Author’s Replies

Replies to Hatzimoysis, Hufendiek and Sievers, Majeed, Gerrans, and Whiting

Tom Cochrane

Flinders University, AU

Cochrane, Tom. 2024. “Replies to Hatzimoysis, Sievers and Hufendiek, Majeed, Gerrans, and Whiting.”

 Journal of Philosophy of Emotion 5, no. 2: 52-61. https://doi.org/10.33497/2024.winter.7. 

Abstract: The concerns of each commentary are addressed in turn. I clarify and defend the claims of The Emotional Mind with regards to the plausibility of automatic responses to representational content, the distinction between emotions and bodily feelings, the influence of social contexts upon emotional responses, the complex issue of whether emotions are modular or form natural kinds, the nature of pain asymbolia, and the nature of emotional authenticity.

Keywords: valent representation, emotion, feelings, mental content, natural kinds, modularity, pain asymbolia, social emotion, emotional rationality

Thank you to Hatzimoysis, Hufendiek and Sievers, Majeed, Gerrans, and Whiting for their generous and thoughtful comments on my book. It is now five years since the book was published, and several more since its main ideas were developed. I find it invigorating to be called once again to its defense. In some cases forceful challenges have been raised. In other cases lacunae are identified. But my overriding impression is that the commentators are sympathetic to the project and interested in seeing how it may tackle various problems that we encounter in the philosophy of mind.

I want to emphasise that control theory is a broad framework for understanding the mind that, historically, we are only just beginning to explore. I believe it to be as powerful a framework for the mind as evolutionary theory is for biology. There is no doubt that, like evolutionary theory, there are many difficulties and complexities that need to be worked out, but the basic approach is incredibly fruitful. My book offers one way to develop the framework, based on the limited considerations and evidence that I could muster. Researchers in disciplines like psychology, robotics, and sociology have been busily developing the approach in other ways.[1] Indeed, I am keenly aware that I only deployed a general control theoretic approach to address long-standing issues in the philosophy of mind, while I am still ignorant of much of the specifics of control theory that have been developed by cognitive scientists. I think philosophers in particular would do well to think more about how control theory can help us make progress on the foundational questions that we care about.

[page 52]

HOW IMMEDIATE IS THE LINK BETWEEN 

REPRESENTATION AND RESPONSE?

The most foundational question that I address in the book is how brain activity gets to meaningfully refer to the world around us. My answer is valent representation: a kind of representation that triggers a response aimed at regulating the presence of its target. This negative feedback loop structure achieves a level of sensitivity that is characteristic of perception, while also being mechanistically and evolutionarily plausible. But does the mind really work like that? Hatzimoysis’s commentary most directly focuses on this issue, and I appreciate the clarity with which he presented the problem and the motivation for my solution. The main thrust of Hatzimoysis’s worry is that, to the extent that my concept of valent representation is similar to Millikan’s theory of pushmi-pullyu representations, it has limited applicability, making it unsuitable as a ground for the mind. Both pushmi-pullyu representations and valent representations appeal to immediate connections between descriptive representations and directive responses, yet it seems that in most behaviour, animals “regularly and spontaneously engage in veracity testing, satisfaction testing, and directive denial with descriptive retention (Bauer 2020), none of which would be even conceivable if descriptive and directive content were not primarily given as clearly distinct” (Hatzimoysis 2024, 21; this issue).

I suppose what I find conceivable differs from Hatzimoysis! First, I must note that detection and response are distinct in the valent representational loop. The detection of the target triggers a physically distinct response. It is because of this distinction that I can then allow for elaborations upon the basic valent representational schema in which interferers from other systems inhibit the response. Indeed, the whole point of having a distinct detection node is to allow for interactions with other representational content (including multimodal content) in a way that seems to me lost in enactivist theories of the mind (Cochrane 2018, 28-30). Yet, I suppose that the impulse to respond remains, ready to drive action as soon as inhibition is removed. In this way, delayed approach or avoidance responses can still get their meaning from regulative interactions.

