You are on Érudit's new platform. Enjoy! Switch to classic view

Articles hors thème

A Model for Defining the Concept and Practice of Translation, from the Perspective of Greimassian Semiotics

  • Rovena Troqe

…more information

  • Rovena Troqe
    University of the Free State
    Faculté de traduction et d’interprétation, Université de Genève, 40, boul. du Pont-d’Arve, 1211 Genève 4, SUISSE

Cover of Traduction et rédaction : un destin lié, Volume 29, Number 1, 1er semestre 2016, pp. 9-261, TTR : traduction, terminologie, rédaction

Article body


Defining the nature of Translation is a thorny matter that is not easily addressed. How does one start to define Translation as an object of study? What are the dimensions involved when translation is initiated and performed? This kind of investigation pertains to the pure, theoretical and general branch of translation studies, as defined by Holmes (2000 [1972], p. 176), and raises the question of the most adequate method employed. Drawing on Greimassian semiotics, this paper aims to respond to Holmes’s call to establish general principles that explain and describe the phenomena of translating and translation in the way they manifest themselves in the world of experience. “The Name and Nature of Translation Studies” by Holmes (2000 [1972]) is a seminal paper that set out to establish the field of research of what was then considered to be a new discipline. The Holmes Map of Translation Studies is necessarily grounded in the mindset of its time, but still introduces valid scientific principles such as the relation between theoretical and applied research and the division between general and partial theoretical translation studies. The domains are interrelated, and applied studies are intrinsically connected to pure theory via descriptive translation studies. Pure theory is based and derives from practice; it describes and explains it, and possibly anticipates practical processes. Holmes’s map could be improved, expanded or modernised, since all the research areas it describes have evolved, creating a beneficial osmosis between sub-branches in the fields of text-type, problem, medium, area and time-restricted partial theories (van Doorslaer, 2007; Vandepitte, 2008). Actually, the progress in more or less related disciplines such as computer science, linguistics, literature and comparative studies, and psycholinguistics, to name but a few, has problematised the definition of translation, highlighting the complexity and stratification of this concept and activity on the one hand, but also gradually diverting attention from genuine theoretical reflection on the other.

Other contributions (Chesterman and Arrojo, 2000; Tymoczko, 2007) have acknowledged the difficulties in encompassing the concept of translation, but have also avoided embarking on concrete definitions, highlighting the fact that the ethnocentric bias would stifle any attempt at providing a universal or objective definition. In general terms, it could be said that reasoning on what translation is varies between intuitive claims, and cultural-time-space-bound and relativistic attitudes. The former easily dismisses the question by referring to translation as an interlinguistic transfer; the latter stresses the cultural-time-space factors that account for a variational normative attitude to what translation must be and what translation products must look like. In both cases, the definitional problem remains. In fact, very few studies have systematically and programmatically evolved into principles, models or theories that may serve to explain or predict translation. The purely theoretical general branch has remained almost underdeveloped—as pointed out by some scholars (Stecconi, 2004, p. 473)—and the challenge to seek and bring about a general translation theory, or at least general principles, has rarely been taken on. Holmes’s criticism, made in the early 1970s, remains valid:

Most of the theories that have been produced to date are in reality little more than prolegomena to such a general translation theory. A good share of them, in fact, are not actually theories at all, in any scholarly sense of the term, but an array of axioms, postulates, and hypotheses that are so formulated as to be both too inclusive […] and too exclusive […].

Holmes, 2000 [1972], p. 178

Situated in the theoretical branch in Holmes’s map, and based on the general epistemic questions that this branch should deal with, this study outlines a new framework consisting of: i) a semiotic, abstract and general definition of Translation; ii) a formalisation of the practice of translation—situating the general and abstract definition of Translation in human activity—and identification of the agents and their interactional semiotic modalities; iii) a description of the specific enunciation situation in translation, and development of relevant translated-text analysis tools.

This theoretical device is applied to a specific translational reality and to a parallel corpus (texts translated from English into Italian and French). The research is set in an interdisciplinary field and draws on Greimassian semiotics. Advantages in adopting a semiotic view are clear: it allows us to conceive translation as a complex-beyond-language activity and provides an effectively applicable method to the study of translation, as a sociocultural, intersubjective and thymic practice involving a passional dimension that shapes and leaves its mark on the translated texts.

