‘A re-envisioning of Psychology itself’

'No man is an island, entire of itself,' wrote John Donne in 1623. This famous line has been quoted often ever since. But, write the authors of a new review paper in Perspectives on Psychological Science, 'it is resonating anew in psychological science'.

Interactions with other people underpin who we are, how we think and how we behave, writes Thalia Wheatley at Dartmouth College, US, and colleagues. For over a century, though, psychologists have mainly focused on studying single individuals. One of the main reasons is that natural social interactions are hugely complex, making them extraordinarily challenging and time-consuming to analyse – and limiting the analysis that can be done.

But new methods and tools, including AI, are now making those complex interactions more possible to analyse. Work embracing these methods is giving us unprecedented insights into all kinds of aspects of our psychological functioning, with the promise of lots more to come.

The benefits of gossip

In a typical day, an average person speaks about 16,000 words, and about 65 per cent of those words are on 'social topics' – about what they were up to at the weekend, for example, or what a mutual friend said at a party. Chat about other people is generally regarded as 'gossip'. Though many of us regularly gossip, its exact role in our lives has not been clear.

But in 2021, a study in Current Biology, in which groups of participants played online games and some participants were able to gossip via a private chat with a partner if they wanted to, while others weren't, revealed all kinds of benefits. (Importantly, in contrast to earlier work, no one was advised to discuss anything in particular via private chat. Rather, as happens in the real world, this gossiping developed spontaneously.)

In that study, Eshin Joly and Like Chang found that gossiping made participants like each other more. It also led them to develop shared impressions of other people, something that other studies have shown strengthens social bonds. And, by allowing participants to learn about the successes and mistakes of others who they couldn't observe directly, it helped them to get better at playing the games.

Human connection 

Most of us will be familiar with the feeling of 'clicking' with another person. In 2022, a paper in PNAS revealed that the speed with which two conversation partners respond to each other is 'a robust and sufficient signal' of whether they have clicked – or not.

The researchers, led by Emma Templeton, found that both pairs of strangers and friends felt more connected when the person they were talking to responded very quickly. What's more, participants who listened to these (unstructured, naturalistic) conversations judged the pairs with faster response times to be more connected. This was despite the fact that they hadn't been told to pay attention to the timing of turns, which suggests that this is a cue that we learn and use automatically.

There's a good explanation for why this cue should be reliable, the team adds. In a conversation, as well as listening, we need to prepare our response to whatever the other person else is saying, we need to judge when they are likely to stop speaking, and we need to decide when exactly to respond and also anticipate their reaction. This is a complex series of tasks that we generally accomplish extremely rapidly, but which we can do more rapidly the better we understand the other person's mind.

However, long gaps are not always a sign that people are failing to connect. In 2023, some of the same researchers behind this study reported other work in a paper titled 'Long gaps between turns are awkward for strangers but not for friends'. They wrote that their findings suggested 'the gaps of friends may not function as "gaps" at all, but instead allow space for enjoyment and mutual reflection.'

They add that since research on conversation is generally done on strangers (because it's easier to recruit strangers than pairs of friends), earlier studies may not have captured what happens in some of our more familiar – and most important – relationships.

Learning and social ties

New work on interacting minds is also revealing how social network structure affects what people learn and remember. Ida Mommenejad, who works for Microsoft Research in the US, has led work on this, and recently published a review of research in the field.

It's long been known that people pass and share information within their social networks, leading to the development of shared, or 'collective', memories, knowledge, and beliefs. But not all the people within a network are connected in the same way. Some human networks contain clusters of tightly knit individuals with strong ties (a group of friends or a work team, for example), and also other people who communicate with different clusters, and so 'bridge' them.

In her own research, Mommenejad has found that the collective memories of a group converged more if people in this network with 'bridge' rather than strong ties exchanged information first. In this study, these 'bridge' people had a direct tie to one other individual, but they didn't have any ties in common. In the real world, Mommenejad explains, this could mean two friends who talked – but their friends weren't friends and didn't to each other.

In their review paper, Wheatley and her colleagues comment that this type of work, involving detailed studies of interactions in a shared social network, is leading to a new understanding of how some members of a network have more influence than others over the shared memories, beliefs, and knowledge of all the individuals.

Human bias

Other research aimed at exploring how we behave in real-world settings is now revealing, in sharp and sometimes uncomfortable detail, the existence and implications of some common biases.

For example, a 2021 study led by Nicholas Camp of body-cam footage from US police officers making everyday traffic stops found that officers used a less unpleasant tone of voice when talking to White versus Black drivers. In the same paper, the researchers investigated how these tonal differences might shape how participants construed the police generally, and found they undermined trust in police departments. The team argues that their data paint a picture of a cycle in which differences in interpersonal aspects of police encounters with the public erode trust in that institution, and perhaps in others.

Machine learning, which can enable the rapid analysis of vast amounts of data, could take this type of work further, Wheatley and her colleagues argue, and identify other patterns in data that would be otherwise hard to detect. As an example, they point to another study, from Cydney Dupree and Susan Fiske published in the Journal of Personality and Social Psychology in 2019, which found that White liberals - but not conservatives - were more patronising towards non-White people than they were towards White people. Black people and other minorities are stereotyped as being of a lower status and less competent, the team writes, and a desire among liberals to affiliate with them may have caused their more patronising attitude, they suggest.

These examples of new findings in everything from how we learn to how we connect with others are just the cusp of a major shift in psychology towards the idea that social interaction underpins our psychological functioning both as individuals and as groups, write Wheatley and colleagues. This shift is badly needed, they argue – and they think it could represent a moment of fundamental change. 'We are not arguing for a new subfield of psychology or for a revamped version of social or collective psychology,' they write. 'Instead, we propose a re-envisioning of psychology itself.'

Of course, we will still need work into how individuals respond to a wide range of experiences. But such a change would mean that the days of researchers treating people as 'islands' would be gone for good.