< previous page page_31 next page >

Page 31
that at such an elementary level the fabula becomes a matter for box 8, a pure opposition of roles; or for box 10, a world structure.
But, if the text does not tell a consistent fabula, there is another way to approach it from a narrative point of view: the text narrates the various steps of its own construction. It is possible to read a scientific essay this way, as does Greimas (1975), where he analyzes a ''discours non figuratif" such as Dumezil's introduction to Naissance d'Archange. Greimas discovers in this scientific text not only an "organisation discursive" (corresponding to box 4) but also an ''organisation narrative," corresponding to a part of box 6 (as far as narrative functions as loss or victory are identified) and to boxes 9 and 10. The text thus appears as the story of a research, with its temporal steps, the modification imposed on the starting situation by an acting subject (the researcheror Science in person).
0.7.2. Forecasts and Inferential Walks
0.7.2.1.
However, the role of the reader does not consist only in choosing the level of abstraction at which to produce the macropropositions of the fabula. The fabula is not produced once the text has been definitely read: the fabula is the result of a continuous series of abductions made during the course of the reading. Therefore the fabula is always experienced step by step.
Since every step usually involves a change of state and a lapse of time, the reader is led to make an intermediate extensional operation: he considers the various macropropositions as statements about events taking place in a still-bracketed possible world. Each of these statements concerns the way in which a given individual determines or undergoes a certain change of state, and the reader is induced to wonder what could happen at the next step of the story.
To wonder about the next step of a given story means to face a state of disjunction of probabilities.
In fact, such disjunctions occur at every sentence of a narrative step, even within the boundaries of a single sentence: "La Marquise sortit à cinq heures . . ."to do what, to go where? But the condition of a neurotic reader compelled to ask Whom? What? at every occurrence of a transitive verb (even though witnessing a profound affinity between sememic structures and narrative ones) is usually neutralized by the normal reading speed.
The 'relevant' disjunction of probability ought then to take place at the junction of those macropropositions the reader has identified as relevant components of the fabula.
In many cases the right clues to establishing these junctions are given either at the text linear manifestation level (subdivision in chapters and

 
< previous page page_31 next page >