Enactivism has been gaining in popularity in a variety of disciplines over the last couple of decades (Di Paolo & Thompson Reference Di Paolo, Thompson and Shapiro2014). Its discourse is centered on a cluster of concepts – autonomy, sense-making, emergence, embodiment, and experience (Di Paolo et al. Reference Di Paolo, Rohde, De Jaegher, Stewart, Gapenne and Paolo2010). Briefly, mind is defined as meaningful activity in the world; it is an autonomous process of sense-making embodied in self-sustaining neural, bodily, and sensorimotor dynamics under precarious far-from-equilibrium conditions. Speculations about internal mental representations are rejected in favor of investigations of the complex dynamics that structure a subject's lived (phenomenological) and living (biological, i.e., both neural and bodily) existence (Thompson Reference Thompson2007).

Nevertheless, a specifically enactive neuroscience is still outstanding. To be sure, some of its key concepts can be traced to the neuroscientific research done by Maturana and Varela in the 1960s and 70s. And after the first comprehensive statement of the enactive approach (Varela et al. Reference Varela, Thompson and Rosch1991), Varela continued to promote a systems approach to neuroscience (e.g., Varela et al. Reference Varela, Lachaux, Rodriguez and Martinerie2001). Moreover, Varela's proposal of integrating this approach with a disciplined study of first-person experience turned into one of its core methods: neurophenomenology (Gallagher Reference Gallagher, Bayne, Cleeremans and Wilken2009). Such methodological recommendations are valuable, but they are not an enactive theory of the brain.

This striking omission has not gone unnoticed. For example, proponents of the predictive processing framework, which aspires to a unified theory of the brain (Friston Reference Friston2010), are taking advantage of this lacuna by attempting to subsume enactivism under its way of thinking (e.g., Clark Reference Clark2012; Seth Reference Seth2014). Although this takeover move is resisted by some enactivists (Froese & Ikegami Reference Froese and Ikegami2013; Roesch et al. Reference Roesch, Nasuto and Bishop2012), as yet there is no clear alternative proposal.

I suggest that Pessoa's (Reference Pessoa2013) integrative approach provides much needed steps in the right direction. He avoids two undesirable extremes that typically characterize other approaches. He rejects the tendency, prevalent in neuroimaging research, to reduce aspects of mind to specific brain regions (pp. 194–95). And neither does he reduce the workings of the brain to one general abstract principle (e.g., a free-energy principle). Instead, Pessoa emphasizes that there is a diversity of neural structures and processes at play; that structure and function are mutually dissociable; and that mental processes are realizable in multiple ways by distributed networks and unavoidably context dependent. This middle way between the extremes of anatomical specificity and computational generality matches with enactivism's insistence on a systems theoretic approach that nevertheless remains grounded in the concreteness of the living organism. And Pessoa is in favor of developing stronger links between neuroscientific data and subjective experience (p. 116). I am also intrigued by several more specific convergences between his approach and the enactive approach.

Pessoa interprets the findings of cognitive neuroscience as demonstrating that perception, cognition, emotion, and motivation do not refer to essentially distinct categories. Instead, they are overlapping aspects of a complex network of mental processes in which one aspect can temporarily become more expressed than the others depending on the circumstances, but in which none can be completely isolated from the others. The enactive approach has come to the same conclusion on the basis of theoretical considerations stemming from the philosophy of the organism and phenomenology of the body, which is why the notion of sense-making has become one of its core concepts (Thompson Reference Thompson2007). Briefly, sense-making refers to the process by which an organism enacts a meaningful point of view on the world, which always contains a mixture of sensation, pre-reflective interpretation, and valuation. Accordingly, every organism is situated in what von Uexküll (Reference Von Uexküll and Schiller1934/1957) once called an Umwelt: a concrete context of significance defined by affordances for interaction, which are shaped by the environment but also by each organism's embodiment, knowledge, and motivations (goals, needs, desires). This supports Pessoa's contention that “vision is never pure, but only affective, vision. A similar point can be made for other sensory modalities” (p. 257).

As Colombetti (Reference Colombetti2014) has argued extensively, an important implication of the enactive theory of sense-making is that affectivity is an essential aspect of mind, and that this primordial affectivity provides the foundation for the origins of more specific emotions, moods, and values. In the words of Varela and Depraz (Reference Varela and Depraz2005, p. 61): “emotions cannot be seen as a mere ‘coloration’ of the cognitive agent, understood as a formal or un-affected self, but are immanent and inextricable from every mental act.” Enactivism therefore provides a suitable interpretative framework – for example, for explaining the finding that emotional networks are among the most widely connected in the brain (Pessoa Reference Pessoa2013, p. 229). In addition, the substantial connections from brain regions implicated in cognitive and emotional processing to even the earliest sensory regions, such as projections from the amygdala and the pulvinar to visual cortex, are to be expected if we primarily live in a meaningful world, and not in some abstract model consisting of meaningless physical facts, as assumed by classical cognitive science. Enactive theory may therefore help us to better understand the role of “nonstandard” pathways to visual perception (Pessoa Reference Pessoa2013, pp. 244–46).

This convergence also provides an opportunity for neuroscience to help resolve a puzzle that has emerged for enactive theory. In short, if sense-making is the default mode of being in the world, such that an organism's point of view is saturated with affordances that are meaningful in terms of its potential actions, then how do we explain the emergence of the detached observer's stance and dispassionate reflection, that is, precisely the hallmarks of higher-level cognition (Cappuccio & Froese Reference Cappuccio, Froese, Cappuccio and Froese2014)? The problem is no longer just to understand how perception, cognition, and emotion are integrated, but likewise under what conditions they can become temporarily separated. This task becomes especially pressing when it comes to explaining the unusual requirements of sense-making in the context of symbolic culture.

For example, when we see a representation of a pipe, as in Magritte's famous painting, what we perceive is a presentation of an imagined pipe that is in fact objectively absent (Seth Reference Seth2014). But from the point of view of most animals (as long as they have not been enculturated), such a representation is experienced differently: either the presented is taken as objectively present (i.e., a pipe is seen, as a result of a failure of veridical perception), or the presented is not even recognized as such. Phenomenologically, to experience a representation as a representation, one has to detach from one's immediate preoccupations with the world so as to bring the underlying material substrate of the potential representation to attention, while neutralizing any preexisting affordances of that medium, in order to then imbue it with renewed meaning in accordance with cultural norms. The revolutionary human abilities based on symbolic cognition, particularly language and writing, are only achievable with this tight yet flexible integration of perceptual, cognitive, and emotional processes. Pessoa (Reference Pessoa2013) did not systematically review species-specific differences of neural integration, but we can hypothesize that in humans these processes are more flexibly integrated. Findings worthy of a closer look include the malleable relationship between prefrontal cortex and the amygdala (p. 115) and the relative isolation of prefrontal cortex from sensory stimuli (pp. 233–34; 255–56).

Pessoa's work also nicely complements the enactive approach to social cognition, which includes the phenomenological claim that we normally directly perceive aspects of others' minds in their bodily expressions; for example, happiness in another's smile. This “direct perception hypothesis” (Froese & Leavens Reference Froese and Leavens2014) stands in contrast to standard theory of mind accounts, which start from the assumption that others' subjective states are perceptually inaccessible and must therefore be inferred (or simulated) based on perception of purely objective surface behavior. Unless we are referring to the abnormal social experience of certain people with schizophrenia and autism spectrum disorder, standard approaches are not phenomenologically convincing (Froese et al. Reference Froese, Stanghellini and Bertelli2013). But it is easy to see how they theoretically made sense when combined with the premise that perception, cognition, and emotion are distinct modules: perception could only ever deliver neutral physical facts, such that cognition then has to figure out what is going on, after which it can evoke an emotional response. However, if perception, cognition, and emotion are tightly integrated, as Pessoa asserts, then there is no longer any reason to assume that perception cannot directly present us with others' states of minds in their meaningful bodily expressions. Conversely, the direct perception hypothesis helps us to better understand the role of cognitive-emotional feedback to early sensory regions: it is required to constitute such a meaningful perceptual world.

