Regional variation and the language of English witness depositions 1560-1760: constructing a 'linguistic' edition in electronic form *

Merja Kytö, Peter Grund & Terry Walker (see other home page)
Uppsala University


This article has a twofold aim: to introduce ongoing work on an electronic edition of English witness depositions from the period 1560-1760; and to demonstrate in two case studies that this edition is particularly appropriate for studies of regional variation in Early Modern English. The first part of the article outlines the background and methodology of the project. It stresses the need for an accurate, large-scale electronic database of witness depositions based on transcriptions from the original manuscripts. These manuscripts originate from a variety of regions across England. In the second part of the article, the two case studies illustrate the importance of region on language use. Regarding the third person neuter pronoun forms (hit, it, 't, and him), the older form hit is only found in the North-west. In the choice between was and were with third person plural subjects, was is only frequent in the North, in particular the North-east.

1. Introduction

Written records of the oral testimony of a witness in a criminal or ecclesiastical trial give a glimpse into the lives and language of ordinary people in the past. These witness depositions are generally rendered in the third person by the scribe recording the testimony, but direct speech is often found when earlier speech events are cited, as is highlighted in example (1), [1] which is from a servant's testimony, recorded in 1758 in north-west England:


This Informant on her Oath Saith That In the
Night between the Thursday the sixth And
Fryday the Seventh day of October last to the
best of her knowledge & belief she heard her
Master Mr William Dobinson Call out Murder
upon which she got up & Came down Stairs,
at the Stairfoot She Saw a light in her Master's
room & heard a Man Say to him you must
deliver it & you Shall deliver it, upon which
this Informant Said Lord Bless us what is the

(National Archives, London. Assizes, Northern Circuit, Criminal Depositions. MS ASSI 45/26/2, the information of Margaret Scott, 1758) [2]

Depositions are of particular interest to historical linguists because they give a hint of what the spoken interaction of the day might have been like, in different regions of England. They also frequently give valuable sociohistorical information such as the sex, age, and occupation/status of those involved.

We are currently working on an electronic text edition of English depositions spanning the two-hundred-year period from 1560-1760, with the intention of making such records more accessible for linguistic research. In this article we will describe the project, and comment on why such an edition is needed, especially as regards the study of regional variation. We will then show in two case studies how the extralinguistic parameters of region and period can have an effect on particular linguistic features in depositions.

2. The project

Our electronic edition will be published on CD-ROM or the Internet, and will contain both a transcription (the so-called 'readable' version) and a version adapted to facilitate computer searches (the so-called 'searchable' version). Both of these will be based on the first-round transcription that we are currently working on. Moreover, the edition will include some manuscript images (enabling the study of scribal hands etc.), a historical and linguistic introduction, textual notes and a glossary, as well as an index containing sociohistorical information on those involved in the cases that the depositions pertain to (for a discussion of these auxiliary features, see Kytö, Walker, and Grund 2007).

We illustrate our working procedure below using an extract from a sailor's testimony, recorded in 1649 in south-east England.

Image 1 (click to enlarge)

Image 1. Essex Record Office, Chelmsford. Colchester Borough, Book of Examinations and Recognizances 1647-1687. MS D/B 5 Sb2/9, detail. Reproduced by courtesy of Essex Record Office.


First-round transcription:

[...] And this informant
reproueing of the ^{s=d=} Annable for calling my Lord
Deputy of Ireland Rogue: The said Nuttle
replyed, I saie ^{Cromwell} hee is A Rogue: And the said
Diglott saied, And ^{Diglott saied} I saie he is A Rogue: And
Thaxston saied if I ^{saie} he be a rogue or an honest
man what can you saie to it. And this
informant saieing to the said Henry Nutton if
you loue the Kings p%tie why Doe you not goe to him
The said Nuttall replyed I can Doe him as good
seruice heere as I can Doe him there:


'Readable' version:

[...] And this informant
reproueing of the ^{sd} Annable for calling my Lord
Deputy of Ireland Rogue: The said Nuttle
replyed, I saie ^{Cromwell} hee is A Rogue: And the said
Diglott saied, And ^{Diglott saied} I saie he is A Rogue: And
Thaxston saied if I ^{saie} he be a rogue or an honest
man what can you saie to it. And this
informant saieing to the said Henry Nutton if
you loue the Kings քtie why Doe you not goe to him
The said Nuttall replyed I can Doe him as good
seruice heere as I can Doe him there:


'Searchable' version:

[...] [$And this informant
reproueing of the ^{s=d=} Annable for calling my Lord
Deputy of Ireland Rogue: The said Nuttle
replyed,$] I saie ^{Cromwell} [§hee§] is A Rogue: [$And the said
Diglott saied,$] And ^{[§Diglott saied§]} I saie he is A Rogue: [$And
Thaxston saied$] [§if§] I ^{saie} he be a rogue or an honest
man what can you saie to it. [$And this
informant saieing to the said Henry Nutton$] if
you loue the Kings p%tie why Doe you not goe to him
[$The said Nuttall replyed$] I can Doe him as good
seruice heere as I can Doe him there:

