This volume comprises a wide-ranging collection of papers on historical linguistics. It pays tribute to Professor Richard Hogg, and reflects the wide scope of Richard's research interests. The papers are empirical in approach, with a focus on methodological issues and innovations providing a coherent theme across the papers. This theme seems to be one of how to handle and interpret incomplete historical data, in order to avoid forcing historical data to fit analytical schema or constructs which are based upon the synchronic analyses of Present-day languages. Several papers make clear that viewing change from a present-day perspective can mislead us.
Consequently, many of the papers are concerned with questions of historical continuities or discontinuities, and with functional rather than formal or theoretically driven explanations of historical phenomena. They reappraise formally based analyses on the grounds that these analyses provide an inadequate fit to historical data. Instead, the volume makes a strong case throughout for viewing historical data in their own terms, from the perspective of the users of those historical varieties. Throughout, the papers are uniformly characterised by very careful and detailed philological scholarship, and clearly demonstrate the insights gained from such work. This seems to be entirely in keeping with the approach advocated and practised by Richard Hogg
The volume is divided into sections, with an introduction to each one highlighting common themes between the papers. This is largely effective. In order to discuss the volume in more detail, I will now discuss each section in turn.
The first section, ‘Metrics and onomastics in older English’, comprises three papers. Geoffrey Russom's paper (‘What explantory metrics has to say about the history of English function words’) is concerned with the question of metrical continuity between Old English Beowulf and Middle English Sir Gawain and the Green Knight. On the basis of a detailed metrical analysis of several texts, Russom argues that Old English metrical principles largely persist into Middle English, in spite of the increase in Middle English function words which might disrupt the metre. An increase in hypermetrical patterns at the end of the Old English period provides evidence for some adjustments to these metrical principles in order to accommodate the unstressed syllables provided by additional function words. However, additional function words do not lead to loss of the Old English metrical principles.
Richard Coates's paper (‘to þære fulan flóde. óf þære fulan flóde: on becoming a name in Easton and Winchester, Hampshire’) focuses on an area which has not received much linguistic attention: onymisation – the process by which a noun phrase with compositional semantic meaning may become a proper name, whose meaning is purely referential. Coates proposes a usage-based account of onymisation. However, this process has few linguistic correlates, making the mechanism difficult to observe and describe, and particularly difficult to date with any accuracy, especially, as Coates argues, as an expression may not become a proper name for all its users at the same time.
Peter Kitson (‘Notes on some interfaces between place-name material and linguistic theory’) examines the evidence of river names for patterns of migration. He argues that recurrent river naming patterns throughout Europe are too widespread to be the results of migration. Instead, he proposes that their distribution is evidence that these names have common European ancestors which pre-date the division of Indo-European into language groups. River names are shared between languages as they emerge, rather than being spread by migration or contact between various groups. Hence, they pre-date many of the patterns of migration previously thought to account for their distribution. The second part of Kitson's paper appears rather disconnected thematically from the first part, almost constituting a separate paper. Through a detailed analysis of just one word, herepaþ ‘army-path/main road’, which occurs frequently in boundary charters, he shows how Old English data can yield evidence of dialectal variation. Specifically, in the case of herepaþ, he argues that syncope of the medial <e> appears to be spreading from the midlands to the south. The prospect of extending this type of work to provide more linguistic evidence for dialect variation in Old English is an intriguing one.
The second section addresses writing practices in earlier English. Both papers in this section are concerned with what orthographic textual evidence can tell us about linguistic variation, and how such evidence should be interpreted. R. D. Fulk (‘Anglian features in late West Saxon prose’) examines textual evidence for Anglian features in Late West Saxon. He considers the possibility that late West Saxon is more heterogeneous than previously thought with many texts comprising a mixture of Anglian and West Saxon features. Anglian features in these texts may represent diatopic, stylistic or even social variation. Through comparison of parallel texts, Fulk argues that not all Anglian features are consciously manipulated by scribes, hence they are not stylistic devices. He also discounts the possibility of scribal practices introducing Anglian features into the manuscripts. Instead, he concludes that Anglian features tend to be eliminated rather than added over time in successive manuscripts of the same text. This is most consistent with many Late West Saxon texts originally being composed in an Anglian dialect.
Roger Lass and Margaret Laing (‘“ea” in early Middle English: from diphthong to digraph’) examine the distribution of the digraph <ea> following the monophthongisation of Old English ea and eo. Instead of disappearing along with the diphthongs it represented, the monophthongisation of Old English ea results in a floating <ea> digraph used to represent different long and short vowels. Through careful philological analysis of a selection of Early Middle English texts, Lass and Laing reconstruct a number of uses for <ea> to represent long and short vowels. However, they conclude that the use of <ea> is so variable across texts that while patterns can be observed, a full description of its use has to take account of different practices in different texts. They argue that the digraph <ea> comes to function as a diacritic, in which the <a> indicates that the vowel to its left is more open (e.g. <e> for [e] but <ea> for [ɛ]).
