Priorities and Restrictions in Translation * Patrick Zabalbeascoa 1 1. A review of priorities and restrictions In this paper I am presenting a performance model 2 of translation for the main purpose of raising awareness of the complexity of the production of translation in trainees (and hopefully, purely intuitive translators as well). I intend to rescue the word priority as used by Nida, at the risk of being regarded as a hijacker, since I also intend to update it by trying to benefit from the insights of functional models, norm theory and descriptive translation studies (DTS). From this revival of the term priority, within a didactic/ evaluative paradigm, I will then provide a personal account of subjectivity, scope, ambition and difficulty, proposing them as evaluative criteria, which is the other primary aim of the paper. I will deal with equivalence as a (variable) characteristic of priorities, although I am fully aware that this is not the only possible account of equivalence, even within this paradigm 3. I assume that the objects translators work with and produce are texts and the constituent elements of texts are verbal and nonverbal signs. Nida and Taber 4 (1969: 14) make the following claim: As a basis for judging what should be done in specific instances of translating, it is essential to establish certain fundamental sets of priorities. Nida defends a system of four priorities; more precisely, they are four pairs of norms (as we now know with the benefit of hindsight) where one norm of each pair is given priority * Dedicated to the memory of André Lefevere. 1 I would like to thank Anna Espunya for her useful remarks. 2 From the translator s point of view, it is a tool for justification hopefully, a contribution towards the justification procedures as proposed by Toury (1995: 37); for the critic and reviewer it can be seen as an evaluation aid (unlike Toury s twoway path which couples justification procedures with discovery procedures, addressing himself first and foremost to researchers and theorists rather than critics). 3 For example in Chesterman (1997: 69): One kind of relation might of course be equivalence [...] of some sort. One translation task might require a translation which gave priority to a close formal similarity to the original [...] Another might prioritize stylistic similarity [...] yet another might higlight the importance of semantic closeness [...] yet another might value similarity of effect above all [...]. 4 Henceforth to be referred to as Nida, for short.
Priorities and Restrictions in Translation over the other. For example, Nida s priority number 4 is forms that are used by and acceptable to the intended [target text] audience have priority over traditionally more prestigious forms. Nida was clearly being prescriptive in his proposal (as was Newmark later in picking it up and turning it, slightly changed, into one of the two methods he presented as his own contribution to the field), and it is interesting to note that his book is included in a series called Helps for Translators. Strictly speaking, it is forcing things a little to regard his model as being applied since he had no awareness of applying a truly theoretical model: nothing like skopos theory, norm theory or descriptive studies was well-known or accepted at the time. Rather, his approach was to make theoretical claims. Here lies, to my mind, the huge contribution that Holmes (1988) 5 makes in his call for a more disciplined approach to the study of translation, especially the initial division between pure (nonutilitarian) studies and the applied extensions. Toury (1995) makes a very good case of explaining the relationship that can be made between pure studies and its applied extensions, and also the distance between applied models and actual practice. An exact contemporary of Nida, Levý (1969: chapter 1), proposed a functional hierarchy that would determine the relative ranking of importance of various aspects of word meaning for the translator to use as a criterion 6. In relation to Nida s words quoted above the reader is left somewhat mystified as to how to interpret priority and who is to establish priorities. I think the idea can be pedagogically interesting if we regard the priorities of translation not as universal or constant (as in to be established by Nida ), but as variable and context-sensitive, to be re-established for each new assignment (by those who are responsible for the translation). In the latter approach, TT priorities are goals and objectives that are intended for the product of translation, as established by the translator 7. For Chesterman (1997: 68) different translation tasks may require balancing acts between different priorities, a point that I overlook here, but in this paper the point is picked up and given centre-stage. In broad terms, restrictions, in this paper, refers to the constraining nature of the translator s circumstances, or conditions 8. Thus, factors such as 5 Especially The Name and Nature of Translation Studies in Holmes (1988). The perceived need for this article to be better known can be seen in the time and effort that several scholars have spent quoting and explaining its contents, most notably Toury. Holmes combines scientific discipline with readability so admirably that it seems almost ridiculous that no one had been capable of coming up with that sort of common sense before. 6 For a full discussion of Levý s hierarchical model, see Gutt (1991). 7 Regardless of whether s/he is (more or less reluctantly) taking on board the priorities of others. 8 See Chesterman (1997: 54) for a distinction between conditions and norms, the 160
