Corps de l’article

Introduction

Whether explicitly asserted or not, discussions on interdisciplinarity have accompanied Translation Studies (TS hereinafter) ever since it claimed recognition as an independent branch among longer-standing academic disciplines. Surprising as it may be, the more recent articles on the interdisciplinary character of TS (e.g., Chesterman, 2002; Gambier, 2006; Martín Ruano, 2006; Snell-Hornby et al., 1994; Snell-Hornby, 2006) raise few new issues: shared conceptual paradigms and terminological uniformity are as absent now as they were in the 1970s.[1]

In this article, although I welcome the call for interdisciplinarity, I do, however, also endorse the idea that the methods of neighbouring disciplines need not necessarily be integrated into one comprehensive research model for TS. If I am convinced that there is no need to back a single common research model, it is not so much because the existing literature convincingly argues the benefits of interdisciplinarity. My basic assertion is much simpler: I would never have been able to understand my data as well had I not embraced simultaneously different research principles and methodologies.

In what follows, I will take my research into the translation of literary irony as the starting point for a discussion on TS’s interdisciplinary nature.[2] I will explain on which points and in what ways my research benefits from more than one research tradition and methodology. After a general outline of the research (section 1), I will home in on two points. The first one concerns the origins of my analytical instrument, which has, as I will show, varying theoretical backgrounds. In that same section, I will illustrate how this analytical instrument works with three examples (section 2). The second point relates to the way I have used different approaches within TS in order to utilise more adequately the particularities of each novel integrated in the corpus (section 3).

1. Outline of the Research

Below, I will briefly outline my research project. I do not claim exhaustiveness; my only goal here is to provide the necessary background information so that my discussion on interdisciplinarity can be an informed one.

In the first part of my research, I developed an analytical instrument for the comparison of translated ironic excerpts, called the “ironic effect.” The aim of the second part of the research was, first, to fine-tune and test this instrument using passages from La tía Julia y el escribidor (2006) by Mario Vargas Llosa (translated in English as Aunt Julia and the Scriptwriter (2007); AJS hereinafter).[3] All the selected passages were compared systematically against the Dutch, (U.S.) English and French translations. For heuristic reasons—the objective at this stage being the testing and fine-tuning of the analytical instrument—the focus was as much as possible on the textual-linguistic make-up of irony in the original and translated texts. In other words, the broader communicative, cultural and/or social nature of both translation process and product were neglected and the text was, somewhat artificially, isolated from its surrounding context. This imbalance was restored in the two subsequent chapters, where the research hypotheses were built on relevant contextual elements and their possible interaction with the ironic effect in the target texts. More specifically, I investigated what possible effect factors such as translation norms, translation policy, translators’ poetics, author’s interference and literary field might have on the ironic effect.

This outline will do for the moment. Let us go back now to the beginning, the analytical instrument called the ironic effect and the underlying premises. Although I am acutely aware that these premises are highly controversial, they will be presented here as though they were not. Summing them up is however essential in that they reflect my understanding of irony and partially explain why I included three different components in the analytical instrument (see section 2). Firstly, irony can be a local and space-restricted device. In other words, it can be understood, as it is in the most traditional way, as a trope, boldly paraphrased as “saying one thing and meaning another.” But irony can also be triggered by structural means; the contrast between (unknowing) characters and (knowing) readers, or by the manifest contradictions between different narrators and/or characters. Secondly, irony is not only a linguistic problem; it can be conveyed through very different means: linguistic, stylistic and textual/functional elements. Irony cannot, therefore, be pinned down to one specific linguistic category or text-level. For the TS scholar engaged in comparative research into original and translated texts, this is an important issue, because it has a bearing on the unit of comparison, which cannot be determined on the basis of a pre-established list of fixed criteria. Last but not least, irony relies on interpretation and is, therefore, not fixed: irony “happens” rather than “is” (Hutcheon, 1994). It is an interpretive move in which the reader, and of course the translator as a reader, takes an active part.[4]

2. The Ironic Effect

The analytical instrument, then, builds on these premises and includes three different components: (1) irony’s semantics, by which I mean the (semantic or logical) mechanism that underlies the incompatible information conveyed through irony; (2) irony’s axiological dimension, i.e., the fact that irony always involves an evaluative dimension or gives information about (negative) attitudinal dynamics and (3) irony’s textual, contextual and/or intertextual signalling.

