Chapter 8Fact, fiction and French flights of fancyMichael Stubbs
University of Trier

Abstract
This chapter discusses a satirical novel: La septième fonction du langage by Laurent Binet, published in 2015. The book is a thriller with a deliberately absurd plot about a search for a lost manuscript which holds the secret of ultimate rhetorical power: the ability to convince anyone to do anything. Although the characters in the novel include some “real people”, such as two former Presidents of France, Stubbs argues that Binet’s characters in general embody an extreme mix of factual and fictional characteristics, and the merciless satire expressed in their mixed ontological status has left many ordinary readers and professional critics uncertain how to evaluate the novel. Stubbs employs various models of analysis, including John Searle’s observations on the logical status of fictional discourse, and contends that, while useful, this approach does not explain in full how readers distinguish fact from fiction, nor indeed how far writers can appropriately go with outrageous caricatures of living persons. In sum, the chapter shows how the novel provides textual problems which have not been solved by either literary scholars or language philosophers.
Keywords
  • caricature;
  • fictional worlds;
  • French intellectuals;
  • intertextual references;
  • ordinary versus professional readers;
  • parody;
  • satire;
  • semiotics
France … a country famous for making an art form out of ridicule …(Le Carré 2016: 249)

1.Introduction

Bill Nash is the author of a comic novel, a book on the language of humour, and a book on rhetoric (Nash 1982, 1985, 1992). I will discuss some sources of humour in a book which I am sure he would have enjoyed: a comic novel whose zany plot involves the search for a manuscript which holds the secret of ultimate rhetorical power.
My chapter is a (slightly) satirical discussion of a (very) satirical novel, La Septième Fonction du Langage by Laurent Binet (2015),1 henceforth SFL. A productive methodological strategy for investigating a topic is to use an extreme example as a case study, and SFL is extreme in several ways. First, since its pervasive mixture of fact and fiction has left some readers uncertain about which is which, it provides good material for a case study of fictional worlds. Second, since this pervasive mixture involves extravagant caricatures of a large number of famous French intellectuals, it illustrates ethical issues involved in parody and satire. Third, since these extravagant caricatures demand detailed knowledge of French cultural life in the 1970s and 1980s, it illustrates the perennial problem of how much general and specialized knowledge readers require in order to (fully?) understand texts.
Perhaps the simplest reason for discussing SFL is that it’s a very funny and enjoyable novel: a tour de force of suspense and intellectual high jinks. But satire – along with caricatures and other forms of exaggeration – is a risky strategy. Authors can never be sure that readers will get the joke, since what is said is not what is meant. So, the novel illustrates many problems of textual interpretation.
There are useful attempts to disentangle “the logical status of fictional discourse” (such as Searle 1976), but they do not explain how readers distinguish fact from fiction, or indeed how far you can appropriately go with outrageous caricatures of living persons. The novel provides textual problems which have not been solved by either literary scholars or language philosophers (and they are not solved in this article, so don’t get too excited).

2.The plot: Spoiler alert!

