Article body

1. Introduction

The development and internationalisation of modern life has led to a huge increase in the demand for translation as a natural part of intercultural communication. In connection with written translation the focus is no longer on literary or Bible translation as an immense number of LSP (Language for Specific Purposes) genres have evolved and need translation. Modern high-tech society with its international cooperation and intercultural communication in business as well as in political and cultural life has led to demands for many different kinds of translation or translation-like activities which often exceed the boundaries of what translation theory traditionally terms translation proper. Numerous efforts have been made to define, exemplify and systematize what constitutes translation proper in the real world, but this is not the case as far as intralingual translation is concerned.[1] In practice we see many kinds of intralingual translation, but more often than not these are merely mentioned in passing by translation scholars. We see easy-readers for children, subtitling for the deaf (Snell-Hornby, 2006, p. 21), new translations of religious texts and the classics (see Steiner, 1975, pp. 28ff on diachronic translations) as well as for instance American versions of British publications. Harry Potter has been published in a special American edition replacing cultural words like biscuits, football, Mummy, rounders and sherbet lemons with cookies, soccer, Mommy, baseball and lemon drops (Hatim & Munday, 2004, pp. 4-5) and Denton (2007, forthcoming) showed that Sue Townsend’s novel of 1982 The Secret Diary of Adrian Mole, Aged 13 ¾ which is a cult best-seller in Britain is far less successful in the US, where its very British cultural codes and slang terms have impeded readers’ enjoyment, thus exemplifying the need for intralingual translation. In addition, highly functional LSP translations (which meet skopoi which differ greatly from those of the source texts), localisation, précis-writing, some kinds of news reporting as well as numerous varieties of expert-to-layman communication (patient package inserts containing information on medicine, tax leaflets based on new legislation, manuals for durable consumer goods, etc.) are all part of modern life, of reality, and the question is where do such activities fit in theoretically? My starting point for this article was to take a closer theoretical look at intralingual translation and how to describe this kind of translation and the strategies involved. I have for a number of years carried out research within medical expert-to-layman communication, in addition to my work within translation proper, and as a translation scholar been intrigued by the many similarities between interlingual and intralingual translation. However, despite Jakobson’s classical definition, intralingual translation or rewording is extremely peripheral to translation studies, more so than it deserves, and the relationship between interlingual and intralingual translation is a neglected area of research, as is a thorough description of intralingual translation; it is next to impossible to find any relevant literature. My initial research into the nature of intralingual translation made me aware of the fact that since the time of Jakobson’s definition, general definitions of translation have become less inclusive. This, I think, is a major setback as there seems to be much to gain theoretically as well as practically by looking for similarities and differences between the various kinds of translational activities carried out. With the ulterior motive of putting intralingual translation (back?) on the map of translation studies and encourage future empirical research, including my own, within this area I shall discuss various suggested definitions of translation and argue for a broader perception of translation and consequently of translation studies as a discipline. Finally, I shall attempt to draw up an open definition of translation which reflects the many-faceted nature of translation. As a starting point, I would like to revisit and examine Jakobson’s classical definition as well as Steiner’s subsequent discussion and application of Jakobson’s work.

2. Communication is Translation

2.1. Jakobson’s Classical Definition of Translation

Translation studies is engaged in the academic study of translation and it is therefore common that works on translation devote chapters or paragraphs to a definition of ‘translation’ as a concept. Intuitively even laymen would know what a translation is and would probably define it in a way which corresponds to the prototypical ‘translation proper’ in Jakobson’s terminology. It seems that also many translation scholars rely on Jakobson’s three kinds of translation for their definitions of what constitutes translation. Jakobson builds on Peirce’s theory of signs and meaning and postulates that “the meaning of any linguistic sign is its translation into some further, alternative sign” (Jakobson, 1959, TSR, p. 114). The implication is that translation is a component in all language transactions and Jakobson divides these transactions into three kinds of translation or “ways of interpreting a verbal sign”:

  • Intralingual translation or rewording is an interpretation of verbal signs by means of other signs of the same language.

  • Interlingual translation or translation proper is an interpretation of verbal signs by means of some other language.

  • Intersemiotic translation or transmutation is an interpretation of verbal signs by means of signs of nonverbal sign systems.

