who is engaged in mental combat with a renegade Time Lord who imposes his own version of reality on him; this ‘representation’ has Dr Who plunging to his death down a cliff face; but half-way down, by a supreme effort of will and imagination, he manages to ‘represent’ himself back into his space-ship.
5 See Pierre Macherey, ‘Lénine critique de Tolstoi’ in Pour une théorie de la production littéraire, Maspero, 1966.