Cognitive and Behavioral Approaches to Language Acquisition: Conceptual and Empirical Intersections

Dale, Rick, The Behavior Analyst Today

The past 20 years have seen research on language acquisition in the cognitive sciences grow immensely. The current paper offers a fairly extensive review of this literature, arguing that new cognitive theories and empirical data are perfectly consistent with core predictions a behavior analytic approach makes about language development. The review focuses on important examples of productive linguistic behavior: word learning and early grammatical behavior. Language experience, through social and other contingencies, influences language development directly. Through these contingencies, the structure of language behavior exhibits a gradual emergence.

Keywords: language acquisition; linguistic behavior; grammar; syntax learning; word learning.

**********

The modern mainstream study of language acquisition has a surprising quality. Despite decades of intense effort collecting observational and experimental facts, debate still rages as to the nature of the language learning child, and perhaps even the nature of language itself. This is somehow made more surprising when we consider the billions of in -home "laboratories" around the world permitting ever-present observation of a child's development. Indeed, language researchers themselves doubtless diligently observe the growth of language in their own children (e.g., Tomasello, 1992). Nevertheless, these observations provide a stark portrait of the debate on language learning: Even given intense observation, individual observers can differ wildly on appropriate theoretical interpretations.

There is, however, one very important consensus among most developmental psycholinguists: There is no substantial explicit negative feedback about grammar in language directed to children. An example of such feedback is presented below, and has two important properties that should be borne in mind. First, the mother must stop the flow of conversation to address the child's error, and secondly, the mother is providing information about language structure in particular - the conversation has ceased to be about the topic of discussion, but instead about the grammaticality of the child's utterance.

Child: Mummy I have toy!

Mother: No, say, "I have a toy."

Brown and Hanlon (1970) famously demonstrated in the Adam, Eve, and Sarah mother-child interaction corpora that parents do not offer such information. In tracing the development of complex sentence types in language production by children, Brown and Hanlon sought the basis on which children establish correct usage of these sentences. They discovered that information based on explicit approval and disapproval is extremely rare. The feedback they did isolate was directed towards semantic and phonological problems, with morphological and syntactic errors almost never eliciting it. Driven by early results such as these, many were eager to identify child-directed speech as altogether chaotic: "A record of natural speech will show numerous false starts, deviations from rules, changes of plan in midcourse, and so on." (Chomsky, 1965, p. 4) This famous quote exemplifies an ideological momentum at that time to characterize the child's task as rife with difficulty and void of a certain "linguistic pedagogy" that behaviorist theories supposedly required. To this day, linguistics textbooks are persistent in presenting this tendency (e.g., Fromkin & Rodman, 1997; see Schlinger, 1995, p. 179 for relevant discussion). Even if it were present, some argued, a number of famous anecdotal examples supposedly demonstrate that children might not use it to change their language anyway:

Child: Nobody don't like me.

Mother: No, say, "Nobody likes me."

Child: Nobody don't like me.

Mother: No, say, "Nobody likes me."

Child: Nobody don't like me.

Mother: No, say, "Nobody likes me."

...

Mother: Now listen carefully, say, "Nobody likes me."

Child: Oh! Nobody don't likes me. (McNeill, 1966)

Child: Want other one spoon, Daddy.

Father: You mean, you want the other spoon.

Child: Yes, I want other one spoon, please, Daddy.

Father: Can you say "the other spoon"?

Child: Other ... one ... spoon.

Father: Say ... "other."

Child: Other.

Father: "Spoon."

Child: Spoon

Father: "Other ... Spoon ."

Child: Other ... spoon. Now give me other one spoon? (Braine, 1971)

Through Brown and Hanlon's analysis and anecdotes of this kind, the theoretical influence of this consensus was quick and severe. For example, it rapidly became assumed by many that any kind of negative evidence is absent in language input (Valian, 1999; Moerk, 2000). Also, formal theorists utilized the absence of negative evidence to support nativist theories of language acquisition (e.g., Wexler & Culicover, 1980), partly on the basis that simple inductive learners would not be able to accomplish the task without negative evidence (a classic example is Gold, 1967). One of the more influential and still-present intellectual tools generated by this debate is "poverty of the stimulus" arguments, a modern version of a kind of reasoning that finds a famous and early expression in Plato's Meno: Precocity, without appropriate input, entails rich epistemological innateness (Chomsky, 1986). This logic continues to exert an influence in current research:

   Nonetheless, we show that 18-month-old infants do have command of
   the syntax of one. Because this syntactic knowledge could not have
   been gleaned exclusively from the input, infants' mastery of this
   aspect of syntax constitutes evidence for the contribution of
   innate structure within the learner in acquiring a grammar. (Lidz,
   Waxman, & Freedman, 2003, p. B65)

The authors refer to this argumentation specifically as poverty of the stimulus "logic." A casual glance at the cognitive science literature reveals its persistent frequency (e.g., Collins, 2003; Crain & Pietroski, 2001; and even outside language in other realms of cognitive development, Spelke, 1994).