Jakobson, 1959, TSR, p. 114

Interlingual translation, or translation proper, is naturally enough also by translation scholars seen as the classic, prototypical kind of translation and many scholars even want to limit research to very restricted definitions of translation proper (Tymoczko, 2005, footnote 3, p. 1096). Even though translation scholars mention and acknowledge Jakobson’s other kinds of translation these are often classified as peripheral or de facto considered of no real relevance to the discipline of Translation studies. Newmark (1981, p. 12), for one, would certainly not include interlingual and intersemiotic translation and even Munday (2001, p. 5) who, having explained Jakobson’s three kinds of translation, states that “[it] is interlingual translation which is the traditional, although by no means exclusive, focus of translation studies” would probably find it a bit of a challenge to provide abundant examples of studies focusing on the other two kinds of translation. It seems to me that his statement is rather an endorsement of Jakobson’s three kinds of translation as relevant for translation studies than a reflection of reality within the field of translation studies. Once Jakobson’s seminal text—with its very broad philosophical and hermeneutic definition of translation—has served its purpose of defining translation, authors quickly move on to the field of translation proper, or to the restricted area of translation proper which has their particular interest. This observation is supported by Tymoczko (2005, p. 1084): “What most translation scholars would like to believe is that the stage of defining translation is essentially over: it would be satisfying to think that the big parameters regarding translation have been sketched out. (…) This task of defining translation is not finished and it will continue to be a central trajectory of translation research in the decades to come.”

And as regards intralingual translation more specifically, Baker, in the preface to the Routledge Encyclopedia of Translation Studies, argues that we have been narrow and restrictive in defining our object of study and expresses concern about the lack of research outside the field of translation proper:

(…) intralingual translation is not such a minor issue as the existing literature on translation might suggest (…) I know of no research that looks specifically at the phenomena of intralingual or intersemiotic translation. We do have classifications such as Jakobson’s, which alert us to the possibility of such things as intersemiotic and intralingual translation, but we do not make any genuine use of such classifications in our research.

Baker, 1998, p. xvii (My emphasis)

We rarely see empirical work within the fields of intralingual, or intersemiotic, translation and only few scholars treat the subject theoretically. Within polysystem theory and the realms of literary translation (Even-Zohar, 1990) we do in fact find some work on intralingual or intersemiotic translation (e.g. Weissbrod, 1998, 2004 and Shavit, 1986). Translation, whether interlingual, intralingual or intersemiotic, is seen as part of the semiotic concept of ‘transfer’ and special focus is on transfer from one culture to another. According to Weissbrod (2004, p. 24), reflecting the views of Even-Zohar, researchers should deal with all these examples of transfer within one theoretical framework as the mechanism of transfer is largely the same in all instances. Also Eco (2001, pp. 75-77, p. 101) discusses intralingual translation, but not systematically and with sole focus on literary translation. Eco argues for a metaphorical reading of Jakobson’s use of the term ‘translation’ to mean interpretation. That is Jakobson’s model should not be read as a model of translation in the literal sense, but of various kinds of interpretation. In other words ‘translation’ in the literal sense is a species of the genus ‘interpretation’ (Eco, 2001, p. 68). Eco proposes a new and more detailed model as an alternative to Jakobson (2001, pp. 99ff). Pym argues that Eco wants translation to remain “translation proper” and that in this “Eco’s experiences run firmly against the trend of contemporary Translation Studies” (2003, p. 254). Relevant for this article, Eco still treats all the different kinds of translation or interpretation as related activities. Intersemiotic translation is not the subject of this article, but for those interested it should be mentioned that some scholars have indeed worked within this field. Petrilli (1992, 2003) contain important work on intersemiotic translation, however, it is important to note that Petrilli does not really work with intersemiotic translation as exemplified in Jakobson’s model, but rather as a phenomenon ever-present in all kinds of (interlingual) translation: “The role of translation is fundamental in the very constitution of the sign, both verbal and nonverbal, in the very determination of its meaning” (Petrilli, 1992, p. 234). Also Eco discusses intersemiotic translation and provides numerous examples of the phenomenon (in Jakobson’s sense) as well as translation between systems other than verbal language, a kind of translation not covered by Jakobson (2001, pp. 67ff).

Jakobson argued for a broad, inclusive definition of translation as a phenomenon fundamental to all language transactions. This line of thought was further elaborated by Steiner in 1975.

