English Historical Semantics PDF

7 English colour terms: A case study C. P. Biggam 7.1 Introduction In Chapter 6, attention was drawn to the importance of basic level cat- egories: that is, those that are especially cognitively salient, are learned first by children, and generally evoke a high level of agreement across speakers. T...

7 English colour terms: A case study C. P. Biggam 7.1 Introduction In Chapter 6, attention was drawn to the importance of basic level cat- egories: that is, those that are especially cognitively salient, are learned first by children, and generally evoke a high level of agreement across speakers. These include categories of animals like Dog and Cat, objects like Chair, Table, Hat and Trousers, and abstractions such as Anger and Fear. Basic level categories are also important in the way we think and talk about colours, a rich and extensive area of the lexicon. This chapter deals with the development of the English colour lexicon and shows in depth how the techniques of historical linguistics can be applied to one area of our vocabulary. It begins by discussing the particular techniques and problems of Historical Colour Semantics. Nothing can be taken for granted in colour studies because, although we tend to believe that our own colour system is the only sensible and possible one, a quick look at other languages of the world shows that there are many different ways to classify colours. Some societies consider green and blue to be different shades of the same colour, for example, and others have to consider the brightness, dryness, shape or size of objects before they can choose the appropriate colour term to describe them. Various unexpected aspects of colour semantics like these can be found in the historical records of English, showing that not only have the words used for colours often changed over the centuries, but so has the way in which English speakers have used those words. 7.2 How to describe colour When the need arises to describe a colour, English speakers normally use a commonly occurring term such as blue or green and, if necessary, qualify it by a term such as pale, bright, dark or dull. The speaker can thus reduce the basic category of, for example, blue to a section of blue 113 114 english historical semantics which may be referred to with phrases such as dark blue or dull blue.1 These smaller areas of hue, such as that denoted by dull blue, can be even further reduced by additional qualifiers which are not specific to colour, producing longer phrases such as very dark blue or slightly dull blue. In all these cases, the speaker has taken a principal colour term (blue), usually referred to as a basic colour term or BCT, and has ‘zoomed in’ on increasingly precise areas of the cognitive category (basic colour category or BCC) that it names. This process is similar to the profiling of domains we observed in section 3.4. English speakers (and others) also use non-­ basic colour terms, which are often the same as, or derived from, words for gems, fruit, flowers or other objects which are considered typical examples of the colour, for example emerald, orange and rose. Such terms are particularly prevalent in fields such as advertising, fashion and poetry. A phrase such as dark blue is somewhat utilitarian; it offers a helpful description of the colour but is a little vague and definitely unexciting. In the world of advertising, there is a need to catch the eye of potential customers, to intrigue them, and to make them look at the product more closely. Similarly, in creative literature the writer is often keen to catch the reader’s attention and focus his or her imagination by the use of what may be called ‘fancy’ colour terms such as taupe, burgundy, tangerine and terracotta. The advertising industry in particular also makes use of evocative expressions, often conveying an indeterminate colour sense, which may encourage feelings such as nostalgia or sophistication in a potential purchaser. Such phrases are often highly imaginative, for example Paris mist, highland falls, April glow and many more. At the other end of the imagination scale are the colour descriptions found in fields such as Vision Science, Psychology, pigment manufac- ture, textile dyeing and others, in which the above approaches would be hopelessly subjective. Everyday terms such as red and pale blue lack the precision needed in such work, and fancy terms are often unclear as to hue. These research fields and occupations require exactitude, so they resort to precisely defined names (sometimes specific to their subject), to codes or numbers on a colour chart, or to accurate measurements of the hue wavelength and other elements of colour. Leaving aside these highly codified colour descriptors and the more manufactured colour expressions of advertisers, how can the colour terms of ordinary English be described? To answer this question, we must consider those constituent parts of colour which play a significant role in English vocabulary. There is no doubt about the most impor- tant element. In ModE, the word colour is used principally (and almost exclusively) of hues, that is the range of observed impressions on the  c. p. biggam 115 visible part of the electromagnetic spectrum. English speakers use BCTs such as red, green, yellow and blue to denote these hues, as well as a range of non-­basic terms such as turquoise, scarlet, violet and crimson. For the majority of English speakers, the words colour and hue are syno- nyms, but in the technical language of colour studies colour denotes all the elements which combine to create a particular colour experience, including hue. Apart from hue, a colour can be described by means of