Skip to main content
Log in

What Do People Think Is an Emotion?

  • Published:
Affective Science Aims and scope Submit manuscript


In emotion research, both conceptual analyses and empirical studies commonly rely on emotion reports. But what do people mean when they say that they are angry, afraid, joyful, etc.? Building on extant theories of emotion, this paper presents four new studies (including a preregistered replication) measuring the weight of cognitive evaluations, bodily changes, and action tendencies in people’s use of emotion concepts. The results of these studies suggest that the presence or absence of cognitive evaluations has the largest impact on people’s emotion attributions, and bodily changes and action tendencies are considered to depend on cognitive evaluations. Implications for theories of emotion (concepts) and the interpretation of emotion reports are discussed.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Similar content being viewed by others


  1. The spoken language section of COCA consists of transcripts of unscripted conversation from more than 150 different TV and radio programs between 1990 and 2019.

  2. The literature on concepts offer different proposals regarding what statistical and/or salience properties determine the weight of a feature (Murphy, 2002; Machery, 2009). Here, we adopt an uncompromising notion of weight as the impact of (the presence or absence of) a feature in categorization, which remains silent regarding what determines this impact.

  3. This has been called the Problem of Parts (Prinz, 2004): The problem of determining which features are essential for emotion and which are not. Accounts that identify emotion with a combination of several features face a different problem, the Problem of Plenty, which consist of explaining how very different elements hang together to create a unitary phenomenon.

  4. An exception is Study 6 in Fehr and Russell (1984), which is further discussed in this section.

  5. Note that the question regarding which features determine emotion attribution (the one that this paper concerns) is different from the question regarding which features determine the intensity of the attributed emotion (Frijda et al., 1992).

  6. These emotions are sometimes called “basic emotions” (Ekman, 1999). Here, we do not take a stand on whether these emotions are basic or not. These emotions were selected merely because they are paradigmatic examples of emotion.

  7. The wording for Cognition builds on Richard Lazarus’s list of the core relational themes for each particular emotion (Lazarus, 1991, p. 122). The expression “seeing as” is taken from Nussbaum (2004, p. 197) and Robert Solomon, who claims “emotion is viewed as a way of seeing something as a thing of a certain sort” (Solomon, 2003, p. 63). The description of the Body element follows the idea that because “emotions may each correspond to several physiological patterns […] it is best to associate emotions with body state prototypes” (Prinz, 2004, p. 72). The prototypical reaction for each emotion was taken from previous research on the topic (Scherer & Summerfield, 1983; Shaver et al., 1987; see “Study 1”). Finally, the wording for Motivation is based on Andrea Scarantino’s list of action tendencies, which in turn builds on the work of other researchers (see Scarantino, 2014, p. 181). Note that some of these action tendencies can be interpreted as behaviors or goals. The expression “feel the urge” is taken from Nico Frijda, who claims that “Introspections produce a wealth of statements that refer to […] impulses to approach or avoid, desires to shout and sing or move, and the urge to retaliate” (Frijda, 1988, p. 351). Furthermore, I would like to thank Julien Deonna, Jesse Prinz, and Mauro Rossi for their feedback regarding the wording.

  8. This and the following effect size interpretations are based on Cohen’s (1988) benchmarks.

  9. Tests concerning a fourth hypothesis (H2: No feature or combination of features is both necessary and sufficient for emotion attribution) can be found in footnotes 10 and 11.

  10. The percentage of participants who disagreed (Emotion Ratings < 4) that BM (.27), CM (.18), and CB (.12) cases were cases of emotion was not significantly higher than .50 (all ps > .05). This is a very minimal test of necessity. Its results suggest that none of these features is necessary for emotion.

  11. The percentage of participants who agreed (Emotion Ratings > 3) that C (.63), B (.33), and M (.54) cases were cases of emotion was not significantly higher than .50 (all ps > .05). This is a very minimal test of sufficiency. Its results suggest that none of these features is sufficient for emotion.

  12. Note, however, that a common objection against Cognitive theories of emotion states that cognitive evaluations can be made in a dispassionate way (Deigh, 1994). Thus, cognitive evaluations might not be highly specific of emotion (vs. non-emotion).

  13. In Study 1, mean emotion ratings in the motivation condition (M) were lower for anger (2.00, SD=.82) than for fear (3.43, SD=1.13) and disgust (3.14, SD=1.34). All Ns < 8. In Study 2, mean emotion ratings in the motivation condition (M) were again lower for anger (3.36, SD=.81) than for fear (4.33, SD=.98) and disgust (3.58, SD=1.00). All Ns < 13.


Download references

Author information

Authors and Affiliations


Corresponding author

Correspondence to Rodrigo Díaz.

Ethics declarations


This work was funded by the Swiss National Science Foundation (research projects 100012_169484 and P2BEP1_200040). I would like to thank Claus Beisbart, Julien Deonna, Charlie Kurth, Anne Meylan, Jesse Prinz, Kevin Reuter, James Russell, Mauro Rossi, and Christine Tappolet for their comments on early drafts of the paper.

Data Availability

Data and materials are available at the Open Science Framework website:

Ethics Approval

Research was approved by the Ethics Commission at the University of Bern.

Conflicts of Interest

The author declares no competing interests.

Informed Consent

Participants provided informed consent to participate in the studies.

Code Availability

Not applicable

Consent for Publication

Not applicable

Consent for Publication

Not applicable

Authors' Contributions

Not applicable

Additional information

Handling editor: Phoebe Ellsworth

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Díaz, R. What Do People Think Is an Emotion?. Affec Sci 3, 438–450 (2022).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: