Abstract
This paper aims to offer an account of affective experiences within Predictive Processing, a novel framework that considers the brain to be a dynamical, hierarchical, Bayesian hypothesis-testing mechanism. We begin by outlining a set of common features of affective experiences (or feelings) that a PP-theory should aim to explain: feelings are conscious, they have valence, they motivate behaviour, and they are intentional states with particular and formal objects. We then review existing theories of affective experiences within Predictive Processing and delineate two families of theories: Interoceptive Inference Theories (which state that feelings are determined by interoceptive predictions) and Error Dynamics Theories (which state that feelings are determined by properties of error dynamics). We highlight the strengths and shortcomings of each family of theories and develop a synthesis: the Affective Inference Theory. Affective Inference Theory claims that valence corresponds to the expected rate of prediction error reduction. In turn, the particular object of a feeling is the object predicted to be the most likely cause of expected changes in prediction error rate, and the formal object of a feeling is a predictive model of the expected changes in prediction error rate caused by a given particular object. Finally, our theory shows how affective experiences bias action selection, directing the organism towards allostasis and towards optimal levels of uncertainty in order to minimise prediction error over time.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
We use the terms “affective experiences” and “feelings” interchangeably.
It seems largely a matter of definition that feelings are conscious: “That there can be no unconscious feelings is still the position of “commonsense.”” (Lacewing 2007, p. 98; see also Clore 1994). However, this is not to say that there are no unconscious emotions or, more generally, unconscious states that are functionally largely analogues to feelings or sub-classes of affective experiences like emotional feelings (Winkielman and Berridge 2004; Lacewing 2007). For perceptual experiences it seems plausible that there are forms of unconscious perception and something similar might be said for affective experiences (say, there might be unconscious affective reactions) (Prinz 2005). However, when making this comparison the aspect of interest is not that perceptual experiences are a form of perception (as is unconscious perception) but that they are a form of experience, and it is a conceptual fact about experiences that they are conscious. Affective experiences should also be distinguished from emotional or affective processing which often does occur unconsciously (see e.g. Mathews and MacLeod 2002). Apart from terminological rationale, there is good reason to believe that fine-grained affective experiences require conscious awareness (Pessoa 2005).
This phenomenal quality often but not always correlates with closely associated but ultimately non-phenomenal properties such as object valence (Colombetti 2005). It is also worth mentioning that the idea of “unconscious valence” is lately gaining ground (e.g. Berridge and Kringelbach 2015). Unconscious valence has a functional profile similar to phenomenal valence in motivating behaviour. Although we think there might be some problems with this idea, arguing the point would take us too far afield. Here, our focus are affective experiences and we are thus concerned with phenomenal valence. As we will see, however, a virtue of the emerging picture is that it can incorporate the idea of unconscious valence.
This conception of the mark of the affective echoes the conception of the mark (or marks) of the mental, the feature (or set of features) “that set characteristically mental phenomena apart from the characteristically physical phenomena” (Pernu 2017, p. 1).
This also points to the connection between affect and learning, an issue to which we will return in Sect. 4.
Note that it is not only that the (weakly) negative valence does not lead to a weakly negative motivation to avoid cigarettes—we rather observe that there is a strong, positive motivation in place that is poorly explained by the implicated negative valence.
There are differing views as to whether or not moods (which we take to be affective experiences) should be considered intentional states. This is an issue that we address in Sect. 5.1.
PP is a process theory whereas the Bayesian brain is a normative principle, a clarification suggested by an anonymous reviewer.
A recent review introduces the possibility that context-independent structural components constraining bottom-up processing constitute a form of bottom-up predictions (Teufel and Fletcher 2020). It is still early to tell if (and how) such a conception would change the PP framework.
Although the constructs of prediction, hypothesis and inference are often used interchangeably in the PP literature, it is helpful to briefly clarify their relations and differences. Technically, a hypothesis is a joint set of predictions. Inference is the updating of the prior to the posterior in the light of the prediction and the prediction error. Predictions are expected states based on Error and previous experience while inference is the process of predicting states based on Error and previous experience. That is, inference is the process that yields new predictions.
Two points need to be qualified here: (1) We experience affect only if the interoceptive predictions in question additionally meet the criteria for supplying the contents of conscious experience. There might be interoceptive predictions that do not meet these criteria, such as those that are computed but discarded because competing predictions are more successful in minimising Error. (2) These interoceptive predictions are normally not all that there is to the affective experience. As affective experiences are usually multimodal, they are co-constituted by predictions from other modalities (e.g. those underlying the visual experience of the bear). It is, however, the interoceptive bit that makes the multimodal experience in question an affective experience.
