Unspeakable sentences: Narration and representation in the language of fiction

by Ann Banfield

First published in 1982, this title grew from a series of essays on various aspects of narrative style; the result is a finished product that melds literary theory with linguistic methodology. It is argued that, where linguistic theory intersects with literary theory, it is narrative that provides the crucial ‘experiment’ for deciding between a communication and a non-communication theory of language and, by extension, of literature. Chapters discuss such areas as subjectivity in direct and indirect speech, the absence of the narrator, and the development of narrative style. With a detailed introduction to the subject, this reissue will be of value to students of linguistics and literature with a particular interest in narrative style and linguistic theory.

Genres: Literary Criticism, Philosophy, Linguistics, Writing

340 pages, Hardcover

First published June 17, 1982

Reviews

[Steven · Rating 5 out of 5]

Amazing book. Great companion to Dorrit Cohn's Transparent Minds. Spent a good chunk of last summer (2023) immersed within this book and also rereading Woolf and Joyce to follow some of Banfield's example sentences in their contexts. A primary learning is that there are grammatically correct ways to say things narratively that can't be said (indicated) with direct speech: subjectivity, narrativity, tense (putting something into the past but narrating from the subjectivity of someone experiencing it in the present).

High level, this book is a theoretical argument (to answer critical attacks on her paper that originally presented this theory) about where this subset of language (represented speech and thought) fits into linguistic models. Six long chapters. In Chapter one, Banfield extends Chomsky's generative model to include subjective expressions. Chapter two explores represented speech and thought (free indirect discourse). Chapter three contrasts represented speech with direct speech. Chapter four takes up the counterpoint to represented speech. Chapter 5 engages and refutes the counterexamples raised by critics of her theories. Chapter 6 defines narrative fiction and describes the historical development of its style. (Proust, Joyce, Woolf, Beckett, Lawrence)

Fascinating study and essential for understanding free indirect discourse.

Update 5/26/2024: Some things to re-read in light of Banfield: Column McCann's story "Step We Gaily, On We Go" from Fishing the Sloe-Black River. Edna O'Brien's A Pagan Place. Stanley Elkin's novella "Her Sense of Timing," from Van Gogh's Room at Arles. Jayne Anne Phillip's story "Mamasita" in Black Tickets.

[Jon · Rating 4 out of 5]

In Unspeakable Sentences Banfield notes Structuralist linguistics failed in attempting to situate literary theory on top of a scientific foundation: "Barthes, maintaining that 'a certain science of literature is possible,' quickly adds, 'if it is to exist one day.' [...] Barthes [later] admitted that the project of a literary science on a linguistic model had failed." (See Jameson's very detailed and sympathetic account in The Prison-House of Language ). It seem to me it was generally concluded that this was the wrong direction from which to approach literary theory, but in Unspeakable Sentences Banfield sees it another way, and instead uses Noam Chomsky's generative grammar theory of linguistics to address the subject. Was she successful?

Well, that's a complicated question. Familiarity, if not expertise, with Chomsky's theory would make Banfield's arguments much easier to follow, at least in the first few chapters where she lays the groundwork for the rest. This makes it pretty tough to judge where she might have fallen short—let's just say I'm no linguist, no "Narratologist," nor even a professional academic (though this is not to say I don't find any of this work debatable, such as her reasoning behind the absence of the narrator... considering her argument an old saw fits well here: the absence of proof is not proof of absence). But just like Structuralism still presents a valuable utility belt (or frame of reference if you prefer) for literary theory to work with, Banfield definitely created a few more tools to work with in this book.

So why "unspeakable sentences?" It is of course complicated, but what Banfield is talking about here are sentence structures that make sense in written narrative but not when spoken. The book is full of examples, such as, "Mother and child then – objects of universal veneration, and in this case the mother was famous for her beauty – might be reduced, he pondered, to a purple shadow." (From Woolf, To the Lighthouse). And the lesson in grammar here makes her argument shine.

One goal Bancroft states for this work is to put the spoken and written word on a level a playing field, against the grain of Plato (who favored the spoken word as many know) and Derrida (who favored the written). Peace between the camps: an admirable goal in itself. But has she succeeded? Personally, I am inclined to say yes, because of the argument outlined immediately above, though maybe she succeeded a little too well (if I was a polemicist on this issue I might even see evidence of the superiority of the written word here).

So was mashing of narrative style into grammar a worthwhile project? I think so, regardless of the misgivings I might have with it. My interest is piqued enough to want to read the book Banfield published in early 2019 subtitled Unspeakable Sentences After Unspeakable Sentences. I must admit I am intrigued over how her arguments have aged in the ensuing 37 years.

[Julianne]

Good lord Ann Banfield is good. This book will change how you think about free indirect discourse forever. Really.