Binet’s novel is a thriller / detective novel / spy story, with an absurd / burlesque / crackpot / drôle / extravagant / farcical / grotesque … zany plot, about a search for a lost manuscript which holds the secret of ultimate rhetorical power: the ability to convince anyone to do anything.
The title of the novel – and the mainspring of the action – derives from Roman Jakobson’s famous model of language, which is summarized in SFL (pp. 135–40). Jakobson (1960) proposed that any speech event involves six constitutive factors. An addresser sends a message about some content to an addressee along some communication channel, encoded in a way which both sides hopefully understand. To each of these factors corresponds a function. Most communication involves all factors, but one function may be predominant. For example, an academic report may be principally concerned with the truth value of its propositions (the referential function), rather than with the addresser’s attitude to this content (the emotive function), or the elegant style of its language (the poetic function). The model has been criticized for seeing meanings too mechanically, as objects which are packaged in words, then sent along tubes to be unwrapped by recipients.
However, all that is relevant here is that Jakobson (1960: 355) also speaks, rather vaguely, of a “magic, incantatory function”. The plot imagines that this seventh magical function, which gives unlimited performative power to its possessor, has been discovered. The secret services of several countries are trying to find an unpublished manuscript which describes how the seventh function works.
The characters include “real people”: leading French politicians, in particular Giscard and Mitterrand, and cultural theorists, including Althusser, Barthes, Culler, Derrida, Eco, Foucault … Kristeva, Sollers and many other members of the intelligentsia who created what Americans called “French Theory”, plus a few well known linguists and philosophers of language, including Jakobson and Searle. Of course, you will probably agree, they are not “really” real people, but a mixture of factual and fictional characteristics who/which exist only in the world of the novel. This mixed ontological status has perplexed both ordinary readers and professional critics.
The relations in the novel between fiction and reality – my main topic in what follows – are complex. Each time I mention characters, I could distinguish between the real world in which we live (W1) and the fictional world of the novel (W2), for example, “W1-Searle” (the “real” Searle) and “W2-Searle” (the fictional Searle in SFL). But this could get tedious and pedantic, and the distinction is not always clear. So, I will generally assume that you are paying attention and can work out for yourself which is which.
The story takes place during the French presidential campaign of 1980–81. In 1980, Valéry Giscard d’Estaing, a centrist politician, was President. In 1981, he lost the election to François Mitterrand, leader of the Socialist Party. The action starts in February/March 1980 with the death of Roland Barthes, and ends after Mitterrand’s election victory in May 1981. To get the plot started, Binet imagines a parallel world which diverges from the real world. In February 1980, on his way home from lunch with Mitterrand (a year before the presidential election), Roland Barthes is knocked over in the street and dies some weeks later. This really is how he died. But suppose it was not an accident, but murder. Suppose someone suspected that Barthes possessed the only copy of the secret manuscript and killed him to gain possession. That is, Binet takes an undisputed fact, adds a bit of counter-factual speculation, and then slides into an ever more fantastic world:
Barthes was killed in a traffic accident.
But what if it was not an accident?
Then someone killed him. But who? And why?
The case is investigated by Commissaire Jacques Bayard, assisted by Simon Herzog, a young semiotician. Their pursuit of the murderer and the manuscript takes them from Paris to Bologna to Cornell University to Venice, then back to Paris and finally to Naples. On their quest, they meet many leading figures of linguistics and semiotics: including, Roland Barthes himself (dying in hospital), Roman Jakobson and Umberto Eco.
The crazy logic of counter-factuals is illustrated when Umberto Eco quotes his mother-in-law: “What would have happened if my son-in-law had not married my daughter?” (p. 292). But then if he hadn’t, he wouldn’t be her son-in-law, so she couldn’t have asked the question. But then again, if he hadn’t, then she wouldn’t be his mother-in-law, so she wouldn’t exist. But then again, if someone else had married her daughter, then he would be her son-in-law, so she could have asked the question of him. But then again, … Trying to work this out is too much for Herzog. To add some semblance of plausibility, Binet follows a strategy common to conspiracy theories, and stirs in a series of coincidences. In 1980, a bomb exploded in Bologna’s train station (p. 255), Althusser strangled his wife (p. 262), Bulgarian spies were targeting dissidents with poisoned umbrellas tips (and two characters, Tzvetan Todorov and Julia Kristeva, are both of Bulgarian origin). One of the final scenes is based on the television debate between Giscard and Mitterrand on 5 May 1981. Mitterrand is “abnormally good” (p. 441). Perhaps he had the manuscript. In an interview published in Partisan2 Binet noted that, “seven years before, [Mitterrand] was terrible in the debate with Giscard, but in the 1981 debate he was good. So I connected the seventh function to his dramatic rhetorical success”. In a conversation with Yuri Andropov (yes, he also appears in the novel), the point is made that “a good lie must be drowned in an ocean of truth” (p. 149).
The title of Barthes’ most famous article, “The death of the author”, is just another coincidence (Barthes 1967a). But other events are outrageously make-believe. In the novel, Searle – who is also searching for the manuscript – attacks and kills Derrida. In the real world they attacked each other’s work, but Searle certainly did not kill Derrida (Derrida died in a hospital in Paris in 2004, not in a cemetery in Ithaca in 1980). By this point, the characters have been joined by the entirely fictional Morris Zapp, who has wandered in from novels by David Lodge.

3.What is the book really (sic) about?

First, it is about the relation between fact and fiction. The opening sentence is: “Life is not a novel”. For readers who were not paying attention, the sentence is repeated later (p. 172). For readers who still have not got the point, Eco asks about the difference between truth in a novel and in real life (p. 248), and Morris Zapp criticizes literary critics who confuse life with literature (p. 312): even fictional characters can make serious statements.
Second, it is about language, rhetoric and semiotics. This is the focus of much of “French Theory”, the topic of the missing manuscript, and the reason for Mitterrand’s success in the television debate. Language and discourse are the raison d’être of the secret Logos Club: two players debate for or against a concept, a jury judges the winner, the loser loses a finger. Eco explains the key idea: “the person who masters discourse […] is virtually the master of the world” (p. 222).
Third, it is an attack on the high priests of postmodernism. Semiotics is explained in fairly straightforward ways, in a seminar on Barthes and James Bond (pp. 38–42) and elsewhere (pp. 14–15, 18–19, 46–49, 83–84). But otherwise, ideas about the power of discourse and the discourse of power – which are central to Foucault’s work – are reduced to a farcically literal form. And the gurus themselves are reduced to physical caricatures, devoid of intellectual credibility. At a party, Lacan utters owl-like cries, Althusser and others all talk nonsense and behave badly (pp. 152–63).
So, we have a mélange of true statements, plausible statements, counter-factuals, coincidences, and total nonsense. But how do readers know that some propositions in SFL are true (Barthes was knocked down in the street and died some weeks later), that some might conceivably be true (The driver was a Bulgarian called Yvan Delahov), and that others are batshit crazy (The murder was masterminded by Kristeva, who is in charge of a gang of killers from the Bulgarian secret service)?
I assume that the proposition about “Yvan Delahov” is invented. On 27 March 1980 Le Monde reported simply that Barthes had died as a result of being hit by a car on 25 February. A search on the Web for “Yvan Delahov” finds only references to Binet’s book.

4.The intertext

Tales of the irresistible persuasive power of the human voice go back to the Greeks and Romans. Plato was suspicious of the subversive power of rhetoric. The seductive song of the Sirens lured sailors to their deaths.3 However, like any post-modernist novel worth its salt, SFL is packed with more references than any normal reader could possibly identify, to other books, to the large world of politics, and to the “small world” of international universities.
Some references are familiar to any reader of detective novels. A death which appears to be an accident turns out to be murder. The death is investigated by two contrasting characters, detective and sidekick: compare Holmes and Watson. Jacques Bayard is a policeman with a low opinion of left-wing intellectuals. The death has a potential connection to this circle, and since Bayard knows nothing of “French Theory”, he recruits Simon Herzog, a young semiotician working at the University of Paris at Vincennes, notorious for its left-wing sympathies.
Some references are more obscure. Pierre Bayard is Professor of literature at Vincennes, and author of Qui a tué Roger Ackroyd? and L’affaire du Chien des Baskerville. To call the policeman after a professor at Vincennes, who has written books about detective stories, is presumably Binet’s private joke.
Some references are untraceable. The name Simon Herzog does not – as far as I know – allude to any real person, but he does have the same initials as Sherlock Holmes. SFL plays with oppositions and reversals. It is the sidekick Herzog – not the detective – who performs amazing feats of rapid inference, by combining Holmes’ techniques of abductive thinking with Barthes’ techniques of semiotics. By the end of the novel, Herzog has had a haircut and acquired a suit and tie, and Bayard has decided that some female members of the intelligentsia are not so bad after all.
Some references are parodies of parodies. In almost every Sherlock Holmes story, Holmes classifies people by their appearance. His trademark skill often involves inferences so detailed as to be a totally implausible running joke. From his behaviour and clothes, Herzog identifies Bayard’s background, including his service in Algeria and his failed second marriage. Elementary! (“C’est très simple”, p. 48). And he interprets the significance of objects in Giscard’s office, including a painting by Delacroix and a book about JFK: “The whole room is saturated with symbols”4 (p. 83).
Some references will be recognised by any self-respecting well-read reader. But for those who have not seen the parallel to a story of “a lost document for which people are killed” (p. 249), Eco sees a man carrying a bunch of roses, and a vision of a poisoned monk flashes through his mind. (The Name of the Rose was published in Italian in 1980, the year in which SFL takes place: just another coincidence.)
On their travels, the heroes are pursued by Bulgarian gangsters, who drive a black Citroën DS. One of Barthes’ most famous essays on the semiology of the bourgeois world is about this cult car: DS = déesse = goddess (Barthes 1957). For readers who might miss this reference it is explained – twice (pp. 54, 109). But readers have to work out other references for themselves. When Althusser hides the manuscript, he thinks of Poe and hides it in plain view (p. 177): The Purloined Letter. When Searle unleashes two savage dogs on Derrida, the scene takes place in a cemetery (pp. 354–57): The Hound of the Baskervilles. The heroes are protected by Japanese guardian angels: is this a reference to Barthes (1970), L’empire des signes, on a “fictive nation”? Could be, I’m not sure. Life is too short to trace all the references which show off Binet’s five years of research for the book (which he emphasized in an interview in Libération5), and he has undoubtedly packed in more references than I have spotted, even with help from Wikipedia.

5.The lists

It is common to find real historical characters in novels. What is different about SFL is
  • there are dozens of W2-caricatures of attested W1-personalities
  • some caricatures – although arguably in poor taste – seem not too far from what we know from biographies of the W1-character (e.g. Foucault)
  • some caricatures embody fantasies which do not correspond in any way to the attested personality of the W1-character (e.g. Searle)
  • some caricatures mix harsh evaluations – which may or may not be justified – with participation in humiliating W2-situations (e.g. Sollers: see below).
So, before I discuss the ontology of factual/fictional characters, we need some pedestrian documentation on their sheer abundance. Here is a list, in chronological order of birth, of the W1-scholars whose W2-caricatures play major roles in the novel:
Roman Jakobson (1896–1982, Russian/American linguist, literary theorist)
Jacques Lacan (1901–81, French psychoanalyst)
Roland Barthes (1915–80, French literary critic, semiotician)
Louis Althusser (1918–90, French Marxist philosopher)
Gilles Deleuze (1925–95, French philosopher)
Michel Foucault (1926–84, French philosopher, social theorist)
Pierre-Félix Guattari (1930–92, French philosopher, psychotherapist)
Jacques Derrida (1930–2004, French philosopher, literary theorist)
Umberto Eco (1932–2016, Italian philosopher, novelist, semiotician)
John Searle (1932–, American philosopher of language)
Philippe Sollers (1936–, French writer and critic; Kristeva’s husband)
Hélène Cixous (1937–, Algerian/French feminist writer, philosopher)
Tzvetan Todorov (1939–2017, Bulgarian/French structuralist, narratologist)
Julia Kristeva (1941–, Bulgarian/French critic, psychoanalyst, Sollers’ wife)
Jonathan Culler (1944–, American literary theorist)
Bernard-Henri Lévy (1948–, aka BHL, French media personality)
Judith Butler (1956–, American philosopher, gender theorist)
Butler is named only as “Judith”, presumably because in 1980 she was a postgraduate student, and not yet world famous, but she is clearly identified as a feminist of Hungarian Jewish origin (p. 317).
All the major figures of “French Theory” appear, sometimes in actual lists: in the enumeration of those attending Barthes’ funeral (p. 108), in the programme of a conference at Cornell (pp. 297–99), and in the poll of the 42 leading French intellectuals which Herzog reads in a magazine (pp. 432–33). This list really was published in Lire6 on 1 April 1981, as “Référendum: Les 42 premiers intellectuels”. Given the importance which the French attach to their intellectuals, I suppose it was not intended as an April Fool’s joke.
Many novels contain a statement such as the following, from Changing Places by David Lodge, the novel in which Morris Zapp first appears:
Although some of the locations and public events portrayed in the novel bear a certain resemblance to actual locations and events, the characters, considered either as individuals or as members of institutions, are entirely imaginary (Lodge 1975: n.p).
Given the lists above and below, such a statement would make no sense in SFL, and none occurs.
The scholars listed above will be mainly known to readers familiar with linguistics, semiotics and literary/cultural theory. Others, who get a brief mention, include Jean-Paul Sartre, who sits in a café with Françoise Sagan, and Noam Chomsky, who appears at the conference, plus Étienne Balibar, Jean Baudrillard, Pierre Bourdieu, Daniel Defert, Jean-Edern Hallier, Hervé Guibert, Mathieu Lindon, Sylvère Lotringer, Jean-François Lyotard, Paul de Man, Jeffrey Mehlman, Serge Moati, Camille Paglia, Roger Peyrefitte, Avital Ronell, and Gayatri Spivak. Further writers and politicians are likely to be immediately familiar only to French readers (I had look several of them up in Wikipedia): Jacques Attali, Robert Badinter, Régis Debray, Laurent Fabius, Jack Lang, Mathieu Lindon, Michel d’Ornano, Michel Poniatowski, and many more.
Michelangelo Antonioni debates at the Logos Club and loses; Monica Vitti bandages his finger (pp. 234–39). Romano Prodi is in the audience. (Why are they there? I have no idea.) Björn Borg, Ivan Lendl and John McEnroe appear on television to contribute comparisons between tennis, semiology and rhetoric (e.g. pp. 377–78).

6.Factual worlds and fictional worlds

So far, my SparkNotesy summaries of the plot, themes and references of SFL cover fun reasons why linguists should appreciate Binet’s novel. They can recommend to their first-year students the sections which read like a very useful Pragmatics for Dummies, on Barthes’ theory of semiotics, Jakobson’s theory of language, and Austin’s theory of illocutionary and perlocutionary (pp. 38–42, 135–40, 249–51, 345).
Time now for more serious stuff. SFL illustrates unsolved problems of textual interpretation, such as the following. The relation between fact and fiction: how do readers distinguish which is which? The relation between text and reader: does SFL make more sense if you’re French? Does the novel provide new examples which a theory of fictional worlds must solve? Can a work of fiction therefore contribute to theory?
Human languages allow us to talk about six impossible things before breakfast. In legends and myths we can talk about things which do not exist (such as unicorns and gods), or about any of the intersubjective realities (such as semiotics and religion) which are created by our talk. But it should be easy, one might think, to keep fact and fiction apart. Shakespeare wrote Macbeth to flatter King James VI, who believed in witches. But even James VI must have grasped the point that three of the actors on stage were not themselves witches, but men pretending to be women pretending to be witches. He must have understood that he and the pretend witches belong to different worlds, as we say, and he would not have had them burned at the stake.
So, no readers of SFL, one might think, should get confused (except perhaps those who visit the museum in Baker Street to see Sherlock Holmes’ pipe and deerstalker). There is the real world in which we live (W1), which provides the geographical background for the action (e.g. in Paris) and several events (e.g. the French presidential election). There is the entirely fictional world (W2), in which fictional characters (e.g. Simon Herzog) take part in fictional events (e.g. a debate at the Logos Club). There are W2-caricatures who say things which their W1-originals actually said or might have said (Derrida lectures on speech acts, p. 336–40), but who are involved in entirely fictional events (Derrida is killed by savage dogs, pp. 354–55). And there are W2-characters who say and do things which they actually said in W2 (Morris Zapp repeats his central principle: “every decoding is a new encoding”, p. 313).
However, it is impossible to list all conceivable combinations of factual/fictional characters/worlds: novels with fictional creatures in fictional settings (Hobbits in J. R. R. Tolkien’s The Lord of the Rings), novels with fictional characters in real historical settings (Edward Waverley in Walter Scott’s Waverley), novels with real characters in fictional settings (Napoleon in Leo Tolstoy’s War and Peace), and fables (in which fictional animals formulate serious moral lessons), not to mention sci-fi, myths, and religious texts (in which events violate the laws of nature). And then many myths and religious texts used to be interpreted as literally true, but are now regarded as, well, myths. The text has stayed the same, but readers’ perspectives have changed. Conversely, fiction can become reality: some countries give official religious status to the Jedi realist movement.7 May the Farce be with you! This list hardly starts to tackle all the fun which can be had with the fiction/reality distinction.

7.What is this chapter really about?

It is about the concept of fictional worlds, and it is time I dealt with the review of the literature. It is not surprising that mainstream linguistics takes almost no interest in the concept, since it is difficult to see how there could be any strictly linguistic features which distinguish factual from fictional statements. It is, however, surprising that literary theory takes so little interest in the concept. You might think that literary scholars would have long ago worked out the exact ontological nature of the fictional characters which inhabit almost every work of art: Hamlet, Mr Pickwick, Godot (or does he really not exist even in the play?) and so on. John Woods (1974) applied concepts from the philosophy of language to fictional texts, and argued that fictional sentences have a distinctive logic. For example, if we insert a fictionality operator such as “according to the story in SFL”, then the proposition “Barthes was murdered” is true. The proposition has, as it were, fictional truth conditions. From the 1970s on it was famously argued by Hayden White that it is impossible to distinguish between history and fiction. Both tell narratives, and all narratives share properties of fictionality, since their specific point of view always entails selection and interpretation. Barthes (1967b) had himself made the same point, so perhaps his ghost cannot complain if the facts of his death are turned into fantasy. More recently, in the area known as “cognitive poetics” and “text world theory”, ideas about possible worlds have been applied to literary texts, initially by Paul Werth, and then by Gilles Fauconnier, Joanna Gavins and others (e.g. Fauconnier & Turner 2002; Gavins 2007; Werth 1999). The basic idea is that we construct mental models based on all the discourse – everyday conversation, newspaper reports, novels, whatever – which we hear or read. It is in the philosophy of language that the concept has been of greatest interest. Overviews often start with work around 1900 on names and other noun phrases. Gottlob Frege distinguished denotation and reference (the name “Sherlock Holmes” has a meaning – outside the fiction – although it does not refer to any spatio-temporal entity in W1). Bertrand Russell, Peter Strawson and others discussed sentences such as “The (present) King of France is bald”, arguing that this proposition is neither true nor false, but defective, since its referent doesn’t exist and its existential presupposition therefore fails. Alexius Meinong, whom Russell much admired, is notorious for proposing an extravagant ontology, arguing that we refer to many kinds of imaginary and abstract objects, which all enjoy some sort of existence (e.g. Sherlock Holmes, Godot, right-angled triangles, the number three, the average Frenchman, Laurent Binet, etc.). From the 1960s on, Jaakko Hintikka, Saul Kripke, David Lewis and others developed work on the semantics of proper names, possible worlds, modal logic and counterfactual conditionals.
You may be relieved that I will not discuss this literature further, beyond recommending the following review articles. Lahey (2014) reviews cognitive approaches to literary stylistics. And Kroon and Voltolini (2011) and Eklund (2015) provide authoritative overviews of work in the philosophy of language. They are excellent overviews, but can do little more than record disagreement about the ontological status of fictional characters. Stating the exact nature of fictional worlds turns out to be much more difficult than my fictional James VI possibly thought (but then he was rather confused about witches). However, W1-Searle and W1-Eco provide helpful discussions.

8.Searle: A pragmatic approach?

One of the clearest starting points is an article by Searle (1976), who asks the paradoxical question: how can words in a novel have their ordinary meanings although readers do not comply with the ordinary rules for interpreting them? The author of a newspaper article, he argues, performs illocutionary acts of assertion that conform to four rules:
  1. Essential rule: the writer is committed to the truth of the proposition.
  2. Preparatory rule: s/he must be able to provide evidence for its truth.
  3. The proposition must not be obviously true to both writer and reader.
  4. Sincerity rule: the writer believes that the proposition is true.
But the author of a novel is not committed to any of these rules. So, what kind of illocutionary act is performed? The answer, says Searle (1976: 324), is obvious:
[The author] is pretending, one could say, to make an assertion, or acting as if [he/she] were making an assertion, or going through the motions of making an assertion, or imitating the making of an assertion […] without any intention to deceive.
Searle is making four points. Words in fiction have their ordinary meanings. The words do not have their ordinary illocutionary force. Fiction is pretence. There is no intention to deceive. This abstract formulation is characteristic of Searle’s approach, but odd in so far as literary language is usually regarded as having some added value, rather than being a use of language which has lost one of its ordinary functions.
Searle makes two further points, both of which could arguably be more precisely formulated. First: “[T]he identifying criterion for whether or not a text is a work of fiction must of necessity lie in the illocutionary intentions of the author” (Searle 1976: 325). Literary theorists talk of the “intentional fallacy” (Wimsatt & Beardsley 1946), namely, that one can never know the intentions of an author. Searle argues that this fallacy is a fallacy: it is absurd to ignore the author’s intentions, since if you think that a text is a novel, you already assume something about the author’s intentions. As a statement about the whole text, this is surely correct. As a whole, SFL does not refer to any reality outside the text. However, we may not know which individual propositions are intended to be taken literally.
Second: “There is no textual property, syntactical or semantic, that will identify a text as a work of fiction” (Searle 1976: 325). However, the position of formal linguistic features in a text may signal fiction. For example, some short stories and novels (only some, of course) start with sentences containing a third person pronoun with no anaphoric referent. Stanzel (1984: 160) argues that this is “inconceivable in a non-fictional text”:8 perhaps not inconceivable, but unlikely. Rather than talking of “textual properties”, it would be better to say that there are no properties of individual sentences which will identify them as fictional. Both of these points illustrate a problem in interpreting SFL. It is abundantly clear that the whole text is fictional, but often unclear which individual propositions are fictional.
Searle (1976: 332) also points out that works of fiction generally contain factual propositions. For example, in SFL the sections on Pragmatics for Dummies can be judged on their accuracy as serious academic statements. In addition, fictional texts are an amalgam of explicit content plus generally accepted beliefs and default assumptions that W2 departs minimally from W1. Much of the action in SFL happens in real streets in real places. We are told, for example, that Roland Barthes lived in the Rue Servandoni, which is close to the Jardin du Luxembourg, and that Cornell is five hours in the bus from New York City. If you’re feeling pedantic, you can check this in Google Maps.
Searle’s approach is essentially pragmatic. There need be nothing in the text itself which signals that we are reading fiction, and no meaning expressed in fiction is distinct from everyday meaning. We need contextual information, such as the title or other device – what Genette calls the “paratext” (Genette 1997) – in order to decide. On the title page of SFL is the word “Novel”.
Finally, it takes two to tango: the reader must recognise the author’s intention and join in the pretence. But what triggers the pretence? How is the reader to know which aspects of SFL have a basis in real world truth? Searle (1976: 332), in his closing sentence, points to a serious gap in literary theory: “[T]here is as yet no general theory of the mechanisms by which such serious illocutionary intentions are conveyed by pretended illocutions.” [Emphasis added.]

9.Eco: A semantic approach?

