Sunday, November 8, 2009


My November 7th blog has a sentence starting with the peculiar "Poetry are..." Of course I know it should be "Poetry is..." What happened is that I had first written "Poetry and novels are...," but when I decided to comment on poetry and novels separately, I bleeped out the "and novels," but forgot to change are to is. Upon reading the post over before submitting it, my brain automatically read it as "Poetry is..." When someone is reading over one's own words, a common error is to read what one intended and not notice the actual words used. That is why even the most practiced of writers require copy editors.

A related phenomenon is that one needs to hear clearly only about 50% of a spoken message in order to decode it properly. All languages are 50% redundant, the redundancy being built into the sound and grammatical systems, as well as in words. For instance, when we hear bake we know that words like cake, cookies, bread, and pie will follow, but beef will not. The only meats we bake are lobster and ham. The word roast is reserved for other meats and chestnuts.

The reason English has two separate words meaning 'place in a hot oven to cook' is to help listeners narrow down what words could be following. That way, if they miss the actual word when decoding the stream of speech, they can fill it in. Both the general context and the use of either bake or roast help the hearer settle on what was actually baked or roasted. The funny thing is that the hearer is not conscious of the fact that he or she did not really hear the name of the food. The hearer fills in a likely option, thinking he or she really heard the word. One nefarious consequence of this "filling in" of words is that people are often sure that someone said something that he or she never said. Most of you have been accused of saying something you never said. I know I have.

Why would language --and it's all languages, not just English-- be built so that hearers are, in effect, guessing what was said? Well, for speed. It has been shown that any series of sounds that come to our ears as rapidly as language does is heard as buzzing or static. Only language sounds can be decoded when they come at people so rapidly. That means that if there is any noise in the environment, people can still decode the speaker's message because of the redundancy. By "noise," linguists mean not only competing sounds, but such things as inattention or suddenly thinking of something else.

Nobody is aware of the fact that while hearing the stream of sounds in languages one knows, one's ears are segmenting out the individual sounds and the brain is deciding what words have been used by the speaker. When this goes awry, people may notice, however. Have you ever had the experience of not hearing a speaker and just as you are asking, "What did you say?" the words suddenly pop into your head.

The corollary to this is when you hear a language you don't know. It sounds like what it is, a stream of sounds, and you often can't even distinguish what sounds have been uttered, much less what words. In your own language, you think you are hearing pauses between words, but actually, it is your brain that is deciding where one word ends and another begins. Spectographs show that there is no pause between each word spoken.

If enough brains mishear where one word ends and another begins, we can get a language change. For instance, originally in English, cooks wore "a napron." However, enough hearers thought they were hearing "an apron," so apron it became, although the related napkin (literally, 'little napron') retained the n. A more modern example is the number of people who are sure their dogs were spaded, not spayed. When they heard "spayed," they thought the word was spade -- not spay plus -ed

Another example of a decoding glitch is when you think someone said one thing and you heard it as another, typically one with the same kind of sound in it, but not the actual sound made.

For instance, once I was buying large, very ripe peaches at a greengrocer's. He asked me,

"You know what to do with those peaches, don't you?"

I answered, "Yes, I put them in port."

"No," he bellowed,"you put them in wine."

"Port is wine," I said.

"Oh," he responded, "I thought you said 'pork.'"

Port and pork are pronounced alike except for the final consonant in each, and those final sounds, t and k are closely related. That's why some toddlers learning to speak may call a kitty a "titty," much to the embarrasment of their parents. The toddlers haven't yet learned to distinguish between those two sounds.