Emotions – preamble to thinking about them ecologically

I’m thinking about emotion.

This is something I’ve done for a long time. What I really want is to move towards an ecological/dynamic view of them, which I have always thought was the thing to do. Emotion researchers have dipped their toes in it. I should dip some more, I think (when I have time to think).

Emotions are global events (I’ve had the hardest time thinking up a good term for this). The right stimulus (it could be external – William James’s bear, a lovely person, an angry face, a weird body sensation, etc etc etc), and there is a rapid organismic reorganization. There are cognitive effects – easier recall of emotional (often emotion congruent) information, redirection of attention, shifts in perception, altered judgments and persuasion, etc. There are physiological effects that are measurable in combinations of heart-beats, sweat, respiration, and pupil dilation among others. Visible physical changes – body posture, facial expressions, changes in vocalization. I sometime think of the organism (including brain and body), as a dynamic system on the edge of chaos, robust against a lot of impact, but can rapidly get rearranged into a different attractor when needed. An emotional event is that re-arranged attractor.

The complex of emotion – which is way more than just the conscious experience of it (which seems to vary individually) has an orienting function for the organism, I sometimes starts lectures about emotion with chemotaxis – the mechanisms bacteria possess that helps them move towards good stuff (light, food) and away from bad stuff (poisons) as a kind of proto emotional mechanism. The Approach/avoidance tendency is probably quite basic. There are claims of brain systems dealing with this (BIS & BAS – also shows up within motivation). Richie Davidson spent a great deal of work just looking at approach/avoidance and their signature in frontal EEG activation. If you look at any aspect of emotion – emotional expression, emotion words, emotion experience, you always find a valence dimension. The positive and the negative. Feeling good or bad. Positive or negative experssions, Positive or negative moods. Things are good for you or bad for you. (There is a rotated version, putting positivity and negativity on orthogonal dimensions, with the poles being presence and absence from Watson & Tellegen, but it is basically quite similar)
So, whatever they are, they are (likely) evolutionarily old, built up on tricks and mechanisms for orienting in the environment over eons of evolution. You study emotions in other mammals (although I completely agree with Joseph LeDoux that we should call what we are looking at survival circuits rather than emotion circuits, in order to not get entirely tangled in the folk psychology – both philosophical and in general – of the concept of Emotion). But, they seem to say something about emotion in human beings also.

To go with the survival circuit notion, the different emotions have functions necessary for survival and thriving. One can think of them as attuned both to the state of the organism (they are a part of the state of the organism) and the state of the surroundings. That is not at all new. I think Descartes had a similar idea (although he thought of it more as a connection between body and soul, if I recall – don’t feel like Whipping out my Oatley right now), but it is certainly the point of the appraisal theorists.

Emotion – expression
I continuously find myself working on facial emotional expression. It wasn’t quite what I planned to do. I wanted to do emotion. Especially what emotional states do, and how emotional states begin and develop and subside, and what happens as they develop. But, I did that with expressive faces. Not really successfully, until we started looking at ambiguous expressions.

Faces, in humans (and possibly in other species too) are communicative, as are bodies, and voices, and touch. We can discern emotion in several modalities (not sure about taste or smell, although there is likely something there too). It appears to be automatically bi-directional. Getting ones emotions activated results in (imperfect) readouts in facial expressions, body postures and vocalization. We also tend to mimic others individuals expressions (in fact it seems just about impossible to truly suppress it, based on Dimberg’s research). Posing our face in expressions (as well as our bodies) evokes that emotional state in us, albeit in a weaker form. (No, smiling is not going to defeat my depression, although there are times I think that slugging persons suggesting I should would bring some grim satisfaction. Long time since that happened, though.) This body of research has been going on for a long time. Facial feedback was pursued by Robert Zajonc, and by Ekman (along with Levenson), and has been looked at by Dimberg, and this long before embodied research became fashionable.

When I started looking at emotions, early 90’s, the debate seemed to be whether emotions and emotional expressions should be conceptualized as more or less universal categories, or more as falling in a space specified by a valence and an arousal dimension (sometimes a dominance dimension also). A great deal of this – when it came to faces, were between Paul Ekman and James Russell. (I briefly mentioned that this supposedly has been quite a brawl, enough that even I have sensed the fury of it, despite me only having access to the journals and their writing. I think I can see its aftermath still). Face stimuli were static photographs. The set I used was developed by Paula Niedenthal, where she had actors being read into the emotion before taking the picture, in order to make the expression more genuine and looking less posed. (I think they were successful). The theoretical conceptualization we used was the one of categories. Not sharply demarcated but fuzzy boundaried categories. This is very much in the Ekman tradition, where he talks about universal expression, and universal recognition (which is also in the Darwinian tradition). The expressions we had for this particular data-set was happy, sad, angry, fearful and neutral. (Perhaps disgust. No Surprise, if I recall).

