Web Translation Projects/Approaches to Translating Dialect

The present project is a presentation of approaches to translating dialect, prepared based on Leszek Berezowski's 1997 work Dialect in translation, which is a descriptive study of dialect translation strategies. It was intended to be a first step in devising a translation theory that accounted for the translation reality. Berezowski's opinion on previous prescriptive translation theories was that they are rarely used in practice.In order to change that, scholars should first conduct a descriptive study, such as Berezowski’s, of strategies used by translators in practice, without prescriptive judgment. Only then will formulating a general theory be possible as a second step of that process..

This project will look at the theoretical assumption on which Berezowski operates, and explain specialist terminology first. Then, some prescriptive translation theories preceding the approach devised by Berezowski will be briefly discussed. The presentation of the strategies described in Dialect in translation will be accompanied by examples, also included in the book. Finally, this page will present some reservations to Berezowski's findings, put forth by another Polish scholar, Krzysztof Hejwowski, which mostly focused on problems with the operational theoretical assumptions and taxonomy.

Terminology
Berezowski uses several terms and theoretical assumption which it may be useful to explain. All definitions are paraphrases of Berezowski and refer to their uses in his book.

Dialect

In Berezowski's view, dialect is a language variety that is relatable only to the particular speaker - it may be their idiolect or a variety characteristic of a larger group (geographical, social etc), but cannot change depending on context (register, setting, discourse function, etc.). Dialect is also relative, which means that it is definable only against the background of the standard language, and depends on it for certain functions (such as formal language, legal and medical jargon, etc).

Dialect markers

Berezowski talks about strategies as ways of preserving social deixis and intertextuality to the extent possible. This is done at the level of retaining or omitting dialect markers in translation. They can be preserved in their entirety or only in part. Dialect markers are those features of language which imply that it is a non-standard variety.

Berezowski lists dialect markers of four kinds, referring to different levels of language - the level of phonetics, morphology, syntax, and lexis:

- phonetic dialect markers, also called "eye dialect" are indications of non-standard pronunciation through spelling

- morphological dialect markers are those choices from the available options of word-building that would not be used in the standard, including the use of honorifics

- syntactical dialect markers refer to a nonstandard choice of syntactic relations

- lexical dialect markers are choices of non-standard, dialectal vocabulary.

Neutralisation

Here, using fewer markers of dialect in the translation than in the original, which makes the translated text less marked for non-standard speech.

Amplification

Here, using more markers of dialect in the translation than in the original, which makes the translated text more marked for non-standard speech. Also used to refer to a bigger amount of information, cultural references or overall meaning in the translated text than in the original.

Translation

Translation can be interlingual (from one language into another language) or intralingual (between different varieties, registers, etc. of the same language) - Berezowski's book makes references to both. Berezowski's study also operates on the assumption that all translation is inherently imperfect, so rendition of the full meaning of the original is never possible and the translator can only attempt to approximate it. This means that Berezowski will discuss each strategy in terms of the precise effect it achieves, both the losses and gains of meaning, but it also means that translations will be described based on how it is done, however imperfectly, rather than comparing it to the original and creating theoretical models of translation.

Social deixis

Because dialect can only be identified in relation to the standard variety of language, and because both the standard and non-standard have social implications, the use of dialect can be seen as pointing to the position of the speaker in the social realm, just as certain elements of language point to the speaker's location in time and space. Berezowski calls this social deixis. Different dialects will have different social deixes, pointing for example to the speaker's provenance, social standing, education, fluency etc.

Intertextuality

The relation between the dialect and the SL or TL reality; the "effect" that the use of dialect achieves. Because the SL and TL realities are different, the translation will tend to modify the intertextuality in varying degrees. It may happen that the translation will introduce intertextuality that was not there in the original text - Berezowski calls "false intertextuality". This happens, for example, when the TL dialect used in the translation has different associations or refers to a different language group than the SL dialect does.

Source language, or SL - the language of the original (also, SL text, SL culture, and so on).

Target language, or TL - the language of the translation (also, TL text, TL culture, and so on).

Earlier proposals
In his book, Berezowski presents several prescriptive translation theories, attempting to answer the question of how to translate dialect from one language to another. Here will be presented three that mark the most progress towards establishing a theory of dialect translation. As the first such proposition, Berezowski cites Nida's stance, stemming from the generative approach, the prescribed method is finding the "closest natural equivalent to the source-language message ". Berezowski describes it thusly: since communication between languages happens via language universals, it should be able to render all aspects of one language into another, in the case of dialects by replacing the SL dialect with a TL equivalent. This was to achieve an "equivalent effec"t. However, on the basis of later research into this method by Wojtasiewicz, we can see that this approach is faulty because it ignores the aspect of intertextuality - that the use of dialect evokes certain associations in readers. He uses the example of substituting the dialect of the Scottish Highlanders with the speech variety recognisable by the Polish reader as used by the inhabitants of the Podhale region. While both refer to a speech variety used by a community living in a mountainous area, the effect that the use of the respective variety has on the SL and the TL audience respectively is nowhere near equivalent - the SL readership will immediately think specifically of the Scottish Highlanders, while the TL readership - of Podhalanie, and not of the inhabitants of the generalised highlands. Such a simple replacement of the SL dialect with a TL equivalent will lead to significant distortions of the intended intertextuality. The second approach described by Berezowski was the functionalist approach, represented by Catford, which looks at the functions that texts and linguistic devices perform. This was the first approach that specified the extent of the word 'dialect' that it operated on, and differentiated it from style and register. This new definition of dialect comprised idiolect, geographical dialect (or dialect proper), temporal dialect, and social dialect. Because this approach stressed the function of dialect use in text, it said that the translator, when rendering an SL dialect, should choose a TL variety that has the same relationship with the TL culture, as the SL dialect has with the TL culture - the aim is to convey the same function in the TL as in the SL text. Catford calls it the relationship between the dialect and the "situation substance." Defining the extent of the term 'dialect', and acknowledging that it stands in a relationship with the situation substance - made it possible for Catford to set criteria for equivalency between the SL and TL variety. For example, if the original uses an SL variety that is characteristic of the capital, the translator should choose a TL variety also associated with the capital, regardless of whether the actual dialectal features (pronunciations changes etc.) are the same. However, this approach is also imperfect for Berezowski because it only looks at factors objective to all countries, such as the same geographical location - the capital, the north, the south and so on, ignoring any subjective associations that the SL and TL readership may have with those varieties. The dialect of the capital in one culture may, for example, hold some level of prestige, whereas the equivalent variety in another culture will be rid of that prestige, or even have negative connotations. In this case, Catford's approach will fall short of its goal because the intended function of dialect use in the SL will not be achieved in the TL; in Catford's own words: "translation fails [...] when it is impossible to build functionally relevant features of the situation into the contextual meaning of the TL text." Finally, the first attempt at a descriptive study of dialect translation, aiming to answer questions about what strategies are used by translators, the conditions for their choice and the side effects they introduce, has been made by Polish scholar, B Sienkiewicz. She looked at the novel as a genre that is set apart from others by the range of speech varieties that it employs apart from the idiolect of the author, a characteristic that is termed heteroglossia. Sienkiewicz found that translators use four different strategies: image for image substitution (using TL varieties that fulfil the same function as the SL variety), approximate variety substitution (using a TL variety that is equivalent in function to the SL only in some respects), neutralisation (forfeiting the SL imagery introduced by dialect by using a standard variety in translation) or amplification (introducing more differentiation between varieties in the translation than there were in the original). Sienkiewicz also attempted to state the conditions under which each strategy was used: amplification was used most in translations of texts narrated in stylistically neutral language, neutralisation was used when translating texts with a high number of different varieties and in cases when non-standard elements clashed with the translator's ethics; approximate varieties were used most when TL did not have a fully equivalent variety, when the translator did not feel competent in a more appropriate dialect, or when stylistic conventions operational at the time prevented them from using one; finally, the default procedure was image for image substitution, used whenever the above conditions were absent. Sienkiewicz's statements, though detailed, were also not a perfect theory, because she used a very limited number of texts (two), so her study was not representative.

The shortcomings of these attempts at creating a dialect translation theory were what created a need for a descriptive study, such as Berezowski's.

Berezowski's strategies
Below is presented a list of the translation strategies observed and characterised by Berezowski on the basis of data from existing translations of original texts written in English dialect/non-standard variety into Polish. This list preserves the original order in which they were presented in Berezowski's book, which reflects their progress from the most neutralising to the ones which preserve the most dialect markers. All of the descriptions and examples are taken directly from Berezowski, while analyses of the provided examples are mine (exceptions will be indicated by footnotes).

Leszek Berezowski's dialect translation strategies:


 * 1) /Neutralisation/
 * 2) /Lexicalisation/
 * 3) /Partial translation/
 * 4) /Transliteration/
 * 5) /Speech defect/
 * 6) /Relativisation/
 * 7) /Pidginisation/
 * 8) /Artificial variety/
 * 9) /Colloquialisation/
 * 10) /Rusticalisation/

Reservations
In 2004, Krzysztof Hejwowski, another Polish scholar, in his book, Translation: A Cognitive-Communicative Approach, made some important reservations to Berezowski's findings. Most of them have to do not with the observed strategies themselves but rather with some of Berezowski's initial assumptions and the categorisation that he introduced. Overall, Hejwowski's comments build on Berezowski's pioneer work and complement it to create a more extensive theoretical framework.

Firstly, Hejwowski comments on Berezowski's use of the term 'social deixis'. While it is appropriate for those uses of dialect which point to the speaker's social identity, for varieties such as Jamaican English, he claims the term 'ethnic dialect' would be more appropriate; in other places, the term is not applicable at all, such as is the case with idiolect, which only points to a specific person's individual language use. The second reservation regards Berezowski's understanding of the term 'strategy'. Hejwowski explains that the word 'strategy' refers to "a plan of action or policy designed to achieve a major or overall aim" - based on this definition, we can observe that some of the "strategies" which Berezowski lists as separate have in fact the same "overall aim", such as colloquialisation and colloquial lexicalisation, or rural lexicalisation and rusticalisation. Berezowski seems to be categorising his strategies based on the level of language they appear on, rather than looking at what the translator is trying to achieve by employing them. Hejwowski lumps together also speech defect and artificial variety, which he both sees as trying to achieve the indication of the distinctiveness of a speaker's language use by employing a variety that is not associated with any existing language groups. Hejwowski replaces the word 'strategy' with the word 'technique' and comes up with the following classification :

- neutralisation

- functional replacement (replacement of SL itemswith functionally relevant TL items, e.g. Warsaw dialect for Cockney)

- stylisation (introducing elements to the TL text that mark it as archaic, regional, idiosyncratic; Berezowski's rusticalisation, colloquialisation, speech defect)

a) realistic (elements taken from existing varieties)

b) artificial (elements invented by the translator)

- relativisation (marking social distane between characters by means of honorifics/forms of address)

- elimination (ommission of problematic items)