Jump to ContentJump to Main Navigation
Show Summary Details
More options …

Journal of Historical Sociolinguistics

Ed. by Rutten, Gijsbert / Auer, Anita / del Valle, José / Vosters, Rik / Pickl, Simon

2 Issues per year

See all formats and pricing
More options …

‘The vsuall speach of the Court’? Investigating language change in the Tudor family network (1544–1556)

Mel Evans
Published Online: 2015-08-28 | DOI: https://doi.org/10.1515/jhsl-2015-0011


This paper considers how micro-level analysis can enrich our understanding of macro-level processes of language change, using a case study of the Tudors. It explores how language use in the Tudor family network relates to the role of the Court in the supralocalisation of innovative forms during the sixteenth century. Using an original corpus of correspondence and other autograph writings, I conduct a comparative analysis of the language of Elizabeth Tudor with her siblings, parents and caregivers. The findings suggest that Elizabeth’s siblings, Mary I and Edward VI, were progressive in changes localised at the Court, but that Elizabeth’s caregivers and peripheral kin may have influenced Elizabeth’s uptake of non-Court-based changes. Using Network Strength Scores to represent the social experiences of Elizabeth and her nearest kin, it appears that Elizabeth’s changing position within the Court network, from a peripheral to more central member, may have played a part in the Court’s catalyst effect for the supralocalisation of innovative forms, and the emergence of an overtly prestigious “norm” in Early Modern English.

Keywords: family language; social network analysis; morphosyntactic variation and change; correspondence; the Tudors

1 Introduction

In The Art of English Poesie, George Puttenham famously advises his readers to aspire to the example of “the vsuall speach of the Court […] and not much aboue” ([1597] 2007: 101). He differentiates the Court from the language varieties used further afield, beyond “the riuer of Trent” or the dialects of the “far” West (2007: 101). The Court, he suggests, write “good Southerne” (2007: 102). Puttenham’s favourable depiction of Courtly language likely reflects, in part, his own ambitions in this social sphere; his appraisal of Courtly English is matched by an effusive dedication to Sir William Cecil, and hyperbolic adulation of Queen Elizabeth I (for a discussion of Puttenham’s life and works, see May 2008). However, Puttenham’s presentation of the sixteenth-century Court as a sociolinguistic domicile, whatever his motivations, appears to have an empirical basis. Nevalainen and Raumolin-Brunberg (2003) suggest that the Court is a catalyst for the adoption of innovative variants. While not necessarily the source of innovation, the Court appears to propagate a change, presumably through the prestige and power of its speakers who socially “endorse” the new forms, which subsequently supralocalise (Conde-Silvestre 2012: 343).

This paper examines the Court’s catalytic role from a micro-level perspective, in order to better understand the structures and individual practices underlying the macro-level trend. One can hypothesise how the networks of members of the Court (educated, geographically mobile) might permit weak ties that foster language change (Milroy 1980; cf. Granovetter 1973). The present study tests this hypothesis, and focuses on the individuals that constitute the Court’s (ideological) core: the Tudor royal family. It looks for possible correlates between the social experiences and networks of the high-status, well-educated Tudor kith and kin, and their language, and reflects on the implications for our understanding of “Courtly language” in sixteenth-century language change in English.

2 The Court: macro- and micro-perspectives

In their analysis of language change in the Corpus of Early English Correspondence, 1 Nevalainen and Raumolin-Brunberg note that “[a] Court advantage proved the single most common pattern in the data”, evident “until the change was completed or nearing completion” (2003: 182). The picture of the Court as a normalising, endorsing force could suggest linguistic homogeneity within what was a relatively small domicile. However, previous micro-level studies show no clear evidence of a unifying Court “standard”. For example, Nevalainen (2002, 2013) compares the linguistic preferences of Henry VIII to that of his scribes, contrasting the king’s holograph letters (love-letters to Anne Boleyn) with official documents issued in Henry’s name (often with a wooden stamp of his signature) (Daybell 2012: 97). In addition to features associated with epistolary style – such as linguistic traces of medieval ars dictaminis in the secretarial letters, including present participles and doublets (Nevalainen 2013: 108; Davis 1967) – Nevalainen finds that Henry’s administrators were generally conservative in their use of morphosyntactic variants. By comparison, Henry VIII shows a mixed behaviour: conservative in some changes, progressive in others, at least when compared with his administrators.

Nevalainen’s analysis of Henry VIII and six of his administrators points to the linguistic heterogeneity within the Tudor court at the start of the sixteenth century. A division might be made on linguistic lines between the King and his professional scribes. Although Nevalainen’s study considers the language of several secretaries, she does not engage with the language of Henry’s kin. This part of the Court requires further examination, particularly since Henry’s language use was generally more progressive than his courtiers. Is the king, as a central member of the Court, a high-status speaker who could “endorse” innovative forms? Is this the situation for other royal members of the Court? Exploring the Tudor family may potentially shed light on the processes underlying the Court’s catalytic macro-level role.

3 The historical family

Studies of the family in contemporary sociolinguistics generally focus on processes of linguistic transmission (see Hazen 2003). A repeat finding is that parents and caregivers have linguistic influence in early years of childhood, after which peers dominate (Kirkham and Moore 2013; see also Eckert 1988, Eckert 2000; Kerswill 1996; Labov 2001; Tagliamonte and D’Arcy 1999; but see Bugge (2010), who finds dialect vocabulary is influenced by family). A speaker’s language is also influenced by their family orientation. Siblings, even when demographically similar, can have very different linguistic profiles throughout their lifetime, aligning towards family or peer-group “for reasons of identity (or non-identity) with other members of the family” (Hazen 2003: 505; see also Kerswill and Williams 2000: 75).

Historical sociolinguistic studies of the family are not (easily) able to investigate linguistic variation from the perspective of transmission, because records are restricted to adolescent (and older) language. Indeed, the parent-child relationship in Early Modern England is, paradoxically, best understood after the child leaves home, because distance entails written communication (Houlbrooke 1984: 178). 2 More propitiously, however, many English manuscript collections are family collections and permit intra- and inter-generational linguistic analyses (e.g. Kallel 2002; Nevalainen 2002, Nevalainen 2009). In the Johnson family corpus, Raumolin-Brunberg (1996) identifies linguistic similarities between Anthony Cave (the senior caregiver), and brothers John and Richard. However, brother-in-law Ambrose Saunders shows atypical linguistic preferences that “shed doubts on [a transmission] hypothesis” (Raumolin-Brunberg 1996: 99). One interpretation of this mixture of homo- and heterogeneity is that Ambrose’s orientation towards the household unit differed to his peers (Kerswill and Williams 2000; see also Bergs 2000, Bergs 2005, discussed below).

4 Social network analysis

Social network analysis (SNA) has proved a productive approach for exploring language change in the family (e.g. Tieken-Boon van Ostade 2005; Sairio 2009). SNA in sociolinguistics is best known for the investigation of linguistic variation in Belfast (Milroy 1980). Since then, the “set of procedures” (Milroy 1980: 46) for SNA has improved our understanding of how individuals’ social relationships influence processes of language change. In brief, SNA models the structural and content components of social groups, looking at the connections between individuals (ties), the extent to which those individuals know each other (density) and the ways in which those individuals interact (multiplexity). Different network configurations promote different kinds of linguistic behaviour. Close-knit networks (high density and multiplexity) are norm-enforcing, resisting linguistic innovation (e.g. the preservation of local non-standard norms). Conversely, loose-knit networks (uniplex ties and low density), are more receptive to innovation, and promote the diffusion of linguistic variants within and across communities (Bergs 2006: 2–3; Conde-Silvestre 2012: 333–334). The family typically constitutes a “cluster in a total network” (Bergs 2000: 245), a denser pocket of network members, which has implications for language use.

The data limitations of historical research present challenges for the application of SNA. Bergs argues that framework adjustments to negotiate such challenges are permissible, so long as these are explicit (2005: 24). He cautions that a distinction needs to be made between structural (density, centrality and clusters) and content (multiplexity, transactional content, reciprocity) components (Bergs 2000: 240). The former can be objectively documented. The latter require greater speculation.

Bergs’ (2000, 2005) study of the Paston network, a fifteenth-century Norwich family, considers primarily structural components, but “evolves into an ego-centric model” (Conde-Silvestre 2012: 340). Bergs looks for evidence of progressive or conservative language use by each family member, and considers the possible correlation with their network position. To overcome inconsistencies in the dataset, Bergs models a posthumous, static network of the Pastons. This strategy allows for a greater quantity of linguistic data to be used, but struggles to capture the dynamics of language use in changing network contexts.

