Everyone needs a strategy to keep up with interesting papers
A year ago, I decided to keep track of all the papers I read over the course of one week. ‘Read’ here, for the most part, actually means skimmed. Although for intense papers, deep reading pays dividends, in most cases, thoroughly taking in every word leads to diminishing returns.
I read 58 papers that week, and plotted my results graphically. The charts confirmed that my reading was dominated by higher-impact specialist journals, Angewandte Chemie and the Journal of the American Chemical Society. Review journals and mid-tier organic chemistry journals soon followed in frequency. Most of my insights were hot off the press, 54% having been published that year. The majority (60%) consisted of methodology articles – the kind that I might use directly in the lab.
Everyone has their own ways of managing all these papers. I use a combination of an RSS reader and a reference manager. I follow a series of journals, authors, and keyword alerts by RSS. Papers swarm in almost constantly, so I try to accept or reject them from my reading list as quickly as possible. I’ve even considered building a machine learning bot to filter out articles that definitely do not match my interests, although I’ll save that for a rainy year. After reading, any paper I have any chance of remembering is saved in my reference manager with a hefty collection of tags to enhance retrieval. I think of this as my personal library. I also use a tool called Pocket as a holding pen to save articles for later perusal.
Doubtless, one week is not sufficient to illuminate general trends about my own reading habits. In those seven days, I read seven papers outside my field of organic chemistry, which I’m sure is more than usual. Although I find it useful to read broadly, learning precise details from scholarly papers is not very helpful without bigger picture familiarity.
Many of the articles (25, in fact) appeared as new alerts in my RSS reader. Most of the others were found by searching for specific terms or transformations. The remaining six came via social media, from friends either directly sending me papers they knew I’d share an interest in, or broadly disseminating what they were reading or writing. More recently, I’ve found my reference manager’s ‘suggestions’ feature is scarily good at predicting what I’ll enthuse over, usually proffering papers that I have already read and loved, but just not saved in my library.
I calculated some crude statistics to look for bias in my reading. Authors whose work I was familiar with made up 64% of the works. Some biases I already knew about – for example, through RSS author alerts. Perhaps unsurprisingly, nearly half the papers involved researchers based in the UK or US. As well as my own biases, these measures reflect how science is published – the most popular articles originate from well-known, established groups at institutions that attract funding. Of course, I am also more likely to familiarise myself with research by researchers I know personally, even when their work is not especially relevant to my own.
Keeping up with the literature can be a chore, a terror, or a joy. In my earliest days of scholarhood, I learned from exemplary mentors to read as much and as widely as possible. I set about the task, unfortunately before learning how to read papers most effectively. After much stress and frustration, I scaled back the number of journals I was trying to read, adding more, gradually, over the years. Matters changed again when my own research intensified, and there were barely enough hours in the day to keep up with even the most relevant works (or luxuries like sleep), never mind the general literature. However, after defecting to industry and the productivity boost that comes with it, 58 papers at the time of my experiment feels like a slow week.
I now realise that era was my zenith of literature reading so far. After all, 58 papers per week totals over 3000 a year – and I’m certain I don’t get through that many at the moment. Mid-way through 2018, I moved into a different lab, and my reading circumstances changed once more. Where I had previously had occasional spare moments, during which I would skim through my RSS reader, these free moments no longer occur in my daily workflow. I’m much happier with many competing tasks on the go at any instant, and unexpected errands frequently materialising. As well as new work meaning different kinds of papers were suddenly much more interesting, my reading habits were forced to transform from ‘mostly in lab’ to ‘almost entirely at home’.
After musing on this, I realise I haven’t checked my RSS reader in several days. 1703 unread articles… it will be a busy Christmas!