Part 6

[33] In a great many cases lisping is due to an over-long tongue; or the tongue may be "tied," in which case the ligature is easily cut.

[34] The back of the tongue may also be raised to some extent; how does this explain the substitution of [w] for [r] which is sometimes heard?

[35] The rule may also be stated thus: _r_ is only heard when a vowel follows in the same or the next word. "Vowel" must here be taken to include [j].

[36] The term "dark" here implies a deep and obscure resonance, with little friction.

[37] Contrast the [l] of _will_ and _willing_ (where its position between front vowels leads to forward formation).

[38] Some pronounce this word [sælv].

[39] Some pronounce this word [fɔlkən].

[40] These are modifications of the Scotch form of the word.

[41] [h] is described as voiceless; but it may also be produced with voice. We have seen that the vocal chords consist of a fleshy and a cartilaginous part: it is possible to let the former vibrate, while the latter is left open, and the breath passing through produces [h]. Try to utter this sound.

[42] The "neutral" vowel [ə], for which see § 38, requires less effort.

[43] Notice the faulty tendency to raise the tongue too high in uttering the first part of this diphthong; see § 40.

[44] However, there is a pronunciation of _are_, intermediate between the emphatic [ɑː] and the unstressed [ə], which may be described as short [ɑ]. The _a_ in the unstressed prefix _trans-_, and the second _a_ of _advantageous_ also have the sound of [ɑ] sometimes.

[45] Especially when it is final.

[46] Many cultivated people pronounce _girl_ as [gɛəl]; but [gəːl] is to be preferred. _Clerk_, _sergeant_ have [ɑː], not [əː]; also _Derby_, _Berkshire_, _Hertfordshire_.

[47] It is heard on the stage and in public speaking generally; in ordinary conversation it suggests the speech of a foreigner, especially if the [ɑ] element of the diphthong is lengthened.

[48] The first part is still more open in a common vulgar pronunciation of _tea_, _please_.

[49] For [ɔː] see the diagram on p. 125.

[50] Exceptions are _borrow_, etc., _work_, _attorney_, etc.

[51] To pronounce this [ə] distinctly in such words as _law_, _saw_ is a mistake.

[52] It is absurd to speak of _fort_ and _caught_, _morn_ and _dawn_ as "cockney" rimes; they are perfectly good rimes in standard English; and a southern Englishman only shows ignorance by speaking of them as bad. Considering, however, that standard English is by no means universal, the would-be poet is advised to avoid these rimes.

[53] For [ö] see the note on p. 68.

[54] In the case of this diphthong as well as in that in _name_, _pail_, etc. (see § 41), untrained singers usually betray themselves by passing too soon to the second part of the diphthong.

[55] For [o] see the diagram on p. 124.

[56] [ö] is [o] pronounced with the whole body of the tongue more forward than usual. To the ear it gives an effect like that of French _eu_ or German _ö_; but for these sounds the lips are rounded.

[57] The pronunciations [wində, pilə] for _window_, _pillow_ are, however, avoided by educated speakers.

[58] [ü] is [u] pronounced with the whole body of the tongue more forward than usual. To the ear it gives an effect like that of French _u_ or German _ü_; but for these sounds the lips are rounded.

[59] For [u] see the diagram on p. 124.

[60] See also § 29.

THE SOUNDS IN CONNECTED SPEECH

=46.= Let us take a familiar nursery rime as an example of simple conversational English; it will serve to give us some idea of the problems which have to be considered when we deal with the sounds of connected speech. This is the rime:

siŋ ə sɔŋ əv sikspəns | ə pɔkit ful əv rai | fɔːr ən tʍ_e_nti blækbəːdzz̥ | beikt in ə pai | w_e_n ðə pai wəz oŭpnd | ðə bəːdzz̥ bi´gæn tə siŋ | wɔzn̩(t) ðæt ə deinti diʃ | tə s_e_t bi´fɔː ðə kiŋ.