A way of resolving this issue might be to draw the line between affective and non-affective (interoceptive) experiences by appealing to the implication of deep (i.e. high-level) expectations associated with an individual’s concerns or goals, arguing that it is specific to affective experiences (cf. Seth and Friston 2016, p. 5). An initial worry with this idea is that there seem to be affective experiences, such as itches or the enjoyment of one’s favourite music, where it is not clear from which high-level expectations they receive support. The suggestion, furthermore, poses the following question: why would the mentioned deep expectations be specific to interoceptive inference? It seems that all kinds of inference should be able to be supported by expectations at higher levels of the hierarchy, and, in fact, they regularly are. Thus, the implication of high-level expectations expressive of concerns does not appear like an IIT-friendly mark of the affective. Thanks to an anonymous reviewer for making us think about this point.
Actions themselves are understood as predictions that dictate bodily behaviour, see Sect. 2.
For two important qualifications see footnote 12.
As an anonymous reviewer helpfully suggested, one might try to accommodate music or humour by reference to cultural (hyper-)priors. We are looking forward to the development of a proposal along these lines. However, we would like to point out a challenge resulting from instances of affective experiences that are similar to e.g. the mentioned aesthetic or humour-related experiences but seemingly unrelated to culture. One is the mentioned sunset. Then there is also the internet phenomenon of “Oddly Satisfying Videos”. These clips show events and actions that typically involve the meticulous manipulation of physical objects such as peeling wood (see the same-named subreddit and YouTube channel). Their audience reportedly experience positive affective experiences watching them. By the same token, why is the experience of watching upward flowing water negatively valenced (i.e. why does it feel wrong)? Here is a demonstration: https://youtu.be/NiOAfQZwn0g It appears doubtful to us that these affective experiences are easily explained in terms of instrumental interoceptive predictions or by reference to cultural priors. A strength of the alternative we are going to offer in Sect. 5, is that it derives the positivity and negativity resulting from such expectation-satisfying/violating (perceptual) experiences from underlying PP-principles rather than from priors with case-specific content.
There is another problem with the appeal to future interoception which has to do with the intentionality of affective experiences. If there are no changes in interoceptive signals right now, how come that there is an affective experience that seems to be about this music that we enjoy—something in the here and now? It appears that if there are no immediate changes in interoceptive signals, there will be nothing interoceptive to infer the causes of, and so affective experiences would be about nothing in the here and now. In contrast: affective experiences are more often than not about the here and now (this is, of course, not to say that there are no affective experiences with future or even past-directed temporal orientations, such as hope, pleasant anticipation or regret).
Interestingly, this sort of thing is proposed to happen by PP-accounts of conditions like chronic fatigue and depression (Stephan et al. 2016). We agree that this seems like a plausible account of these abnormal affective conditions. However, it does not seem like a good model for normal affective states that are our primary focus here. Thanks to an anonymous reviewer for bringing this work to our attention.
It is worth emphasising what it is that is missing in IIT. IIT gives us (control-oriented) predictions of interoceptive changes as the mark of the affective. Thus, upon a bear encounter, instrumental interoceptive predictions will issue physiological changes such as an increase in heartbeat rate. These predictions will make the protagonist of our example enter an affective state. Now, when we read about this situation, we have no difficulty judging that our unlucky protagonist will not only be in some affective state but in a negative affective state of fear. However, the only thing that we seem to get from IIT as the base of affective experiences are counterfactual predictions about physiological changes. What does not straightforwardly fall out of IIT is the intuitive metric by which we determine that the predicted physiological changes are bad for the wellbeing of the protagonist (see also footnote 20). Such a metric is needed because it is not obvious that an instrumental interoceptive prediction taken by itself is intrinsically positive or negative. Intuitively, the instrumental prediction of physiological changes such as a heartbeat increase can be a bad thing in the case of a bear encounter or a good thing in the case of a romantic encounter. In this context, high-level predictions (bear vs. romantic encounter) can plausibly contextualise the counterfactual interoceptive predictions, marking them as positive or negative. Apart from the mentioned issue that high-level predictions are not specific to interoceptive predictions (see footnote 13), there is another problem lurking. We seem to encounter a regress: instrumental interoceptive predictions are plausible reactions to a bear encounter—but now we need the contextualising high-level predictions for the instrumental interoceptive predictions to emerge (as negative) in the first place.
