On the Nature and Treatment of Implicit Information in Literary Translation:
A Relevance-Theoretic Perspective[1]
As an instance of human communication, literary translation operates by certain laws and principles assumed to be built into our human make up. These 'natural laws' of communication give rise to implicit information and are responsible for its special characteristics, such as graded strength of communication and its correlates, including poetic effects. They furthermore determine the interdependence of text, context and successful communication, and limit communicability in incompatible contexts. One important contextual factor consists in what kind of interpretive resemblance the audience expects there to be between original and translation. The ultimate test for a translation is whether or not it achieves with the target audience what the translator intended it to achieve, rather than whether it conforms to some translation-theoretical notion of equivalence.
One of the striking characteristics of literary texts is their density. The art of the author often shows itself in the ability to communicate a richness of ideas, feelings and impressions that are not necessarily expressed in words, but communicated implicitly.
For an example take the title of Stendhal's novel La chartreuse de Parme. Adams (1973) shows that these few words in French bring to mind "La Grande Chartreuse, an immense austere monastery high in the French Alps, where one would find Carthusian monks, vowed to asceticism, penitence, and eternal silence" (9). The collocation of chartreuse with the Italian place Parma therefore shocks the French reader - as if someone had chosen to speak in English of "The Great Basilica at Tuscaloosa". The word chartreuse can set off a rich train of thought: taken as an allusion to the "colour chartreuse (which combines the two colors of the liqueur, green and yellow) [it] may have some special significance through contrast with Stendhal's previous novel, Le rouge et le noir; instead of colors which flatly oppose one another, the later novel is drawn in colors which can, and do, combine. It is to be a novel, as it were, of rapprochement. ..."(10).
The problem for the translator is how to convey this rich meaning in a similarly condensed manner:
Now it's no great problem to write a paragraph explaining, ... the implications of "Chartreuse de Parme", but to find three or four English words which will, in combination, produce something like this effect, is the literary translator's overwhelming dilemma ...(10)
Adams' view of the prospects for solving these problems in translation is somewhat gloomy:
... all the choices open to [the translator] are in various ways and for various reasons impossible. The choice is simply between different ways of murdering the original. (10)
It seems worth asking whether things really are as bad as Adams suggests. Is the only decision left to the literary translator in which particular way he wants to "murder the original"?
In this paper I want to show that recent insights from communication theory can help the literary translator to better cope with the problem of implicit information. These insights will be grouped into two sections; the first section deals with the nature of implicit information; the second brings out some general characteristics of translation that have a bearing on the treatment of implicit information.
Building on the work of Paul Grice (1975), Sperber and Wilson (1986) have proposed a relevance-theoretic model of human communication that includes an explanation for the existence and nature of implicit information. They contend that at heart human communication relies on inference. In human communication meaning is not conveyed by the text alone, but crucially relies on the inferential combination of the text with a context. Coding may or may not be part of any given instance of communication. Since a presentation of the theory as a whole would go beyond the scope of this paper, this study will concentrate on those aspects that are directly related to the phenomenon of implicit information.
In order to understand the nature of implicit information, we need to first have a brief look at relevance - the central notion of the theory.
For an utterance to be relevant at all, two conditions must be met: on the one hand, it must provide some new information - we don't usually consider it relevant to be told things we already know. On the other hand relevant information must in some sense "link up" with other information one already has. Information that does not relate to any such given information seems irrelevant.
Relevance theory captures these two conditions by the technical concept of contextual effect: a contextual effect is a change in one's awareness - more technically, in one's 'cognitive environment' - that has been brought about not by the information in the utterance alone, nor by contextual knowledge we already possessed alone, but by the inferential combination of both. To be relevant at all, an utterance must have at least some contextual effects.
Since contextual effects thus go by definition beyond the information expressed in the utterance, they are necessarily implicit. Furthermore, since contextual effects are a prerequisite for relevance, it follows that implicit information is not incidental but inherent in human communication: to be relevant at all, an utterance must convey some implicit information. Thus in human communication implicit information is a prerequisite for relevance, and it is recovered in the search for relevance.
However, for an utterance to be communicated successfully, it is not enough that it be relevant, but rather that it be optimally relevant in the following sense: it should give the audience an adequate amount of contextual effects, without causing it to spend unnecessary effort. Put somewhat loosely, the utterance should modify the audience's knowledge sufficiently without undue effort.
The role played by optimal relevance is captured in the so-called principle of relevance, which states that whenever a person starts to communicate, he automatically communicates the claim that what he has to say is optimally relevant to the audience. It is assumed that this principle is part of our human psychological endowment, and, as we shall see later on, it is the kingpin in the whole communication process.
To put some "flesh" on these rather technical concepts, let us work through a little example.
(1) a) Mother: How do you like your new headmaster?
b) Daughter: I can't stand show-offs!
In this brief exchange, the daughter does not answer her mother's question explicitly; yet her reply suggests rather clearly i) that she does not like the headmaster, and ii) that she considers him a show-off. None of this information is expressed, that is, formulated in language, - so how is it conveyed?
Focusing our investigation on the daughter's answer, its most immediate context is the mother's question. The principle of relevance leads the mother to expect her daughter's reply to yield contextual effects. The question and answer alone do not yield such effects; this suggests that some further contextual information needs to be supplied. The following information appears to be highly accessible. By choosing to talk about show-offs in the context of the mother's question about the new headmaster, the daughter strongly encourages the mother to relate the two ideas. A very obvious relation would be expressed by the following thought:
(2) The new headmaster is a show- off.
If this thought is indeed the contextual information which the daughter intended the mother to supply, then it should lead to contextual effects. It is not difficult to see that this is indeed the case:
(3) a) Contextual assumption: The new headmaster is a show-off.
b) Content expressed: The daughter does not like show-offs.
c) Contextual implication: The daughter does not like the new headmaster.
Thus we see that adding this fairly obvious thought to the immediate context, makes it possible to derive a contextual effect, which in this instance is a contextual implication. Note that this contextual implication - The daughter does not like the new headmaster - provides the answer to the mother's question; this reconfirms that the right, that is the speaker-intended, context has indeed been supplied.
So this example illustrates that the information left implicit in the conversation is recovered in the search for optimal relevance. Contextual implications and chunks of contextual information needed to establish the optimal relevance of an utterance are called implicatures (Sperber and Wilson 1986:193ff). They are part of the information the speaker intended to convey.
Since there are many different ideas as to what "context" is and since implicatures are context-based, let me briefly point out three essential properties of this notion as understood in relevance theory. First, context is a psychological notion; it is a subset of all the information accessible to a person. It is thus a very comprehensive notion, including the surrounding text or co-text as well as any socio-cultural, historical, situational or other kind of information assumed to be available. So it includes what others refer to as "situation", "setting", etc.
Secondly, context is not "given", but is selected (cf. Sperber and Wilson 1986:132-142).
Thirdly, "the selection of a particular context is determined by the search for relevance" (Sperber and Wilson 1986:141); more specifically, context is that set of information which allows the text or utterance to be "optimally processed" (Wilson and Sperber 1986:144).
Thus, while the notion of context in relevance theory is very comprehensive on the one hand, it is delimited and defined in each instance by the criterion of optimal relevance on the other hand.
The feeling that implicit information is part of the intended meaning of the original writer often gives rise to the demand that the translator make implicit information explicit in the translated text if it cannot be conveyed implicitly. Thus Newmark advises:
If one is translating important information which is likely to puzzle the proposed reader, it is better to write the background into the text to make it meaningful rather than as a note. (1988:148)
However, closer examination shows that implicit information has a number of rather special characteristics which the translator should carefully consider before taking such a step. Let us take a brief look at several of these characteristics.
A common suggestion in the literature is that information is left implicit because it is already known and hence redundant. The avoidance of unnecessary statements can certainly be a motivating factor for leaving information implicit. However, there are other factors involved, and these can be more significant to the literary translator.
