The Meaning of 'Theory' Author(s): Gabriel Abend
Source: Sociological Theory, Vol. 26, No. 2 (Tun., 2008), pp. 173-199 Published by: American Sociological Association Stable URL: http://www.jstor.org/stable/20453103
Accessed: 10/05/2011 08:40
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=asa.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
	
9	
STOR	
American Sociological Association is collaborating with JSTOR to digitize, preserve and extend access to Sociological Theory.
http://www.jstor.org
The Meaning of 'Theory'*
Gabriel Abend Northwestern University
'Theory' is one of the most important words in the lexicon of contemporary sociology. Yet, their ubiquity notwithstanding, it is quite unclear what sociologists mean by the words 'theory,' 'theoretical,' and 'theorize.' I argue that confusions about the meaning of 'theory' have brought about undesirable consequences, including conceptual muddles and even downright miscommunication. In this paper I tackle two questions: (a) what does 'theory' mean in the sociological language?; and (b) what ought 'theory' to mean in the sociological language? I proceed in five stages. First, I explain why one should ask a semantic question about 'theory' Second, I lexicographically identify seven different senses of the word, which I distinguish by means of subscripts. Third, I show some difficulties that the current lack of semantic clarity has led sociology to. Fourth, I articulate the question, 'what ought "theory" to mean?,' which I dub the 'semantic predicament' (SP), and I consider what one can learn about it from the theory literature. Fifth, I recommend a 'semantic therapy' for sociology, and advance two arguments about SP: (a) the principle of practical reason—SP is to a large extent a political issue, which should be addressed with the help of political mechanisms; and (b) the principle of ontological and epistemological pluralism—the solution to SP should not be too ontologically and epistemologically demanding.
1. INTRODUCTION
'Theory' is one of the most important words in the lexicon of contemporary sociology. I am not referring only—in fact, not principally—to the subfield of sociological theory. The words 'theory,' 'theoretical,' and 'theorize' are constantly and consequentially used by all sociologists. For instance, one way of describing what sociologists of social movements do is to say that they develop 'theories' about social movements. What sociologists of the family do is to develop 'theories' about the family. And so on. Moreover, it is a widespread belief that empirical sociological research should be driven or informed by 'theory' Thus, sociology journals tend to reject 'atheoret-icaP and 'undertheorized' papers, as well as papers that fail to make a 'theoretical
*Direct correspondence to: Gabriel Abend, Department of Sociology, Northwestern University, 1810 Chicago Ave., Evanston, IL 60208 (g-abend@northwestern.edu). The origins of this paper lie in an invitation to reflect on the present and future of sociological theory. I am thankful to the organizers of the Junior Theorists Symposium 2005—Mathieu Deflem, Marion Fourcade, and Neil Gross—for this invitation, and to my discussant, Charles Camic. I also benefited from conversations with fellow "junior theorists" Pierre Kremp, Simone Polillo, Isaac Reed, Erika Summers-EfTler, Jonathan VanAntwerpen, and Robb Wilier. I presented a slightly different version of the argument at the 2005 Annual Retreat of the Society for Comparative Research, hosted by Central European University. At this conference I received useful suggestions from my discussant, Jack Goldstone, as well as from Carsten Schneider and Robin Stryker. Finally, I am indebted to Sareeta Amrute, Charles Camic, Mathieu Deflem, Marion Four-cade, Neil Gross, Carol Heimer, Adam Kissel, Donald Levine, Richard Morales, Michael Sauder, Arthur Stinchcombe, Devin B. Terhune, and the Sociological Theory editors and reviewers for their comments and criticisms on earlier drafts of this paper.
Sociological Theory 26:2 June 2008
© American Sociological Association. 1430 K Street NW, Washington, DC 20005
174
SOCIOLOGICAL THEORY
contribution' to the literature (be the paper's subject-matter inequality, education, gender, or culture). Sociologists' business is to 'theorize' about social things. Indeed, that sociology offers 'theories' about social things is arguably what makes it count as a social science (and differentiates it from the accounts of society that laypersons, poets, and journalists offer).
But what exactly do sociologists mean by the words 'theory,' 'theoretical,' and 'theorize'? Their ubiquity notwithstanding, we shall see that it is quite unclear what these words mean in the sociological language. More importantly, we shall also see that this is not at all an abstract philosophical problem, unrelated to the actual production of sociological knowledge. Unfortunately, semantic confusions about the word 'theory' have led to much miscommunication, inside and outside the subfield of sociological theory. Rather than an abstract philosophical problem, this is a practical problem, which—insofar as agreed-upon logical and semantic bases are a prerequisite for any sort of epistemic progress—sociology cannot neglect.
Thus, in this paper I tackle two main questions: (a) what does 'theory' mean in the sociological language?; and (b) what ought 'theory' to mean in the sociological language? I proceed in five stages. First, I very briefly explain why it is a good idea to ask a semantic question about 'theory.' Second, I investigate what 'theory' and some of its inflected forms are taken to mean by different sociologists in different sociological contexts. I find seven senses of the word, which I distinguish by means of subscripts (theoryi, theory2, theory3, etc.). Third, I consider whether the polysemy of 'theory' is a good or a bad thing. As this introductory section suggests, I believe the latter to be the case, so section 4 brings out some difficulties that the current state of semantic affairs has gotten sociologists into. Up to that point, the argument is that the extension and intension of the word 'theory' are unclear, and that this has brought about undesirable consequences. The fourth stage of my argument is to ask what is to be done about this problem. How ought sociologists to use the word 'theory'? This is what I dub the 'semantic predicament' (SP). Then I consider what one can learn from the theory literature, both in terms of missteps to be avoided and insights to be built on. Fifth, I recommend a 'semantic therapy' for sociology, and advance two arguments about SP: (a) the principle of practical reason—SP is to a large extent a political issue, which should be addressed with the help of political mechanisms; and (b) the principle of ontological and epistemological pluralism—the solution to SP should not be too ontologically and epistemologically demanding.
2. A MATTER OF MEANING?
'What is theory?' 'What is a good theory?' 'What is theory for?' These are three questions to which sociologists have certainly given a lot of thought. Let us call them, respectively, the ontological question, the evaluative question, and the teleological question. I would like to raise an apparently more basic and apparently simpler one: 'what does "theory" mean?'1 Let us call this the 'semantic, question' (SQ). I say that SQ seems to be a more basic question, because the other three presuppose some stance on it. SQ seems to be a simpler question as well. Unlike the ontological
'I follow John Lyons's (1977) typographical conventions, although making a few modifications: (i) single quotation marks: 1. for lexemes and expressions (when they are mentioned rather than used); 2. for the citation of sentences; 3. instead of double quotation marks within double quotation marks; (ii) double quotation marks: 1. for meanings; 2. for quotations from other authors; 3. for titles of articles; 4. instead of single quotation marks within single quotation marks; (iii) italics: 1. for emphasis; 2. for titles of books and journals; 3. for words in languages other than English.
THE MEANING OF 'THEORY'
175
question, it does not force us to think about metaphysics, reality, essences, being, and the like. Unlike the evaluative question, it does not involve the tricky predicate '(be) good.' Unlike the teleological question, it does not appear to require that we are clear about the nature and aims of sociology. Rather, considering that sociologists frequently use the word 'theory,' SQ just asks what it is meant by that mark on the paper or string of sounds.
In order to show why it is a good idea to raise SQ and what one may learn from it, let me propose a thought experiment. Suppose a person named Jones uttered the following words: 'It never snows in Chicago in January.' You can have either of two reactions to this statement. One is to think that Jones has a false belief. Since you have actually lived a few years in Chicago, you know all too well that it does snow there in January. Probably Jones got that piece of information from an unreliable source, which she mistakenly took to be a reliable one. In all likelihood, she has never lived there herself. A second possible reaction is to think that Jones is confused about the meaning of the words she is using (for example, you may speculate that her English is bad). Perhaps what she intends to assert is, 'It often snows in Chicago in January,' but has confounded the adverbs of frequency 'never' and 'often.' If this were the case, then Jones would have a true belief. Or, she might have intended to affirm that 'it never snows in Chicago in July' but got the English names of the months wrong. Similarly, the problem might lie in what the word 'Chicago' is taken to refer to. If by 'Chicago' Jones is referring to the largest city in the U.S. state of Illinois, whose mayor is Richard M. Daley, etc., then her statement is patently false. But maybe Jones is referring to a small town somewhere in South America, also named 'Chicago,' where it never snows in January.
The general point is that, as Donald Davidson (1984:142) puts it, '[i]f all we have to go on is the fact of honest utterance, we cannot infer the belief without knowing the meaning, and have no chance of inferring the meaning without the belief.' So, as things stand, it is impossible to determine whether Jones has an untrue understanding of Chicago's weather or an untrue understanding of what the English word 'never' means. To put it without bringing epistemic privileges into play, it is impossible to determine whether the disagreement between Jones and you is one of meaning or belief. And if the former is the case, any discussion you have with Jones about the occurrence of snow in Chicago in January will be futile, and probably frustrating.
I think that something like this is at the bottom of sociologists' unremitting disagreements about 'theory.' Sociology has been plagued by disputes over what a good theory is, what constitutes a theoretical contribution, where theory should go, whether sociology has made theoretical progress, which theoretical paradigm should be favored, what the functions of theory are, what it is for a paper to be an atheoretical one, and so on. These disputes have been framed as being about the nature, features, functions, and future of a certain entity that the word 'theory' is supposed to pick out, not as being about what exactly it is that the word 'theory' picks out. This framing is misguided. I argue that before any progress can be made on those important questions, certain confusions about our words and concepts must be cleared up.
For example, suppose that A, B, and C are three sociologists, and they engage in a discussion about the question, 'how should theory be judged?' Now, when A speaks of 'theory' she mainly thinks of an ongoing dialogue with some classic texts. When B speaks of 'theory' she mainly thinks of the construction of propositions of the form 'if p then q.' And when C speaks of 'theory' she mainly thinks of the development of lexica and schemata with which to talk about the social world. Not surprisingly, A, B, and C find it impossible to come to an agreement about the question under
176
SOCIOLOGICAL THEORY
discussion. Yet the most reasonable interpretation of this situation is not that A, B, and C have a substantive disagreement about how theory should be judged, but rather that they are talking about different things. That they are indeed talking about different things should not be obscured by the fact that they happen to use the same English word (or, to be more precise, that nearly identical phonemes come out of their mouths).
A, B, and C illustrate the kind of problem that many arguments and debates about 'theory' have suffered from: they bypass SQ, and start directly with ontological, evaluative, or teleological questions about the 'object' everyone is purportedly referring to. The circumvention of SQ has brought about much miscommunication. Very much like scientists working under different paradigms, who use the same word— e.g., 'mass'—to express different concepts (Kuhn 1970), oftentimes sociologists literally talk past one another. In a snowball fashion, it has also brought about further semantic and conceptual confusions, as 'theory' (however conceived) is at the heart of several other sociological problems, debates, projects, and institutions.
In brief, there are three reasons why sociology needs a semantic analysis of 'theory.' First, it will improve understanding and communication, without which there can be no productive (indeed, no meaningful) substantive discussion. Second, it will prevent what I call in section 4 the 'Socratic error': the belief that theory is an object out there that our concepts or language can track down. As we shall see, A, B, and C will never find out whose definition really refers and whose definition fails to refer, because there is no real or objective referent for 'theory.' Consequently (third), it will show that how one ought to use the word 'theory' is to a great extent a political or practical-reason problem.
In the next section I identify seven different senses of the word 'theory' in the sociological language. Yet before moving on to this task, I would like to add an important caveat. The problems sociology is interested in can be neither solved nor dissolved by means of conceptual or linguistic analysis. For these are genuine problems, not mere linguistic puzzles. What I argue is just that they would be much more profitably addressed if our words and concepts were clearer. Let me illustrate the point. Consider an imaginary world in which an omnipotent tyrant could force A, B, and C to eliminate the word 'theory' from their vocabularies. From now on, she would declare, in sociology the word 'theory' shall be meaningless; the word 'gavagai' shall mean "ongoing dialogue with some classic texts"; the word 'ravagai' shall mean "construction of propositions of the form 'if p then q'"; and the word 'savagai' shall mean "development of lexica and schemata with which to talk about the social world." In some respects this decree would make things better: A, B, and C will have fewer misunderstandings, will identify with more precision what it is that they disagree about, and will have more fruitful conversations about these issues of discord. Nevertheless, most of the fundamental problems will remain untouched. First and foremost, one would not have made any progress at all on questions such as: Is gavagai (ravagai, savagai, tavagai) beneficial, useful, important, necessary, etc.? If it turns out that all of them are legitimate parts of the discipline, how do they fit together? How should they be weighed? Nor would one have made any progress at all on questions internal to each of these parts, such as: What is a good gavagai (ravagai, savagai, tavagai)? What is gavagai for? What gavagais are true, valid, plausible, 'warrantedly assertible,' etc.? How does one tell a true from a false gavagai? Who are the classics of sociological gavagai? These problems are not an artifact of our linguistic practices, rules, and conventions, and hence cannot be settled by any clarification of what words mean or what things are called.
THE MEANING OF 'THEORY'
177
3. THE MULTIPLE MEANINGS OF 'THEORY'
I would like now to distinguish seven different things that sociologists may mean when they use the word 'theory.' My task in this section is not evaluative but descriptive. Like the lexicographer, I want to give an empirical account of different ways in which a particular word is used by competent speakers of a certain language (see Allan 2001; Cruse 1986; Landau 2001; Sager 2000). Each of these senses is used by some reasonably large number of people whom one can reasonably call 'sociologists.' However, each of them may be considered semantically inaccurate or even incorrect by some other people, whom one can reasonably call 'sociologists' as well. Still, these people will probably not say that it does not make any sense, that the utterance is unintelligible, that the speaker knows nothing about sociology, or that she is out of her mind.
It is crucial to realize the differences between my lexicographic exercise and the usual attempts to distinguish kinds, types, or forms of theory, theoretical approaches, ways of theorizing, etc. These attempts address themselves to the ontological question, 'what are the different kinds of theory?' But this question presupposes an underlying concept of theory, of which there are kinds to be found. For example, you may set out to identify the different 'theoretical schools' that exist in sociology, and come to the conclusion that there are four of them: Si, S2, S3, and S4. However, you are committing a petitio principii here: while it is quite clear what it is about these four things that makes them 'schools,' you are begging the question of what it is that makes them 'theoretical' ones in the first place. It is your own, a priori concept of theory what allows you to tell what is and what is not a theory or a theoretical school. In addition, you are performatively making a strong normative claim. You are indeed taking sides in the very dispute these 'theoretical schools' have over what theory is. In all probability, Si does not see any theory at all in the projects carried out by S2, S3, and S4 (and vice versa). But you disagree: according to you, S2, S3, and S4 are genuine theoretical schools—that is why they are included in your typology. By contrast, my semantic approach does not presuppose a concept of theory. It does not make any ontological commitments or normative claims. It just reports on what different sociologists seem to mean when they use the words 'theory,' 'theoretical,' and 'theorize.'
Theory 1. If you use the word 'theory' in the sense of theory 1, what you mean by it is a general proposition, or logically-connected system of general propositions, which establishes a relationship between two or more variables. As an example, let us consider the 'mass society version' of the 'breakdown theory' of social movements: "Individuals are most likely to join social movements when they have few personal ties within a community and a weak sense of identification with that community" (Useem 1980:357). This theory establishes a relationship between the variables 'likelihood of joining social movements,' 'number of personal ties within a certain community,' and 'strength of sense of identification with a certain community.' But what I would like to stress is that if the theory were put in symbolic form, it would have to be universally quantified. The subject of the sentence is not 'some individuals,' 'late nineteenth-century individuals,' or 'Chinese individuals.' The theory establishes a relationship between those three variables in general, independently of things like time and place.
Likewise, when it is said that 'empirical investigations should make a theoretical contribution,' what is usually meant is that 'empirical investigations should make a theoretical 1 contribution.' For instance, suppose you write a paper about two social
178
SOCIOLOGICAL THEORY
movements that arose during the May Revolution of 1810 in Buenos Aires. For it to be accepted for publication in a mainstream sociology journal, some conclusion about social movements in general must have been drawn. Or, to put it using a ubiquitous and highly-revealing euphemism, the empirical investigation must have 'suggested' some more general conclusion. If there is no such a conclusion, the reviewers would probably point out that the paper is 'atheoretical' or, more benignly, that it is 'undertheorized.'
Theory2. A theory2 is an explanation of a particular social phenomenon. In this sense, if you say that you have a 'theory' about the demise of the Valois dynasty in late-sixteenth-century France, what you mean is that you can offer an explanation of it. This explanation should identify a number of 'factors' or 'conditions,' which individually should pass some sort of counter factual test for causal relevance, and whose interaction effects should be somehow taken into account. (A theory2 could be an explanation of some other kind; I consider only causal explanations because of their hegemony in contemporary sociology.) So, for example, a theory2 of the concept of the self tells one why at a certain point in history such a concept came into being. A theory2 of the stock market crash of October 1929 tells one what brought it about. A theory2 of the victory of the White Party in the 1860 presidential elections in Uruguay tells one why it is that it won.
Several epistemological understandings of social explanation are compatible with the aims of theory2, some of which would see theoryi as a misguided project. By contrast, the deductive-nomological model of explanation posits a special relation between theoryi and theory2. As Hempel (1965:240) puts it, "in empirical science, the explanation of a phenomenon consists in subsuming it under general empirical laws." In this view, then, one or more theoriesi logically entail a certain theory2. For instance, in order to explain why the Uruguayan White Party won in 1860, it is necessary (and sufficient) to appeal to one or more general propositions about voting behavior, electoral politics, party systems, and so on. What is important from my point of view, however, is that even for the deductive-nomological model, theoryi and theory2 remain distinct entities.
Theory3. Like theoryi and theory2, the main goal of a theory3 is to say something about empirical phenomena in the social world. However, the main questions that theory3 sets out to answer are not of the type 'what x causes yV Rather, given a certain phenomenon P (or a certain fact, relation, process, trend), it asks: 'what does it mean that PI,' 'is it significant that PI,' 'is it really the case that PI,' 'what is P all about?,' or 'how can we make sense of or shed light on PI' Thus, one can think of theory3 as a hermeneutical task, even if theoristS3 interpret social reality rather than texts, and they do not necessarily share the philosophical inclinations of the hermeneutical tradition. To put it another way, what theories 3 offer is an original 'interpretation,' 'reading,' or 'way of making sense' of a certain slice of the empirical world. They may shed new light on an empirical problem, help one understand some social process, or reveal what 'really' went on in a certain conjuncture. Unlike theoryi, theory3 does not view P as the value of a variable y, which in turn is related to other variables in such a way that can be described by a function y = F(\], x2,..., x„). Unlike theory2, theory3 may or may not causally explain P.
For example, that you have a theory3 of corruption in twentieth-century Latin American politics does not mean that you have identified its causes. Nor does it mean that you know how this case fits into a system of social laws about politics, corruption, the law, and morality, which specifies the 'conditions under which' corruption occurs. Rather, you may mean that your theory helps understand what
THE MEANING OF 'THEORY'
179
corruption in twentieth-century Latin American politics 'is all about'; that it provides a 'better interpretation' of corruption in twentieth-century Latin American politics; or that it tells a story about corruption in twentieth-century Latin American politics that is illuminating, instructive, helpful, or edifying. Now, the exact nature of these 'interpretations' is difficult to establish a priori, codify, and standardize. Therefore, theory3 may strike one as conceptually vague, methodologically problematic, or just unscientific. While they are rarer in U.S. sociology, theories3 are a staple of some Latin American and European sociological traditions (cf. Abend 2006).
Theory4. The word 'theory' and some of its derivatives are sometimes used to refer to the study of and the students of the writings of authors such as Marx, Weber, Dürkheim, Simmel, Parsons, Habermas, or Bourdieu. These 'studies' are variously described as 'interpretations,' 'analyses,' 'critiques,' 'hermeneutical reconstructions,' or 'exegeses.' They often involve the study of 'meaning,' in either of the two following meanings of 'meaning.' First, they may ask what the author of a text 'really' meant, that is, what she wanted to say when she wrote a particular passage (or essay, book, oeuvre). Most of the authors with which theory4 concerns itself put forward very complex arguments, which they put forward in a very complex style. Thus, their writings lend themselves to discussion about what exactly the argument is. In addition to these discussions, one can reconstruct the logic of these arguments, identify how the question is framed and how this framing affects the range of possible answers, spell out assumptions, crucial concepts, borrowed ideas, etc. Furthermore, most of these authors have produced a large body of work, so theory4 may also consider how their thinking evolved, and, more generally, how they relate to their ancestors and contemporaries. Second, theory4 may ask what the meaning of a certain text is in another sense, namely, what is its significance, relevance, usefulness, what was or is original about it, how it has been drawn upon, etc.