In the next sub-sections, I will expand upon these three components and thereby specify whether and how they build on theories of irony developed in neighbouring disciplines (mainly pragmalinguistics, cognitive psycholinguistics and rhetoric).

2.1. Irony’s Semantics

The first component, irony’s semantics, has been subdivided into three subcategories. The first one, semantic reversal, is close to the classical definition of irony. I am referring here to a trope which involves meaning substitutions, either by semantic inversion or propositional negation. In other words, a literal meaning (the “said”) is substituted by a figurative meaning (the “unsaid”). These local ironies, which function by inverted meanings, are only of secondary interest to the TS scholar insofar as they do not generally trigger major translation interventions. Most of the textual samples recorded in my corpus show that translators almost always translate semantic reversals literally, particularly when the reversed element is a lexical item and restricted in space.

The second subcategory, based on supplementing rather than supplanting a meaning, is closely related to Sperber and Wilson’s account of irony as echoic mention. Their earliest writings were motivated by their dissatisfaction with both the classical rhetoric account of irony and the Gricean understanding of irony as a flouting of the maxim of quality (“try to make your contribution one that is true”). Their examples are convincing; it is indeed difficult to explain what exactly is reversed in an ironic understatement such as “you can tell he’s upset” when coming upon a customer in a shop, blind with rage and making a public exhibition of himself (1992, p. 54). By the same token, it is difficult to perceive irony in an utterance such as “Oh, to be in England / Now that April’s there” (ibid., p. 55), unless the utterance is recognized as a quotation and the main goal of the speaker lies in the pragmatic effect caused by the utterance, rather than its strict semantic value. Therefore, Sperber and Wilson explain both the nature and the reasons why irony is used by referring to the distinction between use and mention. Use of an expression involves reference to what the expression refers to whereas mention of an expression involves reference to the expression itself (ibid., pp. 58-59). In verbal irony, the speaker does not use but mentions another person’s speech thereby making evident a disapproving attitude vis-à-vis another person’s utterance. The last example, the rain in April, would then, when ironically intended, communicate “that the English spring does not always live up to expectations, that the memory of home is not always accurate, that romantic thoughts do not always survive reality” (ibid., p. 55). In my corpus, the “mention” principle is particularly useful to understand passages, be they isolated words, (parts of) phrases, or entire paragraphs, where the axiological dimension stems from contrastive repetitive use, rather than from lexical inversion or propositional negation. Though coming from different theoretical backgrounds, double-voiced discourse approaches (Bakhtin, 1981; Ducrot, 1984; Fludernik, 1993) share with the “echoic mention” principle the possibility of plural speech even when the formal structure does not indicate the precise origin of these speeches and even when these speeches are not neatly distinguishable. If we apply this to literary discourse: an utterance, or a part of it, can be mentioned (not used) by a character whose interest lies then in conveying an attitude rather than communicating something to which s/he does not adhere. A character can be merely mentioning another character’s speech in order to show disapproving distance. In narrative fiction, what determines whether the reader will attribute or not a speech to a given character depends on the way the events, experiences, thoughts, dialogues, settings and characters are described throughout the entire novel.

The third and last subcategory brackets incongruent and ambiguous ironies which cannot be understood in bivalence or binary (either/or) thinking. My understanding of this subcategory is close to the analysis of irony as frame-shifting. A frame is a conceptual metaphor and suggests “a patterned set of connections among ideas” (Ritchie, 2005, p. 290). Frame-shifting, then, is a metaphor “for the activation of a new set of connections and suppression of a previously activated set, when new information makes the new set relevant and the old set irrelevant” (ibid.). The listener or reader is confronted with information belonging to incompatible frames, which s/he should somehow reset.

