Caesar, Chomsky and Comme le Prévoit

I have had formal study in Latin for many years, but I learned more from one teacher than all the rest together.

It was the second year of Latin, in a new school. The first year had been sleepy and boring – Britannia est insula and the like – but this class was full of excitement. The teacher had a simple maxim: ‘work out who’s doing what to whom.’ We learned declensions and verb forms and grammar, but we didn’t treat the Latin as a sort of puzzle, to be deconstructed into ablatives absolute and datives of agent. We actually learned to read the language.

We started that year with Caesar. Things could go badly wrong in the provincial forests if you didn’t know who was doing what to whom: the Gauls could end up slaughtering the Romans, or Caesar’s troops could send their commander on a forced march, instead of the other way round.

What was absolutely forbidden was to translate word-for-word. It was the sense, rather than the syntax, that mattered. I worked harder in that class than I had ever worked before, but by the middle of the year I could look at a reasonably simple sentence – I remember this one in particular:

Ibi perpauci aut viribus confisi tranare contenderunt aut lintribus inventis sibi salutem reppererunt

and know, as a whole and virtually at sight, exactly what it meant – more or less, ‘There [on the banks of the Rhine] a few men, relying on their strength, attempted to swim across, while a few others found their safety in boats’. I had not previously come across linter (a small boat) but given that it was an alternative to swimming across, that the river was likely too wide to jump and too deep to ford, and that a linter had to be found before it could be used, it was clear what was going on.

Whether or not this teacher knew it, he was following the thinking of Noam Chomsky, the celebrated and controversial linguist who taught us to distinguish between ‘deep’ and ‘surface’ structures of a language. The deep structures, at least in Chomsky’s early thinking, represented underlying components (‘kernel sentences’) that were independent of the target language. These were then transformed into English or French or Latin of a given style and register of speech. Hence – returning to Caesar – the following surface forms share a common deep structure.

copias suas Caesar in proximum collem subduxit (‘Caesar leads his forces away to the next hill’)

Copiae suae Caesare in proximum collem subductae sunt (‘His forces were led away by Caesar to the nearest hill’)

even though one is in the active and the other the passive voice. I find it interesting, by the way, that while I was able to read the first sentence instantly, rendering it in the passive was a bit of a struggle, e.g. to ensure that the passive perfect verb agreed in number and gender with the subject. This is largely because my Latin is weak, but I think it also demonstrates that ‘surface structures’ vary hugely across languages.

Chomsky published his influential Syntactic Structures in 1957. The authors of Comme le Prévoit seemed to be following his thinking when they wrote in 1969 that

§5. A liturgical text, inasmuch as it is a ritual sign, is a medium of spoken communication. It is, first of all, a sign perceived by the senses and used by men to communicate with each other. But to believers who celebrate the sacred rites a word is itself a “mystery”. By spoken words Christ himself speaks to his people and the people, through the Spirit in the Church, answer their Lord.

… §7. Thus, in the case of liturgical communication, it is necessary to take into account not only the message to be conveyed, but also the speaker, the audience, and the style. Translations, therefore, must be faithful to the art of communication in all its various aspects, but especially in regard to the message itself, in regard to the audience for which it is intended, and in regard to the manner of expression.

§8. Even if in spoken communication the message cannot be separated from the manner of speaking, the translator should give first consideration to the meaning of the communication.

Anscar Chupungco describes the liturgical translation process in more detail. For our purposes the ‘source’ language is Latin and the ‘receptor’ language English:

This process also aims to discover the relationship between the ‘kernel’ and the other segments of the text, or, in other words, its global historical, cultural, doctrinal, and literary context. The process concludes with the ‘recomposition’ of the text in the new form of the receptor language.

With its particular cultural and linguistic qualities the new form should be able to clothe the content in the same manner as the original form has done. It is implied that the new form possesses equivalent qualities wherewith the content can be adequately expressed. In a way, we can describe the process of translation as the passage of content from one cultural form to another than represents equivalently the original form. A good translation is thus able to produce in the audience of the receptor language the same effect as the source language has on its original audience. In short, it achieves the same purpose as the original text.

