Extended emotion

Abstract

Recent thinking within philosophy of mind about the ways cognition can extend (e.g., Clark, 2008; Clark & Chalmers, 1998; Menary, 2006; Wilson, 2000, 2004) has yet to be integrated with philosophical theories of emotion, which give cognition a central role. We carve out new ground at the intersection of these areas and, in doing so, defend what we call the extended emotion thesis: the claim that some emotions can extend beyond skin and skull to parts of the external world.

Keywords:

• Active Externalism
• Emotion
• Extended Cognition

View correction statement:

Erratum

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

^[1] We emphasize that a clear distinction must be drawn between two senses of ‘collective emotion’, perhaps by retaining the term ‘distributed emotions’ for the claim that emotions can be constituted in a distributed fashion, while applying the term ‘shared emotion’ to emotions that are merely common between the members of a group. Moreover, with respect to the idea of ‘collective emotions’, a referee correctly points out that the term may, as Scheve and Salmela (2014) use it, denote both emotions that are merely “common” between the members of a group and emotions that are constituted by all the (or at least several) members of a group at the same time.

^[2] We will not here discuss the objections facing active externalism, which we use as a premise to our arguments. In brief, however, many of the existing worries point either to the dissimilarity between the inner cognitive processes and the external elements that are supposed to be parts of one's cognitive system (e.g., Adams & Aizawa, 2008; Rupert, 2004), or to the perceptive rather than introspective manipulation of those external elements. Others deny the mark of the cognitive to the alleged extended cognitive processes (e.g., Adams & Aizawa, 2008), or claim that there cannot be a science of active externalism (e.g., Adams & Aizawa, 2008; Rupert, 2004). For a short discussion and reply to most of these objections, see Menary (2006).

^[3] See Bach-y-Rita and Kercel (2003) for a recent review on TVSS.

^[4] Two potential worries might be raised with respect to our distinction between the extended mind thesis and the extended cognition hypothesis. First, as a referee for this journal suggests, one might question the precedent for such a distinction in the literature on the basis that critics of active externalism (e.g., Adams, Aizawa, and Rupert) seem to not distinguish between the two views, by being equally hostile to both. Second, the referee also claims that we should not equate the distinction between cognition and mind with the distinction between processes and states, “as if there are no mental processes and no cognitive states.” With respect to the first point, we would like to insist that, indeed, there is such a clear distinction available in the literature and that a failure to address such a distinction on the part of the critics of active externalism can only undermine their objections to this general approach to the nature of mind rather than set the dialectics or determine how we should understand active externalism. Indicatively, when Clark and Chalmers introduce the extended mind thesis in their original (1998) paper, they write: “So far we have spoken about “cognitive processing,” and argued for its extension into the environment. Some might think that the conclusion has been bought too cheaply. Perhaps some processing takes place in the environment, but what of the mind? Everything we have said so far is compatible with the view that truly mental states—experiences, beliefs, desires, emotions, and so on—are all determined by states of the brain. Perhaps what is truly mental is internal, after all? We propose to take things a step further” (p. 12). Moreover, for a distinction between the extended mind and the extended cognition hypothesis that is very similar to the one we advocate above, see Shani (2013)—though note that Shani refers to the extended cognition hypothesis as “process externalism.” With respect to the referee's second point we would like to clarify that we take the adjectives ‘cognitive’ and ‘mental’ to be identical such that ‘cognitive processes’ and ‘mental processes’ refer to the same things and so do ‘cognitive states’ and ‘mental states’. Nevertheless, as Clark and Chalmers' (1998) quote above indicates, and in line with our distinction in the main text, the extended cognition hypothesis is normally associated with mental/cognitive processes whereas the extended mind thesis is supposed to concern cognitive/mental states. Thanks to an anonymous referee for this suggestion.

^[5] The term ‘commonsense functionalism’ comes from Braddon-Mitchell and Jackson (2007). Thanks to an anonymous referee for drawing this to our attention.

