What’s so “proper” about translation? Or interlingual translation and interpretative semiotics

1.1 Translation and interpretation: Jakobson and Eco

Umberto Eco, in the second part of Experiences in Translation (2001b) entitled “Translation and Interpretation” ^[1] does more than use Jakobson as an introduction to the topic or as a peg on which to hang a discussion about translation. Eco challenges the manner in which Jakobson used the term “translation” stating that in Jakobson’s famous 1959 tripartite distinction the term “translation” is used in a figurative and not in a technical sense. This use parallels one of Peirce’s most frequently cited phrases where he speaks of interpretation as “the translation of a sign into another system of signs” (CP 4.127). Peirce makes use of the word “translation” metaphorically, and not synonymously, argues Eco (Eco 2001b: 71), and since Jakobson was heavily influenced by Peirce’s writings, the metaphorical use was extended. Eco and Jakobson both express respect for Peirce’s work. They share Peirce’s definition of meaning and their own views of interpretation are derived specifically from Peirce’s notion that “a sign is something by knowing which we know something more” (Eco 1984a: 2), which implies the process of semiosis in that meaning is communicated through other signs, other words, ad infinitum.

Jakobson’s “fascination” ^[2] with Peirce’s description of meaning as “translating” a sign into another system of signs, led him to draw a parallel between interpretation and translation that seemed as though he were treating them as synonyms and thereby ascribing the same operation of understanding to each. ^[3] Not coincidentally, the definitions that Jakobson gives of the three types of “translation” all depend on “interpretation” in their explanations. This ambiguity between the two terms leads Eco to ask: “If all three types of translation are interpretations, did Jakobson not mean that the three types of translation are three types of interpretation, and that therefore translation is a species of the genus interpretation?” (Eco 2001b: 68).

Eco’s next step is to demonstrate that, logically, interpretation does not always imply translation. He provides the following argument:

there is a limit to translation, when we are confronted with ‘diversity in the purport of expression’. Having identified this limit, we are forced to say that, at least in one case, there are forms of interpretation that are not wholly comparable to translation between natural languages. (Eco 2001b: 73) ^[4]

This disambiguation of terms has relevant consequences. Not only is interpretation the genus of which translation is a species, and therefore not co-extensive with it, but interpretation also precedes translation in that every act of reading and understanding presupposes an interpretation, and only after that is the decision to translate or not to translate into another system, in this case another language, taken. ^[5]

Eco does more than question Jakobson’s tripartite distinction. In Experiences in Translation (2001b: 98) he proposes an alternative classification of forms of interpretation. His division is also tripartite, but the distinctions are between Interpretation by transcription, Intrasystemic interpretation, and Intersystemic interpretation, where “translation proper” falls into the third category. In addition, this classification expands considerably Jakobson’s first category of intralingual translation or rewording which, according to Eco “covers an immense variety of types of interpretation” (Eco 2001b: 68) that are not made apparent in Jakobson’s classification.

1.2 The lector and the interpres

Just as interpretation and translation have much in common, but are not the same, so do the reader and the translator have much in common but they, also, are not the same. The “reading” or interpretation that the translator provides is, by definition, written, fixed in a new set of signs, in a different language, and it becomes a new text with a particular relationship to the source or original text. ^[6] There is yet another important distinction between the interpretative activities of the reader and the translator: the reader’s understanding of the text in hand is, certainly, given in other signs, as Peirce insisted, but those other signs tend to be in the same language and are therefore instances of rewriting or intralingual translation, whereas the translator “rewrites” in a different language. This is a highly relevant distinction according to both Jakobson and Eco, but less so according to Steiner who uses the same term “translation” for both cases of interlingual and intralingual translation. ^[7]

Neither the reader’s nor the translator’s interpretations are identical, in an obvious sense, to the source text, as Derrida reminds us that in a translation “the letter is mourned to save the sense” (Derrida 2004), and in so doing the sense is embodied in a different letter. Despite this mournful loss, Eco argues that rewriting in the same language actually moves the text further from the source meaning than interlingual translation does. In the case of rewriting “[t]he substance of the expression also changes” as for instance, the word “cat” does not connote the same as Felis Catus, to keep to Eco’s example and consequently, “two substances of the expression are produced’ (Eco 2001b: 84). Dictionary definitions, paraphrases, explanations, commentaries, encyclopedic entries are all examples of rewriting in the same language and they are all examples that change the substance of the expression. ^[8] In the case of translation, on the other hand, the substance of the expression, or the sameness or identity of words in Derrida’s terms, is certainly changed, but the substance of the content, or the sense, is the same ^[9] (or close):

we can turn enunciations in one language into enunciations in another language because, even if synonyms do not exist at the lexical level, two different enunciations can nevertheless express the same proposition. (Eco 2003a: 345, author’s italics, my translation) ^[10]

In order to amplify this argument, Eco (2001b: 85) provides the following example:

1. Go home

2. Please go back to where you habitually dwell.

3. Torna a casa.

where (i) and (iii) encompass the same content and would ring identical to a bilingual speaker, whereas (ii), although within the same language, says something more, much more, than the original phrase.

