1 Introduction
How does linguistic variation at the level of the individual language user interact with language change at the community level? Is there any empirical evidence that an individual's grammar – or more specifically, their understanding of how to use morphosyntactic constructions – can change past adolescence, and are such changes necessarily in line with the community trend? And is there any evidence that certain morphosyntactic constructions are associated in the individual's mind, so that change in one construction can influence the development of the other? In a nutshell, it is these pertinent questions that Lynn Anthonissen sets out to tackle with her first monograph, Individuality in Language Change.
Anthonissen's book is only one of many interesting studies resulting from the ERC-funded Mind-Bending Grammars project at the University of Antwerp, and can be situated more generally as part of a relatively recent wave of historical (socio)linguistic studies that aim to understand and model the interaction of community-level language change and individual-level language use. Yet Anthonissen's work still stands out because she manages to address the topic in such a way that it combines a relatively large-scale quantitative analysis (of thousands of tokens, attested in the linguistic output of thirty to fifty different authors) with a qualitative depth that is otherwise perhaps most commonly found in corpus-linguistic analyses of a much smaller scale.
The book is divided in four parts: ‘Part I: Introduction’, ‘Part II: Language change and special passives’, ‘Part III: Connecting individual and community grammars’ and ‘Part IV: Summary and conclusions’. In what follows, I will provide a description and evaluation of each chapter in order of appearance in the book.
2 Parts I and II: setting the scene
Chapter 1, ‘Introduction’, starts by explaining that, while it has long been common practice in historical linguistics to study language change solely at the community level, our understanding of how linguistic change progresses may in fact benefit from giving individual language users and the role they play in the process ‘the attention they deserve’ (p. 3). The introductory chapter not only introduces the two types of special passive constructions the book focuses on, but also sets out the goals of the book, which all in some way relate to the question how ‘variation and change at the individual level interact[s] with change at the community level’ (p. 7). One aspect of addressing that question is investigating whether the locus of morphosyntactic change is, as argued by more generative accounts, situated in the individual's childhood, or, as argued by usage-based accounts, whether an individual's understanding of how morphosyntactic structures work is continually updated, including during adulthood. Another aspect is looking into whether community-level patterns, such as a correlation between the frequency of two constructions that are presumed to be cognitively associated with one another, can also be shown to take effect in individual usage.
The second part of the book (chapters 2–4) starts with chapter 2, ‘A theory of language change’, which provides further motivation for the study into individuality in syntactic change. This chapter constitutes, in many ways, the backbone of the study, as it introduces the theoretical frameworks and concepts against which it should be understood. The chapter picks out grammaticalization theory, (cognitive) construction grammar and historical sociolinguistics as three research traditions that have substantially informed Anthonissen's work, but have left some of its central research questions underexplored.
For grammaticalization theory and to some extent also construction grammar, Anthonissen points out that ‘the reality of individual grammars is not contested, but has not been thematized either’ (p. 17). Yet grammaticalization – and constructional change more generally, including the changes scrutinized in this book – is believed to be driven by processes such as reanalysis, analogy and inferencing, which crucially ‘originate in the minds of individuals’ (p. 17). Hence, empirical research that examines individual-level language use is not only compatible with, but also a necessary addition to grammaticalization and constructionist theory. Historical sociolinguistics, by contrast, has ‘periodically explored how individual variation relates to sociolinguistic structure’, despite Labov (Reference Labov2006: 5) famously claiming that ‘the individual does not exist as a linguistic entity’. What sets Anthonissen's study apart is that (most) extant work in historical sociolinguistics has focused on linguistic change where new variants compete with (and ultimately replace) older variants, thus largely ignoring constructional changes that are not competition-driven, such as the one considered in this book. Still, Anthonissen elegantly argues that (historical) sociolinguistic research has helped provide support for some of the core assumptions of cognitive construction grammar, which is committed to ‘a usage-based model of language’ where grammatical knowledge is considered malleable even in adulthood.
