Popular Science Monthly/Volume 79/October 1911/The Sounds of ch and j




THE sounds "ch" and "j," as in "church" and "judge," have been represented, on the one hand, as being composed of two sounds, "t" or "d" with "sh" (in phonetic transcription ts and dz) and on the other as being single sounds (in phonetic transcription? and J). Related to this is the question whether the Italian "c" and "g" before "i" or "e," as in "cio" and "gia," are more like "ch" and "j" or like the two forms of "sh."

Records of these sounds were made with a voice-recording apparatus (Fig. 1). This apparatus consists of a mouthpiece into which the words are spoken. The waves thus formed proceed down a tube to a rubber membrane at the end, making it vibrate. A straw lever is attached to the membrane with which it rises and falls in unison and so makes a record on a surface of smoked paper around a revolving cylinder.

Records of typical pronunciations of "ch" and "j" enable us to settle their nature definitely.

Fig. 1. Recording the Voice. The waves from the mouth proceed down the wide tube to the rubber membrane at the end. The vibrations of the membrane are recorded by a straw lever on a surface of smoked paper around a revolving cylinder.

A record for "Mitchell" is shown in Fig. 2. It begins with very faint vibrations for "m"; thereupon follow strong vibrations for the vowel "i." The straight line after the vowel indicates a complete stoppage of breath, that is, the sound is an occlusive. After the occlusion the line rises at first gently and then strongly, as the result of the stoppage being ended and the air rushing out rapidly; this phenomenon is called an explosion. The sound at this point is therefore an occlusive with an explosion. This is the sound "ch" which is indicated by the spelling "tch" in the word recorded. The word ends with strong vibrations for the vowel "e" and fainter ones for "l."

A record of the word "nut" by the same speaker is shown in Fig. 3. It begins with very faint vibrations for "n" followed by stronger ones for "u." The straight line indicates the occlusion for the "t." The strong sharp rise of the line indicates that the "t" ends with a sharp explosion. This explosion is quite different from the more gradual explosion of "ch" as shown in Fig. 2.

A record of the word "nutshell" by the same speaker is shown in Fig. 4. Very faint vibrations for "n" are followed by stronger ones for "u." The straight line indicating the occlusion of the "t" is followed by a very gradually rising line which remains for a time at quite a distance above the base line. This indicates that the occlusion of the "t" was not followed by an explosion, but by a continuous rush of air. This portion of the record is typical of the records for "sh."

A record of final "ch" in the word "atch" is shown in Fig. 5, a record of initial "ch" in "chew" in Fig. 6. Records of "j" differ from those of "ch" in showing small vibrations during the occlusion and the explosion. These are due to the vibrations of the larynx which are present during "j" and not during "ch." The "j" is said to be "sonant," the "ch" to be "surd."

These records and many others from the same person and from other persons (Americans) show clearly that the sounds "ch" and "j" consist of an occlusion with an explosion following it, that the explosion is more gradual than the explosion for "t" and "d," and that the explosion is of quite a different character from the rush of air during "sh." The conclusion is unavoidable that "ch" and "j" are not compound sounds, but simple occlusives with characteristic explosions.

By coating the tongue with ultramarine just before speaking, "ch" or "j" or by any of the other methods of palatography[1] a record of the contact of the tongue with the palate may be obtained. The regions of contact for "ch" and "j" are found to be larger than those for "t" and "d."

The final conclusion is that "ch" (c) and "j" (j) are to be recognized as individual sounds quite distinct from the compound sounds "tsh" and "dsh."

Records by a native of Potenza, Italy, are shown in Figs. 7 and 8.

The record of "Mariuccia" begins with some very faint vibrations from the "m." These are followed by stronger ones for the "a".

Fig. 4. Record of "Nutshell." The straight line showing the occlusion for "t" is followed by a rising line showing the rush of air for "sh." This is quite different from the curve for "ch" in Fig. 2.
Fig. 7. Record of "Mariuccia." For "ci" there is no complete occlusion but a steady emission of air ending with an explosion. The curve differs from that of "sh" or "ch."
Fig. 8. Record of "Adagio." The curve for "gi" shows an almost complete closure followed by an explosion; it is a modification of the curve for "ci" in Fig. 7.
The sudden jolts in the line are the record of the two flaps of the tongue for the "r"; the small vibrations indicate that the "r" was sonant throughout (that is, that the larynx was in vibration throughout). Then follow the vibrations for "i" and the more open "u." The sudden
Fig. 2. Record of "Mitchell." The record for "ch" shows an occlusion with an explosion of a special form.
Fig. 3. Record of "Nut." The explosion for the "t" is different from that for "ch" in Fig. 2.
Fig. 5. Record of "Atch" showing the Final "ch." The end of the occlusion and the form of the explosion are like those of "ch" in Fig. 2.
Fig. 6. Record of "Chew" showing Initial "ch." The explosion of "ch" is the same as in Fig. 2.

descent of the line after "u" indicates that this sound was cut short by some closure in the mouth, namely, by the tongue action for the sound "c." The line, however, does not remain at zero, but rises gradually; this indicates a steady emission of breath and not a complete closure. The partial closure is finally released and the explosion is registered in the sharp upward movement of the line. The sound "c" thus shows a sharp explosion like that of "t" but an incomplete closure. The closure is much greater than that of "sh" and the emission of air is much smaller (Fig. 4). The Italian soft "c" is therefore not an explosive occlusive like "t" or even like English "ch"; it is not a fricative like "sh"; it might be termed a fricative with an explosion. At any rate it is a distinct sound not existing in English.

The record of "adagio" begins with vibrations for "a." The first downward movement of the line corresponds to the sound "d"; this ends with a strong upward movement due to the explosion. Thereafter follow vibrations for the vowel "a." The nearly straight piece of line with faint vibrations is the first part of the soft "g"; it is almost but not quite an occlusion. It ends by the strong upward movement of the explosion. The record ends with the vowel vibrations of "o." Other records from the same speaker show even less occlusion through the soft "g." We seem justified in concluding that soft "g," which is the sonant corresponding to soft "c" is, at least in this part of Italy, not identical with the English "j."

  1. Scripture, "Elements of Experimental Phonetics," chapter XXI.