This article has multiple issues. Please helpimprove it or discuss these issues on thetalk page.(Learn how and when to remove these messages) This articlemay containoriginal research. Pleaseimprove it byverifying the claims made and addinginline citations. Statements consisting only of original research should be removed.(May 2010) (Learn how and when to remove this message) This articleshould specify the language of its non-English content using{{lang}} or{{langx}},{{transliteration}} for transliterated languages, and{{IPA}} for phonetic transcriptions, with an appropriateISO 639 code. Wikipedia'smultilingual support templates may also be used.See why.(June 2020) This articlehas an unclearcitation style. The references used may be made clearer with a different or consistent style ofcitation andfootnoting.(November 2025) (Learn how and when to remove this message) (Learn how and when to remove this message)

Thetraditional English pronunciation of Latin, andClassical Greek words borrowed through Latin, is the way theLatin language was traditionally pronounced by speakers of English until the early 20th century. Although this pronunciation is no longer taught inLatin classes, it is still broadly used in the fields ofbiology,law, andmedicine.^[1]

In theMiddle Ages speakers of English, fromMiddle English onward, pronounced Latin not as the ancient Romans did, but in the way that had developed among speakers of French. This traditional pronunciation then became closely linked to the pronunciation of English, and as thepronunciation of English changed with time, the English pronunciation of Latin changed as well.

Until the beginning of the 19th century all English speakers used this pronunciation, includingRoman Catholics for liturgical purposes.^[2] FollowingCatholic emancipation in Britain in 1829 and the subsequentOxford Movement, newly converted Catholics preferred theItalianate pronunciation, which became the norm for the Catholic liturgy. Meanwhile, scholarly proposals were made for a reconstructedClassical pronunciation, close to the pronunciation used in the late Roman Republic and early Empire, and with a more transparent relationship between spelling and pronunciation.

One immediate audible difference between the pronunciations is in the treatment of vowels. The English pronunciation of Latin appliedvowel sound changes which had occurred within English itself, where stressed vowels in a word became quite different from their unstressed counterpart. In the other two pronunciations of Latin, vowel sounds were not changed. Among consonants, for example, the treatment of the letterc followed by afront vowel was one clear distinction. That is, the nameCicero is pronounced in English as/ˈsɪsəroʊ/SISS-ə-roh, inEcclesiastical Latin as[ˈtʃitʃero], and in restoredClassical Latin as[ˈkɪkɛroː].

The competition between the three pronunciations grew towards the end of the 19th century. By the beginning of the 20th century, however, a consensus for change had developed. TheClassical Association, shortly after its foundation in 1903, put forward a detailed proposal for a reconstructed classical pronunciation. This was supported by other professional and learned bodies. Finally in February 1907 their proposal was officially recommended by theBoard of Education.^[3][4] Adoption of the "new pronunciation" was a long, drawn-out process,^[5] but by the mid-20th century, classroom instruction in the traditional English pronunciation had ceased.

The traditional pronunciation survives in academic and general English vocabulary:

• In a very large body of words used every day:album, apex, area, asylum, axis, basis, bonus, camera, census, cinema, circus, crisis, dilemma, error, focus, genesis, genius, hypothesis, icon, insignia, item, junior, major, medium, minor, murmur, onus, panacea, podium, ratio, sector, stamina, terminus, trivia; as well as such common phrases asad infinitum, et cetera, non sequitur, quid pro quo, status quo, vice versa, etc.
• In academic vocabulary:campus, syllabus, curriculum, diploma, alumnus
• In anatomical vocabulary:aorta, biceps, cranium, patella, penis, sinus, vertebra, vagina, etc.
• In astronomical nomenclature, including the names of planets, moons, asteroids, stars and constellations, such asMars, Io, Ceres, Sirius, Ursa Major, nova, nebula, though many of these are irregular (e.g. expected/ˈlaɪbrə/ for Libra is less common than irregular/ˈliːbrə/)
• In many biblical names:Ananias, Cornelius, Felix, Jesus, Judas, Lydia, Nicodemus, Nicolas, Priscilla, Sergius, Silas, Titus, Zacharias, etc.
• In a number of historical terms and names, particularly those associated with Greek or Roman culture and politics:augur, bacchanal, consul, fibula, lictor, prætor, toga, Augustus, Cæsar, Cicero, Diocletian, Hypatia, Plato, Socrates, Trajan, etc.
• In legal terminology and phrases:affidavit, alibi, alias, de jure, obiter dictum, sub judice, subpoena, etc. In many cases Classical pronunciation is used, however.
• In the specialized terminology of literary studies:codex, colophon, epitome, index, periphrasis, parenthesis, etc.
• In some mathematical terms:calculus, parabola, hyperbola, isosceles, rhombus, vector, etc.
• In medical terminology describing diseases, symptoms and treatments:anaesthesia, bacterium, coma, lumbago, mucus, nausea, ophthalmia, rabies, tetanus, virus, rigor mortis. etc.
• In words and names from classical mythology:Achilles, Argus, Calliope, Gorgon, Myrmidon, Sphinx, etc.
• In many place names:Carolina, Judæa, Annapolis, Dalmatia, Ithaca, Nicæa, Pennsylvania, Romania, Salina, Virginia, etc.
• In some religious terms:Angelus, basilica, Magi, martyr, presbyter, etc.
• In many saints' names:Athanasius, Eugenia, Eusebius, Ignatius, Irene, Januarius, Leo, Macarius, Marcella, Theophilus, etc.
• In certain sporting terms:gymnasium, stadium, discus, pentathlon
• In the taxonomic nomenclature of botany and zoology:phylum, genus, species, chrysanthemum, hibiscus, rhododendron, foetus, larva, ovum, pupa, chameleon, lemur, platypus

In most cases, the English pronunciation of Classical words and names is predictable from the orthography, as long aslong and short vowels are distinguishable in the source. For Latin, Latinized Greek or for long versus short α, ι, υ Greek vowels, this means thatmacrons andbreves must be used if the pronunciation is to be unambiguous. However, the conventions ofbiological nomenclature forbid the use of these diacritics, and in practice they are not found in astronomical names or in literature. Without this information, it may not be possible to ascertain the placement ofstress, and therefore the pronunciation of the vowels in English.

Note that the following rules are generalizations, and that many names have well-established idiosyncratic pronunciations.

Latin stress is predictable. It falls on thepenultimate syllable when that is "heavy", and on theantepenultimate syllable when the penult is "light".

In Greek, stress isnot predictable, but it may be ignored when pronouncing Greek borrowings, as they have been filtered through Latin and have acquired the stress patterns of Latin words.

A syllable is "light" if it ends in a single short vowel. For example,a, ca, sca, scra are all light syllables for the purposes of Latin stress assignment.

Any other syllable is "heavy":

• if it is closed (ended) by a consonant:an, can, scan, scran
• if the vowel is long or adiphthong in Latin, or in the Latin transliteration of Greek:ā, cā, scā, scrā (a long vowel) oræ, cæ, scæ, scræ (a diphthong).

Latin diphthongs may be written⟨æ⟩ or⟨ae⟩,⟨œ⟩ or⟨oe⟩. Long vowels are written with amacron:ā ē ī ō ū ȳ, though this is a modern convention. Greek long vowels are ει, η, ου, ω, sometimes ι, υ, and occasionally α. (Long α is uncommon.) For example,Actaeon is pronounced/ækˈtiːɒn/ak-TEE-on or/ækˈtiːən/ak-TEE-ən. Adiaeresis indicates that the vowels donot form a diphthong:Arsinoë/ɑːrˈsɪnoʊiː/ar-SIN-oh-ee (not *AR-sin-ee).

The importance of marking long vowels for Greek words can be illustrated withIxion, from Greek Ἰξίων. As it is written, the English pronunciation might be expected to be */ˈɪksiɒn/IK-see-on. However, length marking,Ixīōn, makes it clear that it should be pronounced/ɪkˈsaɪɒn/ik-SY-on.

When a consonant ends a word, or when more than a single consonant follows a vowel within a word, the syllable is closed and therefore heavy. (Aconsonant is not the same thing as a letter. The lettersx[ks] andz[dz] each count as two consonants, butth[θ],ch[k], andph[f] count as one, as the pronunciations in brackets indicate.) The English letterj was originally ani, forming a diphthong with the preceding vowel, so it forces the stress just asæ, œ, z, andx do.