The first transcription below the image of the original Colchester manuscript represents our current working transcription, or first-round transcription. The aim of this transcription is to reproduce the features of the manuscript as faithfully as possible with modern typographical means. For example, we have reproduced the lineation, spelling, capitalisation and punctuation. In adopting our transcription principles, we have been influenced by conventions used in e.g. The Helsinki Corpus of English Texts, A Corpus of English Dialogues 1560-1760, and the new edition of the Salem witchcraft records (Kytö 1996; Kytö and Walker 2006; Rosenthal et al. forthcoming). We have used equals signs to signal that something is superscript in the manuscript, as in e.g. "s=d=" for sd (i.e. said). For the crossed 'p' brevigraph representing 'par', 'per', 'por' or 'pur', we have used 'p%'. Text that has been cancelled in the manuscript has been struck through, and supralinear additions have been signalled by enclosing the addition within curly brackets. Our focus in this round of transcription is to record as many aspects of the manuscript text as we can as accurately as we can.

Our first-round transcription will serve as the basis for the two final versions of the transcriptions. The 'readable' version is intended to be as faithful as possible to the original manuscript, and we will thus retain many of the features already present in our working transcriptions. However, in the 'readable' version, we will also reproduce superscript, abbreviations, and where feasible, special characters. For example, we will reproduce the 'p' brevigraph in party as closely as is possible with typographical means (here given as a 'p' with a stroke through it). This version is geared to generally interested readers, who are not necessarily aiming to pursue complex linguistic searches.

Superscript, cancellations and special characters (like the 'p' brevigraph just mentioned) cannot easily be dealt with by search programs. For the 'searchable' version, we will therefore retain some of the coding found in our current first-round transcriptions, and add some other types of codes. Conventions that we plan to keep are the equals signs for superscript and 'p%' for 'par' (or possibly 'per') in this transcription. New coding introduced in this version includes, for instance, the section sign ("[§...§]") for cancellations. We will also distinguish direct speech from the rest of the text by using codes to enclose all text other than direct speech ("[$...$]"). These features will help researchers who are particularly interested in exploring the sections recorded as direct speech, or who want to conduct more complex searches using a tool such as WordSmith.

Our material comes from archives across England: the deposition collections were selected to represent four regions of England (the North, South, East and West) plus the London area (for details, see Kytö, Walker, and Grund 2007). The material, covering the two-hundred-year period 1560-1760, is divided into four shorter periods: the late sixteenth, the early seventeenth, the late seventeenth, and the early to mid-eighteenth century. Our plan is to have 5,000 to 10,000 words from each collection, which will total 200,000-250,000 words for the whole edition. We have collections from church court records (pertaining to offences such as defamation and the breaking of marital contracts) and from criminal court records (with offences such as assault, murder and theft). Map 1 shows the dioceses from which our church court material originates; Map 2 shows the counties from which our criminal court records derive.

Map 1

Map 1. Dioceses represented in the material (church court records).

Map 2

Map 2. Counties and/or towns represented in the material (criminal case records).

3. The need for a 'linguistic' edition

Recently, calls have been made for more linguistically-oriented editions. Such editions would aim at reproducing the original manuscripts more faithfully than critical or eclectic editions do (see e.g. Lass 2004; Bailey 2004; Grund 2006). In a linguistic edition the language of the original manuscript text is not normalised, modernised, or otherwise emended. Rather, the manuscript is reproduced as closely as possible in transcription (for such a printed edition, see e.g. Rosenthal et al. forthcoming, and the discussion in Grund, Kytö, and Rissanen 2004). Such an edition of depositions is particularly desirable. As mentioned above, depositions are a valuable source of data for historical linguists. They give a glimpse of authentic spoken interaction in that they are written records relating to past speech events. They also involve both women and men, especially those of the lower ranks (Sharpe 1980; Gowing 1996). However, those interested in using such material have had to choose between two less-than-ideal alternatives. One option for researchers has been to invest a great deal of time both collecting and transcribing the material. Interpreting the scribal hands can be particularly time-consuming. The image of a manuscript from the diocese of Durham shows a fairly typical example of a late sixteenth-century document. In addition to problems relating to the manuscripts, research based on unpublished transcriptions is difficult for other researchers to duplicate or verify.

Image 2

Image 2. Durham University Library: Archives and Special Collections, Durham. Durham Diocesan Records, DDR/EJ/CCD/1/2, f. 339v. Reproduced by kind permission of Durham University Library and the Durham Diocesan Registrar.

The other option has been to rely on the very few editions which have been published. However, these have generally not been produced with linguistic study in mind, and may not always be reliable, as has been shown by Kytö and Walker (2003) for material from England, and Grund, Kytö, and Rissanen (2004) for depositions from New England.

One of the most important aspects of language for which our edition will provide material is regional variation. Although deposition collections from different areas of England would seem to provide excellent opportunities for investigations of regional differences in Early Modern English, researchers have not always been well served by the currently available editions. For illustration, we will mention three rare or regional features found in our material that have been missed or misrepresented in earlier editions.

The lexical item house-roomth (meaning 'storage') is rare, according to the OED (s.v. house-roomth). It occurs in our material from the East from 1560-1599, as "howse rometh" (underlined in the image). The passage in which it appears is transcribed in example (3). [3]

Image 3 (click to enlarge)

Image 3. Norfolk Record Office, Norwich. Norwich Quarter Sessions Depositions. NCR Case 12a/1c, f. 50r, detail. Reproduced by kind permission of Norfolk Record Office.