=47.= Pedantically precise speech is as much out of place in the nursery as vulgar speech; therefore we do not say, siŋ eĭ sɔŋ ɔv siksp_e_ns.

Notice that the following words have =strong and weak forms=, a weak form being regularly used when they are not stressed:--

_weak_ _strong_ a, an ə, ən ei, æn the ðə (before consonants) ðij ði (before vowels) has həz,[61] əz, z hæz have həv,[61] əv, v hæv had həd,[61] əd, d hæd is iz, z, s iˑz are ɑ(r), ə(r) ɑːə, ɑːr was wəz wɔz were wə(r) wəː(r), wɛːə(r) can kən, kn̩ kæn shall ʃəl, əl, l ʃæl will əl, l wil could kəd kud should ʃəd, ʃd, ʃt, d ʃud would wəd, əd, d wud he hi,[62] i hij she ʃi ʃij her hə(r),[62] ə(r) həː(r) him ɪm him his ɪz hiz we wi wij us əs ʌs you ju, jə juw them ðəm, (əm) ðem your jü(r), jɔ(r), jə(r) juːə(r) of əv ɔv from frəm, frm̩ frɔm to tə (before consonants) tu and ənd, n̩d, ɔn, n̩ ænd or ɔ(r), ə(r) ɔː(r)

The use of strong forms for weak ones in ordinary conversation is undoubtedly a fault, and should be avoided; much of the unnatural reading aloud in our schools is due to this cause. Foreigners who have lived long in England often fail in this respect when they have overcome almost all other difficulties. It is also not uncommon in the speech of colonials.

* * * * *

Notice [p_e_ns], but [sikspəns].

A word which forms the second part of a compound often changes in pronunciation, a weaker form being substituted. Compare _penny_ and _halfpenny_, _board_ and _cupboard_, _come_ and _welcome_, _day_ and _yesterday_, _ways_ and _always_, _fast_ and _breakfast_, _mouth_ and _Portsmouth_, _land_ and _England_, _ford_ and _Oxford_.

The first letter of the second part is sometimes dropped; thus the _w_ in _housewife_ (case for needles, etc.) [hʌzif], _Greenwich_, _Harwich_, _Woolwich_, _Norwich_, _Keswick_, _Warwick_ is no longer pronounced, nor the _h_ in _shepherd_, _forehead_, _Clapham_, _Sydenham_, and in many words beginning with _ex-_, e.g., _exhale_,[63] _exhaust_, _exhibit_, _exhilarate_, _exhort_. The dropping of _h_ in _neighbourhood_ is vulgar.

Sometimes there is a change in the first part of a compound word. Compare _half_ and _halfpenny_, _three_ and _threepence_, _fore_ and _forehead_, _break_ and _breakfast_.

The stress of compounds like _sixpence_ is discussed below ("blackbirds").

=48.= In _pocket_ the second vowel is not middle [_e_], but a very laxly articulated variety of [ɪ], with the tongue only a little higher than for close [e]; see § 38. In the speech of elocutionists the middle [_e_] often appears here. They tell of the [gaːəd_e_n ɔv ijd_e_n], just as they succeed in pronouncing [d_e_vɪl] instead of [d_e_vəl], thus avoiding all offence; for [d_e_vəl] is said by common people, but [d_e_vɪl] only by the polite.

* * * * *

Notice that in _four-and-twenty_ the _r_ is pronounced, as it comes between vowels; but it is mute in _before the King_, where it comes before a consonant, as in the word _forth_. See § 32.

=49.= In _and_ the _d_ is dropped. Here it might be a case of assimilation; that is to say the _t_ which immediately follows, and which is closely akin to it, might have changed it to [t], and the two would have fallen together.

In _sit down_, do you utter both [t] and [d]? If you speak naturally, you probably say [sidaun] or [sitaun]. What is your pronunciation of _hold tight_, _less zeal_?

