I've taken the editors' question seriously and literally. What should sociology look like? What form should it take? what it should be about, what topics we should be discussing, what problems addressing. None of that, just what it ought to look like. When I told Barbara Risman that I was going to take this approach, she correctly anticipated the first part of my answer. Not like a standard journal article, right? Right. The highly ritualized, formulaic journal article is an incredible impediment to thought and communication. These articles are written to satisfy an evergrowing list of increasingly complicated and arcane requirements: Mention all the right references, use the most up-to-date methods, link what you've done up to the approved canon of Big Thinkers, and so on. Since articles are almost always read by several readers and since editors have, for the most part and because of the scrutiny of the boards who appoint them, given up exercising independent editorial judgment, authors reasonably and ritualistically seek to satisfy all the complaints of all the readers. This almost endless task of revise and resubmit'' is usually impossible, and the result looks like it. But that's a common complaint, if not in print then certainly in the halls of conventions and departments. Its solution lies in the reform of the organizations in which articles get read and accepted or rejected, not in teaching sociologists more about writing. I won't pursue this further, but merely leave its implementation to the oppressed. I'm more interested in sociologists' avoidance of a multitude of other ways to communicate what we know or tKink we know the ways that are not the standard journal article. I have a number of recommendations, all of which go toward expanding the range of what sociologists do when they present their results, and suggesting other ways that sociology could, therefore, look. To start with, the visual. One of the marks of an advanced science is the use of photographic and cinematic methods of recording and analyzing data. I don't say that to be cute. It's obvious. Astronomers, for instance, work by making photographs of the sky. Nuclear physicists photograph the collisions that take place in particle accelerators and inspect those images for traces of the unusual events that give them clues to the structure of elementary particles. Biologists rely on photographs made through the electron microscope for their data and evidence. The social sciences have lagged behind the natural sciences in the use of visual materials. For a long time, sociologists viewed visual imagery as the mark of a reformist perspective that, they were taught, was not real science. Real scientists were objective and unsentimental, and photographs seemed to make people sympathetic to the (usually) down-and-out folks contained in the images found in the great social surveys of the early twentieth century, and even in the early volumes of the American Journal of Sociology. A brave band of sociologists, organized as the International Visual Sociology Association, have tried to counter this retrograde anti-visual tendency in sociology. Though their members have produced some wonderful work, and continued to put out the International Visual Sociology Review, they have not influenced the practices of most of the ASA membership. What's so good about visual materials for sociologists? For one thing, photographs, more aptly than words, display social phenomena in context. (Of course, they can be manipulated easily to get rid of context, and photojournalists often strive for that, but it's not built in. ) Displaying context visually can have important empirical and theoretical consequences. Here's a homely example: One of the students in a summer workshop on documentary photography I taught, wanting to combine her work for the course with tending her small children, decided to photograph the of at the beach. She complained that adults' legs kept getting in the way of her attempts to isolate the children's interactions. We finally decided that this apparently unavoidable contextual feature was not a problem. It was a finding: that small children do not, in our society and in that social class, have an independent society. They are always under the surveillance of adults, whose legs therefore always appear in photographs of the of children.