[...] / then I answeryd hym
ageyne then shall I Lose bothe Doble Cranage and howse
for he hathe too butt¤ of Sacke in the howse allredy

[butt¤ = buttes; the symbol '¤' represents the abbreviation in the MS for 'vowel + s'; "too butt¤ of Sacke": 'two casks of wine' (OED s.v. butt n.2 and s.v. sack n.3)]

The previous editor misleadingly gives the compound word as "howse rounts" glossing "rounts" as "(rents?)" (Rye 1905:75), thus obscuring the word's origin and meaning.

Dialectal words may also simply be omitted. A case in point is the dialect word "Shudd" (for shed), also from the Eastern material (OED s.v. shud), [4] which is underlined in the image. The previous editor ends his part-paraphrase part-transcription of the text just one line prior to the line where "Shudd" occurs. In example (4), we continue the transcription to include the text surrounding the word "Shudd".

Image 4 (click to enlarge)

Image 4. Norfolk Record Office, Norwich. Norwich Quarter Sessions Depositions. NCR Case 12a/1c, f. 34r, detail. Reproduced by kind permission of Norfolk Record Office.


[...] and so they wente into the
gardeyne together and there in the same gardeyne in A
lyttle Shudd on the right hande as we go into the same
gardeyne toward¤ Nicholas Sothertons the seid Wy@l@l@m@
vyncente and Bennett Goodwyns ded c@o@m@ytt the moste
abhomynable vyce of whoredome

["Wy@l@l@m@", "c@o@m@ytt": William, commit; the symbol '@' represents a line through or above letters, signalling one or more abbreviation letters]

Our third example is from our late sixteenth-century material from the diocese of Chester, in the North-west. In three depositions, 'they' occurs as a demonstrative determiner or, possibly, a plural form of the definite article (OED s.v. they 5). This usage is rare, and the alternatives 'those', 'the' and even 'these' are more frequent in the collection. An example is highlighted in the image of the manuscript text, which is transcribed in (5).

Image 5 (click to enlarge)

Image 5. Cheshire and Chester Archives, Chester. Archdeaconry of Chester Depositions. MS EDC 2/7, f. 183v, detail. Reproduced by kind permission of Cheshire and Chester Archives.


Ad iij she think¤ her good name is
not hurt biecause they word¤ were spoken
in a furye

The previous editor, Furnivall (1897:123), gives the reading as "the" in all three of the depositions in which this determiner occurs. In the material we have transcribed so far, the use of 'they' as a demonstrative determiner/plural definite article is not found outside the North-west, which may be indicative of regional variation. [5]

4. Regional variation

Our previous examples demonstrate that our material contains dialectally restricted lexical items and, in one case, a rare, seemingly regional, construction. The presence of such features opens up the clear possibility that our material can provide new evidence for other types of linguistic variation between different areas of England in the Early Modern period. We still know comparatively little about regional Early Modern Englishes: they have been difficult to study since edited material that can be clearly anchored in a specific region of England is scarce. The ongoing standardisation of English during the period also means that regional features were being increasingly suppressed in favour of the norm in most formal written records (for a discussion of Early Modern regional variation, see Görlach 1999; Nevalainen and Raumolin-Brunberg 2003).

Our edition has clear potential as a source for regional variation as depositions can be located to fairly specific areas of England. For example, Walker (2007) and Kytö, Walker, and Grund (2007) have offered some evidence of regional variation in depositions with regard to second person singular pronouns thou and you. At the same time, there are of course caveats. Depositions were far from immune to standardisation. On the contrary, since legal writing is a very formal genre, standardisation is only to be expected, and so is perhaps even the removal of "inappropriate" regional language. Although the presence of dialectal vocabulary does show that all regionalisms were not sifted out by the scribe, it is frequently difficult to determine whether the language reported by the scribe is indeed that of the witness: the language may in fact be partly or wholly that of the scribe (see Kytö and Walker 2003; Grund, Kytö, and Rissanen 2004; Grund 2007). The level of training of the scribe probably contributed to the picture: the more highly educated would perhaps have been more discriminating, while less trained scribes may have followed the actual language of the deponents more closely. Unfortunately, there is little information on the scribes, and more studies on the Early Modern legal scribes that worked for justices of the peace and different courts remain a desideratum.

Despite these challenges, witness depositions are a promising, largely untapped source for the study of possible regional variation. In two case studies, one on third person neuter pronoun forms and one on was/were variation, we will demonstrate that our edition will indeed provide suitable material for investigations of Early Modern regional variation.

4.1 The material used for our studies

For our two case studies, we have taken depositions from three areas of England: the North (both North-east and North-west), the South, and the East (the Western material has not yet been transcribed). This material is listed in Table 1. Maps 3 and 4 show the areas represented in each period. The transcription of our deposition collections for the edition is still underway, but we have finalised first-round transcriptions of a large part of our material from the late sixteenth century and the early to mid-eighteenth century. We selected depositions from these two periods to enable us to consider variation across time as well as region. The number of words for each period is around 50,000, giving a grand total of just over 100,000 words for these studies.

