I recently bought and read Samuel R. Delany's The Jewel-Hinged Jaw for the first time, since I wasn't aware it had been reprinted by Wesleyan (way back in 2009) until Patrick Nielsen Hayden mentioned it to me at Viable Paradise this year. Those of you who are longtime readers will know that I have a great deal of respect for Mr. Delany's critical work, even when I don't agree with him, and this remains true of most of the essays in The Jewel-Hinged Jaw. That said, I feel the initial essay, "About 5,750 Words" (written in 1968) establishes a premise upon which Delany constructs much of his critical apparatus (such as his criticism of the opening of The Dispossessed), and I feel that this premise - in light of variations in how people read and recent discoveries about human memory - is fundamentally flawed.
I'm going to skip past Delany's equation of form with content, which I feel is at least partly based on the claims he makes about reading. (The short version of that argument is that there is no separation between style and content in fiction, because the change of a single word in a novel can be "all-important".) Delany argues that:
A story is not a replacement of one set of words by another-- plot-synopsis, detailed recounting, or analysis. The story is what happens in the reader's mind as his eyes move from the first word to the second, the second to the third, and so on to the end of the tale. (Ibid., p. 4)
This might seem reasonable until you realize that Delany means it entirely literally - that he is claiming the transition from "The" to "The red" to "The red sun" is actually a meaningful and significant part of the story process:
[Words] sit in numerous inter- and overweaving relations. The process as we move our eyes from word to word is corrective and revisionary rather than progressive. Each new word revises the complex picture we had a moment before. (Ibid., p. 4)
This claim, and the micro-focused narration of Delany's 2.5 page word-by-word response to reading the sentence "The red sun is high, the blue low," followed by the bald assertion that "Though it ordinarily takes a quarter of a second, and is largely unconscious, this is the process," made me stop and gape in disbelief. Because that is not the process as I understand it, nor is such an understanding of human literacy conducive to a philosophy of writing that properly accounts for human variation and frailties.
I am now going to descend to argumentation by anecdote, which while less than perfectly scientific, is a practice Delany has never flinched from himself. To quote from a later essay in the book, "Reading The Dispossessed" - 'This did happen to me. And it doesn't happen that way.'
When I was younger, I used to read much faster than I do now. I could (and regularly did) read 4-5 paperback books of around 300 pages each in a single day. In general (though of course there were exceptions) I didn't "skim" - instead, I took in multiple sentences and even whole paragraphs at once, digested them, then moved on to the next paragraph or set of sentences. I was
reading holographically, not parsing each sentence a single word at a time - and I know from the comments of my previous post on this topic that I am not alone in having (had) this capability.
Even now, when I read at a much slower pace than I did when I was a teenager, I don't parse things letter by letter (or even one word at a time). In fact, there's cognitive science research indicating that most literate adults don't parse words letter by letter but recognize them (again) holistically, taking in the whole word, recognizing it, and moving on,
even if the letters between the first and last letters are scrambled. And like I mentioned previously, while I don't generally take in whole paragraphs at a time any more, I still don't parse sentences word by word unless they are awkwardly or very precisely phrased. The smallest unit of signification that I actually parse out isn't the word but the phrase, which is a far less granular mode of reading that Delany describes.
My objection to Delany's claim that "The story is what happens in the reader's mind as his eyes move from the first word to the second, the second to the third, and so on..." should hopefully be clear by now. Depending on one's critical views, 'the story' may well be what happens in the reader's mind as they read a text. But given the continuum of ways in which people read, I am entirely unconvinced that the author has as much control over what will be received as Delany seems to imagine. The faster/more carelessly a reader reads, the higher the probability that words/packets will get dropped. And the more packets are dropped, the less likely it is that "the complex picture we had a moment before" is the same complex picture, and that each new word (or phrase, or paragraph) is revising it in the same way as for other readers. (Barthes's notion of "The Death of the Author" seems highly relevant here.)
I am not arguing that literary style can be divorced from content in SF, as the editors Delany was inveighing against in "About 5,750 Words" were. (The two versions of a single sentence that Delany quotes on page 9 of the Wesleyan edition are argument enough that style and cadence and word choice can matter a great deal - and that additional words do not inevitably improve anything.)
I am arguing that it is quite easy to waste vast amounts of time looking for exactly the right word or phrase to express the precise shade of nuance you meant to express, and then have most or all of your readers completely overlook what you meant to do because either, A) the mode of reading they've adopted resulted in them missing or glossing over the word in question, or B) the difference in reading experience between the last few versions of the word/phrase/sentence being minuscule to the point of negligibility.
One further point. At the risk of beating a dead horse until it is a smear on the pavement, Delany proposes in "About 5,750 Words" that "A sixty-thousand word novel is one picture amended fifty-nine thousand, nine hundred and ninety-nine times," (Ibid., p. 5) I have (I hope) already established that it is not in fact a single picture which is received by readers, but a multiplicity of pictures which share certain common elements. In order to completely overthrow this notion of endless emendations, however, let us consider the question of human memory. The claim that a novel is a picture amended on a word-by-word basis contains, implicitly, the notion that the reader can retain a clear memory of both the start point and either all the emendations or their cumulative effect.
I hope none of you will be surprised when I denounce this as hogwash.
In The Paradox of Choice (which I don't have with me, alas), a study on human memory was cited indicating that people's memories of an event were not based on the entire event, but tended to be dominated by the peak experience of the event, and how the event ended. In the context of the study, this manifested as people having more positive memories of a brief period of intense discomfort followed by a long period of mild discomfort, when compared to the same brief period of intense discomfort without any period of mild discomfort. While interesting in its own right, this study is relevant because it provides evidence that people's memories undergo compression, in a manner similar to a highly detailed image file being shrunk to a lower resolution. Details deemed insignificant are lost, and when a reader goes to decompress their memory, they are likely to interpolate missing information from the details they do remember.
How can the notion of a novel as a single image, amended 59,999 times, survive this compression, this constant readerly condensation and summarizing, in which pages and chapters of description and nuance are rendered down to "Tyrion escaped, killing his father on the way out"?
I submit that it cannot.
I further submit that attempts on the part of an author to control their readers' reactions rather than to channel them in a direction are generally doomed, for the reasons I have laid out above. (When such efforts succeed, they do so by invoking cultural assumptions and patterns that are powerful enough to overwhelm both the text and natural variations in readers' responses. Such extra-textural factors are not eternal, and their expiration often render texts and their original appeals incomprehensible to readers.)
To sum up: When we write fiction, we may mean to convey a piercingly clear vision to readers, but variations in reading methods and protocols (as well as our memories) will conspire to transform an author's vision into a statistical distribution of received images and events; a range of readings that share common elements but may well be orthogonal to what the author thought they were transmitting. We aim for lossless communication, but what fidelity to the original vision we do achieve is usually reached through redundancy and a series of approximations.
I am not claiming that a well-placed, precisely chosen word is worthless. Clearly this is not the case.
I am asserting that to attempt to make every word or sentence of a work of any length sing is a grotesque waste of effort, and that when arrayed against the dispersive forces I have mentioned, any individual word or phrase is a leaf facing down a hurricane.
Further (thanks to
mrissa for this point) if the only way to 'truly appreciate' a particular work is to process it word by delicate word, then such a work will inevitably encounter readers who are neurologically incapable of reading it as it was intended.
As writers, we craft works of such fragility at our peril.