The Presidential Address: Measurement and Conceptualization Problems: The Major Obstacle to Integrating Theory and Research

In one sense the theme of this paper is obvious. Sociologists face extremely tough intellectual and practical tasks owing to the ambitious nature of our common objectives and the complex reality with which we deal. These tasks will require a concerted effort of scholars with diverse substantive, theoretical, and methodological interests and persuasions. Yet in many respects we seem badly divided into a myriad of theoretical and methodological schools that tend to oversimplify each other's postions, that fail to make careful conceptual distinctions, and that encourage partisan attacks. Rather than dwelling on these divisive issues within our profession, it is crucial that we learn to resist overplaying our differences at the expense of common intellectual interests. There will obviously be disagreements over appropriate strategies, as well as ideological and subdisciplinary differences. But an idealization of conflict and dissensus is self-defeating. Some conflicts will inevitably occur and, if constructively resolved, may result in benefits to the discipline. But I think there has been too great a tendency to exaggerate these benefits, without recognizing the inherent dangers of endless theoretical and methodological debates and a further fractionating of our field. One particularly disappointing feature of our discipline is that we have not had the productive interplay between theory and research called for so eloquently by Merton (1968) several decades ago. This interplay, if it ever comes about in a systematic way, will require us to grapple with a number of extremely complex problems that I shall merely list before narrowing my remarks to two issues that illustrate the need for analyses that are simultaneously theoretical and methodological. My list is as follows: 1. Reality is sufficiently complex that we will need theories that contain upwards of fifty variables if we wish to disentangle the effects of numerous exogenous and endogenous variables on the diversity of dependent variables that interest us. 2. Many social changes are either very rapid compared to the intervals of observation or are continuous rather than discrete, so that temporal sequences cannot easily be inferred or linked to given historical events. 3. Realistic models of naturally occurring social phenomena must be nonrecursive or contain highly specific assumptions about lag perods or distributed lags. 4. Many important theoretical variables are highly intercorrelated, though perhaps the empirical associations among them will be underestimated due to random measurement errors. Resolving this multicollinearity problem will require a combination of large samples and good measurement. 5. Human actors and social systems tend to be nonhomogeneous with respect "Address all communications to: Hubert M. Blalock; Department of Sociology; University of Washington; Seattle, WA 98100.