Furthermore, the second chapter offers a careful description of the constructionist conceptualization of linguistic knowledge as ‘a hierarchical and dynamic associative network’ – and introduces Schmid's (Reference Schmid2020) Entrenchment-and-Conventionalization model (EC-model), which Anthonissen considers to be ‘one of the first serious attempts at fully integrating the individual dimension into a comprehensive, usage-based theory of language’ (p. 32). Chapter 2, in sum, makes it clear that both cognitive construction grammar and the EC-model will play an important role in the book, as will become evident in later chapters.
Following chapter 2, Anthonissen surveys extant literature on the historical development of the English passive in chapter 3, ‘Systemic change and the rise of the passive’. Subsequently, in chapter 4, ‘The development of special passives’, she narrows her scope further to discuss the rise of the prepositional passive ([SBJ be V-ed (X) P], e.g. She was made fun of) and the ‘Nominative and infinitive’ ([SBJ be VPCU-ed to VINF], e.g. She is rumoured to be rich), which serve as the main subjects of the empirical part of the book. Both chapters offer in-depth overviews of the history of English (special) passives, and critically assess some established ideas (such as the role of reanalysis in the rise of the prepositional passives). A very minor note that could be left here is that chapters 3 and 4 engage only sparingly with the central theme of individuality in language change. Both chapters are, however, supported by several helpful summary/discussion subsections that should help prevent the reader from getting lost in the details.
3 Parts III and IV: community trends, inter-speaker variability and lifespan change in special passives
If sharp distinctions have to be drawn, one could say that the literature survey ends after part II, and the empirical chapters start in part III (chapters 5–8). And indeed, the third part of the book opens with a description of the corpus data in chapter 5, ‘The EMMA corpus’. This chapter is an adapted version of Petré et al.'s (Reference Petré, Anthonissen, Budts, Manjavacas, Silva, Standing and Strik2019) introductory paper on the Early Modern Multiloquent Authors (EMMA) corpus, which includes the written works of fifty authors from five generation cohorts born in the seventeenth century. It is, however, not entirely accurate to draw such a sharp distinction, as the later chapters in part III still engage thoroughly with the theoretical and empirical literature surveyed in part II as well as other empirical studies that do (or do not) align with Anthonissen's findings.
Chapter 6 presents two predominantly quantitative analyses. The first analysis focuses on 5,701 prepositional passives found in samples of the works of all fifty authors included in EMMA. Here, Anthonissen makes a number of observations; she notes, for example, that despite there being quite some heterogeneity in the exact usage rates of authors in each generation cohort, there is still a clear pattern of generational change where ‘successive generations of writers advance the frequency of the prepositional passive beyond that of the earlier born cohort’ (p. 141). Further observations that align quite well with previous historical sociolinguistic work also suggest that periods of change in the usage rate of a particular construction seem to go hand-in-hand with high inter-speaker variability, and the rate with which an individual uses syntactic constructions can, but need not, change across their lifespan (most commonly in the direction of the larger, community-level trend; see, for instance, Baxter & Croft Reference Baxter and Croft2016).
