Terry Winters: Process and Generative Art

Long before my scientific path began, I was drawn to the paintings of Terry Winters. I remember visiting an exhibition of his work at the Whitechapel Gallery in London in 1998 and being bowled over by the complexity and scale of the paintings. The sensation of being absorbed by one of them is reminiscent of spending time with a Jackson Pollock painting, only I’ve always felt that Winters’ paintings have more substance, structure and depth, giving the mind a more concrete space to run around in. It has been noted that “Winters’s interest is in diagramming the new mental spaces created through the merging of technological, architectural, and biological forms” [1].

Terry Winters’ works are not the beginning or end of an idea, but are a journey or process in themselves. Each work in turn takes the artist to new places. He commonly produces large series of works under the same title, often in different media. The ‘Graphic Primitives‘ (1998) collection comprises drawings, paintings, etchings, and woodcuts. Notably, hand-made drawings were scanned into a computer image-editing program for manipulation by Winters. He used the digital tools to layer images, alter the scale of different elements and adjust the composition. He was then able to “pull information out of each line and see the potential in the virtual spaces between the lines” [2]. These digital images were used to guide a laser to incise woodblocks for production of a series of nine woodcuts on Japanese Kochi paper. The blocks were printed using a process pioneered by Pablo Picasso, in which they were first printed in white ink, and the sheet then rinsed in black ink [3]. These prints, made with the aid of a computer, then fed into new, hand-made works.

For Winters, the act of making artworks are the steps in his journey; for scientists, the production of drawings to describe emerging concepts happens alongside reading, writing, and discussion. Nevertheless, like Winters, for scientists drawings are not final thoughts in themselves but form an essential part of the journey. Researchers make drawings to help them to visual and understand new work, and it is the process of making the drawing that aids in their understanding.

Scientists make drawings without any consideration for what they are doing, and consequently do not seek to choose their materials with any importance. Frequently, drawings begin life with a sense of immediacy, being made in whatever media comes to hand: potentially pencil or pen on paper, but the support may also extend to the back of a catalogue, the corner of a newspaper or a napkin [4]. Frequently, blackboards or whiteboards are installed in laboratory and office spaces to capture these early thoughts. These initial drawings are transient by their very nature: the idea may be quickly superseded by another, just as the scrap piece of paper on which the drawing has been made may be lost or discarded, and markings on a white board are easily wiped away. Winters used the computer as a tool in making his collection of works, allowing him to view his drawings in a different way, and to manipulate them in a manner distinct from that which would be possible using more conventional means. Scientists use computers to aid in their drawing processes in two distinct ways: firstly, to make a transient image more permanent such that its dissemination would be permitted through publishing in a scientific paper, or presentation at a scientific conference; or secondly in the visualisation of complex sets of data [4]. In the former example, the computer is unlikely to serve any more of a purpose than to ‘firm up’ the idea; in the later, the computer performs a generative process, permitting the emergence of concepts or ideas not apparent in the raw data itself.

“Generative art focuses on the process by which an artwork is made and this is required to have a degree of autonomy and independence from the artist who defines it” [5]. Whether Winters uses the computer for generative purposes is arguable, and it would require more information as to the exact computer software and methods used in the production of his woodcuts before we could come to this conclusion. A good example of how scientists use generative methods in their drawings is in the visualisation of protein structure. The images below (Figure 1) describe the lipid transfer protein, RdgBβ. The molecular visualisation program, PyMOL, was used to ‘read’ a large set of data describing the location of each of the protein’s atoms in space, and to output the data as a three-dimensional object. Here, the computer uses autonomous methods, producing images in response to a defined set of rules. PyMOL allows the user to view and manipulate a complex molecule in space, render it in a variety of ways, depending on the user’s needs, and to output two-dimensional image files (three-dimensional movie files also possible). This process would not be so easily carried out using the human mind alone. The image in Figure 1D has been modified using Adobe Illustrator with the addition of hand-drawn elements (the red ‘tail’ of the protein, letters denoting two short amino acid sequences, phosphate molecules, and the ‘w’-shaped protein, 14-3-3).

Figure 1: The structure of the lipid transfer protein, RdgB-beta, modelled on the X-ray crystal structure of another family member, PITP-alpha, and visualised using PyMOL software.
(A) The bonds between the atoms in the protein are shown as lines. (B) Molecule rendered to show what the surface of the protein looks like. (C) Protein structure shown as a cartoon with alpha-helices and beta-sheet used to show the different connectivity between the protein residues. (D) Cartoon structure coloured to highlight different parts of the structure: red, alpha-G helix; blue, regulatory loop; green, lipid exchange loop. Phosphatidylinositol ligand has been superimposed into the lipid binding pocket and is coloured black. Further hand-drawn elements have been added using Adobe Illustrator: red, long, disordered C-terminal tail; R, S, P, denote amino amino acids forming a consensus site for 14-3-3 binding; X, stands for any amino acid. The letter ‘P’ in a circle denotes a phosphate molecule, which is covalently bound to each of the serine (S) residues in the protein tail and is required for 14-3-3 binding [6]. Dimeric cup-shaped (‘w’) structure at the bottom of the drawing symbolises the 14-3-3 protein. Figure taken from [4].
The generative process of molecular visualisation enables the researcher to see their molecule of interest in three-dimensional space, and to uncover features of this structure that they might not have been previously aware. Furthermore, knowledge of the structure can be re-interpreted by the scientist into a symbol-like form, easily reproducible in future quick sketches of ideas on the backs of envelopes. The following two images include the simplification of the RdgBβ structure, so that the core domain is represented by a rough lozenge shape (Figure 2), or hexagon containing a phospholipid (Figure 3). Figure 3 is also an example of a drawing made using a computer, where the software serves merely as a means by which to preserve an idea, rather than to uncover any previously unknown concepts or to generate new ideas.

Figure 2: Schematic drawing made whilst thinking about RdgB-beta protein function.
Black biro on notebook page. Figure taken from [4].

Figure 3: Schematic drawing of a possible function for RdgB-beta.
Digital drawing made using Adobe Illustrator. Figure taken from [4].


1. Weinberg, A. D., in introduction to ‘Terry Winters: Paintings, Drawings, Prints 1994-2004’, Addison Gallery of American Art, Phillips Academy, Andover, Massachusetts, 2004, p.14.

2. Teagle, R., ‘Terry Winters and the Aesthetics of Information’, essay in ‘Terry Winters: Paintings, Drawings, Prints 1994-2004’, Addison Gallery of American Art, Phillips Academy, Andover, Massachusetts, 2004, p.161.

3. Wye, D., ‘Artists and Prints: Masterworks from The Museum of Modern Art’, New York: The Museum of Modern Art, 2004, p.250.

4. Garner, K., ‘A Circle of Drawing in Research into Life at the Molecular Level.’ TRACEY Journal: Drawing and Visualisation Research, 2013, in press.

5. McCormack, J., Bown, O., Dorin, A., McCabe, J., Monro, G., and Whitelaw, M., ‘Ten Questions Concerning Generative Computer Art’, Leonardo, 2013, in press.

6. Garner, K., Li, M., Ugwuanya, N., and Cockcroft, S., ‘The phosphatidylinositol transfer protein RdgBβ binds 14-3-3 via its unstructured C-terminus, whereas its lipid-binding domain interacts with the integral membrane protein ATRAP (angiotensin II type I receptor-associated protein).’ Biochemical Journal, 2011, 439, 97–111.

Leave a Reply

Your email address will not be published. Required fields are marked *