Cognitive Lingusistics Through Misheard Lyrics

Originally posted 2008.  Updates printed in bold.

******** 

One of the debates in linguistics is whether there is a grammar embedded in us genetically. One side claims that the marvelously versatile and plastic human brain learns words and speech because that is what the infant is presented with. The brain would adjust itself to some other form of communication equally well - there is little or no brain structure which must be used for language. It happens to be used for language because that is the raw material presented to baby.

The other side, following Chomsky, holds that there are brain structures which are pre-installed for language. The terms Deep Structure, Universal Grammar and Transformational Grammar are used in this context, if you want to go look up the subject more thoroughly. I will note in passing that Deep Structure enjoyed a faddish fame in the 70's (and maybe later, though I wasn't paying attention then), supposedly referring to mythic profundities that underlay music, art, and theater. All the Jungians were very into this, and the Joseph Campbell people fall into this swamp from time to time as well. Deep Structure is actually only a rather banal technical term in linguistics. Steven Pinker calls it a "useful gadget."

The evidence for the Chomskyite view has accumulated over the years, and the current dominant belief is that specific language structures are in the genes. However, the fact that AI seems to be developing grammar based solely on inputs is mucking that up these days. Reply hazy, ask again later, I say.  Linguists are fond of ingenious experiment design to tease out ambiguous evidence, but you can do one experiment in your own head. Contrary to the postmodern view that our expectations hugely affect what we hear, we actually do a lot of hearing, in the sense of the brain dividing long chains of sound up into words, before our expectations get their grubby little hands on meaning. It is true that our expectations have some effect on interpretation, especially at the margins, but these are so far downstream as to be of minor importance.

Human speech has few clear breaks between words. Fluent speakers, especially in informal conversation, run everything together. If you have briefly studied a language and expect to be able to understand it spoken, you find that you cannot pick out a single word of a sentence - you can't even tell where one word leaves off and another begins. Only after much hearing does your brain begin to predigest the sounds into words for you. Then, armed with a clear few words of each sentence, you can begin to interpret based on gesture and context.

This is why receptive understanding of a language one learned when young lasts much longer than speaking that language. Logically, it should be the other way around: we should be able to take a hundred remembered verbs, a hundred adjectives, and a few hundred nouns and make do even decades later. The thousands of words that other speakers might throw at us should be a much harder list to hold in long term memory. But it is not so. Once trained to a language, the brain breaks up the stream of sound into the correct discrete words, even when it has forgotten what those words mean. (Interestingly, this sometimes takes a period of adjustment, as if the mind is searching for the correct language archives, but once found, switches over quickly.)

Misheard lyrics are your at-home evidence. Because singing is unlike speech in cadence, the brain has trouble finding the breaks. It settles for breaking the stream into any words that are close, then sends it on to another part of the brain to figure out the provisional meaning. That pre-meaning-finding part may find a plausible construction, or may have the memory play it back to the decoder brain for a new division. "Scuse me, while I kiss this guy" splits the s from sky and attaches it to the, making it this.  The decoder brain thinks this is fine, because it has found real words to pass on to the pre-meaning-finder. The meaning-finder says "weird, but coherent," and accepts it. Unless we direct even higher parts of the brain to focus attention on the meaning to see if it really makes sense, the provisional meaning lives on in our memory indefinitely.

Note that we do a lot of decoding and storing automatically, completely independent of the meaning of the words. Our expectations of meaning, all our culture and biases have played no part.

Garage bands and psychedelic wannabees did not attempt to sing "Incense and Peppermints" in the 1960's. Why? Because no one could tell what some of the words were. Try it. No peeking. I intentionally did not put up one of the videos that shows the lead singer, as that would provide additional clues that were not present on the radio.

(Ed King, who became the bassist for Lynyrd Skynyrd and wrote "Sweet Home Alabama" is on the far left.)

Even I never got what some of those lines were, and I was a little fanatic about such things. I won't admit to you what I heard then.

When Austin Powers used the song in a movie, he got the words wrong. I have put the incorrect lyrics in italics, with the correct ones in bold. (How do I know which is correct? If you were there in the 60's, you know that "occasions, persuasions" is infinitely more likely than "oh cajun spice, sweats and" You just know. And it's a great example of the brain just making stuff up, sending it along to the memory, and hoping no one will notice that it makes no sense.

Good sense, innocence, cripplin' and kind. mankind
Dead kings, many things I can't define.
Oh Cajun spice, sweats and blushers your mind.
Occasions, persuasions clutter your mind
Incense and peppermints, the color of thyme. time

Who cares what games we choose?
Little to win, but nothing to lose.

Incense and peppermints, meaningless nouns.
Turn on, tune in, turn your eyes around.
Look at your sound yourself, look at yourself,
Yeah, yeah.
Look at yourself, look at yourself,
Yeah, yeah,
Yeah, yeah.

Tune-a by the To divide this cockeyed world in two.
Throw your pride to one side, It's the least you can do.
Beatniks and politics, nothing is new.
A yardstick for lunatics, want one of you one point of view.

Who cares what games we choose?
Little to win, but nothing to lose.