|
Article Excerpt
In the last three decades, there has been a steady flow of empirical papers investigating the role of mental imagery in the creative process. Although a diverse range of inventive protocols has been developed, two standard approaches dominate the research literature. The first of these, the individual differences approach, focuses upon a hypothesized association between self-reported mental imagery and divergent thinking measures (e.g. Forisha, 1978a, 1981,1983; Gonzalez, Campos, & Perez, 1997; Khatena, 1975, 1978). The second, the image generation approach, employs protocols that investigate the emergence of creativity through the visualization of specific forms (Anderson & Heistrup, 1993; Finke, 1990; Finke & Slayton, 1988; Finke, Ward, & Smith, 1992). Of the two methods used to investigate mental imagery and creativity, the individual differences approach has the longest history and relies upon traditional statistical procedures. The aim of the present study is to carry out a systematic empirical investiga tion into the individual differences approach; that is, to assess the hypothesized relationship between self-reported mental imagery and divergent thinking performance across a set of methodologically satisfactory studies. However, prior to this, it is necessary to understand the theoretical and observational rationale for believing that there should be an association between the two variables under investigation.
Anecdotal reports of the role of mental imagery in creativity
A review of the individual differences approach to mental imagery and the creative process suggests that the main stimulus for empirical research of this kind is the anecdotal evidence provided by historically creative individuals (Forisha, 1981; Parrott & Strongman, 1985; Shaw & DeMers, 1986). Reports of 'the association between mental imagery and creativity in the development of history-making ideas abound (Daniels-McGhee & Davis, 1994; Finke, 1990; Mavromatis, 1987; Paivio, 1983; Rothenberg, 1979, 1995; Shepard, 1978a). The best known are presented in Table 1 and suggest that a diverse sample of creative individuals reported using mental imagery in the creative process. This has led to the development of explanatory accounts of the facilitating role of mental imagery, some of which are empirically supported (e.g. Sobel & Rothenberg, 1980). Rothenberg (1979, 1995) claims that creativity is enhanced through homospatial and 'Janusian' imagery, the fusion of pictorial stimuli. He supports this through research into the superimposition of visual stimuli (see Rothenberg & Sobel, 1980). Shepard (1978b) emphasizes the importance of the voluntary manipulation of mental images (e.g. the mental rotation of stimuli). He provides non-empirical support through reports of the use of mental transformations by creative people (Shepard, 1978a,b). Suler (1980) believes that creativity follows from the emergence of an unconscious flow of imaginings (primary process thinking); he also provides anecdotal support for this account.
These explanations suggest that a particular procedural style (or context) of imaging can account for the role of mental imagery in the creative process. Collectively, they suggest that the nature of mental imagery in creativity is multifarious, with styles, forms, contents, and contexts of mental imagery described by creative individuals lacking a common denominator (aside from the reference to mental imagery). The imagery reported by creative individuals may be deliberate (e.g. Einstein, reported in Hadamard, 1945) or prophetic (Coleridge, 1816/1952); synaesthesic (e.g. Debussy and Wagner; see McKellar, 1997) or unimodal (e.g. Loewi; see Ochse, 1990); 'dream-associated' (e.g. Helmholtz; see Shepard, 1978b) or 'wake associated' (Moore, 1952). Mental images, and their reported associations to creativity, are found in such a variety of forms that it would not be reasonable to assume a defining feature of mental images in creativity.
Empirical studies of mental imagery and creativity
Even though reports of the association of mental imagery with creativity vary in style, form, content, and context, the empirical research has adopted two fairly stable protocols. The most frequently employed of these, the individual difference approach, pays scant attention to the explanatory accounts of the use of mental imagery in creativity. Rather, it seems to be grounded in the observations of creative individuals and simple threshold notions. For example, Shaw and DeMers (1986) emphasize a prerequisite intelligence quotient. Campos and perez (1989) note that the importance of being able to control imagery and Forisha (1983) considers cognitive style to be essential to the imagery-creativity relationship. Thus, based on the assumption that mental imagery and creativity are continuous variables, the basic tenet is that differing degrees of imagery ability (and possibly additional prerequisite attributes or contexts) reflect differing creativity abilities.
A review of papers employing the individual differences approach reveals an association principle. That is, the protocol searches for degrees of association between mental imagery and creativity variables. The measures used in the approach are roughly summarized as:
(1) a self-report measure/s of an aspect of mental imagery (normally control or vividness);
(2) a test/s of divergent thinking which provides individual measures of fluency, flexibility, and originality;
(3) an intervening variable that is hypothesized to clarify the association between measures 1 and 2 (e.g. intelligence, verbal ability, cognitive style, instruction to image).
Once the materials have been selected, the tasks are typically presented in a test-room format. Analyses of the measures are based firstly on the assumption that individual differences in self-reported mental imagery are correlated with the sub-measures of divergent thinking and secondly that the association will be stronger when the intervening variable is taken into account. Generally, multiple bivariate correlation is used to assess the hypothesized association, although some authors dichotomize or trichotomize the self-report measures of mental imagery and perform multiple t- and F-tests (e.g. Campos and Perez, 1989). Generally no attempts are made to address error-wise probability rates.
To summarize, the use of mental imagery has been linked to the genesis of creative ideas and products across a range of disciplines. Such reports have provided the impetus for empirical research into an association between the two variables, and two methodological protocols have developed to test the assumption that mental imagery abilities predict performance on creativity tasks. The aim of the present study is to assess the individual differences protocol through a meta-analytic approach.
Method
Literature searches and inclusion criteria
The initial selection was based upon computer databases (PsychLit 1974-1999 and BIDS 1981-1999) and a manually searched target journal (Journal of Mental Imagery). Together, these yielded over 160 articles, of which 58 were deemed to be related to mental imagery and creativity. Of these, 30 were reviews or theoretical articles and 28 were empirical articles. Of the 28 empirical papers, 18 employed the individual differences protocol, and 9 met the selection criteria listed below:
(1) Order of presentation. Only those studies that presented the self-report imagery measures first were used. This requirement was based upon observations of demand characteristics in imagery self-report studies presenting self-report measures after performance tests (McKelvie, 1995; Marks, 1972, 1999).
(2) Response measures of divergent thinking. Given that all of the studies used a self-report measure of mental imagery, only studies which employed performance-based creativity...
|