The idea is that these emotions have different functions. There are cultural issues about how they emerge, but let’s not quite muddy the discussion with that (yet). The expressions have meaning. They signal to others what we are feeling, and we also interpret expressions as signals of that emotion – when asked. A person that looks sad is sad. A person that looks happy is happy. A person who ought to be sad but doesn’t look it (or looks happy) is weird and odd and wrong.

But, it is not mandatory, the connection between experienced internal emotion and facial expression (or other modalities most likely). We do spend time controlling our faces, attempting to moderate what we are signaling to others, possibly to not appear weird, or to conform, or to gain advantage. (As an aside, I notice my kids playing with faking emotions, or hiding emotions. They are not necessarily so good at hiding them – like when they try to lie. Their true emotions seem to be very open and bright. But, I gather that this is not unusual, and I wonder if this is practice to be able to manage their expressiveness, in the way we do expect adults to manage theirs. I recall also, from when I was 12, feeling distressed that I could not mask what I felt, and how vulnerable it made me feel. Especially when either a) I could not stop crying in rage over being humiliated or b) smiling and giggling uncontrollably when someone I had a crush on was near).

Dimensions
Some of what I describe above is likely also applicable to the dimensional account of emotions, promoted by Russell as an alternative to the categorical/universal account by Ekman, and many of his papers really were positioned as alternatives to the Categorical account. Russell was not first with the dimensional account. In the literature it emerge withWoodworth & Schlossberg on facial expression in the 40’s and 50’s, and Osgood devised another version (possibly not in the realm of expressions). A dimensional account also emerged from Peter Lang’s lab (he of the IAPS), and I recall reading a paper in the mid-nineties that these dimensions were what fell out of his data. The Watson and Tellegen dimensional account came out of clinical research, although the axes there are rotated. The two main dimesions (with the exception of W & T) are valence and arousal. A good/bad dimension and an activation dimension. The facial expressions then falls in a circular (circumplex) shape on the plane where the x-axis usually is valence, and the y-axis is arousal. In his critique of Ekman’s categorical account, Russell demonstrated that there were quite a bit of overlap between the categories, if you started pulling together the anthropological record. There was also the fact that you forced the categories by giving people the labels or stories to match (although the categories also emerge from free labeling, and from free sorting). In addition, He (along with Beverly Fehr) showed that when you rate a series of expressive faces, they are not independent of one another. You will judge the next face in the context of the previous expression, and this can push a judgment quite a bit away from what is seen as the stable, focal expression. He also demonstrated that there is quite a bit of learning going on, as kids develop. There is no stable, sprung out of the womb recognition of expressions. The learning emerges (but, he left it unclear what was driving this anyway). His view was that these dimensions were the primitives, and the other specifics emerged out of those.

I think the dimensional account has an important point, but there were always aspects of the research that didn’t seem terribly convincing to me. Like it just did never hit on the explanation that trumped the categorical account. But, neither, really, did the categorical. It seemed to really depend on what you were looking for and in what way.
I do have a slight preference for the more categorical, although I admit there is possibly a biasing aspect from my past in there. (But, I have used both categorical and dimensional measures in my research). That is because, as Paula pointed out even before I started researching, the different emotions have specific meaning and function. This is especially clear among the negative emotions: Being angry is not the same as being sad, is not the same as being fearful, is not the same as being disgusted (although anger and disgust can be conflated, and apparently there are conditions such as Huntington that specifically seems to create an inability to detect disgust, or possibly creates a conflation between the two). Looking at an angry face is likely to elicit a different response than looking at a sad or fearful face. Fear, anger, and disgust also tend to be rather high arousal, but seem to do different things. Anger is an approach emotion, whereas fear is more of a withdrawal emotion.

Photographs and dynamics
Apart from the dimension/categories controversy was the other issue that the expressions were always shown in photographs, and just about always as posed, high-intensity expressions. In real life, expressions are ongoing and dynamic, and usually not high-intensity. This was an old critique, which Ekman Friesen and Ellsworth responded to in their seminal book with a, yes, we know, but honestly, we don’t even know if there is anything stable in our emotion recognition, so let’s go simple, clear and unambiguous first to see if we can.

By now we have a lot of information about processing the static expressive face both in regular people and in different clinical populations, as well as across the planet, because emotion research has been good at the cross-cultural question for a very long time.

In the late 90’s me and a fellow grad-student presented low-intensity expressive avatars to participants. Two frames right after one-another, sometimes with a noisy mask in between. In the first experiment the two frames showed a slightly different degree of intensity. In some trials there was the low followed by the slightly higher – with or without mask. In some it was the slightly higher followed by the low. Without a mask, you see an expression either emerging or waning. In the distrupted condition, you see two neutral faces. If you increase the intensity to where the expression is obvious, albeit not necessarily high intensity, movement doesn’t matter. This work will never see the light of day, as we basically were scooped. In many ways, it was an obvious experiment to do. Before that, we had used morphed expressions – dynamic again, to assess where people see expressions change, and if it differs depending on your emotional state, or other manipulations (it does). There is increasing work on the dynamic expression, because, well, as the early critique said, that is how we see it. But, there needs to be more.

