Procedures
The first step is to study the phonetics of the target language. It is not necessary to learn the language, but a modest degree of familiarity with common words is desirable. The ability to make all the sounds and accurately imitate the pronunciation of words is essential. Generally, this will result in the ability to read the language aloud, slowly, with good pronunciation. Of course, being a native speaker of the language is a great advantage, though a native speaker must guard against thinking of his/her own dialect as
preeminent.
The second step is to compile a complete list of phonemes of the target language. For the purposes of adaptation of Cued Speech the following simple definition of a phoneme can be used, though it lacks the rigorousness of formal (and very elaborate) definitions used by phoneticists: "A phoneme is a family of closely related sounds ordinarily thought of by native speakers as only one, and necessary as a distinct group in order to perceive differences in the meanings of words and phrases." For example, the various shades of the short a vowel in the English word fat, from that used by most Englishmen to the flat short a of southern Mississippi are all thought of as "short a," even though one is aware of the differences among them. Whether one uses the sound as made by an Englishman, or that used in the southern United States, the meaning of the word is the same. Thus, all the various shades of the sound of short a belong to the one phoneme, short a, and are the allophones that belong to the family of that phoneme. Cued Speech is a phonemic system. It does not distinguish between allophones within a phoneme, except in special cases that will be explained later. The test for determining whether two specific sounds belong to different phonemes is the existence of a minimal pair, a pair of words with differed meanings that are identical except for the two sounds in question. For example, the existence of fat and fit affirms that short a and short i are different phonemes. Remember that we are considering only sounds, not spelling. Another example: In many languages the two vowel sounds pull and pool are allophones, that is, they can be used interchangeably without changing meaning. In English they are separate phonemes, as is indicated by the different meanings of pull and pool. In English and German, the short i as in fit, the sound of long e, as in feet in English, ie in German (biegen) are separate phonemes. In most languages (Spanish, French) they are allophones, members of a single phoneme ranging in acoustic quality from short i to English long e, but all spelled as i.
A complete list of the phonemes of the target language can usually obtained from a book on the phonetics of the language. However, it necessary also to take note of allophones of various phonemes, in cases which there may be reason for Cued Speech to distinguish between allophones. For example, in the 21 countries in which Spanish is the major language, there is great variation in the pronunciation of the word yo, which means I. In Castilian, the most prestigious dialect of Spanish, used wide in Spain itself and by some speakers in several of the countries of South America (such as Columbia), the usual pronunciation of the consonant is that of yin yes. But, in Argentina and several other South American countries, and even by many speakers throughout Spain, the pronunciation is like s in pleasure and vision. In Puerto Rico, the prevalent pronunciation is like j in Joe, and many Spaniards use this pronunciation also. These three sounds are allophones; they can be used interchangeably without changing the meaning, and without causing
confusion anywhere. There are two reasons for cueing these allophones differently in Spanish. The first is that their use is so much a matter of culture and pride in some countries that parents will want their deaf children to be able to distinguish and use the pronunciation preferred in that country. The second reason is that two of the three allophones (zh and dzh) are the same on the mouth, but the other one (y) is different. Thus, I designed the Spanish adaptation (on the advice of a committee of Spanish-speaking persons from six different countries) to provide for distinction among these three allophones. In most languages it will not be necessary to distinguish among allophones.
After a complete list of phonemes has been assembled, the next step is to group the vowels into groups assigned to the several hand locations, and the consonants according to the eight or so hand configurations. They must be grouped so that no group contains two phonemes that are too similar in appearance on the mouth. Usually, one begins by arranging the phonemes essentially as they are in English, and then making changes as needed. If in the target language the vowel i represents only a single phoneme (not i and ee as in English), the vowel arrangement of Spanish may be a better starting point. When the list of phonemes is complete, and a trial arrangement is in place, three things should be checked. First, each group
of vowels or consonants should contain at least two phonemes (preferably three), so as to follow the basic principle that the reader must use the information seen on the mouth separate the sounds within a group designated by a specific cue. Second, if there are only two phonemes in a group, one of them should not be a low-frequency phoneme, in order to prevent the cue being interpreted as a "sign" for a specific sound. Finally, a series of trial drafts of the arrangement should be made and tested. For example, in the writer's work with Dr. Anna Metlyuk and Dr. Nadezhda Evtchik, of Minsk, the goal was to produce a workable adaptation that accommodated both Standard Russian and Byelorussian. It was necessary to work through four successive trial drafts, over a period of months, to arrive at one that was as nearly satisfactory as possible.
