I get this question from people in various disciplines who are familiar with the term discourse and are curious how I’m using it within linguistics and in my particular research. It’s good to have a stock answer, though it changes based on who I’m talking to and what I assume about the person.
Generally, I consider discourse to be the use of language in social interaction to produce and consume ideas that are attached to that language. More specifically, I think there are recognizable patterns of language use that point to shared beliefs or ideas, and that people make use of these patterns in order to encode those beliefs, which may or may not be taken up by a listener, depending on how much shared context there is. When I talk about discourses out there, I refer to both the language patterns in use by a group of people and the ideologies (explicit or implicit) that such language patterns carry with them. Some people in the humanities refer to intertextuality– the concept that one idea/way of speaking is always in reaction to, or in dialogue with, other ideas/ways of speaking. This is an important aspect of discourse– how it gets repeated and taken up by others.
Some people ask me if discourse is oral or written, and I feel that we should be inclusive about the scope because much of the distribution of ideas through language occurs in both written and spoken forms (and today, through the visual more and more). Linguistics is usually more concerned with oral language because it’s considered “natural” when produced spontaneously, and communications, cultural studies, and literary criticism are usually more concerned with written language. However, this dichotomy is not necessary when studying discourse because linguistic social interaction occurs through various modes.
A syntactician once asked a discourse analyst if discourse analysis was psychoanalysis. I understand why he asked this question– because ideologies, beliefs, values– these are all cognitive; however sociologists are interested in the same concepts and find ways to observe the ways in which people demonstrate, enact, and reproduce these cognitive concepts. In a similar way, a discourse analyst observes language use, takes into account the social context, and finds systematic patterns that reveal the connection between language structures, the ideas in our minds, and social behavior. So no, discourse analysis is not psychology, but it does attempt to go beyond descriptive linguistics and ask how language functions in interaction.