These issues are particularly pertinent to the translator who (usually) does not want to say “something more” in the target text. The notion of equivalent effect ^[11] has its roots in the observation that the textual surface of a source and a target text can be extremely different yet the deep meaning can be the same. Ironically and perhaps even dangerously, the translator’s traditional tool is that of the dictionary, which, based as it is on definition, has the tendency to say more than the source word or phrase communicates. These issues will be discussed in more detail in the sections below.

2.1 The encyclopedia as regulative hypothesis

The cooperative behavior of the reader that Eco describes in The Role of the Reader (1979b) is the same as that of the translator inasmuch as it is focused on working out and building up the intentio operis. ^[15] It is precisely the intentio operis to which the translator is to be truthful. There is, arguably, a difference of degree between the cooperative readings of a “normal” empirical reader and a reading by a translator. The translator is a very specific kind of reader whose mode of reading is more careful and nuanced, often described in intimate terms such as the translator “caresses” and “cajoles” the text in drawing out its details and subtle meanings. In addition, the translator is not only a special kind of reader but also a particular kind of writer. Edith Grossman, in her compelling discussion of translation, describes the translator’s mode of reading in this way:

In the process of translating, we endeavor to hear the first version of the work as profoundly and completely as possible, struggling to discover the linguistic charge, the structural rhythms, the subtle implications, the complexities of meaning and suggestion in vocabulary and phrasing, and the ambient, cultural inferences and conclusions these tonalities allow us to extrapolate. This is a kind of reading as deep as any encounter with a literary text can be.

(Grossman 2010: 9)

Grossman describes the process as “hearing” the implications and complexities of the work. This contact with the text, this “deep … encounte,” this profound knowledge of the work is not translation itself but the reading that precedes it. All encounters, whether they are ephemeral, instantaneous, and subconscious, or long, deep, and very consciously undertaken, are instances of inferential processes. Eco demonstrates that to interpret a sign is not merely to recognize equivalence, but always requires a dynamic act of inference. He maintains that “inferential processes (mainly under the form of Peircean abduction) stand at the basis of every semiotic phenomenon” (Eco 1984a: 8). The inferential act precedes all others, as Eco explains in Semiotics and the Philosophy of Language: “we deal with both with language and with every other kind of sign by implementing inferential processes. These processes can be defined as interpretative processes. The understanding of signs is not a mere matter of recognition (of a stable equivalence); it is a matter of interpretation” (Eco 1984a: 43).

The close and detailed interpretation that a translator applies follows the same rules as other regular readings except that the translator tends to be more aware since the reading is a preamble to the translation task ahead. There are certainly degrees of awareness since the more familiar an instance of a word or an action, the more automatic our response becomes to the extent that we may become unaware of the inferential process taking place – just like driving a car takes up all our attention when we begin, but becomes almost automatic over the years. Giampaolo Proni (this volume) amplifies this point when he states that “interpretation is always an inferential act, what changes is only the degree of automatism and the certainty of conclusion.” In the case of the translator the intention is that of producing a new text that will, in ideal circumstances, evoke the identical inferential processes as those of the source text. Eco describes this in Experiences in Translation:

Interpreting means making a bet on the sense of a text, among other things. This sense that a translator must find – and preserve, or recreate – is not hidden in any pure language, neither a divine reine Sprache nor any Mentalese. It is just the outcome of an interpretative reference that can or cannot be shared by other readers. In this sense, a translator makes a textual abduction.

(Eco 2001b: 17)

“Textual abduction,” in Peirce’s sense, is based on habit making which becomes an automatic response to a text (or other stimulus) as a consequence of the familiar experience of reading. The reader’s knowledge of other texts, or lack of such knowledge, provides the pragmatic context within which interpretation takes place. This background knowledge determines individual responses and interpretations, as well as the collective or social responses. The passage just referred to above, describing the translator’s interpretative action, continues in this way: “Of course, the whole history of culture assists the translator in making relatively safe bets, in the same way as the whole theory of probability assists the gambler at the roulette table” (Eco 2001b: 17).

