Making sense of making faces
Interview: Dr Lisa DeBruine and Dr Rachael Jack (University of Glasgow)…
Making sense of making faces
Face Facts is an innovative project set up by psychologists Dr Lisa DeBruine and Dr Rachael Jack of Glasgow University to reach out to the general public and share current thinking about how we form our opinions about people based on their face shape and colour and facial expressions. DeBruine's research focuses on kin recognition, facial resemblance and face perception – how humans use facial resemblance to tell who their kin are and how the visual system learns about faces. Jack is interested in human social communication – how people transmit and decode signals (e.g., facial expressions) for social interaction – with a particular focus on cross-cultural communication. Her approach is interdisciplinary, combining psychophysics, social psychology, information theory and social robotics.
Face to face with psychologists Dr Lisa DeBruine and Dr Rachael Jack of Glasgow University, you may feel they are analysing every single feature and every expression you make, for signs of hidden messages or even a family connection. But according to Jack and DeBruine, you needn't be concerned – we actually make up our minds about people without even thinking, just by “reading” their faces.
The first few seconds may be critical when you're deciding to buy a new home, but most of us usually take just a few milliseconds to decide how we feel about people, “without even making an effort,” says Jack. When we look at someone's face, the morphology (form, structure, size and shape), complexion (the colour and texture of the skin) and facial expressions all send out various signals, she explains. For example, from face shape and colour, we see ethnicity, age and gender, and state of health. We also decide social status and sexual orientation, and weigh up potential sexual partners.
We are used to the expression “body language,” but we also use our facial muscles to communicate what we are thinking or feeling; for example, we scrunch up our noses to indicate disgust. And when it comes to interpreting facial expressions, some of us are more “literate” than others. It’s useful to make up our minds about people's intentions (whether someone is threatening or friendly, bored or interested), but sometimes we jump to conclusions or make simple errors, based on the appearance of someone’s face.
Some of the conclusions we reach about people are common to all human beings, says Jack, but some of them differ from place to place, and can subconsciously influence our judgement. There are also cultural “rules” about faces and facial behaviour which become second nature. For example, in some cultures too much eye contact is considered rude, while others think avoiding it is rude. As Jack describes it, people engage in an “eye gaze dance” when they meet, and this sends out a lot of complex signals regarding our status (e.g., superior “looking down on” inferior) as well as our feelings. Sometimes we “second guess” what signals mean, but some characteristics can provoke completely opposite responses. For example, having a suntan is nowadays thought of as desirable and healthy in most Western countries, but used to be associated with “lower-class” people who worked in the open.
Sexual attraction is one of the most complex aspects of facial perception. As women’s hormones change at different stages of the menstrual cycle, there is evidence they are attracted to more masculine or feminine partners. People also view potential sexual partners aslong-term or just one-night stands.
Our facial responses to environment are also important; for example, to signal disgust, fear or danger. And these expressions have evolved on the same path as language – just as grunts become words, so our facial expressions have become more complex over the millennia. “Our facial expressions have their own grammar and syntax,” says Jack.
In recent years, Jack’s research has focused on cultural factors, questioning orthodox thinking on social interaction and facial expressions. Earlier theories (proposed by Ekman in the late 1960s) suggested there were six “universal” facial expressions which all cultures shared, but Jack’s research has challenged this view. According to Jack, the earlier research was more “Western-centric,” but by using a more agnostic, data-driven approach, she has shown clear cultural differences in the facial expressions once considered universal, and that four (not six) facial expression patterns are common across cultures. To do this, she has taken full advantage of a new kind of software which generates a range of random dynamic facial expressions. She then asks people of different cultures which emotion they see in each facial expression: happy, surprised, fear, disgust, anger or sad. By doing so, she can identify the specific dynamic facial expression patterns that individuals in a given culture associate with different emotion categories. She has also recently extended her work to 60 emotions across two cultures – a first in the field. “This new technology makes a huge difference to the scope of our research,” says Jack, “but the most important thing was taking an agnostic approach, to avoid preconceptions about which facial expressions communicate which emotions.” Another major difference was that Jack did not average out the results across participants, which would eliminate individual nuances in face signalling, but focused on the responses of individuals within each culture to understand variance within a population. Thus, Jack’s approach of using psychophysics allows her to arrive at more objective conclusions about these subjective emotions.
