A corpus-based network analysis of onomastic references in 16th- and 17th century British grammar writing

2.1Defining grammar

As noticed by Mitchell, grammar books during the Early Modern English period covered a wide range of language-related material. The term grammar could mean anything from hard-word lists, spelling, pronunciation, synonyms, homonyms, etymology, and Latin- English dictionaries to poetry, logic, rhetoric, and Scripture lessons. For centuries grammar had been part of a classical education, forming the trivium with logic and rhetoric. Because of grammar’s traditional authority, it held the right to make language decisions (1994Mitchell, Linda C. 1994 “Inversion of Grammar Books and Dictionaries in the Seventeenth and Eighteenth Centuries”. In Euralex 94 Proceedings, 548–554.Mitchell, Linda C. 1994 “Inversion of Grammar Books and Dictionaries in the Seventeenth and Eighteenth Centuries”. In Euralex 94 Proceedings, 548–554.: 548–549). This view of grammar is applicable when considering the span of the Early Modern English period ranging approximately from the 16th-17th century, yet the definition of what constitutes a grammar for 16th and 17th century grammar authors differs.

Dons (2004Dons, Ute 2004 Descriptive Adequacy of Early Modern English Grammars. Berlin & New York: De Gruyter. Dons, Ute 2004 Descriptive Adequacy of Early Modern English Grammars. Berlin & New York: De Gruyter. : 4) claims that prior to the 16th century, a book was only referred to as a grammar when it dealt with the Latin language. Additionally, due to being overshadowed by Latin and French for centuries, writing about the grammar of English was considered uncommon even after the publication of so-called vernacular grammars. This is illustrated in Howell’s prologue addressing the reader: “Now, touching this new English Grammar, let not the Reader mistake, as if it were an English Grammar to learn another Language, as Lillie is for Latin” (1662: xxx).

Nonetheless, in the 16th century, authors attempted to define what grammar means for the English language: “For the first and chef pooint in Grammar for English iz too know what part of spech euery word in euery sentenc iz” (Bullokar 1586Bullokar, Wiliam 1586 Brief Grammar for English. London: Edmund Bollifant.Bullokar, Wiliam 1586 Brief Grammar for English. London: Edmund Bollifant.: 18–19).

By focusing on the individual sentence constituents in English, Bullokar is making an early attempt to create an English grammar framework which is distinct from the Latin model and create a bridge between the medieval tradition of Latin-based education to emphasis on the vernacular language. Bullokar emphasizing that grammar is not just concerned with the correctness of language but is a fundamental tool for understanding the structure and function of language using a practical and analytic approach, aligns with the Humanist approaches of the Renaissance period that aimed to study and revive classical antiquity (see Law 2015Law, Vivien 2015 The History of Linguistics in Europe: From Plato to 1600. Cambridge: Cambridge University Press.Law, Vivien 2015 The History of Linguistics in Europe: From Plato to 1600. Cambridge: Cambridge University Press.; Heath 1971Heath, Terrence 1971 “Logical Grammar, Grammatical Logic, and Humanism in Three German Universities”. Studies in the Renaissance 18. 9–64. Heath, Terrence 1971 “Logical Grammar, Grammatical Logic, and Humanism in Three German Universities”. Studies in the Renaissance 181. 9–64. ).

To understand the shift towards viewing grammar as a way of using language in a proper way, it is necessary to contextualise the process of standardisation unfolding throughout this period. In the late 16th century, the codification stage begins, which develops through the 17th century with Early Modern English grammars that are largely descriptive or normative rather than prescriptive, laying down rules in grammars and dictionaries as authoritative guides (Nevalainen & Tieken-Boon van Ostade 2006Nevalainen, Terttu & Ingrid Tieken-Boon van Ostade 2006 “Standardisation”. A History of the English Language ed. by Richard M. Hogg & David Denison, 271–311. Cambridge: Cambridge University Press. Nevalainen, Terttu & Ingrid Tieken-Boon van Ostade 2006 “Standardisation”. A History of the English Language ed. by Richard M. Hogg & David Denison, 271–311. Cambridge: Cambridge University Press. ). The prescriptive stage of standardisation does not fully emerge until the late 18th century and is often marked by Lowth’s Short Introduction to the English Language (1762Lowth, Robert 1762 A Short Introduction to English Grammar. London: J. Hughs.Lowth, Robert 1762 A Short Introduction to English Grammar. London: J. Hughs.). Lowth’s grammar is distinctive for explicitly identifying and criticising grammatical errors, even in the works of respected authors, drawing on earlier codification efforts to establish a norm of correct usage (ibid.). Vorlat (1998Vorlat, Emma 1998 “Criteria of Grammaticalness in 16th- and 17th-Century English Grammar”. A Reader in Early Modern English ed. by Mats Rydén, Ingrid Tieken-Boon van Ostade & Merja Kytö, 485–496. Frankfurt am Main: Lang.Vorlat, Emma 1998 “Criteria of Grammaticalness in 16th- and 17th-Century English Grammar”. A Reader in Early Modern English ed. by Mats Rydén, Ingrid Tieken-Boon van Ostade & Merja Kytö, 485–496. Frankfurt am Main: Lang.: 485–486) proposes a threefold distinction between: (1) descriptive grammars that record language without value judgments and including all language varieties, (2) normative grammars that still focus on language usage but favour how particular social or regional groups use language for pedagogical purposes, and (3) prescriptive grammars that impose rules based on logical or other criteria rather than actual usage. These different approaches shape how the grammar book is written due to the stance the author takes.

In the 17th century, the educational and prescriptive aims of the time began to broaden. This is reflected in Ben Jonson’s The English Grammar, where he states: Grammar is the art of true, and well speaking a Language: the writing is but an Accident (Jonson 1640Jonson, Ben 1640 The English Grammar.Jonson, Ben 1640 The English Grammar.: 35).

Jonson’s emphasis on rhetorical precision, clarity, and refinement in language use, rather than focusing on parts of speech and didactics, shows how grammar begins to be viewed less as a practical tool to systematically analyse linguistic features, but cultivating intellect.