Perhaps, then, affect intensity maps in the following way: the more interoceptive changes are predicted (i.e. the more fluctuation in interoceptive variables), the more positive or negative it is. In fact, this might also offer an account of valence polarity: if high/low interoceptive fluctuation is predicted, then the experience is positive/negative. This idea entails that we would have the most intense positive affective experiences if no interoceptive changes are predicted. However, this appears to obviate the need for interoceptive prediction in the first place and it might be thought to render the resulting experience non-affective. It also seems at odds with the observation that positive affective experiences are often accompanied by a myriad of bodily sensations, which would have to map onto a lot of (instantaneous) interoceptive changes. But then, the suggested criterion would predict that the affective experience with many interoceptive changes would have to be negative. Is it then, perhaps, that the negativity of the many instantaneous interoceptive changes is somehow outweighed by the predicted future interoceptive stability? This would appear at odds with how the bodily sensations appear to one in the moment of the positive affective experience—namely as contributing rather than subtracting from its positivity.
We thank an anonymous reviewer for this suggestion.
This mark of the affective is narrower than the first, i.e. interoceptive predictions simpliciter, but broader than the second, i.e. instrumental interoceptive predictions.
Note that this proposal can preserve the intuitive distinction between “cold” and “hot” interoception by appealing to the distinction between perceptual and active interoceptive inference.
We have mostly focused on single aspects of the IIT machinery, seeking to identify a plausible mark of the affective among them that should give us something resembling the valence of affective experiences. IIT might now rejoin: valence is not to be found in single IIT aspects—but in the interplay of several of those. For instance: active interoceptive inferences will not only yield predictions that are either epistemic or instrumental but predictions that will be both at the same time, expressing a balance between epistemic and instrumental components. Furthermore, these dual-natured counterfactual interoceptive predictions will be contextualised by high-level predictions (see footnotes 13 and 20). We admit that evaluating such a proposal introduces a level of exegetical complexity that we hesitate to take up, leaving it to the proponents of IIT to spell out against the background of the raised issues. To us it is not obvious at first look that this (or a similar) mixed proposal can successfully overcome the raised issues and provide a graded polarised metric which plausibly maps on affective experiences. Also, mixing several components together would introduce the further challenge to explain how varying combinations of relatively heterogenous components map onto valence, a property that appears as a rather unitary fundamental component of affective experiences.
Rate is short for instantaneous rate of decrease of prediction error, which is the negative of the instantaneous rate of change of prediction error.
The early versions of EDT emerged as accounts of aesthetic experience (Van de Cruys and Wagemans 2011), just like early versions of IIT emerged as accounts of interoception. As we will see, this move “away from the body” puts EDT in a better position to address the challenge of the aesthetic cases presented in the previous section, such as listening to Chopin.
More precisely, they are “the content of the joint set of predictions geared towards allostasis” (Seth and Tsakiris 2018, p. 6).
We thank an anonymous reviewer for flagging up this point.
In the most common version of binocular rivalry, one eye is presented with a picture of a house and the other eye is presented with a picture of a face, and conscious experience switches between the two percepts. Hohwy’s PP explanation is that the likelihood of the combined face-house blend hypothesis cannot overcome “the exceedingly low probability that a face and a house could co-exist in the same spatiotemporal location […] so the hypothesis that is selected, and which determines perception, is either the face or the house hypothesis.” (Hohwy 2013, pp. 21–22).
On the other hand, (instantaneous) Rate might be best understood as unconscious valence (e.g. Berridge and Kringelbach 2015) (see footnote 4).
Thanks to an anonymous reviewer for directing us towards this work.
The temporality issue was particular to EDT rather than to IIT. Because on IIT predictions are not only about the states of hidden causes, but about their trajectories over time. Thus, both AIT and IIT are well-placed to explain the phenomenological insight that the here-and-now is always intertwined with the past and the future. We thank an anonymous reviewer for bringing this to our attention.
