Crosslinguistic influences
3.10 FIRST LANGUAGE INFLUENCE BENEATH THE SURFACE: THE CASE OF INFORMATION STRUCTURE
It would be inaccurate to imagine that L1 transfer, whether positive or negative, always leads to a direct translation of an L1 form into the L2 or to a glaring absence or excess of a form. The findings by Williams we have just discussed attest to this. More often than not, crosslinguistic influences yield interlanguage solutions that look very different from surface errors of either commission or omission. Often, knowledge of an L1 results in subtle influences that remain beneath the surface and are easy to miss or can be readily misinterpreted or misdiagnosed. We will illustrate this point with a well-researched case of subtle L1 transfer: information structure. Languages are known to be of two kinds with regard to this feature. Topic- prominent languages organize information in sentences through the statement of a topic, followed by the new information. Subject-prominent languages typically organize sentences around a subject and verb and mark topics only when pragmatically needed and through other exceptional means. Thus, for example, Mandarin Chinese, Japanese, Korean and Somalian are topic-prominent languages, whereas Arabic, English and Spanish are subject-prominent languages. In Japanese, topics (whether they are subjects, objects and so on) are marked with the particle wa. In English, by contrast, if we want to structure a sentence around a topic, we may use a construction such as As for my lost wallet, Peter finally found it, or we may use intonation and pauses (Noisy, those neighbours!). Information structure of the L1 can have a profound influence on grammatical and discursive choices learners make in their L2, but the effects are difficult to diagnose. Findings contributed by William Rutherford and Jacqueline Schachter (Rutherford, 1983; Schachter and Rutherford, 1979) provide a well-known illustration.
Certain non-target-like choices are typical of the essays written by Chinese L1 learners of English, as the following sentence from Schachter and Celce-Murcia (1971) shows:
First language influence beneath the surface: the case of information structure 45 (5) There are so many Taiwan people live around the lake
At first blush, this L2 writer appears to have omitted the relative pronoun (who/that/which) of a relative clause she arguably intended. Her writing teacher may think ‘the problem’ with this wording is caused by a relative clause error. Yet, through a detailed analysis of such interlanguage cases, Schachter and Rutherford were able to show that in fact in these cases L2 writers may be unconsciously attempting to organize information following principles of their L1. If they were right in their analyses, then the Chinese L1 writer in (5) may have meant to say something like ‘Many Taiwanese people, they live around the lake’ or ‘As for Taiwanese people, many of them live around the lake’, which is a topic-prominent way of organizing the information in this sentence. Having discovered that English has an existential construction (there is/there are) that enables information to be structured in this way (first topic, then comment), this writer may be skilfully using it to fulfil an L1-induced preference in this area.
More specifically, and if we expand on the analysis proposed by Rutherford (1983), sentence (5) above can be understood as an intermediate solution placed in the middle of the following developmental continuum, from most L1-like to most L2-like information structure:
(6) a. Many Taiwan people, they live around the lake b. There are so many Taiwan people live around the lake c. There are many Taiwanese people who live around the lake d. Many Taiwanese people live around the lake
If Rutherford and Schachter are correct, then, a writing teacher who offers (6c) as a correction for (6b) may ‘teach’ this L2 writer something interesting about relative clauses in English, but the real benefit for this learner would probably be to realize that, in order to sound more natural in her English writing, she needs to begin topicalizing less (reducing the use of sentences like 6a through 6c) and to organize information around subject–predicate sequences more (using more sentences like 6d).
It should be noted that moving in the reverse direction, that is, away from the information structure of a subject-prominent L1 and towards devices that favour a topic–comment structure in the L2, is also difficult for L2 learners. This is what Jin (1994) found in a study of 46 L1 English speakers who were learning L2 Chinese. Participants at early stages of proficiency had difficulty adopting a topic-prominent orientation to syntax and discourse. The same finding was corroborated with 23 L2 Korean learners from an L1 English background by Jung (2004). These learners initially transferred subject-prominent clause structure from English, and only those at the more advanced levels produced the Korean topic marker (n)un to signal given or known information. The Korean double-nominative construction was even more rare. These topicalization devices were underused by beginning L2 learners even though they are common in the target language (as in khokkili-nun
With time and increasing proficiency, the tendency to transfer the information structure of the L1 in order to frame ideas in the L2 may gradually diminish, but the process may be rather slow. This is what Patricia Duff (1993) found in her longitu- dinal study of JDB, a Cambodian learner of English in Canada. JDB was at a stage before the (6a)–(6d) continuum, in that he had yet to discover the existential English construction. Instead, to preserve the topic-prominent information structure of Cambodian, he used the possessive construction have for both possessive and exis- tential meanings, as in (7), which Duff paraphrased as meaning (7a) or (7b):
(7) Khao Larn Camp the King of Thailand they has a small camp about three thousand people
a. ‘There is a Thai camp at Khao Larn with about three thousand people’
b. ‘The Thai government has a camp at Khao Larn with about three thousand people’
This use of have or has as an invariant form that means there is/there are has been amply documented across SLA studies. At the end of two and a half years of observation, JDB suddenly began using an interlanguage form ’s (pronounced as a voiced /z/), shown in this utterance from Duff’s recordings:
(8) And ’s many rock by the river
Duff interprets this form as a precursor of the existential topic marker there is. That is, after two and a half years of immersion in an L2 environment, JDB managed to advance to a stage that has the potential to evolve into stage (6b) in the developmental continuum presented earlier. While this can be considered an important sign of progress in JDB’s interlanguage, it came after a considerable long time of stable use of has, the form that helped him preserve in the L2 the L1- preferred information structure.
Not only is the process slow, but Carroll et al. (2000) argue that information structure in the L1 continues to exert an important if subtle influence on the L2 even at very advanced stages of proficiency. They noted that English and Romance languages will prefer existential constructions for the introduction of new referents as in There is a fountain on the square, whereas German will favour locational constructions of the kind Auf dem Platz ist ein Brunnen (‘On the square is a fountain’). They found that the existential construction preference was transferred by 10 L1 English speakers even though they were at very advanced levels of L2 German. This L1-influenced way of structuring information in their sentences had consequences for other narrative and discourse choices as well.