Sunday, February 1, 2009

Thinking About Thinker Attributions

Once more you can thank (or perhaps blame) a post by Emma Darwin for impelling me to gnaw on a subject and try to set my thoughts in some sort of order. As usual, Emma’s post touched on a half-dozen interrelated ideas; and, also as usual, my post contains only one idea. Maybe less.

'Thinker attributions’ are the inward equivalent of dialogue tags—that is, ‘she thought’ or ‘he wondered.’ In the Jane Austen era (or in the hands of Patrick O’Brian in our own era), such attributions often followed the conventions of standard dialogue: “Ah-ha, my friend,” he said inwardly, “that caught you off guard…” or “How the devil did she get to the top of that steeple?” he wondered.

Although these usages are still acceptable, many contemporary writers avoid quotation marks around thoughts. Modern readers tend to see quotation marks as stating that something was said aloud, and it can take them a moment to sort out what is being said from what is being thought—especially when inserted in the midst of an exchange of dialogue. Quoted thought becomes even more confusing if a character thinks and speaks in the same paragraph:

“That’s Susan, all right. Who else would wear galoshes with a bikini?” he said. “How the devil did she get to the top of that steeple?” he wondered.

Is that a mess, or what? To be sure, you could make it less clunky by rearranging it, but one of the constraints is that you probably don’t want quotation marks smacking up against one another. It’s simpler by far if you render the thoughts without marks, a more 'modern' style:

“That’s Susan, all right,” he said. “Who else would wear galoshes with a bikini?” How the devil did she get to the top of that steeple? he wondered.


“That’s Susan, all right,” he said. “Who else would wear galoshes with a bikini?” How the devil did she get to the top of that steeple? he wondered.

Ah, I hear you saying, but you don’t even need the thinker attribution. Why not simply:

“That’s Susan, all right,” he said. “Who else would wear galoshes with a bikini?” How the devil did she get to the top of that steeple?

Or even:

“That’s Susan, all right.” How the devil did she get to the top of that steeple? “Who else would wear galoshes with a bikini?” [etc., etc.]

True enough, and many editors and writing teachers now urge the elimination of thinker attributions in almost all situations. Sure, they say, they were needed back in the 19th century, when omniscient narrators slipped from one head to another; but in most fiction nowadays, they can be cut.

Emma’s point (and I couldn’t agree more) was that simply because attributions can be cut doesn’t mean that they always ought to be cut. Her example is wonderfully sly, because it doesn’t even involve a direct thinker tag, but instead nudges you deeper into POV by using “…I looked…” “…I watched…” “…I stared…”, until the moment is vividly, pointedly from the narrator’s eyes. None of these nudges is necessary to relate the story events, at least in the strict sense of information conveyed, but the quality of the reading experience is quite different.

I can think of five occasions when some form of thinker attribution is appropriate, and might even be required.

1. Handling shifts in POV without an explicit break. If you’re going to slither from head to head, there will be a point where you need to anchor the POV in a new consciousness. Sometimes the simplest means is to come right out and say ‘she thought’, but some writers, like Ann Patchett, are so good at piling up little clues that it is hard to point to a moment when you’ve been re-anchored; it’s an accumulation of evidence. (Of course, many of those who counsel you to avoid thinker attributions also tell you never to shift POV except at a break; some will even insist on a new chapter for each POV. Lord knows such an approach is simpler, and will keep the unskilful from hurting themselves, but we're old enough to mess about with fire if we choose.)

2. Modulating psychic distance. I’ve rattled on about psychic distance before. At one end of the spectrum you have narration in the voice of the novel (When the hot Santa Ana winds blow through Los Angeles, tempers turn as brittle and dry as the brush on the hillside…); at the other end, we are so firmly embedded in the mind of the POV character that thoughts can be reported without any question that the belong to the character rather than the novel (Fuck Henry, and fuck this job, and fuck this unending goddamned hot wind.). Explicit thinker attributions (“he thought”) occupy a middle ground in this spectrum. Emma’s indirect “he watched,” etc., clues, push us yet deeper into POV, or re-anchor us there. When dropping down from a large psychic distance to a close one—or even more important, when drawing back from a close psychic distance—thinker attributions smooth the transition, easing the reader effortlessly along the narrative.

