
May 26, 2025 | 7 minute read
Book chapter summary - An Introduction to Discourse Analysis, Second edition (chapter 9), by James Paul Gee
What I read
In this chapter, Gee provides an example of discourse analysis, using examples of teenagers from working class families and from upper-middleclass families.
First, Gee explains the research methodology, which included multiple interviews, along with shadowing the participants. He indicates that he will select two sets of interviews for examination in this section, and will contrast the two. Before focusing on the actual interviews, Gee provides a brief overview of the worldview that has led to his selection of the labels (working class and upper-middle class.) He describes that the focus of the analysis will be on the Identity building task. He recalls the use of “socially situated identities” by briefly returning to the example of a college professor and a middle-school teacher.
Next, he begins the analysis, starting by stating the research hypothesis: that working class teenagers build identities attached to everyday interactions, while upper-middle class teenagers build identities detached from this everyday view and instead focused on a future view.
Gee indicates that a way of exploring this hypothesis is by examining the “I-statements” that the teenagers make in the interviews. These include cognitive, affective, state and action, ability and constraint, and achievement statements. These were selected because they “believed they might be important and interesting.” He indicates that the statements have been quantified, and compared between groups; but before showing his analysis of this data, he indicates that the quant-based information is used “simply to guide us in terms of hypotheses that we can investigate through close scrutiny of the actual details and content.” The results of this numeric data indicates that working-class teenagers make more I-statements related to social and affective experiences, while upper-middle class teenagers focus on knowledge, argumentation, and achievement.
Next, Gee explores these findings in depth through analysis. The working-class teenagers use language that is more transactional and interactional; the others are used for introspection, judging, and assessment. He expands on this through several utterance extracts compared across students in the different groups. One way of analyzing is through looking at action statements. The bulk of one student is focused on achievement activities, while the other is based on everyday activities.
The next part of the analysis explores narratives around social interactions and experiences. The upper-middle class teenagers state their perspectives and make arguments, which Gee indicates are “rhetorically clothing their own very personal interests and concerns.” An example of discussion of racism is provided, where the working-class participant makes overtly racist statements, while the working-class students make more “intellectual” statements that are still just as racist, or hint at racial injustices as being less about institution and more about individuals.
The last part of the analysis looks at how narrative is presented by a participant to provide a theme he is “attempting to instantiate and develop.” The teenager uses the phrase “used to” in order to separate being in a success or achievement context now, from when he was in a judged context before. This is used to illustrate how a teenager tells a story of personal transformation, which then provides a way to participate in a story of the upper-middle class group.
What I learned and what I think
This is the first example of how a discourse analysis plays out that I’ve seen, and to oversimplify it perhaps unfairly, it’s someone thinking deeply about what they saw and heard from research participants, and then summarizing it into a more grandiose (used nicely) view of the world. It is what it says on the label; it’s an analysis of the stories the participants told during their interviews, focused on the content of the stories but also equally on the presentation of the stories through words.
There’s probably a “well, obviously” to this understanding of the method. It’s literally in the name of the method, and it’s what Gee’s been describing over and over, just through different lenses.
Some observations.
First, it’s still unclear to me how to do this, other than what I wrote above – sit and think deeply about things. If that’s the method, that’s absolutely fine with me because it’s what I was taught to do with research data and have tried to do since. I don’t see how that method is acceptable in publishable work, though. A reviewer can’t dispute that “deep thinking” was done, because it’s clear, but they can (and will?) certainly dispute that the work is valuable in any greater sense than the localized thinking. It’s not the old “not a statistically significant sample” or “not controlled experiment” argument, but I could see real and warranted skepticism around conclusions that are drawn.
For example, I know there are narratives being told by upper-middle class teenagers that are contrary, or at least ancillary, to these; one that I know exists over and over is that “I’ve been acting in a way that is expected of me in high school in order to be popular, and I know it, and I present myself that way to trusted adults.” Another is “The pressure to go to succeed by going through advanced forms of education is overwhelming me; I am being pushed around.” These probably didn’t come out of the transcripts, and it’s likely because the researcher didn’t focus on these things during the interviews, because it wasn’t part of the stated hypothesis. It’s a conclusion that maybe emerged because it’s what the researchers were looking for in the first place. So is the point to add color and detail to an already decided assertion? It does that, for sure. But it seems that if you say “this is the way it is,” and then build a research plan to learn about how “this is the way it is,” by asking questions about “this is the way it is,” and then analyzing the data through the lens of “this is the way it is,” then you find out that “this is the way it is.”
Next, it’s unclear how Gee selected Identity as the one of seven building tasks, as compared to any or all of the others. He says, “I want to start with a consideration of building task 3, identities.” He never moves on to others. Is that because it’s just a short snippet of an otherwise much more exhaustive exploration, and this is just the one he picked for the chapter in the book? Or did he pick it for any other reason? Significance is certainly explored and described, but seems to be discounted here. The others don’t come up in the transcripts provided, so maybe they aren’t important and can be excluded just because of their lack of presence?
I also don’t understand the “I’m not juxtaposing things, now I will juxtapose things because it’s valuable here” ideas. He says “I do not focus on two contrasting groups because I think any simple binary distinction exists here… nonetheless, this particular contrast is an important starting place in today’s ‘new capitalist, high-tech, global world.” I’m not sure where that editorializing comes from and why it’s a justification for doing the thing he says he isn’t doing.
Another part that’s confusing to me is the use of quant-based information to make qual-based statements. It’s another of these “I’m not going to do it, now I’m doing it” moments. He says “Now is a good time to make a point about numbers in discourse analysis. The numbers above are not meant to be ‘significant’ in themselves. In fact, discourse analysis, as I have constructed in this book, is not primarily about counting things. We use such numbers simply to guide us in terms of hypotheses that we can investigate through close scrutiny of the actual details and content of the teenager’s talk.” He then goes on to dedicate entire sections to counting things, and to using those to justify conclusions. I actually believe the adding of comments towards a summary is valuable here, but I don’t understand why he explicitly calls it out as being superficial or not the point of the method.
I’m also just continually stuck on content as compared to presentation of content. The analysis focuses (as in the name) on the discourses the participants say about themselves; how they perceive themselves and construct how they want to be perceived or exist. What about how they actually are? The analysis of one of the upper-middle class kids describes that, because the teenager calls out activities like going to Faneuil Hall or doing soccer and tennis and gymnastics, they are telling a certain story about themselves (and that’s true.) But they are also doing those things (or so we can believe, I guess), maybe because they like going places and doing sports. Does that not count, or at least in this method?
Do any or all of the things I’m thinking about matter, if the above point is right – that the point of the method is to think really hard about research data? If that’s true, there aren’t really any rules or even heuristics for the thinking, I would only need a framework of some kind, and the seven building tasks seems as good as any. There are two more examples in the book, and maybe that will help me with some of this weird reflection I’m having on this. And then I need to read something about content and not method.
Download An Introduction to Discourse Analysis, Second edition, by James Paul Gee, here. If you are the author or publisher and don't want your paper shared, please contact me and I will remove it.