Input and Evidence: The Raw Material of Second Language Acquisition (review)

Author Susanne E. Carroll explores here the complex relationship among input, evidence, and acquisition. These issues are first discussed against the backdrop of universal grammar (UG) and then are situated within a theoretical framework of speech processing and production. Next, C provides empirical evidence to support the hypothesis that corrective feedback plays an important role in language acquisition. Finally, attention turns to what input actually is—an area long overlooked, as researchers have accepted a vague conception of input without defining it explicitly. Due to space considerations, I comment only on the main points of C’s argument.

Chs. 1 and 2 lay the groundwork for the book by tackling the problem of terminology and definitions. After an explanation of what a second language acquisition (SLA) theory must explain, the author examines current SLA theories: principles and parameters, the competition model, and the autonomous induction model.

Working against the backdrop of UG, Ch. 3 turns to the problem of representation and development in SLA. The core underpinnings of UG—principles and parameters (P&P)—are discussed in detail and are found to come up short because they do not ade quately manage the problem of representational realism. UG does not have a model of triggers either, a fundamental problem for using P&P as an acquisitional model.

Ch. 4 fleshes out C’s autonomous induction model, which is based on the induction theory put forth by Holland et al. (J. H. Holland, K. J. Holyoak, R. E. Nisbett, and P. R. Thagard, Induction: Processes of inference, learning, and discovery, Cambridge, MA: MIT Press, 1986). The theory, which derives its mental models from Ray Jackendoff’s conceptual structures, uses the condition-action rule in a constrained form. C discusses other constraints on the theory and ends by asserting that other types of induction theories cannot be transferred to language learning, as ‘learning grammatical categories is not concept learning’ (174).

Next, C turns to other constraints on i-learning (i-learning is C’s shorthand for her Autonomous Induction Theory). She asserts that changes to the linguistic system are activated by detection of errors which are not parsable by the current system, prompting restructuring. This restructuring can take place only on-line, in activated representations. These constraints and this type of restructuring account for interlanguage.

According to C, ‘the logical problem of (second) language acquisition’ (243) does not exist. She addresses what she calls the ‘empirical’ problem of SLA—the question of how adult learners acquire language based on the input they receive—concluding that current research is not sufficient to claim adult access to UG.

The next chapter addresses the issue of input. Creating a model for the acquisition of the ‘psychogrammar’ (24) must take into account lexical acquisition, the acquisition of morphosyntactic rules, and the like. C then returns to the problem of evidence, the cause for grammar restructuring, concluding that adults can learn exceptions to grammatical generalizations. The chapter ends with a discussion of the limitations of feedback and correction in i-learning.

Feedback and the role it plays in i-learning is the subject of Chs. 9 and 10. Ch. 9 outlines how feedback is used to draw attention to the need for grammar restructuring. Ch. 10 discusses verbal feedback and the need for a certain level of metalinguistic awareness to be present to make feedback usable. Finally, C argues that while feedback and correction are essentials to the theory of autonomous induction, they are not a substitute for language cognition universals.

This volume is a refreshing new look at a longdiscussed problem: language acquisition and the role that input and negative evidence play. C’s theory of autonomous induction is clearly explained, and empirical as well as theoretical support is provided. The work is readable and accessible to researchers at all levels in linguistics and SLA.