3. Displaying thinking in concurrent time. Thinking is often treated in fiction as if it moves faster than the speed of light, and that usually isn’t a narrative problem. But when the time expended pondering is relevant—when the bomb is ticking, or when the angry lover demands a response—then thinking becomes an action, a time-consuming action, and like other important actions, it needs to be described, and often attributed.

4. Improving rhythm, prosody, or emphasis. Even when we are already in such a close psychic distance that there is no doubt we are hearing the character's unmediated thoughts, a thinker attribution can often add a certain emphasis to the parsing of a phrase.

I’ve earned this money and I’m going to keep it.

scans differently than:

I’ve earned this money, she thought, and I’m going to keep it.

and it’s the writer’s job to decide which fits best, whether a thinker attribution is “necessary” or not. The first example is more intimate, but I find the second to be more emphatic; the hesitation in the middle of the thought is more self-conscious, as if sje jhas paused to cross her arms defiantly across her chest.

5. Doing it because it feels right. Like the centipede who tripped when he thought about coordinating all those legs a-walking, a writer is (or ought to be) doing several things at once. When psychic distance is great a thinker attribution draws us closer, when we are in close a thinker attribution subtly pushes us away, and when we are in a tangle of action or summary, an attribution can make the events more vivid and personal. Attributions can introduce a nuanced pause, or increase the drama of a ticking clock, heighten hesitation, or confirm that we have slid from one point of view to another. And, to be frank, sometimes it just sounds better. Consciously or unconsciously (or, in my case, in my typical state of semiconsciousness) the writer needs to be balancing all of these needs to achieve the desired total effect, and it isn’t the sort of thing that’s amenable to rules.

There's my five. I may have missed some.

Writers need to know that thinker attributions can often be dropped. But knowing they are optional does not amount to a rule that they should be eliminated whenever the sense of the narrative can be preserved without them.

Since when is delivering information in the minimum number of characters the goal anyway? he wondered. What is this, R-U-OK text messaging?


Tim Stretton said...

Yes indeedy!

Sorry to have no more cogent response. Your observations are entirely to the point.

Aliya Whiteley said...

I'm experimenting with omniscient narration at the mo. It's freeing, and why wouldn't a modern reader buy it? What are they, morons?

Tim Stretton said...

I've never understood why omniscient narration gets such a bad rap. It's fashion rather than any kind of technical unwieldiness...

David Isaak said...

Hi, Aliya--Omniscient is great if you have the right story. The last novel I completed was in faux-omnisicent, where a possibly omniscient narrator kept shoving the POVs out of the way and rambling on.

To be sure, some modern readers are morons. But most readers don't notice anything about technique in the first place, and don't have any opinion about it in the second place.

Have fun!

David Isaak said...

Hey, Tim--

Yeah, it's fashion--and the bad rap omniscient gets is almost entirely from writing teachers and other pontificators. Ann Patchett, Ian McEwan, and Larry McMurtry all knock out the occasional novel in omni, and they seem to do just fine.

I do think the technical bar is set a little higher on omni; modulating and re-anchoring POV is something I've seen many newbie screw up terribly.

Then again, an unskillful writer can screw up anything, can't they?

Janet said...

#5 kind of summed up my thinking on it, but I do tend to be less analytical than you about these things.

Whenever I challenge any small element of my writing, I ask myself which version of the sentence best accomplishes what I want that sentence to do, not which one best follows the rules. The rules are useful, but they should be servants, not masters. But we've always seen eye to eye on this point anyway.

David Isaak said...

Hi, Janet--

Yeah, #5 pretty much covers it.

I'm actually a rather intuitive writer when writing. But later I feel a need to sort out what I think. It's the academic in me, I suppose...

Anonymous said...

David, we're obviously on the same wavelength - but we knew that anyway. Great post anyway - love your breakdown of all the excellent reasons for this kind of thing. I'm interested in your riff of my post, since it originated in my thinking not about the attributions at all, but about what I rather pretentiously call the choreography of gaze.

