Spelling pronunciation

Spelling pronunciation occurs when people display their literacy by using a pronunciation that more closely reflects the spelling, under the impression that to do so must be more correct. It is an ongoing phenomenon in English and can be attributed to the increase in literacy over the last half-century. Broadcasters such as the BBC, especially in their news bulletins, are particularly prone to it.

Very often the sounding or not of a particular consonant, which may have traditionally long been silent, is at issue. Some examples are:
 * Whether or not to pronounce both rs in "library" or "February" (should  it end exactly like "January"?)
 * Whether or not to pronounce the t in "often"
 * Whether or not to pronounce the l in "salmon"

In the last case, sounding the l seems to be an American phenomenon, and indeed there are other cases where what might seem to be an arbitrary spelling pronunciation is in fact a genuine difference between British and American English.

Depalatalisation

 * The áccents show stress and pronunciation: see English phonemes

Much spelling pronunciation concerns depalatalisation, usually that of s, c or t when followed by i or u as an 's' sound where a 'sh' is normal. A classic example is íssue: should it be pronounced *íssyûe (IPA /'ɪsju/) or palatalised as *íshûe (IPA /'ɪʃu/)? Clearly, the unpalatalised version *íssyûe reflects the spelling and is of greater antiquity, but is has long been the norm to say *íshûe. Even so, *íssyûe has never quite gone away, as it is perceived that resisting the urge to palatalise to the 'sh' sound is somehow more correct. The same can be heard with tíssue, and of course there is no palatalisation in assûme and consûme: both always have -syûme: a -shûme pronunciation would sound comically uneducated. But foreigners should learn to palatalise words that normally have this pronunciation, if they are not to sound mannered. Nobody says *insyûrence for insûrance; indeed it most often sounds like *inshŏrance (British) or *inshûrance (American).

And sûre itself retains its palatalisation: AmE *shûre, BrE sůre certain = shŏre sea: it does not sound like seŵer. spêcies however has been showing signs of returning from *spêeshíz to *spêessíz or *spêessêez. Meanwhile, Cambôdia, ‘Cambodge’ in French, was for a time in the 1990s Kampuchêa in English: we can see that the -dia represents a palatalisation to a 'j' sound that has been lost in English, and that the K- spelling was an attempt by some to restore it in the form of a 'ch'.

Other words not normally palatalised are: assûme, dûe, ensûe, euthanâsia (-zìə), Galícìa (see below), Parísìan (-zì-), presûme (-zyûme), pursûe, redûce and other words in -dûce, sûicide, sûit (cf. shoôt), synaesthêsia - though anasthaêsia might be (AmE both -nəs-), Tunísìa (BrE *Cheŵ-nízier) and Valéncia. Other normally palatalised words are: amnêsia (-zhə), apprêciâte, Âsia, assôciate, atrôcious, caucâsian and other words in -sian (-zhn), Chrístian (*Kríshchən) and other words in -tian (-shn), Indonêsia, milítia, negôtiate, Patrícia, Përsian, pléasure (-zh-), précious and other words in -cious, préssure (-sh-) and other words in -sure, sôciology, sôldier, substántial, Croâtia (Crô-, and other words in -tia), vísion and other words in -sion (-zhn), vítiate, volítion and other words in -tion (-shn).

Ending -ies lengthening from -íz to -êez
From around the beginning of the twenty-first century, another very strong tendency, perhaps influenced by the spelling pronunciation of some foreign learners, is to lengthen the vowel sound in the ending -ies in nouns, but, interestingly, not in verbs (cárries is still *cárríz). Traditionally the -êez sound has been used only for the plural of words from Greek ending in -is: crîsis, plural crîsês (*crîsêez), but it can now be heard in the plurals of nouns in -y, such as pàrties:

This appears to be the continuation of a change, as we can hear in films from the mid-twentieth century how the -y ending too was once pronounced í. But this new change can be considered unfortunate for listeners, especially learners, as it reduces the distinction between spoken singular and plural to just the z sound (written s).

Sometimes Dâvies can be heard with this pronunciation, instead of being homophonous with Dâvis. The tendency, often heard on BBC News bulletins, may also be spreading to the past tense of verbs ending in -y: accòmpaníed becoming accòmpanìêd; míschíef can also be heard with the spelling pronunciation, and one can also hear intéstíne pronounced like Chrístìne.

Other spelling pronunciations

 * A Spanish-style Colómbia instead of Colòmbia (= Colúmbia British, trademark) as if there were any danger of confusion. Similarly, some affect a Spanish pronunciation of Chílê, which traditionally sounds exactly like chílly. There has also been a slight tendency to pronounce Chílean ‘Chilèan’, perhaps inspired by the Spanish version ‘Chileno’.


 * For Cátalan many people say 'Cátalán', presumably under the impression it is more correct, but they would never do this with Itálian or indeed Chílean.


 * There are various pronunciations of Fébruary: In AmE *Fébûãry, like Jánûãry, and in BrE *Fébry or *Fébyəry, like *Jányəry. The spelling pronunciation pronounces both rs: AmE Fébrûãry, BrE *Fébrəry.
 * In óften, *óffen, some people pronounce the traditionally silent t.
 * A regular pronunciation of âi in agâin and agâinst can sometimes be heard instead of *agén and *agénst. This is perhaps unfortunate, as it can be confused with a gâin: 'celebrating a gain' or 'celebrating again'?
 * toûr alternates with tŏur; BrE pŏor increasingly sounds like poôr, although dŏor remains unaffected.
 * While it is normal to pronounce the n in autúmnal, because of the change of stress from åutumn (*åwtəm), one now sometimes hears also the n of cólumn (*cóllum) in cólumnist.
 * Woòlwich, Gréenwich and Nórwich, which traditionally echo pórridge, can now be heard with final consonant unvoiced, as in Ípswich.
 * The traditional *crézzent may be giving way to a hissing sound in créscent.
 * Some speakers sây sâys and sâid, but most still sây *séz and *séd.
 * If Galícian were as common as patrícian (*pətríshən), one would not hear *Gəlísiən, let alone 'Gəlíthìən' (the latter after the European Spanish th sound of the c). Similarly, if Ossétia were not relatively new to most broadcasters, one would expect it to rhyme with Lucrêtia, *Osêsha: as it is, one is tempted to misspell it 'Osettia'.
 * Weak pronunciations of syllables traditionally with the schwa sound sometimes give way to strong ones: an example is Thaîlánd /taɪlænd/ (sounding like two words, 'tie land') instead of Thaîland /taɪlənd/: one would not expect such a pronunciation for England, Scotland, Ireland, Iceland, Greenland or New Zealand, and it recalls the days of the British Empire: Basûtolánd, Swàzilánd  and Matabêlelánd do indeed usually retain such a pronunciation. And similarly, there is a tendency for people called Jûliet and Márian, with schwa as the final vowel, to pronounce their names more like Juliétte and Mariánne.
 * Shreŵsbury is the spelling pronunciation of the English town long known as *Shrôzebury.
 * Tehràn is traditionally *Té-ràn: the Persian pronunciation, pronouncing the h after a closed [e] sound, is reflected in the spelling, but since the sequence -ehr- does not come naturally in English (despite an alternative spelling Teheràn) there has developed a tendency to lengthen the first vowel to the sound in èh?, making the syllable sound like Tây.
 * The strong form of the indefinite article a (in normal speech always the weak form, schwa, [ə]), pronounced like the letter's name, *ây, can be heard in self-conscious speech: to withstánd â fûture dròught.

Glottal starts
In the opposite direction to spelling pronunciation is a trend to replace the traditional word-linking sounds r and w with glottal stops (\) even when this means leaving silent a written letter. This may also be the result of a foreign learners’ habit, that of pronouncing each word separately, without liaisons. These could be called 'glottal starts', as they attach themselves emphatically to the vowel beginning the second word: with the loss of a linking r sound: and even where written as one word:

But this does not happen with the consonantal 'y' liaison: hîgh úp is still normally pronounced *hiyúp. And conversely, people still tend to join words where the first ends in a with an imaginary, invisible 'r', by analogy with the -er ending: Chîna-América relâtions (*Chinərəmericərəlâshənz).

Other
One often hears Beijíng pronounced with a 'zh' sound instead of a normal j.

Sometimes the names of places in the Spanish-speaking world that have c before e or i, such as Valéncia, are pronounced with the unvoiced 'th' sound, 'Valénthia', echoing the pronunciation of Madrid (not all Spanish has this sound). Also, Nicarágûa is giving way to a more Spanish-sounding *Nicarágwa.

Some English speakers pay homage to a local pronunciation by calling Kósovo 'Kosova'.

Tbilìsi can sometimes be heard with an extra 'i', 'Tíb-', replacing the usual schwa sound before the b (*Təblêessy).

Other languages
Normally, Portuguese final unstressed -o is pronounced as a very short /ʊ/, much like the 'u' in English 'put'. But the name of the euro, the Europe-wide currency that in 2002 replaced the Portuguese escudo, is pronounced by most speakers most of the time without this characteristic. Only when speaking rapidly do the Portuguese say 'yuru'. The reason is no doubt awareness that the word is derived from 'Europa', where the 'o' is very much an 'o' and stressed.

In Italian, there is a tendency, when speaking slowly, to sound the 'i' in words where it was originally introduced merely as a spelling device to soften the preceding 'c' or 'g', as for example in the names Luciano, Giovanni and Giuseppe (compare Luke, John and Joseph).