Imagine the following scene. Harry is just getting acquainted with a nice girl by the name of Caroline. He'd like to suggest to her to see Rambo 17, a movie which has just come to town. However, he does not know whether she'd like that kind of movie and he'd hate to be turned down with his first suggestion. So he might just say:
(4) I hear there's a new movie in town, Rambo 17.
Now by talking to Caroline, Harry communicates the presumption of optimal relevance - that he thinks his utterance will have adequate contextual effects for her. At the same time, he is giving her no indication as to any particular contextual effects he expects her to derive. Caroline might take Harry's remark simply as an indication that he wants to make conversation and that Rambo 17 may be a topic worth talking about. On the other hand, Caroline could take Harry's remark to indicate that he has some interest in this movie. Of course, Caroline could also draw the implication that Harry wants to find out what she thinks about the film, and she might even suspect that he is considering inviting her to this movie.
However, Harry has given her no evidence for believing that he actually intends her to draw any one of these inferences in particular; hence none of these implications is strongly communicated. Whichever of these interpretations Caroline would choose, she would significantly share in the responsibility for deriving that particular meaning. Thus, while the intention to invite her to the movie was within the range of possible interpretations, Harry had given her no evidence to believe that this was the particular thought he wanted to communicate. Hence, if she did not take up that idea, he would not have lost face.
Thus we see that another motivation for the use of implicit information is that it allows the communication of ideas with varying degrees of strength. The communicator can bring to the audience's attention a range of ideas, without necessarily affirming any one of them in particular. Allusions, innuendoes, dropping hints are all devices that exploit this peculiar property of implicit information.
Note that the strength with which implicit information can be communicated covers the full range from very strong to very weak. In example (1), there were two rather strong implicatures: both that the daughter considers the headmaster a show-off and that she does not like him. By expressing herself in this particular way, the daughter gave strong evidence that she intended her mother to derive these implicatures. It is hard to see how she could have been consistent with the principle of relevance without expecting her mother to entertain these particular thoughts. In example (4), the way Harry expressed himself gave no clue to particular thoughts required for optimal relevance. In general, the less evidence the communicator gives that he intended the audience to supply a particular thought, the weaker the communication of that thought.
The fact that implicit information can be communicated with varying degrees of strength is linked to a couple of other significant characteristics. Firstly, the varying degrees of strength with which implicatures can be communicated are closely linked to the range of ideas they can communicate. Recall again example (4) with Harry and Caroline above; there the weakness of the communication resulted from the fact that no single thought could be identified as being necessary for the remark to be consistent with the principle of relevance. Rather, a number of different thoughts were available; hence Harry's remark served to convey a range of thoughts in a weak manner.
Correlating with the strength and range of implicatures is a third feature which is that the degree of responsibility taken by the speaker for that information can vary. For very strong implicatures, the communicator is held fully responsible - as he is for explicit assertions. Thus in example (1), the daughter could hardly deny having communicated the thought 'The new headmaster is a show-off'. However, the weaker the implicatures are, the more responsible the audience becomes for deriving any one of them. In example (4), Harry's remark put the thought he desired to communicate within the range of possible interpretations, without making him fully responsible for it.
These peculiar properties of implicatures - that they can be graded with regard to strength, range, and responsibility - are of crucial importance for what Sperber and Wilson call poetic effects: "Let us give the name poetic effect to the peculiar effect of an utterance which achieves most of its relevance through a wide array of weak implicatures" (1986:222). Metaphors, for example, illustrate the occurrence of poetic effects:
In general, the wider the range of potential implicatures and the greater the hearer's responsibility for constructing them, the more poetic the effect, the more creative the metaphor. ... In the richest and most successful cases, the hearer or reader can go beyond just exploring the immediate context, ..., accessing a wide area of knowledge, ... getting more and more very weak implicatures, with suggestions for still further processing. The result is quite a complex picture, for which the hearer has to take a large part of the responsibility, but the discovery of which has been triggered by the writer. The surprise or beauty of a successful creative metaphor lies in this condensation, in the fact that a single expression ... will determine a wide range of acceptable weak implicatures.(Sperber and Wilson 1986:236f)
For an example, let us look at a passage from the Bible. The text is taken from the Gospel of St. John and describes the scene of the 'Last Supper'. Jesus has just predicted that one of his disciples is going to betray him; Judas has accepted the morsel of bread identifying him as the betrayer. The passage then ends with the following words:
(5) labwn oun to ywmion ekeinos exhlqen euqus; hn de nux.
