Shufflebrain

Experimental psychologists have learned to take great care to eliminate what is called pseudo-conditioning: a positive response by the subject when there's no actual association between two the putative stimuli. One measure is known as the extinction test. After an animal has apparently made a specific association (gets the test right in, say, 90-95 percent of the trials), its performance will drop upon withdrawal of the reward or punishment (US) in direct relationship to the number of trials made with only the conditioned stimulus (CS). If the animal associates the CS with an extraneous variable its performance will not drop off when you terminate the US.

Carl applied the extinction test to every subject. All passed. He also stimulated animals with light alone or shock alone, as an additional control against pseudo-conditioning. In addition, we included eyeless animals in the study. Controls for mental telepathy? You might wonder. But a few years later, excellent evidence came along to show that animals, in fact, have some non-visual perception of light. Our eyeless animals didn't learn Carl's test.[6]

I'd personally raised every animal in the colony from early embryonic stages. All were Ambylstoma punctatumm from two large clutches of eggs (those in a clutch have the same parents). I kept accurate records on individuals; later we ran comparisons of data from the two clutches and eliminated egg clutch (pedigree) as a factor in our findings. I made sure each animal had developed simultaneously through the same embryonic stages. Size and mass? Animals were of the same snout-tail length (in millimeters) and displaced the same volume of fluid (which beats trying to weigh an aquatic animal on the pan of a balance).

Given all the controls Carl had insisted upon, and the numbers of each type of subject necessary to make learning difference statistically significant, he could not train and test the entire group all at one time. Therefore (to avoid the variables implicit in the latter fact), I organized the animals into working squads, each squad with at least one representative of each type of animal. All members of a squad went under anesthesia at the same time and I revived them all together. Squad members lived in individual dishes, but I kept the dishes in the same stack, and where one dish went all the others in the stack went, too. Their spring water came from the same carboy, and even their tubifex worms came from the same culture. I spaced the operation for different squads over a period of days to give Carl some latitude. At the conclusion of the study we ran analyses and found no difference attributable to the particular squad an animal belonged to.

***

Cyclopes dictated when Carl could start the light-shock training. They began using the top-mounted eye about two weeks after surgery. When we dangled a worm above the water, the denizen would ascend hungrily from the depths, "like a submarine surfacing to salvage a free bargeload of beer," as a pharmacologist colleague of ours characterized it. After the transplanted eye began to work, the Cyclopes preferred sight to all other senses. Vision working again, they seemed compulsive about taking a good look at the worm before making a strike. Often, inspection required acrobatics. If a worm dropped below the plane of the salamander's surfaced-oriented visual field, the animal would duck, twist and pirouette on its snout to aim the eye. It would even poke and thrust at a wriggling red mass, using the eye as a prod and reminding the pharmacologist with the beer of "a rhinoceros chasing one of the Three Stooges around a mimosa tree."

Carl's instincts told him to wait a couple of extra weeks before launching the light-shock avoidance training. So that he could carefully control temperature and keep the animals in minimal illumination, I turned over my inner sanctum darkroom to him.

***

Over the years Carl and I have observed a rule: no chewing the fat about the data until they've all been retrieved, cleaned up and run through statistical analyses. It isn't just a matter of introducing bias, which is hard enough to control, but to hedge against a potential hell an experimentalists usually learns early on in the career: letting imagination run wild with a false lead. The pure agony in retracting one's brilliant speculations is akin to the Katzenjammer Kid's dachshund having to surrendering a stolen string of succulent sausages. Thus when the big day finally arrived, I had not a hint of what the returns would say about the one-to-one principle of perception and learning. I did think we'd be making big names for ourselves by showing IQ could be raised with the knife.