For Jack, there were two major breakthroughs involved in the process. First, she realised that the differences between different cultures were masked in the early research, which meant an opportunity to do some new research, taking advantage of the new technology available, including “generative face grammar” software developed by her colleagues at the Institute of Neuroscience and Psychology in Glasgow, Professor Philippe Schyns, Dr Oliver Garrod and Dr Hui Yu (now at the University of Portsmouth). The second breakthrough was that she discovered there are only four facial expression patterns that are common across cultures. “I didn’t set out to get four,” she explains, “but that's what I found in my data across several studies.”
So why does this matter? Apart from learning more about how human beings operate, this knowledge could be useful in designing a new generation of digital tools – for example, what are called “companion robots,” or digital avatars used in automated customer service. Instead of graphics artists designing the facial expressions using their own cultural perceptions, psychologists can help to make the artificial faces seem more culturally sensitive and therefore more “human.”
DeBruine’s path to face research was somewhat unusual: “In psychology, they say that researchers tend to study what they have or what they lack. I was adopted, so family resemblance and how we ‘know’ who our relatives are has always interested me. I didn’t look like anyone else in my family, until my son was born eight years ago. So when I started a PhD with Margo Wilson and Martin Daly, the two evolutionary psychologists who have done the most to bring an evolutionary perspective to the study of family relationships, it was only natural that my project would be on the cues that indicate kinship and influence kin-directed behaviour.”
While doing research at McMaster University in Canada, DeBruine used commercially-available morphing programs to manipulate family resemblance, then later started using special software developed by Professor Dave Perrett at the University of St Andrews, which brought her to Scotland in 2004 to become a “full-time face researcher,” while retaining her interest in kinship.
After 12 years focusing on face perception, including how individual differences in pathogen disgust and regional differences in pathogen exposure influence how we see faces, DeBruine returned to her study of kinship, and won a five-year ERC Consolidator grant to do research into “How do humans recognise kin?” The aim of this new project is to determine “how we perceive the various potential cues of kinship (including facial and odour similarity, co-residence, maternal perinatal association and cognitive knowledge of kinship) and integrate them to modulate our prosocial and sexual behaviour.” And Face Facts has a key role in gathering the necessary data, as well as in public engagement.
On the road
The Face Facts project goes on the road to science festivals and other events, and also has a web site (http://facefacts.scot) where visitors can blend different faces together or use their own “selfies” to transform their faces – feminise or masculinise, or look trustworthy, dominant, heavier or slimmer. You can also look at pictures to decide which baby looks cuter, and who looks healthier or more aggressive. Another popular activity is creating an “average” face by blending multiple faces together and sampling common features. The touring exhibition uses three special pods and nine cameras to capture three-dimensional images of faces, using a system developed by Dimensional Imaging of Glasgow.
According to DeBruine, the Face Facts exhibit has two main aims: to educate the public about the multidisciplinary nature of modern experimental psychology, and to highlight the research on faces being done at the Institute of Neuroscience and Psychology, including the collection of 3D face captures from families in Glasgow, which is part of DeBruine's research into kinship.
Jack adds: “It's a fun way to get people into psychology, and talk about exciting scientific ideas, making complex issues more digestible. We work in the same basic field, but have different specialist interests. Lisa’s work has focused primarily on static face features (e.g., morphology, colour), the biological phenomena and kinship; while I have focused more on dynamic expressions and cultural factors; and so our interests and research findings are highly complementary.”
The technology is much more advanced than it used to be, including the cameras and imaging software, but what makes it work is the “funfair” appeal – like playing with a high-tech hall of mirrors. And the same technology is what enables Jack and DeBruine to do their research, drawing on the vast computing power now available.
“By touring with Face Facts, we are also challenging people's assumptions about what psychology is,” says DeBruine. “Many younger people may have preconceived ideas which limit their choices, and we hope this opens their minds.”
“Psychology can be perceived as fluffy or 'just common sense,' and not as objective or as technical as other mainstream sciences,” says Jack, “but most people are interested in cultural differences, or how we are similar across cultures; which grabs their attention, and gets across the message that psychology has real-world impact.”
After all, to build companion robots which people accept in their everyday life, you first have to understand humans.
The Face Facts Team
The other members of the Face Facts team include psychologists, biologists, computer scientists and anthropologists: Professor Philippe Schyns (Director of the Institute of Neuroscience and Psychology at the University of Glasgow); Professor Benedict Jones; Dr Amanda Hahn; Dr Oliver Garrod; Claire Fisher; Michael Kandrik; Chengyan Han; Dr Ross Whitehead; Dr Iris Holzleitner; Shona Fridh; Chaona Chen; Jiayu Zhan; Danielle Morrison; and Megan Sutherland. The support team also includes Glasgow-based Dimensional Imaging and the Glasgow Science Centre.