Basic BLUE in East Slavonic(1).
AbstractRussian's second BLUE term goluboj `light blue' constitutes a well-known exception to the Berlin and Kay basic color-term typology. If other Slavonic languages do not have a second BLUE term, then the special position of Russian requires explanation; if there is evidence pointing to two basic terms for BLUE, we have a second set of data for investigating the evolution of this unusual color system. The languages genetically closest to Russian, Ukrainian and Belarusian, are examined. Findings of the list task, a simple and elegant test for psychological salience of color terms, provide strong evidence that Ukrainian and Belarusian have also evolved a second BLUE. What is interesting is that the term is not the general East Slavonic term that Russian uses, but a Polish borrowing: blakytnyj `light blue' (Ukrainian), blakitny `light blue' (Belarusian). This opens up the possibility that a category, in this case the second BLUE, may be borrowed separately from the basic term that denotes it. Since the category has been borrowed separately from its term, we have evidence of a different kind that the Russian color system includes a second BLUE: the category is salient enough to be borrowed separately from the term that denotes it.
Introduction
Berlin and Kay's (1969) basic color-term typology, given in Figure 1, has prompted much discussion about color categories among anthropologists, linguists, and psychologists.(2) According to the theory, there is a maximum of eleven basic color terms, and their emergence is universally highly constrained: a term for RED in a language will imply the presence of WHITE and BLACK, etc.(3) The typology, with significant modifications (Kay and McDaniel 1978; Kay et al. 1991; and Kay et al. 1997), has proved remarkably robust.
[Figure 1 ILLUSTRATION OMITTED]
The hierarchy is a constraint on the evolution of basic color terms in a language. Diachronically, languages evolve through stages 1 to 7. Hence a stage 5 language with a BLUE term must have emerged from a stage 4 language that lacked a BLUE but had terms for WHITE, BLACK, RED, YELLOW, and GREEN. (This language will in turn evolve to stage 6, where a basic term for BROWN will be added.) The evolution of color categories is special in that it is monotonic: once a category "opens" in a language, it cannot then subsequently "close." Thus a given language can be assigned to a stage on the hierarchy based on the categories it has, and the categories it lacks.
A well-known problem case for the typology, which was commented on in the original Berlin and Kay monograph, is the Russian color system. It appears that Russian has developed a category that is absent from the typology, the term goluboj `light blue'. In later work Corbett, Davies, and Morgan, using a number of recognized psycholinguistic tests, established the basic status of goluboj.(4) Once it has been established that the Russian color system is unusual, a natural next step is to explore the evolution of the Russian system in the context of its broader language family. We ask the question whether this innovative BLUE category is unique to Russian, or whether it is a property of a color system operating at a higher level, that is, whether it is a broader characteristic of Russian's family Slavonic. If other Slavonic languages do not have a second BLUE term, then the special position of Russian requires explanation. And if there is evidence pointing to two basic terms for BLUE, we have a second set of data for investigating the evolution of this unusual color system?
In section 1 we discuss the notion of basic color term, and how to establish which of the many terms in a language's color vocabulary are the basic ones. Of particular interest is the test for psychological salience of a term. The revised basic color-term typology of Kay and McDaniel (1978) and Kay et al. (1997) is then briefly outlined. We then present the data on Slavonic basic color terms (section 2). We restrict our study to East Slavonic, the branch to which Russian belongs, which is genetically close. The "list task" (a test of psychological salience) was carried out on around thirty native speakers of Ukrainian and Belarusian. In section 3, we discuss these results in the context of what has been established for Russian, particularly concerning the BLUE area.
1. Basic color terms
Of the set of terms denoting colors in a language, there is an identifiable subset that could be described as the "basic" set of terms. The idea of "basic object" belonging to a "basic level of categorization" is found in the cognitive-psychology literature. With color terms, the basic level of categorization follows from physiological reasons, from which constraints on possible categorization can be imposed. Moreover, these determine the evolutionary path a language takes in its development of basic color terms.
1.1. Basic color terms and psychological salience
Humans are categorizors, and objects in the real world are understood partly in terms of what category they fall into. Categories themselves are organized into taxonomies such as the one found in Figure 2. The category LEOPARD is superordinate in relation to SNOW LEOPARD and subordinate in relation to FELINE, which in turn is subordinate to the superordinate category MAMMAL.
[Figure 2 ILLUSTRATION OMITTED]
What is interesting from a cognitive point of view is that within such taxonomies there appears to be some "basic level" of categorization where divisions between members and nonmembers are made most naturally. It is at this level that members of the same category are maximally similar to one another on the one hand, but maximally dissimilar to objects belonging to other categories at that level on the other. In our small animal taxonomy, the basic level is indicated by arrows. Members of the basic level category CAT share many more attributes than members of the superordinate category FELINE. At the same time, the attributes they have in common with members of categories at the same level of categorization, for example leopards and dogs, are few enough so that distinctions are easy to make between members and nonmembers. However, members of subordinate categories such as BURMESE and SIAMESE, while sharing even more attributes, share too many attributes to make obvious distinctions. The idea of "basic level of categorization" and "basic object" (a member of a basic-level category) comes from work in cognitive psychology experiments by Eleanor Rosch and her collaborators, among others.(6)
In the area of color terms, we can also assume that there is a subset of the available terms that could be viewed as the basic terms. Thus in English, among the terms RED, RUST, SCARLET, GINGER, BLUE we pick out RED and BLUE as the basic terms. Both appear to be at a level where it is both natural to identify members of a category (for example types of RED) and contrast them with members of another category at that level (for example types of BLUE). Berlin and Kay (1969: 6-7), working with the idea of basic color term, provide a list of criteria that can be used to characterize the basic terms. First, the term must be shown to be monolexemic, that is, the meaning is not derivable from the sum of its parts. This would rule out sky blue as a candidate for basic status. Second, the color it signifies must not be included in the signification of another basic term. The term scarlet, a kind of RED, cannot be basic. Third, it must apply generally and not be restricted to a limited number of objects, as is the case with blond and ginger, which denote hair color. The criteria, with examples taken from Russian, are given in Table 1.
Table 1. Criteria for basicness Criterion Example of nonbasic term Monolexemic jarko-zelenyj `bright green' Not included in other term alyj `scarlet' Not restricted karij `dark brown (for eyes)'
A fourth test would be whether or not it is psychologically salient. In determining whether or not a color term is basic, it is assumed that there is a correlation between basicness and psychological salience. Evidence for a term being psychologically salient is its prominence in an elicited list, its occurrence in the ideolects of all informants, and the stability of its reference across informants (Berlin and Kay 1969: 6). Tests for psychological salience fall into two broad categories (Corbett and Davies 1997). Behavioral tests include color-naming tasks and color-eliciting tasks. Linguistic tests include textual frequency of the terms and the size of a term's derivational family. The data we present are the result of one of the behavioral tests, the "list task,, where color terms are elicited by asking informants to list as many color terms as they can think of within a specific stretch of time. The frequency of occurrence of a color term across informants and the order in which it occurs on the questionnaires are used as measures of the term's basic status. Higher frequency and greater prominence in the ordering correspond to greater likelihood that the term is basic.