This is in line with the early EDT idea that “prediction errors at the level of style (perceptual ones) sometimes can be resolved on the level of meaning” (Van de Cruys and Wagemans 2011, p. 1053). However, Van de Cruys and Wagemans fail to flesh out this idea, claiming at once that “a central assumption of our theory is that prediction errors are always to some extent emotional, more specifically negative in valence” (p. 1047) and that “this playful and safe as-if context of art, where our guards can be lowered and our actions suspended, allows for the usually negative prediction errors to be enjoyed” (p. 1041), without explaining how negative prediction error suddenly turns into positive phenomenal valence in the context of aesthetic experience.
Insofar, primary affective intentionality is a kind of phenomenal intentionality (Horgan and Tienson 2002). Phenomenal intentionality is the thesis that the intentional content of a state is fixed by the phenomenology or “what it is like”-ness of the state. Thus, what a feeling is about or directed at is (partly) specified by the phenomenology of the feeling.
Note that an expected change in Rate is equivalent to a change in Expected Rate.
Here as well there seems to be somewhat of a gradient. One could make the case that we have something like dedicated fast-track threat detection pathways (for a critical review see Pessoa and Adolphs 2010). Such a case for dedicated sensory pathways is arguably easier to make for feelings such as fear than for feelings such as funniness, (see also Fulkerson 2019).
Of course, also phenomenal valence, i.e. the primary affective (phenomenal) intentionality bit, is a result of inference. The “inference-based” in secondary affective intentionality aims to emphasise that the causes of changes in Expected Rate need to be inferred in turn and that the success of this process is not guaranteed. Consequently, an affective experience cannot fail to have valence (otherwise it simply wouldn’t be an affective experience) while the valence of an affective experience can fail to bind to specific inferred causes (making it lack at least phenomenal object-directed intentionality). This point will become clearer in the following paragraph on moods that have primary affective intentionality but lack (part of) secondary affective intentionality.
Similar things could be said for other feelings that appear relatively undirected, such as feeling tired, relaxed or lascivious. Tentative sketches: In the case of tiredness/relaxation one feels like one’s capability to go on with one’s bodily activity (of Error reduction) is decreasing or smoothly increasing. In the case of feeling lascivious the world appears as populated by mating opportunities, which against a certain model results in PEM.
Here, we use the term action and behaviour interchangeably, reflecting the usage in the PP literature.
An anonymous reviewer brought to our attention this manuscript, which is congenial to the views expressed here. The computational model of Hesp and colleagues can be seen as an extension of the principles behind EDT (Joffily and Coricelli 2013) to the inference of future states of subjective fit (i.e. affective inference) and consequent action selection. Whereas our work is mainly philosophical in nature, their work is mainly computational in nature, but we see this convergence as an auspicious sign.
In the end, AIT is a synthesis and it is therefore potentially compatible with revised versions of both IIT and EDT. A recent paper by Tschantz, Seth and Buckley propose an action-oriented model of goal-oriented and epistemic behaviour based on expected free-energy (Tschantz et al. 2020). A version of IIT in which valence were cast in terms of expected free-energy (in the style of Joffily and Coricelli 2013)—so that, in the analogous terms of prediction, valence could be understood as expected Rate—would be very much in line with AIT (thanks to an anonymous reviewer for suggesting this possibility). Allostasis is a biological imperative, so following the free-energy principle, free-energy minimisation will lead to allostasis. Feelings will guide organisms to minimising Error and, by extension, to allostasis, so there will be a tight connection between affect and interoception in the lines proposed by IIT. The crucial difference between AIT and IIT is that AIT is committed to valence being cast in terms of predictions (of Rate), not directly in terms of interoception.
References
Aristotle. (1986). De anima (on the soul) (H. Lawson-Tancred, Trans.). Harmondsworth: Penguin Books.
Auvray, M., Myin, E., & Spence, C. (2010). The sensory-discriminative and affective-motivational aspects of pain. Neuroscience and Biobehavioral Reviews, 34(2), 214–223.
Bain, D. (2013). What makes pains unpleasant? Philosophical Studies, 166(1), 69–89.
Bain, D. (2014). Pains that don’t hurt. Australasian Journal of Philosophy, 92(2), 305–320.
Barrett, L. F. (2006). Valence is a basic building block of emotional life. Journal of Research in Personality, 40(1), 35–55.
Barrett, L. F. (2014). The conceptual act theory: A precis. Emotion Review, 6(4), 292–297.
Barrett, L. F. (2017). The theory of constructed emotion: An active inference account of interoception and categorization. Social Cognitive and Affective Neuroscience, 12(1), 1–23.
