Historical pragmatics, as the study of language use of earlier periods, needs access to authentic language material of bygone eras. Apart from the last half-century or so, such material has only survived in written form. Spoken language is ephemeral and only survives if it is recorded. In view of this problem that lies at the heart of historical pragmatics, two schools of thought have recently developed. One approach takes a broad view and considers all forms of language pragmatically constrained in multiple and complex ways and therefore worthy of pragmatic analyses. Neither spoken nor written language, in this school of thought, is more deserving of linguistic analyses. All forms, both spoken and written, must be considered in their own complexities of producer intentions, addressee targeting and situational constraints. The second school of thought adopts the traditional predilection of pragmatics with the spoken language and tries to find historical evidence for forms of language that are as close as possible to the spoken language of the past. The spoken language is seen as primary. This new and important book by Culpeper and Kytö squarely sides with the latter approach. Its basic aim is to uncover some of the features of spoken face-to-face interaction of past periods, and, therefore, the nature of data and its relation to spoken language is of paramount importance.
In fact, the entire book is based on a corpus of data that was designed and compiled for exactly this purpose, the Corpus of English Dialogues 1560–1760. It consists of about 1.2 million words drawn from five different genres which are all in complex but fairly direct ways related to spoken face-to-face interaction. Three of them consist of constructed dialogue: comedy drama, didactic works in dialogue form and prose fiction, including a lot of direct speech. And two consist of authentic dialogue: trial proceedings and witness depositions. Culpeper and Kytö spell out a range of criteria that guided them in the compilation of their corpus. They favoured texts that contained direct speech, rather than indirect speech. This means that some types of witness depositions, for instance, had to be excluded while others could be included. Witness depositions concerning shipping, for instance, were mainly concerned with factual information and did not include direct speech, whereas witness depositions concerning witchcraft accusations frequently consisted of direct speech. A second principle excluded texts in verse and, therefore, prevented many comedy plays from being included. The additional criteria concerned attempts to achieve a broad range of social ranks, including sufficient data from both genders, and the texts had to be precisely dateable into one of the five forty-year time slots of the overall time-span of the corpus. Each of the five genres is represented with a word count between roughly 170,000 and 185,000.
The corpus is split into two main types of dialogue: constructed and authentic. Constructed dialogue consists of fictional dialogue invented by an author. Until recently, pragmaticists would not have considered such data at all because it is too far removed from the actual spoken language. The authors of such texts may have aesthetic or didactic principles in mind when they construct dialogue and they may generally be led more by what they think spoken language should be like, rather than what it really is like. Authentic dialogue, on the other hand, consists of real spoken language as it was recorded by a scribe for particular purposes, e.g. of a legal nature. This, too, may appear to be less than perfect because the scribes may have polished what they heard, i.e. they may have corrected speech errors, false starts, repetitions, hesitation markers and so on, and thus would have eliminated much of what present-day pragmaticists are interested in. Thus, neither the constructed dialogue nor the authentic dialogue represents actual spoken language of the past, but the texts compiled in the Corpus of English Dialogues probably get as close to the real thing as we will ever get. For the research objectives of this book such a corpus is crucial because Culpeper and Kytö are not satisfied with the investigation of some general pragmatic features in the English language of the past, but they want to investigate the genuine properties of the spoken language of the past. They pursue these objectives in the following chapters of the book by focusing on specific elements that are known to be typical of the spoken language of today.
The book is structured into four parts. The first comprises two introductory chapters which give an overview of the authors’ position about the spoken versus written language distinction and discuss the five genres of the Corpus of English Dialogues and their socio-historical contexts. The remaining three parts are devoted to the structure of spoken interaction, to what Culpeper and Kytö call ‘pragmatic noise’ and to the social dimension of spoken interaction. Each part starts with an overview chapter and then offers more detailed case studies.
Chapter 3, which opens the part on the structure of spoken interaction, is devoted to general issues connected with speech and writing in a historical context. It takes its inspiration from Koch and Oesterreicher, whose work goes back to the 1980s (for an English presentation of their approach see Koch Reference Koch, Jucker, Fritz and Lebsanft1999). According to Koch and Oesterreicher, it is important to distinguish between the conceptual continuum from the language of immediacy to the language of distance, on the one hand, and the dichotomy of the phonic and graphic realization of language, on the other hand. Such an approach helps to distinguish features that are said to be typical of ‘spoken language’, even if they are regularly attested in writing. According to Koch and Oesterreicher, such elements can be typical of the language of immediacy, and as such they are more likely to occur in, but are not restricted to, the phonic code. They can also be attested in the graphic code. The chapter teases out some of the complexities of the genres included in the corpus, such as the different levels of embedding and the different types of speech reporting in courtroom proceedings or in prose fiction.
The remaining chapters of this part of the book, chapters 4 to 8, deal with more detailed aspects of the structure of spoken interaction and writing.
Chapter 4 focuses on grammatical variation in spoken interaction. Culpeper and Kytö hold the view that spoken and written language share more or less the same grammar, but with considerable differences in the frequency of individual constructions. Chapter 5 turns to lexical bundles, which are – following Biber et al. (Reference Biber, Johansson, Leech, Conrad and Finegan1999: 990, 992) – defined as ‘computationally derived recurrent word-combinations’ (p. 103). Three-word bundles are extremely common and are related to collocational association, while bundles including more than three words are increasingly less frequent and more phrasal in nature. The five most frequent three-word bundles in the Corpus of English Dialogues are I do not, it is not, it is a, I pray you and out of the. Culpeper and Kytö point out that almost all of the most frequent three-word bundles include a verbal element. Among the five most frequent ones only out of the is an exception. What is particularly interesting in this chapter is the different frequencies of bundles in the different genres represented in the corpus. It is noteworthy that these findings were only possible because of a spelling regularization programme that Culpeper and Kytö used in order to make the Corpus of English Dialogues amenable to such analyses.
Chapter 6 is devoted to lexical repetitions, which are compared to present-day repetitions and their functions. While present-day repetitions have generally been analysed in terms of their intensifying and expressive functions, the repetitions in the Corpus of English Dialogues are considered mostly in terms of their interactional and interpersonal dynamics.
Chapter 7 deals with the cohesive function of and. It turns out that clause-level coordination and phrase-level coordination differ in their frequencies in the different genres. And is clearly multidimensional and multifunctional.
The last chapter of this part of the book, chapter 8, is devoted to grammatical variation, and in particular to the third-person neuter possessive its and its variants (of it and thereof); the prop-word one; and periphrastic do in negative declarative sentences. These are all elements known to have changed in the Early Modern period. Culpeper and Kytö, therefore, focus on the question whether speech-related language was a significant factor in these changes, but once again the situation is more complex. For the first two of them speech-related genres were not instrumental in the change, while these genres were at the forefront of the development of do.
The next part of the book, comprising chapters 9 to 12, is devoted to ‘pragmatic noise’, which includes ‘items such as ah, ha, hah, o, oh, ho, um, hum, as well as reduplicative forms like ha, ha or ha, ha, ha’ (p. 199). As such, pragmatic noise comprises many elements that are normally called ‘interjections’, but it also comprises laughter and pause fillers. Pragmatic noise also overlaps with what Biber et al. (Reference Biber, Johansson, Leech, Conrad and Finegan1999) call ‘inserts’. These are elements that one would normally not expect in written language. Pragmatic noise ‘is noise in the sense that the items have developed from natural noises, and, consequently, do not have homonyms in other word classes or always typical phonological structures; it is pragmatic in the sense that the items convey interpersonal and discoursal meanings’ (p. 222). It appears to be difficult to categorize such elements, even if only for presentational purposes. In chapter 10, Culpeper and Kytö suggest a classification on the basis of the vowel involved in the pragmatic noise, i.e. a-related forms ah, ay and alas, o-related forms oh and ho, e-related forms eh, he, hey and hem, and u-related forms um, huh and hum. An additional group contains elements with a word-initial fricative or plosive, e.g. fie, shu, pooh or tush. The classification is based on the hypothesis that these elements embody a certain amount of sound symbolism and that similar sounds, therefore, imply similar meanings.
Chapters 11 and 12 deal with the variability of pragmatic noise in terms of co-occurrence patterns, text-type frequencies and diachronic developments.
The final part of the book, chapters 13 to 15, focuses on the social dimensions of spoken interaction, and in particular on the social roles of the interactants and on gender differences.
The last chapter, chapter 16, finishes the book with a very useful summary and some concluding remarks. It provides an overview table that indicates which of the elements studied in this book are particularly characteristic for which of the five genres represented in the Corpus of English Dialogues. It turns out that none of the five genres is maximally speech-like on all counts, but play-texts apparently have the strongest claim to be more clearly speech-related than the other four genres.
This new book by Culpeper and Kytö is an important and outstanding contribution to historical linguistics. It provides rich insights into the spoken language of the past, and these insights are based on solid empirical evidence. The authors demonstrate what can be achieved through the analysis of a carefully designed corpus even if it is relatively small. The individual studies contained in this book all rely to a large extent on corpus linguistic methods, but here corpus linguistics is far more than just an analysis of pattern frequencies. In fact, throughout the book Culpeper and Kytö combine qualitative and quantitative methods in interesting and novel ways which provide a deeper understanding of the highly complex relationships between spoken interaction and writing. The book will inspire a lot of work in this area, both because of its exciting insights and because of the novel research tools that it introduces and exemplifies.
