Contact Linguistics. Chap

Exercise:
Revisit the discussion of phonemic differences between English and Hindi in Chapter 2. What difficulties would the MDH predict that L1 Hindi learners would face in acquiring L2 English phonology, and vice versa?

Some predictions of the MDH have been borne out in various studies. (See Eckman 1985: 299 for a brief summary.) For instance, it has been claimed that the MDH explains the order of acquisition of relative clauses in L2 English and Swedish by learners of various language backgrounds. The MDH predicts that the order of acquisition of relative clauses should follow the order of accessibility in the NPAH. Also, it predicts that acquisition would be facilitated in cases where the L1 relativization matches that of the L2. Support for this comes from Hyltenstam’s study of L2 Swedish relative clauses by Finnish, Greek, Persian and Spanish learners, referred to earlier. Eckman (1977) also claimed that the data from Schachter (1974) supported the MDH with respect to acquisition of L2 English relative clauses by Arabic, Chinese, Japanese and Persian learners. However, some questions have been raised about discrepancies in various analyses of relative clause formation in the four languages, which cast doubt on Eckman’s conclusion. Further research may clarify the part that L1 influence plays in these cases. Note however that the explanations of acquisitional development offered by the MDH are quite compatible with those that depend on structural principles such as subjacency, as discussed earlier.

Exercise:
What stages would you predict for the acquisition of question formation (both yes/no and wh-questions) in L2 English? (See Braidi 1999:28-32 for some help.) Do you see any similarity with the case of relative clause acquisition? Find out whether there are typological universals that relate to question formation. What order of difficulty would be predicted by the MDH? (See Braidi 1999:85-86; and Eckman et al (1989) for some discussion.)


I.7.5. Constraints on transfer.

The MDH simply notes that some TL structures will be easier to acquire than others based on the degree of typological similarity between L1 and L2. But it offers no explanation for the kinds of L1 influence that may come into play, or the principles that constrain it. For instance, it’s not clear how the MDH would explain why acquisition is not always accurate in cases of closer typological fit or congruence between L1 and TL. A tacit assumption of this hypothesis seems to be that the learner can simply substitute certain L1 properties or structures for matching counterparts in the TL. But this may lead to either positive or negative transfer.

This kind of L1 influence, based on some perceived congruence between L1 and L2 features (cf. Weinreich’s ‘interlingual identifications’), is the focus of Andersen’s (1983) Transfer to Somewhere Principle. This states that:

A grammatical form or structure will occur consistently and to a significant extent in the interlanguage if and only if (1) natural acquisitional principles are consistent with the L1 structure or (2) there already exists within the L2 input the potential for (mis)generalization from the input to produce the same form or structure. (1983b:182)

A frequently quoted example of this is the fact that English learners of L2 French tend to place object pronouns after the verb in their French IL, on the model of their native syntax. Thus they produce structures like *je vois le, lit. I see him instead of the correct French structure je le vois. “I him see.” The Transfer to Somewhere Principle explains this as follows. Post-verbal placement of full NP objects is required in both English and French, and this provides a model for the post-verbal placement of pronouns by English learners of French. By contrast, French learners find no english model for preverbal placement of pronominal objects, hence they do not produce structures like *I him see.

To take another example, it’s often the case that TL sounds that are perceived as quite similar to L1 sounds prove quite difficult for learners to produce accurately. There is much evidence that learners typically fail to accurately reproduce the acoustic properties of TL sounds that have close counterparts in the L1. Flege (1987) showed that advanced English learners of L2 French were in fact more successful in producing French /y/, which has no English equivalent, than French /u/, which differs acoustically from English /u/. In fact, it seems that in general, learners find it harder to master a sound that is closely similar to one in the L1 than a sound that they have to learn from scratch (Leather & James 1991:313). In other words, interlingual identifications based on learner’s perceptions often lead to inaccurate production. These kinds of mis-identification are the phonological equivalent of the ‘false friends’ we mentioned earlier with respect to L1 influence on the acquisition of L2 vocabulary (section I.2.1). Similar phenomena can be found in morpho-syntax, as we saw in section I.2.3. In all these cases, learners seem to be projecting L1 categories onto TL forms.
This kind of transfer, then, is triggered by some type of congruence, whether semantic, functional or structural, between L1 and l2 elements. We will discuss constraints on transfer further, in section II.5.2 below.) It seems clear from all the above that L1 influence is constrained in various ways by general structural principles.


