I. IntroductionFootnote 1
Long before the restoration of the Classical pronunciation of Latin in the nineteenth century, the English language had – and still has – a naturalised, or Anglicised, system for pronouncing Latin, whether Latin prose, poetry or Classical proper names in their full Latin forms.Footnote 2 The legacy of this method in British English is fundamental: almost everyone uses it unconsciously and effortlessly in the standard Received Pronunciation of names such as ‘Caesar’, ‘Cicero’, ‘Euripides’, ‘Lucretius’, and ‘Statius’; or in words and phrases such as ‘circa’, ‘placebo’, ‘appendices’, ‘agenda’, ‘alias’ and ‘sub judice’.Footnote 3 Moreover, though most people do not realise it, there is in fact an underlying and consistent set of rules which operates to yield these pronunciations, and there are, indeed, many other proper names, nouns and phrases throughout our language whose pronunciation is essentially governed by the same Anglicised system.Footnote 4
Of course, in contrast to this, we have the Restored, or Classical, system of pronouncing Latin for the academic study of ancient texts. This reconstructed system, which approximates to what scholars think was the pronunciation of educated speakers of Latin in the Augustan period, was developed from around 1870, and has been widely used in the English-speaking world since the beginning of the twentieth century.Footnote 5 This Classical method has now rightly supplanted the older English system for pronouncing Latin in ancient literature or for philological study; and this was a major achievement of modern scholarship, since Classical texts in the nineteenth century were generally pronounced with a version of the English system – a serious hindrance to the technical study of Latin poetry or philology.Footnote 6
The Reformed pronunciation has also severely eroded the traditional English system in the pronunciation of Classical proper names. Today some classicists can be ignorant of even the distinction between English and Reformed pronunciation, and of the rules for an Anglicised pronunciation of an ancient name.
I intend here to examine the history of the English system used in Britain for pronouncing Latin, and above all the collapse of that system in the nineteenth and early twentieth centuries.Footnote 7 This will include the following: (1) a short review of the history of the pronunciation of Latin in Britain from the Middle Ages until the nineteenth century (Sections III and IV below); (2) a review of the rules and historical development of the English system, in the form it reached by the mid-nineteenth century (Section V); and (3) the reform of that system in the late Victorian era, and its erosion from the late nineteenth century and replacement with the Reformed pronunciation of Latin (Section VI).
II. Transcription
For the pronunciations given in this paper, I have used the standard IPA symbols for the Received Pronunciation (RP) of southern British English.Footnote 8 For stress in English pronunciations the following method is used:
-
ˈ = primary stress on the following syllable, e.g., Cato, [ˈkeɪ-təʊ], with stress on the first syllable ˈkeɪ.
-
ˌ = secondary stress on the following syllable, e.g., Aristophanes, [ˌær-ɪs-ˈtɒf-ə-ni:z], with stress on the syllable ˌær.
III. Latin in England c. 600–1800
Latin in England had a long history, and we can provide a brief survey below of the history of its pronunciation from the Middle Ages to the early nineteenth century. This can conveniently be broken up into three periods, as follows:
(1) Pre-Conquest Latin: This was spoken in England from the conversion until 1066, and was influenced by the spoken Latin of Late Antiquity and early medieval Latin.Footnote 9 The failure to pronounce Classical quantities, except in the penult, was already a feature of fifth-century Vulgar Latin, as was the tendency to make the first vowel of disyllables long, even when such vowels had been short in Classical Latin.Footnote 10 In England, the vernacular Anglo-Saxon languages appear in this period to have had much less influence on the pronunciation of Latin than the Romance vernaculars did on the Continent. Latin, when spoken in England, was probably a better phonetic representation of the written words than the Romance pronunciations, and, in particular, Alcuin of York's reforms in Latin pronunciation at Charlemagne's court, in which Latin was to be pronounced ad litteras (with every letter sounded), may well have reflected what was then regarded as a superior phonetic pronunciation used in England in contrast to the local Carolingian pronunciations on the Continent.Footnote 11
(2) Anglo-Norman Latin 1066–c. 1400: From 1066, British Latin fell under Anglo-Norman and French traditions. It was in this period that the English soft g (IPA [ʤ]) and c (IPA [s]) arose. The medieval tendency had been to pronounce c before front vowels as the combination [ts]. This, by the thirteenth century, was reduced to a simple [s] in English, as in French. In the case of g before front vowels, this had already been transformed into a soft g (IPA [ʤ]) by the twelfth century, though modern French diverged further to the sound of s as in ‘measure’ (IPA [ӡ]).Footnote 12 The diphthongs ae and oe were now pronounced as [e] or [i:], and, indeed, were often reduced to an e in spelling.Footnote 13 Vowels in accented syllables before two or more consonants (excluding the mute-and-liquid combination) were now shortened in pronunciation.Footnote 14 The vowels were not yet affected by the Great Vowel Shift. Thus amicus was pronounced [a-ˈmi:-kʊs] and pater as [ˈpɑ:-tər].
(3) Early Modern Latin c. 1400–1800: From the mid-fourteenth century, English was used in the teaching of Latin in schools – and so the vernacular came to inform a new national pronunciation of the ancient language.Footnote 15 Much the same type of process also occurred in other countries, so that there too national pronunciations of Latin arose in France, Germany, Italy and Holland.Footnote 16 In Britain, most accented antepenultimate syllables now tended to be shortened in pronunciation by taking up the following consonant in syllabification.Footnote 17
The period from 1400 to 1650 saw fundamental changes in English pronunciation itself, part of which was the Great Vowel Shift.Footnote 18 The Great Vowel Shift explains why a number of so-called ‘long vowels’ in Latin words are now in fact diphthongs, even though English tradition has maintained the potentially misleading expressions ‘short’ and ‘long’ vowels when referring to the pronunciations of Latin words. In a complex process after 1400, the Middle English long vowels underwent change, and the [ɑ:], [i:] and [ɔ:] were diphthongised. Thus the long vowels, which in early Middle English retained a phonetic character not unlike the Restored pronunciation of Latin or the later ‘Continental’ pronunciation adopted by some in the nineteenth century (with approximations of Italian vowels), radically diverged to modern English diphthongs by 1800.Footnote 19 Since Latin was naturally affected by concomitant changes, our unusual English system finally emerged as the qualitative and diphthongal changes in the vowel sounds between Middle and Modern English took effect. The most important changes for English pronunciation of Latin, are as follows:
-
The sound of a long a in Middle English went from [ɑ:] > [æ:] > [ɜ:] > [e:] > [eɪ].
-
The sound of a long i went from [i:] > [ɪi] > [əi] > [ᴧi] > [aɪ].Footnote 20
-
The sound of a long o went from [ɔ:] > [o:] > [o:u] > [oʊ] > [əʊ].
Thus by 1600 a Latin word like amicus had changed from [a-ˈmi:-kʊs] to [ə-ˈməi-kəs], then [ə-ˈmᴧi-kəs] and to modern [ə-ˈmai-kəs]. The pronunciation of pater went from [ˈpæ:-tər] to [ˈpe:-tər] to modern [ˈpei-tə]. With the arrival of our modern vowel sounds by the seventeenth to eighteenth centuries, the divergence between the English manner of speaking Latin and the various Continental methods was so noticeable to English travellers that it became the subject of numerous essays.Footnote 21 Some even polemicised against their countrymen's peculiar system, and advocated the adoption of an Italian pronunciation.Footnote 22
An early attempt at reforming the pronunciation of Latin in line with a reconstructed ancient system was made by the humanist scholar Erasmus. During Erasmus's second trip to England from 1509 to 1514, he visited Cambridge and his work later influenced scholars there to take up his ideas on a reformed pronunciation of Latin and Greek.Footnote 23 This attempt in the sixteenth century, inspired by Erasmus, was carried on principally at Cambridge by Professors Thomas Smith and John Cheke. The reforms provoked great controversy and Stephen Gardiner, the Chancellor of Cambridge and Bishop of Winchester, issued an edict in 1542 forbidding the new pronunciations of both Latin and Greek, though this was not enforced until 1554 and it was eventually repealed in 1558.Footnote 24 The influence of the reforms of Smith and Cheke is debated, but it appears to have been far more successful with regard to Greek, rather than Latin, which was still mainly pronounced with the Anglicised system.Footnote 25 By the seventeenth and eighteenth centuries, after the effects of the Great Vowel Shift, the English method of pronouncing Latin had become deeply ingrained.
All the phonetic and accentual characteristics sketched above were thus part of an organic development in spoken British Latin, which stretched back to the Middle Ages.Footnote 26 These principles applied to the pronunciation of Latin text and proper names.