its tone, the amount of black or white which is perceived as being involved with a hue in a visual impression. This sounds like the mixing of a paint colour and, although we know that the colours of many other things have not been mixed literally, we tend to interpret colours intuitively in this way. A hue containing a noticeable white element is usually described as pale, whitish or light, and a hue with a black element is described as dark or blackish. The tone range can also operate independently of hue, when it is known as the achromatic range (meaning literally ‘without hue’). This range has three principal categories, which in English are labelled black, white and grey. Colours can also be described by the purity of their hues. If we look at a hue, for example red, and ask whether it looks greyish to any degree, we may decide that, far from looking greyish, it can be described as pure red or the reddest possible red. Such a colour is an example of a fully-­saturated hue which is at the vivid end of the saturation ­­ scale. If this red hue were combined with a certain amount of grey, it would look duller, perhaps just slightly duller than vivid red, or maybe so grey that the red hue is only just perceived. English speakers tend to call a vivid or saturated hue bright, as in ‘She painted her room a bright green’, but this can be confused with another aspect of colour, called brightness. In Colour Studies, brightness refers to the amount of light involved in a colour impression. This may affect the appearance of an object which produces light, such as the sun or an electric light bulb, or it may refer to reflected light from an object with a shiny surface, or it may simply be an impression of a pale (but not necessarily reflective) colour viewed in a well-­lit environment. You can experiment with various objects to see how different lighting conditions apparently change the colour of an object when viewed, for example, at night, in sunshine, at dusk, or in artificial light. Similarly, a colour on a matt surface looks different from the same colour on a shiny surface. In English we describe such visual impressions by the use of words like shiny, dazzling, glowing and dim. Other aspects, both visual and non-­visual, such as texture, shape, lustre, moisture and softness, also affect a colour impression, and some 116 english historical semantics of these play a significant role in colour naming in other languages (or in scientific research), but they are rarely considered relevant to colour descriptions by English speakers. 7.3 What are basic colour terms? As we have seen in section 7.2, English has both basic and non-­basic colour terms. There are eleven ModE BCTs: black, white, red, green, yellow, blue, brown, grey, purple, orange and pink. Earlier phases of English had fewer BCTs. In ME the BCTs were: blak, whit, red, grene, yelwe, bleu, broun and grei, with a developing BCT purpur(e). ME had no basic terms, therefore, for the BCCs of purple, orange and pink. In OE the basic terms were: blæc (or sweart, which survives in ModE as swarthy), hwit, read, grene, geolo and graeg, with developing BCTs haewe(n) (usually meaning ‘blue’ or ‘greyish blue’) and brun. OE, therefore, had no fully developed BCTs to denote the categories blue, brown, purple, orange or pink. OE and ME had various non-­basic terms and descriptive phrases for these categories, but no BCTs. So what makes some colour terms basic and why is it important to recognise them? A BCT is a colour term which is commonly used in both speech and writing, and which is known by every adult native speaker, but this definition could also be true of several non-­basic terms. There are other tests a colour term must pass before it can be considered a BCT. For example, it must not be a hyponym of another colour word, as ModE emerald, olive and verdant are hyponyms of the BCT green, which denotes the whole BCC of green (see section 3.2.2.1 on hyponymy). Another important test for basicness in Colour Semantics is contextual restric- tion. A BCT must not be largely confined to a particular context such as hair, skin or an animal’s coat. A simple test is to take a doubtful term, such as brunette (a human hair colour) or swarthy (a skin colour) and use them of other things, for example *a brunette book or *a swarthy cupboard. A native speaker of English will reject such phrases, and it soon becomes clear that these words cannot be freely used of every appropriately coloured object, so they must be non-­basic. In some cases, the above considerations may still leave doubts about the basic status of certain English colour terms, suggesting the need for further tests (see Biggam 2012, chapter 3 for further tests appropriate to English). Some examples of further basicness tests appropriate for historical studies are as follows: BCTs are more likely than non-­basic terms to occur frequently in texts; they are more common in colloquial expressions such as sayings, and in metaphors and similes; and there are likely to be morphological vari- ants for BCTs, just as black has the variants blacken and blackish. All these  c. p. biggam 117 factors indicate that BCCs, as denoted by BCTs, have psychological salience over non-­basics in the field of colour. So why is it important to know the BCTs of a language? Their sig- nificance is that BCTs are the linguistic labels for the basic cognitive categories of colour (BCCs) which are used by a particular speech com- munity. The BCCs are the principal divisions of the colour space and can function as abstract concepts among a particular group of people; to take one example, the group may have a concept of red which oper- ates quite independently of specific