12 R. Brownand A. Gilman ‘The Pronouns of Power and Solidarity R. Brownand A. Gilman, "The Pronouns of Power and Solidarity’ nT. A. Sebeok (ed), Spi ln Language, MIT Press, 960, pp- 253-76. Most of us in speaking and writing English use only one pronoun of address; we say ‘you’ to many persons and “you” to one per son. The pronoun ‘thou’ is reserved, nowadays, to prayer and nalve poetry, but inthe past it was the form of familiar address to a single person. At that time ‘you’ was the singular of reverence and of polite distance and, also, the invariable plural. In French, German, Italian, Spanish'and the other languages most neatly related to English there are still active two singular pronouns of address. The interesting thing about such pronouns is their close association with two dimensions fundamental tothe analysis of all social life ~ the dimensions of power end solidarity. Semantic and stylistic analysis of these forms takes us well into psychology and sociology as well as into linguistics and the study of litera ture. ‘his paper is divided into Sve major sections. The frst three of these are concemed with the semantics of the pronouns of address. By semantics we mean covariation between the pronoun, used and the objective relationship existing between speaker and. addressee. The first section offers a general description of the semantic evolution ofthe pronouns of address in eettain Euro- pean languages. The second section describes semantic dilfer- ences existing today among the pronouns of French, German and Italian, The third section proposes a connection between social structure, group ideology, and the semantics ofthe pronoun. The final two sections of the paper are conoermed with expressive style by which we meant covariation between the pronoun used 1. Our study ms anced by a Grantin-Ai-of Reach made by te Ford Foundation to Brow, and the authors gratefully acknowledge this ‘ssstance, 252 Language and Soci! Structures Sin Uaongurge bak sae O Gntoyt ed. Pie Parte Ggiiree Now York. Panguin- (4F2. and characteristics of the person speaking. The first of these sections shows that a man's consistent pronoun style gives away his class status and his political views. Th last section descsibes the ways in which a man may vary his pronoun style from time to time 80 as to express transient moods and attitudes. In this section its also proposed that the major expressive meanings are ‘derived from the major semantic rules. ‘In each section the evidence most important to the thesis of that section is described in detail. However, the various general- ‘ations we shall offer have developed as an interdependent sot froma continuing study of our whole assemblage of facts, and soit ‘may be well to indicate here the sort of motley assemblage this i. Among secondary sources the general language histories (Baugh, 1935; Brunot, 1937; Diez, 1876; Grimm, 1898; Jespersen, 1905; Meyer-Lubke, 1900) have been of litte use because their central ‘concern is always phonetic rather than semantic change. How- ‘ever, there are a small number of monographs and doctoral dissertations describing the detailed pronoun semantics for one oF another language ~ sometimes throughout its history (Gedike, 1794; Grand, 1930; Sohaston, 1904; Schliebitz, 1886), some- times for only a century or so (Kennedy, 1915; Stidston, 1917), ‘and sometimes for the works ofa particular author yme, 1936; Fay, 1920). As primary evidence for the usage of the past we have drawn on plays, on legal proceedings (Jardine, 1832-5), and on letters (Devereux, 1853; Harrison, 1935). We have also learned about contemporary usage from literature but, more importantly, from long conversations with native speakers of French, Italian, ‘German and Spanish both here and in Europe. Our best informa- tion about the pronouns of today comes fron @ questionnaire concerning usage which is described in the second section of this paper. The questionnaire has thus far been answered by the fol owing numbers of students from abroad who were visiting in Boston in 1957-8: 50 Frenchmen, 20 Germans, 11 Italians and ‘wo informants, each, from Spain, Argentina, Chile, Denmark, Norway, Sweden, Israel, South Africa, India, Switzerland, Hok- land, Austria and Yugostavia. ‘We have far more information concerning English, French, IRalian, Spanish and German than for any other languages. Informants and documents concerning the other Indo-European IR. Brown and A. Gilman 253‘address to the principal power in the state and eventually general- {zed to the powers within that microcosm of the state the nuclear family. In the history of languzge, then, parents are emperor figures. It is interesting to note in passing that Freud reversed this terminology and spoke of kings, as well as generals, employers and priests, as father figures. The propriety of Freud's designation for bis psychological purposes derives from the fact that an individual learning a European language reverses the historical order of semantic generalization. The individual's first experience of subordination to power and of the reverential V comes in his relation to his parents. In later years similar asymmetrical power relations and similar norms of address develop between employer ‘and employee, soldier and officer, subject and monarch. We can see how it might happen, as Freud believed, that the later social relationships would remind the individual of the familial proto type and would revive emotions and responses from childhood. In.a man’s personal history recipients of the nonreciprocal V are parent figures. Sinoe the nonreciprocal power semantic only prescribes usage between superior and inferior, it calls for a sodial structure in ‘which there are unique power ranks for every individual. Medieval European societies were not so finely structured as that, and so the power semantic was never the only rule for the use of T and Y. There were also norms of address for persons of roughly equiv- alent power, that is, for members of a common class. Between equals, pronominal address was reciprocal; an individual gave and received the same form. During the medieval period, and for varying times beyond, equals of the upper classes exchanged the smutual V and equals of the lower classes exchanged 7. ‘The difference in class practice derives from the fact that the reverential V was always introduced into a society at the top. In the Roman Empire aly the highest ranking persons had any occa- sion to address the emperor, and so at first only they made use of Vin the singular. In its later history in other parts of Europe the reverential V was usually adopted by one court in imitation of another. The practice slowly disseminated downward in a society. In this way the use of Vin the singular incidentally came to connote a speaker of high status. In later cennuries Europeans ‘became very conscious of the extensive use of V as a mark of 256 Language and Soclal Structures, elegance. Inthe drama of seventeenth century France the nobility and bourgeoisie almost always address one another as V, This js true even of husband and wife, of lovets, and of parent and child if the child is adult, Mme de Sévigné in her correspondence never uses T, not even to her daughter the Comtesse dé Grignan Gobliebitz, 1886). Servants and peasantry, however, regularly used among themselves. For-many centuries French, English, Italian, Spanish, and German pronovr usage followed the rule of nonreciprocal T-V between persons of unequal power and the rule of mutual Vor T. (according to social-class membership) between persons of roughly equivalent power, There was at first no rule differentiat- ing address among equals but, very gradually, a distinction doveloped which is sometimes called the T of intisuacy and the V of formality. We name this second dimension solidarity, and here is out guess as to how it developed, ‘The solidarity semantic ‘The original singular pronoun was T. The use of Vin the singular developed as a form of address ‘0 a person of superior power. ‘There are many personal atributes that convey power. The recipi- ceat of Yay differ from the recipient of Tin strength, age, wealth, birth, sex or profession. As two people move apart on these powier-laden dimensions, one of them begins to say V. In general terms, the V form is linked with differences between persons. ‘Not al differences between persons imply a difference of power. Men are born in different cities, belong to different families of the same status, may attend different but equally prominent schools, may practice different but equally respected professions, A rule for making distinctive use of T and V among equals can >be formulated by generalizing the power semaatic. Differences of power cause V to emerge in one direction of address; differences not concerned with power cause V to emerge in both directions. ‘The relations called older than, parent of, employer of, richer than, sronger than, and nobler than are all asymmetrical. If A is oles than B, B is not older than A. The relation called ‘more powerful than’, which is abstracted from these more specific relations, is also conceived to be asymmetrical. The pronoun usage expressing this power relation is also asymmetrical or non- R, Brown and A. Gilman 257reciprocal, with the greater receiving V and the lesser 7. Now we are concerned with a new set of relations which are symmetrical; for example, attended the same school ox have the same parents or Practice the same profession. If A has the same parents as B, B has the same parents as A. Solidarity is the name we give to the general relationship and solidarity is symmetrical. The corees- ponding norms of address are symmetrical or reciprocal with ‘becoming more probable as solidarity declines, The solidary 7" reaches a peak of probability in address between twin brothers or in a man’s soliloquizing address to himself. Not every personal attribute counts in determining whether ‘two people are solidery enough to use the mutual 7. Eye color oes not ordinarily matter nor does shoe size. The similarities ‘that matter seem to be those that make for like-mindedness or similar behavior dispositions. These will ordinarily be such things 48 political membership, family, religion, profession, sex, and birthplace. However, extreme distinctive values on almost any dimension may become significant. Height ought to make for solidarity among giants and midgets. The T of solidarity can be Produced by frequency of contact as well as by objective simi retainer and as remote as the waiter in a strange restaurant, Extension of the solidarity dimension along the dotted lines of Figure tb creates six categories of persons defined by their rela- tions to a speaker. Rules of address are in conflict for persons in the upper left and lower right categories. For the upper left, power indicates V and solidarity 7. For the lower tight, power indicates T and solidarity V, ‘The abstract conflict described in Figure 1b is particularized in Figure 2a with a sample of the social dyads in which the con- flict would be felt. In each case usage in one direction is un~ ‘equivocal but, in the other direction, the two semantic forces are ‘opposed. The first three dyads in Figure 2a involve conflict in address to inferiors who are not solidary (the lower right category Of Figure 1b), and the second three dyads involve comlict in address to superiors who are solidary (the upper left category in Figure 1b). vi Supatons v arities. However, frequent contact does not necessarily lead to the ‘mutual 7. It depends on whether contact results in the discovery or creation of the like-mindedness that seems to be the core of the solidarity semantic, Solidarity comes into the European pronouns as a means of differentiating address among power equals. It introduces @ second dimension into the semantic system on the level of power equivalents. So long as solidarity was confined to this Jevel, the ‘two-dimensional system was in equilibrium (see Figure 1a), and ittseems to have remained here for a considerable time in all our Janguages. It is from the long reign of the two-dimensional semantic that T derives its common definition as the pronoun of either condescension or intimacy and V its definition as the pro- ‘noun of reverence or formality. These definitions are still current ‘but usage has, infact, gone somewhat beyond them. ‘The dimension of solidarity is potentially applicable to all persons addressed. Power superiors may be solidary (parents, elder siblings) or not solidary (officials whom one seldom sees), Power inferiors, similarly, may be as solidary as the old family 258 Language and Social Structures ua and sldry unl and net slay 7 v VA Supeiorandestdory ‘Supaterandnetzoidey — ¥ ual an eld Equaland not coldary T v at eS TY tteacand bie Iescasdnatsidey YT Figure 1 Tho two-dimensional semantic (a) in equilibrium and (b)-under tension R. Brown and A. Gilman 259‘Well into the nineteenth century the power semantic prevailed and waiters, common soldiers and employees were called T while parents, masters and elder brothers were called V. However, all ‘our evidence consistently indicates that in the past century the solidarity semantic has gained supremacy. Dyads of the type shown in Figure 21 now reciprocate the pronoun of solidarity ‘o¢ the pronoun of nonsolidarity. The conflicted address has been resolved s0 as to match the unequivocal address, The abstract result is a simple one-dimensional system ‘with the reciprocal T for the solidary and the reciprocal for the nonsolidary. Itis the present practice to reinterpret power-laden attributes 80 as to turn them into symmetrical solidarity attributes. Rela- tionships like older than, father of, nobler than, and richer than are now reinterpreted for purposes of T and Vas relations of ‘the same age as, the same family as, the same kind of ancestry a3, ‘and the same income as. In the degree that these relationships hold, the probability of a mutual 7 increases and, in the degree that they do not hold, the probability of a mutual V increases, ‘oxtomer Otter pope TW ty ay te ny lv Water Setter Eon Pant Mester Eder Boor Hh rly 7 rv ryt ‘en Fett savent Youroartxthr Cater ofteer Enoloer iv v v Watar Solsee payee Pact Maser ‘deter " + " ‘Sen Fattulerant Younger » Figure 2. Social dyads involving (2) semantic conflict and (b) their rasolution 260 Language and Social Structures ‘There is an interesting residual of the power relation in the contemporary notion that the right to initiate the reciprocal T ‘belongs to the member of the dyad having the better power-based claim to say T without reciprocation. The suggestion that solid- arity be recognized comes more gracefully from the elder than from the younger, from the richer thas, eam the poorer, from the employer than from the employee, from the noble than from. the commoner, from the female than from the male. Tn support of our claim that solidarity has largely won out over power we can offera few quotations from language scholars. Litteé (4882), writing of French usage, says: ‘Notre courtoisie est méme si grande, que nous ne dédaignons pas de donner du ‘yous et duu monsieur & homme de la condition ta plus vile." Grand (1930) wrote of the Ttalian V: ‘On commence aussi & le donner aux personnes de service, & qui on disait tu autrefois.” We hhave found no authority who describes the general character of these many specific changes of usage: a shift from power 10 solidarity as the governing semantic principle ‘The best evidence that the change has occurred isin our inter- views and notes on contemporary literature and films and, most importantly, the questionnaire results. The six social dyads of Figure 2 were all represented in the questionnaire. {n the past ‘thase would have been answered in accordance with asymmetrical ‘power, Across all six of these dyads the French results yield only I per cent nonreciprocal power answers, the German 12 per cent, the Italian 27 per cent. In all other cases the usage is reciprocal, 8 indicated in Figure 2b. In all three of the languages, address between master and servant retains the greatest power loading. ‘Some of the changes toward solidarity are very recent. Only since the Second World War, for instance, has the French Army adopted a regulation requiring officers o say V to enlisted men, Finally, itis our opinion that a still newer direction of semantic shift can be discerned in the whole collection of languages stud- ied. Once solidarity has been established as the single dimension distinguishing T from V the province of T proceeds to expand. ‘The direction af change is increased in the number of relations defined as solidary enough to merit a mutual Tan, in particular, to regard any sort of camaraderie resulting from common task fr a common fate as grounds for T. We have a favorite example , Brown and A. Gilman 261of this new trend given us independently by several French in- formants. It seems that mountaineers above a certain critical altitude shift to the mutual 7, We like to think that this is the point where their lives hang by a single thread. In general, the mutual 7 is advancing among fellow students, fellow workers, ‘members of the same political group, persons who share a hobby or take a trip together. We believe this is the direction of current change because it summarizes what our informants tell us about the pronoun usage of the ‘young people’ as opposed to that of ‘older people. Contemporary differences among French, Italian and German While T and V have passed through the same general semantic sequence in these three languages, there are today some differences of detailed usage which were revealed by the questionnaire data, Conversations with native speakers guided us in the writing of questionnaire items, but the conversations themselves did not ‘teach us the characteristic semantic features of the three languages; these did not emerge until we made statistical comparison of answers to the standard items of the questionnaire. ‘The questionnaire is in English. Tt opens with a paragraph informing the subject that the items below all have reference to the use of the singular pronouns of address in his native language. ‘There are 28 items in the full questionnaire, and they all have the form of the following example from the questionnaire for French students: 1.(a) Which pronoun would you use in speaking to your mother? T (definitely), — T (probably), — Possibly 7, possibly ¥ — V (probably) ~ ¥ (definitely) — 1.(b) Which would she use in speaking to you? T (@efinitely), — T (probably) — Possibly 7, possibly V — V probably) — ¥ (definitely). — 262 Language and Social Stuctures ‘The questionnaireasks about usage between the subject and his ‘other, bis father. his grandfather, his wife, a younger brother Who is a child, a matried elder brother, that brother's wife, a remote male cousin, and an elderly female servant whom he has known from childhood. It asks about usage between the subject and fellow students at the university at home, usage to a student from home visiting in America, and usage to someone with ‘whom the subject had been at school some years previously. It asks about usage to a waiter in a restaurant, between clerks in an office, fellow soldiers in the army, between boss and employee, army private and general. Tn addition, there are some rather elaborate items which ask the subject to imagine himself in some carefully detailed social situation and then to say what pronoun he would use. A copy of the full questionnaire may be had on application to the authors. ‘The most accessible informants were students from abroad resident in Boston in the fall of 1957. Listings of such students ‘were obtained from Harvard, Boston University, MIT, and the Office of the French Consul in New England. Although we have data from a small sample of female respondents, the preseat analysis is limited to the males. All the men in the sample have been in the United States for one year or less; they come from cities of over 300,000 inhabitants, and these cities are well scat- tered across the country in question. In addition, all members of the sample are from upper-middle-class, professional families. This homogencity of class membership was enforced by the factors determining selection of students who go abroad. The ‘cocasional informant from a working-class family is deliberately ‘excluded from these comparisons. The class from which we draw shows less regional variation in speech than does the working class and, especially, farmers. At the present time we have com- plete responses from 50 Frenchmen, 20 Germans and 11 Italians; many of these men also sent us letters describing their under standing of the pronouns and offering numerous valuable anec- dotes of usage. The varying numbers of subjects belonging to the three nationalities result from the unequal representation of these nationalities among Boston students rather than from national characterological differences in willingness to answer a question- R, Brown and A, Gilman 263naire, Almost every person on our lists agreed to serve as an informant. In analyzing the results we assigned the numbers 0-4 to the five response alternatives to each question, beginning with ‘Definitely "as 0, A rough test was made of the significance of the differ- ‘rices among the three languages on each question. We dichoto= mized the replies to each question into : (a) all replies of either ‘Definitely 7° or ‘Probably 7"; (b) all replies of ‘Definitely ¥" or “Probably ¥” or ‘Possibly V, possibly 7°. Using the chi-squared test with Yates's correction for small frequencies we determined, for each comparison, the probability of obtaining by chance a difference as large or larger than that actually obtained. Even with such small samples, there were quite a few differences sig- nificantly unlikely to occur by chance (P= .05 or less). Germans were more prone than the French to say 7'to their grandfathers, to an elder brother’s wife, and to an old family servant. The French were more prone than the Germans to say T to a male fellow student, to a student from home visiting in America, to a fellow clerk in an office, and to someone known previously as a fellow student. Italians were more prone than the French to say T to.a fernale fellow student and also to an attractive girl to whom. they had recently been introduced. Italians were more prone than the Germans to say 7 (0 the persons just described and, in ad- dition, to a male fellow student and to @ student from home visit- ing in America, On no question did either the French or the Germans show a significantly greater tendency to say T than did ‘the Italians, ‘The maay particular differences among the three languages are susceptible of a general characterization. Let us first contrast Geran and French, The German T is more reliably applied within the family than is the French T; ia addition to the si nificantly higher T scores for grandfather and elder brother's wife there are smaller differences showing a higher score for the Ger- man T on father, mother, wife, married elder brother, and remote male cousin. The French T'is not automatically applied to remote relatives, but it is more likely than the German pronoun to be used to express the camaraderie of fellow students, fellow clerks, fellow countrymen abroad, and fellow soldiers. In general it may be said that the solidarity coded by the German T'is an ascribed 284 Language and Soci ‘Structures solidarity of family relationships, The French 7, in greater degree, codes an acquired solidarity, not founded on family relationship but developing out of some sort of shared fate, As for the Italian Tit very nearly equals the German in family solidarity and it sur- ‘passes the French in camaraderie, The camaraderie of the Italian ‘male, incidentally, is extended to the Italian female; unlike the French or German student the Italian says T to the co-ed almost as readily as to the male fellow student. ‘There is avery abstract semantic rule governing Tand V which is the same for French, Gerinan, and Italian and for many other Tanguages we have studied, The rule is that usage is reciprocal, T becoming increasingly probable and V less probable as the num- ber of solidarty-producing attributes shared by two people creases, The respect in which French, German, and Italian differ from one another is in the relative weight given to various attributes of persons which can serve to generate solidarity. For