As there has never been one consistent unified definition of the term grammar, the present study combines a conventional definition, that is “the rules and conventions of everyday language” (McCarthy 2020McCarthy, Michael 2020 Innovations and Challenges in Grammar. London: Routledge. McCarthy, Michael 2020 Innovations and Challenges in Grammar. London: Routledge. : 4) with the notion of a book which is intended to instruct the learner in the ways of using language according to time-period-specific norms. This view enables both a systematic approach to scrutinizing patterns in language and which linguistic conventions have been broadly accepted as a standardized form, whilst also acknowledging that discursive practices may be shaped by social, cultural, and even political factors at the time when the books were written.

2.2Previous work on 19th-century English grammars

In an effort to trace changing and stable language norms, attitudes towards language, and discursive practices in a diachronic way, the HeidelGram project aims to compile and analyse a corpus of English grammars spanning the 16th to 19th centuries. By combining corpus and network analyses, new perspectives on which figures and schools of thought influenced grammarians of the Early Modern English period are offered and qualitative assumptions on the history of grammar writing may be reassessed.

In a pilot study, a corpus of 19th-century grammars from England (40 texts, approximately 2.6 million words) was compiled and analysed. The selection of the texts was based on the criteria described in Busse et al. (2020Busse, Beatrix, Kirsten Gather & Ingo Kleiber 2020 “A Corpus-Based Analysis of Grammarians’ References in 19th-Century British Grammars”. Variation in Time and Space: Observing the world through corpora ed. by Anna Čermáková & Markéta Malá, 133–172. Berlin: De Gruyter. Busse, Beatrix, Kirsten Gather & Ingo Kleiber 2020 “A Corpus-Based Analysis of Grammarians’ References in 19th-Century British Grammars”. Variation in Time and Space: Observing the world through corpora ed. by Anna Čermáková & Markéta Malá, 133–172. Berlin: De Gruyter. : 137–138), which are revisited in the methodology below. The focus of this study was grammarians’ references to each other in their grammar books and what purpose these references served. The addition of networks to the historical corpus analysis not only provides a new mode of visualization but also carries further its own suite of measures and scores, such as degree centrality. It also allows for investigation of individual authors by means of ego networks.

An example of an onomastic reference of the type ‘application’ is the following excerpt from Abbott’s 1871 grammar: The following table will be sufficient to illustrate Grimm’s law: […] (Abbott 1871, emphasis added).

Based on the frequency and variety of such name-based references, Busse et al. (2020)Busse, Beatrix, Kirsten Gather & Ingo Kleiber 2020 “A Corpus-Based Analysis of Grammarians’ References in 19th-Century British Grammars”. Variation in Time and Space: Observing the world through corpora ed. by Anna Čermáková & Markéta Malá, 133–172. Berlin: De Gruyter. Busse, Beatrix, Kirsten Gather & Ingo Kleiber 2020 “A Corpus-Based Analysis of Grammarians’ References in 19th-Century British Grammars”. Variation in Time and Space: Observing the world through corpora ed. by Anna Čermáková & Markéta Malá, 133–172. Berlin: De Gruyter. introduced categories for the types of references that are being made. Six main categories were established using a grounded theory approach, which are listed and described in Table 1. For more detailed information see (Busse et al. 2020Busse, Beatrix, Kirsten Gather & Ingo Kleiber 2020 “A Corpus-Based Analysis of Grammarians’ References in 19th-Century British Grammars”. Variation in Time and Space: Observing the world through corpora ed. by Anna Čermáková & Markéta Malá, 133–172. Berlin: De Gruyter. Busse, Beatrix, Kirsten Gather & Ingo Kleiber 2020 “A Corpus-Based Analysis of Grammarians’ References in 19th-Century British Grammars”. Variation in Time and Space: Observing the world through corpora ed. by Anna Čermáková & Markéta Malá, 133–172. Berlin: De Gruyter. : 143–144).

For the present study, the reference types are understood more broadly. This means that not only references made to other grammarians will be considered, but also to other types of persons, which are to be categorized. For instance, for the reference type Quotation we decided to not only consider citations from other grammar books, but rather any other works in general, including philosophical texts, the Bible, etc.

Both corpus as well as network analyses allowed for initial findings on the types of references made as well as how the practices of referencing have changed, especially in the sense of evaluation. The analysis of the reference network and reference categories during the 19th century has revealed a marked change of both the quantity and the types of references made around 1850. This change indicates a turn away from an evaluative stance in the early decades to a more descriptive approach later in the century.

This was also reflected in who was being referenced. While most references are made by the grammarians who have published their works before 1850 and to the so-called prescriptivists such as Robert Lowth and Lindley Murray, the later grammarians make significantly fewer references and refer more commonly to contemporary descriptive grammarians in the later decades of the 19th century.

2.3The character of 16th- and 17th-century English grammars

Grammaticography, which refers to producing grammar books and dictionaries for specific languages (see Wolf 2015Wolf, Göran 2015 “English Grammaticography as Discourse Tradition: Comments on 18th-century developments”. In Anglistentag 2014 Hannover ed. by Rainer Emig & Jana Gohrisch, 19–33. Trier: Wissenschaftlicher Verlag.Wolf, Göran 2015 “English Grammaticography as Discourse Tradition: Comments on 18th-century developments”. In Anglistentag 2014 Hannover ed. by Rainer Emig & Jana Gohrisch, 19–33. Trier: Wissenschaftlicher Verlag.; Anderwald 2016Anderwald, Lieselotte 2016 Language between Description and Prescription: Verb and verb categories in nineteenth century grammars of English. Oxford: Oxford University Press. Anderwald, Lieselotte 2016 Language between Description and Prescription: Verb and verb categories in nineteenth century grammars of English. Oxford: Oxford University Press. ), developed in Western language sciences from the Graeco-Latin grammatical tradition (Auroux 1994Auroux, Sylvain 1994 La révolution technologique de la grammatisation. Liège: Mardaga.Auroux, Sylvain 1994 La révolution technologique de la grammatisation. Liège: Mardaga.). Raby and Andrieu (2018)Raby, Valérie & Wilfrid Andrieu 2018 “Norms and Rules in the History of Grammar: French and English handbooks in the seventeenth century”. Standardising English: Norms and margins in the history of the English language ed. by Linda Pillière, Wilfrid Andrieu, Valérie Kerfelec & Diana Lewis, 65–88. Cambridge: Cambridge University Press. Raby, Valérie & Wilfrid Andrieu 2018 “Norms and Rules in the History of Grammar: French and English handbooks in the seventeenth century”. Standardising English: Norms and margins in the history of the English language ed. by Linda Pillière, Wilfrid Andrieu, Valérie Kerfelec & Diana Lewis, 65–88. Cambridge: Cambridge University Press. outline that the Renaissance period was a turning point for the process of grammatication, as prior to the late Middle Ages few scholars were concerned with grammar writing. However, the widespread creation of grammars and dictionaries during this period became feasible due to the metalinguistic framework inherited from Graeco-Latin antiquity, although the use of categories established by the Latin model has often been viewed as “an artificial imposition, as a hindrance to valid descriptions of the vernaculars” (ibid.: 67).

