Today’s issue is about words, and about the slightly peculiar phrases that are constantly passing the lips of scientists. “It seems unlikely that…” “it is reasonable to assume that…” “we can say with a high degree of confidence that…”
Any scientist worth their salt will avoid pronouncing something to be absolutely true[1], and what we get instead are statements like those above. Understandably, this seems strange to some people. It’s not the way that people talk in casual conversation, after all. It sounds as if they have something to hide! Often this is interpreted by the public as being a sign that scientists aren’t really sure about what they are saying. Some people of malicious intent will then insist loudly that any findings of researchers that might impact them negatively cannot be trusted. Examples include climate change deniers (‘so there’s no need to use less oil!’) and politicians (‘no, 90% of New Zealand’s lowland waterways are not polluted!). A scientist might say that a particular aeroplane design has a failure rate in turbulence of one in ten to the eight. What they mean is: the plane won’t fail. I would like to explain in this column why scientists talk this way, and why it is necessary for them to keep doing so.
Now, in most cases I am a strong advocate of scientists exercising humility and being themselves the ones responsible for communicating clearly with the public. But this is one mode of speech which I believe that scientists must not stop using, even though it can be difficult to understand, and I’ll explain why. It takes a wee while to explain, so bear with me.
The source of this turn of phrase is a curious quirk of the scientific method. Science is our best tool for finding out what is fact, but it is only capable of proving general behaviours false – it is not capable of proving them to be true. We can show that Newton’s Laws of motion do not correctly describe objects moving at near-light speeds. But we cannot show that general relativity always gives the correct answer. It is impossible to test every circumstance, even in a well controlled laboratory experiment, so all that we can reliably say is that in all the times we have tested it so far, general relativity has given the right answer[2].
To complicate matters, many of the questions that scientists seek to answer involve processes that are either complex or statistically random. The best that the scientific method can hope for in these circumstances is to give an accurate statistical average. This number won’t be the actual figure found in any particular example set, except by chance or when the sample size is infinite – clearly not a physical[3] circumstance!
So scientists, when answering any question, must give only their best estimate or prediction. To do otherwise would be to lie to the questioner, and as I have said before trust in science rests on a foundation of honesty and reliability. If you compromise that basis of the relationship between science and the public, you can no longer expect people to believe you, and the public loses a valuable tool for predicting the outcomes of its actions. This is why I think it essential that scientists continue to use this language, which implicitly tells the listener that the speaker is not delivering fact, but prediction. (A reliable prediction mind you! – A prediction good enough for you to invest on.) So given that I don’t think it would be right for scientists to change this mode of speech, I do think that the public have a responsibility to understand it. If they want to get the most out of the science they pay for, they are going to have to.
If you are still concerned by the apparent lack of certainty displayed by scientists and engineers, let me assuage your fears by saying that absolute certainty has never been a necessary feature of technology or science. For example, we (humanity) understand aeroplanes with enough certainty to ensure that they very seldom fall out of the sky. We understand biology and agriculture with enough certainty to ensure that enough food is produced every year to feed us all[4]. We understand medicine with enough certainty to extend the average lifespan of humans in wealthy countries for decades longer than ever before. In fact, for any given application of science to our world there will exist some degree of certainty that is sufficient, and it is always possible for scientists to test to these degrees of accuracy. It may not be absolute certainty, but pretty good certainty is enough to work miracles.
[1] Not to be confused with the behaviour of some science teachers.
[2] Actually this statement is not true any more either, but I couldn’t think of another example that was both correct and well recognised outside of specialist fields. There have now been experiments with results which differ from those predicted by general relativity. So you’ll have to imagine that we’re in the nineties or so when you read this statement. My apologies for this!
[3] read: “realistic”
[4] I realise that people still go hungry. The tragedy is that humanity does actually produce enough food to feed all of us. It just isn’t making its way to countries where people are starving because no-one can find a way of making money from that.