Article body

1. Introduction

The goal of this article is to address a theoretical issue arising from the definition of a certain type of explicitation. Explicitation is a phenomenon of the target text stating information in a more explicit form than that of the original (Shuttleworth and Cowie 1997: 55); some explicitations add or change linguistic elements to clarify the original text, others serve to reinforce the original speaker’s attitude (Hirsch 2011). However, as we will show in this article, clarifying the text might in fact affect its global meaning – the sum of “voices” present in the source text (Dascal and Weizman 1987; Weizman and Dascal 1991), in particular when the text’s intention is to remain obscure. In these cases, from a semantic or syntactic point of view, these shifts (Toury 1995) are explicitations. On a deeper level of meaning, though, they can be considered implicitations: they obscure the speaker’s stance and thus render the text’s global meaning more implicit, this global meaning being understood as the implied author’s “speaker’s meaning.”

The notion of implied author, as defined by Booth (Booth 1961) in his Rhetoric of Fiction suggests an imaginary figure who does not have a real biography, but who may be assigned intentions via the text itself, and whose voice does not necessarily coincide with that of the narrator. The most important aspect of this notion for our purpose is that of construing an intended meaning from a text, a process that was analyzed by Dascal and Weizman (Dascal and Weizman 1987; Weizman and Dascal 1991).

As for the notion of speaker’s meaning, it is the third element of the threefold distinction put forward by Dascal (1983) following Grice (1968) between sentence meaning (the context-independent meaning of an utterance), utterance meaning (the conventional meaning of an utterance in a given situation), and speaker’s meaning: what the speaker means to convey by uttering a given utterance in a given situation, and the characterization of the various types and levels of contextual knowledge required.

As argued by Dascal and Weizman (Dascal and Weizman1987; Weizman and Dascal 1991), the global meaning of a text comes through in various ways: the direct meaning of the dialogues, events, characters and narrators described in the books, and implicatures that may be derived from them through the use of textual features. Besides the content itself, it can also be conveyed through the characters’ interpretation of meaning and implicatures, the implied author’s point of view, inconsistencies and other textual signals in the narrator’s speech, the role played by the title and the structure of the narration of the events.

In our investigation of the global meaning, and of explicitations that can affect it, we rely on the text-understanding model postulated by Dascal and Weizman (Dascal and Weizman 1987; Weizman and Dascal1991). Taking into consideration the global meaning of the text and of the different voices within it, our goal is to examine the role played by translated utterances that can be considered explicitations but that obscure the speaker’s stance, and to contribute to the understanding and definition of this specific type of explicitation.

To do so, this article is structured as follows: first, we introduce the concepts of explicitation and implicitation (Section 2), after which we proceed to present some classifications of explicitations (including our own), leading to the category this paper focuses on (Section 3); we then expand on the theoretical issue encountered through the analysis of several examples (Section 4), before offering some suggestions (Section 5) and presenting our general conclusions (Section 6).

2. Definitions

Explicitation and implicitation strategies are the concern of many investigations in the field of Translation Studies, most purporting to provide definitions or explanations of these two phenomena. The following is a brief survey of a few relevant definitions.

2.1. Explicitation and Implicitation

Explicitation, a concept first introduced by Vinay and Darbelnet (Vinay and Dalbernet 1958), refers to the process of making what is implicit in the original explicit in the translation (Klaudy 2003). The concept of explicitation can be interpreted both in terms of the translation process – i.e. as a technique – and in terms of the translation product – i.e. as a textual feature (Pápai 2004: 145).

Explicitation is considered by many investigators to be one of the potential universals of translation (features which are typical of translations in comparison with non-translated texts) (Baker 1993). The concept of translation universals entered the field of Translation Studies in the 1990s at Mona Baker’s initiative (Baker 1993; 1995), but the search for regularities in translation pre-dates the term itself, and goes back to the 1980s with the idea of translation as a third code (Frawley 1984). Translation universals are defined by Baker as “linguistic features which typically occur in translated rather than original texts and are thought to be independent of the influence of the specific language pairs involved in the process of translation” (Baker 1993: 243). A great deal of research has been devoted to translation universals since the 1990s, in particular to interference, simplification, normalization and explicitation – the most researched of all so-called translation universals, defined by Blum-Kulka in 1986 in her formulation of the ‘explicitation hypothesis’ as “an observed cohesive explicitness from SL to TL texts regardless of the increase traceable to differences between the two linguistic and textual systems involved” (Blum-Kulka 1986: 19).

Although a certain consensus exists among researchers on the existence of a widespread trend for a high level of explicitness in translation, the term “universal of translation” to describe explicitation is widely criticized, with a preference for less ambitious terms such as generalization or tendency; the present work is in line with this approach. There is no consensus, however, on the definition of explicitation itself, which makes it difficult to evaluate the relevance of the numerous studies undertaken on the subject, and to draw conclusions on the extent and nature of the phenomenon of explicitation in translation.