His analysis finds no ‘uniform correlational pattern’ between NSS and linguistic usage: no speaker is wholly innovative or wholly conservative (Bergs 2005: 254; cf. Nevalainen et al. 2011). He attributes this finding to methodological constraints (i.e. changing network structures), speaker performance differences, and the varying salience of variables for each individual. That said, he finds that some speakers tend towards innovation, whereas others do not. Their networks provide an explanation. In the Paston family, loose-knit structures do not promote conformity to a supralocal standard because no overt standard existed. Instead, loose networks promote innovation and diversity. Closer-knit groups were norm-enforcing, but for a local norm not shared by other close-knit networks (Bergs 2005: 54–55). Thus, whilst the forces of change are the same, the linguistic consequences are different, reflecting the contemporary macro-level linguistic landscape. On this basis, Bergs hypothesises that the transition from the medieval system of local language norms to that of an overt standard came about some point “before about 1500, or 1600” (Bergs 2012: 96). The evidence of the role of the Court in fostering sixteenth-century supralocalisation of variants supports his hypothesis at a macro-level. Closer analysis of individuals within the Court network, with consideration of their social relationships and identities, may offer further insight into how, and when, this transition came about.

Bergs’ study also finds network structures correlate with sibling language differences. John II is more innovative in his linguistic preferences than his brother, John III. John II was “a travelling bon vivant”, entailing loose-knit networks, whereas John III was more family oriented, “bound up in a multiplex, tight network” (Bergs 2000: 244–245). The Paston brothers’ language is comparable with contemporary sociolinguistic findings (e.g. Kerswill and Williams 2000) and the Johnson corpus (Raumolin-Brunberg 1996). However, this picture is complicated when viewed dynamically. For example, John III moved to London in 1467, after which several innovative features emerge in his letters (not found in John II’s). Conde-Silvestre (2012: 343) suggests that John III’s contact with prestigious individuals and leaders of orthographic change, such as lawyers, is one explanation that requires further investigation. The Tudor siblings offer a comparable, if more complex, example for sibling language analysis.

5 The Tudor family

The royal Tudor dynasty spans over one hundred years (1485–1603), encompassing three generations and five monarchs. Their social status, as heirs, kings and queens, means their writings are relatively well preserved. Yet so far only the language of Henry VIII and his youngest daughter, Elizabeth I, have been explored from a sociolinguistic perspective, albeit with interesting results. Henry VIII’s morphosyntax and spelling show a mixture of innovative and conservative usage, with marked differences to the king’s administrative court-based subjects (Nevalainen 2002, Nevalainen 2013). Elizabeth is more consistently progressive: a linguistic leader of morphosyntactic change (discussed in Section 6), as well as showing innovative practice in spelling (see Raumolin-Brunberg 2005; Nevalainen et al. 2011; Evans 2013). These two Tudors, father and daughter, king and queen, therefore display a general tendency towards progressive linguistic practices, adopting what would become the supralocal norm.

Interpretations of their progressiveness have drawn on demographic and psycho-biographical characteristics. Nevalainen et al. (2011) propose that Elizabeth’s gender and social rank provide relevant corollaries, as Elizabeth participates in both female-led (e.g. replacement of pronoun ye by you) and male-led (e.g. the use of single negation) changes. For the latter, the authors posit that Elizabeth’s elite status, which afforded her an education typical of men, as well as participation in male spheres of literacy, enabled her participation in the early stages of the change. Henry’s social rank likewise promotes participation in changes linked to written, legal domains. More speculatively, a psycho-biographical approach might link Elizabeth’s linguistic non-conformity to her social non-conformity, including (non-feminine) behaviour such as hunting on horseback and her penchant for swearing (Hughes 2006: 54). Her sensitivity to linguistic innovation could arise from her social precariousness; comparable, perhaps, to the middle-class tendency towards hypercorrection in contemporary sociolinguistic studies (see Labov’s 2001: 410 profile of linguistic leader Celeste S).

Henry and Elizabeth are only two members of a family network, conceivable as a cluster – “segments or compartments of networks” built around “ties of kinship, occupation, specific group membership” (Milroy 1980: 50) – in the larger Court network. Clusters often have strong norm-enforcement mechanisms, due to the density of ties, relative to other nodes in a network. To better understand the salience of family network ties for language use at the Tudor court, we need to consider the language of more members of that family, and situate it within its macro- and micro-level contexts.

For this exploratory study, the family network examined is deliberately small. Examining the Tudor dynasty and connections across the sixteenth century would be unwieldy and likely result in inconsistent data and results. This analysis therefore opts for a narrower temporal “snapshot” from 1528 to 1556, with the bulk of the data from 1544–1556. The experiences of the Tudor family are well documented for this period, which encompasses the end of Henry VIII’s reign (d. 1548) including his final marriage to Katherine Parr; the reign of Edward VI (b. 1537, r. 1548, d. 1552); and Mary I (b. 1515, r. 1552, d. 1558). This time-period is formative for Elizabeth, the youngest member of the Tudor family (b. 1533), who maintained her connection to the Court domicile throughout this period and into the latter half of the sixteenth century, when many of the linguistic changes started to supralocalise. As Elizabeth is the only family member to survive into the second half of the century, she is used as the network focus, or ego.

The network snapshot comprises ten informants, including Elizabeth (see Appendix, Table 13). Each can be categorised in terms of their relationship with Elizabeth: nuclear family, caregivers, and peripheral family. Nuclear family “determines [one’s] place in the social order” (Costor 2001: 6), and constitute one’s closest kin. This includes Elizabeth’s father Henry VIII, her step-mother Kathryn Parr, her older half-sister, Mary I and younger half-brother, Edward VI. Edward was a few years younger than Elizabeth, but they shared caregivers and tutors. Mary, on the other hand, was seventeen years older, although the sisters did share the experience of being female heirs. For another perspective on this generation, Edward’s childhood friend, Barnaby Fitzpatrick, is also included. The second category includes caregivers and educators. Arguably, these individuals had more of a “family” role than her blood-kin, in terms of day-to-day interaction during Elizabeth’s adolescence (Borman 2009). Two long-serving household servants, for whom we have some autograph data, are included: Kat Ashley, governess, and Thomas Parry, cofferer. Also included is Roger Ascham, a Yorkshireman, with a university education, who was Elizabeth’s (and probably Edward’s) tutor. I am unaware of extant linguistic data for other caregivers or educators, although better representation of this part of Elizabeth’s household is desirable. Finally, one member of the extended family, Edward VI’s uncle, Thomas Seymour is included. Seymour, a socially ambitious man, had significant socio-political status in the 1540s and 1550s. At a personal level, Seymour became Elizabeth’s step-father (of sorts) when he married Katherine Parr in 1547, and was subsequently a suitor for Elizabeth’s hand in 1548. He was executed at Edward VI’s command in 1549 (see Alford 2002).

The corpus contains transcripts of autograph ego-documents dated to the timeframe 1528–1556. One of the challenges of historical micro-level analysis is the unevenness of the data: Edward VI’s corpus contains over 40,000 words, Kat Ashley’s comprises 885 words. Reconciling these differences, and the linguistic evidence contained within, in a reliable way is difficult. However, small datasets can at least provide positive evidence of an informant’s grammar, as well as illustrating larger trends (Bergs 2000: 244). 3 The documents representing each informant are primarily letters, but also include diaries, religious prose and witness depositions. Ego-documents have played a key role in historical sociolinguistic research. However, in the Tudor family corpus, the “authentic immediacy” (Elspaß 2012: 159) associated with ego-documents is not necessarily to be relied upon, given the informants’ high literacy levels and familiarity with rhetorical tropes and figures (on the analytic potential of ego-documents for historical sociolinguistics see Elspaß 2012). Early Modern letter-writing was a “social-textual” (Daybell 2015: 503) practice, with its own conventions (e.g. Davis 1967; Whigham 1981; Daybell 2012, Daybell 2015). Social relationships were encoded through address forms, rhetorical tropes, stylistic devices (e.g. doublets, participles), as well as materiality and layout. As Daybell (forthcoming) observes of mother and daughter correspondence, the cultural scripts of kinship deference are evidence across letters of the period, “undergirded by the precepts that governed children’s demeanour towards parents”. He posits it is the “nuances and representation of scripts that is telling”, such as address-forms, subscriptions and framing of requests (Daybell forthcoming).

Correspondence with the sovereign and their kith and kin was similarly – and perhaps even more explicitly – shaped by cultural scripts of deference and social hierarchy (Whigham 1981). As such, the corpus correspondence can offer evidence of how each member of the Tudor family conceived their relationship with others in the network, within the macro-level conventions of epistolary expression. This will help to corroborate the relationship categories outlined above. As Taavitsainen and Jucker (2008: 207–208) demonstrate in their analysis of compliments in the history of English, qualitative analysis can rediscover the labels and conceptualisations of social and linguistic phenomena used by speakers in the past. As letter subscriptions are particularly significant for the construction of familial ties, being the part of the letter which best shows “the image the writer has of him/herself in the relationship with the recipient” (Nevala 2004: 128), these are the focus, identifying the address forms and third-party reference therein.

