I used to work with a brilliant programmer who I'll call Bob (not his real name). Bob’s only flaw was with people: he would form an impression of someone and then stick to that first impression — even in the face of new evidence that subtly contradicted the opinion he originally formed.
This habit, of course, led to all sorts of difficulties. Since Bob was a senior engineer at our company, he was ultimately responsible for working with and mentoring many of our junior engineers. In this regard Bob didn’t do very well. If Bob decided you were ‘lazy’, or ‘not smart’ or ‘careless’, then you were ‘lazy’, or ‘not smart’ or ‘careless’ for the rest of your time on his team. In fact, I was often struck by the differences between our evaluations of a particular engineer; sometimes when we sat down to compare notes, it was almost as if we were talking about two completely different people.
I have a friend I’ll call ‘Jess’, who in our gatherings together would often complain about her boss and colleagues to all assembled. In most of those stories, the person who provoked her ire had done something inexcusably bad, and Jess would conclude that this event reflected badly on the company: that her boss was incompetent, her colleagues were clueless, her team was doomed to irrelevancy and so on so forth. The common feature in these stories was that Jess and co were never to blame: Jess would give us elaborate details of how her and posse were all agreed that boss and ‘bumbling idiot colleague of the week’ were completely at fault, and absolutely stupid, and can you believe what he did?! Over time, however, some of us began to notice that such stories emerged at every job Jess had ever worked at; an initial period of bliss would usually give way to a series of ever horrible stories about her boss and colleagues … and we began to wonder if there were more to these stories than meets the eye.
It’s easy to point fingers at other people and say “look at their flaws!” so here’s one that affects me badly. Perhaps this is something that affects you too.
Say that you’re in your car and some schmuck driver cuts in front of you right before you turn at an intersection. You don’t notice the driver speeding up to you from behind, so this surprises you and you slam the brakes to avoid a collision. Your heart’s going a mile a minute. Depending on the time of day, the mental state you’re in, and how hungry you are, the odds are good that you’d let loose with a string of expletives. “YOU IDIOT!” you shout, and lean on the horn, “WHAT KIND OF ASSHOLE DO YOU THINK YOU ARE? WHERE DID YOU LEARN TO DRIVE?”
Months later, you pull off the exact same manoeuvre while rushing to work; someone honks at you long and loud and angry and you think to yourself: “oh, that wasn’t wise of me, I’m not normally like this — I shouldn’t be so late in the future!”
This is the fundamental attribution error in action — that is, when other people mess up, we have a tendency to attribute it to their innate characteristics; when we mess up, we conclude that ‘it’s just this one time’.
I find myself grappling with this often; when I was 19 I caused a motorcyclist to swerve to avoid my car, and then he rode alongside me, waving his fist in rage; after a few seconds of this he sped up and tried to force me to stop by slowing down ahead. I got angry. We were travelling at around 5 km/h, snarled in heavy traffic. I was in a battered old sedan. I drove into his motorbike with my car, causing him to fall, reversed away and sped off. It was road rage, plain as day. My actions caused me no end of trouble that year; my parents ultimately took away my car and sold it off as punishment. I learnt of the fundamental attribution error years later and immediately formed a connection between the idea and this episode. I've never repeated that mistake again.
What’s the conclusion here? You might think that I’m going to talk about the old nut about how “man is very good at finding fault with others, but not with himself”, and I’ll concede that the nut is true, but it isn’t an interesting observation and it isn’t the topic for this essay. The more interesting question to ask in this situation is: ‘what is the mechanism at work here?’ followed by ‘how can I put that knowledge to good use?’
That question is more interesting because it leads to instrumentally useful conclusions. I think that beyond a certain skill-level, emotional intelligence matters — people who can produce accurate judgment of others are more effective at work. In some domains, at least, it doesn’t seem like you can go very far on vocational skill alone.
Narrative and the Human Brain
Ideas are strange things. You start digging into one idea, and then you quickly realise that the idea is connected to every god-damn thing in the universe.
