Requirements for an " Ideal " Bilingual L 1 → L 2 Translation-Oriented Dictionary

The major aim of this article is to outline the requirements for an "ideal" bilingual L1 →L2 dictionary of the general vocabulary specifically designed for the purposes of professional translation. The article challenges three commonly accepted beliefs: (a) a bilingual dictionary equals a translation dictionary; (b) a bilingual dictionary is a source of immediately insertable lexical equivalents of lemmata; and (c) a bilingual dictionary solely furnishes semantic-pragmatic equivalents, whereas a monolingual dictionary always defines meanings of lexical items. It has been claimed that bilingual lexicography should be based on a clear-cut conception of the future reference work specified in terms of lexicographic parameters: (a) the intended user group; (b) the purpose of the dictionary; etc. An "ideal" bilingual L1 →L2 translation-oriented dictionary should be a reference work that is intended to serve the purpose of text production in L2. In the situation of professional translation, L2 text production is subject to two types of constraints: (a) constraints superimposed by the receptor language and culture; and (b) constraints superimposed by the source text. If constraints of the second type cannot be, in principle, envisaged, those of the first type can, and should be accounted for in a bilingual dictionary designed for the professional translator. The article specifies some of the requirements for such a reference work.


Igor Burkhanov
L2-teksproduksie onderhewig aan twee soorte beperkings: (a) beperkings opgelê deur die teikentaal en -kultuur; en (b) beperkings opgelê deur die bronteks geskryf in L1.As beperkings van die tweede soort in beginsel nie voorsien kan word nie, kan en behoort dié van die eerste soort verantwoord te word in 'n tweetalige woordeboek wat vir die professionele vertaler beplan is.Die artikel spesifiseer 'n aantal vereistes van so 'n naslaanwerk.This article forms part of a large-scale project that aims to outline the basic properties of "ideal" reference works intended to serve the purposes of translation and translator training, i.e. the intended user of those reference materials is either a person professionally involved in the process of mediated bilingual communication or a trainee translator.

Sleutelwoorde:
This major objective can be achieved by attaining smaller goals, such as: (a) to specify the notion of professional translation; (b) to highlight the notion of equivalence in translation theory and lexicography; (c) to clarify the correlation between metalexicographic categories "bilingual dictionary" and "translation dictionary"; (d) to evaluate the state of the art in bilingual lexicography with respect to the purposes of professional translation; and (e) to spell out innovations that can be introduced in bilingual dictionaries in order to adjust them for the needs of the translator-user.These matters will be discussed in the appropriate sections of this article.
Pertinent questions will be asked, though not fully answered, in this article.It primarily aims to bring problems to the fore and illustrate the way in which translation-oriented dictionary-making should be developing.We shall mostly concentrate on translation-oriented bilingual dictionaries of the general vocabulary; the issues of terminological lexicography are beyond the scope of this article.
Before we proceed with these matters, the function of the adjective "ideal" in the context of the present article needs to be clarified.As I will try to show, reference works specifically produced for the purposes of professional translation are yet to be designed.The very term "ideal" translation-oriented dictionary" is intended to emphasize this idea.
Firstly, the adjective "ideal" is used here to show that the existing bilingual reference works do not totally satisfy the needs of a practicing translator.Secondly, the considerations below are not meant to be regarded as a description of a particular lexicographic project or projects, but rather as an attempt to outline at least certain basic principles of translation lexicography.Thirdly, a given dictionary, whether existing or intended to be compiled, is, from this viewpoint, an approximation to the non-existent ideal as specified by the require-ments.Fourthly, we are not going to evaluate the feasibility of compiling translation-oriented bilingual dictionaries of a given pair of languages.Such an endeavour will require detailed analysis of the state of the art in bilingual (and monolingual) lexicography in both linguistic communities.

The Notion of Professional Translation and Interlingual Equivalence in Translation Theory and Lexicography
A good deal of inconsistency in metalexicographic considerations concerning bilingual lexicography can be attributed to the variety of interpretations of the notion "translation".In both theoretical publications and introductions to actual dictionaries, it is not always clear what kind of translation the bilingual reference work is supposed to serve, since the notion of translation is erroneously believed to be self-explanatory.