2.2. Steiner’s Hermeneutic Approach to Translation

Peirce/Jakobson’s postulate that the meaning of any linguistic sign is its translation into some further, alternative sign led Steiner to conclude as a natural consequence that “Translation therefore is the perpetual, inescapable condition of signification” (1975, pp. 260-261). This ‘inescapable condition’ is what makes Steiner claim that communication equals translation and also explains why Steiner takes an interest in translation. To Steiner translation is a fundamental within the fields of language and communication: “To study the status of meaning is to study the substance and limits of translation” (1975, p. 414). A theory of translation in Steiner’s sense would be a theory about the operation of language itself and would include all three of Jakobson’s kinds of translation. Steiner himself points out that a theory of translation can of course be limited to translation proper, but he prefers the all-inclusive version “because it argues the fact that all procedures of expressive articulation and interpretative reception are translational, whether intra- or interlingually” (1975, p. 279). Tony Bex points out that Steiner’s thesis can be stated relatively simply as in the following quote from Steiner (1975, p. xii) himself:

After Babel postulates that translation is formally and pragmatically implicit in every act of communication, in the emission and reception of each and every mode of meaning, be it in the widest semiotic sense or in more specifically verbal exchanges. (…) Translation between different languages is a particular application of a configuration and model fundamental to human speech even where it is monoglot.

Bex, 2006, p. 132

Or put very simply in the words of Steiner’s reviewer Woodcock, Steiner claims that “all communication, all reading, all listening, involves an act of interpretation, since no two people speak identically the same language” (Woodcock, 1975, p. 326). To me Steiner’s views and claims are very convincing and are extremely useful in trying to understand and argue for the fundamental nature of translation. But of course Steiner’s conclusions should be treated as hypotheses, as Woodcock sensibly emphasises—as does Steiner himself—we are moving within an area where there are no certainties, an area where we can speculate rather than know and he suggests that we must accept After Babel “for the fertility of its suggestions rather than for the finality of its conclusions” (1975, p. 328). Still according to Woodcock, “…the most original aspect of After Babel is Steiner’s attempt to establish translation not merely as a metaphor for the endless process of interpretation in which speaking and writing and reading involve us, but also as the most crucial functional example of that process” (1975, p. 326).

In other words, translation is both seen as something fundamental involved in all kinds of human communication, but also as a certain kind of human communication—and this is where we approach the more specific field of translation studies. To Steiner translation proper is seen as a heightened case of the process of communication and reception, but relevant for the purposes of this article, Steiner claims that the linguistic problems implicit in interlingual translation are already implicit in all intralingual discourse (Steiner, 1975, p. 414): “The first two categories [rewording + translation proper] are, at crucial points, similar” (Steiner, 1975, pp. 260-261). Again Steiner builds on Jakobson who also discussed the fact that e.g. the challenge of synonymy in intralingual translation resembles the challenge of equivalence in interlingual translation (Jakobson, 1959, TSR, p. 114) (see also Dam-Jensen and Zethsen, 2007). Complete equivalence as well as absolute synonymy is very rare [if it exists at all] and this fundamental dilemma of interpretation is shared by ‘rewording’ as well as ‘translation proper’ (Steiner, 1975, p. 261): “What Jakobson calls ‘rewording’—an interpretation of verbal signs by means of other signs in the same language—in fact raises issues of the same order as translation proper” (1975, p. 414).

3. Translation Studies and the Hermeneutic Approach

In spite of the highly influential texts of Jakobson and Steiner and in spite of their claim of close affinity between interlingual and intralingual translation, translation studies often exclude intralingual translation either deliberately or de facto. Gutt (1991) mentions the fact that not all translation scholars would feel comfortable with a broader definition which would allow for instance summaries and elaborated versions to qualify as translation (TSR, pp. 394-396). Newmark, as a case in point, would consider such activities to be what he terms ‘restricted translation’ falling outside the scope of translation theory proper (Gutt, TSR, 1991, p. 394; Newmark, 1981, p. 12). Newmark is a representative of the most narrow perception of translation which he defines as follows, “to cause what was stated in one language to be stated in another, with the purpose of achieving the semantic and expressive equivalence of both statements” (1999, p. 152—the definition is a translation from the French dictionary Petit Robert). He considers it a “perfectly adequate definition of the basic translation activity, however much modification and differentiation it may require in the case of this or that translation task” (ibid.). Even though Newmark acknowledges that other kinds of translational activities take place in practice he is not interested from a theoretical point of view, but maintains his basic definition and simply adds that “[a]ll the others, whatever you like to call them—surtitling or subtitling [and other translational varieties]—have to use this as a point of reference” (Newmark in Schäffner, 1999, p. 135). Newmark’s definition is very narrow indeed; it includes only translation proper and strongly relies on the much criticised concept of equivalence (see Zethsen, 2004 and Snell-Hornby, 1995). In other words, Newmark’s definition leaves no room for changing skopoi or intralingual translation and it implies that the field of translation studies is not open to insights originating from other kinds of translation than translation proper. Newmark’s definition may be prototypical for him and many others, but a prototype is culture-dependent, i.e. not a universal, static phenomenon (see Tymoczko, 1998). As such, it consequently does not necessarily represent translational reality today. This is discussed by Snell-Hornby (1999, pp. 161-164, 103-120 and 2006, pp. 130-139) who heavily criticises and warns against a too narrow perception of translation (in general, and Newmark’s definition in particular)—“the traditional linguistic transcoding activity” (Snell-Hornby, 1999, p. 164)—and of the tasks of the modern-day translator, which need to be reflected to some extent in our theoretical discussions if they are to remain meaningful. Also Schäffner points out that translational reality exceeds interlingual translation and she claims that it is increasingly agreed within the academic community of translators that “the translator’s responsibilities go well beyond what was traditionally considered a ‘translation proper’” (1999, p. 98). Schäffner relevantly asks whether translation itself is affected by modern developments “or is it rather that more and more activities are added to translation proper? Where does translation stop and something else take over, e.g. technical writing, or desktop publishing?” and whether this means that we would have to redefine the very notion of translation? (1999, p. 100). I think we do need to redefine, even though we might actually ask “redefine what?” as there is no consensus within translation studies about one particular definition.

Having established that translation of one kind or another is fundamental to all language transactions it is time to concentrate on a more specific, though still broad and inclusive, definition of translation. Hermans (1996, p. 1) sides with Steiner’s hermeneutic approach and equals understanding with translation, but also warns: “Once we have reached this point, the point where we understand ‘understanding’ as ‘translation,’ we can broaden our scope. In fact we can broaden it so much that it is hard to see where the end might be. Translation then very nearly becomes the human condition. Every act of understanding involves an act of translation of one kind or another.” I take this to mean that though we basically accept Jakobson’s and Steiner’s hermeneutic approach we do need a more detailed definition if we want to define the particular field of translation studies for research purposes as “it is not possible to proceed with research either abstractly or concretely if scholars do not define or delimit the object of study” (Tymoczko, 2005, p. 1083). Even on the basis of a hermeneutic approach it still makes sense to try to delimit the field of translation studies, the question is in which way? As a starting point I would like to examine a very influential, broad and highly pragmatic definition of translation suggested by Toury (1985, 1995).

4. Toury’s Definition of Translation

The inability to make a satisfactory finite definition of translation and at the same time include all the objects that human societies have identified as translations led Toury to search for a more pragmatic definition (Tymoczko, 1998, p. 3). Toury wanted to tackle the question of why something is regarded as translational, and not why it should have been (Toury, 1995, p. 33). Important for this article Toury set out to extend the range of objects of study to match real-life situations which are regarded to be translational activities. Toury shies away from an absolute definition of translation, but prefers to talk about a “working hypothesis” (1995, p. 32) in order not to get too restrictive and thus counterproductive. His working hypothesis is that all “assumed translations,” i.e. texts which by the relevant culture are taken to be translations, are translations and a translation can therefore be defined as:

all utterances which are presented or regarded as such within the target culture, on no matter what grounds. Under such observation, there is no pretense that the nature of translation is given, or fixed in any way. What is addressed, even in the longest run is not even what translation is in general, but what it proves to be in reality, and hence what it may be expected to be under various specifiable conditions.

Toury, 1995, p. 32

This means that what is assumed to be translation may vary considerably over time or from culture to culture and Toury’s definition will be able to accommodate them all due to its inherent flexibility. More specifically, Toury (1995, pp. 33ff) suggests three conditions for a given text to be a translation: the Source Text Postulate, the Transfer Postulate and the Relationship Postulate. These postulates pose the following requirements:

  1. The Source Text Postulate

    The existence at some point in time of a source text in another culture/language (or as Chesterman (1997, p. 62) interprets Toury “normally in another language”).

  2. The Transfer Postulate

    The translation has been derived from the ST via a transfer process.

  3. The Relationship Postulate

    There is an intertextual relationship between the two texts.

Toury explains his postulates in this way:

If we now proceed to take the three postulates together, an assumed translation would be regarded as any target-culture text for which there are reasons to tentatively posit the existence of another text, in another culture and language, from which it was presumedly derived by transfer operations and to which it is now tied by certain relationships, some of which may be regarded—within that culture—as necessary and/or sufficient.

1995, p. 35

The parameter of quality is irrelevant as regards status as a translation. A “bad translation” is still a translation (Chesterman, 1997, p. 60),[2] that is, Toury’s definition is a category judgement not an evaluative one. The category will be represented by a great variety of different instances ranging from prototypical instances to less typical ones, eccentric ones, peripheral ones etc. (1997, p. 63). One of the great advantages of Toury’s definition is its flexibility, the fact that the categorisation and acceptance of a text as a translation may vary hugely from culture to culture and from age to age, the only requirement being that the text is assumed by its audience to be a translation (a requirement which will be further discussed below). The following two examples show that Toury’s definition is able to handle intralingual translation, but only to a certain extent:

It seems that Toury’s definition is only able to handle intralingual translation in combination with interlingual translation, the crucial factor being the interlingual dimension which means that the text is published and marketed as a translation and thus lives up to the criterion of “assumed translation.”

In relation to the above discussion, Chesterman (1997, p. 62) argues that the last of Toury’s postulates is somewhat problematic and relevantly asks “what kind of intertextual relationships count as translational ones?” In relation to the third postulate there is bound to be enormous variation through time and across cultures. As a way of determining whether the third condition has been fulfilled, Chesterman introduces the concept of “relevant similarity,” though at the same time acknowledging that both ‘relevant’ and ‘similarity’ are in themselves hard to define. Indeed, one has only to think of the problems surrounding the concept of ‘equivalence’ to agree.[3] Chesterman sees ‘intertextual relationship’/‘relevant similarity’ as a concept which can be used as the basis of research within a given culture, in a particular translator, period, etc., that is, ‘relevant similarity’ is not meant as a finite definition, but as a useful working concept for research. And he concludes: “On this view, then, the boundaries of the concept ‘translation’ are ultimately not set by something intrinsic to the concept itself, but by the ways in which members of a culture use the concept” (1997, pp. 62-63).

4.1 Toury’s Definition in Relation to Intralingual Translation

Toury’s definition of translation is very attractive in its pragmatism and flexibility, however, the definition seems to be limited to ‘translation proper’—or translation proper in combination with rewording. As far as the three postulates go they are all equally relevant to intralingual translation:

Especially the requirement of an intertextual relationship, or Chesterman’s relevant similarity, which I would prefer, is a very useful concept when explaining the nature of intralingual translation. However, there are two factors which make Toury’s definition problematic to intralingual translation:

The first factor concerns the requirement of two languages

In his description of the Source Text Postulate (1995, pp. 33-34) Toury writes about the existence of a source text in another culture/language or language/culture. From this it is difficult to see whether he means that another culture suffices or whether both another culture and another language are required. Chesterman (1997, p. 62) interprets Toury’s writings as “normally in another language,” but in the above quote from Toury (1995, p. 35) he specifically mentions translation between two languages: “in another culture and language” (my emphasis). This does of course exclude intralingual translation (unless you apply a very broad definition of ‘language’). Anyway, it would be quite easy to extend the requirement to include different genres, as would be the case in intralingual translation (or mediums in connection with intersemiotic translation), instead of different languages, and therefore the first factor should not constitute a real problem.

The second factor is much more problematic as it concerns the very basis of the definition, the requirement that a text should be regarded as a translation to be categorised as a translation, i.e. the concept of “assumed translation”

It is highly unlikely that the majority of intralingual translations would be ‘assumed translations’ despite their many affinities with ‘translation proper.’ Pym wonders “exactly who is supposed to be doing all the assuming” (2006, p. 4), which is in fact a quite fundamental problem, but one which I do not consider relevant to discuss at length in the present article. The general public would not be sufficiently aware of the fact that these texts (typically e.g. texts within expert-to-layman communication) have been ‘translated’ from another genre and that they adhere to the three postulates. Still they should not be excluded from the definition of translation on these grounds, but effectively they are and if the community of translation scholars were to use Toury’s definition to determine their research interests we may well lose out on valuable insights. It should be emphasised though that Toury himself is by no means interested in excluding any areas of interest, but merely suggests that if a translation scholar works with texts which are not assumed translations in a particular culture but which are still deemed of interest to translation studies it should not deter anybody, only the fact that the texts in question are not regarded as translations should be mentioned and accounted for (Toury, 1995, p. 32). This is of course a very sensible and practical comment, but it does not solve the theoretical problem of how to include intralingual translation in a meaningful definition.

5. Translation as a Cluster Concept

As pointed out by Tymoczko, Toury’s pragmatic definition of translation has been a highpoint in the particular strand of translation studies which attempts to define translation. His definition is “congruent with the notion of translation as a cluster concept,” and among other things it “allows for cultural self-definition and self-representation in the field” (2005, p. 1086). Toury’s definition is undoubtedly a step in the right direction as compared with more rigid and exclusive/finite definitions, but as Tymoczko notes, the definition is “not fully satisfying if we want to know more about the nature of the concept of translation and to be able to say more about its (permeable) boundaries. We might like to know more, for example, about the range of translational phenomena, the sorts of things that enter into decisions by various cultures to identify certain phenomena as translations and reject others as not translations” (2005, p. 1086). In the context of intralingual translation it is quite obvious that Toury’s definition, though attractive, is insufficient as it relies on a concept (‘assumed translation’) which is likely to exclude the majority of intralingual translations. But is it at all possible to define translation in a very inclusive, yet meaningful way? Before I proceed with an attempt to define translation in a way which includes intralingual translation, I shall take a look at the way we, as human beings, tend to categorize the world (see also Zethsen, 1997) and, inspired by Tymoczko (2005) I shall examine what this implies for the task of defining translation.

5.1 Categorisation

Human beings have an innate tendency to categorise and are generally pleased when things can be pigeon-holed. Since the time of Aristotle the classical theory of categorisation has been part of Western culture and involves shared properties as conditions for category membership, strictly objective conditions for category membership and clear boundaries between categories, i.e. no borderline cases. In logic the specification of necessary and/or sufficient conditions is used to determine category membership, to provide precise definitions.

As early as in the 1930s, Wittgenstein noticed that something was wrong with the classical way of categorising. He discovered that there are categories, such as ‘game’ (and ‘language’) in which the members do not share common properties. The category ‘game’ exists on the basis of a whole series of similarities and relationships. Like a family various games are similar to each other, but it is impossible to find a single, well-defined collection of properties common to them all. Some pairs of games may not share any features with each other, but they will share at least one feature with one other game. Consequently, Wittgenstein introduced the very pragmatic and empirically-based (that is, not on formal logic) open concept of family resemblances (also called ‘the cluster concept’) and argued that this is what unites the category of game:

And the result of this examination [of the word ‘game’] is: we see a complicated network of similarities overlapping and criss-crossing: sometimes overall similarities, sometimes similarities of detail. I can think of no better expression to characterize these similarities than “family resemblances”; for the various resemblances between members of a family: build, features, colour of eyes, gait, temperament, etc. etc. overlap and criss-cross in the same way.—And I shall say: “games” form a family.

for a discussion of ‘language’ as a cluster concept see Wittgenstein, 1958/1953, section 66-67

Wittgenstein furthermore pointed out that the classical theory fails to show that there can be good and bad examples of a category. This was followed up in 1969 by the anthropologists Berlin and Kay who, based on a colour study, found out that subjects differed widely when asked to establish boundaries between colours, but that they generally agreed when they had to point out typical examples of various colours. Berlin and Kay introduced the notion ‘focal points’ which refers to typical examples of a category (in 1973, Labov reached similar conclusions when he asked subjects to label more or less cup-like objects). Based on a large number of experiments Rosch (1973) disproved the classical objectivist theory of categorisation and presented her own theory of natural categorisation. According to Rosch human beings categorise by means of prototypes, i.e. many categories are mentally represented by means of schemata of their most characteristic members. Other members constitute borderline cases and are peripheral in nature (a blackbird is a more prototypical bird than a penguin, but a penguin is still a bird).

