Book chapter summary - An Introduction to Discourse Analysis, Second edition (chapter 8), by James Paul Gee

What I read

In this chapter, Gee describes the more tactical or detailed ways that speech is used and is shown in a transcript, and how those details set up a structure for analysis, or “ways into a text.”

Gee starts by explaining how speech is produced, and offered to a listener in “small spurts.” An example is provided (that is then expanded on in the chapter,) of a young girl discussing an interaction related to breakfast. He separates words into function words and content words. Content words are critical for understanding, because they cannot be inferred from a spurt of words on their own, while function words (such as the, those, he, it, on, to, of, some, many, all) act as indicators for guessing content words. The function words are described as “informationally less salient.” The content is the material that is new and unpredictable.

Next, Gee focuses on the stresses and intonations found in spoken language, and how these physical and psychological features of speech indicate information saliency and importance. Stress is presented through loudness, length, and pitch on a word or syllable. Gee identifies the very detailed, nuanced and specific way that stress is presented in a brief example. The most unique or salient or new piece of information is emphasized through a “glide” or “pitch movement.” A focus on this glide as an indicator of salience, combined with an understanding of context, helps a hearer judge salience in related aspects of speech.

Gee describes the way text is presented in a transcript, and relates the “chunking” lines to the idea of how the eye gathers and constructs information in the world. Each fixation in the eye, and each spoken chunk, is an “idea unit” and presented as a line (literally, an individual line of text.) The breakfast example is used to show how, in each line, a new piece of content or salient element is introduced. Because each line is frequently too small or short to capture what someone wants to say, a stanza is used to indicate the boundaries of a block of information. He describes that “Each stanza is a group of lines about one important event, happening, or state of affairs at one time or place, or it focuses on a specific character, theme, image, topic, or perspective.”

A macrostructure captures multiple stanzas into a story with sections. Gee expands on the breakfast example, showing how the example contains a setting, catalyst, crisis, evaluation, resolution, and coda. In showing this example, Gee uses “idealized lines” by removing the various non-language elements like hesitations. He notes that the story itself shows “the basic function of narrative: narrative is the way we make deep sense of problems that bother us.”

Another way of thinking about structure in speech is the distinction between a macro and micro line. Macro lines “tie together two or more lines into something akin to a sentence.” This is the relationship of multiple intonation units are connected to each other. The macro lines give clues to an analysis of how to organize and think about the meaning in the micro lines.

Gee ends the chapter by returning, briefly, to the ideas of tools of inquiry and how the detailed analysis methods described related to those tools. This is an organic set of moves between the elements described (lines, stanzas, etc) and the ideas of the text itself (and themes and meaning that are starting to emerge). Those applied meanings are then tied back to the situational meanings and models that are being attributed to the text.

What I learned and what I think

Even though Gee has repeated that he isn’t interested in providing a “how to” set of steps, this is getting closer to one, and I appreciate that it’s shifting from a way of thinking to a way of doing. I see a great deal in the preliminary content in the first 7 chapters. Now, I want to know what to do with that content, because I am impatient and because I want to practice this. I know I learn by doing things, and while the ideas are getting in my brain, the nuance isn’t concrete.

It still isn’t that concrete, either, about what level of fidelity an evaluation like this should work at. I’m back to methods; design methods are a great way to learn how to do something, but they aren’t how the something is actually done by professionals. I need more nuance on the method here. I have questions;

Gee shows, and then hides, the various tics in language when he discusses idealized lines. What are some rules or heuristics for when to analyze with and without those weird parts of speaking?

He has separated the details of the seven tools of inquiry from this discussion. How do they come back and show up during the analysis? (Maybe that’s coming in the next few chapters.)

Is it “ok” to read the micro lines and absorb the macro lines, but analyze at a stanza level? What’s the heuristic for what level of zoom to explore at? I know it’s a “I know it when I see it” for someone who has done this a lot…

The interviewer snippets that are included show not-wonderful interviewing techniques. Do we care, in the analysis? We typically remove those during design interpretation. Should they come back here? Why?

And, what’s the end-game for a robust and many hour long transcription analysis? I’m imagining I’ve just spent 90 minutes on the phone with someone, talking about creativity. For argument, let’s assume I’ve identified and noted my interpretations related to some or all of the building tools at a stanza level. So, I’ve made meaning about that particular person’s statements and thoughts. I ultimately want to make some meaningful observations about a state of affairs in the world of creativity and design. What’s a good representation of how the stanza explorations lead to a more global statement about a participant (do they?), across participants (do they?), and then to a larger conclusion about a given topic (do they?)

I think I need some published paper examples of this approach in action that Paul feels are well-written and well-structured, and that show the value of the approach and also value in the output.

The next three chapters (the last three chapters) are examples, and that will help me, I hope.