Second, I think Hatzimoysis may be right to argue that some of Millikan’s examples, such as the bee’s waggle dance, are not “double-facing instances of a unitary representation” (Hatzimoysis 2024, 21). The waggle dance example is most helpful for clarifying the idea that meaning is fixed by how other bees respond. The actual psychological mechanics involved in following a waggle dance, or any other communicative signal, are likely to be complex. Indeed, it is for this reason that I don’t consider communicative uses of representation until quite late in my book.

Third, it seems to me that some of our large-scale behaviour shows a reflexive response to representation, such as the blink response and the nociceptive flexion reflex (sudden withdrawal from a source of heat). However, in complex animals like us, I take the primary instances of valent representation to be the homeostatic systems for regulating things like body heat, blood oxygen, or the expulsion of toxins. That is, our foundational valent representations are directed more at the internal milieu than the extra-bodily environment. As far as I’m aware, there are good grounds to treat these systems as following negative feedback principles. 

Note that even our basic homeostatic systems are liable to layers of inhibition and elaboration. Take for instance the system for breathing. Breathing is subject to all sorts of refinements of control, ranging from babies instinctively holding their breath underwater, emotional modulations, talking and singing, all the way up to culturally inherited spiritual practices. Even so, the fact that we can inhibit or modulate our breathing (up to a point) does not at all imply that it isn’t grounded in a basic regulative control loop, or that we don’t possess a reflexive impetus to breathe when suffering oxygen starvation. On the contrary, if the link between detecting low blood-oxygen and inhalation had to be learned or inferred somehow, we could expect it to be far less reliable.

[page 53]

Thus, the control structure of our homeostatic systems is supposed to justify the idea that valent representation lies at the foundations of the mind. But why think that the same control structures are responsible for emotions; states that are more clearly intentional or meaningful in nature? Partly, it is due to the deep continuity we discern between emotional states and other affective states like pain, pleasure, nausea, itchiness, tiredness, and hunger. All of these states have a clear evaluative and survival function that align them with the homeostatic processes. The other big factor is that we frequently observe and introspectively discern the impulsive nature of our emotional responses. Of course, it is true that we inhibit and modulate our emotional responses. But it is no less true that we feel impulses that cry out for release. Even if we hold back our tears, or our fists, we feel our hearts beating faster, the sweat springing to our brows, and the blush of the cheeks that demonstrate that organic nature prefers to prepare and hold back in readiness rather than initiate responses only once the checking is done.

EMOTIONS VERSUS BODILY FEELINGS

Hufendiek and Sievers do a great job of summarising the key ideas of my theory of emotion. Although I’m glad to see that we agree on the major point that bodily responses play a necessary role in fixing emotional content, we disagree on the role of bodily experience in emotion. Recall that I define emotions as valent representations of situated concerns. They are systems for detecting certain kinds of qualities that trigger responses aimed at regulating those qualities. Bodily responses are thus a necessary part of these systems; but I claim that monitoring or conscious feeling of those responses are another layer of control, and are not strictly necessary to emotion even if they typically accompany them.

One of my arguments for why the experience of bodily feelings is not necessary for a conscious emotional state is that there is non-bodily phenomenology, such as the sense of events slowing down or speeding up (Cochrane 2018, 92). Hufendiek and Sievers suggest that this phenomenology could be caused by changes in heart rate or muscle tension, and even “constituted by changes in blood pressure, hormone levels, etc., that we are unable to trace as such, but that still constitute the phenomenological core of emotions” (Hufendiek and Sievers 2024, 26; this issue). It seems to me a bit of an ad hoc defence to propose that this phenomenology is constituted by bodily feelings without our awareness. I think there are also alternate causal explanations for the sense of time slowing down or speeding up, for example by how many cognitive events are occurring per second, or the extent to which we realise that our intentions are in conflict with what is happening, or the way that increased attention enhances perceptual contrasts.[2]