Most U.S. departments of sociology have on their faculty one or more persons who are said to be 'theorists.' The graduate and undergraduate courses they teach are typically called 'sociological theory,' 'social theory,' 'classical theory,' 'contemporary theory,' or something like that. By saying 'theorists' and 'theory courses' in these contexts one generally means theoristS4 and theory4 courses. Finally, note that you can say that you: 'are' a theorist4; 'do' theory4; work in the field of theory4; teach a theory4 class; have written a theory4 paper; or that one of your scholarly interests is theory4. In fact, in most of these contexts it is likely that the word 'theory' be taken to mean theory4. However, you cannot have, have developed, or put forward a theory4. If the expressions 'have a theory,' 'have developed a theory,' or 'put forward a theory' occur in a sentence, the word 'theory' is not being used in the sense of theory4. Strangely, the predicate '(be) a theory4' cannot be applied to the content of a theory4 paper, to the product of the activity of doing theory4, or to the work of a theorist4.
Theory5. A theory5 is a Weltanschauung, that is, an overall perspective from which one sees and interprets the world. Unlike theoriesi, theories2, and theories3, theoriess are not about the social world itself, but about how to look at, grasp, and represent it. That is, theoriess focus on our conceptual and linguistic equipment—for example, the nature of the location from which we look at the social world, the lexicon and syntax by means of which we talk about it, the nature of our conceptual scheme, the categories into which we group things, and the logical relations that there can be between concepts. Obviously, these 'conceptual' and 'linguistic' choices entail, are associated with, or are predicated upon many more general epistemological and ontological views, which theoristss may or may not explicitly articulate. To name but
180
SOCIOLOGICAL THEORY
a few: what can be known, what is worth knowing, what kind of questions can be asked, what counts as good evidence, whom are we talking to, what the social world is made of, what properties can these entities have, how they fit together, and so on.
In its ideal-typical form, the core of a theory5 does not have truth-values. While its actual manifestations often include theories i and theories2, this is not logically necessary. For the thrust of theory5 is, so to speak, syntactic rather than semantic. In other words, theoriess (again, in their ideal-typical form) do not offer propositions about what the social world is like (e.g., 'Education is correlated with political participation' or 'The main cause of the fall of the Roman Empire was an economic crisis'). Rather, they offer a way of looking at it or a way of talking about it (e.g., 'Society can be thought of as a system' or 'The social world can be understood as containing objects with properties and variables with values').2 Indeed, there are interesting similarities between theory5 and Kant's categories of understanding. Theory 5 can be seen as providing one with an a priori framework (scheme, grid, map, net, plan), a framework that is independent from experience, logically prior to any contact with the social world. Thus, it would provide the conditions for the very possibility of experience (or, more conservatively, the conditions for the intelligibility of experience).
When one speaks of 'postmodern theory,' 'poststructuralist theory,' 'feminist theory,' 'queer theory,' 'critical theory,' 'Marxist theory,' 'structural-functionalist theory,' 'network exchange theory,' 'game theory,' and 'rational choice theory,' one is often using the word 'theory' in this sense. Theorys can also occur as an adjective in expressions such as 'theoretical approach,' 'theoretical school,' 'theoretical framework,' 'theoretical perspective,' 'theoretical tradition,' 'theoretical viewpoint,' 'orienting theoretical strategy,' and 'theoretical paradigm.' This last one, of course, draws a parallel between the field of sociological theory and Kuhn's account of the history of science (Ritzer 1980). Another common example is the expression 'theory-laden,' which is often predicated of 'observation' or 'perception.'
Theory6. Lexicographers trace the etymology of the word 'theory' to the late Latin noun 'theoria,' and the Greek noun 'thedria' and verb 'theorem'' (usually translated as "to look at," "to observe," "to see," or "to contemplate"). The connotations of these words include detachment, spectatorship, contemplation, and vision. This etymology notwithstanding, some people use the word 'theory' to refer to accounts that have a fundamental normative component. This usage I identify as theory6. For example, the contemporary projects of 'critical theory,' 'feminist theory,' and 'postcolonial theory' are explicitly normative ones, which usually reject the fact/value dichotomy, and hence the supposedly value-neutral sociological theory. The same is true of a good deal of 'Marxist theory' and 'neo-Marxist theory,' which draw inspiration from Marx's eleventh thesis on Feuerbach. Indeed, the expression 'social theory'—as opposed to 'sociological theory'—can have similar connotations. At least, self-defined 'social theorists' are more likely to do theory6 than, say, theoryi. The basic point for my purposes is that the word 'theory' can be used to refer to a normative, and indeed political, account—a far cry from other senses of it. (The expression 'social theory' transcends the discipline of sociology and has close connections with the humanities—see, e.g., Christian 1988; de Man 1986; Jay 1996; Knapp and Michaels
2Sometimes these statements are presented as propositions about what the social world is like (e.g., 'Society is a system,' 'Society is a text,' 'Education is a variable,' 'Social life is a theatrical performance'). I take these formulations to be ellipses of 'Society can be (profitably) thought of as a system,' 'Society can be (profitably) thought of as a text,' etc. For how could the truth-value of 'Society is a text' or 'Education is a variable' be determined?
THE MEANING OF 'THEORY'
181
1982. In fact, 'social theorists' deliberately try to transcend disciplinary boundaries. Not surprisingly, 'social theory' has been used in a number of different ways. Moreover, these usages, as well as the relationships between 'sociological theory' and 'social theory,' have considerably changed over time—think of 'social theory' as used by Stinchcombe [1968], Merton [1949, 1968], Goode [1973], or Coleman [1990]. Unfortunately, I cannot carry out a thorough analysis of 'social theory' in this paper [see Seidman and Alexander (2001)].)
Theory 7. Many sociologists have written about issues such as the 'micro-macro problem,' the 'problem of structure and agency,' or 'the problem of social order.' This type of work is usually thought to fall within the domain of sociological theory. One may also use the word 'theory' to refer to discussions about the ways in which 'reality' is 'socially constructed'; the scientific status of sociology (value freedom, the idea of a social law, the relations between explanation and prediction, explanation and understanding, reasons and causes, and the like); or the 'relativity' of morality. In these examples the word 'theory' assumes a distinct meaning, which I distinguish as theory 7.
Theory7 projects often include theory4 elements (for instance, a reconstruction of what Marx, Weber, and Dürkheim said about the problem under consideration, or what they would have said had they considered it). Moreover, as a matter of fact, theoristS7 are often theoristS4 as well. Yet, whatever its usefulness or reasonableness, this correlation is not necessary but contingent. Theory? and theory4 remain two different projects. Similarly, while there can be relations between theory5 and theory7, occasionally even relations of necessity, the meaning of the two terms remains distinct. Theory? does not refer to an overarching Weltanschauung, a way of looking at or representing the social world. Rather, it refers to the study of certain special problems that sociology has encountered. Even though it is because of its being in the business of empirically investigating society that sociology has encountered these problems, they are not empirical problems themselves (for example, they cannot be resolved by means of empirical methods). They may be described as 'philosophical' problems, insofar as they call for reflection upon the nature of knowledge, language, and reality, and some sort of conceptual analysis. In fact, most of these problems have been taken up in philosophy as well, usually under different rubrics, in a different voice, and in isolation from the sociological literature. To conclude with a reflexive note, the present paper might be said to be a 'theory' paper mostly in the sense of theory7.
4. THE PROBLEMS OF POLYSEMY
Many natural language words have several different meanings. To take two examples from Spanish, one of homonymy and one of polysemy: 'agujď can mean "needle" or "hand of a clock"; 'banco' can mean "bank" or "bench." Yet Spanish speakers have no problem understanding one another. By considering its syntactic, semantic, and pragmatic contexts, they typically can discern whether any given token of 'banco' means "bank" or "bench." More generally, they are aware of and know how to deal with this feature of the lexical semantics of their language.3 In fact, for certain
3Typical English examples of polysemy and homonymy are, respectively, 'paper' and 'fluke.' As two linguists explain, "[s]trictly speaking, homographs are etymologically unrelated words that happen to be represented by the same string of letters in a language. For example, bass the fish is derived from Old English barse (perch) while bass the voice is derived from Italian basso. Conversely, polysemes are etymologically and therefore semantically related, and typically originate from metaphorical usage" (Ravin and Leacock 2000:2; italics in original).
182
SOCIOLOGICAL THEORY
purposes polysemy is not a liability but a great asset (Levine 1985a). Then why might the multiple meanings of the word 'theory' constitute a problem for the sociological language and sociology in general?
One possible answer would stress the differences between the sociological language and natural languages. Natural languages let one—and, arguably, must let one— express one's feelings, tell jokes, be ironic, do things with words, persuade, speak without saying too much, tell stories, lie, affirm one's identity, and talk of love. By contrast, this argument goes, sociologists are mainly concerned with giving true accounts of the social world. They put forward arguments, which are then subjected to public scrutiny. Their claims are meant to be as precise as possible. Given these aims, the sociological language ought not to be polysemous—all potential sources of ambiguity and fuzziness ought to be eliminated. This is, of course, a version of the classic logical positivist position (e.g., Nagel 1961), and I do not mean to deny that in this context there is something to it. On the other hand, it is clear that speakers of natural languages do a very good job of dealing with cases of homonymy and polysemy. In strict semantic terms, words and utterances may sometimes be ambiguous. Yet, generally speaking, comprehension and communication do not seem to be impaired.
My argument is that in the case of 'theory' the problems stem from the erroneous belief that there is something—indeed, one thing—out there for the word 'theory' to really correspond to. Then, if one used the word to refer to anything but that object, whatever it turned out to be, one would be mistaken. I want to call this the 'Socratic error.' In many of Plato's dialogues, Socrates sets out to answer a question of the form, 'what is XT (see Robinson 1980). For example, in the Republic he asks 'what is justice?'; in the Euthyphro 'what is piety?'; and in the Laches 'what is courage?' (other examples are friendship, virtue, knowledge, and temperance). Socrates' aim is not to find out what things are just, what sort of thing justice is, what is good about justice, or what people mean by 'justice,' but what justice really is. Similarly, sociologists do not take theoryi, theory2, theory^, etc., to be different senses of the word 'theory,' to be treated like different senses of the word 'paper,' 'bass,' or 'fluke.' Rather, they take them to be competing views about what theory (that is, the thing) really is, and hence competing views about the true meaning of 'theory.'4
Here lies the crux of the matter. Many sociologists have said that there are different 'understandings' or 'conceptions' of theory. More to the point, many have said things like 'theory means different things to different people' or 'different people have different definitions of "theory".' Yet most sociologists have failed to draw the necessary implications of this semantic fact. Thus, they still go on to ask Socratic questions about theory, the object (which the word 'theory' supposedly picks out, or should pick out). But recognizing "bank" and "bench" as two senses of the Spanish word 'banco' has important consequences for the sort of questions that one can ask.
