Criterial problems in creative cognition research

Abstract

In creative cognition research, the Romantic view about creative cognition is traditionally rejected in favor of the modern view. The modern view about creative cognition maintains that creativity is neither mysterious nor unintelligible and that it is indeed susceptible to analysis. The paradigmatic objects of analysis in creative cognition research have been creative output and the creative process. The degree of creativity of an output is assessed in accordance with certain criterial definitions. The degree of creativity of a cognitive process is assessed in accordance with certain models of creative cognition, psychometric test measures, and neuroimaging studies that are grounded in certain criteria for assessment. The reliance on criterial definitions and criteria for assessment in analyzing either the creative output or the creative process suggests that creative cognition researchers remain under the sway of the classical, bundles-of-criteria theory of meaning. In this paper, I will critically evaluate the criterial problems that confront both criterial definitions and criteria for assessment before proposing an alternative theory of meaning.

Keywords:

• Bundles-of-criteria theory of meaning
• conceptual pluralism
• creative cognition
• criterial problems
• NSC model of categorization
• prototype theory of meaning

Notes

1. Admittedly, monkeys hitting keys at random for an infinite amount of time might well produce the complete works of William Shakespeare. One might therefore look to add ‘intentional’ (or at least ‘not completely random’) to the laundry list of necessary but insufficient conditions for creativity.

2. There are problems with basing an account of creativity on the necessary and/or sufficient conditions for patentability. Under §101 of the U.S. Patent Act, only processes, machines, articles of manufacture, and compositions of matter are patentable. An idea, an economic theory, a mathematical method, and a computer program cannot therefore count stricto sensu as patentable, no matter how novel, non-obvious, and valuable they might be.

3. Weisberg (2015, p. 111) has himself acknowledged this implication of his non-standard two-criteria definition of creativity. He might therefore have committed himself to a position that would require a fair amount of bullet-biting.

4. These examples of novel yet intelligible sentences have been taken from Boden (2004).

5. The criteria for assessment for the AUT (Guilford, 1967) are: originality, fluency, flexibility, and elaboration. While the criteria for assessment for early versions of the TTCT (Torrance, 1966) closely resembled those of the AUT, later versions of the TTCT (Torrance, 1990) removed flexibility but added resistance to premature closure and abstractness of titles to their criteria for assessment.

6. The criterion problem was first discussed in the late 1950s and 1960s at a series of conferences in Utah (later published in a volume edited by Calvin Taylor and Frank Barron). I am grateful to Mark Runco for having pointed this out to me.

7. For further critical discussion about the relationship between the prototype theory of meaning and face recognition technologies and its implications for cognition in general, see O’Toole (2004).

8. The relevant prototype model that I have in mind is neither the attribute-free prototype model (Posner & Keele, 1968), in which the prototype does not have a clear attribute structure, nor the spatial prototype model (Rips, Shoben & Smith, 1973), in which the prototype is a point in space corresponding to the centroid of an instance cluster. Rather, I am concerned with the featural prototype model, in which similarity to the prototype is determined by counting the number of overlapping features between the prototype and an instance. The featural prototype model is better able to deal with the degrees of abstraction and elaboration that we will expect from creative cognition (when we deal inter alia with creative personality traits, creative mental activities, and creative evaluations). For further discussion, see Hampton (1995).

9. See the similarity measure of Hampton (1979) and the family resemblance score of Rosch and Mervis (1975) for examples of similarity metrices.

10. As our experimental subjects would consist of members of the laity (whom we would presume to lack the domain-specific knowledge of experts), we would hold our prototypes to be standard exemplars (in which statistical information is encoded) rather than best exemplars or category ideals.

11. One could maintain the heterogeneity hypothesis, according to which the class of concepts under which creative cognition researchers look to investigate creativity is made up not of one natural kind but rather of several distinct natural kinds, and reject conceptual eliminativism (as espoused by Machery). One could, for instance, be a conceptual pluralist (Chalmers, 2011; Weiskopf, 2009) or a selective conceptual eliminativist (Taylor & Vickers, 2017). Given the foundational nature of the superordinate concept of ‘creativity’ in human cognition and endeavor, I believe that we are better off maintaining a certain conceptual pluralism about ‘creativity.’

12. For further discussion about prototype theory and scaling from prototypes to category ideals in the context of conceptual pluralism, see Weiskopf (2009).

13. Carruthers (2006) has similarly defended the place for creativity of action in any adequate account of creativity. According to the prototype theory approach, we ought to move away from thinking in terms of necessary and sufficient conditions for creativity (viz., creativity of thought and the transformation of conceptual space) and toward thinking in terms of prototypes of creativity. Prototypes of creative action and prototypes of creative thought might, for example, share a featural overlap that would allow us to characterize them in terms of the superordinate concept of ‘creativity.’

14. In the final resort, and as aforementioned, I am also willing to maintain the heterogeneity hypothesis and conceptual pluralism.