Lindquist et al. mostly use their impressive meta-analysis to contrast their constructionist approach with all other emotion theories, which are credited with a “locationist” account (“discrete emotion categories […] are consistently and specifically localized to architecturally defined brain locales or anatomical networks”; see “Definitions” box in section 3). This is a gross misrepresentation of the architectures of classic emotion theories, many of which suggest that emotion episodes are driven by complex interactions of multiple mental and somatic systems (cognitive, motivational, autonomic, motor) that are not specialized for emotion. This precludes the assumption of emotion-specific brain centers. In consequence, the authors' fundamental premise is untenable, the results of their meta-analysis being consistent with many different emotion theories (Moors Reference Moors2009).

This is particularly true for componential appraisal theories (see Ellsworth & Scherer Reference Ellsworth, Scherer, Davidson, Scherer and Goldsmith2003; Roseman Reference Roseman2011; Roseman & Smith Reference Roseman, Smith, Scherer, Schorr and Johnstone2001), which Lindquist et al. credit with a locationist account (see sect. 2, para. 1). For example, the component process model (CPM) of emotion (Scherer Reference Scherer, Scherer, Schorr and Johnstone1984; Reference Scherer1986; Reference Scherer, Scherer, Schorr and Johnstone2001; Reference Scherer2009b) postulates that the dynamic unfolding of emotion episodes is driven by parallel appraisal processes at different levels of brain organization that sequentially evaluate objects and events for their novelty, intrinsic pleasantness, and goal conduciveness; and for compatibility with self-concept, social norms or values, and the individual's coping potential. These appraisals are content criteria, not specialized mechanisms for emotion, based on the general information-processing machinery of the brain (perception, association, schema matching, prediction error checking, categorization, inference). For example, contrary to Lindquist et al.'s claim (sect. 3, para. 5), novelty checking is not a new, unique, emotion-specific mental operation–there are many studies on novelty processing in cognitive (see Ranganath & Rainer Reference Ranganath and Rainer2003) and affective neuroscience (Balderston et al. Reference Balderston, Schultz and Helmstetter2011).

The appraisal results cumulatively produce changes in several subsystems of the organism, including the motivational system (action tendencies, e.g., approach/avoidance), support system (endocrine, autonomic, and somatic nervous systems), and monitoring system (subjective feeling). These bodily response patterns are fed back recursively into the appraisal process through interoception. This appraisal-driven synchronization of subsystems that are not specialized for emotion, but have specific functions for the organism's different needs, constitutes the emotion episode. The monitoring system integrates proprioceptive information about changes in different systems, producing unique mental representations (qualia). If the subsystem synchronization level is sufficiently high, the emotion process may enter consciousness and engender categorization (which can be nonlinguistic) and–if the context requires communication of the emotion to others–labeling with emotion words or metaphors (Grandjean et al. Reference Grandjean, Sander and Scherer2008; Scherer Reference Scherer2009b). This architecture is “constructionist” in the dictionary sense of the word.

Contrary to Lindquist et al.'s claim (sect. 2, para. 2) that an appraisal approach hypothesizes that “dedicated cognitive mechanisms automatically make meaning of a stimulus and trigger the corresponding discrete emotion,” the CPM includes all levels of processing (from sensory motor to propositional) and does not assume dedicated mechanisms or a limited number of automatically triggered discrete emotions. On the contrary, the CPM allows for an almost infinite number of different emotion episodes, given the recursive, emergent character of the sequential-cumulative configuration of appraisal results and consequent patterning of response systems (Scherer Reference Scherer, Scherer, Schorr and Johnstone1984). This architecture also accounts for the prevalence of blended or mixed emotions (e.g., see Scherer & Ceschi Reference Scherer and Ceschi2000; Siemer et al. Reference Siemer, Mauss and Gross2007).

Also contrary to the authors' claim (sect. 2, para. 2), several attempts have been made to muster neuroscience support for appraisal predictions (Sander et al. Reference Sander, Grandjean and Scherer2005; Scherer Reference Scherer1993; Scherer & Peper Reference Scherer, Peper and Gainotti2001). Furthermore, many CPM hypotheses have been experimentally tested using neuroscience methods. For example, Grandjean and Scherer (Reference Grandjean and Scherer2008), using EEG, chronometrically confirmed the predicted sequence of novelty, intrinsic pleasantness, and goal-conduciveness checks. Dan Glauser and Scherer (Reference Dan Glauser and Scherer2008) showed that the conscious experience of feeling an emotion is associated with specific brain wave oscillations (see Scherer [2009b] for a recent overview of empirical evidence for CPM predictions).

The authors' meta-analysis lends further support to the CPM (whose architecture makes it incompatible with locationist assumptions). For example, low-level relevance detection in the amygdala, interoceptive representation in the insula, integration of subsystems in the orbitofrontal cortex (OFC) or networks relative to attention or conceptualization, are entirely consistent with the model's architecture. As most emotion theories assume, implicitly or explicitly, that in many contexts, emotional experiences are categorized, conceptualized, and labeled (see Scherer Reference Scherer2009b, pp. 1318–24), any evidence for brain mechanisms subserving these operations is equally consistent with those theories. Demonstration of the neural underpinnings of categorization or conceptualization does not privilege any theory on why or when categorizations and conceptualizations lead to specific effects.

One wonders whether the meta-analysis supports specific claims of the authors' theory, which assumes that a “conceptual act” is needed to constitute the emotion. This “definitional act” is immune to empirical evidence of any kind. Contrary to established usage, it limits the term to conceptualized subjective experience or feeling, thereby excluding objectively measurable emotion episodes in nonhuman animals, nonverbal children, and normal individuals who cannot or do not conceptualize such episodes as emotion feelings.

Furthermore, the meta-analysis does not reveal stringent evidence for “core affect,” presumed to be a fundamental feeling architecture organized as an orthogonal valence×arousal system. As valence is generally considered a fundamental organizing criterion for affectivity, any evidence for brain mechanisms underlying valence processing is consistent with virtually all theories of emotion. It does not lend further credence to the monolithic “core affect” notion.

The authors admit that core affect might need to be broken down into more basic mechanisms, such as approach- versus avoidance-related states (sect. 6, para. 5). Such action tendencies (Frijda Reference Frijda2007) are postulated by most emotion theories, and the underlying brain mechanisms are among the first to have been examined (Davidson Reference Davidson1993). The authors also state that the amygdala “realizes” core affect (mostly when “the rest of the brain cannot easily predict what sensations mean,” sect. 5.1, para. 2) by detecting relevance (which they call “salience,” covering only one aspect of relevance). This position, together with emphasis on the brain's “making meaning” of exteroceptive and interoceptive information, comes close to a rudimentary componential appraisal theory, although vague references to context-bound categorization and conceptualization do not facilitate development of testable nomothetic predictions.