It should be mentioned that professional translation differs from other activities that are also loosely referred to as "translation", primarily: (a) "school translation", i.e. finding native-language equivalents of foreign-language lexical items in order to facilitate acquisition of the latter; and (b) "translation for reading" which implies translating parts of a text so that those parts can be read more easily when needed (for a detail analysis see Gile 1991: 188).Thus, the delimiting adjective "translation-oriented" in terms such as "translation-oriented lexicography", "translation-oriented dictionary", etc., is intended to emphasize that the reference works under consideration should be suited for the purposes of professional translation.In our further considerations, unless specified, the term "translation" will be used in reference to professional translation only.
It should also be emphasized that the word "translation" may be used in at least two senses: (a) as an umbrella term designating mediated bilingual communication both in the spoken and written mode; and (b) in the narrower sense referring to translation in the written mode only, cf.translation vs. interpreting (e.g.see Gutknecht 2001).Naturally, bilingual dictionaries are extensively implemented in the process of translating in the second sense, i.e. mediated bilingual communication in the written mode.Conversely, it is difficult to make use of them in the course of consecutive or simultaneous interpreting, for interpreters usually do not have time to consult reference works.
From this it follows that bilingual lexicography primarily serves the purposes of professional translation in the written mode of communication.As for interpreting, reference works of this type, or any other reference materials for that matter, can be used solely in interpreter training.
A matter of primary significance for the issues under consideration is to highlight major aspects of professional translation.This category presupposes a specific form of communication in which there are several primary participants: (a) Communicant1, i.e. a person (or a group of people) who uses the source language (L1) and functions within the cultural setting of the appropri-ate linguistic community; (b) Communicant2, i.e. a person (or a group of people) using the target language (L2), i.e. a different ethnic language that functions within another cultural background; (c) Communicant3, the translator, a bilingual and bicultural person whose job is to negotiate linguistic and cultural discrepancies in the course of bilingual communication. 1  The point to be noted here is that the basic assumption of lexicographers engaged in the production of bilingual dictionaries is that there is at least one obligatory L2 lexical equivalent of the L1 lexical item.It should be realized that this assumption is perfectly in line with the linguistically and/or semiotically oriented models of mediated bilingual communication that view the process of translating as encoding and decoding, i.e. a purely linguistic operation of substituting L1 linguistic signs, particularly lexical items, with corresponding L2 units (e.g.Catford 1965: 20; for a detailed analysis of these models see Burkhanov 2003a: 36-41).One of the consequences of this approach is the assumption of translatability, i.e. inter-replaceability of those signs with no loss of meaning (see, e.g.Petrilli 2001: 278).In practical terms, it means that the target text in L2 can be easily "back-translated" into L1 and the resulting text will be identical to the original.Obviously this assumption contradicts the hard reality of professional translation.Those models, though attractive due to their simplicity, do not account for all the aspects of the multifarious process of translating.
Though numerous and heterogeneous, sometimes even contradictory in nature, more recent translation theories view the process of translating as manipulation of the source text to adjust it to the requirements of the target culture, a specific form of human action, a communicative process which takes place within a given social context, an act of intercultural communication, etc. (see Gentzler 1993, Venutti 2000, Burkhanov 2003a: 41-55).In those theories the major emphasis is put not so much on the semantic-pragmatic content of lexical units, but on the inferencing of utterance meaning on the basis of contextual information, on the specific constraints as imposed by the discourse type, register, etc., as well as on other multifarious culture-specific aspects of communicative interaction.
In view of those developments in the theoretical study of translation, quite a number of experts have come to the conclusion that the very concept of translation equivalence is expendable, if not misleading (e.g.see Snell-Hornby 1988).Simultaneously, those theoretical solutions are, in turn, vulnerable to criticism, since in most cases they fail to draw a demarcation line between translation proper and adaptation.The starting point of our further considerations is that the category of equivalence is indispensable in both translation studies and bilingual lexicography, but it should not be understood as a homogeneous interlingual correspondence between lexical items belonging to L1 and L2.
It is important to note that the notion of equivalence as a major category of translation studies has evolved for the last fifty years.Firstly, various kinds of translation equivalence have been distinguished.It is noteworthy that the term "equivalence" is usually provided with the delimiting adjective: "formal", "dynamic", "semantic", "pragmatic", "communicative", "functional", etc. From this it follows that an L1 linguistic expression may be a semantic equivalent of an L2 linguistic expression, but not pragmatically equivalent to the latter.
Secondly, it has been observed that translation equivalence varies in degree.For instance, according to Kade (1968), four types of equivalence can be distinguished: (a) total equivalence characteristic for standardized specialized terminology; (b) facultative equivalence that presupposes a one-to-many correspondence between L1 and L2 lexical units; (c) approximating equivalence that accounts for one-to-part-of-one correspondence; and (d) zero equivalence observable in the case of culture-bound lexical items.