red objects. (This is why contextual restriction disqualifies a colour term from having basic status.) You may wonder why the BCTs of a speech community are worth researching, since everyone with normal vision knows the visible hues. In fact, the basic categories differ in both number and nature between groups and in many cases even between individual speakers of the same language; for example, there may be differences between the young and the old, between men and women, and between monolinguals and bilinguals. In addition, as we have seen, BCCs also differ between the various periods of a society’s history, as shown by their changing BCTs. Speakers of some languages (including OE and ME) interpret some or all of the purple, pink, orange or brown hues as areas of their red cat- egory, rather than as categories in their own right. Similarly, speakers of some languages regard yellow and green (and other combinations) as areas of a single category. Basic categories such as these, which seem so extensive to ModE speakers, are referred to as macro-­categories or composite categories. Although macro-­categories may seem quite alien to speakers of ModE, they were present in older forms of English and their speakers did not feel the need to have eleven BCTs denoting eleven basic categories as in ModE. We will now consider how and when the basic categories evolved in English. 7.4 The evolution of basic colour categories Historical colour semanticists have a useful resource in what is usually referred to as the evolutionary sequence. Based on some earlier work, it appeared in its more modern form in a highly influential book pub- lished in 1969, Basic Color Terms: Their Universality and Evolution, by Brent Berlin and Paul Kay (see 1969: 4 for the original evolutionary sequence). Aspects of the sequence have always been controversial, and it has been revised and discussed many times since that date. References in this chapter will be to the current version as it appears in Kay and Maffi (1999: 748) and Paul Kay et al. (2009: 11). The sequence claims to display the order in which BCCs may 118 english historical semantics ­ evelop in various languages, as revealed by their BCTs. The contro- d versy mentioned above resulted from early claims that the sequence represents a linguistic universal, that is a feature that we can expect to find in all languages. Its current version, however, is intended to show the acquisition of basic categories by those languages studied in the World Color Survey, a research project at the University of California at Berkeley, which investigated the colour systems of 110 languages from around the world (see Paul Kay et al. 2009 and www1.icsi.berke- ley.edu/wcs/). Even though there remain many languages whose BCCs have not yet been investigated, there is now a great deal of data on this subject, since there have been other multi-­language projects, such as Robert MacLaury’s Mesoamerican Colour Survey (MacLaury 1997), and countless studies of individual languages and language families. Findings from these studies strongly suggest that, although there are differences in the numbers and extent of basic categories in various languages, the historical order of acquisition of the categories is remark- ably consistent. There is a small number of different trajectories (routes) which individual languages have taken along the sequence but, when compared with the theoretically possible number of trajectories, the actual number used is minimal. This suggests that, rather than having no guidance at all on the early history of colour category acquisition, researchers are justified in considering the trajectories on the sequence as being by far the most likely routes taken by communities of speakers in the cognitive development of their colour systems. Let us now con- sider these most likely routes. The two earliest categories acquired involve, firstly, white and the so-­called ‘warm’ hues such as red and yellow, and, secondly, black and the ‘cool’ hues such as green and blue. Because the sequence is concerned only with hues and achromatics (white, black and grey), there is currently no facility for accommodating tone, saturation or brightness in the sequence. These two earliest categories are simply labelled ‘W/R/Y’ (white/red/yellow) and ‘Bk/G/Bu’ (black/green/ blue), often referred to as macro-­white and macro-­black. It seems highly likely, however, that these two concepts represent the opposition of categories such as light/bright/pale/white/warm, and dark/ dull/black/cool, probably best summarised as brightness versus darkness, and based on the prototypes of ‘day’ and ‘night’. As shown below, macro-­white at this stage (Stage I of the sequence) involves the so-­called ‘warm’ colours (red and yellow) and those colours closely related to them such as pink and orange, while macro-­black involves green, blue and sometimes grey and violet. The classic example of a  c. p. biggam 119 Stage I Stage II W/R/Y W Bk/G/Bu R/Y Bk/G/Bu Figure 7.1 Stages I and II in the evolutionary sequence Stage I language is Dani, spoken in Indonesian New Guinea (Biggam 2012: 74). Apart from some apparently rare exceptions, a Stage I language can accommodate all colour impressions in these two basic categories. The evidence shows that, when the first entirely chromatic (hue-­ based) category is developed at Stage II of the sequence, it is most likely to be a macro-­category containing red, yellow and usually other warm hues. It will no longer include white, but brightness may still be involved, as this new macro-­red category is likely to be based on the prototype of ‘fire’. An example of a Stage II language is Ejagham, spoken in the West African countries