Wednesday, October 26, 2016

Secret Science Club Post-Lecture Recap: Snap Judgments

Last night, I headed down to the beautiful Bell House, in the Gowanus section of Brooklyn, for this month's Secret Science Club lecture, featuring Dr Jon Freeman, Director of NYU's Social Cognitive & Neural Sciences Lab.

Dr Freeman kicked off his lecture by noting the importance of transferring information to the general public, especially in light of public funding for research. He then launched into an overview of evolution- the brain evolved to solve certain problems. The most important function of the human brain is getting into others' minds... others' behavior, thoughts, and feelings are complex, and figuring them out is important. Minds are invisible, they can't be directly observed, so any information about them must be gleaned from what information is available, most importantly from facial cues. Facial recognition is important to perceivers- who is a particular individual? Is a particular person a stranger? Evolutionarily, it is critical to determine whether another is friend or foe because of the threat of coalitional aggression. It is important to recognize threats or competition for resources.

Infants have an early ability to recognize faces, and humans have a tendency to see faces where they don't exist. The ability to divine emotional expression is important- is an individual appeasing or posing a threat? An expression of fear is a signal of the likelihood of an impending threat, the recognition of which has an important outcome.

In 1872, Charles Darwin wrote The Expression of the Emotions in Man and Animals, in which he compared human facial expressions to those of other mammals. Paul Ekman built on Darwin's work on 'mapping' emotions and their corresponding facial expressions. He theorized that expressions for fear, anger, joy, sadness, and surprise are genetically hardwired and are spontaneously expressed and recognized. The idea that the expression of emotion is hardwired is an enduring idea.

Robin Dunbar's Social Brain Hypothesis (PDF) turned a lot of thought on brain development on its head, so to speak. Human brains possess a huge neocortical complex. Dunbar noted that the size of an animal's neocortex correlates with the mean social group size of the species. Originally, it was thought that brain size was driven by the development of general intelligence, and social complexity 'spilled over' from general intelligence. In Dunbar's model, the development of social intelligence 'spilled over' into general intelligence.

What are the brain mechanisms for snap judgments? Most of the consequential decisions we make are snap judgments- dating, hiring, electing presidents... In order to determine these mechanisms, Dr Freeman indicated that functional brain imaging, computer mouse tracking, electrophysiology, and computational modelling.

When an individual makes a snap judgment about something, there are typically two dimensions- what is that person's intention and what is that person's ability? Is an individual's intention to be friendly or hostile? Does an individual have the ability to be helpful? Is an individual dominant or not? Is an individual competent or not? Dr Freeman displayed a graph with ability as one axis and intention as the other, and noted that humans have complex emotions regarding the placement of individuals along these axes- trust, disgust, pity. Facial models tend to be judged by subjects using cues- trustworthiness decreases as a facial model frowns. Judgments of trustworthiness subtly overlaps with emotional expression- trustworthiness tends to correlate with joy, untrustworthiness with anger. There is an amazing consistency with subjects' judgments, but are we accurate? Gauging accuracy is difficult even though people demonstrate consistency of judgment. These perceptions matter! How competent a face appears is a predictor of elections. The attractiveness of a face has many consequences. The trustworthiness of a defendant's face is predictive of trial outcomes.

One controversial area of study involves measurement of one's bizygomatic width to determine formidability, it being thought that one's testosterone levels during puberty might predict aggressive behaviors. Do hockey players with a certain facial type spend more time in the penalty box? Both male and female subjects tend to 'puff up' in the lab to express aggression. The 'trustworthiness' markers of an individual's face are the brows and cheekbones.

Much research has been done on how quickly the brain can judge facial expressions and the neural pathways involved in the process. The amygdalae are the structures in the brain which drive rapid responses- Dr Freeman quipped that the amygdalae are fast-acting but 'dumb'. The neural pathway for a quick response involves the amygdalae and the thalamus, using a sub-cortical route. In the longer neural pathway for responses, visual stimuli traveling to the visual cortex in the back of the brain for processing. This long pathway is also quick, even without the shortcut. Functional imaging can help researchers determine these pathways- using MRI, researchers can determine which neurons activate because associated blood vessels dilate in order to provide oxygen and glucose. This hemodynamic response can be tracked to determine brain function. The amygdala was the subject of an 2010 SSC lecture by NYU's Dr Joseph LeDoux... I really stepped up my lecture writeups over the intervening years.

In one experiment, subjects were presented with computer-generated images of faces, with real faces interspersed for the sake of comparison. The subjects were unaware of the faces due to backward msking, these faces were displayed for thirty-three milliseconds, and the images were 'masked' by a display of another image for one-hundred and sixty-seven milliseconds to terminate the visual processing. This was done to determine if subjects needed conscious awareness to perceive trustworthiness. The amygdalae showed a lot of responses to stimuli, and the brain could process untrustworthiness even in cases in which there was no awareness that a face had been shown. This process is beyond our conscious control.

