Classifying Mutability

Cloud Physiognomy
by Lorraine Daston

The essay begins…

When it came to clouds, art and science faced similar challenges of description: how to capture almost infinite variety and variability? Both variety and variability flummox description, whether in words or in images, but not in the same way. Perform the following thought experiment: first, imagine all the species of life on earth arrayed together in their dazzling diversity, all circa ten million of them, from the Lesser Antillean iguana to the figeater beetle, from brain corals to black-capped chickadees, from mosquitoes to barley. That is variety, and clouds have it aplenty: cirrus fibratus intortus, cirrocumulus castellanus lacunosus, altostratus undulatus, and on and on. But viewed on a human time scale, it is a static variety. Evolution rarely proceeds before our very eyes. Now imagine all of these ten million-odd species constantly metamorphosing into one another and into intermediate forms—not just evolution speeded up to cinematic tempo but everything changing into everything else, all at once, not just past forms to present forms but also present to past and this present form to that other one, without regard to taxon or phylogeny. That is variability—the vertiginous variability of clouds.

It is not just the variety of clouds but also their fast-paced variability that eludes description: if the pace of biological evolution is too slow to be perceptible on a human timescale, that of cloud evolution is too swift for the human eye to fix, much less to capture in a net of words and images. Although the skies have been scanned and studied since the meticulous astrometeorological diaries kept for more than six centuries by ancient Babylonian scribes, and weather-watching networks sponsored by scientific societies have been trying to systematize observations since the seventeenth century, it was only at the turn of the nineteenth century that two naturalists, Jean-Baptiste Lamarck in France and Luke Howard in Britain (it is significant that both were steeped in the practices of Linnaean classification) independently and simultaneously proposed cloud classification schemes—but two quite different schemes based on different principles.

For the next hundred years, cloud observers elaborated their own systems, splitting and lumping the original categories according to local weather patterns and individual proclivities. Is it any wonder that almost every scientific publication on cloud classification from Lamarck’s and Howard’s pioneer attempts around 1800 to the latest edition of the International Cloud Atlas in 1975/1987 begins with a tetchy paragraph defending the whole enterprise against skeptics who point to the notorious mutability and evanescence of their subject matter?

After centuries of serving as the metaphor for mutability, clouds began to be classified by genera and species in the nineteenth century, on the model of Linnaean taxonomy. In order to standardize nomenclature, cloud watchers had to learn to see in unison, recognizing cloud types as one would recognize human faces. The analogy between cloud and facial recognition runs deep: in both cases, a few salient features (that aquiline nose, those long wispy streaks) are foregrounded at the expense of a great many others. What the art of caricature is to faces, condensed description was to clouds: a few bold strokes that focused attention on the essential and screened out everything else. Cloud classification depended crucially on description by omission.

LORRAINE DASTON is Director at the Max Planck Institute for the History of Science in Berlin and a regular Visiting Professor in the Committee on Social Thought at the University of Chicago. Her work spans a broad range of topics in the early modern and modern history of science, including probability and statistics, wonders and the order of nature, scientific images, objectivity, quantification, observation, and the moral authority of nature.