Pessoa's proposal also productively relates to another controversial theory of the enactive approach – namely that bodily and environmental (including social) activity can form a genuine part of mental processes (De Jaegher et al. Reference De Jaegher, Di Paolo and Gallagher2010; Thompson Reference Thompson2007). On this view, the brain is conceived of as a “mediating organ” (Fuchs Reference Fuchs2011) of embodied activity in the world rather than as its sole basis (Gallagher et al. Reference Gallagher, Hutto, Slaby and Cole2013). Returning to an issue raised earlier, one criticism of recent attempts to formulate a predictive processing account of enactive perception is that the former is committed to the classical cognitivist assumption of brain-centered internalism (Di Paolo Reference Di Paolo2014; Froese Reference Froese2014), which in the social domain equates with individualism. Admittedly, it is difficult to investigate a distributed view of the mind with the current tools of neuroscience, and even neurophenomenology has not always managed to avoid internalism (Beaton Reference Beaton2013). However, a more embodied and socially situated neurophenomenology seems possible (Desmidt et al. Reference Desmidt, Lemoine, Belzung and Depraz2014; Froese et al. Reference Froese, Iizuka and Ikegami2014).

Pessoa does not take an explicit stand with regard to the limits of mind, but I suspect that he might be sympathetic to this highly contextualized enactive view (see, e.g., p. 202). Crucially, he allows that functional connectivity can play a role in neural processes even in the absence of direct anatomical connectivity (pp. 148–49). It is only a small additional step to extend this concession to include some kinds of extra-neural processes as another type of functional connectivity. For example, there does not seem to be any principled reason for excluding bodily and environmental dynamics from neuroscientific explanations, because externally mediated activity can functionally connect effectors and sensors into a coherent sensorimotor system (Fig. 1).

Figure 1.

Two indirect relationships between function and anatomical structure. According to Pessoa (Reference Pessoa2013, pp. 207–12), a functional relationship between two regions of the brain (R1 and R3) does not necessarily have to be supported by direct structural connectivity (solid arrows), because their functional connectivity can be mediated via structural connectivity through another region (R2). But other forms of mediation are conceivable, including dynamical routes via extra-neural context C1 (dashed arrows). Most straightforwardly, we can think of region R3 as the motor system, region R1 as the sensor system, and context C1 as the body situated in an environment. R1 and R3 are then also functionally connected by sensorimotor interactions.

Relatedly, it has been debated whether the direct perception hypothesis should be cashed out solely in terms of individuals, or whether social interaction can play a constitutive role as well. For example, when we share an intimate moment with a loved one, may we actually be sharing one experience co-constituted by networks distributed across two brains? Several authors have argued that the underlying dynamics of emotional episodes can span two or more individuals (Colombetti Reference Colombetti2014, pp. 66–70), and Pessoa's approach is consistent with this possibility. According to the enactive approach, not every kind of interaction is sufficient. Individuals must co-regulate their activities such that the conditions of success of their actions cannot be reduced to one individual (Froese et al. Reference Froese, Iizuka and Ikegami2014), and the effects are further enhanced by emotional engagement (Schilbach et al. Reference Schilbach, Timmermans, Reddy, Costall, Bente, Schlicht and Vogeley2013). This is reminiscent of Pessoa's thinking about degrees of isolability and decomposability in neural networks (pp. 196–97); for example, that strong integration requires coupled systems that are strongly interacting. Moreover, measures of functional connectivity are considerably strengthened in affective versus neutral contexts (pp. 209–11). Thus, unless there is a principled reason for rejecting the application of Pessoa's approach to inter-brain integration, we can genuinely participate in each other's sense-making.