Structuralist Poetics (5)

Chapter 9, the longest in the book, is "Poetics of the Novel". As Culler says, structuralists have given more attention to novels than to poetry, in particular to the nouveau roman, which has forced a challenge to traditional ways of reading.

Inevitably, S/Z features strongly in the opening remarks. Culler goes on, in the section headed "Narrative contracts" to look at ways in which readers try to naturalise or recuperate the role of the narrator. Part of reading a traditional novel consists in constructing a narrator, and this is equally true whether there is an explicitly postulated narrator (eg in first person narrations) or not (eg in Balzac). He suggests that Benveniste's view of histoire and discours doesn't entirely match the situation, and that things are only complicated by Barthes inclusion (and perversion) of this idea. He looks at what happens when it's difficult to construct a narrator: who do readers then naturalise the novel? He rejects the idea of limited point of view as an explanation; within and among the individual points of view, there is a selection process taking place. In extreme cases, where it is impossible to construct a coherent narrator, we can assume the narrator is deranged in some way; but do we really gain anything from doing so? Closing remark in this section: "The identification of narrators is an important interpretive strategy, but it cannot itself take one very far".

The next section, "Codes", essentially focuses on Barthes' five codes in S/Z, which Culler finds inadequate. In particular, he says that there is inadequate attention to any narrative code, so he moves on to study of plot structure.

In the section simply headed "Plot", Culler looks at various attempts to develop a theory of plot construction. It's clearly very difficult, and there are twin dangers of being over-reductive (reducing all stories to a small number of plots) and over-descriptive (when you're halfway to retelling the story). Vladimir Propp comes out of the discussion pretty well. In his analysis of folk-tales Propp identified certain key functions rather than actions. So running is an action, but running away is part of a function, which might be described as flight, in which the means of flight is in plot terms irrelevant.

Using Barthes, Culler shows that the reader's construction of a plot is provisional and deferred: we don't know what function an action embodies until later, possibly not until the end of the text. In fact, Culler's argument is that understanding of how plots work must focus on the reader's experience. He doesn't argue, but I will, that this means anything with a plot is likely to scriptible; it would be a very dull story that explained the significance of every action.

This is a good section, although a lot of it is spent in discussing and refuting theories of plot. The next one is headed "Theme and symbol". This is about the way a story expresses a wider, deeper, more basic, more generalised meaning. (I'm using lots of adjectives because they are all problematic.) The main suggestion is that the symbolic meaning is expressed by means of contrasts and oppositions. Back with S/Z, there's the contrast between the heat of the salon and the cold of the garden. The discussion is fine as far as it goes but it's hard to resolve: how do we know when we've found the right level of meaning. (If we conclude from Sarrasine that it's more pleasant to be warm than to be cold, we've gone too far.) And why do we apparently need to find a thematic meaning anyway? (Personally I'm less inclined this way than many readers. It's possibly a slightly autistic trait. Asked what a novel's "about", I'm more likely to talk about the plot and the characters, rather than thematic concerns.)

And the section ends with two classes of work that escape these considerations: allegory, where there is probably one true reading of the theme, and works like Finnegans Wake, which seem to deny the possibility of carrying this kind of meaning.

The last section of this chapter is on "Character", which has been underexamined by structuralists, who have doubts about the concept itself. The introductory comments suggest that the concept of character in pre-20th century novels (many of which are still being written!) doesn't work with pre-novel works or with the nouveau roman. The major part of the discussion is about how people in novels relate to underlying roles - as in Propp again. Without coming to any conclusion, the idea is that readers naturalise fiction by assigning pre-defined roles to the characters. The discussion seems to recognise its own inadquacy, and the chapter (at last!) ends with an acceptance that structuralism "does not offer a full-fledged model of a literary system". I'm beginning to think that it never will, and I think that may be what Culler thinks to. Ostensibly he's suggesting that other theorists can develop such a model, but ...