Summary: New meta-analytic research challenges the long-held belief that each emotion produces a single, consistent physiological pattern. Instead, bodily responses to emotions appear to form a diverse population of possible reactions.
Source: Northeastern University.
How do you feel when you’re angry? Tense, jittery, drained? Is it always the same, and is it the same for someone else who feels anger? Most likely, the answer is no—your bodily experience of an emotion can change across situations and differs between people.
New research from Northeastern University demonstrates that the body’s responses during emotions such as anger, fear, sadness, disgust, surprise, and happiness vary widely. These findings overturn a centuries-old assumption that each emotion has a distinct physiological “fingerprint.”
For decades, many psychologists have assumed that specific emotions reliably produce particular patterns in measures like heart rate, blood pressure, and breathing. The new study, led by Lisa Feldman Barrett, University Distinguished Professor of Psychology at Northeastern, together with Karen S. Quigley, research associate professor of psychology, and Erika Siegel, now a postdoctoral fellow at the University of California, San Francisco, shows that this assumption does not hold up when examined across many studies and thousands of participants.
Barrett, Quigley, and colleagues conducted a meta-analysis of more than 200 laboratory studies that measured autonomic nervous system activity and reported bodily responses during induced emotional states in over 8,400 adults. Published in the journal Psychological Bulletin, the analysis found that variation in physiological responses is the norm, not the exception. Co-authors on the study include former doctoral students Paul Condon, Yale Chang, and Molly Sands, and Jennifer Dy, associate professor of electrical and computer engineering.
“Across hundreds of studies we show convincingly that the everyday assumption—each emotion carries a reliable bodily fingerprint—is false,” Barrett said. Instead of a single signature response for each emotion, the data indicate a population of possible physiological reactions that can differ widely by person and situation.
Measures such as heart rate, respiration, and blood pressure fluctuate across episodes classified in the same emotion category, meaning two instances of “anger” can produce very different bodily patterns. This pattern of variability also mirrors prior work showing that facial expressions are far from uniquely tied to single emotions: people scowl while angry some of the time but also use many different facial gestures when angry, and scowling can occur when they are not angry.
Implications for emotion-sensing technology
These findings have important consequences for companies developing devices and applications that attempt to infer emotions from physiological data. Many technologies rely on the premise that a particular pattern of peripheral measures corresponds to a single emotion category—attempting, for example, to detect “happiness” solely from heart rate variability or skin conductance. The study’s results caution that such approaches are likely to be unreliable.
Barrett suggests that emotion recognition technology should move away from a one-size-fits-all model. “Instead of building a single model that reads anger the same way for everyone, developers may need systems that learn how an individual’s physiological responses vary with emotion over time and across contexts,” she said. Personalized, context-aware models are more consistent with the observed variability in how emotions manifest in the body.
Neil Harrison, Wellcome Clinician Scientist and Reader in Neuropsychiatry at Brighton and Sussex Medical School, noted the long-standing debate dating back to Darwin about whether emotional categories have distinct bodily signatures. He called this meta-analysis “an ambitious attempt” to resolve the question and said the results support a re-evaluation of the classical view that links automatic nervous system patterns uniquely to specific emotions.
Source: Allie Nicodemo – Northeastern University
Publisher: Organized by NeuroscienceNews.com
Image Source: Image credited to Northeastern University.
Original Research: Abstract for “Emotion fingerprints or emotion populations? A meta-analytic investigation of autonomic features of emotion categories” by Erika H. Siegel, Molly K. Sands, Wim Van den Noortgate, Paul Condon, Yale Chang, Jennifer Dy, Karen S. Quigley, and Lisa Feldman Barrett, published in Psychological Bulletin, April 16, 2018.
DOI: 10.1037/bul0000128
Northeastern University, “Physiological Reactions to Emotions Are Not Uniform.” NeuroscienceNews, 18 April 2018.
Abstract
Emotion fingerprints or emotion populations? A meta-analytic investigation of autonomic features of emotion categories
The classical perspective holds that distinct emotion categories correspond to specific autonomic nervous system (ANS) “fingerprints.” Under that view, variability within a category is treated as incidental. By contrast, the theory of constructed emotion proposes that an emotion category consists of a population of context-dependent, variable instances that do not require a shared ANS fingerprint. To test these hypotheses, the authors meta-analyzed 202 laboratory studies that measured ANS reactivity during induced emotional states in nonclinical adult samples. Using random-effects, multilevel meta-analysis and multivariate pattern classification, they found increases in mean effect size for many ANS measures across emotion categories, but no clear pattern that distinguished one emotion category from another. Significant heterogeneity was present within categories, with moderate to substantial variability in over half of the observed effect sizes. Experimental factors such as induction method (e.g., film clips versus imagery) explained only a small portion of this variability. Correcting for publication bias further reduced estimated effect sizes and, in some cases, increased heterogeneity. These results align more closely with population-based thinking and principles from evolutionary biology that underlie the theory of constructed emotion, and they suggest new directions for research on the nature of emotion.