Chapter 3 Speech Sounds. Vowels

 3.1. In the OE manuscripts (MSS), long vowels and diphthongs are sometimes marked with and acute accent (´); edited modern texts, the present one included, usually indicate etymologically long vowels with a macron (¯), leaving the short unmarked. Thus gōd, good, but God, God; wītan, to keep, but witan, to know; and so on. Scholars know which to mark or leave unmarked after comparisons made with cognate forms in other Gmc languages and from observing how these vowels later developed in ME and MnE. Another valuable source of evidence is OE poetry, where vowel length is often indicated by the meter (see pp. 274-288).

Classifying Vowels

3.2. Vowels may differ from one another in quantity (i.e. length), quality or both. They are classified as regards quality chiefly by three factors taken together: the degree of openness of the oral cavity, the position of the tongue, and the shape of the lips. In fuller detail:

1. The cavity may be slightly open, half open, or wide open: the vowels produced are accordingly high, mid, or low (compare MnE beat, bait, bat).
2. The tongue may be pushed forward, left at the center, or humped backward within the mouth: vowels are accordingly front, central, or back (compare MnE beat, but, boot).
3. The lips may be more or less pouted or they may be left inactive: vowels are accordingly round or unround (compare MnE boot, beat).

the accompanying diagram 4 shows the vowels of OE.

Diagram 4

________		____Front_______	Central__________		Back ________
		[i:] beat			[u:] boot
High		[y:] Ger kühn
		[ɪ] bit			[υ] put
		[y] Ger küss
		_________________________________________________
		[e:] bait			[o:] boat
Mid			[ə] but
		[ɛ] bet			[ɔ] Brit pot
		_________________________________________________
		[æ:] S-SW US buy			[α:] baah
Low
		[æ] bat	[a] Scot man		[α] US pot

__________________________________________________________

The rounded vowels are: [y:, y, u:, u, o:, ɔ]. the sound [æ:] has no MnE standard keyword but is widely used in Southern and Southwestern U.S. The sound [ə] is found in EWS only as the second element of diphthongs. The vowels [y:, y] lost rounding and did not survive into MnE.

Note that within each of the two pairs [æ:] and [æ] and [y:] and [y] the difference is quantitative, whereas within the other pairs there is a qualitative difference as well.

3.3. Any of these vowels may be designated (described) by its three characterizing features and by whether it is long or short. (In IPA the diacritic [:] following a symbol indicates that it is long.) For example, [α:] is a low back unround long vowel; [y] is a high front round short vowel.

Exercise 1 Following the examples given for [α:] and [y], write designations for the other vowels of Diagram 4.

3.4. A diphthong (Gr di-, two + phthongos, sound) is made by starting in the position of one vowel and moving smoothly and rapidly to the position of another, the shift taking place within a single syllable. MnE has diphthongs in such words as bout [α + υ], bite [α + ɪ], boy [ɔ + ɪ]; the vowels in such a word as chaos do not form a diphthong, however. (The OE diphthongs are described in 4.15.)

Sound Change

3.5. To know the elements of phonetics makes it possible to understand all the regular sound-changes of OE, such as diphthongization, assimilation, palatalization, and others described in later chapters. The sounds of every language are constantly subject to change for several reasons. Though any normal human being can hear and produce a very large number of different sound-features, no language utilizes more than a fraction of those possible. The feature through which one sound is distinct from others similar to it will be accompanied by non-distinctive features which native speakers and hearers learn to ignore or to discount. For example, we recognize nasality as distinctive in the consonant phoneme¹ /n/ because on its presence or absence depends the difference of meaning between pie and pine, or between sea and scene. But if nasality is added to vowels, as it is in some individuals' speech, it does not affect the meaning of what is said; we ignore it and notice only those features necessary to the distinctiveness of the linguistic signal (For example, [mæ̃n] said nasally means the same as [mæn] without nasality). Over the course of time, with many speakings and hearings, a formerly non-distinctive feature may become more prominent, or one that was sporadic or contingent may become more established. The former interrelation of sounds has been altered: there has been a sound-change.

3.6. Such changes are usually the result of the influence which sounds exert upon one another within the sequences of which words are built. It is well known that the [æ] in at is shorter than the [æ] in add - why? Because the vibration of the vocal cords in the first word must stop for voiceless t, whereas in the second vibrations of [æ] do not stop but continue into d, which is a voiced sound. The greater length of the [æ] in add is due simply to its envioronent. (As far as meaning is concerned, this difference of length is quite non-significant.)

3.7. Speech is a continuum. As sounds are spoken, they normally flow into each other with small adjustments that make articulation easier. By itself, [p] (a voiceless bilabial stop) would require an explosion to be heard, yet in context it is not always exploded. In the compound word hop-pole we do not have two explosions: the p of hop makes the bilabial closure and holds it for a moment; then the p of pole makes the explosion. Thus, in fact, two p's have been reduced to one lengthened [p:] which requires less complex articulation.

