An open access publication of the American Academy of Arts & Sciences
Winter 2002

on a new encyclopedia

Neil Joseph Smelser
View PDF

Neil J. Smelser plays a leading role in determining the direction of research in the social and behavioral sciences in America. From 1958 to 1994 he taught in the department of sociology of the University of California, Berkeley; from 1994 until 2001, he was director of the Center for Advanced Study in the Behavioral Sciences, Stanford, California. He currently is chair of the Division of Behavioral and Social Sciences and Education of the National Research Council, and also serves on the council of the American Academy, where he has been a Fellow since 1968.

In late 2001, a new International Encyclopedia of the Social and Behavioral Sciences was published by Elsevier Science. Almost six years in the making, the twenty-four volumes are meant to cover all relevant disciplines under the heading “social and behavioral sciences.” By convention, the more than four thousand entries are presented in alphabetical order, beginning with “Aboriginal Rights” and ending with “Zooarchaeolgy.”

This is the third rendition of such an encyclopedia. The first consisted of fifteen volumes, edited by Edwin Seligman and Alvin Johnson, published in 1931–1935. Then came the seventeen-volume International Encyclopedia of the Social Sciences, edited by David Sills, published in 1968.

Measured by number of words, the new encyclopedia is about double the length of its 1968 predecessor. The coeditors in chief are Paul B. Baltes of the Max Planck Institute for Human Development (Berlin) and I. Of the dozens of academics and others we consulted about the feasibility of a new encyclopedia for these sciences, not one voiced a negative opinion.

Encyclopedias in various forms go back several hundred years and by now are generally regarded as a respected mode of representing and codifying knowledge. Often they have embodied some kind of “integrative impulse”–to symbolize civilizational progress, to express the unity of positive knowledge, to legitimize new areas of scientific inquiry, or to foster interdisciplinary research. More recently, encyclopedias have evolved into “self-contained reference works” meant to present knowledge via brief topical essays and to point readers to related knowledge inside and outside their pages.

Baltes and I justified the need for a new encyclopedia by pointing to the astonishing growth and specialization of the social and behavioral sciences in the past third of a century–we needed a new encyclopedia, we argued, because the older ones were hopelessly outdated.

We nevertheless knew how formidable a task it would be. We attacked the problem in three ways. First, we had thirty-nine separate sections and roughly fifty section editors, in contrast to the 1968 encyclopedia, which had seven associate editors. Second, we included not only the obvious disciplines– anthropology, economics, political science, psychology, and sociology–but also many other fields and traditions of inquiry that incorporate some social and behavioral science research. (I will say more on this point later.) Third, we tried to make the encyclopedia as international as possible, recruiting, in the end, 55 percent of our authors from North America, 35 percent from Europe, and 10 percent from other areas of the world. Readers will ultimately decide whether or not the work is truly comprehensive, but we proudly regard the encyclopedia as a unique asset for students beginning an inquiry, for specialists wishing to gain knowledge of fields other than their own, and for generally educated readers wishing to improve their understanding of the social and behavioral sciences.

At a certain moment, however, it becomes necessary to ask even more basic questions about the capacity to comprehend and present the whole of the social and behavioral sciences in one publication. We observe not only growth and fragmentation in these sciences, but also a crisscrossing of research traditions, an interdisciplinary pursuit of understanding, and new applications of knowledge in many areas. Can all this be tracked, recorded, assembled, and coordinated? Can it be conveyed coherently and usefully by adhering to the conventions of alphabetical presentation and cross-referencing?

Baltes and I believed it could be done, but we knew the risks involved. For one thing, we were attempting to encompass the social and behavioral sciences. The two previous encyclopedias had accentuated the social, but we chose to include parts of evolutionary science, genetics, and neuroscience, in addition to psychology.

We also seemed to be swimming against the tide of contemporary trends in the publication of reference works. There has been an explosion in the number of encyclopedias. lists six thousand of them for sale, and nine thousand. Most of these works, like the fields of knowledge they incorporate, are narrowly focused. They tend to cover subfields of fields (clinical psychology within psychology, higher education within education). One finds such improbable titles as the Encyclopedia of Canadian Music and the Panic Encyclopedia. We believe that the integrative impulse still has value, but in another generation it may no longer be possible, or desirable, to compile a new encyclopedia of the social sciences.

Encyclopedias are, after all, printed compendia of knowledge. A defining fact about a printed book is that it cannot be updated to incorporate new knowledge without considerable effort and cost. Encyclopedias are a species of printed books intended to be “state-of-the-art” representations of some range of knowledge. This understanding is valid only on condition that the art is not forever outrunning itself. However, in our own view, the state of the art is forever outrunning itself. We do not know the precise half-life of printed encyclopedias these days, but we do know that it is continuously shrinking.

Furthermore, the computer and the Internet have given us the capacity to update easily. We are already witnessing some encyclopedias that are exclusively online, with individual entries and groups of entries capable of being replaced continuously and at low cost. Our publishers recognized this by publishing, simultaneously, a hard-copy version (which many libraries and bibliophilic readers still prefer) and a Web version, which will be updated two years hence. This strategy must be regarded as transitional, however, and if the encyclopedia survives as a form in the future, it will surely be in an electronic mode. Still, compiling an encyclopedia does force one to reflect on how to organize the current state of knowledge. Our predecessors divided their world into academic disciplines and chose an associate editor for each discipline. This may or may not have been the correct approach even in their times, but it was clear from the start that we could not proceed in this way. The social and behavioral sciences have become messier, and many areas of inquiry resist ready categorization.

One reason for this messiness is that the development of knowledge in our day is fundamentally uncontrolled. Scientists and scholars are free to go where their curiosity takes them. Knowledge also grows in response to the rise of new social problems and issues, often as defined by national governments and other institutions able to fund research. In recent decades, we have witnessed a spectacular growth in hybrid fields (such as behavioral economics and economic sociology); the rise of interdisciplinary, problem-centered lines of inquiry (for example into the causes of poverty); and the pursuit of parallel substantive lines of inquiry in different disciplines (thus organizational studies conducted by both economists and sociologists).

Given this wealth of new developments, Baltes and I came to feel at moments that we were dealing with the first principle of Greek mythology: in the beginning there was chaos.

How could we create a conceptual architecture that would acknowledge the chaos, yet simultaneously introduce some order? Space forbids a full account, but it is possible to sketch our main strategies, arrived at after extensive reflection.

First, we could not ignore disciplines. They still have a certain logic behind them, and they are the bedrock organizing principle in most colleges and universities, defining the training and identities of professionals and the shape of labor markets. We developed sections and recruited section editors for thirteen disciplines, including the “mainstream” ones listed above and other fields that we judged to have a very strong behavioral and social science component (for example, history, law, and education).

To extend coverage, we also identified eleven areas we called “intersecting”– fields not “in” the behavioral and social sciences, but including some research that could be so described. Examples are behavioral and cognitive neuroscience, health, environmental sciences, and area and international studies. We freely acknowledge some arbitrariness in drawing an exact or even consistent line between “disciplinary” and “intersecting,” but we used the device anyway, as a way of incorporating as many relevant research traditions as we could.

To these intersecting topics, we added five fields that could be best described as “applications” of the social and behavioral sciences (for example, media studies, urban studies, and public policy).

We also took note of a number of areas of work that run through, however unevenly, all of the behavioral and social sciences. Three of these are methodological; we appointed section editors for statistics, mathematics, and computer sciences, and also included articles on the logic of inquiry and research design. Other overarching topics were institutions and infrastructure, ethics of research and applications, history of the behavioral and social sciences, and biographies; we appointed section editors for each of these topics as well.

That is how we got to our total of thirty-nine section editors, to whom we assigned a variable number of individual entries for the encyclopedia. We are confident that this way of casting our conceptual net allowed us to catch most of the fish swimming in the social and behavioral science waters.

In the end, our conceptual efforts at comprehensive coverage, whatever their value, disappear from view, rendered invisible by the encyclopedic principle of listing entries alphabetically, even though we explain our logic in our introduction.

In some ways the virtual encyclopedias of the future will be easier to use than the multivolume compilations we currently have. But time will tell whether the integrative impulse behind these great encyclopedias will survive as well.