Table 1. Material used for the present study.
Date Deposition collections Region Size
Period 1:
1560-1573 Ecclesiastical Courts of Durham N(E) 12,985
1561-1565 Bishop's Court, Chester N(W) 12,980
1566-1577 Winchester Consistory Court Depositions S 11,270
1583 Norwich 1583: Examinations E 3,140
1560-1566 Norwich Quarter Sessions E 10,720
Period 4:
1724-1758 Northern Assizes (Northern Counties) N(E) 12,460
1696-1760 Lancaster Crown Court N(W) 8,580
1751 Oxon Sessions Examinations: Rex v Mary Blandy S 19,300
1700-1754 Norwich Quarter Sessions E 9,140
Total 100,575

Map 3

Map 3. Late sixteenth-century material (1560-99).

Map 4

Map 4. Early eighteenth-century material (1700-1760).

4.2 Third person neuter pronoun forms

The third person neuter paradigm underwent several changes in the fifteenth, sixteenth and seventeenth centuries. One of the most fundamental developments is the replacement of genitive his by its (see Nevalainen and Raumolin-Brunberg 1994). As there is only one clear example of the incoming form its, in the late North-western material, and no unambiguous examples of genitive neuter his, we will focus here on the subject and object forms. Previous research has shown that the old form hit, and the old dative him (used as the object form) disappeared during the sixteenth century in formal written English: by 1600, the originally unstressed form it was the usual form (see e.g. Barber 1997:150; Lass 1999:147). [6] At the same time, the use of the new unstressed pronoun form 't increased in the late sixteenth century.

Table 2 gives the number of occurrences of each of these variant forms (hit, it, 't, and him) in our material in the two time periods.

Table 2. Distribution of third person neuter pronoun forms: raw figures and percentages.
Third person neuter 1560-99 1700-60


221 (94%)

404 (99%)


13 (6%)




4 (1%)




Total 234 (100%) 408 (100%)

The table makes clear that the form it is the predominant form in both periods. Him is not attested at all. The reduced form 't does not appear until the later period, and then only rarely. No regional variation can be detected with 't, as the few occurrences are found in the North, East and South. What is interesting is that, since 't appears to be a feature more typical of speech, the scribe's usage suggests that he is trying to graphically represent the speech of the witnesses. As attested by evidence from A Corpus of English Dialogues 1560-1760 (see Kytö and Walker 2006), this form is common in fictional dialogues, especially in plays, as early as the late sixteenth century, perhaps as an attempt on the part of the dramatists to mimic everyday speech. In written records of authentic spoken interaction (such as trials), 't does not appear until the mid-seventeenth century, and is not particularly frequent in the following hundred-year period. Our evidence from the depositions thus agrees with the patterns in other records of authentic spoken dialogue.

From the perspective of regional variation, hit is probably the most interesting item. In late Middle English, hit was used in some regions but not in others. This is demonstrated by LALME, which primarily records hit in western and southern England (LALME 4:310). Our material reveals a continuation of this pattern. Table 2 shows hit in depositions from the sixteenth century. This usage is illustrated in example (6).


[...] and the said
m=~=garet answerid & said yf dost thou denay
hit then thou art a false theif and apon y=t=
fazakerley answerid & said thou art as like a
hoore as I ["I" written over "a"] a thief

(Cheshire and Chester Archives, Chester. Archdeaconry of Chester Depositions. MS EDC 2/7, f. 61v)

Hit is exclusively found in the collection from the diocese of Chester, one of the regions where hit was most common in late Middle English, as shown by LALME. This pattern suggests that region is indeed a factor in the usage. Some confirmation comes from the fact that the other regions represented by our material fall for the most part in those areas where hit was already less common or not used at all in the late Middle English period, again as shown by LALME.

Although hit occurs in the Chester collection it is still a minority form: hit accounts for 19%, or thirteen of the 70 examples, of the third person neuter forms. Thus, the "new" form it dominates. No clear phonological or syntactic constraints emerge favouring either pronoun form: both forms occur after a vowel sound, and after voiced and unvoiced consonants, and in both stressed and unstressed position. Both forms also occur with the same verbs as subject (be, chance, fortune), and object (hear, know, deny). The instances of hit may thus reflect remnant dialectal usage by some informants. This is of course to some extent speculation, although the distribution of the forms does give us some clues. In depositions, there is always a certain degree of uncertainty about the level of scribal influence in the recording of the spoken testimony, as mentioned earlier. Sometimes there is evidence suggesting that the record is actually a verbatim report, while other depositions leave us in doubt (see Kytö and Walker 2003). If the instances of hit occurred in direct speech, it would be tempting to assign these occurrences to the mouths of the witnesses: it is in direct speech that one might surmise that the scribal influence would be least felt. However, our data is not that straightforward. In fact, the opposite seems to be true: only one of the thirteen examples of hit occurs in direct speech (see example (6)), while the other twelve are in reported speech. At the same time, the distribution is still of significance: eleven of the thirteen instances occur in one court case, in the testimony of four of about forty-four informants in the collection. The key point here is that only one scribe was at work in the collection. The near-absence of hit forms in other parts of the collection suggests that this concentration of hit may actually reflect the witnesses' usage.

