Ilana Yurkiewicz has just written a piece The downside of politicians talking about science which includes the following passage:
That mentality just doesn’t work in science. Those who are new to a subject are intimidated from asking questions and afraid to disagree. Rather than reason through ideas themselves, they are pressured into accepting conclusions presented as settled and thereby indisputable. But the thing is, nearly everything in science is disputable. The nature of discovery means trying to find the absolute truth – and exposing inconsistencies, thinking through how to reconcile them, and critically analyzing data are all ways to get there. We can’t get very far when curiosity and open inquiry – the hallmarks of good science – are stifled. We are touting the bottom line while discouraging the very steps of the scientific method that get us there.She may have been thinking about politicians asking scientists about issues such as global warming but my experience suggest that anyone asking questions about the foundations of any strongly established scientific field is likely to have problems - and I replied giving details of my experiences:
What we have to realize is that science and politics have fundamentally different goals, and it’s damaging to conflate them. In politics, the aim is to convince others that you are right. Scientists, ideally, should be seeking objective truths. To do so, they need to be receptive to dissent and open to the possibility of being wrong. Science thrives when diverse ways of thinking are welcome.
You are right in suggesting that, despite the ideals of science, newcomers with ideas can be intimidated from asking questions – and the more money and careers that are invested in the specialist area the harder it is to even begin to question the establishment view.
The obvious example relates to the internal working of computers, the ultimate incomprehensible black box. Computers were designed to solve mathematical problems involving numbers that can be precisely specified in advance and which humans were either not good at or could not do it accurately or fast enough. Once the first computers were built there was a mad commercial and academic stampede to get on the bandwagon, and anyone who paused to do any “blue sky” research into alternatives was crushed in the stampede to exploit the idea – and I was one of the scientists who paused to ask awkward questions and as a result became a victim of that stampede.
As a newcomer to computers (I started in 1965) I found myself working on one of the largest commercial sales accounting systems then in existence. I had worked on complex manual information systems and suggested that instead of trying to explicitly model every possible option of a task which involved thousand of different products being sold to millions of customers we produced a user-friendly system that “understood sales” and worked symbiotically with the sales staff, allowing the system to dynamically change as the market changed. Forget it I was told – that's research and we have a job to do ...
By chance I moved to a small but imaginative computer company, carrying out market research for the next generation of hardware. Based on my earlier idea I suggested that the stored program computer model was inappropriate if the prime purpose was to dynamically work with humans in an every changing world – and was asked to develop an alternative prototype model. By the time I had got it working the research division was closed down as the result of a merger. While I was able to continue the work (unfunded) at a university for a time – I eventually gave up the unequal battle against the Goliath of the computer establishment. Too much was already invested for anyone to consider alternatives which might throw doubts on the establishment.
Forget whether what I proposed was right or wrong (you can get details on www.trapped-by-the-box.blogspot.com). If you want to build an electronic aid to help humans whose brain evolved as hunter gatherers, who live in a poorly understood world (in the scientific and mathematical sense) do you really think that a good starting point is a model which requires precise predefinition and which uses numbers to represent data, operations on that data, and the addresses at which that data is stored? If in 1945 you had asked the proverbial Irishman the way forward in human/electronic box interaction he would undoubtedly have said “If I was you I would not start from here”
My foolishness was to suggest that, in effect, if you wanted an intelligent information processing tool to work with humans it should (at the innermost level) it should be word (rather than number) based and dynamically open-ended to cope with real world (i.e. not pre-defined) problems. It should also start (like a new-borne child) by knowing nothing – and quickly “learn” what its human master wants it to do. I learnt the hard way that there is no point (if you want a life) in questioning the scientific foundations of a establishment as powerful as the computer lobby – even if it is very obvious that no-one in the early days had time to ask if the algorithm-oriented number-based model they were using was the best one for working with people. I am sure there must have been thousands of other people who have had similar ideas to mine who never got them out of the starting blocks - and as a result had successful careers swimming with the tide.