The 16th century is said to mark the beginning of English grammar writing (Linn 2006Linn, Andrew 2006 “English Grammar Writing”. The Handbook of English Linguistics ed. by Bas Aarts & April McMahon, 72–92. Oxford: Blackwell. Linn, Andrew 2006 “English Grammar Writing”. The Handbook of English Linguistics ed. by Bas Aarts & April McMahon, 72–92. Oxford: Blackwell. ). It is therefore also a sign of a shifted interest in the varieties of English practised at the time which coincides with the onset of standardization processes of English during that century. As for grammar writing, we can witness a transition from the term grammar only referring to Latin grammatical phenomena towards a more inclusive terminology, allowing grammar to also display features and phenomena of the so-called vernacular languages — including the variety of English used. The reason for this development towards English grammar was due to the shift in the cultural climate of the Renaissance, Reformation, and Humanism (Dons 2004Dons, Ute 2004 Descriptive Adequacy of Early Modern English Grammars. Berlin & New York: De Gruyter. Dons, Ute 2004 Descriptive Adequacy of Early Modern English Grammars. Berlin & New York: De Gruyter. ).

The Renaissance period (late 15th to early 17th century) sparked an interest in classical antiquity and a desire for surpassing the achievements from that era by, for instance, refining the English language which was far less standardized in its pronunciation and spelling than Latin. The forerunners of the Reformation movement challenged perceived discrepancies and lack of transparency on behalf of the Catholic Church and yearned for the general public to understand clerical matters. The Bible is translated from Latin into English resulting in those who could read being able to read the Scriptures in their vernacular language. Humanism encouraged use of English for secular purposes. Another contributing factor to the emergence of English grammars was the invention of the printing press with books becoming more broadly accessible to more people than only those capable of reading and writing in Latin and it also led to a slow process of the standardization of orthography (see Blair 2010Blair, Ann M. 2010 Too Much to Know: Managing scholarly information before the Modern Age. New Haven, CT: Yale University Press.Blair, Ann M. 2010 Too Much to Know: Managing scholarly information before the Modern Age. New Haven, CT: Yale University Press.).

A widely held belief among scholars (e.g., Linn 2006Linn, Andrew 2006 “English Grammar Writing”. The Handbook of English Linguistics ed. by Bas Aarts & April McMahon, 72–92. Oxford: Blackwell. Linn, Andrew 2006 “English Grammar Writing”. The Handbook of English Linguistics ed. by Bas Aarts & April McMahon, 72–92. Oxford: Blackwell. ; McCarthy 2020McCarthy, Michael 2020 Innovations and Challenges in Grammar. London: Routledge. McCarthy, Michael 2020 Innovations and Challenges in Grammar. London: Routledge. ; Nevalainen 2006Nevalainen, Terttu 2006 An Introduction to Early Modern English. Edinburgh: Edinburgh University Press. Nevalainen, Terttu 2006 An Introduction to Early Modern English. Edinburgh: Edinburgh University Press. ; Wolf 2011Wolf, Göran 2011 Englische Grammatikschreibung 1600–1900: Der Wandel einer Diskurstradition. (= Arbeiten zur Sprachanalyse, 54). Frankfurt am Main: Lang.Wolf, Göran 2011 Englische Grammatikschreibung 1600–1900: Der Wandel einer Diskurstradition. (=Arbeiten zur Sprachanalyse, 54). Frankfurt am Main: Lang.) is that the first grammar book of English written in English emerged in 1586. It is William Bullokar’s Bref Grammar for English. However, we found that in our conceptualisation of what constitutes a grammar book, namely both a set of rules and conventions governing everyday language and a guidebook instructing learners based on time-period-specific norms, Sherry’s A Treatise of the Figures of Grammer and Rhetorike (1577Sherry, Richard 1577 A Treatise of the Figures of Grammer and Rhetorike. London: Ricardi Totteli.Sherry, Richard 1577 A Treatise of the Figures of Grammer and Rhetorike. London: Ricardi Totteli.) and Mulcaster’s The First Part of the Elementarie Which Entreateth Chefelie of the Right Writing of our English Tung (1582), both of which were published before, could be considered early examples of grammar books. This is because they emphasized the importance of observing and documenting a uniform grammar system, contributing to the standardization efforts of grammar in the Early Modern period. Algeo (1985)Algeo, John 1985 “The Earliest English Grammars”. Historical & Editorial Studies in Medieval & Early Modern English: For Johan Gerritsen ed. by Mary-Jo Arn, Hanneke Wirtjes & Hans Jansen, 191–207. Groningen: Wolters-Noordhoff.Algeo, John 1985 “The Earliest English Grammars”. Historical & Editorial Studies in Medieval & Early Modern English: For Johan Gerritsen ed. by Mary-Jo Arn, Hanneke Wirtjes & Hans Jansen, 191–207. Groningen: Wolters-Noordhoff. states that the Pamphlet and the Bref Grammar for English have mistakenly been treated as two works due to a binding error of the Bodleian copy of the book. The Pamphlet was modelled after William Lily’s (1534) work on Latin grammar, Rudimenta Grammatices (Usmonovna 2021Usmonovna, Latipova Umida 2021 “The History of English Grammar”. International Journal of Engineering and Information Systems 5:2. 194–198.Usmonovna, Latipova Umida 2021 “The History of English Grammar”. International Journal of Engineering and Information Systems 5:2. 194–198.). According to Le Prieult (2015Le Prieult, Henri 2015 “The ‘Stranger’ and the Grammarian: When Early English Grammarians Reached Out”. Caliban 54.309–326. Le Prieult, Henri 2015 “The ‘Stranger’ and the Grammarian: When Early English Grammarians Reached Out”. Caliban 541.309–326. : 321), Bullokar not only departed from a long-standing tradition, which bestowed on Latin the privilege of giving access to knowledge in all fields, he considered the grammar of English as a better passport for Europe. Padley (1988Padley, G. A. 1988 Grammatical Theory in Western Europe 1500–1700. Cambridge: Cambridge University Press.Padley, G. A. 1988 Grammatical Theory in Western Europe 1500–1700. Cambridge: Cambridge University Press.: 230) criticises Bullokar’s Pamphlet for Grammar, stating that it is: “the first in a long line of works assuming that what is appropriate to the description of Latin will be equally appropriate to the description of the mother tongue”.

