The research behind Self-Perception Theory
We usually assume that behavior begins with something internal. You know what you like, what you believe and what matters to you and those attitudes influence what you do. You volunteer because you care about the cause. You spend Saturday morning running because you enjoy running. You keep choosing a particular kind of work because it interests you.
Psychologist Daryl Bem proposed that the relationship can sometimes work in the opposite direction. When our internal attitudes are weak, ambiguous or difficult to identify, we may partly figure out what we think by observing our own behavior and the circumstances in which it occurs.
In other words, we don't always know exactly what we think and then act accordingly. Sometimes we notice what we keep doing and use that behavior as evidence about ourselves.
Bem called this Self-Perception Theory.
THE RESEARCH
Self-Perception Theory emerged partly from a disagreement about how to explain some of the findings associated with Cognitive Dissonance Theory.
One of the most famous examples was Festinger and Carlsmith's 1959 experiment in which participants completed an extremely boring task and were then paid either $1 or $20 to tell another person that the task had been enjoyable. Participants paid only $1 subsequently rated the task more positively than those paid $20.
Festinger interpreted the finding in terms of cognitive dissonance. The participants paid $20 had an obvious external justification for what they had said: they were well paid to say it. Those paid only $1 had less external justification, creating greater inconsistency between their original attitude and their behavior. Changing their attitude towards the task offered one way to reduce that discomfort.
Bem proposed another possibility.
Perhaps people did not need to experience an uncomfortable internal conflict and then change their attitude to resolve it. Perhaps they could infer their attitude by examining the available evidence.
The reasoning would look something like this: I told someone that the task was enjoyable. I was only paid $1, so the money doesn't fully explain why I said it. Perhaps I found the task somewhat enjoyable after all.
To investigate this idea, Bem conducted what he called interpersonal simulations of earlier dissonance experiments. Instead of placing people in the original situations, he gave observers descriptions of the circumstances and asked them to predict the attitudes of the people involved.
The observers had access to essentially the same external information a participant could use when interpreting their own behavior: what the person had done and how much external incentive there had been for doing it.
Remarkably, observers could reproduce important patterns found in the original experiments.
They did not need access to the participant's private psychological state to infer the resulting attitude. They could make the judgment by looking at the behavior and its context.
Bem argued that people might sometimes understand themselves in a similar way.
THE FINDING
Self-Perception Theory proposes that people can infer their own attitudes from observations of their behavior, particularly when their internal attitudes are not already strong or obvious.
Imagine you have started attending a weekly ceramics class. Nobody requires you to go. You continue paying for it, rearrange your schedule to attend and often stay longer than necessary.
You might reasonably conclude: I must really enjoy ceramics.
That conclusion may sound obvious, but notice where the evidence came from. Instead of beginning with a perfectly clear internal statement such as ceramics is important to me, you observed a pattern in your own behavior and used it to understand your preference.
We routinely make similar judgments about other people. If someone spends every weekend volunteering at an animal shelter without being paid or required to be there, you might infer that they care about animals. If someone repeatedly chooses difficult mathematics courses when easier alternatives are available, you might conclude that they are genuinely interested in mathematics.
Bem's argument was that under certain conditions, we can use the same inferential process on ourselves.
Our behavior becomes evidence.
THE THEORY
The circumstances surrounding behavior are crucial to the theory.
If there is a strong external explanation for what you are doing, there is less reason to infer an internal attitude from the behavior.
Suppose you attend a networking event because your employer requires you to be there. Staying for three hours tells you relatively little about whether you enjoy networking. There is already an obvious explanation for your behavior.
If you voluntarily stay an additional two hours after the required portion has ended, however, the behavior becomes more informative.
This is similar to the reasoning we use when interpreting other people. Someone doing something because they are paid, threatened, required or heavily rewarded provides weaker evidence about their personal preference than someone choosing the same behavior without those external pressures.
Self-Perception Theory is therefore not simply I do something, therefore I like it. The inference depends on what other explanations are available.
The theory is also most relevant when your existing attitude is relatively uncertain.
If you passionately hate olives, noticing that you accidentally ate one at dinner is unlikely to make you conclude that you love olives. You already possess a strong attitude and one piece of behavior provides little reason to revise it.
But when your preference is less clear, behavior can provide useful information.
NOT THE SAME THING
Self-perception is not cognitive dissonance
The two theories can sometimes predict similar outcomes, which is precisely why Bem's work became important.
Cognitive Dissonance Theory proposes that inconsistency among attitudes, beliefs and behavior can produce psychological discomfort and that people may change their attitudes to reduce that discomfort.
Self-Perception Theory does not require this internal tension.
Instead, it proposes an inferential process. You observe your behavior, consider the circumstances and draw a conclusion about your attitude.
Later research has suggested that the theories may be more useful under different conditions rather than requiring one to completely replace the other. Dissonance processes may be particularly relevant when behavior conflicts with attitudes that are already important or strongly held. Self-perception can be particularly useful for understanding attitudes that are initially weak or ambiguous.
The distinction matters because Self-Perception Theory is not fundamentally a theory about people rationalizing behavior they secretly know contradicts their beliefs.
It is a theory about how we sometimes discover what those beliefs are.
WHY IT MATTERS
We often treat introspection as the primary route to self-knowledge.
If you want to know what you value, you think about your values. If you want to understand your interests, you ask yourself what interests you. If you are trying to understand what kind of life you want, you imagine different possibilities and notice how they make you feel.
All of that can be useful.
But introspection has limits. People are not always able to produce clear explanations for their own preferences and motivations simply by looking inward. Sometimes the answer to What matters to me? is easier to approach by examining what repeatedly happens in your actual life.
Where does your attention go when nobody assigns it?
What do you keep learning about without needing a reason?
Which activities do you return to even when nobody rewards you for doing them?
Who do you consistently make time for?
What responsibilities do you voluntarily take on?
What do you spend money on when your basic needs are already covered?
None of these behaviors provides a perfect window into identity. Behavior is constrained by money, obligations, opportunity, habit and circumstance. Someone may value travel enormously and rarely travel because they cannot afford it. Someone may spend most of their time working without considering work their highest value.
Context still matters.
But repeated voluntary behavior can provide information that introspection alone sometimes misses.
There is also a more subtle implication. If we use our own behavior as evidence about ourselves, repeated actions can contribute to the stories we develop about who we are.
Someone who repeatedly avoids speaking in groups may eventually conclude I'm a quiet person. Someone who begins exercising regularly may gradually start describing themselves as someone who exercises. Someone who continually organizes social plans may come to see themselves as the person who brings people together.
The identity does not necessarily exist fully formed before the behavior.
Sometimes the behavior helps provide the evidence from which the identity develops.
TRY THIS
Choose one question about yourself that you have had difficulty answering. It could be what you enjoy, what matters to you, what kind of work suits you or what you want more of in your life.
Instead of trying to answer it through another round of thinking, look at your behavior over the last few months.
Focus particularly on things you chose when there was no obvious external requirement or reward. What did you repeatedly make time for? What did you research? What conversations did you keep returning to? What did you continue doing after you technically could have stopped?
Then consider the context. Was the behavior genuinely voluntary? Was there another obvious reason you were doing it? Is the pattern repeated enough to tell you anything useful?
You don't need to turn every behavior into a statement about your identity. A single choice is rarely that informative.
Look for patterns.
Sometimes understanding yourself requires asking what you think and feel.
Sometimes it is worth looking at what you have actually been doing.
References
Bem, D. J. (1967). Self-perception: An alternative interpretation of cognitive dissonance phenomena. Psychological Review, 74(3), 183-200.
Bem, D. J. (1972). Self-perception theory. Advances in Experimental Social Psychology, 6, 1-62.
Fazio, R. H., Zanna, M. P., & Cooper, J. (1977). Dissonance and self-perception: An integrative view of each theory's proper domain of application. Journal of Experimental Social Psychology, 13(5), 464-479.
Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Verbal reports on mental processes. Psychological Review, 84(3), 231-259.