As Wittgenstein pointed out the problem is that many useful concepts are too fuzzy and complex to admit of simple, clear definition. Putnam showed that even the definition of a seemingly simple concept like ‘lemon’ should not rely on necessary and sufficient conditions: “The meaning of lemon, for instance, might be decomposed into such characteristics as: round, yellow, having peel, having a tart taste, and so on. Yet none of these components is necessary: a green lemon is still a lemon, a sweet peel-less lemon is still a lemon” (Putnam in Johnson-Laird, 1983, p. 191). The notions of family resemblances and prototypology help us define ‘lemon’ in a much more meaningful way and no doubt to define a complex and fuzzy phenomenon like translation. The all-important difference from Wittgenstein and Rosch to the classical way of categorising is the absence of necessary and/or sufficient conditions. If we regard translation as a cluster concept, i.e. an open concept (as suggested by Tymoczko 1998 and 2005) our requirements for category membership do not take the form of necessary conditions, but family resemblances.

That the definition of certain concepts needs to be open is simply an inherent feature of such concepts; in other words it is not a flaw or deficiency that a concept cannot be finitely described, neither theoretically nor pragmatically. Also within biology the cluster concept is used (to determine species), and Pigliucci argues in line with Wittgenstein that the apparently unsatisfactory “fuzziness” of complex concepts is not due to our ignorance, it is an inherent feature of the concept and of the reality it is supposed to capture (2003, p. 600).

On the basis of these insights I shall attempt to set up an alternative definition of translation which is able to contain a wide range of translational phenomena, including that of intralingual translation, but which is still narrow enough to be meaningful to the field of translation studies.

6. An Open Definition of Translation

Toury’s well-known definition of translation was definitely a step in the right direction with its flexible and target-oriented approach. However, if we accept that translation can only be accounted for in a meaningful way if we treat it as a cluster concept it does not make sense to attempt a too finite description. As has been pointed out above, Toury’s definition is likely to exclude many intralingual (or intersemiotic) translations, not because of his three postulates, which themselves rely on open concepts, namely ‘source text,’ ‘transfer’ and ‘intertextual relationship,’ but because of two necessary conditions; to constitute a translation, a transfer process must have taken place between two languages/cultures and most importantly the resulting product must be assumed to be a translation by people in general (i.e. factors 1 and 2 as discussed above). In my view neither of these two requirements are necessary conditions for a document/product to constitute a translation and the intralingual translations in the following examples would be excluded:

None of Toury’s three postulates: the Source Text Postulate, the Transfer Postulate or the Relationship Postulate constitute a problem (if ‘text’ is interpreted broadly, i.e. as any oral, written or other semiotic manifestation) to neither intralingual nor intersemiotic translation. What they offer is a description of the fundamental activity of translation, but a description which relies on open concepts and which is therefore able and willing to include a whole range of examples, from the prototypical to the peripheral, depending on the distance in question from the prototypical definition of the concepts involved. Pym criticises the formulation of the three postulates as “hardly elegant” and points out that the relationship postulate is inherent in the source text and the transfer postulate (2006, p. 4). I tend to agree with Pym (though I also think that Toury merely wanted to make the intertextual relationship explicit by awarding it its own postulate) and consequently I shall only work with the first two postulates. I think it is possible to describe translation (and not finitely define it) by means of Jakobson’s three dimensions in combination with Toury’s more specific description of a source text and a transfer. In this way we are able to reach a broad though still meaningful description of translation as relevant to the field of translation studies which does not exclude, in particular, intralingual translation:

  • A source text exists or has existed at some point in time

  • A transfer has taken place and the target text has been derived from the source text (resulting in a new product in another language, genre or medium), i.e. some kind of relevant similarity exists between the source and the target texts.

  • This relationship can take many forms and by no means rests on the concept of equivalence, but rather on the skopos of the target text.

The above constitutes a description of translation which though not as broad as Steiner’s all-embracing hermeneutic approach still sees the phenomenon of translation as much more fundamental to human communication than more traditional translation studies definitions. However, nothing in the description should be considered necessary or sufficient to define translation. The aim of the description is to function as a tertium comparationis when trying to determine family resemblances.

