Thursday 17 May 2018

A Study in Statistics




Stephen Jay Gould (1996). The Mismeasure of Man. Penguin London, pp. 446


‘Art is limitation. The essence of every picture is the frame’.
G.K. Chesterton

Stephen Gould is a rare species in the literary world- a scientist with a flair for essay-writing! Through his monthly column in the Natural History magazine, Gould crossed disciplinary boundaries from evolutionary theory to science history and used diverse tools such as statistics and common sense to drive home his arguments. To a student of biology, he is known for his punctuated equilibrium theory among other things and to the philosopher, for his argument that science and religion were two distinct domains with non-overlapping authorities. It is to the non-specialist, that Gould was at his delightful best, exposing the dangers of applying faulty logic and ‘evidence’ from science to justify social projects of domination and discrimination.

Ranking Intelligence

In this book, Gould examines one specific form of quantifying claim about ranking human beings as an illustration of the larger malady. He discusses the belief that intelligence can be meaningfully abstracted as a single number and graded in a linear scale to denote the intrinsic and unalterable worth of a person. He examines this question from three vantage points- that of the philosophical error of the assumption, social impact of its implementation and the epistemological underpinning of ‘nature versus nurture’ that it evokes.

Characteristic to his method, he goes about dismantling this argument by systematically tracing its historical antecedents. He revisits each scholar as he sets about his experiments and draws his inferences, walking with them through their intellectual journey and exposing where the fault-lines of their logic lay.  He takes us back to the time of feverishly measuring cranium and bodies of corpses in the hope of measuring intelligence and later, tinkering with genes and environment with the faith that the truth about our abilities can be abstracted and distilled in a meaningful way. He demonstrates how even the best intended scholars can be blinded to their own fallacies in methods and measurement and how our implicit beliefs often colour our inferences. By doing this exercise with exacting rigour and immense patience, he shows us how, women, people of colour and people with differential abilities come out on their own, liberated from the burden of historical accusations of intellectual inferiority that was unfair, untruthful and dangerous.

This book is a study in history as much as it is in statistics. The parts of the book where Gould marshals evidence from research methodology (especially on factor analysis) is so lucid that I cannot but recommend it to students beginning their research work using statistical methods. I believe that the greatest contribution of the book is the perspective it gives us regarding what scientific enquiry is all about. Like the message in Chesterton’s quote in the beginning of this essay, science like art, is limited by its frame of reference. It is in acknowledging the limits of enquiry that we can hope to be truthful about our inference.