Barrett, L. F., & Bliss-Moreau, E. (2009). Affect as a psychological primitive. Advances in Experimental Social Psychology, 41, 167–218.
Barrett, L. F., & Simmons, W. K. (2015). Interoceptive predictions in the brain. Nature Reviews Neuroscience, 16(7), 419–429.
Batson, C. D., Shaw, L. L., & Oleson, K. C. (1992). Differentiating affect, mood, and emotion: toward functionally based conceptual distinctions. Emotion, 13, 294–326.
Berridge, K. C., & Kringelbach, M. L. (2015). Pleasure systems in the brain. Neuron, 86(3), 646–664.
Berthier, M., Starkstein, S., & Leiguarda, R. (1988). Asymbolia for pain: A sensory-limbic disconnection syndrome. Annals of Neurology, 24(1), 41–49.
Brown, C. A., Seymour, B., Boyle, Y., El-Deredy, W., & Jones, A. K. (2008). Modulation of pain ratings by expectation and uncertainty: Behavioral characteristics and anticipatory neural correlates. Pain, 135(3), 240–250.
Carruthers, P. (2017). Valence and value. Philosophy and Phenomenological Research, 97(3), 658–680.
Carver, C. S., & Scheier, M. F. (1990). Origins and functions of positive and negative affect: a control-process view. Psychological review, 97(1), 19.
Carver, C. S., & Scheier, M. F. (2001). On the self-regulation of behavior. Cambridge University Press.
Chanda, M. L., & Levitin, D. J. (2013). The neurochemistry of music. Trends in Cognitive Sciences, 17(4), 179–193.
Charland, L. (2005). The heat of emotion: Valence and the demarcation problem. Journal of Consciousness Studies, 12(8–10), 8–10.
Clark, A. (2015). Surfing uncertainty: Prediction, action, and the embodied mind. Oxford: Oxford University Press.
Clore, G. L. (1994). Why emotions are never unconscious. In P. Ekman & R. J. Davidson (Eds.), The nature of emotion: Fundamental questions (pp. 285–290). New York: Oxford University Press.
Colombetti, G. (2005). Appraising valence. Journal of Consciousness Studies, 12(8–10), 8–10.
Corder, G., Ahanonu, B., Grewe, B. F., Wang, D., Schnitzer, M. J., & Scherrer, G. (2019). An amygdalar neural ensemble that encodes the unpleasantness of pain. Science, 363(6424), 276–281.
Corns, J. (2014). Unpleasantness, motivational oomph, and painfulness. Mind and Language, 29(2), 238–254.
Craig, A. D. (2003). Interoception: The sense of the physiological condition of the body. Current Opinion in Neurobiology, 13(4), 500–505.
Craig, A. D. (2009). How do you feel—Now? The anterior insula and human awareness. Nature Reviews Neuroscience, 10(1), 59–70.
Critchley, H. D., Wiens, S., Rotshtein, P., Ohman, A., & Dolan, R. J. (2004). Neural systems supporting interoceptive awareness. Nature Neuroscience, 7(2), 189.
Damasio, A. R. (1994). Descartes’ error: Emotion, rationality and the human brain.
Damasio, A., & Carvalho, G. B. (2013). The nature of feelings: Evolutionary and neurobiological origins. Nature Reviews Neuroscience, 14(2), 143–152.
De Haan, S., Rietveld, E., Stokhof, M., & Denys, D. (2013). The phenomenology of deep brain stimulation-induced changes in OCD: An enactive affordance-based model. Frontiers in Human Neuroscience, 7, 653.
De Sousa, R. (1987). The rationality of emotion. Cambridge: MIT Press.
Deonna, J. A., & Teroni, F. (2017). Getting bodily feelings into emotional experience in the right way. Emotion Review, 9(1), 55–63.
Döring, S. A. (2007). Seeing what to do: Affective perception and rational motivation. Dialectica, 61(3), 363–394.
Duncan, S., & Barrett, L. F. (2007). Affect is a form of cognition: A neurobiological analysis. Cognition and Emotion, 21(6), 1184–1211.
Feldman, H., & Friston, K. (2010). Attention, uncertainty, and free-energy. Frontiers in Human Neuroscience, 4, 215.
FitzGerald, T. H., Friston, K. J., & Dolan, R. J. (2012). Action-specific value signals in reward-related regions of the human brain. Journal of Neuroscience, 32(46), 16417–16423.
Frijda, N. H. (1986). The emotions. Cambridge: Cambridge University Press.
Frijda, N. H. (2005). Emotion experience. Cognition and Emotion, 19(4), 473–497.
Frijda, N. H. (2007). The laws of emotion. OCLC: 938467399. Mahwah: Lawrence Erlbaum Associates.
Friston, K. (2003). Learning and inference in the brain. Neural Networks, 16(9), 1325–1352.
Friston, K. (2009). The free-energy principle: A rough guide to the brain? Trends in Cognitive Sciences, 13(7), 293–301.
Friston, K. (2010). The free-energy principle: A unified brain theory? Nature Reviews Neuroscience, 11(2), 127–138.
Friston, K., FitzGerald, T., Rigoli, F., Schwartenbeck, P., & Pezzulo, G. (2016). Active inference and learning. Neuroscience and Biobehavioral Reviews, 68, 862–879.
Friston, K., Mattout, J., & Kilner, J. (2011). Action understanding and active inference. Biological Cybernetics, 104(1–2), 137–160.
Friston, K., Rigoli, F., Ognibene, D., Mathys, C., Fitzgerald, T., & Pezzulo, G. (2015). Active inference and epistemic value. Cognitive Neuroscience, 6(4), 187–214.
Fulkerson, M. (2019). Emotional perception. Australasian Journal of Philosophy, 98, 16–30.
Gerrans, P. (2015). All the self we need. Open MIND.
Gill, C. (2010). Naturalistic psychology in galen and stoicism. Oxford: Oxford University Press.
Goldie, P. (2002). Emotions, feelings and intentionality. Phenomenology and the Cognitive Sciences, 1(3), 235–254.
Gu, X., Hof, P. R., Friston, K. J., & Fan, J. (2013). Anterior insular cortex and emotional awareness. Journal of Comparative Neurology, 521(15), 3371–3388.
Helm, B. W. (2009). Emotions as evaluative feelings. Emotion Review, 1(3), 248–255.
Hesp, C., Smith, R., Parr, T., Allen, M., Friston, K., & Ramstead, M. (2019). Deeply felt affect: The emergence of valence in deep active inference. https://doi.org/10.31234/osf.io/62pfd.
Hohwy, J. (2011). Phenomenal variability and introspective reliability. Mind and Language, 26(3), 261–286.
Hohwy, J. (2012). Attention and conscious perception in the hypothesis testing brain. Frontiers in Psychology, 3, 96.
Hohwy, J. (2013). The predictive mind. Oxford: Oxford University Press.
Hohwy, J. (2017). Priors in perception: Top-down modulation, Bayesian perceptual learning rate, and prediction error minimization. Consciousness and Cognition, 47, 75–85.
Hohwy, J. (2020). New directions in predictive processing. Mind and Language, 35(2), 209–223.
Hohwy, J., Roepstorff, A., & Friston, K. (2008). Predictive coding explains binocular rivalry: An epistemological review. Cognition, 108(3), 687–701.
Hommel, B. (2004). Event files: Feature binding in and across perception and action. Trends in Cognitive Sciences, 8(11), 494–500.
Horgan, T., & Tienson, J. (2002). The intentionality of phenomenology and the phenomenology of intentionality. In D. Chalmers (Ed.), Philosophy of mind: Classical and contemporary readings (pp. 520–533). Oxford: Oxford University Press.
James, W. (1884). What is an emotion? Mind, 9(34), 188–205.
Joffily, M., & Coricelli, G. (2013). Emotional valence and the free-energy principle. PLoS Computational Biology, 9(6), e1003094.
Kenny, A. (1963). Action, emotion and will. London: Routledge & Kegan Paul.
Kidd, C., Piantadosi, S. T., & Aslin, R. N. (2012). The Goldilocks effect: Human infants allocate attention to visual sequences that are neither too simple nor too complex. PLoS ONE, 7(5), e36399.
Kiverstein, J., Miller, M., & Rietveld, E. (2019). The feeling of grip: Novelty, error dynamics, and the predictive brain. Synthese, 196(7), 2847–2869.
Klein, C. (2015). What pain asymbolia really shows. Mind, 124(494), 493–516.
Kozuch, B. (2018). No pain, no gain (in Darwinian fitness): A representational account of affective experience. Erkenntnis, 85, 693–714.
Kriegel, U. (2014). Towards a new feeling theory of emotion: Towards a new feeling theory of emotion. European Journal of Philosophy, 22(3), 420–442.