Searle asks and answers a variant of his initial question: “But how is it possible for an author to ‘create’ fictional characters out of thin air, as it were? […] By pretending to refer to people and to recount events about them.” (1976: 329). This does not get us much further. However, Eco (2001) sees the ontology of fictional characters as follows. They are intensional semiotic objects (intensional with an “s”). That is, they have no referent in the external spatio-temporal world (they have no truth value), but are created by a bundle of semantic features (which are its truth conditions). In W1 the name “Sherlock Holmes” only denotes. In the W2 of the stories it both denotes and refers. Eco recommends later work by Searle (e.g. 1995: 7–13, 2010: 17–18), which shows that the opposition objective / subjective is ambiguous between epistemic and ontological senses:
epistemically objective and subjective are predicates of judgements
ontologically objective and subjective are predicates of entities.
Epistemology concerns knowledge. Compare two propositions:
(1)
Herzog works at the University of Vincennes.
(2)
Herzog is a better detective than Bayard.
(1) is epistemically objective. It is not a matter of personal opinion. You can check it in the text of Binet’s novel. (2) is epistemically subjective. It is a matter of personal opinion.
Ontology concerns existence. Compare the referents in two subject noun phrases:
(3)
Cornell University is situated about 200 miles from New York city.
(4)
The Logos Society meets in Venice.
The referent in (3) is ontologically objective. You can check it on Google maps. The referent in (4) is ontologically subjective. It is a product of the human mind which created the text.
Woods (1974) proposed that fictional sentences have a distinctive logic, which makes them “bet-sensitive”. If I bet that (1) and (4) are true, and you bet the opposite, then I will win the bet. This distinguishes them from notorious cases such as “The King of France is bald”. There is no point in betting whether this proposition is true or not (unless you put him in a story and preface the proposition with a fictionality operator: “According to the story … the King of France is bald”). Note also that (1) and (4) would be true (in the fictional world) even if no-one had ever read Binet’s novel. This immediately refutes extreme versions of Reader Reception Theory.
One approach to these ontological questions is to say that there are many different kinds of things in the world. As Eco says (p. 247), events which seemed credible to readers of a medieval encyclopedia now seem legendary. Gabriel (2013) develops the view that there is no single reality to which all real things belong. A purely materialist view that the real world consists of the totality of spatio-temporal things misses everything which is not spatio-temporal, but conceptual, including all properties of things, all relations between them, and so on. The Federal Republic of Germany, he points out (he’s German), is not a thing which can be identified with the physical bit of land where it has been located since 1990. It includes its organization into federal states, its concept of democracy, which is enshrined in its Constitution (Grundgesetz), a written document of around 100 pages, and so on.
For Gabriel – and presumably for W1/W2-Eco – it makes no sense to talk of the world as a whole. Existence is a property of a realm or domain. Gabriel therefore talks of a plurality of domains, only one of which is the spatio-temporal domain studied by the natural sciences (which is itself not a unified domain). Lots of other things exist in their own fields of sense: numbers, citizens, republics, money, marriages, toothache, unicorns in fantasy literature, witches in Faust. It would be very strange to argue that there are no witches in Faust (or Macbeth). But there are none in Germany (or Scotland). For something to exist it has to be part of a context. Existence is always relative to particular fields of sense.
Very different kinds of objects, concrete and abstract, are referred to in the subject noun phrases in the following sentences:
  1. François Mitterrand was President of France from 1981 to 1995.
  2. The President of France cannot serve more than two consecutive terms.
  3. The average Frenchman eats almost a whole baguette every day.
  4. The Logos Club is a secret society.
  5. The Logos Club was created by Laurent Binet.
In (a) the subject noun phrase refers to a spatio-temporal entity. In (b) and (c) the subjects are different kinds of abstract things. In (d) the Logos Club is referred to from inside the story, whereas in (e) it is referred to from outside. In terms of “bet-sensitivity” (Woods 1974), adding a fictionality operator (“according to the story”) to (d) would not change the conditions of the bet, but this operator would make no sense for (e). Anyone can understand such sentences, but philosophers of language cannot explain how they work.

10.Searle and Eco (and Gabriel)

So, in a word or two:
Searle argues that authors of fiction make statements, but only pretend to make true statements. This seems rather obvious, but usefully relates to standard positions in the philosophy of language. In the terminology of Speech Acts (Searle 1969), the utterance act is real, but the illocutionary act is pretended. Works of fiction also contain sentences which are intended to be taken literally. But the puzzle remains: how do readers know which is which?
Eco argues that fictional characters are created by words in texts. This seems rather obvious, but usefully relates to standard positions in semantics, concerning the reference and denotation of proper names. In addition, Gabriel proposes that many different kinds of things exist in different contexts: the Logos Club is an abstract entity which exists only in SFL. But the puzzle remains: how do readers know which context to invoke on any given occasion?
Since there need be no textual markers of fiction, it all seems to come down to the reader’s knowledge, either specialist knowledge about “French Theory”, or historical/political knowledge about French presidents, or what anyone on the Clapham omnibus knows about this and that – which brings us back to Searle’s gloomy admission that we have no general theory of the mechanisms involved.
Let us call Searle’s view the “pretence theory” and Eco’s view the “creation theory”. The first says nothing about the purpose of the pretence. The second says nothing about the extent of the creation. The resulting problems of interpretation have been noted by both ordinary readers and also by professional critics of SFL.