In _cupboard_ none but the absurdly precise pronounce the [p].

* * * * *

Assimilation of consonants is common in English, and the more colloquial the speech is, the more assimilation you are likely to find. Assimilation reduces the number of movements which have to be made, and thus represents a saving of trouble; and in colloquial speech we incline to take as little trouble as possible.

The general rule is, that when two sounds come together, those movements of articulation which are common to both are executed once only. Thus in _don't_,[64] the stopping of the passage for [n] also does duty for [t]; it is the opening of the passage which constitutes the [t]. In _stamp_ the closure for [m] also does duty for [p]. In _witness_ the closure for [t] remains for [n], which merely requires the opening of the nose-passage and vibration of the vocal chords.

Utter the word _clean_, and observe whether you produce the [k] in the same way as in _keen_; probably you will find that for the [k] of _clean_ you open the closure only at the sides, leaving the centre of the tongue in contact, ready for the production of [l]. See whether anything similar happens when you say the word _atlas_.

Sometimes a voiced sound makes a neighbouring sound voiced, or a voiceless sound makes a neighbouring sound voiceless. Examples in the nursery rime are [bəːdzz̥] and [beikt]; find similar examples of the _s_ of the plural[65] and the _ed_ of the past participle, and determine in each case whether the final sound is voiced or voiceless. Try to find pairs like _lagged_ and _lacked_, _bids_ and _bits_.

Utter the words _apt_, _act_, and notice carefully when you make the closure for [t]; probably it is earlier than you would have thought. Do you make the [n] closure in _open_ before or after the [p] opening?

In compound words, and in neighbouring words which belong closely together, assimilation is common. When one word ends in a voiceless sound and the other begins with a voiced sound, or _vice versa_, it is usually the second which prevails. Observe _cupboard_ [kʌbəd], _raspberry_ [rɑːzbəri], _blackguard_ [blægɑːd], _bedtime_ [betaim], _hold tight_ [houltait].

Consider the pronunciation of _observe_, _obstacle_, _gooseberry_, _absolve_, _absolute_.

In careless speech [hɔːʃu] is heard for [hɔːsʃu], [lædbru grouv] does duty for _Ladbroke Grove_, and [həsijn] for _has seen_. _Is she_ is regularly pronounced [iʒ ʃi], or [iʃi] in quick conversation.

The nasals frequently change to suit the place of articulation of the _following_ sound, as in _congress_ [kɔŋres], _congregation_ [kɔŋri´geiʃən], _anchor_, _concave_, _conclusion_, _concourse_, _concrete_, _syncope_, _tranquil_, _unctuous_, _pincushion_ [piŋkuʃən], _infamous_ [imfəməs], _Holland Park_ [hɔləmpɑːk]; or of the _preceding_ sound, as in _second single_ [sekŋsiŋgəl], _captain_ [kæpm̩], _open the door_ [oupm̩ ðə dɔː], _cup and saucer_ [kʌpm̩sɔːsə]. The examples from _pincushion_ onwards occur only in distinctly careless speech.

The change of [s] to [z] in _house_, _houses_ [haus, hauziz], shows a different kind of assimilation.

=50.= The dropping of _d_ in _four-and-twenty_ might also be due to the desire to =simplify a group of consonants=; and this will seem the more likely explanation if we notice that the _d_ of _and_ is generally dropped before a consonant, but kept before a vowel. Compare _you and Ida_, _bread and butter_; if you drop the _d_ in the first instance, or utter it in the second, you are equally wrong. Such simplifying is fairly common in educated speech; most people drop the _t_ in _often_, and the _p_ in _empty_ (where it has no etymological justification), and _jumped_; in colloquial speech _don't know_ is [dʌnou]. In quite careless speech you may notice consonants dropped in such words as _acts_, _insects_, but this is clearly a licence which cannot be permitted in the class-room. Indeed these groups of consonants should be articulated with great care. Nothing so quickly gives an effect of slovenly speech as the slurring of consonants, where it is not generally adopted.