1. The Semiotic Paradigm

Originally, most of the scholars who have developed an interdisciplinary approach to semiotics and translation studies have referred to the interpretative semiotic school, as theorised by Charles S. Peirce, and have investigated translation inside the general process of semiosis (Petrilli, 1992, 2015; Gorlée, 1994, 2004; Cosculluela, 2003, Stecconi, 2004; Hartama-Heinonen, 2008). The semiotic research has focused not only on verbal translational processes but also on the importance of translation in cultural contacts and development (Torop, 2014), in communication processes (Petrilli, 2014) and between different sign systems (Gorlée, 2015). The importance of extending the boundaries of translation in order to include problems related to intersemiotic transfers—that take into account visual, audio and multimodal material—has resulted in a growing interest in translation studies that increasingly turn to semiotic tools and approaches. In fact, recent publications draw on semiotics to deal, for instance, with questions related to the resémiotisation across multimodal texts (O’Halloran, Tan and Wignell, 2016), the localisation of videogames (Bernal-Merino, 2016), the intersemiotic transfers between poetry and mathematics (Kempthorne and Donelan, 2016), but also to address broader issues such as the concept of creativity (Aguiar, Atã and Queiroz, 2015) or the ethics of translation (Petrilli, 2016).

The present study falls into the framework of this research and revolves essentially around the difficulties that arise when one is to define what translation is. In order to try to answer this question, we turn to the work of Algirdas J. Greimas (1970, 1976, 1983a, 1983a, 1987), Greimas and Courtés (1979, 1982, 1986) as well as to the Paris School Semiotics, which adopts methods and principles derived from Greimas.

The main tenets of Greimassian semiotics may be briefly summarised as follows: it can be defined as the theory of signification, the theory that renders explicit the conditions for the apprehension and production of meaning, in the form of a conceptual construction (Greimas and Courtés, 1982, p. 292). Greimassian semiotics is based on the following methodological principle: the meaning of any semiotic object unfolds, in a generative path, from the most simple to the most complex, from the most abstract to the most concrete, from a deep immanent[1] semiotic network of relations[2] to surface manifested structures. The generative trajectory of meaning is a theoretical conceptualisation that invests, constructs and manifests any semiotic object. The following representation illustrates the subcomponents of the trajectory—the loci of meta-semiotic construction and generation of meaning.

Figure 1

The generative trajectory

The generative trajectory
adapted from Greimas and Courtés, 1982, p. 133

-> See the list of figures

1.1. Semiotic and Narrative Structures

Deep semiotic structures constitute the most abstract level, the ab quo point of the generative trajectory of the meaning of any semiotic object. They are visually represented as a square, the Semiotic Square (fig. 2), which summons the logical articulation of the minimal units of meaning. These units, s1 and s2, are called primitive terms; they are interconnected through operations and relations—contrariety, contradiction and complementarity—and generate other, more complex units of meaning called metaterms. Terms and metaterms are not defined in a substantial manner but only as points of intersection, as the results of operations and relations in the semiotic square; thus, they are logic positions which can be invested with semantic categories and specific values.

Figure 2

The semiotic square

The semiotic square

-> See the list of figures

As an elementary structure of meaning, the semiotic square defines the fundamental conditions of existence of a concept, an individual or society (Greimas and Rastier, 1968, p. 87). It has been applied to varied semiotic objects, to the description of the articulation of the microsemantic universe of myths such as culture/nature, life/death, euphoric/dysphoric (Greimas and Rastier, 1968; Greimas, 1983); it has also been used for the analysis of universal categories of the imaginary in literary texts (Greimas, 1976); and it has been applied to religious, social, legal and political discourse (Courtés, 1991; Landowski, 1989; Bertrand, Dézé and Missika, 2009). The semiotic square is apprehended as meaningful when it is manifested in the human dimension. The relations and operations of affirming and negating by conjunction and disjunction in the semiotic square are thus converted into human activities (the narrative level) by those who act.

The distinctive feature of this level is the anthropomorphic representation of the operations […]. If you say anthropomorphic you mean interpreting the notion of operation in terms of “doing”. In other words, “doing is an operation that is made specific by the addition of a human classeme” […]. The syntactic operations of affirming and negating by conjunction and disjunction are thus re-written as a syntactic doing.

