Prosody (suprasegmental features)

      So far we have been talking about phonetic features as they apply to single phonetic segments, or phones.  Phonetic features can also apply to a string of several sounds, such as a syllable, or an entire word or utterance. The study of phonological features which apply to groups larger than the single segment, such as the syllable or the word, are known as suprasegmental features. The study of these features is known as prosody.

      The most obvious prosodic feature in language is the syllable. Let's briefly discuss the notion of syllables.  Like all of our other basic linguistic concepts, although everyone knows what a syllable is, the concept "syllable" is difficult to define in absolute terms.

Articulatory definition--a syllable depends on the obstruction of the vocal tract during speech, the nucleus of a syllable may be defined as the point when the airstream is least obstructed--the sonorant peak. /go over which sounds are least sonorant to most sonorant/ 

Acoustic definition.  The nucleus of a syllable is formed by a sonorant peak.  Syllables are groups of sounds that cluster around a sonorant peak;  the peak is usually a vowel, although liquids can also function as a sonorant peak:  Czech and Slovak contain syllabic r/ l.  Glides (or semivowels) differ from real vowels in that they are non-syllabic; they cannot function as a sonorant peak.

Physiological definition.  A syllable corresponds to a chest pulse (main body of air released).  Sometimes two vowels run together, difficult to determine what constitutes a chest pulse:  realtor/ rilter.  (compare to three-fifths, neighbor).  Also, not the words few, new.

Psychological definition.  A syllable is naturally recognized by the mind as a unit.  We speak with syllables, not sounds.  The first phonetic writing systems were syllabaries, each symbol stood for a consonant + a vowel.  Syllables are evident in language errors as well, cf. slips of the tongue usually involve an unconscious and involuntary exchange of sounds from one syllable to the next, such as these Spoonerismsa malt whiskey-- a walt miskeysons of toil = tons of soil.

      However we define the syllable, the sound segments of any language can be perceived by the ear as being roughly divisible into syllables.  This universal is probably physiologically based:  since all humans breathe out while talking for the most part, more air escapes during certain points in the speech act than at others.   

      Although all languages have some sort of syllabicity, the phonetic characteristics of syllables differ across languages.  Languages differ in just what types of sounds can cluster together around a single sonorant peak, or--if you prefer-- around a single syllabic nucleus. 

      In some languages only vowels are syllablic, in others liquids can be.

      Also, each language has specific rules about what consonants or combinations of consonants can begin or end a syllable.

      In some languages, syllables are always open, that is, they always end in a vowel, never a consonant.  (Hawaiian)  Liliukalani, Kamehameha.  On the other hand, every Hawaiian syllable must begin with a consonant. (Aloha spoken as a single word begins in a glottal stop.)

      In other languages, syllables are always closed; they must end in a consonant (Navaho):  Háá'ishah dididiljah.  Let's build a fireTáá diné 'ooljéé'go naaskai'  Three men went to the moon.  (Like Hawaiian, they must also begin in a consonant.)

      Some languages allow a great variety of syllabic structures. In English, for instance, a syllable may consist of only a vowel, or of consonant cluster + vowel + consonant cluster, the most extreme example being the word strengths.  (also, fifths, strengths, streams.)

      Some languages permit consonant clusters that are even more complex than in English: 

      Russian permits complex clusters at the beginning of syllables: vstretit'  to meet ;  but not at the end of words:   Russian syllables are always open except at the end of words:  nash vs. nostraticheskij. Cf. English nos-trat-ic.  A fill vowel is added to break up a cluster at the end of a word.

      Georgian permits very complex clusters at the beginning as well as at the end of words:  mskhlebi pears; brdzola struggle; mdzghneri shit; gvhqavt we have him.

      Tense vs. lax vowels have the following relationship with syllable structure in English.

