Mugglestone - The Oxford History of English

198 terttu nevalainen

colloquial or to literary language? And, having made its way into questions and negative statements, why did it fail, after a promising start, to spread to aYrmative statements as well? The following corpus-based survey oVers some answers to these questions, but will hardly provide a deWnitive account of this intriguing phenomenon.

If we look Wrst at modern English, we can see an interesting asymmetry in the use of do. As Chapter 6 has already outlined, if there is no other auxiliary verb in the clause, do is required with not-negation (as in ‘they did not see it’), with inversion, and especially in questions (as in ‘did they see it?’), and with emphasis (as in ‘they ’’did see it’), as well as acting as a prop-word in reduced clauses (‘they saw it, and we did too’). But apart from the emphatic use, do is not required in aYrmative statements (‘they saw it’) when no other auxiliary is present.

Present-day spoken-language corpora suggest, however, that do can sometimes appear in aYrmative statements even when it is without emphasis. The example below comes from the London-Lund Corpus of British English conversation which was recorded in the 1970s (and which is provided with prosodic annotation). In B’s contribution, the Wrst word, I, is stressed, and so is the third, know. But no prosodic prominence is attached to do, which therefore appears to convey no overt semantic contrast or emphasis. In this text, it instead signposts the speaker’s contribution to the discourse topic, that of smoking. Whatever its speciWc function, in aYrmative statements do is more common in modern spoken-language corpora than it is in written-language corpora.

A:but^I !noticed that :Joseph _went :out for ’quarter of an :h_our# at^_one point#^I’m !sure he ’went for a sm/oke# ( - - laughs) - -

B:^I did ‘know :one _Indian ’who . :i!r_onically# -^learnt to ch/ain’smoke#^in this !c\ountry# (London-Lund Corpus of Spoken English S.1.6.606–612)

Unstressed do is also used to mark habitual action in Welsh English and in the south-western dialects of British English from Cornwall to Somerset and Dorset. But even there do is not the only expression available. In Welsh English the habitual past is indicated by means of the simple past tense and the used to construction, as well as the past tense of do, as in constructions such as ‘He went/ used to go to the cinema every week’ and ‘He did go to the cinema every week’. Examples such as these suggest that in aYrmative statements the use of periphrastic do, as it is called in the literature, might have been quite Xexible in the past as well.

mapping change in tudor english 199

origins of do

Few issues in the history of English have attracted as much interest as the rise of the do-periphrasis. There are some uncertain instances of it from Old English, and more certain data from the end of the thirteenth century onwards, but the periphrasis only gains ground at the end of the Wfteenth century in the texts that have come down to us and in which both emphatic and non-emphatic functions are in evidence.

One of the puzzles in the history of do are the circumstances which give rise to the construction in the Wrst place. As it can be seen to assume an aspectual function expressing habitual action (e.g. ‘He did walk to school every day’) in traditional south-western dialects (see p. 198), some scholars argue in favour of its south-western origins, probably prompted by contacts with the Celtic languages in the area. Others suggest that it may have arisen from contacts between English and Anglo-Norman French. Still others look for its origins in causative constructions of the type the king did write a letter that is in the sense ‘the king had a letter written (by someone)/made somebody write a letter’. Because it is attested in early Middle English poetry, there are also suggestions that it started out as a metrical Wller. None of these accounts is perfectly satisfactory, and not least because of problems of localization.

Let us begin by looking at some corpus evidence from the Wfteenth century, the period when periphrastic do began to gain ground. A comparison of the regional data in the CEEC reveals the following trends. The causative construction dominates, especially at Court, in the Wrst half of the Wfteenth century but becomes very rare after 1500. Good examples can be found in the Signet Letters of Henry V, as in the following extract from a letter of 1419 (kynwolmersh refers to William Kynwolmersh, appointed Dean of St Martin le Grand in London in 1421): ‘We wol ye do make a patent vnder oure greet seel vnto þe said kynwolmersh of þe Deanee of saint martines grande yn London’ ([SIGNET] 116). In contrast, periphrastic do occurs particularly in the City of London and to some extent in the west, but remains relatively infrequent throughout the Wfteenth century. A typical instance appears, for example, in a letter written by Richard Cely to his brother George in 1480: ‘the xxvj day of thys monthe I resauyd ij lettyrs frome you, whon to houre father, another to myselue, the qweche I do whell wndyrstonde, and heyr I sende yow . . .’ ([CELY] 84).

