Such a distinction is largely a false opposition: as if in geographical mapping one were forced to make a choice between latitude and longitude. Both are necessary to locate the features which are being identified: in interviewing, the 'horizontal' narrative or the 'vertical' categories within it. These represent different, essentially complementary, approaches; and there is no one right form of analysis although how interview data have been gathered can point in one direction rather than another. An unstructured interview is usually conducted with narrative or thematic forms of analysis in mind; a more structured interview indicates a categorical analysis.
What exactly do we mean by 'analysis'? A helpful analogy is to think in terms of the kind of nutritional analysis you see on the side panel of a packet of breakfast cereal. This answers the question: of what elements is this made up? which is a kind of categorical analysis. But even at the level of chemical analysis the sum is greater than the parts: the total experience of your cornflakes is not simply an aggregation of the supposedly beneficial ingredients. For one thing, chemicals in combination have a different mode of operation than they do in isolation. And so it is with the categorical or thematic elements of a narrative.
To take the analogy further: some elements, even if they do not loom large, are more important or have greater significance than others. Here, of course, the analogy starts to break down. The 'importance' of a chemical element – effect on health for example – is a relatively objective phenomenon. The importance of an element or strand in an interview is a more subjective evaluation: mainly justifiable in terms of how far it helps us to understand what is being studied.
As noted at the end of the previous chapter the labour of transcription is the beginning of the process of analysis. Nothing takes you so firmly into the detail of the material. Almost unconsciously you start to 'construct' your