Ricoeur, 1989, p. 11

Thus, the narrative structures describe the doings of actants,[3] their performances and the modalities of these performances, as follows. Manipulation is the ‘causing-to-do’ (faire-faire) something, or the action of humans upon other humans with the goal of having them carry out a given programme. The actant Sender causes the actant Subject to do something, to realise a programme, a contract, and to reach something, namely the actant Object-value. Competence is the ‘wanting-to-do’ (vouloir faire), the ‘knowing how to do’ (savoir-faire), and the ‘being-able-to-do’ (pouvoir faire) of the actant Subject. Performance is the ‘doing’ (faire) of the Subject to achieve something and obtain the actant Object-value. Sanction is the ‘knowing’ (savoir) of the Sender regarding the conformity of the Subject’s behaviour and performing. Narrative grammar is translinguistic, since actants, the interaction between them, and the modalities characterising these interactions can be manifested in any semiotic system.

1.2 Discoursive Structures

Discoursive structures are the ad quem point of the generative trajectory. They set into discourse the narrative grammar through the domain of enunciation: actants are manifested as specific actors and characters of specific narrations acting in specific space-time dimensions, bearing and conveying specific themes and figures. Enunciation is defined as the domain which governs the passage from linguistic competence to linguistic performance, from virtual semiotic structures to structures that are realised in the form of discourse. Enunciation is logically presupposed by the very existence of the utterance, which contains traces or markers of an enunciation (Greimas and Courtés, 1982, p. 103). Thus, enunciation presupposes an enunciator—since the uttered discourse reveals the presence of the utterer—; however, the domain of enunciation is a semiotised place of structural conversion between immanence and manifestation, and removes any interpretation that allows for psychological and mentalist approaches to discourse. Greimassian semiotics has been since discussed and applied by many distinguished authors who form the Paris School Semiotics. With this paper, we propose to apply the heuristic method of the generative trajectory and tensive semiotics in order to provide a definition of the object of study and a description of the practice and the product of translation.

2. On the Concept of Translation

The concept of equivalence is an immanent feature in the theoretical discourse on translation and translating; implicitly or explicitly, and regardless of how it has been defined, the term equivalence has influenced and regulated the practice and theory of translation over time, becoming a supermeme (Chesterman, 1997b) and an immanent condition to translation. However, recent developments in the field underline the paradoxical condition that characterises the concept of translation. The undeniable non-equivalence of translation compared to the original, whether conceptual, ontological, pragmatic, semantic, in the medium, in the finality and so on, is evident, and the idea of “difference” is clearly theorised by translation scholars.[4] Difference is another condition immanent to the existence of translation.

Appropriate methodological tools can construct and explain the intuitive paradoxical and apparently contradictory condition of the concept of translation (Troqe, 2014, 2015). Therefore, we propose here to recognise the terms difference and equivalence as simultaneous contraries to be articulated in the semiotic square of Identity (fig. 3a).

Figure 3a

The semiotic square of identity

The semiotic square of identity

-> See the list of figures

According to the internal functioning mechanism of the semiotic square, the terms difference and equivalence are subsumed in the metaterm self. Here, self refers to the category Translation seen as an emerging identity, be it the self identity of the subject (for example, the translator), of the object (for example, the translated text) or of the phenomenon (for example, the practice of translation). The term difference is intended here to signify the contrast: in order to be and emerge as an individual and unique entity, the Translation-self shall first be different from the Original non-self. The translation is a different subject, a different linguistic expression, and a different practice[5]. The term equivalence refers to a condition of derivation; it enables the possibility (and the necessity) that the Translation-self emerge as a reference, an analogy, a simulation or a copy of something else, the Original-non-self. The question of identity represents a crucial epistemic aspect in the study of translation. The fact that in the semiotic square the Translation-self is generated in relation to the other, Original-non-self, makes explicit the confrontation, exchange, resistance or docility, compatibility or incompatibility ruling the existence of those identities.

Greimassian semiotics allows for semiotic concepts to be set in a veridictory framework in order to determine its degree of truth; in this specific case, what Translation is and what it seems to be. Integration of the veridictory modalities (Greimas and Courtés, 1982, p. 369) completes the semiotic square of Translation (fig. 3b below) and allows for Translation to be defined as an identity bearing values (equivalence and difference) that are considered to be a truth (with respect to the countertruth of the Original) based on paradoxality (being different while simultaneously seeming equivalent). This definition of the concept of translation as a semantisation of the semiotic square represents a system—an axiology—, a universe of virtual and abstract values that may refer to singularities that may occur in different forms and in different cultures. In fact, temporality is not a value in the first stage of this theoretical conceptualisation; it comes to be considered in a second stage, when the practice of translation is reified into a specific scenario. Therefore, further investigation into practices and theoretical discourses on translation is required in order to include these practices in the semiotic square of Translation and corroborate the degree of validity of the metadiscourse that it establishes.