Thirdly, and no less importantly, we may note the impact of correlated categories, such as "unit of translation", "translation adequacy" and "translational norms", which not only contributed to the development of translation studies, but also delimited the scope of application of the notion "equivalence" (for details see Burkhanov 2003a: 91-114).
A unit of translation is a functional-semantic category that is not restricted to the linear boundaries of linguistic structures and involves macro-textual parameters.So, the unit of translation cannot possibly be equated to an interlingual correlation describable in terms of one-to-one correspondences between L1 and L2 lexical items; it is a dynamic conceptual configuration spread over a large stretch of the text (Newmark 1988;Rabadán 1991).Simultaneously, the very notion of translation adequacy, as an optimal level of mediated bilingual communication in given circumstances of verbal interaction, implies selecting the most appropriate equivalents out of a variety of linguistic choices in the course of translating (Reiss and Vermeer 1984;Shveitser 1993).Moreover, translation strategies are dependent on historically-motivated cultural, literary and translational norms of the receptor culture (Toury 1980;Heylen 1993).
As for the notion of equivalence in lexicographic theory, Piotrowski (1994: 104) rightly pointed out that, though there are a large number of publications on bilingual lexicography, the notion of bilingual equivalence has not been fully specified yet.The greater majority of metalexicographers and practicing lexicographers assume that this basic category of bilingual lexicography is selfexplanatory, or superficially formulate it in terms of translation equivalence understood as cross-linguistic correlations between lexical items, or ignore the problem altogether.One of a few attempts to fill the gap was undertaken by Wiegand (2003 and earlier publications).Nevertheless, though very interesting, his discussion tends to be limited to only one aspect of the issue under consideration, namely the lexicographic presentation of L1 and L2 lexical items, whereas their semantic-pragmatic equivalence is supposed to be determined at the earlier stages of lexicographic description.
The foregoing brief exposition has demonstrated that there is a considerable gap between the interpretation of equivalence, not to mention similar categories, in translation studies and what can be found in bilingual lexicogra-phy.Bringing the concept of equivalent in bilingual lexicography into conformity with that of translation equivalent as developed in contemporary translation theory will only be profitable for this branch of lexicographic activity.Ideally, the optimal translation-oriented bilingual dictionary should be applicable in the translator's attempts to produce an adequate target text that conforms to the requirements of translational norms, which implies facilitating his/her search of appropriate units of translation.

Bilingual Dictionary and Translation Dictionary: General Considerations
Our major goal in this section is to clarify the notion of bilingual dictionary and juxtapose it with that of translation-oriented dictionary.It is no secret that for the majority of experts in metalexicography those two terms are interchangeable (e.g.see Hartmann and James 1998: 146, Hannay 2003, Swanepoel 2003: 67-69).As has been mentioned, this line of reasoning presupposes that a monolingual dictionary provides definitions, whereas a bilingual dictionary furnishes interlingual equivalents, i.e. for each L1 lexical unit that may be found in the word list as a lemma there is an L2 lexical equivalent (or equivalents) presented in the right-hand side of the dictionary entry.We may say that the longstanding tradition of bilingual lexicography has been based on the assumption that for each and every L1 lexical item there is at least one most appropriate L2 lexical equivalent, and the characteristic features of bilingual reference works are specified in contrast with those of monolingual ones (cf Zgusta 1984: 147): The dictionary [bilingual dictionary -I.B.] should offer not explanatory paraphrases or definitions, but real lexical units of the target language which, when inserted into the context, produce a smooth translation.This is a perfectly natural requirement.
It should be noted that this statement not only contradicts the aforementioned major postulates of contemporary translation theory.Even more importantly, the ever-increasing vast number of reference works intended to meet the needs of different user groups do not fit into the neat classificatory framework presented above.This state of affairs has not escaped the attention of experts in lexicographic theory.Nevertheless, in metalexicography those reference works that display the features of both supposedly basic types are usually regarded as either hybrid works of reference or some sort of deviation from the general principle.For instance, Hartmann and James (1998) distinguish, in addition to a bilingual dictionary and a monolingual dictionary, another kind of lexicographic product: a bilingualised dictionary.A bilingualised dictionary is a reference work which is "based on a monolingual dictionary whose entries have been translated in full or in part into another language" (Hartmann and James 1998: 14).