The third section concerns itself with investigating sociolinguistic and dialectal diversity in earlier English. The uniformitarian principle presupposes that such variation existed in earlier English. Hence this section focuses on the difficulties of applying present-day sociolinguistic and dialectal methodologies to earlier stages of English. These are areas in which the limitations of historical data – both limits on the overall amount of textual data available to us, and limits on how well these extant data represent the speech community as a whole – prove to be most problematic.
Joan Beal's paper (‘Levelling and enregisterment in northern dialects of late Modern English’) takes a macro-linguistic perspective, arguing that processes of dialect levelling and enregisterment play a crucial role in the formation of urban dialects in the north of England during the nineteenth century. She focuses on metalinguistic evidence, representations of dialect in texts, and contemporary commentaries on dialect diversity, arguing that these neglected sources can provide a way to examine processes of change in historical dialects. These sources highlight both perceived dialect attrition and evidence that certain dialects come to be linked with particular social identities or social stereotypes, indicating a high degree of enregisterment.
April McMahon and Warren Maguire (‘Quantitative historical dialectology’) present the results of a new computational method for classifying languages and varieties based on a range of phonological features (for example, height, frontness, rounding and length of vowels). This analysis produces a taxonomy in the form of a network, based on degrees of similarity or differences between the varieties. This kind of analysis has previously been used to examine the phylogeny of languages. McMahon and Maguire argue that it can be extended to examine the relationships between varieties, and that this may shed further light on the historical development of the varieties in question. This method allows examination of several properties simultaneously, and hence mitigates any biases which come about by classifying varieties on the basis of a small number of preselected features. While the results of their pilot analysis largely confirm existing classifications, they argue that this method can provide a much finer grained classification of varieties. When applied to both present-day and historical varieties, the empirical results obtained by this method are impressive, and its potential for further research is an interesting prospect. However, it is not clear whether the analysis replicates the distinctions that speakers make between varieties. It would be interesting to test whether the distinctions made under the analysis are those most salient to speakers.
Terttu Nevalainen (‘Reconstructing syntactic continuity and change in early Modern English regional dialects: the case of who’) focuses on evidence for Early Modern English dialect variation in the domain of morphosyntax, using quantitative techniques familiar from variationist studies of Present-day English, and from her work on historical sociolinguistics. Using data from the Corpus of Early English Correspondence that she subdivides into three dialect areas, East Anglia, London and the Royal Court, Nevalainen argues that relativisation strategies in sixteenth-century subject relatives exhibit dialect variation. The Early Modern English development of relative who with human antecedents is well known, but Nevalainen shows that dialect plays a role in conditioning variation between that and who in this context. She demonstrates that different dialects base their relativisation strategies on different structural and semantic generalisations – that the development of who relatives is not a uniform process throughout the country. The London dialect makes the distinction between who with animate antecedents and that with inanimate antecedents familiar to us from Present-day English, whereas East Anglian writers seem to be generalising wh-relative pronouns across all contexts, at the expense of that. This dialect distinction is levelled in seventeeth-century writing when the London usage becomes supralocalised. The paper makes an important methodological contribution, emphasising that accounts of morphosyntactic change which take little account of variation within the speech community may obscure generalisations and patterns of change in quantitative data.
The fourth section takes early English phonological change as its theme. The introduction focuses our attention on questions of archaism – that is, whether accounts of early English phonology force constructs and analyses on the language that are historically motivated, but irrelevant to a particular synchronic stage of the language under investigation. The papers by Minkova and Britton address this theme.
Donka Minkova (‘Syllable weight and the weak verb paradigms in Old English’) examines the status of the distinction between two classes of Old English verbs: those that have a thematic vowel <-ode> (e.g. getimbrode ‘built’) in the preterite and those that do not. Previous accounts characterise this difference as one of vowel syncope, partly as a function of the weight of the preceding morphological root. However, Minkova shows that the relationship between vowel syncope and root weight is somewhat opaque in Old English, arguing that such an account of the alternation is archaic for ninth- and tenth-century data. Instead, she argues that an interpretation of these data in which the thematic vowel has already been reanalysed as part of the stem leads to a more transparent analysis of allomorphy in the Old English past-tense inflection.
The paper by Nikolaus Ritt (‘How to weaken one's consonants, strengthen one's vowels and remain English at the same time’) takes a more general view of phonological processes in early English over a rather longer time scale. He argues that historically recurrent processes of consonant weakening and of vowel strengthening are both reflexes of English being a stress-timed language rather than a syllable-timed one. Consonant weakening and vowel strengthening are both adjustments to produce feet with a greater degree of isochrony. In this sense, he claims, structural properties of the language can influence the trajectory of its development over several generations of speakers. Ritt's account groups changes which occur at different times and in different contexts in terms of their effects, then posits a shared cause to account for these similar effects. While the account seems very intuitive, and it appears to explain a lot of historical phenomena, it is rather abstract, and, as with all accounts of this type, necessarily glosses over the role of more local factors in individual changes. There is little discussion of potential counterexamples. While Ritt's argument that the typology of English predisposes speakers towards making certain kinds of changes is convincing, it does not – nor is it intended to – explain how particular changes arise out of speaker behaviour in particular contexts at particular points in time.
