Empirical Studies of the Arts

Unlike most forms of artistic expression, the feature film is the collaborative product of many individuals. The comparative impact of these separate contributions was assessed in 2,323 movies nominated for Academy Awards in the major categories. Two criteria of a film's impact were defined (best picture honors and movie guide ratings) along with 16 potential predictor variables (direction, male and female leads, male and female supporting roles, screenplay, art direction, costume design, makeup, cinematography, film editing, score, song, visual effects, sound effects editing, and sound) and five control variables (release date and the genre of drama, comedy, romance, and musical). Multiple regression analyses indicated that between 30 percent and 75 percent of the variance in impact could be explained using a subset of these factors.

The previous experimental proofs of Martindale's preference-for-prototypes model (Martindale, 1984, 1988; Martindale, Moore, & West, 1988) did not use artworks as stimuli except the experiment of Hekkert and van Wieringen (1990a, 1990b, 1992) who applied reproductions of cubist paintings. In our experiment, the category of surrealism was taught to naive subjects by repeated presentations of a large assortment of surrealist paintings. In the main experiment, two experimental groups participated. The members of the first group learned the category of surrealism in four series, scaling 30 paintings selected out of 40 so that the 10 most prototypical paintings were left. These 10 paintings were entered in the fifth series, and they were more preferred than the other 30. A similar procedure was followed in the case of the other experimental group, but here the least prototypical elements were left and entered at the end of the experiment. These 10 paintings were less preferred than the other 30 in the first series. This finding proved that the phenomenon we found is a prototypicality-effect and not a novelty-effect.

The use of sound was studied in several of Tennyson's shorter and better-known poems (e.g., Break, Break, Break and Ulysses) and in In Memoriam A. H. H. Poems were broken down into their component phonemes which were then classified in terms of their emotional character. The emotional character of sounds preferentially employed in each of the shorter poems matched the emotional theme of the poem (e.g., sounds employed in Airy Fairy Lilian and Lady of Shallot were most pleasant, those in Crossing the Bar were least pleasant). The emotional character of sounds preferentially employed in the final segments of In Memoriam revealed an underlying sadness to the poem's close belied by Tennyson's own interpretation of it. Analyses of In Memoriam also highlighted the elegy's frequent transitions from grief to hope and back again. It is concluded that Tennyson used sound (both consciously and unconsciously) to amplify the effects of his poetry.

The sound of the language in a literary text is often thought to contribute to its meaning. We hypothesize that this is due not to fixed or universal phoneme properties, as theories of phonetic symbolism have supposed, but to the use of local phonetic contrasts to elicit meaning. Writers may set an overall range of phonetic tones that are distinctive to a particular text and then introduce significant variations to achieve local effects. In the present study, an analysis of phoneme distributions developed by Miall (2001) and an approach to phonetic symbolism developed by Whissell (1999, 2000a, 2000b) were applied to a Katherine Mansfield short story. Readers' responses to the story were obtained using Semantic Differential ratings. The findings show the influence of phonetic patterns consistent with the hypothesis that phonemic contrasts elicit local changes in feeling tone. The effects of phonetic symbolism, while evident, were much less pronounced.

Gray's (1991) neurobiological systems–the behavioral approach system (BAS) and behavioral inhibition system (BIS)–have been operationalized by Carver and White (1994) in their BIS/BAS Questionnaire. In the present study, 129 undergraduate students completed the BIS/BAS Questionnaire, the Eysenck Personality Questionnaire–Revised (EPQ-R), and Zuckerman's Impulsive Sensation Seeking Scale. They also rated 72 paintings previously divided into two stylistic (Abstract and Traditional) and five content (Erotic, Pleasant, Neutral, Violent, Unpleasant) categories. As hypothesized, scores on the BIS were negatively associated with liking for Unpleasant and Violent paintings; scores on the BAS predicted liking for Pleasant and Neutral paintings, though this was due almost entirely to the BAS Reward Responsiveness sub-scale. Sensation Seeking, EPQ-R Psychoticism, and the BAS Fun-Seeking sub-scale predicted liking for Abstract stimuli, and disliking for Unpleasant and Violent paintings. Our results provide evidence that Gray's constructs can be usefully applied to the area of aesthetic preference.