Contact Linguistics. Chap

Exercise: Examine any situation in which a minority group’s language is under threat from a dominant language, for instance, Spanish in the US, French in Eastern Canada. Try to determine what social factors either promote or inhibit shift in the community of your choice. [Suggested reading: Dressler & Wodak-Leodolter (1977), Seliger & Vago (1991a), Dorian (1989) and Fishman (2001) contain many interesting case studies of endangered languages.]


III.3. The linguistic consequences of L1 attrition.

Studies of the linguistic consequences of first language attrition are particularly interested in the final 3 stages of the process of language death. Sasse (1992a:20) labels these stages as follows:

32. 
    Primary language shift, in which shifting speakers become more proficient in the L2 and use it more often than the AL (Batibo’s Stage 3, above);
    
33. 
    Language decay, characterized by significant structural change in the AL (Batibo’s Stage 4);
    
34. 
    Language death and replacement, that is, monolingual competence in the L2 and total loss of competence in the AL (Batibo’s Stage 5).
    

Each of these stages is characterized by particular linguistic changes in the dying language.

During the period of primary language shift, L2 elements increasingly intrude into the AL, encouraged by frequent code switching and code mixing. L2-dominant bilinguals are the chief agents of the diffusion of L2 lexicon and structure into the AL. Early manifestations of this process can be seen in the various situations of incipient shift that we discussed in Chapter 3. For instance, languages like Prince Edward Island French, L.A Spanish and Pennsylvania German all show evidence of lexical and structural interference from English. This is particularly true of those varieties of the AL’s that are spoken by English-dominant bilinguals.
Change in a threatened language under influence from a dominant L2 involves the same mechanisms and processes as those we saw earlier in these cases of convergence. They include direct transfer of L2 lexical and structural elements as well as the diffusion of L2 structural patterns into the L1. Seliger & Vago (1991b:7-10) discuss many examples of such diffusion in the attriting L1 German and Hungarian of speakers for whom L2 English had become the primary language. L2 rules of word order, preposition stranding, patterns of subordination etc. are adopted into the L1. Loan translations, semantic extensions etc on the pattern of the L2 become common in the L1, whose lexicon as a whole is increasingly forgotten and replaced by L2 items. As Haugen (1978:37) puts it:
“As the learner builds new systems in the language he acquires, he dismantles and reorders the systems of the language he already knows.”

Some languages involved in a long period of gradual shift manage to retain something of their autonomy despite a great deal of influence from an encroaching language. But they usually pay the price of losing a great deal of their former character, evolving into linguistic hybrids that are substantially different from earlier forms of the language. Examples include Ngandi, now becoming extinct as its speakers shift to Ritharngu in Arnhem Land, Australia (Chapter 3, section 6). An extreme case of survival at the expense of massive change is Cappadocian Greek, which has suffered pervasive interference from Turkish at every level of its structure (Chapter 3, Section 7.2). The same might be said of the Kupwar situation where local varieties of Urdu, Marathi and Kannada have all undergone significant structural change due to convergence (Chapter 3, section 7.3). In all these cases, the languages concerned have not become extinct, and continue to meet their speakers’ communicative needs. Those aspects of their structure that were lost are simply replaced by others from the L2, so there is no loss of communicative power. However, a different picture emerges in cases where the AL is undergoing not just interference, but also decay.

III. 3.1. Language Decay.

The decision to give up an AL and cease transmitting it to one’s descendants signals the start of serious attrition in the language. This is the period of decay both in individual idiolects and the communal language. This stage is associated with the emergence of speakers with limited competence in the AL. The degree of proficiency they have depends on how they acquired the AL, and what opportunities they have to use it. Sasse (1992b:77, fn 2) cites Dorian’s (pc) classification of limited proficiency speakers into two broad categories: formerly fluent speakers (sometimes called “rusty” speakers) and semi-speakers. The latter can be further divided into those who acquired the AL as children but abandoned it at school and those who experienced “abnormal acquisition” due to inadequate exposure to the language. The latter are “prototypical” semi-speakers, Semi-speakers are also referred to as “terminal” speakers (Tsitsipis 1989).

Formerly fluent speakers often preserve or can recover much of their original knowledge, so that their language “dies with its morphological boots on” (Dorian 1978). Others (the “forgetters”) have significantly reduced competence, like the remnant speakers of Menomini described by Bloomfield (1927). In fact, such speakers really belong to the category of “semi-speakers.” An example of a semi-speaker who abandoned his AL early is Michal, a child who gradually lost command of Hebrew as he shifted to English between the ages of 2:7 and 4:6 (Kaufman & Aronoff 1991). His dying Hebrew passed through all the stages of attrition typical of decaying languages, as discussed earlier. Semi-speakers of different kinds are the primary producers of decayed language. Their performance in the AL ranges from reasonably fluent speech to distorted output somewhat similar to those kinds of aphasic speech characterized by agrammatism (Menn 1989). The extent of attrition correlates closely with age differences, with younger speakers displaying less competence than older ones. In the final stages of death, such speakers make up the vast majority of the speech community.
Manifestations of language decay include increasing replacement of L1 elements by L2 counterparts, and various kinds of reduction and simplification in the L1 itself. The more speakers come to rely on the L2 for communication, the more their AL disintegrates. This manifests itself especially in loss of morphological and morphophonemic apparatus, such as verbal and nominal inflection and derivation. Reduction in allomorphic variation and over-generalization of forms result in greater paradigmatic regularity. For example, Dimmendaal (1992) reports the loss of nominal affixes marking number and gender in semi-speaker varieties of Kore, a language spoken on Lamu Island off the Kenyan coast. The verbs used by these semi-speakers are also invariable, having lost inflectional markers of tense, aspect and negation, as well as number and person agreement. Sasse (1992b:70-71) also reports similar kinds of morphological simplification in a semi-speaker variety of Arvanítika.
Loss of inflection must sometimes be compensated for by fixed word order to express grammatical relations. This is why semi-speaker varieties of immigrant Finnish in northern Minnesota have fixed word order whereas Finnish word order is variable. Larmouth (1974:359) gives the following examples of differences in the language of first versus fourth generation speakers:

(38) G1. a. Mieheltä kuoli äiti

man-ABLAT died (the) mother
OR b. Miehan äiti kuoli
man-GEN mother died.

G4. Miehen äiti kuoli (Only the genitive construction is used)

Other changes at the syntactic level include the loss of synthetic constructions in favor of periphrastic ones, as well as the loss of subordination mechanisms. For instance, The “conjugating” prepositions of Gaelic are replaced by analytic prepositional phrases in semi-speaker varieties. Dorian (1981) provides examples like the following:

(39) Proficient speaker Semi-speaker Gloss

rium ri mis’ “to me”
dhomh orn mis’ “for me”

Maher (1991) provides a comprehensive overview of all these types of simplification and regularization in semi-speaker varieties found in various immigrant enclave communities. Sasse (1992b:71) gives further examples from Arvanítika, while Schmidt (1985) provides others from Dyirbal.

In phonology too, we find various kinds of leveling in decaying languages. For example, semi-speaker varieties of Kore have lost the opposition between voiceless plosives and voiced implosives, as well as distinctions of length and tongue root position for vowels. Interestingly, though, they preserve tone distinctions, possibly because these may be less marked than segmental distinctions (Dimmendaal 1992:129-30). In general, many of the changes at all levels of structure appear to involve the loss of more marked features or structures and the preservation of less marked ones.