Tuesday, 22 September 2015

Emotions as Psychological Reactions

This post is by Edoardo Zamuner (pictured above), a senior research fellow in the School of Psychology of the University of Auckland, New Zealand. He previously held research and teaching positions at University College London, La Trobe University (Melbourne) and the University of Hong Kong. While he is a philosopher by training, he is a psychologist by trade and persuasion. His current research in psychology focuses on visual perception of faces and facial properties such as gender, expression and personality. His work in philosophy of mind concerns the emotions. Here he summarises his recent paper ‘Emotions as Psychological Reactions’, published in Mind and Language. 

What kinds of mental states are emotions? My paper argues for the view that emotions are reactions to our experience and thoughts, broadly construed. But what exactly is a reaction? And why should we think of emotions as reactions? Talk of reactions is a reflection of everyday causal explanations, just as talk of cause and effect. It would, however, be a mistake to equate reactions and effects. As I show in my paper, we think of reactions as effects with polarity. For example, we view changes such as worsening and improvement as polarly opposed reactions to a treatment. If emotions are reactions, then they have polarity. But what exactly is the polarity of emotions? The paper’s main thesis claims that we can characterise emotional polarity in terms of the distinction between approach and avoidance.

Specifically, emotions come with the tendency to engage with (approach), or withdraw from (avoidance), the objects and situations in reaction to which emotions occur. This tendency can express itself behaviourally, like when anxiety results in procrastination (avoidance), or psychologically, like when joy results in anticipation (approach).

Two main objections can be raised against this thesis. First, emotional behaviours such as freezing in fear and sulking in resentment do not seem to come with any tendency to approach or avoid the objects of fear and resentment. Second, the emotional syndrome commonly known as brooding does not seem to dispose one to engage with, or withdraw from, the object of one’s emotion. My replies involve showing that both objections fail to acknowledge the polarity of the emotions that underlie the phenomena under consideration.

Freezing is known to occur when movement might provoke further attacks from a predator or assailant. This suggests that the emotion of which freezing is an expression comes with the tendency to avoid damage. Sulking, on the other hand, is a behavioural strategy in which seemingly non-relational attitudes (e.g. grudging) and signals (e.g. ‘yes-or-no’ answers) aim to negatively affect another party in a relationship. Since affecting others is a way of indirectly engaging with them, there is a sense in which the emotion that leads to sulking involves the tendency to approach.

Similar considerations apply to brooding. This is an affective state in which lack of action often translates into rumination and obsessive thinking about something that makes one unhappy, angry or worried. As I noted earlier, an emotion’s tendency to approach or avoid its object can express itself not only behaviourally but also psychologically. Brooding seems a case in point since the person who broods over an issue imaginatively engages with it. We may thus view brooding as an emotional syndrome that comes with the tendency to approach and engage with thoughts and images that represent the object of one’s sadness, anger or worry.

At this point, one may wonder about the theoretical gain of the view that emotions are reactions. The answer lies in the explanatory power of polarity. Most notably, polarity can explain some aspects of the intentionality of emotion, understood as the way in which a person views the world as a result of her being in the grip of an emotion. In recent years, new accounts of the intentionality of bodily sensations have been proposed. They begin with the observation that certain bodily sensations tell us what to do or not to do. Tickles, for example, tell us to laugh. Pain, on the other hand, tells us to avoid those movements that would make it worse.

Drawing on this observation, some theorists argue that bodily sensations are part descriptive and part prescriptive. The descriptive part represents a region of one’s body as being a certain way, i.e. as being titillated in the case of tickles or damaged in the case of pains. The prescriptive part, on the other hand, involves a command to do or not to do something. Thus, tickles involve the imperative ‘Laugh!’ and pains involve the negative imperative ‘Don’t do it!’.

This account seems particularly well suited to explain some aspects of the intentionality of emotions such as disgust, surprise and fear. Suppose you experience disgust upon finding a dead bird on your balcony. Besides representing the dead bird as repulsive, your emotion commands you to avoid or stay away from the carcass. It is in this sense that polarity can help explain certain aspects of the intentionality of emotions. Since polarity defines reactions, the claim that emotions are psychological reactions is both plausible and significant.


  1. Hi Edoardo,

    Thanks for the post! I'm curious what you think about other emotions that challenge analyses of valence in terms of approach and avoidance. Lots of emotions have valences that don't map onto their approach/avoidance profiles (although perhaps your notion of polarity is not supposed to capture the notion of valence). For instance: anger, envy, or surprise.