Hufendiek and Sievers later write that “It is hard to see why somebody would want to introduce valent representation as a representation type whose function it is to regulate bodily responses, only then to draw an artificial distinction between the core representation of a situated concern and the representation of bodily responses” (26). Of course, to me, the distinction is not at all artificial. When I first started thinking about emotions, I was persuaded by the Jamesian claim that without bodily feelings there would be no emotion. Yet the distinction between bodily feelings and emotions was forced on me by the consideration that bodily responses can be triggered without our awareness (Cochrane 2018, 90-93), that significant cognitive work is required to appropriately trigger those bodily responses (70-71), that particular objects are part of emotional meaning yet not captured by bodily feelings (93), and that we are perfectly capable of attending separately to our bodily responses and the situation that these responses are directed at (7-8). 

Not only this, but making a distinction between the initial emotion and the monitoring of bodily responses was extremely fruitful in helping to grasp the additional function of coping potential (§4.7), the nature of moods (§4.3), refinements in emotion types (110), the distinct sense of confidence (111), and our capacities for empathy

[page 54]

(ch. 5). I also consider it a genuine reconciliation between cognitivist and non-cognitivist positions on emotional experience that has persisted for more than a century! When a debate persists for so long, I think we must find ways to explain how both sides have grasped a piece of the truth.

SOCIAL NORMS

Hufendiek and Sievers go on to provide an interesting review of the ways in which our emotions are influenced by social norms. Relevant to later points I will address about emotional modularity, Hufendiek and Sievers point out that non-human animals show differential emotional responses in social settings. Our expressive behaviour is also sensitive to social hierarchy. I entirely endorse their argument that “for social animals, the triggered responses may be harnessed for social purposes and may therefore be subject to social influences (such as socially normative regularities)” (Hufendiek and Sievers 2024, 27). This includes the claim that “we become ontogenetically attuned to social norms in a way that shapes our bodily response patterns and the representational content” (27).

It is not the case that, in my view, social norms are merely about self-consciously holding ourselves to certain ideals of propriety (a point I make in passing in chapter 6). Chapter 5 of my book is devoted to analysing how social interactions constitute a distinct level in the control hierarchy of valent representation; one that is developmentally and evolutionarily prior to our conscious reasoning capacities. This control function is achieved by means of our expressive behaviour. And the purpose of expressive behaviour is not merely to conform to social rules about what situations merit what emotions, but to regulate our affiliative status–our levels of intimacy and dominance with respect to other animals. The consequence is that I allow for social versions of emotions like fear and anger (Cochrane 2018, 122). Thus, following one of Hufendiek and Sievers’s examples, one could experience a distinctly social fear, the regulative aim of which is to protect the group from harm.

Another aspect of my account worth emphasising is that I do not think the affiliative control function can be entirely reduced to the individual animal’s neuro-biological machinery since this mechanism relies indispensably on how others react (§5.4-§5.5). Given the reliance of this system on how others in fact react, my view entails that culturally specific patterns of emotional interaction can emerge. I do not claim that there are entire types of emotion that are culturally specific, where say, the individuals in one culture feel awumbuk (207) while the individuals in another culture never do, even if the second culture lacks a term for that emotion. The components out of which our emotions are made, and the mechanisms whereby intimacy and dominance are established, are species wide. Yet styles of affective interaction can vary between cultures (or indeed between small groups) such that one group has a flamboyant manner of social-emotional regulation, where another is mutually restrained, or where one group embraces exquisite refinements in social hierarchy where another group is rigorously egalitarian. Of course, it is also possible for the way we understand situations to differ such that some people find a situation worthy of anger where other people find it worthy of sadness. Allowing for such variations is an indispensable feature of any plausible emotion theory.

NATURAL KINDS

This brings me to Majeed’s (2024; this issue) commentary. Majeed wonders if, given their wide variability and flexibility, the affective states form an innate natural kind. In particular, he is skeptical that grief involves the same sort of causal mechanism as other emotions, which according to Griffiths (1997), is a condition of belonging to a common natural kind. 