The above description of translation is very operational as it attempts to describe what falls within the interests of the field of translation studies and as such should be in line with the recommendations of Tymoczko:

The goal can only be an open definition, one that helps in understanding the nature of many translation processes and products, even if not all translation processes and products share a common core of specific features. The definitional impulse in translation research aims at indicating the extension of the concept translation, mapping some borders or boundaries or limits for the inquiry about translation, even if these borders do not form a closed figure.

Tymoczko, 2005, p. 1086

I do not think that Toury himself would necessarily disagree with a broader definition, description or “working hypothesis,” as he prefers, than his own and as mentioned above he does emphasise (1995, p. 32) that it is perfectly acceptable if translation studies scholars work with texts which are not “assumed translations” as long as they explain why the texts in question are relevant. With the above description no such explanation should be necessary when doing research involving intralingual translations. Toury’s greatest contribution to the definition of translation, as I see it, is the fact that the concept of assumed translation (though I reject it) has made us aware that the norms of translation and translation strategies vary enormously from culture to culture from age to age or indeed from person to person.

However, what varies are the preferred strategies, not the basic phenomenon of translation which is not culture-dependent, but fundamental to human communication, and which remains constant at all times.

7. Conclusion

In this article I have discussed the philosophical background to a broad definition of translation and the field of translation studies. Jakobson and Steiner’s hermeneutic approach highlights the fundamental nature of translation, but also makes it clear that an academic discipline needs to delimit its field. Various translation scholars have tried to define translation with the purpose of defining translation studies, but it seems that this has more often than not led to a too restricted view of the field. It goes without saying that for methodological purposes each individual research project needs to define and frame its own particular field, but this does not mean that we have to set up finite boundaries for the entire discipline: “Although no one frame can suffice to illuminate all translations, a frame that is well chosen may illuminate a significant type of translations or a significant facet of the process of translation. Although no absolutes are possible in the case of a cluster concept, it is possible to illuminate characteristics shared by a significant segment of the category (…)” (Tymoczko, 2005, p. 1090).

Primarily inspired by Tymoczko and on the basis of Wittgenstein’s theory of family resemblances and Rosch’s prototypology, I have attempted to define the discipline as an open field which relies on an open, inherently non-finite, yet describable concept and not on necessary or sufficient conditions or audience assumptions. An open definition is important in not excluding phenomena, such as intralingual translation, which by some scholars may be considered peripheral to the field of translation studies, but which I postulate would be able to provide many useful insights as regards translation proper and perhaps vice versa. As Steiner points out, the two kinds of translation raise issues of the same order and are, at crucial points, similar (1975, pp. 260-261). In the practical and especially the didactic world a too narrow definition of the field only sets an artificial boundary for translators and the jobs they see themselves as able to carry out. The professional translator and cultural mediator of today needs a large number of skills “to qualify as an expert for interlingual and intercultural communication” (Snell-Hornby, 1999, p. 164). I would like to think that intralingual skills are included in the term ‘intercultural communication,’ if we do not limit our definition of ‘intercultural’ to mainly a question of national culture (see Zethsen, 2008a, forthcoming).

In Denmark, and presumably in the entire Western world, there is an ever-increasing demand that expert knowledge be made accessible to the general public. Consumers, patients, taxpayers, etc. no longer tolerate incomprehensible expert texts. There is a huge demand for expert-to-layman translation as most experts find it difficult to write about their field in layman terms. In my view translators are excellently equipped to carry out this kind of intralingual translation and also translation studies would benefit from the additional insights provided by studying this translational activity (even if it cannot be termed “translation proper”). I therefore encourage translation scholars to carry out research within the field. What is needed is an empirically founded thorough description of intralingual translation (of all kinds, not only the kind mentioned here), of similarities and differences between intralingual and interlingual translation, of the translational microstrategies typically employed in the two kinds of translation, etc.

Leech writes that people generally agree about what constitutes a prototypical member of a category—though the distinction is not a sharp one—whereas disagreement and uncertainty is common when it comes to establishing peripheral members of a category (1981, p. 84). This explains why intralingual translation has not been central to the field of translation studies, but I hope that in future the focus of research will be slightly less prototypical.