In Language: An Introduction to the Study of Speech Edward Sapir notes that "no sooner are the old dialects ironed out... or ousted by the spread and influence of the one dialect which is culturally predominant, when a new crop of dialects arise." Within the context of the debate surrounding standard English the actual characteristics of language change tend to be repeatedly elided, while language is regarded as being an agency in its own right. Instead language becomes an arena for socio-political concerns, of which it is merely the symptom rather than the cause.

The most haunting concern within the standard English debate has been the tendency to metamorphose language into a hypothetical object. For instance in studies by Labov, Blum, and Gumperz when subjects were questioned about their usage of standard forms, insisted that their usage was perfectly consistent with received pronunciation, even though this was not to be the case in their actual speech. Similarly upper class Arabic men stressed that they only used high class forms, but in recorded conversations were found to regularly use lower class forms. This cannot be simply reduced to a division between competence and performance, but to indicate that automatic code switching is a natural feature of linguistic communication, within which language use is socially directed. Use of vernacular forms tends to be regarded as a mark of group identity. Labov found copula deletion to be a common feature of Black English vernacular i.e. "he a bad man," a form that is not found in isolated individuals. Further evidence of this can be found in the use of phonetic titles in records by black artists, i.e. "in tha middle," to assert a common black identity. A specific peer group acts as a norm enforcement mechanism, but outside of such environments language use becomes much more diffuse.

Naturally the same argument proceeds in exactly the same manner when reversed. In modern Russia a conservative backlash has developed against the use of English loan words for economic and computing terms (for obvious reasons there are no native Russian equivalents) simply being imported, declined and conjugated as Russian words. In exactly the same manner France has had a powerful debate about the introduction of English words. Accordingly language serves as a social marker, but not without variation. In particular language change appears to have a strong component of social motivation, with subconscious changes appearing to move away from prestige forms while conscious changes move towards them. To take one example of this, Labov found that New York Salesgirls tended to move mimic their customers, and attached a negative prestige to their own accent, leading to the restoration of the post vocalic 'r'. Similarly the Belfast pronunciation of 'a' is moving down the social scale to be unstressed. For linguistic change to occur, speakers belonging to comparatively loose knit social networks tend to be exposed to a wider range of influences and are able to spread such influences into other wise closed environments further, whether this refers to a vernacular or a prestige norm to be adopted. In short, linguistic change tends to be closely related to social status, serving a wide range of functions, by which alterations to a language closely mirror social attitudes, as with the restoration of the post-vocalic r reflected a desire to move up the social scale, while the unstressing of the 'a' to move away from a pretige norm towards a solidarity position. Nonetheless the ability of the individual to code switch indicates that language is a social function, to be used according to the social position of the individual.

In emphasising that language is a self-regulating social function that in a society that has sought to purge dialects of extreme habits of vocabulary and pronunciation, standard serves as a social norm available for usage, rather than as a fixed state. It should be noted that there are difficulties in research, however a study conducted in Reading found that non-standard forms accounted for 55% of casual speech, but only 25% of formal speech. Written standard forms accommodate this flexibility by frequently being no closer to standard than dialect forms, with the pronunciation of certain words being subject modification; hence the fact that Australia and South Africa retain essentially the same written language as England. However it is exactly this flexibility that is the source of complaint. The Newbolt report into The Teaching of English in England was in many ways more explicitly concerned with post-war national unification than the role of language in society; "English children... are surely entitled to be taught... the accepted speech of their own country," thus concluding by contrasting England and other nations. The social context had been made clear in a remark made by Lloyd George in 1918; "an educated man is a better worker... a better citizen. That was only half comprehended before the war." Newbolt himself held the presidency of the English Association, what we now term an educational pressure group, formed to "promote the due recognition as an essential element in the national education." The committee formed by the Government to consider this role was dominated by members of the Association. The social prejudices of their arguments are often painfully obvious; "evil habits of speech contracted in home and street," and "a definite dialect whose speech is disfigured by vulgarisms." Such views are made more explicit by George Sampson's English For the English where was is declared against "the enemy's growing tyranny," a fear largely fostered by the threat posed by the Soviet Union, and the possibility of a communist revolution in Britain.