German, ascribed family membership isthe important attribute; French and Italian give more weight to acquired characteristics. ‘Semantics, social structure and ideology A historical study of the pronouns of address reveals a set of semantic and social psychological correspondence. The non- reciprocal power semantic is associated with a relatively static society in which power is distributed by birthright and is not sub- ject to much redistribution. The power semantic was closely tied with the feudal and manorial systems. In Ttaly the reverential pronoun Lei which has largely displaced the older voi was origin ally an abbreviation for fa astra Signoria ‘your lordship’ and 19 ‘Spanish westra Merced’ your grace’ became the reverential usted. ‘The static social structure was accompanied by the Church's teaching that each man had his properly appointed place and ‘ought not to wish to rise above it. The reciprocal solidarity semantic has grown with social mobility and an equalitarian ideology. The towns and cities have led the way in the semantic change as they led the way in opening society to vertical move ‘ment. In aldition to these rough historical correspondences we have made a collection of lesser items of evidence favoring the thesis. In France the nonreciprocal power semantic was dominant R. Brown and A. Gilman 265until the Revolution when the Committee for the Public Safety condemned the use of V as a feudal remnant and ordered a uni- versal reciprocal T. On 31 October, 1793, Malbec made a Parlia~ ‘mentary speech against V: “Nous distinguons trois personnes pour Je singulier et trois pour le pluriel, et, au mépris de oette regle, esprit de fanatisme, d’orgueil et de féodalité, nous a fait contracter Phabitude de nous servir de la seconde personne du plurie! lorsque nous parlons un seul” {quoted in Brunot, 1927). For a time revolutionary ‘fraternité” transformed all addsess into the mutual Cifoyen and the mutual 14. Robespierre even address- ced the president of the Assembly as ru, In later years solidarity declined and the differences of power which always exist every= where were expressed once more. ‘It must be asked why the equalitarian ideal was expressed in a universal rather than a universal V or, as a third alternative, ‘why there was not a shift of semantic from power to solidarity with both pronouns being retained. The answer lies with the ancient upper-class preference for the use of ¥. There was animus against the pronoun itself. The praaoun of the “sans- culotes" was T (Gedike, 1794), and so this had io be the pronoun, of the Revolution. ‘Although the power semantic has largely gone out of pronoun use in France today native speakers are nevertheless aware of i In part they are aware of it because it prevails in 50 much of the ‘greatest French literature. Awareness of power as a potential factor in pronoun usage was revealed by our respondents’ special attitude toward the saying of T to a waiter. Most of them felt that this would be shockingly bad taste in a way that other norm violations would not be, apparently because there is a kind of seignorial right to say Tto a waiter, an actoal power asymmetry, which the modern man’s ideology requires him to deny. In French Africa, on the other hand, it is considered proper to rec ‘ognize a caste difference between the African and the European, and the nonreciprocal address is used to express it. The European says Tand requires V from the African. This isa galling custom to the African, and in 1957 Robert Lacoste, the French Minister residing in Algeria, urged his countrymen to eschew the practice. ‘in England, before the Norman Conquest, ‘ye" was the second person plural and ‘thou’ the singular. «You’ was originally the 266 Language and Social Structures accusative of ‘ye’, but in time it also became the nominative plural and ultimately ousted ‘thou’ as the usual singular. The first uses of ‘ye as a reverential singular occur in the thirteenth, century (Kennedy, 1915), and seem to have been copied from the French nobility. The semantic progression corresponds roughly to the general stages described in the first section of this paper, except that the English seem always to have moved more freely from one-form to another than did the continental Europeans Gespersen, 1905). In the seventeenth century ‘thou’ and ‘you’ became explicitly {avolved in sociat controversy. The Religious Society of Friends (or Quakers) vas founded in the middie ofthis century by George Fox. One of the practices setting off this rebellious group from the larger society was the use of Plain Speech, and this entailed saying ‘thou’ to everyone. George Fox explained the practice in these words: “Moreover, when the Lord sent me forth into the world, He forbade ‘me to put off my bat to any, high or Jow; and I was required to Thee ‘and Thou all men and worten, without any respect to rich or poor, great or smal!” (quoted in Estrich and Sperber, 1946). Fox wrote a fascinating pamphlet (Fox, 1660), arguing that T 10 one and ¥ to many is the natural and logical form of address in all languages. Among others he cites Latin, Hebrew, Greck, ‘Arabick, Syriack, Aethiopic, Exyptian, French, and Italian, Fox suggests that the Pope, in his vanity, introduced the corrupt and illogical practice of saying V 10 one person. Farnsworth, another carly Friend, wrote a somewhat similar pamphlet (Farnsworth, 1655), in which he argued that the Scriptures show that God and ‘Adam and God and Moses were not t00 proud to say and receive the singular 7. For the new convert to the Society of Friends the universal T ‘vas an especially difficult commandment. Thomas Ellwood has described (3714) the trouble that developed between himself and bis father: But whenever I had oocasion to speak to my Father, though I had no Hat now to offend him; yet my language did as much: for} dust ot say YoUtohim, but rHov or THEE, as the Occasion required, and then ‘Would he be sure tofall on me with his Fists, R, Brown and A. Gilman 267‘The Friends’ reasons for using the mutual T were much the same as those of the French revolutionaries, but the Friends were al- ‘ways a minority and the larger society was antagonized by their ‘Violations of decorum. ‘Some Friends use ‘thee’ today; the nominative “thou’ has been dropped and ‘thee’ is used as both the riominative and (as form- erly) the accusative. Interestingly many Friends also use ‘you’. “Thee” is likely to be reserved for Friends among themselves and ‘you’ said to ontsiders. This seems to be a survival of the solid- arity semantic, In English at large, of course, ‘thou’ is no longer used. The explanation ofits disappearance is by no means certain; however, the forces at work seem to have inchided a popular reaction against the radicalism of Quakers and Leveless and also fa general trend in English toward simplified verbal inflection, In the world today there are numerous examples of the association proposed between ideology and pronoun semantics, In Yugoslavia, our informants tell us, there was, for a short time following the establishment of Communism, a universal mutual TT of solidarity. Today revolutionary esprit has declined and V hhas retuned for much the same set of circumstances as in Italy, France, or Spain. There is also some power asymmetry in Yugo" salvia's ‘Socialist manners’. A soldier says V and Comrade General, but the general addresses the soldier with T and sur= ‘name. tis interesting in our materials to contrast usage in the Afri- ‘kaams language of South Africa and in the Gujerati and Hindi languages of India with the rest ofthe collection. On the question= naire, Afrikaans speakers made eight nonreciprocal power dis- tinctions; especially notable are distinctions within the family ‘and the distinctions between customer and waiter and between boss and clerk, since these are almost never power-coded in French, Italian, German, etc., although they once were. The Afrikaans pattern generally preserves the asymmetry of the ‘dyads described in Figure 2, and that suggests a more static society and a less developed: equalitarian ethic, The forms of address used between Afrikaans-speaking whites and the