Patrick Zabalbeascoa deadline for the translation, choice of translator, (insufficient) revision and copy-editing, structural differences between source text (ST) and target text (TT) languages and other systemic differences, as well as technological limitations and some forms of censorship, are more easily formulated as restrictions than as priorities. It is interesting to note that even a norm theorist of the stature of Toury resorts just as often to the notion and term constraints restrictions as to norms. One of the best examples of what I am trying to show here can be found in Toury (1995), especially in section 3 of chapter 10, Using Revisions to Uncover Constraints ; however, he frequently refers to the translator as restricting himself to certain characteristics intended for the TT that can be conceived of as hierarchically structured. This is what I would prefer to call the translator s priorities, so as to be able to distinguish between priority-oriented restrictions and situations like restricting oneself to such things as not asking anybody else for help, typing with one hand, or finishing the job in 320 minutes flat, for instance. He actually writes that A certain aspiration for adequacy should [...] be added to our list of constraints. This [...] would rank hierarchically lower than most of the constraints listed (1995: 204; italics in original). Chesterman (1997: 78), like Toury, also refers to restrictions in his account of norms: Lefevere distinguishes five constraints which determine the way translators [...] manipulate texts. I think the advantage of referring to constraints is that it is an easier basis for comparing relevant statements of all of those translation scholars who have an awareness of restrictive forces in translation, even if they are unaware of or belong to an earlier time than norm theory and DTS. 2. A new proposal for priorities From a text producer s (e.g. ST writer, speaker, translator) angle, priorities are the formal and functional characteristics that the text will have once it is finished, or, from the user s point of view (especially a professional user/evaluator 9 such as a scholar, editor, or critic who can be expected to see beneath the surface ), priorities are the characteristics that a text is interpreted as having its explicit and implicit aspects. What makes a translation stand out from other forms of text production (that might also be considered as priority-driven) is that its priorities will necessarily reflect the way in which one text is to be regarded as a version 10 of another already existing text, important point being that conditions are a component of a definition of norms according to Bartsch (1987: 76). 9 I.e. what Chesterman (1997: 66) calls a norm-authority. 10 The relation norm, in Chesterman (1997: 69). 161
Priorities and Restrictions in Translation referred to as the ST. I believe that this approach can make sense even in the extreme case of pseudotranslation, since the ST priorities are supposedly reflected in the pseudotranslation, so that speculations as to what the ST would have been like will tend to coincide in many respects. Thus, comparing translations with their STs can be done at the surface level of the solutions and forms actually resorted to, or at a more abstract level of goals and intentions, i.e. the rationale, or priorities, that give rise to the solutions. The comparison may help, if not to appreciate, at least to understand them better. Evaluation of the TT against its ST can be done at either of these levels, or the actual solutions ( surface manifestation) can be evaluated according to their compliance with the text s own priorities. A set of priorities will not necessarily include all of those factors that norm theorists would call norms, only those that the translator establishes (regardless of awareness) as intended characteristics for the text. For example, a norm that forbids footnotes is not considered a priority here because, in any case, the inclusion of footnotes would be a resource or tactic, not a priority in the sense of a goal or objective per se for the TT. In this example, the noninclusion of footnotes, due to banning or any other reason, would be a restriction rather than a priority. Any priorities the translator may have in the larger context of his/her life (such as responsibilities and ambitions other than the translation at hand) are not priorities in the sense used here; they are restrictions if they are shown to hinder the achievement of the best possible result by the best possible means. Neither should what Toury (1995: 58) includes in the concept of preliminary norms be considered as priorities in those cases where they are completely beyond the translator s control. Such norms are also to be considered as restrictions (of a sociocultural dimension), which may explain the absence/presence of certain textual characteristics, or account for the translation or some aspect of it (including challenge the prevalent norm as a possible priority). In this scheme of things, a translation is the result although a single ideal result is not envisaged of an interaction of a hierarchical 11 set of goals for the TT, i.e. its Priorities (or P set). It is assumed that each priority (henceforth P) is conditioned by restrictive circumstances and constraints. The success of a translator for a given task can be measured (e.g. by some normauthority or other) according to the number of conditioning factors that were 12 11 Used in the same sense as in Toury and Levý, although Levý is prescriptive and Toury is not, of course. 12 Assumptions and questions with the verb in the past tense apply for evaluation or discovery of a rationale behind the solutions as they appear in an existing TT. The past verb tenses can be switched to present and future tense and even some modal verbs in certain cases when the model is used by a translator during the 162
