THE SCIENCE:
Research Summaries

How do people work out what other people are thinking?

We have spent a lot of time conducting experiments designed to uncover how people work out what other people are thinking. This is something that scientists don’t really understand – and until we do, we won’t be able to work out why people are worse at understanding what outgroups (people who are not like themselves) think.

We have shown that one of the key processes involved in working out what people are thinking is to work out what type of person they are – what type of personality they have, for example.

WHAT DID WE DO?

We developed a new way of testing people’s ability to work out what others are thinking and feeling, and to work out what kind of person they are. We recorded videos of pairs of strangers having conversations, and measured their personality, as well as what they were thinking throughout the conversations. Then we asked new participants to work out what the people in the videos were thinking.

WHAT DID WE LEARN?

We can measure people's ability to read the traits, thoughts and emotions of others. One of the key factors in people's ability to do this is how similar the other person is to them.

WHAT DOES THIS MEAN?

If outgroups are dissimilar to us, we will be worse at judging all of these things. This work might be especially interesting when we think about how different groups, such as those with and without autism, understand each other.

This research has led to the following papers:

Long, E.L., Catmur, C., Fleming, S., M., & Bird, G. (in press). Metacognition facilitates Theory of Mind through optimal weighting of trait inferences. Cognition, 256, 106042. https://doi.org/10.1016/j.cognition.2024.106042

Long, E.L., Catmur, C., & Bird, G. (in press). The Theory of Mind hypothesis of autism: A critical evaluation of the status-quo. Psychological Review.

Long, M., Cuve, H.C., Conway, J.R., Catmur, C., & Bird, G. (2022). The Interview Task: Novel Theory of Mind Task Demonstrates Representation of Minds in Mental State Inference. Scientific Reports, 12(1), 21133. https://doi.org/10.1038/s41598-022-25490-x

This research was run at the University of Oxford by Emily Long, in collaboration with Dr Caroline Catmur (King's College London) and Professor Geoffrey Bird (University of Oxford) and is funded by The John Templeton Foundation.

What other ways can we use to investigate people's ability to understand others' personality?

Another way we are working out how people understand each other's personality is to use computer games, where people often think of players as having different 'player types', which is a bit like an in-game personality. You can read more about how we are using the game Minecraft in this research here.

This research is being run at King's College London by Lorenzo Cappiello in collaboration with Dr Caroline Catmur and Dr Salim Hashmi.

How do people work out what other people are feeling?

Most research into how we understand what other people are feeling focuses on our ability to interpret facial expressions. However, expressions are ambiguous, so we rely on additional information, like the context someone is in, to narrow down the possible emotions. In this research, we were interested in whether a person’s personality might be another important source of information that we use to determine what they might be feeling.

WHAT DID WE DO?

We asked people to judge facial expressions of emotion both before and after being told about the person’s personality traits.

WHAT DID WE LEARN?

Information about people's personality altered participants’ emotion judgments. The same facial expression was judged to be happier when seen on a person who was described as friendly or altruistic, and as more angry when the person was described as uncooperative or untrusting.

WHAT DOES THIS MEAN?

We use people's personalities to understand what they are feeling.

Next we wanted to understand why some people are better at judging emotions than others; and why we are better at judging certain people's emotions compared to other people's.

WHAT DID WE DO?

We video-recorded pairs of strangers having conversations, and measured their emotional experiences throughout, as well as their personality. Other participants then watched clips of these conversations and had to rate the emotions and personality of the people they observed.

WHAT DID WE LEARN?

People who had a better understanding of the relationships between personality and emotions, and who were more accurate at judging the strangers’ personality, were also better at judging their emotions. This also depended on how similar the participant and stranger were – the more in common you have with someone, personality-wise, the better able you are to judge their traits and therefore their emotions.

WHAT DOES THIS MEAN?

People who are better at judging others’ personalities are better at judging their emotions
We are better at understanding the emotions of people who are similar to us

This research is currently under review, but will be available here as a pre-print shortly.

This research was run at the University of Oxford by Leora Sevi, in collaboration with Dr Caroline Catmur (King's College London) and Professor Geoffrey Bird (University of Oxford) and is funded by The John Templeton Foundation.

How do our past experiences with others affect how we understand new people’s minds?

This study explored how the minds of people that we have previously met influence the inferences we make about new people we encounter.

WHAT DID WE DO?

We introduced participants to a fictional group of characters displaying high levels of the personality trait paranoia.

Then participants met some new characters and had to work out what they would believe, and how they might behave, in everyday situations.

In some tasks, participants had no specific information about these new people, while in others, we provided details about their personality.

WHAT DID WE LEARN?

Whenever participants were given details about a person’s personality, it influenced the inferences they made about that person’s beliefs and behaviour.

But when no specific information about the new person was provided, participants judged the new people's beliefs and behaviour as more paranoid, like the original group of characters they had met.

This suggests that the types of people we’ve previously met can affect how we judge new people, particularly those we know very little about.

This helps to explain why people vary widely in the way they interpret the thoughts and feelings of others, as this appears to depend on our previous experiences and encounters in life.

This research was presented at the January 2024 meeting of the Experimental Psychology Society. The full manuscript for this research is currently being prepared for submission, and will be available here as a pre-print shortly.

This research was run at King's College London by Alicia Mixter, supervised by Dr Caroline Catmur and Dr Bryony Payne (King's College London), and Professor Geoffrey Bird (University of Oxford) and was funded by the Experimental Psychology Society and The John Templeton Foundation.

Creating a new trauma-informed test of mentalising

This study explored how childhood trauma might impact the way people understand others’ thoughts and feelings, a skill known as "mentalising." Early life trauma can shape how individuals interpret and relate to others, often in ways that differ from those who haven’t had similar experiences. To better reflect these unique perspectives, this research focused on developing a new test with the help of people who have experienced childhood trauma.

Most tests that measure how well people understand others’ thoughts and feelings assume there’s a single “right” way to interpret social situations which is often based on what people with more typical life experiences would expect. This new test takes a different approach by recognizing that people affected by childhood trauma may interpret social cues differently. Rather than focusing on a single "correct" answer, it aims to capture the unique experiences of those who have faced early adversity.

WHAT DID WE DO?
We conducted interviews with survivors of childhood trauma. They shared a range of everyday social situations such as times when they felt misunderstood, angry or uncomfortable.

Their stories helped us create a test with different scenarios reflecting the real-life social challenges they have faced.

In this test, the ‘correct’ answer was determined as the way they felt and thought, rather than a pre-determined ‘right’ way of thinking or feeling.

We then administered the new test to individuals with and without experiences of childhood trauma.

WHAT DID WE LEARN?
Survivors of childhood trauma performed better than those without trauma on our new test.

This finding suggests that mentalising is not a one-size-fits-all skill, but instead, people understand others best when they share similar life experiences.

This can help explain why individuals with childhood trauma might feel misunderstood or have difficulty understanding others in a society where most people haven’t had the same experiences.

We hope to make this test available for clinicians working with childhood trauma survivors. It opens the door to a new approach in understanding the mentalising abilities of people with childhood trauma—not as a deficit, but as a difference. This approach can help clinicians to better support trauma survivors by adapting to their unique ways of understanding others.

This research is currently being prepared for submission, but will be available here as a pre-print shortly.

This research was run at King's College London by Mariza Stagaki, supervised by Dr Caroline Catmur (King's College London), Dr Tobias Nolte and Professor Peter Fonagy (University College London, Anna Freud Centre) and was partially funded by The John Templeton Foundation.

How do adults change their language when talking to children?

For children, one of the most important ways of learning about other people's minds is from the language that adults use. We looked at whether, and how, the language that adults use with children changes depending on the child's gender.

WHAT DID WE DO?

We showed adults (both parents and non-parents) pictures of children with masculine, feminine, or neutral expressions of gender. We then asked them to choose the type of language they would use to speak to each child.

WHAT DID WE FIND?

Adults chose to use more mental state language (focused on thoughts, emotions, and desires) with feminine and gender-neutral children than with masculine children. Additionally, gender-neutral children received less elaborated language (explanation about mental states). You can see a video about this study here.

WHY DOES THIS MATTER?

We know that by using mental state language, adults can help children to understand that other people have different minds with different thoughts, emotions and desires from their own. And the studies described below show how mental state language affects how we think about other people. So if adults use different language with different children, this could affect the children's social understanding.

This research has led to the following paper:

Farrell, C., Slaughter, V., Thai, M., Mulvihill, A. (2023). How We Talk to Kids: Adults Prefer Different Forms of Language for Children Based on Gender Expression. Sex Roles, 89, 119–134. https://doi.org/10.1007/s11199-023-01393-7

This research was run at the University of Queensland by Callyn Farrell, in collaboration with Dr Aisling Mulvihill and Professor Virginia Slaughter and is funded by The John Templeton Foundation.

How does the language in the media affect how we think and feel about others?

The media is one of the main ways in which information about others is shared. The media often tells us stories about “out-group members”, for example immigrants or people who have different religious or political views to ourselves. We tested whether the language used in the media to describe these “out-group” members, shaped the way we think about them.

First, we compared the language used by The Guardian (a politically left-leaning news source in the UK) versus The Daily Mail (a politically right-leaning news source in the UK).

WE FOUND THAT...
When out-group members were introduced by The Guardian...