1.2. Constraints on color categorization and the evolution of basic color terms
The physiological mechanisms that distinguish different wavelengths of light, and the interpretation of these distinctions as various color sensations,(7) are assumed to operate similarly across humans. Consistent across the species is the "automatic registration," or perception, of the six elemental colors WHITE, BLACK, RED, GREEN, YELLOW, and BLUE on which theories of universal color categorization rest (MacLaury 1991b: 42). Thus with color categorization, at any rate, there appears to be good physiological grounds for categorization taking the course that it does, and for the same basic categories emerging cross-linguistically. The original Berlin and Kay basic color term hierarchy in Figure 1 has been revised in light of further studies, notably Kay and McDaniel (1978). Instead of viewing category evolution as the successive lexical encoding of new foci, it is taken to be "the progressive differentiation of color categories" (Kay and McDaniel 1978:617). Early stages contain color composites such as the warm composite WHITE/RED/YELLOW, containing multiple foci. Stages evolve along the lines of differentiation, such that a subsequent stage will distinguish WHITE from RED/ YELLOW, and a stage subsequent to this one will distinguish RED and YELLOW. Fuzzy set theory is invoked to capture the fact that membership of a composite category is a matter of degree, where in our example of WHITE/RED/YELLOW focal white, focal red, and focal yellow have the highest degree of membership, and colors "falling in between" the foci have the lowest degree. Evolution of basic color categories progresses along the lines of decomposition of all composite categories into the six elemental ones, corresponding to the six fundamental neural response categories of white, black, red, green, yellow, and blue. The elemental colors are then combined to form compound colors, such as BLUE and RED to form PURPLE. Other compound colours are ORANGE (YELLOW and RED), BROWN (BLACK and YELLOW), PINK (RED and WHITE), and GRAY (BLACK and WHITE). Again fuzzy set theory is used to express this, where a compound category is the fuzzy intersection of two elemental categories, allowing for best and worst examples of the compound category (highest and lowest degree of membership of the category).
Figure 3 gives a version of the revised typology as described in Kay et al. (1997). The general claims are that a cool composite category (BLACK/BLUE/GREEN) is opposed to a warm composite category (WHITE/RED/YELLOW). This opposition of cool and warm takes place at stage 1. Evolutionary stages 2 to 5 deal with decomposition into the elemental colors WHITE, RED, YELLOW (the warm elemental colors) and BLACK, GREEN, and BLUE (the cool elemental colors). Note that bold type denotes changes that occur at the current stage. For example stage 2 records the emergence of the elemental category WHITE and the composite category RED/YELLOW, in addition to the unaffected composite category from the previous stage BLACK/ BLUE/GREEN. What is important for our discussion is how the elemental colors combine to form compound colors. Kay and McDaniel (1978: 638-641) assume that this occurs subsequent to all decomposition, and that BROWN is the first compound color to emerge at stage 6.(8) Important for our discussion is that Russian goluboj `light blue' is speculated to be the combination of the elementals BLUE and WHITE (just as PINK is the combination of RED and WHITE).
Figure 3. Revised Berlin and Kay [WHITE/RED/YELLOW (warm) BLACK/BLUE/GREEN (cool)] Stage 1 > [WHITE RED/YELLOW BLACK/BLUE/GREEN] Stage 2 > [GREEN BLACK/BLUE WHITE RED/YELLOW] Stage 3 > [RED YELLOW WHITE GREEN BLUE/BLACK] Stage 4 > [BLACK BLUE GREEN WHITE RED YELLOW] Stage 5 > [BROWN BLACK BLUE GREEN WHITE RED YELLOW] Stage 6 > [PURPLE ORANGE PINK BLACK BLUE GREEN WHITE RED YELLOW] Stage 7
Languages at the various stages have been identified in the World Colour Survey, and examples are given in Kay et al. (1997). For example, the Niger-Congo language Ejagham spoken in Nigeria and Cameroon is at stage 2, having a term for WHITE, ebare, a term for RED/YELLOW, ebi, and a term for BLACK/GREEN/BLUE, enyaga. On the other hand, Martu-Wangka, a Pama-Nyungan language (Australia) represents a stage 4 language: the warm colors have fully decomposed with piila-piila for WHITE, miji-miji for RED, and karntawarra for YELLOW. However, the cool colors have not completed their decomposition: there is a term for GREEN, yukuri-yukuri, but maru-maru covers both BLACK and BLUE.
2. The Ukrainian and Belarusian list tasks
We examine Russian's unique color system from a diachronic perspective. This entails an investigation of the basic color terms of the other East Slavonic languages, Russian's sister languages, Ukrainian and Belarusian. Slavonic is typically divided into three main branches: East Slavonic, which includes modern-day Russian, Ukrainian, and Belarusian, West Slavonic (Czech, Slovak, Polish, Cassubian, Lower Sorbian, and Upper Sorbian), and South Slavonic (Slovene, Serbo-Croat, Bulgarian, and Macedonian).(9) From around the fourteenth to fifteenth century onward, we can start talking of differentiation among three East Slavonic languages, though Ukrainian as a standard literary language was not adopted until the nineteenth century, and Belarusian not until the twentieth century (Schenker 1995: 74). There are forty-five million speakers of Ukrainian (five million outside of Ukraine), and about seven million speakers of Belarusian (Schenker 1995: 74).
Native speakers of Ukrainian and native speakers of Belarusian were asked to perform the list task as a test of the psychological salience of the color terms of the respective languages. Psychological salience is then used as an indicator of a term's basicness. Two measures were used, and compared. The first measure is the frequency with which a term appears across informants: higher frequency correlates with high salience. The second measure is the "height" of the term on the informant's list of terms. The closer to the top of the list, or the "higher" the term, the more salient the term should be. For a discussion of these measures for psychological salience, see Corbett and Davies (1997), who compare data on American, Japanese, and Russian color terms and statistically match the results with the Berlin and Kay hierarchy. We look first at the results of the Ukrainian list task, then turn to Belarusian.
2.1. Ukrainian list task
The list task was carried out by a native speaker of Ukrainian, who conducted the task in Ukrainian. Thirty-four informants who took part declared Ukrainian as their first language. The informants were from the University of Xarkiv, North East Ukraine.(10) The age group was 18 to 22 years, and all except one were female. About one hundred color terms were elicited. We first present results of frequency of occurrence of terms across informants, and then turn to the height on the list on which the terms are elicited. Recall that higher frequency and greater height on the list correspond to greater likelihood of basic status.
2.1.1. Frequency measure. The frequency order of terms occurring at least five times is given in Table 2. The first two columns give the term and gloss. The occurrences of the term across the thirty-four informants (also recorded separately as a percentage) is then given, and the terms are ranked in frequency order. The table is divided into frequency "zones," delimited by the white space.
Table 2. Rank frequency order of Ukrainian color terms (N = 34) Term Gloss Frequency Rank occurrences % zelenyj green 34 100 1.0 cervonyj red 33 97 2.0 bilyj white 32 94 3.5 cornyj black 32 94 3.5 synij dark blue 31 91 6.5 korycnjavyj(11) brown 31 91 6.5 siryj gray 31 91 6.5 rozevyj pink 31 91 6.5 zovtyj yellow 30 88 10.0 blakytnyj light blue 30 88 10.0 fioletovyj purple 30 88 10.0 salatnyj(12) light green 26 76 12.0 malynovyj raspberry 22 65 13.0 zovtoharjacyj orange 18 53 14.0 bordovyj(13) bordeaux 17 50 15.0 bezevyj(14) beige 12 35 17.0 birjuzovyj turquoise 12 35 17.0 vysnevyj cherry 12 35 17.0 purpurnyj(15) purple 11 32 19.0 oranzevyj orange 10 29 20.0 rudyj red (yellow) 9 27 21.0 holubyj light blue 8 24 23.5 kastanovyj chestnut 8 24 23.5 zolotyj(16) gold 8 24 23.5 burjakovyj beet 8 24 23.5 bolotnyj marsh green 7 21 26.0 kavovyj coffee 6 18 29.0 sribljastyj silvery 6 18 29.0 kremovyj cream 6 18 29.0 sribnyj silver 6 18 29.0 limonnyj lemon 6 18 29.0 zolotisty golden 5 15 34.0 temno-synij dark blue 5 15 34.0 buryj brown 5 15 34.0 lilovyj lilac 5 15 34.0 metalyk metallic 5 15 34.0
We can make a number of general observations, which will be discussed in more detail in section 3. First of all Berlin and Kay's basic color terms are represented by the top eleven ranking terms, with the notable exception of ORANGE. Second, there are two terms for ORANGE, neither of which appears among the top ranking terms (delimited by a line of white space). The first, zovtoharjacyj, a compound term literally meaning `burning yellow', is ranked fourteenth, with a frequency of 53%; the second, oranzevyj, is ranked lower (twentieth) with a frequency of 29%. Third, while two terms for ORANGE fall outside of the top frequency group, there are three terms for BLUE, two of which fall within this group. These are synij `blue', and blakytnyj and holubyj `light blue'.(17) The first two, synij and blakytnyj, both appear in the top frequency zone and have nearly identical frequencies: synij occurs 31 times, and blakytnyj 30 times.