I.7.5. Cognitive principles and IL development.

The structural constraints on IL development that we have considered so far, whether based on general linguistic principles or typological markedness, are often linked to cognitive principles and constraints of a more abstract psycholinguistic nature. Just as we suggested for the early stages of L2A, these cognitive principles also guide the process of hypothesis testing and rule formulation that brings learners closer to the TL system. The principles are reflected in various processing constraints on the emergence of more complex IL structures. Clahsen (1984:221) suggested that structures requiring more processing would be acquired later than those requiring less. This idea was taken up by Pienemann (1997), whose Processability Theory states that “the learner cannot access …. hypotheses which he/she cannot process” (1997:3). This approach attempts to link a processing model of speech production to a model of grammar construction so as to account for developmental stages in SLA. (See discussion in Braidi 1997:123-127.) Research on the links between processing and structural constraints in l2 grammar construction is still very much in its infancy. But it promises to shed light on the central issue of how universal cognitive and linguistic principles interact with TL input and L1 influence to regulate the path of L2A. This is the ultimate goal of a comprehensive theory of SLA.

This brief sketch of L1 influence on L2 acquisition falls far short of revealing the full complexity of the strategies and operations learners employ in producing some approximation to the TL. The learner’s success depends on a number of factors, cognitive, linguistic and socio-linguistic. We have focussed our attention on those cognitive and linguistic principles that regulate the learning process and help learners to maximize ease of perception and production. We will see in the next section how such principles are relevant to cases of group SLA, that is, language shift by an entire community.
Though the sociolinguistic factors that influence individual SLA are also quite important, we will not consider them in any detail here. Such factors include, for instance, the learner’s age and maturation, motivation to learn, degree of social integration and acceptance among TL speakers, and his/her attitudes toward the TL community and its culture. Factors such as these are also relevant to group SLA, to which we now turn our attention.


Section II. Group SLA or Language Shift.


II. 1. Introduction.

As our brief survey of SLA has shown, we have a fairly good idea of how individuals go about learning a second or foreign language. But what happens when an entire community or group acquires a TL as their second language, or shifts entirely to it? Very often, the outcome of such shift is a contact language significantly different from the original TL. Among the examples we will consider here are Irish English (also referred to as Hiberno-English) and Singapore English. Like other outcomes of group SLA, both show striking linguistic peculiarities due in part to influence from ancestral languages, or processes of simplification in the course of shift. How do such contact vernaculars originate? What factors, structural and non-structural, influence the form they take? Questions like these can be difficult to answer clearly, since all we have to go on are the outcomes themselves, and we seldom have the opportunity to observe the actual process of their formation. We therefore have to rely on what we can find out about the linguistic inputs to the contact situation, and the social context of the contact itself, in order to try to reconstruct this process of language creation. The insights gained from the detailed examination of individual SLA are obviously of great value in this undertaking.

