Typical Items Facilitate Fear Learning, Atypical Items Don’t

Have you ever recoiled at something because it reminds you of something else that you’re genuinely afraid of? Research indicates that people have a propensity to generalize their fear — so, for example, a person afraid of doctors might also feel uneasy at the sight of a hospital or medical equipment.

This is a photo of a woman covering her face in fear.Moreover, typical items in a category seem to lend themselves to generalization more than atypical items do. For instance, we’re more likely to generalize information about mice and apply it to bats rather than the other way around, since mice come to mind more easily when we think of mammals.

Bringing these different areas of research together, psychological scientists Joseph E. Dunsmoor and Gregory L. Murphy of New York University wanted to investigate whether we incorporate conceptual knowledge into fear learning.

The researchers analyzed data from 37 participants who viewed pictures of animals, either birds or mammals. While viewing one category of animal, such as birds, participants received a mild electrical shock to the wrist. They viewed three highly typical or atypical examples of the selected category three times each, and were shocked two out of the three times they saw each picture.

Participants also viewed the same number of images from the other animal category but did not receive a shock while viewing those images.

Typical examples of birds included a crow, hummingbird, and sparrow, whereas the emu, ibis, and penguin were atypical examples.  Typical mammals included a bear, horse, and rabbit versus the more atypical aardvark, armadillo, and otter.

Next, participants viewed three new examples from both the bird category and the mammal category.

Throughout the experiment, participants rated how likely they thought it was that they would receive a shock for each picture they saw.

The researchers found that while people generalized their fear of an electrical shock to atypical examples after seeing typical examples, the reverse was not true.  In other words, while learning to fear a rabbit might inspire a fear of otters, learning to be afraid of a penguin did not result in fear of a sparrow.

Dunsmoor and Murphy note that their study makes an important contribution to research on the nature of fear:

“The present study adds novel evidence that typicality, a relatively complex attribute, influences generalization of fear learning and opens the way for further research on how other phenomena of category-based induction…integrate with humans’ acquisition and expression of fear.”

The researchers conclude that understanding how we generalize fear learning may carry practical applications, such as treating combat veterans with posttraumatic stress disorder, who may be afraid of objects categorically related to the warzone in which they fought.

Dunsmoor, J., & Murphy, G. (2014). Stimulus typicality determines how broadly fear is generalized. Psychological Science. DOI: 10.1177/0956797614535401


Good day, and thank you for the article. I am very much a fan of what your organisation is doing.

I run a driver training company and have observed that in the case where a learner driver has had a traumatic experience (i.e. a vehicle accident, whether as a pedestrian or passenger), teaching them how to drive to a point where they are sufficiently competent and license-ready takes much longer than teaching someone with no prior traumatic experience, with the goal to have them obtain a similar level of driving competence. However, I’m not sure where to look for or obtain the research citation that would support this observation. Could you kindly assist me in this regard?

I would very much appreciate it (given the learner driver mindset-based/psychological predisposition awareness we have and offer with our driver training). Eagerly awaiting your response.


Leave a Comment

Your email address will not be published.

Required fields are marked*

This site uses Akismet to reduce spam. Learn how your comment data is processed.