This page has been archived by the Computer Arts Society.

Non-linear elements and the concept of risk


The process of making Computer Art is bound up with the artist’s reaction to changing elements represented on screen and altered by the artist’s actions with the input hardware. The changes invoked by manoeuvring the mouse around the 2D canvas are displayed almost instantaneously, and stored as discrete stages in the work’s development. These can be undone by referring back to a menu that keeps a log of such changes.

The principal limitation here is with storage space, and most programs can only retain a certain number of steps; but within this area, the artist can revert or even combine non-consecutive steps in a non-linear fashion. Pictures can be saved under different names in different states and reworked or even recombined at a later date. Such non-linearity belies the quasi-physical appearance of the computer image.

Although these manipulations have a seemingly physical effect, they can be dismissed by reverting back to the previously stored image. They are thus insubstantial because of their existence as data: their apparent substance consists of their visual solidity and their effects on other elements within the artwork. Although the computer image rests upon a supporting substructure of data, its visual images have apparent or potential physicality; yet they also exhibit movement and transformative properties at odds with our comprehension of physical materials. This may be what makes them so compelling. To Edward Zajec, the computer’s importance is the way it brings together the “organization, structure, and dynamics of a message”:

At the same time, leaving it open to different interpretations and modifications, or better: only with a computer can we untie the constructive aspects of an idea from its material features and observe and articulate them in time through direct interaction.[1]

Thus the visual form can be separated from its physical context, making the computational space quite a different “material” from its precursors. It is precisely the discrete, binary structure of all computer data that not only enables its great flexibility and apparent malleability, but also separates it in one crucial respect from material forms of visual art.

By offering multiple undo options, computer graphics packages can reverse the sequence of pictures and allow for non-linear editing. The discrete existence of each specified object (and, to a lesser degree, each layer or bounded object in a 2D scene) means that it can be lifted out and moved around without modifying the composition of the surrounding objects, and it can also be sent forward or back along the linear sequence of an animation. Thus the computer grants a certain degree of freedom in temporal as well as spatial arrangements.

Gibson considers the underlying reality of the time dimension to be the “sequential order of events” and that the dimension of space is underpinned by “the adjacent order of objects”. Yet the two are not directly comparable, or even analogous; objects can be shifted in space, but events cannot be changed in sequence.[2]

This is where the bounded space of the computer paradoxically offers greater freedom than the wider physical environment. The virtual space removes the linearity of time and the adjacency of objects, at least to the extent that they become independent of each other. Even attributes such as shape and colour can be treated separately.

The “undo” feature makes a great change to an artist’s view of their art, simply because any mistake (or bad design decision) can be reversed. Gentner and Nielsen term the undo function “forgiveness”, with attendant ideas of atonement for faults.[3] As Sue Gollifer and Colin Beardon point out, this is more than just a way of cancelling errors: it makes the design non-linear at a very fundamental level:

[…] the availability of the “undo” function [eliminates] the effect of any operation and [returns] the designed object to its previous state. The “undo” button effectively enables the user to turn time back one unit [or more...] This enables ideas to be tried out in practice without heavy costs. The ready creation and disposal of hypothetical designs is something very new for this type of work [...][4]

This association of experiment with freedom from wasted material makes the “undo” conceptually powerful, and is arguably an artistic adaptation of a design feature intended simply for revoking mistakes. A similar freedom may also be achieved by constantly saving files at different stages in the work’s progress, then returning to them and working in other directions. On the other hand, the removal of risk can lessen the beneficial effects of risk-taking and perhaps the overall virtuosity of the resulting piece: As Gollifer and Beardon discovered:

[…] an important distinction was made by David Pye between “the workmanship of risk” in which “the quality of the result is continually at risk during the process of making” and “the workmanship of certainty” in which “the quality of the result is exactly predetermined before a single saleable thing is made.” [5]

Risk is an inherent part of the virtuosic execution of a piece of art, like Michelangelo sculpting from a single piece of stone. Remove it and the urgency of judgment it requires can fade away into complacency. But there are other opportunities for risk on a computer, mainly related to preservation of data: crashes and disk corruption which can ruin a work-in-progress or erase an entire portfolio. Binkley’s “wily computer-geist” often intervenes at the crucial moment!

[1] Edward Zajec pp52-3, Artist and Computer [1976], Ruth Leavitt

[2] Gibson, J.J. The Ecological Approach to Visual Perception (Boston, 1979)

[3] “The Anti-Mac Interface” Don Gentner and Jakob Nielsen,

[4] “Designers as Users”, Colin Beardon, Sue Gollifer, Christopher Rose & Suzette Worden

Paper presented at Computers in Context: Joining Forces in Design, Third Decennial Conference, Århus, Denmark, 14-18 August 1995.

[5] Beardon and Gollifer, ibid.