One of the signiWcant issues that has been debated in the history of periphrastic do is whether it arose in literary or colloquial contexts. Those who argue for its

200 terttu nevalainen

literary origins suggest that it grew out of the causative function (as in the example in the 1419 Signet Letter discussed above). Conversely, those who are in favour of colloquial origins refer instead to the inXuence of language contact or semantic weakening of the lexical verb do. As we have seen, causative do occurred frequently in oYcial Court correspondence in the early Wfteenth century. But it could also occur in private letters as something of a politeness marker, to indicate that the writer did not necessarily expect the recipient to carry out the request himor herself, as in the following illustration from Margaret Paston’s letter to her husband John in c1453: ‘Also I pray yow þat ye woll do bey a loV of gode sugowr and di. j li. of holl synamun, for þer is non gode in this town’ ([PASTON I] 252).

On the other hand, many instances of periphrastic do in Wfteenth-century London merchants’ letters were rather formulaic, and cannot perhaps be labelled as colloquial (cf. Richard Cely’s use of I do whell understand on p. 199). It seems, therefore, that with periphrastic do the question of colloquial as opposed to literary origins, although useful in cases like provided that, may not be very illuminating. We will return to the issue below.

affirmative and negative do

Periphrastic do clearly gains momentum in the sixteenth century, and interestingly aYrmative statements (its least typical context today) also seem to have played a signiWcant role in the process. In eVect, it looks as though do had the makings of being generalized to all sentence types in Tudor English, had not something interfered with its progress in aYrmative statements. Earlier research suggests that in the sixteenth century the rise of do was being led by interrogatives or questions, as in George Colville’s ‘And doest thou think that such thynges as suYsaunce, and power be, are to be dispysed, or contrarye wyse, that they be most worthy reuerence aboue all thinges’ (from his 1556 translation of Boethius ([BOETHCO] 68–69)). This was followed by negative declaratives, and, at a somewhat slower pace, by the use of do in aYrmative declaratives such as ‘I did mislike the Queenes Mariage’ from Sir Nicholas Throckmorton’s confession of treason in 1554, which will be discussed below on p. 201–2. The non-use of do in interrogatives, and in negative interrogatives in particular, was already much rarer than question-forms which used do, although it could still be found, as the following example, from a 1521 sermon by John Fisher, Bishop of Rochester, illustrates: ‘Seest thou not his eyes, how they bee fylled with blood and bytter teares?’ ([FISHER 1] 400).

mapping change in tudor english 201

However, the fact that aYrmative statements are much more common in communication than negative statements, and especially questions, in fact serves to make aYrmative do numerically the most frequent kind of periphrastic do in texts. We will therefore focus in the following sections on the rise of the periphrasis in aYrmative and negative declaratives. Let us begin with aYrmative do in the multigenre Helsinki Corpus. Figure 7.3 presents the average frequencies of do in aYrmative statements between 1500 and 1710. The development clearly falls into two phases: the use of aYrmative do Wrst increases between the sixteenth and early seventeenth centuries, after which there is a dramatic decline in the latter half of the seventeenth century.5

Focusing on the usage of the sixteenth century, these Wndings could be interpreted to lend more support to the spoken associations of the periphrasis than to the division between colloquial and literary language. The genre with by far the highest average frequency of aYrmative do in 1500–1570, for example, is trial records. While trials cannot of course be called colloquial, they certainly display features of interactive spoken discourse. The use of do is also very common in scientiWc and educational treatises, diaries, sermons, and comedies. By contrast, only a few instances are found in statutes, biographies, the Bible, private letters, travelogues, and histories.

The high incidence of aYrmative do in trials in the corpus evidence is largely due to their clustering in long speeches in the 1554 trial of Sir Nicholas Throckmorton, a diplomat and MP accused of high treason. An extract from this appears on the following page; as can be seen, after the Wrst appearance of do in line 1 in Throckmorton’s confession that he ‘did mislike the Queenes Mariage with Spain ’, the

Fig. 7.3. Periphrastic do in affirmative statements, 1500–1710

5 Figs. 7.3, 7.5, and 7.6 are based on the individual genre scores normalized to 10,000 words provided by the data in Rissanen (1991: 325), Nurmi (1999a: 169), and Meurman-Solin (1993: 262–3), respectively. The Wgures show how many times do could, on average, be expected to appear in every 10,000 words of text in each period.

202 terttu nevalainen

repetition of do is hardly emphatic or contrastive. Instead it could serve as a device to mark the relevance of the actions narrated by Throckmorton in response to the questions being put to him:

I confess I did mislike the Queenes Mariage with Spain, and also the comming of the Spanyards hither: and then me thought I had reason to doe so, for I did learne the Reasons of my misliking of you M. Hare, M. Southwell, and others in the Parliament House; there I did see the whole Consent of the Realm against it; and I a Hearer, but no

5Speaker, did learne my misliking of those Matters, conWrmed by many sundry Reasons amongst you. (1554, State Trials [THROCKM I] 66, C1–C2)

The other genres with high frequencies of aYrmative do also display features of spoken interaction, such as Wrst-person narration and references to the audience. A cluster of aYrmative do’s can be found, for instance, in Robert Record’s 1551 First Principles of Geometrie, in which he justiWes to his readers the necessity of introducing one more category of circles:

Nowe haue you heard as touchyng circles, meetely suYcient instruction, so that it should seme nedeles to speake any more of Wgures in that kynde, saue that there doeth yet remaine ij. formes of an imperfecte circle, for it is lyke a circle that were brused, and thereby did runne out endelong one waie, whiche forme Geometricians dooe call an egge

5forme, because it doeth represent the Wgure and shape of an egge duely proportioned (as this Wgure sheweth) hauyng the one ende greater then the other. ([RECORD] B2R)

Corpora again enable us to trace change through time, and in 1570–1640 the use of aYrmative do picks up in almost all HC genres. The only exceptions are comedies and, again, trials where its usage clearly declines despite the overall rising trend. This apparent deviation has been accounted for by the greater likelihood of the record of spoken language (together with the imitation of this in drama), reXecting changes which were indeed taking place at this time in real spoken interaction. By the last period covered by the HC, 1640–1710, a rapid decline can also be seen in these patterns of usage across the rest of the genres too. Nevertheless, and despite this general pattern, non-emphatic aYrmative do was to persist well into the eighteenth century in many written genres.

In contrast, do continued to advance in negative declaratives in this last HC period, but the process was not completed by the end of the seventeenth century. This is evident if we list all negative declarative sentences with not in the HC, and compare the number of instances which contain do (as in I do not mean) with the corresponding simple Wnite verb forms which are used without do (as in I mean not). Figure 7.4 presents the results, showing a steady increase in the use of do at the expense of the simple Wnite form.

Just as in the earlier discussion of the shifts which can be observed over this time with reference to the singular third-person endings of verbs, the process of do-generalization in negative declaratives was partly one of lexical diVusion. A group of verbs called the know-group (including know, doubt, mistake, trow (‘to believe’), and wot (‘to know’)) lagged behind the general development. Do only began to be associated with these verbs from the seventeenth century onwards. This development can also be observed in the CEEC. As a result, the do-less I know not which appears in the Wrst extract below, taken from a 1547 letter from Queen Catherine Parr to Lord Admiral Seymour, is more typical of sixteenth-century usage than is the I do not know which we can see in the 1572 letter of the humanist and author of The Arte of Rhetorique, Thomas Wilson, to Bishop Parkhurst, and which is given in the second extract below:

My Lord where as ye charge me with apromys wryttin with myne one hand, to chaunge the two yeres into two monethes, I thynke ye have no suche playne sentence wrytten with my hand; I knowe not wether ye be aparaphryser or not, yf ye be lerned in that syence yt ys possyble ye may of one worde make ahole sentence . . . ([ORIGINAL 2] 152)

I do thinke if Mr. Mynne might haue but this moch, he wold be some what satisWed; and how your Lordship can of right denie this moch vnto hym, I do not know. ([PARKHURST] 107).

Overall, the correspondence evidence suggests that men generally used do more than women both in aYrmative and negative statements in the late sixteenth century. However, the gender preference changed in both processes in the seventeenth century, as women took the lead in their divergent developments.

204 terttu nevalainen

the fall of affirmative do

The correspondence corpus can also tell us more about the history of do in aYrmative statements. More speciWcally, it may be used to date the time when its progress came to a halt, and a fall in its frequency began. As shown by Figure 7.3, corpus data suggest that the use of aYrmative do reached its peak between 1570 and 1640. By contrast, earlier Wndings (based on a less controlled genre selection) date the beginning of its fall to the 1570s. In a case like this, diachronic comparisons will be easier to make if they are drawn from genres that can be sampled at shorter intervals. Figure 7.5 presents the development during the crucial period in the correspondence corpus. As this indicates, the CEEC evidence suggests that aYrmative do was used very frequently in the Wrst two decades before 1600, but that its use plummeted during the Wrst decade of the seventeenth century. Do did not recover from this drop but continued to be used at this much more moderate level in the following decades.

But, importantly, there were also regional diVerences in the use of do. If we compare Nurmi’s (1999) Wndings on London, the Royal Court, East Anglia, and the north (see the Further Reading for this chapter), we can see that in the two decades before 1600 aYrmative do was very common among East Anglian writers and those resident at Court, or attached to it, as indeed in Queen Elizabeth’s usage in the following example from 1592: ‘Wel, I wyl pray for you, that God wyl unseal your yees, that to long haue bin shut, and do require you thinke that none shal more joy therat than myselfe’ ([ROYAL 1] 70). It was also commonly attested

206 terttu nevalainen

British English data from the same period of time (cf. Fig. 7.3). Incidentally, the middle period, 1570–1640, shows roughly the same average level of do-use as does the London and the Royal Court in the English data after 1600.

In the selection of thirteen prose genres which are included in the Scots corpus, it is mid-sixteenth century pamphlets, such as William Lamb’s Ane Resonyng of ane Scottis and Inglis Merchand betuix Rowand and Lionis of 1550 (an extract of which is given below), which Wrst display some instances of aYrmative do. The rise of this device in pamphlets may be connected with both Latinate rhetoric and southern inXuence:

Eftir the refusale to restore þe thre or four aikiris of mure, commissionaris of baith þe realmes did proclame þat guid ordour suld be keipit and obseruit, as wes accustummat for guid reule on þe bordouris, and siclik Lord Maxwell, Warden of þe Scottis West Merchis, did proclame guid ordour. ([LAMB] 47).

However, as Figure 7.6 indicates, the frequency of aYrmative do rises slowly towards the end of the seventeenth century in Scots, becoming particularly common in trials and educational treatises. The clustering of do in the passage below (cited from the 1688 trial of Philip StandsWeld for the murder of his father Sir James StandsWeld is reminiscent of the cluster in the Throckmorton trial which was discussed on pp. 201–2:

. . . he did attempt to assassinat, and oVered violence to his fathers person, and did chase and pursue him upon the King’s high way at Lothian-burn, and did Wre Pistols upon his father. And likewayes upon one or other of the dayes, of one or other of the moneth of one or other of the years of God above speciWed, he did attempt to assassinat his father

5for his life, at Culterallors, and did Wre Pistols upon him. ([STANDSFIELD] 4–5)

In general, literary evidence suggests that the do-periphrasis was established in Scots later than in southern English in other sentence types, too. With some dialectal exceptions, contemporary Scots follows general English usage.

linguistic motives for-do

As aYrmative do has attracted a great deal of scholarly interest over the years, there are numerous suggestions as to the motives which triggered its use in texts. But we should not forget that, while aYrmative do is more frequent in texts than is do in the other sentence types (i.e. interrogatives and negatives), when we think of absolute numbers, it is obvious that, even in its peak period, it does not occur

mapping change in tudor english 207

in the majority of aYrmative Wnite clauses. Queen Elizabeth’s use of aYrmative do, for example, occupies the middle range with less than one do in every ten clauses that have no other auxiliary.

We can, for example, see that syntactic conditions motivate the introduction of do to negative declaratives and to clauses which involve inversion such as interrogatives: in these contexts, it provides a carrier for the tense, mood, and polarity of the clause when no other auxiliary is present. This is, of course, also true of aYrmative do. However, many scholars argue that the appearance of do in aYrmative declaratives in the sixteenth century was not so much to do with syntax—that is, with introducing an auxiliary to all sentence types. Instead they suggest that the inXuence of textual and stylistic factors which operate in response to certain structural features (constraints) in the clause could have been more important. These are related to structural complexity and ease of information Xow. An adverbial separating the subject from the verb, for instance, makes the clause harder for the reader to parse. Inserting do into a context like this can facilitate it.

Looking at the Wrst extract below, from George Colville’s 1556 Boethius translation, structural reasons for introducing do are worth considering. Both instances of do here occur in relative clauses, and the second one in particular has several structurally marked features: the subject (thou) is separated from the two clause-Wnal verbs (defyne or detemine) by an adverbial (a litel before). The clause would have become awkward to pronounce with simple past-tense forms of these verbs (*defynedst or determinedst). In her own translation of forty years later (which appears as the second extract below), Queen Elizabeth does not use do-support, but neither does she relativize the second clause. She makes do with a single verb, which she puts in the perfect, and her adverbial phrase comes after the verb:

In the which I do iudge to inquyre fyrste, whether anye suche perWt good (as the same that thou a litel before dyddyst defyne or determine) myght be in the nature of thyngs, that no vayne imaginacion or shadowe deceyue vs, and put vs out of the trewth of the thynge or matter, that we be aboute to talke of. ([BOETHCO] 73)

In which Wrst this I think to be inquyrd of, whither any such good ther be, as thou hast defynd a lyttle afore, among natures woorkes, leste a vayne imagination of thought deceaue us wyde from the truthe of that we talke of. ([BOETHEL] 61)

In the HC, features conducive to structural complexity were found in a large number of aYrmative statements with do, especially in typical written genres. But this was not the case with typical spoken genres such as trials, which displayed few instances of these structurally marked uses.