My paper on genre and data selection

I posted the handout of my paper earlier on. It deals with the methodological issue of data selection, which is never as interesting as discussing findings, but nevertheless I got some very good comments (both good/positive and good/constructive - thanks to the people who commented).

The main issue is this: I have collected 6 different magazines, 3 copies of each. 18 volumes in total. I have to select which texts to analyse from each magazine, and the sample from each magazine has to be comparable across the magazines. What better way to do it than select specific genres to analyse, genres present in all magazines such as the editorial, the 'agony aunt', or the advertisements?

Well I thought that was too restrictive. We all can tell which text the editorial is (it even says so on the top of the page) - but there are texts in the magazines which don't fall into a 'folk categorisation' because people simply don't care enough to stick them in a category. Does this mean that nobody will ever analyse those? Or should I as an analyst decide what genres they belong to? Or should I use a category other than 'genre' for my classification?

Well I ended up with the four broad 'genre categories' or 'function categories' I mention in the handout:
- Promotional
- Advice
- Commentary
- Gossip

(This idea is the outcome of a conversation with Costas Gabrielatos, thanks Costa!)

These categories may consist of different genres. E.g. the 'advice' category includes the 'agony aunt', interviews with doctors, nutritionists or cosmetologists giving advice on health, nutrition or skincare, small bulleted lists, long texts etc.

Comments I got related to my presentation:
1. I neglect form
2. I shouldn't be neglecting form, because it determines the way we see and treat texts (and consequently plays a role in the effect of texts, which is what we are interested in ultimately)
3. Should I be saying that only 'promotion' and 'advice' texts perform a 'directive' speech act? Are (not) all texts in lifestyle magazines 'directive'?

One of my points has been that I am not interested in form but in function, and this is my criterion of categorisation. The question then arises: should I be neglecting form so much? (This was both a comment from the conference and from my supervision meetings before that). To this I have a practical answer and some impractical further questions:

First of all, I am not denying that form is not important for determining genre membership. But my purpose was not a taxonomy of the genres within magazines per se, but to find a way to construct a body of texts which make sense to analyse. In some way. And if I was going to select as a dataset of texts which should share function AND form, this would be too restrictive.

BUT this does not mean that because I ended up with a convenient categorisation (although this is open to further questioning, as in Comment 3 above) I should (appear to) say that in principle genres or forms don't matter, or any such folly. And I probably I SHOULD actually end up classifying the texts within the 'function categories' in genres, although now I am at liberty to include in my anylisis text of unidentified/unidentifiable genre identification as long as I have identified their broad function.

So I'm saying that at this stage what is more important is the fact that the texts in the categories share similarities in terms of function (advice) and the textual devices through which this is expressed, than the variation in form.

But still. Is it? This leads to point 2: If forms determine the way we see texts and the effect of texts on us, have I been missing the wood for the tree? If I want my data selection to be well-motivated, and if I am interested in ideology, shouldn't I be concerned with putting together texts with the same form, belonging to the same genre and consequently having the same kind of effect? (Maybe I am misinterpreting the point of this comment, need to check but for the time being I will pretend I know what I am talking about).

I have to finish this post now but I will continue it soon.