11.Ordinary readers

SFL contains material to please both a mass audience and a Bildungsbürgertum (who understand words such as Bildungsbürgertum). On the one hand, it contains clichés from a pot-boiler: several murders, a car chase through Paris, killer dogs, a secret society, mustachioed thugs who look like mustachioed thugs, and a mafia boss who gets thrown into boiling mud. On the other hand, it contains a lot of semiotic theory, intertextual references, and quotes from French intellectuals.
A relatively new kind of empirical evidence about how books are interpreted is available on websites9 which allow readers to post ratings and reviews of books which they have read. They can contain dozens and sometimes hundreds of commentaries, from a few lines to short essays.
Opinions are divided on whether readers can understand SFL without expert knowledge of “French Theory”. Some readers think that the book remains readable even if not all references are recognized, and/or that readers can be expected to do their own research on unfamiliar characters (most readers would have to consult Wikipedia frequently). Some think that the humour fails if jokes have to be explained all the time, and/or that all these references to French intellectuals are too difficult for non-French readers. Others point out that central concepts in linguistics and semiotics are explained, sometimes at length. After all, Commissaire Bayard knows that he understands nothing of “toutes ces conneries”. That’s why he employs Herzog to translate for him. But – crucially – some readers ask how they can distinguish pastiche from reality if they are not familiar with the cultural background. Neither side seems able to provide knock-down arguments on whether and/or how much real-world knowledge is needed to appreciate the novel.

12.Professional critics

SFL was positively reviewed in many leading French newspapers, but not all the reactions were favourable. A popular French television talk show10 broadcast a debate between Yann Moix and Laurent Binet. Moix (himself a prize-winning novelist) criticized the vulgarity of the situations in which Binet places his characters, especially Foucault and Barthes. From the beginning of the novel, he argued, they are presented in a severely reductionist fashion, with reference to features of their personal lives, but with no indication of their academic importance to readers who may be unaware of their enormous cultural influence. SFL, he argued, fails to successfully combine popular and elitist culture, and reduces important scholars to extreme caricatures. Binet’s defence was that his characters speak, throughout the novel, in their own words, which are taken from their own publications. He claimed to have pulled off the trick of constructing a parody from the words of those who are parodied. Foucault’s publications, he argued, are well known, as are publications by Barthes, including the personal revelations in his Soirées de Paris. However – one might add – this consists of diary entries dated shortly before Barthes’ death, but published posthumously, possibly against his intentions (Barthes 1987). It would be interesting to have Simon Herzog’s semiotic analysis of body language and facial expression in the televised exchange between Moix and Binet.
In the interview in Partisan published a few weeks later, Binet agreed that some readers were shocked by his undignified fictionalization of Foucault, but argued that the scenes are based on what has long been public in biographies, and that they do not detract from his theoretical work. Moix had argued – on the contrary – that this would not be clear to readers unfamiliar with Foucault’s life and work. Again, neither side seemed able to provide knock-down arguments on whether aspects of SFL are legitimate satire or unacceptably offensive.

13.On the ethical status of fictional discourse

Binet reserves his most biting satire for his French characters, who are portrayed as the high priests of an intelligentsia who have renounced coherent critical thought. Umberto Eco (Italian) emerges as the dignified elder statesman of semiotics, Roman Jakobson (Russian-American) is gently mocked for his woollen tie and strange haircut, Jonathan Culler (American) potters around harmlessly organizing a conference, and John Searle (American) … Oh, OK, he turns out to be a murderer. But, apart from that, the foreigners all emerge reasonably unscathed by Binet’s satire. Barthes and Foucault are dead, but others are not. Binet was asked in the Partisan interview whether he was worried about being sued by the targets of his satire. He replied that in France fiction is well protected, that his publisher’s lawyers had approved the book, and that his novel “reveal[s] itself as fiction page after page”. Unfortunately for textual theory, he did not explain how this works, saying only that he had created events which no-one could take as even remotely true. This is true of the whole plot, but not of all the details. For example, Deleuze has “abnormally long fingernails” and a large collection of hats (objectively true?), is obsessed by the tennis on television (invented?), and utters philosophical banalities (Binet’s subjective opinion?). Though Deleuze’s banal philosophical question, about the problem of establishing the borderline between true and false (p. 78), is precisely the question which has led to criticism of SFL by Moix and others.
It is Sollers who is reduced to a particularly harsh caricature. He is described as displaying “histrionic dandyism”, “pathological boasting” and “clown-like” behaviour (p. 152). He gives a vacuous description of a book he is writing: “I cannot conceive of a book better informed, many-layered […] a philosophical, even metaphysical, novel with a cold and lyrical realism”. He is portrayed “as a child”, with a memory span of three minutes, who enjoys Punch and Judy shows (pp. 273–78). At a meeting of the Logos Club in Venice, he talks embarrassing nonsense (“delirious logorrhoea […] mentally disordered […] megalomaniac”, p. 416), loses the debate, and then loses more than just a finger. He had an inaccurate version of the secret manuscript.
Binet – Searle might say – is only pretending to make true statements about Sollers. But pretence is an intentional act. So, what purpose does Binet’s pretence serve? Since we cannot access his ulterior motive, it is unclear which individual propositions are intended to be true. Compare two sentences:
P1. “Sollers is like a child.”
P2. “Sollers is just a funny character.”
Considered as isolated sentences, they contain nothing which indicates whether they are fictional or serious statements. But P1 is from the novel (p. 276) and P2 is from the interview11 in Libération, where Binet was outside the protection of fiction and carried full responsibility for his utterances.
Genette (1991) proposes that a criterion of fiction is that the author is distinct from the narrator, but now author and narrator do not seem clearly distinguished. Binet argued that, as a writer himself, Sollers respects freedom of expression in the novel, and was therefore unlikely to go to court. Nevertheless, it is understandable if W1-Sollers felt unhappy with the portrayal of W2-Sollers, and aggrieved at such an ad hominem portrayal by a fellow novelist. He might at least wish to avoid Binet at the next publishers’ party (or perhaps a character called Binet will receive humiliating treatment in a future novel of his own).
There are objects which don’t exist (e.g. W2-Sollers), but which really do have properties. Sherlock Holmes never existed, but (it is true that) he really was a detective. Unicorns don’t exist, but (it is true that) they really do have just one horn. For a hundred years, philosophers of language have been trying to solve the logical puzzles raised by non-existent objects (Kroon & Voltolini 2011).