The subscriptions in the Tudor Family corpus show a collective commitment to signalling the writer’s familial tie as well as their social connection with the recipient. In letters exchanged between nuclear and peripheral family, the writers share conventions in specifying their blood relationship as well as their social status; thus Mary subscribes her letter to Edward VI: “your majesty’s most humble sister”, and Elizabeth signs a letter to Mary I as “Your Majesty’s obedient subject and humble sister”. Elizabeth’s adolescent correspondence is typical of the use of address and third-party reference seen in the corpus, adhering carefully to socially deferential scripts. 4 For example (spelling standardized):


Master Denny and my Lady with humble thanks prayeth most entirely for your grace praying the almighty God to send you a most lucky deliverance, And my mistress [Ashley] wisheth no less giving your Highness most humble thanks for her commendations (Elizabeth I to Parr, 1548).

Example 1 suggests the correspondents had sufficient shared knowledge to recognise to whom “Lady” and “mistress” referred; a trace, perhaps, of the intimacy between Elizabeth and her stepmother.

Third-party references illustrate how writers conceived their relationships with others; or, how they perceived they needed to depict that relationship, within norms of epistolary writing. Elizabeth’s caregivers, Ashley and Parry, refer consistently to their royal charges using official titles. Elizabeth is “her grace” and “my lady’s grace”. There is no trace of positive politeness that may indicate intimacy or affection of these individuals towards the royal children. In Ascham’s correspondence, Elizabeth is “my mistress” and “my lady” and Edward “my prince”. There is therefore a linguistic distinction in address and third-party reference between blood relations, on the one hand, and caregivers and tutors on the other. Viewed collectively, the letters show that epistolary address conventions had currency in the Tudor network. 5 They also support the structural differentiation between family and caregivers/educators in the posited network.

6 Morphosyntactic variation

Having established the broad structures of the Tudor family network using primary and secondary evidence, I now examine the linguistic progressiveness or conservatism of each informant for eight morphosyntactic changes. I consider the individual preferences in relation to fellow network members, as well as with reference to the macro-level demographic trends, as identified by Nevalainen and Raumolin-Brunberg (2002, 2003).

6.1 Replacement of subject ye by you

In Middle English, ye (example 2) was the subject form and you (example 3) the object. However, you started to generalise to the subject position from the fifteenth century; Bergs (2005: 129) reports that subject-you is found in the third generation of Paston letter-writers, born after 1420. In the sixteenth century, the rate of the change accelerated and completed around 1600. Ye survived in more formal registers, such as religious prose (Lass 1999: 154). A likely contributing factor was the presence of a weakened (unstressed) form of you, which plausibly created ambiguity surrounding form and case distinctions in speach, and forced writers to make a binary selection of form in writing (Lass 1999: 154; Bergs 2005: 121).


but ye should have much ado (Parr, 1548).


When you be at leisure (Parr, 1547).

Socially, the generalisation of you emerges in the middle ranks, led by female informants, with links to the Court and London. CEEC suggests that there was some social awareness – possibly negative – of the new variant, as social aspirers, known to be linguistically sensitive to stigmatized forms, lag in its uptake (Nevalainen and Raumolin-Brunberg 2003: 152).

In the Tudor family data (Table 1), Elizabeth is a categorical user of you (tokens=37), and Edward and Mary use you over 90% of the time. Mary, given her age, is arguably more innovative than Elizabeth in this regard. There is thus no indication of social stigma in this generation of the Tudor family, nor any gender differentiation. It seems likely that, just as Bergs (2005: 126) identified the youngest generation of the Pastons as the most progressive speakers in this change in the fifteenth century, the Tudor siblings also show generational patterns of language change.

Table 1:

Frequency of you in Tudor family corpus.

The data for the older generations supports the macro-level female-led trend (e.g. Parr uses you 45%; total tokens=46). Henry VIII is an exceptional male, who as a 100% user of you is considerably ahead of the curve (Nevalainen 2013). Overall, the generalisation of you appears to be an “inclusive” change found across the Tudor family network, excepting Seymour (18.7%; tokens=32). This may reflect the mid-range stage of the change in the 1540s (Nevalainen and Raumolin-Brunberg 2003: 60). But it could also reflect the network norms of the Tudor family, versus others at the Court. The scribal correspondence of Henry and Elizabeth, for instance, is distinctive for the presence of ye, compared to the royal autograph which uses you (near) categorically. It may be that one factor is stylistic, with the formal, scribal documents of the sovereign classifiable as part of the “special register” that preserves ye (Lass 1999: 154). This raises an interesting conceptual difference between the actual language use of the Tudor monarch, and the language used to symbolise their authority. At a network level, the comparability within the Tudor nuclear family members may support a case for the linguistic influence of this family “cluster” constructing a local, familial norm.

6.2 Negative declarative do

The do+not+verb (e.g. He did not run) construction is first attested in the fourteenth century, and gains ground over the existing post-verbal not construction (e.g. He ran not) in the subsequent three centuries (Rissanen 1999: 270–271). 6 The rise of negative do, as part of the emergence of do periphrasis in English, has been subject to extensive scrutiny (e.g. Ellegård 1953; Kallel 2002; Warner 2005; Culpeper and Kytö 2010). A range of internal and external factors appears to shape its development. Systemically, a set of high-frequency verbs known as the know group (know, boot, care, doubt, mistake, trow, fear, skill and list), resist do when compared to other verb forms (Ellegård 1953; Nurmi 1999), although the small corpus size entails that this factor is difficult to account for in the following discussion.

Socially, negative do was a change from below, linked to London and the Court (Nurmi 1999: 178), and seems a communal, rather than generational change in the sixteenth century (Warner 2005: 270–271). The distribution also shows a “dip” in the late sixteenth century (e.g. Ellegård 1953; Nurmi 1999: 146). Warner argues this is due to the stylistic re-evaluation of the construction, with negative do falling out of favour in more formal, literary and “public” texts (Warner 2005: 271). The Court plays a possible role here, too, this time linked to the arrival of the Scottish King, James VI (James I). Scots English did not use do periphrasis, and thus the prestigious status of the new Scots personnel is a possible external trigger for this stylistic change (Nurmi 1999). The Tudor family corpus captures an earlier stage in the change, when the variant was stylistically associated with more formal text-types. This is notable, given that such texts were the preserve of the literate.

Table 2 shows that Elizabeth and her siblings use a similar frequency of negative do, which is comparable with the macro-level average (c. 20%) (Nurmi 1999: 166). In fact, Elizabeth’s use of do lags slightly behind the expected frequency, in contrast to Edward and Mary (p>0.05). Comparison with older family members suggests, as with the replacement of ye by you, that generational factors are relevant. However, the older users of negative do are the inverse of ye/you. Elizabeth’s male kin, caregivers and educators, Thomas Seymour, Thomas Parry and Roger Ascham, use the form. Ascham is particularly progressive. Although originally from the north of England, his social status (a university man, and social climber) makes it more plausible that his progressive usage arises from positive social associations with the form, as social aspirers show a greater sensitivity to linguistic innovations (Nevalainen and Raumolin-Brunberg 2003: 32–37). Ascham’s social profile supports the stylistic data that situates do within more formal text-types during this period.

Table 2:

Frequency of negative do in Tudor family corpus.

By comparison, neither Henry VIII nor Kathryn Parr use the form, although the lack of evidence cannot be taken as proof they never used the form, of course. Although there are few valid tokens for Henry, its absence from Parr’s writing (12 tokens), including her religious prose, is more striking, given the formal stylistic context. By comparison, Mary, two years younger than Parr, uses do frequently. One possible explanatory factor may be educational differences, and the consequences this had on their networks. Mary received a high level, innovative education from men affiliated with the universities; men who also interacted with do users like Ascham and Seymour. Parr, on the other hand, received a more moderate education in her youth (Somerset 1991: 12–13, 15–16).

The informant data from the Tudor family corpus from the early to mid-sixteenth century therefore suggests that negative do may have been a male-led change in this local environment, and also supports London and the Court as the regional source. At the micro-level, the evidence from these two changes suggests that Mary was as innovative, if not more so, than her younger sister, and that negative do was establishing ground within the youngest generation of the Tudor family.

6.3 Positive declarative do

Positive declaratives represent the only syntactic context in which do failed to generalise in English, not reaching beyond the incipient level of the change (c. 10%) (Rissanen 1999; see example 4, below). The form’s emergence has been attributed to multiple factors. Internally, auxiliary do underwent partial grammaticalisation (Wischer 2008: 141), and it is plausible that positive declarative do developed through analogy with negative and interrogative contexts. Functional motives may also apply, as do disambiguates verb phrases in complex structures (e.g. intervening adverbs; subordinate hypotheticals) (Wischer 2008: 145–146). Do also seems to have specific functions within different text-types. Whilst disambiguating functions apply to more formal writing, more speach-like texts show a “clustering” of do that may have had a foregrounding function for topic or narrative “peaks” (Rissanen 1999: 241; also Rissanen 1991). From an external perspective, region and gender appear to have been significant factors in the rise and fall of the variant. Regionally, do appears to have been prominent at the Court, at least by the end of the sixteenth century (Nevalainen and Raumolin-Brunberg 2003: 175), after which it declined. Men use the form more than women (Nevalainen and Raumolin-Brunberg 2003: 125). In light of the functionality of do in more formal texts, it may be that the greater proportion of men with high levels of education, and professions requiring written language, underpins this trend.

Table 3:

Frequency of positive do per 1,000 words in Tudor family corpus.

The Tudor family corpus shows a general preference for positive declarative do (Table 3). Elizabeth uses do 5.3/1,000 words (33 tokens), higher than the CEEC (2.5) average for the period 1540–1559 (Nurmi 1999: 166). 7 Mary’s correspondence contains a near-identical frequency (5.1; 43 tokens). This contrasts with 1.6 (78 tokens) for Edward, which is perhaps surprising given the male (and Court) bias in CEEC. However, the functions of do may help to explain Edward’s low frequency. The primary data for the Mary and Elizabeth is correspondence. This is deferential and petitionary, drawing on common tropes of supplication (Whigham 1981) that are constructed using complex structures associated with do disambiguation, as well as a stylistic intensity that could promote foregrounding do (Rissanen 1999: 241), e.g.:


Good Mr Secretary, I do thank you with all my heart for the great pain and suit you have had for me. For the which I think myself very much bound to you. And whereas I do perceive by your letters that you do mislike mine exception in my letter to the King’s Grace, I assure you, I did not mean as you do take it. For I do not mistrust, that the king’s goodness will move me to any thing which should offend God and my conscience. But that which I did write was only by the reason of continual custom (Mary to Cromwell, 1536).

By comparison, a letter from Edward to his stepmother works through similar filial tropes, although with less intensity, befitting their school-room status (example 5). Unfortunately, the low number (3) of tokens in Edward’s correspondence does not allow for a reliable assessment of his usage, although the normalised frequency (5.5) is comparable with his siblings.


I have me most humbly recommended unto your grace, with like thanks, both for that your grace did accept so gently my simple and rude letters (Edward VI to Parr, 1546).

These epistolary differences are indicative of the different social relationships each sibling had with their elders at the Court. The socio-textual significance of epistolary prose entails that these relationships required different kinds of letter and, therefore, different stylistic resources for composition. Positive do is rarer in Edward’s chronicle (1.8/1,000 words). This is surprising, given that Wischer (2008) finds do is more frequent in formal text-types in the Helsinki Corpus.

The distribution of positive do in the language of other network members is heterogeneous, although token counts are small in some cases and must be treated with caution. Henry’s correspondence contains seven tokens of do, which gives the highest normalised frequency (4.8) of his generation. Henry’s corpus comprises eight love letters to Anne Boleyn (avg. 130 words), and 2 letters to Cardinal Wolsey (300 words in total). Yet, despite the intensity of his letters to Anne, there is only one token of do, compared to six in the letters to Wolsey.


according to your desire, I do send you mine opinion by this bearer, the reformation whereof I do remit to you and the remnant of our trusty counsellors, which I am sure will substantially look on it. As touching the matter that Sir William Says brought answer of, I am well contented with what order so ever you do take in it (PCEEC, Henry VIII, ORIGIN1_041).

Although the figures are small, the correspondence to Wolsey (example 6) adheres more to formal epistolary style, such as present participles, and wh- adverbials, than Henry’s intimate correspondence with Anne. This may suggest a Court norm in which do was an appropriate marker of formal epistolary writing at this time. Henry’s letters to Wolsey are the earliest in the corpus (dated to the 1520s), suggesting Henry in particular was an early adopter. There is a possible relationship between nuclear family networks and this form, bound up with epistolary conventions and the kinds of letter represented in each informant’s corpus.

6.4 Multiple and single negation

The use of non-assertive forms for sentential negation (example 7) originates in legal texts in late Middle English (Rissanen 2000), supplanting the use of multiple negators (example 8), possibly for reasons of greater semantic precision (Nevalainen 2006: 259; Kallel 2007). Single negation lags in coordinate constructions, especially those with semi-idiomatic expressions such as nor never (Rissanen 1999: 271).


there is no mention made of any plate ornament ready money or of any jewels (Parry, 1550).


never would I speak no wise of marriage (Ashley, 1549).

The legal origin of single negation corresponds with the social trajectory of the change, with upper-ranking educated men the leading group in CEEC (46% in 1520–1559; Nevalainen and Raumolin-Brunberg 2003: 151). Social aspirers show an early uptake of the new construction, suggesting prestigious associations. Many were situated at the Court. Relatedly, women lag behind men in this change throughout the sixteenth century, which can be similarly linked to the professional provenance (Nevalainen 2000: 50).

The macro-level picture is paralleled by the local distribution in the Tudor family network (Table 4). Elizabeth’s near-absolute use of single constructions (93%) is shared with her brother (88%) and her sister (78%), and also most of the older generation. Unfortunately, the token counts for this variable are low and not statistically significant, curtailing interpretations. However, Kathryn Parr (31%, 16 tokens) and Kat Ashley (29%, 7 tokens) use single constructions less frequently than other informants, which may reflect their experiences as moderately educated women rather than professional men. By comparison, both Mary and Elizabeth use the form, and thus participate in a male-led change. Mary and Elizabeth’s education – which involved connections to men based at the universities and legal spheres – plausibly affords links that permit the adoption of the legal innovation.

Table 4:

Frequency of single negation in Tudor family corpus.

Overall, in these four changes the Tudor siblings show evidence of homogeneity in their language use. Despite the fact that the macro-level profiles of these changes are mixed (one from above, three from below, three are male-led, one is female-led), all, it should be noted, were, or would become, fully established at the sixteenth-century Court. This may suggest that, more locally, these variants operated as Courtly norms, with the kinship ties between siblings, as well as the networks arising from their similar and shared experiences in education, promoting shared preferences. For comparison, I now discuss four changes that show greater diversity across the family network.

6.5 Relativization: animate who

Relative marker who is a late development in English relativization, first attested in the Paston letters in 1426 (Rissanen 1999: 294; Bergs 2005: 143; see example 10). During the Early Modern period, who supplanted which (example 9) as the wh- marker for animate subject-position antecedents. Studies suggest who diffused along a cline of referents of semantic salience, from deity, to nobility, intimates and other animates (Nevalainen and Raumolin-Brunberg 2002; Bergs 2005: 147). Stylistically, who emerges in more formal text-types, including subscription formulae in letters (Bergs 2005: 158), modelled on Latin epistolary expression (Rissanen 1999: 293). Socially, CEEC reveals a socially stratified change in the first half of the sixteenth century, with social aspirers (78%) surpassing the middle ranks (74%) and upper ranks (49%) (Nevalainen and Raumolin-Brunberg 2002: 112).

To what extent are these macro-level patterns mirrored in the Tudor family network? Elizabeth and Edward show a preference for animate who, although Elizabeth’s correspondence contains only two relatives with animate antecedents. Edward uses who for 78% (122 total tokens) animate antecedents, and the majority occur in his journal, showing that the form was not restricted to epistolary formulae. By contrast, Mary, however, uses who 57% (total tokens=23) of the time (p>0.05) (Table 5). This is surprising given Mary’s progressiveness in other changes. Edward’s peer, Barnaby Fitzpatrick, who was a core member in Edward’s educatory network, also uses who for animate antecedents (89%; total tokens 11), including contexts outside of closing formulae. If we take Elizabeth’s use of who to be similarly progressive, the main demographic difference that might explain Mary’s conservatism is age (generational change).

Table 5:

Frequency of subject who (%) in Tudor family corpus.

This hypothesis finds further support when the use of who by older network members is considered. In the language of the educated older men, who is infrequent. Roger Ascham’s correspondence contains 10.5% (2 out of 29 tokens) animate who, one of which occurs in closing formula. This suggests, contrary to the macro-level trends, that Elizabeth’s tutor did not attach much social significance to the marking of animacy. This is surprising given the propensity of social aspirers to use the form in CEEC (Nevalainen and Raumolin-Brunberg 2002: 110); a category in which Ascham can be placed. Even more notable is that Henry VIII does not use who at all. Although there are only nine tokens, this includes closing expressions, where one might assume that an indirect reference to oneself, as the monarch, would warrant animacy marking if the writer considered it polite or decorous (example 9). This does not appear to be the case.


Written with the hand of him which desireth as much to be yours as you do to have him (Henry VIII, HENRY8_001, PCEEC).

Who is more established in the language of Elizabeth’s household members, and her peripheral kin (and social climber), Thomas Seymour. Despite the small token counts, all informants use the form outside of closing formulae:


to his wife, who I think knoweth of our matters (Seymour to Parr, 1547).

