Mugglestone - The Oxford History of English

7

MAPPING CHANGE IN TUDOR ENGLISH

Terttu Nevalainen

Davphine. Why? whom do you account for authors, sir Iohn Daw? Daw. Syntagma Iuris ciuilis, Corpus Iuris ciuilis, Corpus Iuris canonici, the King of Spaines bible.

Davphine. Is the King of Spaines bible an author? Clerimont. Yes, and Syntagma.

Davphine. What was that Syntagma, sir? Daw. A ciuill lawer, a Spaniard. Davphine. Sure, Corpus was a Dutch-man.

Clerimont. I, both the Corpusses, I knew ’hem: they were very corpulent authors.

Ben Jonson, Epicoene, or The Silent Woman (1616), II.iii.

NOW, four hundred years on, Clerimont in Jonson’s Epicoene is not too far oV the mark when he thinks Corpusses are authors. Modern corpora (or corpuses) are structured collections of texts, both written and spoken. DiVerent kinds are available for language studies. A multigenre corpus contains a variety of genres, and a single-genre corpus consists of only one, such as personal letters, pamphlets, or newspapers. Both types usually have multiple authors. Singleauthor corpora also exist, with the Shakespeare canon as a case in point. As corpora are usually digitized, it is easy to run searches for words and construc-

tions in the texts they contain.

Over the last couple of decades, electronic corpora have greatly enriched the study of the history of the English language. Giving quick and easy access to a

mapping change in tudor english 179

wide selection of data, they have made it possible to explore how the language was used not only in successive time periods such as Middle and Early Modern English, but also in various genres, and by diverse groups of people. Apart from the Corpus of Old English, which contains all extant texts from that period, most historical corpora consist of text selections. They aim, in essence, to provide a window on diVerent kinds of writing from administrative documents to early science, handbooks, sermons, Wction, drama, and personal letters, to name but a few. The number of extant genres grows with time as literacy improves and new genres come into being, such as the private diary in the sixteenth century, the newspaper in the seventeenth, and the novel in the eighteenth.

In this chapter, historical corpora will be used to shed light on some of the details of how the English language changed during the Tudor era, roughly, in the sixteenth century (although seventeenth-century English will also be considered at various points). As other chapters in this volume have already stressed, period divisions of this kind are arbitrary in that language change rarely if ever coincides with royal dynasties—or indeed with any of the other landmarks commonly found in history books. The time span adopted in this chapter will therefore be introduced not in terms of absolute boundaries but as a core period for the linguistic processes which will be discussed. As these processes partly extend beyond the sixteenth century, the time span could equally well have been labelled ‘the Tudor-Stuart period’. This would have accounted for the fact that what was the Tudor period in England was already part of the Stuart period in Scotland, the linguistic characteristics of which will also be included in our discussion.

As noted in Chapter 5, by the sixteenth century English spelling no longer contained much information that could help us identify a writer’s dialectal background. This is obviously the case in Jonson’s Epicoene, printed in 1616. But the Tudor era also represents the time before prescriptive grammars, and so enables us to see how grammatical changes spread quite unmonitored in the language community, often replacing other, earlier, or more local features as they did so. The use of corpora as a means of investigation importantly enables a close-up of such change, enabling us to map the details of shift and variation in ways which are otherwise impossible. Although often neglected in traditional histories of the language, corpus evidence of this kind is, therefore, extremely valuable, a means of taking us much closer to the ‘real English’ of the day, and the complexities of language history as it was enacted through the usage of a wide range of writers.

In this context, private writings—such as personal letters and diaries—oVer considerable insight into how Tudor English was used by individuals, by women

180 terttu nevalainen

and men, northerners and southerners, and by a range of people from diVerent walks of life. All of these necessarily drew on the English of their time but, in doing so, they often made diVerent linguistic choices where choice was available. DiVerent people and groups of people could hence become leaders of linguistic change, promoting new forms, picking up on-going changes, or avoiding traditional forms such as the second-person pronoun thou (an important shift in Tudor English which we will examine in detail later in this chapter). Such speakers can thereby be seen as instrumental in changing the language of their day as many of the changes they implemented eventually diVused throughout the language community. Many features promoted in Tudor English have also become part of modern English—of both mainstream regional varieties and the standard variety alike.

