On the development of color naming in young children: Data and theory

On the development of color naming in young children: Data and theory

BRAIN AND LANGUAGE 26, 72-93 (1985) On the Development of Color Naming in Young Children: Data and Theory MARC H. BORNSTEIN New York University Am...

2MB Sizes 2 Downloads 86 Views

BRAIN

AND

LANGUAGE

26, 72-93 (1985)

On the Development of Color Naming in Young Children: Data and Theory MARC H. BORNSTEIN New York University Among the semantics of different common attributes of the environment, consistent and correct color naming seems to develop paradoxically late in children, for even young infants discriminate and categorize colors well, color is a salient feature of the child’s world, and children are aware of color as a separate domain, know color terms, and respond to color questions with color names. Several specific anomalies characterize early linguistic development in this domain: Errors in beginning color naming are perseverative or random, there seems to be a minimum age for correct and consistent color naming, developmental rate shows wide individual differences, and acquisition among girls is generally faster than among boys. This essay reviews data that support these observations, and evaluates three traditional explanations for them-including the perceptual salience of color for children, experience and learning in the child, and cognitive developmentagainst a fourth new possibility. It is hypothesized that appropriate color naming may depend on the maturation and integration of specific cortical neurological structures; among several interpretations, the neurological one accounts best for diverse characteristics of early color naming. o 1985 Academic PKSS, II-K. The young child distinguishes, recognizes, and easily matches without the least hesitation the most delicate shades of color, and has nothing to envy in the adult so far as his color sense is concerned; it is the verbalization of his color sense, if we may so express it, in which he is defective. A. Binet and T. Simon (1908/1916, p. 215)

INTRODUCTION When viewing a rainbow or other physically continuous chromatic displays, adults may discriminate literally thousands of nuances,’ but Preparation of this paper was supported in part by a Research Career Development award from the National Institute of Child Health and Human Development; in part by a Guggenheim Foundation Fellowship; and in part by a Visiting Professorship at the Laboratoire de Psychologie Exp&imentale et E.P.H.E. (3eme section, associe au CNRS), Paris. Address requests for reprints to Marc H. Bomstein, Department of Psychology, New York University, 6 Washington Place, Room 1065, New York, NY 10003. ’ In the Dictionary of Color, Maerz and Paul (1930) identify more than 3000 English color terms, a number that still does not match normal discriminative acumen. 72 0093-934X185 $3.00 Copyright Au rights

0 1985 by Academic Press, Inc. of reproduction in any form reserved.

EARLY

DEVELOPMENT

OF

COLOR

NAMING

73

nevertheless naturally partition the spectrum into only a few qualitatively distinct categories. In turn, a small set of common primary color names is sufficient to describe the color space: blue, green, yellow, and red. Aristotle (in W. D. Ross (Ed.), 1913) commented specifically on the dominant qualitative chromatic organization of the rainbow, and Newton (1671-1972) also described the prismatically diffracted photic spectrum in this spirit. Modern psychophysical research, by Beare (1963), Ekman (1963), and Boynton and Gordon (1965), among others, has confirmed the consistency of this basic organization in mature color perceiving, even across a wide variety of cultures and language communities (e.g., Berlin & Kay, 1969; Bornstein, 1973). For an amplified discussion, see Bornstein (1975a, 1985a), Boynton (1975), or Miller and Johnson-Laird (1976). This essay is concerned with how children achieve mastery of the basic fourfold color-name organization of the spectrum. The four basic color names represent a necessary and sufficient, if minimal, description of the psychological reality of chromatic space. When and how does the child formulate this first map of color-to-name correspondences? Perceiving colors veridically is a capacity that is present and functional quite early in human development (Bornstein, 1978; Teller & Bornstein, 1985). Moreover, the basic psychological organization of the spectrum seems to be in place before the end of the first year of life. In the pertinent developmental study, Bornstein, Kessen, and Weiskopf (1976) found that even 4-month-old infants discriminate and categorize spectral wavelengths into the four basic qualitative categories and that infants’ groupings closely match the fundamental adult distinctions. In short, insofar as consistent and accurate color naming presupposes mature perception, discrimination, and categorization of color, modern data support the first of Binet and Simon’s (1908/1916) propositions that color vision per se is well developed in the young child. Perception therefore ought not to bar, inhibit, or confuse language development in the color domain. In this light, it is both curious and arresting that correct and consistent color naming follows a tardy and untidy course in semantic development. Though young children see, discriminate, and even categorize colors appropriately, though they know color names, and though color is salient in their lives, correct and consistent color naming develops rather late, and along the way children commit odd, even gross, errors, next to their naming in other related and comparable domains. This observation is, as Binet and Simon (1908/1916) noted, a part of tradition in developmental psychology (e.g., Bateman, 1915; Gesell, Ilg, & Bullis, 1949; Istomina, 1963; Leopold, 1939; Miller & Johnson-Laird, 1976; Norsworthy & Whitley, 1920; Preyer, 1890; Stern, 1924; Vernon, 1962). In the nineteenth century, James Mark Baldwin (1893), the eminent developmentalist, observed and reported such difficulties in his daughter Helen; and Wilibald A. Nagel

74

MARC H. BORNSTEIN

(1906), the eminent visual physiologist, independently observed and reported similar difficulties in his son Gerhard. In the same tradition, the developmental psycholinguist Cruse (1977) nearly a century later observed and reported that his young son Pierre never used and seemed not to understand color words, though they were relatively frequent in his speech to the child, but at the same time “certain words referring to shapes were all well established, and (more or less) correctly used: round, square, and triangle” (p. 305). Modreski and Goss (1%9) likewise reported that 3- to S-year-olds know form names better than color names. This observation is so compelling that, even in spite of their knowledge that infants and young children accurately discriminate colors, many early investigators mistook the paucity, confusion, and inaccuracy of children’s color vocabulary to reflect some fault with their faculty of color vision (e.g., Preyer, 1890; Shinn, 1909; Stern, 1924; Vernon, 1962). Charles Darwin was among them, if only temporarily. In 1877, as part of the German version of his celebrated Biographical Sketch of a Young Znj’unt, Darwin recorded the following observations: I carefully followed the mental development of my small children, and I was astonished to observe in two or, as I rather think, three of these children, soon after they had reached the age in which they knew the names of all the ordinary things that they appeared to be entirely incapable of giving the right names to the colors of a color etching. They could not name the colors, although I tried repeatedly to teach them the names of the colors. I remember quite clearly to have stated that they are color blind. But afterwards this turned out to be an ungrounded apprehension. (p. 376; my translation)

Generally speaking, questions about the relationship between the psycholinguistics of color nomenclature and the physiology and function of color vision are classic (e.g., Gladstone, 1958; Rivers, 1901; Werner, 1940; Whorf, 1952); indeed some authors list color among the most prominent topics in epistemological debates within the social sciences (Berlin & Kay, 1969; Bornstein, 1975a, 1985a; Conklin, 1973; Lenneberg, 1%7; Segall, Campbell, & Herskovits, 1966). The fact that children perceive categories of hue, but initially color name inconsistently and inaccurately poses yet another set of potentially interesting developmental questions for the interrelationship of brain, language, and thought. An otherwise reasonable surmise from the fact that hue categorization precedes color naming developmentally would be that, in this one realm at least, linguistic identification simply overlays perceptuocognitive organization and thereby enormously facilitates semantic development. Paradoxically, it does not. The main aims of this essay are, first, to review the literature that documents the tardy and problematic acquisition and use of color nomenclature by young children and, second, to evaluate alternative interpretations of the synchronic schism between the perception and semantics of color in childhood.

