a quarterly of art and culture Issue 39 learnIng c US $12 canada $12 uK £7 3 0 5 6 3 5 8 9 8 9 6 6 5 0 Aplysia californica inking. Courtesy Genevieve Anderson. learning Degree Zero dismiss its remains as yet another nasty gob of formless D. Graham Burnett protoplasm coughed up by the deep. But in continuing your stroll, you would have walked past the creature Aplysia californica is a soft-bodied marine invertebrate upon which much of the modern scientific understand- that makes its living nibbling red algae in the kelp beds ing of learning has been built. and rocky reefs of the Pacific coast of the united States. For close to fifty years, Aplysia californica they are meaty, sometimes weighing more than ten has served as the irreplaceable model organism at pounds, and the deep hues of their food give their the center of what is arguably the most significant translucent flesh the mottled marbling of polished tor- twentieth-century research program in the sciences toiseshell. When scrunched up contentedly, they look of mind and behavior. In fact, on the afternoon of 8 a bit like rabbits, on account of two earlike tentacular December 2000, when the distinguished neuroscien- horns that sit atop their faceless heads—a resemblance tist eric Kandel stepped to the podium of Sweden’s that long ago gave rise to the common name for the Karolinska Institute to present his nobel Prize lecture, whole taxon, known as the “sea hares” since antiquity. he flashed up on the screen behind him a huge, photo- When irritated, A. californica release a cloud of acrid, shopped image of a robust Aplysia sporting around its mucosal, purple ink from a shockingly labial cleft that extended neck Kandel’s golden medallion—the most runs down the length of their backs and shelters the coveted prize in science. the mollusk, it was agreed, delicate folds of a feathery, blood-rich gill. they move deserved a little bling on the occasion. feelingly and with a certain dignity, like large slugs— So how did a herbivorous hermaphroditic gas- which is exactly what they are. tropod end up as a Stockholm honoree? the answer Stumbling on a defunct Aplysia in the tide-washed requires a quick turn through the twentieth-century flotsam of a California beach, you would be inclined to science of learning and memory. 59 to begin with, then, ask the most basic question: the product. and for this reason questions about how What is “learning,” anyway? a recent handbook in the we learn have long been entangled with the question of field offers the following definition: how we form and maintain memories. across the first half of the twentieth century, a Learning is the process by which relatively permanent variety of physiologists and experimental psychologists changes occur in behavior potential as a result of went after these problems in different ways. Some, experience.1 like the German psychologist hermann ebbinghaus, worked directly with human subjects and tried to Which is to say, learning is a little like an accelerated identify consistent patterns in the dynamics of informa- mechanism for evolutionary adaptation, one that tion acquisition. In a series of classic experiments that works at the scale of the individual. If you can figure out involved memorizing long sequences of nonsense syl- how reliably to convert new experiences into revised lables, ebbinghaus eventually codified what we now behavior, you’ve just gotten a serious leg up on the busi- call “the learning curve”—a time-sequence plot of the ness of being here, since you’ve effectively acquired decreasing effort needed to master a given task. Other the capacity to redeem some measure of the general investigators, like the russian physiologist Ivan Pavlov unpleasantness of existence—you can now turn your and the american psychologist edward thorndike, bad breaks into lessons. left humans aside and made fundamental contribu- true, there is an existential price to be paid for this tions to the understanding of learning by studying trick, in that now you are stuck with memory, against animals. Pavlov’s dogs, famously, learned to associate which nietzsche (among others) railed very persua- one stimulus (say, a bell) with another (say, a zap in sively. But no memory, no learning, since if you can’t the foot) in what has come to be known as “classical store information, you’ve got no way to make your past conditioning”—a behavioral phenomenon that has bear on your future. Learning is the process, memory is served ever since, for better or worse, as a basic Forty-nine-day-old Aplysia juvenile, nicknamed Claire, that has just completed metamorphosis. Courtesy the National Resource for Aplysia at the University of Miami. 