For the past year, I have been stuck in a research slump. One year ago, I started applying for jobs; and with fellowship applications in the mix too, the whole period from October to December disappeared into the mire of proposals, cover letters and research statements. When I emerged after Christmas, it was into the familiar period of interminable waiting, broken only by a few welcome emails inviting me to interviews and many more not so welcome emails telling me that my research was not worth paying for.

I was spat out at the end of the job cycle in mid-February with one offer and two positions where I was the first runner-up. Having no time or desire to wait for news from those places, I accepted the solid offer, and gladly took some feedback from one of the people who had interviewed for one of the other positions. He told me that the main point against me was that I had not spoken possessively enough about my work; it was not clear what I had done versus what my collaborators had done.

This was a fair assessment, both of my performance in the interview and of my work as a whole. Like many people, I struggle with talking positively about my own work; in fact, I tend to assume that whoever I’m talking to is a better physicist than me and knows more about my own work than I do, which often leads me to talk very dismissively about anything I have done. It comes from a place of insecurity, of course, in much the same way that some people find it hard to take compliments on their appearance. Fortunately, there’s a cure for this, which I’ll get on to later.

Regardless of how I talk about my work, I knew there was a grain of truth in the insinuation that I had not actually done much of the work that my name is attached to. My own honest assessment of my research skill is that I am very good at doing things that other people suggest and then writing about it afterwards; less good at actually coming up with ideas in the first place.

I see this as the crucial transition that one must make at some point post-PhD if one is to survive in academia (or at least, crucial to me). But initially, I didn’t know what to do about it. I spent most of the time between February and the summer fiddling around with inconsequential stuff; trying to read papers, making notes, organising journal clubs, but not actually doing anything.

In August I went to Chicago to attend a conference at KICP. In fact, I was an invited speaker, but my talk was something of a disaster (jet-lag can take some of the blame, along with a big miscalculation on my part about the audience I was pitching the talk for). In general, I’m not a fan of conferences; I find travel inherently tiring, as well as talking about my work with endless new people — see the aforementioned internalised belief that it’s not up to snuff. Coupled with my lack of research productivity, I was not expecting to have a good time.

However, the KICP workshop was, on the whole, above average, thanks mostly to the organisers who selected very good speakers, and also made time for small breakout sessions, which I particularly enjoyed. By the end of the conference I was starting to perk up, and to actually think about what other people were saying. A stray comment by Anowar Shajib struck me (I’ll paraphrase because I can’t remember it word-for-word): “strong lensing constraints on cosmological parameters are interesting because the joint posteriors tend to be orthogonal to those from other probes”.

This got me thinking. During my PhD I worked a lot on interacting dark energy models. I wondered if strong lensing data could provide new insights into these models, by virtue of the different correlations between the cosmological parameters compared to the datasets I had used before.

There was also a breakout session on combining strong and weak lensing information at the likelihood level. I didn’t attend this session, but I did join the Slack channel created afterwards. I realised that it would be relatively simple to implement the TDCOSMO likelihood in Cobaya, and then use it to get constraints on whatever dark energy model I wanted. I could even write up the results as a single-author paper.

Still, I hesitated. Would it be worth it? What would people think?

These are the kind of thoughts that one needs to kill to succeed in research. Or, if you prefer some nuance, these are the kind of thoughts that should be given their due consideration — since there are some ideas which are decidedly not worth pursuing — and then put away, never to be thought about again once the decision is made.

The best advice I ever got as a PhD student was don’t wait for permission. This was told to me by a PhD student a couple of years ahead of me at ICG. He was the sort of PhD student who conducted research like he was already a postdoc. And he was right. I realised that no one was going to say yes, you may do this project — but at the same time, no one was going to say no. I had the power to do what I wanted.

Serendipitously, at the same time that I was thinking about this, I stumbled across the following quote:

Finish something. Anything. Stop researching, planning, and preparing to do the work and just do the work. It doesn’t matter how good or how bad it is. You don’t need to set the world on fire with your first try. You just need to prove to yourself that you have what it takes to produce something. There are no artists, athletes, entrepreneurs or scientists who became great by half-finishing their work. Stop debating what you should make and just make something.

James Clear

(When I saw it, this quote was unattributed; after searching for it just now I’ve found that it’s by James Clear, author of Atomic Habits. I’ve heard a lot about that book and may just have to read it now.)

This manifesto spoke deeply to me and my situation. It was enough to galvanise me into actually working on and finishing the project I had in mind. Suddenly, my research slump was over. I had a clear task in front of me, with decent scientific justification and the prospect of a publishable result. Furthermore, I did it by myself, from conception to execution. Perhaps there was something of a chip on my shoulder from that interview feedback, but if so it was a good thing. I now have a piece of work that I can point to and say without a shadow of a doubt: that’s mine.

So about those thoughts of inadequacy, that little voice which says “your work isn’t good enough and neither are you”. The way to tackle it (I think) is to contradict it as often as you can. And contradiction by action is part of the process, but literally saying out loud to yourself “my work is good and I am a good scientist” is an excellent way to eventually get yourself believing it.

Yesterday I was talking with a friend about an ongoing project, and specifically about some software written by other people which we’d like to replicate for our use-case. I said something like “I don’t mean to brag, but I’m pretty sure I could write that code easily”. Did I say it with a heavy tinge of sarcasm? Maybe. But I still said it out loud. The self-belief was verbally reinforced.

So, I don’t mean to brag, but I’ve written a single-author paper. It’s neat, concise, advertises my implementation of a likelihood which should be of wide interest to the community, and presents a thought-provoking result.

For once, then, something.

---

Update: my paper was recently published in Monthly Notices of the Royal Astronomical Society: Letters.