Pavlovian conditioning is one of the oldest, most frequently studied phenomenon in the field of psychology, learning theory etc., and yet it is still badly misunderstood, not only by folk psychologists, but also by “real” psychologists and other experts, who still see conditioning as the acquisition of one stimulus to evoke the original response to another extit{because of their pairing}. These people have failed to notice that knowledge about conditioning has expanded and that the current understanding contrasts considerably from the traditionalist view. Robert Rescorla, whose work heavily influenced the way we understand Pavlov’s experiment and its implications today, demonstrates how and why the traditionalist understandings of (Pavlovian) conditioning and what it tells us about learning is inadequate: they fail to characterize adequately the circumstances producing learning, the content of that learning, or the manner in which learning influences performance.parA description of conditioning in the literature that Rescorla feels is less than accurate concerns the circumstances which produce Pavlovian conditioning. Often, Rescorla criticizes, descriptions emphasize the contiguity (closeness in time) of two events (unconditioned stimulus, conditioned stimulus) too much. He argues that while contiguity continues to be an essential element, it is neither necessary nor sufficient for the learning of relations: arranging two events contiguously does not necessarily result in learning a relation between the two and not arranging two events contiguously does not hinder the production of an association between them. Instead, he shows that two stimuli (CS/US) with the same contiguity can still differ regarding the “learning effect” they have, due to different information content (information the CS gives about the US).  A CS can fail to be associated with the US when it is equally likely to occur, regardless of the US occurring or not, because in this scenario, the CS provides no information about the US; it is not predictive about the US. Even though contiguity is guaranteed, the timing of the two events is uncorrelated. If, on the other hand, the CS is correlated in time with the US, it will be perceived as informative about the US, resulting in an association between CS and US. To sum up, conditioning is less sensitive to the immediate neighborhood of CS and US per se, but instead, learning is influenced by the extit{background rate}, the base rate of US against which the CS/US contiguity takes place, and the extit{contingency} of two events.parRescorla further mentions that conditioning, ergo learning of relations among events, is not only very sensitive to the informativity the stimuli provide, but also to the nature of the stimuli. Against the common misconception, he shows that not all stimuli are equally associable, implying that some stimuli may be easier to associate with certain signals than others. Older textbooks might give the impression, that extit{any two stimuli could be associated} as conditional stimulus and unconditional stimulus, ignoring instances of preferential learning. However, modern research (like Rescorla, 1980) shows that some stimuli are much easier to associate with a particular biological response than others.In addition, Rescorla complains that psychology textbook descriptions make conditioning sound like a slow and gradual process which requires many repetitions/trials, when in fact “modern conditioning preparations routinely show rapid learning” requiring from 1 to 8 trails.parRescorla feels like not only the circumstances producing learning, the content of that learning, or the manner in which learning influences performance are not captured adequately in many psychology textbooks, he wishes for a fresh image of classical conditioning, since it is not “all spit and twitches” but “intimately involved in the control of central psychological processes, such as emotions and motivation” as well as the immune system, sexual anticipation, tolerance to addictive drugs and much more.

Author