Lecture 12: Phonetics ILecture 12: Phonetics I[This page contains the text of overheads used in Lecture 12, plus some notes on additional points discussed. There is no guarantee that these notes are complete, and they are intended to supplement rather than to substitute for regular attendance and active participation in class.]
A. Single sound can have many different spellings
he e believe ie Caesar ae see ee people eo seize ei seas ea amoeba oe key ey machine i
B. Single spelling can correspond to many different sounds
C. Individual speech sounds are often represented by multiple letters
D. Letters in spelling are often not pronounced
Note the contrast between 'resign' and 'resignation'. When the nominalizing suffix '-ation' is added to the word, the '-g-' is pronounced. Therefore, the presence of the 'g' in the simple verb stem is not just an arbitrary quirk of spelling: the spelling reflects a morphological property of the word in English, which it shares with the form 'resignation'. Why is the 'g' sometimes pronounced and sometimes not? Answer: English does not allow [gn] sequences of sounds, except when the [g] and the [n] occur in different syllables. In 'resign' the 'gn' sequence occurs in a single syllable, and therefore the 'g' cannot be pronounced. In 'resignation', on the other hand, the 'g' occurs at the end of the syllable 'sig' and the 'n' occurs at the start of the syllable 'nay', so both can be pronounced.
There are many other examples in which English spelling reflects morphological regularities rather than phonetic regularities. E.g., the highlighted 'c' in electric/electricity reflects the fact that the words are morphologically related, but the 'c' corresponds to a different sound in each case.
E. Individual letters sometimes correspond to two speech sounds
F. What about more consistent languages than English?
Although some other languages show more regular spelling-sound correspondences, they still do not provide us with a universal way of writing sounds.
The International Phonetic Alphabet (IPA) provides a one-to-one mapping from sounds to written symbols, and can be used for any dialect of any language. [It is useful to more than academic linguists: among the groups that make extensive use of the IPA are musicians, who must be able to sing the lyrics of songs written in languages that they do not know, and actors, who regularly use IPA as a guide in learning to imitate unfamiliar voices and accents.]
Basic problem to be solved: in the brains of language users, sentences of language consist of many neatly packaged units -- sounds, words, phrases etc. -- each of which is clearly defined and independent of the next unit. A speaker has to somehow take his/her linguistic thought, which is packaged into neat linguistic units, and cause the same neatly packaged linguistic units to come to be activated in the brain of the listener. However, speech is produced by the motion of various parts of the mouth, which gives rise to complex patterns of vibration in the air, which gives rise to complex patterns of vibration on the listener's ear drum and patterns of nerve firings in the listener's auditory nerve ... and all of this is not neatly packaged into discrete linguistic units. Therefore, the basic task of speech communication is to translate discrete linguistic units into continuous patterns of mouth movement, and then do decode continuous soundwaves in order to retrieve the discrete linguistic units.
How speech sounds are produced
When airflow is wholly or partially blocked by vocal tract, consonants are produced; when airflow is unimpeded by the vocal tract, vowel sounds are produced. Therefore, the consonant/vowel distinction that we are all familiar with is really a very simple articulatory distinction.
Typically air is expelled through the mouth (oral cavity).
[see notes for next lecture for more specifics, and for explanation of the relevance of the words on the right.]