1How can we best describe the novels of the Nigerian writer Amos Tutuola if not by saying that they are ‘ghost novels’, as suggested by Harold R. Collins?1 Indeed, Tutuola’s fictions are haunted by all kinds of ghosts such as spirits, monsters and satyrs, genii, goblins, demons, imps and gnomes. A precise study of this fictional staff should distinguish among these ghosts those who belong to the complex Yoruba pantheon and those who come from the profane folk tales, but this shall not be our aim here. The ghost that interests us is a metaphorical ghost, the ghost of the unwritten Yoruba, Tutuola’s mother tongue, which lies beneath the lines of the written English.

2Our starting point will be an overview of two studies in which Chantal Zabus recorded Tutuola’s non-standard formulations in English and revealed the Yoruba substrate that lies beneath them, influencing their syntax as well as their lexicon and their morphology. This investigation will lead us to discuss the analogy often drawn by critics between postcolonial writing and literary translation. We shall see that the translation model used to account for the transfer from Yoruba to English must be complicated since no original text is involved in the procedure. While Freud’s model for a psychic translation can help to formulate a hypothesis for the production of the haunted text, Riffaterre’s conception of intertextuality suggests another model to account for its reception. Finally, we shall take a look at Queneau’s famous translation of Tutuola’s first novel, The Palm-Wine Drinkard into French. We shall see how Queneau’s translation tends to rationalize Tutuola’s haunted text. But the translation also involves a creative misreading that gives flavour to the target-text. To borrow Derrida’s terminology, we will see how hauntology becomes hospitality within the translating process.

3Every linguistic survey of Tutuola’s language is indebted to Afolayan for having provided the first detailed commentary on Tutuola’s use of English in 1975.2 But it is Chantal Zabus who first showed ‘the Yoruba layer beneath the scriptural surface’ with the help of an informer, the linguist Yiwola Awoyale. We shall synthesise her two studies published respectively in 19893 and in 2004.4 In the footnote of page 113, Chantal Zabus explains the way she proceeds: ‘Each sample is patterned as follows: English as it is found in the text; The African language calque; the word-for-word English translation of the African language sentence, whenever necessary.’ Chantal Zabus does not mention that the Yoruba counterpart of each quotation from the text is marked by italics and followed by an asterisk, so as to mark its particular status. She also omits to mention the proposal of the standard sentence expected by the Anglophone reader, introduced in her study by ‘instead of’.

4Use of tenses, as in the negated preterit of ‘to be’ (Yoruba: kan)

• ... even a pond did not be near there, ... (PWD, p. 75)

• Kodaadagunomikankosiniegbéibè*

• even pond-water not be at-near there

• Instead of: even a pond was not to be found

5Use of verbs

• when I entered the room, I met a bed which was made with bones of human-beings (PWD, p. 15)

• Sugbonnigbati mo w inu mo babéèdé*

• But at-time that I entered inside room, I met a bed

• Instead of: when I entered the room, I came across/encountered a bed

6Indirect clause structure

• He (Death) asked me from where did I come? (PWD, p. 13)

• Ikeé bi mi nibo no mo(ti) wa?*

• Death ask me at-place be I come from

• Instead of: Death asked me where I came from

7Redundancies (duplication of adverbs)

• the rotten blood was pouring on it often and often (The Witch, p. 13)

• léra-léra*

• often and often

• Instead of: again and again

8Prepositions
Non-standard usage

• the food which had been prepared against the night (PWD, p. 33)

• funalé*

• against

• Instead of: for the night

9Superfluous usage

• we were travelling inside bush to bush (PWD, p. 91)

• latiinuigbodéinuigbo*

• ...

• Instead of: we were travelling inside the bush/from bush to bush

10Absence

• He put it front of the king and sat on it (The Witch p. 81)

• gbé e siwajuoba, mo si jokoolée*

• ...

• Instead of: He put it in front of the king

11Pronouns

12The awkwardly repetitive style may be attributed to the lack of gender in the Yoruba third person singular subject and direct object pronoun and the third person singular possessive adjective

• p. 10 Then I told the old man (god) that I am looking for my palm-wine tapster

• p. 11 when they saw me (bird) on the roof

13Use of adjectives
Nominalization of adjectives

• we met uncountable of them at the gate (PWD, p. 59)

• aimoye*

• that you do not know the number of

• Instead of: we met an uncountable number of them/countless/numerous people

14Use of the adverbial form of the adjective

• this old man was not a really man (PWD, p. 10)

• okunrinarugboyiiki i se eniyanlasan*

15Addition of the deictic pronoun in front of a nominal group

• a distance of ten days journey from this our town (The Witch p. 23)

• iluwayii*

• town our this

16Semantic calques, neologisms

• when it was about two o’clock in the mid-night (PWD, p. 14)

• nigba ti o di nkan bi i aagomeji o ganjoonu*

• when that it became about like clock two deep night

• Instead of: when it was about two (o’clock) in the morning/two hours past midnight

17Both Chantal Zabus and Afolayan agree to consider the relation between the Yoruba ghost text and the written English as a kind of translation. Neither of the two critics believes this translation to be intentional. Chantal Zabus asserts that ‘Tutuola is for the most part inadvertently calquing’ (Zabus 2004: 113) and Afolayan, in the conclusion of her study, claims that ‘some sort of translation (at least psychic) of literature takes place in the process of producing Tutuola’s novel.’ (Afoloyan 1975: 206)

18Of interest here is the common resort to translation as a metaphor to account for the production of the heterolingual text. In fact, this explanation is hardly surprising since a large number of critics draw the parallel between translation and the use of language made in postcolonial literature. Maria Tymoczko, for example, asserts that ‘literary translation provides an analogue for post-colonial writing’, although she is bound to admit that

Significant differences between literary translation and postcolonial literature are obvious and should be addressed from the outset. The primary difference is that, unlike translators, postcolonial writers are not transposing a text.5

19The current models for translation suppose a single transfer from a source to a target text whereas Chantal Zabus describes in fact a very complicated stratification of texts: four layers at least and sometimes more, when different standard re-arrangements are possible. This complexity shows that there is no direct transfer from Yoruba to English but rather a stratification of various possible utterances. Could a model of translation account for this particular case of unconscious loan translation? A first hypothesis could be to take seriously the idea of a ‘psychic’ translation.

20Although Freud never wrote about translation as such, it is possible to infer a theory of psychic translation from his essays about the way memory functions. In an essay entitled ‘Freud et la scène de l’écriture ’ collected in the volume L’écriture et la différence,6 Derrida has shown that Freud has often recourse to similes involving scriptural writing when trying to account for the mechanism of memory. One of the fundamental texts in this respect is the letter addressed to Fliess on the 6th of December 1896. In this letter, Freud introduces the hypothesis of a ‘translation’ involving different periods of the psychic life, thus defining psychic activity as being a process of constant reinterpretation:

As you know, I am working on the assumption that our psychical mechanism has come into being by a process of stratification: the material present in the form of memory-traces being subjected from time to time to a re-arrangement in accordance with fresh circumstances to a re-transcription. Thus what is essentially new about my theory is the thesis that memory is present not once but several times over, that it is laid down in various species of indications.7

21The two main notions developed in the letter to Fliess are those of ‘differed action’ and ‘series’ or ‘stratification’. Within this representation, the original cause gives place to a series of at least three rearrangements, figured as follows in the letter (see fig. 7).

22The notion of ‘differed action’ defies the idea of a continuous time. Didi Huberman, in a fascinating book called L’image survivante. Histoire de l’art et temps des fantômes selon Aby Warburg,8 explains that Freud’s great discovery is that of a discontinuous time that does not flow anymore, but collapses. As proof, Didi Huberman reproduces the following drawing:

23This untimely consideration explains Derrida’s paradoxical formulation when he asserts that within Freud’s theory, the non-originary and the delay become the true origin. The source text is from the start a ‘trace’ or in other words, the source text is self-effacing. This leads us back to our ghosts but also leads to Samia Mehrez’s assertion:

The postcolonial texts, frequently referred to as ‘hybrid’ or ‘métissés’ because of the culturo-linguistic layering which exists within them, have succeeded in forging a new language that defies the very notion of a ‘foreign’ text that can be readily translatable into another language. With this literature we can no longer merely concern ourselves with conventional notions of linguistic equivalence, or ideas of loss and gain which have long been a consideration in translation theory. For these texts written by postcolonial bilingual subjects create a language ‘in between’ and therefore come to occupy a space ‘in between’.9

24The analogy drawn by Freud between a series of transcriptions and mnesic mechanism provides a new model for translation that does not rely on an established source-text. We can thus describe Tutuola’s writing as a procedure of various rearrangements the origin of which is a missing text, the erased trace of the Yoruba language. One major question remains: what is the reader likely to perceive of this latent work?

25The implicit reader of Tutuola’s novel is the Anglophone reader, whose knowledge of Yoruba is highly unlikely. Therefore, it is doubtful that the reader will grasp the Yoruba layering that lies beneath the English surface. How is it possible for the text to suggest another tongue of which the reader does not have the slightest idea? The reader reads a non-standard formulation and all he can do is to connect it with an expected one. But what if it was precisely this gap that produced for the reader the feeling of an absent intertext? Freud gave us an opportunity to imagine a model for the production of the heterolingual text. We shall now consider Michael Rifatterre’s theory of intertextuality to try and understand how the reception of a heterolingual text functions.

26In two major articles, Riffaterre has emphasized the difference between ‘intertext’ and ‘intertextuality’.10 The intertext is the corpus of texts the reader may legitimately connect with the one before his eyes. It is thus a material archive of hypotexts in relation with the text that is being read. On the contrary, intertextuality designates a vertical structure of signification that implies an absent text. The reader perceives an agrammaticality in the text he is reading. These agrammaticalities can be various textual anomalies, such as incongruous elements that disrupt the textual grammar, obscure wordings and phrasings that the context alone does not suffice to explain. Riffaterre assumes that in its withdrawing, the intertext leaves signs of incompleteness to be completed elsewhere. These traces or ungrammaticalities alert the reader to the possibility of an (absent) intertext that can remain unknown without causing the structure to collapse.

27This explains why the translation metaphor is so vivid in the critical literature about postcolonial writing: the heterolingual palimpsests, because they do not correspond to the linguistic norms, create gaps that are filled in with the imagination of another text. Of course, there is no such text and Yoruba, as a language, is no more than a ghost...

28We shall now observe the crucial moment when reading turns out to be also writing, that is, effective translation of the text. May we expect the target text to be a stratification of sources and targets as complex as the so-called ‘original’? The Palm Wine Drinkard was translated into French by the famous Oulipien writer Raymond Queneau in 1953. Queneau was then reader of English for Gallimard. This first publication is introduced by an important, although short, preface in which the translator tells about some difficulties he encountered when translating. Of interest for us here is the affirmation that the translator tried to respect the agrammaticalities of the original:

La traduction comportait quelques problèmes particuliers. L’auteur, par exemple, utilise les conjonctions de la langue anglaise (notamment but et or) d’une façon inhabituelle qui m’a donné bien du souci. D’autre part, j’ai dû résister à la tentation de rationnaliser un récit dont les ‘inconséquences’ et les ‘contradictions’ se glissent parfois dans la structure même des phrases.

29However, it appears that Queneau only preserves the particularities that do not affect the norms of the French language. He keeps the pronouns redundancies and few neologisms. He doesn’t keep any morphological or syntactic deviation from the norm.

30Use of tenses, as in the negated preterit of ‘to be’ (Yoruba: kan)

• ... even a pond did not be near there, ... (PWD, p. 75)

• il n’y avait même pas la moindre langue de terre près de là (p. 74)

31Use of verbs

• when I entered the room, I met a bed which was made with bones of human-beings (PWD, p. 15)

• une fois dans la chambre, je vis que le lit était fait avec des ossements humains (p. 15)

32Indirect clause structure

• He (Death) asked me from where did I come? (PWD, p. 13)

• il (Mort) me demande d’où je viens? (p. 15)

33Prepositions
Non-standard usage

• the food which had been prepared against the night (PWD, p. 33)

• la nourriture qui avait été préparée pour passer la nuit (p. 33)

34Superfluous usage

• we were travelling inside bush to bush (PWD, p. 91)

• Nous marchions à travers la brousse (p. 90)

35Use of adjectives
Nominalization of adjectives

• we met uncountable of them at the gate (PWD, p. 59)

• nous trouvons là une foule innombrable qui nous attend à la grille (p. 59)

36Use of the adverbial form of the adjective

• this old man was not a really man (PWD, p. 10)

• ce vieillard n’était pas réellement un homme (p. 12)

37Semantic calques, neologisms

• when it was about two o’clock in the mid-night (PWD, p. 14)

• vers deux heures après minuit (p. 15)

38Pronouns

• p. 10 Then I told the old man (god) that I am looking for my palm-wine tapster

• p. 12 Je dis donc à ce vieillard (ce dieu) que je cherchais mon malafoutier

• p. 11 when they saw me (bird) on the roof

• p. 12 quand ils me voient (c’est-à-dire l’oiseau) sur le toit

39If we borrow Lawrence Venuti’s terminology, we can say that Queneau’s translation seems to be domesticating: it erases the particularities of the original to make the text look familiar to the target-reader. In other words, Queneau seems to have got rid of the Yoruba ghost that haunted Tutuola’s English. Do we have to conclude that the statements made by Queneau in his preface were mere denials? But then, how can we understand the dithyrambic appreciation of this translation by Vialatte:

Qu’on ne s’y trompe pas : les trois quarts du charme viennent du talent du traducteur. Le Planton du palais écrivait en anglais. Queneau l’a traduit comme un dieu, avec un inégalable fruité. Il ne traduit pas, il réinvente ; il délire et il prophétise comme dans sa langue maternelle dans l’idiome d’un pays qui n’existe jamais. A croire qu’il a passé sa vie dans la Ville-des-Bébés-sans-Jambes.11

40There is an inventiveness in this translation that lies in a ‘creative misreading’:12 Queneau has read Tutuola as an oral story teller, which is not really surprising, but also as an absurd writer.

41It is enlightening to compare Queneau’s translation and another version, proposed in the Anthologie critique de la littérature africaine anglophone: le roman et la nouvelle, edited by Denise Coussy, Jacqueline Bardolph, Jean-Pierre Durix and Jean Sévry in 1983. The two versions differ significantly.

42The non-standard comparison of equality (adverb ‘very’ instead of ‘as’) is translated by Queneau by an unusual formulation (‘très belle, comme un ange’) whereas the translators chose the standard (adjective + adverb ‘comme’) for the Anthologie critique de la littérature africaine anglophone. The following sentence starts with the conjunction ‘so’ that is unusual in the narrative and becomes the colloquial ‘donc’ in Queneau’s translation. The use made of fake logical links is characteristic of Tutuola’s poetic. The odd repetition in the second sentence ‘she went to the market on a market-day’ is reproduced in Queneau’s translation but not in the anthology. Last but not least, the third repetition of the prepositional clause ‘on that market-day’ is the occasion for Queneau of an invention: the choice of the presentative marker ‘voilà’ emphasizes for the French reader the feeling of hearing an oral story teller, whose use of the French language needs not be as formal as the standard used by a writer in a novel.

43The second example shows that Queneau has turned Tutuola’s non standard formulations into absurd situations within the plot.

44The comparison is translated by Queneau as if it were a metamorphosis: the character actually becomes ‘a log of wood’ instead of being compared to it. It is thanks to a creative misreading and not because of the respect of the hauntology within the original that Queneau has produced a translation that is still celebrated. However, the ghosts have not entirely disappeared.

45The Palmwine Drinkard was published in 1952 in Lagos. Queneau read the book the very year of its publication and translated it into French a year after, in 1953. He read the two following books written by Tutuola less than a year after their publication. During this entire period, Queneau was involved in an ongoing reflection about language. This reflexion gave birth to the essay entitled ‘Écrit en 1955’ in which Queneau argues that spoken French is a language radically different from written French. As Marie-Sophie Armstrong puts it:

Queneau viewed ‘Neo-French’ or spoken French to be a language as radically different from French as French is from Latin and he argued that a triple evolution involving vocabulary, spelling, and syntax was necessary if Neo-French was to achieve the status of written language to which it was entitled.13

46The same critic suggests that Zazie dans le métro, the famous novel written in 1959, is both a defence and an illustration of ‘Neo-French’. The opening word ‘Doukipudonktan’, which Barbara Wright translated by ‘How canaystinksotho?’, could stand for the disruption undergone by the French language within this book. As a translator, Queneau did not seem to dare to damage the target language but as a writer, he reproduces Tutuola’s experiment directly into French. To put it another way, Zazie seems to be haunted by the very ghost Queneau had exorcized in his translation. Did not Queneau stop reading Tutuola after having written Zazie?14 Translating Tutuola into French thus seems to have been a necessary detour for the creative transformation of the target language. But if we admit Freud’s theory of an anachronical memory, we can go as far as considering Zazie as a posterior hypotext for L’Ivrogne dans la brousse.