Familiarity and Novelty in Aesthetic Preference: The Effects of the Properties of the Artwork and the Beholder

Abstract

Familiarity and novelty are fundamental yet competing factors influencing aesthetic preference. However, whether people prefer familiar paintings or novel paintings has not been clear. Using both behavioral and eye-tracking measures, the present study aimed to investigate whether the effect of familiarity-novelty on aesthetic preference is independent or dependent on artwork properties (painting content, visual complexity) and viewer characteristics (experience in art). Participants were presented with two images of paintings, one of which was repeatedly presented but was always paired with a new painting in a randomized lateral arrangement. They were asked to indicate which of the two images they preferred with the degree of their preference. Behavioral results demonstrated an interactive influence of painting content and complexity on familiarity-novelty preference, especially alongside the distinction between representational and abstract paintings. Also, the familiarity-novelty preference was modulated by the degree of art experience, for abstract paintings in particular. Gaze results showed the differential effects of painting content, complexity, and art experience echoing the behavioral results. Taken together, the convergent results derived from behavioral and eye-tracking measures imply that novelty is an important feature of aesthetic appreciation, but its influence is modulated by properties of both the artwork and the beholder.

Keywords: art experience; content; familiarity; novelty; paintings; preference; visual complexity.

Figure 1

Stimuli and trial sequence. (A) Examples of images that were used in the experiments. All the paintings that are shown in this figure are copyright-free images that were sourced from WikiArt. The values under the images are the mean raw scores of preference (left) and complexity (right) obtained from the preliminary survey. (B) An illustration of a sample experimental trial. After a fixation period of 1 s, two images of paintings were presented side by side, below which a response bar was presented. Participants were instructed to indicate which of the two paintings they preferred with their degree of preference on a 7-point scale that ranged from −3 to 3. The initial position of the mouse cursor was always on zero to avoid potential bias. There was no time limit for responses.

Figure 2

The time course and summary of familiarity-novelty preference. (A) The time course of familiarity-novelty preference. Time courses of preference values through a subblock were averaged for each complexity condition. The mean relative preference values are shown for each trial from the first to the fifteenth trial. The negative values of the y-axis indicate a preference for repeated (i.e., familiar) paintings, and the positive values of the y-axis indicate a preference for novel paintings. Error bars represent +1 standard error of the mean. (B) Summary of familiarity-novelty preference. Violin plots illustrate the distributions and the descriptive statistics of the individual mean relative preference values from the fourth to the fifteenth trial for each of the six conditions. The cross hairs indicate the mean. The middle horizontal lines indicate the median, and the lower and upper lines indicate the first and third quartiles. Asterisks indicate a significant difference from zero, as per the results of one-sample t-tests (*p < 0.05, n.s., not significant).

Figure 3

Familiarity-novelty preference and art experience in simple abstract paintings. (A) Correlation between the mean relative preference values and the total art experience scores. (B) Correlations between the mean relative preference values and the two subscale scores of art experience: education (left) and activity (right). *p < 0.05.

Figure 4

Temporal changes in familiarity-novelty preference. All trials between the fourth and fifteenth trials were divided into three phases: early (4–7), middle (8–11), and late (12–15) phases for portraits (A), landscapes (B), and abstract paintings (C). All trials between the fourth and fifteenth trials were also divided into four phases: phase 1 (4–6), phase 2 (7–9), phase 3 (10–12), and phase 4 (13–15) phases for portraits (D), landscapes (E), and abstract paintings (F). The mean relative preference values for each phase are shown. The negative values of the y-axis indicate a preference for repeated (i.e., familiar) paintings, and the positive values of the y-axis indicate a preference for novel paintings. Error bars represent +1 standard error of the mean. Asterisks indicate a significant difference from zero, as per the results of one-sample t-tests (*p < 0.05 and **p < 0.01).

Figure 5

Fixation heat map and the ratio of fixation count. Fixation heat map for simple (A) and complex abstract painting (B) trials. A 2-dimensional Gaussian filter was applied to each of the fixations, and the height of the Gaussian was weighted by the duration of the individual fixations. (C) The gaze index (i.e., the ratio of fixation count for novel paintings to that for familiar paintings) was calculated. Violin plots illustrate the distributions and the descriptive statistics of the individual gaze index for each of the six conditions. The cross hairs indicate the mean. The middle horizontal lines indicate the median, and the lower and upper lines indicate the first and third quartiles. The larger a gaze index is than 1, the more frequently a participant had fixated on novel rather than repeated, familiar ones.

Figure 6

The ratio of fixation count and art experience in simple and complex abstract paintings. (A) Correlations between the gaze indices and art experience scores for the simple (left) and complex abstract paintings (right). (B) Correlations between the gaze index and art education for simple (left) and complex abstract paintings (right). *p < 0.05, **p < 0.01.