When I was writing my Putting Mental Models to Practice series I found myself waist-deep in the judgement and decision making subfield of applied psychology. I had not visited this literature for years, and certainly not all together in the same short period — Thinking & Deciding, for instance, was a book that I’d only read bits and pieces of over the past decade.
Reading everything in a period of a few short months made it easier to spot patterns across the books and papers. One of the most interesting patterns that I began to notice was the constant reiteration that our brains are just really good narrative machines.
The human brain is built for stories. We pay attention to stories against our will. We prefer movies to lecture videos; novels to textbooks. We find non-fiction narrative books easier to read compared to non-fiction idea books, and popular science authors know this, which is why they pad their books with anecdotes and background sketches, and splice the exposition of some interesting scientific idea with an irrelevant aside into the childhood of a side-character scientist whose discoveries would change the world of talent/lead us to the discovery of the DNA/illuminate the ways habit holds power over us/give us good sex. Those narrative asides exist as catnip for our distracted brains.
Narrative is good when we’re using it for learning or reading, but it quickly becomes problematic when we are required to make decisions at work. Any time a story we generate matches up to our observations, our brain flags that story as ‘Plausible! I believe this!’ and then shuts down the machinery that is responsible for analysis.
This, of course, explains Bob, and Jane, and me.
It also explains company politics — something that affects our careers more than we’d like to think.
The Career Benefits of Good Character Judgments
I’ve noticed that the people I consider ‘good with people’ tend to not make egregious narrative errors. To use a somewhat awkward phrase, they are ‘well-calibrated judges of characters’. And unlike Bob, or Jane, they rarely peg people into simple categories like “good” or “bad” or “lazy”; at the very least, if they do pass such judgment, it is judgment that is updated over a sufficiently long period of time.
One of the best judges of character I’ve ever worked with was the first HR exec I hired — she was unerringly good at evaluating candidates during our hiring process. This was not to say that she got every judgment right, of course — she was very conservative with her judgments during hiring, for obvious reasons, often deferring from judgment when there was limited information. But when she (occasionally) got things wrong, she usually corrected her impressions within the first two months of the new hire’s probationary period. The thoughtfulness and nuance she imbued every individual assessment she ever did was quite inspiring to me — to this day I model my assessments with the same conservative updating that I saw in action with her.
Skills like this lead to all sorts of career benefits.
Take Bob, for instance. One of the problems that emerged during his tenure was the constraint of who we could assign to his team. Bob’s behaviour was ultimately self-limiting: because he couldn’t work with people he considered ineffective, there was always a chance that a new hire would fail under his management. As a result, Bob’s team was forever lacking in manpower; we could only ever assign a subset of capable software engineers we’d hired under him, out of the fear that they’d churn out based on a bad experience.
Jess’s tendency to embrace negative narratives hurt her more than it did Bob, however. To date, she has not done well in any role that she’s worked at. After an initial period of bliss, her self-generated disgust with her boss and colleagues would cause her to leave; she’s not broken from the pattern yet, and I’ve mostly concluded that her judgments are self-imposed.
(My observation about self-generated narratives apply to myself with regard to Jess and Bob: if their observed behaviour changes, the onus is on me to update my evaluation of their characters).
I’m offering a simplified narrative here, of course. It was occasionally true that Bob’s judgment on a software engineer was accurate, like it was occasionally true that Jess had stumbled into a really toxic work environment. But the percentages matter. Bob’s unwillingness to update his judgments about a subset of candidates lead to very real limitations on who he could work with; Jess being even slightly biased against her boss and colleagues would lead to an eventual self-poisoning against the company.
In the latter situation, it’s not that difficult to imagine how this could happen: pretend that for every potentially negative action you observe at work, you generate a slightly more negative narrative than is strictly necessary (e.g. boss leaves you out when it comes to a plush assignment, but it turns out she was just tired that one time and forgot; you conclude your boss has something against you). As time goes on, your brain continually strengthens this negative narrative of “boss has something against me”, picking only observations that fit. You eventually conclude that your boss hates you — and perhaps this wasn’t true in the beginning, but as you let your negative assessment affect your actions around your boss, what is imagined eventually becomes real.
