Text Analysis I: Interpretive Analysis, Narrative Analysis, Performance Analysis, and Conversation Analysis
There is growing interest these days in the analysis of texts. Little wonder. Most of the recoverable information about human thought and behavior in complex societies is naturally occurring text: books, magazines, and newspapers, diaries, property transactions, recipes, correspondence, song lyrics, billboards . . .
And there’s more—lots and lots more. Artifacts (clay pots, houses, toys, clothing, computers, furniture) and images (family photo albums and videos, slasher films, sitcoms) are texts. Behaviors (making beer, laying out a garden, dressing the dead for funerals) and events (church ceremonies, homecoming games, blind dates) are texts. All these things come to us raw, in qualitative form. We can study them raw or we can code them—turn them into variables—and study the relations among the variables. Both approaches produce insight and understanding.
Enormous corpora of texts are available online—all U.S. Supreme Court opinions, the great works of the major world religions, all of Shakespeare’s writings, every surviving ancient Greek and Latin text, 160 years of the New York Times, to name just a few—with more and more added all the time. Ethnographic texts on a culture-area probability sample of 60 societies around the world are available on line from the Human Relations Area Files (more on HRAF in chapter 19 when we get to content analysis).
Scholars of social change have long relied on longitudinal survey data—the Gallup Poll (continually collected data since 1935) and the General Social Survey (almost every year since 1972) are just two of the hundreds of data sets available for time series analysis. But longitudinal qualitative data are also plentiful. For a window on American popular culture, for example, take a look at the themes dealt with in country music and in Superman comics over the years. Or look at sitcoms and product ads from the 1950s and the 2000s.
In the 1950s, for example, Lucille Ball created a furor when she got pregnant and dared to continue making episodes of the I Love Lucy show. Now think about any episode of Two and a Half Men or 30 Rock. Or scan some of the recent episodes of popular soap operas and compare them to episodes from 50 years ago. Today’s sitcoms and soaps contain many more sexual innuendos.
How many more? If you really wanted to measure that, you could code two representative samples of sitcoms and soaps, one from the 1950s and another from the past 10 years, and compare the codes statistically. That’s content analysis.
Interpretivists, on the other hand, might be more interested in understanding the meaning across time of concepts like “flirtation,” ‘‘deceit,’’ ‘‘betrayal,’’ ‘‘sensuality,’’ and ‘‘love,’’ or the narrative mechanisms by which any of these concepts is displayed or responded to by various characters.
Text analysis is for positivists and interpretivists alike and there is no single method for doing it. Some of the traditions of text analysis include: (1) interpretive analysis,
(2) narrative analysis, (3) performance analysis, (4) conversation analysis, (5) schema analysis, (6) grounded theory, (7) content analysis, and (8) analytic induction. I deal with the first four of these methods in this chapter and with the other four in chapter 19. The first four methods rely mostly on the intuition and erudition of the analyst, while the last four make increasing use of computer programs. Whatever your taste in methods— super-qualitative or super-quantitative, super-inductive or super-deductive—there’s something for everyone in text analysis.
Before we get to text analysis, though, I want to tell you about the great tradition of text collection in anthropology.