of Nigeria and Cameroon (Paul Kay et al. 2009: 227–8). The sequence encodes the macro-­red category as R/Y (red/yellow), now cognitively separate from the category encoded W (white), as shown in Figure 7.1.2 A speech community which has developed a macro-­red category will as a result have reduced the extent of the pre-­existing macro-­white category, necessitating a realignment of category boundaries. In other words, colours which initially would have been included in the earlier W/R/Y category, and which could therefore be referred to using the same BCT, will be referred to by the new terms instead. If a community develops further basic categories over Stages III and IV of the sequence, they are most likely to involve an emphasis on green, yellow or grue. The term grue is simply a convenient way of referring to a macro-­category consisting of green+blue and sometimes +grey as well. There are three possibilities at this stage for the BCTs that are likely to emerge. Some communities evolve a separate category of yellow (Y) at Stage III, having separated it cognitively from macro-­ red (R/Y), as shown in Figure 7.2. The Kwerba language of Irian Jaya (Indonesia), when it was investi- gated in 1978, was in transition from Stage II to Stage III (option 1), with 120 english historical semantics Stage II Stage III (option 1) W W R/Y R BK/G/Bu Y BK/G/Bu Figure 7.2 Stages II and III, first option Stage II Stage III (option 2) W W R/Y R/Y Bk/G/Bu G/Bu Bk Figure 7.3 Stages II and III, second option basic categories which can be coded as W, R and Bk/G/Bu and a yellow category which was basic for the majority of the community but not all of them (Paul Kay et al. 2009: 341–3). Other speech communities evolve a grue category (G/Bu) at Stage III, having separated it from macro-­black (Bk/G/Bu), as shown in Figure 7.3. The Chácobo language of Bolivia is a good example (Paul Kay et al. 2009: 167–8). Yet others evolve an unusual macro-­ category consisting of yellow+grue (Y/G/Bu) (Stage III, option 3), as recorded for the Karajá language of Brazil (Paul Kay et al. 2009: 305–7). The three options for Stage III may change in different ways if speakers of those languages experience a need to add a further colour category, thus developing a Stage IV system. The Stage IV options are shown in Figure 7.4. Those communities with a yellow category may develop either a green (G) or a grue (G/Bu) category from their macro-­ black category (Bk/G/Bu). The Ifugao language of the Philippines is an example of Stage IV (option 1) with two BCTs for green (Paul Kay et al. 2009: 279–83).  c. p. biggam 121 Stage III (option 1) Stage IV (option 1) OR Stage IV (option 2) W W W R R R Y Y Y Bk/G/Bu G G/Bu Bk/Bu Bk Figure 7.4 Alternative developments at Stage IV (option 1) Stage III (option 2) Stage IV (option 2) W W R/Y R G/Bu Y Bk G/Bu Bk Figure 7.5 Development at Stage IV (option 2) Stage IV (option 2) is recorded for the Zapotec language of Mexico, which has a BCT for grue (Paul Kay et al. 2009: 579–81). Those communities which developed a grue category at Stage III develop a new yellow (Y) category at Stage IV by separating it from their macro-­red (R/Y) category, thus reaching Stage IV (option 2) by a different route, as shown in Figure 7.5. The rare languages with a yellow+grue (Y/G/Bu) category at Stage III either develop separate yellow (Y) and grue (G/Bu) categories at Stage IV (option 2 again), or they develop a new blue (Bu) category while retaining yellow+green. The latter situation can be seen in the Cree language of Canada (Paul Kay et al. 2009: 201–6). All the World Color Survey languages judged to be at Stage V, the final stage in the current sequence, have separate basic categories for white, black, red, yellow, green and blue, as do ME and ModE. The brown, purple, orange, pink and grey categories, which appeared at Stages VI and VII of the earlier versions of the sequence, are now 122 english historical semantics considered to be less predictable than the other categories. If developed at all, they emerge at various stages of the sequence, as will be shown below for English. 7.5 The development of colour terms in English In this section we’ll apply the evolutionary sequence to developments in the three main periods in the history of English. In addition to the OED and HTOED, this section uses evidence from the period dictionar- ies: the DOE (in progress, letters A–G available), the MED, the AND, a dictionary of Anglo-­French, and the DSL. These are described in section 4.4 and details of how to access them are given in the References section (pp. 191–2). Because of the nature of colour, the earliest sense of each colour term recorded is usually an adjective (though they can all be used as nouns as well). 7.5.1 Old English (OE: 700–1150) The value of the evolutionary sequence to historical semanticists is that its general principles (see section 7.4), as opposed to its details, provide a set of implications and a relative chronology for the devel- opment of basic categories which may or may not be universal, but which are clearly extremely widespread. In other words, there is a strong likelihood that, for example, a language with a BCT denoting green must have developed a basic category for red at an earlier date. The earlier basic categories evolved over many years, even centuries; several of the OE categories were inherited from the language’s ances- tors, such as proto-­Indo-­European and proto-­Germanic (Biggam 2010: 234–52). Judging as best we can from the surviving records, it appears that OE had six BCCs: white/bright, black/dark, red+, yellow, green and grey. It had thus reached Stage IV in the evolutionary sequence, with an additional category