Chapter 1 Postmodern fiction: theory and practice


Postmodern fiction can be considered as a response to the kind of socio-historical changes discussed in the previous chapter. Larry McCaffrey, for example, argues that with the assassination of President J. F. Kennedy on 22 November 1963, ‘[p]ostmodernism was officially ushered in – at least in the United States – since that was the day that symbolically signaled the end of a certain kind of optimism and naïvete in our collective unconsciousness, the end of certain verities and assurances that had helped shape our notion of what fiction should be’ (McCaffrey, 1986, xii).

There can be no doubt that literary change works ‘symptomatically’, as a result of what occurs at a wider social and cultural level. But if there is any validity in the category ‘postmodern fiction’ it comes from using it for more than simply identifying particular related themes within a group of texts. Likewise, as well as being a symptom of wide-ranging cultural change, postmodern fiction needs to be understood in terms of specific currents within literary theory and practice. This means considering postmodern fiction in terms of form rather than context, assessing how social and cultural change might prompt changes in what fiction does and how it positions its readers to respond to it. This is the aim of this chapter, which is divided into three main sections. The first will consider the postmodern critique of ‘mimesis’ (the practice which informs realism), the second will set out the qualities that typify postmodern fiction as a result, and the third section will explore how all this relates to the key question of reading.

An incredulity towards realism

Realism and modernism

The extent to which postmodern fiction is post-modernist – in other words, the precise ways in which it might be judged as following on from modernist fiction – is, as I suggested in the introduction to this book, a complex issue. Throughout the postmodernism debate there have always been passionately expressed differences of opinion about whether postmodernism really amounts to a departure from modernism or whether it simply continues with concerns originally dealt with by modernist writers. The technical innovations of a writer like Beckett, for example, have been plausibly identified by various critics as both modernist and postmodernist, while modernist attitudes and techniques can be identified in contemporary writers who might otherwise be considered in tune with the postmodern (such as Jeanette Winterson, Ian McEwan, or James Kelman). Instead of becoming immersed in the debate about the relationship between modernist and postmodernist fiction, however, I think that it is more useful here to concentrate on one particular element which both movements share: a dissatisfaction with nineteenth-century realism. More precisely, the modernist approach to literary realism had profound consequences for how those writers who followed after modernism conceived of and wrote their own fiction.

Realism is a mode of production in literature, art and film which attempts to sustain the illusion that the fictional world we view or read about is a plausible version of the real one, replicating how it looks, how people in it behave, the kind of things which happen to them. A realist novel is often described as presenting us with a ‘slice of life’, as if the text has cut out a particular segment of reality – either in the past or the present – and served it to us so we can taste life as it is or was in some other place than our own. In particular, realism depends upon the practice of mimesis, the Greek term for ‘imitation’ (brought into literary theory by Aristotle), the idea that art and literature can reproduce aspects of the real world.

But as its suffix ‘-ism’ suggests, realism is more than an aesthetic practice but a system of belief which revolves around the conviction that the work of art not only is capable of replicating the ‘sensible’ world (i.e. whatever we can experience through our senses), but has a duty to do so. Realist ideology asserts that art and literature should reflect life and the world soberly, in precise detail, so that we can learn from or analyse it rather than becoming swept up by idealistic and escapist flights of fancy.

Though realism has been a part of literature since Chaucer, it is especially associated with the novel, and in particular with the extraordinary pan-European production of rich and profound novels from the 1830s to the 1870s by the likes of Balzac and Flaubert in France, Tolstoy in Russia, and Dickens and George Eliot in England. This fiction demonstrated the realist philosophy to masterful effect. As such it became the principal object of critique for modernist novelists in the early part of the twentieth century. The key impulse behind modernist writing is encapsulated in the poet Ezra Pound's famous command that writers of the age ‘make it new’. Rather than simply extending the possibilities for the novel in the name of producing something different, however, modernist novelists such as James Joyce, Virginia Woolf, and Joseph Conrad believed that subjectivity (the experience of being a human being, or social ‘subject’) ought to be rendered more accurately than it was in the nineteenth-century novel.

Their work consequently deploys techniques which convey how the conscious mind experiences reality not just as something that can be measured by universal norms, but as something deeply personal and particular. The ‘stream of consciousness’ technique, for example, perhaps the most famous modernist innovation in fiction, represents the contents of a character's conscious mind ‘directly’ without being mediated by a narrator. Its effect is typical of modernism in that it limits the traditional (realist) narrator's role as ‘mediator’, whose job is to present the fictional world to the reader by framing it and shaping our responses to it. Instead it plunges the reader directly into the fictional world with only limited guidance or bearings.

Postmodernism and the French nouveau roman

Although what is known as ‘high’ modernism, the literature which represents most consistently and confidently the values of the modernist era, was relatively shortlived (lasting from around 1890 to 1930), it had a profound effect on writers who came after. It ‘destabilized’ the widely held conception of the novel as the form geared up to presenting accurately the nature of the relationship between the individual and society, and prompted an increased self-consciousness amongst novelists about the practice of writing fiction, which often took the form of scepticism about the very function or possibility of realism. Even writers who were happy to continue in the footsteps of the great nineteenth-century realists felt that after modernism they were to some extent required to defend their values and techniques against the model developed by the modernist novel.

To adapt Lyotard's famous statement, we could define postmodern fiction as writing which is shaped in some way by an incredulity towards realism – a state of mind which does not necessarily conclude that representing the postmodern world accurately, realistically, is no longer desirable, but is convinced that the act of representation cannot be performed as unselfconsciously and wholeheartedly as it was in the nineteenth century.

This incredulity is articulated most powerfully by practitioners of the French nouveau roman, or ‘New Novel’, in the mid-1950s, whose approach to fiction was to prove highly influential on both British and American ‘experimental’ novelists in the 1960s and 1970s, and on literary and cultural theorists in France and America in the same period (such as Roland Barthes and Susan Sontag, considered later in the chapter). In her seminal essay ‘The Age of Suspicion’ (1956) one of these writers, Nathalie Sarraute, argued that modernism was pivotal in what she considered a fundamental shift from confidence to suspicion – on the part of both writer and reader – which defined modern literature (Sarraute, 1963). The distinctive feature of the modernist novel, she suggests, referring mainly to the work of Joyce, Kafka, and Proust and its new approach to character and psychology, is that it amounts to a rejection of the pretence that the world depicted in a work of fiction, its characters and its story, is real. This is manifested mainly by the modernist attitude to character. Put simply, for Sarraute, the modern author no longer believes in his or her characters any more – and even if he or she does then the reader no longer will.

Sarraute suggests that the most distinctive features of modernist fiction are an anonymous first-person narrator/ protagonist who is somehow ‘everything yet nothing’, and secondary characters who are consequently ‘deprived of their own existence’ by this narrator (Sarraute, 1963, 58–9). The reader is therefore unable to conceive of these figures as recognizably human, still less to identify with them (in the kind of empathetic relationship which is central to the effect of nineteenth-century realism). The fictional world no longer seems to exist objectively, but comes to seem, as in Proust, wholly the creation of the narrator's ego. This results in a new relation between reader and author, as the reader is no longer able to rely on the author to guide him or her around the world of the novel. In fact, it seems as if the author himself is exploring this world for the first time. Rather than the world of the novel being presented to the reader as if fully created, ready to be inhabited, it is as if both novelist and reader are engaged in exploring fictional territory which is new to both. As a result the creation of the text is the result of a collaboration between author and reader rather than a kind of ‘gift’ presented from one to the other.

Unlike many critics and writers of the time, Sarraute regards this new difficult relationship between author and reader as a measure of the health of the novel rather than an indication of its imminent demise. The same applies to Alain Robbe-Grillet, besides Sarraute the other key figure in the nouveau roman. His collection of essays, For a New Novel, first published in 1955, was even more influential than Sarraute's following its English translation in 1963, and provides a complementary expression of the twentieth-century writer's disbelief in realism. In his essay ‘Time and Description in Fiction Today’ (1963), for example, he notes how nineteenth-century novels ‘are crammed with houses, furnishings, costumes, exhaustively and scrupulously described, not to mention faces, bodies, etc.’ But where the original function of all this description was ‘to make the reader see’, convince him or her ‘of the objective existence – outside literature – of a world which the novelist seemed merely to reproduce, to copy, to transmit, as if one were dealing with a chronicle, a biography, a document of some kind’, description does not work in the same way for the twentieth-century writer – or, he implies, reader. Description used to ‘reproduce a pre-existing reality; it now asserts its creative function’. Where ‘once it made us see things, now it seems to destroy them’ (Robbe-Grillet, 1989b, 146–7).

For Robbe-Grillet, the function of the novel has changed in the twentieth century. Where the realists assumed the novel was a tool geared towards reproducing reality, he insists – in his 1955 essay ‘From Realism to Reality’ – that the novel is not a tool, not something ‘conceived with a view to a task defined in advance’. Its purpose is not to ‘translate’ that which exists before or outside it, not to ‘explain’ or to ‘express’ reality, or to ‘respect the truth’ (Robbe-Grillet, 1989a, 160). The novel is not meant to inform us about reality but to constitute reality – in other words to create an aesthetic world which exists separately from the real world and does not necessarily correspond to it.

