A wealth of anecdotal material
THIS BOOK is a misnomer for "heritage" connotes something far removed from the present. Kautilya's work, for example, might be described as India's heritage. The book, however, presents 10 autobiographical articles by Raghu Raj Bahadur, Debabrata Basu, Vasant Shankar Huzurbazar, Gopinath Kallianpur, Debabrata Lahiri, Pesi Rustom Masani, Keshavan Raghavan Nair, C R Rao, S S Shrikhande and P V Sukhatme. In the Indian statistical firmament, these are all names to reckon with, but none of them belong to the distant and hoary past. Therefore, although, etymologically, the usage of the word "heritage" cannot be disputed, the title of the book could have better reflected its contents.
After the Indian Statistical Institute (ISI) was founded in 1931, the journal, Sankhya, in 1933 and the National Sample Survey in 1950, there was a flowering of statistical talent in India. The large deviation theory and estimation of parameters is linked with Bahadur, statistical inference with Basu, Bayesian statistical inference with Huzurbazar, filtering theory and stochastic differential equations with Kallianpur, sampling with Lahiri, multivariate stationary stochastic processes with Masani, block designs with Nair, the theory of estimation with Rao, combinatorics with Shrikhande and nutrition with Sukhatme.
The contents are heterogeneous and varied, with a wealth of anecdotal material -- Basu finding out from Rao who Abraham Wald was and then discovering a flaw in one of Wald's theorems; Huzurbazar contemplating the idea of giving up research for a Ph.D. degree; Lahiri obtaining a job at ISI because his name figured in the footnote to a paper; Rao being refused a job as a mathematician and then turning to statistics as a last resort; a question by V M Dandekar, "one of the bright students in the class", leading Rao to prove the lower bound to variance theorem. However, all this should not give the impression that the book is light and anecdotal. Far from it. With the exception of Lahiri's article, there is a lot of heavy, technical stuff, which few lay readers will find interesting.
Not surprisingly, P C Mahalanobis, the founder and director of ISI is present in the articles, sometimes explicitly, sometimes implicitly. Reactions to him differ. Basu remarks they used to see very little of him. Huzurbazar had virtually no contact with him, but was impressed by his personality. Nair reminiscences on the pleasant time he had as Mahalanobis' house-guest at Baranagar, where he met Rabindranath Tagore. Rao respectfully refers to Mahalanobis throughout as "the Professor". Sukhatme is much more harsh.
Perhaps Kallianpur's comment on Mahalanobis is the fairest: "It is not possible, in a few lines, to do justice to his complex personality. Undoubtedly a man of great originality, his contribution to statistical thought goes far beyond his published work. Being a physicist by training he valued and, indeed, demanded scientific precision in one's thinking, but was impatient with the level of rigour and abstraction inherent in mathematical research. This led, especially in his later years, to a certain lack of communication between him and the younger, more mathematically minded colleagues. The latter felt, with some justice, that their work was neither understood nor appreciated by the Professor. On his part, Mahalanobis suspected (as he declared more than once) that mathematical abstraction was all too often used as a cover for mediocre work... In affairs of the ISI, however, his word was law, the result was not so much of the force of his personality as the paucity of people in positions of authority under him who could stand up to him in matters of principle. The result of such autocracy was not an increase in efficiency but a kind of controlled chaos."
But indirectly or directly, Mahalanobis had a role to play in the development of statistics in India and in the intellectual development of the statisticians writing in this volume. Therefore, it is fitting that such a book should be published in Mahalanobis' centenary year.
Bibek Debroy is a professor at the Indian Institute of Foreign Trade in New Delhi.