To turn on the TV or go online is to be inundated with claims of certainty: This product will cure your disease, one side of an argument is true and another false, this or that will happen in the New Year.
The enthusiastic certainty with which these claims are made is almost invariably unwarranted for the simple reason that we live in an uncertain world where few things can be described in terms of black and white. In spite of this condition, admitting uncertainty in public discourse is seized upon as a sign of weakness and can be tantamount to ceding one’s point.
In contrast, the concept of uncertainty is fundamental to science, especially when making measurements and analyzing data. Beyond science, acknowledging uncertainty is essential to the pursuit of knowledge. Therefore, unabashed recognition of uncertainty should not be confined to the field or laboratory, but extended to public discourse and recognized as honest, humble and circumspect.
In every field of science, uncertainty begins when the first measurement is made. Use a ruler to make a measurement, and the endpoint will necessarily fall in between two markings. The distance between these markings limits the precision of the measurement, which is represented by a “plus or minus” term to designate the amount of uncertainty present. A measurement without an associated uncertainty is ambiguous and has little practical value.
Uncertainty continues to exert its effects as measurements are grouped into data sets and analyzed for significance. To be significant, experiments must be repeated and results analyzed to determine the probability that the results are due to chance. Various statistical methods may be used, but typically there must be a 95 percent or higher probability that results are not due to chance to be significant.
Like the individual uncertainties reported with each measurement, “error bars” are graphic representations used to compare data sets on a graph or figure. And like individual measurements, a figure that lacks error bars is insignificant; any associations may simply be due to chance.
Far from undermining an argument, acknowledging areas of uncertainty shows that an argument is grounded in reality and not the product of rigid ideology.
Rather than criticizing those who admit uncertainty, skepticism should be directed at those who refuse to admit uncertainty where it exists. Imagine what the world would be like if companies and politicians were required to provide error bars with their claims.
To paraphrase the British mathematician and scholar Bertrand Russell, “The whole problem with the world is that fools and fanatics are always so certain of themselves, and wiser people so full of doubts.”