Although it is hardly difficult to dismiss such views, the arguments of the report itself are rather more sophisticated, and are worthy of serious consideration. For example, the statement that "we do not advocate the teaching of standard English on any grounds of social superiority." However if this was truly the case then why does the report not address the issue of one single speech form should be valued over others in the name of producing "civilised and articulate human beings", if one of them has not been designated with a superior social value. The report also asserts that "we do not ... suggest that the suppression of dialect should be aimed at, but that children... should... become bi-lingual." The problem with this is that the report fluctuates between such arguments, and the view that "if a child is not learning good English he is learning bad English," a statement that leaves no room for the legitimacy of dialect. No-one would argue that the ability of members of a society to communicate at ease is inconsequential, and that a condition of diglossia may well be a means towards this end (by recognising the legitimacy of non-standard forms). What is arguable is that the report implies this may only be attained through state intervention, in spite of the fact that diglossic code switching is already inherent to speech. It is therefore decidedly difficult to decide where and if the report draws the proverbial line between contesting the view that "there are still people... who are inclined to regard a humane education of the lower classes as subversive of public order," and insinuating the view that resurfaced in Quiller-Couch's What Happens in Hamlet which feared the "communistic or corporate era that seems to lie immediately before us."

For a more rigorous examination of these issues we might turn to John Marenbon's English our English: The New Orthodoxy Examined. Marenbon concedes that there is no intrinsic connection between communicative adequacy and the particular forms of a language, as well as conceding that all languages, including dialects such as Cockney and West Indian Creole are equally regular and rule governed. However he refuses to concede that this leads onwards to the conclusion that every language is wholly adequate to the needs of its speakers, on the grounds that "languages develop along with their user's manner of living," so that "Chaucer's English was still inadequate." Accordingly the long term evolution and attained stability of standard makes it superior to dialect, which he states is always in a state of flux. His argument here appears to contradict itself; "it is never clear which of the constructions a dialect speaker is using are grammatical." This effectively cancels out his view that all languages are equally rule governed and grammatical. The argument that "it is never clear exactly which dialect he is speaking, or how consistently he is intending to speak it," has equal application to standard, which is also never used purely. Furthermore, Marenbon also implies that linguistic usage is dictated by a process of conscious selection, rather than largely automatic adaptation. By contrast, there is little evidence to support the evolutionary view of language, or the difference between standard and dialect forms that he infers from it; especially as many of those dialects possess equal ancestry to standard. Furthermore, if Chaucer really is so primitive what reason would he be prepared to recommend that the The Canterbury Tales should be disregarded as being primitive by the standards of our apparently more sophisticated language ?

Marenbon's opinion on the standard English question can then be seen to have a firmly social basis, advocating the teaching of literature, based on our "literary heritage," and concluding with a resounding flourish that "in the future of its language there lies the future of a nation!" However, such constructions are very far from being a solely modern construction, as may be evidenced by Thomas Sprat's History of the Royal Society of London, written in 1667, which states that "purity of speech and greatness of Empire have in all countries still met together." In Locke's Essay Concerning Human Understanding he speaks of limiting the language of the market, "anyone can pretend to reforming the languages of the World... without rendering himself ridiculous." This is of course also the view largely taken by Johnson and Swift. Accordingly Johnson seeks to "preserve the purity and ascertain the meaning of our English idiom." If English did have to be shared between individuals it had to be purified, with the speech of the lower class being identified as the cause of the contamination, so that in his Plan of a Dictionary of the English Language Johnson wishes to "fix the pronunciation of monosyllables." The social basis of his arguments is readily apparent; "that none but a Northern ear could endure." We find to be even more the case in Swift's Proposal for Correcting, Improving, and Ascertaining the English language; where he opines that "it is better that a language should not be wholly perfect than that it should be perpetually changing."

Advocacy of the imposition of a standard language has then historically tended to be inseparable from social status. The problem tends resides with the fact that such arguments tend to work against the nature of language; it is not possible to fix an object which is by definition mutable, precisely because of overlapping social and regional differences. The arguments advocated by Newbolt strongly reflected a fear of the rebellious masses, while Swift's arguments through the positing a period wherein "the English tongue received most improvement," connect his linguistic arguments to a Tory social argument; a nostalgia for the prelapsarian English society, that supposedly predated the revolution.

Discussions of this topic have often tended run into the problem that attempts have simultaneously been made to identify a discrete women's language, and viewing language as an entity that has largely been determined by men, and therefore inherently predisposed towards discrimination against women. One obvious critique of the latter is that it diverts attention away from specifically sexist areas of language. It may well be more effective to take the view that both of these suggestions are intrinsically unhelpful, and that they should be replaced with a differing view of language as a set of interactions between competing discourses.

In his essay The Order of Discourse, Michel Foucault advocates the view that "in every society the productions of discourse are controlled, organised, redistributed, by a certain number of procedures." An obvious example of this is that established discursive regularities remain a significant obstacle; increasing the number of female journalists has not prevented headlines like "three survivors, two of them women," or "people more likely to be influenced by their wives then by the polls." Equally so, as the former Secretary of State for Social Security, Harriet Harmon noted, one particular idiom commonly employed in politics frequently resorts to the usage of an exceptionally masculine and militaristic set of metaphors, whereby elections are "fought" as "campaigns," in spite of the massively increased number of female members of parliament. Nonetheless language is not wholly identifiable with discursive practices, a view also pointed to by Foucault, "what civilisation has ever appeared to be so respectful of discourse as ours?" By concentrating upon what Foucault terms the caesurae that "disperse the subject into a plurality of possible positions and functions," we are able to recognise that there is no viable distinction between language and speakers. Language is based upon individual dictate not consensus choice, thereby frustrating 'authentic' notions of language, but clearly we have to recognise the ability of discourse to influence if not determine usage.

The first and most obvious instance of discursive practice lies within sexist language. Women tend to be defined not as general representatives of humanity, but as categories; whore, slag, mother, virgin, housewife, and so on. There is a veritable excess of sexual definitions, but very few positive designations for women. Dolly began as a pet name, but then began to refer to a prostitute. Similarly broad originally referred to a young woman, while whore was initially a non gender specific term for a lover. The issue of titles is also a prominent one, as terms like mankind have been defended as not being particular to any one gender, and that they can therefore be used in the same way as titles like American or Nurse (as opposed to the German equivalent, where the term for nurse "schwester" is specific to one gender). It should be noted that in the case of the term 'woman,' woman was "taken out of man" (Gen. 2:23), and therefore the man has the pre-eminence. "The head of the woman is the man;" but yet honour is to be shown to the wife, "as unto the weaker vessel." Concepts like this have often been present in the use of the word 'woman' although current usage shows little sign of that. Nonetheless, it is surely not difficult to see why women might be annoyed by the terms like 'woman' and 'mankind;' they wear their etymology on their proverbial sleeves.

However, if one does accept that such terms can be defended as no longer posessing a meaning peculiar to one gender, problems nonetheless emerge. In the sentence "Jon wanted to find a woman," the term lady could not be used as a viable substitute. Equally so, the sentence "Jon refused to be my master, and returned to his wife" quickly demonstrates that the terms master and mistress are not interchangeable. The same is clearly true for both bachelor and spinster, the former referring to a impecunious male, the latter to a female prostitute. If one chooses to use a more extreme parallel, the terms prostitute and gigolo cannot be regarded as being congruent, as gigolo has too many associations of prowess. Put bluntly, there will never be a film entitled American Slag. There is also a further problem, namely that titles like charperson, tend to be applied rather more to women than men.

This is not to indicate that sexist views of languages are not capable of reformation. The determinist argument, can be illustrated by the view that men are perceived to have built positive associations into motherhood, thereby making it difficult for women to acknowledge negative associations pertaining to it. The argument suffers as it leaves it unclear whether are without linguistic resources because they are without power or because they need linguistic resources to increase their power. However the mere fact that many of the titles we have discussed originally possessed unsexualised meanings, indicates that the process can be reversed through the same process. For example the term 'glamour' originally possessed connotations of female witchery, which has altered to something rather more benign, albeit something still largely applied to women.

The issue of definable characteristics in women's speech takes a similar course. It is patently the case that English possesses sex preferential if not sex exclusive terminology, unlike the carib Indians who employ gender phonologies and lexicons. Women tend to prefer standard english rather than dialect forms, and have a greater tendency to ask questions (in one survey three hundred and seventy questions were asked over a period of twelve and a half hours, of which two hundred and sixty three were posed by a female participant). In addition ot this women have been found to make far more use of minimal responses like 'mmh' and 'yeah' and more use of intensifiers like 'really' and 'very'.