The heterogeneity of this data is difficult to explain. The macro-level evidence suggests that those with a high level of education would be more likely to use who. Yet, in the Tudor family network, these are precisely the individuals who use who least frequently. This trend, alongside the generational differences, may suggest that the change accrued social significance relatively late, as reflected in the mixed usage of the social aspirers, Seymour and Ascham, and its absence in letters of Henry VIII. Contemporary sociolinguistic work on adolescent innovation notes the ability of adolescents to expand the social meaning of a variant as well as quantitatively increase its frequency (Kirkham and Moore 2013). Elizabeth’s generation may thus have helped instantiate the social significance of who, which became quantitatively more prominent in the latter-half of the century, especially at the Court (Nevalainen and Raumolin-Brunberg 2002: 118). This accords with the fifteenth-century Paston data, which also suggests a generational shift (Bergs 2005: 184).

6.6 The which and which

The which was a brief competitor with relative marker which, originating in Northern dialects in late Middle English. The influence of French lesquel has also been mooted as an additional factor (Rissanen 1999: 296–297; examples 11 and 12). The which can be conceived as a failed innovation: CEEC indicates the which occurred in 35% of contexts in the late fifteenth century, falling to around 10% in the early-to-mid sixteenth century (Nevalainen and Raumolin-Brunberg 2003: 74).


I am assured my luck shall be good, the which wholly I commit to God’s prudence (Ascham, 1552).


that way only, which God’s glory his prince’s honour, his country’s profit hath pointed him to follow (Ascham, 1552).

Specific syntactic contexts promote the which (namely, prepositional and sentential clauses) which may indicate grammatical specialization, as part of its downward trajectory (Rissanen 1999: 297; Raumolin-Brunberg 2000: 209). Stylistically, the which has a haphazard distribution, found in both formal and informal text-types in the Helsinki Corpus (Raumolin-Brunberg 2000: 216). Individual writers have their own preferences, but nothing clear-cut emerges at the (macro) generic level. By comparison, the social distribution in the early decades of the sixteenth century shows considerable homogeneity: the which accounts for around 10% of tokens in the Court and London (Nevalainen and Raumolin-Brunberg 2003: 176). Women show a greater preference for the which (25%) in the earliest decades of the century (Nevalainen and Raumolin-Brunberg 2003: 128).

The distribution of the which in the Tudor family corpus shows more variation than the macro-level trends (Table 6). Elizabeth uses 18% the which (total tokens=58), higher than the CEEC average of 9% (1540–1579) (p>0.05). Viewed retrospectively, given the failure of the which, Elizabeth is a conservative user. Mary’s correspondence (written 1530s and 1540s) contains a similar proportion of the which (13.5%; total tokens=74), whereas by contrast, Edward (writing 1540s and 1550s) shows a strong preference for which (98.5%; total tokens=273). The siblings therefore correspond with the macro-level gender trends.

Table 6:

Frequency of which in Tudor family corpus.

However, this pattern is complicated when considered in real time. The which only occurs in Elizabeth’s correspondence before 1550, after which it disappears. One notable biographical correlate with this time-period is Elizabeth’s move from Kathryn Parr’s household to Edward VI’s Court, and her tutoring with Ascham (Somerset 1991: 26). Other studies have posited comparable “pivot” points for individual speakers, such as John Paston III’s move to London (Bergs 2000: 245). One can speculate as to the shift in Elizabeth’s networks that might arise from regular contact with the King’s court, and educated, university men.

The other informants’ data offers some support for this hypothesis. The outgoing variant, the which, is infrequent in the language of Henry VIII (5%; total tokens=20) and Kathryn Parr (7.7%; total tokens=39). Roger Ascham, who has biographical cause to use the which, given his Northern roots, is also progressive (only 4.3%; total tokens=70). Comparatively, Seymour, who was born in the south and was socially ambitious, uses the which 14.3% (total tokens 21). Barnaby Fitzpatrick is least progressive (27.8%; total tokens=22). Although sketchy, the distribution may suggest that the Tudor network may have promoted different social meanings for the same linguistic forms within different sub-groups. Mary, Elizabeth and Thomas Seymour are the most consistently progressive language users in the corpus, which gives some cause to suggest that the which may have some positive social meaning (or, at least, was not highly stigmatised) at a local level for them. This supports the interpretation that the loss of the which in Elizabeth’s writing post-1550 arose from her changing connections with the Court, and subsequent adoption of their linguistic norms.

6.7 Replacement of mine/thine with my/thy

The shift from “long” or n-form first-person possessive determiners, mine/thine, to short-form my/thy took place during the fifteenth and sixteenth centuries (Lass 1999: 147; also Schendl 1997). The change was phonologically determined by the initial phoneme of the nominal referent. Consonant-initial contexts were most progressive, with the change complete by 1500 (Nevalainen and Raumolin-Brunberg 2003: 62). Vowel-initial, h-initial and lexical own contexts adopted the new forms post-1500, and these are the focus of the present analysis. More formal text-types maintain the conservative forms, although this stylistic association may be limited to the second half of the sixteenth century (Schendl 1997: 182, 185). CEEC suggests my/thy emerged from below among lower-ranking groups. The Court therefore, as a domicile populated with educated, high-ranking informants, lags behind other domiciles (Nevalainen and Raumolin-Brunberg 2003: 180). CEEC also suggests a female advantage (Nevalainen and Raumolin-Brunberg 2003: 119).

In the Tudor family network, Elizabeth is notably progressive in this change (c. 87%; total tokens=23), both when compared with the macro-level average (43.5% for the period 1500–1539 and 80% 1540–1579 (figures adapted from Nevalainen and Raumolin-Brunberg 2003: 218, Table 2) and her siblings. Mary (35.1%; total tokens=57) and Edward (42.8%; total tokens=21) use the new forms far less frequently. Mary’s conservatism is surprising, given the macro-level gender bias, as well as her progressiveness in other changes studied (see Table 7).

Table 7:

Frequency of my/thy in Tudor family corpus.

The data for the older generations reveals a similarly mixed picture. The data for Henry VIII (21.4%; total tokens = 14) and Roger Ascham (42.1%; total tokens=38) suggests a conservative usage. Both men are, along with Edward, the most educated in the corpus. Conversely, Thomas Seymour – a social climber with a more moderate education – appears very progressive (88.9%; total tokens=18). Kathryn Parr also shows a preference for the new variant (52.5%; total tokens 46). Parr’s usage suggests that she did not perceive my/thy to have especially informal connotations, as she uses the short-form in a letter written to the University of Cambridge in 1546, as well as in her correspondence to Seymour. Finally, in the limited data for Elizabeth’s caregivers, there is positive evidence that Kat Ashley uses the innovative variant, whereas there are no attestations for Thomas Parry. The data again suggests that there may be a distinction between central (royal) Court norms and more peripheral norms in this change, overlapping with gender and educational bias.

6.8 Third person singular verb-ending -eth and -s

The final change for discussion is the development of the third-person singular verb-ending. The change from -eth (interdental fricative e.g. runneth) to -s (alveolar fricative; e.g. runs) in the Early Modern period has a range of associated internal and external factors (e.g. Kytö 1993; Nevalainen and Raumolin-Brunberg 2003; Gries and Hilpert 2010). Internally, verbs with stem-final sibilants promote -eth (Kytö 1993; Gries and Hilpert 2010), and word-initial alveolar fricatives also show a dampening effect on the use of -s (Gries and Hilpert 2010: 310–311), although Gries and Hilpert suggest these factors were only significant in the seventeenth century. Lexical frequency also has a prominent influence, with auxiliary verbs do and have preferring -eth until the late seventeenth century (Ogura and Wang 1996; Gries and Hilpert 2010). Although conventional wisdom posits -eth as the more formal variant (e.g. Jespersen 1905: 194), more recent work suggests that the register of the variable was mixed and – in correspondence, at least – of only moderate significance (Bambas 1947; Nevalainen and Raumolin-Brunberg 2003: 178; Gries and Hilpert 2010: 307). One exception is the gender of an addressee, with the innovative form used more in interactions with the opposite sex (Gries and Hilpert 2010: 309).

CEEC shows an interrupted development at the social level, with -s “[living] on as a common suffix among the lower orders until it gained new popularity after 1620” (Nevalainen and Raumolin-Brunberg 2003: 145), used more frequently by women. The diffusion pattern could reflect social stigmatization prior to 1600, potentially linked to the Northern origins of -s which migrated to the capital but not the Court (Nevalainen and Raumolin-Brunberg 2003: 178). Indeed, -s may have been a regional stereotype; Alexander Gil describes has as a northern form in 1619 (cited in Nevalainen and Raumolin-Brunberg 2003: 179), presumably linked to the lexical lag of have and do in the adoption of -s. The sixteenth-century Court preserved the conservative form, but later adopts the innovative form, preceding supralocalisation.

