<bergen@hawaii.edu>

Department of Linguistics
University of Hawaii at Manoa

Spatial imagery in language understanding

Embodied approaches to language view meaning as critically involving real-world perceptual and motor knowledge accrued through world experience (Lakoff 1987). A recent development of embodied views proposes that understanding language entails internally imagining, or simulating, the content of utterances, using brain mechanisms responsible for performing the described actions or perceiving the described percepts. Experimental evidence supports this notion (Zwaan et al. 2002, Richardson et al. 2003). This paper builds on existing results, investigating several linguistic questions. First, while it's known that concrete, literal language results in spatial imagery, is the same true for abstract or metaphorical language? Second, which sentential constituents yield spatial imagery? Subjects? Verbs? 

Our method followed Richardson et al. (2003), in investigating whether processing different sorts of spatial language produces corresponding visual imagery. Subjects heard sentences encoding downwards (The bottle fell.) or upwards (The ant climbed.) motion. Immediately after each sentence, a circle or square appeared in one of four positions on the screen (up, down, left, or right) and subjects decided as quickly as possible what the shape was. Since visual imagery is known to make use of the visual system (Kosslyn et al 2001), if language understanders perform perceptual imagery while understanding sentences denoting upwards or downwards motion, then hearing sentences encoding upwards or downwards motion should selectively interfere with using the visual system for other purposes. Thus, we expected interference by a sentence encoding motion in a particular direction (up or down) on categorization of a shape in that same part of the visual field.

We tested four types of intransitive sentence: literal up or down sentences (like those above), metaphorical sentences using the same verbs (like The cost fell, and The temperature climbed), abstract descriptions of quantity change (like The cost increased versus The temperature decreased) and finally sentences in which the subject was strongly associated with upness or downness but the verb was not (like The grass glistened versus The sky darkened). 

Both types of concrete sentence (those encoding upwards versus downwards motion and those containing an up- or down-related subject) yielded significant interference (p<0.05) on responses in the categorization task. The abstract sentences yielded no such effect (p=0.72), nor did the metaphorical sentences (p=0.5). This suggests that perceptual imagery is important to literal motion sentence comprehension, but may be less so for metaphorical and abstract language. More broadly, the results support the psychological reality of mental imagery during language understanding and a view of linguistic semantics as an interface between language and embodied simulation.


Kosslyn, S.M., Ganis, G., and Thompson, W. L. (2001). Neural  foundations of imagery.     Nature Reviews Neuroscience, 2, 635 -642 

Lakoff, G. 1987. Women, fire, and dangerous things. Chicago: University of Chicago Press.

Richardson, D. C., Spivey, M. J., McRae, K., & Barsalou, L. W. 2003. Spatial representations activated during real-time comprehension of verbs. Cognitive Science.

Zwaan, R.A., Stanfield, R.A., Yaxley, R.H. (2002). Do language comprehenders routinely represent the shapes of objects? Psychological Science, 13, 168-171.