The majority of this chapter will deal with two important processes of change in Tudor English: one that aVected the third-person singular verbal ending (e.g. he knoweth, which was gradually displaced by he knows), and one that introduced the auxiliary do into English (so that structures such as they know not were gradually displaced by they do not know). Both are critical aspects of change in the English of this time, and they have attracted a good deal of scholarly interest. The evidence provided by electronic corpora is nevertheless able to give us a more rounded picture of both of them, but it has also raised some new questions for further studies. Some of these questions are related to other processes of change as, for instance, in the Early Modern English pronoun system (including the disappearance of the pronoun thou). This process will also be traced in the light of corpus data, and the evidence of change and variation which it can illuminatingly provide.

some historical corpora

The Tudor period from the late Wfteenth to the early seventeenth century (1485–1603) provides us with a rich array of public and private writings, a selection of which has been sampled for the multigenre Helsinki Corpus of English Texts (henceforth referred to as HC). The corpus spans Old, Middle, and Early Modern English, paying attention to both genre continuity and innovation across time. It is organized into shorter sub-periods, two of which—1500– 1570 and 1570–1640—are of particular interest for our study of Tudor English.

mapping change in tudor english 181

Both consist of a matching set of Wfteen genres ranging from the typical formal kinds of writing such as the Statutes of the Realm to more informal kinds such as comedy.1

Most of the genres included in the HC were publicly distributed or appeared in print (autobiographies, handbooks, philosophical and educational treatises, histories, and plays) but, where possible, private writings were also included (such as diaries and personal correspondence). Language composed for oral delivery (such as sermons or plays) was similarly sampled, as were texts originally produced in the spoken medium (such as trial proceedings). Using a selection of materials like this we can, for example, trace back processes of change in the grammar of Tudor English which emanate from the more oYcial written end of the genre spectrum as opposed to those that were Wrst manifested in informal, colloquial texts.

The distinction between oYcial and informal colloquial genres is relevant in that oYcial genres were often modelled on French and Latin which, as Chapters 3 and 4 have noted, had much longer histories in England as languages of the law and administration than was true of Wfteenthor sixteenth-century English. It is clear that many formal features such as complex subordinating conjunctions came into English through these channels. The passage below, for instance, illustrates an early case of provided that (‘on condition that’, ‘if’) in the Statutes of the Realm for 1489–91 as sampled for the HC:

Except and provided that yt be ordyned by the seid auctorite, that the lettres patentes late made by the Kyng to Thomas Lorde Dacre of Maister Foster of the seid forest, stand and be goode and eVectuell to the same Thomas after the tenor and eVecte of the same lettres patentes, the seid Acte not withstondyng. ([STAT2 II] 532)2

Since all the Wfteenth-century instances of this conjunction in the HC come from statutory texts, as do nearly all sixteenth-century cases, a convincing case can be made, based on corpus evidence of this kind, that provided (that) entered the English language through legal and administrative use in the Wfteenth century.

1 Each genre in the HC is typically represented by two texts, and each longer text by two samples, so as to make up the minimum of 10,000 words per genre per sub-period. Letters, trials, and the Bible have been sampled up to 20,000 words per sub-period. The HC is large enough for the study of grammar change, but it may not give a reliable picture of lexical changes, especially with less common words, where a larger corpus is needed. As English spelling was becoming standardized in the course of the Tudor period, only private writings by less educated people and imitation of speech in drama can provide some information on the pronunciation of the time.

2 The corpus examples cited are identiWed by the year of writing/publication, the name of the writer and the text and, in square brackets, the short title of the text in the corpus (HC, HCOS), or the name of the letter collection (CEEC), followed by a page reference. Any emendations such as expansions of abbreviations have been italicized.

182 terttu nevalainen

Provided that is also found in a 1554 trial for high treason, in which Sir Nicholas Throckmorton was accused of conspiring to prevent Queen Mary’s marriage to Philip of Spain, but there, too, it appears in a passage that quotes from an earlier statute:

. . . yet there is another cause to restraine these your strange and extraordinarie Constructions; that is to say, a Prouiso in the latter ende of the Statute of Edwarde the Thirde, hauyng these Wordes: Provided always, if any other Case of supposed Treason shall chaunce hereafter to come in Question or Trial before any Justice, other than is in

5the said Statute expressed, that then the Justice shall forbear to adjudge the sayd case, untill it be shewed to the Parliament to trie, whether it should be Treason or Felonie. (1554, State Trials [THROCKM I] 75.C1)

If we trace this change further, we can see that although the conjunction continues to be favoured in legal language, it also Wnds its way into less formal contexts of use towards the end of the sixteenth century and at the beginning of the next. It can be found, for instance, in Gervase Markham’s Countrey Contentments of 1615, a book on husbandry which gives instructions on farming and housekeeping. The excerpt below comes from a section on exercising horses:

As for the quantity of his exercise it must be according to his foulenes or cleannes; for if he be very foule you must then exercise moderatelie to breake his grease, if halfe foule, halfe cleane, then somewhat more to melt his grease, if altogether cleane; then you may take what you please of him (prouided that you doe nothing to discourage his sprits). ([MARKHAM] 77)

Nevertheless, there are fewer than ten instances of the conjunction in the entire corpus of William Shakespeare’s plays (and none in the The Merry Wives of Windsor which was sampled for the HC). The following example comes from Act IV of The Two Gentlemen of Verona:

I take your oVer, and will liue with you, Prouided that you do no outrages

On silly women, or poore passengers. (IV. i. 69–71)

Using a range of corpora is particularly useful for establishing the processes of change which may be at work, especially when we consider the variety of usages which may concurrently exist within a given period. The Helsinki Corpus of Older Scots (HCOS), for example, follows the period division of the HC in the sixteenth and seventeenth centuries and with a similar spread of genres. Importantly, however, it gives us an opportunity to compare the pathways of change in Scots and southern British (i.e. English) English. So the new

mapping change in tudor english 183

conditional conjunction found in Scottish legal texts in the sixteenth century is, in fact, not the past participle form provided that but the present participle providing (always) that, as can be seen in the following example taken from the HCOS evidence of the 1555 Peebles Records: ‘The inquest ordanis to ansuer Robert Atzin, and ilk ane of the oYcaris, of ane ferlot of meill in this storme to help thair wiYs and barnis, providing allwayis that thai clame na possessioun thairof in tyme cuming’ ([PEEBLES 1] 225). The past participle form is only generalized in Scottish texts in the seventeenth century, presumably under southern inXuence after the Union of the Crowns of Scotland and England in 1603.

The Tudor era is also covered by the Corpus of Early English Correspondence (CEEC), which is speciWcally designed to facilitate the study of social variation in language use. It consists of personal correspondence, private letters written by one person to another. The way this corpus is structured allows great Xexibility in analysing periods as short as twenty years (or even shorter), while its range of female as well as male writers facilitates the investigation of the impact of gender on language change and variation (a feature which is clearly important in language history but one which, as previous chapters have shown, is often hampered by lack of evidence). The CEEC also contains letters deriving from writers of various social and regional backgrounds. When data were sampled for the CEEC, particular attention was paid to letter writers from London, East Anglia (Norfolk and SuVolk), and the North. London writers proper were, in addition, separated from those attached to the Royal Court in Westminster in order to make it possible to compare their language use.

It is a sign of the less formal nature of the CEEC letter corpus that there are no more than half a dozen instances of provided that in the sixteenth-century data (which amounts to almost a million words). One of these comes from a letter written by the Norfolk lawyer Stephen Drury to Nathaniel Bacon, a local JP and future sheriV of Norfolk, in 1583. The Wrst instance of provided, reproduced here in curly brackets, was deleted by Drury himself:

I, thinking yt would come thus to passe {provided} and supposing (as inded yt followed) that Hast would be this day at Aylesham church, provided that Mr Neave who had no notyce of the countermaund should be there to arrest him, who came accordingly to Aylesham churche. ([BACON II] 270)

In this chapter the CEEC will be used to examine changes which spread from less formal language use across the language community, as well as the fundamental role of language users (both men and women) from diVerent parts of the country in shaping Tudor English.

mapping change in tudor english 185

based on the third-person singular indicative endings. North of a line running between Chester in the north-west of England to the Wash in Lincolnshire in the east, the ending -(e)s was used, whereas to the south of this line, the dominant form was -(e)th (typically spelled, with a thorn, as -(e)þ). This situation reXects the disparate geographical origins of the two suYxes: as Chapter 4 conWrms, -(e)s is Wrst attested in northern texts in Old English whereas -(e)th is found in southern and Midland dialects. This old dialect boundary can partly account for some traditional dialects in the south-west retaining -(e)th until the twentieth century. But it does not explain how the originally northern form -(e)s came to be generalized in the south as well.

-(e)s from the north

Let us Wrst look at the sixteenthand seventeenth-century evidence which is provided by the Helsinki Corpus to get an overall picture of how this change unfolded in time. First of all, we learn that zero forms (which were discussed in the previous section), as in the modern regional ‘he like her’, hardly occur at all in most genres, although they are occasionally recorded in private letters. The two instances below come from a 1625 letter of Lady Katherine Paston, a Norfolk gentlewoman (and a descendant of Margaret Paston whose own linguistic usage was discussed in Chapter 5). Katherine Paston is writing to her son, who was a student in Cambridge in the mid 1620s:

thy father haue bine very ill with his owld truble in his Legge so that he haue kepte his bede with it this 5: or 6: days, but now god be thanked it is on the mendinge hand . . . ([KPASTON] 77)

One reason why this zero form did not spread may be that it was also used to signal the subjunctive mood (as in ‘they insist that he go’ i.e. ‘should go’), which continued to be in common use throughout Renaissance English. There may of course also have been dialect diVerences that have gone unrecorded because most rural dialect speakers at this time could not write, and so did not leave any personal record of their language for posterity. We only know that the zero form did not make its way into the supralocal usage that was being established among the literate section of the Tudor and Stuart language community.

This leaves us with the two alternatives, -(e)s and -(e)th, in the third-person singular present-tense indicative. The HC evidence, which comes from both published and private sources, suggests that the use of -(e)s was in fact negligible

186 terttu nevalainen

at the national level in the period 1500–1570; it occurs in a mere 3 per cent of the cases. It was instead the southern -(e)th which was the dominant form in most kinds of writing from the Tyndale Bible to sermons and trial records. Nevertheless, -(e)s continued to spread, and in 1570–1640 it had already achieved a mean frequency of 20 per cent of all the third-person singular present-tense endings over a selection of HC genres (diaries, histories, oYcial and private letters, sermons, and trials).

Average Wgures such as these, however, can only describe a change in progress in very general terms. In order to Wnd out in more detail the kinds of texts (and genres) in which the incoming form Wrst appeared, we need to dig deeper. Here again, corpus evidence proves its value. The HC data, for example, conWrms that there were notable diVerences between genres in the use of third-person endings. A comparison of diaries, histories, and private and oYcial letters reveals that it was in fact only private letters that had any instances of -(e)s to speak of between 1500 and 1570. Typically, it occurred in the letters of northern writers, as in an extract from the following letter which was written c1506 by Dame Isabel Plumpton to her husband Sir Robert (the Plumptons were a Yorkshire gentry family):

Sir, I have sent to Wright of Idell for the money that he promyst you, and he saith he hath it not to len, and makes choses [\excuses\] and so I can get none nowhere. ([PLUMPTON] 198)

But even Isabel Plumpton alternates between -(e)s and -(e)th, as in her use of -s with make and -th with say and have in the second line of this extract. This is, in fact, a general pattern in the data. There are a few verbs, notably do, have, and say, which take the incoming -s ending later than others. As a result when, in the latter half of the seventeenth century, most other verbs have more than 90 per cent of -(e)s according to the evidence of the corpus, do still takes it in only half of the cases, and have in merely one third. Such patterns are common in language change. A change usually spreads gradually to all relevant contexts, but it can also have word-speciWc restrictions and can thereby proceed, just as in the case of -(e)s, by means of a process known as lexical diVusion.

In the next HC period, 1570–1640, the overall use of -(e)s with verbs other than do and have soars to some 80 per cent in private letters, and comes to about one third of the instances in trials and oYcial letters. This pattern of spread from the private, informal end of the genre spectrum is, of course, precisely the reverse of that which we found with the conjunction provided (that) which, as we have seen, Wrst gained ground in formal genres, and only afterwards spread to informal ones in the course of time. Meanwhile, to return to the indicative endings, it was the southern -(e)th form which, becoming associated with more formal registers, soon

mapping change in tudor english 187

gained a distinctly ‘literary’ status in general use. This passage from a sermon against ‘usurie’ (or excessive gains made by lending money) by the ‘silver-tongued’ preacher Henry Smith illustrates a typical context for -(e)th around 1600:

Now, al the Commandements of God are fulWlled by loue, which Christ noteth when hee draweth all the Commandements to one Commandement, which is, Loue God aboue all things, and thy neighbour as thy selfe: as if hee should say, hee which loueth GOD, will keepe all the Commaundements which respect God, and he which loueth his neighbour will keepe all the Commaundements which respect his neighbour. (1591, H. Smith, Of Vsurie [SMITH] B4R)

The approximate date for this wider generalization of -(e)s based on the HC gains direct support from the Shakespeare corpus. In Shakespeare’s early plays, that is those written between 1591–99, the dominant ending with verbs other than have and do is -(e)th, and -(e)s appears in only one Wfth of the cases. In his later plays, however, those written between 1600–13, the situation is reversed, and it is instead -(e)s which is used in the vast majority of cases.

We can follow the process of change even more closely by referring to some of the other corpora which have been discussed above. In the Corpus of Early English Correspondence, for example, the change can be traced within shorter periods and with more data. The CEEC conWrms that -(e)s was infrequent well into the second half of the sixteenth century, occurring on average in less than 10 per cent of all possible cases. Figure 7.1 presents the increasing frequency of -(e)s towards the end of the century and in the Wrst half of the next. It reaches 50 per cent around 1600, when -(e)th and -(e)s are almost equally frequent in personal correspondence:

Yet even these Wgures hide a great deal of variation. If we make a comparison between male and female writers, a systematic diVerence can be seen to emerge between the two sexes and their patterns of indicative usage. Throughout the