      In English, stressed lax vowels only occur in closed syllables, tense vowels occur in either open or closed syllables:

      Tense (word final syllable) (diphthongs, diphthongoid vowels, by, too, way, so, ma

      Lax (closed syllable) = bit, but, full, get, oil/or, and, (also--hard,  Boston pronunc.)  The only lax vowel to occur in an open syllable is schwa:  sofa  Only the schwa and rarely a-sounds are lax vowels which may occur word finally.  Some vocal gestures in mode in american English violate this: na. (meaning "no"), which has a

      Accent or stress is another suprasegmental feature.  It is a meaure of relative volume of sound between syllable peaks.  Auditorily, we hear an accented syllable of a word as relatively louder than the unaccented syllables.  Acoustically, this difference can be measured in decibels.

      Languages differ in how they use stress.

1) In some languages, each syllable is equally stressed or unstressed-- Cambodian

2) one syllable in each word is more stressed.

      The place of stress is fixed on a certain syllable:

1) initial. Finnish, Hungarian and other Finno-Ugric languages

2) penultimate.  Polish,

3) final.  French.

4) complex set of rules.  In Bulgarian nouns and verbs have separate sets of rules for stress placement.  Hopi (phonetic:  first syllable of a two syllable word:  síkwi meat;  in words of three or more syllables, accent falls on the first long vowel:  máamatsi to recognize; or on the first short vowel before a consonant cluster:  péntani to write;  otherwise it falls on the next to last syllable:  wunúvtu stand up)

      The place of stress is random.

1) In Russian the stress is completely random:  xoroshó, xoróshi.

2) In English the stress is more predictable but still random.  Usually a middle syllable of a longer word receives the stress.  In two syllable words stress is rando and often renders differences in meaning:   project/to project,  produce/produce, insult/ to insult.

      Some languages have more than one stress per word:  English is such a language.  In English, words of four syllables or more have a primary and a secondary stresseducátion.

      Many languages lack word stress and instead have phrasal stress, with one stress per syllable of an entire phrase.  Many Turkic languages have phrasal stress, usually on the last syllable of a phrase: 

      Some English compounds have phrasal stress on the first element of the compound.  Phrasal stress often distinguishes meaning in adjective/noun combinations:  White House, white house  greenhouse/ green house.

      Another prosodic feature is pitch, defined as the frequency of vibration of vocal cords).  Pitch is measured in hertzes.  Physiologically, pitch tends to be higher in woman than in men, and higher before puberty than after puberty.  Also, the pitch of women's voices tends to lower with old age; the pitch of men's voices tends to get higher with age.  Despite these physiological, non-linguistic universals, each language  uses pitch distinctions for linguistically meaningful purposes.

If pitch varies over an entire phrase or sentence, we call the different pitch curves by the term intonation.  Intonation conveys the speaker's attitude or feelings.  In other words, intonation has a deictic function in discourse: questions; or a connotative function:  anger, sarcasm, or various emotions.  Intonation can also convey purely syntactic information, as when it marks where a sentence ends.

      If the pitch of a single syllable or word has the effect of influencing the denotative meaning of the word, we call the different pitch distributions by the term tone.  Every language uses pitch as intonation, but only some languages use it as tone.  There are two basic types of tones in tone languages.

Register tones are measured by contrasts in the absolute pitch of different syllables.  Register tones may be high, mid, or low.  (Cf. the IPA symbols for register tones.)  Many West African languages use contrasts of high mid and low tones to distinguish word meaning: Zulu, Hausa, Yoruba.  See text pp 204-205. 

Contour tones are tones involving a pitch shift upward or downward on a single syllable.  Many languages of East and Southeast Asia use contour tones, the best known being Mandarin Chinese.   (Cf. the IPA symbols for register tones.)

Most tone languages use a combination of height and contour to create their tones.  (See text pp 241 for tones in Thai)  Navaho also uses two different types of tones:  register tones on short or long syllables, and contour tones on certain long syllables.  The Navaho word for Goodbye illustrates these tones:  Hágoóne'.

The only European languages that use pitch in the form of tones are Serbo-Croatian, Lithuanian and Latvian, and Swedish.

Tone languages also have intonation, a gradual increase or decrease in pitch over an utterance as well as an increase in general volume of sound on various parts of the utterance to indicate emotion.