Finally, there is an identifiable group of highly frequent nonbasic terms, headed by salatnyj `light green' with a frequency of 76%. It is divided from the other nonbasic terms: the difference between the frequency of bordovyj `bordeaux', the lowest member of this group, and bezevyj `beige', the highest member of the next group, is fifteen percentage points.
2.1.2. Height-on-list measure. Using the list task, a second measure of basicness is the position the color term occupies in the ordered list of terms on the questionnaire. The higher the position, or the nearer the term is to the top of the list, the stronger the evidence that it is a basic term. The highest position is position 1. We give two kinds of data for the height-on-list measure. Table 3 shows which terms occupy the highest place on the list, and how many times a term occupies the first position. Terms are then ranked according to the frequency with which they occur at this highest positon. For example, the top ranking color term is bilyj `white', which appears highest on the lists of ten informants; bottom ranking is zelenyj `green', which a single informant placed at the top of her list.
Table 3. Color terms occupying the highest place on the list Color term Gloss Informants with Rank term at top of list bilyj white 10 1.0 cornyj black 5 2.5 rozevyj pink 5 2.5 zovtyj yellow 4 4.5 blakytnyj light blue 4 4.5 cervonyj red 3 6.0 synij blue 2 7.0 zelenyj green 1 8.0
We note the following. First, the set of terms occupying the highest place on a list (Table 3) is a subset of the terms found within the top frequency zone of Table 2. Thus only Berlin and Kay basic color terms appear first in an informant's list. Second, this set includes the elemental color categories (see section 1.2). In other words, all the elemental colors are being used by at least one informant as the first-choice color term. Third, two BLUE terms, synij and blakytnyj, which appear in the top frequency zone in Table 1, also occupy the highest place on several informants' lists.
As well as looking at terms that occupy the highest place on an informant's list, we can also rank the terms according to their average place on a list across all informants. High-ranking terms will be those whose mean is closest to 1, in other words the first place on the list. This is shown in Table 4, where on average the term cervonyj `red' is between the fourth and fifth term to appear on an informant's list of terms, and as such represents the highest-ranking color term for this measure.(18)
Table 4. How terms are distributed across list positions Term Gloss Average place Rank on list cervonyj red 4.5 1.0 bilyj white 6.5 2.0 zelenyj green 6.8 3.0 rozevyj pink 7.1 4.0 blakytnyj blue 7.3 5.0 synij blue 7.5 6.0 zovtyj yellow 7.6 7.0 cornyj black 8.9 8.0 fioletovyj purple 9.1 9.0 siryj gray 11.4 10.0 korycnevyj brown 12.5 11.0 malynovyj raspberry 13.7 12.0 zovtoharjacyj orange 14.0 13.0 salatnyj light green 15.8 14.0 bordovyj bordeaux 16.1 15.0 holubyj blue 17.0 16.0 oranzevyj orange 17.8 17.0 birjuzovyj turquoise 18.0 18.0 buzkovyj lilac 18.4 19.5 purpurnyj purple 18.4 19.5 bezevyj beige 19.3 21.5 zolotyj gold 19.3 21.5 vysnevyj cherry 19.4 23.5 burjakovyj beet 19.4 23.5 kremovyj cream 19.9 25.0 lilovyj lilac 20.3 27.0 buryj brown 20.3 27.0 kastanovyj chestnut 20.3 27.0 kavovyj coffee 20.6 29.0 bolotnyj marsh green 20.7 30.0 rudyj red-yellow 20.8 31.0 limonnyj lemon 20.9 32.5 metalyk metallic 20.9 32.5 sribljastyj silver 21.0 34.0 sribnyj silver 21.1 35.0
The eleven top-ranking color terms in the frequency measure (see Table 2) are also the eleven top-ranking terms in the list position measure, all Berlin and Kay basic terms. There are differences in ordering among these eleven terms, however. Further, zones paralleling the frequency zones in Table 2 have emerged. The eleven top-ranking terms appear in the first zone, followed by a zone of strong nonbasic terms that include malynovyj `raspberry', zovtoharjacyj `orange', salatnyj `light green', and bordovyj `bordeaux', in turn followed by a zone of weaker nonbasic terms. It should also be noted that within the eleven top-ranking terms appear two terms for BLUE, synij and blakytnyj, the two terms highlighted in our discussions relating to Tables 2 and 3. As in the frequency measure, they behave similarly to one another: synij has an average list position of 7.5 and blakytnyj has 7.3, and they are ranked sixth and fifth respectively. Finally, the mean list position measure matches the frequency measure with regard to the ORANGE category. There are two terms, oranzevyj and zovtoharjacyj, both falling outside of the top zone.
As with the frequency measure, zovtoharjacyj is the stronger term: it is ranked thirteenth and has an average list position of 14, whereas oranzevyj is ranked seventeenth with average list position of 17.8.
Using the list task as a test of psychological salience, there is evidence that Ukrainian has ten of the eleven Berlin and Kay basic color terms, where ORANGE is the missing category. At the same time it has two basic terms for BLUE, synij and blakytnyj `light blue'. We now turn to examine the situation in Belarusian.
2.2. Belarusian list task
The list task was carried out on twenty-eight Belarusian speakers from the Belarus State Economic University in Minsk, all of whom stated they could also speak Russian.(19) Sixteen entered Belarusian as their first language, and twelve as their second language (with Russian first). The questionnaire was written in Belarusian,(20) and all subjects used Belarusian in their answers. The age range was seventeen to twenty-one years. The task elicited 74 different terms.
2.2.1. Frequency measure. The frequency order of terms occurring at least three times among all subjects is given in Table 5. Some of the informants considered Belarusian to be their second language, with Russian as the first. It is important to distinguish the first language speakers, and these are indicated in Table 4 in the column "Bel = 1." The rank ordering of this group is indicated separately, in the last column ("Bel = 1"). There are slight differences in ranking, but these are not significant: both share the top twelve most frequent terms.
Table 5. Frequency order (for "all" n = 28 and "Bel = 1" n = 16) Term Gloss Frequency occurrences % all Bel = 1 all Bel = 1 zjaleny green 28 16 100 100 bely white 27 15 96 94 cyrvony red 27 16 96 100 corny black 26 15 93 94 zouty yellow 25 15 89 94 ruzovy pink 25 14 89 88 sini blue 24 13 86 81 blakitny light blue 24 14 86 88 sery/sery gray 23 14 82 88 fijaletovy purple 22 14 79 88 karycnevy brown 15 7 54 44 aranzavy orange 9 5 32 31 bezavy beige 6 1 21 6 haluby light blue 5 4 18 25 zalaty gold 5 3 18 19 bury brown 4 2 14 13 purpurovy(21) purple 4 2 14 13 lilovy lilac 4 3 14 19 salatvy light green 4 3 14 19 srebrany silver 4 3 14 19 biruzovy(22) turquoise 4 0 14 0 hranatovy bright red 3 3 11 19 zolocisty golden 3 1 11 6 Term Rank all (Bel = 1) zjaleny 1.0 (1.5) bely 2.5 (4.0) cyrvony 2.5 (1.5) corny 4.0 (4.0) zouty 5.0 (4.0) ruzovy 6.0 (7.5) sini 7.5 (9.0) blakitny 7.5 (7.5) sery/sery 9.0 (7.5) fijaletovy 10.0 (7.5) karycnevy 11.0 (11.0) aranzavy 12.0 (12.0) bezavy 13.0 (21.5) haluby 14.5 (13.0) zalaty 14.5 (16.0) bury 18.5 (19.5) purpurovy(21) 18.5 (19.5) lilovy 18.5 (16.0) salatvy 18.5 (16.0) srebrany 18.5 (16.0) biruzovy(22) 18.5 (23.0) hranatovy 22.5 (16.0) zolocisty 22.5 (21.5)
The first observation is that within the top frequency zone all the elemental categories are included, and additionally three compound categories, PINK, PURPLE, and GRAY. It should be noted that the terms sery and sery are taken to be alternates. They are mentioned in the Lexical Atlas (1993-1998) (volume 4, map 4). Very generally, the term sery is dominant in the dialects south of Minsk and part of the central dialects from Mahileu down to Homel', where sery dominates the central and northern dialects (see Mayo 1983:943 for a summary of Belarusian dialects). Treating these forms as alternants is supported by the fact that no subject used both forms.(23) Second, two terms for BLUE have been elicited, sini and blakitny. Both appear in the top frequency zone, and both have exactly the same number of occurrences, twenty-four (86%). The term blakitny means `light blue, sky blue' (cf. Ukrainian blakytny).(24) Third, the term karycnevy `brown' is stranded in a frequency zone between the highest and lowest frequency terms. The difference between it and the lowest of the high-frequency terms, fijaletovy `purple', is 25 percentage points, and the difference between it and the highest of the low-frequency terms, aranzavy `orange', is 22 percentage points. This puts a question mark over the basic status of the term. Finally, there is a term for ORANGE, aranzavy, but it falls outside the second frequency zone. This strongly suggests that Belarusian, like Ukrainian, lacks a basic ORANGE category.
