Not long ago I was minding my business waiting for the subway when all of a sudden, at the other end of the platform, a man started aggressively shoving and yelling at the man next to him. Bystanders quickly separated the two, and the whole thing took place too far away for me to make out what exactly the issue was, but the episode left me thinking: What could I infer about the man who did the shoving?
In turns out, nearly three decades of research in social psychology has addressed this very question: What sorts of things do people spontaneously infer about a stranger based on a single action? In other words, in what ways do people automatically go beyond the action itself and "read in" additional information? Would people naturally assume, for example, that the "shover" in my subway episode was, by nature, an aggressive person?
In now-classic studies conducted in the 1980s, Jim Uleman and colleagues presented participants with a series of sentences in which an actor performed an action that implied a certain trait ("The librarian helped the old lady carry her bags across the street.") Note that the trait "helpful" is implied by the sentence, but not actually stated. Later, the researchers gave participants a surprise memory test for all the sentences that had been presented. On this test, they provided participants with one-word cues for each sentence. Some of these cues were words that were related to particular content in the sentence (e.g., "books" might be a cue for the sentence above about the librarian). These were called "semantic cues." Other cues were trait terms one could have inferred about the actor but were never explicitly mentioned (e.g., "helpful"). These were termed "trait cues". The researchers found that trait cues helped participants to recall the sentences just as well as semantic cues did. What does this mean? It indicates that when people read a sentence about an actor, they do more than just process the words. They also spontaneously attach a judgment about the actor's presumed qualities ("That must be a helpful person."). This binding of the trait to the person means that when you subsequently think about the person, the trait comes along with it. A host of studies throughout the 1980s and 1990s demonstrated that people do this without even being aware that they are doing it.
But what else might people infer from someone's action? Are people limited to attaching adjectives to people? More recent studies have begun to provide evidence that the spontaneous inferences people make based on observing a single action are surprisingly rich and complex, especially considering that much of this occurs outside of conscious awareness.
For example, in recent studies, Andrew Todd, Daniel Molden, Jaap Ham, and Roos Vonk wondered whether people might also simultaneously draw inferences about aspects of the situation that could have contributed the actor's behavior. For example, when I saw the subway platform incident, did I also infer that the shover had been provoked by the shovee? To address this question, in one study, the Todd et al. team showed a series of sentences on a computer screen that implied both something about the actor's trait and something about the situation. In between each sentence, the computer presented a string of letters - in some cases the letter string was an actual English word, in other cases it was nonsense. Participants were instructed to press a "yes" key each time the letter string was an actual word and a "no" key when it was not. The computer measured, in milliseconds, how long it took participants to make these word/nonword judgments.
What did they find? Participants were faster to judge words that were traits implied by one of the sentences than words that were not traits. So far, this is consistent with the Uleman et al. studies: When participants read each sentence, they appeared to go beyond the surface action and infer some underlying trait about the actor - and this made these words that much easier to identify. (After all, this would be the second time in a short while that the mind is encountering this trait.) But the Todd et al. group found something else: Participants also responded faster to words that had to do with the situation than the control words. In fact they responded just as quickly to "situation" words as they did "trait" words. This was evidence that, in fact, people do not only attach an adjective to the actor; features of the situation that could have contributed to the action also instantly leap to mind.
Now, in many cases an explanation that invokes the actor's trait can conflict with an explanation that invokes the situation. For example, did the student fail the exam because of his trait of stupidity or the fact that the exam was really hard? These two explanations have opposing implications for your opinion about the student. So how is it that people seem to be able to activate conflicting explanations for the same action? The answer appears to be that people engage in a multi-stage process. First, they generate a range of plausible explanations before ultimately settling on the one option that best characterizes what's happening. Studies like those of Todd et al. appear to catch people at the exact moment when all the explanatory options are on the table but no leading contender has emerged.
What is the importance of this sort of finding? First, it speaks to the awesome speed and sophistication of our first impressions. (Indeed, most of the posts on this blog illustrate in some form or fashion just how much of our opinion about someone is formed within the first second.) It may even be that natural selection favored those humans who could instantly and automatically generate a rich portrait of another person. Given how vital it must have been for the survival of our prehistoric ancestors to make nuanced and accurate predictions about other people, it makes sense that humans would become quite expert at instantaneously raising and weighing multiple potential explanations for a person's action.
Todd, A.R., Molden, D.C., Ham, J., & Vonk, R. (in press). The automatic and co-occurring activation of multiple social inferences, Journal of Experimental Social Psychology, doi: 10.1016/j/jesp.2010.08.006.
Winter, L. & Uleman, J.S. (1984). When are social judgments made? Evidence for the spontaneousness of trait inferences. Journal of Personality and Social Psychology, 47, 237-252.