2.2. Irony’s Axiological Dimension

The second component of the “ironic effect” covers irony’s axiological dimension. Many scholars have maintained that irony necessarily implies a negative, critical attitude towards its object. The fact that criticizing by praising is by far more frequent than the opposite, praising by criticizing, is related to the possible consequences of the latter. Indeed, utterances such as “I don’t like you at all,” literally conveying a negative meaning but implicating a favourable judgment, are at risk of being misinterpreted. If the speaker’s negative insincere utterance is not recognized as such, s/he is taken as having said something negative (Roy, 1978, cited in Attardo, 2000, p. 976). According to pragmalinguistic theories (Haverkate, 1990, p. 90) the reason behind a possible misinterpretation concerns the fact that praising by criticizing violates two pragmatic conventions simultaneously. Both the Gricean maxim of quality and the maxim of politeness are violated, whereas criticizing by praise only violates the quality maxim.

For the identification of ironic passages in my own research, I have taken as a minimum requirement some degree of emotional involvement. The notion of emotional involvement comprises a wide array of underspecified ironies, ranging from “cool detachment to engaged hostility” (Hutcheon, 1994, p. 40), which I have organized in four categories, listed here from minimal to maximal emotional charge: ludic, distancing, provisional and aggressively corrective effects of irony.[5] For the comparisons of original and translated texts, I take into account the intensity and type of pragmatic effect (i.e., the four categories listed above), but also the target of the irony. As will be shown in example three below, translation interventions and formal shifts can entail a displacement of the irony’s target.

2.3. Irony’s Textual, Contextual and Intertextual Signalling

The third component of the analytical instrument comprises a variety of elements which all suggest that irony is always signalled. This signalling is understood as a broad concept, for what triggers irony can be rather concrete and restricted spatially. Ironies, though, can also be determined by the way the narrative fiction is set up in its entirety. In that case, the novel’s particular configurations (mainly descriptions of events and characters) provide the necessary backdrop that enables an ironic inference. This type of signalling, i.e., immediate co-text and the novel’s textual space, is what I call textual signalling. Contextual signalling refers to those instances where the ironic inference is based on knowledge gained outside the narrative fiction, pertaining to a distinct semiotic sphere. Intertextual signalling, then, is made up of those instances where irony is triggered by reference to another text, which can be, but is not necessarily, a literary text.

Before turning to some examples, I would like to add one final remark concerning the interrelatedness. The splitting up into three components should, of course, not mask the privileged interconnectedness between certain semantic subcategories and pragmatic effects and/or types of signalling. For instance, it is not difficult to see why the analytical potential increases when double-voiced discourse ironic fragments are simultaneously examined from the viewpoint of the emotional charge involved. After all, these types of irony seek to communicate an attitude rather than a lexical item or propositional content. On a methodological level, conclusions about irony’s axiological dimension can stem from previous observations of formal shifts, i.e., source text and target text comparisons. Shifts concerning the nature, frequency and intensity of evaluative textual markers (for instance, lexico-grammatical indicators of modality or vocatives) allow us to draw conclusions on the passage’s critical edge, which can be enhanced, reduced or nullified. This is how most of the conclusions for the axiological dimension in my own research have been reached. But there are of course other ways of investigating irony using the same analytical device. For example, experimental effect studies, testing reader-responses to ironic fragments in different translations (whether in the same or different languages), could offer valuable insights into irony’s axiological dimension.

2.4. Three-Part Instrument: Motivations

I developed the analytical instrument in this way because it allows for two kinds of analysis at the same time. Firstly, on a text-analytical level, the ironic effect allows the identification of the ironic excerpts, because it delimits the unit of comparison in a fairly systematic way. Secondly, as regards translational analytics, by splitting up three distinctive components, the ironic effect enhances the descriptive potential of the shifts between the source and target texts. For instance, by subdividing types of irony according to the underlying semantic mechanism, it will be possible to identify which types of ironies elicit translation interventions and which are less likely to do so.

Next, I will illustrate how the ironic effect works with passages from AJS. The irony in the first example is based on a traditional semantic reversal mechanism which, as has been said before, did not often entail translation shifts in my corpus. Indeed, in all three translations analyzed (Dutch, French and U.S. English) the irony in “exquisita flor” was left unmodified.[6]

As has been said already, the excerpt is taken from AJS, a semi-autobiographic novel relating the story of the 18-year old Marito, a law-student with aspirations as a writer. He will fall in love with his aunt-in-law, Julia Urquidi, and marry her after a hilarious odyssey in Peru’s countryside. In chapter one, though, from which this excerpt is taken, he feels little sympathy for the older woman: she continually makes jokes about his young age and youthful looks. Marito, eager to be considered as “a full-grown man of eighteen” (Vargas Llosa, 2007 [1982], p. 8[7]), does not appreciate these comments at all. He is somewhat forced by his aunt to accompany her to the movies. The idea fills him with dread, but Aunt Julia’s comment, insinuating that his financial situation is miserable, is the last straw. The ironic signalling in “exquisita flor” will be confirmed several chapters later when the narrator mockingly looks back on how he was as an 18-year old and mentions his “deep-rooted Spanish prejudices with regard to the relations between men and women [never allowing] Aunt Julia to pick up a check” (ibid., p. 122). Example one shows an irony based on semantic reversal without substantial translation shifts. In all three translations the ludic charge as well as irony’s target and intensity (i.e., the components that constitute the second dimension of the “ironic effect”) are similar. As far as the third component of the ironic effect is concerned, in all three translations, irony is still triggered textually, i.e., by the information we gather throughout the novel.

Let us now consider examples which entail more prominent translation shifts and might, therefore, prove more revealing for translational analytics. Example two is very similar to the last one and is situated in exactly the same narrative context. During their “forced date,” Aunt Julia explains that she prefers to go out with Marito rather than with older men, because all they are after is sex. Besides, these older men are sparse with romanticism:

Example two shows the same irony mechanism as example one: a classical case of semantic reversal in “muchas gracias.” Clearly, Marito is not grateful for the part he takes in all this: he has to endure her bad jokes about his youth and lack of financial means and is forced to go to the cinema when he would prefer to (try to) write a short story. Let us now consider the (U.S.) English translation, which, unlike what happened in the first example, shows more than one formal shift. The central question here is whether these formal shifts affect the irony’s axiological dimension and/or semantic mechanism:

“The worst thing about being a divorcee isn’t that all men think they’re obliged to proposition you,” Aunt Julia informed me. “Rather, it’s the fact that because you’re a divorcee they think there’s no need to be romantic. They don’t flirt with you, they don’t whisper sweet nothings in your ear. They just come straight out with what it is they want from you, right off the bat, in the most vulgar way imaginable. That really puts me off. That’s why I’d rather go to the movies with you than go out dancing with a man.”

“Thanks a whole lot—I appreciate the compliment,” I said.

“They’re so stupid they think that every divorcee’s a streetwalker,” she went on, not even noticing the irony in my voice. “And what’s more, all they think about is doing things with you.”

Vargas Llosa, 2007 [1982], p. 12

The translated text shows several modifications. Firstly, the free indirect discourse of the mature narrator looking back on how he used to be, is rendered now by direct discourse. Secondly, the semantic inversion in “muchas gracias” has been followed by a second one. Two literal meanings are now inverted: neither “thanks a whole lot” nor “I appreciate the compliment” can be taken at face value. Thirdly, the character’s comment on Aunt Julia’s reaction to his irony is marked by an explicit irony marker (Barbe, 1993): “sin darse por enterada” has been translated by “not even noticing the irony in my voice.” These formal changes modify the textual signalling but do not alter the semantic reversal mechanism nor the axiological dimension.

The two ironies discussed so far are relatively straightforward. Firstly, because the mechanism involved is the one which has persisted most strongly in the history of the word and concept of irony. Indeed, semantic substitution involving antonymic relationship between the said and the unsaid has been part of the basic definition since Quintilian. Secondly, because the analysis of these types of irony requires little contextualisation beyond the purely textual. For the reader, it is enough to be informed about the characters’ relationships and narrated events in order to get the irony. Thirdly, following on from the second point, these kinds of ironies are not attributable to complex interpretive processes.

However, as I will show in example three, not all passages are as straightforward. By this point, the two protagonists are engaged in a complicated love story. Marito suffers from insomnia due to accumulating difficulties: Julia is his political aunt, a divorced woman, fourteen years his elder, unable to marry him because he has not reached the legal age of 21. Due to his insomnia, he arrives early at work—which is very unusual—and starts working. This is how he informs the reader on his work:

Taken as an isolated occurrence, this passage would not even stand out. It does, however, when situated in the overall constellation of the narrative. In the first place, because we know how the mature narrator evaluates his former professional activity: a job with a “pompous-sounding title, a modest salary, duties as a plagiarist, and flexible working hours: News Director of Radio Panamericana” (Vargas Llosa, 2007 [1982], p. 3). The job title does not live up to the job’s content, which consists of “cutting out interesting news items that appeared in the daily papers and rewriting them slightly so that they [can] be read on the air during the newscasts” (ibid.) or retyping “news items from El Comercio and La Prensa, changing adjectives and adverbs” (ibid., p. 10). Neither does the title match the poor quality of his editorial staff, composed by Pascual—who selects news items in function of their bloody crime rather than their informative relevance—and Big Pablito, a complete illiterate. His material working conditions are as pathetic as his editorial staff: they are stuck in a filthy shack, his desk is carted off to give to the accountant and his typewriter to the author of popular radio serials. All this certainly foregrounds the absurdity of the narrator’s working situation or at the very least diminishes its importance. Let us now return to the passage mentioned above and compare it with the U.S. and Dutch target texts:

(U.S.) I was at my desk in the shack at Panamericana earlier than usual that morning, and when Pascual and Big Pablito arrived at eight, I had already written the bulletins, read all the newspapers, and annotated and marked in red all the news items to be plagiarized

ibid., p. 237

(DU) Ik was vroeger dan anders in mijn dakkamertje van radio Panamericana en toen Pascual en grote Pablito om acht uur kwamen, had ik de bulletins al klaar en alle kranten doorgelezen, van aantekeningen voorzien en aangestreept (om plagiaat te voorkomen).

Vargas Llosa, 2006 [1981], p. 263

[I was at my shack at Panamericana earlier than usual and when Pascual and Big Pablito arrived at eight, I had already finished the bulletins and read all the newspapers, annotated and marked (to avoid plagiarism)]

On a strictly formal level, both translators have textually manipulated the source text’s ambiguous reference concerning the plagiarism; both have changed towards more explicitness. The way they have done it, though, clearly indicates that they have attributed irony in a different way. The U.S. translator highlights the causal relationship between the marking in the newspapers and the plagiarizing. Marito prepares the newspapers by marking in red the relevant news items which will be plagiarized by Pascual afterwards (and by big Pablito, if only he could read and write!). Interestingly, the Dutch translator’s textual manipulations lead to a very different reading. Having arrived early, Marito has already prepared several bulletins. Put in other words, he has already plagiarized several news items, retyped them by modifying some adjectives and adverbs. But unlike what happens in the U.S. text, the Dutch Marito marks the newspapers so that Pascual would not plagiarize his own work when preparing the other bulletins.

These formal shifts have affected the irony’s axiological dimension. Both the U.S. and Dutch readings are still edgy but the target of the negative judgment is wholly different: in the U.S. text, the mature narrator exposes the illicit character of his former professional activity, i.e., the fact that he used to plagiarize others’ work altering some adjectives and adverbs. In the Dutch text, though, the mature narrator mocks his younger self and highlights his own absurd demands: he wants to avoid that others plagiarize his work which is itself a flagrant case of plagiarism. This reading would be odd if it did not square with so many other textual samples where the younger self’s pedantry is harshly exposed by the mature narrator. This last reading, though, was privileged only by the Dutch translator.

3. Different Research Approaches within TS

So far, I have expounded on my analytical instrument and exemplified its functioning (section 2). It should be clear by now why I have used mainly pragmatic and cognitive approaches to irony: they yield a better understanding of irony’s semantics and they allow for a more varying range of pragmatic effects that are inherent to the ironic phenomenon.

