Abstract
This paper explores why unreliable narration should be considered as a concept not only applying to single works of fiction, but also to whole series of fiction, and why impersonal (›omniscient‹) narration can also be suspected of unreliability. Some literary genres show a great affinity to unreliable narration. In fantastic literature (in the narrower sense of the term), for instance, the reader’s »hesitation« towards which reality system rules within the fictive world often is due to the narration of an autodiegetic narrator whose credibility is not beyond doubt. Detective stories, in contrast, are usually set in a purely realistic world (in conflict with no other reality system) and typically do not foster any doubts regarding the reliability of their narrators. The only unreliable narrators we frequently meet in most detective stories are suspects who, in second level narrations, tell lies in order to misdirect the detective’s enquiries. Their untruthfulness is usually being uncovered at the end of the story, in the final resolution of the criminalistics riddle (›Whodunnit‹?), as part of the genre-typical ›narrative closure‹.
As the new genre of detective novels emerged at the turn from the 19th to the 20th century, its specific genre conventions got more and more well-established. This made it possible for writers to playfully change some of these readers’ genre expectations – in order to better fulfil others. Agatha Christie, for example, in 1926 dared to undermine the »principle of charity« (Walton) that readers give to the reliability of first person narrators in detective stories – especially when such a narrator shows himself as being a close friend to the detective at work, as it was the case with Sir Arthur Conan Doyle’s famous Dr. Watson, friend to Sherlock Holmes. Christie dared to break this principle by establishing a first-person narrator who, at the end, turns out to be the murderer himself. Thus, she evades the »principle of charity«, but is not being penalised by readers and critics for having broken this one genre convention because she achieves a very astonishing resolution at the end of the case and thus reaches to fulfil another and even more crucial genre convention, that of a surprising ›narrative closure‹, in a very new and satisfying way.
Fantastic literature and detective novels are usually two clearly distinct genres of narrative fiction with partly incommensurate genre conventions. Whereas in fantastic literature (in the narrower sense of the term), two reality systems collide, leaving the reader in uncertainty about which one of the two finally rules within the fictive world, detective novels usually are settled in a ›simply realistic‹ universe. Taking a closer look at a contemporary series of detective fiction, that is, the Dublin stories of Tana French (2007–), I will turn to an example in which the genre convention of ›intraserial coherence‹ provides evidence for the unreliability of the different narrators – whereas with regard only to each single volume of the series, each narrator could be perceived as being completely reliable. As soon as we have several narrators telling stories that take place within the same fictive world, unreliable narration can result from inconsistencies between the statements of the different narrators about what is fictionally true within this universe.
Additionally, the Tana French example is of special interest for narratology because in one of the volumes, an impersonal and seemingly omniscient narrator appears. Omniscient narration is usually being regarded as incompatible with unreliability, but, as Janine Jacke has already shown, in fact is not: Also impersonal narration can mire in contradictions and thus turn out to be unreliable. With regard to Tana French’s novel, I would add that it can also be mistrusted because the utterances of this narration can conflict with those of other narrators in other volumes of the same series. So in the light of serial narration, the old question of whether impersonal narration (or an omniscient narrator) can be unreliable at all should be reconsidered. In the case of narrative seriality, the evidence for ascribing unreliability to one of its alternating narrators need not be found in the particular sequel narrated by her/him but in other sequels narrating about events within the same story world. Once again, narrative unreliability turns out to be a category rather of interpretation than of pure text analysis and description.
Again, Tana French like previously Agatha Christie is not being penalised by readers and critics for having broken this one genre convention of letting her detective stories take place in a purely ›realistic‹ universe because today, genre conventions are merging more and more. Tana French achieves an even more tempting ›narrative tension‹ by keeping her readers in continuous uncertainty about whether a little bit of magic might be possible in the otherwise so quotidian world of her fictive detectives. Thus, the author metafictionally (and, later also overtly) flirts with the genre of »urban fantasy«, practicing a typical postmodern merging of well-established, hitherto distinct popular genres.
Modern Prose As a Multimedia Hypertext (Using the Example of the Table-Talk 1882 Story By Boris Akunin)