[page 55]

In section 7.5, I characterise grief as the product of a loving sentiment (note that I do not say that it is a sentiment). Sentiments are concerns directed at specific individuals enduring through time that we track as doing well or badly in their own right (Cochrane 2018, 171-172). This sort of concern is underwritten by our capacity for generating narratives about these individuals (175-177). It is the idea that different kinds of affective states are unlocked by innovations in representational sophistication that leads me to ascribe hierarchical levels of valent representation (196). Given the distinctive and generalisable nature of human sentiments, I think they count as a natural kind. And since sentiments function according to the basic causal principles of valent representation (we have impulsive responses to regulate how well or badly the targets of our sentiments are doing), they also belong to the wider natural kind of affective states.

The situation is a bit more complex for grief. I argue that the shifting emotional responses we see over the course of the grieving process are “the product of the kind of long-term adaptation I described earlier with regards to the shaping of personality” (174). Majeed does not quote the last part of this sentence when he discusses my position, but it is crucial for avoiding his charge that I have asserted without justification that grief is an evolutionary adaptation. Earlier in that chapter, I talk about how personality develops over the course of a person’s life as they adapt to the opportunities and skills available to them (165).[3] My claim is that a similar sort of gradual adaptation is occurring in grief.

To clarify, the emotions of grief attempt to restore or compensate for a major loss that cannot be restored or compensated. The griever adopts several different emotional strategies (such as bargaining, anger, and sadness) until they eventually adjust their expectations about the status of their beloved, which stops their emotions being triggered so frequently. Does this make grief a natural kind? I think there is a definite mechanism here in which one’s sentiment towards the beloved is “reprogrammed.” Yet, since the process will involve a great deal of conscious re-evaluation, I would expect it to be highly variable between individuals. If that level of variability is compatible with calling it a natural kind, then that’s fine by me. Meanwhile, the emotions we experience in grief are again all valent representations even if they are characteristically unsuccessful, which is why we can broadly incorporate grief into the category of affective states.

NEUROBIOLOGICAL VARIABILITY

Majeed’s commentary then moves onto whether claims about emotional innateness can be maintained in the face of the evidence of neurobiological variability between peoples’ emotions. Majeed is right to point out that the layers of elaboration that I propose upon valent representation give me the resources to allow for a wide range of variability. Specifically, these resources include calculations of coping potential via the monitoring of our bodily responses, expressive signalling and coordination, conscious checking and planning processes, adherence to social norms, sentimental attachments, as well as variable personality strategies. The range of complex affective phenomena that I can capture in this manner is far beyond the simplistic idea of blending basic emotions.

Despite these elaborations, I still claim that at the base of the mind we possess categorically distinct systems of valent representation that are innately tuned to a limited range of stimuli. I also think there are definite systems for detecting losses, gains, obstacles, and other formal qualities characteristic of the emotion types, which trigger responses aimed at regulating those sorts of situations. I only describe these systems as “somewhat modular” because I do not believe them to be informationally encapsulated. Majeed notes how I allow for our emotional systems to be attuned to new stimuli. We start with a certain range of what counts as an attractant or avoidant, which gets broadened by both associative (Cochrane 2018, 82-85) and inferential processes (142-145). I also suppose that thresholds for stimulation can be adjusted in the process of personality development (165). Majeed observes, however, that I am less flexible with regards to the output responses of our (unelaborated) emotions. 

[page 56]

A major reason for positing innate systems of emotional response is just our common observations of stereotypical emotional impulses, despite how often they are inappropriate for dealing with the situations faced by civilised humans (e.g., raging at a slowly loading computer). But are our responses really so stereotypical? Majeed cites Barrett’s work on emotional variability. Majeed (2022, §1) has also argued that we should take this empirical data at face value. I do not follow him here. When I looked into this data my impression was that it was extremely messy, confounded by researchers’ differing concepts of emotional stimuli (which I discuss in Cochrane 2018, 86-87). In particular, I’d expect calculations of coping potential and social-affiliations to result in major differences to our physiological and behavioural responses (and a fortiori to our neurological activity), to a point where different emotion categories are justifiable (as noted in section 3 above, as well as my précis). As such, it is imperative that these factors are controlled for when collecting data about emotion variability. 

