To say that I am not a fan of Donald Rumsfeld would be to dramatically understate the case. But I do find his description of knowns, unknowns, known unknowns and unknown unknowns compelling.
The realm of what is known and what us unknown is ever changing. As we learn new things more becomes known, clearly, but things that we know can change too. Look at thecareas if healthy eating. A couple of years ago I knew that I could only eat a couple if eggs a week as they contained the wrong type of cholesterol. Now it seems I can eat as many as I like. My earlier knowledge has now been surplanted by new information.
I was thinking about this while listening to Womans Hour on my way down to Edinburgh during a piece where a neuroscientist tried to explain the concept of evidential gap that exists between neuroscience research that shows that brains of boys develop differently to those of girls to an educationalist who asserted that this must mean that they should be taught differently.
What we know from research in one field cannot simply be transposed onto a new one without filling that evidential gap. Don’t get me started on the number of alternative therapists who insist that quantum level uncetainty proves that their particular quakery had foundations in science.
But on reading two articles in the New Statesman of 17th May I must question what I think I know on two levels.
Firstly, I know that in order to reduce global warming we must reduce carbon emissions and that means carbon dioxide. But, according to Oliver Tickell, we are missing the bigger picture. Black carbon – soot to me and you – plays a huge role in climate change. He claims that a typical tonne of soot spends about 2 weeks in the atmosphere and causes the same amount of warming as 1600 tonnes of carbon dioxide in 20 years.
This has particular resonance for me as I collect wood to burn on my open fire, thinking that I was being environmentally responsible as I did so.
What I thought I knew I now do not.
And then to Michael Brookes article discussing the first Reith Lecture given by Michael Rees which looked at how confident we can be in the claims made by scientists as they uncover Rumsfeld’s known unknowns.
For example, did you know that in re-runs of published health studies the results were only replicated 5% of the time?
As a student I relied on the term ‘statistically significant ‘ a great deal. Essentially a result is deemed to be significant if there is only a 5% chance that it was generated by random chance. But this only holds true when examining a single variable. If you look at lots of things at once or only decide what it is we are looking for after the data has been collected then any conclusions drawn are unreliable.
This is often why so many published studies reach conclusions diametrically opposed to one another. The world is a very complex place and it is often virtually impossible to isolate a single variable and reach totally sound conclusions.
So even a time served rationalist like me has to accept that scientific evidence may not be all that it appears to be – that what we know, what we have shown to be true, may be unknowable. Maybe everything is an unknown known!