On the basis of the definitions of pertinent notions provided in a dictionary of lexicographic terminology (Bergenholtz et al. 1997) and his own considerations, Tarp (2002) specifies the following three types of dictionaries: (a) a monolingual dictionary -a dictionary with only one object language, i.e. one language described; (b) a monolingual dictionary with a bilingual dimension -a dictionary that has one object language and another description language; and (c) a bilingual dictionary -a dictionary that has two object languages and provides equivalents in the target language for each word and expression in the source language.Obviously "a monolingual dictionary with a bilingual dimension" in the trichotomous division presented by Tarp corresponds to what Hartmann and James refer to as "a bilingualised dictionary".
Those and similar terminological distinctions have been introduced in order to save the aforementioned commonly accepted dichotomy: a monolingual dictionary furnishes definitions, i.e. defines, specifies, explicates meanings of lexical items, whereas a bilingual dictionary provides L2 lexical items characterized by semantic-pragmatic equivalence to the L1 lexical items represented by the lemmata.References to linguistic and cultural anisomorphism (discrepancies between a pair of languages arising due to their semantic, pragmatic, grammatical and cultural differences), which leads to the fact that "translation equivalents are typically partial, approximative, non-literal and asymmetrical (rather than full, direct, word-for-word and bidirectional)" (Hartmann and James 1998: 51), are usually made to pay the tribute to practicing lexicographers' efforts and to illustrate the difficulties they face rather than to draw conclusions that have far-reaching metalexicographic consequences and may lead to constructive solutions in bilingual lexicography.In the following, I will tentatively suggest a different approach that appears to be more realistic and takes into consideration the actual state of the art in dictionary-making.
First and foremost, the monolingual dictionary definition does not always specify the meaning of the entry word in the sense of providing its paraphrase or a sentential explication.One method of lexicographic explanation is based on intralingual lexical-semantic equivalence and consists in enumerating lexical items with similar meanings in the definiens, primarily synonymous equivalents and opposites in negation, particularly those that came to be known as complementaries or non-gradable antonyms.Cf.: It is evident that a definition on the basis of intralingual equivalence does not explicate the meaning of a lexical item, as much as periphrastic and/or explanatory definitions do, and in this respect is not very much different from a typical bilingual dictionary definition based on interlingual lexical semanticpragmatic equivalence.Cf.: legible = разборчивый; четкий; удобочитаемый (о почерке, шрифте и т.п.) (NERD).
The use of synonyms and/or antonyms in negation in monolingual works of reference obviously leads to circularity of definitions and has been severely criticized by experts in metalexicography.Nevertheless, practicing lexicographers know that sometimes it is unavoidable.It is particularly the case when the lexicographer does not make use of explanatory, sentential definitions, and periphrastic definitions on the basis of a hyperonym cannot be provided due to the specificity of the lexical-semantic content of lemmata.Moreover, if from the viewpoint of logic the defining techniques based on intralingual equivalence are commendable, there are purely applied-linguistic advantages in this way of defining.For instance, they afford the possibility to account for the lexical-semantic relations in the lexicon, particularly synonymy and oppositeness in meaning, which may be considered advantageous with regard to particular lexicographic functions 3 ; hence most appropriate for certain types of lexicographic reference works.
Simultaneously, contrary to popular belief, there are bilingual dictionaries that furnish detailed definitions of the source-language lemmata in the target language.Very often those definitions are not only periphrastic, but also of an explanatory character.By way of example we may cite the German-Russian dictionary of synonyms (DRSW) where the discrimination of synonyms and usage specifications are provided in L2, i.e. in Russian.Cf.: Peinlich неприятный тем, что связан с тягостным ощущением неловкости, смущения и т. п. [unpleasant, because it is associated with a painful feeling of being in an awkward, embarrassing situation, etc.]. 4   Meaning specifications of L1 synonyms in L2, in addition to the listing of their L2 equivalents, may be said to be a characteristic feature of the English-Russian dictionary of synonyms (ERDS).Each dictionary article in that reference work provides an explanatory definition in Russian intended to specify the elements of meaning shared by all the synonyms, and a detailed explication of each member of the synonym series with meaning discrimination as well as notes that account for the senses of the synonyms at issue which are close to the meaning under consideration and comments concerning their figurative senses (for details see Burkhanov 2003b).In fact, those lexicographic works were produced in accordance with Ščerba's original idea that a learner-oriented bilingual dictionary should provide definitions of the foreign language lexical items in the learner's native language.
In addition, it should be emphasized that in case of a lexical gap, i.e. the situation when there is no even approximate L2 equivalent of an L1 lexical item whose meaning is often of a culture-specific nature, lexicographers unavoidably have to use a short interlingual paraphrase in a bilingual dictionary.Cf.: casebook = журнал для записи больных, клиентов, посетителей и т.п.[a book in which patients, clients, visitors, etc. are noted] (NERD); buńczuk = (oznaka wladzy [symbol of power]) horse-tail ensign (GPED).
Though defining in L2 is often accompanied by furnishing even approximate L2 semantic-pragmatic equivalents (if there are any), the very existence of the technique of lexicographic specification exemplified above is tell-tale evidence that all bilingual dictionaries not only provide interlingual equivalents, but may, and do define.
The aforementioned lexicographic facts warrant a different approach in typology of reference works in metalexicography.Instead of a binary opposition "monolingual dictionary" vs. "bilingual (or multilingual) dictionary" with supposedly opposing purposes, the differences between the actual reference works can be described in terms of two distinctive features or lexicographic parameters: "number of languages" and "defining technique". 5 The lexicographic parameter "number of languages" distinguishes reference works in which only one language is implemented (monolingual) and those with more than one language (multilingual), the bilingual dictionary being a kind of multilingual work of reference.Thus, the terms "monolingual", "multilingual", "bilingual", etc. within this framework characterize only the number of languages that can be found in the reference work under consideration and are devoid of any connotations concerning its purpose.
The parameter "defining technique" refers to the actual way of specification of lexical meaning irrespective of the language (or languages for that matter) used in the right-hand side of the dictionary.Various types of lexicographic definitions (with regard to monolingual dictionaries, of course) have been discussed in many sources (e.g.see Hanks 1987;Nakamoto 1998;Geeraerts 2003).
To keep this article within manageable proportions, we are not going to consider these matters here 6 .Suffice it to say that, in principle, a definition of any type (or a mixed one) can be implemented in both monolingual and bilingual reference works.By way of example let us consider the basic types: (a) equivalence definition (i) in a monolingual dictionary (on the basis of intralingual equivalence): invent vt … 1 create or design (sth not existing before) … 2 make up, think of … (ALDCE3); (ii) in a bilingual dictionary (on the basis of interlingual equivalence): legible = czytelny (GEPD); (b) periphrastic definition (i) in a monolingual dictionary: coat = an outer garment with sleeves (NHLD); (ii) in a bilingual dictionary: exterminator = środek tępiący (robactwo itd.) [a substance used to exterminate (insects, etc.)] (GEPD); (c) ostensive definition (i) in a monolingual dictionary: cutlery = knives, forks and spoons, used for eating and serving food (ALDCE6); (ii) in a bilingual dictionary: nephew = bratanek; siestrzeniec ['one's brother's son'; one's sister's son] (GEPD); (d) descriptive definition (i) in a monolingual dictionary: muggy.If the weather is muggy, it is unpleasantly warm and damp, and makes you feel uncomfortable or tired (COBUILD); (ii) in a bilingual dictionary: Eid присяга по сравнению с Schwur подчеркивает публичный, официальный характер клятвы … [oath, compared to Schwur, emphasizes the public, formal character of the oath] (DRSW).
The choice of an appropriate combination of the number of languages and defining techniques for a particular reference work should be primarily determined by the intended user and lexicographic function or functions, for instance: (a) to provide potential candidates for translation equivalents (production dictionary for translation purposes); (b) to ensure understanding of the target text (reception dictionary for reading purposes); (c) to facilitate the acquisition of a non-native language; etc.For instance, in the case of two foreign languages studied by students, a possible solution can be a trilingual reference work where there are two object languages and the students' native language is used to explicate similarities and differences between them, thus performing the function of a metalanguage of lexicographic description (Burkhanov 1993).
Though by no means comprehensive, the foregoing discussion has provided sufficient background for the specification of at least certain requirements for translation-oriented bilingual dictionaries of the general vocabulary, which will be our concern in the next section.

Prospective Innovations in Translation Lexicography
A number of experts are very much aware of the fact that the existing bilingual L1 →L2 reference works do not completely satisfy the needs of professional translation and/or are not sufficient for adequate translational performance.
For instance, Tarp (2002) approaches the problem from a different perspective and argues that monolingual dictionaries and bilingual dictionaries "the other way around", i.e.L2 →L1 dictionaries, can better solve certain problems arising in the course of translating from L1 into L2.
