Alerts for Assessing “Biological Constraints” on Learning

J Bruce Overmier,Julia E Meyers-Manor

doi:10.46867/ijcp.2015.28.01.01

Abstract

Many researchers have reported differential rates of learning and inferred selective associations between events reflecting biological constraints on learning that have evolved for each given species. Although we do not doubt that there are such biological constraints on learning, we suggest that some of the many claims may actually be spurious due to use of less than optimal research designs. We propose six methodological and inferential concerns that current researchers and reviewers of past research may find useful.

Highlights

Many researchers have reported differential rates of learning and inferred selective associations between events reflecting adaptive specializations or biological constraints on learning that have evolved for each given species
We accept that differential rates of learning to different stimulus-reinforcerresponse event combinations occur and that these may reflect selective associations between events reflecting adaptive specializations and biological constraints on learning
We think that many simple demonstrations of such differences in learning found when using different CSs or different USs are sometimes all too readily initially characterized as reflecting biological constraints on learning

Summary

Many researchers have reported differential rates of learning and inferred selective associations between events reflecting adaptive specializations or biological constraints on learning that have evolved for each given species. Importantly—but generally ignored by those who cite it—is the fact that the classic Garcia and Koelling (1966) experiment lacked a number of important controls that could have been the basis of the purported selective associations. We will call attention to the importance of six methodological and inferential concerns These include (1) the problem of temporary or permanent shifts in an organism’s state (motivations, attention, etc.), (2) the need for controls for non-associative learning, (3) the lack of attention to the particular behavior selected for observation, (4) the problems that arise from using behavioral baselines that are unidirectional – typically from zero, (5) the need to consider what law of learning is being violated, and (6) advantage of using the power of double dissociation experiments. Just to make our tasks as researchers harder, the methodological issues we raise below are not necessarily independent of one another

Methodological Alerts and Challenges

The Behavior Selected for Observation

Fin Erection

What Law of Learning is being Violated?

Double Dissociation Paradigm

Conclusions