• Exception: aconsonant cluster ofp,t, orc/k plusl orr is ambiguous. The preceding syllable was open in ordinary Latin. However, when a different stress placement was required by poetic meter, it could be treated as closed. Thus the nameChariclo(Chariklō) could besyllabified as eithercha-ri-klō, with an open penult and stress on thecha (normal Latin), orcha-rik-lō, with a closed penult and stress on therik (an optional poetic pronunciation), so both/ˈkærɪkloʊ/KARR-ik-loh and/kəˈrɪkloʊ/kə-RIK-loh are acceptable pronunciations in English.^[6]

If more than two syllables precede the stressed syllable, the same rules determine which is stressed. For example, inCassiopeia (also Cassiopēa), syllabifiedcas-si-o-pei-a, the penultpei/pē contains a long vowel/diphthong and is therefore stressed. The second syllable preceding the stress,si, is light, so the stress must fall one syllable further back, oncas (which coincidentally happens to be a closed syllable and therefore heavy). Therefore, the standard English pronunciation is/ˌkæsiəˈpiːə,-sioʊ-/KAS-ee-ə-PEE-ə, -⁠see-oh-.^[7] (Note however that this word also has an irregular pronunciation in English:/ˌkæsiˈoʊpiə/KAS-ee-OH-pee-ə.)

Whether a vowel letter is pronounced "long" in English (/eɪ,iː,aɪ,oʊ,juː/) or "short" (/æ,ɛ,ɪ,ɒ,ʌ/) is unrelated to the length of the original Latin or Greek vowel. Instead it depends on position and stress. A vowel followed by a consonant at the end of a word is short in English, except that final-es is always long, as inPales/ˈpeɪliːz/PAY-leez. In the middle of a word, a vowel followed by more than one consonant is short, as inHermippe/hərˈmɪpi/hər-MIP-ee, while a vowel with no following consonant is long. However, when a vowel is followed by a single consonant (or by a cluster ofp, t, c/k plusl, r) and then another vowel, it gets more complicated.

• If the syllable is unstressed, it is open, and the vowel is often reduced toschwa.
• If the penultimate syllable is stressed, it is open and the vowel long, as inEuropa/jʊəˈroʊpə/yoor-OH-pə.
• If any other syllable is stressed, it is closed and the vowel is short, as inGanymede/ˈɡænɪmiːd/GAN-im-eed andAnaxagoras/ˌænəkˈsæɡərəs/AN-ək-SAG-ər-əs.

Regardless of position, stressedu stays long before a single consonant (or a cluster ofp, t, c/k plusl, r), as inJupiter/ˈdʒuːpɪtər/JOO-pit-ər.

• Exception: A stressed nonhigh vowel (a, e, o) stays long before a single consonant (or cluster ofp, t, c/k plusl, r) followed by an/i/ or/iː/ee sound (e, i, y) plus another vowel at the end of a word:Proteus/ˈproʊtiəs/PROH-tee-əs,Demetrius/dɪˈmiːtriəs/dim-EE-tree-əs. This is because, historically and regionally, in many of these words thee, i, y is pronounced/j/y and combines with the following syllable, so that the preceding syllable is penultimate and therefore open:/ˈproʊtjuːs/PROH-tyooss.

Traditionally, English syllables have been described as 'open' when their vowel (in English) is long and they are followed by a single consonant followed by another vowel, and as 'closed' in the same environment when their vowel is short. However, it is debated how accurate this analysis is, as in English syllables tend to attract a following consonant, especially when they are stressed, so that all stressed syllables followed by a consonant are arguably 'closed'. Such following consonants are sometimes described asambisyllabic. This effect is especially apparent in some dialects, such as RP, when the consonant in question is /r/, which affects the quality of the preceding vowel. None of this changes the patterns described in this article: The long-short distinction described above is maintained regardless. For example, the 'e' inHera is long regardless of whether it is pronounced/ˈhiːrə/ or/ˈhɪərə/ in a particular dialect, or analyzed as openHEE-rə or as closedHEER-ə. American dictionaries tend to follow the former transcription, and British dictionaries the latter, so when the consonant 'r' is involved the rules for the English pronunciation of Latin words are more straightforward when using the conventions of American dictionaries.

Anglo-Latin includes all of the letters of theEnglish alphabet exceptw, viz.:a b c d e f g h i j k l m n o p q r s t u v x y z. It differs from Classical Latin in distinguishingi fromj andu fromv.In addition to these letters, the digraphsæ andœ may be used (as inCæsar andphœnix). These two digraphs respectively represent mergers of the lettersae andoe (diphthongs, as are Greek αι and οι) and are often written that way (e.g.,Caesar, phoenix). However, since in Anglo-Latin bothae andoe represent a simple vowel, not a diphthong, the use of the single lettersæ andœ better represents the reality of Anglo-Latin pronunciation.Despite being written with two letters, the Greek sequencesch, ph, rh, th represent single sounds. The lettersx and Greekz, on the other hand, are sequences of two sounds (being equivalent tocs anddz).

Anglo-Latin includes a large amount of Greek vocabulary; in principle, any Greek noun or adjective can be converted into an Anglo-Latin word. There is a conventional set of equivalents between the letters of the Greek and Roman alphabets, which differs in some respects from the current mode of Romanizing Greek. This is laid out in the tables below:

Rh is used for Greek ρ at the beginnings of words, e.g. ῥόμβος (rhombos) >rhombus.Rarely (and mostly in words relatively recently adapted from Greek),k is used to represent Greek κ. In such cases it is always pronounced[k] and never[s] (as it might be if spelledc) : e.g. σκελετός (skeletos) >skeleton not "sceleton".

Greek accent marks and breath marks, other than the "rough breathing" (first in the list of consonants above), are entirely disregarded; the Greek pitch accent is superseded by a Latin stress accent, which is described below.

Frequently, but not universally, certain Greek nominative endings are changed to Latin ones that cannot be predicted from the tables above. Occasionally forms with both endings are found in Anglo-Latin, for instance Latinizedhyperbola next to Greekhyperbole. The most usual equations are found below:

Examples:

• Greek ἄγγελος (aggelos) > Latinangelus (γγ > ng, -ος > us)
• Greek ἔλλειψις (elleipsis) > Latinellipsis (ει > i (shortened before ps), ψ > ps)
• Greek μουσεῖον (mouseion) > Latinmūsēum (ου > ū, ει > ē, -ον > um)
• Greek μαίανδρος (maiandros) > Latinmæander (αι > æ, -ρος > er)
• Greek χρυσάνθεμον (chrȳsanthemon) > Latinchrȳsanthemum (χ > ch, υ > y, θ > th, -ον > um)
• Greek διάῤῥοια (diarroia) > Latindiarrhœa (ῤῥ > rrh, οι > œ)

• The lettersb, f, k, l, m, p, v andz have each only one sound, which corresponds to the equivalent IPA symbols/bfklmpvz/.
• The letterj has the single sound/dʒ/.
• The letterr has a single sound,/r/ inrhotic dialects of English. In nonrhotic dialects, it varies according to placement in a syllable. At the beginning of a syllable, it is pronounced/r/. At the end of a syllable, i.e. between a vowel and a consonant, or after a vowel at the end of a word, it is dropped—though not without, frequently, affecting the pronunciation of the previous vowel sound. Ifr occurs at the end of a word after a vowel, and the next word begins with a vowel, it is usually pronounced as the beginning of the first syllable of the next word.Rh andrrh are pronounced exactly liker andrr.
• When followed by a vowel, the combinationsqu (always) andgu andsu (usually) stand for/kw/,/ɡw/, and/sw/ respectively.
• The combinationph is pronounced/f/.
• The combinationth is pronounced/θ/.
• The combinationch is pronounced/k/ in all environments - although in Scottish pronunciation it is pronounced/x/ at the end of a syllable.
• The lettersc, d, g, h, n, s, t, x have different values depending upon surrounding sounds and syllable structure.

The underlying consonantal phonemes of Anglo-Latin are close in most respects to those of Latin, the primary difference being that/w/ and/j/ are replaced in Anglo-Latin by/v/v and/dʒ/j. The sound/θ/th was borrowed from Greek.

Several word-initial clusters, almost all derived from Greek, are simplified in Anglo-Latin by omitting the first consonant:

• βδbd becomes/d/:bdellium
• τμtm becomes/m/:tmesis
• κνcn, γνgn, μνmn and πνpn become/n/:Cnossus, gnosis, Mnemosyne, pneumonia
• ψps becomes/s/:psyche
• κτct and πτpt become/t/:Ctesiphon, ptosis
• χθchth and φθphth become/θ/:Chthon, phthisis
• ξx becomes/z/:Xanthippe

In the middle of words both consonants in these clusters are pronounced (e.g.Charybdis, Patmos, Procne, prognosis, amnesia, apnœa, synopsis, cactus, captor); medialchth andphth are pronounced/kθ/ and/fθ/ respectively, as inautochthon andnaphtha.

The lettersc, d, g, h, n, s, t andx have different sounds (phonemes) depending upon their environment: these are listed summarily below.

The full set of consonantal phonemes for Anglo-Latin is almost identical to that of English, lacking only/ð/.

Environments that condition the appearance of some of these phonemes are listed below:

The change of intervocalic/s/ to/z/ is common but not universal. Voicing is more common in Latin than in Greek words, and never occurs in the common Greek ending-sis, wheres is always voiceless:basis, crisis, genesis.

The most common type of phonemic change in Anglo-Latin ispalatalization. Anglo-Latin reflects the results of no less than four palatalization processes. The first of these occurred inLate Latin, the second in Proto-Gallo-Romance, the third and fourth within the history of English. While the first two palatalizations are universally used in variants of Anglo-Latin, the third and especially the fourth are incompletely observed in different varieties of Anglo-Latin, leading to some variant pronunciations.

• Palatalization 1 affected only the sound oft, converting it to/ts/ when it preceded asemivoweli (at that stage pronounced/j/) and did not follows,t, orx. This/ts/ sound eventually merged with/s/ and was subject to further changes in Palatalization 3. When/t/ followed/s/ or/t/ it did not change; in some cases it might later change to/tʃ/ by Palatalization 4. Note thatt did not change to/s/ before semivowele, but remained/t/ as inconfiteor.
• Palatalization 2 affected the sounds ofc andg, converting them to/ts/ and/dʒ/; the/ts/ arising fromc merged with the/ts/ arising fromt, and both shared further developments of this sound, turning to/s/. When geminate (double), palatalizedcc andgg were affected diversely; only the secondc incc was palatalized, producing the sound/ks/, as insuccessor; but bothgs ingg were palatalized, producing a/dʒ/ sound, as in "exaggerate".
• Palatalization 3 affected/s/ and/z/ of whatever origin, changing them to/ʃ/ and/ʒ/.
• Palatalization 4 affected/s/ and/z/ exactly as Palatalization 3 did, but also affected/t/ and/d/, changing them to/tʃ/ and/dʒ/

Some of the occasions on which palatalizations 3 and 4 fail to take effect should be noted:

• Palatalization 3 fails:asphyxia, Cassiopeia, dyspepsia, excelsior, exeunt, gymnasium, symposium, trapezium. Note that the semivoweli is always pronounced as a full vowel/i/ in these cases. In some dialects Palatalization 3 frequently fails when another/ʃ/ sound follows, as in "enunciation", "pronunciation", "appreciation", "glaciation", "association", with the/s/ sound then generalized to closely related forms ("enunciate", "appreciate", "associate").
• Palatalization 4 fails (in some dialects):cæsura, fistula, pæninsula, pendulum.

Summary

See further the section on the "semivowel"below.

Following all of the above sound changes except palatalizations 3 and 4, "geminate" sequences of two identical sounds (often but not always double letters) were degeminated, or simplified to a single sound. That is,bb, dd, ff, ll, mm, nn, pp, rr, ss, tt became pronounced/bdflmnprst/. However, for the purposes of determining whether a syllable is open or closed, these single consonants continue to act as consonant clusters.

Other notable instances involving degemination include:

• cc developed two pronunciations:
  • before a front vowel (e, æ, œ, i, y)cc is pronounced/ks/, and as it consists of two distinct sounds, is not degeminated.
  • before a back vowel (a, o, u)cc was pronounced/kk/ which degeminated to simply/k/
• cqu/kkw/ degeminated to/kw/
• gg also has two pronunciations:
  • before a front vowel,gg is pronounced/dʒ/ after degemination.
  • before a back vowel,gg is pronounced/ɡ/ after degemination.
• sc before a front vowel was pronounced/ss/, and degeminated to/s/.
• sc andss before the "semivowel" are pronounced/ʃ/

The following combinations, derived from Greek, are also pronounced as single consonants:

• κχcch is pronounced/k/:Bacchus
• πφpph is pronounced/f/:Sappho
• τθtth is pronounced/θ/:Pittheus

The simple vowels of Anglo-Latin (a, æ, e, ei, i, o, œ, u, y) can each have several phonetic values dependent upon their stress, position in the word, and syllable structure. Knowing which value to use requires an explanation of two syllabic characteristics,openness andstress.

Openness is a quality of syllables, by which they may be eitheropen,semiopen,semiclosed, orfully closed.

Fully closed syllables are those in which the vowel in the middle of the syllable (the vocalicnucleus) is followed by at least one consonant, which ends or "closes" the syllable. Vowels in fully closed syllables appear:

• At the end of a word followed by at least one consonant, e.g.plus, crux, lynx.
• In the middle of a word followed by two or more consonants. The first of these consonants "closes" the syllable, and the second begins the following syllable; thus a word likelector consists of the two closed syllableslec andtor. Sequences of three or more consonants may be broken up in different ways (e.g.,sanc.tum,sculp.tor,ul.tra,ful.crum,ex.tra) but nothing depends upon the exact way in which this is done; any sequence of three or more consonants creates a closed syllable before it. The letterx is equivalent tocs, and as such also closes a syllable; a word likenexus is syllabifiednec.sus, and consists of two closed syllables.
• Two successive consonants of identical pronunciation are always pronounced as a single consonant in Anglo-Latin. When such a consonant sequence follows a penult syllable, the syllable counts as closed for the purposes of determining the position of stress:ba.cíl.lus, di.lém.ma, an.tén.na, co.lós.sus; they also prevent a penult syllable fromlengthening, as in the previous examples and alsopal.lor, com.ma, man.na, cir.rus, cas.si.a, pas.sim, glot.tis. They also count as closed for the purpose of determining whether au isopen or closed. In these respects they act precisely like syllable-closing consonant sequences, although they are pronounced as single sounds. (In words likesuccessor the twoc's do not merge, because each of them has a different sound—/k/ and/s/, respectively.)
• Certain sequences of consonants do not close syllables: these include all instances of obstruents (stops and fricatives) followed byr, includingbr, cr, chr, dr, gr, pr, tr, thr. Thus words likesupra andmatrix are syllabified assu.pra andma.trix, and the first syllable of both words is open; likewisea.cro.po.lis, di.plo.ma, de.tri.tus. The sequence/kw/ (spelledqu) also does not close the preceding syllable; i.e., one syllabifiesre.qui.em and notreq.ui.em.
• Sequences of obstruents followed byl are less consistent. The sequencescl, chl, gl andpl do not close a syllable, e.g.nu.cle.us,du.plex with open first syllables; but the sequencesbl, tl, thl do close a syllable, producing the syllabificationsPub.li.us,at.las,pen.tath.lon, with closed syllables before thel.

Semiclosed syllables are closed, unstressed syllables that had been closed and became open due to the merger of two following consonants of the same sound. For the purpose of determining vowel reduction in initial unstressed syllables they count as open.

• Double consonants following an initial syllable containinga, e, i, o merge to count as one consonant:a.(c)cumulator, a.(g)gres.sor, ca.(l)li.o.pe, a.(p)pen.dix, e.(l)lip.sis, co.(l)lec.tor, o.(p)pres.sor, o.(p)pro.bri.um. The first syllables of all these words are only partially closed, and the vowels are reduced.
• The same phenomenon occurs afteru, but note that theu is both closed and reduced:su.(p)pres.sor, su.(c)ces.sor, cu.(r)ri.cu.lum.

Semiopen syllables are syllables that had been closed and unstressed, and that are followed by a sequence of consonants that can stand at the beginning of a syllable. Since instances of obstruents +r orl are already considered open,semiopen syllables are practically restricted to instances ofs + obstruent,bl, and in some cases perhapstl. Vowels in initial semiopen syllables may be treated as open for all purposes except for determining the value ofu, which is still closed in semiopen syllables.

• Whens is followed by a consonant,s syllabifies with the following consonant:a.spa.ra.gus,pro.spec.tus,na.stur.ti.um,a.sphyc.si.a (asphyxia).S also syllabifies with a following palatalizedc (pronounced[s]):a.sce.sis, pro.sce.ni.um. Whens syllabifies with a following consonant, the preceding syllable counts as semi-open. Possible exceptions arepos.te.ri.or,tes.ta.tor.
• Other sequences of consonants fully close an initial unstressed syllable and produce a short vowel:an.ten.na,am.ne.si.a,bac.te.ri.um,mag.ni.fi.cat,mac.sil.la (maxilla),spec.ta.tor,per.so.na,oph.thal.mi.a,tor.pe.do.

See further the section on initial unstressed syllablesbelow.

Open syllables are those in which the nucleus is followed:

• By no consonant at the end of the word:pro, qua.
• By a vowel in the middle of a word :oph.thal.mi.a,fi.at, cor.ne.a,cha.os,chi.as.mus, a.ma.nu.en.sis.
• By only a single consonant in the middle of a word:sta.men,æ.ther,hy.phen,phœ.nix, ter.mi.nus,a.pos.tro.phe.
• By those consonant clusters that do not fully or partially close a syllable In the middle of a word :ma.cron,du.plex,Cy.clops, tes.ta.trix,a.cro.po.lis.

Stress is another characteristic of syllables. In Anglo-Latin, it is marked by greater tension, higher pitch, lengthening of vowel, and (in certain cases) changes in vowel quality. Its exact concomitants in Classical Latin are uncertain. In Classical Latin the main, orprimary stress is predictable, with a few exceptions, based on the following criteria:

• In words of one syllable, stress falls on that syllable, as marked in the following syllables with an acute accent:quá,nón,pár.
• In words of two syllables, stress falls on the first syllable of the word (thepenult, or second from the end): e.g.,bó.nus,cír.cus.
• In words of three or more syllables, stress falls either on the penult or theantepenult (third from the end), according to these criteria:
  • If the penult contains a short vowel in an open syllable, the stress falls on the antepenult: e.g.,stá.mi.na, hy.pó.the.sis.
  • If the penult contains a long vowel; a diphthong; a closed syllable (with any length of vowel); or is followed byz, the stress falls on the penult.
    • Long vowel: cicāda >cicáda, exegēsis >exegésis.
    • Diphthong:amóeba, Acháia, paranóia, thesáurus
    • Closed syllable:aórta,interrégnum,prospéctus,rotúnda
    • z:horízon

Primary stress can therefore be determined in cases where the penult is either closed or contains a diphthong. When it contains a vowel that may have been either short or long in Classical Latin, stress is ambiguous. Since Anglo-Latin does not distinguish short from long vowels, stress becomes a lexical property of certain words and affixes. The fact thatdecorum is stressed on the penult, andexodus on the antepenult, is a fact about each of these words that must be memorized separately (unless one is already familiar with the Classical quantities, and in the former case, additionally with the fact thatdecus -ŏris n. with short-o- syllable became in late Latindecus/decor -ōris m. with long-o- syllable:Dómine, diléxi decórem domus tuæ).

Secondary stress is dependent upon the placement of the primary stress. It appears only in words of four or more syllables. There may be more than one secondary stress in a word; however, stressed syllables may not be adjacent to each other, so there is always at least one unstressed syllable between the secondary and primary stress. Syllables containingsemivowele ori are never stressed.

• If a four-syllable word has primary stress on the antepenult, there is no secondary stress:pa.rá.bo.la, me.tá.the.sis.
• If a four-syllable word has primary stress on the penult, secondary stress is on the first syllable, marked hereafter with a grave accent:à.la.bás.ter, è.pi.dér.mis, sì.mu.lá.crum, prò.pa.gán.da, ùl.ti.má.tum.
• If a 5-syllable word has primary stress on the antepenult, secondary stress is on the first syllable:hìp.po.pó.ta.mus, Sà.git.tá.ri.us, Phì.la.dél.phi.a.

Secondary stress in words with three or more syllables before the primary stress is less predictable. Such words include those of five syllables with penult primary stress, and all words of six syllables in length or longer. The following generalizations about such long words may be made:

• The syllable immediately before the primary stress is never stressed.
• Words produced by derivation from a shorter word convert the primary (and, if any, secondary) stress of the stem into a secondary stress, as long as it does not fall immediately before the new primary stress:é.le.phant- + í.a.sis becomesè.le.phan.tí.a.sis
• Compounds of which the compound element consists of more than one syllable likewise convert the primary stress of their elements into secondary stress:phár.ma.co- + póei.a becomesphàr.ma.co.póei.a.
• If a primary stress is eliminated in compounding or derivation because it would stand next to another stress, secondary stress remains unchanged:pùsillánimus +itas becomespùsillanímitas.
• Single-syllable prefixes and single-syllable compound elements are generally unstressed:ac.cù.mu.lá.tor, im.pè.di.mén.ta, Her.mà.phro.dí.tus
• In other cases where the composition of the word may be unclear, every other syllable before the primary stress may be stressed:a.mà.nu.én.sis, ò.no.mà.to.póei.a. In some cases the third syllable before the primary stress is stressed when the second syllable is light, just as when assigning the primary stress.

Unstressed syllables are all others. They are always adjacent to a stressed syllable; that is, there can never be more than two unstressed syllables in a row, and that only when the first one follows a stressed syllable.

Several sound-changes in Anglo-Latin are due to the presence of the "semivowel", an alteration of certain front vowels. Originally ordinary vowels, they acquired at different points in history the value of the glide/j/ (ay-sound like that in English canyon). Subsequently, their value has fluctuated through history between a consonant and a vowel; the term "semivowel" thus reflects the intermediate historical as well as phonetic position of this sound. The environment in which the semivowel was produced was as follows:

1. The vowel wase (æ, œ), i (ei), ory.
2. The vowel came immediately before a vowel or diphthong.
3. The vowel was not in the initial syllable:e, æ, ei, i andy inrhea, mæander, meiosis, fiat, diaspora, hyæna, did not become semivowels.
4. The vowel was unstressed:e, æ, œ, ei, i inidea, Piræus, diarrhœa, Cassiopeia, calliope, elephantiasis did not become semivowels.

Examples of words wheree, i, y became semivowels include:miscellanea, chamæleon, nausea, geranium, rabies, Aries, acacia, ratio, fascia, inertia, halcyon, polyanthus, semiosis, mediator, Æthiopia, Ecclesiastes.

The effects of the semivowel include the following:

1. Though always in hiatus with a following vowel, semivoweli andy are never pronounced like longi ory (e.g./aɪ/); historically semivowele could also be distinguished from "long e" (formerly[ɛː] or[eː]). In current varieties of Anglo-Latin, semivowels are pronounced in a variety of ways:
   • Most frequently as/i/:labia, radius, azalea, præmium, cornea, opium, Philadelphia, requiem, area, excelsior, symposium, Cynthia, trivia, trapezium. In British Received Pronunciation, the prescribed pronunciation was once/ɪ/.
   • In some dialects or registers of English as/j/, e.g.junior pronounced[ˈdʒuːnjər].
   • Merged with a following-es or-e ending, as inAries, scabies.
   • They are usually deleted following the palatals/ʃ/,/ʒ/, and/dʒ/:Patricia, consortium, Persia, nausea, ambrosia, Belgium.
   • Occasionally a semivowel is retained after a palatal sound:ratio, sometimesElysium. This type of pronunciation is an artificiality, as the sounds/ʃ/ and/ʒ/ resulted from an absorption of the original/j/ in the sequences/sj/,/zj/. The pronunciations with/ʃi/ and/ʒi/ result from a re-introduction of thei sound to conform with the spelling. This pronunciation was, however, recommended by academics, and as such is common in the pronunciation of Anglo-Latin phrases such asab initio, in absentia, venire facias.
2. The consonantt changed to/s/ and then to/ʃ/ before the semivowel arising fromi:minutia, inertia, nasturtium.
3. The sibilants/s/ (includingss, sc, c, andt) and/z/ (usually spelleds) are usually palatalized before the semivowel:
   • /s/ >/ʃ/:cassia, fascia, species, militia
   • /z/ >/ʒ/:amnesia, ambrosia
4. The vowelsa, e, æ, ando in anopen antepenult syllable becomelong if a semivowel appears in the next syllable:
   • radius,Asia, azalea,area
   • anæmia, chamæleon
   • genius, medium, interior
   • odium, cochlea, victoria

This lengthening takes place regularly in antepenultimate syllables. It is less regular in syllables further back. On the one hand, there are words that do seem to lengthen before a semivowel in the next syllable:

• Æthiopia, Ecclesiastes, mediator, negotiator, variorum.

On the other hand, some words have short vowels:

• gladiator, apotheosis, Meleagrus, polyanthus (and other words containingpoly- followed by a vowel).

In general, those words with lengthened vowels in pre-antepenult syllables before a semivowel in the next syllable are those that are derive from a word with a regularly lengthened vowel in an antepenult syllable, e.g.,Æthiopia fromÆthiops ("Ethiopian"),Ecclesiastes fromecclesia ("church"),mediator frommedium,negotiator fromnegotium ("business'),variorum fromvarius ("manifold"). The failure ofgladiator (fromgladius, "sword") to have a long vowel is anomalous.

The most notable distinction between Anglo-Latin and other varieties of Latin is in the treatment of the vowels. In Anglo-Latin, all original distinctions between long and short vowels have been obliterated; there is no distinction between the treatment of a and ā, etc., for instance. However, the subsequent development of the vowels depended to a large degree upon Latin word stress (which was preserved nearly unchanged in the medieval period), and as this was in part dependent upon vowel length, in certain cases Latin vowel length contrasts have been preserved as contrasts in both stress and quality. However, the immediate governing factor is not length but stress: short vowels that were stressed for various reasons are treated exactly like stressed long vowels.

In addition to the merger of long and short vowels, other vowel mergers took place:

• the diphthongsæ andœ merged withe
• the vowelsi andy merged
• the diphthongei (alsoæi, œi), when still written distinctively, in pronunciation was merged withi or (more frequently)e

The merger ofæ andœ withe was commonly recognized in writing. Sometimes forms written withæ andœ coexist with forms withe; in other cases the form withe has superseded the diphthong in Anglo-Latin. Consider the following:

• æon andeon,æther andether,amœba andameba,anæmia andanemia,anæsthesia andanesthesia,cæsura andcesura,chamæleon andchameleon,dæmon anddemon,diæresis anddieresis,encyclopædia andencyclopedia,fæces andfeces,fœtus andfetus,hyæna andhyena,prætor andpretor

The following words are usually spelled withe, though they originally hadæ:

• ænigma >enigma,æquilibrium >equilibrium,æra >era,Æthiopia >Ethiopia,diarrhœa >diarrhea,mæander >meander,musæum >museum,œsophagus >esophagus,pæninsula >peninsula,præcentor >precentor,prædecessor >predecessor,præmium >premium,præsidium >presidium,tædium >tedium

In other cases, particularly names, the forms with the diphthongs are the only correct spelling, e.g.,ægis, Cæsar, Crœsus, Œdipus, onomatopœia, pharmacopœia, Phœbe, phœnix, Piræus, sub pœna.

The sequencesei, æi, œi (distinguished in writing and pronunciation fromej, the vowel followed by a consonant, as inSejanus) are sometimes retained in spelling preceding a vowel. In such cases the sequence is invariably pronounced as a simple vowel, sometimesi (as inmeiosis, pronounced as ifmiosis), sometimese (as inCassiopeia, Deianira, onomatopœia, pronounced as ifCassiopea, Deanira, onomatopea), and sometimes either (Pleiades, commonly pronounced as ifPleades orPliades).

The result was a system of five vowels,a, e, i, o, u. These would subsequently split, according to their environment, into long, short, and (eventually) unstressed variants; and these variants would eventually also be altered based on neighboring sounds. However, in phonemic terms, Anglo-Latin still hasonly five vowels, with multiple allophones.

In addition, there were the diphthongs,ai,oi,ui,au andeu. Of these,ai andau eventually monophthongized,eu merged with the open variant ofu, andyi merged with the "long"i. Onlyoi andui remained as true diphthongs, but both are extremely rare.

The vowelsa, e, i, o each have three primary variants: long, short, and reduced. Each of these may, in turn, exhibit allophonic variation based on features of its phonetic environment, including whether it is stressed, whether it is in an open or closed syllable, where it is positioned in the word, and what consonants are next to it. One of the most common environmental causes of vowel alteration is the presence of a followingr. Vowels altered by a following "r" are called "r-colored".

This is the default value for vowels, observed:

1. In closed monosyllables
2. In stressed closed penult syllables
3. In all antepenult syllables, open or closed, which receive primary stress,except for those lengthened due to a following semivowel
4. In all syllables with secondary stress
5. In fully closed unstressed syllables which immediately precede, but do not follow, a primary or secondary stress (usually in the first syllable of a word), with exceptions for certain prefixes

All short vowels have variants colored by a followingr sound when ther is followed by a different consonant (notr) or by the end of the word. In addition, there is a variant of shorta that only appears after a/w/ sound – chiefly found in the soundqu/kw/. This is a relatively recent phonetic development in English and Anglo-Latin, so it wasn't present in earlier stages of Anglo-Latin.

Exceptionally, monosyllables ending ines are pronounced with the rhyme/iːz/, e.g.,pes, res. This pronunciation is borrowed from that of-es used as anending.

Exceptions to the pronunciation of shorty generally involve prefixed elements beginning withhy- in an open syllable, such ashydro- andhypo-; these are always pronounced with a longy, e.g.hydrophobia, hypochondria. This pronunciation is the result of hypercorrection; they used to be pronounced with a short/ɪ/, as is still the case in the word "hypocrite" and (for some speakers and formerly commonly)hypochondria.

Prefixes may also behave in anomalous ways:

1. The prefixob- in unstressed syllables may be reduced to/əb/, even when it closes a syllable: cf. "obsession,oblivion".
2. The Greek prefixen-, em- in a closed unstressed syllable may be reduced to/ɪn,ən/,/ɪm,əm/:encomium,emporium.
3. The prefixex- in an unstressed syllable may be reduced to/əks/,/əɡz/, despite always being in a closed syllable:exterior,exemplar.
4. The prefixcon-, com- is reduced to/kən/,/kəm/ when unstressed:consensus, compendium, regardless of whether the syllable is closed or not.
5. The preposition and prefixpost(-) is anomalously pronounced with "long o":/poʊst/:post-mortem and cf. "postpone"; also thus in words in whichpost was originally a preposition (postea, postquam) but not in other derivatives, being pronounced with shorto inposterus, posterior, postremo, postridie.

Long vowels are those that historically were lengthened. By virtue of subsequent sound changes, most of these are now diphthongs, and none is distinguished by vowel length—however, the term "long" for these vowels is traditional. "Long" vowels appear in three types of environments:

1. a, e, i ando are long in anopen monosyllable
2. a, e, i ando are long in astressed open penult syllable
3. a, e ando are long when in anopen syllable followed by semivocalici ande
4. a ando are long when they precede another vowel in hiatus;i ande are long in the same environments, but only when they are not semivocalic (i. e., when they are in the initial syllable or receive primary stress). Hiatus may be original, or may arise from the deletion ofh between a stressed and unstressed syllable

Reduced vowels appear inunstressed syllables, except for:

• Closed initial unstressed syllables, which are generallyshort.
• Certainunstressed final syllables.

A variety of possible realizations are available foropen,semiopen, andsemiclosed initial unstressed syllables, including (fore andi) long, short, and reduced variants. Fully closed initial unstressed syllables are alwaysshort.

* A "schwi". It generally conflates with/ɪ/ in RP and with schwa in Australia. Directly before another vowel it may be the 'HAPPY' vowel.

** A rounded schwa. It has a w-like offglide before another vowel the way/oʊ/ does, but otherwise in many dialects it conflates with schwa.

