Much of the literature on art-science interactions is celebratory, which I think is never a good sign, because it shows that people who might want to question the limits and sense of the field are not engaged. I've been writing on this subject intermittently since the early 1990s, and even though the field has grown exponentially in the last five years, I think many of the issues that support the uncritical literature remain in place. For me, what matters most are the concepts and assumptions that underwrite talk about sci-art. Here are four abbreviated examples of what I mean. 

1. The words art and science

Art-science work is becoming more popular each year, and now it's common to find art-science centers in art schools and university art departments, as well as artists in residence in scientific institutions. There is a lot to be said in favor of some of the new forms of sci-art. For example, it would be churlish to claim that projects involving global warming, post-human awareness of animals, Native American land claims, or genetically modified crops are somehow unsuccessful unless they engage cutting-edge science or art. The purpose of such initiatives is to inspire and educate. Art and science aren't really what's at issue: what matters is what the work does in, and for, the world. 

Yet in other cases the words art and science are put under considerable pressure. They end up being asked to do work, and make sense, in unusual ways. Claims are made about new epistemologies, and new kinds of aesthetic or anti-aesthetic experience. Administrators advertise new sorts of interdisciplinary or post-disciplinary practice. This is where I become skeptical, because it's a rare art-science project that can demonstrate a genuine intervention in either scientific or artistic practice. 

2. The words aesthetics and beauty

One of the best texts here is Leo Steinberg's stupendous essay, "Art and Science: Do They Need to Be Yoked?," published in 1986. Steinberg observes that when scientists say their work has aesthetic value, or that it exhibits beauty, sublimity, pattern, symmetry, or elegance, they use those words in very different senses than art historians or theorists. He rehearses the ways he's heard scientists talk about elegance and beauty, as if they are simple properties of pleasing and informative images in science or art. He then cites the history of violence done to artworks, and asks if that's the sort of beauty scientists mean. Now, thirty years later, it is just as helpful to inquire into the meanings of these words wherever they are proposed as common ground between art and science.

3. Claims that Western art and science have been linked since the Renaissance

There is a standard art-science narrative that is implied in the literature on sci-art: it's the story of how Western science and art were first linked in the Renaissance with linear perspective, and continued to be connected in the optical and naturalistic discoveries of the Enlightenment and Impressionism. In this story, modern artists who work with science have a long genealogy in Western art. My principal objection is that the narrative picks just those relatively rare moments when some science was used by some artists. It omits so much of Western art (not to mention non-Western art!) that it cannot count as reasonable history. Artists who are prominent in the standard narrative tend to be minor in art historical accounts. If you were to read only introductory chemistry and crystallography texts, it would seem that M.C. Escher is one of the century’s most important innovators, but he is absent from art historical textbooks. The canon of scientifically interesting artists is a strange one, a fact that should be bothersome because the values that make the art worth studying to begin with are at odds with the values that proceed from inquiries that search for scientific content. 

4. The word collaboration

The majority of contemporary art-science collaborations, I think, are not about science but rather technology. What matters is the display of science—its data, apparatus, institutions, laboratories, and politics—but not science itself, with its mathematical languages, its frequent indifference to visual display, and its protocols of hypothesis and falsification. Art-science projects are commonly presented as interdisciplinary, and they fulfill institutional interests in producing new configurations of practice. But the collaborations are not often interdisciplinary in an interesting sense: they do not produce nameable new practices. They are more like colloids, dependent on the artists and scientists involved, and prone to separating back into their original disciplines. When it comes to art projects that aspire to connect fully historically informed, contemporary art with fully unpopularized, contemporary science, there is a criterion that can help distinguish successful from unsuccessful collaborations. It is this: the science should not just be freshly represented, visualized in an intriguing way, presented in unusual media, or made available to a new public. The science should itself be influenced by the art, so that the collaboration results in the production of new science. Otherwise the collaboration is not symmetrical.

The first essay I wrote on this subject was called "The Drunken Conversation of Chaos and Painting.” The "chaos" of the title was fractal geometry, which was a lure for painters such as Mark Tansey. These days the mathematics and the media have changed, but the conversation is still drunken, which is to say it's intermittently incoherent, but it keeps going because the two sides are besotted with each other—or with ideas they have about each other. As in any lover's discourse, what matters is more the desires each side has, rather than the precision of their language.