This is not the end of the story- there is a certain degree of malleability as subjects get to know individuals and actively judge faces. In one study, subjects were given prior knowledge about a depicted individual- a label indicating that an individual 'left clothing in the dryer' would have an effect on perception while a neutral bit of information 'the tide is usually low that time of day' would not. If context matters, can it change responses? If subjects aren't allowed to get to know individuals, the amygdala plays a greater role. If a subject is given information about an individual, there is a greater neocortical response involved, with the tempoparietal junction playing a major role.

One problem in processing facial expressions is that we don't have a single image of ourselves. Every aspiring actor or model would be able to tell you that headshots communicate something, but are there some dimensions of social perspectives that are immutable? Trustworthiness is malleable, being related to emotional expressions. There are dynamic facial cues and static facial cues- Dr Freeman joked about resting bitch face and its counterpart, resting nice face. While the facial musculature can change these chronic expressions, the underlying skeletal cues don't change. Dr Freeman also mentioned the 'Bouncer Effect', whereby an individual spontaneously assumes a more dominant, tougher looking expression.

Dr Freeman then posed the question, is it a special process to stereotype another? Stereotyping is part of the general social process, the categorical structure which causes us to lump individuals together is adaptive, it streamlines cognitive resources and reinforces cultural associations. Can we act non-prejudicially in light of cultural associations? By measuring computer mouse movements, intentions can be revealed- the trajectory of movement of a mouse between two facial images (e.g. caring vs aggressive, black or white) can reveal implicit bias despite a subject's intentions. In the case of gender bias, female political candidates in conservative-leaning districts have to maintain a certain degree of conventional physical attractiveness in order to win elections, and more attractive candidates are generally seen as more competent.

Dr Freeman then asked, how can our biases change how we see others? Visual information 'fans out' from the visual cortex to other parts of the brain, from the back of the brain to the front- this information is colored by the higher functions of the neocortex, resulting in a compromise between actuality and expectations. The brain is 'too lazy' to do a complete job of processing visual information. He then presented us with this image:




Prior knowledge changes perceptions- the brain incorporates expectations into perception. One wouldn't expect to see an elephant in the beautiful Bell House... expectations form an adaptive context, in social context, stereotyping results. It has been theorized that a fusiform face area is involved in perceiving faces.

In one study to determine if perceptions can be changed with regard to stereotyping, a series of computer generated faces in a range of pigments was displayed. In the middle range, the appearance of the face was racially ambiguous- in these cases, the attire displayed in the image was an important cue. A racially ambiguous image of a man in a suit and tie was more apt to be deemed 'white' while a racially ambiguous image of a man in a mechanic's coverall was more apt to be deemed 'black'. The brain has expectations and will apply them to a face. Men's faces are more likely to be seen as angry, women's as appeasing. At what level are stimuli biased? White subjects are more likely to perceive African-Americans as hostile or angry.

Thankfully, the dorsolateral prefrontal cortex can correct snap judgments. Is ter a pattern code in the fusiform face area for emotion? By adding the dimension of race, can a face's anger or happiness be changed? Visual representation is biased by stereotypes but individual differences matter. Stronger stereotypes predict bias and similarities in the cortices of the brain- these are not hardwired, they depend on the stereotypes. This is an unfortunate consequence of a perceptual system which is meant to carve up a complex world. Dr Freeman asked, can we change vision to reduce bias? Mixed-race individuals form the fastest growing 'racial group' in the United States- how do we process ambiguity? In areas with low minority populations, such as Utah, racial breakdown of mixed-race individuals is unstable, and category shifts are easily achieved.

Experience matters in perception- Dr Freeman noted the split-second decision involved in deciding whether an individual is holding a gun or an innocuous object. If we acknowledge bias, what effect does this have on vision? Is it a wallet or a gun, and is a misperception a matter of black or white? Bias trickles down to effect vision. We automatically make a lot of snap judgments, often beyond our conscious control. Unconscious bias can alter visual 'reality'. We can change to some extent, this involves a shift to an evolutionarily more modern system.

While the bastard in the audience didn't get a question in during the formal Q&A, he did have a brief conversation with the good doctor after the lecture, in which he asked about our perception of deceptive or 'wrong' emotional cues- the insincere or even minatory smile, the false sorrowful face. Dr Freeman indicated that this would be an interesting matter of inquiry. All told, his lecture was thought-provoking, especially in light of police responses to African-American suspects versus white suspects. Amazing how such an interesting lecture can also be sobering.

Here's a quick video of Dr Freeman, explaining the process of making snap-judgments:





Pour yourself a libation and soak in that Secret Science Club experience. Once again, kudos to Dr Freeman, Dorian and Margaret, and the staff of the beautiful Bell House. High fives, all around, people!

2 comments:

Smut Clyde said...

In Dunbar's model, the development of social intelligence 'spilled over' into general intelligence.

Adrian Desmond made the same point in "The Ape's Reflection" (1979), and he didn't present it as a particularly novel insight.

ifthethunderdontgetya™³²®© said...

I assume "the bouncer effect" when Nutkin comes into the yard.

But neither she nor he gives a hoot.
~