EARLY

DEVELOPMENT

OF COLOR

NAMING

75

DATA Naturalistic

and Empirical

Observations

Consider, first, records of normative language development in young children. A few of the earliest observers of natural language function were circumstantially sensitive to the growth of color vocabulary. Wolfe (1890) reported that of the first 300-500 words which 2-year-olds may command none applied to color, and in the same year Preyer (1890) reported that his son could not reliably “discriminate” even red from green on a verbal basis before 25 months. Similarly, Shinn (1909) and Pelsma (1910) did not hear more than one color term from the children they observed until those children reached 22 and 24 months, respectively. It was generally agreed at this time that children did not usually begin to name colors before they were 2 (e.g., Gesell et al., 1949; Istomina, 1963; Leopold, 1939; Stern, 1924; Tracy & Stimpfl, 1909). Indeed, many contemporary records of the first quarter of this century suggested that color semantics did not emerge at all before children attained 3 years (Bush, 1914; Heilig, 1913; Whipple, 1909) or even 4 years of age (Rowe & Rowe, 1913). In any event, early observers seemed convinced that young children only very infrequently used color names spontaneously or aptly when observing or describing. A difficulty that confounds knowing when young children legitimately possess color words turns on the question of how they use color terms, for the very young curiously misuse color terms. Indeed, it is widely recognized that beginning color naming is unsystematic and unstable for a considerable period. Several authorities across a century of observation, including, historically, Binet (1890), Preyer (1890), Winch (1909), Descoeudres (1921), Koffka (1927), Leopold (1939), Synolds and Pronko (1949), Vernon (1962), Istomina (1963), Cruse (1977), and Bartlett (1977), have recorded that in color naming children may first use a single term for all colors (e.g., a made-up term, like lila; or a real color word, like red; or simply the word color) or, if they possess more than one color name, they may apply diverse names haphazardly (e.g., red for red, yellow, and green, and green for all other colors). Bartlett (1977), who studied the individual development of color naming among 33 children between approximately 2+ and 4 years, documented this phenomenon in a quantitative way. She found that before children acquired four color terms they did not accurately apply any of the names they possessed. The curiosity in this phenomenon is that children know to respond to the question What color is it? with a color name, but they consistently err in responding with the proper color name. Sometimes they perseverate with one color name; sometimes they name colors randomly. It is this qualitative point, the persistent and anomalous nature of the young child’s errors in color naming, given the mature state of the sensory apparatus

76

MARC

H. BORNSTEIN

and perceptual organization that under-pins language along with the child’s knowledge of the proper response domain, that is quizzical. Though children may also acquire the names for tastes, odors, or other sensory experiences relatively late, and though they may acquire names for comparable bodily states or emotions even later, innate perceptual organization in these spheres is often lacking, children do not as readily comprehend in these other response domains, children do not usually possess the appropriate vocabulary, and the pattern of errors that characterizes children’s early naming in these domains is neither as consistent nor as telling. Only relatively late-sometime between 4 and 7 years-do color names assume accurate referential meaning and consistently reflect the spectral composition of particular stimuli (e.g., Dale, 1969; Mervis, Catlin, & Rosch, 1975; Raskin, Maital, & Bornstein, 1983; Yendovitskaya, Zinchenko, & Ruzskaya, 1964/1971). Cook (1931) captured childhood’s discrepancy between correctly perceiving and verbalizing about colors quite clearly. He tested one hundred ten 2- and 6-year-olds in naming and in matching reds, yellows, greens, and blues. Younger children matched colors much more accurately than they named the same colors, and although there was still a matching-naming discrepancy, older children were considerably more proficient at both. Forty years later, Rosch (Heider, 1971) confirmed that color matching and recognition abilities still outstrip color naming in early childhood. Consider, second, empirical studies of the development of color naming. In constructing his original Metrical Scale of Intelligence, Binet (1890; Binet and Simon, 1908/1916) asked young children to name colors he pointed to. To obtain credit on this item, children were required to identify correctly all four “fundamental colors, red, blue, green, and yellow.” Binet initially found that 8 years was the norm for this achievement, although he later concluded that half of the 7-year-olds could pass color naming; Goddard (1911) specified the same age. For the standardized American edition of the Binet test, however, Terman (1916) and Norsworthy and Whitley (1920) reported that the average 5-year-old knew all four colors. Many of their immediate successors contirmed stable color naming by 5 years (e.g., Cook, 1931; Gesell et al., 1949; Goodenough & Anderson, 1931). Synolds and Pronko (1949), for instance, asked 74 children between 3 and 8 years of age to name eight colors from the Dvorine Color Perception Testing Charts, and their data exemplify the startling increase in naming accuracy that occurs around 5 years of age (Table 1). More recent studies have tended to confirm this cohort effect with the earliest age of accurate color naming now having dropped by approximately one-half from Binet’s original estimate of 8 years. Percentages vary across studies; nevertheless, the trend is clear, and 4 years seems to constitute some sort of minimum chronological age for correct and consistent color naming (e.g., Anyan & Quillan, 1971; Conrad, 1972; Dale, 1969; Karpf,

EARLY

DEVELOPMENT

ACCURACY

Age

in Years

OF

TABLE 1 OF NAMING EIGHT AT SIX AGES” Mean

3 4 5 6 7 8 a After

COLOR

NAMING

77

COLORS

percentage

correct

0 25 18 77 94 88 Synolds

and Pronko

(1949).

Goss, & Small, 1974; Kimball & Dale, 1972; Kirk, Hunt, & Lieberman, 1975; Modreski & Goss, 1969). In a systematic and comprehensive study, Johnson (1977), for example, assessed color naming cross-sectionally in children 2$ to 4f years of age. On the average, 38% of 2.6-year-olds, 50% of 3-year-olds, and 56% of 3.3-year-olds knew and could correctly apply four basic color names. Color naming distinctly increased after 34 years: 71% of 3.6-year-olds could name four colors, 72% of 4-year-olds, and 79% of 4.3-year-olds. As have others, Johnson found that girls named better than boys at each age. Three additional experimental studies of the development of color perception confirm that young children’s difficulties in simple color naming are pervasive psychologically. In one experiment, Boynton and Gordon (1965), who studied the simple psychophysics of color identification, found that observer MB, Boynton’s own 12-year-old son, misapplied some color terms: Despite the fact that blue and yellow are mutually exclusive visual perceptions (Hering, 1878/1964), MB named a number of the same wavelengths blue, yellow, or some combination of blue and yellow. Adult observers typically find these perceptions so “extremely different” that they are unable even to scale the degree of difference between them (Miller & Johnson-Laird, 1976). In the second experiment, Schiller (1966), who studied decision processes in color naming developmentally, measured how long adults and children of different ages required to label colors with the same basic names blue, green, yellow, and red. He found that decision time declined with age, but that the decline was as great between first and second grade as it was between second grade and college. In the third experiment, Svinicki, Meier, and Svinicki (1976), who studied verbal mediation in memory, showed children a color, labeled it, and then asked them to pick that color out of a group of neighboring colors. Children 3 to 6 years of age were not affected by the use of color names as verbal mediators in color memory, whereas 7-year-olds and adults were. Svinicki et al., who also looked at gender, found girls more affected by verbal mediation than boys.