Over the last fifteen years, many scholars have been basing their definition on the possibility of inference by the reader. Thus, for Becher (Becher 2010b), explicitness is “the verbalization of information that the addressee might be able to infer if it were not verbalized,” while explicitation “is observed where a given target text is more explicit than the corresponding source text” (3).

Implicitation is usually defined as a reversed version of the definition of explicitation in translation. In other words, implicitation takes place when information explicitly expressed in the source text is left implicit in the target text, even though a stylistically equal but more explicit version exists (Ďurinová 2012).

In his article (2010b), Becher urges researchers to abandon the explicitation hypothesis and to replace it with a modified version of Klaudy’s asymmetry hypothesis, also calling for a redefinition of the terms. Klaudy’s influential reformulation of the explicitation hypothesis postulates that for the explicitation hypothesis to be attested, explicitations in one direction must not be systematically compensated by implicitations in the other direction. Thus, for instance, if a trend to explicitate is detected in French-to-English translations, it must be demonstrated that no implicitating trend exists in English-to-French translations.

As Murtisari (Murtisari 2016) aptly shows, the core recurrent issue in the study of explicitation in translation is the lack of consensus in the definition of the phenomenon, which the present article hopes to help narrow down. One of the main dichotomies at the root of critics’ differing perspectives on explicitation is that of linguistic versus cognitivist approaches, a brief account of which is provided below. These two perspectives are also described as text-oriented and reader-oriented perspectives, the former seeing explicitation as a textual property and the latter as an interpretive strategy on the part of the reader.

2.1.1. Linguistic Approaches

The first landmark definition of the concept of explicitation is that of Vinay and Darbelnet (Vinay and Dalbernet 1958), who characterize it as a “a stylistic translation technique which consists of making explicit in the target language what remains implicit in the source language because it is apparent from either the context or the situation” (translated by Sager and Hamer 1995: 342)[1]. It sees explicitation as a linguistic phenomenon, i.e. a textual property, and is therefore a text-based approach. This definition – or a close variation on it – has been adopted by a great number of subsequent studies, such as Weissbrod (Weissbrod 1992), Englund Dimitrova (Englund Dimitrova 2005), and Hansen-Schirra et al. (2007).

It has often been argued that in her formulation of the explicitation hypothesis, Blum-Kulka offers no clear definition of the phenomenon. Consequently, although a number of scholars have based their work on her hypothesis, they usually provide their own definition of explicitation. Øverås (1998), for instance, defines it as “[the] translation process where implicit, co-textually recoverable ST [source text] material is rendered explicit in TT [target text]” (4).

In her survey article on the main approaches to the concept of explicitation, Murtisari (Murtisari 2016) provides a well-rounded description of the linguistic approach, which dominated until the 2000s:

The degrees of explicitness […] [depends] on factors such as encodedness, informativity, specificity, emphasis/focus and topicality in the information packaging. Explicitation based on textual explicitness will consist in a shift to a higher degree of encodedness, informativity, specificity and so on.

Murtisari 2016: 65

2.1.2. Reader-Based Approaches

In contrast with the linguistic approach, Séguinot put forward in 1988 the idea that explicitation was not limited to shifts from implied meanings, but also included any addition or change contributing to making the text clearer, thereby approaching explicitation as a matter of decoding rather than mere encoding (Séguinot 1988).

A more recent development in this direction was the appearance, through the field of pragmatics, of the cognitivist approach to explicitation. This approach actively focuses on explicitation from the point of view of its reception and interpretation by the reader, rather than as a textual property encoded in the source text. Thus, Saldanha (Saldanha 2008) showed that explicitation does not systematically involve implicitness in the source text. Saldanha emphasizes the role of the translator as a cultural mediator, and shows that this role can lead them to perform what Klaudy calls “pragmatic explicitation” (see Types of Explicitation). However, as shown by Saldanha, such explicitations do not necessarily entail greater linguistic explicitness of the target text.

A significant turning point in this approach was the association of explicitation with less interpretive work for the reader. In 2002, Pápai described explicitation as a process favoring easier or more secure interpretation for the reader, before later defining it as “a technique of resolving ambiguity” (Pápai 2004: 145). In turn, Heltai (Heltai 2005) suggested redefining explicitation as increased cognitive processability – or meaning processability –, following Kusztor and Atayan’s (2003) claim that “a linguistic form is more explicit than another one if it reduces the number of possible interpretations and thereby cognitive load” (Heltai 2005, 47). Heltai contrasts this approach with the linguistic one, pointing out that depending on whether a researcher adopts a linguistic or a cognitivist approach to explicitation, they can sometimes obtain different results.