60 laboratory proxy for learning writ large. thorndike, how to run a given maze, and then surgically remove working at Columbia university, studied cats as they various bits of their brains, and they still generally man- tried to figure out how to escape from a locked box. aged to find their way to the cheese. In the end, their Whereas in Pavlovian classical conditioning there performance appeared to have more to do with how is an arbitrary relationship between the stimuli to much of their brains you removed—not what part. But be associated (Bell? electrocution? no logical con- how could one investigate the basic biology of learning nection), in thorndike’s experiments the cats were and memory if it was so difficult to specify where to learning to make associations that weren’t arbitrary look? this sort of thing depressed physiologists, who at all (pull latch, get out of box). this kind of learn- like to find organs for different functions. meanwhile, ing—learning, via payoffs, about the world of causes the psychologists largely took up versions of what and effects—came to be called “operant conditioning.” is called “behaviorism,” a paradigm of mind science Philosophically oriented wags have tended to wonder that black-boxed cognitive activity, and treated organ- just how different these two kinds of conditioning really isms like input/output systems—which did not make are (isn’t classical conditioning simply the operant for much progress on the internal biology of learning. conditioning of animals forced to live in the “environ- Surveying the field in the early 1950s, Lashley, old and ment” of a laboratory?), but put this knotty matter aside. resigned, summed up the situation on a paradoxical What was significant was that all of these learning note: “I sometimes feel, in reviewing the evidence… situations were shown to share many consistent fea- that the necessary conclusion is that learning is just not tures: practice makes perfect; time off tends to reduce possible.”2 competency; reacquisition of capacities is faster than and yet, of course, it was. the question was: how at the first go. and these sorts of generalities could be to learn about learning? quantified, parameterized, and shown to possess reli- In an important book published in 2007, Science able statistical signatures that were conserved across without Laws, angela Creager and a number of other experimental animals and scenarios. historians of science drew attention to the importance By mid-century, then, there was broad consensus of “model organisms” in the history of twentieth- among researchers about the descriptive features of century biology. there are whole fields of investigation the basic behavioral dynamics of learning. But there that seem inextricable from particular creatures, which was nothing even remotely consistent about the various for various reasons (rational foresight, historical con- theories proposed to explain the underlying physiology tingency) end up functioning as obligatory passage of the process. how did learning actually happen? What points for disciplinary development. What would the was going on in the brain that accounted for these twentieth-century study of genetics have been without observable changes in the doings of dogs, cats, mon- Drosophila (the fruit fly)? Inconceivable. What about keys, mice, and army recruits? (the latter were great virology absent TMV (the tobacco mosaic Virus)? Simi- experimental subjects—a perfectly captive audience.) larly difficult to imagine. So profound is this linkage no one could say with any satisfaction. Yes, people between certain kinds of science and certain specific knew a fair bit about nerves and how they transmitted beasts that it has been dignified with a name, the Kuf- signals around the body. Yes, there was some evidence fler-Grundfest principle: “For every biological problem that regions of the mammalian brain were specialized there is a suitable organism in which to study it.”3 to manage particular cognitive, sensory, and motor When, in the early 1960s, Kandel—then an ambi- activities. and yes, there were even some ideas about tious young medical doctor—decided to turn his how to tie brain activity to behavior (one thinks here of research attentions to the neurophysiology of learn- the work of both Donald hebb and Karl Pribram). ing and memory, he took this principle to heart. Cats? nevertheless, exactly how the nervous system Dogs? rats? too complicated. the learned behaviors acquired and stored new information was hotly dis- in question? too intricate. Despite his Viennese origins puted, and even the best arguments remained largely and expansive interest in the nuance and complexity of speculative. among other difficulties, the available Freudian psychoanalysis, Kandel had a biophysicist’s experimental data were exceedingly hard to interpret. appetite for radical reduction. he wanted to study For instance, in a fascinating and discouraging set of learning in an animal built for the very simplest kinds investigations in this area, the most important ameri- of information acquisition and storage. an animal that can learning researcher in the period, Karl Lashley, could be understood as a little laboratory learning- demonstrated that you could take rats that had learned machine: limited behavioral repertoire; large, simple 61 wiring; a resilient metabolism; and, ideally, small teeth scientists were down at the level where a living crea- (no one likes getting chomped by lab animals).