I C McManus and Paul Gesiak: Experimenting with Mondrian: Comparing the Method of Production with the Method of Choice. Poster - Design Computing and Cognition 2014, University College London
Ian Christopher McManus,Paul Gesiak
DOI: https://doi.org/10.31234/osf.io/n6cf7
2020-01-02
Abstract:The Dutch painter Piet Mondrian (1872-1944) is the archetypal modernist, with the paintings of his uncompromisingly austere mature period, comprised of a white background, black vertical and horizontal lines and occasional, areas of red, yellow or blue, being icons of modern design. His paintings also represent a rare opportunity to experiment on aesthetic composition, the paintings being simulated relatively easily on a computer screen, with participants manipulating them interactively.The founding father of experimental aesthetics, Gustav Theodor Fechner, described three research methods in his Vorschule der Aesthetik of 1876, the Method of Choice, the Method of Production and the Method of Use. Most empirical work has used the Method of Choice, whereby participants are shown two or more similar images and choose between them. Our first studies of Mondrian used the Method of Choice. A set of 25 original (‘O’) Mondrian paintings were synthesised on a screen, as also were two pseudo-Mondrians, (P1 and P2). The pseudo-Mondrians were created by randomly moving all of the horizontal and vertical lines in the painting by a small amount (for P1) or by a slightly larger amount (for P2). P1 and P2 therefore had the same broad structure as O, containing the same ‘words’ (components), using the same ‘deep structure’ or ‘syntax’, but with the composition varied only by altering the relative positioning of the lines (equivalent in linguistic terms to a different pattern of stress or emphasis – prosodics). For each of the 25 paintings we created O, P1 and P2, and participants compared O with P1, O with P2, and P1 with P2, making 75 paired comparison judgements overall.In five separate studies, totalling 277 participants, it was clear that original Mondrians could reliably be distinguished from pseudo-Mondrians, with some participants being more sensitive than others. Fechner’s Method of Production allows participants to manipulate the content of an aesthetic object, altering it until they feel that it is most satisfactory. We adapted this method so that participants were presented with a single Mondrian painting on a computer screen. By moving the computer mouse a vertical, a horizontal or both a vertical and a horizontal line in the Mondrian were moved up and down or sideways. The images were constrained so that the moving lines were yoked to other lines of the same directionality, so that several moved at the same time, and lines could never cross over each other (i.e. the syntax remained fixed, with only proportional arrangements being altered). The original Mondrian was always a possible outcome of moving the cursor.The Production studies manipulated 39 original Mondrian paintings, none of which were included in the set of 25 used in the Choice experiments. All of the 84 participants could therefore carry out both the Method of Choice and the Method of Production, the order of the two Methods being chosen at Random (and order having no effect upon the results).The most striking result was that although the 84 participants, as expected, had clear preferences in the Method of Choice for Mondrians over pseudo-Mondrians, using the Method of Production there was no evidence that participants produced images which showed any overall similarity to the original Mondrians.A detailed study of what participants were doing when they were using the Method of Production suggested that the complexity of the task, despite its relative simplicity, was too great for them. Despite only manipulating in a two-dimensional space (or sometimes a one-dimensional space), participants seemed to find the task difficult. The program always started with the cursor at one of the four corners indicating the maximum dynamic range of the cursor, and in many cases participants ended up either on one of the edges of the space or even in a corner. In most cases only a small proportion of the possible design space was sampled before a decision was made.Our paper will consider how people can and cannot make complex aesthetic choices across a range of possible stimuli. In particular we will suggest that much of the problem with the Method of Production is a problem of visual working memory, participants not being able to hold in their heads a range of previous images to compare with the current one which is shown on screen. The Method of Choice, in contrast, makes no demands at all upon visual working memory. Although computer design in principle allows an almost infinite space of possibilities to be explored, in practice the ability to do so usefully seems to be heavily constrained by the cognitive limits of human processing, meaning that design systems need careful ergonomic organisation to prevent such problems.