However, it is crucial to keep in mind that authors such as Bullokar were among the first to write grammars of English in English. Just because they based their grammars on existing Latin grammar books does not mean they assumed the Latin model would be just as applicable to the vernacular. In fact, Algeo (1985Algeo, John 1985 “The Earliest English Grammars”. Historical & Editorial Studies in Medieval & Early Modern English: For Johan Gerritsen ed. by Mary-Jo Arn, Hanneke Wirtjes & Hans Jansen, 191–207. Groningen: Wolters-Noordhoff.Algeo, John 1985 “The Earliest English Grammars”. Historical & Editorial Studies in Medieval & Early Modern English: For Johan Gerritsen ed. by Mary-Jo Arn, Hanneke Wirtjes & Hans Jansen, 191–207. Groningen: Wolters-Noordhoff.: 206) argues that the grammarians at the time were “fully cognizant of the structural differences between Latin as an inflected language and English as an analytical one”. Algeo further claims that the Latinate approach in these early grammars fulfilled a particular purpose. Students learning English grammar through these books were more readily prepared to learn Latin grammar, as a comparable basis was established (Algeo 1985Algeo, John 1985 “The Earliest English Grammars”. Historical & Editorial Studies in Medieval & Early Modern English: For Johan Gerritsen ed. by Mary-Jo Arn, Hanneke Wirtjes & Hans Jansen, 191–207. Groningen: Wolters-Noordhoff.Algeo, John 1985 “The Earliest English Grammars”. Historical & Editorial Studies in Medieval & Early Modern English: For Johan Gerritsen ed. by Mary-Jo Arn, Hanneke Wirtjes & Hans Jansen, 191–207. Groningen: Wolters-Noordhoff.: 203). Moreover, foreigners who already knew Latin were more gently introduced into the structure of English (Algeo 1985Algeo, John 1985 “The Earliest English Grammars”. Historical & Editorial Studies in Medieval & Early Modern English: For Johan Gerritsen ed. by Mary-Jo Arn, Hanneke Wirtjes & Hans Jansen, 191–207. Groningen: Wolters-Noordhoff.Algeo, John 1985 “The Earliest English Grammars”. Historical & Editorial Studies in Medieval & Early Modern English: For Johan Gerritsen ed. by Mary-Jo Arn, Hanneke Wirtjes & Hans Jansen, 191–207. Groningen: Wolters-Noordhoff.: 191).

The purpose of Bullokar’s book was to dismantle the prevalent assumption that English was not rule-governed like Latin. While he used his own reformed spelling system, other traditional grammars at the time were still written in Latin (Linn 2006Linn, Andrew 2006 “English Grammar Writing”. The Handbook of English Linguistics ed. by Bas Aarts & April McMahon, 72–92. Oxford: Blackwell. Linn, Andrew 2006 “English Grammar Writing”. The Handbook of English Linguistics ed. by Bas Aarts & April McMahon, 72–92. Oxford: Blackwell. ). This is seen by Bullokar identifying the differences between the Latin system and the vernacular and adjusting his grammatical descriptions accordingly, for example by listing five cases, as opposed to six which exist in Latin (Raby & Andreu 2018Raby, Valérie & Wilfrid Andrieu 2018 “Norms and Rules in the History of Grammar: French and English handbooks in the seventeenth century”. Standardising English: Norms and margins in the history of the English language ed. by Linda Pillière, Wilfrid Andrieu, Valérie Kerfelec & Diana Lewis, 65–88. Cambridge: Cambridge University Press. Raby, Valérie & Wilfrid Andrieu 2018 “Norms and Rules in the History of Grammar: French and English handbooks in the seventeenth century”. Standardising English: Norms and margins in the history of the English language ed. by Linda Pillière, Wilfrid Andrieu, Valérie Kerfelec & Diana Lewis, 65–88. Cambridge: Cambridge University Press. ; Michael 1970Michael, Ian 1970 English Grammatical Categories and the Tradition to 1800. Cambridge: Cambridge University Press.Michael, Ian 1970 English Grammatical Categories and the Tradition to 1800. Cambridge: Cambridge University Press., 1987Michael, Ian 1987 The Teaching of English from the Sixteenth Century to 1870. Cambridge: Cambridge University Press. Michael, Ian 1987 The Teaching of English from the Sixteenth Century to 1870. Cambridge: Cambridge University Press. ).

Additionally, with the beginning of books being printed in London at the end of the 15th century, there was a rising economic interest in spelling reform and standardization (Robins 1986Robins, Robert H. 1986 “The Evolution of English Grammar Books Since the Renaissance”. The English Reference Grammar: Language and linguistics, writers and readers ed. by Gerhard Leitner, 292–306. Tübingen: Niemeyer.Robins, Robert H. 1986 “The Evolution of English Grammar Books Since the Renaissance”. The English Reference Grammar: Language and linguistics, writers and readers ed. by Gerhard Leitner, 292–306. Tübingen: Niemeyer.), further incentivizing English grammarians to outline language-specific features of English, particularly for the use in pedagogy (McCarthy 2020McCarthy, Michael 2020 Innovations and Challenges in Grammar. London: Routledge. McCarthy, Michael 2020 Innovations and Challenges in Grammar. London: Routledge. : 23; Michael 1970Michael, Ian 1970 English Grammatical Categories and the Tradition to 1800. Cambridge: Cambridge University Press.Michael, Ian 1970 English Grammatical Categories and the Tradition to 1800. Cambridge: Cambridge University Press., 1987Michael, Ian 1987 The Teaching of English from the Sixteenth Century to 1870. Cambridge: Cambridge University Press. Michael, Ian 1987 The Teaching of English from the Sixteenth Century to 1870. Cambridge: Cambridge University Press. ).