4This view makes sense against the background of Plato's metaphysics—for him, justice (and piety, courage, etc.) has an essence (Form, Idea, 'eidosr), which philosophy can discover. But I take it that this would seem a quite implausible ontological assumption to most contemporary sociologists—e.g., that sociological theory has an essence, the Idea of sociological theory, which, if found, would settle the controversy about what sociological theory is. As Robinson (1980:121) points out, there are "several assumptions that must be made if Socrates' question [what is X?] is to be a legitimate question admitting of a true answer. First, we must assume that the word X is univocal. Second, we must assume that the thing X has an 'essence.' Third, we must make some sort of realist assumption about the ontological status of this 'essence.' And, fourth, we must assume that this 'essence' is not a 'primary element' but has a structure that can be explicated; for otherwise we must already know what X is in asking the question in Socrates' sense."
THE MEANING OF 'THEORY'
183
While in Spanish they are referred to by the same word-form, benches and banks are two very different things indeed. Therefore, only if one keeps these senses apart, can one sensibly ask questions such as: what is a good banco!; how many kinds of bancos are there?; what are bancos useful for?; or what is a banco! Likewise, we have seen that the word 'theory' in the sociological language has several senses, each of which tries to pick out a different thing. In fact, the difference between these things—e.g., theory i and theory4—might be almost as significant as the difference between the concept of a bank and the concept of a bench. Consequently, sociologists' arguments about 'theory'—that is, about theory in general; about its nature, methods, quality, aims, and possibilities—are neither as clear nor as fruitful as they should be. And, as one might expect, this state of affairs can lead to unfortunate misunderstandings.
Let me mention two important examples. Can small-N studies (e.g., ethnographic and historical studies) yield 'theoretical' gains? The literature is as vast as it is discordant and inconclusive.5 On a practical level, one often hears ethnographers and historical sociologists complain about the way in which their manuscripts are judged by mainstream sociology journals. Probably the most common situation is the following: one or more referees say that the manuscript fails to make a 'theoretical contribution' or to 'advance theory,' often on the grounds that if your number of cases is one you cannot derive 'theoretical conclusions.' Ethnographers and historical sociologists respond that you can, and that the idea that you cannot reflects a bias in favor of quantitative sociology. And so on. But if my argument is correct, there is no genuine disagreement here. If your number of cases is one, you obviously cannot derive theoreticali conclusions. It is even dubious that a single case can 'suggest' (as we saw above, a very common way of putting it) theoretical i conclusions. What is more, if it is accepted that a theory \ establishes a probabilistic rather than a deterministic relationship, a single case cannot even disprove it (Lieberson 1991, 1992). But from an ethnographic or historical study you obviously can derive theoretical and theoretical conclusions. In fact, sometimes it might be the best or the only way to obtain such 'theories.'
Similarly, sociologists have long quarreled over the proper relationship between 'sociological theory' and 'empirical research' (for instance, 'Sociological Theory and Empirical Research' was the theme of the 2002 mini-conference of the Theory Section of the ASA [Fine 2001]). Some of them argue that 'theory' ought to be 'based on,' 'close to,' 'linked to,' 'derived from,' 'in dialogue with,' or 'intimately related to' empirical research or empirical data. Some others argue that this is not so, or not necessarily so. Again, it is clear that theoryi, theory2, and theory3, ought to be based on empirical data. On what other bases could one possibly make an argument about the conditions under which ethnic minorities tend to assimilate? How else could one work out an explanation of the rise and triumph of capitalism? However, the same is not true of other senses of 'theory'—for example, if one's aim is to rid a sociological concept of some muddles, propose a new standpoint from which to look at social relations, unveil a writer's general presuppositions, or explore the connections between Adam Smith's well-known political economy and his less well-known moral philosophy.
At any rate, the most obvious sign that something is going wrong here is, I think, that sociology does not seem to be any clearer about the ontological, teleological,
5To take the case of historical sociology, recent reflections on its relations to 'theory' include: Adams, Clemens, and Orloff 2005; Calhoun 1998; Goldstone 1998; Kiser and Hechter 1991, 1998; Lieberson 1991, 1992; Mahoney 2004; Mahoney and Rueschemeyer 2003; Paige 1999; Quadagno and Knapp 1992; Skocpol and Somers 1980; Somers 1998.
184
SOCIOLOGICAL THEORY
and evaluative questions than it was 50 or 60 years ago. It is not just that many of the questions about the nature of theory raised in the 1940s and 1950s by people like Robert Merton are still being debated today.6 What is truly puzzling is that they are being debated in analogous terms, starting from analogous points, making analogous assumptions, encountering analogous difficulties, provoking analogous objections, and bringing about analogous factions. And, as then, consensus is not in sight. Indeed, not only is consensus not in sight—it is hard to see what kind of argument could possibly settle the disagreements about what theory is, what a good theory looks like, why theory is important or useful, how theory is related to empirical research, whether there is theoretical progress, and so on. In fact, my claim is that, as things stand, no argument could do that. One cannot find out or establish what theory really is because theory is not an object out there that our concepts or language could track down. As a result, 'what "theory" really means' is not the kind of thing one can find out or establish. Differently put: this lengthy stagnation should suggest that perhaps it is the question what is defective; that perhaps answers do not seem adequate because the question needs to be reformulated. This is precisely one of the aims of the present paper.
5. THE SEMANTIC PREDICAMENT
We have thus far seen that the word 'theory' in the sociological language is rife with lexical ambiguities, and that this brings about undesirable consequences, including pseudo-disagreements, conceptual muddles, and even downright miscommunication. I want now to consider what is to be done about this situation. How can it be improved? How can these undesirable consequences be avoided? How ought sociologists to use the words 'theory,' 'theoretical,' and 'theorize'? Let us refer to this last question as the 'semantic predicament' (SP).
I propose two main arguments regarding SP, which I refer to as the 'principle of practical reason' and the 'principle of ontological and epistemological pluralism.' First, SP is a practical/political problem, which calls for practical/political methods and solutions. A solution to SP must be the outcome of a collective process of discussion and negotiation, which would look more like a community coming up with a political decision, than a physicist, mathematician, or philosopher coming up with a solution to an intellectual problem. Second, a solution to SP should make as few ontological and epistemological demands as possible. In other words, the meaning of 'theory' should not have too much built-in ontological and epistemological baggage.
I proceed as follows. In section 5, I show that the theory literature has generally been at odds with this ontological and epistemological pluralism, partly because of its neglect of the semantics of 'theory' and its resulting Socratic errors. In particular, I criticize the arguments of 'scientific theorists' and the 'types of theory' argument. In order to articulate the principles of practical reason and pluralism, a different understanding of 'theory' is needed, which I find in the writings of Levine and
6See Merton's AJS article "Sociological Theory" (1945), later included as "The Bearing of Sociological Theory on Empirical Research" in the first edition of Social Theory and Social Structure (1949); and his papers "On the History and Systematics of Sociological Theory" and "On Sociological Theories of the Middle Range," which first appeared as chapters in the 1968 edition of the book. See also the following articles, all of which were published in the ASR: Theodore Abel's (1952) "The Present Status of Social Theory"; Herbert Blumer's (1954) "What is Wrong with Social Theory?"; Howard P. Becker's (1954) "Vitalizing Sociological Theory"; and Parsons's "The Role of Theory in Social Research" (1938), "The Position of Sociological Theory" (1948), and "The Prospects of Sociological Theory" (1950).
THE MEANING OF 'THEORY'
185
Alexander. Building on these insights, in section 6 I go on to put forward my two principles.
5.1 Scientific Theory
A group of sociologists, sometimes referred to as 'scientific theorists,' unambiguously reject ontological and epistemological pluralism. For them, a sociological theory is a scientific explanation. If a given thing is not a scientific explanation, then it is not a sociological theory. This is how the argument goes. Sociology is the science of society. It applies the scientific method to social phenomena. Like any science, it offers objective accounts of what the world is like. And it does so by means of scientific laws (see, e.g., Braithwaite 1959:1). Then, sociological theory cannot be an ongoing dialogue with some classic texts or a story that makes sense of a bit of the social world. For scientists do not 'make sense' of things—they establish general laws in order to explain and predict them. Scientists do not enter into dialogues with classic texts such as Aristotle's Physics or Newton's Principia—that would be utterly useless. As Merton (1948:165) famously put it long ago, "[s]chools of medicine do not confuse the history of medicine with current theory, nor do departments of biology identify the history of biology with the viable theory now employed in guiding and interpreting research. Once said, this seems so obvious as to be embarrassing." Or, to take the case of economics, economic theory is definitely not the study of what Smith, Ricardo, or Keynes said or meant. The 'theory' published in economic theory journals (e.g., the Journal of Economic Theory) is filled with Greek letters, lemmata and theorems, not with long quotations and long footnotes. In brief, the argument is that insofar as sociology is a science, the theories of sociology must be analogous to the theories of other, presumably 'more mature' or 'more developed,' sciences. That is what theory is.
To take a classic example, George Homans had no doubt about what theory was and what kind of thing was not a theory but something else. As he put it in his ASA Presidential Address:
The explanation of a phenomenon is the theory of the phenomenon. A theory is nothing—it is not a theory—unless it is an explanation.
One may define properties and categories, and one still has no theory. One may state that there are relations between the properties, and one still has no theory. One may state that a change in one property will produce a definite change in another property, and one still has no theory. Not until one has properties, and propositions stating the relations between them, and the propositions form a deductive system—not until one has all three does one have a theory. Most of our arguments about theory would fall to the ground, if we first asked whether we had a theory to argue about. (Homans 1964a:812; emphasis in original; see also Homans 1964b, 1967, 1974, 1987)
Thus, according to Homans (1964a:811; emphasis added), "we should stop talking to our students about sociological theory until we have taught them what a theory is." Along comparable lines, Jonathan Turner believes that '"theorists' in sociology rarely theorize" (1985:24; see also Turner 1987, 1992, 2001, 2004). In other words, so-called 'theorists' are not in fact theorists. For instance, making a case for "laws" that are simplistic, abstract, and unconcerned with causality, Turner says that "these
186
SOCIOLOGICAL THEORY
are the characteristics of theory as opposed to current empirical descriptions that masquerade as theory" (Turner 1979: 440; emphasis added). Likewise, in sociology
we are typically given a causal model, often complete with statistical razzle-dazzle, on some substantive area. [...] [TJhere is nothing inherently wrong with such models per se; indeed, they represent refined efforts at empirical description. The problem comes only when sociologists begin to believe that such models are theory. They are not and cannot become theory, because they are tied to classes of observables in specific times and places. Theory must be more abstract, for it must transcend particular times and places. (Turner 1979:451; emphasis added)
So, theory is or must be X; theory is not or cannot be Y. X is a "true theory"; Y is a "masquerade" of theory (see also Turner 1985:25). Yet there is an obvious epistemological and methodological difficulty here. How does one know and find out what theory really isl How does one justify an ontological claim of this sort? What objective reasons can one give to someone who claims that theory really is Z? Scientific theorists argue that the method is simple. One just has to look at what natural science theory is and what natural scientists mean by 'theory.' That is what theory in sociology is and what sociologists ought to mean by 'theory' (e.g., Homans 1987:68).
