phonetics

Phonetics

Phonetics is the branch of LINGUISTICS devoted to the study of the events associated with the production of human speech sounds. By extension, it is also the study of the perception of these sounds, and of their physics. Phonetics is therefore anchored in anatomy, physiology, psychology, and neurology. Phonetics, however, is basically not concerned with meaning, and in that respect it differs from all other branches of linguistics.

Traditionally, phonetics has dealt with the positions and activities of the parts of the human body that produce speech sounds, with the transition from one position to another, and with the qualities and direction of the airstream that is emitted when a person speaks. All of these considerations come under the heading of articulatory phonetics. Left out of account are the speaker's brain, which triggers speech acts, and the listener's brain, which interprets the vocal message. Ideally, phonetics should begin with the study of the encoding of the speech sounds in the speaker's brain, and end with the study of their decoding in the hearer's brain.

The Organs of Speech

First the brain issues a command to the lungs to initiate an airstream. Before this airstream can become speech, however, it must pass through, or by, the larynx, pharynx, tongue, teeth, lips, and nose--all of which can modify the airstream in various ways.

The larynx contains the vocal cords, or vocal lips, which can be closed to stop the advance of the air, and which also can be made to vibrate or made taut. The pharynx is a tube at the very back of the mouth, where the throat begins. It, too, can be made taut--the result being a stage whisper.

The mouth has one all-important movable organ, the tongue, which is rooted in the throat but can move in every direction. The front of the tongue is especially plastic and can be controlled with great precision by the nervous system. To produce vowels, the tongue, by assuming different positions within the mouth, creates a great variety of resonance chambers--all different from each other, like the shapes of different musical instruments. To produce consonants, the tongue collaborates with other parts of the mouth to make partial or complete closures, either forcing the air through a narrow constriction (as in the pronunciation of the th sound of breathe) or stopping its progress altogether (as with the d sound of breed).

For some sounds, like the p in pay, the lips are first closed for a fraction of a second, stopping the progress of the air, then opened, releasing the air. To pronounce the f of file, however, only the lower lip is used--positioned against the upper teeth.

Comparison of the initial sounds of pile, bile, file, vile, and while with that of mile shows that the nose, too, plays an important role in speech. The sounds p, b, f, v, and wh are all released through the lips; for the m of mile, however, a valve in the pharynx channels the airstream into the nose.

Families of Sounds

Sounds can be catalogued according to the place where, and the manner in which, they are produced. Thus the initial sounds in pile, bile, file, vile, while, and mile are all labials--that is, they are formed by the lips. The final sounds in hack, hag, and hang are all velars--formed by pushing the tongue against the back of the roof of the mouth, called the velum. The final sounds in writhe, ride, write are dentals, so called because they are formed by bringing the tongue into contact with the upper teeth. Some of these consonants, such as b and d, are called stops, because the airstream momentarily comes to a complete halt; for others, such as v and th, known as aspirants or fricatives, the airstream flows continuously.

It is more important, however, to understand how sounds are made than to give them labels. Besides, a so-called phonetic transcription--series of letters enclosed in square brackets to indicate pronunciation--can provide no more than a rough visible record of the succession of noises that is human speech. (In this article, parentheses will be used in lieu of brackets.)The International Phonetic Alphabet is the best-known system for phonetic transcription but is losing favor because of the unusual symbols it requires.

Voiced and Unvoiced Sounds

The vocal cords are responsible for producing (h)--called a glottal fricative because the space between the vocal cords is called the glottis--and a few other sounds; but they also account for the voiced/voiceless distinction found between sounds produced by other organs. During production of the labial consonants (b) and (v), for instance, the vocal cords vibrate; for (p) and (f), however, they are inert and silent. The vowels of most languages are voiced; but voiceless vowels do exist, as in whispered speech.

Accent or Stress

In phonetics the term accent, or stress, refers to prominence given to a vowel and the consonants that surround it. Such prominence is achieved by increasing the loudness and/or modulating the pitch of the vowel. Some languages, like Japanese, have only pitch accent; but others, like English, utilize both loudness and pitch.

Exotic Sounds

To the phonetician, no sound is exotic if it is produced by the human speech apparatus and used for communication. To the nonspecialist, however, many of the sounds of other languages seem strange.

The glottal stop, produced by the tight closure and sudden release of the vocal cords, plays a prominent role in many languages, including Arabic, German, and many American Indian languages. Although it is absent from standard English, the sound occurs in various regional or social dialects. In many areas of the United States, and in Cockney English, for example, the glottal stop does duty for (t) in certain words, like battle and bottle.

So-called glottalized consonants--stops that are followed by or pronounced simultaneously with a glottal stop--are found in Georgian (in the Caucasus) and in American Indian languages. In languages spoken over wide areas of Africa, certain other sounds--like (p) and (k), or (b) and (g)--can be pronounced at the same time. These are known as simultaneous articulations.

In Welsh, the spelling ll reflects a voiceless fricative caused by air rushing through a channel formed by the tongue and the sides of the mouth or the area above the tongue. The sound also occurs in American Indian languages and resembles that at the beginning of the English words shin and thin. Scots, Spanish, Russian, and German, among many other languages, have a voiceless velar fricative, found in such words as Scots loch, "lake," and German ach, "oh." A small channel permits air to escape at the point where the sound (k) is normally made.

ACOUSTIC PHONETICS

Ideally, it should be possible to establish a one-to-one correspondence between events occurring during sound production in the human body and records of the events as they are intercepted in the air by delicate equipment. In fact, the sound spectrograph allows investigators to intercept sound waves, analyze them, and print visual images representing the main physical events that make up a given sound.