Lacewing, M. (2007). Do unconscious emotions involve unconscious feelings? Philosophical Psychology, 20(1), 81–104.
Lange, C. G. (1887). Uber Gemütsbewegungen. Ihr Wesen Und Ihr Einfluß Auf Körperliche, Besonders Auf Krankhafte Lebenserscheinungen. Ein Medizinisch-Psychologische Studie. T. Thomas.
Lazarus, R. S. (1991). Emotion and adaptation. Oxford: Oxford University Press.
Levy, D. J., & Glimcher, P. W. (2012). The root of all value: A neural common currency for choice. Current Opinion in Neurobiology, 22(6), 1027–1038.
Mathews, A., & MacLeod, C. (2002). Induced processing biases have causal effects on anxiety. Cognition and Emotion, 16(3), 331–354.
McGlone, F., Wessberg, J., & Olausson, H. (2014). Discriminative and affective touch: Sensing and feeling. Neuron, 82(4), 737–755.
Meinong, A. (1917). Uber emotionale Präsentation. Vienna: A. Hölder.
Melzack, R., & Wall, P. D. (1988). The challenge of pain. London: Penguin.
Mendelovici, A. (2014). Pure intentionalism about moods and emotions. In U. Kriegel (Ed.), Current controversies in philosophy of mind (pp. 135–157). New York: Routledge.
Meyer, L. B. (1956). Emotion and meaning in music. Chicago: University of Chicago Press.
Michal, M., Reuchlein, B., Adler, J., Reiner, I., Beutel, M. E., Vögele, C., et al. (2014). Striking discrepancy of anomalous body experiences with normal interoceptive accuracy in depersonalization-derealization disorder. PLoS ONE, 9(2), e89823.
Millikan, R. G. (1995). Pushmi-Pullyu representations. Philosophical Perspectives, 9, 185–200.
Pernu, T. K. (2017). The five marks of the mental. Frontiers in Psychology, 8, 1084.
Perrykkad, K., & Hohwy, J. (2020). Fidgeting as self-evidencing: A predictive processing account of non-goal-directed action. New Ideas in Psychology, 56, 100750.
Pessoa, L. (2005). To what extent are emotional visual stimuli processed without attention and awareness? Current Opinion in Neurobiology, 15(2), 188–196.
Pessoa, L., & Adolphs, R. (2010). Emotion processing and the amygdala: From a ‘low road’ to ‘many roads’ of evaluating biological significance. Nature Reviews Neuroscience, 11, 773–782.
Pezzulo, G., & Cisek, P. (2016). Navigating the affordance landscape: Feedback control as a process model of behavior and cognition. Trends in Cognitive Sciences, 20(6), 414–424.
Pezzulo, G., Rigoli, F., & Friston, K. (2015). Active inference, homeostatic regulation and adaptive behavioural control. Progress in Neurobiology, 134, 17–35.
Pezzulo, G., Rigoli, F., & Friston, K. J. (2018). Hierarchical active inference: A theory of motivated control. Trends in Cognitive Sciences, 22(4), 294–306.
Ploner, M., Freund, H. J., & Schnitzler, A. (1999). Pain affect without pain sensation in a patient with a postcentral lesion. Pain, 81(1–2), 211–214.
Prinz, J. J. (2004). Gut reactions: A perceptual theory of emotion. Oxford: Oxford University Press.
Prinz, J. (2005). Are emotions feelings? Journal of Consciousness Studies, 12(8–9), 9–25.
Prinz, J. J. (2006). Is emotion a form of perception? Canadian Journal of Philosophy, 36(sup1), 137–160.
Prinz, J. (2010). For valence. Emotion Review, 2(1), 5–13.
Proust, J. (2013). The philosophy of metacognition: Mental agency and self-awareness. Oxford: Oxford University Press.
Proust, J. (2015). The representational structure of feelings. In T. Metzinger & J. M. Windt (Eds.), Open MIND. Frankfurt am Main: MIND Group.
Rubins, J. L., & Friedman, E. D. (1948). Asymbolia for pain. Archives of Neurology & Psychiatry, 60(6), 554–573.
Scarantino, A. (2014). The motivational theory of emotions. In D. Jacobson & J. D’Arms (Eds.), Moral psychology and human agency (pp. 156–185). Oxford: Oxford University Press.
Schroeder, T. (2004). Three faces of desire. Oxford University Press.