78 Summary

MARC

H. BORNSTEIN

and Questions

In brief overview, correct and consistent color naming in young children lags far behind the development of color vision and the acquisition of color terms. Naturalistic observations and empirical studies reveal six principal characteristics of linguistic development in this domain: (i) even young infants perceive colors accurately, that is, discriminate, match, and categorize colors; (ii) young children recognize that color is a separate domain of experience and readily identify it as such, and so the question What color is this? regularly elicits a color term (though infrequently not among the very youngest children); and, (iii) children know different color words. However, (iv) children’s color identification is immature in that they reply with incorrect names to colors they see or pick incorrect colors for names they hear; (v) there seems to be a lower age bound for mature color naming that hovers around 4 years; and (vi) there are relatively wide individual differences in the age and rate for children’s achieving mature color naming, but the developmental function for girls is accelerated relative to that for boys. In short, early in life, sensory and linguistic color knowledge seem to coexist, but a proper map connecting names and perceptions is late in developing. The contradictions inherent in immature color naming raise a number of questions. What motivates this perception-naming schism? Why is early color naming characterized by particular error patterns? Why does consistent and correct naming of colors develop only after about 4 years of age? What factors underlie individual differences in development in this sphere? Why do girls achieve mature color naming earlier than boys? THEORY The development of color naming has traditionally been thought to reflect one of three causes: the relative perceptual salience of color for children, their learning and experience with color, or the course of normal cognitive growth. An unexamined fourth possibility is that color naming depends on the maturation and integration of neurological structures that independently subserve color perception and linguistic nominalization. The second part of this essay explores and evaluates these four possible explanations for the several curiosities identified in the early ontogenesis of color naming. Perceptual Salience The British empiricists proposed that certain qualities of objects, such as their shape, were primary since they contribute more to perceptual identification than do other secondary object qualities, such as their color. Two related theories of semantic development follow this logic and have placed a perceptual salience interpretation on late-developing color semantics. Macnamara (1972), for example, suggested that infants and

EARLY

DEVELOPMENT

OF COLOR

NAMING

79

young children preferentially attend “to varying states and activities rather than unvarying attributes” of the environment and that semantic development proceeds along parallel lines; “thus the order of learning would be as follows: names for entities, names for their variable states and actions, and names for more permanent attributes such as color” (p. 4). Werner (1940) and later Vernon (1962) had argued in a conceptually similar vein that “the child is slow to abstract colour as an independent and variable quality” because the child “regards it as an inherent characteristic of certain particular objects” (p. 100). Nelson (1974) has also theorized that children’s first translations of concepts into words reflect more salient “dynamic functional relations” to which identificational features, like color, are only later attached. On this account, color naming is believed to develop late because color is thought not to be perceived as an independent perceptual attribute that is salient, separate, or specially interesting to young children. Contrary to these opinions, however, color is an impressively variable attribute of objects-the same object may vary in color both absolutely and relatively, as in different lighting (although, to be sure, color constancy is an adaptive and widespread perceptual capacity). Moreover, it is clear from empirical studies of infants (e.g., Bornstein, 1975b, 1978; Cohen & Gelber, 1975; Dodd & Lewis 1969; Lewis & Baumel, 1970) and of young children (e.g., Odom, 1972; Odom & Guzman, 1972) that color is a highly salient dimension of visual experience early in development, as or more impressive than shape, number, or position. Further, empirical research by Brain and Goodenough (1929), Suchman and Trabasso (1%6), FamhamDiggory and Gregg (1975), and especially Tomikawa and Dodd (1980) has demonstrated repeatedly that young children classify and sort by color as or more readily than they do by form or by function. Indeed, color’s spatial redundancy may render it an especially potent perceptual cue for children (Colby & Robertson, 1942). Thus, preschoolers have been found to develop and to use concepts based on color as or more readily than they do concepts based on form, number, or similarity of components (e.g., Lee, 1965; Wohlwill, 1957). If the semantics of color are late in appearing, these studies suggest that it is probably not because the correlative perceptual dimension is secondary or inconspicuous to infants and young children. Further, even if color were not prominent for children, the perceptual salience view would still fail to account for several other odd characteristics of early color naming, including the typical error patterns children make, the age minimum for correct naming, differential development, or regular malefemale differences. Learning

and Experience

An alternative young children,

parsimonious explanation of color-naming anomalies in suggested long ago by Baldwin (1895), is that children

80

MARC

H. BORNSTEIN

simply lack proper experience in associating color terminology with color percepts. By 1) years of age children demonstrate that they can map words onto simple concepts (Carey, 1980); perhaps opportunity in color is just wanting or their learning has proceeded incorrectly. This general rationale appeals to a learning-theory view of semantic development that maintains that naming colors depends on children’s experience and reinforcement history. A philosophical exponent of this view is Quine (1973), and behaviorists like Graham (1965) have asserted that colorname learning proceeds as follows: “a child gives the word ‘green’ in the presence of a specifiable set of wavelengths and in the presence of a parent’s reinforcing approval. ‘Naming a color’ means that, at a later time and in the presence of such wavelengths, the person will continue to give the reinforced response” (p. 350). Certainly, this is a textbook (and plausible) account of how color-name learning might proceed, and although no naturalistic data are at hand to confirm or disconfirm the operation of learning mechanisms directly, education and social class data lend some indirect support to an experiential account. First, schooling seems to provide crucial opportunity to acquire and use color names. Bateman (1915) found that 71% of children could name eight colors before they began school, but 89% could do so after just one semester, leading him to conclude that “a semester of school training markedly increases the ability to name colors” (p. 484). In Synolds and Pronko (1949) as well, children of about 6 who were in school named colors more than four times better than did children who were almost 6 but not yet in school (see Table 1). Schiller’s (1966) findings, cited above, support the view that 1 year of school radically facilitates color naming; and, Anyan and Quillian (1971) found that the use of color names improves markedly among school-trained children. Parental socioeconomic and educational status also seem to influence the development and use of color names in children. Anyan and Quillian (1971), for example, found that color naming in children correlated positively with the level of their parents’ educational attainment. As part of an extensive study of social class differences in language skills, Kirk et al. (1975) found that only 25% of 4-year-old disadvantaged Head Start children could correctly name six colors, whereas 79% of age-matched advantaged nursery-school children could. Head Start children named only about three of six colors on average, whereas nursery-school children named five to six. Kirk et al. (1975) used these class differences in naming performance to exemplify the reality of experiential deficits in linguistic achievement in children from disadvantaged backgrounds; that is, they argued that performance differences on a neutral, “most elementary abstraction” like color must reflect “class differences in the opportunities to acquire knowledge and semantic mastery of this knowledge” (p. 302). Learning or experience could contribute to the acquisition of color

