Dimensional schizotypy, autism, and unusual word associations in artists and scientists

Abstract

Studies from a range of perspectives provide evidence for a relationship between creativity and the tendency to mental illness. The present study further examined this issue by administering questionnaires measuring the minor features of psychosis and autism to 31 professional “artists” (visual artists and musicians) and 28 professional “scientists” (biological scientists and physical scientists/mathematicians). The Oxford–Liverpool Inventory of Feelings and Experiences (O–LIFE), the Hypomanic Personality Scale, and the Autism-Spectrum Quotient (AQ), were administered, in addition to a shortened form of the Kent–Rosanoff Word Association Scale. The results provided strong support for the connection of artistic creativity to positive schizotypy and hypomania and the tendency to make unusual word associations, and somewhat weaker support for the connection of scientific creativity to certain components of the autism spectrum.

Introduction

Recent scientific research provides substantial support for the centuries-old view that creativity is linked to psychopathology, with research by personality psychologists frequently employing personality scales designed to assess predisposition to psychopathology within the general population. Numerous early studies of this kind used the Psychoticism (P)-scale from the Eysenck Personality Questionnaire, with individuals in a range of creative pursuits showing elevated scores on the scale (Merten & Fischer, 1999). The P-scale also tended to correlate with performance on tasks purported to measure associative thinking, such as tests of divergent thinking (Woody & Claridge, 1977) and word association (Eysenck, 1993), supporting Eysenck’s (1995) explanation of the relationship as due to the loose, “overinclusive” associative processes characteristic of both schizophrenic thinking disorder and certain aspects of creative thought.

While much of Eysenck’s view remains consistent with contemporary viewpoints, the P-scale itself has been criticized from a number of perspectives, particularly its tendency to measure impulsive and aggressive behaviours more adequately than the minor features of psychosis. An alternative approach has been to employ scales of “psychosis proneness” or “schizotypy” (Fisher et al., 2004) developed to measure the minor features of psychosis. When factor analysed, these scales have been found to produce three or four relatively independent components (Bentall, Claridge, & Slade, 1989). In particular, the distinction in the schizophrenia literature between positive and negative symptoms was paralleled in the schizotypy literature by the distinction between positive aspects of schizotypy, including magical thinking and unusual perceptual experiences, and negative aspects, including introvertive withdrawal and anhedonia. When correlated with creativity, the results obtained largely reflected the aspect of schizotypy being measured by a scale, showing positive correlations with positive schizotypal features and hypomania and, less frequently, negative correlations with negative features (Green and Williams, 1999, Schuldberg, 2000).

More recent studies have also considered the possibility that there may be differences, in degree or in kind, between the general domains of artistic and scientific creativity. In his review of the empirical literature, Feist (1999) noted both the considerable overlap in the personalities of creators in the two broad domains, and the tendency for artists to be distinguished from scientists by such personality characteristics as “emotional lability, impulsivity, nonconformity, and rebelliousness” (p. 284). Simonton (2004), suggested a continuum from normal, paradigmatic science through revolutionary and non-paradigmatic science to formal and, finally, expressive artistic creativity. The continuum was believed to correspond to increasingly unconstrained cognitive processes, increasing openness to experience, and greater incidence of psychopathology.

In a recent study, Nettle (2006) tested the view that both impulsive, nonconforming behaviour patterns, and the features of positive schizotypy, are more likely to be associated with artistic than with scientific creativity. Groups of poets, visual artists, and mathematicians were given a multi-dimensional schizotypy questionnaire, the Oxford–Liverpool Inventory of Feelings and Experiences (O–LIFE; Mason, Claridge, & Jackson, 1995). This questionnaire includes scales measuring the positive features of schizotypy (called Unusual Experiences), and the negative features of schizotypy (Introvertive Anhedonia), as well as a scale measuring attentional difficulties and social anxiety (Cognitive Disorganization). A fourth scale (Impulsive Nonconformity) measures anti-social and impulsive behaviour, and is strongly correlated with Eysenck’s P-scale. Nettle’s main finding was that, after controlling for age and sex, poets and visual artists were higher on Unusual Experiences and Impulsive Nonconformity than non-poets and non-artists, respectively, whereas mathematicians were significantly lower than non-mathematicians on both the above scales. A first aim of the present study was to confirm these differences between artists and scientists using a different sample.

In addition, we noted Nettle’s (2006) argument that the difference between artists and mathematicians on the Unusual Experiences task reflects differences in their “range of associations”. We explored this idea by including in our study a measure of word association. It was hypothesized that more uncommon, and less common, word associations would be made by individuals who are high on positive schizotypy than by individuals who are low, and by artists than by scientists.

Finally, we examined Nettle’s conjecture that his data are consonant with the ideas of writers such as Baron-Cohen that Asperger syndrome/high functioning autism is connected to the “systematizing” abilities found in mathematics and engineering. Specifically, Nettle notes that the “constellation of autism, systematizing and science” is in many respects the opposite to the “constellation of arts, Unusual Experiences and affective and psychotic disorders…” (p. 887). We employed a measure of Asperger syndrome/high functioning autism developed by Baron-Cohen, Wheelwright, Skinner, Martin, and Clubley (2001) to be suitable for non-clinical samples, the Autism-Spectrum Quotient (AQ). We anticipated that scores on the AQ would be negatively correlated with our measure of positive schizotypy, and that scientists, and particularly physical scientists/mathematicians, would show higher scores on the AQ than artists.