Figure 3b

The semiotic square of Translation[6]

The semiotic square of Translation6

-> See the list of figures

As provided in the generative trajectory, where the virtual terms in the semiotic square are converted into the semio-narrative structures, the semiotic square of Translation is also actualised by the actants of practice of translation. This means a shift away from structures (the semiotic square of Translation) to praxis, towards the operations, actions, and modalities involved in translation.

3. On the Practice of Translation

How does translation practice begin, develop and end? In order to answer this question, one needs to identify the roles and forces necessary for accomplishing translation. Semiotically speaking, these roles and forces are actants (Fontanille, 2003), and they are abstract entities defined by the function and place they occupy in a practice. The necessity of constructing and defining actants and actantial functions stems from the semiotic conviction that nothing is given in reality and everything is to be constructed. Construction allows for meaning to emerge and theoretical arguments to be put forward. Now, the specific ambition of the present description of the practice of translation, or rather, of its formalisation, is to provide a representation of the actants that put the translational process into place, regardless of the particular realisation of those actants.

The Initiator is the actant who initiates a translational practice—a translation programme—by inducing into action another actant, the Translator[7], through persuasion or injunction—that is, semiotically, through manipulation. In the case study that we present in this paper (section 5), the Rolex Institute initiates a specific translation programme: the communication department of the Rolex Institute itself generates English texts—short descriptions of scientific projects recipients of the Rolex Award of Enterprise (RAE)—which are then translated by outsourced translators in eleven languages.

The Manipulation of the Initiator is a ‘causing-to-do’,[8] causing the Translator to accept acting in order to do something or to bring something into being. ‘Causing to do’ and ‘accepting to do’ generate a binding relation that establishes a contractual situation between the two actants. With Manipulation, the Initiator constitutes the Translator as a competent actant, who meets all the necessary conditions to carry out a performance in a specific translation programme. Competence is required and precedes performance. The Translator may be manipulated through persuasion or through injunction, and these two situations account for two types of contract: an injunctive translation contract or a permissive one. In the former case, the Translator is a competent actant facing the deontic modalities having-to-do (prescription) or having-not-to-do (prohibition). Injunctive contracts may sound absurd or extreme, but it suffices to think of highly prescriptive or totalitarian situations to see how they could depict actual realities[9]. This is the case, for instance, of the translation flow in the Soviet Union (Zauberga, 2005; Baer, 2006) or the Marxist translations done in China (Gamsa, 2008). In the latter case, the Translator is established as a competent actant, freely choosing to act and interact with the Initiator. This, perhaps, depicts most common realities, where the Translator is modalised through the volitive modalities ‘wanting-to-do’ or ‘wanting-not-to-do’. In the case study, the RAE translation, translators are modalised through persuasion, and their motives are based on compensation (of monetary or/and professional value); however, translators are chosen according to their training and diplomas, which ensures quality and professionalism[10]. We shall see, with the textual analysis, how this specific programme may affect the semiotic product of translation.

Manipulation and the resulting competence set the stage for the contractual instructions underlying the interaction between the two actants of the practice of translation. The contractual instructions may have different and varied forms; they may be explicitly dictated by the Initiator through manipulation or implicitly embedded in the competence of the Translator. The latter case leaves considerable scope to translate for the Translator, as it is set at the stage of practice as a competent actant. In the former case, the Initiator may specifically indicate not only how the Translator should translate but also, above all, what the Translator should translate. The two situations are extremes on a graded scale, from rigid to loose contractual instructions; in the case study, for instance, the Rolex Institute allows the translators a large degree of freedom in translational choices; however, translators also have to deal with specific terminological and layout constraints. Whatever the degree of freedom in the contractual instructions, the Initiator appears to be the gatekeeper of what should or should not be manifested in the identity of the translation as opposed to the identity of the original.