Heltai’s (Heltai 2005) approach is particularly pertinent to our analysis, for it differentiates between the two main meanings given to explicitness by researchers: on the one hand, linguistic explicitness means that most of the message is coded and minimal inferencing is needed; on the other hand, what Heltai calls “true explicitness” (Heltai 2005) involves less ambiguity and easier processing. The two often overlap, but not necessarily: a target text may well be more linguistically explicit without being easier to process, and vice-versa. Heltai (Heltai 2005) suggests studying explicitness at different levels, since non-explicitness at the sentence-level may be explicitness at the text level, and explicitness at the text level may not be a result of explicitness in its sentences. This point will be further analyzed when we discuss our methodological issue.

2.2. Defining Explicitation and Implicitation

We derive our initial definitions from Becher (Becher 2010a), with slight adjustments to accommodate the notion of speaker’s stance and to take into account the fact that explicitation does not systematically involve implicitness in the source text, or more informativity in the target text than in the source text (Saldanha 2008).

Thus, in the following work:

  • Explicitness is the verbalization, clarification or emphasis of information that the addressee (in this case, the reader) might still be able to infer if it was not verbalized, clarified or emphasized.

  • Implicitness is the non-verbalization, non-clarification or non-emphasis of information that the addressee might still be able to infer if it was not verbalized, clarified or emphasized.

  • Explicitation is observed where a given target text is more explicit than the corresponding source text, or where an explicitating shift is used to attain the same degree of explicitness as in the source text.

  • Implicitation is observed where a given target text is more implicit than the corresponding source text, or an implicitating shift is used to attain the same degree of implicitness as in the source text.

Note that here, explicitness and implicitness are seen as products rather than processes, while explicitation and implicitation are viewed as processes.

3. Types

3.1. Types of Explicitation

Since explicitation and implicitation in translation constitute a diverse phenomenon, a number of researchers before us have attempted to provide a categorization of the different types encountered.

Hjort-Pedersen and Faber (Hjort-Pedersen and Faber 2010) divide explicitations into two basic types: addition – the inclusion in the TT of extra lexical elements that either add or repeat meaningful elements; and specification – adding meaning(s) by using lexical elements that are semantically more informative. Similarly, they consider implicitation to be of two kinds: reduction – which involves leaving out meaningful ST lexical elements in the TT; and generalization – which involves using TL lexical elements that are semantically less specific than the ST lexical elements (Hjort-Pedersen and Faber 2010: 243-244).

In 2008, Klaudy came up with a categorization of explicitation that was widely discussed and adopted by other researchers. Her typology divides the phenomenon into four categories which are not mutually exclusive:

  • obligatory explicitation, which is due to differences in the syntactic and grammatical structures of the language combination involved and cannot be avoided;

  • optional explicitation, which is due to “differences in text-building strategies […] and stylistic preferences between languages” (2008: 106);

  • pragmatic explicitation, which is dictated by “differences between cultures” and often leading translators to “include explanations in translations” (2008: 106-7);

  • translation-inherent explicitation, which mostly seems to justify the claim that explicitation is a translation universal: “[t]ranslation-inherent explicitation can be attributed to the nature of the translation process itself” (2008: 107).

However, Becher (Becher 2010b) points out that this last category is also the only one for which Klaudy provides no examples, adding that no study on explicitation has convincingly demonstrated its existence (3).

These are just a few examples of the different typologies provided by translation scholars. On the basis of their insights and of our own research, we provide an initial categorization, which was first introduced in a PhD thesis (Hirsch 2008) under the supervision of Elda Weizman.

3.2. Initial Categorization

Given that researchers are susceptible to obtaining different results depending on the approach adopted, as recalled in 2.1.2, we concur with Heltai’s claim that linguistic and cognitive levels of explicitness should be studied separately. Our approach here being a text-based one, our categorization of explicitations (and implicitations) belongs to the linguistic perspective, and we classify different types of explicitation according to the syntactic, semantic and pragmatic characteristics of the translational shifts observed.

  1. Semantic explicitation – the choice of more specific words in the target text, or the addition of distinctive semantic elements to the original words (Klaudy 1996: 103). Van Leuven-Zwart (Van Leuven-Zwart 1990) calls this category “semantic modification” (78).

  2. Increased register markedness – when the translation changes the register of the original, thus altering the function of the text. Van Leuven-Zwart (Van Leuven-Zwart 1990) describes this category as “stylistic modulation,” and provides examples in which the translation of Don Quixote’s linguistic style into Dutch affects the original register. This in turn produces a probable shift in the protagonist’s characterization and in the text’s functions (Van Leuven-Zwart 1990: 81, 83).

  3. Structural modification resulting in explicitation– any grammatical paraphrase, such as a change in sentence order (Englund Dimitrova 2005: 236-239; Weissbrod 1992). Ďurinová (Durinová 2012) describes cases in which the information implied in the ST is expressed in an explicit fashion in the TT through a rewording of the whole clause, while the change cannot be pinpointed to an exact spot (46).

  4. Strengthening of logical cohesive ties – explicitations that reveal the implicit textual logical ties of the text. Blum-Kulka (Blum-Kulka1986) identifies shifts in the level of coherence and cohesion from source to translation that result in an increase in explicitness in the target text. Those shifts in the level of explicitness can be attributed to grammatical differences or stylistic preferences, but can also be a direct result of the translation process (Blum-Kulka 1986). Examples of such explicitations are the addition of connectors (Blum-Kulka 1986; Englund Dimitrova 2005; Van Leuven-Zwart 1990: 81), replacement of personal pronouns with proper names (Blum-Kulka 1986: 24-29) and the filling out of ellipses (Weissbrod 1992). For instance, in Krogsgaard Vesterager’s analysis (Krogsgaard Vesterager 2017), the elliptical phrases of the source text were explicitated in most target texts in the form of additions (Krogsgaard Vesterager 2017: 115).

  5. Explicitation of contextual knowledge – bridging the gap in the knowledge of target readers, who may not be as familiar as the source readers are with geographical names, conventions or cultural references. Translators will use different strategies to overcome the lack of information, such as the addition of semantic items, footnotes or other forms of explications (Blum-Kulka 1986: 24-29; Weissbrod 1992; Pápai 2004: 155-156; Klaudy 1996).

  6. Reinforcement of the speaker’s stance – the addition of different forms of emphasis in order to accentuate the speaker’s stance (Klaudy 1996: 103). As Van Leuven-Zwart (Van Leuvan-Zwart 1990) points out, specifying subjective components may, among other things, cause the translation to be more emotionally charged, aggressive or overstated even to the point of cliché (Van Leuvan-Zwart 1990: 71).

This last category is the one that can sometimes be at the root of the methodological problems that will subsequently be described. We will thus further illustrate it here with an example from the pseudo-detective novel El Misterio de la Cripta Embrujada (The Mystery of the Bewitched Crypt) by Eduardo Mendoza (Mendoza 1979), and its translation into Hebrew. The book is a parody steeped in criticism of Spanish society after the democratic transition, depicting a hero who leaves a psychiatric hospital in order to solve a murder mystery. In what follows, the first-person narrator describes a soccer match that took place in the mental institution in which he was committed.

In this example, as in many others throughout the book, the “detective” narrator misjudges the situation and praises himself without good cause. Even though this excerpt is taken from the beginning of the book, a stage at which the reader is still unfamiliar with the narrator’s poor judgment, his claim that his passes during the game were largos [long] and precisos [precise] are likely to come across as unrealistic, in light of the fact that the person in question is a psychiatric patient who does not exhibit great physical form.

As the story advances and the narrator repeatedly over-compliments himself and distorts fictional reality in his own favor, the reader acquires enough specific extra-linguistic contextual knowledge to understand how far-fetched these claims were in the first place. These characteristics call the narrator’s reliability into question and do not allow him to be taken seriously: it is as though the implied author (the authorial figure created by the text) is looking over the narrator’s shoulder and criticizing him for his self-adulation and for the false modesty conveyed by the phrase para qué negarlo [why deny it].

The Hebrew translator[2] adds the word lehadhim [amazingly], which has no counterpart in the original; as a result, the detective praises himself even more in the translation. If at this point the Spanish reader is still uncertain about the truthfulness of the narrator’s descriptions, the added Hebrew word constitutes an explicitation of the speaker’s stance, which further results in an explicitation of the text’s global meaning, serving to convince the target reader (even more so than the source reader) that the narrator is unreliable. Also note that in this case, the explicitation also corresponds to more traditional approaches, which group explicitation with addition.

4. Methodological Issue

As mentioned above, the methodological issue to be described here is found in cases where explicitations add or change linguistic elements to clarify the original text, but do not reinforce the original speaker’s attitude, since in the cases in question the text’s intention is to remain obscure.

This problem derives from the variety of definitions and types of explicitations. As noted by several researchers (Pym 2005; Baumgarten, Meyer et al. 2008; Ďurinová 2012), the original definition of explicitation was restricted to cohesive explicitness, but the concept was subsequently extended to include other aspects of language use (Pym 2005: 2; Baumgarten, Meyer et al. 2008: 183; Ďurinová 2012). Séguinot (Séguinot 1988), for example, specifies that

[e]xplicitation can take three forms in a translation: something is expressed in the translation which was not in the original, something which was implied or understood through presupposition in the source text is overtly expressed in the translation, or an element in the source text is given greater importance in the translation through focus, emphasis, or lexical choice.

Séguinot 1988: 108

Séguinot’s notion of shift of emphasis corresponds in many instances to what we call “reinforcement of the speaker’s stance,” which is often difficult to define for it largely depends on subjective interpretation (Ďurinová 2012: 11-12).

Saldanha (Saldanha 2008) responds to the conflicts arising from the many definitions of explicitation by looking at the texts from the point of view of the reader, demonstrating that sometimes, when translators explicitate, the informativeness in the target text remains the same or even less. She further claims that paradoxically, specification does not necessarily involve a shift from implicit to explicit information (Saldanha 2008: 21); in this regard, Ďurinová (Ďurinová 2012) suggests that this depends on the definition of information.