EARLY

DEVELOPMENT

OF COLOR

NAMING

81

naming in children; further, it could be that some children’s linguistic experiences are rich relative to those of others, and in the same vein that girls are systematically advantaged relative to boys (although why would not be clear). However, teaching and drill routines for semantic development are not common in the everyday experiences of children older than about 2 years (Carey, 1980). Moreover, this view does not explain why color-naming errors are patterned in the ways they plainly are. Nor does a learning view explain why consistent and correct color naming appears only at a particular time; if training were all that mattered, theoretically the acquisition and accurate use of color nomenclature ought to have no lower age bound (at least below 14 years or so), and naming ought to benefit from training. However, Kirk et al. (1975) found that as many as 20% of advantaged Syear-olds could still not correctly name six fundamental colors. More importantly, it has been suggested, as by Darwin (1877) and by Binet (1890), and demonstrated empirically, as by Modreski and Goss (1969) and by Bornstein (1985b), that color naming does not benefit much from color-to-name training in children younger than 4 years of age. Children as young as lf years can learn conceptword associations and by that time they have begun to assimilate some vocabulary sets reasonably quickly (Carey, 1980; Cruse, 1977). Rather, the counterfacts to accurate and consistent color naming in early childhood point to a local impediment between the correct association of particular color words with particular colors, leaving intact associations among colors, color words, and the concept of color generally. Cognitive

Development

Among many who have marvelled at the normal child’s peculiar acquisition of color names, at least one cognitive psychologist-psycholinguist has proposed a comprehensive hypothesis linking concepts and semantics. Miller (1977; Miller & Johnson-Laird, 1976) considered the late acquisition of color names in childhood especially curious since, as he observed, color names represent only “a small, isolated suburb of the lexicon without much grammatical structure.” According to Miller, the child’s problematic acquisition of color vocabulary reflects bipartite processes, the first conceptual and the second linguistic. Conceptual development is the more difficult phase for the child for “(a) he must make the appropriate abstraction of color from other attributes of visual experiences; (b) he must establish certain landmark colors; and (c) he must learn to locate all colors with respect to those landmarks.” The linguistic phase necessitates that “(a) he must learn specific uses of color terms in particular contexts; (b) he must isolate the color terms from other words as a contrastive set; and (c) he must learn the referential values of each term” (Miller & Johnson-Laird, 1976, p. 351). In this theory conceptual and linguistic development are thought to progress serially; one essentially

82

MARC

H. BORNSTEIN

involves learning to abstract the appropriate attribute and to anchor color perceptions to some internal frame of reference, while the other involves discovering which words are relevant to that frame and learning which location in the frame goes with which term. Miller’s explanation of the delayed advent of color naming is clearly more embracing and sophisticated than are those offered by perceptual salience or by learning theories. There is much to recommend it. For example, although Miller himself places no great theoretical weight on conceptual step c-the location of nonfocal colors relative to focal onesdata of Mervis et al. (1975) ‘and of Raskin et al. (1983) suggest that, although the location of nonfocal color boundaries may be fixed early, the boundaries themselves only stabilize later. Thus, aspects of perceptual development, as Miller theorized, may play a role in retarding color naming. Moreover, the time required for children to accomplish Miller’s several steps could place a kind of lower limit on the onset of color naming. However, the cognitive view offers no intrinsic reason why mastering the basic set of color words should be more difficult than mastering any other closed word class. In fact, insofar as semantic development in Miller’s theory presupposes the earlier isolation of a “focal” subset of the attribute class to be named, the development of color naming, next to other word classes, ought to proceed apace since there is good evidence to suggest, first, that the color domain, unlike many other attribute domains, presents obvious focal instances to perception and, second, that focal colors are already physiologically salient and meaningful very early in life. For example, Bornstein (1975b) found that 4-month-old infants attend longer to focal colors as a set than to nonfocal colors; and, Rosch and her colleagues (Heider, 1971; Mervis et al., 1975) have found that before 3- to 4-year-olds use basic color names they regularly select focal hues to represent them and that basic color names establish themselves and stabilize around focal colors. These studies indicate that the color system meets Miller’s primary prerequisites. Further, there is no rationale in this system for children’s perceiving colors and knowing color names but simultaneously misnaming in particular patterns; nor is it clear why individuals or the sexes might differ systematically in achieving proper color naming. Can a single explanation account comprehensively for the several characteristics of early color naming? Neurological

Maturation

and Integration

The perceptual salience, learning, and cognitive views of the development of color naming assume that young children are capable of acquiring names for colors, linking referents to semantics, and associating colors to color names correctly. These basic assumptions can be challenged, however, by contrary evidence that suggests that, although very young

EARLY

DEVELOPMENT

OF

COLOR

NAMING

83

children can discriminate, match, and sort colors accurately-and even categorize colors-and although they can associate environmental attributes with names-and even name colors-they may still not be able actually to associate specific chromatic percepts with specific color names correctly or consistently. Visual-verbal associative abilities rest on the integration of specific cortical structures, and a heretofore unexamined developmental hypothesis of color naming is that neurological maturation may act as a rate-limiting factor in ontogeny. On this explanation, color naming peculiarities in very young children might be conceived of as a linguistic symptom manifesting an underlying neurological deficit. Color agnosia is the neuropathological inability to attach consistent and correct nominal identifications to sensory impressions specifically in the color domain, and this condition is known to coexist with normal capacities both to discriminate colors and to use color names. Color agnosia is recognized to have identifiable neurological causes. Although structure-function relations are perforce speculative, it is natural in developmental study to suppose that maturation of the neural substrate enables cognitive skills. The systematic similarities between neuropathological symptomatology on the one hand and normal developmental deficiencies on the other, combined with the plausibility of a maturational perspective, encourage us to explore this possibility further. What is color agnosia? And, can color agnosia account for the diverse and anomalous characteristics of young children’s color naming? Cases of specific color agnosia in adults were reported from time to time in the early neurological literature (e.g., Wilbrand, 1887), although color aphasia as a syndrome separate from an underlying perceptual deficit was regarded with suspicion (Critchley, 1965) until recently. Now, two syndromes involving specific color dysnomia have been distinguished (Oxbury, & Humphrey, 1969). In both types, patients name the colors they see and point to colors that are named for them inconsistently and incorrectly; their color-name associations are often perseverative or random. These symptoms obtain even though both clinical types possess normal color vision-that is, they discriminate Ishihara plates or perform well on other tests of color perception, and they match, sort and categorize, order and arrange, and recognize colors correctly. Both types also know color names and uniformly describe or respond to colors with color names. One syndrome may be a pure aphasic form, and the other a consequence of hemispheric disconnection. The pure aphasic version is believed to result from primary damage to cortical structures specifically involved in color naming, whereas the visual-verbal variety is believed to be a secondary result of anatomical separation of cortex that subserves visual input from cortex that subserves color language. In adult patterns, both types appear to be complicated by some left-hemisphere damage and