And since I wrote that post I've started reading a very rare beast: a lit crit book which actually says all sorts of things to me as a writer, Dorrit Cohn's Transparent Minds: Narrative Modes for Presenting Consciousness in Fiction. Yes, I know, it makes your list of Favourite Titles seem positively snappy, but it IS fascinating. Cohn's subject is how fiction presents the operation of its characters' consciousness, from Tom Jones to Leopold Bloom. One of his points is that thinker attributions are a flag that the narrator is telling/explaining what someone's thinking. The other modes are quoting the thoughts directly (with or without ""s) or narrating the thoughts in free indirect style(a silent form of reported speech, as it were). (Sorry, it's sophisticated stuff and doesn't summarise easily). His point is that this telling/explaining (he calls is psycho-narration, but it's also omniscient or unlimited third-person), which is apparently more distant from the character, actually is the only mode which has the flexibility to express most fully a character's consciousness, because it can encompass things which they wouldn't have words for (c.f. child Maisie in What Maisie Knew) or which are beyond or beneath what they can put into words. Once a character puts it into words it's no longer the experience, it's their take on the experience. But an omniscient narrator CAN put it into words, for us, while the character's consciousness, as it were, goes undisturbed. He quotes Schiller: "When the Soul speaks, alas, it is no longer the Soul."

Maybe an omniscient narrator is the only observer - pace Heisenberg - who actually can observe and measure things without changing them...

David Isaak said...

Hi, Emma--

Cohn's title is actually rather good--if you leave off the subtitle! In truth, I have nothing against academic books per se--it just bothers me that most of them are overwritten in academese to hide the poverty of ideas within.

This sounds like a good book. Perhaps I'll hunt it up.

The ideas about omniscience remind me a little of Wayne Booth's discussion of "Emma" in "The Rhetoric of Fiction," where he shows that book would be lacking an entire dimension if told wholly subjectively--that what makes it work so well is the interplay between Emma's lack of self-awareness and Austen's omniscient narration. (Of course, no Austen fan needs to be told that at least half the pleasure of any of her books is her narrative voice.)

What I like best about what you summarize from Cohn, though, is that he's made a practical point about craft--that thinker attributions are a 'marker' to let the reader know how the consciousness of the book is being modulated. Depending on the vector, they let the reader know we are getting closer or further from full character subjectivity.

As a writer, that's news you can use--but usually we have to figure it out on our own. Whether or not writing can be taught, a good teacher could at least point out examples, from a craft perspective, of how things like this are accomplished.

Come to think of it, the field is too crowded already. Skip that idea.

I like your notion of the "choreography of gaze." Like it so much, in fact, that I'm going to respond to it in another post.

Anonymous said...

Cohn writes reasonably well, too - it's academic prose, but he only uses jargon when he has to, and he can construct a reasonable sentence, so he stays below the bafflement threshold...

He also makes a very clear distinction between first and third person narratives, which of course is fundamental to what we get up to, but an extraordinary majority of critics working in this area - including an otherwise favourite of mine, Meike Bal's Narratology - completely ignore.

It's just occurred to me that the bit of Cohn I was summarising above chimes very well with John Gardner being rude about third-person-limited narratives. The point about Emma is an excellent one, which backs up the value of this kind of narrative.

I'll look forward to your take on the choreography of gaze.

"Depending on the vector, they let the reader know we are getting closer or further from full character subjectivity."

On this one, in my PhD I've extended Gardner's Psychic Distance to talk about the Psychic Range of a narrative - the distance from the farthest away to nearest into the characters heads it goes. Very useful way of thinking about it, I find.

David Isaak said...

Hi, Emma--

I ordered Cohn's book off Amazon yesterday. (CLearly an academic book: costs too much.) That's the first time I've ever heard the name "Dorrit" outside Dickens.

You're right that most of the critics seem to ignore the real machinery of the whole thing. They'd be read more widely if they had something intelligent ot say about craft.

Your dissertation sounds as if it hacking deep into unexplored territory. Will it be published, or do you have thoughts of revamping it into a book?

Anonymous said...

Yes, it's beastly expensive, I'm afraid.

The PhD is practice-based, so A Secret Alchemy is 70% of it, and the rest is a critical commentary on the writing of it, with excursions into the history of the genre and other hist fic with parallel narratives - questions of voice, structure, turning historical fact into fiction, and so on. Of itself it wouldn't make a book, because so much is perforce about the writing of the particulary novel, but it has been extrememly thought provoking, not least to make me sort out what I really think about all sorts of writing things I burble on about in forums and on my blog, and turn it into respectable academic prose complete with references. There is a book I want to write about writing some day, and I'm sure that'll be a major part of it, but at the moment there's a novel (or two, or three) that I want to write more.