I am going to try and play with a few ideas in this post, but big picture, at their nexus lies a simple question: Do emotions have affordances, and if so, what are they? So, let’s start with affordances, but in a round-about way.
Imagine that you have been placed in a room. Two strings hang down from the ceiling a certain distance apart. In the room is also a chair and a pair of scissors. Your captor tells you that if you can grab ahold of both strings, they will let you free. The only problem is that the strings are positioned such that if you clutch one of the strings and walk towards the other, that other string remains just out of reach. What do you do?
Here’s another one. Below are nine dots arranged in a square. Your task is to take a pen or pencil (or use your mouse pad) and draw no more than 4 lines that connect all of the dots.
Or here is one more. You are a chimp living at the Anthropoid Station in Tenerife during WWI. A German researcher named Wolfgang Koehler has suspended a treat just out of your reach. Sometimes the treat is dangling up in the air, and crates are strewn about your enclosure. Sometimes the treat is on the ground beyond the enclosure’s fencing. In these cases there are sticks available. What do you do?
These types of problems are known as “insight problems.” Insight problems tend to have what is called an “ah-ha” moment. The person doing the puzzle repeatedly bangs their head against a wall, until suddenly the solution emerges as if in one conceived whole. The chimp might reach out towards the suspended treat. It might become visibly frustrated and hurl crates about the enclosure. What Wolfgang Koehler observed, though, was that sometimes after these displays, a chimp might pause and look back and forth between the crate (or stick) and the treat. After this pause, the chimp would suddenly go over to the crate, drag it under the treat, climb up, and pull the treat in. Or the chimp would grab the stick, reach out, and pull the treat into the enclosure. The learning (i.e., achievement of the desired outcome) was not gradual, but sudden. Take the problem of connecting the dots, remember that no one said that your lines couldn’t extend outside of the implied boundaries of the square. Or take the problem of the strings, no one said that a pair of scissors has to be used for cutting. Scissor handles can be tied to strings and used as a weight to start the string swinging.
There is a lot that could be said about insight problems and insight learning, levels of selection and units of selection. I wrote an entire master’s thesis-equivalent on the topic! And that was many, many years ago prior to much additional research that has come out since. For the purposes of this post, though, let’s just appreciate that one common aspect of insight problems is something known as “functional fixedness.” This is the idea that certain objects and situations “suggest” certain actions / behaviors. These suggestions can be more or less “strong.” We see a pair of scissors and the suggestion is, “Scissors are for cutting!” We see nine dots arranged in a square, and we assume that the implicit edges of the arranged dots are a boundary. For chimps, crates aren’t for dragging. Maybe for climbing on top of, but for dragging?? Really? And sticks. They are for poking. Maybe scratching. But for extending one’s reach in a generic sense? For Koehler, who belonged to a school of psychology known as the “gestalt psychology,” the rules (what I have termed “beliefs” in earlier posts) that constructed perceptions were central to how the environment made available different courses of action.
The functional fixedness that defines many types of insight problems essentially says that sometimes perception very strongly suggests particular actions. We see a button, we press it. We see a chair, we sit on it. We see a book, we open it (or throw it across the room). We see chalk and a chalk board, we write or draw on the board. Our perception isn’t just a description of the world, but also a call to action. It is this connection between perception and action that lies at the heart of the concept of affordances.
Or let’s come at this a different way by considering reflexes. We can think of reflexes as a particularly simple call to action, although reflexes, modal action patterns, and reflex chains can be highly complex. In this case were are not talking about “emitted behavior” or skills and habits that we have learned through trial and error, but rather “elicited behavior.” The suggestion from perception is an imperative: thou shalt use the scissors to cut! Only it’s not scissors but puffs of air to the eye, taps to the patella, the smell (to a male dog) of a female dog in heat, a viral invasion of a body, memories of a traumatic event, or the soft cloth to the motherless monkey… In each of these cases, perception mandates action, which is to say that the action feels involuntary. The eye blinks. The foot shoots up. The male dog jumps the fence and pursues the female. The immune system responds. The skin conductance goes up and the pituitary gland releases hormones. The soft texture evokes attachment / feelings of comfort and safety. And yes, emotions can be as much of a response as the blink of an eye – i.e., emotional responses are contained in an animal’s “affordances” as much as a stick may be used for poking and scratching and reaching and throwing and whittling. Perception suggests these behaviors, but also the behaviors of anger, contentment, sadness, stress, and humor. Emotions are responses and therefore suggestions for action made by our perceptions.
Zooming back out, though, the point here is that perception is not just a description of a world (and remember, exteroception perceives the world outside the body, while interoception perceives the world OF the body). Perception is also a weighted suggestion of all sorts of an actions — suggestions with differences in probability, urgency, intensity, and obviousness. Take reflexes. It might be reasonable to assume that our strongest call to action will be elicited by naturally occurring stimuli. A parental gull returns from foraging. The gull chicks peck at the red dot at the base of the parent’s yellow beak. The parent regurgitates food for the chicks to eat. Red dot on a yellow beak suggests pecking, or since we are talking about a reflex, red dot on yellow beak compels pecking in the hungry chick. Except it turns out that there are other perceptions that compel even more pecking. Affordances can be hacked. It turns out that more compelling than the naturally occurring red dot at the base of a parent’s yellow beak are several red lines drawn at the tip of a yellow rod. This, this is the better parent, at least if determined by the vigor of pecking made by the hungry gull chicks (Tinbergen & Perdeck).
The finding that the “best” stimulus for eliciting a reflex might not be a naturally occurring stimulus is captured by the idea of a supernormal stimulus. Basically, a supernormal stimulus is to reflexes what functional fixedness is to insight learning. Glue a long, white feather to a male grass finch’s head, and it becomes very attractive to females (Burley & Symanski). Provide a nesting oyster catcher with an unnaturally large egg, and it becomes the preferred egg to brood. Provide a female platyfish with a male to which you’ve affixed an artificial “sword tail,” and that males become highly desirable, compared to the normally swordless males (Basolo). There are many examples of supernormal stimuli that have been researched, and like visual illusions they tell us something about how our minds construct their beliefs about the world. In the case of affordances – that recursive network of perceptions to actions — whether we are aware of the whispered suggestions or not, our minds are an ecosystem of hows-to-engage.
Now, though, I would like to pivot. Normally, affordances are thought of in terms of exteroceptive perception, or perception of the world “out there.” A chair, the smell of baked bread, an approaching bear “suggests” certain types of behavior… and remember, emotions are a behavior. But we also have perceptions of the world “in here.” That it, we have feelings. Emotions in other words are both a suggested action, but also a perception. So, a reasonable question is what sorts of affordances are attached to these interoceptive perceptions. What are the affordances of anger, of delight, of boredom, of hunger, of body aches and a runny nose? I am going to argue for a couple of possibilities. First interoceptive states / perceptions broadly structure sets of behavior into behavior systems. Second, interoceptive states suggest certain types of memory recall (and yes, “remembering” is an action). Third, interoceptive states are fundamentally responsible for turning exteroceptive description into interpretation. That is, just as our brains actively fill in the blind spots of our exteroceptive, visual world, one of the actions called up by our interoceptive states is projection. We’ll get to all of these in the next blog post.