2020 February: The Psychology of Characters

Introduction to Personal Metrics, performance/experiment held at MARs, Goldsmiths College. 21 Jan 2020:

Hello, welcome everyone, to this Art Research Installation, Personal Metrics, in which I’m hoping to conduct an experiment with your help.

Some of you may know by now that my practice is concerned with using my own personhood as a kind of philosophical laboratory. I make performance, text and films in which I try to write myself into different characters to see what kind of powers or capabilities I can acquire by experimenting with speaking through an altered voice and manner. So to some extent my practice is an exaggerated version of what we all do in the everyday practice of being a person – we take on slightly different characters to respond to different social or even non social situations. By going about this deliberately in practice, what I’ve found is that by acting in a way that is just slightly non-habitual, just slightly outside of the way I normally behave, I start having thoughts pop into my head that have never occurred to me before. I get ideas, or follow hypothetical trains of thought that my usual self would find totally absurd, or which I would flat out disagree with. My working theory at the moment is that character is that thing which makes some thoughts thinkable, and others unthinkable. Assuming a character is like taking on a certain shape, and allowing the input from the world to travel through that shape like a ball in a pinball machine. What comes out of this machine, the trajectory of the pinball, or the expression of the character, is directly impacted by the shape traced by a performative instance through the grooves of character. There is this moment of action, of gesture, of speech, of being, which is in fact articulated by a kind of software more akin to clothing than a person. Character is wearable, changeable, but this is not merely a kind of promiscuity of selfhood. The characters we choose to wear, or which we habitually fall into, directly guide the range of thought open to us. So ‘going into character’ as someone slightly different can have some practical everyday applications, for instance it’s very good for brainstorming. It’s very good for ‘thinking outside the box’.

Through my interest in personhood and character, especially because as an artist I am deliberately, artificially reconstruing and tampering with these things, I’ve always had an interest in artificial intelligence. And in doing my research I’ve often wondered whether personhood or character was quantifiably measurable in any way. I’ve always been conscious of the Turing test for instance, which presumes that a machine can only be recognised as a full-fledged person by another person – that if a human being were sufficiently fooled by a machine acting like a human, that machine would essentially have achieved what’s referred to as general artificial intelligence. But until now I’ve completely neglected the fact that there already exist a variety of metrics for measuring various aspects of our own personhood. The field of psychometrics covers quantitative methods for a wide variety of areas of human psychological capability and character, from IQ tests to personality tests.

One of the reasons this probably hasn’t occurred to me before is that I’ve always had a suspicion about such tests; they’ve always felt a quite dodgy. But I confess I don’t know much about the research underpinning them, or the mathematics of scoring what seem to be fictional qualities. What I do know is that both personality tests and IQ tests are used in a range of professional contexts, with actual consequences. That would suggest that the belief in the wisdom of personality tests is prevalent enough to be consulted at a price. Employees are sorted and assigned roles according to these metrics, and the testing and evaluation service is often conducted by private companies.

The most common types of research-backed personality test are self-report questionnaires and observer-report questionnaires. A combination of the two is recommended as the most accurate test. This means that you fill out a questionnaire on your own behalf, and then have someone else also fill out a questionnaire about you. The two are balanced into a final score of numerous character traits – but also it is common that any discrepancies between how you assess yourself and how others assess you, are brought up as a topic of discussion with your employer.

I’d like to see what happens when such tests are applied to fictional characters; particularly the ones I’ve come to know in my practice. So tonight I’m going to perform short speeches by three different characters. After each character, I’d like us all to fill out this personality test based on our impression of the character. It should take about 12 minutes to fill out each one. My hope is that this process, although possibly a little laborious, will reveal some interesting problems and insights for discussion afterwards. I have also already completed the questionnaire for each character myself – both as a self-assessment, while being in-character, and as an observer assessment, just considering my character as an other. I’ll share with you the results I have so far afterwards, but for now let’s try it out ourselves. When I collect your responses, I hope to collaborate with the researchers who designed this test, and compile the results into a representation of your impression of each character combined versus their self-assessments. This should create a map of each character’s personality make-up, as if they were persons.

Skype session with S

I’m about to meet up on Skype with S, a psychology researcher from Liverpool University. I was interested in her team’s research (Francis, K.B., Terbeck, S., Briazu, R.A. et al., 2017) into the ways in which people respond to a moral dilemma (specifically, scenarios in which committing a harmful action using ‘personal force’ could result in a utilitarian outcome), depending on whether they have to respond to the dilemma with a ‘judgement’ (filling out a questionnaire on hypothetical scenarios) or an ‘action’ (using VR to simulate the scenario and then be met with the choice to respond physically, with haptic feedback of the action). Their research showed that there was a significant discrepancy between what group 1 said they would do (judgement) and what group 2 did (action) in a simulated environment – significantly more people endorsed the utilitarian option in the VR action condition.

