Lydia Davis, “Form as Response to Doubt,” talk given at New Langdon Arts, San Francisco, 20 November 1986

Doubt, uneasiness, dissatisfaction with writing or with existing forms may result in the formal integration of these doubts by the creation of new forms, forms that in one way or another exceed or surpass our expectations. Whereas repeating old forms implies a lack of desire or compulsion, or a refusal, to entertain doubt or feel dissatisfaction.

To work deliberately in the form of the fragment can be seen as stopping or appearing to stop a work closer, in the process, to what Blanchot would call the origin of writing, the centre rather than the sphere. It may be seen as a formal integration, an integration into the form itself, of a question about the process of writing.

It can be seen as a response to the philosophical problem of seeing the written thing replace the subject of the writing. If we catch only a little of our subject, or only badly, clumsily, incoherently, perhaps we have not destroyed it. We have written about it, written it and allowed it to live on at the same time, allowed it to live on in our ellipses, our silences.

Doesn’t the unfinished work tend to throw our attention onto the work as artifact, or the work as process, rather than the work as conveyer of meaning, of message? Does this add to the pleasure or the interest of the text?

Any interruption, either of our expectations or of the smooth surface of the work itself — by breaking it off, confusing it or leaving it actually unfinished — foregrounds the work as artifact, as object, rather than as invisible purveyor of meaning, emotion, atmosphere. constant interruption, fragmentation, also keeps returning the reader not only to the real world but to a consciousness of his or her own mind at work.

Here is Maurice Blanchot on Joseph Joubert: ‘What he was seeking — this source of writing, this space in which to write, this light to circumscribe in space — …made him unfit for all ordinary literary work…’ — or, as Joubert said of himself, ‘unsuited to continuous discourse’ — ‘preferring the centre to the sphere, sacrificing results to the discovery of their conditions, and writing not in order to add one book to another but to take command of the point from which it seemed to him all books issued…’.

We can’t think of fragment without thinking of whole. The word fragment implies the word whole. A fragment would seem to be a part of a whole, a broken-off part of a whole. Does it also imply, as with other broken-off pieces, that enough of them would make a whole, or remake some original whole, some ideal whole? Fragment, as in ruin, may also imply something left behind from a past original whole. In the case of Friedrich Hölderlin’s fragments, the only parts showing of a madman’s poems, the rest of which are hidden somewhere in his mind; or the only parts showing of a logical whole whose logic is unavailable to us, fragments that seem fragments only to us, and seem to him to make a whole — for there is only a thin line between what is so new to us that it changes our way of thinking and seeing and what is so new to us that we can’t recognize it as a coherent thought or piece of writing, i.e., can’t see the connections the author sees or even sense that they are there. Or fragments that seem to him to make a whole and to us eventually, also, to make a whole, though from a different angle.

Or, as with Stéphane Mallarmé’s fragmentary poems for his dead son, A Tomb for Anatole, the fragment is something left from some projected whole, some future whole, i.e., these are fragments destined one day to be pieced together with other elements to make a whole; or they are the fragments of ideal poems shattered by grief; fragments comparable to the incoherent utterances of voiced grief: inarticulateness being in this case the most credible expression of grief. No more than a fragment could be uttered, so overwhelming was the unuttered whole. In the silences, the grief is alive.

Roland Barthes justifies his own early choice of the fragment as form by saying that ‘incoherence is preferable to a distorting order.’ In the case of Mallarmé, inarticulateness might seem preferable to articulateness when it comes to expressing a grief that is unutterable. Mallarmé failed to transcend his grief; he remained inside it, and the ‘notes,’ too, remain inside it. They become the most immediate expression, the closest mirroring, of the writer’s emotion at the inspiring subject, the writer’s stutter, and the reader, witnessing the writer’s stutter, is witness not only to his grief, but also to his process, to the workings of his mind, to his mind, closer to what we might think as the origins of his writing [via].