When Do I Start Knowing What I’m Doing?

xkcd #451: Imposter

Nah, we’re all too busy worrying about whether we’re experts in our field. (Source: xkcd)

When my advisor informed her assembled advisees that I was the group’s “machine learning expert,” I nearly choked. I thought I had a pretty good idea of what expertise looked like. An expert possesses a deep, intuitive understanding of his or her subject. An expert exudes confidence in his or her abilities and reputation. An expert fields detailed questions without batting an eyelid. What an expert most certainly does not look like, I thought, is a clueless amateur of a Ph.D. student.

My lofty image of expertise was not my own invention – our society has an unhealthy tendency to fetishize experts. We see the degree of knowledge possessed by professors and analysts and TED speakers as almost mystical. We speak in awed whispers of their brilliance and intuition. And of course, the praise is often well-deserved; I don’t mean to suggest that there is no such thing as expertise. But the way we idolize experts does great damage to experts and novices alike.

I experienced the troubling fruits of this mindset within my own academic career. As an impressionable undergraduate in computer science, I absorbed the zeitgeist of expert-worship. I saw tools that could extract knowledge from webpages or predict how a political conflict story would end, and I believed the graduate students who built them must be superhuman masters with an unfathomable depth of understanding. As I slogged through several undergraduate research projects, I kept waiting for the moment when I would “get it” – when I would understand how these wizards knew what algorithm to try or how to rigorously evaluate a system. But the day I started graduate school, I was still hoarding the deep, dark secret that I was not one of those superhuman masters. I had no idea what I was doing.

If anything, my first two years of grad school only reinforced my sense of cluelessness. I nodded along sagely as my research advisor suggested “extending generalized linear models of neurons to include hidden variables,” desperately trying to work out what a generalized linear model was and where its variables might be hiding. When I read papers, my eyes glazed over at the unfamiliar statistics equations and the mystifying results graphs. Even at the conference where I published my first paper, I squinted muzzily at each speaker’s slides, glancing around nervously to check whether anyone else was as confused as I was.

Of course, I was not alone in this experience; what’s remarkable about it is that it is so common as to be unremarkable. Three weeks into my program, one of my fellow students joked over dinner, “So, how’s your impostor syndrome doing this week?” Yes, that was it – impostor syndrome. The persistent belief that no matter how much you accomplish, your success is all a sham – a mixture of luck and overly kind assessment – and it’s just a matter of time before it all collapses and you’re exposed as a faker. The problem runs rampant among academics, who are finally beginning to publicly discuss the phenomenon and its causes. Even with this growing recognition of the problem, though, one of the factors that I found most important has also been one of the least-discussed: the insidious myth of expertise.

The way our society idolizes experts prevents most people, even the most qualified, from ever believing they are one. We paint experts as godlike creatures living lives of clarity and certainty, with canonically correct solutions leaping fully-formed into their minds. Given this image, I could only assume that, flawed as I knew myself to be, I couldn’t possibly be a member of that exclusive club. I looked up starry-eyed at peers and professors and thought, “I must not be a competent professional, because I, unlike these masters, am riddled with doubt and ignorance.”

The myth of expertise creates a heavy burden for experts, but the damage extends to novices, too. For many people, the very image of a superhuman researcher, or for that matter a superhuman lawyer or singer or goalie, could easily stop them from even bothering to try. I was fortunate to grow up with a family and a cultural identity that encouraged me to succeed academically, which counterbalanced my self-doubt enough to push me through. But if we depict experts as hyper-advanced life-forms, then many novices who are less lucky will always stay novices; they’ll be no more motivated to strive for expertise than a fish would be to morph into a human.

Recently, though, I had something of a revelation. It came at my second conference a few months ago, where I once again sat through countless talks on computer science and language. This time, the experience was noticeably different: each time I met a new colleague, not only could I understand the outline of their research, but I could guess the techniques and resources they were using, and identify some of the challenges they would face. The uncertainty and ignorance hadn’t gone away – there were still plenty of things I didn’t understand. But with a bit of a shock, I realized that somewhere along the line, without my noticing, my accumulated experience had made me conversant in my field.

And if that had turned out to be possible, I began to wonder whether certainty was what separated me from the pros, after all. It was not, I realized, that they intuited everything with certainty; they all had their own doubts and questions, too. The questions that bewildered me had bewildered them long ago, and they had since moved on to bewilderments that hadn’t even occurred to me yet. They simply had more experience – experience that happened to give them confidence on the particular questions where I was still ignorant. With a little more experience, perhaps I, too, could move my confusion to a more profound level.

I have no idea what I'm doing (Chemistry Dog)

This may not be what a scientific expert looks like, but it sure is what scientific experts feel like.

