Contact Linguistics. Chap

Exercise:
Compare any two case studies of koiné formation – for example, Siegel’s (1987) study of Fiji Hindi; Omdal’s (1977) study of Høyanger Norwegian; or Kerswill & Williams’ (2000) study of Milton Keynes, England. What processes of change are involved in these kinds of dialect convergence? What part do structural and social factors play in facilitating or impeding koiné formation? Discuss the relative importance of children and adults in this process.
7.2. Situations in level 5 of Thomason & Kaufman’s borrowing scale.

The situations that Thomason & Kaufman (1988) assign to level 5 of their borrowing scale include Wutun, a Chinese dialect that has undergone extensive structural change under Tibetan influence, and Asia Minor Greek, which has changed significantly under Turkish influence. These situations are characterized by significant changes in syntax as well as morphology, phonology and lexicon. Wutun, for instance, experienced the loss of phonemic tones, the development of several new cases, and changes in word order, to mention just a few.

Asia Minor Greek underwent similar changes at all levels of structure, including a variety of changes of a sort typically associated with substratum influence. The Cappodacian variety of Greek in particular has been subjected to far more Turkish influence than those in other areas such as Silli and Phárasa (also in Central Asia Minor), while Greek varieties in Pontus on the western coast show much less influence. Dawkins’ (1916) study showed that Cappadocian Greek had converged considerably toward Turkish – in his famous words, “the body has remained Greek, but the soul has become Turkish.” Just a few examples will suffice here to illustrate this. In phonology, vowel harmony is found on Greek suffixes attached to Turkish-derived words, and various morphophonemic rules from Turkish have been introduced into the Greek dialect (Thomason & Kaufman 1988:218). In morphology, we find such developments as an agglutinative pattern of inflection on nouns and verbs – a feature of Turkish, not at all typical of Greek. For example, the Greek genitive suffix -yu (as in spityu “of the house”) was reinterpreted as an agglutinative suffix and extended to all nouns. The model for this was the genitive suffix –in (as in ev-in “of the house”), which was used with all nouns in Turkish (Sasse 1992b:66). Finally, in syntax we find a frequent use of SOV order on the Turkish model (where Greek has SVO) and other kinds of structural diffusion.
Once more, treating the changes in languages like Wutun and Asia Minor Greek as instances of structural borrowing in the strict sense is problematic. With respect to the latter language, Thomason & Kaufman (1988:216) argue that “if Turks did not shift to Greek, all the interference must be due to borrowing.” But this overlooks the possibility that bilinguals, especially those that were Turkish- dominant, played a key role. In other words, the dramatic changes seem rather to be the result of substratum influence under the agency of sl-dominant bilinguals who reinterpreted rl grammar on the model of the sl. This would bring Asia Minor Greek (and Wutun) in line with other cases of massive structural change in an ancestral language, due to continuing shift toward an external language. In Chapter 7, we will consider other cases of language obsolescence due to ongoing shift.
Interestingly, Thomason & Kaufman offer this very explanation for similar kinds of structural change in (other) situations that involve language shift. For example, they discuss "changes in Ethiopic Semitic from VSO, Aux-Verb, Noun-Adjective (probably) and Head Noun-Relative Clause word orders with prepositions to SOV, Verb-Aux, Adjective-Noun and Relative Clause-Head Noun word orders with postpositions to match the patterns of the substrate Cushitic languages." (1988:131). Though these changes have been installed in Ethiopic Semitic as a maintained language, Th. & K. attribute them to substratum influence on the Semitic learnt by Cushitic speakers. Such developments make one wonder whether many such changes that have been ascribed to “borrowing” may not have involved an intermediate stage of substratum influence. Even in cases where native speakers of the recipient language were the agents of change, it is arguable that they must have had a substantial degree of bilingualism in the source language. Hence the cause of change in these cases must have been substratum influence, via processes of transfer or imposition, as is typical in shift. Once more, the distinction between borrowing and substratum influence becomes blurred in these cases.


7.3. Convergence in Kupwar.

If extensive linguistic diffusion can take place across geographically contiguous communities, it is not surprising that convergence also occurs in single speech communities made up of different linguistic groups. A well-known example is the village of Kupwar situated in southern India to the north of the Mysore border (Gumperz & Wilson 1971). Here, three distinct languages, Urdu and Marathi of the Indo-European family, and Kannada of the Dravidian family have converged extensively as a result of diffusion over a long period of intimate contact.