* * * * *

In ordinary speech numerous instances occur of this tendency to simplify groups of consonants, _d_ and _t_ being the sounds most frequently dropped.

_d_ is not pronounced in _handkerchief_ [hæŋkətʃif], _handsome_ [hænsəm], _Windsor_ [winzə], _Guildford_ [gilfəˑd], _Ingoldsby_ [iŋgəlzbi], _Wednesday_ [wenzdi].[66]

The _d_ in _friends_, _grandfather_ is also often dropped; and, in very careless speech, the _d_ of such words as _old_, _cold_, _child_, _thousand_, _kindness_, _landlord_.

_t_ is not pronounced in _christen_ [krisən], _glisten_ [glisən], _hasten_ [heisən], _listen_ [lisən], _moisten_ [mɔisən], _apostle_ [ə´pɔsəl], _bustle_ [bʌsəl], _castle_ [kɑːsəl], _epistle_ [i´pisəl], _gristle_ [grisəl], _hustle_ [hʌsəl], _ostler_ [ɔslə], _pestle_ [pesəl], _rustle_ [rʌsəl], _thistle_ [θisəl], _trestle_ [tresəl], _whistle_ [wisəl], _wrestle_ [resəl], *_Westbourne_ [wesbən], *_Westminster_ [wesminstə], _Christmas_ [krisməs], _chestnut_ [tʃesnət], _coastguard_ [kousgɑːd], _often_ [ɔːfən], _soften_ [sɔːfən], _mortgage_ [mɔːgidʒ], *_directly_ [di´rekli], *_exactly_ [i´gzækli], _postpone_ [pous´poun], _waistcoat_ [weiskət, weskət], _bankruptcy_ [bæŋkrəpsi].

In very careless speech the _t_ of such words as _slept_, _swept_, _wept_ is dropped; also in _acts_, _facts_, _insects_, _sects_.

Notice the French _rosbif_, _bifteck_.

_th_ is not pronounced in _asthma_ [æsmə], _isthmus_ [isməs] and (carelessly) in _depths_.

_p_ is not pronounced in _empty_ [emti], _jumped_ [dʒʌmt], _tempt_ [temt], _attempt_ [ə´temt], _contempt_ [kən´temt], _peremptory_ [pər´emtəri], _symptom_ [simtəm],[67] _sapphire_ [sæfaiə], _Sappho_ [sæfou].

_c_ is not pronounced in _corpuscle_ [kɔː´pʌsəl], _muscle_ [mʌsəl], _victuals_ [vitl̩z], _indict_ [in´dait].[68]

So many educated speakers say [ɑːst] for _asked_, that this pronunciation must be regarded as no longer incorrect.

What is the usual pronunciation of _next station_?

In careful speech the simplifications marked with an asterisk are avoided, as also such pronunciations as [ail dʒʌssij] for _I'll just see_, [difikl̩ kwestʃn̩z] for _difficult questions_. The omission of [k] in the pronunciation of _arctic_ and _antarctic_ and of [g] in _recognise_ is generally regarded as faulty.

* * * * *

Unfamiliar groups of consonants at the beginning of words are simplified by dropping the first sound; notice the simplification of

_bd_ in _bdellium_; _chth_ in _chthonian_; _gn_ in _gnaw_, _gneiss_, _gnome_, _gnostic_; _gz (x)_ in _Xerxes_ [zəːksijz], _Xenophon_; _kn_ in _knee_, _knit_, _know_, etc.; _mn_ in _mnemonic_; _phth_ in _phthisis_ [θaisis], also [taisis]; _pn_ in _pneumatic_, _pneumonia_; _ps_ in _psalm_, _pseudo-_, _Psyche_, _psychic_ [saikik], _psychology_, etc.; _pt_ in _ptarmigan_, _Ptolemy_; _sw_ becomes _s_ in _sword_ (observe also _answer_); _wr_ in _wreck_, _write_, etc.