14.In lieu of a conclusion

Satire and parody are risky strategies. One reason why readers may not get the joke is that real world historical knowledge varies widely. According to a 2008 poll of 3,000 British teenagers,12 20 per cent believe that Sir Winston Churchill is a fictional character, and almost 60 per cent believe that Sherlock Holmes really existed and lived at 221B Baker Street. Well, they were just ignorant teenagers, you might say, but experts can also be misled.
In 1996, Alan Sokal, a physics professor at New York University, published a long spoof in Social Text, a journal of postmodern cultural studies. In a jargon-ridden parody of an academic article, with over 200 references and over 100 footnotes, he argued that “physical ‘reality’, no less than social ‘reality’, is at bottom a social and linguistic construct”. The hoax is well documented on the Web.13 Just a few points are relevant here. (a) Sokal’s article quoted many of the French intellectuals who appear in SFL. (b) Although the article – as Binet might say – revealed itself as a hoax on page after page, this was not recognized by the journal editors. (c) Major scholars could not agree about the implications. Steven Weinberg was “amused by the prank” (The New York Review of Books, 8 August 1996), and Thomas Nagel thought it was a “delicious hoax” (The New Republic, 12 Oct 1998), but Stanley Fish thought it was a bad joke which can “erode the foundation of trust on which science is built” (New York Times, 21 May 1996). (Stanley Fish was director of the press which publishes Social Text. Biographical knowledge about an author is not always irrelevant to understanding a text.) And SFL is not the first time that “French Theory” has been ridiculed – in sometimes abusive, burlesque, controversial, disrespectful, exorbitant, farcical, grotesque, hyperbolic, … pernicious, polemical, preposterous, provocative ways. Here, for example, is a much quoted denunciation by Paglia (1991: 204, 206, reprinted 1993), from which, believe it or not, I’ve omitted the rude bits:
… the death-by-sludge French theorists […] Lacan: the French fog machine […] a twerpy, cape-twirling Dracula dragging his flocking stooges to the crypt […] Derrida: a Gloomy Gus one-trick pony, stuck on a rhetorical trope […] Neo-Foucaldian New Historicism: a high-wax bowling alley where you score points just by knocking down the pins.
In 2016 and 2017, across British and American politics, an increasing number of people had increasing problems in accurately identifying increasingly fluid boundaries between fact and fiction, information and entertainment, and news and satire. In 2016 a satirical poem on German television14 triggered an international wrangle between lawyers, media commentators and European politicians. Many such cases demonstrate the power of language to persuade, confuse and insult, and deserve careful case-by-case analysis by language philosophers. Given just how confused people get about fact and fiction, Binet’s confidence – that readers will recognize that his novel “reveal[s] itself as fiction page after page” – is optimistic.
I hope I have encouraged you to (re)read SFL. The novel raises fascinating puzzles about fictional worlds. And I have not revealed everything about the plot by any means. I have not told you who found the manuscript, what happened to it, or whether Simon Herzog and his girlfriend (who?) live happily ever after – after the book ends … ha! ha!

Acknowledgements

For useful comments on previous versions of this article I am grateful to Joanna Channell, Andrew and Naomi Hallan, Gabi Keck and Julia Wurr. I am also grateful to three literary colleagues, Ralf Hertel, Markus Müller and Heiko Zimmermann, who discussed with me the ontology of fictional characters.

Notes

1.Page references are to the 2016 Livre de Poche edition. Quotations are my translations of the French original.
3.Jim Stansfield reminded me of a modern variant: a Monty Python sketch about a joke which is so funny that anyone who hears it dies laughing.
4.In Revue de la Presse (August 2017) is an analysis of Emmanuel Macron’s official photograph, which will hang in town halls across France. He stands in front of a desk on which are a clock and books, framed symmetrically between the flags of France and of the European Union, an open window at his back. The author comments that the photo is “saturé de symboles”.
6.The list is reproduced in an article by Pierre Bourdieu (who is at place 36!): « Le hit-parade des intellectuels français ou qui sera juge de la légitimité des juges. » Actes de la recherche en sciences sociales. 1984. 52–53: 95–100. http://​www​.persee​.fr​/doc​/arss​_0335​-5322​_1984​_num​_52​_1​_3336.
8.Stanzel gives as an example the opening sentence of James Joyce’s short story A Little Cloud: “Eight years before he had seen his friend off at the North Wall.” I think it would be possible, however, to find many newspaper articles which start thus in medias res.
10.The programme “On n’est pas couché” was broadcast on France 2 on 7 November 2015. See an extract at https://​www​.youtube​.com​/watch​?v​=ZcMeq​_OkCY4. For another disagreement about SFL on a French talk show, and whether it is justifiable to ridicule a complete generation of French intellectuals, see https://​www​.youtube​.com​/watch​?v​=MsrbJbn8Xlk.
11.The quote continues: « Sollers est rigolo en tant que personnage. La façon dont il parle de lui-même dans ses livres, dont il est le sujet principal, permanent, de sa propre œuvre, c’est très drôle. ».
13.The original article by Sokal and a great deal of further material, is available at http://​www​.physics​.nyu​.edu​/faculty​/sokal/.
14.The poem and the subsequent political kerfuffle are well documented on the Web. Search for “Jan Böhmermann” or “Böhmermann affair”.

References