I’m interested in the instinctive response I felt, and I think others attending S’s lecture at Goldsmiths felt, upon hearing the results; that this discrepancy, if taken to be more widely applicable, marked a kind of failing of humans. It reminded me of the discrepancy often found in self-report versus observer-report personality questionnaires; where there is a mismatch between how one perceives themselves and how others perceive them. In that case, it seems hard to answer: who is right? And what would it mean to be right about a person’s personality? Back to S’s work, I wondered how we should interpret, or judge, this tendency. To judge that discrepancy becomes a second-order moral judgement – how good or bad is it that humans judge that they will do one thing and then do the opposite thing in a simulated environment?

This research attempted to record the responses of humans in general, a tendency across numerous persons. But there was also some probing into individual difference, as participants in one of these morality experiments were asked to take personality tests before engaging with the dilemma. Actually, these participants took the same test I used for my fictional characters – the HEXACO-60 (except they were self-reports, plus they took another test specifically designed to measure the psychopathy trait). It turned out that the psychopathy trait had a role in determining the simulated action condition response, with a positive relationship between psychopathy scores and the power of the force put into performing the utilitarian action, as well as the probability of taking the utilitarian decision.

I’ve become now really interested in the prevalence of personality questionnaires. At a glance, they look rather simplistic. The HEXACO-60 organises persons in terms of 30 dimensions. The psychopathy trait is characterised by 2 dimensions: ‘callousness’ and ‘impulsiveness’. And yet, such knowledge, about an individual’s relative position on these scales seems to afford significant predictive power, in certain situations. The most dramatic example of this seems to be Cambridge Analytica’s use of quick personality tests published for free use on Facebook, which meanwhile allowed the company to profile those that took the test and connect that profiling to data harvested from the Facebook account of those individuals, as well as all of their friends. Such profiling has allowed the company to microtarget individual profiles with political advertising emotionally tailored to sway the individual towards endorsing a particular political party (or simply disencouraging some from voting at all). They particularly targeted people scoring highly or predicted to score highly on neuroticism, and for them designed ads that were emotionally fearmongering of a certain party. Cambridge Analytica did this around the world before working on the Brexit and Trump campaign, and continually found they were successful on whichever side of an election they found themselves on.

In my own work, I encounter characters in the endeavour to make fiction. I create characters with a set of mannerisms, statements and behaviours, that then seem to index their personhood. But through that process I’ve begun to notice that really I am flexing a muscle that we all seem to be exercising to some degree in the daily practice of being a person – slipping into different characters to accommodate different circumstances. More recently, I conducted an experiment in which I performed three different characters for a small audience, and asked them to fill out observer-report questionnaires after each performance. I was interested in the data, but also interested in what the process of filling out those reports would feel like for the participants.


Now, having just spoken to S, I reflect that I wanted to find out more about how psychologists see this idea of personality; what they think personality and personality traits are. I told S about the scepticism felt by my art colleagues when I showed them the results for each of the characters in relation to a median. How they worried about the idea of characterising persons in such apparently simplistic terms. They resisted the violence that such metrics imposed on the person – on the dynamic nature of being – reflecting that those measurements might change from one moment to the next; when we are on a period, taking an exam or running late for work. And yet, we experience that there is some predictive power in the interpretation of these results. But what property does this metric reflect? Surely there are not 30 parameters in our brains, each set to the measurement that reflects our behaviours. S told me that, certainly, my colleagues were right to be sceptical. Often scientific models do exchange some complexity for the economy afforded by simplification. And it doesn’t seem as though most psychologists ultimately believe that 5 traits can capture a whole person’s being. But it does seem to capture something, when we speak of being able to predict certain behaviours based on a person’s answers to a questionnaire. I wonder what psychologists ultimately think they are talking about when they talk about personality traits. Although agreeing that personality was more fluid than our popular, day-to-day understandings of it seem to suggest, she did say that ‘behaviour is influenced by personality and the environment’. That seems intuitively true, yet I wonder what personality is when regarded separately from behaviour and the environment, so as to influence behaviour. Is it a collection of neural pathways? Is it genes? Where is this thing that influences behaviour, especially when behaviour is that thing which reveals to us (and to them) for the first time something about someone’s personality? ‘You’ve shown your true colours’, a teacher once said to me in primary school, scolding me for something I didn’t do. Here are some things we spoke about.