The “safe bets” or abductions depend on the pragmatics of contextual knowledge. To return for a moment to the notion of the “proper” in translation, we recall that the definition brought both the individual ownership of, in this case, the mental process of inference, and also the collective sense according to which a “proper” act, in this case of translation, “respect[s] recognized social standards or conventions.” In Interpretation and Overinterpretation (1992a) Eco describes the fact that a text can sustain multiple readings and that the singular interpretation that the author may provide is only one among many that the intentio operis determines to be acceptable. The shared nature of knowledge, effectively described in Stanley Fish’s term the “interpretative community,” underlines the collective nature of meaning and the social practice of language which establishes the shared, social nature of interpretation. The readers bring their cultural expectations to the text however, unlike Fish, Eco insists that the validation, the limits and the parameters of textual interpretation are to be found within the text itself, and not within the community. Eco writes that:

when a text is produced not for a single addressee but for a community of readers – the author knows that he or she will be interpreted not according to his or her intentions but according to a complex strategy of interactions which also involves the readers, along with their competence in language as a social treasury. I mean by social treasury not only a given language as a set of grammatical rules, but also the whole encyclopedia that the performances of the language have implemented, namely the cultural conventions that the language has produced and the very history of the previous interpretations of many texts, comprehending the text that the reader is in the course of reading.

(Eco 1992a: 67–68)

The postulate of “a social treasury” or an “encyclopedia” that holds “the cultural conventions” as well as the “history of previous interpretations” and other relevant knowledge, becomes part of the interpretation, of the sense-making process of reading. The notion of encyclopedia is a key notion in Eco’s interpretative semiotics. It is to be understood as a regulative theoretical hypothesis and not as a lexical model. ^[16] Eco describes it as no less than “the global semantic competence of a given culture” (Eco 1984a: 68), and in the Italian version of Semiotics and the Philosophy of Language, as the “book of books, and library of libraries,” a sort of key to culture of the kind his character Casaubon in dreamed of Foucault’s Pendulum. ^[17]

3.1 The cultural turn: semiotics and translation

The turning point for the study of translation, not yet called Translation Studies, was the 1976 conference in Leuven, Belgium. The self-reflective nature of the conference led to the redefinition of the study of translation as an independent discipline, emphasizing its many points of contact with cultural, political, and literary areas of study that bring into relevance the role of cross-cultural influence through translation, histories of textual manipulation, as well as of intertextual influences and intersemiotic factors, in addition to traditional linguistics.

By the 1980s there was much discussion on the nature of translation. Much of the previous scholarship on translation had centered on notions of the impossibility of translation, with poetry as the primary example of impossibility at the formal level, as well as on various issues of equivalence, which prompted Bassnett (2002 [1980]) to comment that “equivalence” must be the most overused term in the study of translation. These concerns were no longer of central interest to the study of translation which, by the 1990s, came to see translation within a broader context, as highlighted by the oft-repeated phrase “the cultural turn in Translation Studies,” a term that was rapidly adopted and that opened up new vistas for exploration, vistas that were always present but not fully perceived. The cultural turn in translation studies was part of a paradigmatic shift that took place in the study of various disciplines in the humanities in the late 1980s and early 1990s and, as Bassnett points out, “has altered the shape of many traditional subjects” (2007: 16), which inevitably include the study of translation.

Questions to do with textual difference, so pressing in the 1980s and early 1990s, have now been supplanted by questions to do with cultural difference, including racial, ethnic, gender or sexual difference. This is because questions that are now asked by theory no longer have to do with a priori conditions of translatability, but with a posteriori ideological and cultural factors that affect, not just translation, but also the translator.

(Kuhiwczak and Littau 2007: 5)

The relevant context is not only that of the two texts, but also that within which the translator operates. The opening up of translation to issues of culture and the “shared global knowledge” coincides with seeing translation within an encyclopedic structure, as Siri Nergaard and Umberto Eco point out when describing semiotic approaches to translation: “Translation therefore does not simply involve substituting single terms with their alleged synonyms, nor does it involve comparing sign-systems per se. Instead, it involves confronting textual situations against a background of different (partial) encyclopedias, that is, of specific forms of socially and culturally shared knowledge set in different historical situations (Baker 1998: 219).