2.2.2. Height-on-list measure. As with Ukrainian, we examine which of the terms appears at the top of an informant's list, and how many times it appears here across all informants. Table 6 shows for example that cyrvony `red' was listed first by sixteen informants.
Table 6. Color terms occupying the first position Color term Gloss Informants with Rank term at top of list cyrvony red 16 1.0 corny black 4 2.5 bely white 4 2.5 ruzovy pink 2 4.0 zouty yellow 1 5.5 sery gray 1 5.5
As with the Ukrainian data only Berlin and Kay's basic categories are represented. These are both elemental, for example cyrvony `red', and compound, for example ruzovy `pink'. Only one elemental category is "missing" as it were: BLUE.
In Table 7 we give the average list position of all the terms appearing at least three times: in other words, the place a term occurs on an informant's list when all informants' lists are taken into consideration.
Table 7. Average list position for Belarusian Term Gloss Average place Rank on list cyrvony red 2.4 1.0 bely white 4.1 2.0 corny black 4.3 3.0 zjaleny green 5.6 4.0 blakitny blue 6.2 5.5 zouty yellow 6.2 5.5 ruzovy pink 6.8 7.0 sini blue 7.6 8.0 fijaletovy purple 9.5 9.0 sery gray 10.4 10.0 karycnevy brown 10.8 11.0 aranzavy orange 12.3 12.0 sery gray 12.6 13.0 haluby light blue 13.1 14.5 lilovy lilac 13.1 14.5 srebrany silver 13.4 17.0 bezavy beige 13.4 17.0 biruzovy turquoise 13.4 17.0 zolocisty gold 13.6 19.0 salatavy light green 13.7 20.0 hranatovy bright red 13.8 22.0 bury brown 13.8 22.0 purpurovy purple 13.8 22.0 zalaty gold 14.6 24.0
From Table 7 we can note the following. First, the eleven top-ranking terms according to frequency (Table 5) match the top eleven according to average list position. These are all Berlin and Kay basic color categories. Note how karycnevy `brown' again appears at the bottom of this ranking, but, unlike the frequency measure, there are grounds for including it along with the other putative basic terms, as it is close to the term ranked above it (sery `gray') but sharply divided from the term ranked below it (aranzavy `orange'). Hence from this second measure there is evidence that it is basic; this underlines the fact that both measures, frequency and list position, need to be taken into account. Second, two terms for BLUE, sini and blakitny, appear within the top-ranking zone, paralleling the frequency measure. Finally, as with the frequency measure, the term aranzavy falls outside the top-ranking group of terms, suggesting again that Belarusian lacks the category ORANGE.
To summarize, we have evidence from the list task that Belarusian has all of the Berlin and Kay basic color terms except for ORANGE. In addition to the basic BLUE term sini it has the term blakitny `light blue', which also appears to be basic.
3. Discussion: East Slavonic BLUEs
Our investigation into the East Slavonic languages Ukrainian and Belarusian was prompted by a curiosity of the Russian color system, namely that Russian has two basic BLUE terms. In this section we outline the Russian color system, based on Corbett and Morgan (1988) and Davies and Corbett (1994).. and then discuss the color systems of Ukrainian and Belarusian with regard to the BLUE category. Our findings constitute strong evidence that these other East Slavonic languages have also innovated a second BLUE term. There are two major questions that arise from our findings. First, is it possible that the second BLUE term is a borrowing of a category separate from the term that denotes it? And second, what can we learn about the evolutionary path of the second BLUE category? Answers to these questions shed light on Russian's development of a second BLUE category.
3.1. Russian basic color terms: an outline
Russian has innovated a second BLUE category: Russian has two basic terms, sinij `dark blue' and goluboj `light blue'. That both terms are basic has been confirmed by Corbett, Davies, and Morgan in a number of psycholinguistic experiments. These have included behavioral tests, such as the list task and color-naming tasks, as well as linguistic tests, such as examination of textual frequency. In Appendix 1 of Corbett and Davies (1997) they conclude from their findings that Russian has twelve basic color terms, with two BLUE terms. They list them as follows:
belyj `white', cernyi `black', krasnyj `red', zelenyj `green', zeltyj `yellow', sinij `dark blue', goluboj `light blue', koricnnevyj `brown', fioletovyj `purple', rozovyj `pink', oranzevyj `orange', seryj `gray'
Some of their findings are discussed in Corbett and Morgan (1988) and Davies and Corbett (1994), which we briefly look at in relation to sinij and goluboj.
3.1.1. Russian list task. Corbett and Morgan (1988) report on a list task carried out on seventy-seven native speakers of Russian in Moscow. The results are given in Table 8.(25) These show that both BLUE terms appear within the top twelve terms (sinij is ranked highest, with a frequency of 99%, and goluboj is ranked fourth equal with a frequency of 95%). Moreover, there is a clear cut-off point between the claimed Berlin and Kay basic terms and the highest-ranking nonbasic term: there is a 17-point difference between rozovyj `pink' and sirenevyj `mauve', showing that the two BLUEs are comfortably within the basic group.
Table 8. Results of the Russian list task (Davies and Corbett 1994: 73) (N = 77) Term Gloss Frequency Rank occurrences as % sinij dark blue 76 99 1.0 krasnyj red 75 97 2.5 zelenyj green 75 97 2.5 zeltyj yellow 73 95 4.5 goluboj light blue 73 95 4.5 cernyj black 71 92 6.0 fioletovyj purple 69 90 7.0 oranzevyj orange 67 87 8.0 belyj white 66 86 9.5 koricnevyj brown 66 86 9.5 seryj gray 60 78 11.0 rozovyj pink 53 69 12.0 sirenevyj mauve 40 52 13.5 salatovyj salad green 40 52 13.5 bordovyj bordeaux 38 49 15.0 malinovyj raspberry 37 48.5 16.5 bezevyj beige 37 48.5 16.5 birjuzovyj turquoise 34 44 18.0 limonnyj lemon 21 27 19.0 purpurnyj purplish red 20 26 20.0
3.1.2. Russian textual frequency test. Another measure of psychological salience of a term is its textual frequency.(26) Corbett and Morgan (1988) calculated textual frequency using Zasorina's (1977) frequency dictionary of a corpus of one million tokens. Their results are given here in Table 9, ranked according to frequency. Again, both sinij and goluboj appear well within the high-frequency group, more evidence that they are basic terms.
