The three issues we have paid attention to in this chapter are the timing, consistency and social directionality of the gender differences observed. As to the first, it appears that gender affiliation can only be detected when a change has passed its incipient phase and reached the 20-30 per cent level of overall frequency. This is what we shall call, following Labov (1994), the new and vigorous stage in the life cycle of a change in progress. It reveals the gender affiliation of you, my/thy, its, the object of the gerund, third-person singular -s, do in negative statements and the prepositional WH-phrases, all with female advantage. But in a couple of cases, gender differentiation may also be detected earlier. This is the case with two male-dominated processes, the loss of multiple negation and inversion after initial negators. More data are needed to confirm the trend but it is not unexpected at a time when supralocal changes led by men are typically channelled through learned and professional usage.
Another clear trend that emerges from the data is that the gender affiliation of a process remains constant from the new and vigorous stage on until its (near-)completion. This is the case with all except three of the changes we have examined. Crossing-over is only seen with the changes involving periphrastic do and, much less clearly, with relative adverbs. A simple crossing-over from male to female advantage may not, however, be what happened with do either. A sudden break is put on its diffusion in the first decades of the seventeenth century when it is curbed in both affirmative and negative statements. In negatives do is later revived and becomes generalized, but in affirmatives the brief revival in 1620-39 is followed by its slow demise. In both cases women take the lead after the disruption of what had been male-oriented processes. Some scholars have argued (e.g. Kroch 2001) that whatever major disruption there was in the diffusion of do must have been a reflection of a sudden restructuring in its syntactic properties. Following Nurmi (1999a) we have offered dialect contact as a likely cause for the changes in the sociolinguistic profiles of the two processes in the early seventeenth century.
The third trend that appears from the data is female advantage in language change regardless of the social embedding of the process. Some of the changes we have examined, such as the determiners my and thy and the third-person singular -s, proceed from the lower literate end of the social hierarchy rather than from the topmost ranks. The gentle- and noblewomen leading these processes in our data must have adopted them early and diffused them further within their own ranks. In certain cases their usage was even more advanced than that of middle- or lower-rank men. This could also be observed with the object of the gerund, which did not show any distinct pattern of social embedding. Interpreting these findings following Milroy and Milroy (1993), who regard gender as a socially more decisive variable than social class, we could argue that gender difference came before social status in Tudor and Stuart England, too. But before we can really make this claim, we shall have to account for the possible effects of register variation. This is a topic that we shall return to in Chapter 9.
In cases like the spread of subject you and prop-word one, with no great difference between the upper and middle ranks, female advantage is even more pronounced. These processes appear to have progressed from below in terms of social awareness: there is no difference between the City of London and the Court in the rise of you, for instance (see 8.4.1.).
Where the opposite is the case, however, and evidence from other genres suggests that a conscious process was under way, the change was consistently led by men. A case in point is the disappearance of multiple negation, which was promoted by male professionals and systematically led by men in the upper and middle sections of society. Here emerges the chief difference between Tudor and Stuart England and the present day: late medieval and early modern Englishwomen did not promote language changes that emanated from the world of learning and professional use, which lay outside their own spheres of 'being'. However, with this major exception confirming the rule, Labov's 'long-standing cultural pattern' was already clearly in evidence.