Context
Right now, there is a renewed interest in context effects. There’s an excellent summary on how one should think of various context effects in this Frontiers paper. But, much of this renewed interest is due to findings that, in some instances (in real life) facial expressions are not as informative as, say, bodily expression. This is very much a second wave, as the influence of context on emotion-expression interpretation were looked at in the 80’s and 90’s. In fact, there was likely an older look at context, as one of the paradigms (Goodenough & Tinker) was taken from their 1931 paper. And, even earlier than that were Kuleshovs lost footage where he paired the neutral expression of a famed actor with different emotional scences, and claims that viewers expressed awe at the appropriate emotional expressiveness of that actor. This has frequently been cited as an existence proof of how context can influence interpretation of emotional expressions both in the psychology of emotion as well as in film proper (e.g. Hitchcock). We are working on re-examining this, as it has been woefully underexamined for something so influential. In light of this have been reading up on the context effect research again (the older corpus is quite small) and my take is that the early focus then was very much on finding emotionally unambiguous and strong stimuli to try to tease apart the relative importance of context (often in the form of short stories, but also in the form of other expressive faces), and emotional expression. (I actually had a couple of bachelors students replicate one of them, with rather interesting effects). The interest now is not so much in going for clean, but in pursuing the more messy, but more as in real life context effects, away from the stylized towards the more realistic.

I have been pursuing a slightly different angle than the above version, but clearly it falls within context effects.
I usually take this example (picking on my former doctoral student Farhan – he’s from Pakistan). Imagine meeting him, day time, at the university, slightly knit brow. No big deal. Now imagine meeting him after dark, in one of the dodgier areas of Malmö, same knit brow expression. Well. Of course, I know Farhan, so that would more signal the possibility of letting my guard down (yes, a familiar face – now we are two. Wonder what he is thinking about). But, if I didn’t?

Now, imagine it was not Farhan but Erik. Or, imagine a 6 year old Farhan (or Erik) rather than somewhere between early 20’s and early 40’s. Or, 60+. Would things change if it wasn’t Farhan or Erik, but Fatumeh or Ellen? Also, does it matter if you are in a good or bad mood? If you yourself are male or female, belong to majority or minority group? If you have never met people who are called Farhan or Fatumeh (but plenty of Erik’s and Ellen’s), or that your surroundings include plenty of both kinds (and more)? And, the answer to all these questions are, well, yes. I have some data. Others have data too. I think there is more that needs to be done because the effect sizes are smallish. Mostly these are done in labs, in front of a computer monitor, with static images (some exceptions).
But, I’m unhappy about the conceptualizations.

And, it is here I want to come in with Affordances and ecological psychology. I’m far from first thinking that ecological psychology has a place in the more social realm. Leslie Zebrowitz have had that idea for social psychology for a while, and I first came across using affordances in a social psychological manner, especially for facial expressions, from Social Evolutionary psychologists like Kenrick and Neuberg, addressing questions very similar to those I summarized right before (In fact, I came across their research when trying to understand my own results). .

Affordances?
How to analyze this from the questions posted on Andrew & Sabrina’s blog…

1. What is the task to be solved?
2. What are the resources available to solve the task?
3. How might these resources be assembled so as to actually solve the task?
4. Does the organism actually do what your described in 3?

I think I immediately come into issues with the massiveness of emotions. It is an interesting supercategory. I think our intuitions is that feelings and emotions are a category or belong together somehow (phenomena that change the shape of the strange attractor temporarily?). But, the individual emotions (if you think from a categorical perspective) seem to solve very different kinds of tasks. And, in some instances perhaps solving multiple tasks.
But there is an evaluation, a valence perspective most clearly. They are involved in determining, Is it good or bad. In which way is it good or bad? Should I withdraw or approach, and if such, in which manner? Freezing, running away, moving towards for affiliation, moving towards to challenge. A (general) task to solve is then the need for any organism to know what is advantageous for survival, and what is disadvantageous for it, and to respond appropriately. Without needing unnecessary deliberation (the deliberation that humans are capable of seems more to be used for overriding other responses. Perhaps)

I just reminded myself of Frank Herbert’s game theoretical analysis of the function of emotion as commitment devices, at least in a social species like us. (Emotions/survival circuits clearly predates humans, and possibly even social species. Some of our social emotions also seem to exist in other social species where hierarchies are impoartant, but perhaps not in non-hiearchical ones. Me and my masters students amused ourselves with considering shame in cats).
So, I’m coming up against the problem of having multiple possible tasks that could be the one to focus on (while knowing a lot about emotion, and theories surrounding emotion).

But, perhaps the place I should start (where there is some precedence) is with the facial expression/physical expression, and the dynamic between perceiver, sender and context.

(I may go back and do better references for what I pulled together here. But, Oatley, Jenkins & Keltner’s textbook covers a lot of it in a secondary fashion)

Advertisements

About asehelene

... because if I'm in a room with a second person, I want to be reasonably sure I'm the crazier one.
This entry was posted in Emotion and tagged . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s