The 17th century represents an eventful time for the development of the English language and its instruction. The English language expanded to areas of language use where the classical languages had previously been dominating, especially in the late 17th century. This led to increased standardization in language usage (Nevalainen 2006Nevalainen, Terttu 2006 An Introduction to Early Modern English. Edinburgh: Edinburgh University Press. Nevalainen, Terttu 2006 An Introduction to Early Modern English. Edinburgh: Edinburgh University Press. : 42; Milroy & Milroy 2012Milroy, James & Lesley Milroy 2012 [1985] Authority in Language: Investigating Language Prescription and Standardisation. 4th ed. London & New York: Routledge.Milroy, James & Lesley Milroy 2012 [1985] Authority in Language: Investigating Language Prescription and Standardisation. 4th ed. London & New York: Routledge.; Nevalainen & Tieken-Boon van Ostade 2006Nevalainen, Terttu & Ingrid Tieken-Boon van Ostade 2006 “Standardisation”. A History of the English Language ed. by Richard M. Hogg & David Denison, 271–311. Cambridge: Cambridge University Press. Nevalainen, Terttu & Ingrid Tieken-Boon van Ostade 2006 “Standardisation”. A History of the English Language ed. by Richard M. Hogg & David Denison, 271–311. Cambridge: Cambridge University Press. ), which, together with other sociopolitical developments, sparked a shift in favour of English being recognized as a separate academic discipline (Beal 2004Beal, Joan C. 2004 English in Modern Times. London: Arnold.Beal, Joan C. 2004 English in Modern Times. London: Arnold.: 102). Authors were addressing a broader audience of intellectuals and scholars curious about language. There is also a gradual shift towards standardization in comparison to the 16th century, since 17th-century writers such as Jonson and Butler focus more on correctness of forms and codifying norms in English. Due to more books being published, there is also a broader range of topics that are being addressed in the grammars published in the 17th century such as more advanced syntax (e.g. Aickin 1693Aickin, Joseph 1693 The English grammar. London: John Lawrence.Aickin, Joseph 1693 The English grammar. London: John Lawrence.) and pronunciation (e.g. Miege 1688).

English grammars written in the 17th century were still said to be “heavily influenced by their models — grammars of Latin” (Algeo 1986Algeo, John 1986 “A Grammatical Dialectic”. The English Reference Grammar: Language and linguistics, writers and readers ed. by Gerhard Leitner, 307–333. (= Linguistische Arbeiten, 172). Tübingen: Niemeyer.Algeo, John 1986 “A Grammatical Dialectic”. The English Reference Grammar: Language and linguistics, writers and readers ed. by Gerhard Leitner, 307–333. (=Linguistische Arbeiten, 172). Tübingen: Niemeyer.: 309) and “ancient Greek and Roman literati were still considered authoritative in terms of linguistic understanding” (McCarthy 2020McCarthy, Michael 2020 Innovations and Challenges in Grammar. London: Routledge. McCarthy, Michael 2020 Innovations and Challenges in Grammar. London: Routledge. : 24), although there is more recognition of the autonomy of English and scholars are beginning to have access to more contemporary works.

3.1Corpus design

As part of the larger HeidelGram corpus, five grammars from the 16th century and 17 grammars from the 17th century were selected. These numbers represent the few English grammars that were extant in the 16th century as well as the marked increase in publications in the 17th century (see Michael 1970Michael, Ian 1970 English Grammatical Categories and the Tradition to 1800. Cambridge: Cambridge University Press.Michael, Ian 1970 English Grammatical Categories and the Tradition to 1800. Cambridge: Cambridge University Press.: 151; Yáñez-Bouza 2016Yáñez-Bouza, Nuria 2016 “Early and Late Modern English Grammars as Evidence in English Historical Linguistics”. The Cambridge Handbook of English Historical Linguistics ed. by Merja Kytö & Päivi Pahta. Cambridge: Cambridge University Press. Yáñez-Bouza, Nuria 2016 “Early and Late Modern English Grammars as Evidence in English Historical Linguistics”. The Cambridge Handbook of English Historical Linguistics ed. by Merja Kytö & Päivi Pahta. Cambridge: Cambridge University Press. : 166–168). The texts were selected in accordance with the selection criteria for the full corpus using higher- and lower-level strata. More specifically, the higher-level stratum of popularity was considered first. This means that the number of published editions, distribution, and mentions in the primary (i.e. the grammars themselves) and secondary literature (such as Görlach 1998Görlach, Manfred 1998 An Annotated Bibliography of Nineteenth Century Grammars of English. Amsterdam & Philadelphia: John Benjamins. Görlach, Manfred 1998 An Annotated Bibliography of Nineteenth Century Grammars of English. Amsterdam & Philadelphia: John Benjamins. ; Alston 1965Alston, R. C. 1965 A Bibliography of the English Language from the Invention of Printing to the Year 1800. Vol. 1. Leeds: Arnold and Son.Alston, R. C. 1965 A Bibliography of the English Language from the Invention of Printing to the Year 1800. Vol. 11. Leeds: Arnold and Son.) were considered to assess the influence a grammar may have had at the time. Among these popular works, the lower-level strata were considered to provide variety in function, audience, and text type (Busse et al. 2020Busse, Beatrix, Kirsten Gather & Ingo Kleiber 2020 “A Corpus-Based Analysis of Grammarians’ References in 19th-Century British Grammars”. Variation in Time and Space: Observing the world through corpora ed. by Anna Čermáková & Markéta Malá, 133–172. Berlin: De Gruyter. Busse, Beatrix, Kirsten Gather & Ingo Kleiber 2020 “A Corpus-Based Analysis of Grammarians’ References in 19th-Century British Grammars”. Variation in Time and Space: Observing the world through corpora ed. by Anna Čermáková & Markéta Malá, 133–172. Berlin: De Gruyter. : 137). Variety in function was achieved by including teaching and scholarly grammars, whereas different text types included the representation of the grammar material in e.g. prose, catechisms, and verse. The audience of the grammars varied as they were intended for novice and advanced learners, youth and adults, natives and foreigners. This stratified sampling approach ensures that the sample includes as much variety within the genre of grammar writing as possible, which leads to a higher degree of representativeness (see Biber 1993Biber, Douglas 1993 “Representativeness in Corpus Design”. Literary and Linguistic Computing 8:4. 243–257. Biber, Douglas 1993 “Representativeness in Corpus Design”. Literary and Linguistic Computing 8:4. 243–257. , Egbert et al. 2022Egbert, Jesse, Douglas Biber & Bethany Gray 2022 Designing and Evaluating Language Corpora: A practical framework for corpus representativeness. Cambridge: Cambridge University Press. Egbert, Jesse, Douglas Biber & Bethany Gray 2022 Designing and Evaluating Language Corpora: A practical framework for corpus representativeness. Cambridge: Cambridge University Press. ). We follow the definition of genres as “categories of texts which are determined by both formal and functional criteria” (Busse 2015Busse, Beatrix 2015 “Genre”. The Cambridge Handbook of Stylistics ed. by Peter Stockwell & Sara Whiteley, 103–116. Cambridge: Cambridge University Press.Busse, Beatrix 2015 “Genre”. The Cambridge Handbook of Stylistics ed. by Peter Stockwell & Sara Whiteley, 103–116. Cambridge: Cambridge University Press.: 103). The inclusion of grammar books that serve different purposes enables us to have a broader overview of who the English grammar scholars considered to be experts on the topic at the time, and how they interacted with one another.