Table 9. Textual frequency of sinij and goluboj Color term Gloss No. of Rank frequency occurrences cernyj black 473 1 belyj white 471 2 krasnyj red 371 3 zelenyj green 216 4 sinij dark blue 180 5 goluboj light blue 137 6 seryj gray 116 7 zeltyj yellow 109 8 rozovyj pink 49 9 buryj brown 31 10 korycnevyj brown 23 11 fioletovyj purple 22 12 oranzevyj orange 15 13 bagrovyj crimson 13 14 lilovyj lilac 12 15
3.1.3. Russian color-naming task. Davies and Corbett (1994) present their results of a color-tile-naming experiment conducted in Moscow involving 54 native speakers of Russian. The 65 color tiles used were a representative sample of the color space. For exact details of the stimuli see Davies and Corbett (1994: 69-71). The results constitute strong evidence that both sinij and goluboj are basic terms. Table 10 gives their results (based on Table 4 of their paper).(27) They are using consensus across informants of a term and the tile it may denote as a measure of the term's basicness. In the table, terms are ranked according to consensus, which is calculated as the ratio of the number of times a term is used and the number of tiles the terms has been used to name. The top-ranking term is belyj, since all of its 54 occurrences are used to name a single tile, showing 100% consensus across the 54 informants. From the table we see that according to the tile-naming consensus measure, both goluboj and sinij are ranked within the first twelve terms (ranked ninth and eleventh respectively).
Table 10. Tile naming and consensus ranking Color term Gloss No. of No. of No. of occurrences/ occurrences tiles no. of tiles belyj white 54 1 54.00 koricnevyj brown 148 5 29.60 seryj gray 245 9 27.22 zeltyj yellow 131 5 26.20 zelenyj green 314 12 26.17 oranzevyj orange 178 8 22.25 krasnyj red 108 5 21.60 cernyj black 43 2 21.50 goluboj light blue 126 6 21.00 rozovyj pink 161 9 17.89 sinij dark blue 181 12 15.08 salatovyj salad green 45 3 15.00 sirenevyj mauve 158 11 14.36 fioletovyj purple 172 13 13.23 malinovyj raspberry 81 7 11.57 birjuzovyj turquoise 35 7 5 Color term Consensus rank belyj 1 koricnevyj 2 seryj 3 zeltyj 4 zelenyj 5 oranzevyj 6 krasnyj 7 cernyj 8 goluboj 9 rozovyj 10 sinij 11 salatovyj 12 sirenevyj 13 fioletovyj 14 malinovyj 15 birjuzovyj 16
In conclusion, these findings should be viewed as strong evidence that Russian has two BLUE terms, sinij meaning `dark blue' and goluboj meaning `light blue'.
3.2 Ukrainian and Belarusian BLUEs
The results of the list task for both Ukrainian and Belarusian strongly suggest that these languages have two basic terms for BLUE. Ukrainian has synij `blue' and blakytnyj `light blue', and Belarusian has sini `blue' and blakitny `light blue'. This in the context of both languages having all the Berlin and Kay basic color terms, except for ORANGE. Table 11 gives the behavior of the two BLUE terms for Ukrainian and Belarusian (see previous tables). They are noted for how they perform in frequency, mean list position, and the possibility of appearing at the top of an informant's list of colors. For the frequency and mean list position measures, the ranking of the term is given, showing that for both languages the two BLUE terms appear among the top-ranking terms, using either the frequency measure or the list position measure.
Table 11. Comparing the two BLUEs in Ukrainian and Belarusian Frequency Ranking Mean list Ranking position Ukrainian (N = 34) synij 31 6.5 7.5 5.0 blakytnyj 30 10.0 7.3 6.0 Belarusian (N = 28) sini 24 7.5 8.6 8.0 blakitny 24 7.5 6.2 5.5 Heads a color list? Ukrainian synij yes blakytnyj yes Belarusian sini no blakitny no
The table also shows that for each language the two BLUE terms behave similarly. For the frequency measure, Ukrainian synij and blakytnyj differ by a single occurrence. And in Belarusian, sini and blakitny occur the same number of times, sharing the same rank position (seventh equal). When we move to mean list position, we see that the two terms in Ukrainian are nearly identical and are ranked side by side. This is less clear for Belarusian from the table, yet if we go back to Table 7, we see that sini is separated from blakitny by only two terms, zouty `yellow', which shares blakitny's mean list position of 6.2, and ruzovy `pink', which has a list position of 6.8. Finally, considering whether the two terms appear at the top of a list, we see that where this possibility is open for one term it is also open for the other. Thus for Ukrainian both synij and blakytnyj appear at the top of the list, but in Belarusian no informants listed either sini or blakitny as the first term.
3.3. The second BLUE category in East Slavonic
The findings are suggestive of a second BLUE category not only for Russian but also for its sister languages Ukrainian and Belarusian. We examine the implications of claiming that there are two BLUEs in East Slavonic.
3.3.1. The second BLUE: separating category from term. Russian's second term for BLUE, goluboj, is a form that goes back to Old Russian, the East Slavonic ancestor language (eleventh to fourteenth centuries). It is derived from the word for pigeon golub'. It originally meant `gray, blue' and was restricted in its use, denoting colors of animals (Baxilina 1975: 194). Later this term took on an additional meaning, `light blue', and could refer to fabrics and materials (examples found from end of the sixteenth century; see Barxudarov 1977: 70). Examples of its present-day use and meaning are found in the literature of the seventeenth and eighteenth centuries onward (see Baxilina 1975: 195-197). The term is found in both Ukrainian and Belarusian, indicating that it has been inherited by these two languages from Old Russian, that is, Common East Slavonic, the ancestor language of Russian, Ukrainian, and Belarusian. However, unlike the psychological salience tests for Russian (section 3.1), which clearly show that goluboj is basic in Russian, our tests for Ukrainian and Belarusian indicate that though this term is elicited, it is not basic in either language. How the term performs is charted in Table 12. The performance of Russian goluboj in Corbett and Morgan's list task (see Table 8) is included for comparison.
Table 12 shows how the term performs on the frequency and mean list position measure, for both languages. The ranges for high-frequency terms and high-list-position terms are given for comparison. When we compare the frequency of Ukrainian holubyj (column 1) with the frequency range covered by the most frequent Ukrainian color terms (column headed "Frequency of high-ranking terms"), we see that it falls well outside of this range (24% compared to the bottom-of-the-range figure 88%). Belarusian haluby similarly falls well short of the range of Belarusian high-frequency terms. We find the same situation obtaining for both languages with the mean list position. Ukrainian holubyj falls outside of the 4.5-12.5 range of high-ranking Ukrainian terms (column headed "Mean of high-ranking terms"), and similarly Belarusian haluby. Finally, for both languages the term is never ranked higher than fourteenth, and the lowest ranking is twenty-third equal (Ukrainian holubyj). This is in striking contrast to Russian goluboj, ranked fourth equal in the list task according to the frequency measure.
Thus a term used for the second BLUE category in Russian, which goes back to East Slavonic, is not found fulfilling this function in the other East Slavonic languages. Instead an alternative term blakytnyj (Ukrainian)/blakitny (Belarusian) is used for the second BLUE. This term is a borrowing of Polish bletkitny, glossed in contemporary Polish dictionaries as simply `blue' and noted by Rothstein (1993: 753) as the poetic synonym of the basic BLUE term niebieski. However, it is a relatively old borrowing, being found in the Ukrainian and Belarusian word stock from the sixteenth century, where it is used chiefly to describe fabrics and garments.(28) According to Zareba (1954: 47-49) the semantics of Polish blekitny took the following course: name of a kind of material
(Late Latin blanchetus `panni genus albi, candidi'), then the color describing that material `light, unsaturated', then a term used to describe specifically light-blue material. Later it was used, alongside modry, as the general blue term before niebieski took over as the basic term (beginning from the eighteenth century). It came into Ukrainian and Belarusian in the sixteenth century as a term for describing fabrics, with the meaning `light blue'.(29) There are two important facts to note about the borrowing of this term. First, it was not a borrowing of the basic BLUE term: Polish had modry as basic BLUE, and only later did blekitny begin to compete for this slot, at any rate after the borrowing had taken place. Second, nor was this a borrowing of a basic second BLUE term: there is no evidence that Polish has ever had a second BLUE, for which blekitny could have been the basic term; rather, blekitny lost the `light blue' meaning over the course of time and took on the more general meaning. We are therefore left with the prospect that, under the sway of the Russian color system, Ukrainian and Belarusian borrowed Russian's innovative category, the second BLUE, but not Russian's basic term for that category, golutboj.
