《Developmental Origins of Scientific Thinking.doc》由会员分享，可在线阅读，更多相关《Developmental Origins of Scientific Thinking.doc（18页珍藏版）》请在三一办公上搜索。
1、JOURNAL OF COGNITION AND DEVELOPMENT, 2000, Volume 1, pp. 113129Copyright 2000, Lawrence Erlbaum Associates, Inc.Developmental Origins of Scientific ThinkingDeanna KuhnDepartment of PsychologyTeachers College, Columbia UniversitySusan PearsallDepartment of PsychologyTeachers College, Columbia Univer
2、sityIdentifying the developmental origins of scientific thinking, as well as its endpoint, provides an essential framework for understanding its development. The origins of scientific thinking are claimed here to lie in attainments in epistemological under- standing, beginning with the understanding
3、 achieved at about 4 years of age that as- sertions generated by human minds are distinguishable from an external reality against which they can be compared. Despite this achievement, children between 4 and 6 years of age exhibit an epistemological category mistake regarding the source of knowledge.
4、 They confuse a theory making it plausible that an event occurred and evi- dence indicating that the event did occur, as the source of their knowing that the event occurred. Appreciation of this distinction develops rapidly during this age range and reflects increasing mastery of an epistemological
5、understanding we argue to be of foundational status for the development of scientific thinking, defined here as the con- sciously controlled coordination of theory and evidence.Authors of several recent review chapters (Haith & Benson, 1998; Keil, 1998) make the observation that the field of develop
6、mental psychology has not been well served by a focus on identifying competencies at earlier and earlier ages. Such stud- ies reveal little about the developmental processes involved in attainment of theRequests for reprints should be sent to Deanna Kuhn, Box 119, Teachers College, Columbia Univer-
7、sity, New York, NY 10027. E-mail: dk100columbia.edu114KUHN AND PEARSALLcompetency, or about how the competency identified at an early age differs from or evolves into the competency in its more mature forms. Researchers efforts, it is ar- gued, would be more productively focused on understanding dev
8、elopmental pro- cess and the states of “partial accomplishment” (Haith & Benson, 1998) that mark its progress.Studies of partial accomplishment need to be undertaken within a conceptual framework that identifies both developmental origins and endpoints. Scientific thinking, the topic of this article
9、, is problematic in this respect, as neither origins nor endpoint have been clear. The intellectual skills associated with scientific thinking are ones valued in rigorous thinking more broadly (Kuhn, 1996). It is, thus, unclear whether it is the cognitive competencies of the professional scientist(w
10、hich only recently have become a focus of empirical study) or of the mature lay person or intuitive scientist that are to be taken as an endpoint (Klahr & Carver,1995; Kuhn, 1995). A clearly conceptualized endpoint narrows the search for ori- gins and together they define what it is that develops.Ou
11、r objective here is to identify the developmental origins of scientific think- ing. A number of claims have been made regarding early forms of scientific think- ing detectable in young children, a few based on experimental data (Ruffman, Perner, Olson, & Doherty, 1993; Sodian, Zaitchik, & Carey, 199
12、1) and many oth- ers in more educationally oriented literature highlighting young childrens skills in activities such as observation, description, and classification. Our interest in pur- suing the question of developmental origins centers on whether the early child- hood years are characterized by
13、any cognitive achievements specific and central to the development of scientific thinking, possibly to the extent of defining its es- sence. Classification, for example, is a more general cognitive capability with roots evident in infancy (Langer, 1980), and which is clearly necessary for scien- tif
14、ic thinking, but at the same time does not define its essence. Are there, in con- trast, early attainments that represent the rudiments of scientific thinking and provide the foundation for its further development?SCIENTIFIC THINKING AS THE COORDINATION OF THEORY AND EVIDENCEThe conceptualization of
15、 mature scientific thinking that has guided our work is one sufficiently broad to encompass skilled thinking both within and outside of profes- sional science. It is also broadly compatible with the conceptualizations of others who have studied the topic (see DeLoache, Miller, & Pierroutsakos, 1998;
16、 Klahr,1999; Kuhn, Garcia-Mila, Zohar, & Andersen, 1995; Moshman, 1998, for review). The essence of mature scientific thinking, we claim, is the coordination of theory and evidence in a consciously controlled manner. The qualifier “in a consciously controlled manner” is essential because the thinkin