The cornerstone of the following considerations is that an "ideal" L1 →L2 translation-oriented dictionary is always designed for text production.A brief and by no means exhaustive account of the requirements for a reference work of this kind presented below follows from this assumption.
First and foremost, even the slightest discrepancies in the semantic and/or pragmatic aspects of the source-language lexical item and its target-language equivalent or equivalents should be explicitly specified in the dictionary entry.Since all the cases of semantic-pragmatic lexical equivalence of a facultative, approximating or zero character may cause translation errors, they deserve the translator's attention and have to be described in the dictionary entry.Conversely, total equivalence does not have to be specifically noted, since, due to long-standing lexicographic convention, the lack of any commentary, usage note and/or lexicographic indicator of whatever kind will be correctly interpreted by the intended user.The translation-oriented bilingual dictionary will thus furnish exhaustive information concerning semantic-pragmatic discrepancies between interlingual equivalents.
It is evident that this matter is closely interrelated with the issue of lexicographic representation of pragmatic, culture-specific and encyclopaedic information in a bilingual dictionary of the general vocabulary, which has already been touched on by practicing lexicographers (e.g.Sundström 1996).
This goal can be achieved by the unrestricted use of periphrastic and/or explanatory definitions in L2 of the conceptual content and usage of L1 lexical units.It should be emphasized that this technique of lexicographic presentation is undoubtedly underused in contemporary bilingual dictionaries, which may be attributed to the aforementioned generally-accepted belief that lexicographic works of this kind are supposed to provide lexical equivalents only.Thus, paraphrases and explanatory definitions in L2 should not be "smuggled" into a translation-oriented bilingual dictionary, but treated as an efficient technique of explicating the semantic-pragmatic content of L1 lexical items.
Nowadays the translator is supposed to be intuitively aware of the potential chances of meaning that may arise in the course of translating (particularly unavoidable "overtranslation" and "undertranslation"), if the recommended interlingual equivalents are implemented.It is obvious that lexicographic specification of potential complications that may appear in the process of transposition of the source text into another language and culture will be only appreciated by practicing translators.
To cite but one example: The aforementioned Polish lexical items bratanek 'one's brother's son' and siostrzeniec 'one's sister's son' have the same English translation equivalent.The unavoidable use of nephew as an only equivalent of any of those two words may lead to undertranslation in the course of rendering a Polish text into English.The translator can and should compensate for the deficiency in the other parts of the target text.From this viewpoint, a useful innovation in a Polish-English translation-oriented dictionary can be specifications of such instances of partial equivalence.
Simultaneously, information of this kind is indispensable in the course of translating from English into Polish.The fact that the L1 lexical item nephew should be accounted for by one of its two "interlingual co-hyponyms", i.e. either bratanek or siostrzeniec, warrants the necessity to check the whole text and/or other available materials in order not to commit a translation error.A good solution in cases of approximating equivalence of this kind can be the use of lexicographic indicators like WARNING or ATTENTION intended to highlight those interlingual lexical-semantic discrepancies.It is important to note that an extensive lexicographic treatment of such cases is not required in a bilingual L1 →L2 reception-oriented dictionary designed for reading L1 texts.
If the available translational solutions may cause undue domestication of the target text, the translator can resort to borrowing.The use of this technique of translating has far-reaching lexicographic consequences.Among others, it warrants the necessity to suggest transliteration in a translation dictionary when the L1 linguistic community uses a different alphabet (or the same alphabet, but enriched by letters that cannot be found in L2 writing) or a different writing system.This is particularly the case when culture-specific terms are accounted for.The already mentioned Polish word buńczuk may serve as an example.
Firstly, the short interlingual paraphrase "horse-tail ensign" provided in a dictionary does not seem to be sufficient.In fact, the lexical item at issue was used in two senses; and, to serve the purposes of professional translation, the dictionary should provide explications of both.One sense designated a horse tail used as a decoration on the helmet or girdle.The second denoted a wooden shaft with a ball on top and a cross-bar with horse tails attached to it.It was used as a symbol of authority vested with hetmans, Polish commanders-inchief, Cossack leaders (cf.Russian бунчук), Turks (cf.Turkish bunğuk), Kazakhs, and Tatars.
At first sight it seems irrelevant if, in the course of translating into English a Polish historical novel or a museum booklet, the word will be borrowed in its original spelling or will be transliterated in accordance with the rules of English orthography as bunchuk.Nevertheless, adapted spelling seems to be preferable in this case, since, as has been shown above, the same culture-specific term may be used in the source texts written in other languages.
Generally speaking, it seems appealing to suggest that the more potential lexical-semantic equivalents are provided either in the entry or by means of cross-referencing (ideally including interlingual hyperonyms and/or paraphrases, and suggested loanwords), the better the reference work will serve its purpose.Naturally, in view of the foregoing considerations, none of those solu-tions may prove to be applicable in a particular instance of translating.Nevertheless, with a larger range of L2 equivalents there is a better chance of furnishing a more adequate lexical item for a particular context, not to mention the possibility of leading the translator to the right choice by association.
For instance, Birkenhauer and Birkenhauer (1989: 91) maintain that literary translators primarily need monolingual dictionaries with onomasiological arrangement of their word list.Cf.: A dictionary that would present its material in our ideal way -as a network of interrelated meanings, ranging from true synonyms to mere neighbouring concepts that just share one or two semes -would be invaluable for translators (and other professional writers).For only in such a network does the translator have the chance of hitting upon potentially new equivalents that can do more justice to the phonological, syntactic, and semantic constraints of his original text.
It is difficult to contradict the statement that the translator's activity is similar to that of an L2 writer.Nevertheless, there is a very significant difference, namely: the original writer has to comply with only those constraints, which are imposed by the linguistic peculiarities of L2 and its cultural setting.As for the translator, his/her linguistic choices are additionally subject to the impact of the constraints superimposed by the source text.
In view of the considerations above, Snell-Hornby's proposals seem very promising (Snell-Hornby 1990: 222-224).She believes that the lexicographer should aim at the specification of lexical items both paradigmatically (intralingually) and contrastively (interlingually), i.e. both against other members of the appropriate lexical fields in the source language and in contrast with lexical equivalents of the target language.In her opinion, the major aim of translation lexicography is to design a contrastive dictionary of synonyms.
Moreover, I would go on to claim that interlingual semantic-pragmatic equivalence of lexical items is not sufficient to produce an adequate translation.From the standpoint of semantic typology of languages, it is clear that an L1 lexical item can be translationally equivalent to a linguistic sign of a different kind: a morpheme, a functional word, a syntactic structure, or a linguistic expression of a complex nature.For instance, the Russian functional word ведь and its Polish equivalent przecież are pragmatically equivalent to a tag question with falling intonation in English (for details see Burkhanov 1997 and2002).
Another instance of interlingual semantic-pragmatic correspondences can be exemplified by the designation of the notions 'DYING' and 'KILLING'.These two concepts are lexicalized in English and practically all European languages, cf.die vs. kill.In the Turkic languages, the notions at issue are designated by the opposition of two word forms of the same lexeme, cf.ölmäk vs. öldürmäk (Azerbaijani); ölmek vs. öldürmek (Turkish).Bilingual dictionaries have not been very efficient in the lexicographic representation of semantic-prag-matic equivalence of units of various symbolic levels, i.e. lexical items, morphemes and syntactic structures.
Translation-oriented bilingual dictionaries of the general vocabulary should be compiled within the framework of an overall applied-linguistic description of both languages, with semantic typology of languages being the theoretical basis of such a description.In practical terms, it will mean that in cases when the semantic-pragmatic content of a source-language lexical item is not lexicalized in the target language, the equivalent inflectional word form(s) and/or syntactic structure(s) (with appropriate intonation patterns, if needed) should be furnished in the bilingual translation-oriented dictionary.
Moreover, it should be clear from the exemplification above that a dictionary specifying intralingual and interlingual lexical-semantic correspondences is not sufficient.Probably an optimal solution is a reference work featuring semantic and pragmatic correlations of not only lexical units but also those of all the symbolic units within semantic domains of various kinds, i.e. frames, scripts and semantic fields.In this case, the reference work under consideration will not be a dictionary of synonyms, but a bilingual onomasiological (ideographic) dictionary provided with an accompanying communicative grammar.
However, the discussion concerning translation lexicography should in the first place focus on the problems of selection of lexicographic (and grammaticographic) data and adequate description of appropriate linguistically and/ or culturally significant facts enabling the user to produce L2 texts.Matters of lexicographic presentation, i.e. whether lexicographic information should be accounted for in an alphabetically-organized word list, in contrasted synonyms groups, or within an onomasiological classificatory scheme seems to be a question of editorial decision.Of course, an onomasiological arrangement of lexical items seems very appealing for a bilingual dictionary designed for L2 text production.Nevertheless, it seems feasible to present the same lexicographic data in a semasiologically organized word list.