Patrick Zabalbeascoa identified and taken into consideration; and according to how each individual factor was dealt with and accounted for within a global pattern, i.e. a specific set of priorities (Ps) for each new translation. When no solution can be found that will simultaneously account for all of the goals initially established for the TT, the translator can decide which characteristics have priority over which others. In other words, P sets are organized as hierarchies, as an operational principle, in order to justify and account for what is lost, gained, omitted, added, substituted, and compensated for. Each P implies the potential existence of its opposite (e.g. P = hilarious effect implies potential P = dead serious effect ). No two Ps appearing in the same P set can be each other s opposites, e.g. P m = metaphorical effect and P n = non-metaphorical effect cannot appear in the same P set. Although two contradictory Ps cannot be included in the same P set, they may appear in different sets (i.e. for a different translating task, even if it is of the same ST). So, a P that is incompatible with one set (or with many P sets), is not necessarily incompatible with another, especially if the textual and/or contextual circumstances seem to favour its presence. Even P = be informative is not universal. For example, when translating nonsense poetry, or certain kinds of advertising jingles (where the top P is get the public to buy the product ) or jokes (where the top P is laughter-eliciting ), the translator may not be interested in the ST information content at all, but will aim at conveying other aspects which address why the ST is being translated (conveying effects and impressions rather than information). As a theoretical principle (momentarily considering translation outside its social reality), translators are free to decide what Ps to include and their hierarchical arrangement. But in the real world many of a translator s decisions are predefined, coming from various sources of prescription (editors, censors, laws, academic institutions, etc.), which often materialize as the translation order, the style book, etc., especially in the case of staff translators; or they will be determined by other contextual factors. In such cases, the translators only job is to remember at all times what the Ps are, and do their best to fulfil (be faithful to) them. From the point of view of evaluation or criticism, translators cannot be held entirely responsible for an unjustifiable P set if it was predefined by someone of authority and imposed upon them. Ps that are insensitive, indifferent (see P parameters below), to equivalence, giving rise to casual equivalence or difference are not specific to the domain of translation 13 and belong to the principles of such areas process. 13 See Chesterman (1997: 69): Neither the accountability norm nor the communication norm are specific to the translation process. 163
Priorities and Restrictions in Translation as: (a) prescriptions for good writing (grammaticality, referential accuracy, clarity, etc.); (b) principles of logical thinking, coherence and rhetoric; (c) norms of social behaviour and interaction (cooperative principles): politeness, censorship; (d) technical considerations, as in the case of translation that requires some form of technological support; (e) principles of sound business practice. ST and TT priority sets may not coincide entirely, because TT priorities do not always mirror ST priorities. There may be differences in their nature (they may not be the same ones), number (there may be more or less), or rank. Scope and subjectivity (see evaluative criteria below) are two concepts which derive from this observation. 3. The variables of translation Variation in translation can be mapped, among other means, by using the concepts of Priorities (Ps) and Restrictions (Rs). The acceptance of variation does not compromise standards of excellence (which are in fact fixed sets of Ps and Rs). Different strategies and solution types 14 that may be used to fulfil the Ps and overcome the Rs also help to account for different renderings. From this standpoint, unforeseen solutions might be fully justifiable. ST-TT relationships may vary considerably in nature, degree and number from case to case. This means that decisions are made as to how and why the two texts will be different as well as similar. Full awareness of this kind of decision-making 15 is also a means of understanding, often controlling, what is lost and/or gained according to the direction the translation takes. Before looking into how and how well a given text was translated it is important to know why it was translated (e.g. as opposed to writing a piece of original work, translating a different text, using the ST for something other than translation), what was translated (as opposed to what was not translated), what for, who for and a number of questions to do with the viability conditions of the project. Contextualization variables of the (existing or prospective) TT include: the commission for the translation, the translator s identity and 14 I prefer to use strategies to refer to behavioural patterns that are not renderings, such as reading strategies, documentation strategies, etc.; solution types, rather than techniques, refers to the common features of a specific set of solutions; solutions is a nonevaluative term for renderings, the actual textual make-up of the translation, and it stresses the problem-solving, decision-making, nature of translation from the translator s perspective. 15 As stated in Toury (1995: 86): decision-making and the factors which may have constrained it. 164
