Skewing between languages and cross-language equivalences

Review of Frameworks for the Representation of alternative Conceptual Orderings as Determined by Linguistic and Cultural Contexts (Part #11)

[Parts: First | Prev | Next | Last | All] [Links: To-K | From-K | From-Kx | Refs ]

Language is not a neutral vehicle as is frequently assumed by those concerned by information questions. The processes of thinking are affected by the categories and forms available in the language used. This phenomenon is known as the 'Whorfian hypothesis' which has been formulated as: 'observers are not led by the same picture of the universe, unless their linguistic backgrouds are similar or can in some way be calibrated' (25, p. 65). There are some frequently cited examples of this phenomenon, such as the 200 Eskimo words for 'snow', languages which have no separate words for 'blue' and 'green' or others which use verbs in preference to nouns. 

Unfortunately such striking examples have proved of little more than anecdotal significance in affecting thinking about information. It has apparently not yet been possible to clarify in any systematic way how category formation and patterns of categories differ between languages in terms of their effects on information transfer between languages. There are many studies of isolated cases, many comparative studies of languages, but the implication for cross-cultural information transfer does not seem to have been brought out. Indeed Howard Gardner in a recent study of human congnition dismisses 'rabid whorfians' because research has not established the absence of lexical equivalents to be of major significance (25). He is however careful to point out that such research has tendend to focus on tangible phenomena and not on abstracts and generic concepts. It is in this area that major pitfalls appear to lie as is indicated below. 

One of the major difficulties in clarifying such implications is the natural tendency to focus on the words used to convey significance in one language as compared to those used in another. The focus immediately becomes word-oriented rather than concept-oriented. This tendency is reinforced by one school in the long-standing debate as to whether concepts can be usefully considered asindependent of the words by which they are communicated. By locking onto words, for which it is assumed that equivalents in other languages exist or can be found, the embarrassing conceptual problems indicated by the Whorfian hypothesis can be conveniently ignored. 

Clearly it is not possible in this paper to do more than explore ways of drawing attention to this problem and its potential implications for information use. As a shortcut, for purposes of illustration, the assumption has been made here that differences between languages, beyond those of simple substitution of equivalent words (dog = chien = Hund, etc), are indicative of significant differences in the actual handling of concepts. 

The task of this section is therefore to note such non-trivial differences and infer, where possible, how these may affect information handling between languages/cultures. For this exercise it has been found that the organization of material provided by linguistic studies is not as helpful as those concerned with the non-trivial problems of translation, especially between unrelated languages, and more so when the emphasis is on the meaning dimension. The recently-published study used for this purpose is Mildred Larson's 'Meaning-based translation: a guide to cross-language equivalence' (26). 

The forward notes: 

'In the last thirty years a body of literature on translation theory, strongly influenced by modern linguistics, has appeared to support and give academic respectability to the new profession of the non literary translator. Some of these books... have been too philosophical and abstract to relate at all closely to the translator's mundane problems. Some have leaned towards contrastive linguistics. Others have tended to tie translation too closely to linguistic theory. Nida's were the first to deal at all practically with the cultural as well as the manifold linguistic problems of translation'. Larson's study is also significant because of the remarkably wide range of languages from which examples are taken. They include Asian, African, Amerindian, Australasian and European languages. 

A summary of relevant points raised in Larson's study is presented in Annex 2. Of special interest are the following : 

• in order to restate information in a second language, it may be necessary to 'unpack' the concepts and then 'repackage' them in a manner appropriate to the concept scheme of that language. 
• Whilst languages tend to have equivalent terms for specific tangible phenomena, it is increasingly difficult to find equivalents for the generic concepts by which they are ordered as such concepts increase in abstraction. 
• There are many situations in which the concept is unknown within the framework of the receptor language. 
• In addition to the difficulty of finding lexical equivalents is that arising from radically different approaches to ordering the relationships between concepts in propositions and propositional clusters. 
• languages tend to differ significantly in the manner and rate at which it is appropriate to introduce new information into an information exchange. 

Whilst such issues are of interest in their own right in considering the problems of information overload and information underuse. Annex 2 is of even greater instance as a model of the kinds of issues affecting information transfer between sub-cultures or jargons within the same language. It highlights issues which may seriously affect the exchange of information between 'neighbouring' disciplines or schools of thought. In particular the prime concern of a translator to ensure that information is presented in an appropriate and natural manner, in order that it should not be perceived as artificial or facilitate misunderstanding, may be related to the difficulty encountered when information from one school of thought is presented to another. Whilst 'comprehensible', it may easily be viewed as artificial and even experienced as distasteful. 

Of great interest too are the manner in which different languages order sets of concepts in very different ways. It could well be argued that a fundamentally different worldview is built into a language which, for example, favours presenting purpose at the beginning of any grammatical structure in contrast to one in which the purpose statement is presented following the main clause. What for example is the impact on a worldview of presenting the principal verb at the end of a propositional cluster, as in German ? 

Such issues suggest ways of looking at the ordering of information favoured by different disciplines, professions and schools of administration. Whilst it is easy to recognize the extreme distaste with which those with a literary background view the highly ordered documents produced by international administrations (e.g. numbered paragraphs and resolutions), mathematicians (e.g. numbered equations), or numbered propositions (e.g. in some studies of logic or philosophy), Annex 2 suggests other possible differences in ordering which may also hinder rather than facilitate the use of information. 

A series of exploratory studies have been carried out by the Union of International Associations on the implications of such issues for exchange of information within the international community concerning international organizations. These culminated in a symposium on the question (27). Two particular points of emphasis were : 

• whether the neutrality of the English use of 'non' (as in nongovernmental) was preserved on translation into other languages, or whether it acquired pejorative connotations (as in antigovernmental) (28). 
• whether the conceptual distinctions implied by such sets as: cross-, pluri-, multi-, inter-, and trans- (as in multinational), were maintained or lost in other languages (29). 

In both cases the question was to what extent a pattern of distinctions (as opposed to isolated concepts) could be maintained during translation. A back translation technique was used employing mainly non-indo-european languages. The results suggested that such patterns of distinctions were not maintained, even (to some extent) between languages as close as English and French. 

[Parts: First | Prev | Next | Last | All] [Links: To-K | From-K | From-Kx | Refs ]