Meanwhile, Majeed’s (2022) paper outlines a way in which a person’s emotional reactions could gradually take on modular characteristics, making them intra-personally stereotypical, while inter-personally variable. This view can thereby undercut inferences from stereotypical processes to innate modules. Majeed argues that the systems we are born with are “domain relevant” rather than “domain-specific,” meaning that they are only biased towards the processing of certain sorts of information rather than dedicated to them: “Such biases in turn can lead to the gradual development of brain circuits progressively selected to process inputs of that kind” (§3.2). This is an interesting proposal, though I’d like more detail on how a system tracks what counts as domain relevant. To my mind, having a few innate concerns and responses that can then be elaborated is more mechanistically plausible than an initially ambiguous set of concerns and responses. Moreover, I want to note again that my categorical emotion systems are refinements of our fundamental homeostatic processes, which as far as I’m aware, involve innately specified inputs and outputs. The way that my approach aligns our overall mental economy with these basic biological processes gives it the important theoretical virtue of scope.

PAIN ASYMBOLIA

Gerrans also denies that emotions are natural psychological kinds when they say that “there is no specialised system whose role is to transform feelings of bodily arousal to feelings of danger” (Gerrans 2024, 38; this issue).  However, Gerrans is referring to our experiences of the emotional bodily responses here. I only make nativist categorical claims about our simplest emotional systems, which I sharply distinguish from the experience of emotional bodily responses. It was because I recognized that the experiences of our bodily responses do not intrinsically present emotional meaning (as Gerrans notes) that I added a condition that their categorical distinctions are due to socially influenced conceptualisation (§4.6). As such, I am not at all committed to there being specialised neural systems for imposing categorical meanings onto feelings. I am, however, committed to specialised neural systems for detecting that, say, something of value has been lost. I suppose these detections to be achieved by contrast representations (e.g., “good before, bad now”) that use structural features of cognition to capture content in a non-linguistic way available to non-human animals (73-75). I think these contrast representations are innate because I do not see how they can be learned.

Meanwhile, Gerrans has much more to say about pain asymbolia, a condition in which individuals are aware of pain, but seem not to experience it with suffering and do not display the usual aversive responses. As Gerrans notes, I infer from pain asymbolia that pain has distinct sensory and affective components, and that pain asymbolia involves the loss of the affect. 

However, Gerrans argues that even if individuals with pain asymbolia lack affect, this cannot explain their lack of aversive responses, since individuals with frontal lobotomy maintain exaggerated aversive responses to noxious stimuli with reduced or absent affect. In addition, pain asymbolia is most associated with damage to the posterior insular, yet this neural structure seems to be a hub for integrating bodily self-representations. Gerrans

[page 57]

thus proposes that pain asymbolia is caused by an incapacity to track the self-relevance of pain, which has the downstream effect of eliminating suffering, a function more closely associated with the anterior insular (which serves as a hub for coordinating bodily self-awareness with our various adaptive responses).

There is a lot that I like about  Gerrans’ account of pain asymbolia. Elsewhere in the book (Cochrane 2018, 182n1) I also endorse  Gerrans’ (2015) account of depersonalisation as the result of losing the capacity to integrate a sense of self from the monitoring of our bodily responses. I also appreciate  Gerrans’ attempt to allow that my general theory of emotional bodily feelings has the resources to accommodate pain asymbolia, and of course it’s great that he is also working within a broadly control-theoretic framework. I need to make a couple of clarifications about my view of affect, however. 

First, I identify affect with neither emotions nor the feeling of emotional bodily responses, but only with the sense of intrinsic unpleasantness or pleasantness. This is not how the frontal lobotomy studies use the term “affect,” which for them refers to emotion more generally. Second, I do not directly associate unpleasant affect with our aversive bodily responses (unlike imperativist accounts of pain). I claim, rather, that affect is an urgency signal, the main function of which is to draw attention to and thereby prioritise existing aversive responses (Cochrane 2018, 62). Thus, lack of affect (in my sense) is compatible with the presence or absence of emotional responses. I can even allow that reflexive withdrawal responses may be intensified if emotions are lost since it is a general feature of my account that higher-level valent representations can inhibit lower-level valent representations (e.g., 187).