To illustrate this point, Robbe-Grillet gives the example of how he himself had succumbed to the temptations of what he calls ‘the realistic illusion’. While writing his 1955 novel The Voyeur he was at one point struggling to describe exactly how seagulls fly and decided to use a trip to the coast of Brittany to refresh his memory: ‘But from the first gull I saw, I understood my error: on the one hand, the gulls I now saw had only very confused relations with those I was describing in my book, and on the other it couldn't have mattered less to me whether they did or not. The only gulls that mattered to me at that moment were those which were inside my head’ (Robbe-Grillet, 1989a, 161–2). Even though the gulls he was describing bore a relation to reality, they had taken on their own reality, or rather, they had become ‘somehow more real because they were now imaginary’ (162). So while it will always be natural for a novelist to try to make the world in his novel as real as possible – to guard against comments from readers such as ‘a jealous husband doesn't behave like the one in [Robbe-Grillet's novel] Jealousy’ or ‘your soldier lost In The Labyrinth isn't wearing his military insignia in the right place’ – at the same time his or her chief intention ought not to be to transcribe but to construct (Robbe-Grillet, 1989a, 162).

Where postmodern fiction begins

Readers of fiction will doubtless respond to such claims about a disbelief in realism in different ways, but in making them Robbe-Grillet and Sarraute figured as more than just apologists for the nouveau roman. In fact they were representative of a much more widespread anxiety amongst contemporary writers, not least British ones.

A similar disbelief in realism is the motivating force behind John Fowles's parody of nineteenth-century realism, The French Lieutenant's Woman, published in 1969 (and analysed in Chapter 4). It is expressed most directly in the extraordinary authorial outburst which constitutes the novel's notorious chapter 13. Having spent the previous twelve chapters portraying, in more or less faithful nineteenth-century realist style, a detailed world inhabited by characters the reader can identify with, Fowles suddenly interrupts the action to inform the reader that his story is ‘all imagination’, its characters ‘never existed outside my own mind’, and, because he lives ‘in the age of Alain Robbe-Grillet and Roland Barthes’, his novel cannot adhere to the traditional model (Fowles, 1996, 97).

Another example is B. S. Johnson, whose introduction to his collection Aren't You Rather Young to be Writing Your Memoirs? (1973) articulates his conviction that the realist tradition was fundamentally dishonest, not only in attempting to present a reality quite unlike the one that actually pertained in the late twentieth century, but in doing so pretending that there were ‘real’ people in the novel, and that all their actions could somehow be linked together. In his own fiction, he makes it clear that the only real character is himself, acknowledging all others bluntly as authorial puppets. He resists the temptations of a conventional plot, out of a certainty that ‘[t]elling stories really is telling lies’ (Johnson, 1990, 153). A close parallel to Fowles's chapter 13 comes in Johnson's 1966 novel Albert Angelo. Just over ten pages before the end, a paragraph is begun in conventional narrative style – ‘Albert lazed at his drawingboard before the great window. Nearly seven weeks’ summer holiday lay ahead of him…’ – before the author abruptly intervenes: ‘–oh, fuck all this LYING!’ The next chapter, appropriately titled ‘Disintegration’ continues where it left off, its jumbled language illustrative of the breakdown of the realist paradigm: ‘– fuck all this lying look what im really trying to write about is writing not all this stuff about architecture trying to say something about writing about my writing im my hero’ (Johnson, 1964, 167–8).

Postmodern fiction, then, is rooted in the response of a range of writers and critics from the mid-1950s to the early 1970s to the way the modernist novel transformed the possibilities of fiction, most notably in their inability to conceive of realism without some degree of suspicion or disbelief. Fowles and Johnson are typical of these writers (as are British contemporaries such as Angela Carter and J. G. Ballard, and American writers like John Barth, Robert Coover, William Gass, and Donald Barthelme) and the label ‘postmodern’ is applicable to them because of the way their own fiction is underpinned by a scepticism towards realism. By contrast, the actual novels produced by Sarraute, Robbe-Grillet, and the other nouveau romanciers have tended to be regarded by critics (for good reason, I think) as more a late-flowering of modernism than the beginnings of postmodernism ‘proper’. Postmodern fiction tends to be marked by an ambivalence towards realism rather than a desire to reject it outright.

To put it differently, it would be a mistake to regard postmodernism as the ‘opposite’ of realism, or as the equivalent of ‘experimentalism’ in fiction. Andrzej Gasiorek has exposed the simplistic nature of such polarized classifications, noting, à propos the ‘realism v. experiment’ debate of the immediate post-war decades in British fiction, that even the most apparently experimental writers were actually engaged in reconceptualizing realism rather than destroying it and, conversely, the most apparently traditional writers were busily revitalizing an older form rather than simply imitating it (Gasiorek, 1995).

In any case, as Robbe-Grillet put it, ‘[a]ll novelists believe they are realists. None ever calls himself [sic] abstract, illusionistic, chimerical, fantastic, falsitical’. Realism, he suggests, is not ‘a theory, defined without ambiguity’ but an ideology ‘brandished’ by novelists against one another in order to champion their own efforts to depict ‘the real’. ‘[I]t has always been the same’, Robbe-Grillet writes: ‘out of a concern for realism each new literary school has sought to destroy the one which preceded it’ (Robbe-Grillet, 1989a, 157–8). This was true of modernism. For all their formal innovations it is clear that Joyce and Woolf were striving to depict everyday life in a manner more appropriate to their age than that of the nineteenth-century novel. In a different way it is true of postmodernism.

But one thing postmodern fiction is trying to be ‘realistic’ about, and grapples with at some level within its pages, is what fiction is and what its function should be. Postmodernism rests on the assumption that fiction – no matter how realist or experimental – is always, to use Robbe-Grillet's terms, a matter of ‘constructing’ rather than ‘transcribing’. Transcription is in fact impossible because the act of representing something external to the text actually ensures that a separate, aesthetic version of it is created in the pages of the novel, and therefore in the minds of the writer and reader – as in the example of Robbe-Grillet's gulls. Postmodern writing recognizes, either implicitly or explicitly, that it is no longer possible to indulge in the kind of pretence about the possibility of ‘transcription’ which is central to realism.

A cautionary note is required here. For it is important not to assume that realism is a ‘simplistic’ kind of writing, nor a kind of naïve but necessary step on the way to the production of the ‘perfect novel’, nor that its authors were devious or manipulative ideologues. Nineteenth-century realist writers were the products of a particular worldview, just as postmodern writers are. In any case, if we look at it from the perspective of critics such as Robbe-Grillet or Sarraute, the achievement of the great realist writers is actually very difficult: making prose a kind of transparent window through which we can view the represented world of fiction is actually far less natural than producing fiction which foregrounds its own existence as text. The richness and symbolic potential of language means that it is hard for language not to draw attention to itself, whether through an unusual metaphor, a complex formulation, or an unusual choice of word.

We must also be careful not to conclude that the reader of a realist text is naïve or gullible for swallowing the realist illusion wholesale. Of course every reader of a realist text knows that the world they ‘enter’ as they read is not ‘real’. Realism works because of a tacit contract established between reader and author: the reader agrees to ‘suspend disbelief’ and play the game that what he or she is presented with is ‘real’, on condition that the author makes the illusion as convincing as possible.

In fact it is this contract, this game, which postmodernism spoils. More precisely postmodernism objects to two key pretences which sustain the realist text:

(1) that the fictional world it creates exists in its entirety, is analogous to the real world, and that writing is, consequently, ‘referential’;
(2) that the story it tells is ‘natural’ and ‘singular’, a matter of the narrator simply mediating an ‘existing’ story.

My view is that postmodern writing does not reject these pretences on ideological grounds (though many postmodern writers are suspicious of the ideology behind realism), as did late-1970s critics like Catherine Belsey and Colin MacCabe who objected to the way ‘classic realism’ falsified the reality of the relationship between the individual human being and society (Belsey, 1980; MacCabe, 1974). Rather it opposed them on what we might call theoretical or aesthetic grounds. What I mean by this is that postmodern writers seem to assume that there is something fundamentally illogical and untenable about the way the story is transmitted to the reader in realism. To suggest why, we can now consider in more depth the theoretical problems with these two realist pretences.

The fictional world is not like ours

Writing fiction – any fiction, no matter if it is realist, modernist, or postmodernist – is, by definition, an act of creating a world. The world of a realist novel is to be imagined as analogous to our own. Although fictional, it is peopled by characters who might just as well exist in our world, in settings and amidst objects which we could plausibly find there, and who are subject to events and situations which could justifiably occur in ours – even though they tend to be more dramatic or symbolic than in our world.