This discussion is starting to veer into dangerous “is discrimination real or is it self-imagined?” but that’s part of the point, I think: you have to get things like this right if you want to do well in your career. It’s not enough to prevent yourself from forming judgments of people too quickly; you also need to protect yourself against forming your judgments of people too slowly, so that you can take action where it’s warranted.
Or, to put this in concrete terms: “detecting discrimination where there is none is just as harmful as being the subject of discrimination.” (And I say this as someone with experience of systematic discrimination). You don't want to fall into either end of the spectrum.
What To Do About This?
There are two sides to this problem of character judgment. The first side is to figure out how to avoid such errors of judgment when you're doing the judging. The second side of this problem is when you are subject to someone else’s bad judgment — perhaps when you are managing someone like Jess, or when you are embroiled in company politics.
In both these situations, it’s helpful to understand the underlying thinking that leads us to make such mistakes. The academic literature has many names for the biases involved in this effect:
- Myside or confirmation bias — our tendency to favour possible explanations that are already strong, and our tendency to stop searching for (or ignore) evidence that contradicts our chosen beliefs.
- The narrative fallacy — our tendency to look at a series of facts and weave an explanation into them.
- The primacy effect, or, violating the order principle — the order principle states that “the order in which we encounter new evidence should not affect the strength of our final beliefs.” Or, as psychologist Jonathan Baron puts it: “when the order doesn’t matter, the order shouldn’t matter”. People tend to stick to the first conclusion they arrive at; subsequent information doesn’t seem to have as much of an effect on their judgments — hence the name ‘primacy effect’.
- Irrational belief persistence — specifically, the tendency for us to stick to our beliefs, whatever they may be.
Instead of asking you to memorise a list of biases though, I think it's easier to understand the underlying mechanisms of thought that leads to such errors of judgment.
I opened this essay with the observation that our brains are just really good narrative machines: we’re good at coming up with stories, and we’re highly attuned to accepting them as fact. What’s happening here is that we see a handful of behaviours and embrace the first compelling narrative that’s generated for us by our brains. We then stick to that narrative, and either we stop actively seeking contradicting evidence, or we discount the evidence that may cause us to change our minds.
The solution that works for me is simple but difficult to do: whenever you find yourself generating an explanation for someone’s behaviour, generate multiple narratives at the same time. Your boss neglecting to give you the plush assignment could well be because she’s got something against you, or it could be because she forgot. You want to hold both explanations as possibilities in your head, and not commit to either until you have more evidence.
It’s on this note that I find myself in a bit of quandary. My solution to the primacy effect is to generate explanatory narratives very conservatively. I tend to hold no opinion and make no judgment until I think I’ve had enough evidence; every new observation causes me to update previous ones. This means that I’m more effective when I do make my judgments, but it also means that I’m exposed to the primacy effect once I've reached some conclusion. I find that updating existing beliefs to be as difficult as everyone says it is; I have no special techniques to solve for that problem.
I’m curious to know if it’s possible to hold an opinion without fear of belief persistence; so far, most of what I’ve read tells me that the psychologists familiar with this area are sceptical it can be done. Jonathan Haidt, for instance, argues that the best way to get around judgment biases are to create social environments that encourage productive disagreement. He goes so far as to say that academia is one notable example and that — for all its flaws — it does a pretty good job of getting to the truth.
I think my solution of ‘generate multiple narratives’ is in fairly good company though; it’s in line, for instance, with Cognitive Behavioural Therapy’s schtick where the first thing the therapist teaches you is ‘theory of mind’ — the ability to step back from one’s own thoughts and analyse it like you would another person’s.
The big idea here is that what your mind tells you and what is true in reality are two different things; it’s common for people without exposure to this type of thinking to conflate the two, and to act on erroneous judgments like “everyone hates you” and “you’re just going to fail”.
Dealing with Bad Judgments
Once you understand that errors in character judgment often stem from this sort of narrative generation, it becomes easier to deal with other people doing this to you.
How is this so? One skill that’s commonly associated with ‘emotional intelligence’ is the ability to model people accurately in your mind. If you see someone acting out on erroneous judgments, it’s easier for you to respond if you have this intuitive understanding of how such judgments are created in the first place. What’s most likely to have happened is that the person saw you do something, and instead of considering multiple explanations, immediately leapt to a conclusion that now causes them to act badly against you.