As far as I’m concerned, I only needed the case of pain asymbolia to demonstrate the distinction between the sensory representation of bodily damage and intrinsic unpleasantness, which it still does.[4] If asymbolia is more directly explained by the loss of self-representations than affect per se, then I think I could endorse this. Note that in my view, unpleasant affect is a higher-level representation that combines the representation of one’s response with the representation of the target (in this case, bodily damage). One represents oneself as failing to regulate an avoidant.[5] Thus, the incapacity to monitor one’s responses to bodily damage would have the downstream effect of eliminating unpleasant affect from bodily damage.

I would like to sound a note of caution, however. It seems that while pain asymbolic individuals lack the fear of bodily harm, they retain other emotions. For instance, in one account the pain asymbolic individual became angry when called a liar and a thief (Klein 2015, 501). Given that emotions like anger are intact, the monitoring of their emotional bodily responses, and therefore a general capacity for bodily self-representation, may remain intact as well. Thus, accounts of pain asymbolia need to stay focused on problems with emotion and self-representation particularly linked to bodily harm. Finally, it is tricky to draw robust conclusions about pain asymbolia because we have so few detailed reports of these individuals’ experiences. Hopefully, future reports will check if they can experience intrinsic unpleasantness in other sorts of ways (e.g., due to failure of a project), and whether their attention is still grabbed by bodily damage even if they don’t suffer it (which would seriously undermine my model of affect).

TRUSTING YOUR GUT

Finally, I turn to Whiting’s (2024; this issue) commentary, which is focused on the epistemology of our emotions. Whiting is particularly interested in our emotional self-awareness and how this may feed into our rational decision making. Certainly, it is part of my view that the awareness of our emotional bodily responses is a guide to our emotional state. These bodily responses carry information about how we have appraised the situation that we might not have consciously realised (refer to Cochrane 2018, 96; where I describe this as one sort of mood).

[page 58]

At the same time, I do not think that trusting your gut is good generalisable advice, as Whiting seems to agree. Specifically, one’s emotional bodily feelings are just one source of information that can feed into a complex decision-making process. Related to this, Michael Brady’s (2013) book is an excellent discussion of the extent to which one can trust one’s emotional reactions, and in my book I approvingly quote his view that emotional experience is best treated as a stimulus to try and further understand the cause and basis of one’s emotion (Cochrane 2018, 142). What is perhaps more distinctive to me is the view that because rational thinking is a motivated process, it is itself an elaboration of affective processing. Even the rational regulation of one’s emotions can be understood as a meta-emotional state. This raises interesting issues about whether it is always rational to regulate one’s emotions, which I pick up in a later article (Cochrane 2021). If one does not particularly want to be a hyper-rational person (because for instance, it tends to undermine one’s exuberance), then it will be irrational to continually wonder if one’s emotion is rational. Doing so would effectively be worrying about something that you don’t care about, thus irrational in the sense that it is not emotionally true that something has occurred that threatens your concerns. Thus, epistemic (and not just prudential) rationality can depend on one’s larger concerns about life.

Related to this, Whiting makes the nice point that in a social situation, advising someone to trust their gut may be a way to avoid giving specific advice that makes us partially responsible for a disastrous decision. We want the other person to make some kind of authentic emotional decision by their own lights. So, I think the most important issue here is how we discern what our most authentic attitudes are. In chapter 7 (Cochrane 2018), I introduce character as the final “boss level” of the affective life. This is the level at which we most explicitly decide what really matters to us, and the kind of person we want to be. As I see it, our character commitments are most clearly settled by what we love, and the way that our loves can demand that we regulate our emotions and even our rational thinking processes. In particular, in section 7.7, I describe how character commitments dictate our tendencies to consciously reflect on situations, engage in long-term planning, and the extent to which we ought to adhere to social normative standards.

Overall, in keeping with the overall hierarchical organisation of the mind that I have outlined in the book, I do not think that the most authentic emotional state is one that is hidden deep inside one’s gut, and that therefore has to be brought into the light. The authentic emotional life is rather the highest level of elaboration that we achieve as our various levels of regulative control contribute to large scale determinations about how we want to live. 

[page 59]

Notes

[1] For instance, refer to the recent Interdisciplinary Handbook of Perceptual Control Theory (Mansell, de Hullu, Huddy, and Scholte 2023).

[2] I discuss some of these affective influences on consciousness in Cochrane (2023).

[3] In fact, my account of personality traits resembles the developmental modularity that Majeed (2022) outlines in his paper, more on which shortly.

[4] I also do not treat affect as modular since I regard it as a general management system closely related to attention and consciousness.

[5] Connected to this, I was struck by Gerrans’ citation of Velasco and Loev’s (2020) similar view that affective feeling signals the success or failure of regulatory activity.

[page 60]

References

Brady, Michael S. 2013. Emotional Insight. Oxford, UK: Oxford University Press. 

https://doi.org/10.1093/acprof:oso/9780199685523.001.0001.   

Cochrane, Tom. 2018. The Emotional Mind: A Control Theory of Affective States. Cambridge, UK: Cambridge University Press.

_____. 2021. “Reason to Be Cheerful.” Review of Philosophy and Psychology 12 (2): 311–27. https://doi.org/10.1007/s13164-020-00480-y. 

_____. 2023. “Consciousness, Attention, and the Motivation-Affect System.” Journal of Consciousness Studies 30 (7): 139–63.  

https://doi.org/10.53765/20512201.30.7.139. 

Fernandez Velasco, Pablo, and Slawa Loev. 2021. “Affective Experience in the Predictive Mind: A Review and New Integrative Account.” 

Synthese 198 (11): 10847–82. https://doi.org/10.1007/s11229-020-02755-4. 

Gerrans, Philip. 2015. “All the Self We Need,” Open MIND Kap. 15(T): 1–19. http://doi.org/10.25358/openscience-282.

_____. 2024. “Representation and Regulation in Emotional Theory.” Journal of Philosophy of Emotion 5, no. 2: 36-43. 

https://doi.org/10.33497/2024.Winter.5. 

Hatzimoy, Anthony. 2024. “Valent Representation: Problems and Prospects.”  Journal of Philosophy of Emotion 5, no. 2: 17-23.

https://doi.org/10.33497/2024.winter.2. 

Hufendiek, Rebekka and Christine Sievers. 2024. “Valent Representations, Bodily Feelings, and Social Norms.”  Journal of Philosophy of Emotion 

5, no. 2: 24-29. https://doi.org/10.33497/2024.Winter.3. 

Klein, Colin. 2015. “What Pain Asymbolia Really Shows.” Mind 124 (494): 493–516. https://doi.org/10.1093/mind/fzu185. 

Majeed, Raamy. 2022. “Does the Problem of Variability Justify Barrett’s Emotion Revolution?” Review of Philosophy and Psychology. 

https://doi.org/10.1007/s13164-022-00650-0. 

_____. 2024. “Cochrane’s Nativism.”  Journal of Philosophy of Emotion 5, no. 2: 30-35. https://doi.org/10.33497/2024.Winter.4. 

Mansell, Warren, Eva de Hullu, Vyv Huddy, and Tom Scholte, eds. 2023. The Interdisciplinary Handbook of Perceptual Control Theory, Volume II. 

Cambridge, MA: Academic Press.

Whiting, Elizabeth L. 2024. “Is Trusting Your Gut a Good Idea? Implications from The Emotional Mind.”  Journal of Philosophy of Emotion 5, no. 2: 

44-51. https://doi.org/10.33497/2024.Winter.6.

[page 61]

Tom Cochrane © 2024

Author email: tom.cochrane@flinders.edu.au