This concluding chapter summarizes the main insights gained in this book. Emotion theories show substantial agreement in their working definition of emotions in terms of the typical and apparent properties of emotions (intensional format) and the prototypical emotions (divisio format) they want to see explained. Their constitutive explanations range from narrow to broad. The most pronounced differences reside in the causal-mechanistic explanations they propose. Five fault lines stand out centering on the questions whether (a) emotions are caused by a stimulus-driven or goal-directed process, (b) the origin of feelings is central or peripheral, (c) emotions are caused by special-purpose or general-purpose mechanisms, (d) emotions do or do not form a scientific set, and (e) the variety in the set of emotions is best captured with discrete emotions or dimensions. The chapter also examines whether there is any hope for a unifying theory integrating several theories or a general framework for conducting emotion research. Starting from a unifying theory that combines a maximum of mechanisms, a pruning stage is required based on empirical research. A speculation about the result of this pruning exercise, inspired by the goal-directed theory advocated in this book, is presented, ultimately leading to the demystification of emotion.

This chapter applies the demarcation-explanation cycle to emotion theories. In Stage 1, theories propose working definitions of emotion listing (a) properties of emotions (intensional format) and/or (b) prototypical emotions (divisio format). Typical and apparent properties of emotions are that they have ontogenetic and phylogenetic continuity, are caused by important events, have mental aspects (Intentionality, phenomenality) and bodily aspects (somatic responses, subtle and coarse motor responses), and are characterized by heat (valence, intensity), automaticity, control precedence, and irrationality. Prototypical emotions are fear, anger, sadness, joy, and so on. In Stage 2, theories offer constitutive and/or causal-mechanistic explanations for emotions. In Stage 3, the explanations are validated in empirical research. In Stage 4, they form the basis for the eventual scientific definitions of emotions that allows (a) demarcating the set of emotions from other sets (intensional format) and (b) partitioning the set of emotions into subsets (divisio format). The demarcation-explanation cycle forms the basis for a novel typology of emotion theories in which the theories are placed on several axes. The axes correspond to questions that arise at various stages in the demarcation-explanation cycle and the positions of theories on these axes correspond to the answers they formulate to these questions.

This chapter discussed the general precursors of the families of emotion theories to be discussed in the ensuing chapters: Darwin (1872) and James (1890b). Darwin (1872) proposed an evolutionary and eventually also a mechanistic explanation for emotional behavior, particular facial expressions. Initially goal-directed processes grow into habits (learned emotion-response links) and instincts (innate emotion-response links). Today these instincts are no longer of use, which is why they appear in weakened form as facial expressions. James (1890b), seeking a causal-mechanistic explanation for emotions themselves, proposed that stimuli activate instincts (innate stimulus-response links) and habits (learned stimulus-response links), which when enacted produce somatosensory feedback, felt as emotions. James’s (1872) theory accounted for the properties of ontogenetic and phylogenetic continuity, phenomenality, bodily aspect, heat, control precedence, and irrationality, but has difficulty delivering the right kind of (world-directed) Intentionality, and even the right kind of heat. James (1872) rejects partitioning the set of emotions into discrete subsets. Empirical evidence pro and contra James’s (1872) theory is discussed.

This chapter discusses psychological constructionist theories (PCTs). PCTs reject the existence of affect programs as special-purpose mechanisms for the phenomena called emotions. PCTs come in a two-factor version, endorsed by Schachter (1964) and Barrett (2017b), and a multi-factor version, endorsed by Russell (2009). Two-factor PCTs propose that emotions result from the combination of (a) diffuse bodily feelings and (b) a construction process that binds these felt bodily responses to external stimuli and produces a labeled feeling. The multi-factor PCT proposes that the phenomena that people call emotions are composed of many components. The categorization of these components does not result in an emotion per se, but an emotional self-ascription. Two-factor PCTs can deliver discrete emotions. The multi-factor theory does not deliver discrete emotions, but can nevertheless make sense of them. Empirical research that tests two-factor PCTs is discussed.

This chapter discusses stimulus evaluation theories (SETs), which foreground a process of stimulus evaluation (or appraisal). A first brand, called evaluation-first SETs, include appraisal theories in psychology and (quasi-)judgmental and perceptual theories in philosophy. Theories in this brand differ in the role they confer to stimulus evaluation (constituent-only, causal-only, constituent-causal), the granularity of the output of this process (molar, molecular, molar-molecular), and the system required to produce it (in terms of representational format and Attitude, operations, and operating conditions or automaticity). The transition from stimulus evaluation to the other components is accounted for by evaluation-response links that are innate—in the biological version—or learned—in the non-biological version. The latter version splits again in a summary and elemental sub-version. A second brand, called embodied SETs, postulate that stimulus evaluation acquires its heat after bodily responses set in, or they assume an indivisible evaluation-response connection. SETs outperform the previous theories in their capacity to account for the (world-directed) Intentionality of emotions. Most of them can also account for ontogenetic and phylogenetic continuity (except judgmental theories), phenomenality, bodily aspect, heat, control precedence, and irrationality. Molar SETs deliver discrete emotions, purely molecular SETs do not. Empirical research that tests SETs is discussed.

This chapter discusses evolutionary theories in psychology (Ekman & Codaro, 2011; Levenson, 2011; Panksepp, 2012; Keltner, Tracy, et al., 2019), also known as motivational theories in philosophy. Building further on the bridging work of McDougall (1908), these theories propose the following causal-mechanistic explanation of emotion. Stimuli activate affect programs (another word for instincts specific to emotions), which in turn give rise to somatic and (subtle and coarse) motor responses and feelings. Some evolutionary theories are hybrids with stimulus evaluation theories in that they postulate that the affect programs is preceded by a stimulus evaluation process. The pure emotions caused by affect programs get mixed with learning and computation, which create flexibility on the input side (by elaborating the number of stimuli that can trigger affect programs) and on the output side (via emotion regulation and planning). Evolutionary theories account for the properties of ontogenetic and phylogenetic continuity, phenomenality, bodily aspect, heat, control precedence, and irrationality, but the non-hybrid version has difficulty accounting for the right kind of (world-directed) Intentionality. Evolutionary theories partition the set of emotions into discrete emotions. Empirical evidence pro and contra the existence of affect programs is discussed.

This chapter presents an idealized path towards theory development (for all kinds of theories), called the “demarcation-explanation” cycle. The cycle comprised four stages: (a) the provisional demarcation of the explanandum in a working definition, (b) the proposal of different types of explanations, (c) the validation of explanations in empirical research, (d) and the proposal of a scientific definition flowing from these explanations. The chapter also takes a closer look at (a) different types of definitions (working vs. scientific, intensional vs. extensional/diviso) and ways to evaluate the adequacy of scientific definitions, (b) different types of explanations (constitutive, causal, mechanistic, teleological) and levels of analysis, and (c) ingredients of mechanistic explanations such as representations, operations, operating conditions (related to automaticity), and related notions of dual-process/system models, rationality, and cognition.