Choosing What To Read

Blog archive

Like many people, I spend a lot of time reading. Since there is much more that I could read than there is time for, I have to make choices about what I should read. In this short post I’m going to outline what my reading choices have become and why. My aim isn’t to suggest that anyone else should make the same choices – indeed, they’re intensely personal – but, rather, that thinking about such choices is worthwhile.

Unlike many people, these days I intentionally read very little fiction. Saying that I’m a literary philistine [1] out loud is close to a taboo and for many years I downplayed it. These days, I’m much less worried about this admission, in part because the alternative I have gravitated towards has given me greater personal satisfaction and, I hope, insights: history.

Personally, I find that history contains many more surprises than authors can get away with in fiction. We have a strong sense of what is plausible: this constrains writers, even in supposedly fantastical settings, and readers tend to dislike violations of it [2]. Our sense of what is plausible is in large part defined by our life experience — but most of our lives take part within surprisingly narrow parameters.

For example, since our day-to-day lives are fairly ordered and predictable, we assume that important people can carry out ordered and predictable plans — but the real world is complex, messy, and contingent in a way that makes this impossible. Most of us don’t know (and I hope will never know!) what war is like and that makes us surprisingly resistant to descriptions of the real thing. We assume that morality is absolute and evolves in a straight line [3], so we fail to realise that people in the past held fundamentally different assumptions about morality [4]. And so on.

History, I’ve slowly discovered, is full of stories that seem utterly implausible, until you realise that they’re true. The charge of the light brigade was based on a misunderstood command, something which turns out to be rather common in military history. Our idea of a fairytale castle is based on the whim of an eccentric, profligate Bavarian King who later died in mysterious circumstances. The 2000 year old Pantheon is still the largest unsupported concrete dome in the world, and we haven’t quite worked out how to produce concrete of such quality [5]. The man who led the armed uprising that caused the Costa Rican civil war won, abolished the army, voluntarily stepped down as dictator after 18 months, served two much later terms as president, and at 65 years old stood on the tarmac with a machine gun until a plane’s hijackers surrendered.

Once I realised that history actively expanded my thinking, I shifted nearly all the time that I once spent reading fiction to history. This was not as easy as it sounds. Every time I read about a new historical era, or location, I spend considerable time baffled by the new setting. In other words the “setup” costs are considerable. But, for me, the results have been more than worth it.

Much of my reading, though, is not historical. Like many people I take an active interest in the world around me. For many years I read a great deal of news, nearly all via newspapers [6]. However, probably due to the effects of the internet, I slowly noticed that the quality of journalism in newspapers was in decline.

I also noticed how depressing the “news” generally is: as the cliche goes, bad news sells, so that’s what tends to get written about. I realised that not only was I spending hours a week reading about horrible events (murders, corruption, war, etc.) but that I was being heavily influenced by this — I thought the world was a worse place than it really is [7].

Now I skim the news on a daily basis, but I filter out everything that seems “transitory”, particularly if it’s depressing. Instead, I’m looking to identify and understand longer term trends, not (unless they’re major) one-off events. The most consistent publication I’ve found that matches what I want is the Economist. I understand, and can largely correct for, its biases. Since its writing about areas I know well is consistently good, I am reasonably confident that generalises to those areas I don’t know well and wish to learn more about.

I read a small number of opinion pieces in newspapers and a moderate number of blogs in the same sort of vein. At one point I was carefree in such reading, but I realised that I was in danger of only reading those pieces that confirmed, positively or negatively, my existing world-view [8]. At best this was a waste of my time, and at worse it was making it less likely that I would revise incorrect opinions. These days I select the people I read carefully, but not for their ideology — rather, I’m looking for thoughtful people who can challenge my thinking in a carefully justified way. Not many weeks go by without this causing me, even if only slightly, to adjust an opinion.

Perhaps the major portion of my reading relates to my job. Although I’m formally an academic I read, mostly daily, a large number of non-academic programming articles and blogs. Most obviously, these keep me up-to date on real-world trends. Less obviously, I find them an invaluable source of inspiration for identifying problems.

It’s easy as an academic to propose things one already knows as solutions, but it’s much harder to know if they solve problems anyone might care about. Since I continually look for problems, I occasionally find a (sometimes partial) match with something I know about. My hit rate for this is low. I notice a match a handful of times a year at most, and in the past I’ve often gone for much longer than that without a single match. Despite that, I still consider this time exceptionally well spent.

I also read academic papers, mostly in the field of programming languages. I no longer consistently follow any conferences or journals. Partly this is because they’re too intermittent and I prefer to space my reading out over time. But, mostly, it’s because peer review is inherently conservative and tends to reject work in unfamiliar areas. When reading research speculatively, I am more interested in increasing the breadth of my knowledge than I am the depth [9].

Over time, I’ve settled upon checking daily two or three arXiv lists (e.g. cs.PL). Based almost entirely on the title, I’ll skim perhaps 3-5 papers a week this way, and read in some detail perhaps 2-3 a month. I don’t do as good a job at this as I probably should, and I find it the easiest sort of reading to skip when I feel busy, but I have still found this very helpful over time.

Finally, I do a certain amount of reviewing as part my job. I used to do a lot more of this. Slowly, I realised that not only was I doing more of this than was useful to me, but in so doing I was denying other people the opportunity to do so [10]. I still perform slightly more reviews than I receive reviews, but I keep a fairly strict cap on it.

So there you have it: the “algorithm” I use to pick what I read. It’s very much a reflection of my biases, interests, and needs, and I don’t recommend that anyone follows this algorithm directly. For example, while fiction doesn’t do much for me, it might do a great deal for you — neither you or I is “right” or “wrong”. However, what I hope you’ve got from this post is a sense that knowing what you like to read is useful, and knowing why you like to read it doubly so. I’ve found it very useful to reflect on this, and perhaps you might too.

Footnotes

Comments