As soon as Judas had taken the bread he went out. And it was night.
(John 13:30[2])
All we want to focus on here is that last little remark - only three tiny words in the Greek: hn de nux, yet communicating a wealth of thoughts. As Leon Morris (1971) points out,
'Night' is more than a time note. In view of the teaching of this Gospel as a whole it must be held to point us to the strife between light and darkness and to the night, black night, that was in the soul of Judas. (628)
How does such a rich interpretation come about? By choosing these words, John promised his readers adequate contextual effects. However, he refrained from giving them particular guidance as to what these effects might be, thus inviting them to explore and exploit the richness of the cognitive environment he shared with them. Part of that environment would be the recurring references to the concept of 'night' throughout the Gospel. A reflective reader might well be expected to draw on any of those passages and look for new implications they might yield with the current text. For example, the reader might think back to a statement of Jesus in chapter 9:
(6) ercetai nux ote oudeis dunatai ergazesqai.
Night is coming when no one can work. (John 9:4)
Linking that statement to the current passage would suggest that now that time had come; night had fallen on the work of Jesus, as he had expected; it was coming to an end.
The reader's general knowledge would naturally relate the concept of 'night' to that of 'darkness'. He might recall that Jesus had talked about people walking in the dark:
(7) kai o peripatwn en th skotia ouk oiden pou upagei.
The man who walks in the dark does not know where he is going.
(John 12:35)
This might lead to the recognition that Judas did not know where he was going - that he was in fact going to his own destruction.
None of these thoughts are strong implicatures yet they are there, because it seems not at all unreasonable for the author to have expected his readers to derive them. So the reader could go on exploring the richness of this passage, and the more contextual information he would supply, the more responsible he would become for the implicatures he derived.
To summarise this section, we pointed out three characteristics of implicit information:
- implicit information can be communicated with varying degrees of strength
- implicit information can convey a potentially open-ended range of thoughts
- the responsibility for implicit information can be shared between communicator and audience to varying degrees.
In combination these characteristics can lead to poetic richness and beauty.
The three characteristics of implicit information which we just surveyed distinguish it clearly from information communicated explicitly. Such information is strongly communicated, definite and the communicator is fully responsible for it. Since implicit and explicit information differ so significantly, it is likely that the explication of implicit information will change the meaning of the translated text.
A good illustration of these problems is found in Nobuyuki Yuasa's (1987) discussion of a haiku by Basho, a poem typically consisting of seventeen syllables, arranged in three lines of five, seven, and five syllables respectively:
(8) Furuike ya
kawazu tobikomu
mizu no oto[3]
According to the information given by Yuasa, a literal translation would run something like this:
(9) An old pond
A frog jumps in
The sound of water
I shall not attempt to analyse this poem in detail. Rather, what is of interest here are the explications Yuasa uses in his particular translation into English.
The first line of the poem consists of two words furuike ya. Yuasa points out that furuike "... is a compound noun consisting of the adjective furushi meaning 'old' and the noun ike meaning 'pond'" (233). However, Yuasa goes on to explain that this word communicates much more than the idea of an old pond. In fact, it conveys an impression of a whole landscape and also what he calls "the presence of the poet by the pond" (233), presumably in the garden near Basho's cottage. Feeling that it "... was the responsibility of a translator to say more than 'an old pond'", Yuasa chose to add the word 'silence' which "... suggests by implication the presence of a listening ear" (234).