of grey. It has five fewer basic categories than ModE, but in addition the content of the categories differs from that of the modern language. The white category is not restricted to one end of the achromatic scale but also includes types of brightness (see section 7.2), and similarly the black category includes darkness. The red category is denoted here as red+ because it includes at least some shades of purple, pink, orange and brown in addition to red. grey includes not only achromatic shades but also various dull hues (Biggam 1998: 86–9). OE also appears to have had some relatively salient terms for the concepts of blue and brown but these were almost certainly  c. p. biggam 123 not basic categories until the ME period. So what were the BCTs which labelled the OE BCCs? Several of the OE BCTs will look familiar to speakers of ModE: white/bright was denoted by hwit (ModE white); red+ by read (red); yellow by geolu (yellow); green by grene (green); and grey by græg (grey). However, the OE colour words do not necessarily denote entirely familiar colour categories, since at least hwit, read and græg cannot always be accurately translated by white, red and grey. Also, the black/dark cat- egory was named by two principal terms which appear to have been in competition. The older term sweart (ModE swarthy) is cognate with other Germanic terms for black, such as Modern German schwarz, but OE blæc (ModE black) was at least equally salient and, judging by later developments, was probably taking over the pre-­eminent position for- merly held by sweart. Although OE shows signs of having incipient blue and brown catego- ries they do not appear to have developed a basic status. The principal term for blue was hæwen, but this word appears to have been strongly associated with grey (Biggam 1997: 128–9). OE brun, the principal term for brown, has long caused astonishment among students of the lan- guage since it can be glossed as ‘dark’ as well as ‘brown’, and was used to translate Latin terms meaning ‘black’, ‘purple’, ‘red’ and ‘dun’ (DOE brun, adj. 1). Most difficult of all to understand is the use of brun in poetry to mean (apparently) ‘shining’ or ‘gleaming’ of metalwork such as swords and helmets (DOE brun, adj. 3). Over the years, scholars have struggled to explain how a word could mean both ‘dark’ and ‘gleaming’; some of the resulting suggestions show how colour terms must be studied in their cultural context, in this case considering aspects of Anglo-­Saxon social and military life. One explanation of the contradictory meanings of brun was offered by Tremaine (1969). He suggested that early medi- eval weapons were literally dark in tone because of a surface treatment known as ­­‘browning’, which involves allowing a thin layer of rust to develop and then polishing the metal to create a brownish colour. This technique has been used in modern times to prevent the sun glinting off the barrels of firearms, thus revealing one’s presence to the enemy; such a use seems unlikely in the Anglo-­Saxon context, which involved the proud display of elite gear such as swords and helmets. Another suggestion concerns the process used in ironworking, which involves heating the metal and quenching it in water (Walker 1952). The blacksmith must choose the right moment to quench a sword-­ blade or it may not have the correct hardness and flexibility. That right moment is determined by the changing colour of the metal in the fire: 124 english historical semantics a brown colour indicates the correct temperature. A further explana- tion, made by the present author, relies on Colour Semantics. I have suggested that brun, and related words in other languages, had their origin in one of the several ancient terms connected with fire. With this history, the word could easily indicate both darkness and brightness, depending on context, since fire involves both the brightness of flames and the darkness of scorched and burnt material. Such a combination of different semantic features is typical of the early macro-­categories (for a fuller discussion of brun, see Biggam 2010: 256–9). Although OE had fewer BCTs than ModE, lacking basic categories for blue, brown, purple, orange and pink, that does not mean Anglo-­ Saxons found it impossible to refer to these hues. They could use non-­ basic terms and expressions, compound terms, and phrases constructed from other basics or non-­basics (such as geoluread ‘yellow-­red’), as well as descriptions and comparisons with referents such as blood, dyes, flowers and more (for example, geole crog ‘saffron yellow’ and rosen ‘made of roses, rose-­coloured’). It is possible to predict that OE would have developed a purple category after the blue and brown categories. This is because there are several surviving colour words which, at least sometimes, appear to indicate a purple hue, whereas the scarcity of expressions for orange and pink suggests a lower salience for those categories. The following terms are among those which researchers have suggested can at least partly denote purple: basu, brunbasu, felleread (an error for pelleread?), hæw- mænged, pællen, scirbasu, purpuren and purpul. This collection of non-­basic colour terms, some closely connected with textiles, can be interpreted as an awakening interest in a colour which had probably been consid- ered either a type of blue (violet shades) or of red (red-­purple shades). However, although it may seem obvious to us that OE purpuren was destined to become a BCT for purple at some future date, this was by no means guaranteed, since the term principally referred not to a colour but to an expensive textile known in Latin as purpura, which was pro- duced in several colours. 7.5.2 Middle English (ME: 1150–1500) As we saw in Chapter 2, the principal lexical difference between OE and ME was the considerable vocabulary adopted from or via French after the Norman Conquest of 1066. In spite of the prestige of Anglo-­French, the six BCTs already established in OE survived with little more than relatively minor changes in pronunciation, spelling and meaning, evolving into the ME words whit, blak, red, yelwe, grene and grei. These  c. p. biggam 125 spellings are used as headwords in the MED; ME spelling was extremely variable (see section 2.3) and a range of other spellings can be found in the sources. ME speakers had a tendency to separate the hue element of colour from the tone, saturation and brightness elements whereas OE speakers had tended to combine them in the semantics of each colour term. The dominance of hue in certain ME terms, especially in BCTs, was at least encouraged by certain cultural innovations of the later Middle Ages such as banners, livery, and the display of heraldry on coats-­of-­arms, all of which encouraged the development and use of strong dyes and paints. Cognitively, a tendency to regard white, black and grey as addi- tional hues probably began in this period, thus weakening the involve- ment of these three categories with brightness, darkness and dull hues respectively (there were plenty of non-­basic terms to deal with these concepts). The ModE speaker would, however, still notice some dif- ferences from his or her own basic categories. For example, ME had no basic categories for purple, pink and orange, so it is likely that ME red, and probably yelwe too, could indicate an orange hue, while red could also still refer to some purples and to pink. During the ME period, a basic category for blue developed, with considerable help from French. French had a colour term bleu which indicated unnatural skin colour such as the paleness of a sick, wounded or elderly person, and the blueness of bruising. For reasons which are not fully apparent, English speakers found no use for the ‘pale’ sense of this word, but adopted it to indicate blue in preference to the OE word hæwen. Even though hæwen could denote very blue things such as blue gemstones and woad (a blue dye), its ME descendant haue was narrowed to indicate a bluish or grey colour, most often in connection with textiles. It was eventually lost to southern English, but survived for longer in northern England and even longer in Scotland, where it denoted bluish, leaden, dull, pale and greenish colours into the twenti- eth century, functioning as a hyponym of blue (DSL, haw adj.). Hæwen is a clear example of how speakers of the same language in geographical proximity do not necessarily have identical colour semantics. The second salient but not basic colour term of OE (brun) eventually evolved into the brown BCT in ME. There was no competition with French terms in this case since Anglo-­French brun was very similar in pronunciation, written form and meaning. It had a primary sense of ‘dark, brown’ and a sub-­sense of ‘shining, burnished’ (AND brun, adj. 1 and 3). It is noteworthy that the same senses of ‘dark’ and ‘shining’ dis- cussed above occurred in both the OE and Anglo-­French words. BCTs for purple, orange and pink seem not to have emerged during 126 english historical semantics this period. ME had a colour term purple, but it did not become basic until EModE. It also had a noun orange referring to the fruit or the tree, but no adjectival colour usage is recorded. It is difficult to determine the situation concerning pink. There was certainly no BCT, but there are several non-­basic ME terms which could be interpreted as ‘pale red’ or ‘pink’, many of which have the rose as their prototype. The rose in question is Rosa canina L., variously called the ‘wild-­’, ‘hedge-­’ or ‘dog-­ rose’, the flower of which is usually a shade of pink (although it can also be white). ME ‘rose’ words often have primary or secondary senses referring to ‘oil of roses’ (also known as ‘attar of roses’), an oil distilled from rose petals, and chiefly used in making perfumes. Thus there are ME terms such as rosen ‘made of or consisting of roses; rose-­coloured, rosy’ (MED adj. a and b) and roset ‘made of a distillation of red or white roses’, also used as a noun denoting ‘rose-­colour’ (MED adj. a and noun d). Other adjectives which can function as colour terms in this category are rosi ‘rose-­coloured, rosy’ and roseat ‘roseate, rosy’ (MED). 7.5.3 Modern English (ModE 1500–) From EModE onwards, the colour vocabulary of English increased enormously, as a glance at the HTOED colour categories reveals. Travellers to the New World discovered dyes such as logwood and some types of cochineal, while Renaissance artists experimented with pigments to introduce new effects to their paintings. Much later, syn- thetic dyes were introduced, beginning with so-­called ‘mauveine’, a purple shade, in 1856. In the same century, the development of indus- trial processes capable of producing identical items which could only be distinguished by their colours encouraged the proliferation of colour terms to identify and market such products. The twentieth century saw the rise of the mass fashion industry with its regular announcements of ‘this year’s colours’. In periods like the 1960s, colour, especially vivid hues, seemed to dominate the cultural scene. All of these factors moti- vated the coining of new colour terms. The burgeoning of the interior décor industry, which has a never-­ending supply of subtle mixes of hues and tones, also brought colour to the forefront of modern minds, result- ing in a torrent of new colour words and phrases. It has been estimated that Modern British