Your job is now two-fold: you need to correct this judgment, and you need to do it quickly. The nature of narrative generation is that time will strengthen what they currently believe to be true; if you do not act when the narrative is new, you run the risk of never being able to change their minds.
I want to reiterate this point, because I think it’s so important. When I was teaching management to my replacements at my last company, I often dealt with the natural human tendency to avoid difficult conversations, usually done with the misguided hope that ‘it would all go away’. It’s so common to think “oh, as time goes by and they get to know me, they will see that I am not like that”, and to let the erroneous judgment go. But that’s not how this works. You cannot assume that people are willing to update their appraisal of you. Most people don’t. Most people are like Bob — stubbornly sticking to the first compelling narrative they find.
So how do you get them to reassess you? The first step is to understand that most of the time, people only update their beliefs if you act in a way that is utterly and discontinuously inconsistent with their understanding of your character. Myside bias suggests that most people will ignore all your smaller actions — even if (or especially when) those actions contradict their conclusions. Second, you should also understand that you cannot directly challenge their assessment of you. Doing so implies that they are bad judges of character, and often makes things worse!
What I’ve found to have worked for me is a three step process:
First, I determine the event that has led them to their (erroneous) evaluation of me. This isn’t always simple. With a subordinate, I could schedule a one-on-one meeting and ask questions. But with a peer manager, or with my boss, this is not so easy. The simplest method I’ve discovered is a labelling technique I stole from Chris Voss’s Never Split The Difference — I say something along the lines of “It seems that you think I’m a very lazy person. After all, I didn’t deliver the last project on time.”
And then I wait in silence.
The person I’m talking to will invariably respond in one of two ways: either they’ll say “yes, and I think that reflects badly on you”, or they will say “no, I don’t think that” — both of which opens the door to talking about their judgment of me.
(The reason this method works is because most people are compelled to respond to an assertion made about their person; the silence creates a void that increases their need to address that assertion. For more on this technique, I’d recommend buying and reading Voss’s book.)
If you’re fairly sure you’ve got the underlying cause of the judgment down, then your next step is to address it directly. Don’t say “I think you’re wrong because that assessment of me is inconsistent with my actions P, Q and R” — this implies they’ve committed an error of judgment, and — depending on the person you’re dealing with — might put them on the defensive. Instead, say something like “I accept that you think this of me. What can I do to change your opinion? I’ve managed to deliver on P, Q, and R, but what would you like to see going forward?”
It’s often helpful to follow up this meeting with sporadic updates as a third step: for example, if you’ve done something small that is inconsistent with their prior judgment of your character, you can bring this to their attention by dropping them a note soon after you have performed such an action, saying “Thank you for your feedback. I’m still working on it, as you can see from example P, which I’ve just delivered.” or some other similar pronouncement. This increases the impact of your evidence, small as it may be on its own.
(I don’t mean to say that you should always do this update one-on-one — a more advanced version of this technique would be to loudly proclaim your action in a public sphere internal to your company, one where the person you’re dealing with is sure to notice.)
I think the meta lesson here is that understanding how our brains deal with narratives can often be instrumentally useful: if you know how people come up with erroneous conclusions, then you can use this machinery to your advantage. One of my personal principles is to ‘clarify at the source’ — by which I mean that a large number of interpersonal conflicts may be avoided if you force yourself — and others! — to update their narratives as closely as possible to the source of the conflict.
Of course, now that I’ve finished this essay, I realise that there’s a neat trick involved: everything I’ve written about in this post is already known to man. If we take a step back and squint our eyes at the topic a little, what I’m saying is nothing more nor less than ‘first impressions matter, so make them count’.
On one hand, this makes me a little disappointed — I’ve spent 3000 words on something that my grandfather could’ve told me if I’d ask. But on the other hand I think this is slightly encouraging. The decades may pass but human nature remains the same; for those of you who are geeks, like me, this means that we have a real chance of getting good at dealing with people in our own lifetimes.