Privatisation of higher education

Studia Anglica Posnaniensia 55s2 (2020): 549–555 doi: 10.2478/stap-2020-0030 REVIEW Millenia of language change. Sociolinguistic studies in deep historical linguistics. By Peter Trudgill. Cambridge: Cambridge University Press, 2020. Pp. 164. Reviewed by Anna Rogos-Hebda (Faculty of English, Adam Mickiewicz University in Poznań) In his Millenia of language change. Sociolinguistic studies in deep historical linguistics Trudgill, indeed, digs deep in linguistic history. He digs so deep, in fact, that his search for explanations takes him millenia away, sometimes as far back as the Mesolithic and the New Stone Age. Departing from the conviction that “some linguistic features take a very long time […] to develop” (2020: 1) Trudgill looks for the ultimate roots of a number of language change processes. In doing so, he applies the tools of sociolinguistic typology, for to his mind the organisation of the then societies, clearly different from that of most contemporary ones, cannot have been irrelevant to the way linguistic innovations spread. Each of the eight chapters, then, throws light on the role of social factors in promoting or hindering language change in different parts of the world, at different times in history. What aspects of social organisation could be instrumental in encouraging a particular outcome, in the sense of either simplification or complexification of language structure, and how they possibly contribute is the topic of Chapter 1 “Prehistoric sociolinguistics and the Uniformitarian hypothesis”. At the beginning of the chapter Trudgill rightly observes that uniformitarianism, according to which “the general processes and principles which can be noticed in observable history are applicable in all stages of language history” (Hock 1991: 630), only works as a methodological principle with regard to those linguistic features “due to the nature of the human language faculty” (2020: 7–8). Surely, there must be, however, aspects of language structure whose shape or nature depends on (the interplay of a number of) social parametres. Indeed, Trudgill goes on to describe a range of phenomena he sees as determined by such factors as “arbitrary human invention” (Trudgill 2020: 9, after Blust 2012) and small 550 Review community size, the latter resulting in non-anonymity, non-optimality, density of social networks, and large amounts of communally shared information. He mentions, for example, an intentional arbitrary swap in the nominal system of Laro/Laru spoken in Kordofan, central Sudan, whereby the masculine nouns have turned feminine and the other way around, allegedly for the neighbours of the Kordofians to get confused (Trudgill 2020: 10, after Schadeberg 1981). Community size may have been responsible for the fostering of non-optimal object-initial word orders found in “small communities […] more vulnerable to drift away from optimal states” (2020:11–12, after Nettle 1999) as well as a personal pronoun system, attested in Onya Darat (western Borneo), with pronouns reflecting generational affiliation among interlocutors (2020: 11, after Tadmor 2015). The density of social networks and a considerable amount of communally shared information, in turn, are considered by Trudgill conducive, respectively, to the implementation of unexpected sound developments (e.g., unusual fortitions in peripheral dialect areas) and the development of overelaborate deictic systems (e.g., 31 personal pronouns in !Ora, a now extinct Khoisan language, vs. mere 6 forms in Finnish). While “the uniformitarian principle is basically correct”, Trudgill concludes, caution is necessary in applying “the present to explain the past” (2020: 16, 7), for Mesolithic and Neolithic tightly knit communities “provided a social matrix which allowed linguistic phenomena to develop which are most unlikely to arise today in our own modern at-a-distance societies” (2020: 8). The question of social factors influencing linguistic structure is discussed further in Chapter 2 “From Ancient Greek to Comanche: On many millennia of complexification”. Musing about the extent to which social structure impacts the typological characteristics of a language, Trudgill focuses on the issues of structure simplification and complexification as the two outcomes of linguistic contact. On the basis of divergent behaviours of genetically related varieties, namely Afrikaans vs. the Netherlandic dialects of Belgium and the southern Netherlands, he explains how the scale and type of contact determine at which end of the complexity spectrum a language lands. Simplification typically results from “short-term contact with other communities speaking different languages” (2020: 35), due to imperfect adult second-language learning with concomitant removal of linguistic L2-difficult features such as “irregularity, syntactic agreement and grammatical gender” (2020: 20). Complexification, on the other hand, seems attributable to “[l]ong-term contact between communities where small children become bilingual”, speakers transferring from L2 features of grammar and phonology (additive borrowing) (2020: 21). Of the two trends, the latter, Trudgill notes, has been on the decrease, perhaps for the past 2000 years. The complexity of the kind found in Ancient Greek, West Greenlandic or Comanche takes “many millennia rather than centuries” to Review 551 develop (2020: 32) and a hospitable environment. That environment, however, i.e., the sociolinguistic conditions required, is becoming harder and harder to come by, given the size of the world population and the rate of short-term adult language contact these days. The most complex languages are spoken in small, low-contact, non-industrialised tribal communities, after all, and those are increasingly rare. The notion of additive borrowing, i.e., the acquisition of a morphologically expressed grammatical category from another language, is brought up again in Chapter 3 “First millennium England: A tale of two copulas”. The matter at hand is the formal and functional distinction between Old English beon and wesan, both meaning ‘to be’, unattested in other Germanic languages and no longer present in English itself (2020: 45, after Wischer 2011). Trudgill (2020: 3) lends a sympathetic ear to Vennemann’s (2010a) hypothesis concerning the role of Vasconic, possibly originally spoken in the Franco-Cantabrian Refugium, in the transfer of that feature to the north-western European languages, including (Old) English. If widely criticised as heavily speculative, the hypothesis ties in nicely with Trudgill’s claim that nearly all complexification, of which the adoption of a two-copula system is an example, requires “a long-term, co-territorial contact situation between social groups involving childhood bilingualism” (2020: 39). A possibility of that kind of (lengthy) cohabitation taking place first between the speakers of Proto-Vasconic and Proto-Celtic, then between the Celts and the Romans in mainland northwestern Europe and Lowland England, and – finally – between the Romano-Britons and the Germanic tribes in Britain cannot be ruled out in the light of Schrijver’s research on the use of British Latin in the Lowland zone (Trudgill 2020: 41, after Schrijver 2002) and Filppula and Klemola’s (2014) paper re-evaluating Celtic influences in English. The abandonment of the distinction between the habitual/consuetudinal and non-habitual be in early Middle English is attributed by Trudgill (2020: 49, after Lutz 2009) to contact with Old Norse. The role of language contact in the history of English is what Chapter 4 “The first three thousand years: Contact in prehistoric and early historic English” discusses at length. Beginning with language change processes which did not operate in English per se, but nevertheless led to its emergence affecting the structure of its ancestors, Trudgill considers the nature of subsequent encounters with particular language varieties against the move of a “highly synthetic fusional inflecting language” that Old English was towards “a much more isolating type of morphology” found in Middle English (2020: 56). Simplification, as he claims (2020: 51), typically results from adult, imperfect, language learning inherent to short-term (not infrequently tumultuous) contact situations. Trudgill believes (2020: 55–56, after Morris 1973 and McMahon 2011) that after the period of initial British dominance over the Anglo-Saxons, leading to bilingualism on the 552 Review part of the invaders, from the 7th century onwards adult linguistic contact in the Highlands, the Midlands and the South-West between the now subjugated Celts and the speakers of Old English would lead to a degree of pidginisation through imperfect second language-learning. A case for substratal contact with Late British as instrumental in the simplification of Old English is, in Trudgill’s view (2020: 62), much stronger from a sociolinguistic-typological perspective than the case for adstratal contact with Old Norse, involving “long-term co-territorial cohabitation and intermarriage” or French, whose speakers were somewhat isolated due to their social standing. In Chapter 5 “Verner’s law, Germanic dialects and the English dialect ‘default singulars’” Trudgill (2020: 67) dissects what he calls “a widespread perception in the international English-linguistics community” regarding the status of plural was as a “vernacular primitive” or, better yet, an “angloversal”. Specifically, he argues against viewing was-generalisation in vernacular varieties of English as representing the principle of “the default singular” at work. The distinction between was and were as the preterite forms of the verb to be goes back a long time to Verner’s Law, a process conditioned by the mobility of the PIE accent, which inhibited lenition after stressed vowels (Page 1998). While this conservative alternation is very much alive in Standard English and Dutch, the mainland dialects of North Frisian and a number of Dutch dialects, in the remainder of the Germanic language family the s/r distinction has been levelled out, languages opting either for the s- or the r-paradigm (Trudgill 2020: 74). If the s-variant were a default singular, Trudgill argues, one would expect it to predominate in all the varieties (2020: 68). Yet, despite its prevalence in most of the colonial Englishes, it is r-generalisation that the majority of Germanic vernaculars have regularised towards (Trudgill 2020: 74). Therefore, the principle of “default singular”, if operative in English, indeed, “has no explanatory value whatsoever in this case”, Trudgill concludes (2020: 68). Rather, we are dealing with a regularising trend, with world English leaning decidedly towards was simply due to the fact that it happened to be the norm “in the heavily populated English southeast” at the time English-speaking colonies were established (Trudgill 2020: 76). Chapter 6 “Deep into the Pacific: The Austronesian migrations and the linguistic consequences of isolation” tackles the impressive expansion of the Polynesian languages, a millenia-long reduction in the consonant inventory of some and the role of sociolinguistic factors therein. Five thousand years of colonisation of the Pacific by the Austronesians ended around 1400AD with the settlement of the Chatham Islands by migrants from New Zealand. As a result, the Austronesian language family covers an enormous area stretching from Hawai’i to the South Island of New Zealand, and from Madagascar to Rapa Nui (Trudgill 2020: 78–80). Trudgill observes that the consonant inventories in Review 553 distant albeit related Hawai’ian and Rurutu (spoken in the island of Rurutu in the Austral archipelago) contain as few as eight phonemes, following thousands of years of consonant loss beginning with Proto-Malayo-Polynesian (2020: 81). Trudgill links the said reduction to two factors, namely small community size and isolation, but the cause-effect scenario is different from the ones hitherto presented. In previous chapters simplification was described as resulting from adult language contact through faulty L2 learning and subsequent pidginisation (2020: 84). This kind of explanation, however, does not seem adequate in the case of relatively isolated, low-contact varieties that Hawai’ian and Rurutu are assumed to be. With Haudricourt, Trudgill (2020: 84, after Haudricourt 1961) therefore turns to the isolation-triggered impoverishment as well as to “large amounts of shared information” (2020: 88) due to small community size as potential forces behind the reduction in question. In Chapter 7 “The Hellenistic Koiné 320BC to 550AD and its medieval and Early Modern congeners” Trudgill disposes of two major fallacies, i.e., the Monogenesis Fallacy and the Identity Fallacy, concerning colonial varieties of western European languages (2020: 89). Having talked the reader, step by step, through the formation of the Hellenistic Koiné, Trudgill uses this example to show how unnecessary it is to reach for the notion of identity in explaining the motivation behind the emergence of new varieties in colonial situations and how unreasonable it is to believe in their monogenetic origin. Given the logistics of colonialisation and the mechanisms at work when a number of dialects are brought into close contact, a new mixed dialect is bound to be born, Trudgill (2020: 91, 100) argues. He supports his claim with an overview of a range of (chronologically distant) colonial koinés from all over the world, from colonial Arabic of the 7th century through Australian English to the new Polish mixed dialects of the post WWII era. The picture that emerges is one of there being no other way for a colonial concoction of dialectal features but to lead to the establishment of a new variety. The final chapter, Chapter 8 “Indo-European feminines: Contact, diffusion and gender loss around the North Sea”, is an attempt at finding a sound, sociolinguistically grounded explanation for the discrepancy among Germanic languages regarding the number of genders present. While Norwegian and Icelandic retain the original (i.e., Indo-European non-Anatolian) three-way contrast, Standard Swedish and Danish, Bergen Norwegian, northern Dutch, northwestern Low German, West Frisian and the island dialects of North Frisian show the masculine – feminine syncretism, with English having lost grammatical gender altogether. Not unexpectedly, Trudgill sees geography and linguistic contact as “crucially involved in the loss of the feminine” (2020: 113), though it is a very specific type of contact that he means, of course, namely one involving post-critical threshold learning. In that kind of learning semantically redundant, 554 Review cross-linguistically dispensable phenomena such as grammatical gender (Trudgill 2020: 109, after Hickey 2000 and Dahl 2004) are among the first to be removed. Geographically, the area affected by gender reduction is the western seaboard of the North Sea and the eastern seaboard from the Rhine Delta up to Western Jutland, i.e., the zone of intense maritime Hanseatic trade contacts. Trudgill proposes that it was trade relations between urban centres and the associated influx of foreigners that fuelled the spatial diffusion of the (linguistic) innovation from the kernels in Britain and northern Holland (2020: 127). Contact between communities as the ever-present driving force behind linguistic simplification or complexification is the leitmotif of Trudgill’s book. Sociolinguistic factors, such as community size, network density, or the extent of isolation, serve Trudgill as the common denominator for language change processes operative in varieties distant in space and/or in time. Exploring the extent to which sociolinguistic typology could help make sense of the outcomes of those processes, Trudgill casts his net wide both in terms of chronology and geography, which is admittedly an advantage. Approachable narrative and compelling argumentation make it easy for the reader to yield to the enticing illusion that a universal explanation lies at hand, even if the author does not explicitly make that promise anywhere. Would it not be nice, after all, if all elements of the puzzle that historical research not infrequently resembles, finally fell into place? Yet, there are fragments when the degree of speculation requires of the reader a considerable amount of openness, for example when Trudgill quotes Vennemann on the vocabulary of Proto Germanic being influenced by contact with Semitidic (2020: 53). It is also difficult at times to escape the impression that some evidence has been used selectively to help further the author’s agenda. This is the case, perhaps, when – discussing the influence of Britonnic/Late British upon Old English – Trudgill first uses the notions of long-term contact and (child) bilingualism to justify borrowing from Britonnic of the progressive aspect (2020: 55), only to use the same arguments two pages later as causing noticeable simplification in the structure of English. He pins the simplification onto the now different sociolinguistic context (2020: 56), but even with the post-600AD Germanic dominance it does make one wonder where the need for reduction(s) came from, if already between 420 and 600 the speakers of Old English achieved what Trudgill calls “competent bilingualism” (2020: 56). Also, why that shift in power relations would somehow “prioritise” adult bilingualism, with its drive towards simplicity, over child bilingualism involved in additive borrowing is unclear. Of perhaps lesser weight, yet not unnoticeable, are occasional spelling errors which, in the case of place names like “Szczeczin” [sic!] for the Polish city of Szczecin or “Madgedurg” [sic!] for Magdeburg do stand out. Millenia of language change. Sociolinguistic studies in deep historical linguistics is, nevertheless, a well written, well balanced and a much needed book. Review 555 The times of fake news and post-truths seem to offer ample opportunity for researching aspects of contemporary discourse(s), yet Trudgill’s book points to the ongoing need for rethinking historical linguistic approaches to language change and proves that insights from other disciplines can inform research on historical data in multiple ways. REFERENCES Filppula, Markku & Juhani Klemola. 2014. Celtic influences in English: A reevaluation”. Neuphilologische Mitteilungen 115 (1). 33–53. Hock, Hans Heinrich. 1991. Principles of historical linguistics (2nd edn.). Walter de Gruyter.