Interlanguage Variation in Theoretical and Pedagogical Perspective

H.D. Adamson

• Publisher: Routledge
• Year: 2009
• ISBN:
  9781135600822
• Edition: 1
• Topic: Education
• Subtopic: Education General

In this book H.D. Adamson reviews scholarship in sociolinguistics and second language acquisition, comparing theories of variation in first and second-language speech, with special attention to the psychological underpinnings of variation theory. Interlanguage is what second language learners speak. It contains syntactic, morphological and phonological patterns that are not those of either the first or the second language, and which can be analyzed using the principles and techniques of variation theory. Interlanguage Variation in Theoretical and Pedagogical Perspective:

• relates the emerging field of variation in second language learners' speech (interlanguage) to the established field of variation in native speakers' speech
• relates the theory of linguistic variation with psycholinguistic models of language processing
• relates sociolinguistic variation theory to the theory of Cognitive Grammar
• suggests teaching applications that follow from the theoretical discussion

At the forefront of scholarship in the fields of interlanguage and variation theory scholarship, this book is directed to graduate students and researchers in applied English linguistics and second language acquisition, especially those with a background in sociolinguistics.

I
Variation in Native Speaker Speech

1
Variation Theory

Introduction—The Cartesian Mind

The Logical Problem of Learning

Two of the oldest questions in philosophy are the ontological question and the epistemological question. The ontological question asks, “What is the nature of reality?” The epistemological question asks, “How do we know what we know?” Since Plato, philosophers have observed that we know more than we have evidence for. As Bertrand Russell put it, “How comes it that human beings, whose contacts with the world are brief and personal and limited, are nevertheless able to know as much as they do know?” (quoted in Chomsky, 1986, p. xxv). An example of something we know instinctively, without sufficient evidence, is that physical objects continue to exist when no one is looking at them. But the claim that objects do not disappear when no one is looking is not refutable by any observation, as Hume pointed out. The problem of how we know things based on limited evidence has been called the logical problem of learning. The epistemological question and the logical problem of learning were debated by philosophers during the seventeenth century, who came up with some surprisingly modern answers. As Chomsky (1999, p. 36) observes:

These 17th century thinkers speculated rather plausibly on how we preserve the objects around us in terms of structural properties, in terms of our concepts of object and relation, cause and effect, whole and part, symmetry, proportion, the functions served by objects and the characteristic uses to which they are put. We perceive the world around us in this manner, they argued, as a consequence of the organizing activity of the mind, based on its innate structure and the experience that has caused it to assume new and richer forms.

John Locke and Rene Descartes proposed two different answers to the epistemological question, and the debate between their intellectual descendants continues to this day. Locke believed that the mind is like a blank slate and that all of the ideas that it eventually contains are supplied by experience. He wrote:

Let us then suppose the mind to be, as we say, white paper void of all characters, without any ideas. How comes it to be furnished? Whence comes it by that vast store which the busy and boundless fancy of man has painted on it with an almost endless variety? Whence has it all the materials of reason and knowledge? To this I answer, in one word, from EXPERIENCE. (quoted in Pinker, 2002, p. 5)

Descartes, on the other hand, believed that if all human knowledge were based only on experience, there would be no way to be certain about the truths of mathematics, science, or anything else because no two people’s experience is the same. However, he believed that we can be certain of at least one thing: the existence of our own minds: “I think, therefore I am.” A corollary of this proposition is that the mind can also know its own ideas or representations. Knowledge consists of grasping what these ideas are, and working out the connections between them. Many of our ideas, Descartes believed, are imposed on our minds by the force of logical necessity. Such innate ideas included the basic concepts of mathematics and geometry, and, indeed, the idea of God. Other seventeenth-century philosophers took a middle position. Kant agreed with Locke that knowledge of the world is gained through sensory experience, but argued that this experience must be organized by principles that are inherent in the mind. Leibniz expressed the same idea this way: “There is nothing in the intellect that was not first in the senses except the intellect itself” (quoted in Pinker, 2002, p. 34).

