Inês Morais

[Versão em português / Portuguese translation]


Ignorance, coupled with lack of charity, is the source of all prejudices.

In 2010 I published in the journal Disputatio a review of Peter Kivy’s book The Performance of Reading (1). The book promised a revisionary account of the activity of reading literature, one that clashes with current common-sense views. It is prudent to listen to common-sense and learn from tradition, but, at times, widespread, even ingrained beliefs are just falsehoods. Entire communities can be, and sometimes are, mistaken, deluded, for instance when they misinterpret the evidence or when they miss or ignore some detail that matters. And often, mistaken views breed more mistaken views, so it is important that philosophy uncovers and corrects mistakes. Kivy’s book aimed at countering what he considered to be an error in common-sense views of literature reading. At the time, I claimed that Kivy’s arguments against common-sense were not compelling enough for us to abandon common-sense. I was wrong.

My assessment of Kivy’s book was incorrect. I have re-read the book and thought again and more liberally about it – and I radically changed my mind. I now think that not only Kivy’s general view is persuasive, but that it also sheds light on at least one issue in the ontology of literature and in ontology generally that the common-sense attitude overlooks. Here I hope to correct my mistake and do justice to the book, suggesting also a productive employment for its main theory in the ontology of literature (and music) and in ontology in general. 

Kivy begins by considering the history of literature and he notes that literature has been, for the majority of its past, a performing art: “all fictional literature […] has been, for almost all of its history, a history of literature as performance” (p. 18). He opens his history with the Homeric poems, “the oldest texts in the western literary canon that are more or less widely read” (p. 7). He emphasises that the poems were performed to an audience, making them closer to music. It is important to note that Kivy endorses, as is well-known, a Platonic theory of musical works: the works are universals or types; the performances are instances or tokens (p. 2). “[A] version of the work” (p. 3) will be a token of the type. Kivy’s adopted ontology of music is relevant because he will suggest that literature, too – and not just dramatic literature but also modern novels – will be amenable to a similar ontological treatment. His main question is stated on page 4: “It would appear that the novel is a type. But what are its tokens?”

Kivy dismisses the thought that the copies of the novel could be its tokens, just like musical scores are not, for him, the tokens of musical works. He then considers the example of dramatic works: “The tokens of the types Hamlet and Ghosts are their performances” (p. 4). His main thesis will be that “the novel is a reading art, and so it trivially follows that the tokens of the type Pride and Prejudice are its readings” (ibid.). He takes ‘reading’ to be “an act or an activity: it is an action performed by a reader” (p. 5). His main thesis in the book will be that the readings are, themselves, performances, so that “the ontology of literary works is the type/token ontology of musical and dramatic works” (ibid.). That is, just like the tokens of musical and dramatic works are performances, so are the tokens of read literary works.

Kivy’s view is supported by historic considerations. The history of literature, Kivy explains, has been a history or works performed. Silent reading is a fairly modern phenomenon, also because in ancient times texts were written without separations between words (scriptura continua), which made it necessary to read them aloud to facilitate understanding (p. 16). Kivy notes that the first recorded mention of silent reading is in Saint Augustine’s Confessions (4th century AD), interestingly when the author alludes to his conversion: “I had put down the book containing Paul’s Epistles. I seized it and opened it, and in silence I read the first passage on which my eyes fell” (p. 15). Because Augustine emphasises that the reading was made in silence, this suggests that the practice was not common, Kivy concludes. 

Kivy’s proposal is thus to extend the work/performance ontology, available to musical and dramatic works, to silently read works, that is, the works of modern literature such as novels. Reading silently is “a performance of a different kind, a silent performance, but clearly recognizable as a performance” (p. 18). Therefore the history of literary reading can be seen as a continuum, from reading aloud to silent reading, that is, from performing aloud to performing silently to oneself. And the novel can be seen in a continuum, too, with dramatic works, both falling into the category of performing art forms, like music.

One difficulty with Kivy’s proposal might be that, unlike musical works, in which the appreciation of a performance can be neatly separated from the appreciation of the work, with silently read works this separation is more difficult to make. Whereas in music it makes sense to say that we like a work but not a particular performance of it, it is not plausible to say of a literary work that we do not like a certain silent performance/reading of a work we like: it only makes sense to refer to the work itself as an object of appreciation and to its (semantic) interpretations. Kivy seems to always see a neat distinction between work and (silent) performance as objects of appreciation: “there are two objects of appreciation here: the literary work and the reading performance of it, just as, in a musical performance, there is work appreciation and performance appreciation” (p. 21). I think here Kivy may be pushing the analogy a little too far. In my view, it is reasonable to see the reading as a form of performance in the sense that the reading is an interpretation, a ‘translation,’ a perspective of the work, enacted by the reader. The reader is a performer in the sense that he ‘enacts’ the work. However, I believe it is unnatural to see the silent performance as a focus for appreciation in the same way as a musical performance is appreciated. In the literary case, all we judge, when we judge a reading of a work, is the interpretation (the view) it offers of the work, not the action of reading itself. In the musical and dramatic cases, we judge the action of enacting, of performing, the work.

More fruitful, in my opinion, is Kivy’s contemplation of the reading/performance more simply as a version of the work: “A performance is a version of the work performed” (p. 61). In this sense, a performance displays an understanding, a view, an interpretation of the work. And, more significantly for ontology, this examination of the reading/performance as a version of the work leaves open the possibility of a revisionary conception of the type/token distinction itself. Let me explain by pointing to what I find difficult in the distinction, when we think of works of literature. Kivy begins with the assumption that “the novel is a type” (p. 4), to then make the suggestion that the readings/performances are the tokens. But what is this type, then? What is the novel? One way to answer is by saying that the novel is, or exists in, every (read) version of it. However, it seems plausible to say that a novel exists even when it is not read (imagine a novel that no-one will ever read). The novel exists, even though it is never ‘performed,’ in Kivy’s sense. Nonetheless, the author’s reading of his own work, while writing it, is already one version, so all novels have at least one token (reading). But another answer is perhaps more illuminating. We can say that the work exists only in its instances/versions, in the sense that the versions are all we have of the work. The work, the type, is only an abstraction, or perhaps it translates a manner of speaking that is not clearly separable from talk about tokens. We read the work when we read one of its versions, the work exists in its versions. Perhaps literature is teaching us that we do not have to always distinguish precisely between tokens and types.

The analogy Kivy finally proposes is that “of a silent reading of a novel and the silent reading of a musical score” (p. 63). For Kivy the performance “is a story telling in the mind’s ear” (ibid.). Again, the analogy with music is clear. Just like performances are ‘versions’ of works, each reading of a novel enacts a version of it. Kivy offers an eloquent example from music, that of Brahms’s Variations on a Theme by Haydn, which (he explains) exists in two versions, 52a, for two pianos, and 52b, arranged for orchestra. Kivy elucidates that these are “different versions of the same work” (p. 76). The example is enlightening, because each version (52a and 52b) can be seen as a type (of which each performance will be a token) or as a token (of the type Variations on a Theme by Haydn). So, again, the demarcation between type and token seems at least imprecise. Kivy’s view invites us to question a distinction from general ontology.

Thanks to Peter Kivy for his book. It presents a concise and convincing history and analysis of literary reading, from its most famous inception with the Homeric poems to the modern novel, a relatively new art form which no longer exhibits the auditory, more musical, features of the literature of the days of old, but which preserves (or so Kivy argues) its essentially performative nature.

(1) Morais, Inês. The Performance of Reading: An Essay in the Philosophy of Literature, by Peter Kivy. Disputatio – International Journal of Philosophy, Vol.III, no. 28, May 2010. Available at