Yet of course this fictional world is not the real world; it is ‘heterocosmic’, a ‘world in itself, independent, complete, autonomous’ (Bradley, 1926, 5). Indeed the very nature of the author's and reader's ‘contract’ is in fact an implicit admission that there is a difference between the two. Somewhat paradoxically, mimesis emphasizes the distinction between real and fictional worlds as much as it points to the similarity. As the postmodern theorist Brian McHale puts it, ‘For the real world to be reflected in the mirror of literary mimesis, the imitation must be distinguishable from the imitated…A mimetic relation is one of similarity, not identity, and similarity implies difference’ (McHale, 1987, 28).

In fact McHale's theory shows that the relationship between the heterocosmic world and the real world is more complex than this, for the heterocosm is not as autonomous as it might seem. It has in fact, he says, citing the narrative theorist Hrushovski, a ‘dual referential allegiance’ (McHale, 1987, 29). For it to appear realistic, to ‘work’ in the same way as our world, the fictional world is a continuum, with a consistent internal field of reference, meaning that things in the world – from objects to the causal links between events – relate plausibly to each other. But at the same time sustaining the realist illusion requires that it also point outside this to an external field of reference, namely our world – ‘the objective world, the body of historical fact or scientific theory, an ideology or philosophy, other texts, and so on’ (29). This dual referential logic is easily demonstrated in the classic realist novel when the same elements exist in both the real and the heterocosmic world. The village of Middlemarch, in George Eliot's 1871 novel of that name, for example, is fictitious, though it is set in the Midlands, which is a real part of England. The region is thus part of both the internal and the external field of reference in each text.

The apparent autonomy of the fictional world is also compromised by the fact that, as the phenomenological literary theorist Roman Ingarden points out, it requires the consciousness of the reader to actually bring it into being through the act of imagining it. It both does and does not exist ‘by itself’. Middlemarch, that is, only exists when the reader imagines it, but he or she cannot imagine it without Eliot's descriptions. The world presented in a novel is also unlike the real world in that its objects may be ambiguous, for example, if the sentences describing them are ambiguous. Objects can also be presented with what McHale terms ‘an emotional “coloration”’ making them more than simply objects amongst others in the fictional world, but privileged ones (McHale, 1987, 32). Furthermore, while the reader may reasonably assume that Middlemarch is a complete, independently functioning ‘world’ including places not actually described in the text as well as other inhabitants it doesn't mention, it is also strangely incomplete. Ingarden argues that the world fiction projects is not a complete and fully realized one but partially indeterminate. The elements which exist are only those which have been described in the language of the text so that ‘[i]t is always as if a beam of light were illuminating a part of a region, the remainder of which disappears in an indeterminate cloud but is still there in its indeterminacy’ (McHale, 1987, 31). To illustrate Ingarden's point McHale uses an episode in Gilbert Sorrentino's postmodern novel, Mulligan Stew (1979), in which two characters explore the house their author has created for them and are disturbed to find that the rooms outside the living room are simply not there, and a staircase leads only into a kind of ‘haziness’. ‘All houses in fiction are like this,’ McHale points out, ‘partly specified, partly left vague’ (McHale, 1987, 32). Usually, however – in most realist fictions – the narrator and author of the text ensure that neither the reader nor the characters are aware of this. They remain silent about it.

Narrative is not ‘natural’

Fictional worlds, then, are more complex than realist texts acknowledge. So is narrative. The established structuralist definition of narrative is that any narrative is divided into three parts: the story (a set of events), recounted in a discourse (a process of narration) in which the events of the story are selected and arranged in a particular order (the plot) (Chatman, 1978, Genette, 1980, Rimmon-Kenan, 1983). Narrative is really a two-way process of construction by which the writer assembles events into a particular order, and so does the reader. The logic is most obviously illustrated in the detective story, where the writer tells ‘what really happened’ – though not in a straightforward way, but by using artifice to disguise or omit crucial details, conversations, and events. The reader's job is to ‘recuperate’ the tantalizing fragments of the story, to piece it all together, but s/he must not be allowed to do so too easily or the suspense will be lost. Nor must the author obviously withhold crucial details from the reader or s/he will feel cheated. The fundamental rule of the detective story is that the truth must be concealed but must also be present in disguised form, to be ‘fair’ to the reader (Bayard, 2000). Consequently the author of the detective story must engage in a process of concealment and distraction in order to put up a smokescreen.

But analysing how detective fiction works underlines the fact that reading all fiction is to be subject to similar manipulation. All narrative constantly involves artifice: telling a story is not an innocent act, involving a natural sequence of events which are simply ‘extracted’ and recounted. Rather narrative involves selection, organization, and interpretation on the part of the narrator. Similarly, reading is not a simple matter of ‘receiving’ a narrative passively, but requires a certain degree of activity: for example, we respond to repetitive elements in the text, such as events and symbols.

Nothing is natural in a narrative. As Roland Barthes makes clear in his 1966 essay, ‘Introduction to the Structural Analysis of Narratives’, ‘everything, down to the slightest detail, [has] a meaning…Even were a detail to appear irretrievably insignificant, resistant to all functionality, it would nonetheless end up with precisely the meaning of absurdity or uselessness: everything has a meaning, or nothing has’ (Barthes, 1977a, 89). As well as the details in a narrative this total ‘functionality’ applies to the sequence of events. ‘The mainspring of narrative’, Barthes argues, is the confusion between ‘consecution and consequence’, where ‘what comes after’, in a temporal sense, is interpreted by the reader as being ‘what is caused by’ (Barthes, 1977a, 94). He remarks that if a telephone rings in a narrative – unlike in everyday life – then whether or not it is answered is significant in some way to the narrative. Every event, chance or otherwise, is meaningful. Even if it is just a chance wrong number (as happens in Paul Auster's postmodern detective story City of Glass) it will have causal effects. As Barthes says, ‘what is noted is by definition notable’ in narrative (Barthes, 1977a, 89). Narrative is a world in which nothing is ‘natural’, everything is significant either because it has been arranged that way by the author/narrator or because whatever has been arranged will create particular effects.

The overall functionality and partiality of narrative, the fact that the narrative process is always a matter of rhetoric, always subjective, is suppressed or disavowed by realism. Narrators select and interpret continually, deciding the order in which to place narrative events, how to describe them and the narrative world. No matter how objective a narrator claims to be, he or she is inevitably partial. But a consequence of this is that a narrator cannot recount every event which occurs during the time of a given narrative. Time simply contains too many events for this to be possible. Narrative, as Peter Brooks has said, is therefore always a perspective on a story rather than a record of every single event (Brooks, 1993, 105).

There is an important parallel here with McHale's insistence (via Ingarden) that the represented world is incomplete and unrealistic, and must therefore be responded to differently than the real world. There are elements of the story which are indeterminate, which may be filled in in the reader's act of ‘concretizing’ the text. This is demonstrated by Pierre Bayard's Who Killed Roger Ackroyd? (1998), a brilliantly eccentric re-reading of Agatha Christie's The Murder of Roger Ackroyd, which argues outrageously that Hercule Poirot actually gets it wrong in the case which makes up that novel and points to a new suspect. His analysis of one genre, detective fiction, shows how all narrative trades on the logic of the ‘lie by omission’, the principle that to preserve the mystery in a detective story, elements of the story have to be left out, and are only included at the end. This is why Sheppard, the famously unreliable narrator of The Murder of Roger Ackroyd, is able to fool us for so long. As Bayard says, he always tells the truth. It is just that he omits crucial details from his account – such as the murder itself. But the brilliance of Bayard's book is that it shows how because this is a property of all narrative, given that no narrative, however impartial or detailed, is able to contain every event that occurs, it means that every story is potentially opened up to the kind of self-confessedly ‘deluded’ reading Bayard performs on The Murder of Roger Ackroyd. Without necessarily going so far as Bayard, it follows that we can reasonably assume that just as there are parts of the house in Sorrentino's Mulligan Stew which remain undescribed but which can plausibly be assumed to exist, so there are elements of every story which are unnarrated but can plausibly be assumed to have taken place.

Modernism, referentiality, and the avant-garde

What are the implications of all this? The fact that the fictional world and the narrative are indeterminate and subject to potentially infinite expansion effectively mean that realism is unstable, incapable of doing what it claims. The critique is delivered succinctly in the following passage from Barthes's ‘Introduction to the Structural Analysis of Narrative’. Though he is writing from the perspective of narrative theory rather than seeking to classify movements within literature, still less champion experimental or postmodern writing, Barthes's words make explicit the implied grounds for the postmodern incredulity towards realism:

Claims regarding the ‘realism’ of narrative are therefore to be discounted.…The function of the narrative is not to ‘represent’, it is to constitute a spectacle still very enigmatic for us but in any case not of a mimetic order.…Narrative does not show, does not imitate; the passion which may excite us in reading a novel is not that of a ‘vision’ (in actual fact, we do not ‘see’ anything). Rather it is that of meaning, that of a higher order of relation which also has its emotions, its hopes, its dangers, its triumphs. ‘What takes place’ in a narrative is from the referential (reality) point of view literally nothing; ‘what happens’ is language alone, the adventure of language, the unceasing celebration of its coming.

