The Case of the Missing Faces

As I’ve often mentioned,  I’ve been working for quite some time on a study of the photographs of Walter Freeman. Freeman, a Washington, D.C., based physician, was the world’s foremost lobotomist; it’s estimated that he lobotomized some 3,500 people.

He was also a prolific and dedicated photographer. He almost invariably took photos of his patients before and after the procedure, acquiring reams of these images over the course of his career. In a chapter of my book, Depth Perception, I argue that Freeman was participating in a much longer-standing tradition of psychiatric photography, one that claimed that the human face could reveal the depths of the soul. (You can see a recorded version of the story of Freeman’s photographs here.)

I felt fairly confident that I understood why Freeman took the photos that he did. But if I was right, I thought, then the portrait should rise in popularity as the favored mode of evidence for psychiatrists during the early-to-middle part of the 20th century, because Freeman’s psychobiological mode of thinking was most prominent from about 1930 to the mid-1950s.

It occurred to me that the American Journal of Psychiatry would be a good place to test this theory, since it was the preeminent journal of the psychiatric profession for most of the twentieth century. So I obtained PDFs of every journal article published in the AJP between 1910 and 1970. Since the PDFs don’t distinguish between texts and images, I needed to find a way to extract only the images from the journal articles. The answer was a Python script written by Chris Adams that uses the OpenCV computer-vision library to automatically extract figures from pages.

The result was about 3,000 images, which I then hand-categorized. (After trying many different techniques for this, the best method I hit upon was to tag them using Picasa and then use EXIFextractor to extract the images’ EXIF metadata, including the tags, into a CSV.)

The results were … not what I was expecting.

(Data is here.)

The first thing I noticed about this graph is that charts and graphs have always been the dominant mode of evidence in the AJP. I hope to unpack this in a later post.

The second thing I noticed is that images in general reached a peak around 1930-1940. Why?

I think the profession was calibrating itself. The burgeoning discipline of neuroscience, propelled by the work of Harvey Cushing, meant that those who studied the human psyche had to accommodate new ways of thinking about the brain, and new modes of evidence. Cushing’s own favored mode of evidence was to gather many different kinds of images together — faces, graphs, charts, histological slides — to demonstrate the relationship of brain tissue to observable behaviors.

The same data, with the “charts” category removed, so it’s easier to distinguish among the categories. Data is here.

The second thing I noticed about these graphs is that not only was there no particular rise and decline in the presence of patient photographs, there just weren’t many at all. I really thought there would be more, given the huge collections of patient photographs not just I but other historians have started to turn up in physicians’ collections.

So … what were doctors doing with all these patient photographs?

I thought back to my time in various archives. Cushing’s patient photographs, I remembered, were mounted on lantern slides. (In fact, that’s why they’re so well-preserved.) Charcot’s patient presentations are legendary, and included not just live patients but lantern slides as well. I’d seen this in Freeman’s archive, too, although I hadn’t been quick-witted enough to catch it. The notes I was thumbing through were lecture notes.

He was presenting patient photographs onstage to audiences of other doctors; they all were.

Of course. I’d often seen the evidence in programs from medical meetings. Doctors routinely gave talks accompanied by slide decks, just as they do today. This must have been where the patient photographs came into play, not in the pages of medical journals.

Which leads to another question: What’s the significance of these two different rhetorical modes, one for written work and the other for live performance?

In fact, as I’ve shown elsewhere, physicians routinely participated in elaborate methods of display — but they also policed these displays very carefully, particularly during the early- to mid-part of the 20th century, to keep out outside observers.

The most extreme case of this was the American Medical Association’s scientific exhibit, a little-known but quite important part of medical meetings for most of the 20th century. Physicians set up booths — not just posters, but literal booths — filled with models, anatomical specimens, and many different kinds of media, to demonstrate procedures and argue for particular points of view. Freeman himself loved to declaim from his booth like a carnival barker, waving a clicker to attract people to his exhibit.

But physicians didn’t let laypeople into this spectacle, and I think it’s because the scientific exhibit bore an uncomfortable resemblance to the fairground and other scenes of vociferous salesmanship.

The brand of medicine that circulated in journal articles is studiously dispassionate, illustrated with graphs and charts (like mine!) that seduce with the promise of hard data. But the medicine that physicians performed for each other in seminar rooms and exhibit halls was much more performative, harkening back to the patient presentation tradition, which has always been as much about spectacle and showmanship as it is about the bare facts of a case.

Patient photographs, I think, were part of this spectacle. The human face is irresistibly compelling but difficult to interpret, its meaning inchoate. Far better, then, to stick to graphs and charts when the medium of communication could circulate, possibly ending up in lay hands. But inside the lecture hall, the story was different; doctors held forth to each other with appeals that could look as much like sales pitches as case histories.

I need now to go back and edit my chapter on Freeman to reflect this new observation. But I’m happy to do it, since even though I couldn’t see it at first, this conclusion reinforces something that I think is important to recognize, not just about Freeman but about science in general: that for all of science’s rigor, and its important advances, these disciplines rely at their heart on humans’ ability simply to persuade each other, and in sometimes improbable ways.