Another paradox linked to the manifold definitions of explicitation, and the paradox we focus on here, stems from the fact that indirectness, obscurity and deliberate ambiguity can all form part of a speaker’s stance. Contexts in which a speaker may want to remain obscure thus include cases where the aim is to avoid censorship; the use of irony or humour; situations in a detective novel where the reader’s pleasure derives from the detection of subtle clues; and/or the recourse to ambiguity for aesthetic reasons, such as non-resolution and openness calling for reader participation (see Eco’s 1985 concept of open fabula[3]), including narrative unreliability.

In such a context of deliberate obscurity, erasing inconsistencies and filling gaps in the speaker’s discourse often results in resolving ambiguity. Pápai defines explicitation as “[a] technique of resolving ambiguity, improving and increasing cohesiveness of the ST and also of adding linguistic and extralinguistic information” (Pápai 2004: 145). However, indirectness and obscurity would then form part of the speaker’s stance, characterized by deliberate ambiguity. Resolution of ambiguity can thus correspond to explicitation as it has been traditionally described, but in the process, by obscuring the speaker’s stance and thus the global meaning, it can correspond to implicitation as well. Thus, if we define both linguistic and semantic explicitation on the one hand, and explicitation of speaker’s stance on the other as different forms of explicitation, it results in some cases of explicitations also being, at the same time, implicitations at a higher level (that of the text’s global meaning).

Note however that, following Kamenická (Kamenická 2007), we consider cases where the speaker’s stance is no longer co-textually recoverable in translation not as cases of implicitation, but rather of omission. In other words, for there to be implicitation rather than omission, the speaker’s stance must be recoverable in other ways, either in the segment in which implicitation takes place – i.e., the speaker’s stance is less obvious but still perceptible – or at the level of the text’s global meaning. In the case of unreliable narration, for instance, if the narrator’s unreliability is only expressed in one place in the source text and this unreliability is no longer perceptible in the target text, the speaker’s stance has been omitted rather than implicitated. If, however, the reader has the means of identifying this unreliability in the same or other parts of the text, the speaker’s stance has been merely implicitated.

Let us take again the case of narrative unreliability to further analyze and illustrate the paradox presented above. Booth (Booth 1961) was the first critic to put forward a definition for narrative unreliability in a literary work: “[a] narrator is reliable when he speaks for or acts in accordance with the norms of the work (which is to say the implied author’s norms), unreliable when he does not.” (Booth [1961] 1983: 158-59). In Booth’s landmark definition, unreliable narration corresponds to a perceptible gap between narrator and implied author. This creates a narrative distance between them, and a secret collusion between implied author and implied reader – “[t]he audience presupposed by the narrative itself” (Chatman 1978: 150).

In Example 1, we had a straightforward explicitation of speaker’s stance, which emphasized the unreliability of the narrator and did not obscure the text’s global meaning – on the contrary, it highlighted it. In some cases, however, an explicitating shift from another category (i.e. linguistic shifts as opposed to narratological ones) can obscure the speaker’s stance and the text’s global meaning; this global meaning can still be recovered both at sentence level and more generally in the novel as a whole, but its visibility is reduced. These shifts are explicitations from a semantic or syntactic point of view, but they can also be considered implicitations because they make the speaker’s stance more implicit. Such are the occurrences described below.

4.1. A Pale View of Hills

A Pale View of Hills is British writer and Nobel Prize laureate Kazuo Ishiguro’s first novel, published in 1982. It is characterized by its narrator’s unreliability, suggested by a wide range of more or less prominent textual signals scattered across the novel. In A Pale View of Hills, the narrator – a Japanese woman named Etsuko – recounts the visit of her younger daughter to her home in England shortly after her elder daughter Keiko committed suicide. This present-day part of the narrative alternates with memories of Etsuko’s past in Japan. She reminisces, in particular about her friendship with her neighbour Sachiko, and with her little girl Mariko. What is at stake in the unreliable narrative mode is first and foremost the question of who the different characters are: the novel is ridden with hints that Etsuko and Sachiko might be one and the same, or at least that Sachiko stands on some level for the narrator’s past back in Japan, and Mariko for Etsuko’s dead daughter Keiko. In this passage, the narrator has gone looking for Sachiko’s daughter Mariko, who has run away, upset by the fact that her mother is planning to take her to America the following day. However, the novel also suggests an alternative reading in which Sachiko and Etsuko (the narrator) are one and the same person, and in which Mariko stands for none other than Keiko.

In the source text, two main points are worth noting:

  1. The overlap between Mariko and the deceased Keiko, which is the subtler of the two clues in this passage, is expressed through the fact that all references to Mariko are made indirectly, through less specific terms such as the little girl.

  2. In this passage, the narrator’s use of pronouns points more clearly than at any other point in the text towards her unreliability. Instead of using the expected you when referring to Mariko and Sachiko and their imminent journey, she uses the pronoun we, revealing the overlap between the characters of Etsuko and Sachiko on the one hand, and of Mariko and Keiko on the other hand.

In the Spanish translation, both of these textual signals pointing to the narrator’s unreliability are affected by translational shifts.

  1. In two places, the translator uses Mariko’s name rather than an equivalent of the source text’s less explicit wording:

    1. forthe little girlwas before me is translated as ya que Mariko estaba antes mí [since Mariko was before me]

    2. After a long silence, she said: “I don’t want to go away” is translated as Tras un largo silencio, Mariko dijo: “No quiero irme” [after a long silence, Mariko said: I don’t want to go away].

We argue that these shifts constitute linguistic explicitations, since proper names are more specific than pronouns: it falls under the category of changes of logical cohesive ties, which includes the replacement of personal pronouns with proper names (Blum-Kulka 1986, 24-29).

However, these shifts also carry out an obscuring of the speaker’s stance, which is to remain ambiguous as to the little girl’s identity. On a deeper level, this is therefore an implicitation as well, since the hint to an alternative reading is lost: part of what Blum-Kulka calls “the text’s meaning potential” (Blum-Kulka 1986: 23) has disappeared.

  1. The Spanish translator chose to normalize what he presumably saw as an error on the author’s part, or as some eccentricity that he felt was perhaps too far removed from the norms in force in Spanish translation. Every reference to the first-person plural is replaced with the second-person plural, as though the narrator was indeed speaking unambiguously about Sachiko and Mariko and not, possibly, about herself and her own daughter:

    1. if you don’t like it over there, we can always come back is translated as si aquello no te gusta, regresaréis inmediatamente [if you don’t like it over there, you [plural] will come back right away]

    2. But we have to try it and see if we like it there. I’m sure we will is translated as Pero primero hay que ver si aquello os gusta. Estoy segura de que os gustará. [But first it’s necessary to see if you [plural] like it. I’m sure that you [plural] will like it].

There is thus an explicitating shift, here again in the form of a change of logical cohesive ties. At the surface level, there is a greater cohesion and coherence of the text, as the switch in pronouns is a jarring element in the source text. However, this apparent inconsistency of the English text plays a crucial role in the narrative strategy: the implied author uses it as a way to point the implied reader to an underlying version of fictional reality. This linguistic explicitation corresponds to an implicitation of one of the narrative layers of the story, and of the text’s global meaning. We thus see that linguistic and narratological explicitness do not necessarily go hand in hand, as linguistic explicitation can also result in implicitation of the text’s intent.

Incidentally, when it comes to the cognitivist approach, or what we call cognitive explicitness, the situation is even more complex, which tends to confirm that linguistic and cognitive explicitation can, and often should, be dissociated. While this passage is certainly easier to process without the jarring use of pronouns present in the source text, it can be argued that at the scale of the whole text, the shift performed in the Spanish translation can result in more confusion than clarification, i.e. in decreased meaning processability (Heltai 2005). Indeed, while this passage contains one of the most blatant clues as to the overlap between the two sets of characters – the switch in pronouns – it is only one of many passages pointing to an alternative version of fictional reality. In terms of text intent, this scene is meant to come as a confirmation rather than a revelation of this overlap. As the author himself points out, “[he] intended with that scene for the reader to finally realize, with a sense of inevitability, ‘Of course, yes, she’s finally said it’” (Mason 1989:338).

However, still from a cognitivist perspective, the nature of the shift – increased or decreased meaning processability – depends on the real reader’s idiosyncratic ability to pick up on other, often more subtle clues present before and after this passage in the text, and which are by no means all erased in the Spanish translation. The shift performed in this passage, which we have described as linguistic explicitation and narratological implicitation, can thus result in either increased or decreased meaning processability depending on different readers’ individual readings of the text. This underpins Heltai’s observation that explicitness at the sentence (or here, passage) level may not correspond to explicitness at the text level.

4.2. The Remains of the Day

The Remains of the Day, Kazuo Ishiguro’s third novel, received the Booker Prize and is Ishiguro’s best-known work. The first-person narrator, Mr. Stevens, is a British butler serving in a grand English mansion. The narration takes the form of a diary kept by the butler during a short trip he takes across the country, on his employer’s suggestion. The very class-conscious butler is wearing an outfit and driving a car that were both lent to him by his employer, and the people he comes across regularly mistake him for a gentleman. Stevens never tells them that they are under the wrong impression, and is clearly enjoying the situation while also feeling self-conscious about the misunderstanding. He repeatedly resorts to ambiguous phrasings to avoid letting on that he is not in fact a gentleman while also avoiding plainly lying. In this passage, his employer’s Ford has broken down and Stevens is obliged to stay with people from the nearby village, who are flattered to be around someone of such importance as they think him to be. Someone asks him about his employer’s car, assuming that it is his own. The narrator fails to correct this mistake, while using a deliberately ambiguous turn of phrase:

It is plain that the speaker is being intentionally obscure, presumably because of his mixed feelings on the situation he finds himself in. The linguistic ambiguity lies in the fact that the main clause of the conditional form is missing: the phrase “then yes, it is mine” – or a similar equivalent – is implied but not spelt out, which allows Stevens to stay nominally clear of a lie without admitting that the car is not really his. This corresponds to what Heltai (2005) calls a grammatical ellipsis, itself subdivided into three subcategories: “[t]he ellipted element must be recoverable either from other parts of the text (textual ellipsis), or knowledge of syntactic rules (structural ellipsis) or the situation (situational ellipsis)” (n.p.). Here, the ellipsis is a structural one.

The Spanish and Turkish translations of the novel both carry out an explicitating shift of the passage:

The main clause left implicit in the source text is spelt out both in the Spanish and the Turkish translations, with respectively [yes, it is mine] and [yes]. This passage from implicature to explicature corresponds to another explicitation of logical cohesive ties, more specifically the filling out of ellipses (Weissbrod 1992; Heltai 2005). By spelling out what the narrator is only implying in the source text, these two parallel translational shifts both implicitate the narrator’s stance, which is precisely one of obscurity and deliberate ambiguity. Therefore, here again, explicitation on a linguistic level (filling out the ellipse and resolution of ambiguity) also corresponds to implicitation on a narratological level, as the narrator’s inner struggle is obscured and with it the text’s global meaning.

In Example 2, as in Example 3, the explicitation carried out according to our initial categorization (explicitation of logical cohesive ties) thus goes hand in hand with what this categorization views as implicitation (obscuring of the speaker’s stance). In terms of cognitive explicitness, in both cases, the shifts also amounts to what Heltai calls “true explicitation,” which “lies in less ambiguity and easier processing” (n.p.).

4.3. The Silent Patient

As mentioned above, the speaker may wish to remain obscure in a range of literary contexts, such as in detective novels or thrillers, where the reader is intentionally misled (at least up to a certain point), and can only reflect in retrospect on the subtle hints the narration supplied. Ambiguity then forms part of the speaker’s stance and is a crucial component of the pleasure the reader derives from the book.

In Alex Michaelides’ debut psychological thriller, The Silent Patient, the reader follows a narration by a therapist obsessed with uncovering the motive of a woman who killed her beloved husband, while the patient herself refuses to speak, expressing herself only through her diary and paintings. As the story of the counseling progresses, the therapist’s backstory is uncovered: he was the one responsible for forcing the patient to murder her husband because the patient’s husband was having an affair with his wife. The backstory, which appears to be occurring concurrently, actually happened in the past, and its link to the murder is only unveiled in the last pages of the book.

This surprising twist in the plot builds on the narrator providing just enough information to make the plot advance without revealing unnecessary details, thus creating intentional obscurity. While the deliberate ambiguity of the original narrator is explicitated in some instances in the Hebrew translation, it is not always the case, and the shifts in explicitness carried out in the translation are not always consistent with the underlying narrative revealed at the end of the book. Consider the following two examples:

In the source text, the narrator comments on the apparently unaccountable history of the patient he is attempting to treat, presenting the common perception about her mental state without disassociating himself from said opinion. Only at a very late stage does the reader understand the logic of the patient’s actions, a logic that was at least partially available to the narrator all along.

Here, the choice of the modal model verb must is of great significance. This modal can express two main types of modality: epistemic, when it expresses a strong likelihood (e.g. “Your umbrella is wet, it must have rained”); and deontic, when it expresses a constraint or obligation stemming from the speaker (e.g. “It is late, you must come home now!”). In the English source text of the passage under consideration, the more obvious function of the modal must is its epistemic function, indicating a high degree of probability, although however, deontic modality is also hinted at, in particular through the use of had to in the rephrasing of the following sentence: She had to be. This ambiguity forms part of the global meaning of the text by hinting at the narrator’s ambivalent attitude. The Hebrew version, however, presents a literal translation of the original: hi hayevet lihyot metorefet [she is obliged to be mad], also using a model verb in Hebrew. However, this modal verb – hayav/hayevet – is used differently from the English modal must, and expresses more unambiguously obligation: while the narrator’s unreliability can be recovered at the scale of the novel, it is less perceptible in this particular passage. A possible solution to reproduce the double meaning of the source text could have been to use an adverb in Hebrew for must, for instance betah [for sure, definitely, it has to be], instead of the modal hayevet for had to. The use of the modal verb hayevet in Hebrew obliterates the crucial doubt of the source text.

