In Brief

If you eat peanut butter and have an allergic reaction, and then eat peanut butter and jelly and have the same allergic reaction, you might conclude that you're allergic to the peanut butter, not the jelly. However, if you break out in hives after eating jelly alone, and again after eating peanut butter alone, you would probably conclude that you're allergic to both.

Rats too can make such inferences, according to a new study published in the Journal of Experimental Psychology: General (Vol. 135, No. 2). The finding runs counter to existing theories that associative processes are automatic and not reasoned, says Tom Beckers, PhD, one of the study authors and a postdoctoral fellow at the University of Leuven in Belgium.

In the study, the researchers trained rats to expect a shock after they heard a low tone. They then simultaneously played another tone-this one higher pitched-with the low tone before shocking the rats. Later, the animals showed fear to the low-pitched tone, but not the high-pitched one-a well-studied effect known as forward blocking.

Many have theorized that forward blocking occurs because a given unconditioned stimulus-such as a shock-has only so much power to be linked with other stimuli, says Beckers. However, a second part of Becker's experiment suggests that more may be going on than classical conditioning. Rather, the rats may be engaging in reasoning similar to that of a human observing a food allergy, he says.

Like the person who eats peanut butter alone and jelly alone and suffers allergic reactions to both, the second group of rats underwent pretraining, in which they learned that two different stimuli-a flashing light and a buzzer-both preceded the same electric shock. Then, the animals went through the same procedure as the first group of rats, with first a low tone, then low and high tones being played simultaneously before a shock. However, these pretrained rats did not show forward blocking-they responded fearfully to both the low and the high tones, in contrast to the first group that feared only the low one.

This suggests that the animals learned during pretraining that signals before an electric shock do not have an additive effect, and they applied that knowledge to a second set of circumstances, Beckers says. The results may be controversial among behaviorists who believe that effects like forward blocking in rats are automatic and cannot be altered by thoughts, he notes.

"It does indeed suggest some sort of cognitive processing is involved, some sort of reasoning-like process," he says. "We cannot exclude that an associative explanation will be found some day, but at the very least, our results suggest that rats and humans are not so different in the way they handle these kinds of puzzles."

--S. Dingfelder