Patrick Zabalbeascoa environment, assumed expectations, function(s) and context of the TT. In short, knowing the criteria for selecting the ST and the viability conditions, justification, functions and systemic position of the TT. Variables such as deadline, financing, technical or material limitations, as well as ST-TT differences at a number of levels (e.g. language structures, assumed knowledge and moral values of text users, and intertextuality) are Rs (i.e. constituents of the R set) that cannot be adequately classified as Ps. For example, if the purpose of the translation is to make money, then that can be regarded as a P; presumably there will be evidence of this in the text. But if the client s intention is simply to pay the lowest possible price for the job, that is not a P, it is an R that may explain why certain Ps are not satisfactorily reflected in the TT. Different norms will produce different results. Standards of acceptability for text production in the TT environment also change, including the possibility that translations might be judged by different standards from original texts. A translator works under prescriptive pressure from various sources, and these may be more or less obvious. One way of broadly classifying the variables of translating is the following: (1) the factors of the people involved in translation in any way, playing a professional or commercial role, a political or social one, an individual or collective one, e.g. as patrons, translators, editors, text users, critics, etc.; (2) the contextual factors of the ST, of the TT and of the translating process, e.g. the translator s working environment, means and materials; (3) language and communication factors; (4) tools and norms for text analysis and production. 4. The parameters of restrictions Rs are the circumstances that prevent an absolute, universal, constant type of relationship between the ST and its TT (a one-to-one relationship) from being anything but wishful (or prescriptive) thinking. The nature of the Rs actually present has a limiting effect on the number and the nature of the Ps; in turn, the Ps determine the relevance of the Rs, thus creating a tension between the two. Different types of texts and assignments involve different sets of Rs. Some Rs are very common in many different situations, such as the difference in assumed knowledge of the target users of the ST and TT. In the model presented here each R may be characterized according to the following parameters: source, force, level, and range (although the list is not intended as definitive). The parameter of source is for pinpointing the origin and explanation of the presence of the Rs, e.g. whether they come from textual features of the ST 165
Priorities and Restrictions in Translation and the TT (textual Rs), or from their contexts and the conditions of the translating process (contextual Rs). Examples of textual restrictions: ST weaknesses, such as (assumed) omissions, errors, unintended ambiguity and vagueness. Complex semantic and discursive networks in the ST, resulting in several possible interpretations of individual segments of text and of the text as a whole (symbolism, metaphor, irony, parody, etc.). Relevant use of formal or structural aspects of language and communication in the ST. Advertisements and literary texts, especially poetry, provide good examples of this point. TT genre and channel of discourse involving strict rules for text formats and presentation, or substantially different ones from the ST. Examples: subtitled versions of audiovisual texts, summary translation, contracts, diplomas and certificates. Differences in the intertextual relationships of the ST and the TT, such as when there is a lack of parallel texts for the TT (one reason would be that the literary genre the ST is associated with has no tradition in the target community). Examples of contextual restrictions: Differences between ST and TT intended text users (readers, listeners, viewers ) in any aspect related to group or community membership (language and communication systems, sociohistorical factors and traditions, publishing policies, etc.). Low social status of translation; a lack of appreciation for the usefulness and difficulties involved in translating. Insufficient investment in the field and financing of projects. Professional Rs, such as when the employer does not provide enough time, means (i.e. materials, references, choice of translator and other persons to do the job), or incentive (pay, job satisfaction, acknowledgement from others); or the client imposes restrictive conditions for the TT: e.g. total amount of space provided for the text, style book prescriptions, lack of clear criteria, underestimating the value and/or complexity of translation (by, for instance, imposing unrealistic demands). Limited translator competence: inadequate or insufficient training, lack of experience, excessive subjectivity, among others. Generally speaking: (a) a poor appreciation of the importance of any of the textual and contextual factors, or of available alternatives; (b) inability to find satisfactory solutions 166
Patrick Zabalbeascoa according to the goals or Ps and Rs if that is the approach despite an awareness of the problems that are present and the type of solution required. translator competence is closely related to the notion of untranslatability, so it becomes necessary to point out for whom, and according to which criteria, certain texts or textual elements are untranslatable. A luckier or more skilful translator with a greater awareness of potential alternatives (i.e. greater competence), might come up with an unforeseen satisfactory result for the task at hand. Not all Rs are necessarily of equal force. This parameter, unlike source, is one of degree, e.g. on a scale of restrictive force we might speak of three degree labels: strong, weak, and nil (an R is branded as irrelevant when it hardly matters whether it is present or not since it does not affect the actual P set; in this sense sexual taboos are usually irrelevant when translating operating instructions for a coffee grinder). It is important to point out that each P is competing, as it were, with the other Ps. If their structure is hierarchical then we can say that a P is restricted, among other circumstances, by the force of higher-ranking Ps. Level and range specifications are useful since not all of the segments and features of the text are necessarily affected by the R. Rs may appear as some form of systemic or textual differences on various levels that might be labelled (the categorization and naming of labels throughout will depend on the practical needs of the moment and the sophistication of the translator or scholar) as lexical, morphological, metaphorical, lip-synchrony (in the case of dubbing), rhyming structures, etc. Range specifies which parts of the text if not all of it, globally are affected locally by each R. 5. The reversal of restrictions The absence (or noticeable weakening) of a given R during the translating process cancels out its effect (R moves towards nil on the force scale), which brings about a series of favourable circumstances to be exploited by the translator. I call this phenomenon restrictions reversed. Here are some examples. 1) Specific characteristics of the ST, or certain aspects of it, may either point very clearly to the most adequate kind of solution, or otherwise provide the translator with a greater degree of choice; e.g. STs with little semantic or pragmatic ambiguity. 2) A translator may resort to a number of P-oriented tactics or techniques not disallowed by the R set, e.g. (a) textual tactics, either verbal (adaptation, 167
Priorities and Restrictions in Translation compensation, paraphrase, deletion, modulation, increased redundancy), or nonverbal (pictures, gestures, mimicry, sounds); (b) paratextual tactics such as footnotes and glossaries; (c) extratextual tactics such as warnings or justifications appearing independently from the text. 3) Many of the translator s problems are proportionally reduced (the Rs become weaker or disappear) the smaller the (contextual) differences between the ST and TT languages, cultural contexts, social groups, historical periods, etc. 4) The potential dangers of misinterpretation and knowledge gaps may disappear on some levels if the translator also happens to be the initiator, the ST author or if s/he has easy access to either/both of them, or if s/he feels intense empathy towards the writer. Likewise, the translator may have an excellent knowledge of relevant aspects of the TT s intended end users. 5) The better the translator is (and the better suited to the task at hand) the weaker the Rs related to personal and human limitations (the translator, not being unique, perfect or infallible, is a variable to be included in the R set). Factors of translator competence are to be found in the domains of formal training, experience, general and specialized knowledge/skills, reliability, etc. 6) The physical and material conditions of translation are also Rs, therefore, more time, more means, more incentives, and more cooperation all help to weaken the respective Rs with a view to producing a better translation. This is often a matter of money, patience and/or willpower. 6. Establishing the priorities of a translation Many, if not all, of (at least) the language- and culture-specific features of the ST usually disappear or need to be changed. Are translators aware at all times of exactly which features are going to disappear and why? Do they act consistently and coherently when applying techniques for retaining those features that (according to the acceptability norms of the moment) cannot afford to be left out of the TT? The aim here is to show the importance of being able to establish context-sensitive Ps, for the simple reason that they are not necessarily the same from one translation to the next. An alternative to resorting to a specific intuition (or knack) for translation is an awareness of one s decisions (or rather, the grounds on which they are made), and their consequences, in order to be in a position to justify what one has done. Ploughing through a mental checklist and working hard on textual analysis and word-hunting is a useful fallback strategy for those times when inspiration is 168