In the semiotic square of Translation, the Initiator acts as a generator—since it initiates translation—but also as a filter, allowing elements from the Original-non-self to be visible and possibly shape the identity of the Translation-self. The manipulation establishes the Translator as a virtualised[11] actant, because it ‘has to’ (or ‘has not’) or ‘wants to’ (or ‘wants not to’) act but has not yet done so. Manipulation demands competence, and competence establishes the Translator as an actualised actant under the modalities of possibility (‘being-able-to’[12]) and knowledge (‘knowing’). These two modalities refer to contingency and actual skills, to the possibilities and know-how, to the experience or sensitivity of the Translator. Performance establishes the Translator as a realised actant, that is to say, as a performing actant specifically with regard to the contractual instructions. The modalities of the performance are: ‘causing-to-know’ an identity, an Original non-self through the creation of another identity, that is, by ‘causing-to-be’ the Translation-self. The outcome of the Translator’s performance is evaluated, judged and sanctioned by the Initiator. With the Sanction, the Initiator acts as a verifier, examining the Translator’s performance and assessing its compatibility with the contractual instructions. The sanction is ruled by the alethic modalities of necessity (‘having-to-be’) and impossibility (‘having-not-to-be’), that is, what is necessary and appropriate and what is impossible, not acceptable for the translation to come into being. Thus, the sanction is not merely an act of scrutiny but also a regulatory act that adjusts the outcome of the Translator’s performance to what translation should be and should look like. In the present case study, we shall see how the idea of Translation is generated and how sanction is performed when the original text is also originated by the actant that initiates the translational programme. Seeing translation practice in terms of manipulation, competence, contractual performance and sanction entails a few consequences.

Norms and normativity (Toury, 1995; Chesterman, 1997a) result from structured interactions between actants[13]. First, the Translator’s performance may be governed by internalised norms shared by a community (Hermans, 1996). However, as a competent actant, it has large room for action and may choose whatever norm it regards as fitting the situation. Second, the situation is regulated by contractual instructions—whether explicit or implicit—and by the assertive sanction of the Initiator.

The ethical dimension, another sensitive topic in translation studies (Chesterman, 1997a; Pym, 1997), is regulated in the translation practice from the beginning; in fact, it is regulated even before it happens, in the motivational drive, which is the manipulation of the Initiator. Acceptance of the manipulation, endorsement of specific contractual sanctions and agreement to a sanction are all acts that require ethical decisions by the Translator. Ethical decisions may not actually result in a smooth acceptance, endorsement and performance. In fact, semiotics allows and theorises confrontation between actants in terms of polemical contractual relations (Fontanille, 2003).

The target public is not, in the present perspective, considered to be an actant. The Initiator defines the audience of potential readers of the translation. The issue of the public is inscribed in the contractual instructions and may motivate the Translation’s performance and the Initiator’s sanction, but it always remains a projection of the idea that the two actants have of it. The target public may be instructed by the translation contract and by the sanction of the Initiator. Thus, if the Initiator does state what the target public is, then those statements may affect the performance of the Translator and should appear in the outcome of that performance.

Translation is a practice bearing specific values or systems of values. By manipulation, performance and sanction, the Initiator and the Translator act to shape specific ideas, values on how to act, and what the translation should be like. They may refer to the ideas of fidelity, adequacy, re-contextualisation, reformulation, fluency, etc. Whatever values may be named, they fall into the general concepts of the semiotic square of Translation. According to the specific practice that shapes the dynamics of self/non-self, the translation may seem more or less equivalent and more or less create the illusion of similarity. The label “translated by” may be seen, for instance, as a guarantee or an illusion of equivalence; translation may be more or less different, and more or less hide the secrecy of alterity. Semiotic objects labelled as “translation” in most cases hide their being different from the source text, and above all, hide the reasons of this alterity. This may be the case of dubbing or news translation, for instance. These objects are manifested as translations, and their immanent and profound (or necessary) difference must be unsaid, in order for the specific object to function—if one assumes, for instance, that unveiling the recreation process in dubbing spoils the magic of a movie, or that unveiling the transformative pre-editing and post-editing in news translation (Troqe and Fontanille, 2015) could undermine trust in the information source.

The interactions between the Initiator and the Translator of contractual instructions, the Initiator’s manipulation and sanction, and the competence and pragmatic action of the Translator are all elements that require the introduction of a theoretical actant. This actant situates the action, the work and position of the research in the translation studies field. It is a sui generis actant because it operates exclusively on the level of theoretical sanction. The sanction of the actant Researcher is an epistemic act on the Translator's performance, and the conditions that put it into being, steer it and manifest it. It is an interpretative act, allowing a recognition of modalities and authorities shaping the translation practice. Now, three actantial positions and their actions may be brought together in a new formalisation of the translation practice, the semiotic model of Translation.

Figure 4

The semiotic model of Translation

The semiotic model of Translation

-> See the list of figures

The actant Researcher is virtualised by the modality ‘wanting’ (wants to generate scientific discourse about translation), is actualised by the modality ‘knowing’ (knows how to generate scientific discourse through semiotic competence), and is realised in the metatheoretical discourse it generates about the theory itself. In the semiotic model, the Researcher occupies an observational stance and operates on three levels:

  1. the semio-logic level, through the recognition of logical conditions arising in the semiotic square of Translation;

  2. the modal-ideological level, through the identification of the manipulative and sanctionatory elements that emerge in a specific contractual situation; in the present case study, we have adopted the stance of the researcher and tried to investigate the modal and motivational dimension of the RAE’s translational programme by carrying out interviews and fieldwork with the initiator and the translators;

  3. the textual level, through the identification and explanation of textual elements: recognition of the translational choices attributable to the contractual situation and performance. For the RAE’s case study we collected a parallel corpus and analysed it in the semiotic perspective that allows for textual marks directly ascribable to the actants of the RAE translational practice.

4. Enunciation in Translation

Greimassian semiotics defines enunciation as the linguistic domain that is logically presupposed by the very existence of the utterance. Enunciation governs the passage from deep and virtual semiotic structures (narrative grammar) into structures realised in the form of discourse (linguistic manifestation), as described in the generative trajectory (fig. 1). The conversion of semio-narrative structures into discourse is called semiotic competence and is exercised by the empirical instance of discourse production, the subject of enunciation. The subject of enunciation is not to be considered as having an anthropomorphised role, but rather, a position adopted by the actant performing the enunciation (Greimas and Courtés, 1982, p. 103). Enunciation[14] is considered to be a double predication: it is an existential predication, since it is concerned with appearance and disappearance, i.e. the modal component, virtualisation, actualisation and realisation, of utterances and of semiotic forms in the field of discourse. It is also an assumptive predication, since it assumes the utterances by making them present with respect to the position of the subject of enunciation; in other words, the act of assumption is, in fact, the act by which the instance of discourse makes its position known with respect to what comes about in its field (Fontanille, 2003). Thus, the utterance—the enunciated enunciation—is a simulacrum of the enunciation: it logically implies it and bears elements that are organised by it, such as deictic structures and modes of existence. In translation, the subject of the enunciation is mainly the Translator; nevertheless, the Initiator may also assume an enunciative position.

Research of the enunciation in translation may be done in two directions. First, analysis of the enunciation may focus on the enunciative praxis, that is, the translation in process, such as interpreting or vocalised translation. This kind of approach has been adopted by research on interpreting and by research investigating processuality or process-related factors (think-aloud monologue or dialogue protocols, eye-tracking, or key-logger software). These studies deal with the issue of translating as a problem-solving and strategic activity that involves professionalism (since studied subjects are often professional translators, or professional vs. non-professional subjects), timing, the degree of automaticity or of routineness involved in the translation process. Often, the question of product quality remains partially answered, and sometimes the theoretical framework is weak (Bernardini, 2011, p. 242). Second, an analysis of the enunciation may also be done through an inquiry of the utterance; because of the very fact that utterance is the enunciated enunciation, it necessarily carries elements pertaining to the enunciation itself. This approach is typical of translation criticism, where translation scholars and critics investigate a translator’s translation choices and systematize findings in a theoretical framework. Criticism studies especially focus on the products of the translation, and concentrate findings on the translational choices and their effects (Hewson, 2011). From the semiotic point of view, both the translational enunciative praxis and the ensuing translation utterance have to be embedded in a broader vision of translation, one that considers translated texts as a semio-linguistic utterance of an enunciation arising in a specific semiotic environment, where a competent actant is manipulated in order to perform according to given instructions and axiologies. Thus, in the present proposal, enunciation in translation is the mechanism that sets the values of the semiotic square of Translation and the modalities and action by actants in the semiotic model of Translation in a linguistic discourse.