With respect to proper names, it was John Walker, the English actor and lexicographer, who attempted to systematise the rules that governed the English pronunciation of Classical names, in his Key to the Classical pronunciation of Greek, Latin, and Scripture proper names, published in London in Reference Walker1798. A second edition of Walker's Key was published in 1804, and a third in 1807. By 1815, a fifth had appeared, and the Key was regularly incorporated into early nineteenth-century dictionaries. Walker's work was also improved by Reverend W. Trollope, Lancelot Sharpe and Thomas Swinburne Carr.Footnote 27
IV. Early and middle nineteenth century
The modern, developed English system for pronouncing Latin did not attain its final form until the late nineteenth century. We can speak justifiably of the modern English system for pronouncing Latin by about 1860, which applied not just to Latin texts from the ancient or medieval periods, but to countless names from Graeco-Roman biography, religion, mythology and geography, both in the Classics and in English literature.Footnote 28 Furthermore, in mathematics, logic, physics, biology, anatomy, medicine, philosophy and law, there were hundreds of borrowed Latin words whose pronunciations also essentially followed the rules of the English method.Footnote 29 In nineteenth-century science, for instance, the English pronunciation was applied to numerous technical words taken directly from Latin, as well as to Greek and Latin neologisms, particularly the names of species.Footnote 30 In law, a vast set of Latin phrases and terms used the English system, many of which we have to this day.Footnote 31
Even in the nineteenth-century study of Classics, an English system that had developed over many centuries was used for the pronunciation of Latin prose and poetry. Although the principles of this traditional system used for ancient texts were admittedly more inconsistent than those applied to Classical proper names, the differences were minor, such as certain exceptions were made in compound verbs and some monosyllables.Footnote 32 From the mid-Victorian period, however, this latter system was itself changed by reformers who decided to pronounce the long vowels of Classical quantities with English vowel sounds, rather than Restored ones (the so-called ‘Victorian’ system, see Section VI below).Footnote 33
It will also be readily apparent to readers of English poetry that the original Anglicised pronunciation was standard before the 1890s: those ignorant of it will not appreciate the rhymes in pre-twentieth-century English poems, nor will they understand the role that Latin played for centuries in English literature.Footnote 34 In order to illustrate this point, I provide below a sample of lines from poetry with pronunciations of Latin, where those are made to rhyme with English words. My examples range from the seventeenth to the nineteenth centuries. It should be noted that these transcriptions essentially reflect modern RP pronunciation and not strict phonetic transcriptions of seventeenth and eighteenth-century sounds (e.g., the final r was generally pronounced in these centuries, but I have not transcribed it here).
They are Fools that Reason thus,
Communis Error facit Jus [kə-ˈmju:-nɪs ˈer-ə ˈfeɪ-sɪt ʤᴧs].
(Robert Dixon, Canidia, or the Witches (1683), XVI.718–19)To what Absurdities will you hale us?
Semel malus semper præsumitur esse malus
[ˈsi:-məl ˈmeɪ-ləs ˈsem-pə prɪ-ˈsᴧm-ɪ-tɜ: ˈes-i: ˈmeɪ-ləs].
(ibid. XVI.724–5)Culpam in Jovem rejicias [ˈcᴧl-pəm ɪn ˈʤəʊ-vəm rə-ˈʤɪʃ-əs],
Say ’twas the Gods, not you that were vicious.
(ibid. XVI.1073–4)To prove that Peers should never vary,
Nor Leges Anglice mutari [ˈli:-ʤi:z ˈæɳglɪ-si mju:-ˈteə-ri]Footnote 35 .
(Thomas Tickell, For England's Injured Church and Law (1730), lines 79–10)Watches, when first invented – seek 'em,
In brother Trusler's Vade Mecum[ˈveɪ-di: ˈmi:-kəm].
(William Mason, King Stephen's Watch. A Tale (1782), lines 13–14)Your subtile spirits, or your mighty,
Your aqua fortis, aqua vitae [ˈæk-wə ˈfɔ:-tɪs, ˈæk-wə ˈvaɪ-ti:].
(Thomas Holcroft, Human Happiness, or the Sceptic (1783), I.151–2)With nonsense in verse can elate and delight ’em,
And gives them variety ad infinitum [æd ˌɪn-fɪ-ˈnaɪ-təm].
(Anthony Pasquin, The Children of Thepsis (1789), Part II.795–6)But peace to her – her Emperor and Diet,
Though now transferred to Buonapartè's ‘fiat’ [ˈfaɪ-ət].
(Byron, The Waltz: An Apostrophic Hymn (1813), lines 55–6)But that, indeed, was hopeless as can be,
Quite a poetic felony ‘de se’[di: si:].
(Byron, The Vision of Judgement (1822), XCIV.751–2)And had told all the party a great booming lie, he
Cook'd up, that ‘the fête was postposed sine die’ [ˈsaɪ-ni: ˈdaɪ-i:].
(Thomas Ingoldsby, The Lay of St Cuthbert (1840), lines 74–5)So unless you're a dunce, You'll see clearly, at once,
When you come to consider the facts of the case, he,
Of course never gave her his Vade in pace [ˈveɪ-di: ɪn ˈpeɪ-si:].
(Thomas Ingoldsby, The Lay of the Old Woman Clothed in Grey (1840), lines 263–5)Your thoughts, opinions, freely state ’em.
Then, here they follow seriatim [ˌsɪə-ri-ˈeɪ-tɪm].
(Robert Story, Reply to an Epistle from Mr. Gourley (1857), lines 31–2)May thy hand be fierce as Ate's [ˈeɪ-ti:z]
Fighting for our old Penates [pe-ˈneɪ-ti:z]
(Lady Wilde, The Old Man's Blessing (1864), lines 13–14)Pert Nelly Gwynne, gay Molly Davis,
And many another Rara Avis [ˈreə-rə ˈeɪ-vɪs].
(Thomas Ingoldsby, The Country Seat (1881), 18–19)Hired witnesses with tales shall ply ’em
For only two pounds two per diem [pɜ: ˈdaɪ-əm].
(Winthorp Mackworth Praed, The Contested Election (1888), lines 277–8)Footnote 36We can now examine the rules of the English system in the next section.
V. The English system and its development
In this section, I present the principles of the English system from rules 1 to 18 below, in the form the system had reached by the mid to late nineteenth century, but with some comments on its subsequent development in the twentieth century.
A. Accentuation and syllabification of Latin words in English
The English system has traditionally had fixed rules for pronouncing Latin and Classical proper names, and when these rules are learned one can, in nearly all instances, consistently apply them to attain the correct pronunciation.Footnote 37
The point to be borne in mind is that the system for syllabifying proper names for an English pronunciation is very similar to, but not identical with, the Classical syllabification. The fundamental difference consists in the treatment of antepenultimate syllables, and it is necessary to emphasise that this is the crux of the system. In general, however, the normal Classical conventions used to accent and syllabify a word still apply, and the following principles are valid:
1. Accent
1.1 Determining the accent: primary stress
(a) The rules for accenting words are the same as those in Classical Latin: the penult is accented if it is long, either by position or by nature.
If the penult is not long, then the accent goes on the antepenultimate syllable. Disyllabic words are naturally always accented on the first syllable.
In the traditional system used by classicists, a penultimate syllable is accented if it is
-
(i) long by nature, i.e., if it contains a long vowel or a diphthong, or
-
(ii) long by position, i.e., if the vowel is followed by two consonants (excluding mutes and liquids), or if it is followed by the customary double consonants, z, x or consonantal i (= j in older English words like Ajax).Footnote 38
But note that, in syllabification of words for English pronunciation, only the x is a strict double consonant equivalent to ks, and this will shorten the preceding vowel in pronunciation (but neither the z nor j does this).
In English, a mute-and-liquid combination is generally treated as a single consonant, though in Latin the vowel preceding this combination became part of a ‘common’ syllable.Footnote 39 The mute and liquids are generally bl, br, cl, chl, cr, chr, dr, fl, fr, gr, pl, pr, tr, and thr.Footnote 40 Note that ch, ph, rh (= r), and th are treated as single consonants.Footnote 41
(b) A knowledge of the ancient quantity of the penultimate vowel in a Latin word is thus of great importance to the pronunciation of that word in English. For Latin and Greek proper names, this has traditionally been obtained through scansion, but can easily be done today by consultation of a standard lexicon (e.g., Lewis and Short) or dictionary (e.g., The Oxford companion to Classical literature).
(c) In Greek-derived words, one can often fix the accent by referring to the quantity of the penultimate vowel, before applying the traditional Latinised transcription system.Footnote 42 Thus, if a penultimate vowel is long in Greek (e.g., a diphthong, an eta, omega, long iota, alpha or upsilon) or if it is followed by two consonants (excluding the mute-and-liquid combination), then it will be accented on that penult in the Latinised form, e.g., Σιγήρος = Si-ˈge-rus; Ποτείδαια = Pot-i-ˈdae-a; Ἡράκλειτος (with diphthong in the penult) = Her-a-ˈcli-tus; Κασσάνδρα = Cas-ˈsan-dra; Κύκνος = ˈCyc-nus.Footnote 43
If the penult is not long or not followed by two consonants other than a mute-plus-liquid combination, then the accent goes on the antepenult, e.g., Λέβεδος = ˈLeb-e-dus; Ἑκάτη (with short alpha in penult) = ˈHec-a-te; Τάναϊς ‘ (with short alpha in penult) = ˈTan-a-is; Νιόβη = ˈNi-o-be.Footnote 44
(d) Though it is vital to know the quantity of the penult in order to fix the accent, the other Classical vowel quantities are generally ignored in the traditional English pronunciation: this is the salient characteristic of English system.Footnote 45
1.2 Secondary stresses
(a) In polysyllabic words, a secondary stress falls on the second syllable before the primary accent, if this is the initial syllable or if it is long by position or nature.
If the second syllable before the primary stress is not long, then the secondary accent goes on the third before the primary stress, e.g., Himantopodes (ˌHim-an-ˈtop-o-des), Laomedontiades (La-ˌom-e-don-ˈti-a-des), Larentalia (ˌLar-en-ˈta-li-a).