Mixed sex conversations tend to be dominated by men, who make the majority of interruptions, whereas the number of interruptions made by each participant is markedly more symmetrical in same sex conversations. Men tend to employ a number of strategies for controlling a conversation, for example altering the topic, whereas the more typically female strategy of asking tag questions (questions inobtrusively inserted at the end of a sentence) may well be designed to circumvent this. Women only succeeded in introducing new topics into the conversation for 36% of the time in one survey, until tag questions where deployed, at which point the success rate increased to 72%. This raises the rather thorny possibility that linguistic differences may be attributable to women using co-operative strategies where men use combative tactics. This could prove to account for features like the lower occurrence of swearing in women's speech (although this is dependent on context), and the greater prevalence of compliment exchanges in women's speech. Janet Holmes has analysed a corpus of four hundred and eighty four compliment exchanges in New Zealand, thereby establishing that 51% were between women, while a mere 14% were between men. 25% were given by men to women. Conversely men were more likely to make blunt declarations than women.

Clearly the interpretation of these differences is a matter of considerable difficulty, particularly as these features are only associated with women's speech and are not necessarily intrinsic to it; not all women use hyper correct standard forms. The first point to be made at this stage is that there has been a tendency to view women's language as a set of deviations from a male norm. As such, statistical interpretation can be made an extremely subjective matter, a rough parallel to the idea that cockney glottal stops are nasal. Men's speech is seen as normal, and it is only when women's speech deviates from this that it becomes an issue (although arguably attempts to elevate a separate women's speech to the status of a ghetto are no less problematic). One way around this has been to stress social networks rather than stratification models, and therefore parole rather than langue. It may well be the case that social segregation produces many of these differences, as a large amount of female employment is in the service sector which requires usage of standard english to a far greater extent than traditional male occupations. Women therefore have to switch linguistic codes rather more sharply than would be necessary amongst the tightly knit networks which stress solidarity among male participants, and which therefore may create far more unity among their members. It is particularly well worth noting that there is no evidence of innate conservatism on the part of women; in Madagascar it is the men who demarcated as the guardians of revered speech traditions.

Such an approach certainly merits further extension, to consider that gender is not an entirely independent category; it is bisected by social divisions, since men still tend to hold higher status positions than women. Nonetheless when this is taken into account the results become rather more unpredictable. In one survey male subordinates received over three and a half times more assent than their female bosses, and were also more successful at being interrupted less often. On the other hand, another survey undertaken by Janet Holmes sought to cross-reference gender with social status. With this, facilitative tags were used by 75% of women and 45% of men of higher power status, but were not used by those of lower power status of any gender. The same was true of 'softeners' used by 10% of women and 7% of men of higher power status. Conversely modal tags, requesting information of which the speaker is uncertain were increased by 10% in the case of women of lower power status, and 11% for women of lower power status. Gender differences remain but have been strikingly altered, since it starts to emerge that the supposed characteristics of women's speech have emerged elsewhere, for instance in the language of parents and children.

The question of power then takes centre stage, and serves as an obvious explanation of why it is with no difference between the male and female vocal tract, differences of pitch are nonetheless associated with gender. Boys belong to hierarchically organised groups in which language is made to serve the maintenance of power. Girls are excluded from this, and tend to form close friendships instead, with a chiastic dichotomy of collaborative and competitive models becoming established. However women are clearly moving outside the co-operative stereotype, by adopting an assertive style in group interaction, and prosodic features more typical of men i.e. falls rather than rising intonation patterns, use of non-standard accents and swearing. Some very pertinent examples of this can be found in popular lyrics. Swearing has historically been common among male artists, but has increasingly been appropriated by women. To this, it may be objected that women are taking on masculine models and not vice versa. It may be replied that there is no need for women to permanently adopt anything, but rather to appropriate certain previously unavailable registers for certain very specific contexts. A co-operative model of speech is not only a stereotype in its own right, but is also blatantly unsuitable for some public contexts in exactly the same manner in which a competitive model is inappropriate for private conversation.

1. The term spinster having the following etymology: 'a woman of evil life and character; - so called from being forced to spin in a house of correction' and not merely a single elderly lady, which was the meaning that came to predominate later on. Bachelor also has certain notions of being impecunious or not of full status (hence bachelor of arts in academic parlance, as opposed to master of arts), but also shifted in meaning to concentrate on being unmarried.