17、g of even very young childrenSCIENTIFIC THINKING115has been claimed to involve the construction of theories, as a means of understand- ing the world, and their revision in the face of evidence (Wellman & Gelman,1998). It does not follow, however, that young children have explicit awareness of their
18、theories or of the fact that they are undergoing revision. Indeed, it is unlikely that they do, and research we review later in this article supports this assumption. Mature scientific thought, on the other hand, is characterized by a now widely accepted postpositivist philosophy of science as invol
19、ving the examination and in- terpretation of evidence within a theoretical framework that shapes all phases of scientific activity (Kitcher, 1993). Thus, both young intuitive scientists and mature professional scientists make use of both theory and evidence in their thinking. That is not where the d
20、ifference between them lies. Rather, the difference is that in the case of the mature scientist, the coordination of theory and evidence is carried out under a high degree of conscious control (and therefore explicit, consistent, anddemanding criteria).Accordingly, the development in scientific thin
21、king believed to occur across the childhood and adolescent years might be characterized as the achievement of increasing cognitive control over the coordination of theory and evidence. This achievement, note, is metacognitive in nature because it entails mental operations on entities that are themse
22、lves mental operations. In this respect, Inhelder and Piagets (1958) characterization of scientific thinking as involving “operations on operations” is correct. Contrary to their depiction of second-order operations as emerging at adolescence, however, it has subsequently been recognized that metaco
23、gnitive thinking about ones own thought begins to develop much earlier(Brown, 1997; Kuhn, in press-a)a fact highlighted by studies examined in this article and, in particular, those coming from research on theory of mind (see Flavell, 1999, for review). Empirical research on scientific thinking, in
24、contrast, has shown that it may not develop fully even by adulthood (Kuhn et al., 1995). One of the purposes of this article is to connect these two bodies of researchone em- phasizing early competence and the other later lack of competence.Empirical studies of scientific thinking skills can be divi
25、ded into two categories corresponding to two broad phases of scientific activity: the investigative, in which experiments are designed and evidence sought and the inferential, in which the re- sulting evidence is interpreted. Inferential skills entail a path from evidence to the- oryof interpreting
26、various patterns of evidence, drawing conclusions, and exhaustively considering alternative conclusions consistent with the evidence. In- vestigative skills, in contrast, entail a path from theory to evidence, in particular in the design of experiments that have the potential to yield informative ev
27、idence bearing on the theory. As Klahr (1999) notes, few researchers have investigated this full range of skills within a single study. Nonetheless, as our own research and that of a number of others (Amsel & Brock, 1996; Klahr, Fay, & Dunbar, 1993; Koslowski, 1996; Kuhn, in press-b; Kuhn, Amsel, &
28、OLoughlin, 1988; Kuhn et al., 1995; Kuhn, Schauble, & Garcia-Mila, 1992; Penner & Klahr, 1996; Schauble,116KUHN AND PEARSALLTABLE 1Four Types of Theoretical ClaimsTypeExampleSupporting EvidenceDisconfirming EvidenceT1. Category claimPlants are living things.Plants share the category characteristics
29、of living things.Plants lack some category characteristics of living things.T2. Event claim This plant died.The plant has turned brown and is not growing (i.e., the plant no longer shares category characteristics of living things).The plant continues to show the category characteristics of living th
30、ings.T3. Causal or explanatory claimT4. Explanatory system claimThe plant died because of inadequate sunlight.A multivariable process of photosynthesis maintains plant life.Plants not exposed to sunlight die, and most plants exposed to sunlight remain alive (i.e., covariation between alleged causal
31、antecedent and outcome).Plants not exposed to the multiple conditions required for photosynthesis do not remain alive.Some groups of plants not exposed to sunlight remain alive (i.e., lack of covariation between alleged causal antecedent and outcome).Plants not exposed to the multiple conditions req
32、uired for photosynthesis remain alive.1990, 1996) has established, weaknesses are common among lay adults, as well as adolescents and children, with respect to both sets of skills. The investigation of origins may offer some insight into later lack of development of scientific think- ing, a possibil
