I’m reading The Idea Factory1¹ The Idea Factory: Bell Labs and the Great Age of American Innovation; Gertner; Penguin Books; 2012., and somehow, through it, stumbled over an article by Shockley about individual productivity in research.2² On the Statistics of Individual Variations of Productivity in Research Laboratories; Shockley; Proceedings of the ire; 1957. Shockley was a brilliant theoretician. Among many of his accomplishments, he took the transistor – which at the time had barely been invented – and, on his own3³ In a stark contrast to the usual openness and sharing that dominated the Bell Labs culture., improved its design such that it became practically useful.

But that’s not what we are going to talk about here. In that article on individual productivity, we’ll look at three things Shockley does:

1. He sets up rate of publication as a proxy for research productivity, and makes an intuitive argument for its accuracy as a proxy.4⁴ I actually don’t disagree with this operationalisation. I have found that rate of commits is a rather good proxy for individually contributing developer productivity in any organisation I’ve been in. It’s rough, but surprisingly, it seems to work fairly well.
2. He observes through multiple sets of data that a researchers’ publication counts follows a skewed distribution – perhaps log-normal, perhaps power law.5⁵ This means there are a few researchers that contribute most of the productivity. Call them 10× researchers, if you will.
3. He speculates a little on why that might be.6⁶ He uses two completely different arguments, one combinatorial and one probabilistic. He doesn’t attempt to propose one – or any – of them are the correct explanation. He just throws the ideas out there.

In terms of hard, quantitative science, this is junk. There’s nothing here that helps decision-making. The actually trustworthy kernel of the article is the second point, but that’s also just that: an observation, nothing more.

At first I was a little surprised a genius like Shockley would be okay with producing drivel like this.

But then I realised – reading this article gave me at least three different data gathering ideas. Maybe this is how science sometimes has to progress: small, incremental observations, rather than large conceptual leaps. Maybe Shockley’s article plants a seed in me, the outcome of which in turn sparks something in someone else, and five people down the line something really useful is discovered?

Well, probably not. But if there’s at least a small chance, the cost of publishing one’s observations isn’t exactly large. So do it. You never know who you might inspire.