With reference to the case study, the RAE’s translation programme, we shall see that in order to generate (‘causing-to-be’ and ‘causing-to-know’) an utterance of translation, the Translator, as a subject of enunciation, acts with reference to the value of equivalence and simulates the utterance of the original. It is known that the Translator’s utterance does not simply repeat that of the original. From the semiotic perspective, this appears to be obvious, since the translation enunciation is performed by different actants and occurs in a different time, space, and place and has a different motivation compared to the enunciation of the original. This is a simple recognition of the value of difference. Thus, the presence of equivalence and difference in the translation utterance, as translational markers, not only gives information about choices in translation but correlates those choices to the textual presence of the actants of translation. The translational markers represent tangible data that refer to the presence of the actants involved in the translation practice and can be concretely analysed, as in the case of the RAE’s translation project.

5. The Rolex Awards for Enterprise

The texts presented here as a case study describe scientific projects financed by the Rolex Institute in its philanthropic programme Rolex Awards for Enterprise (from 1978 to 2008, biannually). The data collected during the fieldwork at the Rolex Institute show that the translation was done by the same French and Italian translators over time; therefore, it was chosen to constitute a parallel corpus made up of original English texts and their translations into French and Italian. Original texts and translations were provided by the Rolex’s translation project manager; the corpus was composed of 109 original texts and the same number for the French and Italian translations, published in the timespan 1978-2008. Interviews with the translators allowed us to evaluate the semiotic environment they work in, as well as the specific manipulation and contractual instructions set up by the initiator. The original texts were produced by Rolex, and Rolex is at once the actant, performing persuasive manipulation over the translators, and sanction, over the outcome of their performance—the translated texts. In this case, the translators are manipulated under the modality ‘wanting to’, and are competent under the modalities ‘being-able-to’ and ‘knowing’ how to perform. As for the contractual instructions, these are loose and generate a wide scope of action for the translators.

The textual analysis was conducted by looking for translational markers in the translated texts, and different levels of semio-linguistic categories were taken into account: extent and intensivity, isotopic and thymic categories. Isotopy refers to the iterativity of units of content (semic values[15]) that ensure the homogeneity and coherence of the thematic level (dominant themes) and the actantial level (subject, object, antiactant, etc.) in a narration. According to the semiotic tensive model (Fontanille and Zilberberg, 1998), each semic value is determined by two functions: extent and intensity. Extent operates on the quantitative, spatial and temporal range. Extent sorts (classematic extent), expands (expansion extent) and condenses (synthetic extent) semic values pertaining to actantial or thematic isotopies. It also concerns the aspectuality, that is, the evaluative adjectives and adverbs, spatial and temporal deictics that refer to the instance of enunciation. Intensivity operates on the thymic level and increases or decreases euphoric or dysphoric semic values. Variations in these categories in the translated texts, compared to the original texts, are to be imputed to the subject of the translation enunciation.

A few examples extracted from the translated texts are reproduced below. Example 1 shows a variation coded as shift in aspectuality and an increase in thymism. Examples 2 and 3 show variations coded as extent (the translations expands the original) and increase in thymism. Examples 4 and 5 show a variation coded as classematic extent: the original “tolerance” in English becomes “love” in the Italian translation; the “dedicated scientist” is “armed with the passions for science” in the French translation. Example 6 shows a variation coded as synthetic extent since the translation condenses the original and slightly reduces the thymic dimension.

Table 1

Examples 1 to 6

Examples 1 to 6

-> See the list of tables

In order to gather quantitative data, translated texts were annotated according to the type of variations, using QDA Miner. This program allows for the creation of codes and assignment of these codes to parts of translated texts that manifest variations in the extent or intensity levels. Variations were detected in both the French and the Italian translations, but comparing the two corpora, codes referring to variations were different in quantity and nature. The total words in the French translation (fig. 5) is always higher than the total words in the Italian translation and in the original English texts. Even in the case of two Romance language translations, the texts translated in French are wordier than the Italian translations.

Figure 5

Word count in the original and translated texts (1978-2008)

Word count in the original and translated texts (1978-2008)

-> See the list of figures

However, the number of words in coded segments—i.e. in the segments coded as variations—in the French translation is always lower compared to the number of words in coded segments in the Italian translation (fig. 6).