Table 2 lists the grammars that were selected for the 16th- and 17th-century components of the HeidelGram corpus sorted by year of publication.

Following common practice in earlier works, for these texts, first editions, or the earliest available editions, of the books have been traced and scanned. Further processing of the digitized texts is described in the methodology Section (4.1).

Within the HeidelGram corpus, and therefore also in these sub-corpora, a whole text approach is followed. This means that the grammar texts are included in the corpus in full length. Although the grammars vary significantly in size, it is essential to include the full text in order to attain a holistic understanding of the genre of English grammar writing. Selecting same-size samples for the benefit of balance would lead to the loss of domain characteristics (Hunston 2008Hunston, Susan 2008 “Collection Strategies and Design Decisions”. Corpus Linguistics ed. by Anke Lüdeling & Merja Kytö, 154–167. Berlin: De Gruyter.Hunston, Susan 2008 “Collection Strategies and Design Decisions”. Corpus Linguistics ed. by Anke Lüdeling & Merja Kytö, 154–167. Berlin: De Gruyter.: 165–166). The drawback of the different length of the texts are mitigated through normalized frequencies in the analyses.

3.2Corpus compilation

While it is not the focus of this research paper, we want to briefly address some of the crucial steps in the compilation of the HeidelGram corpus. For the corpus data, PDF scans of the first editions, if available, of the grammar books were collected from online archives and through library scans. Where possible, OCR was used to transcribe the texts into machine-readable format. For each grammar text, two corpus files were created, a clean text file and an XML-annotated text file. However, as mentioned above, the historical nature of these texts complicates the compilation process. In lieu of a standardized printing process or standardized spelling at the time period under investigation, the books exhibit a vast range of irregularities, be it in fonts, page structure, figures, spelling, or even special characters. Thus, manual transcription and correction was necessary to allow for clean transcriptions of the grammars. The raw texts were then manually annotated for visual, structural, and project-specific features, such as typeface, text placement, and references to personal names, respectively. Annotations are added in TEI-adjacent XML and the full guidelines will be published alongside the corpus once fully compiled. For example, in Howell (1662)Howell, James 1662 A New English Grammar. London: T. Williams, H. Brome, and H. Marsh.Howell, James 1662 A New English Grammar. London: T. Williams, H. Brome, and H. Marsh. a reference to Ben Jonson is made, which is annotated as: “[...] it cannot be in the compas of human brain to compile an exact regular Syntaxis thereof, <person>Mr. Ben. Johnson</person> a great Wit, who was as patient as he was elaborat [...].”

These manual processes of transcription and annotation were performed by the project’s team members, that is the present authors and trained student assistants. The project-specific annotation of references to personal names is crucial to the extraction process of onomastic references outlined in the following section.

In order to have a broad overview of all references to persons in the corpus, we decided to tag any mentions of a person when annotating the corpus and this was only filtered out once the concordances were extracted and it was apparent that it was indeed a reference to a specific individual.

Throughout the process of compiling the corpus, occasionally there were instances where the words were illegible. In these cases, later editions of the book and facsimiles were consulted in order to ensure the corpus was transcribed as accurately as possible. However, with historical texts there are times where it is very difficult to decipher whether there was an error made by the printer or the author. Howard-Hill (2006Howard-Hill, Trevor H. 2006 Early Modern Printers and the Standardization of English Spelling. London: Modern Humanities Research Association.Howard-Hill, Trevor H. 2006 Early Modern Printers and the Standardization of English Spelling. London: Modern Humanities Research Association.: 16) outlines the importance of comprehending that “early modern printers–unlike present-day printers–did not follow the spelling of their copies. Consequently, the printed works of spelling reformers and dictionary-makers used spellings that were not sanctioned by their authors”. Such variations in spelling could also occur within onomastic references. Additionally, different realizations of referring to the same person occurred. For these reasons, we decided to collect all instances of an author’s name as it occurred in the text during the extraction stage, and in later stages of the analysis, determine whether they refer to the same person.

3.3Working with historical (corpus) data

Diachronic data, by definition, means there is an element of change or stability occurring over the course of a set time frame. We therefore refer to Bakró-Nagy’s explanation on the specificity of historical linguistic data as:

the sum of statements made about a conceptionally described subset of linguistic phenomena described with specific preconceptions in mind. Data are inconsistent in the sense that their truth value is not constant, that is, they can change as a result of possible change in our knowledge of processes of change or of past states, or even induce processes of chain-like reinterpretation.(Bakró-Nagy 2010Bakró-Nagy, Marianne 2010 “The Data in Historical Linguistics: On utterances, sources, and reliability”. Sprachtheorie und germanistische Linguistik 20:2. 133–195.Bakró-Nagy, Marianne 2010 “The Data in Historical Linguistics: On utterances, sources, and reliability”. Sprachtheorie und germanistische Linguistik 20:2. 133–195.: 136)

Fischer addresses the question of whether data in historical linguistics should be considered in terms of formalist grammar change or language change and suggests that:

[a] historical linguist who bases himself too exclusively on a particular theory may come to suggest an interpretation of the historical data that can only be called an oversimplification of its complex nature or even worse can lead to the neglect of relevant and by no means incidental facts.(Fischer 2004Fischer, Olga 2004 “What Counts as Evidence in Historical Linguistics?” Studies in Language 28:3. 710–740. Fischer, Olga 2004 “What Counts as Evidence in Historical Linguistics?” Studies in Language 28:3. 710–740. : 713)

One of the main challenges of working with historical data is the lack of standardized lexical forms or structural norms. Aside from variations in spelling and diacritical marks i.e., macrons, use of blackletter font can make digitized copies difficult to visually inspect and interpret, particularly if the scans are of a lower quality.

Jenset and McGillivray (2017)Jenset, Gard B. & Barbara McGillivray 2017 Quantitative Historical Linguistics: A corpus framework. Oxford: Oxford University Press. Jenset, Gard B. & Barbara McGillivray 2017 Quantitative Historical Linguistics: A corpus framework. Oxford: Oxford University Press. claim that unlike contemporary texts, historical texts present an additional challenge of potentially missing metadata such as information about the author, title, publisher, publication date, etc. This is particularly important when drawing comparisons between different editions of the grammar books. Moreover, additional data about the authors themselves might show, for example, whether scholars with a similar background might refer more often to other authors, or not.

Focusing on a specific genre, in this case grammar books, meant that in addition to collecting information about the author’s academic endeavours, we also added information in a header above each grammar transcript about the target audience and what topics, i.e. which aspects of grammar, are covered in the book. Unlike literary works, the authors often addressed their intended readers in the preface and specified what the purpose of the book was. For instance, grammars could be addressed to younger audiences or adults, and the purpose of the text could be the use in a classroom or for self-learning at home. Additional information about the authors, such as their education and their date of birth and death, was collected in separate metadata files for each author in the corpus. All additional information on the authors and texts were collected using the relevant secondary literature, bibliographies such as the ones by Görlach (1998)Görlach, Manfred 1998 An Annotated Bibliography of Nineteenth Century Grammars of English. Amsterdam & Philadelphia: John Benjamins. Görlach, Manfred 1998 An Annotated Bibliography of Nineteenth Century Grammars of English. Amsterdam & Philadelphia: John Benjamins. and Alston (1965)Alston, R. C. 1965 A Bibliography of the English Language from the Invention of Printing to the Year 1800. Vol. 1. Leeds: Arnold and Son.Alston, R. C. 1965 A Bibliography of the English Language from the Invention of Printing to the Year 1800. Vol. 11. Leeds: Arnold and Son., and online sources such as the Eighteenth-Century English Grammars database online (ECEG) (Yáñez-Bouza & Rodríguez-Gil 2013Yáñez-Bouza, Nuria & M. E. Rodríguez-Gil 2013 “The ECEG Database”. Transactions of the Philological Society 111:2. 143–164. Yáñez-Bouza, Nuria & M. E. Rodríguez-Gil 2013 “The ECEG Database”. Transactions of the Philological Society 111:2. 143–164. ).

3.4Extracting onomastic references using corpus and network analysis

Once the data were fully transcribed and annotated, the annotation for onomastic references enabled an automated extraction of all references to personal names. This extraction process will be the focus of this methodological section. For this purpose, we have built a custom Python script, which will be made available on the project website once the corpus is completed. The onomastic references were automatically extracted using project-specific annotation together with a concordance window (150 characters to the left and right) and the source text from which it was extracted. An example of such an extracted concordance can be seen in Table 3. For each source text the original year of publication as well as the year of publication of the edition in the corpus precede the name of the grammar author, e.g. 1662–1662-Howell indicates that Howell’s grammar was originally published in 1662 and the edition in the corpus is the first edition. The concordance is extracted as in the original text and can therefore contain spelling errors, such as in the case in Table 3, where Ben Jonson’s last name is misspelled.

The resulting concordances were then manually coded by two independent raters, that is two members of the project team. Two rounds of inter-rater reliability coding were performed, first for the reference types and second for the person types. A normalized version of the person’s name was added in this stage as well. This workflow is visualized in Figure 1.

Many references had to be excluded from further analysis based on the following criteria:

• mentions of names within an example,

• mentions of the author’s, printer’s, bookseller’s name on the book cover,

• mentions of names recurring in a boilerplate or other structural element of the book,

• mentions of names in advertisements or other parts of the book that are not part of the original grammar text,

• ambiguous references that could not be clearly identified,

• or mentions of names in non-English language sections of the book.

These onomastic mentions are not considered as references. References to titles that were held by persons e.g. the King of England were excluded because they are not name-based and have been held by different people throughout the years. Thus, only those name-based references were considered which were presumably made intentionally by the authors themselves and which we were able to identify and analyse.

The grounded theory approach (Bryant & Charmaz 2007Bryant, Anthony & Kathy Charmaz 2007 “Introduction: Grounded Theory Research: Methods and Practices”. In The SAGE Handbook of Grounded Theory ed. by Anthony Bryant & Kathy Charmaz, 1–28. London: SAGE. Bryant, Anthony & Kathy Charmaz 2007 “Introduction: Grounded Theory Research: Methods and Practices”. In The SAGE Handbook of Grounded Theory ed. by Anthony Bryant & Kathy Charmaz, 1–28. London: SAGE. ) was used both when analysing the type of person referenced and establishing what type of reference was used, building on our previous work for the 19th-century grammar books (Busse et al. 2020Busse, Beatrix, Kirsten Gather & Ingo Kleiber 2020 “A Corpus-Based Analysis of Grammarians’ References in 19th-Century British Grammars”. Variation in Time and Space: Observing the world through corpora ed. by Anna Čermáková & Markéta Malá, 133–172. Berlin: De Gruyter. Busse, Beatrix, Kirsten Gather & Ingo Kleiber 2020 “A Corpus-Based Analysis of Grammarians’ References in 19th-Century British Grammars”. Variation in Time and Space: Observing the world through corpora ed. by Anna Čermáková & Markéta Malá, 133–172. Berlin: De Gruyter. : 143–144). The extraction of concordances enabled us to analyse not only who the persons being referenced were, such as Benjamin Johnson in Table 3 or other figures such as Cicero, God, and John Milton. It also provided us with the context in which they were being referred to, which would not have been possible had we only worked with search terms. The independent raters conducted two different assessments — for both the persons and how they were being written about. These assessments were then quantified using inter-rater reliability (IRR) scores. Both Cohen’s Kappa and Krippendorff’s Alpha were used for the evaluation. These metrics allow for a quantification of agreement, i.e. whether the raters applied the same categories consistently. A high score in these metrics therefore ensures applicability and robustness of the categories. After evaluating the IRR scores, those categories where the raters’ allocations did not align were re-evaluated within the team according to the category descriptions, resulting in the final categorization of authors and references. ’

