The development of argument analysis skills in children.

• Author: Kline, Susan L.

Researchers across a number of fields have focused their efforts on learning how children master the social and communication skills necessary to manage conflict. For instance, sociologists and linguists are accumulating descriptions of children's disputing practices (e.g., Goodwin, 1990; Grimshaw, 1990; Maynard, 1986; Miller, 1986). Educational researchers are studying techniques that improve children's critical thinking skills (for reviews see Baron & Sternberg, 1987; Chipman, Segal & Glaser, 1985; Nickerson, Perkins & Smith, 1985). Such research efforts, while laudable, are nevertheless constrained by the lack of extensive evidence identifying the extent to which argumentation skills naturally develop, including how children learn to analyze arguments and apply procedures of adjudication. While communication researchers have initiated study of this problem, much remains to be learned (e.g., Benoit, 1984; Haslett, 1983; Katriel, 1985; Willbrand, 1981; Yingling & Trapp, 1985). As a response the present study provides an examination of one aspect of the problem, namely, the development of the ability to analyze arguments.

Speech communication scholars have long recognized that the ability to analyze evidence and reasoning is crucial in formulating one's individual beliefs (e.g., Hample, 1980, 1981), and in communicating persuasively with others (e.g., McCroskey, 1969; Miller & Nilsen, 1966). Yet a number of studies show that adults often reason erroneously about arguments. For instance, adults often reason fallaciously about syllogisms,judging only the appearance features of the syllogism's premises or conclusion (Jackson, 1982; Steinfatt, Miller, & Bettinghaus, 1974). Many adults also reason incorrectly about conditional propositions (Ray & Findley, 1984). Moreover, most scholars recognize that analyzing arguments is not just making judgments about logical form but also knowing how evidence and reasoning are used by arguers (e.g., Kellerman, 1980; Miller, 1966). A functional approach to judging arguments recognizes that argument is public and functions to produce agreement between people within a shared system of knowledge about how to make and have arguments. Hence a functional approach to judging arguments focuses on learning how people come to know what others count as evidence and reasoning, and how people come to know what criteria others use to analyze evidence and reasoning to reach reasoned judgments about argumentative claims.

Our direction in this study is to evaluate children's analyses of arguments with a framework that is able to detect theoretically-based changes in the qualitative features of argument analyses. Specifically our framework integrates Toulmin's (1958) theory of argument with research in social cognitive development that is interpreted with Werner's (Werner and Kaplan, 1963) conception of developmental change. Several lines of work in social cognitive development are interpreted with Werner's Orthogenetic principle (1948), which posits that with age and attendant social experience thought becomes more differentiated, abstract, and integrated. A consequence of this principle is that with age reasoning about social concepts such as fairness, morality, and friendship should and do change in qualitatively distinct ways, with each transformation replacing less functional modes of reasoning with more functional and adaptive modes of reasoning.

Arguments are conceived by Toulmin (1958) as social products and organic wholes, situated in fields of discourse that evolve distinct standards for determining what counts as data and what are acceptable inference principles to use as warrants.

Combining Wernerian theory with Toulmin's ideas, one would expect that children's skill in analyzing arguments includes learning to differentiate arguments within an ongoing stream of discourse, learning to differentiate and use criteria to evaluate evidence and reasoning, and learning to evaluate evidence and reasoning in ways that constitute an adequate refutation, given the expectations of proof within a given argument field. Developmental changes in evidence analysis should be marked by progressive changes in viewing evidence as a distinct component of an argument, with argument soundness not synonymous with one's subjective evaluations, but determined by the valid application of field-relevant criteria. Similarly, developmental changes in reasoning analyses should be marked by progressive changes in viewing warranting premises as distinct components of an argument, with cogency determined by the valid application of field-relevant and field-invariant criteria. Hence children's argument analyses should reflect progressive, systematic changes in the ability to recognize that arguments are not mere evaluations or facets of ongoing behavior, nor are arguments mere depictions of objective facts in the world, but rather that arguments are encoded as units that possess their own criteria for creation and evaluation. Our research is thus aimed at detecting developmental changes along these lines.

