Latin pronunciation

The Latin pronunciation is the phonetics of classical Latin reconstructed by linguists , as it was pronounced by educated speakers in the times of Cicero and Caesar . It differs both from the school pronunciation of Latin taught in today's classroom and from the traditional German pronunciation of Latin .

History and reconstruction

The Latin pronunciation common in the Middle Ages and the early modern period had in many respects deviated from the classical standard that was insufficiently known at the time. In the 19th century an attempt was made to bring essential features of the scientifically developed classical pronunciation back to their right. By the beginning of the 20th century, the classical Latin pronunciation of ⟨ᴄ⟩ or c as [k] was already common in large parts of Germany , as is also attested by Quintilian (1st century AD) as a standard pronunciation . Due to a decree of the National Socialists , the school and university pronunciation of Latin was temporarily "Germanized" again in the 1930s, so that Latin words were read according to the German orthography and pronunciation tradition. In the meantime, the classic pronunciation of some phonemes (such as ⟨ᴄ⟩ or c as [k] and the separate pronunciation of the digraph ⟨ᴀᴇ⟩ or ae as [aɛ]) has become more widespread both internationally and in Germany. A classic pronunciation (pronuntiatus restitutus) that is consistently faithful to the original (e.g. pronuntiatus restitutus) is still in the minority when the ⟨ᴠ⟩ is pronounced next to vowels or v as [w].

The norm according to which Latin texts are pronounced depends on the context and the personal decision of the speaker. The reconstructed classical pronunciation described here is not generally binding. Rather, z. B. with liturgical texts and consequently also with sacred vocal music, the traditional German or - if the composer was / is Italian - the Italian pronunciation is appropriate. For school lessons, on the other hand, school pronunciation enables a middle ground between the classic ideal and the learner's pronunciation habits. These pronunciation variants are not "wrong", but arise from another tradition that follows the development of Latin over two millennia into an idiom that is used almost exclusively in writing, also in the adaptation of the pronunciation to the vernacular.

The reconstructed classical pronunciation, however, can claim to offer a more authentic access to classical Latin texts. It also facilitates international communication. It is true that it is sometimes argued against them that an error-free pronunciation of Latin in view of the many uncertainties is not achievable and is not necessary in a language without native speakers. An approximation can be achieved without further ado and is helpful for studying several university subjects (Indo-European Studies, Classical Philology, Romance Studies, Linguistics, Ancient History, Archeology, Papyrology and others).

Note on the spelling

The following applies in the following overview:

According to common usage today, the vowel lengths are given as a line above the letter .
The approximately phonetic pronunciation according to the system of the international phonetic alphabet ( list of characters ) is between square brackets.

Vowels

A careful distinction must be made between long ( productus ) and short ( correptus ) vowels. There are no simple rules for the length ( quantity ) of a vowel in a given word. Although rules can be given for the length of end vowels, there are numerous exceptions. With the quantity of vowels that are not in the final position, one must rely on the information in a dictionary.

The vowel quantity plays a role above all in Latin poetry, because the meter is not based on the stress of the syllables, as in German, but on the syllable quantity, which again depends on the vowel quantity: If the vowel is long, so is the syllable “Naturally” long ( syllaba natura longa ), but the reverse is not true. But quantity also plays a role in the simple meaning of the word, just as in German ⟨Bann⟩ [ban] means something other than ⟨Bahn⟩ [baːn] : ⟨ᴍᴀʟᴜꜱ⟩ malus "bad" and ⟨ᴍᴀ ´ ʟᴜꜱ⟩ mālus "apple tree “Only differ in the quantity of vowels, which is usually not marked in the traditional texts. In the example mentioned, one can understand what is being talked about through the context, but there are cases in which this may not be possible. For example, ⟨ꜰᴏᴅɪᴛ⟩ fodit “he digs” and ⟨ꜰᴏ ´ ᴅɪᴛ⟩ fōdit “he has dug” differ only in the difference between [ɔ] and [oː].

Lyric texts can be used to determine the length of the vowel, whereby one can deduce the length of an otherwise short syllable from the known meter. However, the problem arises that this is only possible with otherwise short syllables. If the syllable is position long ( syllaba positione longa ), that is, if the vowel is followed by two consonants, it is always metrically considered a long syllable, regardless of whether the vowel is spoken long (which would also make the syllable "naturally long") or not. In such cases, for example, on the basis of linguistic considerations, conclusions can be drawn about the quantity of a vowel, but such conclusions are always subject to uncertainties, which are then also subject to the information in dictionaries.

Deviations from the pronunciation of German

The distribution of the vowel lengths and abbreviations deviates from the rules that a German speaker is used to following. Please note in particular:

Long vowels can appear in any word syllable, not only in stressed syllables: ⟨ʀᴏ ´ ᴍᴀ ´ ɴꟾ⟩ Rōmānī "the Romans" = [ roːˈmaːniː ], vīdī "I have seen" = [ ˈwiːdiː ].
Can in a long vowel a Doppelkonsonanz (which as in Italian lengthened to speak is) follow without the foregoing vowel is shortened attended: Stella "Star" = [ stela ].
In contrast to German, stressed vowels in open syllables are not always lengthened, compare the Latin words and the pronunciation of the German loanwords in:
- Latin globus "Kugel" = [ ˈɡlɔbʊs ] versus the German Globus [ ˈɡloːbʊs ] (but compare the derived German word global )
- Latin pink "rose" = [ ˈrɔsa ] opposite the German rose [ ˈʁoːzə ]
- Latin Venus [ ˈwɛnʊs ] versus the German Venus [ ˈveːnʊs ]

Pronunciation of single vowels

Long ē is to be articulated in a semi-closed manner: ēmī [ ˈeːmiː ] ‚I bought 'as in German take [ ˈneːmən ]
Short e is more like half-open, compare emere [ ˈɛmɛrɛ ], “buy” as ä in German guests [ ˈɡɛstə ], but never weakened to [ ə ] in unstressed syllables
Long ō is half-closed, compare Rōma [ ˈroːma ] ‚Rom 'as o in the German bean [ ˈboːnə ]
Short o is rather half-open, so Latin bonus [ ˈbɔnʊs ] 'good' as o in the German Bonn [ bɔn ]
The short i is to be pronounced almost closed, compare it [ ɪt ] 'he / she goes' as i in the German Please [ ˈbɪtə ]
The same applies to short u , compare Latin humus [ ˈhʊmʊs ] 'earth' as u in the German river [ flʊs ] or the u of the final syllable in German humus [ ˈhuːmʊs ]

Sonus Medius

In classical times the short vowels i and u before labial consonants (b, m, p, f also in Latin, v) probably been spoken than [ʏ], such as the above in the German kiss [kʏsn̩]. Since this sound represents a mixture of i and u , so to speak , there was disagreement about the spelling. This is how the words doc u mentum , opt i mus , lacr i ma are also written: doc i mentum , opt u mus , lacr u ma . This so-called Sonus Medius developed from a historical short / u /, which later became a front tongue vowel through vowel reduction. In the vicinity of labial consonants, this sound has also been moved forward, but has probably retained a certain rounding.

Diphthongs

Besides the very rare ui and ei, there are only four diphthongs in Latin :

au as in Latin aurum "gold" = [ ˈaʊ̯rʊm ] as au in the German house [ haʊ̯s ]

In pre-Christian times, ae is still clearly diphthong, i.e. a monosyllabic sequence of a short [ a ] and a [ ɛ ], Latin maestus "sad" = [ ˈmaɛstʊs ], almost like ei in German mostly [ maɪ̯st ]

The rare, almost only in Greek loan words occurring eu is a monosyllabic series of short, half-open [ ɛ ] and [ ʊ ], that Europe "Europe" = [ ɛʊ̯roːpa ] as it is, for example, is pronounced in Italian today, but on definitely not a German eu like in Europe [ ɔʏ̯ˈʁo ] pa ]

oe corresponds more or less to the German diphthong eu , so it is a monosyllabic sequence of a short, half-open [ ɔ ] and a [ ɛ ], compare Latin poena "punishment" = [ ˈpɔɛna ] almost like eu in German Europe [ ɔʏ̯ˈʁoːpa ], but it is by no means to be pronounced as [ ø ].

According to the Frankfurt classical philologist Axel Schönberger , the spellings <ae> / <ai> and <oe> / <oi> have been in use since at least the third century BC - as well as in the western Greek alphabet of Boeotian Greek and thus also in large parts of the Magna Graecia of Italy - for monophthongs . The approach of a diphthong pronunciation for the older Latin is wrong.

Consonants

All consonants written twice are pronounced lengthened, compare Latin crassus "thick" = [ ˈkrasːʊs ], repperit "he / she has found" = [ ˈrɛpːɛrɪt ].

In classical times c always corresponded to an unaspirated German k , that is, Latin cinis "Asche" = [ ˈkɪnɪs ].

f : It is assumed that it was pronounced like German [ f ], but this has not been proven; The pronunciation [ ɸ ] would also be possible .

gn is articulated like a German sequence of velarem ng plus n , thus Latin īgnis "fire" = [ ˈiːŋnɪs ]; a vowel in front of gn is legally always long.

h was in the classical period at least from the upper layers and the letters even as a German [ h ] (or at most slightly weaker) articulates; among the lower classes it was already silent in the time of Catullus (84–54 BC).

i is pronounced in the vicinity of vowels as [ j ], compare

at the beginning of the word: Latin iūstus "just" = [ ˈjuːstʊs ] like j in German just [ ˈjʊst ]
in the interior of the word between vowels most likely as a long [ jː ], compare eius "his / her" = [ ˈɛjːʊs ],
inside the word between consonant and vowel: as a sequence [ ij ], compare fīlius "son" = [ ˈfiːlijʊs ] (three syllables)

l had two different pronunciations:

before [ i ] and as a long ll : like German [ l ], compare
- fīlius "son" = [ ˈfiːlijʊs ]
- bellus "beautiful" = [ ˈbɛlːʊs ] similar to German Elle = [ ɛlə ], but with an extended l
in all other cases: as velares [ ɫ ], also called “ Meidlinger L ”, compare cūlus “ass” = [ ˈkuːɫʊs ]; lūna "moon" = [ ˈɫuːna ], similar to l in English well "gut" [ wɛɫ ] or kölsch kölsch [ kœɫʃ ]

m was pronounced like German [ m ], but was largely silent at the end of the word except for monosyllabic words; In some cases, the vowel before it was pronounced nasally , compare Latin Rōmam ( accusative singular) = [ ˈroːmã (m) ] or especially in less careful pronunciation = [ ˈroːma ]. At the end of the syllable, m , if followed by a consonant, was assimilated to it (similar to the anusvara of Sanskrit , in the interior of the word); compare the spellings eandem or 'hanc , with the latter word n standing for the velar sound [ ŋ ]. This was probably also the case when a word with m at the end is followed by a word with a consonantic beginning; compare Latin tum dīxit = [ tʊnˈdiːksɪt ] or [ tʊnˈdiːsːɪt ] and so on. Hence Cicero's warning to clearly pronounce cum nos - with cum as a conjunction of a subordinate clause - otherwise it would sound like the accusative form of cunnus " shame ".

n is pronounced as in German, except in the consonant groups ns and nf , where in classical times, if at all, it was only articulated very weakly, while the vowel in front of it was certainly more or less clearly nasalized and always pronounced lengthened, compare Latin ānser “goose” = [ ˈãːnsɛr ] or even [ ˈãːsɛr ] similar to an in the French pantalon “pants” = [ pɑ̃taˈlɔ̃ ].

p is always pronounced unaspirated , so as in French , compare Latin pūrus "rein" = [ ˈp⁼uːrʊs ] as in French pur = [ p⁼yːʀ ], but not as in German pur = [ pʰuɐ ]. According to some, p is also used as a mere orthographic sign after m to indicate a de-nasalized pronunciation of the m : Latin sūmptus "taken" = [ suːmtus ]. In terms of linguistic history, however, such a phenomenon of p admixture after the consonant m can also be found in other languages; see. on this: "Thompson". Therefore, it cannot simply be assumed that it should not be spoken in Latin under any circumstances.

qu is the labialisierte velar [⁠ K ⁠] , so Latin quis "who" = [ kʷɪs ].

r was not rolled with the uvula but with the tip of the tongue (as it is today in Italian or Spanish and still often in Bavaria and parts of Switzerland), compare Rōma "Rome" = [ ˈroːma ] as in Italian.

The exact articulation of s is controversial:

It was probably always articulated as a voiceless consonant, definitely at the beginning and end of a word or before a consonant, compare sōl "sun" = [ soːɫ ], thus as s in the English cent = [ sɛnt ]. It is possible, but not proven, that a simple s in the interior of a word could become voiced between vowels, perhaps pink "rose" = [ ˈrɔsa ] or [ ˈrɔza ];
After all, it is likely that the s , whether voiceless or voiced, was not pronounced like [ s ] or [ z ], but like [ ʂ ], which corresponds to Swedish rs . Correspondingly, the Latin īnsula “island” would probably have sounded like [ ˈĩːnʂʊɫa ].

Like p , t is unaspirated , compare the Latin tālis “such” = [ ˈtaːlɪs ] with the German thaler = [ ˈtʰaːlɐ ].

ti is pronounced in classic pronunciation as a non-aspirated dental with the following i or j [ tɪ ] or [ tj ].

u next to a vowel (or, in graphically highly modernized texts, v ) is not used as [ v ] like w in German wine = [ vaɪ̯n ], but as [ w ] ( labialized voiced velar approximant ) like w in English well "gut" = [ wɛɫ ] pronounced; compare Latin uespa (also written vespa ) "wasp" = [ ˈwɛspa ] (or [ ˈwɛʂpa ]) like w in English wasp [ wɒsp ]; Latin uallum (or vallum ) "Wall" = [ ˈwalːʊ (m) ] like w in English wall [ wɔːɫ ].

In Latin there are no glottal beats ("crackling sounds"), as many native speakers of German automatically put in front of every vowel initial sound without them being aware of it or hearing the crackling sounds they speak. Since almost all Germans speak these pops involuntarily in front of a word beginning with a vowel, this results in a strong German accent, especially in all Romance languages and Latin. As a makeshift, the German can pronounce all these words with the beginning h (cf. the difference between German "Hain" and "ein"), which avoids the crackling sound and is much closer to the original pronunciation (but of course it is also wrong in the end).

From all of this it follows that, for example, the name of the famous dictator, Caesar , was pronounced roughly like [ ˈkaesar ] or [ ˈkaezar ] in Latin during Caesar's time . This pronunciation does not match the school pronunciation [ ˈkʰaɪ̯zar ] or [ ˈkʰɛːzar ] or [ ˈtsɛːzar ].

Stress rules

Stress mark in a Latin missal

There are two problems to be distinguished when stressing Latin words:

Which syllable is stressed

In polysyllabic words, the so-called word accent usually falls on the penultimate or third from last syllable ; in rare cases it is also on the last syllable.

An emphasis on the last syllable is therefore entirely possible in polysyllabic words; This mainly applies to words whose last syllable has been omitted , for example: adhuc or vidistin (= vidistine ), but also final stresses to avoid homonym conflicts such as érgō (conjunction) vs. ergô (postposition) or círcum (accusative singular) vs. trailing circúm (preposition). According to ancient evidence (see Schönberger 2010) proper names such as Maecēnās were given final stress : Maecēnās . A final emphasis on case forms, on the other hand, is always unhistorical and wrong. The accentuation of Latin words can change within phonetic words as, in certain cases, proclamation or enclise occurs. There is no change in accent in verse, since Latin poetry has nothing to do with the emphasis on words, but is based on a fixed sequence of short and long syllables that leave the natural word accents unchanged.

The decision as to which syllable is to be emphasized in polysyllabic words depends solely on the penultimate syllable (so-called panultia rule ):

The penultimate syllable is stressed when
- the word is two-syllable, for example in Rō-ma "Rome" = [ ˈroːma ];
- the penultimate syllable has a long vowel, compare Rō-mā-nus "Roman" = [ roːˈmaːnʊs ], can-dē-la "lamp" = [ kanˈdeːɫa ]; if no vowel lengths are given in a text, the so-called quantity of the vowel, i.e. its length or shortness, cannot be seen in the body of the word;
- the penultimate syllable has a diphthong , for example in in-cau-tus "careless" = [ ɪŋˈkaʊ̯tʊs ];
- the penultimate syllable is “closed”, that is, ends with a consonant followed by at least one (initial) consonant of the last syllable, compare ter-res-tris “earthly” = [ tɛrˈrɛstrɪs ], a-man-tur “she are loved "= [ aˈmantʊr ]. As in the Latin cerebrum "brain" = [ ˈkɛrɛbrʊ (m) ], the combination of a muta ( b ) and a liquida ( r ) does not cause a syllable ( br ), so the third from last syllable is stressed.
In most other cases, the third last syllable is stressed, compare exer-ci-tus “Heer” = [ ɛkˈsɛrkɪtʊs ], exer-ci-tu-um ( genitive plural ) = [ ɛksɛrˈkɪtʊʊ (m) ].

Nature of the accent

There is no agreement among the experts on the question of how the stressed syllable was emphasized. Some assume that, in contrast to German and similar to ancient Greek, Latin was a language with a melodic accent , in which the stressed syllable is not characterized by an increase in volume, but by a change in the vocal tone . Word for ro-ma = [ Roma ] Accordingly, the first would be More the stressed syllable ro- with other pitch spoken were as the following -ma . One of the arguments for this is that the Latin metric was quantitative and not accentuating and that the typical effects of a dynamic accent such as regular syncopation and attenuation of unstressed vowels are missing. W. S. Allen (see literature ), however, thinks that the facts speak for a dynamic accent. Axel Schönberger (see literature ) compiles the statements of ten ancient specialist writers (from Quintilian to Priscian), all of which testify to a melodic accent in Latin, whereby the treble could only last one more, so that the word Rō-ma follows accentuating is: Rôma (in Mora notation : Róòma ), while Athēnae has an acute accent : Athénae (in Mora notation : Atheénae ). Schönberger also assumes that there was never an initial accent in archaic Latin, but rather a melodic accent prevailed from the Indo-European basic language to Classical Latin, which in older Latin could recede to the fourth from the last syllable; the claim of a temporary initial stress in Latin is an extremely unlikely hypothesis without sufficient evidence.

Sources of the reconstruction

The attempts to reconstruct the Latin pronunciation are essentially based on a combination of the following sources, information and circumstantial evidence:

explicit statements of the ancient Latin grammarians on pronunciation
Rules of the Latin metric
Development of the pronunciation of Latin words in living follow-up languages
Verses, rhymes, theater and chants (ductus)
Information, comparisons and criticism of errors, for example, from Quintilian and Aulus Gellius
Compare similar sounding words
Comparison with root word
Reproduction of Latin words in ancient Greek and vice versa
Doodles written in phonetic transcription due to ignorance on the walls of ancient houses
Spelling mistakes in surviving original scripts
Preservation of the old phonetic level in borrowed words
Reproduction of the apex (characters) in classical inscriptions

literature

William Sidney Allen: Vox Latina. A guide to the pronunciation of classical Latin. 2nd edition. Cambridge University Press, Cambridge et al. 1978, ISBN 0-521-22049-1 .
Vera UG Scherr: Handbook of Latin Pronunciation. Performance practice vocal music. Classic - Italian - German. With extensive phonetics of Italian. Bärenreiter, Kassel et al. 1991, ISBN 3-7618-1022-9 .
Axel Schönberger: Priscian's depiction of the syllable-bound pitch mora accent in Latin (= Bibliotheca Romanica et Latina 13). Latin text and annotated German translation of the book on the Latin accent. Valentia, Frankfurt am Main 2010, ISBN 978-3-936132-11-3 .
Axel Schönberger: On phonetics, prosody and phonotactics of Latin according to the description of Priscian. In: Millennium . Vol. 11 (2014), pp. 121-184.
Lothar Steitz: Bibliography on the pronunciation of Latin (= Phonetica Saraviensia 9, ISSN 0721-6440 ). Institute for Phonetics of the Saarland University, Saarbrücken 1987.
Wilfried Stroh : Arsis and thesis or: how did you pronounce Latin verses? In: Michael von Albrecht , Werner Schubert (Hrsg.): Music and poetry. New research contributions. Dedicated to Viktor Pöschl on the occasion of his 80th birthday (= sources and studies on music history from antiquity to the present 23). Lang, Frankfurt am Main et al. 1990, ISBN 3-631-41858-2 , pp. 87-116.
Frederic William Westaway: Quantity and accent in the pronunciation of latin. Cambridge University Press, Cambridge 1913, digitized

Web links

Phonology of Latin and Vulgar Latin

Individual evidence

^ FW Westaway: Quantity and accent in the pronunciation of latin. Cambridge 1913, p. 43ff

↑ In Latin inscriptions that have survived since the 1st century BC Long vowels marked with apex , but these markings are not always reliable. Before long vowels were denoted by doubling (e.g. PAASTOR for pāstor ).

↑ See Johan Winge: Vowel Quantity - Where your Dictionary is Wrong. 2007, accessed March 14, 2020 .

^ William Sidney Allen: Vox Latina. A guide to the pronunciation of classical Latin . 2nd Edition. Cambridge University Press, Cambridge 1978, ISBN 0-521-22049-1 .

↑ Axel Schönberger: On phonetics, prosody and phonotactics of Latin according to the description of Priscian . In: Millennium 11, Issue 1 (2014), pp. 121–184.

[1] FW Westaway: Quantity and accent in the pronunciation of latin. Cambridge 1913, p. 43ff