The second word, ya, does not have a lexical equivalent in English. Here it "... cuts the stream of thought for a brief moment, thereby indicating that the poem consists of two thoughts half independent of each other" (234). Yuasa's translation of the two words of the first line then reads:
(10) Breaking the silence of an ancient
pond
Thus he 'explicated' the words 'breaking the silence' in an attempt to bring
out more of the implicated meaning.
In the light of what we just said about the nature of implicit information, these additions can be seen to influence the interpretation of the poem significantly in several respects. Firstly, in the original the idea of 'silence' by the pond appears to have been a fairly weak implicature; the use of the word 'silence' itself in the translation, however, communicates this idea very strongly.
Secondly, the explicit mention of 'silence' sets the reader off along a certain path of interpretation: he will first of all turn to this concept and look for contextual implications arising from the thoughts and ideas associated with it. The original, by contrast, does not guide the reader in this way but leaves him the freedom to look for contextual effects in any direction he might find profitable. As a result, the interpretation of the translation is likely to be more focused, and consequently narrower than that of the original.
Thirdly, without access to the original, the reader of the translation is likely assume that the original author was fully responsible for the thought of 'silence' - that Basho himself wanted to make sure that the reader paid attention to this idea. This, however, is not the case.
Seeing that the explication of implicit information tends to change the original meaning of the text, one might conclude that it is just one way of "murdering the original" (Adams 1973:10) and should, therefore, be avoided. However, matters are more complex than that and I would like to briefly highlight a number of other factors that have a bearing on the matter. Each of these factors would deserve further discussion.
Any act of communication, and that includes translation, is subject to certain 'laws' that are built into our human nature. The most basic 'law' is that the only way by which a thought can be successfully communicated is by consistency with the principle of relevance. It is the only means we have for conveying our own thoughts to others and for discovering their thoughts. Hence no matter what meaning a translator might intend to convey by his translation, he can only expect it to be successful to the degree that his meaning is consistent with the principle of relevance for his audience.
This is a fundamental consideration, and yet one that is often overlooked, especially when dealing with implicit information. A translator may make his translation very explicit, presenting the information to the target audience in an easily digestible manner - but may still remain generally unsuccessful simply because the intended meaning itself is not felt to be optimally relevant to the target language audience. The cause could be that the intended meaning lacks on the benefit side - it simply does not yield sufficient contextual effects; alternatively the translation could fail on the cost side: the amount of effort required for processing all that information may simply be more than is warranted by the resulting contextual effects. The translation could obviously also suffer from the combined effect of shortcomings in both areas.
The central point here is: making it all clear and plain does not necessarily guarantee a successful understanding of the translated text. The crucial factor is whether or not the target audience can arrive at the intended meaning through consistency with the principle of relevance.
The second major point is that relevance is necessarily context-dependent. Consequently consistency with the principle of relevance does not exist in the abstract but is always relative to some cognitive environment or context. This context dependence of communication is often recognised, but its consequences are not often seen very clearly. The particular consequence that interests us here is that a message - or some part of it - that was communicable in one context may simply not be communicable in another. Note that this may not be a matter of how that message is presented but of the message itself: unless that message turns out to have adequate contextual effects in the cognitive environment of the target language audience, it cannot be communicated, regardless of how it is expressed, whether, for example, information is explicated or left implicit.
This has far reaching implications for the translator. It means, for example, that even before deciding to translate a text, a translator will do well to assess whether the text he intends to translate is likely to be communicable - in whole or in part - , given the cognitive environment of the target audience. Many seemingly inextricable 'translation problems' are, in fact, problem of communicability and could be anticipated ahead of time by a comparison of the cognitive environments of the original and target audiences respectively.
While much has been written on the issue of translatability, I am not aware that this issue of communicability has been given much attention. In fact, a very common assumption seems to be that the main problem of the translator has to do with language: how to find the equivalent ways of expressing the meaning of the original in the target language. However, as relevance theory helps us to realise, that is only one part of the problem; mismatches in contextual knowledge can prove just as serious problems as mismatches in linguistic resources.