Schwartenbeck, P., Passecker, J., Hauser, T. U., FitzGerald, T. H., Kronbichler, M., & Friston, K. J. (2019). Computational mechanisms of curiosity and goal-directed exploration. Elife, 8, e41703.
Seth, A. K. (2013). Interoceptive inference, emotion, and the embodied self. Trends in Cognitive Sciences, 17(11), 565–573.
Seth, A. K. (2015). The cybernetic bayesian brain. In T. Metzinger & J. M. Windt (Eds.), Open MIND. Frankfurt am Main: MIND Group.
Seth, A. K., & Friston, K. J. (2016). Active interoceptive inference and the emotional brain. Philosophical Transactions of the Royal Society B: Biological Sciences, 371(1708), 20160007.
Seth, A. K., Suzuki, K., & Critchley, H. D. (2012). An interoceptive predictive coding model of conscious presence. Frontiers in Psychology, 2, 395.
Seth, A. K., & Tsakiris, M. (2018). Being a beast machine: The somatic basis of selfhood. Trends in Cognitive Sciences, 22(11), 969–981.
Stephan, K. E., Manjaly, Z. M., Mathys, C. D., Weber, L. A. E., Paliwal, S., Gard, T., et al. (2016). Allostatic self-efficacy: A metacognitive theory of dyshomeostasis-induced fatigue and depression. Frontiers in Human Neuroscience, 10, 550.
Tappolet, C. (2000). Emotions et Valeurs. FeniXX.
Teroni, F. (2018). Emotionally charged—The puzzle of affective valence. In C. Tappolet, F. Teroni, & A. Konzelmann Ziv (Eds.), Shadows of the soul: Philosophical perspectives on negative emotions (pp. 10–19). New York: Routledge.
Teufel, C., & Fletcher, P. C. (2020). Forms of prediction in the nervous system. Nature Reviews Neuroscience, 21, 231–242.
Tschantz, A., Seth, A. K., & Buckley, C. L. (2020). Learning action-oriented models through active inference. PLoS Computational Biology, 16(4), e1007805.
Tye, M. (1995). Ten problems of consciousness: A representational theory of the phenomenal mind. Cambridge: MIT Press.
Tye, M. (2008). The experience of emotion: An intentionalist theory. Revue Internationale de Philosophie, 243(1), 25–50.
Uhelski, M. L., Davis, M. A., & Fuchs, P. N. (2012). Pain affect in the absence of pain sensation: Evidence of asomaesthesia after somatosensory cortex lesions in the rat. Pain, 153(4), 885–892.
Van De Cruys, S. (2017). Affective value in the predictive mind. In T. Metzinger & W. Wiese (Eds.), Philosophy and predictive processing. Frankfurt am Main: MIND Group.
Van de Cruys, S., & Wagemans, J. (2011). Putting reward in art: A tentative prediction error account of visual art. i-Perception, 2(9), 1035–1062.
Wiese, W. (2018). Experienced wholeness: Integrating insights from gestalt theory, cognitive neuroscience, and predictive processing. Cambridge: MIT Press.
Winkielman, P., & Berridge, K. C. (2004). Unconscious emotion. Current Directions in Psychological Science, 13(3), 120–123.
Winkielman, P., Berridge, K. C., & Wilbarger, J. L. (2005). Unconscious affective reactions to masked happy versus angry faces influence consumption behavior and judgments of value. Personality and Social Psychology Bulletin, 31(1), 121–135.
Acknowledgements
We would like to thank José Araya, Valérian Chambon, Andy Clark, Marco Inchingolo, Solène Le Bars, Elisabeth Pacherie, Agostino Pinnapintor, George Neish, Takuya Niikawa and Nura Sidarus for their helpful comments on previous drafts of this paper. We would also like to thank two anonymous reviewers for their insightful and constructive comments and suggestions, which helped us to significantly improve the paper. This work was supported by the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie Grant Agreement Number 675415; and the Agence Nationale de la Recherche under Grant Agreement Numbers ANR-17-EURE-0017 (FrontCog) and ANR-10-IDEX-0001-02 (PSL).
Author information
Authors and Affiliations
Corresponding authors
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Fernandez Velasco, P., Loev, S. Affective experience in the predictive mind: a review and new integrative account. Synthese 198, 10847–10882 (2021). https://doi.org/10.1007/s11229-020-02755-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11229-020-02755-4