4.3 Was/were variation

Research has shown that complex systems of verbal agreement and internal as well as external constraints influence the usage of was and were in different regions. The majority of studies have looked at variation in Modern English (e.g. Tagliamonte 1998; Anderwald 2001; Britain 2002). One important exception is Nevalainen (2006), who has considered was/were variation in letters from London, East Anglia, and the North in the period 1460-1680. But, apart from letters, and the areas studied by Nevalainen, there is to our knowledge no research into the development of this variation in Early Modern English. Our edition can thus help provide some further clues to the historical picture of was/were variation. Depositions are actually particularly well-suited for a study of aspects of this variation: due to the prevalence of reported speech, the use of the past tense, particularly in the third person, is a characteristic of this genre. We will focus here on was and were with third person plural subjects. There are several reasons for this limitation. The third person plural is the main context for was/were variation in our material. Although third person singular subjects made up the bulk of our original data, only two examples of were occurred where was would have been expected. First and second person subjects were limited to six examples: five instances of the now standard English forms I was or we/you were, and one thou was, a combination that has been shown to be common in the north, especially in Middle English, but not elsewhere (Tagliamonte 1998:156-157). Not surprisingly, this one example is found in a deposition from the North-east, dated 1734. [7] Due to the scarcity of was/were variation outside the third person plural context, we will hence concentrate on variation there. To contextualize our results, we first provide a brief overview of the findings in previous studies.

In the Middle English Northern dialect, was occurred frequently with plural subjects, particularly with existential there and in relative clauses, but seldom adjacent to a plural personal pronoun subject (Forsström 1948:207). With regard to other dialects, was is used very occasionally in the plural in the West Midlands dialect (Forsström 1948:167). Pietsch (2005:128, 150) argues that the Northern was/were variation developed in analogy to the Northern Subject Rule, according to which present tense verbs took -s with all subjects except I, we, you, and they when these were verb adjacent. Applied to was/were variation in the third person plural, this would predict: the words was spoken, they were married but they, he understood, was married. Pietsch (2005:128-129) states that the Rule has become more variable in Modern English: the type of subject and position of subject act as constraints encouraging rather than predicting a particular form. With regard to was/were variation, this means that subjects other than they take was variably, but was is favoured with all non-adjacent subjects. While was never seems to have been as regular as the present tense -s form with plural subjects in the North, in recent times was/were have been more resistant to the standard English pattern, especially with third person plural subjects (Pietsch 2005:150-151).

As early as the sixteenth and seventeenth centuries, the Northern -s/was with plural subjects spread south, and is found in e.g. Shakespeare, as shown by Visser (1963:72). He also shows that throughout the history of English particular types of plural subject have encouraged -s/was irrespective of region, such as existential there, and two subjects linked by and (Visser 1963:74, 80-81). Like Forsström (1948) for the Northern dialect in Middle English, both Nevalainen, for Early Modern English, and Tagliamonte, for Present-day York, found that existential there was the most typical subject with non-standard was, and personal pronouns the least typical, with NP subjects somewhere in-between (Tagliamonte 1998:162, 181; Nevalainen 2006:362-363). Nevalainen (2006:362-363) furthermore shows that was with plural subjects was particularly common in the North in the fifteenth and sixteenth centuries, but less so in London and especially East Anglia. Was declined in the seventeenth century, notably with NP subjects, and the difference between the three regions evened out.

In our depositions material, there are 158 examples of were and 34 examples of was with third person plural subjects. In other words, was only comprises 18 percent of the 192 examples. As some collections exhibit no variation, Table 3 exclusively presents the data for those collections and regions in which was occurred. We will consider the results region by region.

Table 3. Distribution of was/were with type of third person plural subjects by period and region of collection: raw figures and total percentages.
Type of third
person plural
1560-99 1700-60
was were was were was were was were was were















plural NP















singular NP + singular NP















existential there















relative pronoun















collective noun















pronoun viz NP + NP












































E = Eastern (Norwich)
NE = North-eastern (Durham diocese, period 1560-99; Northern Counties, period 1700-60)
NW = North-western (Chester diocese, period 1560-99; Lancashire, period 1700-60)

The most striking trend, and at the same time perhaps the most expected, is that apart from one solitary example, was does not occur with a third person plural subject in the Southern material (from Winchester and Oxford, excluded from Table 3) or the Eastern material (from Norwich) in either period. This conforms to the findings of previous research (e.g. Nevalainen 2006), which places was used in the third person plural primarily in the North. The one exception is example (7). It occurs in one of the collections from late sixteenth-century Norwich, which otherwise patterns according to the now standard paradigm. Perhaps what is at work here is the principle of proximity, or attraction, suggested by Quirk et al. (1985:757; see also Nevalainen 2006:364). Was would in this case be influenced by the adjacent singular noun phrase, "horsebacke", rather than the actual subject "strypes".


[...] And that the strypes that
Keteringhm@s wyfe ded geve vpon his horse hed w=th=
a waster as he sat on horsebacke was the only cawse [...]

(Norfolk Record Office, Norwich. Norwich Quarter Sessions Depositions. NCR Case 12a/1c, f. 42r)

In the late sixteenth-century North-eastern collection, from Durham, was makes up as much as 61% (or twenty instances) of the examples. Was is used with all types of plural subjects except for they, in accordance with the Northern Subject Rule. The usage is illustrated in examples (8) and (9).