In addition, the choice to repeat this modal verb as the translation of both must be and had to be tampers with the narrative complexity of the original text. This explicitation appears even less logical after the twist in the plot, where the reader finds out that the narrator knows much more than he let on, and that he is actually responsible for the woman’s actions. Such is also the case in the following example:

These lines are also taken from the first part of the book, where the narrator describes his plans to help Alicia and solve the mystery surrounding her. At this stage, the reader is ignorant of the fact that a great part of the story is actually known to the therapist; his narration is thus a careful combination of questions that touch only upon a part of the story the narrator does not actually have access to – namely, Alicia’s feelings. The therapist-narrator being the one who actually instigated the murder, something the reader only finds out at the end of the story, it is consistent with the narrator’s unreliability that he should only ponder Alicia’s feelings or state of mind rather than what actually happened.

However, the Hebrew translation adds a whole sentence to the source text: ma kara shegaram la laharog oto? [what happened that made her kill him?]. This alters the narrator’s stance by explicitating the core question whose answer is supposedly unknown to him, but which he has actually had access to all along. Hence, while the narrator in the original text retains crucial information but avoids straightforwardly lying, this explicitating addition in the translation results in an implicitation of the speaker’s stance of deliberate ambiguity.

5. Proposed Solution

In light of recent developments in research on explicitation in translation, it appears that neither linguistic explicitation (see for instance Saldanha 2008) nor narratological explicitation (highlighting the speaker’s stance, see Lessinger 2019) are satisfactory descriptions for the widespread and multi-facetted phenomenon described by researchers as explicitation. The fact that explicitation at one level can co-exist with implicitation at another – an issue which is explored in this article and illustrated in the examples – testifies to the fact that the initial categorization we proposed after Hirsch (Hirsch 2008) cannot fully account for linguistic explicitation either.

It follows that, in order to resolve the paradox described in this article, narratological explicitness – the degree of prominence of the speaker’s stance – should be analysed separately from linguistic explicitation. As shown in this article, while highlighting the speaker’s stance can overlap with linguistic explicitation, it is not systematically the case. Our revised categorization of explicitness thus excludes explicitation and implicitation of the speaker’s stance:

  1. Semantic explicitation

  2. Increased register markedness

  3. Structural modification resulting in explicitation

  4. Strengthening of logical cohesive ties

  5. Explicitation of contextual knowledge

This narrowing of the definition to linguistic explicitness means that in all the examples studied, the shifts observed do correspond to linguistic explicitation.

As for narratological explicitness, we suggest that it should generally be studied as a different phenomenon. Even within the paradigm of narratological shifts, the study of explicitness is susceptible to giving rise to a paradox similar to that explored in this article, i.e. narratological explicitations on one level that are also narratological implicitations on another. This paradox can in turn be resolved by speaking not in terms of explicitation and implicitation but of reduction and highlighting of complex narrative voices[4]. Thus, Taivalkoski-Shilov (Taivalkoski-Shilov 2006) shows that free indirect discourse tends to be shifted to reported discourse in translated texts, which Chesterman assimilates to the potential S-universal “reduction of complex narrative voices” (Chesterman 2010: 41). In the case of more complex narrative pacts where the deeper meaning is at odds with the surface meaning – for instance because of the speaker’s deliberately ambiguous stance, as in the examples presented – the erasure of one of the text’s layers also corresponds to a reduction of complex narrative voices. Conversely, giving more prominence to narrative complexity, which is also a common scenario (see Lessinger 2019) corresponds, in our view, to a highlighting of complex narrative voices.

Thus, with the exception of the first example (which corresponds to both linguistic explicitation and highlighting of narrative complexity), in the examples studied here, the shifts observed correspond to both linguistic explicitation and a reduction of narrative complexity.

6. Conclusion

The different examples studied in this article illustrate the paradox that lies at the heart of the phenomenon of explicitation taken as a broad category, including not only linguistic and cognitive but also narratological considerations: explicitation on one level can correspond to implicitation on another level of analysis. The analysis of these examples serves to underpin our proposed redefinition of the phenomenon. As has been pointed out by many Translation Studies scholars (see for instance Becher 2010b; Murtisari 2016), the study of explicitation in translation is seriously undermined by the difficulty in defining the phenomenon and by the inconsistent definitions adopted by the different studies.

We argue that while linguistic, cognitive and narratological explicitness are all relevant angles of approach to explicitation, they should not be studied as one and the same phenomenon. In particular, we advocate studying narratological explicitness from the angle of the more specific phenomenon of “reduction of complex narrative voices” (Chesterman 2010: 41). The results presented in this article also support Heltai’s notion that cognitive explicitness should be explored at different scales, as microtextual explicitation can correspond to macrotextual implicitation and vice-versa. Moreover, unlike linguistic and narratological explicitation – which can be studied from a source-text-based perspective –, when studying shifts in cognitive explicitness, only a reader-based approach can yield reliable results. This discrepancy further reinforces our claim that it would be more beneficial to study all three types of explicitness separately.