What is this thing called science? Eppendorf envy, hospitality sciences, and medicine.
I am half-way through reading The Reflective Practitioner, by Donald Schon. I will write more about it another time, but a topic he deals with— the relation between science and the professions— surfaces in a couple of articles / letters in this week’s Lancet. For instance, John Wennberg and colleagues, refer to a subject called ‘delivery science’. I do not think they mean Amazon, but they might, because I have argued (article pdf) that a lot of clinical research should be outsourced to Amazon (and I am a big fan of Wennberg who should have been awarded a Nobel). Also in the Lancet, an article on Rifat Atun, quotes him in a similar vein. He does not use the term ‘delivery science’, but when talking about cross-country comparative studies, and the study of the effects of demographic change, he is quoted as saying we need a science to understand such things. What sort of science is this?
What of Schon? Well, he documents how science is used by professions to maintain or develop status. I think he is right about this, and he is also right about how little we know about the sort of knowledge that underpins many areas of expertise. Schon quotes one of my heroes, Herbert Simon:
Engineering schools have become schools of physics and mathematics; medical schools have become schools of biological science; business schools have become schools of finite mathematics.
I like this, and used this quote in an article I wrote a long time back. I have also poked fun at the way subjects such as dermatology or pharmacology, have been re-branded ‘dermatological sciences’ or ‘pharmacological sciences’. I find this just plain silly. With apologies to Freud, and penis envy, I call it eppendorf envy.
The issue underlying all of this, is that there are large areas of rationality that underpin medical practice, but which are not science. I am quite happy with physics and biology being science, but economics, ethics, and most clinical practice is not science. RCTs are useful, but doing them is not usually science, rather they are just exercises in logistics: they are important, but require, as a statistician once said to me, the sorts of professional skills that doctors undertake every time they see patients. You just shouldn’t count them as scientific enquiry. If you develop new statistical analytical methods, count these as scientific, but not when others wheel them out to get a job done. If I invent a new drug, I am doing science; if I prescribe the same drug to a patient, I am practising medicine.
The problem Schon identifies— as does Herb Simon— is that we seem terrified to admit that something can be worthwhile and meaningful, but not science. Science is a subset of rationality. Much of medicine is about design: building things, with certain constraints, often on the basis of expert opinion, and tinkering, as well as more formal studies.