[Anscar J Chupungco OSB, “The Translation of Liturgical Texts”, in Chupungco, ed, Introduction to the Liturgy (Liturgical Press, 1997).]

Chomsky’s theories have been controversial – the subject of the ‘linguistics wars’ of the 1970s and 1980s – and he himself eventually moved away from his earliest notion of ‘deep structure’. Nonetheless, I think that no language scholar would argue that syntax and underlying content are fundemantally the same thing.

This is where Liturgiam Authenticam goes so badly wrong: it seems to abandon the idea of the liturgy as a medium of communication. As best I understand it, Lit Auth asserts that the Latin of the Mass has not only a ‘sacral’ vocabulary but also a ‘sacral’ syntax that has to be followed in every language into which it is rendered. In a tangled way, it seems explicitly to contradict the idea that Chupungco sets out above:

§19. The words of the Sacred Scriptures, as well as the other words spoken in liturgical celebrations, especially in the celebration of the Sacraments, are not intended primarily to be a sort of mirror of the interior dispositions of the faithful; rather, they express truths that transcend the limits of time and space.

In other words, it is unimportant that the faithful understand the words as they are spoken; it is the words themselves, not their meaning, that ‘express truths that transcend the limits of time and space.’

Hence we get the remarkable §29 of Lit Auth:

It is the task of the homily and of catechesis to set forth the meaning of the liturgical texts, illuminating with precision the Church’s understanding regarding the members of particular Churches or ecclesial communities separated from full communion with the Catholic Church and those of Jewish communities, as well as adherents of other religions – and likewise, her understanding of the dignity and equality of all men. Similarly, it is the task of catechists or of the homilist to transmit that right interpretation of the texts that excludes any prejudice or unjust discrimination on the basis of persons, gender, social condition, race or other criteria, which has no foundation at all in the texts of the Sacred Liturgy. Although considerations such as these may sometimes help one in choosing among various translations of a certain expression, they are not to be considered reasons for altering either a biblical text or a liturgical text that has been duly promulgated.

Liturgiam Authenticam therefore agrees that correct understanding is important – but apparently this understanding doesn’t have to occur during the liturgy itself. That is why I have heard more than one homily explaining that, while we say ‘for many’, what we really mean is ‘for all’.

I can only imagine that, as the Mass is translated on these bizarre principles to languages like Japanese and Turkish, where the surface structure differs wildly from that of English, more such problems will emerge. Eventually the language becomes a sacral babble, one that requires the faithful to carry along a hand missal. In which case, why not stick to the Latin? That would certainly make the work of music directors easier.

I think that Chomsky, Comme le Prévoit and my second-year Latin teacher had the right view of translation. But if this view is no longer licit, then we should say Mass in Latin and not bother translating. Perhaps this was the intent from the start. Caesar might well have approved.



  1. I think the implied model of the sacral syntax – the model being we are talking to God, so it only matters if God understands, and the understanding of the faithful is at best a marginal concern – is fundamentally undermined by the re-sacramentalization of the faithful over the past century (that is, the decisive shift away from the common praxis of sacramental participation for centuries before Pius X).

  2. How are the translations into non-Western languages such as Japanese being done? Are they directly Latin to Japanese or Latin to English to Japanese? In other words, are the Japanese to be given a translation of our translation?

  3. The reference to “sacral babble” recalls for me an episode of the first “Star Trek” TV series, in which Kirk et al. end up on a planet colonized by a group of early space explorers. They protected and proclaimed a sacred text and repeated sacred words for which they had lost the meaning, remembering only that the words were sacred. The sacred words began something like this: “Ik plek aleeance . . .” It was not until an old, torn American flag was brought into the room that Kirk realized that the colonists were trying to recite the Pledge of Allegiance. Is this what we’re headed for with our sacral language?

  4. Karl, I agree entirely — though some would solve this problem by calling for a return to the marginal role of the faithful.

    Brigid, Liturgiam Authenticam is explicit in §24 that translations must not be produced from other translations. You have to go back to the Latin.

  5. This comment is nothing new, profound, or argumentative from me who has only 2 years of high school Latin!
    But I wanted to say how much I heartily agree with Chomsky, Chapungo, Comme le Prévoit and Day! Thank you!

  6. I agree with Jonathan- I was taught the same as he was.
    We translated six books of The Aeneid. Had we translated word for word it wouldn’t have made any sense! Sometimes one sentence was an entire paragraph! But what really grinds me is when they don’t even follow their own rules and pretend “esse” is the same as “adstare” and add “we pray” where it doesn’t exist in the Latin. LA is a made up translation process. It’s annoying.

    1. ES –
      (Tongue only slightly in cheek)
      And what, pray, is wrong with one sentence being an entire paragraph? Yes, even in English. Such is a challenge to compose and great fun to read. Old history books and such are full of them. Many scholarly books by PhD’s of today would, grammatically, bore to tears a literate person of even a hundred years ago.

      1. MJO, there’s a very simple reason why paragraph-long sentences, 80-word sprawlers like some of the prayers in the new translation, don’t belong in the English liturgy. It is because the liturgy is a means of communication, not a puzzle to be deconstructed or something quaint and fun to read.

        Back to that little statement in Comme le Prévoit:

        §5. A liturgical text, inasmuch as it is a ritual sign, is a medium of spoken communication. It is, first of all, a sign perceived by the senses and used by men to communicate with each other. But to believers who celebrate the sacred rites a word is itself a “mystery”. By spoken words Christ himself speaks to his people and the people, through the Spirit in the Church, answer their Lord.

        Now you may decide that this is wrong, that the text is, first or even entirely, a sign used by humans to communicate with God, that human understanding of the Liturgy is unimportant. So be it. But in that case, the rational course is to go back to the Latin.

        The massive cost of translations — time, money and agony — could have been saved.

  7. I was introduced to the classics through 7 years of Latin and 4 years of Greek, all intensively taught with the same objectives and mental processes Jonathan Day experienced. Details have faded, but the benefits remain. The aim was to discover and understand very accurately what the originator had said and meant, keeping in mind his purpose, world, era, and audience. The word-for-word, syntax-copying approach suggests to me an oddly inadequate notion of what language does internally and externally to those who are using it. An interesting aspect of the literalist approach is its remarkable similarity to what dumb machines can do to a paragraph today without active human interaction.

  8. I’m confused. I thought that one of the reasons for the new English translation (and an excuse for many of its problems) is that certain language groups have no Latin scholars and so will be making their translations from the new English translation (horrors). I think the languages in question were in the Philippines and Africa but I could be mistaken.

    1. There is a directive in LA (cited above) that all translations are to be made directly from the Latin. As with many ecclesial directives, the experience and process generated by the realities of lived experience (countries with few or no people who understand Latin sufficiently) cause the translation-from-a-translation to occur.
      Some did speculate that it was an “open secret” in the Vatican that the English translation would be the one used in many places (English is the new Latin) as the “base” translation from which to work.

      1. If this were the justification for the nature of the English translation then a simple solution would be to have two English translations: a translator’s version and a liturgical version.

  9. J. Day: “In which case, why not stick to the Latin? That would certainly make the work of music directors easier.”

    Well we don’t have to stick to Latin in total but if we regularly experienced more Latin at Mass we’d make things easier for the musical directors you mention and we’d have the added benefit of actually working to implement V2’s SC which directs us to sing or say in Latin the parts of the Mass that pertain to us. It seems there is much work to do in that regard.

  10. Yes, my undergrad degree was in linguistics — a first at University of Toronto. I knew at first reading of LA during a break from editing the National Bulletin on Liturgy that we were in deep trouble. Those who drafted the document might know several languages but their product was evidence that they had no mastery of language science. It’s not the lack of Latin scholars that is the problem. Language structures vary wildly. How can you demand a mirror of Latin structure when, for example, a language has no prepositions?