The replacement of -eth by -s has been explored at the micro-level, also, with Early Modern writers showing evidence of communal (lifespan) change. Raumolin-Brunberg (2005) observes the rise of -s in Elizabeth’s adult years (1580–1600). The Tudor family corpus (Table 8) shows that Elizabeth’s upward trajectory in adulthood is preceded by a progressive adolescence: 29.7% with lexical verbs and 13.9% with grammatical verbs (total tokens=73). By comparison, Edward is a 100% user of -eth (tokens=167) and Mary also uses -eth for lexical verbs without exception. The single token of auxiliary does in Mary’s correspondence is of uncertain authenticity. The transcription is based on a lost manuscript, although the eighteenth-century editor does use hath/doth elsewhere (Hearne 1716: 143). Barnaby Fitzpatrick, who was schooled alongside Edward and Elizabeth, also uses -eth exclusively. This suggests -eth was the family network norm, meaning Elizabeth is orientating away from her siblings in this change. Importantly, Elizabeth’s letters include -s when writing to both genders, suggesting that she did not discriminate on this basis. Likewise, her siblings’ letters include mixed-gender interactions, but this clearly does not promote the innovative variant.

Table 8:

Frequency of -s (%) with lexical verbs and have/do in Tudor family corpus.

The preference for -eth is also evident in the older nuclear family. Neither Henry VIII (tokens=16) nor Kathryn Parr (tokens=67) favour the innovative variant in their autograph writing, exemplifying the Court avoidance identified in CEEC. Elizabeth’s caregivers and educators show a more progressive usage. Although the token counts are low, and must be treated with caution, the presence of -s in the language of Parry and Ashley offer positive evidence of their usage (contrasting with the absolute use of -eth in the royal informants). Thomas Parry’s data includes lexical -s and a rare instance of sixteenth-century does (Raumolin-Brunberg 1996: 106). Kat Ashley also uses -s, including two examples of grammatical verb has: ‘this sorrow has made it much worse’. The evidence of grammatical verbs could signal a localised norm favouring innovation (or, at least, the lack of stigmatization) in this change within Elizabeth’s household in the larger Court network (Bergs 2000). The presence of has in Elizabeth’s language permits speculation that her connection with Ashley and Parry (and her household more broadly), may promote the innovative forms in her idiolect.

However, the data for -eth and -s affords a different picture when viewed dynamically. Although lexical verb -s increases over her lifespan (Raumolin-Brunberg 2005), Elizabeth’s use of has is truncated. There are no examples in her writing after 1550, the same cut-off point as the which. The hypothesised social stigma of -s, and specifically Gil’s presentation of has as a Northern (and thus non-Courtly) form, could explain the decline of has in Elizabeth’s idiolect, if we see her changing location to Edward’s Court and her education impacting her network position and re-aligning her linguistic values from norms of her peripheral household to the norms of the central Court. In support of this, Ascham (an educated social climber) uses -s a progressive 44% for lexical verbs, but grammatical verbs always take -eth (total tokens=131).

7 Discussion

I now wish to assess what, if any, patterns of linguistic variation can be identified in the Tudor family network, and how informants’ network positions may correlate with their language use. Bergs (2005, 2012) hypothesises that the impact of network type on language change may be differently configured in past societies. He suggests that the absence of an overtly prestigious national standard entails that norm-promoting structures were locally configured. Tight-knit networks plausibly promoted their own set of linguistic norms, whereas individuals of loose-knit networks could “develop their own, personal linguistic systems” and “change and adopt their verbal behaviours in deliberate acts of identity” (Bergs 2012: 95; also Bergs 2005: 55). The sixteenth century has been viewed as a transition point in the development of the standard (e.g. Wright 2000; Bergs 2012), and the Court’s role in the supralocalisation of subsequent standard forms suggests that the Tudor family network could shed light on the specifics of this process.

Each informant constitutes a member of the Tudor Court in a broad sense, but with important idiosyncratic differences. It would be unwise to equate the experiences of Kathryn Parr with Edward VI, even if they did reside in the same household during the 1540s. To try and quantify these differences in experience, I follow Bergs (2005) in calculating Network Strength Scores, based on eleven criteria that indicate the number of ties, their density and the multiplexity, for each informant (Appendix, Tables 10, 11 and 12; Bergs 2005: 55–80; Conde-Silvestre 2012: 340–341). The NSS is calculated by dividing number of ties by the combined total of density and multiplexity scores. Higher scores indicate looser, uniplex networks. In previous studies, such networks promote linguistic innovation, with these members acting as bridges between different networks (e.g. Milroy 1980; Bergs 2005). Bergs acknowledges that scoring these criteria involves a degree of subjectivity, but “NSS can still serve […] as orientating statements. They can give an idea of whose overall life mode […] is more advantageous for changes” (Bergs 2005: 262). It is from this perspective, then, that the scores for the Tudor family network members are examined (Table 9).

Table 9:

Network Strength Scores for the Tudor Family Network (following Bergs 2005). Informants ranked by posited category.

The Tudor NSS show a fairly restricted range: Barnaby Fitzpatrick scores the highest (+7) and the lowest is Kat Ashley (–9). This partly reflects the components used to calculate the network scores, which rely on general, structural information (necessarily, given the limitations of historical evidence), rather than capturing more content (transactional) information which would be more refined.

More surprising, however, is that a positive score does not show a consistent correlation with a more progressive informant. For example, Edward (+6) is progressive in four of the eight changes, whereas Elizabeth is progressive in all but one change (loss of the which), despite having a negative score (–2). There are several plausible reasons. The first is methodological. It is not certain that the NSS components are relevant for the sixteenth-century Tudor court; it was of course devised for the Paston letters (Bergs 2005). Further testing and refinement of the NSS components is desirable, although factors such as gender, marital status, travel and formal offices have a justifiable relevance.

However, following the principle that network analysis is a heuristic device, the relationship between the language use of informants and their social networks offers a more intriguing correlation, if the kinds of linguistic change, and their Courtly status, is explored. All informants in the Tudor family network are progressive in at least one change, and more typically across multiple changes, supporting the Court’s macro-level significance. However, an informant’s network and posited sub-group (nuclear, household, peripheral) seem to correlate with the extent to which the variable had salience at the Court at a macro-level. This potential connection is most convincingly seen in relation to the two changes that emerged from below and lag at the Court: possessive determiners my/thy and the third-person singular verb-ending -s. The most central (and prestigious) members of the Court – Henry VIII, Kathryn Parr, Edward VI – are conservative users, whereas peripheral Court members (Elizabeth, Ascham, Parry and Seymour) are more progressive. Changes with a more progressive trajectory have a reversed distribution: in the replacement of ye by you, Parry and Seymour are conservative users, whereas Edward, Henry, Parr and Ascham are more progressive. This is not a hard and fast trend – micro-level heterogeneity is something of a truism (cf. Bergs 2005) – but it does suggest that core members of the Court had linguistic normative pressures perhaps different to those in more peripheral sub-groups. Better populating the networks at the Tudor Court, expanding to include professional men and their families will allow this interpretation to be tested.

Regarding the Court’s macro-level catalysing role for supralocalisation, the presence of exceptionally high-status individuals (the monarch) may have tempered linguistic innovation (cf. Conde-Silvestre 2012: 343). Sociolinguistic network analysis suggests that if a new form is accepted by central members, it generalises, whereas forms rejected by high-status members stand little chance of becoming the norm. It makes sense, therefore, that Elizabeth, Ascham, Parry, Ashley and Seymour, show more innovative language use, particularly in relation to changes not yet underway at the Court. These informants occupied more peripheral positions, in relation to the central court members (Edward VI, Henry VIII), and could better adopt innovations from outside this domain.

Two individuals in particular support this hypothesis: the second-most progressive language user in the corpus, Thomas Seymour, and Elizabeth, the most progressive speaker. Socially, Seymour’s character seems to accord with Labov’s (2001) non-conformist linguistic leaders; he was socially ambitious and successful, prior to his fall in March 1549. Somerset (1991: 20) describes him as “an exceptionally attractive bachelor … whose shallow intellect and dangerous streak of instability were hidden beneath a formidable charm”. His linguistic innovativeness, and general awareness of language, is suggested by his meta-linguistic comments; for example, he apologises that: “I never over read it [the letter] after it was written wherefore if any fault be I pray you hold me excused”. The trope of “bad writing” is more typical of feminine expression, and it is unusual to find it in male to female correspondence (Daybell 2012: 47). The social aspirer tends to have weak ties that encourage innovation and early adoption of forms (Nevalainen and Raumolin-Brunberg 2003: 152), and, indeed, Seymour’s social mobility saw him shift from Wiltshire, to the Court, to Parr’s household in the space of a decade. He can therefore be viewed as a bridge, a link between different networks and a route through which linguistic change can diffuse.

To an extent, Elizabeth is similarly mobile and ambitious. She has local status within her household, shown through the formulaic, respectful address forms, but had less significance than other network members in the period considered here. During her father’s reign she was repeatedly removed and reassigned to the succession, and viewed primarily as European marriage material (hence her education). It was only in later adolescence (post-1550) that her social trajectory moved her into more central positions at the Court, and, upon her accession in 1558, that she became its central member.

