The aim of the editors of this volume is to outline possible future paths for the still young field of historical sociolinguistics, an area that has been very much shaped by the ground-breaking work of Terttu Nevalainen and Helena Raumolin-Brunberg. The point of departure is a programmatic article by Terttu Nevalainen (‘What are historical sociolinguistics?’, Nevalainen Reference Nevalainen2015) that highlighted the notion of ‘layered simultaneity’ for historical sociolinguistics, that is, the availability of various micro- and macro-levels of contextual meaning at any given point in time and the different ways in which they spread across time and space.
The chapters in this volume set out to materialise this notion with regard to methodological innovations, the analysis of new (that is, un- or underexplored) data, and the development of new theoretical approaches. Accordingly, the chapters fall into three sections, focusing on methodology, new data and theory.
The emphasis of the four chapters of the first part is on methodological innovations. Layered simultaneity can here be seen as various ways of combining and visualising linguistic data and sociolinguistic metadata. In ‘Exploring part-of-speech frequencies in a sociohistorical corpus of English’, Tanja Säily, Turo Vartiainen and Harri Siirtola test part-of-speech frequencies as a tool for studying genre evolution, sociolinguistic change and colloquialisation in the Parsed Corpus of Early English Correspondence, covering the period between c.1410 and 1681. The basic idea behind this study is that part-of-speech frequencies can be seen as an indicator of structural complexity and a more or less involved style. The authors can actually demonstrate a trend towards colloquialisation that is led by the gentry. This finding is combined with strong evidence for gendered variation with regard to the proportions of nouns and personal pronouns. However, from a methodological perspective, the results of part-of-speech frequencies prove to be rather mixed. While they can, in principle, be used for measuring colloquialisation and gendered styles, they are – at least with the present parts-of-speech categorisation – not reliable and detailed enough to go beyond a heuristic tool and must be complemented by other methodological tools (for example, close reading).
Susan Fitzmaurice, Justyna A. Robinson, Marc Alexander, Iona C. Hine, Seth Mehl and Fraser Dallachy (‘Reading into the past’) report on the so-called Linguistic DNA project. The methodological innovations of this project evolve from the combination of the fields of historical semantics / lexicography and corpus linguistics, making use of the huge text database EEBO-TCP in combination with the Historical Thesaurus. Basically, the Linguistic DNA project focuses on a data-driven and bottom-up study of conceptual change during the Early Modern period. The data-driven nature of the investigation is shown by the fact that the conceptual entities studied (the so-called ‘discursive concepts’) are not constructed on lexemes but are the result of statistical analyses based on sets of associated words, mostly co-occurrence clusters. The emphasis of the article is on three case studies, with rather preliminary results that mainly point out future improvements of the software but also the range and potential of future applications (especially for historical sociolinguistics). For example, starting from the Thesaurus category ‘virginity’ and its near-synonyms, the authors work out a general discursive concept that covers the association of unmarried people and marriage, but also, in the context of the particularly religious Early Modern discourse, the more specific discursive concept ‘virgins in marriage to Christ’. The project seems to be still in its infancy but it will certainly be an exciting task to compare the computationally developed discursive concepts with the categories of the Historical Thesaurus.
In the third chapter in this section, ‘Ireland in British parliamentary debates 1803–2005: Plotting changes in discourse in a large volume of time-series corpus data’, Helen Baker, Vaclav Brezina and Tony McEnery approach another huge data base, the Hansard Corpus of British parliamentary debates, presenting two new statistical techniques for studying and visualising historical data, Meaning Fluctuation Analysis and sparklines. The first technique measures changing collocations of a given item and shows points where the number of shared collocations decreases and meanings drift apart. These points can be taken as indicators for shifts of discourse. Sparklines are a convenient tool that give (visual) summaries of individual collocation frequencies over time. In contrast to the first tool, they are mainly useful as a shorthand display when referring to individual collocates. The aim of the present case study is to trace the collocational change of the word Ireland in the Hansard Corpus. The Meaning Fluctuation Analysis tool is taken as an objective heuristic indicator for collocational change and – in this case – changes in parliamentary discourse about Ireland. In fact, the authors can identify two major troughs in the graph, which they can closely associate with two historic events, the Great Famine (1845–8) and the establishment of the Irish Republic (c.1920–40). Checking the individual concordance lines, they show that parliamentary discourse reveals many of the topics associated with the historical background of the two events. Closer analysis reveals that the collocates shown by the first trough (associated with the Great Famine) are rather short-lived, whereas the second trough (Irish Republic) seems to have produced a couple of long-term collocations (for example, northern and southern). The authors can also nicely demonstrate that the troughs in the graph produced by the Meaning Fluctuation Analysis in no way coincide with the frequency of the term Ireland in the corpus. One advantage of this method seems to be that, once a relevant point for discourse change has been indicated by the software, the individual concordance lines for the years can be checked qualitatively.