The existing bilingual dictionaries are often criticized for an insufficient description of the syntagmatic potential of lexical items.It should be emphasized that this requirement primarily concerns translation-oriented reference works.Reception-oriented bilingual dictionaries are not to account for those linguistic phenomena.
Thus, one of the requirements for a translation dictionary should be extensive information concerning the syntagmatic properties of the lexical items of the target language, i.e. their syntactic and lexical-semantic valence and collocational range.As I have mentioned elsewhere, certain regularly co-occurring words may seem to be free word combinations from the standpoint of a native speaker of a language but clearly display the characteristics of set phrases from the non-native speaker's viewpoint.Nevertheless, they are language-specific, and as such, have to be noted in a learner's dictionary designed for a nonnative speaker-user (Burkhanov 2003c: 110).For instance, in the Slavonic languages the lexical-semantic correlate of be in prison literally means 'sit in prison', cf.сидеть в тюрьме (Russian) or siedzieć w więzieniu (Polish).There is even more reason to assume that those phrases should be featured in the word list of a bilingual translation-oriented work of reference.
A translation-oriented reference work should not only account for the syntagmatic potential of all the lexical items, i.e. their lexical-semantic and syntactic valence and collocability.It should also represent language-specific linguistic expressions that Mel'čuk (1995) designated as pragmatemes, i.e. pragmatically constrained expressions which, semantically, are free word combinations but cannot be substituted by synonymous linguistic expressions in certain sociocultural contexts; for instance the inscription on perishable groceries specifying the expiry date: best before … .Its equivalents vary from language to language, cf.à consommer avant … ['for consuming before'] (French), haltbar bis … ['keepable until'] (German), sporzyć przed … ['to consume before'], срок годности до … ['period of usability until'] (Russian), etc.It seems reasonable to assume that such expressions should be included into the word list of a bilingual L1 →L2 translation-oriented dictionary.
In general, translation-oriented lexicography should be developing in the direction of accounting for complex units of translation rather than cross-linguistic correspondences between individual lexical items.
It should be clear from the considerations above that in the meantime a practicing translator has to compensate for the deficiency and have at his/her disposal a set of bilingual and monolingual lexicographic works of reference which not only provide target-language lexical-semantic equivalents, but also, and at least no less importantly, contain lexicographic data concerning targetlanguage synonym discrimination, collocations and usage.

Concluding Remarks
The majority of the existing bilingual dictionaries are compiled on the basis of the postulated cross-linguistic semantic-pragmatic correspondence between lexical items of an obligatory character.It has been demonstrated that the bilingual dictionary cannot, in principle, always be the source of immediately insertable translation equivalents.The foregoing considerations are not intended to undermine the very raison d'être of bilingual dictionaries.They aim to emphasize that bilingual lexicography should be based on a clear-cut conception of the future reference work specified in terms of a number of lexicographic parameters, the principal ones being "intended user group" (foreign language learners vs. professional translators, etc.) and "the purpose of the dictionary" as a utility tool (text production in L2 vs. text comprehension in L1, etc.).Of primary importance in this context is the specification of the lexicographic function(s) the bilingual work of reference is intended to fulfil.
It has been claimed that a translation-oriented bilingual L1 →L2 dictionary should be a reference work intended to serve the purpose of text production in L2.In the situation of professional translation, L2 text production is subject to two types of constraints: (a) constraints as superimposed by the receptor language and culture; and (b) constraints as superimposed by the source text written in L1.If constraints of the second type cannot be, in principle, envisaged, those of the first type can, and should, be accounted for in a bilingual dictionary designed to satisfy the needs of the professional translator.Certain requirements for an "ideal" bilingual translation-oriented reference work have been specified in this connection.
What has not been considered in this article is to what extent the requirements for a translation-oriented bilingual dictionary overlap with those for other types of bilingual reference works which are designed to perform different lexicographic functions, for instance: bilingual L1 →L2 learner-oriented dictionaries or bilingual L1 →L2 reception-oriented dictionaries.These are issues of primary significance and they still await principled solutions.Nevertheless, these two participants of the process of translating are of no importance for our present purposes.

2.
The cases of intralingual and interlingual equivalence are underlined in the definitions.

4.
Literal translations of lexical equivalents are provided in square brackets.
They should be described with regard to other ways of lexicographic explication of meaning: exemplification, pictorial representations, etc.

Bibliography Dictionaries
also mentions the Source Language Reader in the case of communication in the written mode and the Client, i.e. a person or an organization, who or which commission translator's service, as an indispensable element of professional translation in both modes.