  • S spoke to a colleague before talking to me, who she will hopefully put me into contact with. He is a personality psychologist and knows of experiments in which personality questionnaires were used on fictional characters and compared with those of real people. These tests found patterns in the tests on fictional characters that differed to patterns on real people. Hopefully she’ll ask him to forward me this work soon.
  • The paper written by S and her research team discusses whether the discrepancy in endorsements of using personal force to achieve a utilitarian outcome in ‘judgement’ versus ‘action’ conditions was in part due to differing ‘frames of reference’ that apply to judgment and action respectively. I was interested in this – “According to these accounts,” they write, “action choices are driven by egocentric perspectives and unlike judgments, may overlap with self-interested motivations as individuals consider the self-relevant consequences of their actions. Judging, on the other hand, relies on allocentric considerations influenced by cultural norms.” So judging might be a fundamentally different frame of mind to action. Ultimately, the two frames of reference reminded me of frames of reference in narrative. In judgment, the protagonist of the narrative is a generalised notion of ‘humanity’; in action, the ‘self’ is the protagonist. This also reminded me of my dual approach to filling out personality tests for my characters (I did both a self report, in character, and an observer report, as ‘myself’). I went through the series of questions as an onlooker (judging), and then as a protagonist (actor), and indeed the results were different. Is this discrepancy, either in moral judgment versus action or in self-report versus observer-report of personality, a margin of error – a failure of some kind? Or is it evocative of some fundamental alteration of the narrative being evaluated?
  • Attribution Theory: S told me that this area of research looked at how one might judge a situation differently to an observer based on the fact that they do not see themselves in that situation. She gave me the example of a person who slips in a puddle – the person is more likely to blame the environment for her slipping, whereas an observer is more likely to blame her characteristics (‘e.g., ‘she’s clumsy’). Furthermore, she spoke to me about how mirrors are employed in therapeutic contexts, to draw attention to one’s self. Arguments between couples might be filmed and played back to them to help change their frame of reference. She said that it has been found that people often act more ‘morally’ when they see themselves (e.g. in a mirror) while behaving. The effects of of seeing oneself whilst acting (as, S pointed out, we were now doing over Skype) was interesting to me because I deliberately create feedback loops as part of my method of getting into character. Seeing and hearing myself is integral to persuading myself that I am someone else.
  • Personality Complexity: Is a term that is different to self complexity but I am not entirely sure how yet.
  • Self Complexity: Is a psychological trait. When people score low on this scale, they believe that what makes them who they are are only a few distinct roles (e.g. ‘I am a student’). When they score highly on the scale, they see themselves as being made up of a lot of different attributes and roles. Good mental health has been associated with high levels of self-complexity and vice versa for low levels. She reckoned that my practice enabled me to practice a sense of higher self-complexity, of being able to be many different things, and that’s when she began to think that aspects of what I am doing to create characters might have benefits in a therapeutic setting.
  • We discussed briefly whether personality questionnaires might have more predictive power for some behaviours, and not others. In forensic psychology for instance, S said, using a combination of tried and tested psychometrics have not been very useful in determining the likelihood of an offender re-offending in parole cases.
  • S spoke to me of the power of situation and social influence in shaping a person, citing Zimbardo’s prison experiment as a famous example of dramatic transformation of behaviour and what seemed to be the deeper makeup of their being. Or Milgram’s experiment. This had a lot to do with the roleplay involved – and we spoke of how role seems to be very integrated with the makeup of personhood (certainly this is backed up by a lot of ethnographic work, which I came across cited in Frow, 2015).

Finally, after listening to how I make my characters, S seemed to think that aspects of my method might have benefits in a therapeutic setting. She proposed we collaborate on an experiment: she would be able to access participants who could be instructed on changing aspects of their voice or mannerism (in the privacy of their home, or any empty room), and we could assess their mental health before and after using these behavioural methods. I’d never thought of trying to instruct others to try what I do. Suddenly it seemed we could conduct an experiment in which both of us could benefit from the data in similar ways. S mentioned that we could try and arrange it so that we could get videos of the participants, and it struck me that that could end up being quite a striking art exhibition: Having videos of persons before and after adopting this light acting method, and showing these videos side by side. My motivation in my research is to explore and demonstrate the ‘fluidity of personality’ as S called it – as I have always had a fluid relationship to identity and being, given my multicultural background, which S also could relate to. Such an art exhibition would show this taking place in the bodies of different persons; evidence of the exploration of self-scope. Meanwhile, S could research the therapeutic dimension, and see if the practice might train self-complexity and contribute to a sense of well-being. Of course, we must work out the details – not least how I could translate some of the principles of my method into simple instructions – but then, S has access to and experience with participants, ethics procedures and waiver forms that I don’t have. Working with her could be a great learning experience for me, and also open artistic opportunities through inviting others to try my process.

Plan Achievements/Notes
Week 1 (3-9)
Skype with S
Create PT visualisations
Write up PT experiment
Design ‘therapeutic’ experiment

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s