Thursday, July 18, 2013

When the answer is, "It's hard to say...."

Often in science, physical as well as biological, when one asks an important question, the answer given by the expert who may be expounding on his/her research or on the state of play in the field, is "Oh, it's hard to say!".

Examples would be how much risk will be associated with a given environmental exposure, how much life expectancy will change by the year 2050, how many planets there are per galaxy, or how high ocean levels will be by 2050, and so forth.

Science is stuck with what we know today, and we don't know what we'll know tomorrow.  But our society has high regard for experts (perhaps, experts favoring a given person's personal interests or viewpoint), and of course in many areas we need to make policy.

But we scientists are proud and we don't advance our careers by saying what the truth is in these circumstances:  "We don't know!"   Most of the time, that is how we should properly translate "It's hard to say."  Or, perhaps, better, "It is currently impossible to know that."  That is the honest answer and if we did our jobs better we'd be saying that, clearly, more often.

One might respond to this by arguing that while we don't literally know the answer, we at least can give some estimate of it, some approximation.  So the typical hedging answer isn't all that wrong.  Indeed, often the expert in question would say something like  "It's hard to say, but based on current data it will be about xxx...."  Or they'll give a range of possibilities ("it will likely be between xxx and yyy...").

Now, there is nothing ethically wrong with giving such vague or approximate answers, but in many if not most situations, the expert doesn't really even know these xxx and yyy values and is just giving his/her personal opinion.  Sometimes the answer is simply a very wild guess (how many planets in the stars in an average galaxy?), with almost no connection to real data.  How did upright posture evolve?  We typically can't predict the future values of causal variables, or really know what they were in the past (as in evolutionary adaptation reconstructions), or what new technology might let us see.  Often, we can't predict such things even in principle.  Yet we see such things said almost every day in the popular media and even in Discussion sections of science journal articles and the like.

Science is largely about what we don't know.  We should acknowledge that so that the public is aware of it, and so we keep ourselves aware of it.  Instead, too often we act as if "hard to tell" means we basically know but not very precisely, and too often that kind of connotation is used for self-interested purposes.

It is unnerving to realize how much we don't know, even if it is rather inspiring to realize how much science has learned, and even how much for the first time just in our own lifetimes.  But a more sober, slow and carefully considered examination of what we don't know--and why we don't know it--might lead to even more inspiring attempts to push ahead.

Of course, it's hard to tell how well that might work!

No comments: