IVA.31 General Statements.

A psychological theory has been well developed which posits that memory representations consist of prototypes and variations (deflections) of a prototype. This prototype or “schema theory”* has been used to account for memory effects with non-spatial stimuli (eg. stories, members of categories, music), and spatial stimuli (eg. drawings, visual scenes) and also for the overall memory of bodily movements (see below).
__________
* Some researchers use the term “schema” (Edmonds et al., 1966a; 1966b; Evans, 1967; Schmidt, 1975), others use the term “prototype” (Neumann, 1974; 1977; Strange et al., 1970) and some use the two terms interchangeably (Posner, 1969; Posner and Keele, 1968; 1970). Generally, a “prototype” refers to the most usual example, or typical features of a class of items, and “schema” includes the prototype plus the range of parameters used to derive variations of that prototype (Edmonds and Evans, 1966a; Franks and Bransford, 1971).
__________


Bartlett (1932) was one of the first modern psychologists to popularise the notion that items are not remembered in terms of particular features but are abstracted into prototypes which are representative of important attributes shared by members within a particular category of items. Individual members of the category are identified by noting their variation away from the prototype. Bartlett (1932, pp. 199-202) referred to this abstracted prototypical memory representation as a “schema” and asserted that body movements are not stored in memory as individual items but that similar movements are organised together into a unitary prototype. The notion of schematic representations was also applied to an overall theory of motor learning by Schmidt (1975; 1976; 1982) which is described as “currently [the] dominant psychological theory of motor learning” (Jordan and Rosenbaum, 1989, p. 753).

The prototype structure of memory representation provides a degree of cognitive economy since a single prototype and small set of variable parameters can represent a large group of actual items. This problem of limited memory capacity can be referred to as the “storage problem” which must be addressed by theories of motor learning (Cummings and Caprarola, 1986, p. 51). Memory representation as a prototype and deflections reduces the demand on memory capacity. “The schema would be stored only once and each stimulus or instance would be stored by noting only those aspects which deviated from the schema” (Evans, 1967).

A prototype structure of memory representation also allows new, never before experienced items to be recognised or produced by judging the degree to which they fit together with, or deviate from, existing prototypes. The ability to recognise and respond to new, never before experienced stimulations can be referred to as the “novelty problem” which must also be addressed by theories of motor learning (Cummings and Caprarola, 1986, p. 51). For example, Bartlett (1932, p. 202) and Schmidt (1975, p. 230) argue that in a tennis or a cricket game a stroke is never an exact repeat of a previous stroke, neither is it an absolutely new stroke, but it is a product of all the past experiences of similar strokes. Novel movements can be executed and recognised which are not exact duplicates of past movements, but are derived by creating variations of prototypical movements stored in memory. In the psychological literature this process of deriving variations of a prototype has been referred to as “corrections” (Evans, 1967), “distortions” (Posner and Keele, 1970), “deviance”, “structural changes”, “transformations” (Franks and Bransford, 1971) and “deviations” (Evans, 1967; Evans and Edmonds, 1966; Rosch, 1975a).

IVA.32 Psychological Effects Indicative of Prototypes.

Certain experimental effects have been identified which indicate that the items in question are mentally represented as prototypes and deflections.

IVA.32a Prototypes perceived and recalled fastest.

Items which are prototypical of a class of items will be perceived, categorised, and recalled faster than items which are less prototypical. This has been found for verbal semantic categories (Loftus, 1973; Rips et al., 1973; Wilkins, 1971), colours (Berlin and Kay, 1969; Rosch, 1973; Rosch-Heider, 1972), numbers (Armstrong et al., 1983), abstract visual shapes (Posner, 1969, p. 65; Posner and Keele, 1968), geometric forms (Rosch, 1973), random dot patterns and body stick-figures (Rosch et al., 1976), actions within an event or “script” (Barsalou and Swell, 1985; Galambos and Rips, 1982), and the angle from which an object is best viewed (Palmer et al., 1981).

IVA.32b Prototypes learned first.

Items which are most prototypical of a category will be learned before other less prototypical items. This has been found for colours (Rosch, 1973), the semantic meaning of a verbal paragraph or story (Anderson and Pichert, 1978; Bower, 1976; Bransford and Franks, 1971; Bransford et al., 1972), geometric forms (Rosch, 1973), random dot patterns, and body stick-figures (Rosch et al., 1976).

IVA.32c Prototypes recalled first.

Items which are most prototypical of a category of items will be recalled before other items which are less typical. This has been found for verbal semantic categories (Battig and Montague, 1969; Bousfield and Sedgewick, 1944), random dot patterns and body stick-figures (Rosch et al., 1976), and the angle from which an object is first imagined to be viewed (Palmer et al., 1981).

IVA.32d Prototypes recalled more accurately.

Items which are more prototypical of a category will be recalled better (ie. more often, with greater accuracy) than items which are less prototypical. This has been found for colours (Rosch-Heider, 1972), actions within an event or “script” (Lichtenstein and Brewer, 1980), objects and their locations in a visual scene (Mandler and Parker, 1976; Mandler and Ritchey, 1977; Mandler and Stein, 1974; Von Wright et al., 1975), and positions of chess pieces (Chase and Simon, 1973; Goldin, 1978; Holding and Reynolds, 1982).

IVA.32e Prototypes serve as reference points.

Items which are prototypical of a class of items serve as reference points for other less prototypical items. For example, a less prototypical item is conceived to be “almost like” a prototype, rather than the prototype being conceived as “almost like” a non-prototype. This has been found for patterns of dots and angles between line segments (Wertheimer, 1923, pp. 78-79), colours, line orientations and numbers (Rosch, 1975a) and locations in a well-known town (Sadalla et al., 1980). Reference points in the kinesphere were experimentally probed as part of this research (see IVA.110).

IVA.32f Perceptual/memory bias toward the prototype.

The most characteristic effect is that prototypical items tend to be perceived/remembered even if they were never actually experienced. When learning a new set of items Subjects will later believe they have seen a prototypical item when in fact they had not. When Subjects are presented with a well-known category or context this creates expectations that prototypical items will be present. This expectation encourages perceptual bias in which actual items may be perceived to be more similar to the prototype than they actually are.

This bias toward the prototype has been found for stories (Anderson and Pichert, 1978; Bartlett, 1932, pp. 83-85), colours (Medin and Shoben, 1988), abstract visual or kinesthetically traced shapes (Franks and Bransford, 1971; Lee, 1985; Neumann, 1977; Solso and Raynis, 1979), patterns of dots (Posner and Keele, 1968; 1970; Strange et al., 1970), an ambiguous letter within a word (Eysenck and Keane, 1990, p. 50; Smyth et al., 1987, p. 7), indecipherable sounds within an auditory sentence (Warren and Warren, 1970), the meaning of a word within a particular context (Anderson and Ortony, 1975; Barclay et al., 1974; Light and Carter-Sobell, 1970), the significance of body movements within a particular context (Poizner, 1983), actions within an event or “script” (Abbott et al., 1985; Bower et al., 1979; Bower and Clark-Meyers, 1980; Schank and Abelson, 1977; Walker and Yekovich, 1984), the identity and locations of objects in realistic visual scenes (Brewer and Treyens, 1981; Friedman, 1979; Labov, 1973; Mandler and Parker, 1976; Palmer, 1975), line drawings of objects (Bartlett, 1932, pp. 178-183), locations of dots within a circle (Huttenlocher et al., 1991), orientation of line segments (Weintraub and Virsu, 1971; 1972), orientation of arm positions (Clark and Burgess, 1984), and orientation or angle of geographical directions (Byrne, 1979; Moar, 1978; Moar and Bower, 1983; Ross et al., 1970).