For a start, we can assume that the same structural principles and processes that operate in individual SLA are at work in group SLA as well. Hence we would expect that the various kinds of innovation and creative restructuring found in individual SLA would have their counterparts in the outcomes of group shift as well. But of course the two situations are not identical. The differences between them might be characterized in terms of Weinreich’s well-known distinction between “interference in speech” and “interference in language.” The former he compares to sand carried by a stream, while the latter is “the sedimented sand deposited on the bottom of the lake” (1953:11). In other words, contact-induced changes in individual production is variable and ephemeral, while such changes in language are fixed and permanent. One of our tasks is to explain why some features are selected or conventionalized as part of the communal grammar, while others are discarded.
The creation of new contact vernaculars involves a stage of continuing interaction and competition among individual interlanguage grammars that is eventually resolved into a shared communal system. The actual resolution depends on a variety of sociolinguistic factors, including the demographics of the groups in contact, the extent of inter-group vs intra-group interaction, the length of contact, the power relationship between the groups, their attitudes toward each other, and so on. In order to explore the origins of contact vernaculars, we must have the clearest possible picture of all these factors, as well as of the linguistic structures of the languages involved at the time of the contact (Thomason 1993). Before we explore these issues further, it may be useful to examine particular outcomes of shift and their settings.
In chapter 1, we outlined two broad types of situation in which group language shift can occur. The first of these involves settings in which minority groups form part of a larger host community speaking a different language. Members of such groups usually become bilingual in the host language, or shift entirely to it, often by the third generation. In most of these cases, children born in the third generation acquire native proficiency in the host language. These cases will not concern us here. When demographic and other sociocultural conditions are right, some immigrant groups may develop their own distinctive version of the TL, which they preserve as an in-group vernacular and a badge of separate identity. The variety of American English known variously as African American Vernacular English or Ebonics is such an example. So too are some varieties of English spoken by Chicano and other Latino groups in the US.
The second type of situation can be found in (former) colonies where conquerors have introduced their own language, but have remained a minority. In such cases, the indigenous community acquires its own version of the foreign language, either replacing their ancestral language with it, or employing both in different domains. A well-known example of the second type of language shift is Irish English.


II. 2. Irish English.

English was first introduced to Ireland in the 12^th century by settlers who spoke various regional dialects of English. The major input seems to have come from southern and southwest midlands English (Kallen 1997:140). The contact between English and Irish resulted in a distinctive form of medieval Irish English with substantial influence from Irish. This early contact variety no doubt had some influence on the emergence of modern Irish English. For the next several centuries, English remained very much a minority language in Ireland, with most of the population maintaining Irish as their only language. The English-speaking population grew steadily over the period 1500 to 1700, due to migration of settlers from other parts of Britain, as well as to education and informal second language learning among the Irish (Odlin 1992:182).

British colonial settlements were established in Ulster (Northern Ireland) and eastern Ireland. Large numbers of settlers were introduced to Ulster from Scotland and northern England, while eastern Ireland attracted settlers from various regions of England, especially the south and southwest midlands. Another important factor in the spread of English was the seasonal migration of Irish workers to other parts of Ireland as well as Scotland and England (Odlin 1997a:11). Over the next two centuries, English gradually spread at the expense of Irish, establishing itself first in the towns and eventually throughout most of Ireland. Odlin (1997a) provides a fuller account of the sociohistorical background.
The differences in the settlement patterns as well as the patterns of contact in general resulted in the emergence of two broad varieties of Irish English, Northern and Southern. The former includes the variety known as Ulster Scots, originally introduced by Scottish settlers, as well as two other varieties, mid-Ulster and south Ulster English (Kingsmore 1995:20). The latter two, though they share many features with Ulster Scots, seem to have been more influenced by regional varieties of southern English. Despite their differences, northern and southern Irish English have a great deal in common due to shared superstratal input and substratum influence from Irish (Odlin 1997:29). In keeping with practice in Contact Linguistics, I will henceforth use the term “substratum influence” as a synonym for L1 influence.
The varieties of Irish English that most interest us are those spoken as everyday informal vernaculars, particularly by persons of rural and/or working class background. These non-standard varieties share various characteristics that include continuities from early Modern English dialects as well as features due to Irish substratum influence and other processes of contact-induced change. The effects of such change are most noticeable in parts of Ireland once having a high degree of bilingualism, and in conservative rural speech (Harris 1984:305).
Examples of early Modern English dialectal continuities in Irish English include the variable absence of subject pronouns in relative clauses, as in (20) and the use of for to as a complementizer where Standard English uses to, as in (21)

(20) I know a girl works in that shop.

(21) It’s not easy for to do that.

Structures like these are still found in some contemporary regional dialects of England and Scotland, and Irish English presumably adopted them from the ancestors of these dialects.