(Barthes, 1977a, 123–4)

Naturally, this ‘truth’ makes itself felt in postmodern fiction through aesthetic practice rather than carefully prepared theoretical critique. The best example is the work of Samuel Beckett, whose fiction in many ways represents the beginning of postmodernism in literature. His Trilogy of 1951–7 (Molloy, Malone Dies, The Unnamable) is preoccupied by the efforts of a narrator to retain control of the world of his narrative. As a result each novel repeatedly reminds its reader that the fictional world is not equivalent to the real world. At one point in Molloy, for example, the narrator tries to explain in precise detail – in the manner of realist narration – what happened when Molloy deals the charcoal burner ‘a good dint on the skull’ with one of his crutches:

This is how I went about it. I carefully chose the most favourable position, a few paces from my body, with my back of course turned to it. Then, nicely balanced on my crutches, I began to swing, backwards, forwards, feet pressed together, or rather legs pressed together, for how could I press my feet together, with my legs in the state they were? But how could I press my legs together, in the state they were? I pressed them together, that's all I can tell you. Take it or leave it. Or I didn't press them together. What can that possibly matter? I swung, that's all that matters…

(Beckett, 1994a, 84)

It is a passage that comically and powerfully questions the need for narrative to adhere to the details of the real world. The narrator's words force us simultaneously to imagine the fictional world he is presenting and thus affirm its existence while also to question his reliability in describing it. If, as it is logical to assume, there exist elements of a world and a narrative about which we are not directly informed, then the potential for interpreting the fictional world and story becomes infinite and beyond the author's control.

We must remember here that it is not only postmodern writing that is driven by the flaws in realist assumptions about its fictional world. Modernist fiction frequently problematizes the referential, producing language which refers to itself as much as to any fictional world. A prime example would be Finnegan's Wake, the last work by the greatest modernist writer, James Joyce, a profound influence on Beckett. This novel forsakes the referential almost completely as it is written in a private language, a code which has to be cracked in order to understand it. The complexity of Finnegan's Wake's language means that even though there is a fictional world, characters and a kind of discernible plot, its words refer directly only to other words. The reader is continually conscious of this as s/he reads because s/he must translate or decode the private language in order to get to the represented world and narrative.

Finnegan's Wake thus offers one response to the pretences of realism: abandoning altogether the illusion that a novel's language refers to anything other than itself. But there is a danger here, and this is one reason why Finnegan's Wake can be regarded as the apex of fictional experimentation in the novel: after it there is nowhere to go. Joyce's novel would seem to conform to Eco's theory about the pattern replayed throughout literary history by which the avant-garde fights back against the way the past ‘conditions us, harries us, blackmails us’ (Eco, 2002, 111) by destroying or cancelling it. It is the culmination of modernism's destruction of the realist model of George Eliot or Balzac, problematizing the referential function taken for granted by the previous generation of novelists until ‘it has produced a metalanguage that speaks of its impossible texts’ (111). As an impossible text Finnegan's Wake heralds, in Eco's terms (though he doesn't mention Joyce's novel), the limit-point of the avant-garde cancellation of the past.

Postmodern fiction – which in Eco's theory is the solution to the avant-garde impasse, not the moment of impasse itself – subscribes as much as Finnegan's Wake to the critique of realism summarized by the passage from Barthes quoted above. However, because it must keep producing art rather than lapse into silence, and also because postmodernism has a strong desire to analyse contemporary reality (which in one way all the postmodern authors featured in this book do) it does not abandon the referential function but preserves it ironically, as Eco suggests, interrogating it while continuing to use it, continually examining the complex nature of the represented world and the narrative in fiction.

What postmodern fiction does

Without taking things as far as Finnegan's Wake, how does postmodernism remind us of the complexity and instability of the fictional world and the loaded functionality of narrative? To answer this question we need to consider in more detail what postmodern fiction actually does that distinguishes it from other kinds. This section, then, introduces ‘metafiction’, the most distinctive formal practice employed by postmodern writers (as demonstrated most persuasively by Patricia Waugh in her book Metafiction: The Theory and Practice of Self-Conscious Fiction [Waugh, 1984]), but first examines two different and very influential attempts to establish a ‘poetics of postmodernism’, by Linda Hutcheon and Brian McHale respectively, the two theorists who have made the most enduring impact on theories of postmodernism in fiction.

A poetics is the practice of breaking down a series of related texts into their constituent parts so that a taxonomy of features can be built up. These features can then be regarded as the essential properties of a particular genre or grouping of texts. While postmodern fiction is too diverse to be a genre, examples of it do, as I have been suggesting, share certain key features.

The value of both Hutcheon's and McHale's descriptions of how postmodern fiction works is that each preserves a flexibility which highlights what is distinctive about much of the fiction produced in the post-war period without suggesting that it is unique to this period. What distinguishes postmodern fiction is the degree to which these elements are present. Postmodernism does not do anything fiction has not done before; it is not innovative in this respect. It is rather, to return to the point about fiction's ‘dominant’ which I made in the Introduction (and which is drawn from McHale's appropriation of Jakobson's concept), a question of which elements are most obsessively considered.

Hutcheon and the ‘double-coding’ of postmodern fiction

Hutcheon's insistence about postmodernism's inherent doubleness (introduced in the Introduction), its adherence to the logic of ‘both…and…’ rather than ‘either…or…’ offers a useful way of conceiving of how the postmodern engages with realism and modernism. Postmodernism, she argues, is simultaneously ‘intensively self-reflexive and parodic, yet it also attempts to root itself in that which both reflexivity and parody appear to short-circuit: the historical world’ (1988a, x). It is an aesthetic practice which results from the paradox created at moments when ‘modernist aesthetic autonomy and self-reflexivity come up against a counterforce in the form of a grounding in the historical, social, and political world’ (1988a, ix).

She takes as her model for this postmodern architecture, explored by other theorists of ‘double-coding’ such as Charles Jencks and Paolo Portoghesi. Postmodern architecture, we recall from the previous chapter, is a kind of ‘text’ which those who look at or inhabit can ‘read’ as a kind of commentary on previous forms of architecture and aesthetic practice, while also being, much more practically, something which has to function in the real world at a particular point in history. Obviously, this latter point does not apply to fiction, whose ‘function’ is much more mysterious than architecture. Nevertheless, according to Hutcheon, postmodern fiction exhibits a similar need to engage in dialogue with past and present, the aesthetic and the political.

She is quick to point out that the doubleness she explores is not unique to the postmodern era, and can be found in Don Quixote or Shakespeare. However, she points to two significant differences between earlier times in which the self-reflexive and the historical combine in a similarly paradoxical way and the postmodern. First, postmodernism constantly treats this with a distinctive ‘attendant irony’ (Hutcheon, 1988a, x). Irony enables writers to continue working within particular discourses while simultaneously managing to contest them. Second, the presence of irony in contemporary art is ‘obsessively recurring’ (Hutcheon, 1988a, xi) – in other words postmodernism in fiction does what previous fiction has done, but at an extreme level.

For Hutcheon, postmodern fiction is both referential and self-reflexive; at the same time a preservation of some realist values and a shattering critique of them. Postmodern writing and art is frequently parodic and this is natural, given that parody is similarly double, ‘for it paradoxically both incorporates and challenges that which it parodies’ (Hutcheon, 1988a, 11). This doubleness relates to the position of the reader, who is invited to interpret the work freely, on the one hand, but on the other is required to adhere to the imposed logic of the writer's (political) agenda.

For Hutcheon one of the distinctive characteristics of the postmodern is its combination of theory and aesthetic practice. Postmodern fiction is undoubtedly art – that is, an aesthetic object which entertains and amuses us, makes us think, etc. – but at the same time engages in a form of theorizing. We might call this ‘theory in practice’, partly didactic, but also a form of writing which invites the reader to consider the theoretical questions it raises for themselves. How it goes about this is clearest when we examine (in Chapter 4) ‘historiographic metafiction’, the category of fiction Hutcheon regards as most representative of postmodernism.

Hutcheon's theory about the doubleness of postmodernism is a useful way of describing the relation between modernism and postmodernism. Rather than regarding postmodernism as an absolute break with modernism or realism we can argue that it both breaks with modernist conventions and continues with them. It certainly does not seek to put into practice a wholesale rejection of realism. Indeed she argues that ‘the ideology of postmodernism is paradoxical, for it depends upon and draws its power from that which it contests. It is not truly radical; nor is it truly oppositional’ (Hutcheon, 1988a, 120).

McHale's poetics of postmodernism

While Hutcheon's poetics of postmodernism is ambitious in its attempt to construct ‘a flexible conceptual structure which could at once constitute and contain postmodern culture and our discourses about and adjacent to it’ (Hutcheon, 1988a, ix), Brian McHale's is, on the face of it, more modest. He approaches postmodernism purely from the perspective of a literary historian, acknowledging that this means his system's ‘explanatory scheme is entirely internal to the literary-historical dynamics and does not respond in any systematic way to larger historical developments’ (McHale and Neagu, 2006, n.p.) His description of what postmodernism does is essentially that postmodern fiction foregrounds the problems or inconsistencies regarding the creation of a fictional world which we explored above – elements which are integral to every work of fiction but are passed over or suppressed in some, such as realist ones. To explain how, he sets up a useful comparison between modernism and postmodernism which is underpinned by Jakobson's idea of ‘the dominant’, ‘the focusing component of a work of art’, shifting through literary history which ‘rules, determines, and transforms the remaining components and guarantees the “integrity” of the structure’ (McHale, 1987, 6).

