Language and conceptual representation: A cross-linguistic study of motion 

Languages differ widely with respect to the ways in which they encode
motion events. Previous research has identified two main groups depending
on the way motion is lexicalized (Talmy 1985, Aske 1989, Jackendoff
1996). In some languages (e.g. English, German), manner of motion is
typically encoded in the verb, while direction of motion information
appears in prepositional phrases. In other languages (e.g. Spanish, Modern
Greek), the verb usually encodes the direction of motion, while the manner
information is encoded in gerunds or prepositional phrases:

(1)	The man ran down the stairs. 
(2)	O andras	katevike	tis skales	trexontas.  (M. Greek)
	the man	descended	the stairs	running

It is an interesting question whether these cross-linguistic differences
(also documented in subsequent psycholinguistic research - see Slobin
1996, Naigles et al. 1998), have an effect on the way speakers of
different languages reason about motion events. We report the results from
two studies that were designed to investigate this possibility. 
We compared the performance of monolingual English and Modern Greek
speakers (a) in non-linguistic tasks involving motion events, and (b) in
their linguistic descriptions of these same motion events. The first study
included preschool and elementary school children and adults. Speakers
were asked to describe a series of pictures depicting various motion
events. Later, subjects were shown a second set of pictures, similar to
the originals but differing in either manner or direction of motion. We
then tested recognition memory of the subjects for the second set of
pictures.
Participants in the second study were 8-year-old children and adults. They
were shown three short motion events in a picture-book format. The first
event ('the sample')  resembled each of the other two events in terms of
either manner or direction of motion. Participants were asked to select
the story in which the actor was 'doing the same thing' as in the sample
item. After the task was completed, subjects gave linguistic descriptions
of all motion events. 
It was hypothesized that, if cognitive processes are affected by
language-specific patterns, the English-speaking population should be more
likely to pay attention to manner of motion information than the
Greek-speaking population in both the memory and categorization tasks. We
found no evidence supporting this neo-Whorfian assumption. Even though the
two linguistic groups differed significantly in terms of their encoding
(linguistic) preferences (in accordance with the previous literature),
their performance in the non-linguistic tasks was identical. In the memory
experiment, there was no differing tendency, across languages, to preserve
either the path or manner information more reliably. In the categorization
experiment, both groups again performed identically: Language did not
predict whether individual subjects would choose the path-identical option
or the manner-identical option. The results are inconsistent with the
strongest (and most controversial) version of linguistic determinism,
according to which language necessarily plays a causal role in both the
time-course and the end product of cognitive development.