(b) A further secondary stress also occurs in very long words. The rules for determining it are the same as for the secondary stress, but with the first secondary stress as the starting point, i.e., a further secondary stress falls on the second syllable before the secondary accent, if this is the initial syllable or if it is long by position or nature, etc. E.g., Claudiopolitani (ˌClau-di-ˌop-o-li-ˈta-ni), Polymachaeroplagides (ˌPol-y-ma-ˌchaer-o-ˈplag-i-des).
1.3 Exceptions due to the recessive accent
Movement of the accent has been a regular phenomenon in English, owing to the so-called ‘recessive accent’, the ‘tendency to place the … stress early in the word’.Footnote 46 Hence there are some words which are no longer accented as they would be in Classical Latin. Even in Walker's time, the words ‘orator’, ‘senator’, ‘auditor’, ‘minister’, and ‘plethora’ all had the accent on the antepenultimate syllable, rather than on the penult (as in Classical Latin). It is, generally, the movement of the accent from penult to antepenult that has occurred, and under rule 2.2 below this usually makes a closed syllable.
Even some proper names are affected by this trend, e.g., Amazon (originally A-ˈma-zonFootnote 47 but now ˈAm-a-zon), Alexandria (A-ˌlek-san-ˈdri-a but now ˌAl-ek-ˈsan-dri-a [x = ks]), Thalia (Tha-ˈli-a but now sometimes ˈTha-li-a), Chalcedon (Chal-ˈce-don but now often ˈChal-ce-don).Footnote 48
However, such exceptions are generally proper names well known by educated speakers of English.
2. Syllabification
Classicists have traditionally used an orthographic system of syllabification for Latin words and Classical proper names based to some extent on Classical principles, from which one can ascertain the sound of vowels in stressed syllables. But it should be noted that this orthographic system of syllabification must not be confused with phonetic syllabification.
2.1 Number of syllables
For syllabification, there are as many syllables in a Latin word as the number of vowels or diphthongs. A syllable that ends in a consonant is closed, and a syllable that consists simply of a vowel or diphthong or that ends with a vowel or diphthong is open. In most cases, any single consonant (or a mute-plus-liquid) between unaccented vowels is joined with the following vowel or diphthong.Footnote 49 Two consonants, excluding a mute-plus-liquid, are broken. Thus we have the following (which only show syllabification, not phonetic values): Me-ˈzen-ti-us, ˈCi-mon, ˈCae-sar, ˈBac-chae, ˈO-tho, ˈMar-cus, ˈMem-mi-us, O-ˈri-on, Pe-ˈrin-thus, Da-ˈri-us, Vi-ˈtel-li-us, ˈNer-va, Mi-ˈle-tus, Eu-ˈter-pe, Pla-ˈtae-a, Ca-ˈmil-lus, Mar-ˈcel-lus, Ly-ˈsan-der.
A minor exception exists for bl, gl and tl.Footnote 50
-
A bl is divided when after a u, e.g., ˈPub-li-us, Pub-ˈlic-o-la.
-
A tl is divided after an a, e.g., ˈAt-las, At-ˈlan-tis, At-ˈlan-tes.
-
A gl is divided after an a, i, or o, e.g., Ag-ˈla-o-phon, Ag-ˈlau-ros, Ag-ˈla-i-a, ˈsig-lum, Trog-ˈlod-y-tae.
2.2 Antepenultimate syllables and secondary stressed syllables
There is, however, a major exception to rule 2.1, one which is different from the syllabic principles of Classical Latin: the syllabification of the antepenult and secondary stressed syllables.
In accented antepenultimate or secondary syllables, the traditional English system of syllabification of Latin words has the following consonant included in that syllable (not joined with the next) to form a closed syllable, and two consonants are divided, including a mute and a liquid. This will shorten the vowel in pronunciation: vowels in closed, stressed syllables are, as a rule, short. One should note, however, that this written tradition of syllabifying Latin words in this way is at variance with the actual syllabification of modern phonetic/phonemic transcriptions, where the intervocalic consonant in such circumstances often goes with the following syllable. But the shortening of the vowel is in line with the English trisyllabic shortening rule, a centuries-old characteristic of the English language which has existed since late Middle English, and possibly before 1100, as noted by Chomsky and Halle.Footnote 51 Thus we have ˈCic-e-ro, ˈRom-u-lus, ˈSen-e-ca, ˈNic-i-as, ˈNem-e-sis, ˈLyd-i-a, ˈSal-a-mis, ˈHec-a-te, Bi-ˈthyn-i-a, ˌAr-is-ˈtoph-a-nes (with secondary, accented syllable closed), ˌAr-is-ˈtar-chus, Eu-ˈrip-i-des, Thu-ˈcyd-i-des, He-ˈrod-o-tus, ˌAr-is-ˈti-des.Footnote 52
2.3 Exceptional open antepenults
There are only two cases where the antepenult is open, as follows:
(a) The vowel u
The vowel u is exempt from rule 2.2 when in an accented, antepenultimate syllable (and in accented secondary syllables), and the single consonant or mute-plus-liquid goes with the following vowel, just as in rule 2.1.
But, when u is followed by bl, or two consonants other than a mute-plus-liquid, these are usually divided (even when the u is not accented).Footnote 53 Thus we have ˈJu-pi-ter, ˈU-ti-ca, ˈRu-bi-con, ˈCu-bri-tus, ˈSu-tri-um, ˈU-ra-nus, but Sub-ˈlic-i-us, ˈPub-li-us, ˈTul-li-a, ˈMum-mi-a, ˈUl-pi-us, ˈUr-bi-cus, ˈTus-cu-lum.
(b) The endings -ius, -ium, -ia, -ie, -ii, -ya, -yia, -eia, etc.
An accented antepenultimate syllable with a, e, ae, oe, o, or u (but not i or y) followed by a single consonant (or a mute and a liquid) and then a penultimate i, y, or e (or ae or oe) plus another vowel in the final syllable is not closed but open, i.e., the consonant (or mute-plus-liquid) goes with the following syllable: An-ˈto-ni-a, Hy-ˈpa-ti-a, ˈLae-vi-us, Ma-ˈcro-bi-us, Pal-ˈla-di-um, Ar-ˈca-di-a, Au-ˈso-ni-us, ˌBac-cha-ˈna-li-a, ˈClo-di-us, E-ˈge-ri-a, ˈFa-bi-a, Pom-ˈpo-ni-us, ˈRe-gi-a, Si-ˈdo-ni-us, I-ˈo-ni-a, ˈU-bi-i.
But note that i or y in antepenultimate syllables is exempt from this, and will be short, as in 2.2 above, e.g., ˈClyt-i-a, Nu-ˈmid-i-a, Bi-ˈthyn-i-a, Il-ˈlyr-i-a, ˈClin-i-as, Tri-ˈphyl-i-a.
2.4 Other stressed syllables
As noted above, all the rules expressed in 2.2 and 2.3 apply fully to secondary stressed syllables, e.g., ˌMax-i-ˈmi-nus, ˌPo-ly-ˈae-nus, ˌPhil-e-ˈde-mus, ˌTir-i-ˈda-tes, ˌHe-li-ˈop-o-lis, ˌHe-li-o-ˈgab-a-lus. Sometimes, however, anomalous pronunciations have arisen by ignoring the rules governing syllabification of secondary syllables, e.g., ˌPol-y-ˈae-nus for ˌPo-ly-ˈae-nus, ˌPol-i-or-ˈce-tes for ˌPo-li-or-ˈce-tes.
2.5 Double consonants
Although double consonants should be carefully syllabified according to the principles discussed above, note that, when one finally pronounces a word, one of the double consonants will be suppressed in pronunciation, in accordance with this standard characteristic of English, e.g., Apollo (A-ˈpol-lo but pronounced [ə-ˈpɒl-əʊ]), Hippias (ˈHip-pi-as but pronounced [ˈhɪp-i-æs] or [ˈhɪp-i-əs]), Thessalia (Thes-ˈsa-li-a but pronounced [θe-ˈseɪ-li-ə]), Assyria (As-ˈsyr-i-a but [ə-ˈsɪr-i-ə]).
2.6 Exceptions due to irregular closed syllables
Some words which should be accented in an open syllable have come, through long-standing English usage, to be closed by taking up the following consonant,Footnote 54 e.g., Diana (originally Di-ˈa-na but now Di-ˈan-a), Patroclus (Pa-ˈtro-clus but now Pa-ˈtroc-lus)Footnote 55 , Cleopatra (ˌCle-o-ˈpa-tra but now ˌCle-o-ˈpat-ra), Mithras (ˈMi-thras but now ˈMith-ras), atrium (ˈa-tri-um but now sometimes ˈat-ri-um), macron (ˈma-cron but now ˈmac-ron).
Note that a closed syllable shortens the vowel, e.g., Diana (once pronounced as [daɪ-ˈeɪ-nə] but now [daɪ-ˈæn-ə]). Though this has occurred in some proper names and Latin-derived words, these are generally well known.
B. Vowel length and pronunciation
3. Vowels
3.1 Sounds
The long and short vowel sounds in English pronunciation are as follows:
-
long a = IPA [eɪ], as in ‘take’, ‘make’, ‘fake’.
-
short a = IPA [æ], as in ‘mat’, ‘sat’, ‘hat’.
-
long e = IPA [i:], as first e in ‘scene’.
-
short e = IPA [e], as in ‘set’, ‘bet’.
-
long i = IPA [aɪ], as in ‘site’, ‘kite’.
-
short i = IPA [ɪ], as in ‘sit’, pit’.
-
y = i.
-
long o = IPA [əʊ], as in ‘wrote’, ‘no’.
-
short o = IPA [ɒ], as in ‘not’, ‘hot’.
-
long u = IPA [ju:] as in ‘mute’ (the semivowel [j] tends to appear before long u).
-
short u = IPA [ᴧ], as in ‘cup’, ‘puck’.
Because of the Great Vowel Shift (as described above in Section III.3), the so-called ‘long’ a, i and o are now diphthongs in English, and not really long vowels, though the convention has been to continue to call them ‘long’ vowels in the traditional English pronunciation of Latin. For example, the modern pronunciation of the long a began in the sixteenth century during the Great Vowel Shift, became common after 1650, and then standard English in the early eighteenth century.Footnote 56 In earlier centuries, the long a had taken a number of values, including the a of ‘father’.Footnote 57
3.2 Long vowels
3.2.1 Vowels are long when in an open, stressed position
(a) In the examples I give below, the word, its syllabification, and then its pronunciation in IPA symbols are given, e.g., Tigris (ˈTi-gris [ˈtaɪ-grɪs]), Numa (ˈNu-ma [ˈnju:-mə]), Delos (ˈDe-los [ˈdi:-lɒs]), Plato (ˈPla-to [ˈpleɪ-təʊ]), Draco (ˈDra-co [ˈdreɪ-kəʊ]), Darius (Da-ˈri-us [də-ˈraɪ-əs]), Psyche (ˈPsy-che [ˈsaɪ-ki:]), verbatim (ver-ˈba-tim [vɜ-ˈbeɪ-tɪm]), datum (ˈda-tum [ˈdeɪ-təm]), stratum (ˈstra-tum [ˈstreɪ-təm]), Lucius (ˈLu-ci-us [ˈl(j)u:-si-əs] or [ˈl(j)u:-shəs] or [ˈl(j)u:-shi-əs]), Musonius (Mu-ˈso-ni-us [mju-ˈsəʊ-ni-əs]), genius (ˈge-ni-us [ˈʤi:-ni-əs]), Petronius (Pe-ˈtro-ni-us [pe-ˈtrəʊ-ni-əs] or [pə-ˈtrəʊ-ni-əs]), Deucalion (Deu-ˈca-li-on [d(j)u:-ˈkeɪ-li-ən]), Palladium (Pal-ˈla-di-um [pə-ˈleɪ-di-əm]), Pausanias (Pau-ˈsa-ni-as [pɔ:-ˈseɪ-ni-əs]), alias (ˈa-li-as [ˈeɪ-li-əs]), Utica (ˈU-ti-ca [ˈ(j)u:-ti-kə]), Phaëthon (ˈPha-e-thon [ˈfeɪ-ə-θən]), Niobe (ˈNi-o-be [ˈnaɪ-ə-bi]).
(b) The vowels i, e, and o are long when in final position (although the long e can be IPA [i:] or [i]), e.g., Io (ˈI-o [ˈaɪ-əʊ]), Parmenio (Par-ˈme-ni-o [pɑ:-ˈmi:-ni-əʊ]), Circe (ˈCir-ce [ˈsɜ:-si]), Phoebe (ˈPhoe-be [ˈfi:-bi]), sine die (ˈsi-ne ˈdi-e [ˈsaɪ-ni ˈdaɪ-i:]), alibi (ˈal-i-bi [ˈæl-ɪ-baɪ]).Footnote 58
(c) In open, unstressed position before another vowel, a, e, or o can be long, for euphony, e.g., Laomedon (La-ˈom-e-don [leɪ-ˈɒm-ə-dən]), Laertes (La-ˈer-tes [leɪ-ˈɜ:-ti:z]), Danaus (ˈDan-a-us [ˈdæn-eɪ-əs]), Heroopolis (ˌHer-o-ˈop-o-lis [ˌher-əʊ-ˈɒp-ə-lɪs]), Meroë (ˈMer-o-e [ˈmer-əʊ-i:]), Arsinoë (Ar-ˈsin-o-e [ɑ:-ˈsɪn-əʊ-i:]).
(d) An e is long in a final, closed syllable ending in s, e.g., Hades (ˈHa-des [ˈheɪ-di:z]), Aristophanes (ˌAr-is-ˈtoph-a-nes [ˌær-ɪs-ˈtɒf-ə-ni:z]).
(e) When a or e occurs as a single word in Latin phrases, it is pronounced long, e.g., a priori (a pri-ˈo-ri [eɪ praɪ-ˈɔ:-raɪ]), a posteriori (a pos-ˌte-ri-ˈo-ri [eɪ pɒs-ˌtɪə-ri-ˈɔ:-raɪ]), e pluribus unum (e ˈplu-ri-bus ˈu-num [i: ˈplu:-ri-bəs ˈ(j)u:-nəm]).
(f) (i) The o in the Latin preposition and adverb ‘post’ and its compound verbs was pronounced long, e.g., postpono (post-ˈpo-no [ˈpəʊst-ˈpəʊ-nəʊ]).Footnote 59
(ii) In certain compound verbs, a prefix that normally has a short vowel will keep the short vowel in pronunciation, e.g., redeo (ˈred-e-o [ˈred-i-əʊ]), ineo (ˈin-e-o [ˈɪn-i-əʊ]).Footnote 60
3.3 Short vowels
(a) Vowels are short when in a closed, stressed syllable, e.g., Appius (ˈAp-pi-us [ˈæp-i-əs]), Cunaxa (Cu-ˈnak-sa [k(j)u:-ˈnæk-sə]), anno Domini (ˈan-no ˈdom-i-ni [ˈæn-əʊ ˈdɒm-ɪ-naɪ]), animus (ˈan-i-mus [ˈæn-ɪ-məs]), Demophon (ˈDem-o-phon [ˈdem-ə-fɒn] or [ˈdem-ə-fən]), Megabyzus (ˌMeg-a-ˈby-zus [ˌmeg-ə-ˈbaɪ-zəs]), Vettius (ˈVet-ti-us [ˈvet-i-əs]), Tibullus (Ti-ˈbul-lus [tɪ-ˈbᴧl-əs]), Catullus (Ca-ˈtul-lus [kə-ˈtᴧl-əs]), Tullia (ˈTul-li-a [ˈtᴧl-i-ə]), Dolabella (ˌDol-a-ˈbel-la [ˌdɒl-ə-ˈbel-ə]), Domitilla (ˌDom-i-ˈtil-la [ˌdɒm-ɪ-ˈtɪl-ə]), Philocrates (Phi-ˈloc-ra-tes [fɪ-ˈlɒk-rə-ti:z]), Lycidas (ˈLyc-i-das [ˈlɪs-ə-dæs] or [ˈlɪs-ə-dəs]), Cicero (ˈCic-e-ro [ˈsɪs-ə-rəʊ]), Romulus (ˈRom-u-lus [ˈrɒm-jʊ-ləs]), Seneca (ˈSen-e-ca [ˈsen-ə-kə]), Nemesis (ˈNem-e-sis [ˈnem-ə-sɪs]), Lydia (ˈLyd-i-a [ˈlɪd-i-ə]), Pericles (ˈPer-i-cles [ˈper-ɪ-kli:z]), Thucydides (Thu-ˈcyd-i-des [θju-ˈsɪd-ɪ-di:z]).
(b) In the words tibi, sibi, ibi, quibus, tribus, Paris, eram, eras, erat (and the other imperfect and future forms of the verb ‘sum’), the penultimate vowel was pronounced short, an exception to the normal rule.Footnote 61
3.4 Unstressed vowels
(a) The vowels a, e, o, and i are generally short or obscure, when in medial, open and unstressed syllables, and are often merely pronounced as the schwa [ə], e.g., Salamis (ˈSal-a-mis [ˈsæl-ə-mɪs]), Molossus (Mo-ˈlos-sus [mə-ˈlɒs-əs]), Orion (O-ˈri-on [ə-ˈraɪ-ən]), Persepolis (Per-ˈsep-o-lis [pɜ-ˈsep-ə-lɪs]).
(b) In open, unaccented syllables, u generally has the sound of IPA [u] or [ʊ],Footnote 62 e.g., Lucretius (Lu-ˈcre-ti-us [lju-ˈcri:-ʃəs]), Bibulus (ˈBib-u-lus [ˈbɪb-ju-ləs] or [ˈbɪb-jʊ-ləs]), nebula (ˈneb-u-la [ˈneb-ju-lə] or [ˈneb-jʊ-lə]), Corbulo (ˈCor-bu-lo [ˈkɔ:-bu-ləʊ]).
(c) The final a is pronounced as the schwa [ə]. This convention has existed for over two centuries: ‘A, ending an unaccented syllable, has the same obscure sound as in the same situation in English words’ (Walker, Key, 23),Footnote 63 e.g., Electra (E-ˈlec-tra [ɪ-ˈlek-trə]), Atalanta (ˌAt-a-ˈlan-ta [ˌæt-ə-ˈlæn-tə]), Macedonia (ˌMac-e-ˈdo-ni-a [ˌmæs-ɪ-ˈdəʊ-ni-ə]).
4. An a in the combination quadr-
When a occurs in the Latinate prefix quadr-, it often has the sound of IPA [ɒ] (the short ‘o’ as in ‘not’), though this occurs more frequently in common nouns rather than in proper names, e.g., quadra (ˈquad-ra [ˈkwɒd-rə]), quadratus (qua-ˈdra-tus [kwɒ-ˈdreɪ-təs]).Footnote 64
5. The i or y
Note that Classical Latin y was a vowel representing the Greek upsilon (υ), and had roughly the sound of the German u with umlaut (IPA [y]).Footnote 65 In English, however, y is treated exactly as if it were an i, so y = i in all instances. Thus a long ‘y’ = a long ‘i’ and a short ‘y’ = a short ‘i’.
5.1 Intermediate i/y
In an unstressed, open and medial syllable, when followed by a vowel, i or y generally has an intermediate sound represented by IPA [i]. Walker himself noted that ‘the last syllable but one of Fabii, the Horatii, Curiatii etc. is pronounced as if written Fa-be-i, Ho-ra-she-i, Cu-re-a-she-i’ (Walker, Key, 22), e.g., Antonius (An-ˈto-ni-us [æn-ˈtəʊ-ni-əs]), Fabius (ˈFa-bi-us [ˈfeɪ-bi-əs]), ratio (ˈra-ti-o [ˈreɪ-ʃi-əʊ]).
5.2 An i/y in an initial, unstressed and open position
(a) With a vowel beginning the next syllable
An i or y in an initial, unstressed and open syllable, when followed by a vowel in the next syllable, is usually long, e.g., Iambe (I-ˈam-be [aɪ-ˈæm-bi]), Dione (Di-ˈo-ne [daɪ-ˈəʊ-ni]).
(b) With a consonant beginning the next syllable
When a consonant (or a mute and liquid) follows the i or y in the next syllable, there was originally no consistent rule for determining the length of the vowel in such instances, a state of affairs which Walker (Key, 23) noted at the end of the eighteenth century:
[the i or y] is sometimes long and sometimes short, when ending an initial syllable not under the accent, as Ly-cur-gus, pronounced with the first syllable like lie … and Lysimachus with first syllable like the first of legion.Footnote 66
Even at the end of the nineteenth century, it was admitted that consistency was still absent.Footnote 67
Arguments are sometimes heard in Classics departments in which it is asserted that the length of an i or y in this position in a Latin proper name when pronounced in English should be determined by its quantity in Latin (e.g., that there should really be a short i in the first syllable of Tiberius, because the i is short in Latin). But this type of argument is quite misplaced, for in English pronunciation euphony and convention have traditionally governed the length of an i or y in such a position. Hence tradition gives us Tiberius (Ti-ˈbe-ri-us [taɪ-ˈbɪə-ri-əs]), Hymettus (Hy-ˈmet-tus [haɪ-ˈmet-əs]), Lyceum (Ly-ˈce-um [laɪ-ˈsi:-əm]), Lycurgus (Ly-ˈcur-gus [laɪ-ˈkɜ-gəs]), Lysander (Ly-ˈsan-der [laɪ-ˈsæn-də]), Isocrates (I-ˈsoc-cra-tes [aɪ-ˈsɒk-rə-ti:z]), Lysistrata (Ly-ˈsis-tra-ta [laɪ-ˈsɪs-trə-tə]), but Vitellius (Vi-ˈtel-li-us [vɪ-ˈtel-i-əs]), Tibullus (Ti-ˈbul-lus [tɪ-ˈbᴧl-əs]), Philostratus (Phi-ˈlos-tra-tus [fɪ-ˈlɒs-trə-təs]), and Vitruvius (Vi-ˈtru-vi-us [vɪ-ˈtru:-vi-əs]).Footnote 68
6. An i as a semivowel: accented ai, ei, oi, and yi
6.1 Original rule
When the combinations ai, ei, oi, or yi take the accent on the first vowel and are then followed by another vowel, the i changes to the semivowel [j] (the yod, as in ‘yellow’) and the first vowel becomes long,Footnote 69 e.g., Veii (ˈVe-ii [ˈvi:-jaɪ]), Achaia (A-ˈcha-ia [ə-ˈkeɪ-jə]),Footnote 70 Pompeius (Pom-ˈpe-ius [pɒm-ˈpi:-jəs]), Cocceius (Coc-ˈce-ius [kɒk-ˈsi:-jəs]), Pleiades (ˈPle-ia-des [ˈpli:-jə-di:z]), Apuleius (ˌAp-u-ˈle-ius [ˌæp-ju-ˈli:-jəs]), Graius (ˈGra-ius [ˈgreɪ-jəs]), Dryopeia (ˌDry-o-ˈpe-ia [ˌdraɪ-ə-ˈpi:-jə]), Ilithyia (ˌIl-i-ˈthy-ia [ˌɪl-ɪ-ˈθaɪ-jə]), Veiento (Ve-ˈien-to [vi:-ˈjen-təʊ]), Maia (ˈMa-ia [ˈmeɪ-jə]).
Walker did not formulate this principle, though it was current in some words during his time.Footnote 71 It was not until the nineteenth century that this rule was properly developed, and was widely applied until the twentieth century: hence the older pronunciation of words such as Hygeia [haɪ-ˈʤi:-jə] and Pompeii [pɒm-ˈpi:-jaɪ]. However, the word ‘Pleiades’ has always been an important exception to this rule, and the pronunciation [ˈplaɪ-ə-di:z] has been used alongside [ˈpli:-jə-di:z] since the eighteenth century.Footnote 72
6.2 After the late nineteenth century, the IPA [j] came to be simply elided in modern pronunciation of such words, e.g., Hygeia [haɪ-ˈʤi:-ə], prosopopeia [ˌprɒs-ə-pə-ˈpi:-ə], mythopoeia [ˌmɪθ-ə-ˈpi:-ə].
7. Lengthening and diphthongisation of a, e, i/y, o, and u before r
The combination of a vowel plus r before a consonant, or sometimes even before a vowel, often changes the quality of the vowel (e.g., diphthongises the e or a).
7.1 The ar before a consonant
(a) When ar comes before any consonant other than r, when the a is accented and sometimes when it is unaccented, the a is pronounced as IPA [ɑ:] (r is silent), e.g., Marcellinus (Mar-cel-ˈli-nus [ˌmɑ:-sə-ˈlaɪ-nəs]), Hipparchus (Hip-ˈpar-chus [hɪ-ˈpɑ:-kəs]), Parthenius (Par-ˈthe-ni-us [pɑ:-ˈθi:-ni-əs]), Sparta (ˈSpar-ta [ˈspɑ:-tə]).
(b) Pronunciation of a in an open syllable
In cases where a should be long in an open, accented syllable, but is followed in the next syllable by r plus another vowel, English tends to pronounce the a as the diphthong [eə], e.g., Apollinarus (A-ˌpol-li-ˈna-rus [ə-ˌpɒl-ɪ-ˈneə-rəs]), Marius (ˈMa-ri-us [ˈmeə-ri-əs but now often ˈmɑ:-ri-əs]Footnote 73 ), lares (ˈla-res [ˈleə-ri:z]), Pharos (ˈPha-ros [ˈfeə-rɒs]), Charon (ˈCha-ron [ˈkeə-rɒn]Footnote 74 ), denarius (de-ˈna-ri-us [dɪ-ˈneə-ri-əs]).
7.2 The e, i/y, or u followed by r
When e, i/y, or u are followed by r in a closed syllable and are followed by any consonant except r, always when accented, though often when not under the accent, the vowels are pronounced as IPA [ɜ:] (r is silent), e.g., Pergamum (ˈPer-ga-mum [ˈpɜ:-gə-məm]), Euterpe (Eu-ˈter-pe [ju-ˈtɜ:-pi]);Footnote 75 Curtius (ˈCur-ti-us [ˈkɜ:-ʃ(i)-əs]), Turnus (ˈTur-nus [ˈtɜ:-nəs]), Dirce (ˈDir-ce [ˈdɜ:-si]), Hyrcania (Hyr-ˈca-ni-a [hɜ:-ˈkeɪ-ni-ə]).
7.3 Pronunciation of e in an open syllable
When a long e in an open, accented syllable is followed in the next syllable by r plus a vowel, the e is pronounced as IPA [ɪə], e.g., Tiberius (Ti-ˈbe-ri-us [taɪ-ˈbɪə-ri-əs]), Ceres (ˈCe-res [ˈsɪə-ri:z]), Chimera (Chi-ˈme-ra [kaɪ-ˈmɪə-rə] or [kɪ-ˈmɪə-rə]).
7.4 Pronunciation of o
(a) When followed by a consonant other than r, an or is pronounced with the long vowel IPA [ɔ:] (r is silent), e.g., Orcus (ˈOr-cus [ˈɔ:-kəs]), Hortensius (Hor-ˈten-si-us [hɔ:-ˈten-si-əs]).
(b) In cases where o is long in an open, accented syllable, but followed in the next syllable by r plus a vowel, the o is pronounced as IPA [ɔ:], e.g., Apollodorus (A-ˌpol-lo-ˈdo-rus [ə-ˌpɒl-ə-ˈdɔ:-rəs]), Sertorius (Ser-ˈto-ri-us [sɜ:-ˈtɔ:-ri-əs]).
