Imagine a creature that is like humans in every way, except that it has no enduring inner self. The behaviour of this hypothetical species – which we will call Homo vacuus – would be driven not by their private beliefs and inclinations, but rather by knee-jerk responses to their immediate environment. Lacking an enduring web of preferences, attitudes and proclivities, the contents of the H vacuus character would be filled by whatever situation they find themselves in. As a result, their behaviour would vary dramatically as they moved from one situation to the next.
Scrutiny of H vacuus would reveal a host of curious behaviours. Their attitudes towards others would be shaped by random factors that unconsciously influence their thoughts. They might feel ‘warmer’ toward others when they handle a hot cappuccino, and ‘colder’ when clasping an iced latte. Their political decisions would hinge on seemingly innocuous details. For instance, turnout in their elections would surge if they were affirmed as being ‘voters’, rather than merely people who vote. And on morality, their views on topics from abortion to pornography could depend on something as mundane as whether they had washed their hands, making them feel ‘cleansed’ and in the mood for judging others.
Given the inherent fluidity of H vacuus, trying to predict their behaviour beyond the present moment would be pure folly. H vacuus would be so mercurial that when they wed, they would effectively be marrying a stranger. In the end, being the pawns of powerful environmental forces would compromise their ability to effectively plan and adapt. Others would recognise the impossibility of holding them responsible for their actions, as they could refuse to accept responsibility for their behaviour by claiming: ‘I was just responding to the situation!’ At the end of the day, religious, ethical and judicial systems would be largely irrelevant.
In light of the many challenges facing H vacuus, it may be surprising to learn that this fictitious creature is based on an implicit assumption about human nature suggested by the work of numerous influential behavioural scientists. This assumption is that if you were to peer into the contents of people’s selves, you would discover little that isn’t susceptible to the random influences of the situation. Indeed, the peculiarities of H vacuus life described above – including the ideas about ‘warm’ attitudes, election turnout and moral cleansing – are based on claims that have been made about real humans.
We label this view of human psychology the empty-self metaphor. Although it has fostered an erroneous understanding of human nature, the empty-self metaphor has inspired a large following. These followers embrace an optimistic view in which human foibles can be readily vanquished through systematic manipulation of the social environment. However alluring it may be, we suggest that it is time to replace this Panglossian view with one that acknowledges the importance of enduring psychological structures.
The empty-self view has a long history in psychology, arguably tracing its origins to the behaviourist movement at the beginning of the 20th century. The behaviourists were almost entirely focused on ‘the situation’, eschewing speculation about what they considered the ‘black box’ of the mind. This emphasis on the situation was influential with founding social psychologists, who defined social psychology as the study of how individual thoughts, feelings and behaviours ‘are influenced by the actual, imagined, or implied presence of other human beings’.
An emphasis on the power of the situation, and its propensity to overwhelm or displace personal agency, has dominated social psychology ever since. Examples abound. In the 1970s, the author of self-perception theory argued that, rather than beliefs guiding behaviour (as common sense might imply), situations cause behaviour, and people generally infer their own beliefs by taking note of how they behave. Others made the corollary argument that personalities effectively do not exist. In fact, the term ‘fundamental attribution error’ was coined to describe an allegedly flawed tendency to attribute someone’s behaviour to their personality. Elsewhere, including in sociology, impression management research and the dramaturgical movement suggested that all the world’s a stage, with people portraying whatever character best suits a given scene. This focus on the situation reached a fever pitch with the ‘situationists’, who explicitly placed the influence of the situation on human behaviour far above any possible internal motivations.
Although situationism has lost some of its lustre, the empty-self metaphor has remained at least implicitly popular within certain circles. This is particularly true in the field of social priming, which deals with the impact of seemingly inconsequential environmental cues on human behaviour. In examples of this kind of research, seeing words related to old age seemed to make university students walk more slowly, and plotting dots on paper further apart appeared to make people feel more distant from their family.
Classic evidence for the primacy of the situation has lost force
In recent years, however, the evidential tide has turned against the empty self. Social priming and other areas of social psychology have struggled in the wake of the ‘replication crisis’ – the unfortunate realisation that a surprising number of psychological findings could not be independently replicated. Many such findings are likely to have been false positives, stemming from tendentious statistical decision-making and other questionable research practices. The authors of one report found that when a number of social priming studies were carried out again, only 17 per cent of the findings were successfully replicated. A much smaller proportion of empty-self-style research has been dismissed as simply fraudulent – such as the celebrated finding that people are more honest when filling out forms if they sign at the beginning, rather than at the end.
