Is ordinary language spoken in a kind of meter?
I do not mean to suggest that we speak in iambic pentameter, but there seems to be a tendency to break syllables in a sentence into symmetric groups of two and three and to place a stress on single syllable words.
For example, consider, “I went to the store yesterday.” This would be broken into “I went – to the store – yesterday” with stresses on “went” and “store”.
Here is another example. “It is just impossible to do.” This would be broken up as “It is just – impossi – ble to do,” with stresses on “just” and “do.” If, on the other hand, you say, “It’s just impossible to do,” we get “It’s just – impos – sible – to do,” again with stresses on “just” and “do.”
Surely someone somewhere has noticed this and given it further study. Anyone trying to get robots to computers to speak naturally would have to take this into consideration. Did anyone here watch the television show Jeopardy when IBM’s Watson was on? How natural was the computer’s speech?