A Rebuttal of Pylyshyn’ s Criticism of the “Continuity Thesis between Seeing and Thinking”

Yingjin Xu

2013-01-01

Abstract:There are two basic positions with respect to the relation be-tween visual perception ( seeing ) and inference ( thinking ): according to the“Continuity Thesis between Seeing and Thinking”, visual perception is in principle penetrable by information ( with semantic content ) , or itself is nothing but consequence of some tacit form of inferences.By contrast, ac-cording to the “Discontinuity Thesis between Seeing and Thinking”, visual perception is impenetrable by high-level inference, or in other words, the visual module works in a manner which is independent of the central belief system.The target theorist of this article is Pylyshyn, who argued that any attempts to reject the afore-defined“Discontinuity Thesis”will inevitably re-sult in a “pure descriptivist view of visual representations”, which cannot explain why human’ s visual system can efficiently track the targeted visual objects.Hence, the “Discontinuity Thesis”, as he concluded, should be accepted.Nonetheless, his argument has mistakenly assumed that the so-called“pure descriptivist view of visual representations” can be only formal-ized by first-order predicate logic or the like, which is clumsy at diachroni-cally recording visual representations.I will further propose another way to record visual representations, according to which the pragmatic weight of each description will be recorded, and consequently, descriptions without e-nough weights are supposed to contribute too little to the fulfillment of the task of tracking targeted objects.If this is the actual way in which the vision system works, semantic contents will still play an indispensable role in tracking objects.Unfortunately, Pylyshyn’ s theory doesn’ t preclude my competing hypothesis, which is still on the very track of the“pure descrip-tivist view”.