Perceptual learning is the learning of perception skills, such as differentiating two musical tones from one another or categorizations of spatial and temporal patterns relevant to real-world expertise. Examples of this may include reading, seeing relations among chess pieces, and knowing whether or not an X-ray image shows a tumor.
Sensory modalities may include visual, auditory, tactile, olfactory, and taste. Perceptual learning forms important foundations of complex cognitive processes (i.e., language) and interacts with other kinds of learning to produce perceptual expertise.Goldstone, R. L., Steyvers, M., Spencer-Smith, J. & Kersten, A. (2000). Interaction between perceptual and conceptual learning. In E. Diettrich & A. B. Markman (Eds). Cognitive Dynamics: Conceptual Change in Humans and Machines (pp. 191-228). Lawrence Erlbaum and Associates Underlying perceptual learning are changes in the neural circuitry. The ability for perceptual learning is retained throughout life.
Basic sensory discrimination Laboratory studies reported many examples of dramatic improvements in sensitivities from appropriately structured perceptual learning tasks. In visual Vernier acuity tasks, observers judge whether one line is displaced above or below a second line. Untrained observers are often already very good with this task, but after training, observers' threshold has been shown to improve as much as 6 fold. Similar improvements have been found for visual motion discrimination and orientation sensitivity. In visual search tasks, observers are asked to find a target object hidden among distractors or in noise. Studies of perceptual learning with visual search show that experience leads to great gains in sensitivity and speed. In one study by Karni and Sagi, the time it took for subjects to search for an oblique line among a field of horizontal lines was found to improve dramatically, from about 200ms in one session to about 50ms in a later session. With appropriate practice, visual search can become automatic and very efficient, such that observers do not need more time to search when there are more items present on the search field. Tactile perceptual learning has been demonstrated on spatial acuity tasks such as tactile grating orientation discrimination, and on vibrotactile perceptual tasks such as frequency discrimination; tactile learning on these tasks has been found to transfer from trained to untrained fingers. Practice with Braille reading and daily reliance on the sense of touch may underlie the enhancement in tactile spatial acuity of blind compared to sighted individuals.
In the natural world Perceptual learning is prevalent and occurs continuously in everyday life. "Experience shapes the way people see and hear." Experience provides the sensory input to our perceptions as well as knowledge about identities. When people are less knowledgeable about different races and cultures, people develop stereotypes. Perceptual learning is a more in-depth relationship between experience and perception. Different perceptions of the same sensory input may arise in individuals with different experiences or training. This leads to important issues about the ontology of sensory experience, the relationship between cognition and perception.
An example of this is money. Every day we look at money and we can look at it and know what it is but when you are asked to find the correct coin in similar coins that have slight differences we may have a problem finding the difference. This is because we see it every day but we are not directly trying to find a difference. Learning to perceive differences and similarities among stimuli based on exposure to the stimuli. A study conducted by Gibson's in 1955 illustrates how exposure to stimuli can affect how well we learn details for different stimuli.
As our perceptual system adapts to the natural world, we become better at discriminating between different stimuli when they belong to different categories than when they belong to the same category. We also tend to become less sensitive to the differences between two instances of the same category. These effects are described as the result of categorical perception. Categorical perception effects do not transfer across domains.
Infants, when different sounds belong to the same phonetic category in their native language, tend to lose sensitivity to differences between speech sounds by 10 months of age. They learn to pay attention to salient differences between native phonetic categories, and ignore the less language-relevant ones. In chess, expert chess players encode larger chunks of positions and relations on the board and require fewer exposures to fully recreate a chess board. This is not due to their possessing superior visual skill, but rather to their advanced extraction of structural patterns specific to chess.De Groot, A.D. (1965). Thought and choice in chess. The Hague, Netherlands: Mouton.
When a woman has a baby, shortly after the baby's birth she will be able to decipher the difference in her baby's cry. This is because she is becoming more sensitive to the differences. She can tell what cry is because they are hungry, need to be changed, etc.
Extensive practice reading in English leads to extraction and rapid processing of the structural regularities of English spelling patterns. The word superiority effect demonstrates this—people are often much faster at recognizing words than individual letters.Wheeler, D. D. (1970). Processes in the visual recognition of words (Doctoral dissertation, University of Michigan, 1970)" Dissertation Abstracts Internationals 31(2), 940B.
In speech phonemes, observers who listen to a continuum of equally spaced consonant-vowel syllables going from /be/ to /de/ are much quicker to indicate that two syllables are different when they belonged to different phonemic categories than when they were two variants of the same phoneme, even when physical differences were equated between each pair of syllables.
Other examples of perceptual learning in the natural world include the ability to distinguish between relative pitches in music, identify tumors in x-rays, sort day-old chicks by gender, taste the subtle differences between beers or wines, identify faces as belonging to different races, detect the features that distinguish familiar faces, discriminate between two bird species ("great blue crown heron" and "chipping sparrow"), and attend selectively to the hue, saturation and brightness values that comprise a color definition.
Brief history The prevalent idiom that "practice makes perfect" captures the essence of the ability to reach impressive perceptual expertise. This has been demonstrated for centuries and through extensive amounts of practice in skills such as wine tasting, fabric evaluation, or musical preference. The first documented report, dating to the mid-19th century, is the earliest example of tactile training aimed at decreasing the minimal distance at which individuals can discriminate whether one or two points on their skin have been touched. It was found that this distance (JND, Just Noticeable Difference) decreases dramatically with practice, and that this improvement is at least partially retained on subsequent days. Moreover, this improvement is at least partially specific to the trained skin area. A particularly dramatic improvement was found for skin positions at which initial discrimination was very crude (e.g. on the back), though training could not bring the JND of initially crude areas down to that of initially accurate ones (e.g. finger tips). William James devoted a section in his Principles of Psychology (1890/1950) to "the improvement in discrimination by practice". He noted examples and emphasized the importance of perceptual learning for expertise. In 1918, Clark L. Hull, a noted learning theorist, trained human participants to learn to categorize deformed Chinese characters into categories. For each category, he used 6 instances that shared some invariant structural property. People learned to associate a sound as the name of each category, and more importantly, they were able to classify novel characters accurately. This ability to extract invariances from instances and apply them to classify new instances marked this study as a perceptual learning experiment. It was not until 1969, however, that Eleanor Gibson published her seminal book The Principles of Perceptual learning and Development and defined the modern field of perceptual learning. She established the study of perceptual learning as an inquiry into the behavior and mechanism of perceptual change. By the mid-1970s, however, this area was in a state of dormancy due to a shift in focus to perceptual and cognitive development in infancy. Much of the scientific community tended to underestimate the impact of learning compared with innate mechanisms. Thus, most of this research focused on characterizing basic perceptual capacities of young infants rather than on perceptual learning processes.
Since the mid-1980s, there has been a new wave of interest in perceptual learning due to findings of cortical plasticity at the lowest sensory levels of sensory systems. Our increased understanding of the physiology and anatomy of our cortical systems has been used to connect the behavioral improvement to the underlying cortical areas. This trend began with earlier findings of Hubel and Wiesel that perceptual representations at sensory areas of the cortex are substantially modified during a short ("critical") period immediately following birth. Merzenich, Kaas and colleagues showed that though neuroplasticity is diminished, it is not eliminated when the critical period ends. Thus, when the external pattern of stimulation is substantially modified, neuronal representations in lower-level (e.g. primary) sensory areas are also modified. Research in this period centered on basic sensory discriminations, where remarkable improvements were found on almost any sensory task through discrimination practice. Following training, subjects were tested with novel conditions and learning transfer was assessed. This work departed from earlier work on perceptual learning, which spanned different tasks and levels.
A question still debated today is to what extent improvements from perceptual learning stems from peripheral modifications compared with improvement in higher-level readout stages. Early interpretations, such as that suggested by William James, attributed it to higher-level categorization mechanisms whereby initially blurred differences are gradually associated with distinctively different labels. The work focused on basic sensory discrimination, however, suggests that the effects of perceptual learning are specific to changes in low-levels of the sensory nervous system (i.e., primary sensory cortices).Fahle Manfred (Editor), and Tomaso Poggio (Editor), Perceptual Learning. The MIT Press (2002: p. xiv) More recently, research suggest that perceptual learning processes are multilevel and flexible. This cycles back to the earlier Gibsonian view that low-level learning effects are modulated by high-level factors, and suggests that improvement in information extraction may not involve only low-level sensory coding but also apprehension of relatively abstract structure and relations in time and space.
Within the past decade, researchers have sought a more unified understanding of perceptual learning and worked to apply these principles to improve perceptual learning in applied domains.
Characteristics
Discovery and fluency effects