English has at least 8,000 colour terms (Kornerup and Wanscher 1978: 140). We have seen that ME speakers probably had eight basic colour categories; the remaining basic categories of purple, orange and pink developed in ModE. After an erratic development from a textile term to a generally red term, ModE purple took on its familiar sense of a  c. p. biggam 127 hue range at the intersection of red and blue from about the fifteenth century. However, it takes time for a speech community to develop an abstract basic category after the earliest recognition of a particular colour sense, so the emergence of purple as an English BCT would have happened later. The adjective purple developed from the OE noun purpure, referring to expensive and prestigious cloth (not always purple in colour) and the clothing made from it. The first recorded occur- rence of purple in an adjectival colour sense is 1415, but its early use is restricted to textiles. However, the amethyst gemstone is described as having ‘a purpill colour’ in a text dating to sometime before 1500, indicating a loosening of the contextual restriction to textiles and dyes (OED purple, adj. 2a). ModE orange developed a colour sense from its ME meanings of the orange fruit and tree. The earliest recorded colour sense of the adjec- tive is from 1532 (OED orange, adj. a) but, as with purple, its development towards basic status would have taken time. The last basic category to develop among English speakers was pink. The earliest recorded colour use of the adjective pink dates to 1607 (OED pink, adj.2 II. 2a), and, once again, it would have required time to become a BCT. Indeed, for some (British) English speakers, it may still be considered a pale red, not quite basic. Colour terms and concepts in English are not restricted to purely descriptive uses: they also have metaphorical, contrastive and symbolic roles in the language, such as the use of red to indicate importance in written texts. They feature in similes, for example as white as a sheet, usually said of a terrified, ill or shocked person’s face. The phrase evokes a white exemplar, a sheet, which in an ideal world should be spotlessly clean and crisp. Its referent, a person’s face, cannot literally be such a pure colour; the meaning is something like ‘a face much paler than normal for a white person’. The use of the white BCT creates a compact expression packed with cultural, non-­literal and contrastive meaning. Colour terms often take on metaphorical meanings which operate in particular societies to convey a strength of meaning which the standard word for the concept may lack. Thus a person may be called yellow, which implies cowardice in many varieties of English, and is somehow more damning than the straightforward word cowardly. Similarly, feeling blue seems more depressing than simply feeling sad. Most of these meta- phorical uses of colour terms involve BCTs, as their long history in the language has allowed their meanings to develop and change. 128 english historical semantics 7.6 The changing nature of a basic category: blue A BCC has not only at least one BCT but also several non-­basic terms indicating only parts of the category or the use of that colour in only a particular type of context. This means that, even if the BCTs of two languages are easily translatable, the composition, nature and extent of their categories are likely to be different. It is important to be aware that ModE and earlier English had markedly different categories, and this difference has implications for our understanding of historical records and our appreciation of older literature. The adjectives of the blue cat- egory throughout the history of English provide an interesting example. OE included the following adjectives which denoted a simple blue or partly blue meaning: blæwen, glæsen, hæwen and wæden (Biggam 1997). Blæwen probably denoted a dark blue in the context of textiles and dyes, wæden was used in the same context, and glæsen is likely to have denoted a pale blue shade, especially used of eyes. The principal blue term, hæwen, appears in several compounds, such as blæhæwen ‘dark blue, dark grey’, grenehæwen ‘greenish blue, greenish grey’ and swearthæwen ‘dark blue’, allowing reference to a variety of shades involving blue. The blue category in ME was influenced by French more than any other colour category, and its vocabulary changed completely. Other categories usually acquired a large number of new colour terms from French (and some from Latin) while also retaining many English words. Even though we must make allowance for the smaller number of records surviving from Anglo-­Saxon England, it seems apparent that the number of colour terms increased considerably in ME. The ME adjectives denoting simple blue or partly blue meanings are: azure (asur), blae/blo, blue (bleu), glawke (glauk), haw (haue), inde, perse, plunket, sap- phire (saphir), sapphirine (saphirine), venet, violet and wachet (HTOED; MED headwords are given in brackets where their spelling differs). These words covered a wide range of meanings. For example, a vivid, strong blue could be saphir or saphirine; a pale blue could be wachet; a dull, greyish blue could be perse or plunket; and a livid blue (often referring to bruising) could be blo or, in the north of England or Scotland, blae or haw. In surviving ME records, the contexts of several blue terms appear to be dominated by descriptions of clothes, textiles and dyes, especially inde, plunket and wachet. The principal blue term, blue (bleu), could be quali- fied by other terms, such as bright (bright or vivid), dep (deep, intense), inde (dark), light (bright), sad (dark) and violet (purplish), but they did not