Many other distinctive lexical and structural characteristics of Irish English are also direct continuities of features found in earlier English dialects, including literary varieties. They include demonstratives like them ‘those’; yon/thon ‘that/those further away’; use of concord –s with both singular and plural 3^rd person subjects (the women knows; the men is/was); reduction of “strong” verb paradigms from a 3–way to a 2-way distinction (go/went/went vs S.E. go/went/gone); multiple negation (I never said nothing to nobody); positive anymore used in the sense of ‘nowadays’ as well as ‘from now on’ (Wool is expensive anymore; We can do our homework on this [desk] anymore, can’t we?). Example sentences are from Kallen 1997:153). For a more detailed account of these and other continuities from earlier English, see Harris (1993).
Many lexical items that are now rare or obsolete in British English can still be found in contemporary Irish English. Some of these came from Scotland and northern England, for example palatic ‘drunk’’ skite ‘drinking session, party’, etc. Others came from southern England, for example ledge ‘a row of mown hay’, pill ‘river tributary’ etc. (Kallen 1997:146-147).


II. 2. 1. Substratum influence on Irish English.

There seems to be general agreement that, while Irish English is based on mid-17^th century English dialects, many of its features are due to influence from Irish, the first language of its learners. As Bliss (1972:63) put it:

This seventeenth century English was acquired gradually and with difficulty, by speakers of Irish; and in the process of their acquisition of it they modified it, both in pronunciation and syntax, toward conformity with their own linguistic habits.

This L1 or substratum influence is apparent at all levels of Irish English structure. In phonology, the following are among the characteristics can be found, especially in more conservative rural speech. Interdental fricatives /Q/ and /D/ are realized as affricates, that is, they are co-articulated with alveolar or dental stops as /tQ/ and /dD/ respectively. Velar stops tend to undergo palatalization more frequently than they do in other varieties of British English. Alveolar fricative /s/ is produced with fortis articulation. The labio-dental fricative /f/ and /v/ have the allophones /F/ and /B/ respectively. All of these have counterparts in Irish (Odlin 1997:30).

The morpho-syntactic features ascribed to Irish influence include tense/aspect categories like ‘hot news’ perfect and Habitual be; the imperative expressed by let, and peculiar uses of prepositions, as in Hugh is buried with years (= has been buried for years) (Harris 1993: 171-73). These features are not exact replicas of those in Irish, but represent compromises between English and Irish structure. For instance, the overall organization of the tense/aspect system of Irish English is a blend of continuities from both source languages. I.E. preserves earlier English perfect constructions like the following:

(22) She’s nearly her course finished. (Harris 1984:307)

‘She’s nearly finished her course.’

It also preserves other categories such as present, past and future tense, progressive and past habitual aspect, etc. But, as noted above, the ‘hot news’ perfect and the present habitual are clearly due to Irish influence.¹⁵ The former is expressed by a construction consisting of be + after + V-ing, as in the following:

(23) She’s after selling the boat

She’s just sold the boat.’

Irish has a very similar construction, exemplified in the following:

(24) Tá sí tréis an bád a dhíol.

Be+nonpast she after the boat selling. Harris 1984:319).

As can be seen, the Irish English construction conforms to English rather than Irish word order, but the Irish influence is clear. Similarly, I.E makes a distinction between the simple present copula and habitual be, as in the following:

(25) a. She’s here now.

b. She be’s here often. ‘She’s often here.’

The same distinction is found in Irish, as in:

(26) a. Tá sí anseo anois.

Be+nonpast she here now.
b. Bíonn sí anseo go minic
be+nonpast+hab. she here often

I.E. also employs do as a present habitual auxiliary with be and non copular verbs, as in the following examples from Harris (1986:176):

(27) a. They do be fighting among other.

‘They’re usually fighting among themselves.’
b. He does help us. He does plough the field for us.
‘He helps us. He ploughs the field for us.’

Present habitual do appears to derive from simple periphrastic do in 17^th century English dialects, especially those of the southwest, which were a major input to Irish English. The reanalysis of do as a habitual auxiliary seems to have been motivated both by the fact that it often conveyed habitual meaning, and the fact that Irish had a distinct habitual category that provided a model for its reinterpretation (Harris 1986:180). A similar reinterpretation of do occurred in Bajan (Barbadian English) for much the same reasons (See Chapter 9, section 4).

Several other distinctive features of Irish English syntax can also be attributed to compromise between English and Irish structure, and to congruence between the inputs from the two sources. For instance, I.E has a distinctive use of and as a subordinating conjunction, which appears to have counterparts in British English dialects (Filppula 1991, Odlin 1992). However, Odlin (1992) argues that these constructions are modeled primarily on similar constructions in Irish which are introduced by agus, the Irish equivalent of and. The following examples from Odlin (1992:186-7) illustrate:

(28) a. Irish. Agus é san IRA, phós sé Albanach buí

And him in-the IRA married he Scot yellow
“He married an orange [Protestant] girl while in the IRA.

b. Irish E. The sergeant ran for his life. And he going out over the wall, he hit against a tomb.

Irish substratum influence is also apparent in other areas of I.E. grammar, including cleft constructions (e.g., It was painting I was) and use of reflexive pronouns where SE doesn’t permit the (e.g., I don’t remember himself) (Odlin 1997b:40-41) Also of interest are various idiomatic expressions (Odlin 1991), and the use of sorrow and devil as negators, as in the following examples (Odlin 1995, 1996).

(29) a. Divil a one ever I seen.

“I never saw one.”
b. Sorra one o’ them was equal to Charlie.
“Not one of them was equal to Charlie.”

Odlin (1997a:14) suggests that these forms of negation are the result of combined Scots superstrate and Irish superstratum influence.

II. 3. “Indigenized” Englishes and similar contact varieties.

We have used Irish English as an illustration of the kinds of changes that characterize contact vernaculars that arise in scenarios of group shift involving influence from the group’s native language(s) and other processes of change. Similar kinds of restructuring can be observed in all of the “new” or “indigenized” Englishes that emerged in other colonial settings. We include here other situations in the British Isles that involved contact between (speakers of) English and indigenous languages like Welsh and Scottish Gaelic, out of which emerged contact varieties such as Welsh English and (especially Highland) Scots. Interestingly, these varieties are seldom highlighted in discussions of the ‘new’ colonial Englishes, though they clearly form part of this category (Kachru 1992:231). English remained confined to the British Isles until the end of the 16^th century. As a result of colonial expansion from the early 17^th century on, English came into contact with Niger Congo languages in Africa, Indic and Dravidian languages in India, and Sinitic and Austronesian languages in South-east Asia and the Pacific. Contact varieties of English emerged in all these settings, serving as lingua francas to ethnically and linguistically diverse populations.

We will not be concerned here with those colonial settings that are associated with the emergence of pidgin and creole languages. These will be dealt with in subsequent chapters. Our concern is with situations where the spread of English was linked to the spread of English-medium education, for instance in Ireland, India, Southeast Asia, parts of Africa and areas of the Pacific such as Fiji and Papua New Guinea. This occurred during the course of the late 18^th to 19^th centuries, in the later phases of British colonial expansion, when the British acquired trading and political control of these areas. In addition to India, countries like Singapore, Malaysia and African nations like Kenya and Nigeria are good examples of such situations. It was socially and economically rewarding to learn English, and local varieties of the language became prestigious second languages for the more educated and elite sections of the community. In the course of time, the new Englishes were increasingly adopted as everyday vernaculars for use in inter-group, and later, in-group communication.
Of course, English was not the only language to be subjected to these forces of change in colonial settings. Similar contact varieties of French, Spanish, Portuguese and other languages emerged in their respective colonies. Non-European languages have also undergone similar processes of change under conditions of shift. Taiwanese Mandarin, heavily influenced by Taiwanese, a southern Min dialect of Chinese, is one such example. All these situations share in common the fact that the shifting group has a strong sense of its autonomy and separate identify vis à vis the TL group. This motivated the emergence of a distinctive local variety of the TL that serves as an in-group vernacular, and symbolizes the group’s ethnic or national identity.