McHale's theory is that the ‘dominant’, or ‘focusing component’ of modernism is ‘epistemological’, or to do with knowing:

[M]odernist fiction deploys strategies which engage and foreground questions such as …: ‘How can I interpret this world of which I might be a part? And what am I in it?’…What is there to be known? Who knows it? How do they know it, and with what degree of certainty? How is knowledge transmitted from one knower to another, and with what degree of reliability? How does the object of knowledge change as it passes from knower to knower? What are the limits of the knowable? And so on.

(McHale, 1987, 9)

The dominant of postmodernism, by contrast, is ‘ontological’, to do with being:

[P]ostmodernist fiction deploys strategies which engage and foreground questions like …: ‘Which world is this? What is to be done in it? Which of my selves is to do it?’ Other typical postmodernist questions bear either on the ontology of the literary text itself or on the ontology of the world which it projects, for instance: What is a world? What kinds of world are there, how are they constituted, and how do they differ? What happens when different kinds of world are placed in confrontation, or when boundaries between worlds are violated? What is the mode of existence of the world (or worlds) it projects? How is a projected world structured? And so on.

(McHale, 1987, 10)

McHale's choice of terms, the ontological and epistemological, can be slightly misleading, for if you ‘google’ them or look them up in a dictionary profound philosophical debates will be cited that do not seem to relate to the kind of things McHale is writing about. What is at issue in his theory is not really a matter of philosophical fundamentals, but the question of what we are invited to do as we read a text: what are the interpretive solutions available to us as we read particular examples of fiction? His answer is that some of these are to do with knowing (epistemological ones) and some are to do with being (ontological ones). More precisely, the key to understanding McHale's theory is to remember that the epistemological and ontological questions which various texts invite us to ask are to be asked of the fictional world – for example, raising questions about the fictionality of the work which creates this world.

McHale argues that detective fiction is ‘the epistemological genre par excellence’ (McHale, 1987, 9) because the detective story typically revolves around precisely the kind of questions he thinks are representative of modernism. Solving the mystery in a classic detective story is an epistemological task as detective and sidekick examine the evidence to determine what can be known about a particular crime. There is no doubt about the fact that a crime has occurred. A postmodern detective story, on the other hand (and there are plenty of examples of these, as I shall show in Chapter 7) is one in which the epistemological questions ‘tip over’ into ontological ones, and the detective is frequently left asking disturbing questions such as: was there a crime? If so, in which world did it occur?

For McHale the key genre which exemplifies the ontologically dominated text is science fiction. The reason for this is that science fiction stages ‘“close encounters” between different worlds, placing them in confrontation’ and, as a result, ‘foregrounds their respective structures and disparities between them’ (McHale, 1987, 60). Actually, McHale argues, confronting our known world with something which comes from outside or beyond that world is something that fiction does by definition. Every single work of fiction, no matter the genre, includes characters who did not exist and events which did not happen in the real world. This is of course even true (perhaps it is especially true) of realistic modes of fiction. Nineteenth-century realism, for example, for all its commitment to replicating the world we know, is really asking us to inhabit a parallel version of our world in which unreal yet plausible characters and events feature.

This leads McHale to make a persuasive parallel between the historical novel and sci-fi, because the former actually involves quite a radical ‘violation of ontological boundaries’. It tends to fictionalize the lives of real people, such as Napoleon or Richard Nixon. Yet ‘[t]raditional historical novels strive to suppress these violations, to hide the ontological “seams” between fictional projections and real-world facts. They do so by tactfully avoiding contradictions between their versions of historical figures and the familiar facts of these figures's careers, and by making the background norms governing their projected worlds conform to accepted real-world norms’ (McHale, 1987, 16–17). This is why sci-fi, according to McHale ‘serves as a source of materials and models for postmodernist writers’ and he mentions a number which we will cover in this book: Burroughs, Vonnegut, Pynchon, ‘even Beckett and Nabokov’ (McHale, 1987, 16).

Metafiction

Metafiction is the main technical device used in postmodern fiction. It may be defined as ‘fiction about fiction’ – fiction, that is, which is about itself or about fiction rather than anything else. John Barth described his novel Giles Goat-Boy (1966) as one of those ‘novels which imitate the form of the Novel, by an author who imitates the role of Author’ (Barth, 2002, 145). Metafiction is fiction (and other kinds of art such as film or visual art) which is ‘self-conscious’, that is, aware of itself as fiction (as if it has its own consciousness), or ‘self-reflexive’ or ‘self-referential’ fiction, that which reflects on or refers to itself as a work of fiction rather than pretending it is offering the reader an insight into the real world. More precisely we might define metafiction as fiction that in some way foregrounds its own status as artificial construct, especially by drawing attention to its form. The effect of metafiction is principally to draw attention to the frames involved in fiction, which are usually concealed by realism.

Patricia Waugh begins her discussion of frames in Metafiction by quoting from the OED definition of ‘frame’: a ‘construction, constitution, building; established order, plan, system…underlying support or essential substructure of anything’. She goes on to claim that both modernism and postmodernism, ‘begin with the view that both the historical world and works of art are organized and perceived through such structures or “frames”. Both recognize further that the distinction between “framed” and “unframed” cannot in the end be made’ (Waugh, 1984, 28). Waugh's statement makes the crucial link between postmodernism as a socio-cultural phenomenon – as discussed in the Introduction to this book – and postmodernism as an aesthetic practice. We suspect that our lives are ‘framed’, not necessarily in any sinister sense (though the connotations of paranoia do have a persistent relevance in postmodern fiction, as we shall see in due course) but because we experience the world as mediated through a range of discursive and narrative constructs, especially from culture, media and advertising.

The most obvious way of thinking of the ‘frame’ in a work of art is to consider the actual frame in a realist painting, a landscape for example. Its function is to show us what is represented in the picture. But it also provides the boundaries of what we can see of the pictured world. The implication, as with the fictional world created by any narrative, is that what we can see within the frame is not the totality of the represented world but only a small section of something much larger. The implication is that if we could peer through the frame and look to either side of the picture or above and below, we would see other aspects of the landscape, potentially in an unbroken continuum that accords with Euclidean laws of geometry. An equivalent of the frame in a painting would be the camera in a film or a TV programme. The ‘natural’ assumption here too – that is, the one proffered by realist forms of art, which still predominate in film and TV drama – would be similar, that the camera allows our gaze to roam around a fully realistic and natural world. But framing is not a natural nor an innocent act, and it is the job of the painter, those involved in the mise-en-scène in film or TV (director, cameraman, editor, etc.), or the author to decide which parts of the represented world to frame and how and when to do it. These visual examples make it easy to acknowledge that it is not possible to present any world or narrative without some degree of framing. This calls to mind a rather paranoid-sounding remark by the philosopher Friedrich Nietzsche: ‘When we are confronted with any manifestation which someone has permitted us to see, we may ask: what is it meant to conceal? What is it meant to draw our attention from?’ (Nietzsche, 1964, 358). Acknowledging the importance of framing in fiction means recognizing that anything we see in art is because we’re allowed to see it by the author. What is more, in order to do this some kind of artificial device is used.

In fiction the frame is really the narration of the text – the element which takes place at the extra-diegetic level. Narrative framing is the means by which the fictional world is made accessible to the real world, and as such is a kind of portal, a viewing screen through which the reader can observe and ‘enter’ the fictional world – or at least the elements of it determined by the author – and experience what takes place there (the narrative).

Waugh identifies three main types of contemporary metafiction. The first is where a particular convention of the novel is upset, for example the way Fowles subverts nineteenth-century omniscient narration in The French Lieutenant's Woman (1969). The second type is where a text parodies a specific earlier text or recognizable fictional mode, such as John Barth's parody of the eighteenth-century novel in The Sot-Weed Factor (1960), while the third type is where a text, such as Richard Brautigan's Trout Fishing in America (1967), attempts in a more subtle manner to create alternative linguistic structures or evoke established forms by inviting the reader to draw on his or her knowledge of traditional literary conventions (Waugh, 1984, 4).

The effect of metafiction is often to foreground and problematize the act of ‘framing’ in both fiction and the real world. In Muriel Spark's 1953 novel The Comforters, for example, the young novelist Caroline becomes aware that she is a character in someone else's novel because she can hear noises from upstairs which turn out to be Spark tapping on the keys of her keyboard, writing Caroline's ‘life’. Throughout Beckett's Murphy (1938) – which on the face of it preserves realist staples such as well-rounded characters and linear plot – there are a number of references to the practice of writing and reading which remind the reader that the world of Murphy is not real: for example, when the narrator comments ‘All the puppets in this book whinge sooner or later, except Murphy, who is not a puppet’ (Beckett, 1973, 71).

Other metafictional texts actively indulge in what the sociologist Erving Goffman calls ‘frame-breaking’, where the frames through which the fictional world is presented to the reader are actually dismantled or shattered. The most powerful way this is accomplished in metafiction is when the author of a fiction suddenly ‘intrudes’ into the fiction itself, seeming to move from the extradiegetic to the diegetic level (although technically this is an impossibility because the diegetic version of the author can be no more than an avatar, and the extradiegetic version remains during the intrusion). The phone call to a wrong number which starts it all in Paul Auster's City of Glass asks for a ‘Mr Paul Auster, of the Auster Detective Agency’ (Auster, 1988, 7). When reading this line, having been briefly immersed in a more or less realist, intriguing story, the reader of City of Glass is jolted into confusion, realizing that the author of the book has suddenly appeared within its pages. Depending upon how aware s/he is of the author Paul Auster the reader may quickly turn to the front cover to check the name.

Frame-breaking therefore involves foregrounding the machinery which perpetuates the illusion of fiction, ‘baring the device’ as the Russian Formalists called it, or to put it in less theoretical terms, creating the same effect as Toto the dog does in The Wizard of Oz when he pulls back the curtain to reveal the shabby ‘wizard’ creating an illusory world for his own pleasure. Its effect is ‘ontological’, in Brian McHale's terms, showing ‘[w]hat happens when different kinds of world are placed in confrontation, or when boundaries between worlds are violated’ (McHale, 1987, 10).

Since, as Waugh is careful to state, frames are part of all fiction, it follows that the frame-breaking effect of metafiction is not simply ‘postmodern’. In fact it can be found in numerous examples of modernist fiction and also the seventeenth- and eighteenth-centuries. In his discussion of novels which imitate the Novel by authors who imitate the role of Author, John Barth points out that ‘this sort of thing’ is ‘about where the genre began, with Quixote imitating Amadis of Gaul, Cervantes pretending to be the Cid Hamete Benengeli (and Alonso Quijano pretending to be Don Quixote), or Fielding parodying Richardson’ (Barth, 2002, 145).

This indicates that Ian Watt's familiar ‘rise of the novel’ thesis – that the novel was realist from its very inception in the eighteenth century because of the way social, economic, and cultural factors placed a new emphasis on collective experience (Watt, 1957) – is not as reliable as it may seem. Indeed there is a case to be made for the postmodern novel being, rather than a kind of mischievous destruction of the ‘mainstream’ realist tradition in the novel, an attempt to return the novel to its original course following a prolonged detour into realism. Two pre-nineteenth century novels in particular are worth considering for a moment because of the extent of their metafictional parody and their influence on some of the writers featured in this book.

The first is Miguel de Cervantes's Don Quixote (1605; 1615). Published a century before the first novels in the early eighteenth century, this novel seriously challenges Watt's ‘rise of the novel’ thesis, revealing it as partial and anglocentric. Besides its layering of authorial frames, the ‘sort of thing’ noted by Barth, it features narrative digressions, references to other works by Cervantes and to its own developing story, and discussions by characters about the rules of writing fiction, all of which draw attention to the narrative frames involved in presenting the story to the reader. The second example is Laurence Sterne's novel The Life and Opinions of Tristram Shandy (1759–67) which is apparently unable to tell its hero's story straightforwardly but disorders the sequence of events (his autobiography starts with the moment of his conception, and he does not actually get ‘born’ until Volume 4) and accompanies them with a range of typographical conceits, such as diagrams, blank pages, lines of asterisks and dashes, and variations in typeface.

Both texts parody the conventions of the new form, the novel – in Cervantes's case before it has even properly begun. In particular they draw attention, as do later postmodern excursions into metafiction, to the difficulties of presenting ‘reality’ in fiction, mocking the idea of narrative ‘development’, where one event leads ‘naturally’ to another.

If metafiction is not a new departure for the novel but one of its essential properties, the question is raised as to why it should begin to return so obsessively in postmodernity. Waugh's answer is a persuasive one. She argues that postmodern fiction, as theory-in-practice, seeks to demonstrate that reality as we experience it is always mediated through discursive frames. It thus complements the kinds of sociological and psychoanalytic theory we considered in the previous chapter. Metafiction reminds us that narration is a form of media. Narrative functions in a similar manner to the vast media apparatus which presents reality to us – TV and cinema, the internet, 24-hour news, newspapers, magazines, etc. Or, to turn it around, metafiction reminds us that the media relies upon the techniques and effects of narrative.

The political implications of this comparison are especially evident in the many examples of metafiction produced in the United States in the late 1960s and 1970s by the likes of Barth and Robert Coover (see Chapter 3). This was the age of the assassination of JFK and the Watergate scandal, and for all their playful ‘fabulation’, the work of these writers was underpinned by a serious commitment to exposing the fact that everyday American reality was manufactured, not by an author, but by the authorities (i.e. the government). As Malcolm Bradbury has put it, ‘What distinguishes postmodern metafiction is its relation to the public and social fictions that surround it, and its attempt to find a mode of discovery and exploration within them’ (Bradbury, 1992, 204).

The metafictional staging of a clash between real and represented worlds encourages us to pursue the implications to their logical conclusion: fiction is fictional, but no more so than reality. When it collapses the distinction between fiction and reality by causing the ‘outside world’ to seep into the world of the novel, metafiction is more subtly demonstrating that the obverse also happens, the fictional world intrudes into the real world. This might seem far-fetched until we consider the effect of having real historical personages feature in a work of fiction as they frequently do in postmodern writing. Jorge Luis Borges's short stories often set references to fictional authors alongside references to obscure real ones and confuse the reader as to whether the fictional ones are in fact real and the real ones fabrications. Similarly, the British author D. M. Thomas creates such a believable version of the psychoanalyst Sigmund Freud in his novel The White Hotel (1981), through a blend of quotations from his work and private letters and perfectly pitched counterfeit ‘Freudian’ documents, that it is difficult to read about the real Freud after reading Thomas's novel without confusing the two.

How to read postmodern fiction

What postmodern fiction does repeatedly is prevent us from passively entering the fictional world by constantly reminding us that it is a fictional world, that fictional worlds are complex, and that the way authors deal with fictional worlds might teach us something about the real world. It is clear from each of the theoretical perspectives summarized in the previous section that postmodern fiction continually challenges the reader. A postmodern text is one which – at some level at least – is aware of its own status as something we read, an aesthetic object. It doesn't pretend its world is the real world or its narrative is natural, and ensures that we cannot do the same, inviting us, indeed at times requiring us, to reconsider our relationship with the world of fiction and the story it tells. Self-conscious writing, in other words, produces self-conscious reading.

What this means is that postmodernism in fiction is not simply a matter of how authors write, but how readers read. One way in which we can conceive of postmodern literary theory and practice is as a clarion call not to writers but to readers to do things differently. Modernist literary innovation is often summed up through the poet Ezra Pound's command to writers to ‘make it new’. Postmodernism might be characterized by a more implicit but just as insistent demand to ‘read in a new way’.

Susan Sontag and the ‘erotics’ of reading

An important diagnosis of the changing conditions surrounding art in postmodernity (though she doesn't use the term) was put forward by the American cultural critic Susan Sontag in two essays written in the mid-sixties, ‘Against Interpretation’ (1964) and ‘One Culture and the New Sensibility’ (1965). The second essay in particular focuses usefully on the aesthetic consequences of major shifts in the late-twentieth-century way of life. Postmodernity, she argues, is characterized by the extreme growth of science and technology, a greater sense of ‘extreme social and physical mobility’, ‘crowdedness’, physical speed and speed of images, and the mass reproduction of art objects. This means that the function of art has consequently changed. Art is now less ‘a technique for depicting and commenting on secular reality’ and more ‘an instrument for modifying consciousness and organizing new modes of sensibility’ (Sontag, 2001, 296). In other words, in the late twentieth century, it is not so much what art is about that matters (i.e. not its subject nor ideas) but what it does and how this makes us respond to it. This also means a decline in the importance of the novel. When the function of art was to comment on secular reality, it was natural that the novel, a form which depicts the workings of social reality more powerfully than any other, perhaps even cinema, reigned supreme. Yet once the emphasis is placed on how things affect the consumer of art, it meant that fiction is superseded by modes such as cinema, visual art, and music, each of which can mirror the tumultuous late-twentieth-century social conditions in a way that impacts more powerfully and directly on the client.

Sontag nevertheless regards some writers as in tune with the new climate. She mentions Burroughs and Beckett, whose work is uninterested in offering rational commentary on social reality. But the examples she gives from visual art make her argument about what art ‘does’ rather than ‘says’ most clear. ‘Yellow and Orange’ by Mark Rothko, for example, a canvas upon which are simply two rectangular panels, one positioned above the other, the top one hazy yellow, the one below smudged orange, cannot be ‘interpreted’ in the manner of other modernist forms of visual art, such as work by Picasso. However obscure and challenging, Picasso can be explained. By contrast, it is difficult to say what Rothko's painting is ‘about’. Rather it just is; it doesn't refer to anything – except the colours ‘yellow’ and ‘orange’. So, if we wish to claim that ‘Yellow and Orange’ is ‘about’ anything we can only make the obvious claim that it is about yellow and orange.

A literary parallel might be B. S. Johnson's 1973 short story (though this label is unsatisfactory), ‘A Few Selected Sentences’. It has no obvious narrative but presents us, as the title suggests, with twenty-seven short sections of prose, some no longer than a line or a word (‘Life.’) Confronted by such a text, skilled readers naturally start looking for patterns, such as the preoccupation throughout the sections with death and sex, or the repetition of the phrase ‘Someone has to keep the records…’, and conclude that these suggest that Johnson's text is a textual record of fundamental elements of Life: sex, death, etc. (Johnson, 1989, 285).

But it is impossible to go beyond this very general and unsatisfactory reading (every text could be said to be ‘about life’) without going outside the text and considering Johnson's own life (he committed suicide in the year this was published). Johnson's text seems to resist any effort to unify the disparate elements by establishing links between them or to impose an overall explanatory narrative. So in the terms of Sontag's theory it seems we really have to read the text precisely as its title tells us to, as simply ‘a few selected sentences’ which may or may not entertain or inform us, and which cause us to reflect on the story's form not content: for example, to think of the different kinds of texts in our culture, how they mediate our reality, how they interact together, even how Johnson ‘selected’ the sentences or whether he composed them himself.

Sontag's essay ‘Against Interpretation’ contends that the new emphasis in postmodern art is on form rather than content and explores further how we should respond to it. The consumption of art is about pleasure rather than ‘edification’, fun and wit rather than seriousness or morality. One of the happy consequences of this is that enjoying art becomes an egalitarian matter rather than something snobbish. Sontag's argument is in tune here with others by prominent cultural critics of the day, most notably Leslie Fiedler, who argued in 1972 that one of the defining features of postmodern art is that it ‘closes the gap’ between high and low culture (Fiedler, 2002), in other words incorporating the forms of pop culture into its work. Sontag insists that postmodernism is not a dropping of standards but a ‘new way of looking at the world and at things in the world’ (Sontag, 2001, 303).

But it is not a simple matter of suddenly learning to take pleasure in art instead of wasting our time worrying about how to interpret it. Central to Sontag's theory is the recognition that art is a challenge. It is not something that we can now just sit back and enjoy, like a fine wine, because part and parcel of this new approach to art is that art becomes formally more complex. This is how it creates its effects. Sontag acknowledges the difficulty of postmodern art by arguing:

The commonest complaint about the narratives of Beckett or Burroughs is that they are hard to read, they are ‘boring’. But the charge of boredom is really hypocritical. There is, in a sense, no such thing as boredom. Boredom is only another name for a certain species of frustration. And the new languages which the interesting art of our time speaks are frustrating to the sensibilities of most educated people.

(Sontag, 2001, 303)

Realizing that one is not bored after all may not prove much consolation for those who do find reading Beckett or Burroughs something of a chore. But perhaps the key phrase in this provocative statement is ‘the sensibilities of most educated people’. We are educated to think of fiction in a particular way – specifically, as if it were realist fiction, and to judge all fiction according to its standards. Readers of modernist or avant-garde fiction often express their boredom or frustration (which are of course the same thing, Sontag insists) by simply saying that it is not like a ‘traditional’ novel – a neutral statement on the face of it but one which, in this context, actually conveys a withering criticism.

However, Sontag's argument is directed not so much at everyday readers of fiction, educated or otherwise, but at that specific group of specialist readers known as literary critics. It follows, she suggests, that a new sensibility in art really needs a new approach to criticism. This means that criticism becomes a matter of analysing the form rather than the content of a work of fiction. Sontag explains this by calling at the end of ‘Against Interpretation’ for an erotics rather than a hermeneutics of fiction. By this she means that we concentrate on examining how fiction creates specific kinds of enjoyment in readers rather than trying to figure out a narrative, as if it is a matter of cracking a particular code and reading the hidden message (Sontag, 2001, 13).

Now we have to acknowledge that there is something very much of its time in this argument, something unmistakably ‘1960s’ about the whole idea of art modifying consciousness and reading narrative being liberating and erotic. It happens that there is an odd ‘sexual turn’ around this time in theories of reading, most obvious and unconvincing in Robert Scholes's comparison of reading to the sexual act (Scholes, 1979). However, Sontag's use of the term ‘erotic’ is more complex, more Platonic, than Scholes's, and does not claim universality but only to be an analysis of contemporary art.

Roland Barthes and the ‘writerly’ text

Sontag's description of postmodern art and literature and her argument about how we ought to respond to it provides an interesting complement to the pronouncements by novelists at the time, such as Johnson, Ballard, and Carter, about how unsuited realism was to present-day reality. What she advocates also corresponds to the dynamic approach to reading championed by the poststructuralist theorist Roland Barthes in the late 1960s and 1970s, the ‘goal’ of which, jouissance, is the parallel to that of Sontag's ‘erotics’ of reading.

Even though Barthes never used the term ‘postmodernism’ and is not usually included in the list of postmodern thinkers, his approach to creatively reading fiction exemplifies the spirit of the postmodern version of ‘making it new’, especially in three works published in the late 1960s and early 1970s: ‘The Death of the Author’ (1968), S/Z (1970) and The Pleasure of the Text (1975).

The most important of them is his obsessive book S/Z, in which Barthes reads, line by line, an entire realist short story by Balzac, Sarrasine (1830). The result is a commentary over twice the length of the original text. Determined to dissolve the realistic text's illusion of unity, Barthes sets about ‘disrupting’ the Balzacian text, ‘manhandling [it], interrupting it’ (Barthes, 1975, 15). This involves making the text ‘plural’ by ‘cut[ting] it up’ into ‘a series of brief, contiguous fragments’ which Barthes terms ‘lexias’, each of which is ‘a stage or space in which we can observe meanings’ (Barthes, 1975, 13). S/Z is a detailed account of how one reader (Barthes himself) interprets the meanings generated by the text, like a ‘real-time’ documentary about a reading (Barthes, 1990).

The violence Barthes does to Sarrasine is far from gratuitous. He doesn't seek to elevate criticism over literature out of vanity but, as in other of his works, is dedicated to showing how an apparently transparent and non-political act like reading can become part of an ethical movement geared towards unmasking the repressive ideology created by the capitalist system. His cutting-up of Sarrasine is supported by an ideology which is most polemically expressed in ‘The Death of the Author’. Here Barthes argues that the idea of the author as the individual solely responsible for the meanings in the text is no longer tenable. The author is not the creator of his text but only a mediator – that is, someone who selects and organizes various discourses available in ‘the general text’ (in other words, language and literary convention) rather than producing anything ‘original’, in the conventional sense of this term. The outcome – as the famous last line of Barthes's essay has it – is that the birth of the reader must be at the cost of the death of the Author’ (Barthes, 1977b, 148).

S/Z provides a more systematic account of how a reader is ‘born’. Its supporting theory develops a binary model by which we can measure the level of activity which a particular work of fiction demands of its reader. The translated terms are a little confusing to us, but the concepts are useful when it comes to explaining how postmodern fiction works on its reader.

At one end of the pole is the lisible or ‘readerly’ text, so termed because it is a text which tries to confine the reader to a role as reader, one who is guided to interpretation by the narrative itself. At the other end – and this is the zone which brings Barthes himself most pleasure as a reader (as he expounds in The Pleasure of the Text, the work which is a follow-up of sorts to S/Z) – is the scriptible or ‘writerly’ text, one which does not have a single ‘closed’ meaning. Instead, with this text, readers are obliged to produce their own meanings from fragmentary or contradictory clues, thus effectively writing the text themselves (or at least co-producing its meanings). A ‘readerly’ text is a text which reads itself for us, as it were, meaning that its rhetoric guides us to respond to it passively, as consumers. A ‘writerly’ text, by contrast, is one in which the reader becomes ‘co-writer’; it is he or she who is responsible for activating its meanings. The ‘readerly’ text is finite, whereas the writerly text ‘exists nowhere’, as what it ‘is’ depends upon how it is read at any one time. Put more simply, the ‘writerly’ text is one which is less conventional, less realist, and requires us to ‘work’ in order to make sense of what is going on within its pages.

A postmodern work of fiction is an example of a writerly text. However, it is important to understand that Barthes's definitions are really rhetorical models: the closest thing to ‘readerly’ texts are nineteenth-century realist narratives, such as those produced by George Eliot, Leo Tolstoy, and Honoré de Balzac. But in using a Balzacian short story as the way of putting his theory into practice, Barthes effectively demonstrates that every text is ‘writerly’; it's just a question of degree. Similarly Pierre Bayard's reading of Christie demonstrates that a definitively readerly form of fiction, the classic detective story, can in fact be made writerly.

In his 1971 essay ‘From Work to Text’ Barthes sets up a similar binary based on two ways of conceiving of a work of fiction, in which the equivalent of the readerly text is the ‘work’ and the equivalent of the writerly one is ‘text’. The shift of emphasis, though, is reminiscent of Sontag's point about how postmodern writing is ‘frustrating to the sensibilities of most educated people’ for it makes it clear that setting out to respond to a text in a ‘writerly’ rather than ‘readerly’ manner is not something that comes naturally because of the way we are taught to read.

The Work is something that is displayed, can be seen in bookshops, libraries, exam syllabi, etc. It is a finished product which is ‘closed’, finite and only ‘moderately symbolic (its symbolic runs out, comes to a halt)’. The Text, by contrast, is demonstrated rather than displayed; it exists only in the movement of discourse, is experienced only in an activity of production by its reader. It consequently ‘goes to the limit of enunciation (rationality, readability, etc.)’, challenges received opinions, is intertextual, and ‘radically symbolic’ (Barthes, 1977c, 157–8).

Though it may seem like it (and there is perhaps some truth in this), Barthes is not trying to establish a canon of postmodern or avant-garde fiction by suggesting that some texts are ‘better’ (more open) than others. Rather he is concerned with how institutions, educational and legal, school us into a particular limited practice of reading. Where once we were taught rhetoric, and composition, since ‘the coming of democracy’ we have been taught in schools to read well – and reading here means consuming rather than playing. This institutionalized method of reading brings ‘pleasure’ but it is not the jouissance Barthes most values. Unlike simple ‘pleasure’, jouissance moves one away from their fixed subject position and all the ideological and authoritarian operations it involves. Pleasure merely confirms it (Barthes, 1975).

‘Paranoid reading’ and ‘rhizomatic reading’

Sontag's and Barthes's arguments suggest that there is a case to be made for a text's identity as ‘postmodern’ being determined by the act of reading rather than writing. There is some value in this, though it would clearly be stretching things to call certain texts postmodern – like Sarrasine, for example – simply because of how we choose to read them. Nevertheless, for all their differences in form, postmodern texts generate a self-consciousness about how we read which is in tune with the ideas about reading put forward by Sontag and Barthes. To complement their insistence that postmodern fiction invites us to concentrate on form rather than content, we need to conclude this chapter by considering two other kinds of reading practice which are relevant to the postmodern fiction considered in this book.

The first of these is what some critics (Siegel, 1976; McHale, 1992) have termed ‘paranoid reading’. This concept might usefully be considered in relation to Sarraute's notion of the ‘age of suspicion’, and also in terms of Paul Ricoeur's idea of the ‘hermeneutic of suspicion’, something that became dominant in the twentieth century – on a wider cultural level than simply literature – as a result of the influence of Marx, Nietzsche and Freud. These thinkers developed ‘a method of interpretation which assumes that the literal or surface-level meaning of a text is an effort to conceal the political interests which are served by the text. The purpose of interpretation is to strip off the concealment, unmasking those interests’ (Ricoeur, 1970, 33).

Similarly, when interpreting fiction we bring to texts a certain logic, one predicated on the idea of reading ‘beneath the surface’. We assume that texts are really saying something different deep down, that the various patterns and repetitions are not coincidences, that symbols and images are not there just to please us but because they mean something. This is a kind of reading practice which does not allow for significance, which assumes that everything must mean something. Mark Siegel argues that paranoia ‘is the condition under which most of modern literature comes to life: the author relies on the reader to find correspondences between names, colours, or the physical attributes of characters and other invisible qualities of those characters, places, and actions, while to do so in “real life” would clearly be an indication of paranoid behaviour’ (Siegel, 1976, 50). Brian McHale has built upon Siegel's insight to provide a kind of historical context for this modern approach to reading, arguing that the approach to criticism is in fact instituted as a kind of critical orthodoxy during the era of modernism, namely through its academic counterpart ‘practical criticism’, a technique taught in university English departments (variations continue to be taught even now) which involves ‘close reading’ of the text and is governed by the assumption that what it presents us with directly is only a cipher for what it is really saying deep down. To really understand a modernist or postmodernist work of art, McHale argues, we have to work on this assumption.

Postmodern fiction displays a typically double-coded attitude to paranoid reading. Unlike modernism, where paranoid methods are ‘rewarded’, postmodern works of fiction are ambivalent about the practice. As we shall see in some of the readings which follow, postmodern texts frequently invite readers to interpret them in a paranoid manner – and this is undoubtedly part of their appeal – but some deliberately frustrate their attempts or demonstrate that determining a final meaning is more complex than it seems.

The second reading-practice relevant to postmodern fiction is more playful and accepting of open-ended interpretation than paranoid reading. Postmodern narrative involves us in a process of conjecture. This is a notion I have adapted from Umberto Eco's discussion of detective fiction in his short postscript to his own postmodern detective novel, The Name of the Rose. The appeal of detective fiction, he suggests in Reflections on The Name of the Rose, is not because of some morbid fascination with the corpse or even because the genre represents the victory of order over ‘the disorder of evil’. Rather, he argues, the appeal is that of conjecture: a hypothesis formed through speculation (usually without hard evidence) that, moreover, works through narrative. The crime story presents us with a set of events which we conjecture (usually by following the lead of the detective, the reader's more enlightened avatar in the world of the text) have a logic to them, which can explain what happened and the criminal responsible.

Eco's model for how conjecture works is the labyrinth. Some labyrinths, he argues, are straightforward. The ‘classical’ version ‘does not allow anyone to get lost: you go in, arrive at the center, and then from the center you reach the exit’, while ‘the mannerist maze’ – the kind popular in theme parks and the gardens of stately homes – is a structure which, if unravelled, is ‘a kind of tree, a structure with roots, with many blind alleys. There is only one exit, but you can get it wrong’. But the labyrinth Eco is most interested in is what he calls the ‘rhizomatic’ maze (Eco, 1985, 57).

The metaphor of the rhizome (a botanical term for a kind of plant stem) is used by the philosophers Gilles Deleuze and Félix Guattari in their book A Thousand Plateaus: Capitalism and Schizophrenia (1980) to enable us to picture the relation of things to other things in philosophy, language, the arts, and social sciences – according to principles of connection, heterogeneity, multiplicity, and rupture – in a way that provides an alternative to a straightforward linear ‘surface–depth’ model of interpretation which is central to paranoid reading. Unlike the roots of a tree ‘which plots a point, fixes an order’ the value of the rhizome is that ‘any point…can be connected to anything other, and must be’ (Deleuze and Guattari, 1987, 7).

They suggest that language works in this way. While it can be understood in ‘tree-like’ terms if we work with a basic Saussurian signifier–signified model, Deleuze and Guattari's point is that because language is a living phenomenon, there are always numerous other meanings suggested, other traces of usage which deepen the meaning. Interpreting a text can work the same way. We can look for one underlying meaning which can make sense of what is ‘above’ ground, or we can accept that any text is likely to generate a whole string of meanings which are only connected by the fact that they are triggered by the same element of the text. This explains the appeal of the metaphor for Eco's comparison of labyrinths: ‘The rhizome is so constructed that every path can be connected with every other one. It has no center, no periphery, no exit, because it is potentially infinite’ (Eco, 1985, 57). An obvious comparison now would be with the internet, which can be imagined as a labyrinth with a rhizomatic structure. The internet ‘starts’ and ‘ends’ nowhere, or rather you can enter and exit from anywhere and go from any one place to any other. The rhizome, according to this model, means that connections can be made between things which are otherwise unconnected.

Eco's analogy suggests that we can understand the process of interpreting different kinds of text according to the idea of being lost in a maze. The classic detective story – for example, Conan Doyle's Sherlock Holmes stories or Agatha Christie's ‘clue-puzzle’ version – can be regarded as ‘mannerist mazes’ which require us to arrive at the conclusion, to establish the final narrative, via a lengthy process of trial and error. Reading a detective story involves going through a process of entertaining various hypotheses regarding the crime. The process is made particularly vivid in TV/cinematic adaptations of Agatha Christie's novels, in the convention where, as the detective hypothesizes about a particular scenario or suspects claim to remember an event, the episodes their words conjure up are presented on screen as if they really happened. The televisual mechanism is particularly powerful here as it is able to present the conjecture as if indistinguishable from the real.

Realist fiction and classic detective fiction typically closes down the conjectural possibilities at a certain point in the narrative – certainly by the conclusion, when the true narrative becomes apparent and replaces all the other potential versions of the truth. Postmodern detective fiction, on the other hand, has shown a fondness for keeping the possibilities open – in the manner of Borges's story ‘The Garden of Forking Paths’ (considered in the next chapter). Once activated they are not closed down. This is also the logic of Bayard's Who Killed Roger Ackroyd?, in which he refuses to accept that the paths opened up throughout Christie's novel have to remain closed once Poirot privileges one of them over the other.

Beyond detective fiction, however, I think the parallel Eco draws between conjecture in narrative and the rhizome has a much wider significance. As we shall see, postmodernist narrative frequently works according to this model. The classic example is Nabokov's Pale Fire (see Chapter 3), a text which opens a number of conjectural possibilities and leaves it impossible for us to close any of them down. Texts like this suggest that the ideal reader of postmodern fiction is less a detective and more of a traveller in space, ready to encounter different, co-existent, worlds.