usually form compound terms, as had been the case with OE hæwen. In ModE, the BCT blue is the superordinate of a veritable army of terms which have been used or are still used for blues in the modern  c. p. biggam 129 period; a mere selection is given here. Simple terms include sapphiric (vivid/fully-­saturated), cyaneous, cobalt, lazuline (dark, strong), azurine, periwinkle (pale), cerulean, lazure (pale and bright), turquoise, aquamarine (greenish), gunpowder, Silurian (greyish), hyacinthine, violaceous, lavender (purplish). Colour phrases and compound terms abound, including electric blue, powder-­blue, coal-­blue, lead-­blue, sky-­blue, indigo-­blue; many of these terms can be further qualified by words such as dark and pale (see section 7.2). However, not content with these verbal riches, the fashion and paint industries invent fanciful words and phrases to name each shade, often replacing them regularly as they become well known. Blue examples, taken at random from paint charts, include forget-­me-­not, king- fisher, Oxford blue, Indian Ocean, Cambria, sea satin, moon shadow, Trafalgar and porcelain blue. 7.7 Summary Leaving aside the non-­basic colour terms and concentrating on hues, we can summarise the major changes to colour categories and terms that have taken place in English as follows. The basic (cognitive) colour categories (BCCs) of OE, ME and ModE are: OE: (1) white/bright, (2) black/dark, (3) red+, (4) yellow, (5) green, (6) grey ME: (1) white, (2) black, (3) red+, (4) yellow, (5) green, (6) grey, (7) blue, (8) brown ModE: (1) white, (2) black, (3) red, (4) yellow, (5) green, (6) grey, (7) blue, (8) brown, (9) purple, (10) orange, (11) pink The BCTs of OE, ME and ModE are: OE: hwit, blæc/(sweart), read, geolu, grene, græg ME: whit, blak, red, yelwe, grene, grei, bleu, broun ModE: white, black, red, yellow, green, grey, blue, brown, purple, orange, pink 7.8 Conclusion Colour is an integral feature of our world, and is deeply engrained and widely applied in the cognition and language of English speakers. However, as this chapter has shown, English BCCs and BCTs have changed over the centuries and non-­basic colour vocabulary has grown. These changes inevitably affect how the modern reader understands historical texts. For example, he or she must consider how much of 130 english historical semantics the modern concept of pink lies behind earlier uses of words meaning ‘red’, ‘pale red’, or ‘the colour of the rose’; whether an earlier use of a ‘grey’ word could mean ‘dull blue’ or ‘dull green’ rather than ‘grey’; and whether an earlier use of a ‘brown’ word could mean ‘shining, flashing’ rather than ‘brown’. Such possibilities also have implications for the meanings of other contemporary colour terms, since a society’s conceptualisation and naming of colour constitutes an interlocking and interdependent system: where one category changes, others must change to accommodate it. However, bearing in mind the differences of colour usage among contemporary speakers, Colour Semantics can often provide the modern reader with insights into the age, gender, lin- guistic expertise or locality of an earlier writer. The potential of Colour Semantics for providing cultural glimpses into past societies is only beginning to be explored. Exercises 1. The colour adjective pink is problematic in many ways, including its derivation. Use the OED or an etymological dictionary to examine the candidates for its origin, and decide which you think is most likely. 2. Make an examination of the HTOED adjective category Blue (01.04.09.07.06 in the print edition). For a selection of words, con- sider such factors as the dates when the words are first recorded, their origin, their referents in the external world and any contextual restrictions they may have. Further information about the words and their origins is available in the OED. 3. Use the OED to find metaphorical uses of the English BCTs. What are the earliest uses you can find? Have these uses survived into ModE? When have new uses entered the language, and can they be related to cultural developments at the time? 4. Use the resources at your disposal to try to establish when purple became a BCT in English. You may find the information about spelling variants and etymology at the head of the OED entry useful, especially if you plan to use other resources. Further reading Biggam (2010) offers further information on topics covered in this chapter, while Biggam (2012) is the most comprehensive account of Historical Colour Semantics available. Burnley (1976) and Biggam (1993) deal with aspects of colour in ME. Three chapters in Anderson  c. p. biggam 131 et al. (2014) use HTOED to examine the development of colour, espe- cially in metaphor: Alexander and Kay; Anderson and Bramwell; and Hamilton. There is also a keynote chapter by Biggam on the history of colours. There is a section on OE colour terms in the online package Learning with the Online Thesaurus of Old English, available at http:// www.gla.ac.uk/schools/critical/research/fundedresearchprojects/ learningwiththeonlinethesaurusofoldenglish/ For quick reference, a handy dictionary of OE is A Concise Anglo-­Saxon Dictionary ( 1960, ed. J. R. Clark Hall). Notes 1. Small capitals are used throughout this chapter to indicate seman- tic categories (as opposed to words or phrases). This convention is usually followed in Colour Studies. 2. The category codes used in the current evolutionary sequence are the following: W (white), Bk (black), R (red), Y (yellow), G (green) and Bu (blue). Macro-­categories are shown by linking these codes with forward slashes, as in W/R/Y (macro-­white). Roman numerals are used to identify stages in the sequence.