The analysis then continues by looking into the productivity and degree of schematicity of the prepositional passive per author, as Anthonissen points out that different authors may have a different understanding of which verb types the prepositional passive may extend to. Investigating such a claim is, however, a substantial challenge. Indeed, as Anthonissen aptly points out (p. 155), when sample sizes differ in number of tokens per author, simple measures such as type frequency counts or type–token ratios are deceptive, and ‘an author's oeuvre may not reflect the full extent of linguistic knowledge this author has’ (p. 153). Still, the chapter manages to make some convincing claims by relying on dispersion measures, which show that the prepositional passive is more evenly distributed across different verbs (and can hence be considered more schematic) for later-born authors. It is also suggested that individual authors may also have idiosyncratic semantic preferences in the verb–preposition (V–P) combinations they passivize, but these findings should be treated with some caution: the statement that an author such as Thomas Pierce (compared to other authors) has a preference for prepositional passive types that convey judgement and approval (e.g. look (up)on, judge of, approve of) does not directly consider that Pierce may be more likely to write about these topics in general, and perhaps exhibits a similar preference for these verbs in other constructions. That there may be more to consider here is at least partially confirmed later, in the second analysis of chapter 6, which looks into a whopping 30,521 V–P combinations that can be used in both active and passive constructions. Here, it is pointed out that to some extent ‘the frequency of a V–P combination in the prepositional passive is correlated with the general frequency with which V and P occur’ (p. 161), and ‘some authors show a moderate correlation between their use of [idiosyncratic] V–P [combinations] in the prepositional passive and their general usage frequency of that collocation in the active’ (p. 163).
Finally, while there is nothing about chapter 6 that does not interest me, the most appealing sections are presented at the end, when Anthonissen discusses the ‘effects of the social landscape’ in which the authors are situated (pp. 164–72). Granted, subsection 6.5.1 ‘Clustering usage profiles’ could have contained some more detailed methodological information on how the numerical information that served as input for the authors’ usage profiles was standardized and fed to the clustering algorithm, and the results of subsection 6.5.2 ‘Correlating usage and social network proximity’ are inconclusive. Yet Anthonissen does present a type of quantitative (and data-driven) investigation of well-known sociolinguistic concepts (such as communities of practice and social network proximity) that is still far too rare in historical corpus analyses, which makes these sections a particularly interesting read.
Turning to the second type of special passive this book sets out to study, chapter 7 (‘The nominative and infinitive’) follows a structure that is comparable to chapter 6. First, after a brief introduction, the data sample on which the quantitative analysis presented later in the chapter is based is discussed. Here, no less than 10,901 attestations of the nominative and infinitive construction (NCI) were retrieved from the written works of thirty different authors (six per generation cohort). These examples were subsequently manually classified into different semantic-pragmatic usage types. Notably, the discussion of the semantic-pragmatic usage types engages very thoroughly with previous literature on the history of the English NCI (e.g. Noël Reference Noël2008), carefully arguing that ‘one should distinguish between two rather than three usage types of the NCI’ (p. 198). The chapter then continues to examine patterns of variation and change in how the thirty selected authors use the NCI, both in terms of overall frequency shifts (which appear to be less frequent overall, and usage frequencies are more likely to increase with authors who write in specialized genres such as science, medical or legal texts) and in terms of which usage type of NCI individual authors prefer.
Finally, chapter 7 continues to show that there is also substantial inter-speaker variability in usage type preferences for specific verbs, and presents a more qualitative analysis of these lexically specified NCIs with say (e.g. It is said to be a lie). Based on this analysis, Anthonissen argues that ‘speakers may recognize and exploit the potential of particular uses of a construction that are licensed by the community but [are] not particularly representative of the construction per se, especially if – as in the case of the say-NCI – interspeaker variability is great’ (p. 215). At this point in the book, then, it has become abundantly clear that even within the relatively homogeneous group of authors represented in the EMMA corpus, it cannot easily be maintained that ‘individual variation is reduced below the level of linguistic significance’ (Labov Reference Labov2012: 265), and that the relation between individual-level and community-level change may not be capturable by relying on a simple binary distinction of ‘stability’ and ‘change’.
