“Show your working”, they told me at school, and the principle is sound. Later on, a mathematical proof isn’t accepted unless all the stages are published. And a scholarly text cites all the author’s sources. In PhD-writing days, the literature review was a seemingly endless labour, but also a pleasure, in its way. I kept the card index (two large boxes!) that was one of the tangible products of my historical and literary thesis project for a good few years afterwards.
But outside the academy, the bibliography and notes can easily look otiose. The massive lists of works at the back of chunky non-fiction books serve largely rhetorical purposes. (Look at all the stuff I’ve read! Take me seriously!). A smaller, indicative list would usually suffice – though granted it is sometimes illuminating when someone is advancing a controversial view to be able to check whether their sources really say what is claimed.
The combination of the internet, and fast search engines, is surely a game-changer, though. A decade ago, I thought carefully about this when I put out The Rough Guide to the Future. It was a presumptuously wide-ranging, fact-filled tome. But all Rough Guides wanted was a brief suggestion for further reading in each chapter. That seemed fine to me, as it was already possible to write enough details of any key work into the text to allow folks to google it up if they wanted to.
I’ve seen a few more people adopt that stance since – Christopher Potter’s Earthgazers is a recent example I’ve just been reading, and he talks about this explicitly. There’s a list of selected further reading, plus a stripped down bibliography and set of notes that simply includes things he judged might otherwise be hard to identify precisely.
I reckon that’s about right. I didn’t do that with my last book, which cited a big list of recent papers about the microbiome. That was partly because it was a synthesis of very recent science and I thought the field would move on rapidly, so it was important that people who wanted could see exactly when something was published, and so find out more easily whether it had been superseded. I now wonder if I need have bothered, and a few simple attributions in the text would have done the trick.
Alternatively, one can now consign the bibliography to the web, as Stewart Brand did for Whole Earth Discipline a few years ago. That poses a risk, though. You create another job for yourself after you’ve finished the book. Will you have the motivation to finish it? Brand clearly didn’t. The weblink to the notes and references is still up, but they peter out just over half way through, leaving four chapters unannotated. Tsk.
All of which leads up to a just published book, my Cracking Neuroscience. This looks nice – it was a pleasure to work with the designers to turn a stack of words into a nicely illustrated book. But the format has no references at all. And I’m torn. I don’t think it really needs them. On the other hand, I worked hard on the text, and read a lot of contemporary journal papers, as well as quite a few classic papers and books. Do I need to impress that on everyone by publishing a list? I guess not. Pretty well everything I write about can be checked easily on the web.
But I do have this folder on my hard drive which has, let’s see, 370 odd papers and articles in it. So if anyone happens to read the book, and can’t find a source for something I’ve asserted, feel free to email me and I’ll happily dig it out. I’ll say that here because I now wish I’d said something to that effect in the book. Next time…