groups of ‘coloreds’ and ‘blacks’ are especially interesting. The Afri- kkaaner uses 7, but the two lower castes use neither T'nor V. The intermediate caste of ‘colareds’ says Meneer to the white and the 268 Language end Social Structures “blacks? say Baas. It is as if these social distances transcend ‘anything that can be found within the white group and so require their peculiar linguistic expressions. ‘The Gujerati and Hindi languages of India have about the ‘same pronoun semantic, and it is heavily loaded with power. ‘These languages have all the asymmetrical usage of Aftikaans and, in addition, use the nonreciprocal T and V between elder brother and younger brother and between husband and wife, ‘This sruly feudal pronominal pattecr ic consistent with the static Indian society. However, that society is now changing rapidly and, consistent with that change, the norms of pronoun usage are also changing. The progressive young Indian exchanges the ‘mutual T with his wile. In our account of the general semantic evolution of the pro- ‘nouns, we have identified a stage in which the solidarity rule was limited to address between persons of equal power. This seemed to yield a two-dimensional system in equilibrium (see Figure 1a), sand we have wondered why addvess did not permanently stabil- ie there. Its possible, of course, that human cognition favors the binary choice without contingencies and so found its way to the suppression of one dimension. However, this theory does not account for the fact that it was the rule of solidarity that tri umphed. We believe, therefore, that the development of open societies with an equalitarian ideology acted against the non reciprocal power semantic and in favor of solidarity. It is our suggestion that the larger social changes created a distaste for the face-t0-Tace expression of differential power. ‘What of the many actions other than nonreciprocal T and ¥ Which express power asymmetty? A vassal not only says V but also bows, lifts his cap, touches his forelock, keeps silent, leaps tw obey. There area large number of expressions of subordination Which are patterned isomorphically with T and V. Nor are the pronouns the only forms of nonreciprocal address. There are, in addition, proper names and titles, and many of these operate today on a nonreciprocal power pattern in America and in Europe, in open and equalitarian societies. In the American family there are no discriminating pronouns, but there are nonteciprocal norms of address. A father says ‘Tim’ to his son but, unless he is extraordinarily ‘advanced’, he R. Brown and A. Gilman 268he ee “a be be - i & oes not anticipate being called “Jack” in reply. In the American South there are no pronouns to mark the caste separation of Negro and white, but there are nonreciprocal norms of address. ‘The white man is accustomed to call the Negro by his frst name, but he expects to be called ‘Mr Lezeee’. In America and in Europe there are forms of nonreciprocal address forall the dyads of asymmetrical power; customer and waiter, teacher and student, father and son, employer and employer. Differences of power exist in a democracy as in all societies. ‘What is the difference between expressing. power asymmetry in pronouns and expressing it by choice of title and proper name? It seems to be primarily a question of the degree of linguistic compulsion. In face-to-face address we can usually avoid the use of any name or title but not so easily the use of a pronoun. Even if the pronoun can be avoided, it will be implicit in the inflection of the verb. ‘Dites quelque chose’ clearly says vous to the French- ‘man. A norm for the pronominal and verbal expression of power ‘compels a continuing coding of power, whereas a norm for titles ‘and names permits power to go uncoded in most discourse, Is ‘there any reason why the pronominal coding should be more ‘congenial to a static society than to an open society? ‘We have noticed that mode of address intrudes into conscious- ness. as.a problem at times of status change. Award ofthe doctoral degree, for instance, transforms a student into a colleague and, ‘among American academics, the familiar first name is normal ‘The fledgling academic may find it difficult to call his former teachers by their first names. Although these teachers may be ‘young and affable, they have had a very real power over him for ‘several years and it will feel presumptuous to deny this all at ‘once with a new mode of address. However, the ‘tyranny of democratic manners’ (Cronin, 1958) does not allow him to continue comfortably with the polite ‘Professor X’. He would not like to be thought unduly conscious of status, unprepared for faculty rank, a born lickspittle. Happily, English allows him a respite, He can avoid any term of address, staying with the un- committed ‘you’, until he and his addressees have got used to the new state of things. This lineuistc rite de passage has, for English speakers, a waiting room in which to screw up courage. In a fluid society crises of address will occur more frequently, 270 Language and Social Structures than in a static society, and so the pronominal coding of power differences is more likely to be felt as onerous. Coding by title and name would be more tolerable because less compulsory. ‘Where status is fixed by birth and does not change each man has enduring rights and obligations of address. A strong equalitarian ideology of the sort dominant in America works to suppress every conventional expression of power asymmetry. If the worker becomes conscious of his unrecipro- cated polite address to the boss, he may feel that his human dig- nity requires him to change. However, we do not feel the full ower of the ideology until we are in a situation that gives us some claim to recsive deferential address. The Americen profes- sor often feels foolish being given his ttle, he almost certainly ‘will not claim it as a prerogative; he may take pride in being on a first-name basis with his students. Very ‘palsy’ parents may invite their children to call them by first name, The very President of the Republic invites us all to call him ‘Tke*, Nevertheless, the differences of power are real and are experienced. Cronin’ has suggested in an amusing piece (Cronin, 1958) that subordination {is expressed by Americans in a subtle, and generally unwitting, body language. ‘The repertoire includes the boyish grin, the deprecatory cough, the unfinished sentence, the appreciative siggle, the drooping shoulders, the head-scratch and the bottom- waggle.” Group style with the pronouns of address ‘The identification of style i relative to the identification of some constancy. When we have marked out the essentials of some action - it might be walking or speaking a language or driving a ‘car - we can identify the residual variation as stylistic. Different styles are different ways of ‘doing the same thing’, and so their ‘identification waits on some designation of the range of perform- ances to be regarded as ‘the same thing’. Linguistic science finds enough that is constant in English and French and Latin to put all these and many more into one family = the Indo-European. It is possible with reference to this con- stancy to think of Halian and Spanish and English and the others 8 so many styles of Indo-European. They all have, for instance, ‘wo singular pronouns of address, but each language has an R. Brown and A. Gilman 274individual phonetic and semantic style in pronoun usage. We are ignoring phonetic style (through the use of the generic T and ¥}, ‘but in the second section of the paper we have described differ. ‘ences in the semantic styles of French, German and Jtalian. ‘Linguistic: styles are potentially expressive when there is covariation between characteristics of angoage performance and characteristics of the performers. When styles are ‘interpreted’, language behavior is functionally expressive, On that abstract level where the constancy is Indo-European and the styles are French, German, English ana Kalian, interpretations of style ‘must be statements about communities of speakers, statements of ational character, social structure, or group ideology. In the last section we have hazarded a few propositions on this level. ‘tis usual, ia discussion of linguistic style, to set constancy at the level of a language like French or English rather than at the level ofa language family. In the languages we have studied there are variations in pronoun style that are associated with the social status of the speaker. We have seen that the use of V because of its entry at the top of a society und its diffusion downward was always interpreted as a mark of good breeding. It is interesting to find an organization of French journeymen in the generation after the Revolution adopting a set of rules of propriety caution- ing members against going without tie or shoes at home on Sun- day and also against the use of the rmutual T among themselves (Perdiguies, 1914). Our informants assure us that V and T still function as indications of class membership. The Yugoslavians have a saying that a peasant would say Tto a king. By contrast, a French nobleman who tured wp in our net told us that he had said T to n0 one in the World except the old woman who was his ‘nurse in childhood, He is prevented by the dominant democratic ideology from saying T to subordinates and by his owa royalist ideology from saying it to equals. In literature, pronoun style has often been wsed to expose the pretensions of social climbers and the would-be elegant. Persons aping the manners of the class above them usually do not get the imitation exactly right. They are likely to notice some point of difference between their own class and the next higher and then extend the difference too widely, as in the use of the “elegant” broad [a] in ‘can* and ‘bad’, Moliére gives us his ‘présieuses 272 Language and Social Structures Tidicules’ saying V to servants whom a refined person would call In Ben Jonson's Everyman in his Humour and Epicoene such ‘true gallants as Wellbred and Knowell usually say ‘you’ to one another but they make frequent expressive shifts between this form and ‘thou’, whereas such fops as Jobn Daw and Amorous- La-Foole make unvarying use of "you". ‘Our sample of visiting French students was roughly bomogene- ‘ous in sotial status as judged by the single criterion of paternal ‘occupation. Therefore, we could not make any systematic study of differences in class style, but we thought it possible that, even within this select group, there might be interpretable differences of style. It was our guess that the tendency 10 make wide or narrow use of the solidary T would be related to general radicalism or conservatism of ideology. As a measure of this latter dimen sion we used Eysenck's Social Attitude Inventory (1957). ‘This is a collection of statements to be accepted or rejected concerning a variety of matters ~ religion, economics, racial relations, sexual behavior, etc. Eysenck has validated the scale in England and in France on Socialist, Communist, Fascist, Conservative and Liberal party members. In general, to be radical ‘on this scale is to favor change and to be conservative is to wish to maintain the status quo or turn back to some earlier con dition. We undertook to relate scores on this inventory toan index cof pronoun style. AAs yet we ave reported no evidence demonstrating that there exists such a thing as a personal style in pronoun usage in the sense of a tendency to make wide or narrow use of 7. It may be that each item in the questionnaire, each sort of person addressed, is an independent personal norm not predictable from any other. A child learns what to say to each kind of person. What he Jearns in each case depends on the groups in which he has ‘membership. Perhaps his usage is a bundle of unrelated habits. Guttman (Stouffer, Guttman, et al., 1950) has developed the technique of Scalogram Analysis for determining whether or not a collection of statements taps a common dimession. A perfect Guttman scale can be made of the statements: (a) I am at least 5° tall; (6) I am at least 5’ 4” tall; (©) I am at least 5’ 7" tall; (@ Lam at least 6” 1" tall; (©) Iam at least 6° 2” tall, Endorsement of @ more extreme statement will always be associated with en- R. Brown and A. Gilman 273dorsement of alllessextreme statements. A person can be assigned ‘single score ~a, bo, d, ore ~ which represents the most extreme statement he has endorsed, and from this single score all his individual answers can be reproduced. If he scores c he has also endorsed a and b but not d or e. The general criterion for scal- ability isthe reproducibility of individual responses from a single score, and this depends on the items being interrelated so that endorsement of one is reliably associated with endorsement or rejection of the others. "The Guttman method was developed during World War Il for the measurement of social attitudes, and it has been widely used Perfect reproducibility is not likely to be found for all the state- ments which an investigator guesses to be concerned with some single attitude, The usual thing is to accept a set of statements as scalable when they are 90 per cent reproducible and also satisfy certain other requirements; for example, there must be some statements that are not given a very one-sided response but are accepted and rejected with nearly equal frequency. ‘The responses to the pronoun questionnaire are not varying degrees of agreement (as in an attitude questionnaire) but are rather varying probabilities of saying T or V. There seems to be ro reason why these bipolar responses cannot be treated like yes (oF no responses on an attitude scale. The difference is that the scale, if there is ane, will be the semantic dimension governing the pronouns, and the scale score of each respondent will rep- resent his personal semantic style Tis customary to have 100 subjects for a Scalogram Analysis, bbut we could find only 50 French students. We tested all 28 items for scalability and found that a subset of them made a fairly ‘good scale. It was necessary to combine response categories so a8 to dichotomize them in order to obtain an average reproduc lity of 85 per cent. This coefficient was computed for the five termediate items having the more-balanoed marginal frequen: cies. A large number of items fell at or very near the {9 ex tremes. The solidarity or T-most end of the scale could be defines by father, mother, elder brother, young boys, wife or lover quite as well as by younger brother. The remote or V-most end could be defined by ‘waiter’ or ‘top boss’ as well as by ‘army general’. The intervening positions, from the T-end to the Wend are: 274 Language and Social Structures the elderly female servant known since childhood, grandfather, a male fellow student, a female fellow student and an elder brother's wife. For each item on the scale a T'answer scores one point and a V answer no points, The individual total scores range from 1 to 7, which means the scale can differentiate only seven semantic styles. We divided the subjects into the resultant seven stylistically homogeneous groups and, for each group, determined the aver~ ‘age scores on radicalism-conservatism. There was a set of almost perfectly consistent differences. ‘In Table 1 appear the mean radicalism scores for each pro- noun style, The individual radicalism scores range between 2 and. 13; the higher the score the more radical the person’s ideology. ‘The very striking result is that the group radicalism scores dupli- cate the ofder of the group pronoun scores with only @ single reversal. The rank-difference correlation between the two sets of scores is .96, and even with only seven paited scores this is a very significant relationship. ‘There is enough consistency of address to justify speaking of a personal-pronoun style which involves @ more or less wide use of the solidary 7, Even among students of the same socioeconomic level there are differences of style, and these are potentially ‘expressive of radicalism and conservatism in ideology. A French- ‘man could, with some confidence, infer that a male university student who regularly said T to female fellow students would favor the nationalization of industry, free love, tril marriage, the