Auxiliary deletion in the informal speech of Welsh–English bilinguals: A change in progress
Introduction
Although, as Silva-Corvalán (1994:1) points out, structuralist views of language precluded the observation of change in progress, methodological advances in the latter part of the twentieth century have made this possible. By now the ‘apparent time’ construct (cf. Bailey et al., 1991) has become widely accepted. In the words of Bailey et al. (1991:241), “[t]he basic assumption underlying the construct is that, unless there is evidence to the contrary, differences among generations of similar adults mirror actual diachronic developments in a language”. In this paper we shall use the apparent time construct to examine the evidence for change in the use of the auxiliary verb bod ‘be’ in Welsh speech, while paying due attention to “evidence to the contrary”. We shall then consider the cause of the changes established, weighing up the possible contribution of internal and/or external (contact-induced) factors.
There has been considerable discussion in the literature on the mechanisms which cause language change, particularly on the distinction between internally motivated change, where the change is a result of internal restructuring of the grammar, and externally motivated change, where the grammar of one language is influenced by the grammar of another language in contact with it (see the discussions in e.g. Farrar and Jones, 2002, McMahon, 2004, Backus, 2005, Treffers-Daller and Mougeon, 2005, Mougeon et al., 2005).
The process of internal language change is argued by Croft (2000) to consist of speakers selecting one of three options when speaking: (i) they use an established form (“normal replication”), or (ii) they use a new form which they have not used before (“innovation”), or (iii) they choose a recently innovated form over an established form (“propagation”). Normal replication is the conservative choice and inhibits language change, whereas innovation by itself is not sufficient to create a lasting change in a language (since there is no certainty that the same speaker will not choose normal replication the next time she has to select a form). Propagation of an innovation by sufficient speakers, however, can result in that innovation eventually becoming itself an established form, perhaps unsettling the ‘old’ form, and thereby changing the language. Thomason (2003:694) notes that, in principle, “any feature that can appear in a single person's speech at any time… can turn into a permanent change in the entire language”. Internal language change is essentially a regularization of the variation found in human speech.
Externally motivated change, on the other hand, is precipitated by language contact (e.g. Weinreich, 1953:1; Winford, 2003:23). This change takes the form of parts of a language's grammar being lost, added to, or otherwise modified or restructured. For example, Finnish has shifted from SOV to SVO under influence from English and other Indo-European languages (Thomason, 2001:11), and Central Coast Salish has mostly shifted from VSO to SVO, probably under the influence of English (Shaw et al., 2008).
Several authors have in fact pointed out that it is frequently difficult to identify whether a particular change in a language in a contact situation is due to internal or external factors. Backus (2005:314) argues that “internal and external factors often conspire” and Thomason and Kaufman (1988:61) use the term “multiple causation”, illustrating this with the example of the fricatives [f] and [v] in Middle English, which were formerly allophones but later became distinct phonemes, due in part to the borrowing from Norman French of words with initial /f/ and /v/ which were minimal pairs. Thomason and Kaufman argue that this was due to both the external influence of French and an internal process, where the influence of French “generally anticipated or reinforced” existing internal processes of sound change (1988:124).
In order to determine whether a particular change has been accelerated by language contact it is convenient to have a community of monolingual speakers of the language in question as a reference point. However, as Backus (2005:312) points out, many languages in contact with a dominant language no longer have monolingual speakers, and this is true of Welsh in Wales. Welsh is dominated by English not only because the majority of the inhabitants of Wales are monolingual English speakers, but also because all adult speakers of Welsh also speak English.
In this paper we examine one apparent change in progress in Welsh, namely the deletion of finite auxiliary forms of the verb bod ‘be’ (AuxD) in Welsh informal speech, and discuss whether or not the change is due to internal factors, external factors, or a combination of both—or, indeed, whether or not such a separation of causal factors is even possible—without adopting a priori the ‘either–or’ mentality criticized by Farrar and Jones (2002).
Welsh, a Celtic language, is spoken primarily in Wales, where language contact with English dates back to the time of the Anglo-Saxon settlers in Britain in the 5th and 6th centuries AD (Davies, 1994). As Thomason and Kaufman (1988) note, bilingualism is usually a prerequisite for any major kinds of externally motivated language change (e.g. word order or morphology, rather than mere changes in vocabulary), and Welsh–English bilingualism has been common in Wales since at least the 19th century. According to the 2011 UK Census, 19% of the Welsh population of 3.1 million people can speak Welsh. Although Welsh monolinguals existed in the past, all Welsh speakers are now bilingual in English to varying extents. Thomason and Kaufman (1988:50) propose a typology of language change that notes the link between the intensity of bilingualism in a language community and the extent and types of language change observed. Following their framework, we suggest that the situation of the Welsh language falls into the category of “intense bilingualism”, so, according to Thomason and Kaufman's criteria, changes attributable to language contact of practically any kind could occur, and such changes can be extensive. We will examine AuxD with this prediction in mind.