Modern psychology has found considerable evidence that supports Kant’s and Leibniz’s position that the mind has innate ways of organizing experience. In a famous set of experiments Spelke, Vishton, and von Hofsten (1995) demonstrated that three- to four-month-old infants have a fairly well-developed ontological theory. They understand what an object is, that objects normally move along a continuous trajectory, and that objects cannot disappear from one place and reappear in another. The method used to demonstrate these facts is extraordinary. Before they can crawl, infants can turn their heads to observe their surroundings, and they like to watch new or unexpected things. They are easily bored and will turn their heads to see new things and will look longer at things that are unexpected. By timing how long infants will look at a scene before turning their heads, researchers can infer what the infants consider new or unexpected. Using this methodology, Baillargeon (1995) showed that infants do not expect one object to pass through another. The babies he studied were fascinated by an animated scene in which it appeared that a panel placed in front of a cube fell flat, right through the space that the cube should be occupying.

Another well-studied example of innate mental organizing principles (which is examined in more detail in the appendix) comes from research on how languages name colors. At first glance color naming systems seem to be remarkably diverse. The language of the Dani of New Guinea has only two color terms (which roughly correspond to “light” and “dark”), whereas English has eleven basic color terms. According to Berlin and Kay (1969), a color term is basic if it is a single morpheme, not derived from another color term (like reddish-brown), and uniquely names a region of the color spectrum. The eleven basic color terms in English are:

black	red	yellow	brown	purple
white		green		pink
		blue		orange
				gray

Berlin and Kay (1969) discovered a remarkable fact about the color-naming systems in all of the languages they studied. If a language has a color term on the right side of the chart above, it will also have the terms to the left of it. Thus, if brown is a basic color term in a language, that language will also have terms for yellow, green, blue, red, white, and black. The reason for this universal order among color systems is the human perceptual apparatus. People perceive six purest colors: red, yellow, green, blue, white, and black. The maximum contrasts between these colors determine which colors will be named in a system with a particular number of terms. For example, if a system has three terms, they will be roughly light, dark, and reddish, which divide the color spectrum up into three maximally contrasting regions. As discussed in the appendix, social factors are also important in how a language names basic colors, but these factors are constrained by principles of color perception and sensory processing that are innate in human beings.

The Logical Problem of Language Acquisition

One aspect of the logical problem of learning involves how we are able to learn a language. The problem applies to first language acquisition in a straight-forward way. How can we explain the fact that children know grammatical patterns that they have not encountered in the input they receive? Evidence of such knowledge was found in a classic experiment by Crain and McKee (1986). They showed that three-year-olds understand that in the sentence, “He was eating pizza while Kermit the Frog was dancing,” he must refer to someone other than Kermit the Frog, say Cookie Monster. This finding is surprising because the reason he must have a referent that is not mentioned later in the sentence is very abstract. At first glance one might think that the reason is fairly simple: a pronoun cannot precede its referent. But this is not the case, as shown by the sentence, “While he was dancing, Kermit the Frog was eating pizza.” The actual reason involves a principle of Universal Grammar (UG) called Binding Principle B, which states that a pronoun cannot refer to an NP that it c-commands. C-command, roughly put, means this: In a tree structure, node A c-commands node B if there is a node C that directly dominates node A and also dominates node B. For example, in the case of the tree on page 8, the first NP (Marsha) c-commands VP and every node within VP. Similarly, V (picked) c-commands Part, the final NP, and every node within the final NP. Chomsky claims that children know Binding Principle B and the other principles of UG innately. Chomsky’s answer to the logical problem of first language acquisition, then, is that children can use grammatical knowledge that they could not have figured out from input because they are born with it.