Accordingly, one might want to replace the phrase 'sociological theory' with the phrase 'history of sociological theory' when referring to analyses and exegeses of classic texts (as Merton [1948, 1968a] and others have proposed). Or, like Randall Collins (1998:7), one might argue that "there is no sharp dividing line between theory and empirical research: if the research is important enough, it gets called theory." Pushing this logic even further, one might repudiate the existence of a theory specialty, journals, university courses, and ASA section, since "substantive relevant theoretical ideas ... should be and often are thoroughly incorporated into our substantive classes, publication organs, and section interests" (Chafetz 1993:1-2; see also Reed 1952:166).
Is the scientific theorists' argument persuasive? Let us first note that they often make the Socratic error. But this is not an insurmountable problem, because their argument could be consistently recast as being about what the word 'theory' ought to mean (rather than about what theory really is). However, there is a second, more serious problem with their equation of sociological theory with scientific explanation. It is obviously not my task here to take sides in the endless epistemological dispute over whether the social sciences should be modeled after the natural sciences. And this is precisely why I disagree with the scientific theorists' solution to SR
According to their reasoning, a stance on that epistemological dispute is the crucial premise from which a conclusion on the meaning of 'theory' is drawn. Now, if it turned out to be true that sociology must be all about coming up with general laws that look as much as possible like physical laws, then scientific theorists would be on the right track. Undoubtedly, the answer to 'what ought "theory" to mean?' cannot be divorced from the answer to 'what is sociology?', 'what is the business of sociology?' or 'what is sociology all about?' Unfortunately, we still cannot agree upon what sociology is all about. This is partly why there are so many competing understandings of theory and uses of the word 'theory' in sociology in the first place. So the scientific theorists' argument turns out to be circular! If one could convince all sociologists about what the true business of sociology is, SP would not be a major issue. The real challenge is to propose a well-founded solution to SP, which does not
THE MEANING OF 'THEORY'
187
depend on having a prior agreed-upon understanding about the nature of sociology. This solution should be as independent as possible from that other very thorny issue.7 It is perhaps ironic that, for all their aversion to 'intractable' 'philosophical' issues, the scientific theorists' argument is not at all independent of or orthogonal to them. On the contrary, it is logically dependent on a definite (and not particularly moderate) stance on one traditional philosophical issue: what kind of knowledge about human beings it is possible to attain.
5.2 Types of Theory
One staple of the theory literature—as well as of theory syllabi, conferences, and textbooks—is the 'types of theory' argument. In bypassing the semantics of 'theory,' this argument makes a different sort of mistake: it does allow for a certain kind of pluralism, yet at the unacceptable cost of conceptual incoherence. This is how the argument goes. There exist several 'types,' 'kinds,' 'styles,' or 'varieties' of theory and of theorizing (or theoretical 'perspectives,' 'paradigms,' 'viewpoints,' or 'approaches'). That is, theory comes in many varieties, in the same sense that tables come in many varieties: glass coffee tables, pool tables, solid oak dining tables, and so on. Still, all of these objects can be correctly described as articles of furniture having a flat horizontal surface supported by one or more vertical legs. Likewise, 'theory' is a legitimate class or (non-natural) kind. Theoriesi, theories2, etc. are in some ways different, but they still share some important properties. The word 'theory' refers to all the objects that share these important properties. These properties allow one not only to tell whether any given thing belongs to the class, but also to compare the worth of those that do, even when one has to compare the worth of, say, a theory i vis-á-vis a theory5.
The problem with this argument is that it is unclear what those important common properties are. How do these purported types hang together? What exactly is that thing they are types ofl Let us try to define 'theory' in such a way that we can capture all of the things sociologists refer to when they use the word. That is, this definition must be such that it would be semantically correct to say that someone who attempts to figure out what Marx really meant, someone who tries to resolve the problem of structure and agency, and someone who wants to state the conditions under which revolutions occur, are engaged in the same activity. Or, to put it slightly differently, that a commentary on Marx, an articulation of the relations between structure and agency, and a law-like statement about revolutions, are all members of the same class, or instances of the same concept.
Is it possible to pull off such a definition? I do not think so. In contrast to the case of 'table,' the definiens would be so broad as to both sound empty and be of little use. For example: 'sociological theory is that which sheds light on human behavior and social relations,' or 'sociological theory is that which helps one better understand, make sense of, etc. empirical data or the empirical world.' Not only do these two broad definientia sound a bit empty; they are in fact not broad enough, as they still leave out some of the seven meanings of 'theory' identified above. As an example, consider Martin Albrow's (1999:41) statement: "[n]o research can work without theory, even when it denies it has any. For theory simply means the connection of
7Obviously, one cannot see these two issues as completely orthogonal. If it turned out that the subject-matter of sociology is the behavior of the fairies that ultimately cause social action, and its method turned out to be introspection, the solutions we are currently prepared to suggest would become completely implausible.
188
SOCIOLOGICAL THEORY
ideas. Even counting is theory." But if 'theory' means "the connection of ideas," the class would be excessively large and the concept not really useful. According to this definition, 'theory' can be applied to any meaningful sequence of words whatsoever; indeed, 'atheoretical argument' is analytically contradictory.
It seems, then, that there is no reasonable and informative set of necessary and sufficient conditions for the application of the word 'theory' (nor is there a 'prototype' in Rosch's sense [Lakoff 1987; MacLaury 1991]). Nevertheless, there is a more sophisticated version of this sort of argument, based on Wittgenstein's (1953) notion of 'family resemblance.' Camic and Gross (1998:455), faced with the task of writing an Annual Review of Sociology piece on sociological theory, chose to take this route.
In discussing contemporary sociological theory, we consider an area with notoriously fuzzy boundaries. In identifying what works fall within these boundaries, we borrow Wittgenstein's notion of "family resemblances"... It is such family resemblances that constitute the contemporary theory field: criss-crossing similarities in terms of analytical issues and problems, intellectual ancestory and points of departure, vocabulary and style of argument, self-identification (calling one's own work "theory"), institutional membership (belonging to theory sections of sociological associations), group adoption (having one's contribution embraced by theorists), and more. No fixed cluster of these traits defines a family member nor makes it possible to track down all the stepchildren, distant cousins, and black sheep. Nonetheless, after one has spent some time among family members, it is not difficult to recognize the different branches of the family tree.
The notion of 'family resemblance' has been the subject of much debate, and is clearly not without its problems (see, e.g., Baker and Hacker 1980). For example, it makes it a bit mysterious how one can correctly use words (and, even more, how one can learn to correctly use words), whose rules of use cannot be fully articulated. To put it in Camic and Gross's terms, how exactly is it that after spending "some time among family members," the task becomes "not difficult"? What exactly happens to one after some time has passed? (Of course, the answer here cannot be that you have learned to imitate your mentors, so that you now know under which conditions your making certain noises will not elicit puzzled looks. For this answer would make the notion of 'family resemblance' pointless.) Furthermore, it might be argued that, say, theory i and theory7 cannot be convincingly said to resemble one another, except for their linguistic isomorphism.
More importantly, from the point of view of SP Camic and Gross's strategy is not very helpful. For the notion of 'family resemblance' only perpetuates the fuzziness around the use of 'theory.' Indeed, perpetuating these kinds of fuzziness is precisely its point! If one knew the necessary and sufficient conditions for the application of a term, there would be no need for such a notion. Sometimes it might be useful to hold things together in this fashion. Sometimes it might be the only way of making sense of ordinary language (or, for that matter, of the sociological language). Yet some other times—"for [some] special purpose," as Wittgenstein (1953:§69) says— one may need to draw a more precise boundary. Thus, in order to avoid the above-mentioned confusions, misunderstandings, and dead-end discussions, the extension and intension of 'theory' should be more precisely specified. To put it another way: the Wittgensteinian's typical argument is that one need not worry about the concept of a game, its logical structure or correct analysis. English speakers know how to use
THE MEANING OF 'THEORY'
189
the word, they understand one another, and that is all there is to it. But in our case this is not so. Ours is a case of a language gone awry. Speakers of the sociological language have significant disagreements about the use of the word 'theory.' Yet, at the same time, all of them are members of the community; they all, as it were, share a form of life. So the Wittgensteinian cannot help one here.
5.3 Taking Semantics Seriously
Unlike the previous two arguments, Levine and Alexander can provide the basis for a satisfactory articulation of the principles of practical reason and ontological and epistemological pluralism. Unlike the scientific theorists and other Socratic views, they take the semantics of 'theory' seriously. Thus, they do not try to authoritatively impose one or another meaning as the true one, nor do they try to establish what theory really is. Unlike the types of theory argument, they do not presuppose the existence of theory as a kind.
In fact, Levine has already offered a properly-semantic analysis of 'theory.' In "Social Theory as a Vocation," he disambiguates "the four distinct meanings of the term" (1997:2; emphasis added), and, as customary for a semantic analysis, uses numbers (subscripts) to distinguish between its senses:
The most common associations to the term theory reflect four distinct meanings of the term. The term can be construed in the sense of theory i—abstract or rational, as contrasted with empirical; theory2—general, as contrasted with particular; and theory3—contemplative, as contrasted with practical. There is also a function (theory4) that might be termed exegetical, as contrasted with heuristic. We might take a big step toward clarification simply by insisting that these meanings be kept distinct ... (Levine 1997:2; emphasis in original)
Even tough social science concepts are essentially contested (1985b:17; 1997:4; cf. Gallie 1964), Levine still believes that "[cjritical exploration of the meanings of key terms and their implications offer[s] a ... commendable route toward intellectual sophistication and clear thinking"; indeed, the "codification of the plurality of standard meanings can itself be a boon" (1997:4). This is what Levine began to do, even though briefly, concerning the meaning of 'theory'; and this is also what he had done at more length concerning the meaning of 'rationality' (1985c; see also 1965:5-11, 15-17; 1985a; 1985b; 1995). The present paper take up this task, following his suggestion that keeping the meanings of 'theory' distinct would be "a big step toward clarification."
Now, we have seen that many sociologists argue as though theory were an object out there that our concepts or language could track down. But the principles of practical reason and ontological and epistemological pluralism require that this Socratic position be transcended. In his four-volume work, Theoretical Logic in Sociology, Alexander forcefully rejects the Socratic ontology by means of an essentially semantic understanding of 'theory.'
In the first volume, Positivism, Presuppositions, and Current Controversies, Alexander presents a diagram that represents 'the scientific continuum and its components.' This diagram
clarifies ... the relative character of the theory/data split. That "data" is a thoroughly relative formulation can be illustrated by the fact that as social
190
SOCIOLOGICAL THEORY
Metaphysical environment
Empirical environment
h
■ c
o
o
2
Figure 1. The Scientific Continuum and its Components Source: Alexander (1982a:3).
scientists we continually treat as data the more general "scientific" formulation of those around us—others' propositions, models, classifications, and general assumptions about the empirical world. But it is also clear that "theory" is just as much a designational convenience. (Alexander 1982a:2-3)
'Theory' and 'data'—note Alexander's use of quotation marks—are just relative and convenient linguistic formulations. Moreover, 'theory' and 'data' are not qualitative but quantitative distinctions. Indeed, his decisive claim is that given a certain point in the continuum, one applies the word 'theory' to whatever is on its left and the word 'data' to whatever is on its right.
Although data and theory are, thus, commonly equated with qualitative positions on the more specific and general sides of the scientific continuum, it is more correct to understand them as quantitative distinctions: every formulation "leftward" of any given point of focus is called theory and every statement "rightward" of that point is claimed as data.
[...] If these elements actually were completely qualitatively differentiated, they would represent "concrete" distinctions. They are, instead, "analytic" distinctions, separations established for the convenience of scientific discourse, made to facilitate communication and not to establish ontological qualities. (Alexander 1982a: 3-4)
Thus, Alexander explicitly rejects the Socratic ontology and makes an essentially semantic point. It is a mistake to think of 'theory' and 'data' as if they were ontological qualities, which are part of the fabric of the world and hence one can pin down and refer to once and for all. Rather, this is just a convenient analytic distinction that facilitates communication. There is nothing for theory to really be. 'Theory' is a relative term, which can be more or less analytically useful, which can cause more or less confusions, which can result in better or worse communication, etc.
This point is also evident in another key feature of Alexander's diagram, whose value can be better appreciated vis-a-vis Merton's typology in his article "Sociological Theory." Merton writes: "The phrase 'sociological theory' has been used to refer to at least six types of analysis which differ significantly in their bearings on empirical
THE MEANING OF 'THEORY'
191
research. These are methodology, general orientations, conceptual analysis, post factum interpretations, empirical generalizations, and sociological theory" (Merton 1945:462; see also Merton 1945:462-63, 1968b:39). At first glance, this statement appears to take the semantics of 'theory' seriously. Merton speaks of linguistic elements such as 'phrase' (and, later on, 'term'), linguistic relations such as 'usage' and 'reference,' puts quotation marks around 'sociological theory' to indicate mention rather than use, and seems to imply that the expression 'sociological theory' has several senses. However, Merton quickly forgets about word meanings and senses, and goes on to show that some of these 'six types of analysis' are not really theory. For example, speaking of the first and second types—'methodology' and 'general orientations'—he says: "we should distinguish clearly between sociological theory, which has for its subject matter certain aspects of the interaction of men and is hence substantive, and methodology, or the logic of scientific procedure"; and "[m]uch of what is described in textbooks as sociological theory consists of general orientations toward substantive materials" (Merton 1945:463, 464). Most revealingly, Merton refers to the last of his six types as 'sociological theory.' So, sociologists have used the term 'sociological theory' to refer to six different things only one of which turns out to be sociological theory (without quotation marks)!
The contrast with Alexander could not be greater. There are 10 terms in Alexander's continuum ('ideological orientations' is added in Alexander 1982a:40 and 1987:7). Remarkably, the term 'theory' is not one of them. Because of its being a relative rather than an ontological quality, 'theory' cannot be precisely located on the continuum. Instead, the use of the word 'theory' is a function of the particular point of reference chosen, the particular interests and goals of the enterprise, and so on. Furthermore—and this is crucial for my argument below—Theoretical Logic makes a strong case for ontological and epistemological pluralism. Alexander believes that general presuppositions are of the utmost importance, he shows why this is so, and he chooses to focus on them in the subsequent three volumes on Marx and Dürkheim, Weber, and Parsons. But he does not believe that it would be wrong to apply the word 'theory' to models, classifications, or laws, let alone that those things are not really theory. In fact, the tendency "to reduce—to 'conflate'— theoretical argument to one or another particular set of nonempirical commitments" is one of the main antagonists of the whole project (Alexander 1982b:xviii; see also 1982b:373).
6. SEMANTICS, POLITICS, AND PLURALISM
Armed with these lessons from Levine and Alexander—and aware of the missteps of other writers—I am now in a better position to address the semantic predicament (SP). What ought the word 'theory' to refer to in the sociological language? More generally, by what criteria is this to be determined?8 I argue that, while prima facie this may appear to be a technical semantic issue, here sociology is faced with a practical/political problem. This is the principle of practical reason. Then I argue for the above-mentioned principle of ontological and epistemological pluralism. Yet
8 Methodologically, the question is by what criteria one determines what any given word ought to refer to in a language like the sociological language. What is the extension of 'revolution,' 'social movement,' or 'ethnicity' (i.e., to which objects can you correctly apply these terms)? How about 'rational action,' 'globalization,' 'status,' or 'moral value'? How are semantic disagreements to be adjudicated? These are critical methodological questions, which sociology is far from having an adequate answer to.
192
SOCIOLOGICAL THEORY
before presenting these two principles, I make the practical suggestion that sociology needs a semantic therapy in order to clarify some of its conceptual confusions.
6.1 Semantic Therapeutics
If my argument up to this point is correct, the first thing sociology needs is some sort of semantic therapy. All sociologists should be fully aware that their disagreements about theory have a semantic dimension, which has important effects on the appropriateness and forcefulness of different kinds of arguments. If this point became common sociological wisdom, that would surely amount to a step forward. For instance, no theory discussion would forget that there are many senses of the word 'theory' and no real referent or true meaning; that the many things that the word 'theory' is used to express are quite different indeed; or that the ontological, evaluative, and teleological questions in their customary form are problematic. Full consciousness of these facts would just dissolve numerous problems and disputes— namely, those that are ultimately caused by semantic vagueness. Further, it would clarify those (also numerous) problems and disputes that would still persist, pinpoint with more precision what the dispute is about, make discussion easier, and ultimately make substantive progress possible.
Therefore, first, I suggest that a semantic therapy regarding the meaning of 'theory' should be part of university courses called 'sociological theory,' theory textbooks, certain Theory Section activities, and other appropriate forums. Second, I propose that each time sociologists engage in a debate about 'theory'—indeed, each time they use the word 'theory' in a potentially confusing context—they make it clear whether they mean theory], theory2, theory3, etc. For instance, if a journal referee rejects a paper on the grounds that it is 'undertheorized,' or that it needs 'more theory,' she should unambiguously say if she means theory2, theory4, or theory? (rather than hiding behind this semantic ambiguity). Authors should indicate whether their projects make a contribution to theory 1, theory3, or theory5 (rather than ambiguously claiming that they are 'very theoretical' or 'driven by theory' as a badge of epistemic significance).
Unfortunately, this is only a first step forward toward making things better. For, as sociologists know full well, words are sites of power (e.g., Bourdieu 2001). The meaning of 'theory' is intimately related to very real institutional resources, careers, funding, prestige, status systems, sociology's public relevance, and so on. Moreover, 'theory' and 'theoretical' are 'thick' terms (Williams 1978, 1985), that is, they can simultaneously do descriptive and evaluative work. Specifically, in numerous contexts applying the expressions 'theoretical,' 'relevant for theory,' 'informed by theory,' and 'theory-driven' to a paper, project, scholar, or line of research is not just to say that they are of a certain nature, but also implies a favorable judgment. While there is only one sign, one English word-form 'theory,' there are many people who want to use it in different ways, and who have a practical interest in doing so. There is only one Theory Section in the ASA, whose practical interest is to enlarge its membership and ascendancy.9 Useful and necessary as it is, no semantic therapy will do by itself. For I want to argue that SP is to a large extent a political problem.
9Likewise, there is only one journal called Sociological Theory, a fact that was patent in a recent debate in the pages of the newsletter Perspectives (Adams 2006; Kidd 2006; Perrin 2006; Sanderson 2006a, 2006b; Wilkes 2006).
THE MEANING OF 'THEORY'
193
6.2 The Principle of Practical Reason
The lexical semantics of a completely formal language is determined by its inventor. As Hilary Putnam (1975:55; emphasis in original) puts it, "[a] formal language has, after all, an inventor, and like any human being, he can give commands. Among the commands he can issue are ones to the effect that 'If you want to speak my language, then do thus and so'." Putnam's point is that if you do not do thus and so, then you may be speaking some other language, but you are not speaking mine. Suppose the question arises, 'what does the word "malabia2614" mean in your language?' The method to answer this question does not involve argumentation, discussion, and negotiation. The correct answer can be found in the dictionary I wrote for my language. That is what 'malabia2614' really means.
The lexical semantics of natural languages does not derive from a process of explicit argumentation, discussion, and negotiation either. Nor can it be established by an authority such as the OED, a summit of linguists, or philosophical analysis. To put it in Wittgenstein's terms, what it is to follow a rule (linguistic or otherwise) is fixed by social practice. It cannot be fixed 'privately.' This is why Humpty Dumpty's stance is preposterous. Rather, word meanings derive from the very complex social and linguistic processes studied by historical lexicology, lexicography, and semantics (Brinton and Traugott 2005; Keller 1994; Labov 1994, 2001). The question, 'what does the word "paper" mean in the English language?,' is an empirical one. Lexicographers have empirically established that at present the word 'paper' has a certain number of senses. Those are the things you can correctly refer to by using the word 'paper.' That is what 'paper' really means. The question, 'what ought the word "paper" to mean in the English language?,' just does not make sense.
By contrast, the nature of the sociological language (and other comparable semi-formal languages) is such that the meaning of its terms is partially fixed by the social practices that occur within a relatively small and self-contained community of speakers. In our case, the community of speakers of the sociological language has a relative autonomy to agree upon what it is to follow this linguistic rule (i.e., under which conditions it is semantically appropriate to use the word 'theory'). Its autonomy is relative because there are external linguistic, social, and institutional constraints on what can count as a reasonable agreement—for instance, a linguistic constraint is the meaning of 'theory' in the English language. Presently, though, the use of 'theory' in the sociological language is not constant across users. Then, the question becomes how to establish standards of validity and merit for arguments, methods, and evidence that aim at settling this semantic divergence.
