|
What follows is not a specific description of each project, but rather the overarching questions of current interest in the lab. At any one time there are several ongoing projects that examine various aspects of each larger question.
What, Exactly, is a "Word?"
J.J. Gibson wrote in 1966, [Words] "...are components of the environment and sources of stimulation...But they are the most difficult of all components or stimuli to define in scientific terms" (p. 28). People in my workgroup believe this to be true, but that's what makes our work so exciting and challenging.
Is a word an object? Is a word an event? The inadequacy of current models of word recognition makes even these basic questions difficult to answer.
What Does it Mean for a Word to "Mean" Something?
A word means what it means because members of a language community have implicitly agreed that this is what it means.
If we look up the word "home" in the dictionary, we will find several meanings listed:
A place where one lives; a residence.
The physical structure within which one lives, such as a house or an apartment.
A dwelling place together with the family or social unit that occupies it; a household.
The place, such as a country or town, where one was born or has lived for a long period.
The native habitat, as of a plant or animal.
There are more entries, relating to games, computers, and so on. But they only go part way toward some of the central connotations of the word "home," having to do with security, Mom & Dad, the smell of cookies baking, and the sights/sounds of a Christmas celebration.
We have done some work in my lab trying to see whether there are universals of connotation, and if so, how they might relate to the process of recognizing a spoken or written word. Our initial studies used the framework of Osgood and colleagues, in which there are three dimensions: Evaluation (good vs. bad), Potency (strong vs. weak), and Activity (fast vs. slow). We did indeed find that these variables affect the time it takes listeners to recognize words.
Subsequent studies have focused on the dimensions of Danger and Usefulness, because these seem more easily relatable to the concept of "adaptiveness for survival," which we believed to be driving our initial results.
What Should We Make of Semantic/Emotional Effects in Word Recognition?
In most models of word recognition, semantic information (i.e. meaning) does not become available until after a word has been recognized. This view is intuitively pleasing: How, after all, can we know what a word means while we are still trying to figure out what the word is?
In several published studies, we have found that people's subjective ratings of the Danger and Usefulness of words have an effect on how quickly those words can be recognized. This would seem to be just the kind of semantic effect ruled out by most models of word recognition. Perhaps even more intriguing, in one of our recent studies we found that men (but not women) were faster to recognize words with higher ratings on Danger.
Our accumulating data suggest two possibilities. First, all researchers misunderstand the nature of the data provided by common experimental tasks (word shadowing, lexical decision, and perceptual identification in noise). A more likely possibility (in my view) is that aspects of semantics are having a measurable influence during the recognition process.
I believe our data to be most consistent with an "embodied" approach to perception and cognition. There are several variants of embodied approaches, but the essential idea is that perception is constrained by the kinds of bodies we have, and what we can do with them. There is assumed to be a tight link between perception/cognition and muscle movements, a link which is completely absent from traditional models of word recognition.
Michaels and Carello (1981) asserted, "Obviously evolution cannot build in or preprogram the ability to detect the ecological significance of spoken words; words vary from language to language" (p. 79). But couldn’t evolution build in the ability to learn the ecological significance of the spoken words to which we are exposed?
What Are the Effects of Morphological Structure in Word Recognition?
Another major focus of work in my lab has been on the influences of morphological structure on language processing. Investigation of the auditory processing of prefixed words has been a particular emphasis, although we have also done some work with visually-presented items (and suffixed words, as well). We have done work in both English and Dutch on these projects. One current study being done in collaboration with a colleague in Canada examines visual processing of compound words.
What Can We Learn From Cross-Linguistic Language and Memory Experiments?
I've done a couple of studies comparing English and Dutch, and we have also done work with colleagues in Taiwan, Hong Kong, and Canada. One current study is examining memory for text in English-French bilinguals. Cross-linguistic studies tell us whether the phenomena we observe apply generally to the human cognitive and language systems, or are tied more specifically to the structure of a particular language.
|