C. Diphthongs
In English, the Latin diphthongs are generally pronounced as one of the long vowels.Footnote 76
8. ae and oe
Two approaches to the pronunciation of the diphthongs ae and oe have existed over the past two centuries.
8.1 Original rule
(a) The original rule expounded by John Walker was that both oe and ae should follow the example of e, in every position.Footnote 77 Hence an ae or oe in an open, stressed syllable was to be pronounced as a long e or IPA [i:], e.g., Laetus (ˈLae-tus [ˈli:-təs]), Croesus (ˈCroe-sus [ˈkri:-səs]), Naevius (ˈNae-vi-us [ˈni:-vi-əs]), Philopoemen (ˌPhil-o-ˈpoe-men [ˌfɪl-ə-ˈpi:-mən]).
(b) An ae or oe in a closed, stressed syllable was pronounced like a short e (IPA [e] as in ‘bet’), e.g., Oedipus (ˈOed-i-pus [ˈed-ɪ-pəs]), Daedalus (ˈDaed-a-lus [ˈded-ə-ləs]), Aeschines (ˈAes-chi-nes [ˈes-kɪ-ni:z]), Aeschylus (ˈAes-chy-lus [ˈes-kɪ-ləs]).
8.2 Ae and oe long in all positions
The system of 8.1 prevailed in the early and middle eighteenth century in Britain and America. But, towards the end of the century, British Classical scholars contended that the Latin diphthongs ae and oe were to be vocalised as IPA [i:] in almost every position, e.g., Oedipus [ˈi:d-ɪ-pəs], Daedalus [ˈdi:d-ə-ləs], Aeschines [ˈi:s-kɪ-ni:z], Aeschylus [ˈi:s-kɪ-ləs]), Maecenas ([mi:-ˈsi:-nəs] or [mi:-ˈsi:-næs]), quaestor ([ˈkwi:s-tə] for [ˈkwes-tə]).Footnote 78 Beginning in the 1880s, this reform was part of the ‘Victorian’ English pronunciation of Latin, and therefore sought to mark Classical quantities or diphthongs with English sounds (see Section VI below). The new sound became common in the 1890s, and then standard English in the early twentieth century.Footnote 79 Thus a divergence between British and American pronunciation occurred – one which is still extant. But some Victorian traditionalists were strongly opposed to the change. As early as the 1880s, conservative opinion railed against the new pronunciation.Footnote 80 In America it was derided by scholars, amongst whom was one Professor George Hempl. He declared that
[the long e] has, indeed, already introduced into English learned circles such absurdities as Æschylus and Œdipus [ˈi:s-kɪ-ləs and ˈi:d-ɪ-pəs], and, what is more, it has even tainted the pronunciation of learned English words – witness those linguistic pearls æsthetics and asafœtida.Footnote 81
Nevertheless, the process that he disparaged did indeed give us the customary long e in aesthete [ˈi:s-θi:t] and aesthetic [ˈi:s-θet-ɪk], and other words,Footnote 82 pronunciations which are not standard in America, where a short e still predominates.Footnote 83 But British English now uses IPA [i:] for these Latin diphthongs, especially when they are in accented syllables.
9. au
9.1 In English, the Latin diphthong au has customarily been pronounced as IPA [ɔ:], even when it occurs in an unstressed syllable,Footnote 84 e.g., Augusta (Au-ˈgus-ta [ɔ:-ˈgᴧs-tə]), Faustinus (Faus-ˈti-nus [fɔ:s-ˈtaɪ-nəs]).
9.2 The au of the Greek suffix -aus in proper names is not treated as a long vowel in English; rather, the two vowels are pronounced distinctly, e.g., Menelaus (ˌMen-e-ˈla-us [ˈmen-ə-ˈleɪ-əs]).
10. eu
10.1 Eu as long u
The Latin diphthong is pronounced as IPA [ju:] (as in ‘mute’), even when it occurs in a closed or unstressed syllable,Footnote 85 e.g., Eudoxus (Eu-ˈdok-sus [ju:-ˈdɒk-səs]), Eustathius (Eus-ˈta-thi-us [ju:s-ˈtei-θi-əs]).
10.2 The Greek suffix -eus in Classical proper names
(a) Original rule
The Greek suffix -ευς , with the epsilon-plus-upsilon diphthong, appearing in Latin proper names, was originally pronounced with separate vowels in English,Footnote 86 e.g., Orpheus (ˈOr-phe-us [ˈɔ:-fi-əs]), Odysseus (O-ˈdys-se-us [ə-ˈdɪs-i-əs]).
This was the rule of Walker, and it has continued into modern times.
(b) Suffix -eus with a long u
From the nineteenth century, the tendency arose to pronounce the eu as IPA [ju:],Footnote 87 e.g., Orpheus (ˈOr-phe-us [ˈɔ:-fju:s]), Odysseus (O-ˈdys-se-us [ə-ˈdɪs-ju:s]).
This change arose as part of the ‘Victorian’ reform of English pronunciation of Latin (see Section VI below), to mark the Classical diphthong. Though this pronunciation is still widespread in modern British English, it has always existed alongside the alternative and original pronunciation by which the vowels are separate.
D. Greek-derived diphthongs
11. ai (αι)
11.1 ai as English long i
(a) Since the alpha-and-iota diphthong in Greek was regularly transcribed as ae in Latin,Footnote 88 no difficulty regarding its pronunciation normally arises. Very rarely the diphthong, or an ai from separate Greek vowels, is treated as a diphthong, when before another vowel, and given the sound of [aɪ], a sound which was once the value for the alpha-plus-iota diphthong in the English system for pronouncing Greek,Footnote 89 e.g., Achaia (A-ˈchai-a [ə-ˈkaɪ-ə], from Ἀχαΐα, where ai is mistaken for the Greek diphthong).
11.2 For the alternative way of pronouncing words of this type, see rule 6 above.
12. ei (ει)
The Greek epsilon-and-iota diphthong occasionally appears as ei in some Latinised words, rather than as the more usual Latin transcription, a long e or i. The ei can be pronounced as a long i (as in ‘fight’ or IPA [ai]), a sound used for ei in the traditional English system for pronouncing ancient Greek,Footnote 90 e.g., Elegeia (ˌEl-e-ˈgei-a [ˌel-ə-ˈʤai-ə]), Pleiades (ˈPlei-a-des [ˈplai-ə-di:z]).
For the alternative way of pronouncing words of this type, see rule 6 above.
It is curious that, though the long i (IPA [ai]) for Greek-derived ei was common even in the eighteenth century, Walker condemned it as a scholarly eccentricity: ‘scholars who are fond of displaying their knowledge of Greek, will be sure to pronounce Elegeia, Hygeia … Antheia, Deiopeia, with the diphthong like the noun eye’.Footnote 91 From the late nineteenth century, there began a tendency to sometimes write the Greek-derived diphthong ei in Classical proper names, e.g., Cleitarchus for Clitarchus (both pronounced as [ˈklai-tɑ:-kəs]), Cleitus for Clitus (both vocalised as [ˈklai-təs]), Heracleitus [ˌher-ə-ˈklai-təs], Peisistratus [pai-ˈsɪs-trə-təs], Poseidon [pə-ˈsai-dən].
The name ‘Cleisthenes’ is now a special case. The word was normally spelt ‘Clisthenes’, and was traditionally pronounced as [ˈklɪs-θə-ni:z]. Today the vowel of the accented syllable has changed to [ai],Footnote 92 with pronunciation as [ˈklais-θə-ni:z].Footnote 93
E. Consonants
13. Soft c and g
13.1 C and g
(a) The consonant c is IPA [s] and g is IPA [ʤ] before i, y, e, ae, and oe, but hard in all other cases,Footnote 94 e.g., Cecrops (ˈCe-crops [ˈsi:-krɒps]), Gyges (ˈGy-ges [ˈʤaɪ-ʤi:z]), Leucippus (Leu-ˈcip-pus [lju:-ˈsɪp-əs]), but Gordias (ˈGor-di-as [ˈgɔ:-di-əs] or [ˈgɔ:-di-æs]), Galinthias (Ga-ˈlin-thi-as [gæ-ˈlɪn-θi-əs] or [gæ-ˈlɪn-θi-æs]).
(b) In a double c, the first c is hard when before a second soft c, e.g., Cocceius (Coc-ˈce-i-us [kɒk-ˈsi:-jəs]), Tricce (ˈTric-ce [ˈtrɪk-si:]).
(c) A gg is reduced to one soft g, if the second g is soft, e.g., Agger (ˈAg-ger [ˈæʤ-ə]), Aggenus (Ag-ˈge-nus [ə-ˈʤi:-nəs]).
13.2 Sc
Sc before i, y, e, ae, or oe is reduced to IPA [s] in pronunciation, e.g., Scylla (ˈScyl-la [ˈsɪl-ə]), Scipio (ˈScip-i-o [ˈsɪp-i-əʊ]), Sciron (ˈSci-ron [ˈsaɪ-rɒn]), but Scamander (Sca-ˈman-der [skə-ˈmæn-də]), Scopas (ˈSco-pas [ˈskəʊ-pæs]).
14. ch
The Latin ch derives from the Greek chi (χ), and in English pronunciation ch is pronounced as IPA [k],Footnote 95 e.g., Charon (ˈCha-ron [ˈkeə-rən] or [ˈkeə-rɒn]), Charybdis (Cha-ˈryb-dis [kə-ˈrɪb-dɪs]).
15. Difficult initial combinations: Bd, Cn, Chth, Ct, Gn, Mn, Ps, Pn, Pt, Phth, and Tm
15.1 A number of Latin words of Greek etymology contain difficult initial combinations of consonants, and the convention, since the time of Walker, has been to suppress the first consonant in pronunciation,Footnote 96 e.g., Bdelius (ˈBde-li-us [ˈdi:-li-əs]), Cnidus (ˈCni-dus [ˈnaɪ-dəs]), Chthonia (ˈChtho-ni-a [ˈθəʊ-ni-ə]), Ctesias (ˈCte-si-as [ˈti:-ʃ(i)-əs]), Ctesiphon (ˈCtes-i-phon [ˈtes-ɪ-fən]), Gnomon (ˈGno-mon [ˈnəʊ-mɒn] or [ˈnəʊ-mən]), Mnemosyne (Mne-ˈmos-y-ne [nə-ˈmɒz-ɪ-ni:]), Psyche (ˈPsy-che [ˈsaɪ-ki:]), Pnytagoras (Pny-ˈtag-o-ras [naɪ-ˈtæg-ə-rəs]), Phthia (ˈPhthi-a [ˈθaɪ-ə]), Tmolus (ˈTmo-lus [ˈməʊ-ləs]), Tzetzes (ˈTzet-zes [ˈzet-zi:z]).
However, tm- is sometimes an exception, e.g., tmesis [ˈtmi:-sɪs].
15.2 An exception is Tl-, which has traditionally been pronounced fully, e.g., Tlepolemus [tlə-ˈpɒl-ə-məs].
16. S
16.1 A final s after e has the sound of a z,Footnote 97 e.g., Damocles (ˈDam-o-cles [ˈdæm-ə-kli:z]).
16.2 A final s after n changes to z, Tiryns (ˈTi-ryns [ˈtaɪ-rɪnz]), Labens (ˈLa-bens [ˈleɪ-benz]).
16.3 In some cases, on the analogy of English, intervocalic s changes to IPA [z], e.g., Caesar (ˈCae-sar [ˈsi:-zə]), Arethusa (ˌAr-e-ˈthu-sa [ˌær-ə-ˈθju:-za]).Footnote 98
17. A t, c, or s as IPA [ʃ], [ʒ], [z]
When a t, c, or s is preceded by the accent, and followed by an i or y (and eu in Walker's systemFootnote 99 ) plus another vowel in the final syllable, the s can change to IPA [ʒ] or [z], the c to IPA [ʃ], and the t change into IPA [ʃ], and the i or y is sometimes elided.Footnote 100
17.1 T or c
(a) In the original rule, a t or c always changes to IPA [ʃ] (as in ‘should’), e.g., Statius (ˈSta-ti-us [ˈsteɪ-ʃəs]), Dacia (ˈDa-ci-a [ˈdeɪ-ʃ(i)-ə]), Propertius (Pro-ˈper-ti-us [prə-ˈpɜ-ʃəs]), Pontius (ˈPon-ti-us [ˈpɒn-ʃəs]), Latium (ˈLa-ti-um [ˈleɪ-ʃ(i)-əm]), Cappadocia (ˌCap-pa-ˈdo-ci-a [ˌkæp-ə-ˈdəʊ-ʃ(i)-ə]), dementia (de-ˈmen-ti-a [də-ˈmen-ʃə]), ex gratia (ex ˈgra-ti-a [eks ˈgreɪ-ʃ(i)-ə]), minutiae (mi-ˈnu-ti-ae [mɪ-ˈnju:-ʃi-i:] or [maɪ-ˈnju:-ʃ(i)-i:]).
(b) Some names, and often Latin phrases or nouns, were excluded from this rule in the course of the twentieth century, owing to the influence of the Restored pronunciation, e.g., Byzantium [bɪ-ˈzæn-ti-əm] or [baɪ-ˈzæn-ti-əm] for [baɪ-ˈzæn-ʃəm],Footnote 101 Actium [ˈæk-ti-əm] for older [ˈæk-ʃəm], a fortiori [eɪ ˌfɔ:-ti-ˈɔ:-raɪ] for older [eɪ ˌfɔ:-ʃi-ˈɔ:-raɪ].
(c) In other aberrant cases, c is sometimes changed to an s, e.g., Caecias [ˈsi:-si-əs] for older [ˈsi:-ʃ(i)-əs], Lycia [ˈlɪs-i-ə] for older [ˈlɪʃ-(i)-ə].
17.2 The S
(a) Original rule
When s is preceded immediately by an accented vowel and is followed by i/y plus another vowel in the final syllable, the s was originally changed to the sound of IPA [ʒ],Footnote 102 e.g., Artemisia (ˌAr-te-ˈmis-i-a [ˌɑ:-tə-ˈmɪʒ-i-ə]), Aspasia (As-ˈpa-si-a [æs-ˈpeɪ-ʒ(i)-ə]).
In the late nineteenth and early twentieth centuries, this was superseded by (b).
(b) The more recent tendency has been to change the s to a z, e.g., Aspasia (As-ˈpa-si-a [æs-ˈpeɪ-zi-ə]), Elysium (E-ˈlys-i-um [ɪ-ˈlɪz-i-əm]).
This is now the general rule in English pronunciation of Classical names.
(c) In some cases, as Walker noted, an s sometimes changes to IPA [ʃ], particularly when preceded by a consonant, e.g., Asia (ˈA-si-a [ˈeɪ-ʃə]), Persia (ˈPer-si-a [ˈpɜ-ʃə]), Athanasius (ˌAth-a-ˈna-si-us [ˌæθ-ə-ˈneɪ-ʃəs]).
Note that double s was once governed by this rule, e.g., Cassius (ˈCas-si-us [ˈkæʃ-əs]).
However, British English now excludes the double s from this rule,Footnote 103 e.g., Cassius (ˈCas-si-us [ˈkæs-i-əs]), Cassiopeia (ˌCas-si-o-ˈpe-i-a [ˌkæs-i-ə-ˈpi-(j)ə]), Cassiodorus (ˌCas-si-ˈdo-rus [ˌkæs-i-ə-ˈdɔ:-rəs]).
In some few words, the [ʃ] in such a position survives in American pronunciation.Footnote 104
17.3 Traditional exception: t unchanged
(a) A t is unchanged after s, t, or x. E.g., Sestius (ˈSes-ti-us [ˈses-ti-əs]), Sextius (ˈSek-sti-us [ˈsek-sti-əs]), Mettius (ˈMet-ti-us [ˈmet-i-əs]), Ostia (ˈOs-ti-a [ˈɒs-ti-ə]), Hestia (ˈHes-ti-a [ˈhes-ti-ə]), Attius (ˈAt-ti-us [ˈæt-i-əs]).
(b) Note that Greek derivatives terminating in -tion or -tyon did not originally conform to this rule, but took the normal sound of t (as in ‘rate’), e.g., Gration (ˈGra-ti-on [ˈgreɪ-ti-ən]), Metion (ˈMe-ti-on [ˈmi:-ti-ən]), Eurytion (Eu-ˈryt-i-on [ju-ˈrɪt-i-ən]), Polytion (Po-ˈlyt-i-on [pə-ˈlɪt-i-ən]), Amphictyon (Am-ˈphic-ty-on [æm-ˈfɪk-ti-ən]), and Sotion (ˈSo-ti-on [ˈsəʊ-ti-ən]).Footnote 105
Two nineteenth-century exceptions were Theodotion [ˌθi:-ə-ˈdəʊ-ʃən] and Hephaestion. The name ‘Hephaestion’ was first pronounced as to rhyme with question, i.e., [he-ˈfes-ʧən],Footnote 106 but later as [he-ˈfes-ti-ən]. In the late nineteenth century, the pronunciation changed to [he-ˈfi:s-ti-ən], by 8.2 above.
18. x
18.1 Initial x
An initial x is pronounced as a z,Footnote 107 e.g., Xenophanes (Xe-ˈnoph-a-nes [ze-ˈnɒf-ə-ni:z]).
18.2 Medial x
(a) Medial x is pronounced as the double consonant ks: Xerxes (ˈXer-xes [ˈzɜk-si:z]).
(b) When the Latinate prefix ex precedes an accented vowel, it sometimes has the sound of [gz] (especially in Latin common nouns), e.g., exemplum (eg-ˈzem-plum [ɪg-ˈzem-pləm] or [eg-ˈzem-pləm]), exempli gratia (eg-ˈzem-pli ˈgra-ti-a [eg-ˈzem-plaɪ ˈgreɪ-ʃ(i)-ə]). In other cases, the x is still pronounced as ks: exeat (ˈek-se-at [ˈek-si-æt]), exeunt (ˈek-se-unt [ˈek-si-ᴧnt]).