Minna Nevala and Anni Sairio (‘Discord in eighteenth-century genteel correspondence’) look at discourse about discord in eighteenth-century correspondence, using the Corpus of Early English Correspondence Extension and the Bluestocking Corpus. Their procedure is largely qualitative analysis. Starting with the relative frequencies of politeness and discourse words in the corpus, they look at the social meaning of these terms and how conflict situations were described and evaluated.
The second section of this collection presents studies about new (that is, un- or underexplored) data for historical sociolinguistic research. In ‘Competing norms and standards: Methodological triangulation in the study of language planning in nineteenth-century Finland’, Taru Nordlund and Ritva Pallaskallio look at standardisation processes in nineteenth-century Finnish. As a basis for their analysis, they offer fascinating new data: short letters sent by readers to Finnish newspapers, with the linguistic changes and ‘corrections’ made by the editors before they published the letters. In addition, the authors use the meta-level discussions found in nineteenth-century grammars and language guides plus data from published newspapers in nineteenth-century Finland. In two case studies involving Finnish nominal morphology (variation in the morphological stem and allomorphy in a case suffix) they show that the newspaper editors were astonishingly conservative in their revision processes and that the resulting newspaper data reveal changes in morphological form that had advanced much more than contemporary prescriptive discourse might suggest. Nordlund and Pallaskallio conclude that morphological change may tend to be determined by language-internal factors that language users are seldom aware of.
‘Relativisation in Dutch diaries, private letters and newspapers (1770–1840): A genre-specific national language?’, by Andreas Krogull, Gijsbert Rutten and Marijke J. van der Wal, presents Dutch diaries and private letters from the eighteenth and nineteenth centuries, which the authors compare to contemporary published newspaper texts (the Going Dutch Corpus, newly designed to investigate the interplay of language planning and patterns of variation and change). The corpus patterns not only according to genre but also according to different regions, with an additional differentiation between centre and periphery. The authors investigate relativisation in Dutch, with a special focus on the neuter relative pronoun and against the background of language planning and prescriptivism in the Netherlands around 1800. Their results are quite surprising. It turns out that language-internal factors (definiteness of the antecedent), normative discourse and the distribution across region and gender cannot satisfactorily account for the usage pattern found in the data. It seems that genre is the decisive factor, with letters and diaries showing a great difference in patterning. This distribution suggests a strong contrast to the idea of a homogeneous national language that was propagated in prescriptive discourse about eighteenth- and nineteenth-century Dutch.
The third contribution in the data section (‘“A graphic system which leads its own linguistic life”? Epistolary spelling in English, 1400–1800’) shows how small, carefully transcribed manuscript corpora of letters can be used to test the reliability of the large Corpus of Early English Correspondence with regard to the study of orthography. As is well known, the CEEC is mostly based on edited editions of letters, where any editorial interference cannot be excluded with absolute certainty. This makes large-scale investigations of spelling potentially unreliable. Samuli Kaislaniemi, Mel Evans, Teo Juvonen and Anni Sairio focus on two frequent variables, <l/ll> (in always and in formations with -ful) and < ie/ei> (in receive, believe and friend), tracing long-term trends in the history of English private spelling and comparing these results with the smaller manuscript corpora. It turns out that in the fifteenth and sixteenth centuries idiolectal variation prevails, followed by change determined by generation in the seventeenth century. It is only in the eighteenth century that sociolinguistic analysis becomes possible, when the forms of the emerging standard can be contrasted with non-standard spellings against the background of the available sociohistorical data. Interesting insights are, among others, that men are ahead of women (with gender differences vanishing at the end of the period), that middle- and lower-class informants generally seem to follow the emerging norms more closely and that London leads the standardisation process in every period. The comparison with the small corpora by and large corroborates the CEEC results. So it seems that the CEEC is a fairly reliable tool for the study of long-term trends in high-frequency orthographical features. As with the other two chapters in the data-centred section, this contribution not only offers new and interesting resources but also produces novel and fascinating insights.
The last section of the book deals with theoretical aspects, under the heading of bridging gaps and new challenges. Martin Hilpert's chapter, ‘Historical sociolinguistics and construction grammar: From mutual challenges to mutual benefits’, seeks to bridge the gap that is usually perceived between the fields of historical sociolinguistics and construction grammar. For example, whereas construction grammar aims at the individual speaker, speaker knowledge and the functional and formal properties of constructions, historical sociolinguistics is primarily concerned with groups of speakers and processes and forces that shape language change in a communal setting. Hilpert seeks to transform the theoretical and methodological differences into mutual challenges, which then should turn out to be mutual benefits. For example, analysing the form and meaning of a construction could be supplemented by consideration of its use in conversational practice, in varying contexts, registers and varieties, tracing its spread across different domains in space and time. This enriched analysis is illustrated with a corpus-based study of an idiomatic scheme usually called the sarcastic much? construction.
