Warren Buffett, perhaps the greatest investor of all times, prefers his holdings to have what he calls a `moat’; some form of protection that sets a business apart from the competition, similarly to how a moat protects a castle from intrusion.1 The economic moat should function by making it impossible for rivals and new entrants to outcompete the established one. It can be a cost advantage, a well-known brand, a cemented connection with customers, infrastructure already in place, or special know-how, possibly a patent portfolio.
If you believe in keeping science open, then you should be wary when you sense something similar happening in research.
A PhD Comics strip expresses a similar effect for the individual PhD student as a rising then plunging curve of irreplaceability.2 As long as you’re the world’s top expert in your (hopefully not too obscure) topic, you’re super-indispensable. As soon as you’ve documented your work, you’re expendable.
I’ve come across a few professors who are authorities in their fields and command great respect by possessing skills their colleagues don’t. Most of them are honest and it is not my intention to call individual cases out but to warn you to recognise a certain pattern.
When somebody goes `it is more of an art than a science’ telling about their subject, it can mean different things. A topic might be too big, even inherently unwieldy to be summarised in a few easy-to-follow steps. This saying was how Philip Maini introduced the mathematical modelling of natural phenomena to us. But normally the saying means that the person spent time playing around with their new method but cannot yet summarise and communicate how to use it in general, in what cases the method applies and when it fails.
The danger is, it might be tempting for some to become the guru in a field by intentionally obfuscating some technique or trick which could otherwise be presented transparently. Just recall the so-called Shirky principle: `Institutions will try to preserve the problem to which they are the solution.’ I imagine it is possible to build a scientific career on developing a new method and reaping the rewards for applying it to many problems, but not giving others access to the method itself.
The latter might even happen by chance, if the first publication of the new methodology (for fear of rejection) omitted its weaknesses. Over time, the authors can develop a good feel for the applicability of the result, publish the success stories (but never the constraints of applicability), while their colleagues are frustrated by the repeated application of the methodology to cases it cannot solve.
In an ideal world, the goal of a researcher should be to further scientific knowledge by the greatest extent possible. This must have priority over one’s ego and longing to land a good job. Therefore, respect should be reserved for the ones whose achievements foster the work of others, and not for those who form bottlenecks in the flow of knowledge.
Donald Knuth wrote in a foreword to a book3, `Science is what we understand well enough to explain to a computer. Art is everything else we do.’ I say:
Long live who turns art into boring science!
1The first written mention seems to be in Warren Buffett’s Letter to Berkshire shareholders for the year 1986.
2How irreplaceable are you? PhD Comics, 20th March 2013.
3A=B by Marko Petkovšek, Herbert S. Wilf and Doron Zeilberger, A K Peters/CRC Press, 1996. The book is also available for free download.