Several studies carried out in different cultures have pointed to the existence of a broad underlying factor responsible for the clustering of attitudes relating to many areas of social life: moral, political, religious and philosophical (Bagley et af., 1970; Wilson, 1970; Schneider & Minkmar, 1972). This factor has been variously labelled 'fascism', 'authoritarianism', 'rigidity' and 'dogmatiSm', with appropriate shifts in emphasis, but the less value-toned term 'conservatism' has more recently gained currency.Reviewing his own findings concerning correlates of conservatism, Wilson (1975) proposed that certain attitudes such as anti-hedonism, punitiveness, dogmatism, superstition, conventionality, conformity, militarism and resistance to science were organized together in the conservative because of an underlying generalized susceptibility to anxiety in the face of uncertainty. Wilson suggested that individuals with low self-esteem and feelings of insecurity would perceive the world as threatening, changeable and treacherous and would seek to minimize both stimulus and response uncertainty. Thus, conservatives conform to the social order because this reduces response uncertainty and they dislike and seek to avoid stimuli with elements of ambiguity, complexity, change and individuality. Wilson et af. (1973) investigated this proposed relationship between conservatism and aversion to stimulus uncertainty by studying art preferences in conservatives and non-conservatives. They predicted that conservatives would show a preference for simple over complex paintings and for representational over abstract paintings, but that the simple-complex dimension would be the primary discriminator. Their results confirmed this hypothesis.The Wilson et af. experiment is extended in the present study to investigate the relationship between conservatism and stimulus uncertainty using poems in place of paintings as stimuli. It was found necessary to reinterpret the dimensions used to classify the paintings in the original study: the abstract-representational dimension could not readily be applied to poetry and was replaced by the dimension 'modern-traditional' which, as Wilson ef af. point out, is generally related to it. The simplicity+mmplexity dimension, defined in the case of paintings by reference to the 'number and concentration of different elements', could not be applied directly to poems in this way. In the case of poetry, it seems likely that the descriptions 'simple' and 'complex' might be applied to poems on the basis of more than one criterion (see Discussion). Rather than erect a single-probably overrestrictive-criterion of 'complexity', or even a combination of such criteria, it was decided in this exploratory study simply to select poems which appeared intuitively to be clearly 'simple' or 'complex' and to subject this categorization to validation in two ways: firstly, by having the categorization process repeated 'blind' by a specialist in English literature and, secondly, by asking the subjects themselves to rate...