Patrick Zabalbeascoa not forthcoming. This is even the preferred method during the learning process and training period when the translator s automatisms are still to be developed. From the translator s position, establishing Ps includes understanding and interpreting the ST, most probably by means of a detailed translationoriented textual analysis. An in-depth study of a text-as-st includes knowing the reasons for the choice of the text to be translated. A P and R approach stresses the benefits of setting clear, coherent, realistic targets for the TT. The targets are the final arrangement of the Ps once the Rs have been identified and taken into account. It is important to emphasize that certain Ps will only be feasible if certain Rs are absent or very weak; further, each P is restricted by any other Ps of higher rank 16. An experienced translator gets through this (P and R establishing) stage very quickly in nonchallenging assignments that have a lot in common with previous work. ST analysis and interpretation, often called the decoding stage of the translating process, may include (especially if it is translation-oriented) a twofold process of: (1) interpreting the text and identifying a P set in it; (2) assessing the quality, communicative effectiveness and authority of the ST (e.g. measuring the ST writer s success in achieving intended goals to consider the relative importance of the textual elements actually used). Answers to questions regarding the need(s), purpose(s) and goal(s) of a translation appear in the P set as global Ps, global 17 indicating that these Ps apply throughout the text as a whole. This initial contextualization develops into a contrastive analysis to discover as many relevant factors as possible regarding ST and TT perception (including interpretation) and production (and even distribution). This will serve as a broad base for establishing possible ST-TT relationships and for building up a detailed account in terms of Ps and Rs. This is where applied linguistics and the application of models and discoveries from other disciplines may contribute most notably. 7. The parameters of priorities TT priorities can be characterized according to the following parameters: source, rank, range, function, form and equivalence. Source is the same as for the Rs, but in particular the question is, who decided it is to be a P (e.g. translator, client, other)? To what extent are these direct sources role-players in the translating process agents of more distant or less tangible sources (academic, political, religious institutions, business interests and so on)? 16 A concept that already appeared in references to Toury and Levý (above) and that is further presented and defined in the next section. 17 A more detailed explanation of global and local is provided below. 169
Priorities and Restrictions in Translation Rank 18 is the importance of each P with respect to the other members of the P set: e.g. a hierarchical set scaling down from top to nil passing first through high and low, such that top and high are mandatory, low is desirable, not mandatory, and nil means a potential P does not belong to the P set for the TT in question. High Ps (strong requirements/prohibitions) are either aspects/items that must appear in the TT, or aspects/items that must not appear or must be avoided. From among the high Ps, one may sometimes stand out as the top P. Low Ps (weak requirements/prohibitions) are aspects that should appear in the TT whenever possible, or aspects that should be avoided as much as possible. Possible here means as long as all higher Ps are accounted for as well. The parameters of level and range are essentially the same as for the R set, except that level is broken down here into function, form and equivalence. Range: Ps and Rs can be seen as either global or local, depending on whether or not they hold their rank (or force) throughout the text. Global Ps are operative throughout the text. Typical global Ps will have to do with questions of text types, discourse and overall intention. Local Ps are of limited range, operative only over a given segment of text; there is a temporary re-shuffling of the P hierarchy, in answer to a specific R that is equally limited (local) in its range. The functional level of a P (informative, aesthetic, expressive, didactic, humoristic, etc.) accounts for any aspect or item in terms of its internal textual function (cohesive, exemplifying, concluding, etc.) or its contextual function (interpersonal, pragmatic, social, etc.). The level of form considers formal textual elements (such as repetitions, nonverbal constituents, layout, delivery in recited texts) as motivated in certain places, usually subordinated to some function or other, which must then be identified. Equivalence a concept that has come under a considerable amount of attack from some theorists 19 although it has too strong an intuitive base to be lightly discarded can be seen as a label that may be used to further qualify certain Ps for a given task. We will not consider any type of equivalence as a P but will use the concepts of intended equivalence, intended difference (nonequivalence), and indifferent to equivalence as potential characteristics of any P. Rather than speak of equivalent effect, for instance, as a potential P, we will say that once it is decided that X (e.g. X = comic effect) is a member of the P set, X is then labelled yes, no or indifferent for equivalence: whereby yes = intended equivalence (degree of similarity would then have to be specified), this means that X is a P for the TT precisely because it was a P for the ST as 18 Rank shows how priorities are competing forces, whereas the forces of the restrictions have an accumulative effect. 19 See e.g. Snell-Hornby (1988). 170