In ‘A lost Canadian dialect: The Ottawa Valley 1975–2013’, Bridget L. Jankowski and Sali A. Tagliamonte analyse two vernacular features (verbal -s in the verb BE with third-person plural subjects and preterite come) in Ottawa Valley English, a rural Canadian dialect in recession. They are interested in the development of the two features as the rural dialect faces increasing encroachment of standard Canadian English. The study comprises a combination of two trend studies that were conducted from 1975 till 1981 and 2012 till 2013 respectively, covering informants who were born in the periods between 1884 and 1928 and 1917 and 1969. The combination of the two studies allows an interesting comparison between real-time and apparent-time investigations. The results show that the two features develop quite differently. Verbal -s is slowly dissipating; in the second study it is found only in fossilised contexts, with no social stratification. Preterite come, on the other hand, is still maintained and shows age-grading in both corpora. It seems that older people use the form more frequently as a marker of social networks and of identification. The authors conclude that verbal -s reveals a long-term stagnation of a dialect feature, with a neutralisation of social features and a concentration of linguistic constraints, whereas preterite come prevails as a robust feature with local and rural associations.
Lieselotte Anderwald (‘“Vernacular universals” in nineteenth-century grammar writing’) looks at what she calls the ‘non-standardisation’ of some vernacular universals (multiple negation, adverbs without -ly, the construction you was and constructions with there is/was with plural subjects) in nineteenth-century British and American prescriptive grammar books, tracing the various processes of stigmatisation. In her corpus-study she arranges the negative epithets according to the categories ‘education’, ‘mistake’, ‘social evaluation’ and ‘logic’. The interesting results show that the features are stigmatised differently and to a different extent (with multiple negation in first position and there is/was last) and that American grammar books stigmatise more than British grammars.
The last contribution, ‘Revisiting weak ties: Using present-day social media data in variationist studies’ by Mikko Laitinen, Jonas Lundberg, Magnus Levin and Alexander Lakaw, addresses present-day non-native English in the Nordic region. Its innovative aspect is the application of present-day social media in a sociolinguistic study involving the traditional network model. Looking at the user data included in micro-blog messages on Twitter (covering about 200,000 informants), the authors have access to a range of small ethnographic networks, which can be defined as weak or strong(er). In this framework, they can show that innovations spread not only in weak-tie networks but also in networks showing slightly stronger ties (if they comprise at least 100 to 130 users). The authors invite researchers to apply the innovative use of big and often freely accessible data to sociolinguistic studies, both synchronic and diachronic. It seems to me that the application in historical sociolinguistics is rather problematic since this method is very much limited to the contemporary conditions of social media, and what they call innovations are not clearly related to linguistic variables in sociohistorical linguistics.
Reviewing the whole array of the contributions collected in this volume, the first reaction that comes to mind is astonishment at the sheer range of different topics, methods, languages and periods covered here. This certainly deserves great respect and shows what a large, innovative and heterogeneous field historical sociolinguistics has become. Generally, all the chapters represent a high level of expertise. They are well researched and well written, also considering possible limitations of the methodology and alternative hypotheses. They include instructive examples, adequate illustrations of statistical procedures and extremely attractive (colour) figures and graphs.
However, given the great diversity of topics, the thematic coherence of the three sections of the book (methods, data and theory) is not always maintained. It is most plausible in the first three contributions of the first section. But with the fourth contribution one may wonder what this particular study of discord in eighteenth-century correspondence specifically contributes to methodology (despite the interesting insights it offers). On the other hand, some of the chapters offering new data (for example, the studies on language planning in nineteenth-century Finland and on Dutch relativisation) contribute fascinating new insights on methodology and theory. The third section appears to be the least coherent, with only Hilpert's chapter offering a clearly theoretical focus. The rest is not as obviously innovative in theory, comprising a purely descriptive study on nineteenth-century grammar books, a fascinating empirical dialect study with insightful implications and a new methodology resting on social media. Having said that, it seems extremely difficult to me to find an alternative, more appealing arrangement. Perhaps this heterogeneousness is a sign that historical sociolinguistics is, after all, a very young field and still full of creative development.
What, then, may be future paths for historical sociolinguistics? Against the background of the present collection and Nevalainen's idea of layered simultaneity, the editors suggest some ‘key components’ (p. 15) for forthcoming developments in their introduction: first of all multidisciplinarity, that is, a focus on studies and approaches to language use that bridge the different levels of analysis and cross the boundaries of various disciplines. In addition, they invite multilinguistic cooperation, combining the results of sociohistorical analysis in various philologies and revealing similarities and differences of (historical) use in various languages. Lastly, they call for rich multiple data sources that help to reconstruct contexts of language use of past centuries. In fact, all these aspects are found in the present collection. And perhaps during the next decades these different paths may form (more) clearly discernible patterns.
