There should be a one-to-one correspondence between a speech sound and the symbol used to represent it. A symbol should always represent the same sound, regardless of the language being transcribed. A sound should always be represented by the same symbol.
Since there are many more speech sounds than there are letters of the Roman alphabet, many new symbols had to borrowed or created. Most were created by modifying the letters of the Roman alphabet in some way, for example, by using small capitals as in [, , ], turning the letter backwards or upsidedown as in [, , ], or by extending the tail of the letter or adding some other flourish, as in [, , ]. In some cases, the association resorted to less familiar letters from some of the Roman-based alphabets of Europe, e.g., [, , , ], or letters of the Greek alphabet, e.g., [, , ]. In only a few cases were entirely new symbols like [] or [] created.
The number of speech sounds that the human vocal tract can produce is vastly greater than the number of unique symbols that any group would want to invent or any person would want to memorize. In order to solve this problem, the IPA includes a number of diacritics, small accent marks that can be placed above, below, or next to a major symbol to indicate that the sound it represents in this case is slightly different from the sound it normally represents. Some examples:
A page from the "Maître
phonétique", the journal that the International
Phonetic Association used to
publish entirely in its phonetic alphabet.
We'll be using IPA for the rest of this course.