[...] / he saith y=<t>=
the word¤ was layd to his y=e= said th@ chardg w=ch= culde
not then deny them bifor wyttnes

["th@" = Thomas's]


[...] / he saith y=t= he belyvith y=t= they were maried
togither but this ex=~= was not at the mariadge

["ex=~=" = examinant; the symbol '~' signals an abbreviation in the form of a flourish at the end of 'x']

(Durham University Library: Archives and Special Collections, Durham. Durham Diocesan Records, DDR/EJ/CCD/1/2, ff. 97v-98r)

However, particularly notable is that a large proportion of the was examples occurs with 'singular NP + singular NP', and to a lesser extent with existential there. Both have been said to encourage was regardless of region (e.g. Visser 1963:74, 80-81). The 'singular NP + singular NP' environment is especially favourable to was: twelve of the twenty instances of was occur in such contexts (exemplified in (10)). Perhaps a version of the principle of proximity is at work here, the verb form being governed by the closest NP only and not the two combined.


this ex saith that the said Rauff ogle & John Rosse was bothe
in the churche off Stanington as is ar=l@= /

["ar=l@=" = articulated, mentioned in the articles of the suit]

(Durham University Library: Archives and Special Collections, Durham. Durham Diocesan Records, DDR/EJ/CCD/1/2, f. 339r)

Unlike the sixteenth century North-eastern collection, the eighteenth-century material comes from several Northern counties, although most of the data for this eighteenth-century North-eastern collection is from Yorkshire. Our eighteenth-century North-eastern collection echoes the decline of was with NP subjects that was reported by Nevalainen (2006) for the seventeenth century. Were is now the form that accounts for almost two thirds (63%) of the examples. At the same time, there is less data in this period. There are also some striking patterns in the distribution. Several depositions in the same case of manslaughter, from Harden in the West Riding of Yorkshire, actually account for five of the seven was examples. As illustrated in example (11), the evidence seems to point towards a generalisation of was in the Harden case. Although the subjunctive is not treated in this article, it is interesting to note that was is also used where subjunctive were might be expected in the third person singular.


[...] that their was some Words
between Sharp & Rycroft but the first Words this Witness heard
Distincktly was
that Rycroft told Sharp if he wo=d= not be quiet he
(Rycroft) wo=d= make him (Sharp) and gave the Child to another Man
and then took Sharp by his Coat Neck and gave him a Shake [...]
{Sharp after they was parted set his hand,} upon his Knees & bended his body as if #
he was hurt

(National Archives, London. Assizes, Northern Circuit, Criminal Depositions, 1724-1758. MS ASSI45/26/1, f. 115v (1))

Was is consistently used with all third person plural subjects in the Harden case, even they. A striking exception is singular NP + singular NP, which occurs twice, both times with were, as shown in example (12). Actually, these same three introductory lines are used in two depositions. Perhaps the scribe even copied this preliminary information from the first deposition.


Says that on Saterday the second day of July 1757 there
was a Cockfighting at Harden Beck when John Sharp
and John Rycroft were
both present

(National Archives, London. Assizes, Northern Circuit, Criminal Depositions, 1724-1758. MS ASSI45/26/1, f. 115r (1))

In the eighteenth-century North-eastern collection as a whole, this type of subject no longer seems to favour was: only one out of six instances takes was. This is in stark contrast to the earlier collection from Durham, but, again, the number of examples is very low.

Our North-western material shows quite a different pattern from that of the North-east. In the late sixteenth-century depositions from the diocese of Chester, the percentage of was (6%) is the same as for the contemporaneous Eastern collection, from Norwich. In the Chester collection, were is clearly the norm with all third person plural subjects (with the possible exception of existential there). Neither non-adjacency nor non-pronominal subject seems to encourage was. In the eighteenth-century North-western material, from Lancashire, were (73%) is again preferred: was occurs in only two depositions. Again, there is very little data: although was is proportionally more common in the eighteenth than in the sixteenth-century North-western material, there are only four instances. Example (13) is from the deposition that accounts for three of the four occurrences of was. Two of the was examples occur in conjunction with relative pronouns. This is an environment where was was found to be particularly common in the Northern dialect in Middle English (Forsström 1948:207). Nevalainen (2006:364) suggests that Quirk et al.'s (1985) proximity principle may be applicable to relative pronoun subjects: as they are not marked for number, relative pronouns may promote a singular verb, irrespective of the number of the antecedent. Countering the assumption that this is an environment favouring was is the fact that were also occurs after a relative pronoun in this collection.


[...] Then Neighbours came, and found y=t= two of her
Petticoats was
pull'd of w=ch= was next her body, and foulded together
And laid vnd=~= her, and y=e= Pocketts turned y=e= other side out, and all things
gone w=ch= was in them

(National Archives, London. Palatinate of Lancaster: Crown Court: Depositions PL27/2, the information of Robert Emott)

Interesting to note is that the location of the Lancashire deposition quoted in (13) and that of the Yorkshire case where was is preferred are only about twenty miles apart, whereas the Lancashire material with were is from further west, and the other North-eastern material from further east or north.

Our results indicate that there is regional variation with regard to was/were with third person plural subjects. Only were is found in the Southern and Eastern material, apart from one example of was in an early Norwich collection. Hence it is the Northern material, especially the North-east, that essentially accounts for the was examples. This agrees with Nevalainen's (2006) finding that was with plural subjects was most common in the North, but not common in East Anglia. There are also regional differences between the North-east and the North-west. The was/were variation in the late sixteenth-century North-eastern material seems to pattern along the lines of the Northern Subject Rule, but the North-west patterns similarly to the Eastern collection from Norwich, with a clear preference for were. The sixteenth-century Durham collection strongly favours was with 'singular NP + singular NP', a common type of subject in the depositions, but perhaps surprisingly, other collections almost exclusively prefer were. They clearly encourages were in all collections, although rare instances of they + was do occur. More data is needed to determine whether existential there and non-adjacency of subject encourage was (as argued in other studies). [8] Investigating the period between those considered here will also reveal a clearer chronology of development.

5. Conclusion

Our article had two main purposes: to introduce our ongoing work to produce a 'linguistic' edition of Early Modern English witness depositions in electronic form; and to argue for depositions (and hence our edition) as a promising source for studies of regional variation. The key methodological issue for us is to make available a 'linguistic' edition that faithfully reproduces the language of the original manuscripts. At the same time, our edition will be geared to facilitate advanced computer searches. Our second purpose is illustrated by two case studies: one study of third person neuter pronoun forms (especially hit), and one of was/were variation with third person plural subjects. Both studies point to regional variation, thus highlighting that our edition has great potential for investigations into the comparatively unresearched field of Early Modern regional variation.

As we have pointed out elsewhere (Kytö, Walker, and Grund 2007), in our work we combine our philological and editorial aims with principles of modern corpus compilation, striving at a new type of text edition that will also serve as a computerised corpus. In its final form, our edition will thus provide access to previously unexplored material as well as a new methodological perspective.


* We gratefully acknowledge the funding of this project by the Swedish Research Council (Dnr. 421-2004-1310).

[1] Note that it is not always completely clear where the line should be drawn between 'direct speech', 'indirect speech' and other levels of discourse. Thus the possibility of considering the word "Murder" part of the third person narrative cannot be ruled out.

[2] In our transcription, we aim to follow the manuscript as closely as possible. We retain the manuscript punctuation and capitalisation. Some of our other principles are explained further below and after individual examples.

[3] Henceforth, examples are from our first-round transcriptions, from which we will later produce the readable and searchable versions.

[4] The OED records shud as a now obsolete dialectal word used in East Anglia, Yorkshire, Derbyshire, and Herefordshire.

[5] The OED gives this form as "Now dial." (s.v. they 5), but does not specify which dialectal areas. The word 'thae' ('t-h-a-e'), meaning 'those', used as both a demonstrative determiner and a demonstrative pronoun, is said to be a feature of Scottish and Northern English: however, our material does not attest 'they' as a demonstrative pronoun, nor is the spelling 't-h-e-y' for 'thae' recorded in the OED. The use of the personal pronoun form 'they' as a demonstrative pronoun is said to be a southern feature (see e.g. Wakelin 1972:116-117).

[6] Hit is still found in some dialects today, for instance, in the southern United States (Algeo 2003:15).


[...] this Deponent said to the
said Alice thou Wicked Jade thou has hang'd thy Child what did
thou think on ^{was thou sensible}"

(National Archives, London. Palatinate of Lancaster: Crown Court: Depositions PL27/2, the information of Isabell Rossall)

[8] The finding with reference to Modern English dialects (Anderwald 2001:19; Britain 2002:16) that polarity plays an important role in was/were variation (i.e. the tendency for generalised was in positive statements, and generalised were in negative statements) could not be tested, due to the predominance of affirmative statements in our data. In depositions, what happened is more likely to be described than what did not happen. Cf. Nevalainen (2006:360), who suggests that generalisation of were in negative statements is a more recent phenomenon as there was no clear evidence for it in her Early Modern English material.


A Corpus of English Dialogues 1560-1760. 2006. Compiled by Merja Kytö (Uppsala University, Sweden) and Jonathan Culpeper (Lancaster University, England).

The Helsinki Corpus of English Texts. 1991. Helsinki: Department of English, University of Helsinki.

OED = The Oxford English Dictionary. Online ed. 18 April 2007.


Algeo, John. 2003. "The origins of Southern American English". English in the Southern United States (= Studies in English Language), ed. by Stephen J. Nagle & Sara L. Sanders, 125-209. Cambridge: Cambridge University Press.

Anderwald, Lieselotte. 2001. "Was/were variation in non-standard British English today". English World-Wide 22(1): 1-21.

Bailey, Richard W. 2004. "The need for good texts: The case of Henry Machyn's Day Book, 1550-1563". Studies in the History of the English Language II: Unfolding Conversations (= Topics in English Linguistics, 45), ed. by Anne Curzan & Kimberly Emmons, 217-228. Berlin & New York: Mouton de Gruyter.

Barber, Charles. 1997. Early Modern English. 2nd ed. Edinburgh: Edinburgh University Press.

Britain, David. 2002. "Diffusion, levelling, simplification and reallocation in past tense be in the English Fens". Journal of Sociolinguistics 6(1): 16-43.

Forsström, Gösta. 1948. The Verb 'To Be' in Middle English: A Survey of the Forms. Lund: Gleerup.

