ABSTRACT
The phonologies of different languages are in many respects very similar, to the extent that some features appear to be part of every language. These cross-linguistic similarities are in large measure due to the ergonomics of the speech process. Speakers, and languages, prefer distinctions that are easy to perceive and easy to produce. The difference between [t] and [n] is very clear to the perceiver and not too difficult to make for the producer. The contact made by the tongue tip and rims with the upper gums is the same, while the velum is lowered for [n], opening up the nose at the back, and raised for [t], trapping the air behind the oral closure. In the openvelum position, very little effort is needed to bring the vocal folds somewhat closer together than during breathing in order for them to start vibrating. This is so easy that voicing during sonorant consonants and vowels has been called ‘spontaneous vibration’. By contrast, if we block the egressive flow by closing off both the mouth and the nose, it will take some effort to create a sufficiently powerful air stream through the glottis to make them vibrate at all: we are pumping more air into a small, closed pocket of air. Speakers would be well advised therefore not to be too eager to vibrate their vocal folds while their vocal tract is significantly obstructed, as maintaining an air pressure difference across the glottis takes some effort. (You can increase the pressure from the lungs, or make more room by pulling up the velum for [d, ] or blowing out your cheeks for [b].) That is why almost all languages have a dental/alveolar [t] and [n], but only 64% have the voiced counterpart of [t] ([d]), and less than 0.5% the voiceless counterpart of [n] ([ nunderring]).1 It would be extremely improbable to find a language that had no [t, n], but did have [d, nunderring]. Another speaker interest is to duplicate contrasts. If a language has the vowels [i, e, a, o, u] and you find it has a nasalized [o˜], you can bet it also has [a˜], and probably there will also be [e˜]. The speaker’s phonetic routine of nasalization is exploited so as to maximize contrasts with the same velum lowering gesture. But here too the hearer is not forgotten. For high vowels like [i, u], nasalization does not have a whole lot of acoustic effect, and many languages, like French, therefore leave them out of their subsystem of nasal vowels (system ‘gaps’).