Patrick Zabalbeascoa well; no = intended difference, meaning that X is a P for the TT even though it was not a P for the ST; indifferent (equivalence not considered), meaning that X is P for the TT regardless of whether it was or was not a P for the ST. As for the actual surface end-result (more or less accurate representations of the Ps), TT renderings can be described as instances of intentional equivalence (or difference), casual equivalence (for indifferent), or unjustifiable equivalence for unaccountable or unaccounted solutions, or ones contrary to the assumed P set. Examples of areas for establishing priorities 1) Information load. (a) Information items: quantity and type of information. (b) The balance of explicit and implicit information. (c) The relative importance of information and content with regard to other aspects of the text, i.e. its rank. (d) Relevance of the information to the intended text user. 2) Decoding and text processing. How important is it that the text be immediately understandable, and unambiguous in its intention and/or in its meaning (e.g. road signs vs metaphysical poetry)? 3) Text-user response. Is there an intended single specific response or a free personal one? Is the response meant to be immediate or delayed? What is the degree of coaxing and hoaxing on the part of the text producer? 4) Aesthetic elements (related to text-user response). Is the text, or any part of it, art for art s sake, or are the aesthetic elements mostly instrumental in achieving other Ps? 5) Formal aspects. Genre, discourse, register, text-type membership traits. What kind of formal aspects are involved (rhyme, repetitions, total number of words, letter size, use of colours, etc.) and how are they related to the text s functions? 8. Evaluative criteria Ps and Rs help to justify or account for: (a) the various strategies resorted to in the process and the solutions (renderings) used in the resulting version; (b) TTs as texts that respond to a variety of needs and goals dependent on the circumstances of their production and reception. Hatim and Mason (1990: 160) say that translators may not intrude arbitrarily or inadvertently. I would like to interpret Hatim and Mason s use of arbitrarily as meaning unjustifiably, which is usually what happens when the translator is not fully aware of all of the factors involved. We will thus rephrase 171
Priorities and Restrictions in Translation this statement by saying that there is a difference between unjustified manipulation and P-oriented manipulation. 172
Patrick Zabalbeascoa TT ambition and scope Ambition counts the number of Ps for a given translation, regardless of equivalence labels. An example of limited ambition might be the characteristic P set of summary or gist translation. On the other end of the ambition scale we find TTs that are rich in their expressiveness and show many-layered meanings and effects (some of which may be of intended equivalence). A TT is more ambitious the higher the number and rank of its Ps. The scope of a TT results from adding its ambition to the fulfilment of the ambition, i.e. the degree to which the Ps are clearly rendered and accounted for in the TT solutions. Ambition can be used as a measure of difficulty, although it need not always be a yardstick for quality, since a TT may be considered too ambitious. Subjectivity Subjectivity measures mismatches in the hierarchical orders of the P sets of the ST and TT. A TT becomes more subjective as the number and importance of the mismatches grow (the importance of mismatches can be seen by looking at differences on the rank scale), becoming more objective the greater the coincidence, giving rise to more Ps and renderings ticking off yes for equivalence. A TT is more subjective the greater the differences with regard to its ST, both in the respective ranking position of each P and the solutions used, regardless of whether the differences were intentional or unintentional (casual). A translator s intention(s) may be quite different from the ST author s. This can be made to surface, on the scale of subjectivity, by identifying Ps in the TT that were absent in the ST and vice versa, and/or differences in the rank of shared Ps. As in the case of ambition there is no (nonprescriptive) automatic relation between subjectivity and quality. Difficulty Individual Ps taken separately have at least as many solutions available as any combination of Ps in a given P set, i.e. more often than not, the greater the number of Ps, the smaller the number of possible solutions that can do justice to all of them to the same degree, especially if the Ps are of very different natures. Hence, we might say that difficulty is increased. A translation becomes more difficult as the Ps increase in number, especially those of higher rank, i.e. the more ambitious the TT is. The greater the number and force of the Rs, the more difficult it will be for the translator to fully satisfy expectations. 173
