The case of the literary genre

A post which results from chasing the hare which started with music at reference 1.

There are lots of things out in world which we try to analyse, to classify. So animals analyse into species. Chemicals analyse into elements. Ill health analyses into disease. The economic activity of people analyses into occupation, sometimes glossed as social class. The criminal activity of people analyses into statistics of crime. One might attempt to sort all the features of a once glaciated, mountainous area into a short list of typical glacial features, things like horns, glacial troughs, hanging valleys, alps, corries and moraines. All analyses on which a lot of thought and ink has been spent. And certainly in this last case, plenty of features are not going to fit any list very well, be it ever so clever.

Figure 1

Indeed, only in the first two of these examples, are the groups into which we analyse, the species and the elements, reasonably clear cut, with Lavoisier’s effort from the middle of the eighteenth century snapped above. Otherwise, it is much more a matter of taste and judgement, with plenty of grey areas.

In the case of elements, clear cut but close: from the carbon atom with its six protons, one gets to that of the very different nitrogen by adding just one proton, two neutrons (usually) and one electron; or from the gold atom with its seventy nine protons, one gets to that of the very different mercury by adding just one proton, no neutrons (usually) and one electron.  While there is rather more in the way of genetic material distinguishing, for example, a lion from a tiger. That is to say, more absolutely – but who is to say what the story is relatively speaking?

Some people try to analyse feelings and emotions into something reasonably discrete and definite, try to produce a definitive list of same, often mapped onto a circular disc or a square involving the two dimensions of valence – that is to say good or bad – and arousal. Things like affection, anger, hunger, joy, pain and shame.

While here we hark back to the time when people bought books from nearby shops, owned and run by booksellers - and we think of literary genres. A lot of what follows is also applicable to libraries – public or other – and librarians, but our focus here is on those bookshops. Modest bookshops, not like the huge Foyles shop, formerly of Charing Cross Road, where things might well have been rather different.

We assert that the issues raised here read across to other domains, inform those other domains.

We argue from the corner opposite to that of reference 3, which argued in favour of cross-sectional or multi-dimensional analysis of consciousness. We see no inconsistency here, it all being a matter of what one is trying to do.

Figure 2

In what follows, what was the real world of bookshops has been condensed into a simple story to make a point. We do not think that the this simplification destroys the points being made.

So we have books coming from publishers, in through the back, into the stockroom. From there they are moved out into front of shop. Customers come in through the front, find the book they want, pay and leave the same way as they came in. So, much like any other shop.

But how do we organise things front of shop so that our customers can find what they want?
Books and bookcases

Figure 3

All books are printed by publishers. Publishers organise their offering into series, an obscure example of which might be the many volumes of the ‘Oxford History of Oriental Textiles’. Each book has an author and there is usually a large number – thousands – of copies of each book printed. We put aside books which have editors and more than one author and authors who figure in more than one series. We put aside editions and printings.

Figure 4

Despite the complications suggested in the figure above, we suppose that booksellers organise front of shop into cases, with cases made up of shelves. We further suppose that all the cases can be considered to be along the lines of that on the far left in the figure above and that the books in a case are ordered from left to right, from top to bottom. As the word is on the page, at least in all European languages.

Many booksellers, as a matter of experience, wisdom and habit, found it convenient to classify books first into type, that is to say fiction or non-fiction, and then into one of around twenty genres – ‘Historical Romance’ or ‘World War II’ to take an example from each type. Any one genre is either fiction or non-fiction. We suppose that genres occupy the whole of a number of adjacent cases (a case genre) or the whole of a number of adjacent shelves within a case (a shelf genre). In either case, that number may be one.

Figure 5

Books on the shelf, within a genre, are organised alphabetically first by author, then by title. Something like the non-fiction case in the figure above – a case in a bookshop located in an area where a lot of people are interested in the ancient world, by which is usually meant ancient Greece, ancient Rome and their various colonies and empires. A Eurocentric view of ‘ancient’.

Around twenty seems to be a number that works. Customers can look at a list of them by the front entrance and without fuss work out in which genre they are likely to find what it is that they want and where that genre is to be found in the bookshop.

We note that some modern bookshops, perhaps Waterstones, have moved away from this a bit. Most fiction (dimension=type) is kept in just the one genre, in one big alphabetical series, with just a few other fiction genres, like children’s books (dimension=audience) and large print (dimension=format), kept separately.

Twenty genres

Now if one was a literary theorist one might well regard the genres as multi-dimensional, with the first dimension being new/second hand/antiquarian, the second dimension being fiction/non-fiction and with the candidates for the other dimensions being things like best seller status, format, length, place of composition, place of setting, price band, publisher, subject matter, target audience, time of composition and time of setting  – a lot of which dimensions are going to be categorical and unordered, not at all like temperature or age, which can be treated simply as numbers. Note that best seller status is also time dependent, varies with time – a complication which we put aside.

But this is not going to do in a bookshop, where what is wanted is a straightforward, short list of genres, without bringing dimensions into it. Some books might be reasonably assigned to more than one genre, although in practise, a book is assigned to just one genre in a bookshop and customers have to make the best of it. The list is likely to be rather eclectic, with some focussed on one dimension and some on another. There is merit in allowing list to vary from one bookshop to another, to take account of varying circumstances – although there is also merit in all bookshops using the same genres in order to reduce customer confusion.