In the remainder of this article, I will narrow down the discussion to issues within TS and discuss the advantages of embracing methodologies traditionally associated with different translation turns (Snell-Hornby, 2006). I will do so by specifying the research aims and by listing the research traditions I have integrated for the analysis of the other two novels of the corpus: Tres tristes tigres (Cabrera Infante, 2005) and La invención de Morel (Bioy Casares, 2007).

For Three Trapped Tigers (TTT hereinafter), I started off with a predictive hypothesis (Chesterman, 2007a) related to external conditions of the translation process. More particularly, three factors were taken into account. Firstly, the translation circumstances are similar, for indeed, both the French and U.S. translators collaborated closely with the author Cabrera Infante, whereas there was virtually no collaboration with the Dutch translator. Secondly, there is Cabrera Infante’s reputation as a highly interfering author, who is keen to advertise that a translation of one of his novels is an excellent opportunity for him to remodel that novel (cited in Guibert, 1973, pp. 409-411). Thirdly, there was the proximity of the publication dates of the translations: the French and U.S. translations were published in 1970 and 1971 respectively, whereas the Dutch translation was published more than 25 years later.[8] These circumstances backed my hypothesis which was that the French and American translators’ interventions would be comparable and that their translations would show substantial differences compared to the Dutch one, which had been translated with virtually no collaboration and more than 25 years later. My data did, however, not corroborate this. What is more, there were remarkable analogies between the translators’ interventions in the Dutch and U.S. texts.

The initial predictive hypothesis was then modified with a series of explanatory hypotheses. At this point of the research, social factors intervened as possible explanatory elements for the observed cultural phenomena (see Pym, 2006). What follows is a non-exhaustive list of some of these explanatory elements (see De Wilde, 2010a, for a more elaborate discussion):

  1. A glance at the professional careers of the translators taught me that they were all inexperienced as literary translators. This ruled out the influence of the translators’ experience on the translation product which could have been essential, considering the linguistic, thematic and intertextual complexity of the novel.

  2. Judging from the conditions of the respective literary fields, the Dutch translation was not an easy project, financially speaking. Unlike the French and American translations, the first Dutch translation (1997) appeared long after the apogee of the Latin-American Boom literature. It has been argued that the volume and complexity made it a very unprofitable project for a very small Dutch-speaking market. My research showed that this translation project emanated from the particular interests of two agents active in the Dutch literary field: the chief editor and the translator himself.

  3. Both the Dutch and the American translators stressed their fascination for the puns in the original. These remarks partially explain the particular interest for the recreation and addition of new puns observed in both the Dutch and U.S. translations.

  4. However, the analysis of “metatranslative documents” (Gouanvic, 2007, p. 26) revealed that the French translator was worried about restrictions imposed by the French language and literary tradition. This might explain why the French translation deviates much less from linguistic norms and literary conventions.

  5. The metatranslative documents also showed that the translators had very divergent opinions as regards their role and (in)visibility, or their (in)fidelity to original writing.

  6. Finally, a summary of the life and aesthetic preferences of the author himself proved revealing for the different kinds of collaboration that possibly took place between the author and the French and American translators.

These explanatory factors should, of course, be situated in a broad picture of “multifactorial conditioning” (Pym, 2006). In human science, it is indeed difficult to represent causality as a simple linear chain (Chesterman, 2007b, p. 175); nor should one slip into a stricter interpretation of “causality” than can be justified (see Chesterman, 2008a, for different types of causality in TS). But, broadening the research scope and multiplying the questions asked, I have undoubtedly reached an alternative understanding of the data.

For the other novel under scrutiny, La invención de Morel (IM hereinafter), let me stress first of all that the nature of the material used to hypothesize was completely different. Unlike the TTT chapter, which took text-external conditions collected in metatranslative documents as its points of departure, the hypothesis for IM started off from textual findings observed in the (original and translated) primary texts. Textual comparisons of the original text and the three translations pointed at diverging operational norms (Toury, 1995, pp. 53-69). These micro-level tendencies were linked to diverging initial norms for the three translations under scrutiny. From there, I assumed that the Dutch and French translations could be catalogued as globally adequate (in that adherence to source norms was observed), whereas the U.S. translation is globally acceptable (given its adherence to target norms). These categories are of course used with precaution. Firstly, they do not reflect absolute regularities. Secondly, and more importantly, their usefulness does not exceed a purely heuristic level. These textual findings were supported then by Venuti’s claim (1995) that Anglo-American translations tend to be fluid and reader-friendly, in that they mask their status as translations and adopt lexical, syntactical and stylistic options frequently used in the target culture.