abolition of capital punishment, and the weakening of national- istic and religious loyalties, ‘What shall we make of the association between a wide use of T and a cluster of radical sentiments? There may be no ‘sense’ toit at al, that i, no logical connection between the linguistic practice and the attitudes, but simply a general tendency to go along with the newest thing. We know that lef-wing attitudes are more likely to be found in the laboring class than in the professional class~ cs. Perhaps those offspring of the professional class who sym- pathize with proletariat politics also, incidentally, pick up the working man’s wide use of T without feeling that there is any- thing in the linguistic practice that is congruent with the ideology. (On the other hand perhaps there is something appropriate in R, Brown and A. Gilman 275the association, The ideology is consistent in its disapproval of Darviers between people: race, religion, nationality, property, ‘marriage, even criminality. Ait these barriers have the effect of separating the solidary, the in-group’, from the nonsolidary, the ‘out-group’. The radical says the criminal is not far enough “out” to be killed; he should be re-educated. He says that a nationality ‘ought not to be so solidary that it prevents world organization ‘rom succeeding. Private property ought to be abolished, industry should be nationalized. There are to be no more out-groups and in-groups but rather one group, undifferentiated by nationality, religion, or pronoun of address. The fact that the pronoun which is being extended to all men alike is T, the mark of soli darity, the pronoun of the nuclear family, expresses the radical’s intention to extend his sense of brotherhood. But we notice that Table 1 Scotes on the Pronoun Scale in Relation to Scores ‘on the Radicalism Scale Group mean Group pronoun score radicalism score 530 666 6a 7.83 683 383 975 the universal application of the pronoun eliminates the discrim- ination that gave it a meaning and that gives particular point 1 an old problem. Can the solidarity of the family be extended 30 widely? Is there enough libido to stretch so far? Will there per- hhaps be a thin solidarity the same everywhere but nowhere 50 strong as in the past’? ‘The pronouns of address as expressions of transient attitudes Behavior norms are practices consistent within a group. So long the choice of a pronoun is recognized as normal for a group, its interpretation is simply the membership of the speaker in that 276 Language and Social Structures group. However, the implications of group membership are often very important; social class, for instance, suggests @ Kind of family life, a level of education, a set of political views and much besides. These facts about a person belong to his character. They are enduring features which help to determine actions over many years. Consistent personal style in the use of the pronouns of address does not reveal enough (o establish the speaker's unique character, but it can help to place him in one or another Tare category. Sometimes the choice of @ pronoun clearly violates a group ‘norm and perhaps also the customary practice of the speaker. ‘Then the meaning of the act will be sought in some attitude or emotion of the speaker. itis as ifthe interpreter reasoned that variations of address between the same two persone must be caused by variations in their attitudes toward one another. If two men of seventeenth century France properly exchange the V of upper-class equals and one of them gives the other 7, he ‘suggests that the other is his inferior since it is t0 his inferiors that a man says T. The general meaning of an unexpected pro- noun choice is simply that the speaker, for the moment, views his relationship as one that calls for the pronoun used. This Kind of variation in language behavior expresses a contemporaneous feeling or attitude. These variations are not consistent personal styles but departures from one’s own custom and the customs of 4 group in sesponse to a mood. ‘As there have been two great semantic dimensions governing, and V, 30 there have also been two principal kinds of expressive ‘meaning. Breaking the norms of power generally has the meaning that a speaker regards an addressee as.his inferior, superior, of equal, although by usual criteria, and according to the speaker's ‘own enstamary usage, the addressee is not what the pronoun implies. Breaking the norms of selidarity generally means that the speaker temporarily thinks of the other as an outsider or as an intimates it means that sympathy is extended or withdravsn, ‘The oldest uses of and V to express attitudes seem everywhere to have been the Tof contempt or anger and the V of admiration ‘or respect. In his study of the French pronouns Schliebitz (1886) found the first exannples of these expressive uses in literature of the twelfth and thirteenth centuries, which is about the time that R. Brown and A. Gilman 277the power semantic crystallized in France, and Grand (1930) has found the same thing for Italian, In saying 7, where V is usual, the speaker treats the addressee like a servant or a child, and assumes the right to berate him. The most common use of the expressive V, in the early materials, is that of the master who is exceptionally pleased with the work of a servant and elevates, him pronominally to match this esteem, Racino, in his dramas, used the pronouns with perfect semantic ‘consistency. His major figures exchange the V of upper-class ‘equals. Lovers, brother and sister, husband and wife - none of them says Tif he is of high rank, but each person of high rank hhas a subordinate confidante to whom he says T and from whom hhe receives V. It is a perfect nonreciprocal power semantic. This courtly pattern is broken only for the greatest scenes in each play. ‘Racine reserved the expressive pronoun as some composers save the cymbals. In both Andramaque and Phédre there are only two expressive departures from the norm, and they mark climaxes of feeling. espersen (1905) believed that English ‘thou’ and ‘ye” (or ‘you’) were more often shifted to express mood and tone than were the pronouns of the continental languages, and our com- parisons strongly support this opinion. The ‘thou’ of contempt ‘was so very familiar that a verbal form was created to name this expressive use. Shakespeare gives it to Sir Toby Belch (Twelfth Night) in the lines urging Andrew Aguecheek to send a challenge ‘thou thou’st him some thrice, it shall not be amiss.” In life the verb tumed up in Sir Edward Coke's attack on Raleigh at the latter's trial in 1603 Jardine, 1832-5): ‘All that he did, was at thy instigation, thou viper; for I thou thee, thou traitor.” ‘The T of contempt and anger is usually introduced between persons who normally exchange V but it can, of course, also be used by a subordinate to a superior. As the social distance is greater, the overthrow of the norm is more shocking and gener- ally represents a greater extremity of passion. Sejanus, in Ben Jonson's play of that name, feels extreme contempt for the emperor Tiberius but wisely gives him the reverential V to his face. However, soliloquizing after the emperor has exited, Sejanus begins: ‘Dull, heavy Caesar! Wouldst thou tell me..." In 278 Language and Social Structures Jonson's Volpone Mosca invariably says ‘you" to his master ‘until the final scene when, as the two villains are about to be carted away, Mosca turns on Volpone with ‘Bane to thy wolfish nature.” Expressive effects of mach greater subilety than those we have described are common in Elizabethan and Jacobean drama. The ‘exact interpretation ofthe speaker's attitude depends not only on, the pronoun norm he upsets but also on his attendant words and actions and the total setting. Still simple enough to be unequivocal is the ironic or mocking ‘you said by Tamburlaine to the captive Turkish emperor Bajazeth. This exchange occurs in Act IV of Marlowe's play: ‘amon: Here Tuk, wit thou have a dean tener? ‘Pastonvane! Soft Ss you mat be ite; too much eating will make you surfeit. “Thou” isto be expected from captor to captive and the norm is upset when Tamiburlaine says “you", He cannot intend to express admiration or respect since he keeps the Turk captive and starves him, His intention is to mock the captive king with respectful addres, implying a power thatthe king has lost. ‘The momentary shift of pronoun directly expresses a momen tary shift of mood, but that interpretation does not exhaust its ‘meaning. The fact that man has a particular momentary attitude oF emotion may imply a reat deal about his characteristic cise position, his eadiness for one kind of feeling rather than another, Not every attorney general, for instance, would have used the abusive “thou’ to Raleigh, The fact that Edward Coke did so suggests an arrogant and cholerie temperament and, in fat, many made this assessment of him Qardine, 1832-5). When Volpone spoke to Celia, a lady of Venice, he oweht to have said “you” but te began at once with thee . This violation of decorum, togethe: with the fact that he leaps from his sick bed to attempt rape of the lady, helps to establish Volpone’s monstrous character. His abnormal form of address is consistent with the unnatural im ages in his speech. In any given situation we know the sort of People who would break the norms of address and the sort who Would not, From the fact that a man does break the norms we R. Brown and A, Gilman 279infer his immediate feelings and, in addition, attribute to him the general character of people who would have such feelings ‘and would give them that kind of expression. With-the establishment of the solidarity semantic a new set of expressive meanings became possible - feelings of sympathy and estrangement. In Shakespeare's plays there are expressive mean ings that derive from the solidarity semantic as well as many ‘dependent on power usage and many that rely on both conno- tations. The play Two Gentlemen of Verona is concerned with the Renaissance ideal of friendship and provides expecially clear expressions of solidarity. Proteus and Valentine, the two Gentle men, initially exchange ‘thou’, but wher they ‘ouch on the sub- ject of love, an which they disagree, their address changes to the * you" of estrangement. Moliére (Fay, 1920) has shown ws that a ‘man may even put himself ata distance as does George Dandin in the soliloquy beginning: “George Dandin! George Dantint Vous aver fait une soitise ...” In both French and English drama of the past, T and V were marvelously sensitive to feelings of approach and withdrawal. In terms of Freud’s striking amocba metaphor the pronouns signal the extension of retraction of libidinal pseudopodia, However, in Freneh, German and Italian today this use seems to be very ‘uncommon. Our informants told us that the 7; once extended, is almost never taken back for the reason that it would mean the ‘complete withdrawal of esteem. The only modern expressive shift we have found is a rather chilling one. Silverberg (1940) reposts that in Germany in 1940 a prostitute and her client said du when they met and while they were together but when the libidinal te (in the narrow sense) had been dissolved they resumed the mutual distant Sie ‘We have suggested that the modem direction of change in ‘pronoun usage expresses a will to extend the solidary ethic to ‘everyone. The apparent decline of expressive shifts between Tand V is more difficult to interpret. Perhaps i is because Europeans hhave seen that excluled persons or races or groups can become the target of extreme aggtession from groups that are benevolent within themselves. Perhaps Europeans would like t0 convince themselves that the solidary ethic once extended will not be ‘withdrawn, that there is security in the mutual 7: 280 Language and Social Structures References Bavaw, A.C. (935), A History of the English Language, New York. BRUNOT,F. (1927), La penste ela lange, Pars, BRUNOT,. (1937, Histoire de alone frangaise, Pais, BYawe, $1572% $1. (1936, "Shakespeare's ee ofthe progoun of| addres’, diseration, Catholic Univesity of America, Washington. ‘Cur rexatn, B.1880),"Du plural de respecten Latin", Revue de philolgie, vo. 4, pp. 129-39. (Cont, M. (1958), "The tyranny of democratic manners! New ‘Republic vol. 137, pp. 12-14. Duventux, W.B.(1853),Licrand Letters ofthe Devereux, Brlsof use, ithe Reign of Elizabeth, James Tand Charles 1, 1540-1645, Londen, uz, F.- (1870, Grammaire des langues romanes, Pati Euuwoon, (1714), The History of the Life of Thomas Elwood, ‘London. Estaica, R.M.,and SpunveR, (1946) Thro Hers to Language, ‘New York Brsencx, H.J.(1957), Sense and Nonsense in Pychology, Penguin. FARNSWORTH, R. (1655), The Pure Longuage ofthe piri af Truth fr" Thee' and Thou nts Place...» Londo Fay, P-B.(1920),'Theusa of "tu and vous” ia Mote’, Uiversty “f California Publications in Modern Pilology vol. 8,09. 227 86. Fox, G. (160), A Bale Door for Tecchon ad Professors io Learn Pliraland Singer, Condon. Gerke, (1799), Ober Du snd Si in der deutschen Sprace, Bet. GRAND, C.(1930)," 7 90h lt rude des pronomsallocuoires Haters, Tage. Gains, J (1898), Dewrsche Grammati, vo. 4, Gite, HARRISON, G-B.(cd.)(1935), The Letters of Queen Eizbett, London. Sanne, D. (1832-5), Criminal ial, ol, 1-2, London. Susennsey, 0. (909), Growth and Siuctre ofthe English language, Leipag. JoRNETON, O.M. (1904), “The we of"ela,"“ei" and “Ia a polite forms of adétes in Italian’ Modern Philotagy, ol 1, pp. 459-15. KENNEDY, A.G.(I915),The Pronoun af Adare English Literatur of the Tureenth Centar, Stanford University Press. Livre, B (1882), Dictionnaire dia laraue francaise O18, Psi MavERLOaxe, W. 190), Grammaire des languerromanes, V6.3, Paris MOLLER FEF. (1914), The wses ofthe plural of reverencein the eters of Pope Gregory”, Romanic review, vol. 3, pp- 6889. Pexpiouren, A. (1910), Memoiresd'unconpapnon, Moni. Scuitpres, V- (1886), Di Person der Anredeinder/ranrSizhon ‘Spracke, Breslau, Sezvenseno, W.V. (1940), On the psychological siraiieance of “Du” and"Si, Prpchoanalytie Quarterly, vol 9, pp. 508-25, R. Brown and A. Gilman 281StrpsroN, R,0. (1917), The Use of Yeinthe Function of Thou: A Study (of Grammar and Social Intercourse in Fourteenth: Century England, Stanford University res. StoUrrER, §.A., GUTTMAN, Li ef a. (1950), Measurement and ‘Prediction, Priceton University Pres 282 Language and Social Struct 13 W. Labov ‘The Study of Language in its Social Context ‘Bicep from W. Labo, "The study of ‘Studion Generale vol 23,1970, pp 65-65 guage inits social context’, ‘Sociolinguistic structure We may define a sociolinguistic variable as one which is correlated with some nowinguistic variable of the social context: of the speaker, the addressee, the audience, the setting, ete. Some lin- Biistic features (which we will call indicators) show a regular distribution over socio-economic, ethnic, or age groups, but are used by each individual in more or less the same Way in any con- text, Ifthe social contexts concerned can be ordered in some kind of hierarchy (like socio-economic or age groups), these indicators cean be said to be stratified. More highly developed sociolinguistic variables (which we will call markers) not only show social distribution over socio-economic, ethnic, or age groups, but are can be ordered along a single dimension according 10 the amount ‘of attention paid to speech, so that we have stylistic as well as social stratification. Early studies such as those of Fischer (1958) or Kuéera (1961) observed linguistic variables only one dimension at atime, but more recent studies (Labov, 1966; Wolfram, 1969; ‘Anshen, 1969) look at the interrelation of both dimensions. A stable sociolinguistic marker: (th) One of the most general and simple sociolinguistic markers in English is (th): the phonetic form of the voiceless interdental {ricative /6/ in thing, thick, etc. The prestige form is universally ‘he fticative, while affricates and stops are stigmatized. The i ‘uence of other languages without this interdentalfricative may seinforve the development of the stop form in various large cities ‘of the United States, in Anglo-Irish and in NNE (Non-standard = Negro English); but we also find this sociolinguistic variable in a great many other rural and urban arcasin England and the United W, Labov 283,