- 1. As mentioned in Chapter 1, much of the earlier research on sociohistorical linguistics has been based on textual variation alone. The rich sources of information offered by social and cultural historians have not been taken advantage of. For instance, the bibliographies of Romaine (1982a) and Milroy (1992) do not contain any works on social history. However, we are hardly exaggerating when we claim that familiarity with relevant linguistic communities is one of the prerequisites for successful sociolinguistic analysis.
- 2. In her discussion of sociohistorical linguistics and the Observer's Paradox, Tieken-Boon van Ostade (2000c) offers a very different interpretation of the concept. She does not limit her attention to the role of the researcher observing people's linguistic behaviour but includes in the workings of the Observer's Paradox factors that prevent passages of the 'completely unmonitored' style - the term borrowed from Chambers (1995: 18) - being written down. The article contains interesting discussions, e.g. Fanny Burney's habit of taking notes about the language she heard around her, but it is difficult for us to see how, for instance, letter-writing conventions would be connected with the Observer's Paradox in the traditional sociolinguistic sense of the term. We would rather argue that every genre, including the spoken conversation which sociolinguists usually observe, has its own conventions, and therefore completely unmonitored speech is only a relative concept (see Chapter 9, below).
- 3. The information about informants' social backgrounds that is incorporated in modern multi-genre corpora such as the British National Corpus (BNC) makes it possible to study sociolinguistic variation in a broader variety of data than only spoken conversation.
- 4. The following variables have been studied in the variationist framework: mine and thine vs. my and thy, possessive determiner its vs. of it and thereof, object and subject of the gerund, third-person singular suffix -th vs. -s, periphrastic do in affirmative statements, inversion after initial adverbs and negators, relative markers, and indefinite pronouns with singular human reference. The references are given in Chapter 4.
- 5. The dimensions in Biber (1988: 121-210) are as follows: (1) Involved versus Informational Production, (2) Narrative versus Non-Narrative Concerns, (3) Explicit versus Situation-Dependent Reference, (4) Overt Expression of Persuasion, (5) Abstract versus Non-Abstract Information and (6) On-Line Informational Elaboration. They have been further refined in subsequent research.
- 6. To be accurate, a few private letters have been found from the first decade of the fifteenth century, viz. the letters of Lady Zouche (Payne & Barron, eds. 1997). Their late discovery by the project team placed them in the CEEC Supplement, and they are not used in this study. The odd-looking end year 1681 can be explained by the fact that we decided not to cut some interesting series of letters at 1680.
- 7. 1558 is the year of Elizabeth I's accession to the throne and is, of course, quite arbitrary from the linguistic point of view.
- 8. The literature gives the following numerical information about the Civil War. The New Model Army was 22,000 men strong in 1645. In four years of struggle, about 100,000 men were killed (Briggs 1985: 141). Approximately 150-200 country houses, 10,000 houses in cities and towns and 1,000 in villages were destroyed (Porter 1997: 66). For women's experiences of the period on both sides, see Hudson (ed. 1993).
- 9. Burke (1992a: 62-63) suggests that the rival models should be treated as complementary rather than contradictory ways of viewing society. According to him, the debate between two outstanding social theorists, Karl Marx and Max Weber, is 'complicated by the fact that the two men were trying to answer different questions about inequality. Marx was especially concerned with power and with conflict, while Weber was interested in values and life-styles. The class model has become associated with a view of society as essentially conflictual, minimizing solidarities, while the model of orders has become associated with a view of society as essentially harmonious, minimizing conflict.'
- 10. The status inflation of the middle ranks, causing considerable problems for social classification in research, has been commented on in various studies (e.g. Brooks 1994a, 1994b). Sabine Johnson's husband John, a wealthy merchant operating in foreign trade, acquired an estate in Northamptonshire, and so fulfilled one of the gentry criteria, viz. landownerhip.
- 11. In previous studies, especially in Nevalainen & Raumolin-Brunberg (2000a), we only compared three different areas, London, East Anglia and the North. It seems, however, that the geographical area of London consisted of too heterogeneous a collection of informants to form one unit. The creation of a new category, the Court, to comprise the royal family and upper-ranking government officials, has given us the opportunity of comparing the language of two very different kinds of inhabitants of (greater) London.
