I was intrigued by the idea presented in Preface To Plato that within the Homeric epics the narrative proceeded paratactically, that is by one scene following another without there necessarily being any causative chain:

But it can fairly be generalized that the saga considered from the standpoint of a later and more sophisticated critique is essentially the record of an event-series, of things happening, never of a system of relations or of causes or of categories or topics (Preface, p.173).

 
Havelock goes on:

[The events] are remembered and frozen into the record as separate disjunct episodes each complete and satisfying in itself, in a series which is joined together paratactically. Action succeeds action in a kind of endless chain. The basic grammatical expression which would symbolize the link would be simply the phrase ‘and next…’ (Preface, p.180).

 
This got me thinking about good old cause and effect. Would it be possible to create meaningful and coherent paratactic narratives where scenes were simply butted up against each other without obvious causation? Probably, but I think what would be happening is that the audience would, in effect, supply their own chains of causation: simply by placing these scenes in order would imply some kind of relationship, causal or otherwise. Seymour Chatman has said:

But the interesting thing is that our minds inveterately seek structure, and they will provide it if necessary. Unless otherwise instructed, readers will tend to assume that even “The King died and the Queen died” presents a causal link, that the king’s death has something to do with the queen’s. We do so in the same sense in which we seek coherence in the visual field, that is, we are inherently disposed to turn raw sensation into perception (Story & Discourse, pp.45-46).

 
Perhaps the best conclusion we can come to is provided by Abbott:

[T]here are narratologists who require a clear causal sequence as an essential defining feature of narrative, though [I am] defining narrative as “the representation of events,” whether bound together by a clear sequence of causation or not. A quest story, for example, can include many events that come after one another without causal connection (first the knight sinks into a bog, the he is set upon by wild rodents, then his pants catch on fire…), yet it would be difficult on that score alone to say that it is not a narrative. Here is an instance where the term narrativity may help. For, if the sense of causation is not a defining feature of a narrative, it is so commonly a feature that we can say that its presence increases narrativity (Cambridge Introduction To Narrative, p.38).

 
Of course this hinges on what he admits is the “disputed term” narrativity. If you use this word it ultimately implies that you accept there is no ‘edge’ to narrative, no place where you could say that at this point it stops being a narrative and becomes something else. Which I do accept.

I think.

References
Abbott, H. Porter (2002) The Cambridge Introduction to Narrative. Cambridge: Cambridge University Press.
Chatman, S. (1978) Story and Discourse. Ithaca, London: Cornell University Press.
Havelock, E. (1963) Preface To Plato. Cambridge, London: Belknap Press.