Skip to main content Skip to Search Box

Definition: Discourse Analysis from Dictionary of Information Science and Technology

an approach to the analysis of naturally occurring talk that considers the social actions and consequences achieved by the words’ use, and the resources which participants employ to achieve these actions (Pedersen & Smithson, 2010)

Summary Article: Discourse Analysis
from Encyclopedia of Case Study Research

The term discourse and its adjective, discursive, are not meaningful research terms unless modified by additional descriptors. One can say something about Foucauldian or Derridean discourse analysis, critical discourse analysis, narrative discursive approaches, and so forth, but one should be wary of any research claiming merely to be “discursive” and any methodology claiming only to analyze “discourse.” As a term in common parlance since at least the mid-16th century, discourse is legitimately used to describe any speech, any conversation, and any process of reasoning. It may even still be used in its original etymological sense, meaning “to run to and fro.” Anyone whose research is spoken or written—that is, anyone— may claim with some justification that such research is discursive.

Conceptual Overview and Discussion

Approaches to social science research claiming to be discursive in a specifically methodological and/ or epistemological sense began to appear in social science theorizing in the 1980s, usually in scholarship claiming either to be deconstructive, with reference to the work of Jacques Derrida, or genealogical or archaeological, with reference to the scholarship of Michel Foucault. Derrida and Foucault do not stand alone, but they, far more than their colleagues (including Deleuze and Guattari, Kristeva, Spivak, Baudrillard, and others) have become popular in the English-language social sciences. Much of this discursive work has claimed to be, or has been labeled by others, poststructuralist or postmodernist. Such work can be understood to represent one variety of anti-essentialist, or social constructionist, thought.

Conceptually, such work is grounded in Saussurian linguistics.

From this perspective, one can speak meaningfully of discursive systems as an object of analysis. When used this way, a discursive system has much in common with a paradigm as described by Thomas Kuhn in The Structure of Scientific Revolutions. In both cases, the reference is to a system of interlinked assumptions, values, beliefs, and interpretations forming a specific common sense about what is real, true, good, and important. The main difference between these terms would be that calling such common sense a discursive system directs one to look at this system through a linguistic lens, considering text as the data for analysis that will yield information about the underlying beliefs and their consequences. One must be careful not to automatically assume that the term discourse is being used in this way when one encounters it because, like the term paradigm, it has been used in a plethora of ways, some incompatible with each other, some questionable in their own right.

Broadly speaking, these approaches to scholarship are all heirs to the phenomenological philosophical tradition. One broad cluster of positions within this tradition, a cluster primarily indebted to Hegel, attempts to derive from interpretive reality an underlying, hermeneutic truth. Recent heirs to this tradition would include Hans-Georg Gadamer and Jürgen Habermas. Another, perhaps broader cluster emerging from this tradition abandons the idea of objective truth for more constructionist/ anti-essentialist assumptions within which reality is treated less as an objective thing to be discovered and more as a complex, multiply determined entity whose meanings are socially produced. The distinctive quality of discursive approaches to research would be reliance on the root metaphor of textuality.

If reality is a social product, it follows that the structure of a society's language contains evidence of the process by which that reality is produced and that determines the boundaries to what can and cannot be real within that common sense. When Foucault referred to his early work as archaeological, it was in this sense. Like an archae-ologist digging through an ancient city, a “city of language” (a metaphor originating with Ludwig Wittgenstein), Foucault was exploring in language the sedimented evidence of the assumptions; the values; the common sense through which, for instance, a phenomenon such as madness could have one set of meanings in one era and a contradictory set of meanings in another.

Textuality as a Root Metaphor

In a review of the organizational culture literature in the early 1980s, Linda Smircich distinguished between scholars who treat culture as a variable and those for whom it is a root metaphor. For the latter group, culture becomes the lens through which sense data are interpreted. Culture is not a topic among others; instead, all reality is significant to the extent that it is represented in cultural terms. From this perspective meanings are, axiomatically, cultural meanings. Analogously, for a Freudian psychologist sexuality is a root metaphor; psychodynamic processes are interpreted through a lens of sexuality. Marxian theories of surplus labor value appropriation constitute a root metaphor for Marxists. It is in this sense that textuality can form a viable root metaphor for analysis of social reality as discourse, as discursive systems—as text. It is not that everything is really text but that everything that has meaning functions textually and can be studied through the root metaphor lens of discourse.

One might, for instance, look at the discourse of universal human rights in Western history. Whatever good and bad has been done in the name of these values, there is an identifiable set of core statements and beliefs that powerfully affect actions, from political colonization of much of the world in past centuries to the recent American war on terror. The discourse of human rights has provided a language of social progress, of the “white man's burden,” of liberating people and of democ-ratizing. The relationship among knowledge, power, and action in these highly influential events can be profitably studied by seeking the data contained in text as evidence. Although there have been multiple approaches to discursive analysis, all tend toward one of two poles on an essentialist-anti-essentialist epistemological continuum. Empiricist approaches, including, for instance, Peircean, Chomskian, and critical realist perspectives, are grounded in the assumption that language is a reflection of an underlying reality. The role of analysis, then, is to clarify language to make it a nondistorting conduit for that underlying reality. Moving toward the opposite end of the continuum, many interpretive, poststructuralist, postcolonial, and related perspectives assume a more constructivist view of reality; that is, although physical reality is presumed to exist, we have no direct evidence of this reality except as it is mediated through socially constructed interpretations. Therefore, reality, understood as meaning, is literally a product of language. Consequently, one cannot seek “true” meanings. One can only elaborate how meanings are produced, foregrounded, marginalized, and shaped by social interaction that leaves its evidence in language.

For example, one might consider the tension between Frankfurt School critical theorists and French poststructuralists. Although the Germans, such as Habermas, conceded a high degree of constructivism, ambiguity, and interpretation to our knowledge of reality, they remain devoted to the search for transcendent meanings that reflect, if imperfectly, reality. We might contrast this with the major works of Foucault, who repeatedly attempted to show how fundamental notions such as insanity, crime, or sexuality changed so profoundly from one cultural epoch to another that there could be nothing about the notions that could be said to have meanings that transcended the situational.

For example, if one were to study representation of women in popular magazines from a feminist perspective, a Habermasian analysis would be directed toward showing how the underlying truth of gender equality is masked and distorted by representation. A more Foucauldian analysis would be epistemologically unable to claim to know an underlying truth; it would be restricted to showing competing representations and analyzing the differing consequences of differing representations. One might then advocate one representation over another, but this must be acknowledged as a value-based act. It cannot be legitimated with reference to objective truth or reality.

Speech Acts and Rules Governing Speech Acts

One can study the structure of a language, or one can study how individuals and groups use a language. Either is a legitimate object of study, but it is important to distinguish between the two. The latter is about what is done in a specific instance. The former is about what can be done in any instance. For instance, suppose one conducted participant research with a group of salespeople with attention to the uses of humor by informants. It is possible that analysis of one's field notes would reveal patterns in the identities of the individuals who were the object of the jokes. A group of unhappily married men might be well stocked with humor reflecting negatively on wives. Another group might consistently denigrate recent immigrants, and so forth. Broadly speaking, there are two different constructs we might use this data to analyze.

On the one hand, this research might provide empirical data about the intentions and values of individuals. This might contribute to answering questions such as “What portion of salespeople in this formerly male-only occupation are prejudiced against female ‘salesmen’?” or “How do people feel about the recent influx of immigrants from Ethnia?” On the other hand, our research might not be about individual perceptions and intentions at all. Suppose a happily married, profeminist man started work with a group of older salesmen who bonded in their free time by telling jokes that reflected negatively on wives and family life. The new guy must choose between being untrue to his values on the one hand and risking being an out-cast on the other. In this case, the misogyny of the work group may not even involve intentionality or self-consciousness on the part of those participating. It is structural. Our research interest might be the process of social construction through which misogynistic jokes are normalized as expected behavior and legitimated as inoffensive. This situation was not produced by the study's participants, although they participate in perpetuating it. This is a structure preexisting, and largely independent of, their speech acts, one that they entered and one that can be studied as an entity in its own right.


Analogously, one might look at the discourse of “productivity and efficiency” in organizations, with its origins in large-scale industrialization and its present contradictions in emerging postindustrial economies. This is a theme researched discursively in Jacques's Manufacturing the Employee, which seeks to discursively analyze the pernicious legacy of the industrial within the postindustrial from a perspective indebted to Foucault. Similarly, one might study masculine bias within management discourse as did Joyce Fletcher by inquiring how the discursive system makes it easier to successfully speak of military and sport images than of more work-relevant feminized images related to nurturing, supporting, and communicating—even when these feminized qualities are precisely what is stated by management as being needed for organizational effectiveness.

The metaphor of text has not been limited to books, articles, and things commonly thought to be texts. Any systematically encoded practices can be read as texts. Verbal speech can be considered text. Street advertising, graffiti, signage, television, architecture, and more can be read. For instance, one might “read” particular forms of music videos, such as gansta rap, in order to determine the structure of that particular language in terms of its symbols and meta-messages structuring what can be said. What can be said, for instance, regarding the appropriate role of women (be a “ho” or a “bitch”) or the meaning of success (“bling,” coercive “gangsta” power). This constitutes a language in the sense that a music video producer can—and must—work with a specific vocabulary of symbols whose meanings are already familiar to speakers of this language: viewers. If these symbols are combined according to the rules of the language, they produce intelligible meanings. If they are not, they either produce unintended meanings or they are simply unintelligible. Even subversion of the rules, ironically, demands observance of the rules.