    These sometimes elicit approach and sometimes elicit avoidance and sometimes don't elicit any sort of behavior. They are mixed. (Sadness and grief look mixed on your account as well, given that they cause avoidance of the painful stimulus but brooding and sulking are approach behaviors.) However, if emotions are reactions, I don't think you can allow emotions to have mixed polarity. Reactions by your definition have either positive or negative polarity. It seems to me that a problem with thinking that emotions are reactions is that emotions have *many* behavioral consequences and action tendencies. For example, fear commands attention. When you're scared of an oncoming dog, you open your eyes wide and find it difficult to attend to anything else. Is attentional capture a form of approach? This seems no less metaphorical than brooding being a form of approach. Does this make fear an emotion with positive polarity? I note that you switch in your post from talking about emotions having polarity to talking about emotional behaviors like brooding having polarity. Is polarity a feature of emotions, or of emotional behaviors?

    I'm also curious how you characterize approach and avoidance. Saying that brooding and sulking are forms of approach is pretty counterintuitive to me, and I can only understand the claim as being metaphorical. The problem is that nearly any individual emotional behavior can be metaphorically interpreted as a form of approach or metaphorically interpreted as a form of avoidance. For instance, male rats soil themselves when a cat scares them. Is this a way of driving off the cat (so a kind of avoidance) or a way of negatively interacting with the cat (so a kind of approach, like sulking)? I don't really know how to handle these kinds of questions on your account... either answer seems acceptable.

    1. [I'm posting this on Edoardo's behalf.]

      Hi Richard,

      Thanks for your questions. Here are my replies.


      Polarity and valence are different notions in my account. I take valence to be a feature of the phenomenology of emotion. I extensively discuss this issue in my article.

      It may be helpful to think of polarity in the following terms. The representation “x is dangerous” can be the content of the mode "fear-with-approach" or the mode "fear-with-avoidance", hence the idea that polarity may explain aspects of the phenomenology and behavioural expression of emotion.

      Panic and stage fright may be viewed as examples of fear-with-avoidance. On the other hand, the thrill and suspense someone experiences when watching a horror movie or going on a rollercoaster ride may be viewed as examples of fear-with-approach. Similarly, disgust upon seeing a dead animal comes with a tendency to avoid, whereas the type disgust we experience when taking a second look at the carcass, almost with curiosity, that type of disgust comes with a tendency to approach.


      I am open to the possibility that some emotions may have mixed polarity. If we think of grief as an emotion that unfolds over time and goes through stages, we allow for the possibility that different stages may have different polarities. During the early stages of grief, people often avoid dwelling on memories of the deceased. Later in the process, however, they become less avoidant and the same memories can be a source of comfort. It is then that grief involves a tendency to imaginatively approach and engage with its object.

      My account also allows for the possibility that different tokens of the same emotion type can have different polarities. Some tokens of fear involve avoidance; others involve approach. As noted above, panic and stage fright may be viewed as examples of fear-with-avoidance, whereas thrill seekers and adrenaline junkies may be viewed as people who experience a type of fear that motivates them to approach frightening situations.


      I do not switch from talking about emotions having polarity to talking about emotional behaviours like brooding having polarity. I explicitly say that brooding is “an affective state in which lack of action translates into rumination and obsessive thinking about something that makes one unhappy, angry, or worried.”


      My use of the terms “approach” and “avoidance” is not meant to be metaphorical. For example, in the case of brooding over x, I characterise approach as a form of imaginative engagement with x. The brooding person broods because she obsessively rehearses thoughts and images that are about, directed at or make reference to the object of her emotional state. These are ways for the brooding person to imaginatively approach the object of her emotion.

      In your example of the male rats, we need to consider the function of the behavioural reflex, like in the case of freezing. My guess is that it functions to elicit disgust in the predator. This seems a form of approach in which the rat engages with the predator through a behaviour, expression or reflex that serves the function of negatively affecting the predator.

  2. Thanks for the responses, Edoardo! Cheers!

    Would you allow for individual emotional episodes tokens that have mixed polarity at a single time? A fear episode, for instance, might simultaneously have two different consequences in action: the subject will attempt to escape the target of the fear (avoidance), and at the same time, the subject will engage with the target in attention and thought (approach).

    1. [I'm posting this on Edoardo's behalf.]

      According to my account, polarity is a feature of the intentionality of emotion. This account claims that the representational content of emotion involves two components: descriptions and imperatives. I take polarity to be an aspect of the imperative side of the content. With this picture in place, I propose the following analogy. Just as you can give complex orders that command both approach and avoidance (e.g., "Fight but don't get killed"), so we can have emotions the representational content of which involves complex imperatives that command both approach and avoidance (e.g., "Run and stay focused", "Run and don't look away" or "Run and don't get distracted", to use your example). The emotion you describe might have this kind of imperative content. While the imperative "avoid!" partly explains the agent's behaviour, the imperative "approach!" determines the way in which the agent's mind is directed at the object of her emotion. This, in turn, partly depends on her being compelled to keep her attentional focus on that object.


Comments are moderated.