1. People had more empathy for out-group members, reporting that they felt worse for them, and were more willing to help them.

2. People thought about members of out-groups as being individuals, each with their own mind, rather than thinking of everyone in the group being the same.

3. People judged out-group members to be more rational and more trusting.

We then tried to pin down what the difference was between the language used by these two newspapers that might explain these results.

We found that it is very important whether or not the language used to describe the out-group members uses a lot of mental state language or not (i.e., words that describe someone’s thoughts, feelings, desires, and intentions). When people read about out-group members that contained a lot of mental states, those people were more likely to think about the out-group members as individuals, considering what sort of mind they might have.

This work is important because it shows that the type of language people use when referring to others can influence how we think about their minds, and whether we treat them as individuals.

This research is currently under review, but will be available here as a pre-print shortly, and you can watch a video about this research here.

This research was run at King's College London by Dr Bryony Payne, in collaboration with Dr Caroline Catmur (King's College London) and Professor Geoffrey Bird (University of Oxford) and is funded by The John Templeton Foundation.

How do people learn to be empathic?

How people learn to have empathy for others, especially those who are dissimilar from themselves, is one of the most pressing issues facing society. In this series of experiments we investigated how we could encourage people to have empathy for an alien species – the Cloods!

WHAT DID WE DO?

We showed people cartoons of "aliens" that we called "Cloods". We tried different ways to encourage people to empathise with the Cloods. We let them watch other people’s empathy for them, we described how the Cloods were feeling using words, or we let participants feel the Cloods’ pain themselves, in a variety of ways. We then measured how much empathy people felt for the Cloods and how much they wanted to help them.

WHAT DID WE FIND?

We were able to encourage people to feel empathy for the Cloods – even though they are a fictitious alien species – and it seems that language is a great way to do so. Just by explaining to people that the Cloods were in pain, they felt empathy for them and wanted to help them.

This research is currently under review, but will be available here as a pre-print shortly.

This research was run at the University of Oxford by Dr Zoe Pounder, in collaboration with Dr Caroline Catmur (King's College London) and Professor Geoffrey Bird (University of Oxford) and is funded by The John Templeton Foundation.

How does the language used by pre-school educators affect children's social understanding?

We know that when parents use mental state language (talking about thoughts, beliefs, feelings and desires) with their children, the children develop a better ability to understand other people's minds. What is not currently known is whether the language used by other caregivers has the same effect.

WHAT DID WE DO?

We observed pre-school educators working with children during story time and during a Duplo block-building task. We measured how much mental state language the educators used. Then we measured how good the children were at understanding other people's mental states.

WHAT DID WE FIND?

Surprisingly, we did not find a link between the amount of mental state language that the educators used, and the children's understanding of other people's minds.

WHAT DOES THIS MEAN?

It may be that educators do not play as large a role as parents in the development of children's mental state understanding; or it may be that in this study, because the tasks were carried out in small groups of children, educators were unable to tailor their mental state language to each individual child's level of development. That would make it hard to detect an effect of the educators' language on each child's ability, because the language used wasn't aimed at individual children in the same way as it might be by the parent.

This research has led to the following paper:

Mulvihill, A., Armstrong, R., Casey, C., Redshaw, J., Scarinci, N., & Slaughter, V. (2023). Early childhood educators' mental state language and children's theory of mind in the preschool setting. British Journal of Developmental Psychology, 41, 227–245. https://doi.org/10.1111/bjdp.12449

This research was run at the University of Queensland by Dr Aisling Mulvihill in collaboration with Professor Virginia Slaughter and is funded by The John Templeton Foundation.

How well do different groups of people recognise others' faces, and can other people take this into account?

People are used to grouping other people on the basis of their gender, politics, ethnicity, religion, and geography. However, humans also take account of how other people differ in their mental abilities too. For example, people change the way they speak to children to make it easier for them to understand, and people even change the way they speak to older adults to compensate for how their hearing is likely to have changed. In these studies we wanted to see whether people will take into account the way other people see the world - and we chose to look at how people recognise other people's faces.

WHAT DID WE DO?

First we looked at how people recognise faces, breaking this down into their ability to perceive faces, to work out whether two faces are the same or different (‘face matching’) and to remember faces. We mapped out these abilities in individuals with autism, individuals with developmental prosopagnosia (lifelong inability to recognise faces) and in super-recognisers (individuals with exceptional face recognition).

After learning about each of these groups’ face abilities, we taught other people about them and asked them to pick helpers for different tasks that required one or more of these face skills.

WHAT DID WE FIND?

Individuals with developmental prosopagnosia struggled with all three of these face processing abilities, while super-recognisers excelled at all three abilities. Individuals with autism had difficulties with face perception and face memory, but had no difficulties with face matching.

