The Early Career of Charles F. Stevens (1934-2022)
I am William H. Calvin, the first graduate student of the late Charles F. Stevens (1934-2022), a neuroscientist late of the Salk Institute. This is my 15 minute talk on 5 May 2023 at the memorial conference.
Chuck arrived in the summer of 1963, a year before finishing his Ph.D. in 1964 at the Rockefeller. He soon recruited me to work with him on the stochastic processes underlying synaptic noise and their effects on the interspike interval variability of cat spinal motoneuron’s rhythmic firing.
Chuck must have started at Harvard in 1952. Chuck’s Harvard major was Experimental Psychology.... Then from 1956 to 1960, Chuck was off to Yale for his MD. He spent the summer of 1959 recording from squid axon at MBL. He was looking at the variability of the spike threshold, no less....By 1963, his persona was already that of a creative, enthusiastic yet carefully spoken, basic scientist. He seemed equally at ease with physics, math, and biology.
Chuck worked with Keffer Hartline at the Rockefeller on self-inhibition in the Limulus eye. His thesis title was ‘A quantitative theory of neural interactions: theoretical and experimental investigations.’ It laid out a mathematical theory of computation in the nervous system.... Chuck’s 1966 book, Neurophysiology, A Primer, seemed straight out of the years at the Rockefeller....
It took me until the 1980s to master the mindshift needed to think probabilistically about Darwin’s evolutionary process, about allopatric and sympatric speciation as population sizes boomed and crashed, about genes going to fixation. My invited piece for Nature, “The brain as a Darwin Machine” in 1987, started up a dozen years of thinking about the layer-two neocortical circuitry which offered the possibility of making copies of spatiotemporal firing patterns, thanks to exciting triangular mosaics about a half millimeter wide. It could, I thought, serve to define a Hebbian cell assembly that cloned itself.
While I began with the connectionist darwinism that Gerry and Chuck were imagining, I soon switched from an anatomical to a physiological darwinism. I began using my evolutionary biology to ask what it would take for a cortical circuit to be suitable for running a true Darwinian process—say, on the time scale of shaping up a novel sentence to speak.
The immune response had already provided a darwinian example of how to improve quality within weeks with differential reproduction of the more successful antibodies. How about a Darwin Machine in a patch of cortex capable of producing high-quality physiological function—and on the timescale of thought and action?
That turned into my 1996 monograph, The Cerebral Code, from MIT Press. I tried to make Darwin’s process more abstract, stripping it of biological specifics such as DNA and recombination and haplotypes. I identified six essentials for a Darwinian process:
"a pattern that copies with occasional variation, where populations of the variants compete for a limited workspace, biased by a multifaceted environment, and [where most of] the next round of variations [come] from the more successful of the current generation [heredity]."
All six are essential if the recursion is to keep improving quality. You cannot, as many do, leave out copying or heredity. Otherwise your process will dead-end. What people were calling Darwinian was usually just competition for a limited workspace, with synaptic pruning biased by a multifaceted environment....
Chuck consistently wandered off the neuroscience reservation. I, too, have had such a wandering-off-the-reservation experience. My 2019 book cover shows that I have been captured by the climate problem. I have been camping out with the climate scientists part-time since 1980, initially because our ancestor’s brain size tripled during the back and forth of the ice ages. Evolutionary speed-ups usually happen when climate change keeps repeating. I wanted to know what there was in our ancestors’ environment and ecology that could promote brain enlargement.
Gradually I learned a lot of atmospheric science and oceanography.... In 1997 the editor of The Atlantic phoned me repeatedly, trying to persuade me to undertake the history of Gulf Stream failures.
So here I am at age 84, writing about how to stay away from slippery slopes and then back out of the danger zone for more climate flips. For example, in 2008 the number of billion-dollar windstorms jumped to six times the 1980-2007 baseline. This is no longer gradual change. Climate has become unstable. Neurophysiologists are familiar with gradual change becoming unstable and then going “pop”.
In closing. Last October, Chuck demonstrated another model we might all hope to emulate someday (but not soon). Chuck died both unexpectedly and in the middle of the night. That’s far better than the alternatives. May we all prosper, despite no longer being able to try out our ideas on Chuck.