^[6] This does not mean that the hypotheses of extended and distributed cognition are incompatible with common-sense functionalism, or that they are anti-functionalist on the whole. In so far as a cognitive process is a function, these two hypotheses are compatible with functionalism. For more details see Palermos (2014b).

^[7] While this paper was first published in 2010, it has been available online since 2006. The “glue and trust” criteria, however, had already made their appearance in Clark and Chalmers (1998), although the phrasing was somewhat different. Also note that, in Clark and Chalmers (1998), the authors consider a further criterion: “Fourth, the information in the notebook has been consciously endorsed at some point in the past, and indeed is there as a consequence of this endorsement” (p. 17). As the authors further note, however, “the status of the fourth feature as a criterion for belief is arguable (perhaps one can acquire beliefs through subliminal perception, or through memory tampering?),” so they subsequently drop it.

^[8] Farkas (2012), for example, gives the additional example of Lotte who has downloaded 37 volumes of the history of Europe with a quick search function from a source she completely trusts, and Lynch (2014) worries that, by the extended mind logic, much of the information online should count as our dispositional beliefs.

^[9] For a sample of some recent work at the intersection of extended cognition and mainstream epistemology, see Carter (2013); Carter and Kallestrup (forthcoming); Carter, Kallestrup, Palermos, and Pritchard (2014); Carter and Palermos (forthcoming); Carter and Pritchard (forthcoming); Giere (2012); Kelp (2013); Kirchhoff and Newsome (2012); Menary (2012); Palermos (2013, 2014a); Palermos and Pritchard (2013); and Pritchard (2010).

^[10] There are a few exceptions. Stephan, Walter, & Wilutzky (2014), though they focus on embedded emotions, consider the possibility of extended emotions in section 6 of their paper. See in particular their case of Arnold, an autistic individual who is equipped with a headset camera connected to a computer supplying him with real-time online information about the emotional states of his interaction partners (Stephan, Walter, & Wilutzky, 2014, p. 74). While the information fed into the program causally affects Arnold's responses, his responses do not similarly play a role in shaping or informing the information supplied to him by the external device. Contrast this with the notebook cases (Clark & Chalmers' case of Otto, as well as INSULT^* and GLOSSOPHOBIA^*), where the causal interaction is reciprocal. See also Greenwood (2013, p. 420) for an argument that specifically human emotional ontogenesis is a world-to-brain transcranial achievement. As with the case of Arnold, Greenwood's claims about emotional ontogenesis fall short of the stronger claim we motivate here, which is that emotions themselves will (either on EMT or HEC) supervene on parts of the world. See here, also, Krueger (2014), who nicely taxonomizes a range of ways emotions might be extended, arguing for the view that “music … becomes part of the extended vehicle needed to bring about certain emotions” (p. 6). However, Krueger's argument seems to establish just that music would, under certain circumstances, be an enabling condition for certain emotions. Constrast this with the more radical claim that emotions themselves can supervene on music. While Krueger considers how emotions might be extended in the stronger sense we defend (what Krueger calls the “hypothesis of individually extended emotions”), Krueger does not go so far as to defend the view. Slaby (2014) by contrast does defend such a position, though only in the context where it is EMT (rather than HEC) that must be assumed. Given that HEC is easier to motivate than EMT, our proposal demonstrates a simpler way to the conclusion, and further, unlike Slaby's proposal, does not rely on a further need to motivate “phenomenal coupling.” Likewise, Colombetti and Roberts (2015), while they argue that one who endorses EMT should accept more generally extended affective states, motivate their argument (as Krueger does) via EMT, rather than HEC.

^[11] See James (1890). James's account of emotion also features in the well-known James-Lange theory of emotion.

^[12] See Hutchinson (2009, p. 61).

^[13] For example, one of the most common objections is that the view implausibly explains the rationality of emotions with reference to the rationality of standard propositional attitudes (i.e. beliefs or desires) in terms of which cognitivists often advert when characterizing the judgments that feature in emotions (e.g., Elster, 2003; Goldie, 2000; Lyons, 1980). Other dissenters focus on so-called “recalcitrant emotions” (e.g., Brady, 2009; D'Arms & Jacobson, 2003). For example, cognitivists struggle with the fact that it seems possible to judge that it is safer to fly than it is to drive and nonetheless find oneself paralyzed by a fear of flying, and that it is possible to believe that someone has bad intentions without having a corresponding emotional response. Some defenders of cognitivism try to respond by arguing that the judgments featuring in emotion are more like that of perception than that of beliefs; others draw on the potential similarity between perceptions and emotions to support the idea that the content of either might not be limited to propositional content. See for example Charland (2002), Goldie (2000), Peacocke (2001), and Tappolet (2003).

^[14] It is worth noting here that the often-cited recalcitrant emotion examples (if they hit the mark) simply show that not all emotions can be assimilated to propositional attitudes—as a result, some version of cognitivism may still be used as one of the building blocks for an extended emotions thesis on which some emotions extend beyond the skull and skin.

^[15] We could also envision an even weaker version than the one we explore here, according to which cognitions are necessary antecedents of emotions. The extended emotion thesis, however, cannot be motivated by a view that is this weak, but rather, only by views on which the cognitions are either exhaustive of, or an essential ingredient of, emotion. Thanks to a referee for raising this point.

^[16] See Scarantino (2010) for a helpful diagram of such examples.

^[17] Some other cognitivists do branch out from a general claim about emotions as judgments and suggest that emotions should be seen as affect-laden judgments (e.g., Broad, 1971; Lyons, 1980).

^[18] See Kenny (1963, pp. 171–186) for the locus classicus of this discussion.

^[19] Thanks to an anonymous referee for this journal for requesting clarification on this point.

^[20] As Sander, Grandjean, and Scherer remark, appraisal “requires the interaction between many cognitive functions and their underlying neural circuits in the process of comparing the features of stimulus events to stored schemata, representations in memory and self-concept, and expectations and motivational urges of high priority” (2005, p. 56). See also Leventhal and Scherer (1987), as well as Scherer's (2001) component process model of appraisal mechanisms in emotion. For a dynamic systems modeling approach to cognitive mechanisms of emotion, see Lewis (2005).

^[21] See Scarantino (2010, p. 744) for a helpful discussion of this point.

^[22] For dynamic approaches to the kind of appraisals that feature in emotion, see Roseman and Smith (2001) and Scherer (2001, 2009).

^[23] See, for instance, de Sousa (2013). The taxonomy between what we are calling “strong” and “weak” cognitivist views is aimed only at illuminating how cognitivists positions could be (on HEC) committed to viewing some emotions as extended. We recognize that this distinction glosses over many other differences between cognitivists views, though these distinctions won't be relevant to our argument.

^[24] See de Sousa (2013, section 4).

^[25] This is the point de Sousa (2013, section 4) makes.

^[26] We might envision, for the weak cognitivists, further explanations for how emotions might be conceived of as extended. For instance, it would be open to a weak cognitivist to hold that emotional responses include actions, or at least tendencies to action, and could thus be conceived of as extended in this further sense. We agree that such options would be available for the weak cognitivist (who is not committed to understanding emotions exclusively, but only partially, in terms of relevant cognitions), though it would be incumbent on such proposals to provide argument for why these non-cognitive elements count as metaphysically extended in the sense that cognitive processes are claimed to extend by proponents of HEC. Thanks to an anonymous referee for this journal for noting this possibility.

^[27] See Ellsworth and Scherer (2003).

^[28] See de Sousa (2013) as well as Eysenck and Keane (2005, chapter 18) for helpful overviews.

^[29] An important contrast here is with dimensional theories of emotion, according to which emotions can be distinguished vis-à-vis abstract sensation, or qualia.