As a result, the languages have developed virtually identical constituent structures and grammatical categories, leading to an extraordinary ease of translation among them. Even their phonologies have become quite similar. Hence a structure in one can be converted to any of the others by simply substituting the corresponding morphs, one for one. Gumperz & Wilson found it possible to postulate a single syntax for all three languages, though they are kept distinct by different lexical items and grammatical morphs.
According to Gumperz & Wilson, the changes in the Kupwar languages become apparent when they are compared with the respective (standard) varieties spoken outside the village. The most striking changes are found in Kupwar Urdu (KU) which we can use for purposes of illustration. In the first place, we find changes in the KU system of gender agreement, to conform to patterns in Kannada. For instance, the KU system of grammatical gender is changed to one of semantically-determined gender. Hence words with non-human referents which are assigned feminine gender in Standard Hindi/Urdu (HU) are reassigned to the masculine category, as example (9) illustrates.

(9) HU w´hã n´dii a-ii

KU hwa n´di ay-a
there river came "There was a flood."

Gumperz & Wilson also describe other changes in the gender system, such as the loss of gender distinctions in the verb “be” and in future tense constructions. Another morphological change is the introduction of an inclusive/exclusive "we" distinction in KU, on the model of both Kannada and Marathi.

There are also several changes in word order, for instance in complex sentences with purpose clauses on the model of Kannada, as in example (10)

(10) HU. wo [bhœs c´rane-ke liye] g´y-a th-a

KU o g´e t-a [bhœs carn-e-ko]
KM tew gel hot-a [mhœs car-ay-la]
he go Past-Agr [bufalo graze-Obl-to]
"He went to graze the buffalo")

Other examples of (morpho-)syntactic change in KU include the use of compound verbs and the use of wh-questions with 'should' on model of Marathi and Kannada. Finally, there are some examples of what appears to be direct borrowing of morphemes from Marathi. For example, the dative suffix -na replaces the native Urdu suffix -ko, as in (11).

(11) KU hwa-si h´m-na bula ne-ko p´wne ae

there-from us-to calling-for cousins came
“Our cousins came from there to call for us.”

Various questions have been raised concerning the validity of Gumperz & Wilson’s findings. For instance, the study is based on a corpus of only 10,000 words, and very scant illustration (one supporting example) is provided for each of the 16 convergent changes described. Moreover, Gumperz & Wilson fail to locate convergence in Kupwar within the larger context of widespread convergence in the surrounding region. As a reviewer of this book pointed out, “regional varieties of Urdu, Kannada and Marathi would have provided more appropriate reference points than the standard varieties, possibly revealing that convergent change in Kupwar itself is less impressive than assumed.” Various other criticisms have been leveled against the study.

Despite this, the situation remains of some interest as an example of extreme structural diffusion (whether in Kupwar itself, or in the region as a whole). The precise motivations - social and linguistic - for these developments are difficult to pin down. One factor seems to be the extensive code-switching that has been practiced in the village for generations. As we shall see in Chapter 5, code-switching is a powerful vehicle for diffusion of structural and other features across languages. Other motivations may lie in economy, in that speakers need only learn a single grammar for inter-group communication. But we still do not know the details of the convergence process, that is, the agency and precise mechanisms of the changes. As Gumperz & Wilson acknowledge,

To what extent the adaptation in each language has been initiated by its home-group speakers under the influence of their knowledge of the languages of others, and to what extent adaptation has been initiated by others using a language to its home-group speakers, or in the case of Marathi, to others, we are not able to say. (1971:269)

The preservation of separate vocabularies might be explained by the need to preserve distinct group identities and a separate code for in-group and familiar use. Such cultural pressures to keep some semblance of autonomy for languages that are converging can be found in other situations as well, including that in Northwest New Britain, the subject of the next section.

7.4. Convergence in Northwest New Britain.

The Northwestern area of New Britain in Papua-New Guinea (NWNB) has a long history of contact between the earlier established Papuan or Non-Austronesian (NAN) languages and the Austronesian (AN) languages introduced in the distant past by seafaring groups from other parts of the Pacific. At present, only one of the original NAN languages, the isolate Anêm, continues to survive in the area. It is in contact with eight AN languages, including the isolate Amara, four languages of the Bibling subgroup (Aria, Mouk, Lamogai and Tourai), three belonging to the Bariai subgroup (Kabana, Kove and Lusi) and Tok Pisin, the general lingua franca of Papua New Guinea. The following is a sketch of the main languages involved: