Do Narratives Require Narrators? II: Ontological Gap Arguments
Narrators and the connection to narrative events
We are currently looking at the prospects of narrative positivism, i.e., the view that every narrative requires an at-least-implicit narrator in order to be a narrative at all. In the first part, we looked at the analytic argument, which is based on the idea that this can be known a priori. Here we will look at another argument, usually known as the ontological gap argument or the argument from means of access, although, as will be clear, we think that this is more a family of arguments than a single argument.
The ontological gap argument is a harder argument than the analytic argument to pin down precisely, but it is clear that every version has at least two major stages:
(I) The first major stage proposes that rational understanding of narrative (or, more narrowly, fiction, if that is particularly in view) requires that there be an explanation of how we are able to know the things we are being told in the narrative.
(II) The second major stage proposes that such an explanation requires an implicit narrative agency as part of the narrative (or more narrowly fiction) itself, where this requirement is understood as a presumption, or as the best default assumption, or as part of an inference to the best explanation, or some other such thing.
Exactly how each of these stages are developed can vary from version to version, which is why this should be considered more a family of arguments than a single argument. Andrew Kania, whose discussion gives us the ‘ontological gap’ name for this family of arguments, calls Stage (I) ‘the positive stage’ and Stage (II) ‘the negative stage’; we will stick with just Stage (I) and Stage (II). However, Kania notes several problems with both stages.1
The obvious problem with Stage (I) is that narratives, particularly fictional narratives, can be incoherent in a wide variety of ways, including ways that affect precisely this question. In a fictional story, there is no reason why the narrator alone would be immune to fictional shenanigans. Kania gives an example of a novel that purports to be a diary whose only copy was destroyed in a fire. Likewise, Huckleberry Finn is a long, literarily complex novel narrated by an illiterate boy who is work-shy.2
An example we ourselves would give is the extremely complicated narrative structure of the Cao Xueqin’s classic, A Dream of Red Mansions (also known as The Red Chambers Dream and The Story of a Stone). In this work, we start with a clear statement of the work as being authored as a way of allegorizing some of the authors dreams and illusions about the world. It then, however, attributes the source of the story to a magical Stone and a Fairy; both the Stone and the Fairy are incarnated into the human world. The Stone is incarnated as Jia Baoyu; Baoyu means ‘Precious Jade’, because the boy is born with an engraved jade stone in his mouth. It later turns out that the words engraved on the jade stone are the complete story of what will become A Dream of Red Mansions. After many generations a Taoist monk, known as the Taoist Void, finds an inscription on a large stone that tells the whole story of the Stone’s origin and its incarnation into the human world, the same story on the jade stone; he then gets into an argument with the Stone about whether the story is publishable. The Stone convinces the monk, who copies it out entirely, but (having by this point changed his name to The Passionate Monk), changes the title from The Story of a Stone to Record of the Passionate Monk. However, when the Taoist Void copied the story, he made an effort to verify it, which he did by finding one of the characters in the story, Jia Yucun, who verifies the story, but under conditions that suggest that it may all just be Jia Yucun’s dream. Indeed, there is some reason to think that the meeting between Jia Yucun and the monk is itself just something that Jia Yucun is dreaming. Advising the monk on how to publish it, he recommends that it be given to Cao Xueqin, who rewrites it five times and edits it, re-titling it The Twelve Beauties of Jinling. However, when the Taoist Void copied the story, he made an effort to verify it, which he did by finding one of the characters in the story, Jia Yucun, who verifies the story. (We are in fact simplifying considerably.)
Every part of the tale seems to have a narrator, but it’s quite clear that Cao Xueqin is deliberately muddying the levels of narration (as seen in the fact that the story is apparently eternal, and predestined, and written by Cao Xueqin). This is entirely in line with a work whose most famous comment is that when false is taken as true, true becomes false and when unreal is treated as real, real becomes unreal. The story has narrators, but foils all attempts to use them for the kind of narrative understanding required by Stage (I). Indeed, it does so in a way that is so humorously absurd that it is clearly deliberate: the story has at least five narrators, but all of them are telling the same story, a story in which it is deliberately unclear what is real and what is not, and which is a story that has apparently always existed and is also a dream of one of its characters and is also authored by Cao Xueqin. To the philosopher who thinks that every fictional narrative needs a narrator so that we can understand how we have access to the events of the story, A Dream of Red Mansions gives more narrators than anyone could ever want in a mix in which it is usually impossible to tell for sure who is narrating what, or whether any of the narrators we are explicitly given are actually narrating the story as we actually have it, or how the story being narrated actually relates to the events that are narrated.
Thus there is reason to doubt that rational understanding of narratives requires understanding how we ‘access’ the events in the narrative; there are exemplary narratives in which part of the point is to make unclear how we do.
Kania also notes that it is unclear whether merely identifying a narrator could actually meet the standard that Stage (I) sets. How does the narrator know any of this? How are we connected to the narrator? These are questions that are often simply not addressed in the narratives themselves, and that readers do not always treat as important.
With respect to Stage (II), Kania notes that fictional narratives often do not have the kinds of clear ‘ontological levels’ that it requires. Many narratives are from a narrative perspective that has no clear relation to the story being narrated, as with third-person omniscient narrators who also comment on the narratives. More importantly, contrary to what Stage (II) requires, it seems that no posited narrator can actually do what it requires. We have, on the one hand, an extra-narrative author and audience, and on the other, a bunch of events in the narrative; merely positing another element in the narrative (the narrator) will not bridge the two. It’s all on one side. The gap between what is narrated and those receiving the narrative cannot be bridged by simply complicating the narrative side of the gap.
These concerns seem to apply quite generally to all versions of the ontological gap argument. Thus ontological gap arguments do not seem to establish that narratives require narrators in order to be narratives. If narrative positivism is true, it must be for some other reason.
Andrew Kania, “Against the ubiquity of fictional narrators,” The Journal of Aesthetics and Art Criticism, Vol. 63, No. 1, (Winter 2005) pp. 47-54. It’s perhaps important to note that Kania considers the argument as specifically concerned with fictional narratives. However, we agree with Derek Matravers in Fiction and Narrative (Oxford University Press [Oxford: 2014]) that none of the issues raised are actually specific to fiction rather than relevant to narratives in general.
An example that is commonly given in the literature is Bram Stoker’s Dracula, which purports to be a bunch of different letters, recordings, diary entries, etc., brought into a narrative unity, and, it is commonly said, it is uncertain who is doing it. Dracula is a good example of how complicated ‘the narrator’ of a work can be, but there is really no question of who is the narrator at any given point. Each individual component, of course, is narrated by its fictional author (Jonathan Harker, Dr Seward, etc.), and there are multiple indications throughout the story that the materials overall are gathered and organized by Mina Harker, who is transcribing and typing it all, and the note at the end indicates that Jonathan and Mina are responsible for the publishable form, perhaps with input from the other surviving members of the group. If anything, Dracula seems to be an example that shows why someone might find ontological gap arguments attractive, since the work is explicitly structured by its narrator-complexity and knowing our ‘means of access’ to each event plays a significant role in understanding that event’s role in the narrative.