In fact, the same holds true for becoming an expert manager, animal trainer, or anything else. When extremely experienced people say they still have much to learn, they’re not just trying to look humble; that really is the most salient experience of being a relative expert. To be an expert is not to live with certainty and confidence, but to see an ever-advancing horizon of new and daunting questions and tasks, none of which you know you’re qualified to handle, but any of which would have stopped you in your tracks two years earlier. As science and multimedia expert Jeff Lieberman said at ComSciCon 2014, “What craft comes down to is getting comfortable with not knowing what you’re doing…. Do your craft constantly while standing in the unknown.”

I don’t believe the discomfort of impostor syndrome is a necessary part of being any kind of expert. We just need to reiterate, verbally and in writing, that we all have our uncertainties, and that there is no magic “expert-level confidence” badge that the “real” experts earn. If we all agree to humanize our experts – to admit that, as The Guardian’s Oliver Burkeman recently wrote, “everyone is totally just winging it, all the time” – we’ll all be a lot better off.

So let me be the first to say: I still feel like I have no idea what I’m doing, and maybe I’ll always feel that way. But I know more than I did twelve months ago – my confusion has migrated to a higher level. I’m learning. And that will have to be enough.


6 thoughts on “When Do I Start Knowing What I’m Doing?

  1. Thank you for this article, Jesse! Grad school can be a very hard time particularly due to the lack of moments of objective and comparable self-evaluation and the corresponding amount of subjectiveness and loneliness that involves much of our work. I have also felt clueless at the beginning of my PhD studies and feel like gradually – after a lot of time – I begin to understand how things work and that even such vast amounts of research topics and ideas hoovering around us can end up forming a big picture with a structure that we can somewhat understand and discuss. I think the fact that our difficulties and failures are so obvious to us while our role model’s are so hidden does hinder our own progress and ambition. I have noticed that learning of the occasional failures and missteps of my own mentors and role models have made me feel closer to them, in the sense that I feel I can also achieve feats like theirs and at the same time that I empathyze more with them and don’t feel as ‘scared’ of showing and discussing the vulnerabilities of my knowledge and my work. Sadly, those moments mostly happened involuntarily. Perhaps a good idea would be to organize a meeting or a talk where Professors and/or grad students share some of their bad options, failures and missteps.

    See you and the rest of the gang in a few weeks in Pittsburgh! 😉



  2. Your blog post brought a couple of opposing ideas to mind.

    Years ago I encountered a segment on a public radio show claiming that young people’s experience with video games may prepare them for career success. One of the reasons put forth is that gamers learn that if you put in enough hours trying to get past an impediment to achieving a goal, you are usually successful. This makes them resilient in the face of setbacks and gives then a healthy skepticism about expert pronouncements, as the difference between themselves and an expert is only a matter of hours of experience. As more and more members of society have gaming experience, perhaps experts will be hard-pressed to find a pedestal to stand on.

    If I look at the debates about climate change today, I see cynical people enriching themselves by denigrating science and its expert practitioners to the detriment of society at large. If society reduces what little remaining respect it holds for scientific expertise, the results could be catastrophic.

    Liked by 1 person

    • Yeah, that’s an interesting point. When I was writing this, I was a bit concerned that it’d sound like I was saying we shouldn’t trust experts. And on climate change and other significant global issues, we certainly should trust the best information we’ve got. Ultimately, though, I don’t think this is an argument for trusting expertise less; it’s just an argument for humanizing experts (scientific and otherwise) more – for discussing more publicly the uncertainties and inner struggles that expert scientists and improv performers and whatnot have.

      As for the specific issue of trust in scientific expertise, there’s actually quite a bit of social science research demonstrating that people do respect science as an enterprise – they just conclude that the particular scientists whose recommendations disagree with their values must be bad scientists. (This is a particularly strong thread in the work of Dan Kahan at the Cultural Cognition Project.) Given that, I think humanizing the scientific experts can only help – it makes them more relatable and harder to dismiss as crackpots and politically motivated idiots.


  3. This hits very close to home. As an undergraduate researcher, I wallow in feelings of inadequacy about my skills, my personal presentation, and about my research in general. Being the lowest on the totem pole has probably contributed to my feelings of anxiety — I’m sometimes asked at inopportune times to clean large piles of glassware or go fetch a heavy, dangerous nitrogen tank.

    But I see my own imposter syndrome, my own persistent thoughts of worthlessness, reflected in the grad students, who often describe their work as “derping around” and are quick to admit they have “no fucking idea” what they’re doing. Even my calm, Jedi-like PI has told me that research involves a lot of (read: mostly) failure.

    So, I realize, my transformation from a young chemist to a wizened one may not be limited to gaining more lab skills, and becoming more adept at analyzing data and presenting before my peers — no, it may also include coming to terms with my imposter syndrome, looking it in the eyes and saying, “you don’t define me. My research is meaningful and my paper kicks ass.”

    Thank you for this blog post.


  4. You absolutely nailed this! I feel like this all the damn time, especially when international students rely on me to interpret texts for them even when they’re studying a completely different subset of my discipline. I of course help as much as I can – I can’t imagine how bad this ‘imposter syndrome’ must be from an international student’s perspective.


  5. Pingback: Life Jackets | Linguistics PhD


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 )

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