One hypothesis, therefore, is that it is Elizabeth’s transition, along with her household and others of her peripheral network who moved with her, which partly underlies the Court’s catalytic role. Elizabeth was, for the first twenty-five years of her life, on the peripheries of the central Court network: a position that appears to have allowed her to negotiate her linguistic identity using both Courtly variants (e.g. you, positive do and single negation) and non-Courtly forms (e.g. my/thy, -s). However, from 1558 Elizabeth moved from a subordinate position in the network to become its ideological centre as Queen. Just as Nurmi (1999) has speculated that the arrival of the Scottish King, James I, triggered a shift in the evaluation and use of periphrastic do, we might speculate that Elizabeth’s change in network position signalled a reconfiguration of the sociolinguistic norms previously operating at the Court. It would be this variety, or this idea of a variety, that became salient to Puttenham and all aspiring Courtiers in the latter half of the sixteenth century.

8 Conclusion

This paper explores the language of the Tudor Court from a micro-level perspective, in order to better understand the processes of morphosyntactic language change in the sixteenth century, and the Court’s catalytic position in supralocalisation. The application of social network analysis, as a heuristic device, reveals some correlates between network type and linguistic behaviour, and suggests that the central family members of the Court did help to shape the local norms that would later become the standard. As an exploratory study, there are limitations. These are primarily related to the dataset, which is limited in size (i.e. word counts) and scope (i.e. number of informants). Future work on enriching the picture sketched here, drawing on the archives of other Court-based informants such as the Cecils or the Dudleys, would help to test the presented hypotheses and interpretations. The study, due to limitations of space, has also only considered morphosyntactic variation. Other dimensions of language use, such as spelling, present further resources to develop our understanding of language use within the Tudor family network, to refine our methodologies for historical sociolinguistic network analysis, and to improve our understanding of processes of language change in Early Modern English.


I wish to thank the two anonymous reviewers and the editors for their thoughtful and astute feedback on a previous version of this paper. All remaining errors are my own.


  • Alford, Stephen. 2002. Kingship and politics in the reign of Edward VI. Cambridge: Cambridge University Press. Google Scholar

  • Bambas, Rudolph C. 1947 [1998]. Verb forms in -s and -th in Early Modern prose. In Mats Ryden et al. (eds.), A reading in early Modern English, 65–71. Berlin: Frankfurt am Main. Google Scholar

  • Bergs, Alexander. 2000. Social networks in pre-1500 Britain: Problems, prospects, examples. European Journal of English Studies 4(3). 239–251. Google Scholar

  • Bergs, Alexander. 2005. Social networks and historical sociolinguistics: Studies in morphosyntactic variation in the Paston Letters. Berlin: Walter de Gruyter. Google Scholar

  • Bergs, Alexander. 2006. Analyzing online communication from a social network point of view: Questions, problems, perspectives. Language@Internet 3, article 3. 

  • Bergs, Alexander. 2012. The uniformitarian principle and the risk of anachronisms in language and social history. In Juan Manuel Hernández-Campoy & Juan Camilo Conde-Silvestre (eds.), The handbook of historical sociolinguistics, 80–98. Oxford: Wiley-Blackwell. Google Scholar

  • Borman, Tracy. 2009. Elizabeth’s women: The hidden story of the virgin queen. London: Random House. Google Scholar

  • Bugge, Edit. 2010. The role of “family” in intergenerational transmission of Shetland dialect vocabulary. In Robert McColl Miller (ed.), Northern lights, Northern words: Selected papers from the FRLSU conference, Kirkwall, 2009, 70–97. Aberdeen: Forum for Research on the Languages of Scotland and Ireland. Google Scholar

  • Conde-Silvestre, Juan Camilo. 2012. The role of social networks and Mobility in Diachronic Sociolinguistics. In Juan Manuel Hernández-Campoy & Juan Camilo Conde-Silvestre (eds.), The handbook of historical sociolinguistics, 332–352. Oxford: Wiley-Blackwell. Google Scholar

  • CEEC = Corpus of Early English Correspondence. 1998. Compiled by Terttu Nevalainen, Helena Raumolin-Brunberg, Jukka Keränen, Minna Nevala, Arja Nurmi and Minna Palander-Collin at the Department of Modern Languages, University of Helsinki.

  • Costor, Will. 2001. Family and kinship in England: 1540–1800. Harlow: Pearson Education. Google Scholar

  • Culpeper, Jonathan & Merja Kytö. 2010. Early Modern English dialogues: Spoken interaction as writing. Cambridge: Cambridge University Press. Google Scholar

  • Daybell, James. 2012. The material letter in Early Modern England: Manuscript letters and the culture and practices of letter-writing, 1512–1635. London: Palgrave. Google Scholar

  • Daybell, James. 2015. Social negotiations in Correspondence between Mothers and Daughters in Tudor and Early Stuart England. In Women’s History Review 24(4). 502–527. 

  • Davis, Norman. 1967. Style and stereotype in Early English letters. Leeds Studies in English 1. 7–17. Google Scholar

  • Eckert, Penelope. 1988. Adolescent social structure and the spread of linguistic change. Language in Society 17. 183–207. Google Scholar

  • Eckert, Penelope. 2000. Language variation as social practice. Oxford: Blackwell. Google Scholar

  • Ellegård, Alvar. 1953. The Auxiliary ‘do’: The establishment and regulation of its growth in English. Stockholm: Almqvist and Wiksell Google Scholar

  • Elspaß, Stephan. 2012. The use of private letters and diaries in sociolinguistic investigation. In Juan Manuel Hernández-Campoy & Juan Camilo Conde-Silvestre (eds.), The handbook of historical sociolinguistics, 156–169. Oxford: Wiley-Blackwell. Google Scholar

  • Evans, Mel. 2013. The language of Queen Elizabeth I: A sociolinguistic perspective on royal style and identity (Transactions of the Philological Society Monograph Series 46). Oxford: Wiley-Blackwell. Google Scholar

  • Granovetter, Mark S. 1973. The strength of weak ties. American Journal of Sociology 78(6). 1360–1380. Google Scholar

  • Gries, Stefan Th. & Martin Hilpert. 2010. Modelling diachronic change in the third person singular: A multifactorial, verb- and author-specific exploratory approach. English Language and Linguistics 14(3). 293–320. Google Scholar

  • Hazen, Kirk. 2003. The family. In Jack K. Chambers, Peter Trudgill & Natalie Schilling-Estes (eds.), The handbook of language variation and change, 1st edn. 500–525. Malden, MA: Wiley-Blackwell. Google Scholar

  • Hearne, Thomas. 1716. Sylloge Epistolarum, a variis Angliae principius scriptarum. Oxford. Google Scholar

  • Houlbrooke, Ralph A. 1984. The English family: 1450–1700. London and New York: Longman. Google Scholar

  • Hughes, Geoffrey. 2006. Encyclopedia of swearing: The social history of oaths, profanity, foul language, and ethnic slurs in the English-speaking world. Armonk, NY, USA: M.E. Sharpe, Inc. [Accessed 8 April 8 2015. ProQuest ebrary]. Google Scholar

  • Jespersen, Otto. 1905. The growth and structure of the English language. Oxford: Blackwell. Google Scholar

  • Kallel, Amel. 2002. The age variable in the rise of periphrastic “do” in English. Reading Working Papers in Linguistics 6. 161–185. Google Scholar

  • Kallel, Amel. 2007. The loss of negative concord in Standard English: Internal factors. Language Variation and Change 19. 27–49. Google Scholar

  • Kerswill, Paul. 1996. Children, adolescents and language change. Language Variation and Change 8. 177–202. Google Scholar

  • Kerswill, Paul & Ann Williams. 2000. Creating a New Town koine: Children and language change in Milton Keynes. Language in Society 29. 65–115. Google Scholar

  • Kirkham, Sam & Emma Moore. 2013. Adolescence. In Jack K. Chambers & Natalie Schilling (eds.), The handbook of language variation and change, 2nd edn. 277–296. Malden, MA: Wiley-Blackwell. Google Scholar

  • Kytö, Merja. 1993. Third-person present singular verb inflection in early British and American English. Language Variation and Change 5. 113–139. Google Scholar

  • Labov, William. 2001. Principles of linguistic change: Social factors. Oxford: Blackwell. Google Scholar

  • Lass, Roger. 1999. Phonology and Morphology. In Roger Lass (ed.), The Cambridge history of the English language, 56–186. Cambridge: Cambridge University Press. Google Scholar

  • May, Steven W. (ed.). 2004. Elizabeth I: Selected Works. New York: Washington Square Press. Google Scholar

  • May, Steven W. 2008. George Puttenham’s Lewd and Illicit Career. Texas Studies in Literature and Language 50(2). 146–172. Google Scholar

  • Milroy, Lesley. 1980/1987. Language and social networks. New York: Blackwell. Google Scholar

  • Nevala, Minna. 2004. Address in Early Modern correspondence: Its forms and socio-pragmatic functions (Mémoires de la Société Néophilologique de Helsinki 64). Helsinki: Société Néophilologique. Google Scholar

  • Nevalainen, Terttu. 2000. Gender differences in the evolution of Standard English: Evidence from the Corpus of Early English Correspondence. Journal of English Linguistics 28. 38–59. Google Scholar