Such theoretical consequences took root before further thorough analyses of language-learning corpora, records of parent-child interaction, were conducted. They were quickly taken as foundational in many theories of language acquisition (Moerk, 2000). Some felt that this axiom was adopted far too hastily given the small size of the dataset and the nascence of our quantitative instruments (Bates & Carnevale, 1993; Moerk, personal communication). To those who opposed this momentum, it is not that surprising that mothers should have goals in mind for their children's development that lie outside finicky pedantry about language structure. Language learning is embedded in social contingencies, where everything from phonology to syntax, comprehension and production, are all part of complex dynamics among caregivers, the wider social environment, and the language -learning child. Nevertheless, many linguists since Chomsky have assumed that learning to understand sentences proceeds by "setting a grammar" through exposure to "sentence examples." Many find this idealization egregious, neglecting the requisite richness of the learning context in which children find themselves (e.g., Halliday, 1975).

An apocryphal yet common interpretation of Skinner's Verbal Behavior is that language acquisition hinges mostly on explicit negative feedback of this kind (e.g., Sokolov & Snow, 1994). In early language behavior, the "specifications upon which reinforcement is contingent are at first greatly relaxed" (Skinner, 1957, p. 29), and gradually establishing contingencies "arranged between a verbal response and a generalized conditional reinforcer," influencing responses that have "dynamic properties similar to those which it would have acquired if it had been severally followed by all the specific reinforcers at issue." (p. 53) Skinner's framework makes ample room for subtle control. Any bout of language behavior is indeed richly controlled by environmental contexts. Skinner clearly acknowledged this by devoting an entire chapter to it. From the contingencies dictated by simpler reinforcers, early "ends" of conversational "means," such as food and warmth, contingencies based on generalized patterns of behavior can develop and mutually constrain the further development of language. Skinner thus motivates the expectation that parents do not have to stop their children with every grammatical anomaly. In a recent review of these issues, Schlinger (1995) articulates the modern behavior analytic approach to development and language acquisition, and addresses common misconceptions about reinforcement. For example, mainstream textbooks on language learning often invoke "criticisms of reinforcement as a viable language learning process [that] are based on misunderstanding of the concept itself." (p. 181) The author argues that an appropriate conceptualization of learning in a behavior analytic framework does accommodate facts about language learning. The conclusion in the budding cognitive sciences that behavior analysis has little to contribute to understanding language was premature, if not false (see also Bijou & Ribes, 1996).

The current paper reviews recent research in developmental psycholinguistics relevant to behavioral approaches that aim to elucidate the complexity of human learning and behavior. The recent spate of debate concerning higher-order operant approaches would benefit from the examination of current evidence to further refine theory - and support it. In hope of contributing to this effort, I argue below that extensive empirical cognitive work on language acquisition is quite supportive of a behavioral account. Whether one adopts naming (Horne & Lowe, 1996), Sidman's (Sidman & Tailby, 1982; Sidman, 2000), or relational frame theory (Hayes, Barnes-Holmes & Roche, 2001), these data should guide the evolution of theories based on equivalence and other potential derived behavior (e.g., "syntax" as sequential classes, Lazar, 1977; Wulfert & Hayes, 1988; relations, Hayes et al., 2001; relations of relations, Stewart, Barnes-Holmes, & Roche, 1999).

I will begin by laying out some predictions of the behavioral perspective. These predictions will be very broad, expressed in general patterns of verbal behavior that would be expected from a behavior analytic perspective. Next, I offer a selective review of two crucial stages of language development, word learning and particularly early grammatical behavior. These two stages are often attributed to the emergence of truly creative and productive behavior.

Predictions

I will not venture from the common and perhaps "trendy" dichotomy (Michael & Malott, 2002) between the function and structure of the child's language behavior. From a Chomskyan perspective, accounting for language acquisition requires no sense of function whatsoever--such issues are thought to be peripheral to language acquisition (Chomsky, 1986) and not needed for understanding the learning of language structure (Wexler & Cullicover, 1980). Some psycholinguists sometimes make strong statements that "... to a surprising degree, language is the product of the young human brain, such that virtually any exposure conditions short of total isolation and vicious mistreatment will suffice to bring it forth in every child" (Newport & Gleitman, 2002, p. 685). The child is setup for language, and need not necessarily use it extensively to acquire it. Rather from a behavioral perspective, "input" is crucial. Skinner's (1957) own definition requires control by contingencies in the surrounding verbal community. From here, controlling variables of acquisition can become richly multidimensional (Skinner, 1957, chap. 9), perhaps taking on many forms, from such primary reinforcers as food (e.g., Issacs, Thomas & Goldiamond, 1960) to generalized conversational reinforcers such as "yes" and "uh huh" (Greenspoon, 1955; see Risley, 1977, for an excellent early review). More recently, collateral reinforcement in a game task (Liebermann et al., 1998a, 1998b; …

Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Cognitive and Behavioral Approaches to Language Acquisition: Conceptual and Empirical Intersections. Contributors: Dale, Rick - Author. Journal title: The Behavior Analyst Today. Volume: 5. Issue: 4 Publication date: Fall 2004. Page number: 336+. © 2007 Behavior Analyst Online. COPYRIGHT 2004 Gale Group.

This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.