Table 12. Performance of Ukrainian holubyj and Belarusian haluby Frequency Frequency of Rank Mean (as %) high-ranking list terms position Ukrainian holubyj 24 100-88 23.5 17.0 Belarusian haluby 18 100-79 14.5 13.1 Russian goluboj 95 99-69 4.5 -- Mean of Rank high-ranking terms Ukrainian holubyj 4.5-12.5 16.0 Belarusian haluby 2.4-10.8 14.5 Russian goluboj -- --
This opens up the possibility that a category, in this case the second BLUE, may be borrowed separately from the basic term that denotes it. However, the separation of category from term in the process of borrowing is not the usual course of events. As an example, we can consider Ukrainian korycnevyj (alternant korycnyavyj), which from the evidence of the list task looks to be the basic term for BROWN (both from the frequency measure, Table 2, and the list-position measure, Table 4). This is a borrowing from Russian korycnevyj that has latterly become the basic BROWN term (see its strong performances in the Russian list task and color-naming task, Tables 8 and 10). Ukrainian, on the other hand, has had several BROWN terms: brunatnyj, cinamonovyj, and the East Slavonic term buryj (Table 13). Priestly (1987) notes an additional one, bronzovyj (the term that emerged during his list task carried out on 12 Ukrainian speakers resident in North America). Dictionaries dating before the first half of the twentieth century give one of these as the primary meaning.(30) The confusion surrounding the category prompts Shevelov (1993: 992) to claim that for BROWN "there is no generally accepted term." Most of these terms are elicited in the list task, and their performances are shown in Table 11. Clearly, the borrowed term korycnevy is the strongest candidate among them for basic status.
Table 13. Contending BROWN terms in Ukranian (N = 34) Term Frequency Rank (among occurrences % BROWN terms) korycnjavyj/korycnevyj 31 91 1 buryj 5 15 2 brunatnyj 2 6 3 bronzovyj 1 3 4 cinamonovyj 0 0 5
In the case of BROWN, then, a term has been borrowed from Russian, along with its basic status. In other words, the term has not been separated from the category it denotes, making the situation with the second BLUE look particularly striking.
However, this appears less striking if we consider that terms and categories may well be separated in the history of a language. Shields, writing on Indo-European color terms, notes that for RED Indo-European has *reudh- yielding Greek eruthros, Latin ruber, etc., but among the Indo-European languages no cognates for WHITE and BLACK are to be found (Greek has leukos, Latin albus, etc.). He notes that this is not evidence against Berlin and Kay but simply a consequence of "lexical replacement":
... it is well known that members of the "basic core vocabulary" (like colour terms) of a language are lost and replaced by other (generally semantically related) forms as time passes (Shields 1979: 143).
In other words, the evolution of the categories operates independently of the shifts in basic status of the terms themselves. We do not need to look far for an example of lexical replacement in Slavonic. We can consider the basic category RED in East Slavonic. In Ukrainian and Belarusian the term used is cervonyj and cyrvony respectively, a basic term confirmed by the results of the list task (see Tables 2 to 7). Russian, however, has the basic term krasnyj (confirmed as basic by the tests carried out by Corbett, Davies, and Morgan; see Tables 8 to 10). Yet the ancestor language Old Russian has cbrvlenb (alternate cbrvenb), a form inherited by all three East Slavonic languages, used for basic RED by Ukrainian and Belarusian, but not by Russian. Moreover, the Russian krasnyj is fairly recent as a color term, not denoting basic RED until the end of seventeeth century (Baxilina 1975: 80). Separate terms may at first suggest that Russian developed the RED category independently from Ukrainian and Belarusian, but the facts turn out otherwise. Common Slavonic had cbrvlenb/cbrvenb for basic RED (see Herne (1954: 32-48; Schenker 1993:111), but later Russian used a different term to cover this same category.
One final question is why this separation of category and term should have taken place in the borrowing of the second BLUE from Russian. A possible answer is found in the term goluboj itself. As we noted, the East Slavonic term originally meant `gray (blue)' and was a restricted term, used to denote the color of animals. This original meaning has been inherited by the other East Slavonic languages, Belarusian and Ukrainian. By the time Russian had innovated the second BLUE category, Belarusian and Ukrainian already had available a term with a much more appropriate semantics, blakytnyj `light blue'.
In sum, the influence of Russian on its sister languages Ukrainian and Belarusian has led to the spreading of Russian's unusual color system. The innovative second BLUE category has been adopted throughout East Slavonic. Moreover, the very nature of this process says something about the category itself: since the category has been borrowed separately from its term, we have evidence of a different kind that the Russian color system does indeed include a second BLUE, a category absent from Berlin and Kay: the category is salient enough to be borrowed separately from the term that denotes it.
3.3.2. The second BLUE and category evolution. Kay and McDaniel's (1978) revision of the Berlin and Kay typology is important for the evolutionary predictions of the typology (see section 1.2). Later stages involve the combination of elemental categories to yield new compound categories. The development of compounds other than the original four cited in Berlin and Kay is therefore theoretically possible, such that the evolutionary process may be taken beyond the basic eleven categories:
There is no apparent reason to believe that the process will not continue, extending basic color term lexicons beyond their present 11 terms (Kay and McDaniel 1978: 640).
The second BLUE in Russian, goluboj `light blue', is given as an example of a possible twelfth term, the combination of BLUE and WHITE. Corbett and Davies have established the basic status of goluboj, and the Ukrainian and Belarusian findings here constitute strong evidence of a second BLUE in the other members of East Slavonic. In Russian goluboj came to denote basic LIGHT BLUE at a stage after BROWN (denoted by buryj) and before the evolution of the later compound color categories: the earliest compound term rozovyj `pink' is borrowed into Russian from German in the eighteenth century (Vasmer 1986), a century after goluboj acquires the meaning `light blue'. For Belarusian and Ukrainian, evidence from the list task suggests that LIGHT BLUE emerged before ORANGE. Belarusian aranzevyj falls just outside the top ranking group of colors for both the frequency and list-position measure. Ukrainian has two terms, zovtoharjacyj and oranzevyj, but neither appears within the top-ranking group for either measure.
The indication is that the second BLUE may emerge before the completion of the set of eleven categories cited in Berlin and Kay. In other words, it does not have to be the "twelfth basic color term" as Kay and McDaniel speculate (1978: 640). There is some evidence that besides Russian (and its East Slavonic sister languages) a few languages have a second BLUE term. An interesting question is whether in such cases the term is the twelfth basic term, or whether it has emerged ahead of the recognized eleventh Berlin and Kay basic terms. Languages which may have a second BLUE term include Guatemalan Spanish (Harkness 1973), Nepali (Bolton et al. 1980), and Italian (Vincent 1983).(31) One particularly strong contender is Turkish, which has been extensively investigated using the color listing and naming task. The term for elemental BLUE in Turkish is mavi. Ozgen and Davies (1998) explored the possibility of a putative second BLUE term, lacivert `dark blue', which if basic would be viewed in the Kay and McDaniel framework as a compound category made up of elemental BLUE and BLACK (Ozgen and Davies 1998: 951). The listing and naming tasks were carried out on a substantial sample of children and adults from Istanbul and Fethiye.(32) For the color-naming task, two measures were used as indicators of basicness, frequency of a term used to name a tile, and agreement among informants that the choice of a term to name a tile is correct. The findings are summarized in the form of general ranking of terms when all these meaures have been computed. For both the child and adult samples there is evidence that Turkish has the eleven recognized Berlin and Kay basic color terms, ranked 1 to 11. At the same time, the term lacivert ranks twelfth for both groups (see their Table 8, Ozgen and Davies 1998: 943). The tentative conclusion is that if lacivert is basic, its basic status is emergent, whereas the basic status of the other eleven terms is firmly established. Hence for Turkish the second BLUE comes after the recognized eleven basic categories have developed.
