Commentary on J. R. Hurford article entitled “The Neural Basis of Predicate-argument Structure”.

Abstract: 58 words
Main Text: 934 words
References: 211 words
Total Text: 1203 words

Perceiving and Describing Motion Events

Shulan Lu
Psychology Department and Institute of Intelligent System
The University of Memphis
3693 Norriswood, Psychology Building
Memphis, TN 38152
U. S. A.

901 678 2145
slu@memphis.edu

Donald R. Franceschetti

Physics Department and Institute of Intelligent System

The University of Memphis

Memphis, TN 38152

U. S. A.

901 678 5257
dfrncsch@memphis.edu

Abstract

According to Hurford, PREDICATE (x) is correlated with deictic object variables during event perception. This claim is inconsistent with some core literature on the perception of motion events. We point out that the perception of events involves the activation of the modal properties and amodal properties of underlying event structure, for which the Hurford account fails to account.

Arguing that PREDICATE (x) is a schematic representation of object properties and spatial information, Hurford approached the problem by dissecting the components of the predicate logic and correspondingly how attention is partitioned when an event is perceived. For instance, variables in FOPL tend to be instantiated by whole objects. Hurford reviewed the relevant literature and did find evidence supporting this position. This raises a question: How do we perceive and describe events?

Imagine a man living 50,000 years ago. How does he survive? He tracks, follows, and hunts animals. He tries to avoid some animals. He and his people definitely need to have certain variables that convey the features of the animals. More importantly, he needs to perceive and interpret the motions of the objects around him. How do people describe motion events? Talmy (1975; 2000) suggested that there are the following four semantic components of motion events: (a) the Figure: an object moving or located with respect to another object (the Ground); (b) the Motion: the presence per se of motion or location in the event; (c) the Path: the course followed or the site occupied by the Figure object with respect to the Ground object; and (d) the Ground. Furthermore, there is an external co-event that often bears the relation of Manner or Cause to the motion event. The following is Talmy’s own example that demonstrates all four of the semantic components and the notion of co-event (2000; pp. 26):

	Manner	Cause
Motion	The pencil rolled off the table.	The pencil blew off the table.
Location	The pencil lay on the table.	The pencil stuck on the table (after I glued it).

In the above example, the pencil is the Figure, and the table is the Ground. Off and on encode the Path (or a site). The verbs in the top row encode the motion, and the verbs in the bottom row encode the location. Besides the states of Motion, the Manner is expressed by rolled and lay, while the Cause is expressed by blew and stuck.

Cross-linguistic studies have suggested that these semantic components are part of the conceptual structure despite the fact that different languages have different surface forms (Talmy, 2000). For instance, English language tends to use prepositions when expressing the Path, while Spanish tends to package the Path into the verb. This indicates that the perception and interpretation of motion events is more than the mapping of the predicate logic and the activation of the relevant components in the brain. The perception of motion involves the segmentation of motion.

How do we partition the stream of motion? Zacks and Tversky (2001; pp. 10) proposed that “The building blocks of events should be temporal units in which the figure, motion, path, and ground are constant. A change in any of these features of the situation constitutes a new atomic event”. As the hunter sees a deer, the hunter quietly approaches the deer. The deer seems to be noticing something. It turns its body and looks around. Very likely the change in the motion of the deer is perceived as an event. The hunter hides himself in the bushes, holding his breath. As the deer resume its activity, the hunter creeps toward the deer again. This time the deer spots the movement. The deer runs. The hunter chases the deer. The deer suddenly takes a left turn. Here a change in path occurs. Quickly the deer gets into the forest. Here a change in Ground occurs. The hunter sees that the deer is running, therefore, the hunter accelerates his speed. Zacks and Tversky (2001) suggested that changes in Figure may not lead to a new sub-event. However, the following example suggests that the change in Figure could give rise to a new event. The deer runs into its heard, and the heard is running away from the hunter. This suggests that the perception of events involve perceiving the changes and segmenting the stream of information. Zacks and his colleagues (2001) identified a network of brain regions tuned to perceptually salient event boundaries under both passive viewing and active segmentation of videotapes of goal-directed everyday activities. The segmentation involves the differentiation of the aspects associated with motion, which an account of updating a few variables of object profile could not account for. By examining the patients who are incapable of perceiving the motions, studies in neuroscience suggests that the perception of motion involves more than the striate visual cortex (Bear, Connors, & Paradiso, 2001).

When the hunter sees the heard, what is his mental representation? According to predicate argument structure, his representation is heard (deer).  However, we tend to treat the heard itself as an object. The argument here is that the perception of events is more than the output of object variable system (Hurford’s major claim). The perception of events involves the perception of the underlying structure of an event, which is amodal (Gibson & Spelke, 1983). For instance, viewing motion events could activate the perception of causality (Michotte, 1943 / 1964; Scholl & Nakayama, 2002). The hunter throws the flint toward the deer. To capture the deer, the hunter needs to know the angle of the flying flint and the velocity. There is evidence showing that people tend to spontaneously construct hierarchical structure of a mundane goal directed events such as hunting during the online perception of the event (Zacks, Tversky & Iyre, 2001). In summary, this suggests that the perception of events not only involves the perception of the modal properties but also amodal underlying structure of the incoming information. This is where the current account falls short.

References

Bear, M. F., Connors, B. W., & Paradiso, M. A. (2001). Neuroscience: Exploring the brain. Baltimore, MD: Lippincott, Williams, & Wilkins.

Gibson, E. J., & Spelke, E. S. (1983). The development of perception. In P. Mussen, J. H. Flavell, & E. M. Markman (Eds.), Handbook of Child Psychology: Cognitive Development (Vol. 3). San Francisco, CA: Jossey Bass.

Michotte, A. E. (1963). The perception of causality [T.R. Miles & E. Mile, Trans]. New York: Basic Books. (Original work published 1946)

Scholl, B. J., & Nakayama, K. (2002). Causal capture: Contextual effects on the perception of collision events. Psychological Science, 13, 493-498.

Talmy, L. (1975). Semantics and syntax of motion. In J.P. Kimball (Eds.), Syntax and semantics (Vol.4, pp.181-238). New York, NY: Academic Press.

Talmy, L. (2000). Toward a cognitive semantics. Cambridge, MA: MIT Press.

Zacks, J. M., & Tversky, B. (2001). Event structure in perception and conception. Psychological Bulletin, 127, 3-21.

Zacks, J. M., Braver, T. S., Sheridan, M. A., Donaldson, D. I., Snyder, A. Z., Ollinger, J. M., Buckner, R. L., & Raichle, M. E. (2001). Human brain activity time-locked to perceptual event boundaries. Nature Neuroscience, 4, 651-655.

Zacks, J. M., Tversky, B., & Iyer, G. (2001). Perceiving, Remembering, and Communicating Structure in Events. Journal of Experimental Psychology: General, 130, 29-58.