LANGUAGE AND THOUGHT
A. Effects of Spatial Language on Spatial Cognition
How might language affect our spatial representations? Does it create new representations? Does it help solidify old ones? Does it serve as an attentional gate? All of these hypotheses are presently being considered in our field. We are especially interested in the question of whether language can serve as the "attentional glue" for fragile spatial representations-- those that have properties that are not robustly represented in the visual system, and are retained in a fragile manner by our visual-spatial working memory. We are currently carrying out experiments designed to test whether language can facilitate the binding of color and location, and whether this facilitation is achieved by language's function as a general attentional modulator, or as a specific carrier of relational information. Banchi Dessalegn can be contacted for further information on this project.
B. Cross-linguistic Studies of Spatial Cognition and
Spatial Language
How do we manage to talk about space? In order to do so, there must be some
connections between two systems of knowledge--one linguistic, and one non-linguistic.
One possibility is that these two systems of representation are identical--
that is, whatever is represented non-linguistically is also represented by language.
Although this is intuitively plausible, there is a great deal of evidence showing
that languages do NOT encode everything in our system of spatial knowledge.
Still, it seems likely that there is some partial mapping between spatial language
and our other ways of knowing about space-- otherwise, it would be hard to imagine
how we manage to talk about space at all! What is the nature of this mapping
between spatial cognition and spatial language?
Universals in Spatial Language and Spatial Cognition
(Munnich, Landau,
& Dosher, 2001; Munnich & Landau, 2002)
One way of answering this question is to look for universals in spatial language--
those kinds of spatial relationships that are
encoded by ALL languages of the world-- and see whether their organization correspond
to the organization of our non-linguistic
spatial knowledge. Some theoretical and empirical work suggests there are such
correspondences, called "homologies". For
example, Hayward & Tarr (1995) found homologies between spatial memory and
spatial language in English-speaking adults.
Below on the left, you can see the pattern of performance shown by adults in
a non-linguistic memory task, and on the right, the
pattern of labelling shown in a language task, when adults give spatial
terms to name these locations. Darker areas represent
better memory in the non-linguistic task and more frequent use of basic terms
like "above" and "below" in the language task.
But now we confront an interesting problem: Although languages of the world
tend to encode of the same kinds of spatial relationships, there are also striking
cross-linguistic DIFFERENCES in the ways that languages choose to encode spatial
relationships. If there are homologies, then how do these differences arise,
and what effect do they have on other aspects of spatial
organization? Some have suggested that cross-linguistic differences may actually
lead to cross-cultural differences
in our ways of knowing about space. You may recognize this as a version of Whorf's
hypothesis-- the idea that differences in languages create differences in cognition.
Ed Munnich's research (Munnich, Landau, & Dosher, 2001) examines the relationship
between non-linguistic
spatial memory and spatial language across adult native speakers of English,
Korean, and Japanese, using Hayward & Tarr's methods. Although the three
languages code some spatial relationships in the same way, there are other clear
differences. For example, in English, we can use the word ON for relationships
in which some kind of force, such as gravity, holds one thing in contact with
another. e.g The book is ON the table. We cannot use the term ABOVE, for example,
for such a relationship. But both Japanese and Korean possess a single term
that can be used to express relationships covered by the separate English terms
ON and ABOVE.
What do you think? Should such a cross-linguistic difference lead to a difference
in spatial memory? If English clearly
distinguishes between ON and ABOVE relationships, but Japanese and Korean do
not, then should that make it easier for English
speakers to distinguish between these relationships in memory? If you'd like
to know the answer, please download our paper, "Spatial
language and spatial memory: What is universal?", Cognition, 2001.
Cross-linguistic acquisition of verbs (Kim, Phillips,
& Landau)
Using another approach, Meesook Kim's thesis (Kim, 1999; Kim, Phillips, &
Landau, 1998) examined how native adult speakers
of English and Korean encode locational events such as filling, pouring, etc.
Native adult speakers of these two languages encode
the events quite differently. For example, for English-speakers, one can say
either "John poured the juice into the glass" or "John
filled the glass with juice", but not *" John filled juice into the
glass" or "John poured the glass with juice". Thus "pour"
must
occur in one kind of syntactic frame; and "fill" in another. But the
rules for Korean are somewhat different, permitting the
"figure frame" for both verbs: That is, it is grammatical to say the
equivalent of "John filled juice into the glass".
How would a child learn these differences in the ways that their native language
expresses spatial events? It seems unlikely that
young children learning English vs. Korean differ in their observation-based
understanding of events in which pouring or filling
take place. It seems more likely that children must learn the patterns of their
language from linguistic input they receive (in this
case, the syntactic frames which occur together with pouring and filling events)
and from an abstract analysis of the typological
nature of their language. How early is this knowledge acquired by children?
What are the pertinent properties of the linguistic
input provided to children by their parents? What kinds of knowledge about the
universal properties of language must children
begin with, in order to support their learning? Meesook addressed these questions
in her 1999 Ph.D. dissertation.
Read it to find out the answers!