Rethinking Linguistic Relativity

Studies in the Social and Cultural Foundations of Language

by John J. Gumperz (Editor), Stephen C. Levinson (Editor)

Linguistic relativity is the claim that culture, through language, affects the way in which we think, and especially our classification of the experienced world. This book reexamines ideas about linguistic relativity in the light of new evidence and changes in theoretical climate. The editors have provided a substantial introduction that summarizes changes in thinking about the Sapir-Whorf hypothesis in the light of developments in anthropology, linguistics and cognitive science. Introductions to each section will be of especial use to students.

Genres: Linguistics, Language, Nonfiction, Science, Psychology, Reference

500 pages, Paperback

First published July 13, 1996

Reviews

[Manny · Rating 4 out of 5]

Sometime in early December, we were having an agreeable picnic with our friends P and J in the Adelaide Botanical Garden when conversation turned to the subject of the Sapir-Whorf hypothesis. We all had opinions to offer, and on some points we didn't agree, but after some discussion at least one thing became clear: despite the fact that Sapir-Whorf is very famous, none of us really knew anything about it. I decided I needed to reduce my embarrassing ignorance. After some search, I ordered a used copy of this edited collection of papers, which arrived just in time to be redefined as a Christmas present. I've now finished it.

First of all, what is the Sapir-Whorf hypothesis? People aren't even completely agreed on this, but the following quote from Whorf does a good job of presenting the key claim:

We cut nature up, organize it into concepts, and ascribe significance as we do, largely because we are parties to an agreement to organize it in this way - an agreement that holds throughout our speech community and is codified in the patterns of our language. The agreement is, of course, an implicit and understated one, but its terms are absolutely obligatory; we cannot talk at all except by subscribing to the organization and classification of data which the agreement decrees.

The above is typically summarised by saying that what we think is determined by the language we use, but it is not clear what this actually means; the book was helpful in improving my grasp of the issues. First of all, it is not the case that all our thinking is linguistic. As Feynman points out somewhere, imagine any complex machine whose functioning you're familiar with. You'll probably see some kind of picture in your head showing how the various parts fit together, and you won't be able to describe most of the picture accurately with words. So to that extent, thought isn't linguistic. A couple of the chapters here explored that theme further. Herbert Clark gives many examples showing how our understanding of what words mean is generally dependent on their context: for instance "muddy" in "muddy water" means "mixed with mud", but in "muddy windshield" it means "spattered with mud". No one ever told you this, but you figure it out because of what you know about the behaviour of mud, water and windshields. Charles and Janet Keller's chapter has another nice take on this, where they walk us through protocols recording how an expert knife-maker takes an order for a new knife and then forges it. The knife-maker needs words to communicate with the client about what they want, but most of the thinking is in terms of visual concepts having to do with shapes and processes: the words are mainly used to reference the images.

On the other side, there is some really interesting data suggesting that language can influence thought in basic ways. The most striking example came from Stephen Levinson's chapter on Tzeltal, a Central American language I hadn't come across before. Instead of our "in front, behind, left, right", the basic orientational system in Tzeltal is "uphill, downhill, sideways": left and right aren't differentiated, and it turns out that Tzeltal speakers have great trouble distinguishing between an object and its mirror image, even after they have received special training. Melissa Bowerman's chapter also looked at locative expressions, and was almost as interesting. If you're a native speaker of English, you probably think the distinction between "on" and "in", used as locatives, is obvious. But in fact, even languages as close as Dutch do things differently; Finnish and Korean do it very differently. There are some effective diagrams contrasting "apple in bowl", "handle on pan", "bandaid on leg", "ring on finger", "fly on wall", "picture on wall" and "cup on table". The on/in distinction turns out to be very largely a linguistic artefact.

The most thought-provoking passage of all, though, came from John Lucy's chapter; Lucy is an expert in this area. Though long, I can't resist the temptation to reproduce it in its entirety:

An example of ideologically mediated structural-functional interplay in the referential domain appears in the intellectualization or rationalizalion of the standard language (i.e. the language of public life and the workplace) in the West as analyzed by Bohuslav Havránek (1932). The functional goal of intellectualization or language forms is to make possible precise, rigorous, and, if necessary, abstract statements capable of expressing a certain complexity of thought. Such language forms may be required for legal, bureaucratic, or technical purposes and reach their fullest elaboration as a functional type in scientific discourse wherein lexical items approximate concepts and sentences approximate logical judgments. In such a case, we must recognize that the decision to use language as an aid to thought and action in this way is itself a cultural achievement and not something to be taken as given. Although the primary rationale for such forms of speech lies in the practical need for standardization in a large and complex social formation, an important secondary rationale lies precisely in the perceived advantages of such speech forms in supporting more precise, "accurate" thinking.

The intellectualization of the standard language manifests itself in the lexicon not only by a simple expansion of the vocabulary but also by changes in the structural relations among words. In order to provide unequivocal words, special distinctions, and abstract summarizing terms, new words must be created or old words adapted — words to express relationships such as existence, possibility, necessity, relations of causality, finality, parallelism, and the like (e.g. unsubstantiated). This entails a specialization of word formative patterns to express abstracted concrete events by a variety of forms such as substance of quality, verbal nouns, verbal adjectives, participial expressions, etc.

Intellectualization also affects the grammatical structure of the language. This is manifested in a preference for nominal groupings brought about by combining nouns with attributes or by nominal predication using empty verbs, a preference for the normalized sentence with clear formal differentiation of the subject and predicate, and a desire to achieve parallelism between the grammatical form and underlying logical structure — for instance by the expanded use of the passive voice. Finally, there is a preference for a tightly knit and integrated structure of sentences and compound sentences with an elaborate hierarchy of superordination and subordination expressing different relations of causality, finality, parallelism, and the like: this tendency also manifests itself in a certain specialization of conjunctions.

This intellectualized or rationalized language sacrifices everyday intelligibility for accuracy. General intelligibility and clarity cannot be the gauge of the accuracy of expression of a mathematical work or a legal document. Where everyday language achieves definiteness of reference by a combination of language conventions and appeal to situation, this rationalized language seeks to achieve a definiteness solely by use or an elaborate set of decontextualized conventional forms, that is, forms defined and codified so as to be generally valid rather than situationally contingent. Ultimately, speakers will require elaborate training or formal schooling in these conventions in order to be able to understand the code and use it to achieve the goals it was designed for. Socially disadvantaged speakers who lack the presupposed language skills, for example those from lower-class strata or minority language communities, may be closed out of certain occupational spheres. Indeed, socially advantaged speakers who control these language patterns may have privileged access to such spheres despite real deficiencies in qualifications in other respects.

Such an ideology can become widespread or even dominant in a culture generally by being valued and therefore analogically extended outside of the sphere(s) in which it first developed. This, again, is a cultural achievement and not necessary. Such a characteristic or dominant linguistic ideology will embody a culturally and historically specific world-view which may appear arbitrary, admirable, or foolish from the outside. Thus Bloom (1981) reports Chinese speakers' reluctance to accept or participate in the theoretical, context-independent mode of discourse characteristic of the West which they regard as amoral in some contexts.

I have seen related arguments advanced by feminist philosophers, but not as concisely or powerfully.

I've given you the bits I liked most. Not all of the book was this engaging, and I thought there was too much about American languages. (I should say that I know nothing about American languages, so this may just be my prejudice speaking). I'm still confused about Sapir-Whorf, but my confusion is at a considerably higher level. Next, I'm going to read an author that many of the contributors cited with approval, and whom I've somehow managed to avoid until now. I'll have more to say when my copy of Vygotsky's Thought and Language arrives.

[Rebecca]

I'm still on the introduction ... er .... so far, all I have to say for this book is that it's dense....

[Sally · Rating 3 out of 5]

This is a collection of academic essays addressing linguistic relativity. I'm not invested enough in linguistics to find the theoretical debate too gripping, but some of the examples from non-Indo-European languages were interesting. I wouldn't read the book for those examples, however, as they form a very small portion scattered sparsely through the book.

[Othman · Rating 2 out of 5]

eeeeh

[Paula Santoro]

Sapir-whorf or whatever