So we suppose that experience, wisdom and habit have resulted in a list of around fifteen fiction genres which generally work in the book trade and which can be whittled down a bit in any particular bookshop. Perhaps something like: Children, Classic, Collectible, Contemporary, Detective, Drama & Poetry, Experimental, Fantasy, Horror, Large print, Mystery, Romance, Science fiction, Thriller, War and Western. With collectible only being applicable in a bookshop which did second hand. It might be appropriate to raise very popular authors – people like Agatha Christie, Ian Fleming, Elinor Glyn, J. K. Rowling or Simenon, at least in their day, now past – to being genres in their own right.

Customers are familiar with this sort of list, often displayed at the entrance, and quickly find out in which part of the bookshop they should be looking.

Stocking up

The book trade then looks about for information about their market, their intended customers. Information which is available by area, information which predicts the buying habits in particular areas and which can be used by individual booksellers in choosing their stock, in deciding how much space to give to the various genres. In choosing what sort of a shop they want to be, in striking a balance between stocking a good, broad range of books, so that people can feel they are going to a good bookshop, perhaps stocking books which interest the bookseller and certainly stocking what people will actually want to buy.

They find that age, sex and social class predicts demand by genre pretty well. More precisely that each combination of age, sex and social class has a distinct buying profile, a buying distribution by genre. The combination of age, sex and social class predict genre. And as luck would have it, the decennial census provides statistics about age, sex, social class and area.

Variations in the structure of the population will then drive the sort of bookshops which one will find in any particular area.

So for the book trade, the population is four dimensional. The four dimensions of age, sex , social class and area are enough to do their business. Four dimensions which strike a useful balance, a workable balance between the availability of data and the relevance of data to the job at hand.

A conclusion which depends on their being significant variations in that structure across the country. The interest would be much reduced if that structure was pretty uniform.

Note that the choice of dimensions for population is driven by the job at hand, by the requirement at hand. While there may well be choices which fit a lot of requirements, it is unlikely that one size is going to fit all. We associate to the dictum of computer system designers that design must always be driven by requirement – noting that this was a dictum that we have had trouble with in the past, asserting that good requirements are sensitive to the possible, to the design. A two way stretch.

Note also that coming up with the list of genres is something of a black art.

Statistical footnote

Figure 6a

Figure 6b

As it happened, we failed to turn up a suitable analysis of population by age, sex, social class and area on the Internet. The sort of analysis which we imagined would have been included in one of the printed volumes of, for example, the 1971 Census of the UK. As far as we have been able to ascertain, facsimiles of these volumes are not available online and while a lot of data is available online, a fair amount of work is involving in finding it, downloading it and turning it something presentable in the present context.

Furthermore, the social classes still used in conversation – variations on upper, middle and working classes – have been replaced for statistical purposes by something more complicated based on occupation, with this last further complicated by the growth of self employment, in which occupation is not as neatly categorised and certainly not as well known as it is for most people in employment.

The rather limited analysis presented in the two figures above is drawn from CASWEB, a data service under the UK Data Service umbrella, to be found at reference 5.

Figure 7

While the incomplete analysis above was turned up much more quickly and easily from the US Census Bureau, with the down side that some of the occupational groups are more industry than occupation, let alone social class.

Conclusions

From all of which we draw two conclusions.

First, a list of fiction genres which is useful in the book trade is going to be a rather eclectic list, drawing on a number of the dimensions of analysis that someone of more theoretical turn of mind might be interested in. It is going to be soft in the sense that one person may well prefer one list to another, not like the hard list of the elements of chemistry, about which, there is no argument at all, at least not in the lower reaches. People are not going to agree on which genres some books belong to. The list is not going to be ordered, for who is to say which genre should come first and which should come last? And it is certainly not going to be much like temperature, when one can say that this temperature lies between those two others, that this temperature is greater than that temperature.

Indeed, it seems likely that books are not going to be the only population for which there is a need for some simple classification which supports the needs of its users – with the theoretical purity of that classification being unimportant. Such users might be unimpressed by, uninterested in all the meta’s – as in words like meta-self-consciousness – which crop up in one of the dimensions of reference 2, noticed at reference 3.

Second, the number of dimensions we want to ascribe to some population or other depends on our interest in that population. While some populations will clearly be well described, for most purposes, by some widely supported, small number of dimensions – for a lot of populations this will not be the case. It will all depend on the point of view, on the job in hand.

References

Reference 1: What music makes us feel: at least 13 dimensions organize subjective experiences associated with music across different cultures – Alan S. Cowen, Xia Fang, Disa Sauter, and Dacher Keltner – 2020.

Reference 2: Four-Dimensional Graded Consciousness – Jakub Jonkisz, Michał Wierzchoń and Marek Binder – 2017.

Reference 3: https://psmv4.blogspot.com/2020/03/classifying-consciousness-again.html.

Reference 4: https://en.wikipedia.org/wiki/List_of_writing_genres. Writing genres more generally, including things like memoirs, non-fiction and newspapers.

Reference 5: https://www.ukdataservice.ac.uk/.