The logical problem of learning a language applies differently to adult second language acquisition because in this case it is more difficult to claim that learners know grammatical patterns that they have not encountered; after all, they already know a whole language. Therefore, it is not clear what role, if any, UG plays in adult second language acquisition. Three positions are usually distinguished. The fundamental difference hypothesis (Bley-Vroman, 1990) claims that UG is not involved at all in second language acquisition. The evidence cited for this position is that, unlike first language acquisition, second language acquisition is often not successful, as implied by the failure of many language teaching programs to produce fluent speakers. The second position is that UG works for adults in much the same way that it works for children. Supporters of this view (Schwartz and Sprouse, 1996; Shi, 2003) point out that there are many cases of successful adult language learning and that the logical problem of language acquisition applies to these cases just as it applies to children. Therefore, UG must be at work. The third position is that only the principles and parameters employed during first language acquisition are available to aid the second language learner. As an example, consider another principle of UG, the null subject principle, which states that a sentence subject either must appear in the surface structure or need not appear in the surface structure. Like many UG principles, the null subject principle comes with an associated parameter, which can be set to plus or minus. In Spanish, the null subject parameter is set to plus, which means that Spanish allows null subject sentences, that is, it allows sentences without a surface subject, such as: Tengo un libro “I have a book” (literally “Have a book”). In English and French, on the other hand, the null subject parameter is set to minus, which means that these languages require that sentences have subjects on the surface, except in special cases like commands. Those who say that only the principles and parameters settings employed in the first language are available in learning a second language would predict that a speaker of French will be able to learn this aspect of English easily because English has the same setting, but a speaker of Spanish will have more difficulty because Spanish has the opposite setting.

The Developmental Problem of Language Acquisition

Gregg (1996) makes a distinction between the logical problem of language acquisition and the developmental problem. The answer to the logical problem, he maintains, must involve a comprehensive theory of language, like UG. Such a theory is a property theory, which addresses the question, “How is acquisition possible?” A property theory not only explains how language is acquired but also goes a long way toward explaining what language is. “A property theory describes the components that constitute the system, and their interrelations” (Gregg, 1996, p. 51). The developmental problem of language acquisition, on the other hand, involves a transition theory, which addresses the question, “Why does child language or interlanguage change from state A to state B?” UG is not a transition theory. For example, notice that Binding Principle B sets limits on how children construct an internal grammar of pronoun reference, but it does not describe the course of this acquisition in detail, predicting, for example, which pronouns a child will use first. A transition theory of pronoun reference might complement the UG property theory of pronoun reference by describing the order in which a particular speaker or group of speakers acquired pronouns, coupled with an explanation of why that order occurred, an explanation that might involve, for example, the frequency of pronouns encountered in the input.

Gregg (1996) maintains that an adequate transition theory must be associated with a property theory, and that would, no doubt, be desirable. Unfortunately, however, we have no agreed-upon property theory. The property theory that Gregg endorsed in 1996, the Government and Binding version of generative grammar, was in the process of changing fundamentally even as he wrote. Therefore, it seems reasonable to investigate a transition theory of language development independently of any particular property theory, keeping in mind that the two kinds of theories must eventually be compatible. I will claim in the next section that Variation Theory, a way of modeling change in linguistic systems, can help to explain the developmental aspects of first and second language acquisition, and thus serve as part of a transition theory.

Generative Grammar

Variation theory was developed in the 1960s and 1970s, largely by William Labov (1966, 1969, 1972a, 1972b), to explain the relationship between different varieties of the same language. Labov and his colleagues took a special interest in nonstandard varieties, especially African American English, whose speakers often use both nonstandard forms and standard forms in the same discourse, for example, “He play basketball everyday” and “He plays basketball everyday.” It should be noted that similar variation is also common in standard varieties of English (“He’s playing basketball now” and “He is playing basketball now”). As we ...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Preface
• Phonetic Symbols Used in the Book
• List of Figures and Tables
• Part I Variation in Native Speaker Speech
• Part II Variation in Nonnative Speaker Speech
• Part III Variation in Theoretical Perspective
• Part IV Variation in Pedagogical Perspective
• Appendix: Variation and Change in Color Semantics
• Notes
• References