The variation in the value of the initial open unstressed vowel is old. Two different types of variation can be distinguished; the older use of a "long" vowel fori, y, o (and their variants); and more recent variations in the value of the reduced vowel.

No completely general rule can be laid down for the appearance of an initial unstressed long vowel, although such vowels must have appeared before the shortening of geminate consonants, as they are restricted to fully open syllables. The most general tendency is for long vowels to appear wheni andy are either preceded by no consonant or byh, e.g.,idea,isosceles, hyperbola, hypothesis. The prefixesin andsyn never have long vowels:inertia, synopsis.I andy also tend to be short when the next syllable contains ani ory, short or long:militia, divisor.

O is a little less likely to appear with a long value in this location; or, at any rate, it is harder to distinguish the long value from the reduced vowel.

Unstressede andi in open syllables had merged by the early 17th century; their reduced reflex is often transcribed[ə], but by many speakers is still pronounced as a high front lax vowel, distinct from the[ə] derived froma, here transcribed ⟨ᵻ⟩. For such speakers, the first syllables inDemeter andDamascus are pronounced differently.

Unstressedo, also often transcribed[ə], is by many speakers pronounced with considerable lip-rounding, here transcribed ⟨ᵿ⟩.

The partially closed initial unstressed vowels began asshort vowels, but were later reduced.

These are the same sounds as in the preceding chart, but without the option of the "long" vowels and much less rounding of theo.

proscenium does not fall in this group, apparently because felt to bepro+scenium.

All vowels in medial unstressed syllables are reduced to/ə/ or/ɪ/, regardless of whether they are in open or closed syllables.

The pronunciation of the letteru does not depend upon stress, but rather upon whether the syllable in which it appears is open or closed. There are no "long" and "short" variants of either type ofu, but there are reduced andr-colored variants of both types.

The underlying sound of openu is/juː/; it shares developments with the homophonous diphthongeu, which can however appear in closed syllables.

The sound[j] in/juː/ and its variants is deleted in various environments:

• After palatal consonants (/ʃtʃʒdʒ/), whether original or resulting from the merger of/j/ and the preceding consonant, in both stressed and unstressed syllables; e.g./dʒ/:junior, Julius, Jupiter, cæsura, educator, spatula, fistula

After the following consonants when they precedeu in an initial, final, or stressed syllable:

• /r/ and/l/:rumor, verruca, luna, Lucretia, Pluto, effluvium
• /s/,/z/ (except in conservative speech):super, superior, Vesuvius, and variablyZeus

In some dialects, particularly of American English,/j/ is deleted after all dental/alveolar consonants when they precedeu in an initial, final or stressed syllable:

• /d/,/n/,/t/ and/θ/:duplex, caduceus, medusa, nucleus, lanugo, tutor, Thucydides
• For some speakers,/juː/ is pronounced[ɪu] following these consonants.

/j/ is not deleted in the following environments:

• Whenu is the first letter of the word or follows/h/:uterus, humerus
• Following a vowel:Ophiuchus
• Following labials/pbfvm/:pupa, furor, nebula, uvula, musæum
• Following velars/kɡ/ :cumulus, lacuna, Liguria
• When it is in an interior unstressed syllable not following a palatal consonant,/j/ remains after a single consonant even when it might be deleted in a stressed syllable:amanuensis and cf. "cellular, granular", for some speakers "virulent".
• After a consonant cluster/j/ may or may not be deleted:pæninsula, cornucopia

Closedu appears only in closed syllables, except for instances of the prefixsub- before a vowel. It has reduced andr-colored variants, as shown below.r-coloration only appears when ther is followed by a different consonant (notr) or the end of the word.

Diphthongs in Anglo-Latin are distinguished from simple vowels by having no long or short variants, regardless of position or syllable type. The only diphthongs that are at all common areau andeu. For variations in the pronunciation of the latter, seeOpenu.Au is, rarely, reduced in an unstressed syllable to[ə]:Augustus pronounced as if "Agustus". Such words may be pronounced with the full value of the diphthong, however.

Note thatui is generally disyllabic, as infruc.tu.i, va.cu.i, tu.i. The monosyllabic wordscui andhuic were traditionally pronounced/kaɪ/ and/haɪk/.

In general,ua,ue,ui,uo, anduu come immediately afterq,g, ors.^[8] However, when they occur in the same syllable, as insuavitas,questus,anguis,aliquot, andequus, they are not considered diphthongs (except in the exceptional pronunciations ofcui,huic, and the above examples).^[9] Accordingly, when these letter combinations occur in the penultimate syllable, it is not necessarily stressed (e.g.,équitas, notequítas). Note that their second vowel may become reduced, r-colored, or take on a different quantity even in a monosyllable.

The pronunciation of the final syllables of polysyllabic words do not always correspond to what might be expected from the constituent phonemes. Some endings also have more than one pronunciation, depending upon the degree of stress given to the ending.

Three types of endings can be distinguished:

The first class consists of vowels alone, i.e.-a, -e, -æ, -i, -o, -u, -y. In this class, the vowels are generallylong, but -a is always/ə/.^[10]

Words deriving from Greek long-ê (η) end in/iː/ unless assimilated, such assimile ending in/i/ (the 'HAPPY' vowel).

In the wordsmihi,^{[citation needed]}tibi,sibi, by an old tradition, the finali was pronounced like finale above (i.e., as if spelledmihe, tibe, sibe).

A late and purely academic pronunciation distinguished final-ā from -a by pronouncing the former like "longa",/eɪ/: for instance, Oxford professorA. D. Godley rhymedRusticā and "day". That this was not the usual pronunciation can be told from such forms ascirca, infra, extra, in absentia, sub pœna, all of which have an originally long final vowel: circā, sub pœnā, etc. This use is distinct from the older tradition (in use in the 17th-18th centuries) had madeall finala's "long", regardless of their Latin length.

The second class consists of vowels followed by consonant clusters such asns, nt, nx, ps, x. In this class, the vowels are always short, except foru, which may be reduced to[ə].

The third class consists of vowels followed by the consonantsl, m, n, r, s, t. The treatment of these endings is inconsistent. Generalizations include:

1. All vowels are reduced before finalr for/ər/:Cæsar, pauper, triumvir, Mentor, sulfur, martyr.
2. All vowels are reduced to/ə/ beforel:tribunal, Babel, pugil, consul.
3. Except sometimes beforet,a is reduced to/ə/ before any of this class of consonant:animal, memoriam, titan, atlas.
4. All instances ofu are reduced to/ə/ before any of this class of consonant:consul, dictum, locus.

The remaining endings are:-at, -em, -en, -es, -et, -im, -is, -it, -on, -os, -ot. Of these,-em, -im, -is, -it, -on, -ot have two possible pronunciations, one with a short vowel and one with/ə/. Final-es and-ies are alike pronounced/iːz/.^[11] Final-eus, when derived from Greek -ευς-eus (as in Proteus) rather than from -εος-eos with the-os changed to-us in Latin, may be pronounced as a single syllable with a diphthong (juːs), or as two syllables with a long (though unstressed)e followed by-us (iː.əs). However, even when pronounced as two syllables,-eus counts as a single syllable for the purpose of determining vowel length – that is, the syllable preceding the-eus ending is considered the penult, just as happens in derivatives ending in-ian – though the placement of the stress shifts. E.g. ἸλιονεύςĪlioneus/ɪliˈoʊniːəs,ɪˈlaɪənjuːs/il-ee-OH-nee-əs, i-LYE-ə-nyoos.

* The vowel of-is, -it is a "schwi". It generally conflates with/ɪ/ in RP and with schwa in Australia.

This last pronunciation of-os is the expected one; however, in the masculine accusative plural, where the ending is historically -ōs, the academic prescription was the pronunciation/oʊs/.^[12] Such an ending is not found in English loan words or proper names.

English adjectives formed from Greek and Latin roots often end in a suffix-an or-ic added to theoblique stem, sometimes retaining a precedingthematic vowel. These produce generally predictable sound changes in the stem though, depending on its source or simply due to confusion, English-ean may be either stressed or unstressed. The first derives from the Latin adjectival suffix-ānus, though it has far wider application in English than in Latin. (Pythian, for example, translates the Latin adjectivePythius.) The other suffix,-ic, derives from the Greek adjectival suffix -ικος-ĭkos.^[13]

The fact that these suffixes are added to the oblique stem is relevant with words of thethird declension whose stems end in a consonant that alters or disappears in the nominative case, as is apparent in such English noun–adjective pairs as Pallas ~ Palladian, Mars ~ Martian, Venus ~ Venerian, and indeed from non-adjectival derivatives of these words such as 'palladium'. Some roots have more than one oblique stem, in which case they may have more than one adjectival form in English.