84

MARC

H. BORNSTEIN

attendant right visual-field hemianopia. Kinsboume and Warrington (1964), Wyke and Holgate (1973), Netley (1974), and Basso, Faglioni, and Spinnler (1976) have described aphasic cases; Geschwind and Fusillo (1966) and Stachowiak and Poeck (1976) have described disconnection syndromes; and Oxbury et al. (1969) have described examples of both. These syndromes are characterized by “remarkable specificity” of the color-naming defect (Goodglass, 1980; Oxbury et al., 1969), although to be sure the defect is not complete or total. That is, error rates in naming colors seen or in identifying colors named in both clinical types are reported to vary between 33 and 66%, rather than to achieve 100%. Further, the aphasia may be distinguished from the disconnection syndrome in that color aphasics frequently cannot name from memory the correct color that a thing is (Oxbury et al., 1969). Young children’s peculiar color naming resembles these syndromes in detail. But what evidence is there to support the proposition that very young children might actually be color agnosic? Consider the ways in which color perception and naming function, and then how differential maturation and integration of cerebral structures concerned with perception and naming of colors might play a role. Color is processed along the retinogeniculostriate to prestriate to inferotemporal pathway of the visual system. Along this course chromatic analysis refines and specifies considerably. Whereas broad-band “cones” and “color-opponent” cells have been identified at the retina and lateral geniculate nucleus, respectively, more narrowly tuned “multiple-color” and “color-specific” cells have been identified in striate and prestriate (V4) cortex, respectively (DeValois & DeValois, 1975; Gouras, 1974; Zeki, 1980). Multiple-color and colorspecific cell types differ in capacity and service in important ways: Multiplecolor cells contribute to discrimination of pure color change (even in the absence of brightness differences) but are not excited by particular colors; as DeValois and DeValois (1975) observed, these cells “discriminate colors yet cannot tell which color is which” (p. 142). By contrast, colorspecific cells, further upstream, are sensitive to and identify particular narrow-band wavelength ranges. The existence of functionally distinct multiple-color and color-specific cell types segregated to different cortical regions suggests that color discrimination and correct and consistent color identification or naming could co-occur independently, a precondition for color agnosia. The syndrome then manifest would include awareness of the dimension of color, ability to discriminate color differences, and application of color differences to detect form, all in parallel with the use of color names, but it would also include impairment of the capacity to associate specific color names with specific colors. Since cortical maturation continues at least until puberty (e.g., Conel, 1939-1959; Kinsboume & Hiscock, 1983; Lenneberg, 1%7; Peiper, 1961/1963; Trevarthen, 1974), it is possible that the early immaturity of cerebral structures in

EARLY

DEVELOPMENT

OF COLOR NAMING

85

striate and prestriate cortex manifests itself in such a “color agnosia.” On this account, young children could discriminate colors and name color names, but not name colors or point to named colors correctly or consistently; that is, color cum name activities would be perseverative or random, until neurological maturation were achieved. An alternative possibility related to the disconnection syndrome echoes the normally slow maturation of interhemispheric communication in childhood. Studies of unilateral cerebral damage in patients (e.g., Boller & Spinnler, 1967; Scotti & Spinnler 1970) and tachistoscopic assessments in normal individuals (e.g., Pennal, 1977) have tended to center color discrimination in the right cerebral hemisphere, and studies of commisurotomy (e.g., Trevarthen & Sperry, 1973) and unilateral cerebral damage in patients (e.g., Albert, Reches, & Silverberg, 1975; Stachowiak & Poeck, 1976) have tended to center color naming in the left cerebral hemisphere (see, too, De Renzi & Spinnler, 1967). Lesions in the left occipital cortex and splenium usually occasion disconnection to the right hemisphere visual system from the angular gyrus, a significant verbalvisual association area (e.g., Vincent, Sadowsky, Saunders, & Reeves, 1977), and this dissociation is almost invariably accompanied by color dysnomia. In early normal development, late maturation of the splenium of the corpus callosum between the right visual area of cortex and dorsal and ventral association cortex would mean functional “disconnection” of the angular gyrus from visual input (Innocenti & Caminiti, 1980; Kinsbourne & Hiscock, 1983; Salamy, 1978; Trevarthen, 1974; Yakovlev & LeCours, 1967). Indeed, it has been hypothesized that the two hemispheres are thus anatomically disssociated early in life: In summary, then, the ordering of the developing cerebrum is of such a nature that it allows our hypothesis that the young infant is to some extent a “splitbrain.” This would suggest that interhemispheric communication is slight at birth, and increases with age, with good communication seen for all conditions starting around the ages of two to three. (Gazzaniga, 1970, pp. 131-132)

On this account, young children may have intact color vision, and could learn color names and relations on a purely verbal basis (even as Marmor (1978) has shown that the blind learn them in the absence of color experiences), but color vision and color naming would coexist independently resulting in color dysnomia. As callosal structures and cerebral association areas mature and integrate, and the hemispheres functionally interconnect, color vision and color naming, among other capacities, are brought into congruent association. There are of course many unknowns to detract from this otherwise attractive hypothesis, not the least of which is the problematic nature of brain-behavior relationships. Nevertheless, the neurological hypothesis accounts reasonably well for the six principal characteristics of color

86

MARC H. BORNSTEIN

naming in young children. (i-ii) The hypothesis admits that young children can discriminate among colors, categorize colors qualitatively, and possess awareness of a dimension of color. (iii) The hypothesis also accepts that young children may use color names. The neurological hypothesis states, however, (iv) that associations between specific color names and specific chromatic stimuli would be impaired in early life. In particular, it suggests that very young children’s spontaneous color naming ought to be incorrect and inconsistent, as we have seen it is. Moreover, color-word associations ought to be difficult for children to learn. This is a key symptomatological failure, and commensurately very young children have been found to experience considerable difficulty in learning to associate particular colors with particular names correctly and consistently (e.g., Binet, 1890; Darwin, 1877; Modreski & Goss, 1969). In an example study, I asked sixteen 3-year-olds to learn both shape-name and color-name associations to the same criterion. Stimuli in the two domains were equally simple, and equally arbitrary proper names were used as paired associates. Sets of names and task order were completely counterbalanced, and children were observed by experimentally blind testers. Significantly more children learned shape naming than learned color naming; children learned shape-name pairs almost twice as quickly on average as color-name pairs; and children made onethird more errors in learning color-to-name associations than in learning shape-to-name associations (Bornstein, 1985b). As if to capture this very spirit, at the turn of the century Nagel (1906) termed early childhood “Farbendummheit.” The neurological hypothesis could also account for (v) the lower age limit as well as for (vi) wide individual differences in children’s eventually achieving correct and consistent color naming. Four years is, as we have seen, roughly the minimal age for accurate color naming. Gazzaniga (1970) speculated that 3 years would be a minimum for good interhemispheric communication, and pertinently, Salamy (1978) found that it was not possible to obtain reliable ipsilateral somatosensory-evoked potentials in children younger than 3.5 years. These times confirm the same approximate age minimum for development of corpus callosal transmission. Progressive myelogenesis afterwards helps to account for individual developmental differences.* Finally, the neurological hypothesis accommodates the finding that color naming develops consistently faster in girls than in boys, since females are generally more advanced in physical * This hypothesis also helps to explain why Schiller’s (1966) 1st graders may have experienced the least color-word interference in naming colors, and why Svinicki et aL’s (1976) 3-year-olds were unintluenced in color identification by experimenter-provided color names. Both studies indicate a tendency in young children to dissociate perceiving from naming colors.

EARLY

DEVELOPMENT

OF

COLOR

NAMING

87

and neurological maturation (Butfery & Gray, 1972; Kinsbourne & Hiscock, 1983; Waber, 1977). After 4-5 years of age, most children associate at least the basic colors and their names correctly and consistently. In addition to its other virtues, the neurological hypothesis predicts, first, that when proper color-name associations emerge they ought to be achieved initially for focal colors, which match neurophysiological sensitivities, and only later for nonfocal colors, but, second, that focal color terms should not be acquired in any particular order. Several studies support these additional predictions. Children tend to know focal color terms before they know terms for nonfocal colors, and, despite some claims following Berlin and Kay (1969) that they might know or use particular color terms first, children seem to acquire basic color names in essentially arbitrary order (Heider, 1971; Johnson, 1977; Kirk et al., 1975; Nagel, 1906; Preyer, 1890; see too, Whitfield, 1981). Especially telling longitudinal studies of individual children fail to bear out any ordered acquisition among basic terms: Bartlett (1977), who punctiliously observed 33 children between 2f and 4 years of age, found that orders of acquisition and use of color terms were wholly idiosyncratic. If there are two color dysnomic syndromes, one aphasic and the other disconnection, can we tell whether children are normally color aphasic or, alternatively, that they normally suffer interhemispheric disconnection? Unfortunately, the two syndromes are not themselves altogether well differentiated in the adult neurological literature. There is no visuosensory deficit in either, as there is none in children, and patterns of color naming and color identification anomalies are similar in the two. Color dysnomia in both patient populations is also remarkably specific, and both show a high degree of language function outside their color naming defects; indeed, as Oxbury et al. (1969) observed, “color anomia is the only known form of specific anomia” (p. 856). Only the fact that patients with a disconnection syndrome can tell what color an object normally is separates them from aphasics who generally cannot. This difference would constitute a significant distinction and inform diagnosis in children; however, it is not failsafe. First, the dearth of general color knowledge in the aphasic type is not total; Oxbury et al. (1969) reported that aphasics cannot name the color an object normally is “most of the time,” that is, some of the time they can. Second, the fact that very young children may not always genuinely know what colors things normally are further confounds differentiation on this basis. Unfortunately, the specificity of color dysnomia complicates diagnosis further since children, like clinical patients, could well be normatively color dysnomic in the absence of other aphasias. We can only hypothesize that the color naming anomalies of childhood reflect a lack of neurological integrity. Only additional developmental study, based on systematic adult

88

MARC H. BORNSTEIN

work, might disambiguate whether children in their immature state are more like one or the other of these syndromes.

neurological

CONCLUSIONS Color naming has attracted the attention of physicists, physiologists, psychologists, linguists, anthropologists, and philosophers because much is known about the physics of color and the physiology of color vision, because color names represent a small and circumscribed lexical set, and because the linguistics and the physics of color ought to correspond. Indeed, color naming has recommended itself on these bases as a nearly prototypic domain in which to explore the nexus of brain and language. Certainly, many developmental neuropsycholinguists have regarded the color question in this light, viz., as a staging area in which to explore how children master semantics. This review has endeavored to show that the normative development of color naming is unfortunately not straightforward, but in fact rather convoluted and complex. Theories why this is so are plentiful and themselves more complicated than previously suspected. Color naming is not unique, however, in being developmentally problematic; many other nominalizations of sensations and emotions may be late in developing or may follow peculiar ontogenetic courses. Color naming is singular in that the perceptuocognitive apparatus on which it depends is fully functional in infancy, so that the anomalous ontogeny of color naming is surprising. The normal development of color naming is described by six salient characteristics. Among several theories offered to account for them, including perceptual salience, learning, and cognition, a speculative hypothesis related to neurological maturation and integration acquits itself well. In brief, two variants of color dysnomia have been identified in adult neuropathology, one the consequence of purely verbal aphasia, and the other the result of a verbal-visual disconnection syndrome. Patients who suffer these disabilities possess intact visuosensory functions and a color vocabulary, but they name and identify colors in perseverative or random ways. Very young children show formal parallels to these dysnomias, and the possibility is explored that normal neurological immaturity entails one or another condition. REFERENCES Albert, M. L., Reches, A., & Silverberg, R. 1975. Hemianopic colour blindness. Journal of Neurology,

Neurosurgery,

and Psychiatry,

38, 546-549.

Anyan, W. R., & Quillian, W. W., II., 1971. The naming of primary colors by children. Child Development, 42, 1629-1632. Aristotle 1913. De coloribus. In W. D. Ross (Ed.), The works of Aristotle (T. Loveday & E. S. Forster, trans.). London/New York: Oxford Univ. Press. Clarendon. Vol. 6. Baldwin, J. M. 1893. A new method of child study. Science (Washington, D.C.), 21, 213214.

EARLY

DEVELOPMENT

OF COLOR NAMING

89

Baldwin, J. M. 1895. Mental development in the child and the race. New York: Macmillan. Bartlett, E. J. 1977. The acquisition of the meaning of color terms: A study of lexical development. In P. Smith & R. Campbell (Eds.), Proceedings ofthe NATO conference on the psychology of language. New York: Plenum. Basso, A., Faglioni, P., & Spinnler, H. 1976. Non-verbal colour impairment of aphasics. Neuropsychologia, 14, 183-193. Bateman, W. G. 1915. The naming of colors by children: The Binet test. Pedagogical Seminary, 22, 469-486. Beare, A. C. 1%3. Color-name as a function of wave-length. American Journal of Psychology, 76, 248-256. Berlin, B., & Kay, P. 1969. Basic color terms: Their universality and evolution. Berkeley: Univ. of California Press. Binet, A. 1890. Perceptions d’enfants. Revue Philosophique, 30, 582-611. Binet, A., & Simon, T. 1908/1916. [The development of intelligence in the child] (E. S. Kiffe, trans.). In H. H. Goddard (Ed.), The development of intelligence in children. Baltimore: Williams & Wilkins. Boller, F., & Spinnler, H. 1967. Visual memory for colors in patients with unilateral brain damage. Cortex, 3, 395-405. Bornstein, M. H. 1973. Color vision and color naming: A psychophysiological hypothesis of cultural difference. Psychological Bulletin, 80, 257-285. Bornstein, M. H. 1975a. The influence of visual perception on culture. American Anthropologist, 77, 774-798. Bornstein, M. H. 1975b. Qualities of color vision in infancy. Journal of Experimental Child Psychology, 19, 401-419. Bornstein, M. H. 1978. Chromatic vision in infancy. In H. W. Reese & L. P. Lipsitt (Eds.), Advances in child development and behavior. New York: Academic Press. Vol. 12. Bornstein, M. H. 1979. Effects of habituation experience on posthabituation behavior in young infants: Discrimination and generalization among colors. Developmental Psychology, 15, 348-349. Bomstein, M. H. 1981. Psychological studies of color perception in human infants: Habituation, discrimination and categorization, recognition, and conceptualization. In L. P. Lipsitt (Ed.), Advances in infancy research. Norwood, NJ: Ablex. Vol. 1. Bornstein, M. H. 1985a. Infant into adult: Unity to diversity in the development of visual categorization. In J. Mehler & R. Fox (Eds.), Neonate cognition: Beyond the blooming, buzzing confusion. Hillsdale, NJ: Erlbaum. Bomstein, M. H. 1985b. Color-name versus shape-name learning in young children. Journal of Child Language, in press. Bornstein, M. H., Kessen, W., & Weiskopf, S. 1976. Color vision and hue categorization in young human infants. Journal of Experimental Psychology: Human Perception and Performance, 2, 115-129. Boynton, R. M. 1975. Color, hue, and wavelength. In E. C. Carterette & M. P. Friedman (Eds.), Handbook of perception, New York: Academic Press. Vol. 5. Boynton, R. M., & Gordon, J. 1965. Bezold-Brticke hue shift measured by color-naming technique. Journal of the Optical Society of America, 55, 78-86. Brian, C. R., & Goodenough, F. L. 1929. The relative potency of color and form perception at various ages. Journal of Experimental Psychology, 12, 197-213. Buffery, A. W. H., & Gray, J. A. 1972. Sex differences in the development of spatial and linguistic skills. In C. Ounsted & D. C. Taylor (Eds.), Gender differences: Their ontogeny and significance. London: Churchill. Bush, H. D. 1914. The vocabulary of a three-year-old girl. Pedagogical Seminary, 21, 125-142.