^[30] Consider, for instance, Scherer's suggestion that: “One can use the analogy of a radar antenna to refer to the fact that organisms constantly scan their environment (and their internal state) to detect and re-evaluate changes. … Consequently, we expect events or internal changes to trigger cycles of appraisal running through the evaluation checks proposed here until the monitoring subsystem signals termination of or adjustment to the stimulation which triggered the appraisal episode” (1987, p. 10).

^[31] For an explicit denial of the occurrence condition, see Posner (1992), who identifies emotions in part with reference to non-conscious, subcortical states.

^[32] For support of the view that evaluation affects the duration of emotion, see Verduyn, van Mechelen, and Tuerlinckx (2011).

^[33] See also Prinz (2004a) for a defense of the view that such evaluations can be embodied.

^[34] See Feldman, Cohen, Hamrick, and Lepore (2004) for a study showing that one's anticipation of public speaking correlates with increased cardiovascular response.

^[35] Implicit memory is a form of “non-declarative” memory. See here Sutton et al. (2008).

^[36] While this case relies on a psychological abnormality to generate the HEC, it's not the case that (for instance) only in rare cases of amnesia like Brian's in GLOSSOPHOBIA^* would it be possible to generate the extended emotion thesis on psychological appraisal accounts. Compare here with the parallel point, in section 2.2, with respect to the case of INSULT^*.

^[37] Of course, this line might be submitted with respect to cognitivism no less, and the reply would be that the extended emotion thesis follows from (super) weak cognitivism. We raise the line of thought with respect to appraisal theories because we are envisioning, in descending order of strength, the kinds of positions that would suffice to generate the extended emotion thesis, and the superweak PAT would be the weakest.

^[38] Note that our contribution has been to show, on strictly philosophical grounds, that the extended emotion thesis is substantially easier to reach and that, perhaps, it is a view that is, in effect, more plausible than initially expected. A separate and very worthy issue, in light of our conclusion, however, as noted by a referee for this journal, is this: What do some additional extended emotions look like, and what kinds of real-life shapes might they take? While this is admittedly an important applied question, elaborating on it falls outside the scope of the paper's philosophical objective. Nevertheless, in footnote 39, we briefly consider some possible real life examples that would qualify as extended emotions by, at least, the paring of HEC and weak PAT.

^[39] In relation to footnote 38, imagine the case of Vivi who suffers from supraventricular tachycardia and has been advised by her doctor to avoid exceeding a heart rate of 200 beats/minute. To cope with her problem, Vivi recently bought an Apple Watch that can monitor and display her heart rate. This device allows her to continuously keep an eye on her heart rate every time she goes running. Provided her heart rate is well under the suggested limit, she judges the situation to be fun and exciting. Meanwhile, if the watch tells her that she is getting close to 190 beats/minute, then she judges the situation to be precarious and must therefore regulate her speed, such that on the basis of the feedback she receives from the watch she can judge the situation to be a safe and enjoyable one. Once the screen shows 190 beats/minute, however, then she judges the situation to be dangerous until the feedback she receives from her watch is once again a safe one. This seems to be a bona fide case in which the external artefact plays a constitutive role in the agent's emotional appraisal of the situation as exciting or frightening. More formally, the agent interacts mutually with her watch, which means that, according to dynamical systems theory, the watch plays a constitutive part in the cognitions involved in her emotional appraisal of the situation, thereby resulting in what may be plausibly construed (even on weak PAT) as a case of extended emotions.

Additional information

Notes on contributors

J. Adam Carter

J. Adam Carter is Research Fellow, AHRC Extended Knowledge Project, Eidyn Research Centre, University of Edinburgh.

Emma C. Gordon

Emma C. Gordon is an independent researcher.

S. Orestis Palermos

S. Orestis Palermos is Research Fellow, AHRC Extended Knowledge Project, Eidyn Research Centre, University of Edinburgh.