- Every boy is riding a pony.
- A boy sang every song.
- Een jongen heeft ieder liedje gezongen.
Comparative study of the processing of universal quantification in SLI and impaired adult SLA
The central research questions for this PhD project concern two related phenomena: (a) the semantic interpretation of distributive determiner universal quantifiers and (b) the scope ambiguity of such quantifiers in sentences containing more than one quantified NP. The first question is simply how SLI /L2 learners understand sentences such as (i). This is a less trivial issue than it may seem: between the ages of 4 and 6, normally developing children will often assign a strikingly nonstandard interpretation to such sentences in experimental contexts: roughly 25% of the time they take a sentence like (i) to be false of a state of affairs in which, say, each of 3 boys is riding a different pony and a fourth pony has no rider at all (see Philip & Lynch 1999).
The second question concerns the degree to which SLI/ L2 learners perceive sentences like (ii) to be ambiguous as to the relative scope of the universally quantified NP. Adult L1s differ in this regard. The English sentence in (ii) is fully ambiguous; it has a (slightly marked) reading under which it asserts that every song was sung by a (possibly) different boy. In contrast, the Dutch sentence in (iii), which is a close translation of (ii), completely lacks this scopal ambiguity; under its only grammatical reading, it asserts that every song was song by the same boy.
Both the nonstandard interpretation of distributive universal quantifiers and the availability or absence of scopal ambiguity with such quantifiers raise important theoretical issues and have spawned a series of competing competence-theoretical hypotheses. For example, the nonstandard child interpretation of (i) described above has been attributed by a number of researchers to hypothesized difficulties applying certain universal pragmatic principles. Others have attributed it to a difficulty maintaining or interpreting the A-bar chains of LF movement. These alternative hypotheses make different predictions for SLI/L2 learners.
As regards the scope ambiguity in sentences like (ii), a large family of alternative syntactic hypotheses attributes the availability of this ambiguity in languages like English, and its absence in languages like Dutch, to core syntactic properties of the respective grammars.
The general aim of this project is to disentangle the linguistic and nonlinguistic factors underlying the semantic interpretation of universally quantified sentences by examining and comparing, in off-line and on-line experimental studies, the comprehension of such sentences in L1 and L2 by language learners with three different kinds of language learning impairment: (a) children and teenagers clinically diagnosed as Specific Language Impaired (SLI), (b) normal adults having severe difficulty in second language acquisition (SLA), and (c) adults engaged in SLA who are classifiable as SLI on the basis of psycholinguistic and nonlinguistic tests in their L1 developed to identify "hidden SLI" in adults (see Tomblin et al. 1992). These three populations are targeted because, SLI individuals having difficulty acquiring an L1 and normal adults having severe difficulty acquiring an L2 both appear to suffer a general deficit in language processing capability, albeit in different ways and to different degrees (for SLI, see Leonard 1999; for SLA, see Gass & Selinker 1994).
Some Key References
Gass, S.M. and L. Selinker (1994) Second Language Acquisition: An Introductory Course. Lawrence Erlbaum: Hillsdale, NJ.
Leonard, L.B. (1998) Children with Specific Language Impairment. MIT Press: Cambridge, MA. Lawrence Erlbaum: Hillsdale, NJ.
Philip, W. and E. Lynch (1999) Felicity, Relevance, and Acquisition of the Grammar of Every and Only. Proceedings of 24th Annual Boston University Conference on Language Development. Cascadilla Press.
Tomblin, J.B. , Freese, P., and N. Records (1992) Diagnosing Specific Language Impairment in Adults for the Purpose of Pedigree Analysis. Journal of Speech and Hearing Research, 35,832.