Priorities and Restrictions in Translation For example, if the top priority is to communicate the most salient pieces of information included in the ST and all of the other Ps are low, then the task will be simpler and more straightforward than if the translator decides that the P set includes five high Ps such as these: (a) one complete syntactic unit per line (labelled indifferent for equivalence); (b) a strict rhyme scheme (labelled yes for equivalence); (c) exactly eleven syllables in every line (labelled no for equivalence); (d) include certain metaphors (labelled yes for equivalence); (e) adapt the text so that it can be sung (yes for equivalence) according to a certain tune (no for equivalence). Each one pushes the translator towards a different ( best ) solution, and only in a few places will most translators be capable of finding a compromise solution, and even then it might well be the result of happy coincidence rather than some carefully planned method. Quality A translation is better, the better it fits in with the criteria of its evaluation, which are variable. It all depends on the kind of TT that we want to (or are able to) produce. Evaluative criteria may (or may not) include ambition, scope, and subjectivity; we can even evaluate the translator s choice of Ps. In comparing two versions of the same ST, we might say that one is better than the other because it has greater scope, or, from a different angle, that it has more satisfactorily fulfilled all or most of its objectives regardless of whether the other translation had set for itself more Ps (was more ambitious). A TT may be regarded as worse than another if it has a higher number of unjustifiable solutions. 9. Concluding remarks Ambition and subjectivity scales are proposed as an alternative to evaluating translations on a scale with literal at one end and free at the other. I think it is important to realize that ambition and subjectivity are nonevaluative as concepts. One becomes normative only when one fixes what degree of ambition one expects, or how much subjectivity one is willing to tolerate. Other than that, the most normative aspect of my proposal is that it demands that a translator be able to justify his or her solutions in terms of intended goals and objectives in a given set of circumstances. Hopefully, this means that subjectivity and ambition can be applied to descriptions (e.g. for comparative purposes) as well as evaluations. 174
Patrick Zabalbeascoa Several traditional concepts of translation theory (e.g. method, equivalence, faithfulness) can be redefined in terms of Ps. Literal translation is often ambiguous because it may be regarded as a P, a method, or a type of solution. The P and R model regards so-called translation methods as labels for a group of Ps that are recurrent or prescribed in different sets. Thus, for the interlinear method of translation, the high global P = refrain from rearranging the order of the words as much as possible, especially from one line to the next, even at the expense of target language naturalness. In this case naturalness is a low P. For the communicative method, the high global P = to produce on its readers an effect as close as possible to that obtained on the readers of the original (Newmark 1988: 39), with lexical equivalence at the bottom of the hierarchy. From a nonprescriptive point of view there is no universally valid method. The appropriateness of a method will always depend on which factors affect the process and its product. So, more than a theoretical method-label, what translators and critics need is help in raising their awareness so that they can recognize and assess the relative value of the factors that are operative in each case. It is unfair to evaluate a translation without considering all of the intervening factors, including translators motivations, goals, constraints, and the sociohistorical circumstances. So, I believe that critics and researchers can benefit from applying this kind of approach to the versions they have chosen to analyse. I also believe that much would be gained by the development of more stylebooks or similar sorts of guidelines spelling out as specifically as possible the Ps and Rs of certain types of texts for specialized translators. These stylebooks could include a list of available and forbidden solutions and suggestions for finding the necessary sources of reference, such as specialized databases, dictionaries and parallel texts. References Bartsch, Renate 1987 Norms of Language, London: Longman. Chesterman, Andrew 1997 Memes of Translation, Amsterdam: John Benjamins. Gutt, Ernst 1991 Translation and Relevance, Oxford: Blackwell. Hatim, Basil and Ian Mason 175
Priorities and Restrictions in Translation 1990 Discourse and the Translator, Harlow: Longman. Hewson, Lance and Jacky Martin 1991 Redefining Translation. The Variational Approach, London: Routledge. Holmes, James S. 1988 The Name and Nature of Translation Studies., in R. Van den Broeck (Ed.), Translated! Papers on Literary Translation and Translation Studies, Amsterdam: Rodopi, pp.67-79. Lefevere, André 1992 Translation, Rewriting, and the Manipulation of Literary Fame, London: Routledge. Levý, Jiří 1969 Die literarische Übersetzung: Theorie einer Kunstgattung, Frankfurt: Athenäum. Newmark, Peter 1988 A Textbook of Translation, London: Prentice-Hall. Nida, Eugene 1964 Towards a Science of Translating, Leiden: E.J. Brill. Nida, Eugene and Charles Taber 1982 (1969) The Theory and Practice of Translation, Leiden: E.J. Brill. Nord, Christiane 1991 Text Analysis in Translation, Amsterdam: Rodopi. Snell-Hornby, Mary 1988 Translation Studies: An Integrated Approach, Amsterdam: John Benjamins. Toury, Gideon 1995 Translation Studies and Beyond, Amsterdam: John Benjamins. Zabalbeascoa, Patrick 1995 Levels of Prescriptiveness in Translation, in Ian Mason and Christine Pagnoulle (Eds.), Cross-Words, Liège: University of Liège, pp.23-29. 176
Patrick Zabalbeascoa 1996 Translating Jokes for Dubbed Television Situation Comedies, The Translator 2:2, pp.239-258. 1997 Dubbing and the Nonverbal Dimension of Translation, in Fernando Poyatos (Ed.), Nonverbal Communication in Translation: New Perspectives and Challenges in Literature, Interpretation and the Media. Amsterdam / Philadelphia: John Benjamins, pp.327-342. 177