陆续

In short, when welook at writing as a process, we’re really looking at a complex literateactivity that includes reading and writing, feeling and thinking, speaking and listening, observing and acting.

For text analysis, this means that it takessome effort and discipline to move beyond questions of what things meanto questions of what they do and how they mean. You might find that asyou analyze a text you slide back into the natural stance of an everydayuser, thereby losing your analytical orientation. As your attention focuseson what a text means, especially insofar as that meaning seems unproblematic or mundane, your attention goes past the text to think about thewriter or the ideas. When you notice this happening, step back and remindyourself of the analytic question(s) that you want to answer, remind yourself to attend to the means that the text deploys, and ask how the words ofthe text and their organization are producing the effects you are perceiving.At first, it might be easiest to take an analytic stance to texts that you arenot deeply engaged in or familiar with, texts that you can look at coolly andfrom some distance. However, if you can learn to see even the most familiartexts as strange objects worthy of close analytic attention, you will start tosee the real benefits of text analysis, for you will understand texts in totallynew ways. Breaking out of the natural attitude toward texts and textualpractices is crucial to textual inquiry and analysis.

Conceptual Versus RelationalContent analysis has traditionally relied on two basic methods: conceptual analysis and relational analysis. In conceptual analysis, a concept is selected, coded, and counted for its presence in a text or corpus (set of texts).In relational analysis, the process goes one step further: it identifies a number of concepts and then examines the relationships among them. Alsocalled “concept mapping,” it is especially common in cognitive studies involving the construction of mental models (Carley & Palmquist, 1992). Inconceptual analysis, meaning is assumed to reside in individual concepts,whereas in relational analysis, meaning is understood to derive from the relationship among concepts. The key step in both types of content analysisis the coding of concepts, for it is here that the analyst exercises considerable subjectivity. For example, if one wants to count all references to “children” in a set of texts, should that include infants, offspring, youngsters, teenagers, kids, dependents, and youth, or only some of these terms? The answerto such a question depends on the goals of the research. That is, the codingscheme used in content analysis should be determined based on its abilityto shed light on the question that drives the research, on its ability to givethe study validity; and it should be set forth explicitly, so as to facilitateinterrater reliability.Quantitative Versus QualitativeSince the 1970s, content analysis has tended to combine quantitative(“objective”) and qualitative (“impressionistic”) approaches: “The hallmarkof modern content analysis has become the maintenance of a balance between the objective and implied aspects of textual data” (Roberts, 1989, p.148).