In sum, we appear to have three distinct color systems with reference to the development of the second BLUE category. In the first system, represented by Russian, this category is the first compound category to emerge after BROWN. In the second system, represented by Ukrainian and Belarusian, the second BLUE category is not the last compound category to emerge (ORANGE is at best emergent). Finally, Turkish represents the third system, where the second BLUE category has waited for the other recognized categories to be established. These three systems are naturally accounted for in the typology if we consider the second BLUE to be a compound category, which, following Kay and McDaniel, is either the intersection of BLUE and WHITE (Russian sinij, Ukrainian blakytnyj, Belarusian blakitny) or the intersection of BLUE and BLACK (Turkish lacivert). It is assigned to the set of compound colors that develop from stage VII. A property of this set is that there is no ordering within the set, reflecting the different orderings in the three systems described above. At the same time, a precondition is the emergence of BROWN, which is met by each of our examples.
4. Conclusions
In examining whether the innovative second BLUE category is unique to Russian, or whether it is a broader characteristic of Russian's family, Slavonic, we have produced some evidence, using a recognized psycholinguistic test, that both Ukrainian and Belarusian have two basic terms for BLUE. The first term is inherited from East Slavonic sinij denoting `(dark) blue', and the second is a borrowing from Polish blekitny, which in both languages means `light blue'. Though Russian has influenced the development of the Ukrainian and Belarusian color systems in this way, the Russian basic term itself, goluboj, has not been borrowed. It appears that a category, in this case the second BLUE, may be borrowed separately from the basic term that denotes it. Thus the evolution of categories operates independently of the shifts in basic status of the terms themselves. The separation of the term from the category is further evidence, and evidence of a different kind, that Russian does indeed have two BLUE categories, since the category is salient enough to be borrowed separately from the term that denotes it. Finally, among the languages that have this category, its development differs with respect to other compound categories except for BROWN, which is always prior. This can be accounted for by assigning it to the compound color categories that evolve from stage VII. Thus in principle it may be the first stage VII compound category to emerge, as in Russian, or the last, as in Turkish, or somewhere in between, as in Ukrainian and Belarusian.
Received 27 June 2000 Revised version received 25 October 2000
Notes
(1.) Versions of this paper were read at the Linguistic Circle of Oxford, the Second Slavic Linguistics Society at Berkeley, and the British Association of Slavonic and East European Studies at Cambridge, and all helpful comments from the floor are gratefully acknowledged. I would also like to thank Greville Corbett for valuable suggestions on earlier drafts, Ian Davies for valuable comments and help with the statistical analysis, and the helpful comments of two anonymous referees. The research reported here was supported by the ESCR (grant no. R000237845) and I am grateful for their support. Correspondence address: Department of Computing, University of Surrey, GB-Guildford GU2 7XH. E-mail: [email protected].
(2.) See for example the papers in Hardin and Maffi (1997).
(3.) Uppercase is conventionally used for color categories, as distinct from color terms.
(4.) A summary of their work on the Russian color system, mentioning the key references, can be found in Appendix 1 of Corbett and Davies (1997). Moreover, their evidence that the PURPLE category is not fully established in Russian indicates that the second. BLUE is not the twelfth term, as Kay and McDaniel (1978) speculate. We will return to this in section 3.3.2.
(5.) The Slavonic language family is genetically close and therefore provides an interesting area for the study of color category evolution. See Hippisley and Gazdar (2000) for discussion, and a computable model for the reconstruction of Proto-Slavonic basic color terms (and the stages leading up to it) that assumes the Slavic basic color terms outlined in Comrie and Corbett (1993).
(6.) See Rosch et al. (1976), Rosch (1978), and discussions in MacLaury (1991a).
(7.) See Wooten and Miller (1997) and Abramov (1997) for good summaries of the physiology and psychophysics of color sensation.
(8.) Note, however, that the emergence of compound colors after the decomposition of the composites into elements is a tendency rather than an absolute rule. Kay et al. (1991: 18) report on the possibility of BROWN and PURPLE appearing before complete decomposition has taken place.
(9.) See Schenker (1993: 60-61, 114) for the historical perspective.
(10.) Thanks to V. M. Pavlenko from the Psychology Department at Xarkiv University, who translated the questionnaire into Ukrainian and carried out the list task at Xarkiv.
(11.) Note that the alternant korycnevyj is counted together with this term.
(12.) The alternant salatovyj is counted together with the term.
(13.) Note also the alternant bordo, an indeclinable adjective.
(14.) Note also the alternant bez, an indeclinable adjective.
(15.) This includes the alternant purpurovyj.
(16.) This includes the alternant zolotovyj.
(17.) These are the glosses given in the major Ukrainian dictionaries, for example, Bilodid (1970) and Hrincenko (1907-1909). The main Ukrainian-English dictionary, Andrusysen and Kret (1995), glosses both blakytnyj and holubyj as `sky blue'.
(18.) The mean is calculated in such a way as to compensate for the potential distorting effect of low-frequency terms having high means because a few informants (perhaps just one) said the term early. The calculation gives all subjects a score for all terms. If a subject does not say a given term, then it is assumed that they would have eventually given it if they carried on. So, they are given a score of their total number of terms plus one. In most cases this has little effect on the relative values of the mean position scores.
(19.) Thanks to Alexander Povalyeav, who coordinated the task.
(20.) Thanks to Arnold McMillin for the Belarusian translation of the questionnaire.
(21.) Note that some informants had the alternate purpurny.
(22.) Note the alternant biruza, an indeclinable adjective.
(23.) Note that sery was used more frequently in our sample, appearing sixteen times next to the nine appearances of sery.
(24.) See for example the gloss in Sudnik (1996) and Atraxovic (1977).
(25.) See Corbett and Morgan (1988: 72-75) for the original table and detailed discussion.
(26.) See Hays et al. (1972), who use textual frequency as a measure of psychological salience for Russian, among other language,;.
(27.) Davies and Corbett (1994: 78).
(28.) See Kersta (1994--), Mel nicuk (1982--), and Rudnyc'kyj (1962-1972) for Ukrainian, and Martynau (1978--) and Zuruuski (1982--) for Belarusian.
(29.) For Ukrainian, see Kersta (1994--), Mel'nicuk (1982--), and Rudnyc'kyj (1962--1972). For Belarusian, see Zurauski (1982--).
(30.) See Hrincenko (1907-1909), who gives brunatnyj; the 1929 Academy Dictionary, which gives einamonovyj as cited in Shevelov (1993: 992); Kmicykewytsch and Spilka (1912), a German-Ukrainian dictionary, which gives buryj.
(31.) Cited in Davies et al. (1995: 18). Catalan, as reported by Davies et al. (1995), represents yet another possibility for the emergence of the second BLUE. The two terms blau mar' `navy blue' and blau cel `sky blue' have strikingly high scores in the list task, performed on a sample of forty children and forty adults. The conclusion is that they are the highest of the nonbasic terms, suggesting that a compound BLUE category is emergent. However, these terms are not monolexemic and are included in the signification of the established basic term blau, failing two of Berlin and Kay's basicness criteria, so anyway their basic status is problematic.
(32.) For the list task eighty children and 153 adults were consulted. Fewer were consulted for the naming task (seventeen children and thirty-three adults).
References
Abramov, Israel (1997). Physiological mechanisms of color vision. In Colour Categories in Thought and Language, C. L. Hardin and Luisa Maffi (eds.), 89-134. Cambridge: Cambridge University Press.
Andrusysen, Constantine Henry; and Kret, Yakiv Nikolajevic (1995). Ukrajins'koanhlijs'kyj slovnyk. Toronto: University of Toronto Press. (First printed [1955]. Saskatoon: Gospel Press.)
Atraxovic, K. K. (1977). Tlumacal'ny slojnik belaruskaj movy, 5 vols. Minsk.
Barxudarov, S. G. (ed.) (1977). Slovar' russkogo jazyka XI-XVII vv tom 7. Moscow: Nauka.
Baxilina, N. B. (1975). Istorija cvetooboznaenij v russkom jazyke. Moscow: Nauka.
Berlin, Brent; and Kay, Paul (1969). Basic Color Terms: Their Universality and Evolution. Berkeley and Los Angeles: University of California Press.
Bilodid, I. K. (1970). Slovnyk ukrajins'koji movy, 11 vols. Kiev: Nukova dumka.
Bolton, Ralph; Curtsi, Anne; and Thomsa, Lynne L. (1980). Nepali color terms: salience on a listing task. Journal of the Steward Anthropological Society 12, 309-322.
Comrie, Bernard; and Corbett, Greville G. (eds.) (1993). The Slavonic Languages. London: Routledge.
Corbett, Greville G.; and Davies, Ian R. L. (1997). Establishing basic color terms: measures and techniques. In Colour Categories in Thought and Language, C. L. Hardin and Luisa Maffi (eds.), 197-223. Cambridge: Cambridge University Press.
--; and Morgan, Gerry (1988). Colour terms in Russian: reflections on typological constraints in a single language. Journal of Linguistics 24, 31-64.
Davies, Ian; and Corbett, Greville G. (1994). The basic color terms of Russian. Linguistics 32, 65-89.
--; Corbett, Greville G.; and Margalef, Jose Bayo (1995). Colour terms in Catalan: an investigation of eighty informants concentrating on the purple and blue regions. Transactions of the Philological Society 91, 1, 17-49.
Hardin, C. L.; and Maffi, Luisa (eds.) (1997). Colour Categories in Thought and Language. Cambridge: Cambridge University Press.
Harkness, Sara (1973). Universal aspects of learning color codes: a study in two cultures. Ethos 1, 175-200.
Hays, D. G.; Margolis, E.; Naroll, R.; and Perkins, D. R. (1972). Color term salience. American Anthropology 74, 1107-1121.
Herne, Gunnar (1954). Die Slavischen Farbenbenennungen: eine semasiologisch-etymologische untersuchung. Uppsala: Almquist and Wiksell.
Hippisley, Andrew; and Gazdar, Gerald (2000). Inheritance hierarchies and historical reconstruction: towards a history of Slavonic colour terms. In Papers from the 35th Annual Regional Meeting of the Chicago Linguistic Society, part 1, 125-140. Chicago: Chicago Linguistic Society.
Hrincenko, B. D. (1907-1909). Slovnyk ukrajins'koji movy. Kiev: Kievskaja Starina.
Kay, Paul; Berlin, Brent; Maffi, Luisa; and Merrifield, William (1997). Color naming across languages. In Colour Categories in Thought and Language, C. L. Hardin and Luisa Maffi (eds.), 21-58. Cambridge: Cambridge University Press.
--; Berlin, Brent; and Merrifield, William (1991). Biocultural implications of systems of color naming. Journal of Linguistic Anthropology 1 (1), 12-25.
--; and McDaniel, Chad K. (1978). The linguistic significance of the meanings of basic color terms. Language 54, 610-646.
Kersta, R. (1994-). Slovnyk staroukrajins'koji movy XVI--pergoji polovyny XVII st. Kiev.
Kmicykewytsch, Volodymyr; and Spilka, V. (1912). Deutsch-Ukrainisches Worterbuch. Cernivtzi: Ukrainischer Verlag.
Lexical Atlas (1993-1998). Leksicny atlas belaruskix narodnyx gavorak [Lexical Atlas of Belarusian National Dialects], 5 vols. Minsk: Belarusian Academy of Sciences.
MacLaury, Robert E. (1991a). Prototypes revisited. Annual Review of Anthropology 20, 25-53.
--(1991b). Social and cognitive motivations of change: measuring variability in color semantics. Language 67, 34-62.
Martynau, V. (ed.) (1978-). Etymalahicny slounik belaruskaj movy. Minsk.
Mayo, Peter (1983). Belarusian. In The Slavonic Languages, Bernard Comrie and Greville G. Corbett (eds.), 887--946. London: Routledge.
Mel'nycuk, O. S. (1982-). Etymolohycnyj slovnyk ukrajins'koji movy. Kiev.
Ozgen, Emre; and Davies, Ian(1998). Turkish color terms: tests of Berlin and Kay's theory of color universals and linguistic relativity. Linguistics 36(5), 919-956.
Priestly, T. M. S. (1987). On basic color terms in Early Slavic and Ukrainian. Annals of the Ukrainian Academy of Arts and Sciences 15 (39-40), 243-251.
Rosch, Eleanor H. (1978). Principles of categorization. In Cognition and Categorization, Eleanor Rosch and Barbara B. Lloyd (eds.), 27-48. New York and London: Wiley.
--; Mervis, Carolyn B.; Gray, Wayne D.; Johnson, David M.; and Boyes-Braem, Penny (1976). Basic objects in natural categories. Cognitive Psychology 8, 382-439.
Rothstein, Robert A. (1993). Polish. In The Slavonic Languages, Bernard Comrie and Greville G. Corbett (eds.), 686-758. London: Routledge.
Rudnyc'kyj, Jaroslav (1962-1972). An Etymological Dictionary of the Ukrainian Language. Winnipeg: Ukrainian Free Academy of Sciences.
Schenker, Alexander M. (1993). Photo-Slavonic. In The Slavonic Languages, Bernard Comrie and Greville G. Corbett (eds.), 60-121. London: Routledge.
--(1995). The Dawn of Slavic. Newhaven and London: Yale University Press.
Shevelov, G. Y. (1993). Ukrainian. In The Slavonic Languages, Bernard Comrie and Greville G. Corbett (eds.), 947--998. London: Routledge.
Shields, Kenneth, Jr. (1979). Indo-European basic colour terms. Canadian Journal of Linguistics 24, 142-146.
Sudnik, M. P. (1996). Tlumacal'ny slounik belaruskaj litaraturnaj movy. Minsk.
Vasmer, Max (1986). Etimologiceskij slovar' russkogo jazyka, 4 vols. Moscow: Progress.
Vincent, J. M. (1983). Categorizzazione e strategie di denominazione dei colori: aspetti metodologici e problemi relativi all'inglese e all'italiano. In Linguistica e anthropologia, Atto del XIV Congresso Intrernazionale della SLI, Roma, Grupp di Lecce (ed.), 161-171. Rome: Bulzoni.
Wooten, Bill; and Miller, David L. (1997). The psychophysics of color. In Colour Categories in Thought and Language, C. L. Hardin and Luisa Maffi (eds.), 59-88. Cambridge: Cambridge University Press.
Zareba, Alfred (1954). Nazwy barw w dialektach i historii jezyka polskiego. Warsaw: Polish Academy of Sciences.
Zasorina, L. N. (ed.) (1977). Castotnyj slovar' russkogo jazyka: okolo 40,000 slov. Moscow: Russkij jazyk.
Zurauski, A. I. (1982--). Histarycny slounik belaruskau movy. Minsk: Belarusian Academy of Sciences.
ANDREW HIPPISLEY University of Surrey
Printer friendly
Cite/link
Email
Feedback
| |
| Author: | HIPPISLEY, ANDREW |
|---|---|
| Publication: | Linguistics: an interdisciplinary journal of the language sciences |
| Geographic Code: | 4EXSV |
| Date: | Jan 1, 2001 |
| Words: | 11272 |
| Previous Article: | Synchronic variation as a result of grammaticalization: concessive subjunctions in German and Italian(*). |
| Next Article: | Grammatik der Zaza-Sprache. Nord-Dialekt (Dersim-Dialekt). |
| Topics: | |

Printer friendly
Cite/link
Email
Feedback