Figure 6

Word count in the coded segments (1978-2008)

Word count in the coded segments (1978-2008)

-> See the list of figures

Focusing on the nature of the coded segments, in the French translations a total of 3383 codes were found: 21.4% of the codes referred to the category of aspectuality and 10.8% to euphoric intensivity. In the Italian translations, a total of 7318 codes were found: they referred mostly to the aspectual category (10.7%) and to euphoric intensivity (8.5). Compared to the French translations, the number of codes in the Italian translations was much higher and represents deep variations which are classematic in extent (7.7%), synthetic in extent (5.5%) and expansion in extent (4.4%).

For the years 1984 and 1993, a similar number and type of codes in both translations was reported. In particular, translations show increasing variations in enunciative dimension (transformation of direct speech into indirect speech; see ex. 7 and 8 below), or generation and suppression of information, compared to the original content. These findings are explained by a specific and punctual need for uniformity and harmonisation of the texts by Rolex in those years. This shows concretely how the sanction of the Initiator operates and how translations were edited internally in order to conform to specific needs for harmonisation or clarity. Below are some examples that show how translations were edited by the translator (on request by Rolex) or by proofreading societies, again upon request by the Rolex Institute. Examples 7 and 8 show transformation from direct into indirect speech, while example 9 shows how new information is generated compared to the original text.

Table 2

Examples 7 to 9

Examples 7 to 9

-> See the list of tables

In this particular translational practice, it can be said that the looseness of the instructions and the manipulation of the professional translators have mostly solicited the value ‘equivalence’ in the semiotic square. In both translations, equivalence is found in the deep grammar narration compared to the original: the same actants, same programmes, same themes and same plots are narrated in the French, Italian and English texts. The value ‘difference’ in the semiotic square of Translation is present as well, although to a moderate degree. In fact, in the discoursive structures of the translations, variations in extent and intensity not only point to the presential status and sensitivity of the translators, but to two particular ways of performing and thus participating in the construction of the concept Translation-self: the French translated texts show stability in the performance and in the realisation of the equivalence while the Italian translated texts display an increasing presence of the Translator and an escalation of the translational markers showing difference. The value ‘difference’ is also manifested in the act of sanction by the Initiator of the translation.


The present discussion has been situated in the general theory of translation. It has attempted to take up the challenge set by Holmes and brings forth a proposal whose ambition is not to provide universal answers but a sound theoretical framework for the study of translation and a scientific method to describe, analyse and perhaps predict it. The semiotic square of Translation is an abstract definition of the concept of Translation, and the semiotic model of Translation is based on a systemic theory of the phenomenon of translation. They are both as generic so as to be applied to different practices, in different space-time situations: the square articulates the fundamental conditions of the concept of translation, namely, the confrontation of identities in the dynamic of difference and equivalence; the model draws generalities on the actants that carry out translation in the human arena.

It is true that languages refer differently to the concept of translation and that it is difficult to find common ground, since historical and sociocultural criteria determine translation as an entity. However, if a relativistic stance is taken to the extreme, then one must identify definitions (or norms) that do not only take the space-time dimension into account, but also cultural, subcultural, minorities, communities, subcommunities, individual and contextual criteria. This means that the definition of translation may go down increasingly specific pathways. Conversely, one could state that to an increasingly specific determination of translation corresponds an increasingly general definition. Thus, universality here is considered as an abstract and impersonal generality that not only ends with a concrete application, but embraces different ways of defining and practicing translation. It is not the lowest common denominator of differences in culture and space-time, but the construction of the meaning that these differences generate.

There is, of course, an ethnocentric bias when specific categories belonging to an epistemic stance (as the one chosen here) are applied, although the generality, abstractness, and minimalism on which the proposal is founded should guarantee its applicability to any object of study that involves, generates and refers to the process of identity emergence and construction. However, more studies should be conducted to prove its pertinence. Eventually, the present theoretical framework should be seen as an epistemological approach that seeks to overcome impressionistic views in translation studies, according to which any research would necessarily be marked by the goals and interest of the researchers. Deconstructive statements like these are obvious, concern any scientific domain and do not bring further insight into the research. In fact, if difficulties exist in defining translation, these are not justified by the fact that the meaning of translation is more context-bound than others, but by the complexity that that meaning generates. One could now back down and give in to the so called anti-essentialist positions or try to deal with complexity by means of the heuristic approach, as conceived by A. J. Greimas and developed in this proposal.