90

MARC H. BORNSTEIN

Carey, S. 1980. The child as a word learner. In M. Halle, S. Bresnan, & G. A. Miller (Eds.), Linguistic theory and psychological reality. New York: Plenum. Carter, R. L., Hohenegger, M. K., & Satz, P. 1982. Aphasia and speech organization in children. Science (Washington, D.C.), 218, 797-799. Cemacek, J., & Podivinsky, F. 1971. Ontogenesis of handedness and somatosensory cortical response. Neuropsychologia, 9, 219-232. Cohen, L. B., & Gelber, E. R. 1975. Infant visual memory. In L. Cohen & P. Salapatek (Eds.), Infant perception: From sensation to cognition. New York: Academic Press. Colby, M. C., & Robertson, J. B. 1942. Genetic studies in abstraction. Journal of Comparative Psychology, 33, 385-401. Conel, J. L. 1939-1959. The postnatal development of the human cerebral cortex. Cambridge, MA: Harvard Univ. Press. Vols. l-6. Conklin, H. C. 1973. Color categorization. American Anthropologist, 75, 931-942. Conrad, R. 1972. Form and color as short-term memory codes in preschool children. Psychonomic Science, 27, 225-226. Cook, W. M. 1931. Ability of children in color discrimination. Child Development, 2, 303320. Critchley, M. 1%5. Acquired anomalies of colour perception of central origin. Brain, 88, 71 l-724. Cruse, D. A. 1977. A note on the learning of colour names. Journal of Child Language, 4, 305-311. Dale, P. S. 1969. Color naming, matching, and recognition by preschoolers. Child Development, 40, 1135-1144. Darwin, C. H. 1877. [A biographical sketch of a young infant.] Kosmos, 1, 367-376. De Renzi, E., & Spinnler, H. 1967. Impaired performance on color tasks in patients with hemispheric damage. Cortex, 3, 194-217. Descoeudres, A. 1921. Le de’veloppement de l’enfant de deux d sept ans. Paris: Editions Delachaux et Niestlt. DeValois, R. L., & DeValois, K. K. 1975. Neural coding of color. In E. C. Carterette & M. P. Friedman (Eds.), Handbook of perception. New York: Academic Press. Vol. 5. Dodd, C., & Lewis, M. 1969. The magnitude of the orienting response in children as a function of changes in color and contour. Journal of Experimental Child Psychology, 8, 296-305. Ekman, G. 1963. Contributions to the psychophysics of color vision. Studium Generale, 16, 54-64. Famham-Diggory, S., & Gregg, L. W. 1975. Color, form, and function as dimensions of natural classification: Developmental changes in eye movements, reaction time, and response strategies. Child Development, 46, 101-114. Gazzaniga, M. 1970. The bisected brain. New York: Appleton-Century-Crofts. Geschwind, N., & Fusillo, M. 1966. Color-naming defects in association with alexia. Archives of Neurology, 15, 137-146. Gesell, A., Ilg, F. L., & Bullis, G. E. 1949. Vision: Its development in infant and child. New York: Harper L Row (Hoeber). Gladstone, W. E. 1858. Studies on Homer and the Homeric age. London/New York: Oxford Univ. Press. Goddard, H. H. 1911. Two thousand normal children measured by the Binet Scale of Intelligence. Pedagogical Seminary, 18, 232-259. Goodenough, F. L., &Anderson, J. E. 1931. Experimental child study. New York: Century. Goodglass, H. 1980. Disorders of naming following brain inquiry. American Scientist, 68, 647-655. Gouras, P. 1974. Opponent-colour cells in different layers of foveal striate cortex. Journal of Physiology, 238, 583402. Graham, C. H. (Ed.). 1965. Vision and visual perception. New York: Wiley.

EARLY

DEVELOPMENT

Heider, E. R. 1971. “Focal” Psychology,

91

OF COLOR NAMING

color areas and the development of color names. Developmental

4, 447-455.

Heilig, M. R. 1913. A child’s vocabulary. Pedagogical Seminary, 20, 1-16. Hering, E. 1878/1964. [Outlines ofa theory ofthe light sense] (L. M. Hurvich & D. Jameson, trans.). Cambridge, MA: Harvard Univ. Press. Innocenti, G. M., 8r Caminiti, R. 1980. Postnatal shaping of callosal connections from sensory organs. Experimental Brain Research, 39, 381-394. Istomina, Z. M. 1963. Perception and naming of color in early childhood. Soviet Psychology and Psychiatry,

1, 37-45.

Johnson, E. G. 1977. The development of color knowledge in preschool children. Child Development, 48, 308-311. Karpf, R. J., Goss, A. E., & Small, M. Y. 1974. Naming, selection, and ordering of color (“hue”) by young children. Journal of General Psychology, 90, 297-314. Kimball, M. M., & Dale, P. S. 1972. The relationship between color naming and color recognition abilities of preschoolers. Child Development, 43, 972-980. Kinsboume, M., & Hiscock, M. 1983. The normal and deviant development of functional lateralization of the brain. In P. H. Mussen (Ed.), Handbook of child psychology, Vol. II, M. M. Haith &J. J. Campos (Eds.), Infancy anddevelopmentalpsychobiology. New York: Wiley. Kinsboume, M., & Warrington, E. K. 1964. Observations on colour agnosia. Journal of Neurology,

Neurosurgery,

and Psychiatry,

21, 296-299.

Kirk, G. E., Hunt, J. McV., & Lieberman, C. 1975. Social class and preschool language skills: II. Semantic mastery of color information. Genetic Psychology Monographs, 91, 299-316. Koffka, K. 1927.

The growth of the mind: An introduction to child psychology. New York: Harcourt, Brace. Lee, L. C. 1965. Concept utilization in preschool children. Child Development, 36, 221228. Lenneberg, E. H. 1967. Biological foundations of language. New York: Wiley. Leopold, W. F. 1939. Speech development of a bilingual child. Evanston, IL: Northwestern Univ. Press. Lewis, M., & Baumel, M. H. 1970. A study in the ordering of attention. Perceptual and Motor

Skills,

31, 979-990.

Mcnamara, J. 1972. Cognitive basis of language learning in infants. Psychological Review, 79, 1-13. Maerz, A., & Paul, M. R. 1930. Dictionary of color. New York: McGraw-Hill. Marmor, G. S. 1978. Age at onset of blindness and the development of the semantics of color names. Journal of Experimental Child Psychology, 25, 267-278. Mervis, C. B., Catlin, J., & Rosch, E. 1975. Development of the structure of color categories. Developmental Psychology, 11, 54-60. Miller, G. A. 1977. Spontaneous apprentices. New York: Seabury. Miller, G. A., & Johnson-Laird, P. N. 1976. Language and perception. Cambridge, MA: Harvard Univ. Press. Modreski, R. A., & Goss, A. E. 1969. Young children’s initial and changed names formcolor stimuli. Journal of Experimental Child Psychology, 8, 402-409. Nagel, W. A. 1906. Observations on the color-sense of a child. Journal of Comparative Neurology

and Psychology,

16, 217-230.

Nelson, K. 1974. Concept, word, and sentence: Interrelations in acquisition and development. Psychological

Review,

81, 269-285.

Netley, C. 1974. Colour aphasia: A case report. Cortex, 10, 388-394. Newton, I. 1671-1672. New theory about light and colors. Philosophical the Royal

Society,

Transcations

of

80, 3075-3087.

Norsworthy, N., & Whitley, M. T. 1920. The psychology

of childhood.

New York: Macmillan.

MARC H. BORNSTEIN

92

Odom, R. D. 1972. Effects of perceptual salience on the recall of relevant and incidental dimensional values: A developmental study. Journal of Experimental Psychology, 92, 285-291.

Odom, R. D., & Guzman, R. D. 1972. Development of hierarchies of dimensional salience. Developmental

Psychology,

6, 271-287.

Oxbury, J. M., Oxbury, S. M., & Humphrey, Brain,

N. K. 1969. Varieties of colour anomia.

92, 847-860.

Peiper, A. 1%1/1963. Cerebralfunction in infancy and childhood. New York: Consultants Bureau. Pelsma, J. R. 1910. A child’s vocabulary and its development. Pedagogical Seminary, 17, 328-369.

Pennal, B. E. 1977. Human cerebral asymmetry in color discrimination.

Neuropsychologia,

15, 563-568. Preyer, W. 1890.

The mind of the child. New York: Appleton. Quine, W. V. 1973. The roots of reference. LaSalle, IL: Open Court, Raskin, L. A., Maital, S., & Bornstein, M. H. 1983. Perceptual categorization of color: A life-span study. Psychological Research, 45, 135-145. Rivers, W. H. R. 1901. Vision. In A. C. Haddon (Ed.), Reports on the Cambridge anthropological expedition to the Torres Straits. Cambridge: Cambridge Univ. Press. Vol. 2. Rowe, E. C., & Rowe, H. N. 1913. The vocabulary of a child at four and six years of age. Pedagogical Seminary, 20, 187-209. Salamy, A. 1978. Commissural transmission: Maturational changes in humans. Science (Washington, D.C.), 200, 1409-1411. Schiller, P. H. 1966. Developmental study of color-word interference. Journal ofExperimental Psychology, 72, 105-108. Scotti, G., & Spinnler, H. 1970. Colour imperception in unilateral hemisphere-damaged patients. Journal of Neurology, Neurosurgery, and Psychiatry, 33, 22-28. Segall, M. H., Campbell, D. T., & Herskovits, M. J. 1966. The influence of culture on visual perception. Indianapolis, IN: Bobbs-Merrill. Shinn, M. W. 1909. Notes on the development of a child. Berkeley: Univ. of California Press. Stachowiak, F.-J., & Poeck, K. 1976. Functional disconnection in pure alexia and color naming deficit demonstrated by facilitation methods. Brain and Language, 3, 135143. Stem, W. 1924. Psychology of early childhood. New York: Holt. Suchman, R. G., & Trabasso, T. 1966. Color and form preference in young children. Journal

of Experimental

Child

Psychology,

3, 177-187.

Svinicki, J. G., Meier, S., & Svinicki, M. D. 1976. Stimulus label and generalization in children as a function of age. Journal of Experimental Child Psychology, 21, 282-288. Synolds, D. L., & Pronko, N. H. 1949. An exploratory study of color perception of children. Journal

of Genetic

Psychology,

74, 17-21.

Teller, D. Y., & Bomstein, M. H. 1985. Color vision and color perception in infancy. In L. B. Cohen & P. Salapatek (Eds.), Handbook of infant perception. New York: Academic Press. Terman, L. M. 1916. The measurement of intelligence. Boston: Houghton Mifflin. Tomikawa, S. A., & Dodd, D. H. 1980. Early word meanings: Perceptually or functionally based? Child Development, 51, 1103-l 109. Tracy, F., & Stimpfl, J. 1909. The psychology of childhood. Boston: Heath. Trevarthen, C. 1974. Cerebral embryology and the split brain. In M. Kinsboume & L. Smith (Eds.), Hemisphere disconnection and cerebral function. Springfield, IL: Thomas. Trevarthen, C., & Sperry, R. W. 1973. Perceptual unity of the ambient visual field in human commissurotomy patients. Brain, 96, 547-570.

EARLY

DEVELOPMENT

93

OF COLOR NAMING

Vernon, M. D. 1962. The psychology of perception. Baltimore: Penguin. Vincent, F. M., Sadowsky, C. H., Saunders, R. L., & Reeves, A. G. 1977. Alexia without agraphia, hemianopia, or color-naming defect: A disconnection syndrome. Neurology, 27, 689-691. Waber, D. P. 1977. Sex differences in mental abilities, hemispheric lateralization, and rate of physical growth in adolescence. Developmental Psychology, 13, 29-38. Werner, H. 1940. The comparative psychology of mental development. New York: Harper. Whipple, G. M. 1909. The vocabulary of a three-year-old boy. Pedagogical Seminary, 16, l-22. Whitfield, T. W. A. 1981. Salient features of color space. Perception & Psychophysics, 29, 87-90.

Whorf, B. L. 1952. Collected papers on metalinguistics. Washington, DC: Foreign Service Institute. Wilbrand, H. 1887. Die Seelenblindheit als Herderscheinung. Weisbaden: Bergmann. Winch, W. H. 1909. Colour preferences of school children. British Journal ofPsychology, 3, 42-65.

Wohlwill, Journal

J. F. 1957. The abstraction and conceptualization of Experimental

Psychology,

Wolfe, H. K. 1890. On the color-vocabulary

of form, color, and number.

53, 304-309.

of children. Nebraska

University

Studies,

1,

205-234.

Wyke, M., & Holgate, D. 1973. Colour-naming defects in dysphasic patients. A qualitative analysis. Neuropsychologia, 11, 451-461. Yakovlev, P. I., & LeCours, A. R. 1967. The myelinogenetic cycles of regional maturation of the brain. In A. Minkowski (Ed.), Regional development of the brain. Oxford: Backwell. Yendovitskaya, T. V., Zinchenko, V. P., & Ruzskaya, A. G. 1964/1971. [Development of sensation and perception] (J. Shybut & S. Simon, trans.). In A. V. Zaporozhets & D. B. Elkonin (Eds.), The psychology ofpreschool children. Cambridge, MA: MIT Press. Zeki, S. 1980. The representation of colours in the cerebral cortex. Nature (London), 284, 412-418.