There are several reasons for supposing that argument analytic skills should develop in ways consistent with our framework. A first reason comes from the substantial work of Damon, Selman, Turiel and others who show that reasoning about social concepts becomes more differentiated, abstract, and integrated with age. Children's reasoning about social conventions proceeds from viewing conventions as simple descriptions of empirical uniformities to viewing conventions as normative regulations that enable social systems to function effectively (e.g., Shantz & Shantz, 1985; Turiel, 1983). Children's reasoning about friendship, for instance, develops from conceiving friendship in concrete and reciprocal terms to conceiving friendships as evolving social systems (e.g., Damon, 1977; Selman, 1980). Across these studies social reasoning becomes more differentiated, abstract, and integrated with age. It is plausible to suggest that reasoning about arguments develops in a similar fashion, for like social conventions and friendship conceptions, arguments are social constructions that coordinate people's interests and regulate their conduct. Just as social conventions facilitate social interaction and social systems by regulating behavioral norms, so arguments function to regulate consensus. Thus we would expect that children's construal of arguments should develop toward seeing arguments as social objects whose features function rationally to regulate peoples' points of view.

A second line of work suggesting that argument analytic skills should develop systematically with age comes from the work of Berkowitz and his colleagues (Berkowitz & Gibbs, 1983, 1985; Berkowitz, Oser & Althof, 1987), who have analyzed children's moral discussions. Berkowitz and his colleagues find increases in the ability of six- to eighteen-year-olds to summarize, critique, and extend each other's reasoning in discussions over moral dilemmas; older children were better able than younger children to detect when their discussion partner's arguments contained faulty data or reasoning. We expect that these developmental shifts represent not merely changes in analyzing features of moral reasoning, but changes in the way children analyze the evidence and reasoning of arguments in general.

Two other lines of work which suggests that argument analytic skills should develop systematically is that of Kitchener and King (1981; King, Kitchener, Davison, Parker & Wood, 1983) on the development of "reflective" judgement, and the work of Kuhn, Amsel & O'Loughlin (1988; Kuhn, 1988) on scientific thinking skills. Kitchener and his colleagues interviewed high school and college students on their opinions about issues in science, politics, history, and religion, and found that students' understanding of what it means to justify one's beliefs develops systematically, from considering adequate justification as based on knowledge of authorities, to justification based on generalized rules of evidence and inquiry. Kuhn et. al (1988) contend that the development of scientific thinking involves the differentiation and independent evaluation of theory and evidence; their studies show systematic changes in the ways children identify and evaluate scientific evidence, and use evidence to test scientific theory. If the process of relating evidence to a theory is similar to the verbal process of using data and reasoning to establish argumentative claims, then we would expect similar developmental changes in argument analytic skills.

Finally, work from Delia, Clark and their colleagues on the development of communication skills shows that children's functional communication skills develop systematically, consistent with a Wernerian framework (see the reviews of Delia & O'Keefe, 1979; O'Keefe & Delia, 1985). For instance, constructivist research has shown that children's persuasive and comforting messages become more differentiated and listener-adapted with age (Burleson, 1982; Clark & Delia, 1976; Delia & Clark, 1977; Delia, Kline & Burleson, 1979). Children's ability to reason about their message choices also becomes more differentiated and listener-adapted with age (Burke & Clark, 1982; Burieson, 1980). It seems plausible to suppose that argument analytic skills develop like these message production and analysis skills.

The present study was undertaken to analyze developmental changes in two specific argument analytic skills, skill in analyzing evidence and skill in analyzing reasoning. Three age groups spanning diverse levels of cognitive development were chosen for the study and two tasks were created, one in which subjects judged and analyzed evidence, the other in which subjects judged and analyzed reasoning. The arguments we chose for analysis are in van Eemeren and Grootendorst's (1984) terms, "simple" in structure (containing single data-claim units) and come from...