For instance, the 1998 video Pretty Fly (For a White Guy) by the Offspring, places a clueless, white suburban male at the center of an otherwise-expected language of rap video symbols, the gangstas, the guns, the bling, the women tricked out as bitches and ho's. With just the one symbol subverted—a rap-per who is nerdy rather than cool/threatening— but leaving the language otherwise intact, the video is made humorous.

In contrast, imagine trying to make a rap video encouraging young men to embrace Islam; dress conservatively; treat themselves and women conti-nently; abstain from alcohol, drugs, and guns; and devote time to spiritual study and bettering the community. There is no music video language for this message. There are heavy metal videos, rap videos, and rock chick videos, all with their own dialect of the language, but there is no vocabulary within this language for representing Islam. One can do a video for the rock band Shihad, but spiritual shihad is unrepresentable, except as foolish. To communicate that message, one would have to either fit Islam into the lexicon of music video language (at the expense of the message), or one would have to get that language community to learn another language. This is done, for instance, by outreach programs such as San Francisco's Black Muslim Bakeries, where people are drawn for the food but through this interaction have an opportunity to observe reality interpreted through a different language—the discourse of Islam— within which it is the vocabulary of the music video that is unrepresentable, except as degenerate.

Critical Summary

It is quite important at the present time to question whether the object of discourse analysis (or discursive analysis) is the speech act or the linguistic structure determining which speech acts are meaningful, because they are two different objects, and it is common for authors to fail to specify which they are studying. This potential confusion is compounded by a growing trend to call one's analysis discursive when it is grounded in methods and techniques that used to be called interpretive, ethnographic, narratological, or dramaturgic. Given the very broad set of meanings the term discourse has had, this labeling cannot be termed incorrect, but it is confusing and can be misleading. Within organizational studies, for example, most of the researchers who have worked under these headings have concentrated on the intentionality of individuals or groups, not on the structural conditions underlying those speech acts. Such work has not introduced methodological or theoretical innovations but has simply attached a more fashionable (albeit less informative) name to an established approach. To further complicate the matter, these approaches can be used as a basis for structural analysis, such as that of the anthropologist Claude Lévi-Strauss, so one cannot create a neat correspondence between theoretical position and method.

At present, one can assume only that research claiming to be discursive will, in some way, work with the metaphor of text. Beyond that, one cannot assume any specific methodology, epistemology, or object of analysis, because these are subject to variation. This is a problem only if one assumes discourse to be a discrete theoretical or methodological position, which it is not. Like qualitative research, discourse analysis is a term applied to heterogeneous positions that appear incoherent if lumped together but that constitute several distinct clusters of coherency if studied in their own right.

See also

Constructivism, Deconstruction, Genealogy, Langue and Parôle, Signifier and Signified

Further Readings
  • Berger, P. L., & Luckmann, T. (1967). The social construction of reality. New York: Doubleday.
  • Culler, J. (1986). Ferdinand de Saussure (rev. ed.). Ithaca, NY: Cornell University Press.
  • Eagleton, T. S. (1996). Literary theory: An introduction (2nd ed.). Minneapolis: University of Minnesota Press.
  • Fletcher, J. K. (1999). Disappearing acts: Gender, power and relational practice at work. Cambridge, MA: MIT Press.
  • Grant, D., Hardy, C., Oswick, C., & Putnam, L. (2004). The Sage handbook of organizational discourse. London: Sage.
  • Jacques, R. (1996). Manufacturing the employee: Management knowledge from the 19th to 21st centuries. London: Sage.
  • Jones, C., & Munro, R. (Eds.). (2005). Contemporary organization theory. Oxford, UK: Blackwell.
  • Jones, C., & ten Bos, R. (Eds.). (2007). Philosophy and organization. London: Routledge.
  • Kuhn, T. S. (1977). The essential tension: Selected studies in scientific tradition and change. Kuhn, T. S. (1996). The structure of scientific revolutions. Chicago: University of Chicago Press. (Original work published 1962).
  • Smircich, L. Concepts of culture and organizational analysis. Administrative Science Quarterly 28 : 339-358., 1983.
  • Stager Jacques, Roy
    Copyright © 2010 by SAGE Publications, Inc.

    Related Articles

    Full text Article discourse analysis
    The Penguin Dictionary of Psychology

    In linguistics and related disciplines, the analysis of ‘units’ larger than a sentence; e.g. in writing, an analysis of paragraphs, in speech,...

    Full text Article Discourse analysis
    Reader's Guide to the Social Sciences

    Billig Michael , Ideology and Opinions: Studies in Rhetorical Psychology , London and Newbury Park , California : Sage , 1991 ...

    Full text Article Discourse analysis
    Key Concepts in Journalism Studies

    Literally, the analysis of discourse . More specifically, discourse analysis (DA) is the analysis of what people do with text and talk....

    See more from Credo