Similarly, an unfamiliar group at the end of a word is simplified, usually by dropping the last sound; notice--

_ln_ in _kiln_ (some do not drop this _n_); _mb_ in _bomb_, _catacomb_, _climb_, _comb_, _dumb_, _hecatomb_, _lamb_, _limb_, _plumber_, _succumb_, _tomb_; _mn_ in _autumn_, _column_, _condemn_, _contemn_, _hymn_, _limn_, _solemn_.

(The opposite tendency is found in vulgar speech, where [vɑːmint] is said for _vermin_, [draund] for _drown_.)

* * * * *

Observe _drachm_ [dræm], _yacht_ [jɔt], _impugn_ [im´pjuwn], _physiognomy_ [fizi´ɔnəmi], _diaphragm_ [daiəfræm], _paradigm_ [pærədaim], _phlegm_ [flem], _sign_ [sain], _feign_, _reign_, _foreign_, _benign_.

=51.= In [rai] we have a diphthong. It is worth noting that the English diphthongs [ai, au, ɔi, ei, ou], etc., all have the =stress= on the former element.

* * * * *

_Blackbirds_ and _black birds_: in the spelling we distinguish these by writing the first as one word, the second as two. What difference is there in the sounds? If you listen carefully, you will find that the second vowel in the compound word is just a little shorter than in _birds_ standing alone, and that in _blackbirds_ the opening of the closure for [k] is not heard, while in _black birds_ it may be audible. The chief difference, however, lies in the =stress of the compound word=. _Blackbirds_ is an example of descending stress [>], _black birds_ is pronounced with level stress [=], perhaps with ascending stress [<].

Take the following compound words or groups, and classify them according to their stress:--

_Sixpence_, _rainbow_, _good morning_, _looking glass_, _moonshine_, _bravo!_, _twenty-four_, _twenty-four men_, _High Street_, _London Road_, _waterspout_, _right of way_, _undo_, _Mr Jones_, _Park Lane_, _season ticket_, _sunflower_, _Hongkong_, _steel pen_, _Chinese_, _hallo!_ _bill of fare_, _earthquake_, _sea wall_, _Bond Street_, _Grosvenor Square_, _fourteen_, _Hyde Park_.

Try to deduce some rules from these examples. It has been said that level stress contrasts, and uneven stress unites the ideas expressed by the compound words; do you agree with this?

Notice what difficulty our level stress gives to the German; he will utter _steel pen_, _Hyde Park_, etc., with descending stress. Do so yourself, and observe how strange it sounds.

Notice the difference in stress of

Substantive or Adjective Verb _absent_ _to absent_ _accent_ _to accent_ _consort_ _to consort_ _converse_ _to converse_ _desert_ _to desert_ _prefix_ _to prefix_ _present_ _to present_

Substantive Verb _proceeds_ _to proceed_ _produce_ _to produce_ _project_ _to project_ _protest_ _to protest_ _rebel_ _to rebel_ _record_ _to record_ _refuse_ _to refuse_

Substantive Adjective _compact_ _compact_ _instinct_ _instinct_ _minute_ _minute_

Some words of two syllables have the stress on the first or the second syllable according to their place in the sentence. Consider the accent of the italicised words in the following sentences: They sat _outside_. An _outside_ passenger. Among the _Chinese_. A _Chinese_ lantern. His age is _fifteen_. I have _fifteen_ shillings. Some fell by the _wayside_. A _wayside_ inn. Try to find a rule governing these cases.

The stresses in a sentence are considered in § 54.

* * * * *

_When_ would be pronounced as voiceless [ʍ] by some, hardly by a southern English nurse saying the rime (§ 46). Notice how the tongue moves forward as the [n] passes over into the [ð] in _when the_.

_Was_ is in the weak form because it is quite unstressed; but notice: [wɛːə ju riːəli ðɛːə? j_e_s, ai wɔz].

