Bring Out The Translator - We Need to Talk to a Scientist

Nick Clarke

Topics:

Data Science

Despite the best efforts of many generations, you get the feeling there are some global problems so entrenched as to be beyond all hope. And no, I’m not talking about world peace, hunger or equal rights. The really big challenge is how to make a scientist remember that when they want non-scientists to listen to them, they have to change how they talk.† How hard can it be – it’s not rocket science, which is of course why they struggle. The rise and rise in the profile of data science, or analytics, or whatever it’s called this week, in business makes this perennial communication barrier a really big deal.

I was compelled to write today’s blog by hearing a news item debating the dangers of air pollution. At one point the expert witness scientist made a bold assertion as to his preferred solution, and was challenged by the host as to the realism of his idea, given the cost implications. His reply was that it “depended upon what you were titrating that difficulty against.” When the host gently suggested that this was not a commonly used word, you could hear the barely concealed irritation in the scientist’s voice in having to adjust his settings for the common man or woman in the street. (By the way, he meant it depended on how you wanted to balance different interests.)

The result of this exchange, of course, is that a neutral listener is more likely to question whether the wider position taken by the scientist is, for the sake of a better phrase, grounded in the “real world”. And whether his position is taken from a perspective that matches their own concerns and point of view. It only takes an odd word or unusual phrase to quickly alienate and appear out of touch. And my experience tells me that people, being people, the inner scientist is largely unchanged from one generation to the next.

Of course each generation has had their brilliant science communicators and popularisers, none more so than at the present. But that’s a bit different. They do two things brilliantly well. They show that science is important, and that it can be accessible, cool and fun. But they are not the people working day in day out on public policy, or drug discovery, water purification or manufacturing. They don’t help build up the trust between the person brandishing the numbers in their research, and the director who has to justify the investment choices to the board. If they can’t establish a common language, how likely is it that the best decision will be made?

The language of science, numbers and rigour is wonderful, and exists for a good reason. It holds subtle concepts with the precision needed to do science. But when you take the great things you have done using that language out into the world, and want to make a difference, you are no longer doing science. You need a different language, fit for a different purpose. A scientist doesn’t need to be taught this different language, because it’s the language of everyday life. I suspect the struggle is accepting the sacrifice in precision which making that switch inevitably requires.

I look at the clarion call from the business media for teams of mathematicians and statisticians to crunch the big data, find the insights and through them direct the future of business as a back seat data-driver. And I wonder, how will the trust between the data scientists and the investors be established? What language will it be done in? Is there a common perspective of what an insight actually is, and which are important? The larger, more complex and unstructured the big data becomes, the greater the uncertainty in interpretation and the bigger the trust barrier to action becomes. It will take a layer of expert translators, conversant in both languages to bring the two sides together, and allow the value in that data to see the light of day.

Now there’s a balance of interests to be titrated for you!

† Personal Declaration: Hi, my name’s Nick and I’m a scientist. Sometimes I forget that other people are not scientists. Such as when I talked to my young niece about the Newtonian spectrum instead of the colours of the rainbow. Still gets a laugh out of the family, 10 years later.