We have, then, a community, which has a relative autonomy to make a semantic choice and come to a semantic agreement. While it is not the case that any answer is as good as any other, there is no one true or correct answer either. This choice has to be made in a particular social context. Moreover, it has not only epistemic but also institutional and economic consequences. Thus, we have a community faced with a typically political or practical-reason problem. On my account, the six defining characteristics of SP as a political problem are the following:
(1) There are several parties, which differ in their preferences, desires, interests, values, beliefs, and practices. (For example, some people use the word 'theory' in such-and-such a way, some people have such-and-such conception of what sociology is all about, some people like and some people dislike postmodernism, rational choice, etc.).
194
SOCIOLOGICAL THEORY
(2) These preferences, desires, interests, values, beliefs, and practices are subject to public scrutiny and rational appraisal. One can demonstrate that a party's stance is false, unreasonable, confused, pernicious, or insane in such a way that any rational person ought to accept it. (For instance, if someone said that 'theory' ought to be roughly synonymous with 'method' or with 'cheesecake'; or that 'what does "theory" mean?' is a question for a quantum physicist.)
(3) However, there is no one single correct set of preferences, desires, interests, values, beliefs, and practices. There can be legitimate differences here. Crucially, if these differences lead to practical conflicts, these cannot be arbitrated from outside the political field (as one could do in the case of a disagreement over whether the first name of the current President of Uruguay is Tabare or Nestor).
(4) The problem has a set of factual initial conditions, which must be taken into account: a certain distribution of power, status, and resources; certain existing institutions (such as university courses called 'Sociological Theory' and theory journals); the histories of the relevant concepts, ideas, and institutions; and a host of other empirical facts about sociology and its social context.
(5) Procedurally, a satisfactory solution must be arrived at through characteristically political mechanisms. These can include rational argumentation and discussion, persuasion, aggregation and weighting of preferences, etc. Unlike those countries that have constitutions, these mechanisms are not already in place in the relevant community (the speakers of the sociological language). Therefore, they must themselves be the subject of a political conversation. Thought experiments such as ideal-speech situations, social contracts, or veils of ignorance are presumably excluded, because SP is an effectively practical problem that calls for an effectively practical solution.
(6) Substantively, a satisfactory solution must take all the different legitimate stances into account. Again, if a political mechanism generated a solution that excludes all but one point of view, one should be able to reject the outcome from outside the political field. The solution must both depend on internal political considerations, and external standards of reasonableness, acceptability, justice, fairness, effectiveness, applicability, and the like. These standards are an external check on the outcome of the political process.
Thus, my argument is that SP is in the domain of 'practical reason' (as opposed to 'theoretical reason,' as these terms are typically understood in philosophy—see Audi 1989). The central aim of practical reason is to guide action. Its chief concern is not what the world is like but with what one is to do. However, the notion of practical reason brings to the fore the role of reason in this kind of situations. Agents engage in reasoning, reflection, deliberation, be it individually or collectively. The end-products of this process—courses of action—might be more or less reasonable (or unreasonable), and even more or less rational (or irrational). This reasonableness or rationality is a function of many factors, the subjective preferences of the actors being only one of them. Thus, there are better and worse answers to the question, 'what ought "theory" to mean in the sociological language?' But one should not expect something like a true and exact solution to a mathematical problem (which a mathematician may resolve in the abstract, on her own, by means of armchair—or bathtub—reflection). Rather, one should expect something like a fair and reasonable solution to a political problem, arrived at by means of political mechanisms for collective decision-making. This solution cannot be discovered; it must be created. An individual person cannot (and this 'cannot' is a logical one) resolve it. Sociology's
THE MEANING OF 'THEORY'
195
discussion about SP should be conscious of and not embarrassed by its political nature.
6.3 The Principle of Ontological and Epistemological Pluralism
My second principle concerns the content of a solution to SP. Given my claim about SP's political nature, I present it as an invitation to discussion. I believe that a satisfactory solution to SP should make as few ontological and epistemological demands as possible. The set of conditions under which the word 'theory' can be correctly used should not have too much built-in ontological and epistemological baggage. I call this the 'principle of ontological and epistemological pluralism.'
The reason why I advocate this principle is, very roughly put, the following. Suppose sociologists made a certain picture of the world or idea about what can be known a prerequisite for something being a sociological theory at all. Consider some examples. We may demand that theories be underlain by the assumption that "the social world consists of fixed entities with variables attributes" (Abbott 1988:169). We may require that causality be taken to be the cement of the universe, the most important relation that can hold between two entities. Or, we could build into the definition of 'theory' the idea that social processes are regulated by laws of nature. Alternatively, we may demand the belief that the distinction between text and reality is misleading, or even the belief that there are no such things as 'reality' and 'objectivity.' Or else, we may demand the assumption that nothing exists but what can be actually observed or otherwise grasped by our senses, thereby denying existence to such 'mysterious' things as causality and similar 'underlying theoretical mechanisms' (Steinmetz 2005). In any of these scenarios, only to the extent that you shared the required ontology or epistemology, could you be said to have a theory of the social world. You could have other things about the social world—opinions, views, beliefs, ideas—but not a theory. By definition, that particular ontology or epistemology would be obligatory for one to be allowed to enter a theoretical discussion, make a theoretical contribution, or theorize at all.
There are two main problems with this approach. First, whatever 'sociological theory' turns out to mean, its meaning will not be identical with the meaning of 'ontological theory' or 'epistemological theory.' In other words, it is reasonable to assume that sociologists' theories do not mainly concern themselves with what it is for something to exist and what things actually exist. Arguably, the kind of stuff that sociological theories concern themselves with is (or, at least, is related to) the kind of stuff sociology is principally interested in—society, social relations, social interaction, social change, etc. Ontology and epistemology have a different subject-matter. Naturally, they also have their own methodological and substantive debates, some arguments that are generally seen as defeated and some arguments that are generally seen as valid, methodological rules, conceptions of epistemic progress, and so on. All of this is outside the jurisdiction and competence of sociology. Thus, sociology is not in a position to adjudicate—nor should it be interested in adjudicating— between competing ontological and epistemological arguments. Insofar as it is not clear how these controversies will turn out, sociological theories should be cautious and not commit themselves to one or another view. And the strongest commitment one can in this respect make is a semantic one—that the very meaning of the word 'theory' carry such ontological or epistemological baggage that some 'theories' turn out not to be theories but misuses of the word (as scientific theorists would have it).
196
SOCIOLOGICAL THEORY
Second, the consequence of this policy for sociology would be a very impoverished discussion. This is the typical argument for pluralism. Dialogues about society, social relations, social processes, etc. between parties that differ in their ontology and epistemology are potentially enlightening, and hence one should want that they be conducted within the realm of 'theory.' Any of these parties may have things to contribute to the discipline of sociology and the subfield of sociological theory, so it would be unwise to draw a semantic boundary that excludes but one of them. Let me illustrate this point. Would it be correct to apply the (sociological language) word 'theory' to Charles Taylor's (1989) Sources of the Self} Well, it is easy to make a semantic argument to the effect that it would not. If by 'sociological theory' you mean something like 'natural science theory,' and thereby build into the meaning of 'theory' the ontology and epistemology of 'scientists naturalism' (Steinmetz 2005), then Taylor does not advance a theory of the self. In fact, he has been one of the most influential opponents of that sort of views (Taylor 1985). Nevertheless, many sociologists would find the outcome of this semantic argument—namely, the a priori segregation of Sources of the Self from sociological theory and sociology—both highly implausible and terribly unfortunate.
I would like to conclude with three points of clarification. First, I am not objecting to any particular type of ontology or epistemology, but to any ontologically or epis-temologically demanding solution to SR My argument is neither for nor against the metaphysics that realism, 'scientistic naturalism,' or postmodernism subscribe to. I illustrated my point using Taylor's Sources of the Self, but I might as well have used, say, Coleman's (1990) Foundations of Social Theory. Second, individual sociological arguments (as opposed to sociology itself or sociological theory itself) may—and many times willy-nilly must—implicitly or explicitly favor, say, a certain ontological conception about what the social world is made of. But my point is that: (a) those individual sociological arguments should not conceive of this as a strong ontological commitment; (b) sociology or sociological theory itself should not have this kind of commitments; (c) they definitely should not be built into the very meaning of the word 'theory' in the sociological language. Finally, I wrote above that "a satisfactory solution to SP should make as few ontological and epistemological demands as possible." As the italicized locutions 'as few ... as possible' and (in the following sentence) 'too much' indicate, my argument is not that one should not make any ontological and epistemological demands. For example, consider a worldview which believes in occult witches who have causal powers over social processes, and which sees the examination of witchcraft substance as the only path to knowledge. We may agree that a sensible rule of the sociological language would be something like this: the application of the word 'theory' is restricted to objects that do not involve this type of ontology and epistemology.
REFERENCES
Abbott, A. 1988. "Transcending General Linear Reality." Sociological Theory 6:169-86.
Abel, T. 1952. "The Present Status of Social Theory." American Sociological Review 17:156-64.
Abend, G. 2006. "Styles of Sociological Thought: Sociologies, Epistemologies, and the Mexican and U.S.
Quests for Truth." Sociological Theory 24(l):l^tl. Adams, J. 2006. "Why Pluralism Still Makes Sense for Sociological Theory: Reply to Stephen Sanderson."
Perspectives 28(3): 1-3.
Adams, J., E. Clemens, and A. S. Orloff. 2005. "Social Theory, Modernity, and the Three Waves of Historical Sociology." Pp. 1-72 in Remaking Modernity: Politics, History and Sociology, edited by J. Adams, E. Clemens, and A. S. Orloff. Durham, NC: Duke University Press.
THE MEANING OF 'THEORY'
197
Albrow, M. 1999. Sociology: The Basics. London: Routledge.
Alexander, J. C. 1982a. Positivism, Presuppositions, and Current Controversies. Vol. 1 of Theoretical Logic
in Sociology. Berkeley, CA: University of California Press. -. 1982b. The Antinomies of Classical Thought: Marx and Durkheim. Vol. 2 of Theoretical Logic in
Sociology. Berkeley, CA: University of California Press. -. 1987. Twenty Lectures: Sociological Theory Since World War II. New York: Columbia University
Press.
Allan, K. 2001. Natural Language Semantics. Maiden, MA, and Oxford: Basil Blackwell. Audi, R. 1989. Practical Reasoning. London and New York: Routledge.
Baker, G. P. and P. M. S. Hacker. 1980. Wittgenstein: Understanding and Meaning. Chicago: University of Chicago Press.
Becker, H. [P.]. 1954. "Vitalizing Sociological Theory." American Sociological Review 19:377-88.
Blumer, H. 1954. "What is Wrong with Social Theory?" American Sociological Review 19:3-10. Bourdieu, P. 2001. Langage et pouvoir symbolique. Paris: Seuil.
Braithwaite, R. B. 1959. Scientific Explanation: A Study of the Function of Theory, Probability and Law
in Science. Cambridge: Cambridge University Press. Brinton, L. J. and E. Traugott, eds. 2005. Lexicalization and Language Change. Cambridge; New York:
Cambridge University Press. Calhoun, C. 1998. "Explanation in Historical Sociology: Narrative, General Theory, and Historically
Specific Theory." American Journal of Sociology 104:846-71. Camic, C. and N. Gross. 1998. "Contemporary Developments in Sociological Theory: Current Projects
and Conditions of Possibility." Annual Review of Sociology 24:453-76. Chafetz, J. S. 1993. "Sociological Theory: A Case of Multiple Personality Disorder." Perspectives 6(1): 1-2. Christian, B. 1988. "The Race for Theory." Feminist Studies 14(l):67-79. Coleman, J. S. 1990. Foundations of Social Theory. Cambridge: Harvard University Press. Collins, R. 1998. "Randall Collins' Top Bunch." Perspectives 20(3):7-8. Cruse, D. A. 1986. Lexical Semantics. Cambridge and New York: Cambridge University Press. Davidson, D. 1984. Inquiries into Truth and Interpretation. Oxford: Clarendon Press; New York: Oxford
University Press.
de Man, P. 1986. The Resistance to Theory. Minneapolis, MN: University of Minnesota Press. Fine, G. A. 2001. "Message from the Chair: Sociological Theory and Empirical Research." Perspectives 24(4): 1, 4.
Gallie, W. B. [1956] 1964. "Essentially Contested Concepts." Pp. 157-91 in Philosophy and the Historical
Understanding. London: Chatto & Windus. Goldstone, J. A. 1998. "Initial Conditions, General Laws, Path Dependence, and Explanation in Historical
Sociology." American Journal of Sociology 104:829—45. Goode, W. J. 1973. Explorations in Social Theory. New York: Oxford University Press. Hempel, C. G. 1965. Aspects of Scientific Explanation, and Other Essays in the Philosophy of Science. New
York: Free Press.
Homans, G. C. 1964a. "Bringing Men Back In." American Sociological Review 29(6):809-18.
-. 1964b. "Contemporary Theory in Sociology." Pp. 951-77 in Handbook of Modern Sociology,
edited by R.E.L. Faris. Chicago, IL: Rand McNally.
-. 1967. The Nature of Social Science. New York: Harcourt, Brace & World.
-. 1974. Social Behavior: Its Elementary Forms. New York: Harcourt, Brace, Jovanovich.
-. 1987. "Behaviourism and After." Pp. 58-81 in Social Theory Today, edited by A. Giddens and J.
H. Turner. Stanford, CA: Stanford University Press. Jay, M. 1996. "For Theory." Theory and Society 25(2):167-83.
Keller, R. 1994. On Language Change: The Invisible Hand in Language. Translated by B. Nerlich. London; New York: Routledge.
Kidd, D. 2006. "Close and Open-Form Approaches to Social Theory: A Pedagogical Rejoinder to Sanderson." Perspectives 28(3):5-6.
Kiser, E. and M. Hechter. 1991. "The Role of General Theory in Comparative-Historical Sociology." American Journal of Sociology 97(1): 1-30.
-. 1998. "The Debate on Historical Sociology: Rational Choice Theory and Its Critics." American
Journal of Sociology 104:785-816.
Knapp, S. and W. Michaels. 1982. "Against Theory." Critical Inquiry 8(4):723-42.
Kuhn, T. S. [1962] 1970. The Structure of Scientific Revolutions, 2nd ed. Chicago: University of Chicago Press.
198
SOCIOLOGICAL THEORY
Labov, W. 1994. Principles of Linguistic Change. Volume 1: Internal Factors. Cambridge, MA; Oxford, UK: Blackwell.
---. 2001. Principles of Linguistic Change. Volume 2: Social Factors. Maiden, MA; Oxford, UK:
Blackwell.
Lakoff, G. 1987. Women, Fire, and Dangerous Things. Chicago: University of Chicago Press.
Landau, S. I. 2001. Dictionaries: The Art and Craft of Lexicography, 2nd ed. Cambridge; New York:
Cambridge University Press. Levine, D. N. 1965. Wax and Gold: Tradition and Innovation in Ethiopian Culture. Chicago: University of
Chicago Press.
-. 1985a. "Ambiguity and Modernity." Pp. 20^43 in The Flight from Ambiguity: Essays in Social
and Cultural Theory. Chicago: University of Chicago Press.
-. 1985b. "The Flight from Ambiguity." Pp. 1-19 in The Flight from Ambiguity: Essays in Social
and Cultural Theory. Chicago: University of Chicago Press.
-. 1985c. "Rationality and Freedom, Inveterate Multivocals." Pp. 142-78 in The Flight from Ambiguity: Essays in Social and Cultural Theory. Chicago: University of Chicago Press.
-. 1995. Visions of the Sociological Tradition. Chicago: University of Chicago Press.
-. 1997. "Social Theory as a Vocation: Engaging with Future Challenges." Perspectives 19(2): 1—8.
Lieberson, S. 1991. "Small N's and Big Conclusions: An Examination of the Reasoning in Comparative Studies Based on a Small Number of Cases." Social Forces 70(2):307-20.
-. 1992. "Einstein, Renoir, and Greeley: Some Thoughts about Evidence in Sociology." American
Sociological Review 57(1):1-15.
Lyons, J. 1977. Semantics. (2 volumes.) Cambridge and New York: Cambridge University Press.
MacLaury, R. 1991. "Prototypes Revisited." Annual Review of Anthropology 20:55-74.
Mahoney, J. 2004. "Revisiting General Theory in Historical Sociology." Social Forces 83(2):459-89.
Mahoney, J. and D. Rueschemeyer, eds. 2003. Comparative Historical Analysis in the Social Sciences. Cambridge, UK: Cambridge University Press.
Merton, R. K. 1945. "Sociological Theory." American Journal of Sociology 50(6):462-73.
-. 1948. "Discussion." American Sociological Review 13:164-68.
-. 1949. Social Theory and Social Structure. Toward the Codification of Theory and Research. Glencoe,
IL: Free Press.
-. 1968a. "On the History and Systematics of Sociological Theory." Pp. 1-38 in Social Theory and
Social Structure. 1968 enlarged ed. New York: Free Press. -. 1968b. "On Sociological Theories of the Middle Range." Pp. 39-72 in Social Theory and Social
Structure. 1968 enlarged ed. New York: Free Press. Nagel, E. 1961. The Structure of Science: Problems in the Logic of Scientific Explanation. New York:
Harcourt, Brace & World.
Paige, J. M. 1999. "Conjuncture, Comparison, and Conditional Theory in Macrosocial Inquiry." American
Journal of Sociology 105:781-800. Parsons, T. 1938. "The Role of Theory in Social Research." American Sociological Review 3:13-20.
-. 1948. "The Position of Sociological Theory." American Sociological Review 13:156-64.
-. 1950. "The Prospects of Sociological Theory." American Sociological Review 15:3-16.
Perrin, A. J. 2006. "Integrating Theoretical Sociology: A Reply to Sanderson." Perspectives 28(3):3-5 Putnam, H. 1975. Mind, Language, and Reality. Cambridge and New York: Cambridge University Press. Quadagno, J. and S. J. Knapp. 1992. "Have Historical Sociologists Forsaken Theory? Thoughts on the
History/Theory Relationship." Sociological Methods and Research 20(4):481-507. Ravin, Y. and C. Leacock, eds. 2000. Polysemy: Theoretical and Computational Approaches. Oxford: Oxford
University Press.
Reed, S. W. 1952. "The Present Status of Social Theory: Discussion." American Sociological Review 17(2): 166-67.
Ritzer, G. 1980. Sociology: A Multiple Paradigm Science. Boston: Allyn and Bacon.
Robinson, R. 1980. "Socratic Definition." Pp. 111-24 in The Philosophy of Socrates. A Collection of
Critical Essays, edited by G. Vlastos. Notre Dame: University of Notre Dame Press. Sager, J., ed. 2000. Essays on Definition. Amsterdam; Philadelphia, PA: Benjamins.
Sanderson, S. K. 2006a. "Reforming Theoretical Work in Sociology: A Brief Reply to My Critics." Perspectives 28(3):9-ll.
-. 2006b. "Reforming Theoretical Work in Sociology: A Modest Proposal." Perspectives 28(2):
Seidman, S. and J. Alexander, eds. 2001. The New Social Theory Reader. London: Routledge. Skocpol, T. and M. Somers. 1980. The Uses of Comparative History in Macrosocial Inquiry. Comparative Studies in Society and History 22:174—97.
THE MEANING OF 'THEORY'
199
Somers, M. 1998. " 'We're No Angels': Realism, Rational Choice, and Relationality in Social Science."
American Journal of Sociology 104:722-84. Steinmetz, G. 2005. "Scientific Authority and the Transition to Post-Fordism: The Plausibility of Positivism
in U.S. Sociology Since 1945." Pp. 275-323 in The Politics of Method in the Human Sciences. Positivism
and its Epistemological Others, edited by G. Steinmetz. Durham and London: Duke University Press. Stinchcombe, A. L. 1968. Constructing Social Theories. New York: Harcourt, Brace & World. Taylor, C. 1985. Philosophy and the Human Sciences. Cambridge and New York: Cambridge University
Press.
-. 1989. Sources of the Self: The Making of the Modern Identity. Cambridge, MA: Harvard University
Press.
Turner, J. H. 1979. "Sociology as a Theory Building Enterprise: Detours from the Early Masters." Pacific
Sociological Review 22(4):427-56. -. 1985. "In Defense of Positivism." Sociological Theory 3(2):24-30.
-. 1987. "Analytical Theorizing." Pp. 156-94 in Social Theory Today, edited by A. Giddens and
J. H. Turner. Stanford, CA: Stanford University Press. -. 1992. "If Not Positivism, Then Why Is Sociology Important?" Canadian Journal of Sociology
17(1):54-61.
-. 2001. "Sociological Theory: Editor's Report." Perspectives 23(2):l-2, 7-8.
-. 2004. "Is Grand Theory Dead and Should Sociology Care? A Reply to Lamont." Perspectives
27(2):2, 11-12.
Useem, B. 1980. "Solidarity Model, Breakdown Model, and the Boston Anti-Busing Movement." American Sociological Review 45:357-69.
Wilkes, C. 2006. "A Modest Proposal Indeed." Perspectives 28(3):7-9.
Williams, B. A. O. 1978. Descartes: The Project of Pure Enquiry. Hassocks: Harvester Press.
-. 1985. Ethics and the Limits of Philosophy. Cambridge: Harvard University Press.
Wittgenstein, L. 1953. Philosophical Investigations. Translated by G. E. M. Anscombe. Oxford: Basil Black well.