Roles of Attention in Perceptual Learning from Perspectives of Psychophysics and Animal Learning

Article excerpt

The role of attention in perceptual learning has been a topic of controversy. Sensory psychophysicists/physiologists and animal learning psychologists have conducted numerous studies to examine this role; but because these two types of researchers use two very different lines of approach, their findings have never been effectively integrated. In the present article, we review studies from both lines and use exposure-based learning experiments to discuss the role of attention in perceptual learning. In addition, we propose a model in which exposure-based learning occurs only when a task-irrelevant feature is weak. We hope that this article will provide new insight into the role of attention in perceptual learning to the benefit of both sensory psychophysicists/physiologists and animal learning psychologists.

In 1963, Eleanor Gibson defined perceptual learning as follows: "Any relatively permanent and consistent change in the perception of a stimulus array following practice or experience with this array will be considered perceptual learning" (p. 29). Nowadays, the definition of perceptual learning is broader and is generally referred to as improvement on a perceptual or sensory task by practice or experience, which includes cases accompanied by no perceptual changes (Fahle, Edelman, & Poggio, 1995; Watanabe, Náñez, & Sasaki, 2001).

Perceptual learning occurs in all the sensory systems, such as the visual (see, e.g., Ahissar & Hochstein, 1993, 1997; Ball & Sekuler, 1987; Beard, Levi, & Reich, 1995; Crist, Li, & Gilbert, 2001; Dosher & Lu, 1998, 1999; Fahle et al., 1995; Fine & Jacobs, 2002; Fiorentini & Berardi, 1980; Fiser & Aslin, 2002; Furmanski, Schluppeck, & Engel, 2004; Herzog & Fahle, 1998, 1999; Karni & Sagi, 1993; Koyama, Harner, & Watanabe, 2004; Poggio, Fahle, & Edelman, 1992; Ramachandran & Braddick, 1973; Schoups, Vogels, & Orban, 1995; Schoups, Vogels, Qian, & Orban, 2001; Schwartz, Maquet, & Frith, 2002; Watanabe et al., 2001), the auditory (Amitay, Irwin, & Moore, 2006; Bao, Chan, & Merzenich, 2001; Demany, 1985; Polley, Steinberg, & Merzenich, 2006), the olfactory (Bende & Nordin, 1997), the gustatory (Owen & Machamer, 1979), and the tactile (Dinse, Ragert, Pleger, Schwenkreis, & Tegenthoff, 2003; Sathian & Zangaladze, 1997).

Perceptual learning is characterized by three distinctive aspects. First, perceptual learning is often very specific for lower level attributes of the stimulus. It is well known that visual perceptual learning is highly specific to stimulus features, such as retinal location (Ahissar & Hochstein, 1997; Crist, Kapadia, Westheimer, & Gilbert, 1997; Fahle & Edelman, 1993; Fiorentini & Berardi, 1980; Karni & Sagi, 1991; McKee & Westheimer, 1978; Poggio et al., 1992; Saarinen & Levi, 1995; Sagi & Tanne, 1994; Shiu & Pashler, 1992), spatial frequency (Fiorentini & Berardi, 1980; Sowden, Rose, & Davies, 2002), and orientation (Fiorentini & Berardi, 1980; Poggio et al., 1992; Schoups et al., 1995; Shiu & Pashler, 1992). Psychophysical studies of visual perceptual learning have shown that there are cases in which performance on detection or discrimination tasks is improved only with respect to features of trained visual stimuli, such as orientation, motion direction, and location. Perceptual learning also tends to be specific to the trained eye. For example, monocular training of a vernier discrimination task improved task performance to a significantly larger degree for the trained eye than for the untrained eye (Fahle et al., 1995). Second, it takes time for perceptual learning to be formed. In many cases, it takes a few days (thousands of trials per day) to see the slightest increase in performance. Third, perceptual learning is persistent. Once perceptual learning is formed, it lasts for months or years (Karni & Sagi, 1993; Watanabe et al. …