Karl Benjamin

Vertical Stripes #5, Karl Benjamin, 1960

I’m most interested in Benjamin’s stripe paintings. It is interesting that, as noted by Claudline Humblet, that colour values in his paintings are non-systematic (diverging from other stripe painters) such that each stripe has its own autonomy. I was thinking about this in the context of the autonomy of each of the stripes in the current sketch of the Zombie Formalist. The ZF stripes are totally independent and have no relation to one an other; in some cases one stripe may cover up another entirely.

Benjamin’s process provides an interesting precedence for the ZF as only thinking about each image component in isolation and not being concerned with overall composition or colour constraint. The value of the overall composition is determined solely by the viewer (as modelled by the machine), so why not provide as much variety as possible?

This also may influence the generative approach where I’ve been torn about using an evolutionary method. With an evo approach, there is some stability in the structure of successive generations. Another option is to use a critic (that models the viewer’s sense of value) that is allied to constantly regenerated paintings. This way there is no successive development over time; each painting is unique, fleeting and does not persist through successive generations. Interestingly, Claudine Humblet writes that Benjamin has used chance operations in his colour selection, which fits very well a non-evo approach where each composition has maximal autonomy. This is also consistent with Benjamin’s process orientation that “accepts the sometimes unexpected results that arise from the process” (Claudine Humblet, The New American Abstraction (1950–1970)).

At this point I have three directions of aesthetic exploration for the ZF: (a) Increasing the number of stripes in the current system significantly (I counted as many as 25 stripes in one of Benjamin’s compositions); (b) Use the sine-wave approach to increase the range of possible number of stripes; (c) a sine-wave approach where the parameters are constrained to force a grid-like relation between the different sine-waves. The latter would cut against the autonomy of the sine-waves, but it’s worthwhile exploration. Maybe I’ll completely change my mind once I get to Gene Davis!