After having been treated separately in their own chapters, the two special passive constructions are finally brought together in chapter 8, ‘Networks in adult cognition’. Using a data set of 3,556 prepositional passives and 2,602 NCIs found in the written works of thirty authors, Anthonissen reveals that ‘authors who use the NCI frequently will also use the prepositional passive frequently, and vice versa’ (p. 225), and that the degree of dispersion/schematicity of the NCI also correlates with that of the prepositional passives for individual authors (p. 233). The special passives can, in short, be considered closely associated nodes in the mental grammars of the authors under scrutiny. Interestingly, further quantitative analysis also reveals that the association between the special passives persists for all but four authors, who appear to dissociate them as they age. And as was the case in the previous two chapters, the quantitative analyses are again also nicely complemented by more qualitative observations on why certain authors, such as Robert Boyle, appear to lead the use of special passives, while others, such as Nathaniel Crouch, are ‘deviant and unpredictable’ (p. 228).
In terms of its contents, chapter 8 contains some of the most interesting empirical findings (with potentially important theoretical implications) presented in the book, but it also requires the most familiarity with (cognitive) constructionist theory. Up until chapter 8, the case studies in part III indeed refer to theoretical concepts introduced in part II, with the dominant focus perhaps being on concepts relating to (the development) of the special passives and, to a slightly lesser extent, constructionist concepts such as productivity and schematization. In chapter 8, however, the scales shift, and the previously discussed network model of cognitive construction grammar and Schmid's discussion of Entrenchment (and EC-model) are brought to the fore again. Towards the end of chapter 8 – and also in part IV of the book, which consists only of chapter 9, ‘Conclusion’ – great care is given to explaining the relevance of the constructionist conceptualization of grammar as a network of symbolically, syntagmatically and paradigmatically associated constructions to understand the results of the empirical case studies (and vice versa).
4 Conclusion
Returning to the list of questions this book review started with, I believe it is a fair assessment of Individuality in Language Change to say that it delivers on what it promises to do: it gives its audience a glimpse into how linguistic variation at the level of the individual language user interacts with language change at the community level, offers some empirical evidence that morphosyntactic constructions can be subject to lifespan change (most likely in the direction of the community trend), and shows that (changes in) special passive constructions are associated in the individual's mind. It should be noted, however, that while there is a lot of emphasis on the notion of individual language, there is much more to gain from this book. In fact, its title, Individuality in Language Change, may not fully do justice to its contents. On the one hand, the title is perhaps too narrow in scope, as several of its chapters offer a detailed overview of the development of English passive constructions, and how they are best analysed in historical corpus data. Thus, the relevance of this book in fact transcends the domain of individual language. On the other hand, the title's reference to ‘language change’ makes its scope also a tad too large, and undersells the fact that the book is one of the first elaborate attempts to home in on morphosyntactic lifespan change (rather than lexical and or phonological change) in such great detail.
To conclude this review, I briefly wanted to highlight one of the main reasons why I found myself increasingly interested in focusing on individual-level language variation and change. As Anthonissen briefly points out in the introductory chapter, there is a divide between usage-based and generative models of language change: in the latter, it is suggested that morphosyntactic change past adolescence is unlikely or even impossible. Yet, while Anthonissen elegantly demonstrates the merits of adopting a usage-based perspective, her findings do not necessarily prove that the assumptions of the generative model are wrong (and accurate as Anthonissen tends to be in her assessment of her results, she also acknowledges this). In fact, generative and usage-based models of morphosyntactic change alike agree that frequencies of linguistic forms can change even in adulthood, which is what Anthonissen's findings also clearly demonstrate. Where the two approaches disagree, however, is that morphosyntactic lifespan change can also be qualitative or ‘systemic’ in nature (e.g. Meisel et al. Reference Meisel, Elsig and Rinke2013). Unfortunately, such systemic lifespan changes have not been attested through the case studies presented in this book (for an example of qualitative/systemic change, see e.g. Fonteyn & Petré Reference Fonteyn and Petré2022). Yet Individuality in Language Change does manage – and this is why it should find its way to your bookshelf – to provide an extensive discussion of the theoretical background against which the importance of such quantitative and qualitative lifespan shifts should be understood, while also presenting a set of case studies that can be considered exemplary of how research questions on morphosyntactic variation and change at the individual level should be approached in corpus linguistics.
