I began my undergraduate education at Antioch College in physics, learned that I like mathematics better, and switched. I was interested in the study of society, but did not take to economics or sociology; anthropology was intriguing but somehow not right for me, and I did not know geography existed.
In 1964, I discovered demography in the form of draft chapters of Nathan Keyfitz's An Introduction to the Mathematics of Population and realized at once that this was my kind of social science. I spent a year in graduate school at the population studies center of the University of Michigan (1967-68), switched to demography at UC Berkeley in 1969, and completed my Ph.D. there in 1972. I came to the East-West Center as a fellow shortly after and have been here ever since. [Note added August 2004: I left the East-West Center at the end of June 1998.]
At the Center, it quickly became clear that methods of estimating levels and trends of fertility and mortality would be my natural area of specialization. Data for most Asian countries was so limited that answering even the simplest factual questions about the demographic situation often required sophisticated technique. My background in mathematics and mathematical demography gave me a powerful comparative advantage. Getting useful information from demographic data may be like producing gold from an underground mine: it requires very substantial effort in extraction and processing. Thus began an enduring theme in my work, the invention of techniques for extracting useful information from demographic data:
1980 'Estimating infant mortality rates from child survivorship data', Population Studies 34(1):109-28. Online
During these years, I focused on the formal structure of demographic data and estimation problems. But I learned that teaching students techniques did not enable them to use those techniques effectively. There was an unexpectedly large gap between understanding the mathematics of an estimation procedure and knowing how to use it in practice. From the mid 1980s, I began systematically teaching the use of methods to draw useful concluslons from demographlc data, together with the concepts and mechanics of those methods:
1991 'Child survivorship estimation: methods and data analysis', Asian and Pacific Population Forum 5(2-3):51-55, 76-87. Online
Partly as a result of this teaching experience, I became intensely interested in demographic transitions in Asia. This led to an intellectual departure. Without having been taught in any explicit way, and without any conscious consideration of merits, I had been working as a positivist, meaning essentially the position that one can understand the situation by looking at the numbers without great attention to interpretive context. On recognizing this position explicitly, I found it unsupportable. Much more had to be learned about broad social and historical context of the demographic transition to understand what happened:
1994 'Fertility decline in East Asia', Science 266 (2 December): 1518-23. Online
My methodological focus also shifted from estimation of standard demographic measures to inventing new ones to facilitate understanding, measures from which inferences might be made about behavior of individuals and families:
1987 'Period parity progression ratio measures of fertility in China', Population Studies 41(3):77-102, with Jingyuan Yu. Online
Interest in demography arose independently of any concern with 'population problems'. Recognizing the importance of Malthusian issues, I find the field's preoccupation with them tiresome and constricting. Demographic phenomena are intricately and inseparably part of the fabric of society as a whole. Demography has a role to play throughout the social sciences in areas that have nothing particularly to do with population problems, as superbly shown by Wolf and Huang's Marriage and Adoption in China, 1845-1945. At the same time, I want to see demography develop further as an intellectually independent discipline. The overall level of technical competence in the field is far lower than it ought to be. There is a wide gap between a handful of people who know a lot and a large number who know next to nothing. ‘A little learning is a dangerous thing; drink deep or touch not the Pierian spring’.
As to the mixture of pure intellectual and applied social concerns in demography, note Alfred Marshall's statement in the introduction of Principles of Economics
Though we are bound, before entering on any study, to consider carefully what are its uses, we should not plan out our work with direct reference to them. For by doing so we are tempted to break off each line of thought as soon as it ceases to have an immediate bearing on that particular aim which we have in view at the time: the direct pursuit of practical aims leads us to group together bits of all sorts of knowledge, which have no connection with each other except for the immediate purposes of the moment and which throw but little light on one another. Our mental energy is spent in going from one to another; nothing is thoroughly thought out; no real progress is made.