INTERLANGUAGE BEFORE GRAMMATICALIZATION: THE BASIC VARIETY OF NATURALISTIC LEARNERS

We know much about the acquisition of an L2 when learners begin from zero L2 knowledge and without the aid of instruction, that is, when interlanguage

development must proceed solely under naturalistic circumstances. A main contribution to this knowledge comes from a large-scale investigation conducted in the 1980s under the auspices of the European Science Foundation (ESF) and led by Wolfgang Klein and Clive Perdue (Perdue, 1982; Klein and Perdue, 1997). The ESF project had two remarkable strengths: it was longitudinal and also crosslinguistic.

Forty adults who were immigrants in five different European countries were studied over two and a half years, as they acquired one of five L2s: Dutch, English, French, German and Swedish. The resulting data amounted to a dense and longitudinal corpus of 15,000 pages of L2 oral transcriptions, spanning 30 months of study (Klein and Perdue, 1997). The combinations of L1 and L2 investigated were determined in part by the political realities of immigration in European countries at the time, which explains the narrow range of L2s included. This was more than compensated, however, by the purposeful ‘two L2s by two L1s’ design, shown in Figure 6.1. You can see how the crosslinguistic insights about interlanguage development yielded by such a design can be deep. In the end, however, the ESF study uncovered strong universal patterns rather than large crosslinguistic particularities.

Over the two and a half years, all learners in the project showed evidence of developing a rudimentary but systematic and fully communicative interlanguage system that was called the Basic Variety by Klein and Perdue (1997). Its main characteristics are summarized in Table 6.3. In a nutshell, the Basic Variety can be described by recourse to a few simple principles for how utterances must be structured (i.e. phrasal constraints) and how constituents must be ordered and information organized along pragmatic and lexical resources (i.e. semantic and pragmatic constraints). The Basic Variety shows no evidence of grammaticalization of resources, that is, it makes little use of morphology or subordination. A good case in point is the expression of temporality, which in the absence of morphology needs to rely on pragmatic and lexical resources, as summarized in Table 6.3 (see also Table 6.5).

We may wonder what, if anything, may make naturalistic learners move beyond the Basic Variety, that is, beyond the reliance on pragmatic and lexical principles and towards the grammaticalization of language resources. Klein and Perdue (1997) posited that it may be the increasingly more pressing need to express complex thoughts and the challenge of putting ideas into words when the concepts

English L2 German L2 Dutch L2 French L2 Swedish L2

Punjabi L1 Italian L1 Turkish L1 Arabic L1 Spanish L1 Finnish L1

Figure 6.1

Interlanguage before grammaticalization 123 Table 6.3

The Basic Variety summarized (Perdue, 1982; based on Klein and Perdue, 1997)

Area of analysis Main findings

Lexicon • Mostly noun-like and verb-like words, plus a few adjective-like and some adverb- like words

• There is no inflection, words occur in an invariant form

• Closed words (demonstratives, pronouns) are only a few and mostly with lexical rather than grammatical meaning

• Other closed words (articles, conjunctions and pronouns) are rare if at all present Phrasal constraints on • There are no complex structures, no syntactic movement

utterance structure • Utterances can be verbless or with a non-finite (non-conjugated) verb; utterances with finite or conjugated verbs never occur

• Verbless utterances are made up of a noun phrase followed by another constituent • Non-finite utterances can appear in three basic forms: (1) Noun phrase plus verb

(plus optional noun phrase); (2) Noun phrase plus copula plus another element; (3) Verb or copula plus noun phrase

• With verbs of saying and giving, three arguments can appear (e.g. the sayer, the hearer and the said); the Basic Variety does not have four-argument utterances • ‘Controller first’: The noun phrase with the highest degree of control (more

agentivity, depending on the semantics of the verb) comes first (there is no real concept of ‘subject’ in the Basic Variety)

• ‘Focus last’: The Basic Variety structures information mostly through word order; topic comes first, focus comes last

• Maintained information can be marked by zero anaphor (or rarely a pronoun) but only if the entity maintained is highest in control and no other agentive entity is in the same utterance; otherwise full noun phrases are used

Temporality • Tense and aspect are marked pragmatically and lexically, not grammatically (an invariant form or ‘base form’ of each verb is used: the bare stem, an infinitive, sometimes–ing form for English)

• The pragmatic marking is realized through the principle of chronological order: ‘recount events in the order they occur’ (e.g. I went home and had dinner versus I had dinner and went home)

• The lexical marking is realized through a rich repertoire of adverbs • Calendric (Sunday) and anaphoric (after, before) adverbials are abundant • Anaphoric (yesterday), frequency (always) and durational (two hours) adverbs are

less developed

• Adverbs denoting two reference points (again) are absent Pragmatic constraints on the organization of information in connected text Semantic constraints on case role assignment

involve conflicting semantic and pragmatic conditions. This meaning-making pressure apparently worked for two-thirds of the 40 naturalistic learners investigated in the ESF project. However, it did not seem to be enough for the remaining third, who were seen to reach a plateau after a year and a half, sometimes earlier, and had not progressed beyond the Basic Variety by month 30, at the conclusion of the study.

6.9 PATTERNED ATTAINMENT OF MORPHOLOGICAL ACCURACY: THE