Analytical sensitivity in the study of rapid facial mimicry: evidence from orangutan play

Rapid facial mimicry (RFM) is the rapid (<1 s), automatic replication of others’ facial expressions, thought to promote emotional sharing and social coordination. Although RFM has been reported in several species, its detection in non-human animals relies on different analytical approaches whose relative sensitivity has never been directly compared. Here, we examined how methodological approaches affect RFM detection in orangutans ( Pongo pygmaeus and Pongo abelii ) during social play. Using the same naturalistic video dataset, we applied three established approaches that differed in their control conditions, behavioural comparisons and analytical design: (1) a detectability-based comparison of perceived versus non-perceived expressions, (2) a comparison between communicative playful facial expressions (Play Face/Full Play Face) and a morphologically similar but non-communicative action (Attempt to Bite) and (3) a comparison between display and matched neutral scenes using multiple statistical procedures. Evidence for RFM emerged only under the second approach: orangutans were significantly more likely to produce playful facial expression after perceiving a Play Face than after perceiving an Attempt to Bite. No evidence for RFM was found using detectability-based or display/neutral comparisons. These findings show that the detection of RFM in orangutans is highly sensitive to analytical design, particularly to the choice of control condition. The emergence of mimicry when communicative and non-communicative facial actions are contrasted suggests that communicative relevance is key in eliciting RFM. Our findings therefore complement existing methodological frameworks by demonstrating that the application of multiple analytical approaches can reveal differences in analytical sensitivity and improve the interpretation of RFM across biological contexts.