In our own research, we’ve compared studies that were or were not premised on the empty-self assumption that behaviour is unexpectedly shaped by subtle, seemingly irrelevant environmental factors, as opposed to pre-existing beliefs and dispositions. We found that studies reflecting an empty-self view were far less likely to hold up.
Classic evidence for the primacy of the situation has also lost force. Take, for instance, Stanley Milgram’s obedience studies, in which participants were asked to administer electric shocks to someone until the victim seemed to be unconscious or dead. These classic studies have been taken as evidence that authority can override individual will. However, subsequent research has found that personal traits associated with diligence and helpfulness predicted an increased willingness to administer the electric shocks. This reframes the experiment as demonstrating not mindless obedience, but often wilful collaboration – participants who gave the shocks were ostensibly helping an experimenter in the service of ‘science’.
Consider also the Stanford prison experiment, where students acting as guards in a fake prison (with others acting as prisoners) appeared to succumb to their temporary role, becoming belligerent. Attributing this behaviour to the roles the students were assigned is jeopardised by several design flaws, including a selection bias that likely caused dispositionally authoritarian individuals to be overrepresented in the study. The upshot is that participants’ unique, personal identities seem to have mattered far more than the authors of these classic examples of situational influence would have us believe.
The idea that the human mind is so fickle should have attracted more scepticism than it has often received. First and foremost, it is at odds with other evidence. The entire field of personality psychology – which investigates the unique qualities of people that persist across situations – informs us that the self is not empty at all. Measures of personality such as the Big Five traits (openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism) are remarkably stable and predictive of behaviour and long-term life outcomes. In fact, research suggests that the average effects of personality and situations on behaviour are similar in magnitude. Elsewhere, there is ample evidence for an enduring self: for example, self-esteem is largely steady over the lifespan, and attempts to break habits frequently fail, highlighting that personal tendencies impede many people’s efforts at self-change.
It is not difficult, though, to understand the appeal of the empty-self metaphor. The idea that our thoughts and behaviours are dictated by banal aspects of daily life, such as the temperature of our coffee, offers a tantalising glimpse at ‘hidden’ causes of behaviour. This is probably why social priming occupied a full chapter of Daniel Kahneman’s bestselling book Thinking, Fast and Slow (2011), why mentalists like Derren Brown attract large TV audiences by showing how the ‘power of psychology’ can be used to control and befuddle, and why exciting ‘counterintuitive’ findings are published in top journals.
The empty-self metaphor also lends itself to an admittedly admirable progressive agenda (popular among social psychologists) that is focused on addressing inequality in the social environment and avoiding victim-blaming. If the plight of the have-nots is primarily attributable to the harsh social conditions that society unwittingly thrusts upon them, then the obvious solution is to change those social conditions. Ultimately, the empty-self metaphor teases the provocative possibility that society could be righted if only we determine how to control the social environment. But reality is much messier.
There is considerable evidence for human consistency, agency and a remarkable propensity to shape our situations just as they shape us
The good news is that psychology appears to be slowly moving toward increased acceptance of what’s called ‘interactionism’: the view that the person and the situation interactively dictate behaviour. Many people have heard of nudging, in which subtle changes to the presentation of choices can systematically influence behaviour (eg, placing fruits and vegetables more conspicuously, to increase healthy consumption). The idea of nudging has proven enormously popular, with multiple governments around the world setting up ‘nudge units’, but initial results were lacklustre. Happily, research has moved to second-generation, interactionist nudges that tailor the intervention to the beliefs and preferences of the individual. For example, recent research suggests that personalised nudges to make healthier food purchases might be more effective than generic ones.
Interactionism likewise paves a more realistic path forward for social change. Rather than assuming that one size fits all, an interactionist approach acknowledges that individual differences inform the way people respond to social situations. Any social movement to reduce harmful behaviour or improve wellbeing needs to meet people where they are at. That means acknowledging that different individuals will respond to the same policies and political messages differently, and in accordance with their own perspectives.
Rebuking the empty-self view recognises that, far from being the dithering shells that psychology studies have often suggested we are, most of us are loyal to a reasonably clear and coherent self-concept. We do not ‘marry strangers’, nor do our relationships hinge on irrelevant details of the moment. On the contrary, there is considerable evidence for human consistency, agency and a remarkable propensity to shape our situations just as they shape us. Although imputing this degree of agency and stability to the self may lack the intoxicating simplicity of Homo vacuus, we are confident that it will pave the way to a deeper understanding of Homo sapiens.