Atomic Nerds

This article dispenses good advice to writers, which is to ignore all advice from well-meaning writers to go out and listen to the way people talk and then do your best to reproduce it in your dialogue. He goes over the basics of why this is a bad idea- when people talk, they add all sorts of verbal tics as they try to keep the pace of talking up with the train of thought, which tends to jump all over the place rather than staying nicely on track like an actual train- and reproducing this in the writing would drive the reader crazy. You can see why this is so more or less instantly whenever you read a transcript of an interview set down by a particularly faithful transcriptionist; all the ums and ahs and half-finished thoughts that cut off and start over will drive most readers halfway around the bend before they finish.

What the article doesn’t mention is that there are actually a couple of neurological reasons* for this. For one, the brain processes language and all other noise in two completely different fashions; language lives mainly in the left brain, noise mainly on the right. (Music processing has some degree of overlap with language, but also has a lot of its own neural circuitry.) The brain’s language-processing is heavily geared to picking out meaningful information and wiping out nonsensical or confusing part, but this filter doesn’t work when you read. Therefore, listening to someone mildly prone to umming and ahhing and repeating themselves or repeating certain phrases may not even be noticeable at all, whereas a completely accurate transcription of the same dialogue would be absolutely maddening to read. You can also see relics of this effect in reading literature old enough that it was meant to be spoken or sung rather than read; the constant repetition of key phrases like “grey-eyed Athena” in The Odyssey is irritating to the reader, but goes over much better when read aloud, as it’s just another helpful marker for the brain to pick up on to make following the story easier for the listener.

Speaking of music, you can also see this language/noise neurological effect simply by picking up a new musical genre. Ever been told by your parents- or told your kids- some variant on “That’s not music, that’s noise”? That’s because, neurologically, you’re processing it differently. I couldn’t understand a goddamn word in a Green Day, Nirvana, Pearl Jam, or any death metal song until I picked up the lyrics sheet and read along with the music- and gained not only the ability to understand what that singer was singing specifically in any subsequent song by them, but sometimes the ability to parse the entire genre if the style was specific enough, as it is with metal’s “death grunt” lyrics. Once my brain had ANY foothold to find language patterns in the “noise”, it happened automatically.

Television and movies, since they depend on a rapid-back-and-forth of dialogue between speakers, are mostly scripted as though they were to be read much as written dialogue is- constructed to minimize the amount of verbal “noise” the viewer has to filter. It is necessarily less complex (which is why so many quotes from film come across as shallow and sound-bitey compared to similar quotes from books and other writings), but follows all the same rules. This phenomenon is probably why I can’t stand talk radio or televised interviews- all the verbal tics of normal speech are intact, which my brain reads as “noise”, so that they sound largely incoherent to me. This probably wouldn’t be true if I spent more time listening to them and the filter kicked back in properly, but since I don’t expect it from those formats, I don’t filter it correctly.

The second neurological twist is that, due to the different mechanisms of picking out and parsing information when read rather than spoken, it’s possible for a reader to digest much more complex or awkward sentences and phrasing than it is for a listener. The mind has a short-term memory “buffer” in which words and phrases are kept while reading through a sentence; punctuation marks like the semicolon I just used, or the commas bracketing this clause, are there to keep the buffer relatively unburdened. If I stripped all the punctuation out, you’d have to read the sentence over several tiimes to figure out which words went with which chunks of sentence and thus what exactly the actual meaning of the sentence is. Pauses, gestures, and vocal emphasis help fulfill this function in speech, but they’re easier to misinterpret- especially if the speaker is reading what he’s saying as he says it rather than speaking naturally. If a speaker gets even a fraction as complex as a writer can expect to be able to, he’ll exceed his audience’s memory buffer and leave them wondering what the hell he just said, since they can’t go back and read it over.

If you’ve watched any political debates- which ones and which years don’t matter much except to give some clue as to the respective candidates’ styles- you’ll notice the exaggerated gestures on the part of the candidates. They’re not trying out for a Karate Kid sequel, they’re trying to insert visual punctuation into their answers. Alternatively, over-emphasis works as well. (McCain and Obama favor the karate-chops, Palin likes to put UPtilt on her words for EMphasis, as though she were often asking a QUEStion.)

Speaking of the debates, despite the fact that Palin technically spoke in a more complex fashion (though I’ll note that the same differences between the way we process spoken and written language mean this distinction is utterly meaningless), I had a much easier time following her than I did Biden. Reading the transcript, I have a much easier time following Biden’s answers than I did at the time- he was overflowing my memory buffer too often, whereas the same repetitive sentence structure that helped me follow Palin while listening now jumps out as irritating to me.

If you read the texts of famously effective speeches- like Reagan’s Wall speech or the Gettysburg Address- it becomes easy to see that any given clause, denoted by the punctuation in the textual version, lends itself extremely readily to a lengthy pause on the part of the speaker and doesn’t depend much on previous clauses to parse. In the Gettysburg Address especially, they’re very nearly complete sentences or understandable sentence fragments in and of themselves.

Good speakers know not to speak like they write, and good writers know not to write like they speak.

*Recommended reading: Mind Hacks. Go ahead and buy the book if the blog tickles your fancy.

Posted Science: we'll fuck you up, current events |