The late Derek Britton reappraises the loss of geminate consonants in Middle English (‘Degemination in English, with special reference to the Middle English period’). He argues that so-called geminate consonants in Middle English do not represent long consonant phonemes, but only occur across syllable boundaries. Their loss, therefore, represents a change in the phonotactic rules of Middle English rather than a change in the phonemic inventory. Britton demonstrates that minimal pairs of lexical items differing only in having a single or a geminate consonant are already quite rare in twelfth-century Middle English. He then goes on to argue that most of these contrasts at the level of lexical items are lost following open syllable lengthening and loss of schwa in final unstressed syllables, concluding that no additional mechanisms are necessary to account for the loss of geminate consonants.
The final section comprises two papers on historical syntax by Olga Fischer and by Anthony Warner. Both pursue a strong empirical focus, with both using large-scale historical corpora to identify generalisations and patterns of change. Crucially, both seek to explain change from the perspective of users of the historical varieties themselves, drawing insights from domains of language other than syntax to explain the behaviour of these speakers.
Taking her earlier work as a starting point, Fischer (‘The status of the postposed ‘and-adjective’ construction in Old English: attributive or predicative?’) makes a convincing case that we should distinguish weak and strong adjectives in Old English as attributive and predicative in terms of their information status (weak conveys given information and strong new information). This morphological and functional distinction also correlates with the adjective's position. Fischer argues that preposed adjectives are attributive, conveying given information, and postposed adjectives predicative, conveying new information. In this paper, she seeks to extend this analysis to coordinated adjectives in the structure a sick man and wounded, specifically asking whether the adjective in the second conjunct is attributive to a following empty noun (pro), or whether it is predicative. A detailed examination of these constructions in the York–Helsinki Corpus of Old English supports Fischer's classification of adjectives. She looks at the presence of a determiner in the second conjunct, the information status of the postposed adjective, and the strictness of coreference between the first and second conjuncts. Where there is no determiner in the second conjunct, the postposed adjective behaves predicatively, introducing new information. Where there is a determiner in the second conjunct, however, the adjective behaves as an attributive adjective, conveying given information. Fischer argues that a null nominal element (pro) is only required in this second type of construction. Fischer's work clearly argues that a single uniform analysis of this construction is insufficiently nuanced, and a poor fit to the data. Furthermore, she demonstrates the role pragmatic factors such as information structure play in structural generalisations in Old English, and by doing so, clearly exposes the inadequacy of a purely formal syntactic analysis. This is something those of us who work on Old English syntax need to consider carefully.
Warner (‘do with weak verbs in early Modern English’) continues the theme of much of his recent work, examining the spread and regulation of periphrastic do in sixteenth- and seventeenth-century English. In this paper, Warner shows that tense emerges during the late sixteenth century as a highly significant constraint on the frequency of periphrastic do. However, this role only becomes clear when the data, from an electronic version of Ellegard's database, are highly circumscribed. Warner's results demonstrate the benefits of paying close attention to the details of individual lexical items, collocations and constructions in order to construct an appropriate variable context for quantitative analysis. The resulting data, including only a subset of regular third-person singular verbs, show a higher incidence of do in the past tense than in the present tense. Warner rejects explanations for this pattern based on grammatical reanalysis or changes in the (sociolinguistic) evaluation of variants, and instead pursues an account based on the misperception of past tense inflections as present tense ones. He argues that once syllabic past tense inflections [ɪd] are reduced to [d], then they are harder to perceive and hence more likely to be misperceived as present tense [z]. This has two possible consequences. First, as speakers misperceive past tense forms without do as present tense forms, their estimate of the frequency of do is higher in the past tense than in the present. Warner argues they may reproduce this difference in their subsequent usage. Second, speakers may utilise past tense did to disambiguate the tenses more clearly. Warner focuses especially on negative declaratives, where he argues that the presence of a following [n] (the word not) provides a context in which [t/d] is weakly articulated, hence prone to misperception. It remains to be seen if tense has a similar effect in other contexts, such as interrogatives (for which Warner's data prove inconclusive) or affirmative declaratives. It is also puzzling why the effect of tense should achieve significance in the late sixteenth century, given that this is precisely when the distributions of do in negatives, affirmatives and interrogatives diverge. Might this be connected to Warner's earlier findings concerning change in evaluation of the sequence do not, or to Kroch's hypothesised grammatical reanalysis of do at around this date?
In conclusion, while it is clear that individual papers make important contributions to long-standing questions or problems within their respective fields, the contribution of the volume as a whole is a methodological one. It demonstrates the benefits of building theories or explanations from a strong empirical foundation, and of explaining change in terms of historical speakers’ behaviour, rather than in terms of abstract theoretical constructs. In this sense then, I think this volume continues an approach to historical data of which Richard Hogg would have been proud.