The obtained frequency and category information is represented in network graphs (see Figure 2 and 3 below), which then allow for further analysis. This representation stores the number of references made from the sources to the targets, as well as the attributes of the targets and the connections. In this case, the graphs represent the number of references made from the grammar authors in the corpus to the persons being referenced as well as the information about the person and references types. The representations were achieved by processing and visualizing the data using a custom R script, which takes a so-called intercitation matrix as input. The scripts for reference extraction and network visualization can be found in the appendices. The citation information is processed using the igraph package (Csárdi et al. 2025Csárdi, Gábor, Tamás Nepusz, Vincent Traag, Szabolcs Horvát, Fabio Zanini, Daniel Noom & Karsten Müller 2025 igraph: Network Analysis and Visualization in R. R package version 2.1.4. Csárdi, Gábor, Tamás Nepusz, Vincent Traag, Szabolcs Horvát, Fabio Zanini, Daniel Noom & Karsten Müller 2025 igraph: Network Analysis and Visualization in R. R package version 2.1.4. ). Both the width of the edges as well as the size of the nodes have been linearly scaled for legibility. The positionality of the nodes is determined by a physics force calculation. This is employed to ensure legibility only, meaning that there is no further purpose to the nodes’ positions.

4.1Categories of persons

A first result of this study is the categorization system of the different kinds of persons or figures referred to by the grammarians in the data set. The seven person categories that emerged from the data are listed and described in Table 5.

It is interesting to note here which types of figures were considered authoritarian on the use of the English language. Not only are grammar authors, poets, and literary authors referred to, but also famous scholars from other disciplines as well as famous orators, politicians, and religious figures. We will see in the following sections that these other scholars and famous figures are even more prominently referenced than linguistic or literary scholars.

4.2Network visualization

The network in Figure 2 displays the references made by the five grammarians in the 16th- century component of the corpus, while the network in Figure 3 displays the references made by the grammarians in the 17th-century component of the corpus.

The rectangular nodes represent the grammars in the corpus, while the size of their node is normalized to provide a representation of the size (i.e., word count) of the grammar text. The circular nodes represent the authors and figures referred to in the grammars. The colours of the nodes refer to the types of persons while the colours of the edges refer to the reference category applied. Thicker edges display larger amounts of references of this type made to this figure. Please note, that this network visualization introduces some room for misinterpretation. For instance, there is no meaning to the positionality of the nodes. Also, larger nodes do represent longer grammars or more references to certain authors, but these sizes are normalized, i.e., a node of double size does not represent a grammar twice as long or twice as many references made. Nevertheless, the network is a meaningful representation of the data, as it allows the observation of patterns such as interconnectedness, centrality, and isolation, as will be elaborated in the following section.

4.3Network observations

Figure 2 clearly shows that two of the 16th-century grammar authors, Richard Sherry and Richard Mulcaster, are the main contributors to the reference network. The other grammars contain very few, in the case of William Bullokar and Edmund Coote, or even no onomastic references at all, in the case of Gabriel Meurier. In that sense, Bullokar and Coote are isolated, Meurier does not occur, and Sherry and Mulcaster are central in the network.

It is also interesting to note that, just as was expected in advance, none of the grammar authors refer to each other. Generally, little references are made to other grammarians of any language. The authorities referred to most in these texts go back to ancient scholars, politicians, and other orators, who are particularly famous for their eloquence in the Latin tongue. Similarly, this can also be seen by means of indegree centrality, which counts the number of incoming edges to a particular author node (Prell 2012Prell, Christina 2012 Social Network Analysis: History, Theory, and Methodology. London: SAGE.Prell, Christina 2012 Social Network Analysis: History, Theory, and Methodology. London: SAGE.: 99). Scaled from zero to one, the ten highest indegree centrality scores throughout the 16th-century network are presented in Table 6.

The scores again confirm our observations. Across all five grammar books considered here, the most central figures of reference were not grammar authors, but rather famous orators such as Cicero, political figures such as Caesar, or famous scholars from ancient Greece or Rome, such as Quintilian or Plato.

Closer inspection of the 16th-century network further indicates that Sherry and Mulcaster are not only the main contributors to the network, but they also display the most prominent interconnectedness, i.e. they are connected through persons they both refer to. Both authors frequently reference Caesar, Quintilian, and Cicero, whom they consider great orators and rhetoricians. However, the network also implies that both grammarians considered different figures as authoritative with regard to language, since both exhibit clusters of persons that none of the other grammarians refer to.

The indegree centrality scores for the 17th-century component of the corpus (Table 7) shows a similar image. Rather than including grammar authors or contemporary scholars, most persons referenced are political figures, poets and literary authors from ancient Rome or Greece, such as Cicero, Seneca, Ovid, and Virgil. This finding confirms that 17th-century grammars of English were still largely impacted and influenced by the Latinate tradition.

The network of the 17th-century references in Figure 3 further displays similar patterns as the one for the 16th century in Figure 2. Most grammarians in the 17th-century data refer to only few other persons, while two grammarians stand out by referring to many different figures, namely Hewes and Wilkins. We can also observe that, as before in the 16th century, there are only few persons that are referred to by multiple grammar authors. Instead, for most grammarians, there is a cluster of persons only they refer to while none of the other authors do.

Quantifiable scores, such as the centrality scores above, enable us to quickly recognize the most central persons referred to within the network. Especially once the networks grow even larger and more complex, quantitative measures will be crucial in order to track influence and referencing activities.

4.4Frequency observations