The above considerations functioned as the premise for the predictive hypothesis adopted in the IM chapter. I assumed, then, that the overall ironic effect of the U.S. translation would be reduced due to target norms adequacy (mainly lexical and syntactical unequivocalness). In-depth analyses of the original text and the different translations did indeed corroborate this hypothesis.

In summary, the approach in the TTT chapter is much more akin to the critical understanding proper to sociological approaches which focus on the social conditions of the translator’s interpretive act and pay attention to the plurality of implicated agents (translators, mediators, readership). As for the IM chapter, even though the description of translation norms does not constitute the final goal of the research, the way the translated texts were approached has much in common with research within Descriptive Translation Studies that focuses on the texts as products of norm-governed behaviour. While combining the ways I look at these data is not problematic, one should not mask: (1) the purport of the questions asked, (2) the intellectual horizon the questions spring from and (3) the scholar’s intellectual preferences.

The remarks on the genesis and nature of the hypotheses threaten to overlook one important feature, namely that the identification of the ironic passages in both the original and translated texts engages with a particular way of approaching the literary text. In that sense, my own approach is greatly indebted to critical hermeneutic methods. Certainly, the hermeneutic method is considered particularly appropriate for only a limited set of human sciences, such as, precisely, literary analysis. This does not take away though that all kinds of scientific observations—including those pertaining to the natural sciences—are first “interpreted,” i.e., they are given meaning in the light of a specific theory (Chesterman, 2008b, p. 55). Undoubtedly, in my case, there have been pre-existing, and often implicit, theories or expectations guiding my selection and categorization of the ironic passages. In fact, the very design of my corpus arises out of an act of interpretation (Crisafulli, 2004, p. 28). I can reduce my own personal subjectivity in the analyses of the three novels by adhering to an existing tradition which has used irony as a critical instrument. But I can never ignore the fact that all these previous readings are based on exactly the same procedure: they comprehend texts from a preconceived understanding of different relevant issues. None of the analytical tools used—in my case mainly narratology, stylistics, text-linguistics—are free from subjectivity. There should at least be recognition of, and margin for, reflection on the scholar’s subjective stance. After all, what I interpret as ironic is always a scholar’s meaning and will depend on the meaning of the literary work in its entirety.

Conclusion

In the first part of the article, I gave an outline of the ironic effect, an analytical instrument designed for the analysis of irony in translated literary corpora. I have shown the advantages of the ironic effect by means of three excerpts from AJS. In the second part of the article, I detailed how my own research incorporated very different research questions and varying research traditions: cognitive and pragmalinguistic understandings of irony, hermeneutic methods for literary analysis, narratology, stylistics and text-linguistics for text-comparative approaches and translation sociology for translation conditions.

It will, therefore, not come as a surprise that, in my opinion, there is margin for an inclusive and flexible TS methodology. This, however, implies that we agree first on the way theory and methodology are understood, i.e., as means of understanding. These means of understanding can be—but are not necessarily—ways of verifying empirically observed data. From this viewpoint, theory has no ontological status and is nothing more than a functional notion: it is a form of understanding (Chesterman, 2007a). In that sense, methods are mere instruments that allow us to develop, apply and test theories which help us to understand observational facts. Given the complexity and the variety of translation-related phenomena, which all, to some extent, deserve to be understood, it is impossible to do this according to a single shared methodology. Although the advantages of interdisciplinarity—their potential as integrative and holistic fields, allowing for syntheses and cross-fertilizations, the view of wider patterns and interrelations (Chesterman, 2002, p. 4)—should not mask its potential risks, such as fragmentation of the field or superficial application of borrowed concepts (ibid., p. 5), this does not mean that differing methodologies cannot coexist. As I have shown here, the conclusions in my own research would certainly be less rich and conclusive had I restricted my scope to one single research tradition.