- 12. Sabine herself lived in Glapthorne, Northamptonshire, on the estate the family had acquired, and is not included among Londoners (see note 10, above).
- 13. The choice of the four areas primarily depends on our interest to investigate linguistic variation in regional terms. However, the availability of the material in the end dictated the selection. The amount of data was also decisive in delimiting the northern area. As far as the west of the country is concerned, unfortunately there is not sufficient chronological continuity among the letter collections that we have been able to get hold of.
- 14. There are varying estimates concerning the number of inhabitants in London. The sources include: Beier & Finlay (1986: 3), Finlay & Shearer (1986: 39), Boulton (1987: 3), Rappaport (1989: 61), Dyer (1991: 33), Keene (2000: 97-98) and Barron (2000: 396-397).
- 15. So common were these visits that exceptions have been mentioned as curiosities. It was, for instance, said about one of our northern informants, Henry Clifford, Second Earl of Cumberland (d. 1570), that he withdrew from the public life after the death of his first wife so that he went to London only three times during the last 20 years of his life (Hoyle ed. 1992: 20).
- 16. The CEEC contains some letters written by women with classical learning, such as Margaret Roper (1505-1544; daughter of Sir Thomas More), Lady Anne Bacon (1528-1610), Queen Elizabeth I (1533-1603) and Viscountess Anne Conway (1631-1679).
- 17. The compilation process and principles of the CEEC have also been discussed in the following publications: Nevalainen & Raumolin-Brunberg (1994b); Raumolin- Brunberg (1997); Keranen (1998b); Laitinen (2002). In 2001 a project for tagging and parsing the CEEC was launched in cooperation between the Universities of Helsinki and York.
- 18. Earlier versions have been used in pilot studies. Some of the raw data in this book also go back to the earlier versions, the cases being duly documented in the text. A sampler version called the Corpus of Early English Correspondence Sampler (CEECS) has been released in the second ICAME CD-ROM containing the material that is no longer subject to copyright (Nurmi ed. 1998; Nurmi 1999c, see also Nurmi 2002).
- 19. Jukka Keranen has edited the letters of William Fawnte (1998a), and Minna Nevala a number of letters by East Anglian women (2001). Jukka Keranen, Terttu Nevalainen and Arja Nurmi have re-edited the Marshall letters from the original manuscripts, and Terttu Nevalainen has edited the letters of Elizabeth, Queen of Bohemia, from State Papers 81. (For details, see Appendix III.)
- 20. We would like to thank Dr Arja Nurmi and Dr Minna Palander-Collin for letting us use the list of texts they compiled for their doctoral dissertations. The Supplement contains material from some modernized editions on the one hand and letters included after the completion of the 1998 version on the other.
- 21. In order to help researchers handle the varying authenticity of the corpus material, a code was attached to each letter. Most of the CEEC letters belong to categories A and C.
The authenticity codes are as follows:
A = autograph letter in a good original-spelling edition; writer's social background recoverable
B = autograph letter in a good original-spelling edition; part of the writer's background information missing
C = non-autograph letter (secretarial work or copy) in a good original-spelling edition; writer's social background recoverable D = doubtful or uncertain authorship; problems with the edition, the writer's background information, or both.
- 22. The background information has been collected into a Sender Database, which contains the following parameters for each informant: name, title, year of birth and death, year of first and last letter, sex, rank, father's rank, place of birth, main domicile, education, religion, number of letters, number and kind of recipients, number of words, letter contents and quality, source collection, career and migration history (for coding details, see Nevalainen & Raumolin-Brunberg 1996b; Raumolin-Brunberg 1997).
- 23. Nine letters signed by the Privy Council have been included in the figures for men. They correspond to the small number of official letters signed by the Regents that have also been selected, mostly to serve as material for comparison.
- 24. The people whose writing 'career' crossed the 20-year-period boundaries have been counted separately for each period, which means that, given a long life as a letter writer, one person may have been counted three times. In analysing the social status of those whose position changed, we have assigned each person the status they had at the end of each 20-year period.
- 25. The small amount of edited correspondence representing the early part of the sixteenth century certainly depends on various factors, some of them probably accidental. However, the shortage of surviving correspondence may also have connections with the educational standards of the time (Heal & Holmes 1994: 258).