The benefits of body-language mimicry have been confirmed by numerous psychological studies. And in popular culture, mirroring is frequently urged on people as a strategy -- for flirting or having a successful date, for closing a sale or acing a job interview. But new research suggests that mirroring may not always lead to positive social outcomes. In fact, sometimes the smarter thing to do is to refrain.

In a study to be published in a forthcoming issue of Psychological Science, Piotr Winkielman and Liam Kavanagh of the psychology department at the University of California, San Diego, along with philosophers Christopher Suhler and Patricia Churchland, also of UC San Diego, note that in real-life situations there are often observers to the mirroring that takes place between two people. This led them to wonder whether mimicry sometimes comes at a reputational cost. Are there cases in which an observer might actually think less of a person for mirroring the behavior of another?

Results of three experiments suggest that mimicry is more nuanced than previously thought and not, the authors write, "uniformly beneficial to the mimicker."

"Mimicry is a crucial part of social intelligence," said Winkielman, UC San Diego professor of psychology. "But it is not enough to simply know how to mimic. It's also important to know when and when not to. The success of mirroring depends on mirroring the right people at the right time for the right reasons. Sometimes the socially intelligent thing to do is not to imitate."

Participants in the study were asked to watch several staged and videotaped interviews. Some saw videos in which the interviewer was cordial and others saw videos in which the (same) interviewer was unfriendly. The people being interviewed in the videos either mirrored the interviewer's simple mannerisms, such as chin-touching or leg-crossing, or they did not. After watching each video, participants evaluated the interviewee on general competence, trustworthiness and likability.

Despite the fact that the participants were not instructed to watch for mimicry and reported no awareness of it, it still influenced their evaluations: Interviewees who mimicked the unfriendly interlocutor were judged to be less competent than those who didn't. That is to say, in the eyes of the outside observers, the imitators of the undesirable model incurred reputational costs -- their unconsciously observed mirroring registered as a kind of error.

In a second corollary experiment, participants were exposed to the same videos but with the interviewer obscured. In other words, they couldn't see any evidence of mimicry, and the results support the researchers' hypothesis: It is not merely interacting with negatively perceived people that has a social cost; you pay a price for aligning with them through body language.

Interestingly, an additional experiment showed that the reputational cost of mimicking an unfriendly interviewer disappeared when participants read positive information about that interviewer -- i.e., that he was engaged in humanitarian work -- before watching the video. "It's almost as though mimicry of a condescending interviewer was forgiven when he was judged to be good at heart," Winkielman said.

Our social lives are incredibly complex, said Winkielman, and in order to build or maintain relationships we have to keep in mind a variety of factors. "It's good to have the capacity to mimic," he said, "but an important part of social intelligence is knowing how to deploy this capacity in a selective, intelligent, context-dependent manner, and understanding, even implicitly, when mirroring can reflect badly on you."