The reason that singers today sing Latin as if it were Italian is because of a somewhat long history of pronunciation reforms set out by the Roman Church. It wasn’t until November 1903, however, that the Italian vernacular pronunciation of Latin became recommended as a standard for the entire church. This conversion happened as a result of a Motu proprio from Pope Pius X that advocated for the rejection of big, theatrical music in the services of the church and for the reintroduction of plainchant and simpler tunes. It was again more solidified in November 30, 1928, when Pope Pius XI wrote, “Not content like our predecessors of happy memory, Pius X and Benedict XV, simply to approve this pronunciation of Latin, we ourselves express the keenest desire that all bishops of every nation shall endeavor to adopt it when carrying out the liturgical ceremonies.” This stance toward Roman Latin lasted from 1903 until Vatican II in the 1960s, when the vernacular languages became standard, and thus no discussion really continued to be made within the Church on how one should pronounce Latin.
According to Harold Copeman in 1990, some lands were slow to adopt the new Roman Latin, such as Germany, where there still thrives a wonderful German variety of pronouncing Latin, partially based on German history and partially based on influence from Classical Latin. In France, even as late as Fauré and Poulenc, Roman Latin had not yet been applied to French speakers. In Fauré’s Requiem or Poulenc’s Gloria, French Latin should still be used. The outrageous accentuation of the text in the Gloria should be evidence enough that Poulenc was not thinking in terms of Roman Latin, which has word stresses on specific syllables, but was thinking in terms of French Latin, which, like classical Latin, is entirely based on vowel length, rather than on syllable stress. In French Latin, the text flows in a constant stream of syllables, none particularly more important than another, just some shorter or longer.
Once we get to Duruflé, though, who did seem to advocate for Roman Latin, we should approach the music differently. When French singers sang his music, they were still flavoring the Latin with their French tendencies. For instance, the vowels were much more variant than the five appointed vowels of Roman Latin. According to Copeman’s interactions with a Latin professor at the Catholic University at Angers, the letter u could either be [y] or [u], on might be [õ] or [ɔn], -ti- could be either [ts] or [s], e might be either [ɛ] or [e], etc. It still wasn’t exactly Roman Latin, per se.
Whereas every region had its own national pronunciation of Latin before the Roman Church in the twentieth century advocated for a universalized Roman pronunciation, after Roman Latin became commonplace, every region continues to have national pronunciations of the Roman Latin. The United States and England are no exception. I will call this pronunciation system Anglo-Roman Latin.
In ARL, the following variances from Roman Latin exist:
- a can be pronounced either [a] or [ɑ]
- au is sometimes pronounced [ɑ] or [ɔ].
- c before a, o, or u is pronounced [kʰ], rather than [k]
- e can be pronounced either [ɛ] or [e], or even [eɪ]
- l can be pronounced either [l] or [ɫ]
- p is pronounced [pʰ], rather than [p]
- s sometimes has no voice between vowels
- t is pronounced [tʰ], rather than [t]
- u can be pronounced either [u] or [ʉ]
- z can be pronounced either [z] or [dz]
- Vowel-initial elements can sometimes begin with a glottal stop, as in [ʔɛʔɑs] for eas
Thus, in the United States, you are likely to hear the word salvator as [sɑɫvɑtʰɔɾ], rather than [salvatɔɾ]. The word et is quite prominently pronounced [ɛtʰ], rather than [ɛt]. Sanctus is particularly different, sometimes pronounced [ˈsɑŋktʰʉs].
Actually, in 1983, the Jesuit history professor and choral scholar Richard H. Trame advocated for a kind of Anglo-Roman Latin in the American Choral Director’s Association’s magazine, Choral Journal [vol. 23, no. 5 (January 1983), 29]. (As an aside, it is in his memory that Morten Lauridsen wrote his Ubi caritas adaptation). He wrote,
Rather than an insistence on making the vowel sounds approximate the Latin of a Roman aristocratic ecclesiastic, the projection of Latin’s broad and short vowels should be rendered with moderation relative to the vowel characteristics of the American Language.
That is, while there seems to to have been plenty of interest in the twentieth century in creating a standard pronunciation for sung Latin, the problem of national phonetic inventories has always gotten in the way of that to the point that advocating for a new kind of national Latin comes about. Latin has always been pronounced variably, and I think it always will be.