I reject Lindquist et al.'s implicit claim that all emotion theories other than constructionist ones subscribe to a “brain locationist” approach. The neural mechanisms underlying relevance detection, reward, attention, conceptualization, or language use are consistent with many theories of emotion, in particular componential appraisal theories. I also question the authors' claim that the meta-analysis they report provides support for the specific assumptions of constructionist theories.

Because of the complexity of human emotional responses, invariants must be sought not in the responses themselves, but in their generating mechanisms. Lindquist et al. show that functional locationism is a theoretical dead end; their proposed mechanistic framework is a first step toward better models of emotional behavior. We caution, however, that emotions may still be quasi-natural perceptual types.

Lindquist et al. remark that not all fear instances lead to heightened amygdalar activity and, instead, point to roles of the amygdala in detecting “motivationally salient “or “emotionally impactful” stimuli. By reviewing research on the amygdala's functions in episodic-autobiographical memory, we further emphasize the involvement of the amygdala in coding the subjective relevance and extracting the biological and social significance of the stimuli.

This commentary (1) raises the question about the possible conflation of core affect with the neural representation of interoceptive changes in regard to whether biological value is subpersonal or must be experienced, and (2) proposes that Wundt's third dimension of core affect – strain-relaxation – can be accounted for in the target model under a generalised predictive model of attention.

Parental brain responses to baby stimuli constitute a unique model to study brain-basis frameworks of emotion. Results for baby-cry and picture stimuli may fit with both locationist and psychological constructionist hypotheses. Furthermore, the utility of either model may depend on postpartum timing and relationship. Endocrine effects may also be critical for accurate models to assess mental health risk and treatment.

Emotional reactions are rather flexible, due to the schema-like organization of complex socio-emotional situations. Some data on emotion development, and on certain pathological conditions such as alexithymia, give further support for the psychological constructivist view put forward by Lindquist et al. Narrative organization is a key component of this schematic organization. The self-related nature of narrative organization provides scaffolding to the contextual dependency of emotions.

In their analyses of the neural correlates of discrete emotionality, Lindquist et al. do not consider the numerous drawbacks to inferring psychological processes based on currently available cognitive neurometric technology. The authors also disproportionately emphasize the relevance of neuronal activation over deactivation, which, in our opinion, limits the scope and utility of their conclusions.

Lindquist et al. provide a compelling summary of the brain bases of the onset of emotion. Their conclusions, however, are constrained by typical fMRI techniques that do not assess a key ingredient in emotional experience – timing. We discuss the importance of timing in theories of emotion as well as the implications of neural temporal dynamics for psychological constructionism.

Lindquist et al.'s meta-analysis focuses on adult humans; the authors' emotion model might be strengthened by considering research on infants and animals, highlighting the importance of the limbic system. Reliance on the James–Lange theory is questionable; emotions typically occur instantaneously, with dubious dependence on bodily feedback for affect. Stronger evidence for localization might be obtained using more precise emotion terms and alterative localization methods.

In our response, we clarify important theoretical differences between basic emotion and psychological construction approaches. We evaluate the empirical status of the basic emotion approach, addressing whether it requires brain localization, whether localization can be observed with better analytic tools, and whether evidence for basic emotions exists in other types of measures. We then revisit the issue of whether the key hypotheses of psychological construction are supported by our meta-analytic findings. We close by elaborating on commentator suggestions for future research.