  11. Regarding the translation into Japanese of RM3 and other liturgical texts, depending on who is translating a particular text, it seems that the work is done initially either from the Latin or the English and then cross-checked; with other Roman documents, depending on the language abilities of the translator, the translation in other European languages is consulted, eg. French or Italian. We aren’t blessed with a surfeit of latinists, and now given the language used in the English translation of RM3, it is going to be a long haul. Some years back they were asked to provide a ‘literal’ English translation of the Japanese translation of the Latin. I presume that was to show how close the Japanese translation was to the Latin. Given the fact that grammatically there are quite a few structural differences between English and Japanese more than a few questioned the purpose of the whole exercise. Over the years, while teaching English and Religion, I’ve done my share of translation work from Japanese to English, and a literal translation doesn’t result in a coherent sentence, you have to ask what the writer want to say, and then find a way of expressing that in English, while respecting the meaning of the key vocabulary items in the Japanese: going from Latin and/or English applying the same principle produces the best result.

  12. I particularly like Chupungco’s assertion that a translation should produce the same effect on its hearers/readers as did the original language on its hearers/readers. This is what I have tried without success to stress here many times before, with our Latin scholars insisting that it couldn’t/shouldn’t be done. When I read a Latin collect, for example, I feel engaged in a very gracious, worshipful and duly deferential act and tone (effect) towards One whose sanctity is beyond comprehension. The old translation failed miserably at this… it Didn’t Even Try. The new translation tries but all to often results in almost comical amateurishness. But at least it tries. Cranmer did it. Surely, if we really wanted to, we could repeat his success in a gracious 21st century idiom.

    1. MJO, I have a lot of sympathy with your perspective here. If the 1970s translation failed in any area, it was in getting the register of the prose wrong. I don’t think that the translators were trying to convey a rather casual tone, but to my ears, that’s what they did. I believe that they were trying to convey “a very gracious, worshipful and duly deferential act and tone”, but that they often failed to do so.

      The new translation, in contrast, simply mimics the grammatical structure of the Latin. It is the new translators who “Didn’t Even Try”. They might as well have fed the texts into Google.

      I also agree that Cranmer succeeded in this difficult task. So, to a large extent, did the 1988 translators. It can be done. But I think it does require deconstructing the Latin text, working out its logical structure, and then conveying that in the right register in proper English. Mechanical, word-for-word translation of the Latin will not do.

  13. I never thought that a PTB contributor could weave Noam Chomsky into a post. Excellent!

    Jonathan: I think that Chomsky, Comme le Prévoit and my second-year Latin teacher had the right view of translation. But if this view is no longer licit, then we should say Mass in Latin and not bother translating. Perhaps this was the intent from the start. Caesar might well have approved.

    Julius Caesar certainly would have approved of Liturgiam authenticam.

    His memoirs’ literary simplicity and limited themes accentuate ideological value. Caesar is “easy to read” because the same semantic and syntactic constructions appear consistently. Caesar’s limited “Chomsky kernel” symbolic set can be rearranged almost infinitely to generate endless messages with similar content. Perhaps Caesar could battle Pompey into eternity through endless strands of Caesar-esque computer-generated Latin. As Jonathan notes, the Caesar kernel/symbolic set becomes easily recognizable and manipulated after even a limited time.

    Liturgiam authenticam employs an anachronistic and purposefully limited kernel. A strict adherence to the semantics and syntax of the Missale Romanum has produced a less-understandable English liturgy. Perhaps this result stems from a strict adherence to a linguistic-symbolic kernel which sharply diverges from a contemporary English (or Japanese, or Turkish) frame. Caesar’s memoirs are not “alive”. Today’s readers can read Caesar though a limited linguistic kernel because this kernel is static and matched to a particular time in history. Catholic worship is forever alive and evolving.

    Comme le prévoit, though an admirable attempt to create a new linguistic kernel on top of the Latin substructure, cannot fully merge late Latin into the stream of living vernacular. Only vernacular compositions can employ a much more flexible linguistic kernel and frame.

  14. Seeing Chomsky in your headline, Jonathan, certainly grabbed my attention – and the subsequent discussion is proving stimulating. But I’m surprised no-one has yet mentioned Eugene Nida. Pre-Chomsky he may have been in many respects (a bit like me – I learned my lingusitics back in the sixties), but I’m sure his views informed Comme le prévoit more directly – and, being an engaged christian himself, he had much more to say explicitly about the translation of sacred texts. His Towards a Science of Translating is still worth reading even after nearly half a century.

  15. Michael, thank you very much for the pointer to Eugene Nida. I knew of him, but had no idea his work was as interesting as it is proving to be. He is a fine read!

    He cites Chomsky in Toward a Science of Translating and seems to use a lot of Chomsky’s concepts (e.g. ‘surface structure’) in his writing. More important, his thinking follows Chomsky’s in many respects.

    Here are two quotes I picked out. The first is from The Theory and Practice of Translation (Nida and Taber)

    … reproducing style, even on a formal level, may not result in an equivalence, and it is functional equivalence which is required, whether on the level of content or on the level of style.

    and the second is from Toward a Science of Translating. He is discussing the translators of the King James version, where

    a text that could have been a series of tasteless compromises turned out to be a remarkably fine translation, owing to the unusually good sense the translators whowed in matters of exegesis and their extraordinary sensitivity to the style of speech appropriate in public reading. They were refreshingly better than the Rheims-Douay translators, who floundered in a morass of awkward literalness and ecclesiastical verbiage.

    The new translation of the Mass has returned us to that morass.

  16. Jonathan: If the 1970s translation failed in any area, it was in getting the register of the prose wrong. I don’t think that the translators were trying to convey a rather casual tone, but to my ears, that’s what they did. I believe that they were trying to convey “a very gracious, worshipful and duly deferential act and tone”, but that they often failed to do so.

    I know that a lot of people think they got the register wrong, and I used to at one time. But the fact is

    (a) that their motives were similar to those behind SC: to strip down the meaning to its bare essentials without the profusion of rococo ornamentation which conceals the clean lines and which we are now, unfortunately, saddled with, and

    (b) more importantly, it is possible for skilled presiders to pray and give great depth to those 1970 Order of Mass/1973 Missal texts. Regrettably, many people have never witnessed this, but I have and it’s a cause for rejoicing. It’s like the sun coming out from behind the clouds.

    Of course it’s a matter of attitude and technique, rather in the same way that a good presider who takes enough care and time can make the 2010 texts prayable and even spiritually fulfilling. Bad presiders won’t be able to, good ones will, in either translation. I think, though, that 2010 requires much more in the way of mental gymnastics, with a resultant loss of prayer-capacity for the presider himself; and many priests do not see why they should have to make the effort with a Latinate text which, in line with the dictates of LA, is deliberately alien (read “sacral”) and so cannot of itself inculturate, and at the same time lose any sense of prayer for themselves.

    1. I think the effort to strip down to bare essentials was fundamentally mistaken; we lost euphony and musicality of too many texts. That was a fault of CLP, but it is no less a fault of Liturgicam Authenicam (which has additional fundamental flaws), and so that does not mean we need all an effete Latin syntax, however, which is drained of its native gravitas when clumsily converted into English as it has been. Texts that are repeated weekly or encountered frequently can bear well a richer array of idiomatic usage and vocabulary, as can texts that have especially rich ritual purposes. I can hope we will some day have translation principles and praxis that recognize this. Still in search mode.

  17. “Liturgiam Authenticam” sometimes tells us that in liturgy we seek to convey a divine message that transcends individual languages, but elsewhere (20, 56) it tells us that what we really need to convey is Latin words: “The original text, insofar as possible, must be translated integrally and in the most exact manner . . . Any adaptation to the characteristics or the nature of the various vernacular languages is to be sober and discreet.”
    My nominee for most shocking example of this perversion is “On earth peace to people of good will,” a robotic word-by-word rendering of “In terra pax hominibus bonae voluntatis,” which is itself a translation (even though LA 24 forbids twice-removed translation). “People of good will” is inconsistent with Luke 2:14. The phrase ordinarily signifies that the “good will” of the people is theirs, not God’s, although New Testament scholars have all but unanimously affirmed the opposite (as in the NAB’s “On earth peace to those on whom his favor rests”).

Leave a Reply

Your email address will not be published. Required fields are marked *