=52.= In _opened_, observe carefully how the consonants are articulated, and put their action down in writing.

How many syllables are there in _opened_, _bubbles_, _chasms_, _mittens_?

Probably you have no difficulty in understanding and answering this question, but if asked to describe =a syllable= you might hesitate, for it is not easy.

Utter [ɑ] and then [t]; which carries farther, which has greater fulness of sound or sonority? If you wished to attract the attention of some one, and were only allowed to utter one of these two sounds, you would prefer [ɑ] without hesitation. Why is [ɑ] more sonorous than [t]? Because, whereas [t] is only a brief noise, in [ɑ] the current of breath is rendered musical by the vibration of the vocal chords, and has a free passage through the wide open mouth. Indeed [ɑ] is the most sonorous of all sounds. It is clear that voiced sounds are more sonorous than voiceless, vowels than consonants, continuants than stops. The liquids and nasals stand between vowels and consonants in point of sonority; they are voiced and with either a fair passage through the mouth or a free passage through the nose. A good deal naturally depends on the force and the pitch of the sounds; a whispered [ɑ] may not carry so far as a forcible [s].

Now if a sound with good carrying power has for its neighbours sounds that do not carry far, it helps them to be heard; notice how such weakly sonorous sounds as [t] or [p] occurring in the words of a song are quite clearly heard at the other end of a large concert hall. They are carried along by the full sounding vowels, as the greater volume of air employed causes more pressure, and hence a more forcible and louder release. It is the sounds of greater sonority that carry the syllable, which term is also applied to a vowel standing alone, or beside other vowels of practically equal sonority. In English, the syllable is generally carried by vowels; sometimes also by liquids and nasals, which are then called =syllabic=.[69]

Rules for dividing words into syllables are given in most grammars, and are required for writing and printing; but they do not always represent the actual state of things. When a consonant comes between two vowels, it really belongs to both syllables. In _leaving_ we pronounce neither _lea-ving_ nor _leav-ing_.

From the phonetic point of view we may think of words and groups of words as consisting of a series of sounds of varying sonority. We may indicate the sonority very roughly by lines; if we connect their top ends, we shall obtain a curve. Thus the word _sonority_ might be represented as follows (no attempt is here made at scientific accuracy).

The curves will represent a series of waves; and each of these waves is a syllable.

* * * * *

_Began_: notice the quality of the vowel in the first, unstressed syllable of this word. It is higher than any real _e_ sound, and is very laxly articulated. It occurs also in _before_, _enough_, _inquire_; find other words in which it occurs. Is it the same sound as the second vowel in _lily_?

_To sing_: read the sixth line quite naturally and see whether you say [tə] or [tu]; get friends to read it, and find out what they say.

_When you wish to ascertain how a friend pronounces some particular sound, do not tell him what this sound is, or he may pronounce it not naturally, but in what he believes, or has been told, is "the correct pronunciation."_

Try to ascertain the pronunciation of these sentences: _What are you going to do to-morrow morning?_ _I'm going to answer letters._

_Wasn't that_: _was_ is here in the strong form (§ 47); are weak forms found at the beginning of a sentence? Notice the syllabic [n]; also the simplification of the group of consonants by the omission of [t]. What is the weak form of _that_? When is it used?

The remaining words present nothing of special interest.

=53.= We may now consider the =stress of the sentence=. For this purpose it is sufficient to consider the most sonorous part of each syllable, generally speaking a vowel. We may distinguish stress and absence of stress, which we can designate by the signs / and ×; extra strong stress will be //, and secondary stress \. The first line of Sing a Song of Sixpence, will then run:--

// × / × // \

Here "sing" and "six" have the strongest stress; "song" has ordinary stress.

Secondary stress is given to that syllable of a word which is stressed, but has not the chief stress; thus the stresses in _energetic_ may be written \ × / ×.