In what follows we describe the phenomenon of AuxD in Welsh, using examples from our spoken corpus, Siarad, which we describe in section 4. We then review previous research on this phenomenon before presenting our own analysis of variation in AuxD and our interpretation of change in progress. We discuss internal and external contact-induced causes of the change, and the extent to which the change reflects convergence of Welsh towards English.
Section snippets
Auxiliary deletion in Welsh
In this section we describe in more detail the phenomenon of Auxiliary deletion (AuxD) in Welsh. Auxiliary deletion in informal spoken Welsh involves the deletion of the initial finite auxiliary verb in periphrastic constructions when that auxiliary is a form of bod ‘be’. An example of such a clause is given in (1) below.1
Review of the literature on auxiliary deletion
Having outlined for the reader what AuxD in Welsh involves, we now turn to a survey of the limited previous work on this phenomenon, mostly in Welsh but also in other languages. In one of the earliest comprehensive studies of spoken, as opposed to literary, Welsh, Jones and Thomas (1977) highlight the difference in choice of +A or −A based on style. In their list of different Welsh inflected verb forms in the appendix to their book (1977:399), they describe the form which they cite as ’ti as a
Materials and methodology
In order to measure the frequency of AuxD in the speech of Welsh–English bilinguals and to assess the evidence for change in progress, we will examine data from our Welsh–English corpus, Siarad (www.talkbank.org/BilingBank/Bangor), a database containing recorded and transcribed informal conversations from Welsh–English bilinguals. The data were collected by a team from Bangor University from 2005 to 2008, as part of a project8 funded by the Arts and
Extent of AuxD in the data
Of the 643 B2S clauses produced by the 28 speakers, 598 (93.0%) clauses showed a deleted auxiliary (−A). Fig. 1 shows these results.
The number of B2S clauses produced by an individual speaker ranged from 1 (Derwena) to 76 (Arfon), with speakers producing a mean 22.96 B2S clauses. Fig. 2 gives the proportion of B2S clauses with −A produced by each speaker individually.
These results clearly show that these speakers in these data delete the auxiliary verb in the great majority of the B2S clauses
Age variation and apparent time
Applying the apparent time method (e.g. Bailey et al., 1991, Labov, 2001, Chambers et al., 2004) to these data allows us to identify whether this age variation is a sign of language change in progress. According to this method, where synchronic data shows that the speech of younger groups is different from the speech of older groups, it can be inferred that there is a change in progress, given the assumption that speakers do not change their speech patterns as they age (though see below for a
Conclusions
In this paper we have reported on the evidence for a change in progress involving the deletion of the auxiliary bod ‘be’ in Welsh in constructions with second person singular subjects. Overall, deletion of this auxiliary was highly frequent in the data we analyzed. We have found that all speakers in our data delete this auxiliary at least half of the time. We found a slightly higher tendency to delete the auxiliary in monolingual than in bilingual clauses, possibly because code-switching
References (48)
Welsh-English code-switching and the Matrix Language Frame model
Lingua
(2006)Codeswitching and language change: one thing leads to another?
International Journal of Bilingualism
(2005)- et al.
The apparent time construct
Language Variation and Change
(1991) Real and apparent time in language change: late adoption of changes in Montreal English
American Speech
(2004)A new approach to Welsh negation. Ms
(2005)- et al.
Welsh negation and grammatical theory
(2001) - et al.
The development of finiteness in early Welsh
Journal of Celtic Language Learning
(2005) - et al.
The syntax of Welsh
(2007) - et al.
Convergence as an emergent property in bilingual speech
Bilingualism: Language and Cognition
(2004) Sociolinguistic Theory: Linguistic Variation and its Social Significance
(2003)
The Handbook of Language Variation and Change
Explaining Language Change: An Evolutionary Approach
A history of Wales
Using the Matrix Language Frame model to measure the extent of word-order convergence in Welsh-English bilingual speech
Internally and externally motivated change in language contact settings: doubts about dichotomy
Introduction
Deletion through movement
Natural Language and Linguistic Theory
A grammar of Upriver Halkomelem
Cognitive effects of bilingualism: digging deeper for the contributions of language dominance, linguistic knowledge, socio-economic status and cognitive abilities
International Journal of Bilingual Education and Bilingualism
Convergence and creolization: a case from the Indo-Aryan/Dravidian border
The acquisition of Irish: a study of word order development
Journal of Child Language
The licensing powers of mood and negation in spoken Welsh: full and contracted forms of the present tense of bod ‘be’
Journal of Celtic Linguistics
Cited by (9)
Doubling: An error or an illusion?
2016, Bilingualism‘If we don’t study the language, the history will be lost’: motivation to learn Welsh in Argentine Patagonia
2022, International Journal of Bilingual Education and BilingualismCrowdsourcing the Paldaruo Speech Corpus of Welsh for speech technology
2019, Information (Switzerland)Age variation and language change in Welsh: Auxiliary deletion and possessive constructions
2017, Sociolinguistics in Wales