Furthermore, we found that other people were able to take into account each group’s pattern of mental strengths and weaknesses in these abilities.

This research has led to the following papers:

Stantić, M., Pounder, Z., Bate, S., Catmur, C., & Bird, G. (in press). Individuals who are ‘super recognisers’ show superior performance on independent measures of face perception, face memory and face matching. Psychonomic Bulletin & Review.

Stantić, M., Brown, K., Ichijo, E., Pounder, Z., Catmur, C., & Bird, G. (2023). Independent Measurement of Face Perception, Face Matching, and Face Memory Reveals Impairments in Face Perception and Memory, but Not Matching, in Autism. Psychonomic Bulletin & Review, 30(6), 2240-2249. https://doi.org/10.3758/s13423-023-02304-3

Stantić, M., Pounder, Z., Bate, S., Susilo, T., Catmur, C., & Bird, G. (2022). Individuals with developmental prosopagnosia show independent impairments in face perception, face memory and face matching. Cortex, 157, 266–273. https://doi.org/10.1016/j.cortex.2022.09.012

This research was run at the University of Oxford by Dr Zoe Pounder, in collaboration with Dr Mirta Stanti ć (Royal Holloway, University of London), Dr Caroline Catmur (King's College London) and Professor Geoffrey Bird (University of Oxford) and is funded by The John Templeton Foundation.

How does this relate to...

How well can we understand people who differ from ourselves?

Societies are becoming more polarised, partly because people often misunderstand the beliefs and intentions of those who are different from themselves. To understand why these misconceptions occur, we investigated how good people are at thinking about other people’s minds.

Our research tested this by asking 256 people from the US to predict the social and political beliefs of others. We presented participants with statements such as “immigrants are good for society” and asked them to what extent they agreed. We then showed them the responses of other people who had been asked the same question.

If the participant said they strongly agreed that immigrants are good for society, they would be presented with someone who said they strongly disagreed with this. This would make them “out-group” to the participant because they disagree. Now aware that the other person disagreed with them on one belief, we would then ask them to predict this other person’s opinion on a different topic, such as “everyone should have access to legal abortion” or “it is right that same-sex marriage is now legal and acceptable”. The participant might assume that people who are anti-immigration are also anti-abortion or against same-sex marriage.

Each time someone predicted what the other person thought, we asked them to state how confident they were that their prediction was correct. We found that people were consistently very confident that their predictions were accurate (75% confident) but, for out-group members, they were wrong more than 60% of the time. In fact, the more confident people were that they could understand, the more likely it was that they were wrong. This showed that people think they are much better at understanding out-groups than they actually are.

Anti-Brexit March in London, quite a few.webp

WHAT THIS TELLS US

1. People are much worse at knowing and understanding the views of people they disagree with, or people that differ from themselves.

2. The reduced ability to understand people they disagree with, was not due to being less bothered to think about those people. It was because we have a poorer understanding of how their minds work, perhaps because we do not interact with them as much.

3. People were unaware that they were much worse at understanding people they disagreed with - they were very confident that their understanding was correct, even when it was wrong.

Payne, B., Bird, G., & Catmur, C. (2024). Poorer representation of minds underpins less accurate mental state inference for out-groups. Scientific Reports, 14, 19432. https://doi.org/10.1038/s41598-024-67311-3

This research was completed at King's College London by Dr Bryony Payne in collaboration with Dr Caroline Catmur (King's College London) and Professor Geoffrey Bird (University of Oxford) and was funded by The John Templeton Foundation.

THE SCIENCE:
Research Summaries

How do our past experiences with others affect how we understand new people’s minds?

Creating a new trauma-informed test of mentalising

WHAT DID WE DO?
We conducted interviews with survivors of childhood trauma. They shared a range of everyday social situations such as times when they felt misunderstood, angry or uncomfortable.

Their stories helped us create a test with different scenarios reflecting the real-life social challenges they have faced.

In this test, the ‘correct’ answer was determined as the way they felt and thought, rather than a pre-determined ‘right’ way of thinking or feeling.

We then administered the new test to individuals with and without experiences of childhood trauma.

WHAT DID WE LEARN?
Survivors of childhood trauma performed better than those without trauma on our new test.

This finding suggests that mentalising is not a one-size-fits-all skill, but instead, people understand others best when they share similar life experiences.

This can help explain why individuals with childhood trauma might feel misunderstood or have difficulty understanding others in a society where most people haven’t had the same experiences.

How does the language in the media affect how we think and feel about others?

WE FOUND THAT...
When out-group members were introduced by The Guardian...

How does this relate to...

How well can we understand people who differ from ourselves?

WHAT THIS TELLS US

The UK 2024 Riots

Our mental health

American Politics

THE SCIENCE: Research Summaries

THE SCIENCE:
Research Summaries