The cultural and the linguistic relationship between the two texts is not the only factor of interest in the study of translation, but the manner in which translators are influenced and formed within particular cultural and historical moments, their particular encyclopedias, also play a dynamic role in the interpretation and translation of texts. The relationship is a triadic one, with the newly produced text as yet another sign, or interpretant, to be further analyzed in due course. There is no final or ultimate or perfect translation. Each translation of a text, such as the many translations into English of La Divina Commedia for instance, showing themselves to be products of their own age forging a dynamic relationship with Dante’s original.

The translator has the rather strange job of having to translating what is not there. The pragmatic spaces in the text which are the unwritten assumptions of a common cultural reality shared by the Model Author and the Model Reader through the textual fabric of the intentio operis. The empirical reader changes over time, and the unspoken context also changes in that it becomes increasingly unfamiliar and a visible obstacle to understanding, which often prompts translator to make use of additional notes to fill in those elements that the modern reader would no longer share with the Model Reader as created in the fifteenth century. The context of production and that of reception move further away from one other as time passes, so those semantic and pragmatic elements of the context of production become more visible as they become as more impenetrable. The past is another culture and the interpreter who works with older texts must be competent in the historical, cultural and intertextual encyclopedia of daily life in the past, as well as its language. Temporal distance and cultural distance function in a similar manner.

In addition, the translated text also reveals features that place it within its moment of construction, the habits and norms of translation practice which date the translation. These features in the target text also become more apparent as time goes by.

3.2 Equivalent effect

Eco’s discussion of equivalent effects in translation parallels the cultural turn in translation theory: the move away from linguistic notions of equivalence while highlighting aspect of the encyclopedic model of interpretation. The emphasis is not on equivalence as substitution, as in the dictionary model, but of equivalent effect in the translated text which aims to produce the same “inferential walks” for the reader of the translated text as the reader of the source text would encounter. The intention is usually, though not exclusively (since texts, even translations of texts, can be put to many uses) to say no more or less than the source but to produce the same cognitive, cultural and intertextual effects in the mind of the reader of the new text. The result is that the two texts might seem very different to one another at the surface level, but be more similar in the less obvious, deep-level effects created in the reader. Eco argued that “the concept of the sign must be disentangled from its trivial identification of coded equivalence and identity,” and therefore from the dictionary model, since “the semiosic process of interpretation is present at the very core of the concept of sign” (Eco 1984a: 1). In Semiotics and the Philosophy of Language, Eco elaborates on the distinction between the dictionary model, which he presents as “a bidimensional tree,” and the encyclopedic model, which is “able to represent the global semantic competence of a given culture.” He goes on to describe it further, stating that, “Such a global representation is only a semiotic postulate, a regulative idea, and takes the format of a multidimensional network that has been described as the Model Q” (Eco 1984c: 68). ^[19] And finally, he collapses the very model of the dictionary into that of the encyclopedia:

The tree of genera and species, the tree of substances, blows up in a dust of differentiae, in a turmoil of infinite accidents, in a nonhierarchical network of qualia. The dictionary is dissolved into a potentially unordered and unrestricted galaxy of pieces of world knowledge. The dictionary thus becomes an encyclopedia, because it was in fact a disguised encyclopedia.

(Eco 1984c: 68)

Any expectation of synonymy or bidimensional correspondence or substitution is effectively blown apart.

In Experiences in Translation as well as in Mouse or Rat?, Eco provides a range of examples taken from his own novels, which amplify the point of equivalent effect rather than literal equivalence. In describing the various solutions used by his translators Eco explains that: “In all these examples I invited the translators to disregard the literal sense of my text in order to preserve what I considered to be the “deep” one, or the effect it had to produce” (2003b: 70–71).

In The Name of the Rose (1980a), for instance, the frequent use of Latin, which was intended to provide the aura of the medieval church, was effectively translated into old ecclesiastic Slavonic to convey the cultural religious equivalent for Russian readers who, belonging to the Greek Orthodox Church do not associate Latin with the Church (2003b: 98–99). In Foucault’s Pendulum (1988b) a literary reference to a line from Giacomo Leopardi’s L’Infinito, the famous al di là della siepe, was substituted by William Weaver with the line “Like Darien, Diolallevi remarked …” alluding to John Keats’s “On First Looking into Chapman’s Homer” – which, incidentally is a poem about the effect of a text in translation – which, not only preserves the poetic reference and the idiosyncratic tendency of the characters to pepper their conversations with literary references, but Leopardi and Keats were also contemporaries and carry equal literary stature in both cultures (2003b: 67). In The Island of the Day Before (1994d) the colors of fish and other natural phenomena are conveyed in Eco’s original in lists containing many terms for colors following the decision, on Eco’s part, to avoid repetition of any of the terms within each single description. The translators were told that they could change the colors as long as they preserved the notion of variety as the intended effect; i.e., that of a multitude of hues and not of precise colors or precise numbers of shades of those colors (2003b: 68). Clearly, equivalence is not dependent on surface resemblance but rather on an interpretation of the intention that the text conveys, and of the Model Reader’s similar experiences in the reading of the original and the translated version.