33、ity we return to.What Is a Theory?If the coordination of theory and evidence is regarded as the essence of scientific thinking and we wish to identify its developmental origins in young children, it is essential, as a first step, to define our terms precisely. What counts as a theory, what counts as
34、 evidence, and what must one do to coordinate them? What counts as a the- ory has in fact been the topic of debate in cognitive development literature(Wellman & Gelman, 1998), with the term used in different ways within different research traditions.In Table 1 we portray four possible uses of the te
35、rm theory, each of which has been implicit in one or more lines of research, beginning with the least stringent(T1) and progressing to the most stringent (T4) definition of the term. In the sec- ond column we give examples of each theory type, and in the two final columns we give examples of relevan
36、t evidence capable of supporting or disconfirming the theory. Evidence is defined as empirical observations distinguishable from the the- ory and bearing on it.SCIENTIFIC THINKING117T1 and T2 are simple knowledge claims, often referred to as beliefs. The two types are not entirely distinct from one
37、another, as many T1 claims can be con- verted to T2 claims and vice versa. For example, the T2 claim This plane crashed is convertible to the T1 claim This plane is in a category of planes that have crashed. In contrast to the first two types, T3 and T4 theoretical claims include an explanation of w
38、hy the claim is correct. T3 claims do so by invoking a single factor as cause. T4 theoretical claims are also explanatory, but in the more complex form of invoking a system of interconnected factors as an explanation.Some researchers have reserved theory to refer to theories of the T4 type, claiming
39、 that even very young childrens theoretical knowledge has internal co- herence and is structured around a set of core assertions interconnected by causal principles (Carey, 1986; Wellman & Gelman, 1998), even though the child lacks explicit awareness of the theory. In this analysis, in contrast, al-
40、 though noting the distinctions among them, we treat all four types in Table 1 as lying on a single continuum and therefore warranting a common label. Our justi- fication for doing so is that, although they differ in complexity, each makes a claim that is potentially falsifiable by empirical evidenc
41、e.1 This common charac- teristic, we believe, is sufficient to link them along a single continuum. We refer to all four types by the single term theoretical claim (thereby encompassing both the claim label more common to the T1 and T2 types and the theory label more common to the T3 and T4 types). E
42、ach of the four is falsifiable by the kinds of evidence indicated in the third and fourth columns of Table 1. Again, the re- quirement for treating an observation as evidence is that it stand apart from the theoretical claim and bear on it.Linking the four types of theoretical claims in Table 1 is i
43、mportant to our objec- tive of identifying developmental origins of scientific thinking. Presumably, the coordination of theory and evidence that we define as scientific thinking will be observable first with respect to the more elementary theory types (T1 or T2) and the simplest kinds of evidence a
44、ssociated with them. Only later might a child be- come capable of reasoning about (as opposed to reasoning with) more complex theories and about how evidence bears on them. Complex instances of theoryevi- dence coordination involve multiple interrelated theoretical claims and multiple kinds of evide
45、nce bearing on them. Here we are asking simply how theoryevi- dence coordination could occur in its most rudimentary form.1Although it can be argued that unfalsifiable theories have legitimacy and value, we do not take on that debate here. Also, in our connecting a range of claims from simple assert
46、ions or beliefs to complex theories, we do not mean to deny the importance of distinctions among them, as well as other related con- structs, such as hypothesis, prediction, and explanation. To address these distinctions would take us away from our purpose, which is to identify the common characteri
47、stic the types in Table 1 share as claims that can be examined in the light of evidence.118KUHN AND PEARSALLCoordinating a Theoretical Claim With EvidenceThree requirements must be met if it is to be claimed that one is engaged in the coor- dination of theory and evidence. First, the theoretical cla
48、im must be recognized as potentially false. If it is not falsifiable, empirical observations could be incorpo- rated within it, but they do not stand apart from it and bear on its correctness. The second and third requirements follow from the first: Evidence must be recognized as the means of falsifying a theoretical claim, and, finally, theoretical claim and evi- dence must be recognized as distinct epistemological categoriesevidence must be di