3.8. One of the the commonest sound-changes is assimilation, in which one sound or sound-feature becomes more like or similar to another near it. The past tense of have in OE is hæƒde (ƒ representing the sound [v]). By ME this has changed to hadde, the [v] becoming [d] by assimilation to the following [d].² (As with hop-pole the double letter represents length, not two explosions.) During most of the ME period hadde remains disyllabic, but by early MnE it has become had and is now even further reduced to 'd in such condensed forms as he'd promised, they'd arrived. Such progressive simplification occurs gradually enough to not destroy the function of the linguistic signal. Many features of OE show the effects of assimilation; one special type, umlaut or mutation, has left interesting traces in MnE. (See Ch 11.)

3.9. Very important sound-changes result also from the feature calles stress, which is simply the differential physical force exerted in producing syllables. This feature typifies the Gmc branch of IE; it has been in operation throughout the history of English and continues in force today. Every English word of two or more syllables places considerably more stress on one syllable than on the other or others. At least three distinctive degrees of stress exist in OE: primary, secondary, and weak. In the word gūð-cyning, war-king, they are found rspectively in the first, second, and third syllables: ˊ ˋ ˣ.

3.10. Strong stress tends to preserve sounds: weak stress lays them open to change. A striking example of stress working in cooperation with other phonetic factors to simplify the linguistic signal is furnished by the history of our MnE word lord. It goes back to Prehistoric OE *hlāƒ-ward, loaf-guardian, i.e. the master of a household in his capacity as distributor of food. As seperate monosyllables, hlāƒ and ward (in WS weard) normally took primary stress. Joined together as a compound word, the second element had to have less stress than the first, thus ˊˋ. With a further reduction of stress to ˊˣ, further change ensued: between the voiced sounds represented by ā and w, the sound represented by ƒ became voiced: phonetically, [f] to [v]. This is a type of assimilation. Further, the rounded semivowel represented by w was simplified altogether out of existence but left a ghostly reminder of itself in the rounding of the second vowel from a to o. Thus by the time of our historical OE records, *hlāƒ-ward had become *hlāƒord, and its etymology was doubtless obscure to most OE speakers. In ME we find the word as loverd: now the voiceless hl- of OE³ has been voiced (perhaps through assimilation to the following vowel) as [l]; the unstressed vowel has lost rounding, hence o is now spelled e, which probably represents [ə].⁴ Finally, internal v, occuring in an unstable position between vowels, disappears; the vowel of the primary-stressed syllable is preserved, the vowel of the weak-stressed syllable is lost, and a monosyllable is the result: lord. Through a similar if less sweeping series of changes, PrehOE *hlāƒ-dige, bread-kneader, became MnE lady.

3.11. One of the most sweeping effects of the stress-differential in changing a language may be observed toward the end of the OE period. Strong stress remains on the base syllable of words; prefixes and suffixes accordingly become weaker by contrast and tend to be at first reduced, then entirely lost. Inflectional syllables, coming at the ends of words, are especially subject to this erosion; indeed, it is their widespread disappearance more than anything else which marks the boundary between OE and ME.

3.12. Sound changes are not of equal importance. Those which affect an entire category of sounds, or which even produce a realignment in the structure or system of the language, are obviously the more profound. Others may affect only a few words or may operate for a limited time then be overcome by countervailing forces. The final outcome of any sound-change is also greatly subject to such nonlinguistic factors as the prestige of one dialect over others (hence its spread at their expense) or serious dislocations in society because of war, plague, economic collapse, foreign influence. The language reflects the society: traditionalism fosters linguistic conservatism; social change fosters change in the language. Since it is probably true that English society has undergone more cultural change than any other in Western Europe, it is not surprising that the English language should have been less conservative than any other. Old English, richly varying in its dialects and everywhere reflecting the many changes and influenes it underwent during the six hundred years in which it was the vernacular language of England, may be seen as a microcosm of the whole English language - though ephasis will be placed, in this Grammar, on the relatively circumscribed and stable stage of EWS.

FOOTNOTES

¹A phoneme is a minimal unit of distinctive sound-feature which contrasts structurally with all other phonemes in the same language or form of speech. Differences in meaning are signaled by this distinctiveness of the phoneme. Phonemic symbols are regularly put between virgules: /n/, etc. See further Appendix II.

²In fuller detail: [v] was already voiced, like [d]; its assimilation consisted in its partial spirant closure becoming full stop closure and its position of articulation moving from labio-dental to dental.

³See further below, Ch 4, footnote 11.

⁴The change of OE ā > ME o in the first syllable reflects the isolative change (i.e., one occuring without reference to an immediate phonetic envioronment) by which every OE ā became ME o (phonetically [α:] > [ɔ:]) - compare OE bāt, ME bote (MnE boat). Most of the changes hitherto exemplified in the development of *hlāƒ-ward have been combinative changes (i.e. those conditioned by an immediate phonetic environment).

16