Furnivall, Frederick J., ed. 1897. Child-Marriages, Divorces, and Ratifications, &c. in the Diocese of Chester, A.D. 1561-6. (= Early English Text Society, Original Series, 108.) London: Early English Text Society.

Görlach, Manfred. 1999. "Regional and social variation". The Cambridge History of the English Language. Vol. III: 1476-1776, ed. by Roger Lass, 459-538. Cambridge: Cambridge University Press.

Gowing, Laura. 1996. Domestic Dangers. Women, Words, and Sex in Early Modern London. Oxford: Clarendon Press.

Grund, Peter. 2006. "Manuscripts as sources for linguistic research: A methodological case study based on the Mirror of Lights". Journal of English Linguistics 34(2): 105-125.

Grund, Peter. 2007. "From tongue to text: The transmission of the Salem witchcraft examination records". American Speech 82(2): 119-150.

Grund, Peter, Merja Kytö & Matti Rissanen. 2004. "Editing the Salem witchcraft records: An exploration of a linguistic treasury". American Speech 79: 146-166.

Kytö, Merja, comp. 1996. Manual to the Diachronic Part of the Helsinki Corpus of English Texts. Coding Conventions and Lists of Source Texts. Helsinki: Department of English, University of Helsinki.

Kytö, Merja & Terry Walker. 2003. "The linguistic study of Early Modern English speech-related texts: How 'bad' can 'bad' data be?". Journal of English Linguistics 31(3): 221-248.

Kytö, Merja & Terry Walker. 2006. Guide to A Corpus of English Dialogues 1560-1760. (= Studia Anglistica Upsaliensia, 130.) Uppsala: Acta Universitatis Upsaliensis.

Kytö, Merja, Terry Walker & Peter Grund. 2007. "English witness depositions 1560-1760: An electronic text edition". ICAME Journal 31: 65-85.

LALME = McIntosh, Angus, M. L. Samuels & Michael Benskin. 1986. A Linguistic Atlas of Late Mediaeval English. 4 vols. Aberdeen: Aberdeen University Press.

Lass, Roger. 1999. "Phonology and morphology". The Cambridge History of the English Language. Vol. III: 1476-1776, ed. by Roger Lass, 56-186. Cambridge: Cambridge University Press.

Lass, Roger. 2004. "Ut custodiant litteras: Editions, corpora and witnesshood". Methods and Data in English Historical Dialectology (= Studies in Language and Communication, 16), ed. by Marina Dossena & Roger Lass, 21-48. Bern: Peter Lang.

Nevalainen, Terttu. 2006. "Vernacular universals? The case of plural was in Early Modern English". Types of Variation: Diachronic, Dialectal and Typological Interfaces (= Studies in Language Companion Series, 76), ed. by Terttu Nevalainen, Juhani Klemola & Mikko Laitinen, 351-369. Amsterdam & Philadelphia: John Benjamins.

Nevalainen, Terttu & Helena Raumolin-Brunberg. 1994. "Its strength and the beauty of it: The standardization of the third person neuter possessive in Early Modern English". Towards a Standard English 1600-1800, ed. by Dieter Stein & Ingrid Tieken-Boon van Ostade, 171-216. Berlin & New York: Mouton de Gruyter.

Nevalainen, Terttu & Helena Raumolin-Brunberg. 2003. Historical Sociolinguistics: Language Change in Tudor and Stuart England. (= Longman Linguistics Library.) London: Pearson Education.

Pietsch, Lukas. 2005. " 'Some do and some doesn't': Verbal concord variation in the north of the British Isles". A Comparative Grammar of British English Dialects: Agreement, Gender, Relative Clauses (= Topics in English Linguistics, 50.1), ed. by Bernd Kortmann, Tanja Herrmann, Lukas Pietsch & Susanne Wagner, 125-209. Berlin & New York: Mouton de Gruyter.

Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech & Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London: Longman.

Rosenthal, Bernard, Gretchen A. Adams, Margo Burns, Peter Grund, Risto Hiltunen, Leena Kahlas-Tarkka, Merja Kytö, Matti Peikola, Benjamin E. Ray, Matti Rissanen & Richard B. Trask, eds. Forthcoming. The Records of the Salem Witch-Hunt. Cambridge: Cambridge University Press.

Rye, Walter, ed. 1905. Depositions Taken Before the Mayor and Aldermen of Norwich 1549-1567. Norwich: The Norfolk and Norwich Archaeological Society.

Sharpe, J. A. 1980. Defamation and Sexual Slander in Early Modern England: The Church Courts at York. (= Borthwick Papers, 58.) York: Borthwick Institute for Historical Research.

Tagliamonte, Sali. 1998. "Was/were variation across the generations: View from the city of York". Language Variation and Change 10: 153-191.

Visser, F. Th. 1963. An Historical Syntax of the English Language. Part 1: Syntactical Units with One Verb. Leiden: Brill.

Wakelin, Martyn F. 1972. English Dialects: An Introduction. London: Athlone Press.

Walker, Terry. 2007. Thou and You in Early Modern English Dialogues: Trials, Depositions, and Drama Comedy. (= Pragmatics and Beyond New Series, 158.) Amsterdam & Philadelphia: John Benjamins.