Why research findings cluster like galaxies, with expanses in between
Why do research findings cluster together? In writing a review of a paper the other day I used the word ‘interesting’ about five times in two paragraphs. Looking back on it a few hours later, I realised that in fact the paper wasn’t at all ‘interesting’, in the sense of being important, ground-breaking, or contributing to human knowledge. No, what I meant was that it was using similar methods to those that I use, but in a slightly different way, that I hadn’t thought of. So it was interesting to me – but that wasn’t the sense that I’d used it in the review, where I’d implied that it was objectively interesting, to humanity.
I was on a plane recently, and reading the local newspaper, about Kenyan politicians. If ‘interesting’ is defined as the difference in knowledge (the change between knowledge before and knowledge after) then I would have found that paper much more interesting than my normal English newspaper. Because, rather than telling me about politicians that I mostly know about already, it was telling me about a set of politicians that I don’t know about at all. But even though it was teaching me a lot, I found it pretty hard going.
I think that, when people use the word ‘interesting’, what they really mean is ‘this maps on to something that I know a lot about already, and adds to my knowledge just a little bit’. I find it interesting to read about my favourite football club Arsenal in the morning *because* I already know so much about Arsenal. Which is why people like to have hobbies, and why newspapers will bend over backwards to publish a newspaper article about Kim Kardashian, or Facebook, for example. Because people enjoy reading more about things that are already familiar.
So what does this mean for research? In reviewing research proposals, and papers for publication, the word ‘interesting’ is commonly used as a value judgement by other researchers – and these judgements are the criteria by which it is decided which projects get funded, and which don’t. But if I’m right, and the word ‘interesting’ means what I think it does, then researchers judge the value of work not on how much of an increase it makes to human knowledge (or to my knowledge, as in the Kenyan newspaper example), but rather on how similar it is to things they know about already.
At a conference recently I counted five separate symposia that were all addressing one particular research question: how do infants perceive actions, and what constitutes an action, from an infant’s perspective? This is a legitimate question, to be sure, but it’s one of many areas in my field that, to be honest (and please don’t kill me for saying this), I sometimes think gets, well, perhaps a bit more attention paid to it than it should do, given its objective importance. And I know that many other researchers can similarly think, in their fields, of topics that are similarly over-researched.
So why do research findings cluster together like this? Well, when there are a lot of academics already active within a particular field, then it’s easier to find reviewers who find your proposal ‘interesting’ (according to the definition used above). And it’s easier to find other authors willing to cite your findings, so bumping up your h-factor… So, once a research field gets started, it increasingly attracts more people to it. Like stars clumping together into galaxies.
I think everyone agrees that, as researchers we ought to be scurrying away from each other – looking out the questions that nobody else is researching, at all. (In the same way as in business – where there is an active selectionary pressure to search out space that nobody else is occupying.) But science, and science funding, says it’s built like that – but it often isn’t, actually. We like company. This is why, I think, the scientific landscape looks the way it does. A few spots of light, in which multiple stars are tightly clustered together. With huge, unexplored expanses of blackness in between.
PS If anyone believes that there aren’t any areas of blackness in their fields – well, come and work in mine! Stress in children is a massively under-explored area – I’ve written about some of the big areas where we don’t know much here, here and here.