Formants

Phoneticians now know and can represent the two principal formants, or concentrations of energy, for each vowel. A formant may be low (few cycles per second) or high. Each vowel has many formants, but the two lowest ones--the so-called first and second formants--are essential for its identification. Both the absolute position of each formant and the relative position of the two principal formants to each other depend on the shape of the resonance chamber--that is, on the position of the speech organs at the moment when the vowel is being pronounced and recorded for analysis. Thus (i) has a low first formant and a high second one; {a} has a higher first formant than (i), but a lower second one than {i}; and both formants of (u) are low.

The formant method of characterizing sounds can also be applied to consonants, because they, too, are the correlates of specific physiological events--of energy passing through resonance chambers of specific shapes. The sound spectrograph is also highly useful in studying such features of speech sounds as voice, length, nasalization, and aspiration. A few other features, like pitch (high versus low), require other equipment.

Recent Developments

Once the techniques of working with the spectrograph were perfected and the methods for reading or interpreting the printouts were thoroughly understood, investigators at Haskins Laboratories in New York devised an ingenious method of reversing the entire process. Instead of converting a sound input into a visual image, they simply drew formants and used these as input for a device that converts visual images into sounds. The investigators varied the position of the formants and played the acoustic output to many subjects who were asked to interpret them. The experimenters thus further improved both machinery and theory.

Electromyography

Another branch of present-day experimental phonetics is electromyography, which is a procedure for investigating muscular activity in the body during the production of speech sounds. By introducing cathodes into the skin, more can be learned about how the muscular and nervous systems are used in speaking and, by implication, in hearing and decoding human language. Some investigators, for example, believe that the hearer involuntarily mimics the speaker.

The science of phonetics is now a vital part of the investigation of the physical chemistry of hearing, the diagnosis and correction of speech disorders, the teaching of speech to the deaf, or the study and treatment of aphasia (the partial or complete inability to speak).

Robert Austerlit

PHONOLOGY

To say that a language's phonology involves the deployment of that language's phonetic resources within the framework of its morphology and syntax is virtually tantamount to saying that a language's phonological system cannot be identified with either its phonetic or morphosyntactic system but rather mediates between those systems. This situation can be illustrated by a few English words: mopper, mop, slobber, pop. First, the plural suffix -s is pronounced differently in moppers and mops, like the z of booze in the former but like the s of moose in the latter. Moreover, these differences in pronunciation of the plural -s are not idiosyncratic facts about the words mopper and mop (as, for example, could be claimed for dice as the plural of die), but rather bespeak a pervasive regularity of English. The z pronunciation of -s is the norm for nouns ending in a voiced sound--that is, a sound articulated with concomitant vibration of the vocal cords (see PHONETICS). Similarly, the s pronunciation of -s is the norm for nouns ending in a voiceless sound--a sound made with the vocal cords at rest.

The preceding discussion might suggest that phonology is not necessary, and all that is needed is a correlation between the morphological fact that English has a morpheme, the plural suffix -s, and the phonetic facts that this morpheme has two pronunciations--z following voiced sounds and s following voiceless sounds. Exactly the same z-s pattern, however, is found in two other morphemes in addition to the noun plural -s: for the third-person-singular present-tense suffix -s (slobbers, like moppers; pops, like mops) and the possessive suffix -'s (the pronunciation of mopper's is identical to that of the plural moppers, and likewise the pronunciations of mop's and the plural mops are the same). Thus three morpheme-pronunciation statements now must be formulated--one for each of the three morphemes involved--even though ostensibly the same pattern is in some way involved for each of the three cases. It is in large part situations like these that have led linguists to posit the existence of a phonological level of language organization.

Phonological Rules

Rather than relating the three suffix morphemes of the previous paragraph directly to their pronunciations, they can be said to share an abstract phoneticlike symbolization--a phonological representation--which will arbitrarily be called X. Then one pronunciation statement can be formulated for X--a phonological rule--to the effect that X is pronounced as s following a voiceless sound, but as z following a voiced sound.

The manner in which phonological rules are formulated and the names and symbols used vary considerably from linguistic school to linguistic school and from theory to theory. The phonological element that serves as input to the rule, symbolized by X in the above example, is variably called a morphophoneme, underlying segment, or phoneme. Despite differences of other sorts, all theories of phonology recognize the importance of distinctiveness in the organization and function of sound systems, normally by taking phonemes to be distinct from one another, with non-phonemic differences in sound following from phonemic distinctions. Thus it usually assumed that mob and mop differ distinctively (phonemically) in the difference b = p, while the difference in vowel length follows from that (the pronunciation of o being longer before b than before p).

Segmental and Suprasegmental Phonology

Segmental phonology is the phonology of vowels and consonants; suprasegmental or prosodic phonology involves phenomena such as stress (intensity) and tone (pitch). An accentual pattern involves the deployment of suprasegmentals within a word (for example, the stress differences between the noun insert--with stress on the first syllable--and the verb insert--with stress on the second syllable--), whereas an intonational pattern involves suprasegmentals within the framework of a sentence (for example, all the words in Mary worries Martin are accentually stressed on the first syllable, but the stress in Martin is intonationally most prominent). Because the sentence characteristically constitutes the framework for intonation, and because sentences are fundamentally syntactic constructs, intonation is one phonological phenomenon whose domain goes beyond morphology.

GO TO THE INTERNATIONAL PHONETICS ALPHABET