VI. The late nineteenth century
It was, curiously, not until the last decade of the nineteenth century that serious study of the original Anglicised system for Latin was undertaken, and even then this was done because of the threat posed to it by the new reforms and competing systems.Footnote 108 For the universality of the dual English system – one for proper names and a very similar one for nouns, phrases or texts – was irreparably broken in the late nineteenth century, when different systems of pronouncing Latin came into prominence. By 1890, the Restored, or Classical, systemFootnote 109 had been introduced, but was not widely adopted for the secular study of Classics until the early twentieth century. The Restored system, however, was only one of a welter of methods, which included
-
(1) the Continental method,
-
(2) the ‘Roman’ or modern Italian method (modern Church Latin),
-
(3) the traditional English method, and
-
(4) the ‘Victorian’ method, which marked some Latin quantities with English vowel sounds.Footnote 110
We should first consider the role of Latin in British churches, for methods (1) and (2) are best discussed in this context. In the small Catholic Church in Britain, the English pronunciation of Latin had always had a rival in the peculiar ‘Continental’ system which adopted English sounds for the consonants, but so-called ‘Continental’ sounds for the vowels.Footnote 111 The scholar F. Brittain ascribed the form of the Continental method to the conservatism of the English Catholic clergy, who had retained some of the Middle English vowel qualities, sounds which were similar to those of modern Italian, even if the evidence for this theory is not conclusive.Footnote 112
But, by around 1871, Catholic communities in Britain had adopted a new method, owing to the zeal of newly converted English Catholics who adopted the ‘Roman’, or modern Italian,Footnote 113 system, a process which was strongly opposed by Catholic traditionalists.Footnote 114 We should also note here that – though it is frequently forgotten – modern Church Latin did not exist outside Italy as a universal ecclesiastical, or international, pronunciation of Latin until the early twentieth century.Footnote 115 Thus it was only in 1870, with the First Vatican Council, that there occurred significant impetus for the adoption of the Italian pronunciation of Latin in the wider Catholic Church. Some years later, Pope Pius X (1903–1914) strongly encouraged this trend, both in his Motu proprio of 1903, which dealt with pronunciation of sacred music, and in a letter to the Archbishop of Bourges of 1912.Footnote 116 Although the older Continental system survived until about 1885 amongst English Benedictines (at Downside and Ampleforth) and English Jesuits,Footnote 117 modern Italian/Church Latin eventually prevailed in the Catholic Church in England.
In the Anglican Church, there was, in contrast, a strong tendency towards the use of the English pronunciation of Latin. From the sixteenth century onwards, in the singing of Anglican hymns, anthems and other rites, Latin had been pronounced in the same way as secular texts, even though, by the 1880s, the Continental system had been adopted in some services.
These opposing systems naturally led to a state of confusion – and this can be seen from reminiscences of contemporaries. When Bertrand Russell attended Cambridge in the 1890s, he found that the Vice-Master of Trinity College
stuck to the English pronunciation of Latin, while the Master adopted the Continental pronunciation. When they read grace in alternate verses, the effect was curious, especially as the Vice-Master gabbled it, while the Master mouthed it with unction.Footnote 118
By the late 1890s, the Classical system was introduced into some Anglican churches, and the Italian pronunciation of Latin into others. This ‘Church’ Latin made very great progress in the twentieth century,Footnote 119 and it eventually effected minor changes in the English pronunciation of secular Latin.
The early version of our modern Reformed system for pronouncing Latin also made its appearance in the Victorian era, after decades of vociferous debate amongst philologists about the original sounds.Footnote 120 But surprising resistance hindered attempts to reform British universities and schools.Footnote 121 Some even recoiled with horror from the sounds of Restored Latin, when it first appeared.Footnote 122
The famous reformers of the 1870s had included Professors Munro and Palmer, whose Syllabus of Latin pronunciation was published in 1872.Footnote 123 Their system was introduced into University College (London) by Robinson Ellis, but had made little progress in other institutions by 1880.Footnote 124 A second attempt at reform came in 1886 with a report by the Cambridge Philological Society which again recommended the widespread adoption of the new system. In America, the reform was accepted with comparative alacrity, and had made great progress even by 1900.Footnote 125 In 1907 the British Board of Education approved the Reformed syllabus. With the imposition of an official policy in schools funded by the government, the final victory in Britain came in the 1920s, though the adoption of the reform in public schools was not uniform or consistent.Footnote 126
But the system the Reformed pronunciation replaced in the 1920s was not the original English one. In fact, a spirit of dissatisfaction with the pure English method had been abroad from the 1870s, and some Victorian teachers of Latin prose and poetry had made an attempt to mark the long and short vowels of Classical quantities by using English vowel sounds, rather than Restored ones.Footnote 127 The new and reformed English pronunciation which resulted was later called the ‘Victorian’ system (method 4 above). A student of this system, Percy F. Frankland, recalled its chief characteristic:
I look back … to the instruction in Latin which I received at University College School under the distinguished headmastership of Thomas Hewitt Key … who was an outstanding Professor at University College, of which the school, as an adjunct, was his creation. I must be among the few surviving who had the privilege in the seventies [i.e., 1870s] of being introduced to the classical world under his inspired mentorship. The essential feature of Key's method was his insistence on a knowledge of the quantity of the vowels and its emphasis in their pronunciation, while retaining their English character [my emphasis]. The knowledge of quantity he valued as essential to a proper understanding of the etymological structure of words …. It appears to me that by the rigorous pronunciation according to quantity many of the objections urged against the ‘old’ or English pronunciation would disappear.
(Frankland (Reference Frankland1938) 8)Some examples are given below of pronunciations of Latin words under the Victorian system (VS) with original English pronunciations (OEP) following:
-
sidera. VS [ˈsaɪ-də-rə]; OEP [ˈsɪd-ə-rə].
-
nomina. VS [ˈnəʊm-ɪ-nə]; OEP [ˈnɒm-ɪ-nə].
-
stamina. VS [ˈsteɪ-mɪ-nə]; OEP [ˈstæm-ɪ-nə].
-
Lydia. VS [ˈlaɪd-i-ə]; OEP [ˈlɪd-i-ə].
-
alias. VS [ˈæl-i-əs]; OEP [ˈeɪ-li-əs].
In the first four words, in which the antepenultimate vowels had been short in the previous English system, the new pronunciation marked the long vowels in the initial syllables of the Classical Latin sidus, nomen, stamen, and Lydia. In the last example, the short Classical quantity of the initial a in alias produced the corresponding short sound in English, though this vowel had been pronounced long under the old rules.Footnote 128 A later variant of the Victorian system even adopted a consistently hard pronunciation of the consonants g and c, as in Classical Latin.Footnote 129
A trend also began to pronounce the Latin diphthongs ae and oe as [i:] (i.e., as the vowel in ‘see’ or ‘keen’). Hence the names ‘Aeschylus’ and ‘Oedipus’, which had previously been heard in British English as [ˈesk-ə-ləs] and [ˈed-ɪ-pəs], were now pronounced as the modern [ˈi:sk-ə-ləs] and [ˈi:d-ɪ-pəs]. Furthermore, the Greek diphthong ‘eu’ (epsilon plus upsilon) in the suffix -eus in Classical names was now pronounced as IPA [ju:], to mark its Classical value, even though the traditional system had pronounced the vowels of this ending separately (e.g., Orpheus was now heard as [ˈɔ:-fju:s] alongside the older [ˈɔ:-fi-əs]).
Curiously, both of these reforms survived to be applied to the pronunciation of Classical proper names in modern RP, but the general ‘Victorian’ system itself was ephemeral and, indeed, is largely forgotten today.
But the new reforms, inspired partly as they were by the criticisms of philologists – though they were an attempt to deal with the major failing of the English method, its tendency to ignore Classical quantities – quickly threw the old system into great confusion.Footnote 130 The main consequence was that many variant pronunciations arose. Hence we can understand that the sheer chaos into which the venerable English system had degenerated by c. 1900 made it an easy target for advocates of the Classical pronunciation.Footnote 131
Moreover, the renewed emphasis on pronouncing Classical vowel quantities with English sounds – and the variants which resulted – left some rather confused about the original English system and its rules, especially for proper names.Footnote 132 Thus the scholar T. E. Page, a schoolmaster, opposed the Reformed pronunciation of Latin, but thought that the name ‘Silius’ should take the diphthong [aɪ] (as in ‘high’) in the initial syllable, i.e., with pronunciation as [ˈsaɪ-li-əs], because the i was long in Latin.Footnote 133 An anecdote, which illustrates the same point, is told by Moore Smith,Footnote 134 who was a student in the 1870s. He recalled that
one master not famous for discipline, if a boy said ‘scĭlicet’ [ˈsɪl-ɪ-set], had a stock joke, ‘Well you are a silly set’. This being known, the pronunciation ‘scĭlicet’ [ˈsɪl-ɪ-set] was always used and the joke that followed greeted with uproarious applause.
This story is explicable in the following way: the master was using the new ‘Victorian’ English pronunciation of Latin, which sought to mark certain Classical quantities with English sounds. The pronunciation that the master wanted was [ˈsaɪ-li-set],Footnote 135 in order to mark the long i in the initial syllable of Classical Latin scilicet. The student gave, either intentionally or not, the long-standing older English pronunciation [ˈsɪl-i-set],Footnote 136 and was rebuked with the master's joke. The schoolmaster T. E. Page, as we saw, had used the principles of the same Victorian system in his reasoning, when he demanded [ˈsaɪ-li-əs] for normal English [ˈsɪl-i-əs].
VII. Conclusion
The English system of pronouncing Latin has a long history and its legacy is pervasive. We have seen above how the system broke down in the late nineteenth century, owing to increasing recognition of its limitations, and the confusion caused by competing systems. The transitional period in which Victorians attempted to reform their English system to mark Latin quantities without rejecting the sounds of the English language has been neglected in modern scholarship. This paper has shed new light on that era and how the Reformed pronunciation of Latin triumphed by the 1920s.