A semiotic approach to translation, based as it is on the encyclopedia as regulative hypothesis and on the intention of the text as a parameter for acceptability of interpretation and translation, is, in its method, close to the sense-for-sense type of translation technique. This is the type of translation solution, which allows for substantial changes at the surface level in order to provide the reader with equivalent effects in the form of similar abductive processes as the reader of the original, or source text, will have experienced.

One of the greatest difficulties in translation is to avoid that which comes naturally through the principle of interpretation, according to Peirce’s, which states that “a sign is something by knowing which we know something more” (Eco 1984a: 2), since there are so many ways in which a translator may say something more than the source text. In Mouse or Rat? Eco argues that it is through the process of negotiation that the translator manages the balance of sometimes saying more and sometimes less than the source text. A translator has to accept that losses will be sustained at some level, but with careful balancing of effects, the translator can also gain in other places. A translated text is a negotiated one that, ideally, provides its new readers with an understanding of the experience of reading the source text.

3.3 Shakespeare’s implants

Eco begins both his books on translation, Experiences in Translation, Mouse or Rat,? and Dire la stessa cosa, with a discussion based on long examples of machine translation. They are amusing, as the title of the opening chapter of Mouse or Rat?, “The Plants of Shakespeare,” indicates. The translations bear a reasonable level of resemblance to the original but they require an added effort of interpretation by the reader to understand that the homonym “works” was translated by the computer software with reference to one context, the use of the term as in “public works” and not in a literary context as in Shakespeare’s oeuvre. The selection of context, which is the result of the immediate cultural and historical association of the name Shakespeare with the writer of the 1623 Folio edition of the collected dramatic texts, immediately puts the reader right – but the corpus of digitized information that the computer program has access to, and through which it selects words, clearly did not provide enough examples of this collocation to make the pragmatic choice evident to the translation software.

The blatant error, the slightly comic outcome based on the knowledge of the various underlying contexts, the understanding that any half-competent culturally socialized person would not make such a mistake, and would, if faced with the two options of impianti and opere in a dictionary, have absolutely no problem in selecting the right word in this context, that any negotiation between the two options was immediate and evident – these considerations highlight the encyclopedic knowledge that a person, consciously or unconsciously, brings to the specific act of translation. Such negotiated moves are an integral part of the act of interpretation. A reader is likely to be unconscious of the selections being performed since competent, efficient, and economic readers will have internalized responses based on the habits of numerous previous textual encounters to guide them. The more experienced the reader, the more detailed the nuances of the particular token of the type will become evident. A scholar of English sixteenth century dramatic texts will be far more aware, and will easily pick up intertextual references, textual anomalies, customary phrases, and the occasional fake, than the casual reader who will have to put more conscious effort into reading.

The reader of the badly translated text who might find the impianti error amusing will have interpreted the error by working backwards in order to recognize that the semantic area covered by the homonym “works” in English is covered by two terms in Italian. The act of disambiguation on the part of the reader is likely to be immediate and unconscious, the dual contexts apparent, and the sentence, as Eco intended, will elicit a slight reaction of amusement on the part of the reader.

This reaction kills two theoretical birds with one stone. The first is Eco’s argument that the underlying cultural encyclopedia is essential to successful translation and the related point that a dictionary model will not provide the pertinent information to disambiguate the terms (unless, of course, as all modern dictionaries do, the term is embedded within a sentence in order to provide a minimal encyclopedic context). And second, it brings to mind the conscious negotiation that is an integral part of translation activity since, as terms do not have equivalent and corresponding synonyms in all natural languages, choices have to be made and losses and gains endured in the imperfect act that translation will always be. Furthermore, translation shares with interpretation the fact that despite the multiplicity of acceptable variants in the translation of any given text, such a thing as a bad or unacceptable translation does exist. The involuntary smile at Shakespeare’s “implants” confirms this.