Stressed vowel plus -an and unstressed -ian

The forms of these suffixes depends on the first vowel of the nominal suffix in Latin or Greek. The resulting forms are carried over into Latin from Greek, and into English from Latin. English adjectives in -ean, where thee is inherited from the Greek, are stressed on the suffix, but there are also adjectives in -ean where thee originates in Latin, and those are not stressed.

For example,

• GreekSamos > adjectiveSamios > LatinSamius > English 'Sā́mian'
• GreekPeriklēs > adjectivePerikleios > LatinPericlēus > English 'Pericléan'
• GreekLēda > adjectiveLēdaios > LatinLēdæus > English 'Ledaean' or 'Ledéan'
• Greek noun and adjectivePtolemaios > LatinPtolemæus > English 'Ptolemaean'
• GreekArgō > adjectiveArgōos > LatinArgōus > English 'Argóan'
• GreekSapphō > adjectiveSapphōios > LatinSapphōius > English 'Sapphóian'

Because thei of the suffix -ian makes an exception to the usual shortening of antepenultimate vowels, forms such as 'Samian' above are pronounced with a long stressed vowel.

Many Greek nouns ending in-o- take adjectives like those ending in-ē- instead, and some may take either form, in which case there may be two adjectives in English. An example isSisyphus:^[13]

• GreekSīsyphos > adjectiveSīsyphios > LatinSīsyphius > English 'Sisýphian'
• (or, alternatively) > adjectiveSīsypheios > LatinSīsyphēus > English 'Sisyphéan'

In the opposite direction,ē-vowelAristotelēs produces 'Aristotḗlian' alongside expected 'Aristoteléan'.

The finaliota subscript in Greek feminine words ending in -ῳ-ōi is frequently omitted in Latin, – if an 'i' appears in English it may be taken directly from the Greek, – but words that end in plain -ω-ō in Greek may also have Greek adjectives in-ōios if the 'i' was historically present.

Unstressed -ean

There is another suffix spelled -ean in English, but this one unstressed. It corresponds to the unstressed Latin adjectival suffix-ĕus and does not derive from Greek. An example is LatinHerculēs > Latin adj.Herculĕus > English 'Hercúlean', alongside GreekHēraklēs > Greek adj.Hērakleios > LatinHēraclēus/Hēraclīus > English 'Heracléan'. The stress assignment of the 'Herculean' has been affected by 'Heraclean' (or perhaps simply by the ambiguity of having two suffixes spelled '-ean'), so the mixed form 'Herculéan' is now also heard.^[14]

Adjectives in -ic

The Greek suffix -ικος-ĭkos replaces the vowels of nominal endings apart from retaininga orō, which will be stressed in English. Stress assignment (and the corresponding vowel quality) of English follows the Latin, as though the suffix were-icus in English as well and the last syllable of the stem were the antepenult when it English it's actually the penult. Thus English 'Homeric' is pronounced with a short stressed penultimatee, when in other English words from Latin ane in that position would be pronounced long. (Cf.botanic, which has the same stressed 'a' and same vowel quality asbotanical.)

For example,

• GreekHomēros > adjectiveHomērikos > LatinHomēricus > English 'Homĕ́ric'
• GreekAristotelēs > adjectiveAristotelikos > LatinAristotelicus > English 'Arístotĕ́lic'
• GreekAlkaios > adjectiveAlkaikos > LatinAlcaicus > English 'Alcáic'
• Greekhērōs > adjectivehērōikos > Latinhērōicus > English 'heróic'
• Greekēchō > adjectiveēchōikos > Latinēchōicus > English 'echóic'

Thus 'Aristotelian' is pronounced with a longe, but 'Aristotelic' with a shorte.

A number of Greek nouns ending inē (which historically is a longā) form adjectives in -aic, such as 'Cyrenaic' (GreekKȳrēnē > adj.Kȳrēnaikos > LatinCȳrēnaicus).

This sectiondoes notcite anysources. Please helpimprove this section byadding citations to reliable sources. Unsourced material may be challenged andremoved.(March 2011) (Learn how and when to remove this message)

This article'sfactual accuracy isdisputed. Relevant discussion may be found on thetalk page. Please help to ensure that disputed statements arereliably sourced.(March 2011) (Learn how and when to remove this message)

Latin as traditionally pronounced by English speakers is part of the living history of spoken Latin through medieval French into English.

Three stages of development of Anglo-Latin can thus be distinguished:

Note: The English pronunciation of Latin varies with accent as much as English itself, as the two's phonological systems are inseparably connected. For convenience's sake, the list below will end with approximately areceived pronunciation accent. As the traditional pronunciation of Latin has evolved alongside English since the Middle Ages, the page detailing English's phonological evolution from Middle English (in this case, from the Middle-English-Latin pronunciation roughly midway through the 1400-1600 section) can give a better idea of what exactly has happened, and this is just an overview.

Latin from the period when its orthography and grammar became standardized through to the pronunciation changes of Late Latin, while it was still a living language. Changes that took place in this period included:

• the merger off andph as[f]
• the change in pronunciation ofv (formerly[w]) to[v] and ofj (formerly[j]) to[dʒ].
• the merger ofi andy as[i]
• the merger ofe, æ andœ as[e]
• the change of non-initial, unstressed, prevocalic[i] and[e] to[j]
• the loss of distinctions of vowel length (merger of all long and short vowels)
• the palatalization oft to[ts] before[j]

Latin spoken in the context of Gallo-Romance and French from approximately the 6th to the 11th-12th centuries. During this period, Latin became a primarily written language, separated from the ordinary spoken language of the people. While it escaped many of the changes of pronunciation and grammar of Gallo-Romance, it did share a few of the changes of the spoken language. This was for the most part a period of stability.Changes in this period included:

• the palatalization ofc andg to[ts] and[dʒ] before front vowels
• the voicing of intervocalics to[z]
• the fronting ofu to[y]
• the restoration (based on spelling) of the vowels[i] and[e] from[j]

Latin spoken in the context of English from the 11th/12th centuries to the present. This last stage provides the greatest and most complicated number of changes. It starts with the displacement of the native pronunciation of Latin under the Anglo-Saxon kings with that used in the north of France, around the time of the Norman conquest in 1066. The English and French pronunciations of Latin were probably identical down to the 13th century, but subsequently Latin as spoken in England began to share in specifically English sound changes. Latin, thus naturalized, acquired a distinctly English sound, increasingly different from the pronunciation of Latin in France or elsewhere on the Continent. For example,Arthur, Prince of Wales andCatherine of Aragon corresponded for two years in Latin, but when they met in 1501 they found that they could not understand each other's spoken conversation, because they had learned different Latin pronunciations.^[15]

Some phases of development in this third stage can be reconstructed:

• The adaptation of the French sounds to English:
  • [s] was substituted for[ts]. (The French sound changed at about the same time; however, Anglo-Latin did not share related French simplifications such as[dʒ] >[ʒ].)
  • the vowels were given the valuesa[a],e[ɛ],i[i],o[ɔ],u[y]
  • [u] was substituted for[y] in closed syllables, e.g. lux[lyks] >[luks]. [y] didnot substitute out as [u] in unstressed open syllables, as it would have if the [y]~[u] distinction was solely length based:saeculum[sɛkylum]
  • At some point in the Middle English period, the distinctly French vowel[y] was substituted with the English diphthong[iw]. It is unclear when exactly this happened, and may have always been the English pronunciation of French (and consequently Latin)u.
• Stressed open penultimate vowels were lengthened, creating the short/long contrasts:

a[a]:[aː],e (alsoae,oe)[ɛ]:[ɛː],i (alsoy)[i]:[iː],o[ɔ]:[ɔː],u[u]or[iw]:[iw]

• Merger of unstressed open[ɛ] with[ɪ]
• Non-syllable-initial, unstressed, prevocalic[ɪ] became[j] (a change almost identical to that of Late Latin)
• Lengthening of the first of two vowels in hiatus
• Lengthening ofe[ɛ],i[ɪ], oro[ɔ] in pretonic initial syllables
• Diphthongization of[iː] to[ɛi]
• Lengthening of vowels in open syllables before[j] in the next syllable
• Raising of[ɛː] and[ɔː] to[eː] and[oː].
• Degemination of geminate consonants
• Palatalization of[s] and[z] before[j]
• Fronting of[ɐː] to[aː]

• Monophthongization ofai to[aː] andau to[ɔː]
• Change of[j] to[ɪ] (later >[i]) in many words, restoring original syllabicity.
• Change of frontedu ([iw]) to[juː]
• Palatalization of[tdsz] before (usually unstressed)[juː] (later >[jə])
• Lowering (and unrounding in the case of[ʊ]) of short[ʊ],[ɔ] to[ʌ],[ɒ]
• Tensing of shorte[ɛ] to[e]
• Former longi[ɛi] becomes[ai]
• Fronting and raising of shorta[a], longa[aː], and longe[eː] to[æ][eː][iː], creating the new contrasts:a[æ]:[eː],e[e]:[iː],i[ɪ]:[ai],o[ɒ]:[oː]
• Beginning of vowel reductions to[ə].
• Shorte,i andu before R; i.e.[ɛɹ],[ɪɹ] and[ʌɹ] are all merged to[ɜɹ]. (the NURSE merger)
• Shorta[æ] is lowered and lengthened beforer in a closed syllable, ultimately reaching modern[ɑː]

• Breaking of[eː] and[oː] to diphthongs[ei] and[ou], and then laxing to[eɪ],[oʊ] and the latter further to[əʊ]
• In non-rhotic accents, syllable-finalr[ɹ] becomes a vowel[ə̯]. The vowel-R sequences simplify somewhat:arC[ɑə̯] >[ɑː],orC[ɒə̯] >[ɔː],erC, irC, urC[ɜə] >[ɜː].
• Continued reduction of unstressed vowels to[ə].
• Shortening ofa in a stressed open penult syllable, e.g., data (not for all speakers)

A similar situation occurred in other regions, where the pronunciation of the local language influenced the pronunciation of Latin, eventually being replaced with reconstructed classical pronunciation. In German-speaking areas, traditional Germanized pronunciation of Latin is discussed atDeutsche Aussprache des Lateinischen(in German), with reconstructed classical pronunciation atSchulaussprache des Lateinischen(in German).

In Rattigan's playSeparate Tables the following conversation takes place between Major Pollock, who has been lying about his background, and Mr Fowler, a retired schoolmaster:

Pollock: . . . Still, those days are past and gone. Eheu fugaces, Postume, Postume.

Fowler:(Correcting his accent) Eheu fugaces, Postume, Postume. Didn't they teach you the new pronunciation atWellington?

Pollock: No. The old.

Fowler: When were you there?

Pollock: Now, let's think. It must have been nineteen eighteen I went up—

Fowler: But they were using the new pronunciation then, I know. . . .^[16]

InGoodbye, Mr. Chips, Chips says: "Well, I—umph—I admit that I don't agree with the new pronunciation. I never did. Umph—a lot of nonsense, in my opinion. Making boys say 'Kickero' at school when— umph—for the rest of their lives they'll say 'Cicero'—if they ever—umph—say it at all. And instead of 'vicissim'— God bless my soul—you'd make them say, 'We kiss 'im'! Umph— umph!"

In a fictional caseRex v. Venables and Others inA. P. Herbert'sUncommon Law, after a barrister uses phrases such as "ooltrah weerayze" (ultra vires) and "preemah fakiay" (prima facie) with the new Classics pronunciation he was taught at school, the Lord Chief Justice says "You are not to be blamed, Mr. Wick. But I am bound to make it clear to you, to the rest of your gallant generation and to the generations that come after, that His Majesty's judges will not permit the speaking of the Latin tongue after that fashion in the King's Courts. I cannot hear you, Mr. Wick, for the very good reason that I cannot understand you. We are using different languages. ... The bitter conclusion is, Mr. Wick, that you must go away and learn to pronounce the Latin tongue correctly, according to the immemorial practice of your profession."

• Andrews, E. A., and S. Stoddard, 1836.Grammar of the Latin Language for the Use of Schools and Colleges. This popular Latin grammar printed toward the end of the period when Anglo-Latin pronunciation was still commonly taught in schools, devotes a section to the rules of the pronunciation. While somewhat scattershot in its approach, it reveals several otherwise inaccessible details of the traditional pronunciation.
• Brittain, Frederick (1955).Latin in Church -- the History of its Pronunciation. Alcuin Club Tract (second ed.). London: Mowbray & Co. Mainly devoted to establishing the authenticity of Anglo-Latin against the claims of the Italianate pronunciation.
• Campbell, Frederick Ransom, 1888.The Language of Medicine. Chapter II, pp. 58–64 describes the pronunciation used in late 19th-century scientific and medical Latin.
• Copeman, Harold (1992).Singing in Latin, or Pronunciation Explor'd. Oxford.{{cite book}}: CS1 maint: location missing publisher (link). History and advice on singing Latin written or performed in the British Isles, France, Germany, Italy, Spain and Portugal.
• Dobson, E. J., ed., 1957.The Phonetic Writings ofRobert Robinson. Includes a phonetic transcription of a Latin poem representing the English pronunciation of Latin c. 1617, the direct ancestor of the later Anglo-Latin pronunciation.
• Ommundsen, Peter."Pronunciation of Biological Latin".
• Owen, Andrew,Pronouncing the Pater Noster in Modern English Latin taken from Copeman (1992) page 279
• Rigg, A.G. (1996). "Anglo-Latin". In Timothy J. McGee (ed.).Singing Early Music: The Pronunciation of European Languages in the Late Middle Ages and Renaissance. Bloomington: Indiana University Press. pp. 46–61.ISBN 0253210267.
• Robinson, D. H., revised by Hannah Oliver, 1903.The Latin Grammar of Pharmacy and Medicine. Chapter I, pp. 7–11, describe the pronunciation of pharmaceutical and medical Latin.
• Sargeaunt, John (1920).The Pronunciation of English Words Derived from the Latin. Society for Pure English Tract. Oxford: Clarendon Press.
• Sturmer, Julius William, 1908.Rudiments of Latin. pp. 9–17 describe early 20th-century scientific Latin pronunciation.
• Walker, John (1798).Key to the Classical Pronunciation of Greek, Latin, and Scripture Proper Names. Hopkins and Earle, no. 170, Market-street. Fry and Kammerer, printers.editions:97Ks2reWxwAC. Although this handbook is mostly devoted to establishing the position of the accent in Classical names used in English, it also includes an essay setting out some of the rules and regularities in the Anglo-Latin pronunciation.
• TheWestminster Schoolpronunciation of Latin^[17][18]
• Perseus Greek and Latin dictionaries. The most complete Greek and Latin dictionaries available online, they include the entire 9th edition of Liddell & Scott'sA Greek–English Lexicon. Greek headwords use the following conventions to remedy the defective orthography for vowels:α^ ι^ υ^ for shortα ι υ (adapted to Latin asă ĭ y̆) andα_ ι_ υ_ for longα ι υ (adapted to Latin asā ī ȳ).

• Law Latin

• Forms of Latin
• English phonology

• Webarchive template wayback links
• CS1 errors: ISBN date
• Articles with short description
• Short description is different from Wikidata
• Articles that may contain original research from May 2010
• All articles that may contain original research
• Pages with non-English text lacking appropriate markup and no ISO hint
• Pages with non-English text lacking appropriate markup from June 2020
• Wikipedia references cleanup from November 2025
• All articles needing references cleanup
• Articles covered by WikiProject Wikify from November 2025
• All articles covered by WikiProject Wikify
• Articles with multiple maintenance issues
• Pages with Latin IPA
• Pages with plain IPA
• All articles with unsourced statements
• Articles with unsourced statements from July 2020
• Articles needing additional references from March 2011
• All articles needing additional references
• Accuracy disputes from March 2011
• All accuracy disputes
• Articles with German-language sources (de)
• CS1 maint: location missing publisher