The third major point is that translation is a kind of reported speech or, in relevance-theoretic terms, an instance of the interpretive use of language, as I have attempted to show in my book Translation and Relevance (Gutt 1991). Two points are of special significance here. First, in interpretive use, a text is presented in virtue of its interpretive resemblance to another text. Second, interpretive resemblance is defined in terms of shared thoughts, or more specifically in the case of translation, of shared explicatures and implicatures.
To begin with the second point, the fact that interpretive resemblance consists in the sharing of explicatures and implicatures, makes it possible to arrive at detailed and concrete comparisons of translation and original, without any need to resort to the use of general categories or terms, such as 'literal' or 'free'.
The significance of the first point is that the foundational relation between a translation and its source text is not one of equivalence, but of interpretive resemblance. On the one hand this entails freedom for the translator: there simply is no fixed, translation-theoretic norm of equivalence that he needs to fulfil in order to produce a "good" translation. The openendedness of the notion of resemblance gives new space to move especially for texts that for one reason or another seem to be outside the reach of equivalence.
On the other hand, it is important for the translator to realise that in interpretive use generally the principle of relevance has the effect of leading to an expectation of faithfulness in the following sense: any text that is presented as an instance of interpretive use automatically purports that it resembles the original text closely enough in relevant respects. This means that for a translation to communicate successfully it is essential that it lives up to this claim. The seemingly vague parameters "closely enough resemblance" and "in relevant respects" are determined by the expectations of the audience. This recognition brings to the fore that the success of a translated text is crucially dependent on the expectations of the target audience.
This is the fourth major factor which the translator needs to consider: the expectations the target audience has of the target language text. These expectations are, in fact, part of the context which the target language brings to the text, and they are crucial for the success or failure of the communication act as a whole. Whether or not the translator will succeed in communicating what he intends to get across will significantly depend on what the audience thinks is being attempted in the translation. Given that there is no general consensus on any clearly defined notion of translation and that there are as many notions of translation as there are readers of translations, this seems to be an area where the translator better be as explicit as possible.
Applied to the area of implicit information as such, the translator needs to carefully consider what the expectation of his target audience are in this respect, whether the audience would prefer the explication of implicit information even if it alters the meaning of the original, or whether it would prefer it left implicit, even if this meant considerably more effort for recovering the meaning. It is at this point that notions, conventions, norms, trends etc. of translation play their role.
This brings us to the fifth and last point - one which I have already tacitly assumed above: This is that the central concern of the translator is not to achieve some fixed standard of 'equivalence', but rather successful communication. While the communication-oriented approach to translation is not new - Nida's book Towards a Science of Translating (1964) being one of the key works in this development - those earlier translation theories were impaired by the inadequacies of the semiotic or "code model" of communication (cf. Sperber and Wilson 1986:4ff) which they were explicitly or tacitly built on. (For an evaluation see Gutt 1991:66-94.) Sperber and Wilson's relevance theory advanced our understanding of human communication considerably by showing the crucial roles plaid by inference and relevance. In their framework, for an act of communication to be successful means "... to have the communicator's informative intention recognised by the audience" (Sperber and Wilson 1986:161). Applied to translation as an instance of interlingual, interpretive use of language, this means that a translation communicates successfully when the target audience recognises what the translator intended to communicate, and unsuccessfully when it fails to do so.
But if all the translator needs to be concerned about is getting the target audience to recognise whatever he intended to communicate of the original - is there no constraint on what he as a translator rather than, for example, a commentator, should attempt to communicate? Does not this mean that all norms of translation are abandoned and that the translator can communicate whatever he pleases?
The answer is yes and no. On the one hand, the translator is free to do what he likes in his translation. At least as far as any general matters of translation theory are concerned, there are no fixed norms - such as equivalence - that he needs to meet. On the other hand, the translator has to be aware of the causal interaction of text and context; as we just mentioned above, he is not producing the text in a vacuum, but for some audience whose context, including its expectations of what a translation should communicate, will causally affect the success or otherwise of his translated text. Seeing that the term "translation" means different things to different people, spelling out his intentions in an introduction to the translated text - rather than relying on the label "translation" alone - may reduce the risk of the translated text getting misinterpreted.
Being explicit about his intention does not, however, necessarily mean that the audience is going to share the translator's views, especially if his views clash with some socially accepted norms. For example, if the audience wants a translation that leaves implicit information implicit, it will appreciate being warned in the introduction that the translator chose to explicate implicit information extensively. At the same time, however, there is no guarantee that the audience will therefore abandon its preference for a less explicit translation; it may turn to a translation that fulfils its expectations better. In that case, the translator on his part may want to look for a different audience.
In conclusion, let me point out that relevance theory predicts an interesting ambivalence concerning the notion of successful communication. As I mentioned above, the audience has no criterion other than consistency with the principle of relevance for identifying what a communicator intended to convey. This predicts the possibility of the following situation. Reading a text, the audience arrives at an interpretation consistent with the principle of relevance - that is, an interpretation felt to yield adequate contextual effects without unnecessary effort. The audience will therefore assume that it has arrived at the speaker-intended interpretation of the text, and that successful communication has taken place. Yet it is thinkable that, due to mismatches in the contextual information used, the interpretation arrived at by the audience may not be the one the communicator had in mind. It just happened to be consistent with the principle of relevance. This explains how - sometimes very extensive - misinterpretations can occur and even go unnoticed, both in intralingual and interlingual communication.
The problem is particularly likely to remain undetected in translation situations where the target language audience has no access to the original. Indeed, it seems to be not uncommon, that a target audience fails to perceive altogether that a translated text is, in fact, a translation. It may read it as a target language original and feel that it did fulfil its expectation of optimal relevance. Depending on the translator's informative intention, this may or may not disconcert him: if he intended the target language text to be recognised as his translation of the original, he may feel frustrated since all credit will go to the original author, whose name appears on the front cover, rather than to him as translator. On the other hand, he may console himself concluding that his translated text did not contain "translationese" renderings that gave it away as a translation - which is no mean achievement either.
Adams, Robert M. 1973. Proteus, His Lies, His Truth: Discussions of Literary Translation. New York: Norton.
Grice, H. Paul. 1975. 'Logic and Conversation'. Reprinted in A. P. Martinich (ed.), The Philosophy of Language. Oxford:Oxford University Press, pp. 159-70.
Gutt, Ernst-August. 1991. Translation and Relevance: Cognition and Context. Oxford: Blackwell.
Morris, Leon. 1971. The Gospel According to John. Grand Rapids: Eerdmans.
Newmark, Peter. 1988. Approaches to Translation. Hemel Hempstead: Prentice Hall.
Nida, Eugene
A. 1964. Toward a Science of
Translating: with Special Reference to Principles and Procedures Involved in
Bible Translating. Leiden:
Brill.
Sperber, Dan, and Deirdre Wilson. 1986. Relevance: Communication and Cognition. Oxford: Blackwell.
Yuasa, Nobuyuki. 1987. "'The sound of water': Different versions of a hokku by Bashô". William Radice, and Barbara Reynolds, (eds.) The Translator's Art: Essays in Honour of Betty Radice. Harmondsworth: Penguin, 1987. pp. 231-40.
[1] Paper prepared for the Symposium on the Pragmatics of Literary Translation at the Åbo Akademi University, Åbo/Turku 10-12.11, 1992.
[2] In this and the following quotations from the Gospel of John, the Greek text is taken from Kurt Aland, Matthew Black, Carlo M. Martini, Bruce M. Metzger, and Allen Wikgren (eds) The Greek New Testament. 3rd edition, Stuttgart, United Bible Societies, 1975. The translation is: Holy Bible. New International Version, New York International Bible Society; published by Hodder and Stoughton, London 1979
[3] From the anthology Haru no Hi, 'Spring Days'. See Yuasa 1987, p. 231.