  • Nevalainen, Terttu. 2002. What’s in a royal letter? Linguistic variation in the vorrespondence of King Henry VIII. In Katja Lentz & Ruth Möhlig (eds.), Of Dyuersitie & Chaunge of Langage: Essays presented to Manfred Görlach on the occasion of his 65th birthday (Anglistische Forschungen 308), 169–179. Heidelberg: Universitätsverlag C. Winter. Google Scholar

  • Nevalainen, Terttu. 2006. Negative concord as an English “Vernacular Universal”: Social history and linguistic typology. Journal of English Linguistics 34. 257–278. Google Scholar

  • Nevalainen, Terttu. 2009. Grasshoppers and blind beetles: Caregiver language in Early Modern English correspondence. In Arja Nurmi, Minna Nevala & Minna Palander-Collin (eds.), The language of daily life in England (1400–1800). 137–164. Amsterdam: Benjamins. Google Scholar

  • Nevalainen, Terttu. 2013. Words of kings and counsellors: Register variation and language change in early English courtly correspondence. In Esther-Miriam Wagner, Ben Outhwait & Bettina Beinhoff (eds.), Scribes as agents of language change, 99–119. Berlin: De Gruyter Mouton. Google Scholar

  • Nevalainen, Terttu & Helena Raumolin-Brunberg. 2002. The rise of the relative who in Early Modern English. In Patricia Poussa (ed.), Relativisation on the North Sea Littoral (LINCOM Studies in Language Typology 7), 109–121. München: LINCOM Europa. Google Scholar

  • Nevalainen, Terttu & Helena Raumolin-Brunberg. 2003. Historical sociolinguistics: Language change in Tudor and Stuart England (Longman Linguistics Library). London: Longman. Google Scholar

  • Nevalainen, Terttu, Helena Raumolin-Brunberg & H. Mannila. 2011. The diffusion of language change in real time: Progressive and conservative individuals and the time-depth of change. Language Variation and Change 23(1). 1–43. Google Scholar

  • Nurmi, Arja. 1999. A social history of periphrastic DO (Mémoires de la Société Néophilologique de Helsinki 56). Helsinki: Société Néophilologique. Google Scholar

  • Ogura, Mieko & William S-Y. Wang. 1996. Snowball effect in lexical diffusion: the development of –s in the third person present singular present indicative in English. In Derek Britton (ed.), English historical linguistics 1994, 119–141. Amsterdam & Philadelphia: John Benjamins. Google Scholar

  • PCEEC =Parsed Corpus of Early English Correspondence, text version. 2006. Compiled by Terttu Nevalainen, Helena Raumolin-Brunberg, Jukka Keränen, Minna Nevala, Arja Nurmi & Minna Palander-Collin, with additional annotation by Ann Taylor. Helsinki: University of Helsinki and York: University of York. Distributed through the Oxford Text Archive.

  • Puttenham, George. (2007 [1589]). The Art of English Poesie. Middlesex: Echo Library.Google Scholar

  • Raumolin-Brunberg, Helena. 1996. Apparent time. In Terttu Nevalainen & Helena Raumolin-Brunberg (eds.), Sociolinguistics and language history. Studies based on the Corpus of Early English Correspondence (Language and Computers 15), 93–109. Amsterdam & Atlanta: Rodopi. Google Scholar

  • Raumolin-Brunberg, Helena. 2000. Which and the which in Late Middle English: Free variants? In Irma Taavitsainen, Terttu Nevalainen, Päivi Pahta & Matti Rissanen (eds.), Placing middle English in context (Topics in English Linguistics 35), 209–225. Berlin & New York: Mouton de Gruyter. Google Scholar

  • Raumolin-Brunberg, Helena. 2005. Language change in adulthood: Historical letters as evidence. European Journal of English Studies 9(1). 37–51. Google Scholar

  • Rissanen, Matti. 1991. Spoken language and the history of do-periphrasis. In Dieter Kastovsky (ed.), Historical English syntax (Topics in English Linguistics), 321–342. Berlin: Mouton de Gruyter. Google Scholar

  • Rissanen, Matti. 1999. Syntax. In Roger Lass (ed.), The Cambridge history of the English language. Volume III 1476–1776, 187–326. Cambridge: Cambridge University Press. Google Scholar

  • Rissanen, Matti. 2000. Standardisation and the language of early statutes. In Laura Wright (ed.), The development of Standard English, 1300–1800: Theories, descriptions, conflicts. (Studies in English Language), 117–130. Cambridge: Cambridge University Press. Google Scholar

  • Sairio, Anni. 2009. Language and letters of the Blue-stocking network: Sociolinguistic issues in eighteenth-century epistolary English. (Mémoires de la Société Néophilologique de Helsinki 75). Helsinki: Société Néophilologique. Google Scholar

  • Schendl, Herbert. 1997. Morphological variation and change in Early Modern English: My/mine, thy/thine. In Raymond Hickey & Stanislaw Puppel (eds.), Language history and linguistic modelling: A Festschrift for Jacek Fisiak on his 60th birthday, Volume I. Language history, 179–191. Berlin & New York: Mouton de Gruyter. Google Scholar

  • Somerset, Anne. 1991. Elizabeth I. London: Phoenix Press. Google Scholar

  • Taavitsainen, Irma & Andreas Jucker. 2008. Methinks you seem more beautiful than ever: Compliments and gender in the history of English. In Irma Taavitsainen & Andreas Jucker (eds.), Speech acts in the history of English, 195–228. Amsterdam and Philadelphia: John Benjamins. Google Scholar

  • Tagliamonte, Sali & Alexandra D’Arcy. 1999. Peaks beyond phonology: Adolescence, incrementation and language change. Language 85(1). 58–108. Google Scholar

  • Tieken-Boon van Ostade, Ingrid M. 2005. Of social networks and linguistic influence: The language of Robert Lowth and his correspondents. In Juan Camilo Conde-Silvestre & Juan Manuel Hernández-Campoy (eds.), Sociolinguistics and the history of English: Perspectives and problems. International Journal of English Studies 5:1. 135–157. Murcia: Universidad de Murcia, Servicio de publicaciones. Google Scholar

  • Warner, Anthony. 2005. Why DO dove: Evidence for register variation in Early Modern English negatives. Language Variation and Change 17. 257–280. Google Scholar

  • Whigham, Frank. 1981. The rhetoric of Elizabethan suitors’ letters. Publications of the Modern Language Association 96. 864–882. Google Scholar

  • Wischer, Ilse. 2008. What makes a syntactic change stop? On the decline of periphrastic do in Early Modern English affirmative declarative sentences. Studies Anglia Posnaniensia 44. 139–154. Google Scholar

  • Wright, Laura (ed.). 2000. The development of Standard English, 1300–1800: Theories, descriptions, conflicts (Studies in English Language). Cambridge: Cambridge University Press. Google Scholar


Table 10:

Ties scores (following Bergs 2005).

Table 11:

Density scores (following Bergs 2005).

Table 12:

Plexity scores (following Bergs 2005).

Table 13:

Details of the Tudor family corpus.


  • 1

    CEEC is a socially representative corpus of English letters spanning the period c.1410–1681 and comprising 2.7 million words, designed for sociolinguistic research; see Nevalainen and Raumolin-Brunberg (2003: 43–49). 

  • 2

    This is, by contrast, the stage of the family relationship perhaps least understood and critiqued in modern sociolinguistics (Hazen 2003: 506). 

  • 3

    Although Bergs (2005) argues that scribal correspondence preserves the named author’s morphosyntactic preferences in the Paston letters, the differences evident between royal scribal and autograph correspondence (e.g. Nevalainen 2013) means that extending the corpus to scribal documents is problematic. More research is needed into the stylistic features of royal scribal documents in this period. 

  • 4

    In later life, Elizabeth rejected convention, awarding her courtiers intimate nicknames that presumably had an in-group function relevant to the performance of their services to the queen (Earl of Leicester was her “eyes”; Robert Cecil was her “elf”) (May 2004). 

  • 5

    It also illustrates the difficulties encountered when trying to analyse historical family relationships: even the most intimate relationships are only hinted at, and establishing a sense of daily interactions and significant relationships is stymied by the homogeneous system of titles and address. 

  • 6

    A third construction, pre-verbal not, shows a brief appearance in Early Modern English, but never gained substantial ground outside of colloquial contexts (Rissanen 1999: 271). 

  • 7

    It is conventional to document positive do in normalised frequencies, due to the difficulties of identifying non-do declaratives in untagged corpora. 

About the article

Published Online: 2015-08-28

Published in Print: 2015-09-01

Citation Information: Journal of Historical Sociolinguistics, ISSN (Online) 2199-2908, ISSN (Print) 2199-2894, DOI: https://doi.org/10.1515/jhsl-2015-0011.

Export Citation

©2015 by De Gruyter Mouton. Copyright Clearance Center

Comments (0)

Please log in or register to comment.
Log in