What Is the Best Way to Learn Latin?

The Complete Transcript

Folio 1r of the Roman Vergil: The Eclogues

2016 saw the publication of two important and exciting books about how to learn Latin. The first, in March, was Eleanor Dickey’s Learning Latin the Ancient Way: Latin Textbooks from the Ancient World (Cambridge University Press), a study of ancient materials used to teach Latin to Greek speakers in the Roman empire. It was followed in October by Reginald Foster’s Ossa Latinitatis Sola/The Mere Bones of Latin (Catholic University of America Press), a Latin textbook using the legendary Vatican Latinist’s teaching methods.

The following is a conversation between Dickey (ED) and Daniel Gallagher (DG), Foster’s longtime student and successor in the Office of Latin Letters at the Vatican. The conversation is led by Michael Fontaine (MF), Associate Professor of Classics at Cornell University.

This is the complete, unedited transcript of the conversation.

MF: Eleanor, your new book is a revelation! It shows that ancient Greeks learned Latin the way we learn modern languages. They memorized made-up dialogues — dialogues that illustrate stereotypical Roman culture — and only then went back and analyzed each word for its grammatical function. By contrast, a reader opening Reginald’s book might be surprised that he advocates an entirely different method: he insists on total philological mastery. It seems completely different, but it obviously works, too. Do you see Reginald’s method as a total break from “the ancient way” (as your title aptly puts it)? Or do you see continuities?

ED: Definitely I see continuities! The most striking one is in Reginald’s frequent suggestion that students should practice manipulating Latin sentences by making small grammatical changes, such as from singular to plural, active to passive, or present to past. This type of linguistic exercise was popular in antiquity; it was called chreia. I haven’t put it in my book because I have no firm evidence of it being used for Latin, but it was very widespread for Greek, and I strongly suspect that it was used for Latin too. Students would take a sentence and run it through a prescribed set of changes, often so that one particular word in it was put into every possible form in turn:

Diogenes (nom. sing.) taught his students well.
Diogenes’ (gen. sing.) students learned well.
I said to Diogenes (dat. sing.) that he had taught his students well.
I said that Diogenes (acc. sing.) had taught his students well.
Diogenes (voc. sing.), you taught your students well.

Compare p. 205 of Reginald’s book, where we find (the Latin for):

I received two letters from you yesterday.
I shall receive two letters from you tomorrow.
I had received two letters from you the day before yesterday.
Two letters from you were received by me yesterday.
He said yesterday that Cicero had received two letters from him the day before yesterday.
He said yesterday that Cicero would receive two letters from him tomorrow.

The main difference here is that Reginald’s changes are less systematic, and therefore all his sentences are sensible. The ancient chreiai usually descend into farce eventually, because in order to put ‘Diogenes’ into every possible form you need to include the dual and the plural:

The two Diogeneses (nom. dual) taught their students well.
All the Diogeneses (nom. plural) taught their students well.

So Daniel, what I’m wondering is whether Reginald knows about the chreiai, or whether he had the same idea independently? In the latter case it might not be a continuity — but it still might be one, in the sense that the very nature of the way Latin and Greek work means that this type of exercise is an obvious way to develop your language skills. Most people nowadays wouldn’t notice that for themselves, but to the ancients it was obvious, because they grew up speaking these languages. The utility of such exercises might have been obvious to Reginald in the same way as to the ancients, because of his intense familiarity with the way Latin works.

And, Daniel, I’m also wondering what you think of the characterization of Reginald’s book as insisting on “total philological mastery.” I would say it doesn’t do so at the early stages; the basic grammar is spread out over three years of study, so for quite a long time students would not have the ability to identify every form they see. This contrasts sharply with some intensive courses, where students learn all the forms in one semester (or even in five weeks flat, in the summer school of the CUNY Latin/Greek Institute).

DG: I agree with you, Eleanor, that Reginald’s method does rely heavily on chreiai. The examples you offer highlight this point. One of the first things Reginald teaches his students is to recognize the difference between anima, for example, as a subject form, and animam as an object form. But he also quickly points out — even before the students have learned the ablative case (or any other case, for that matter) — that anima is not always necessarily a subject form (let’s leave aside macrons for now). From the outset — and I think Ossa illustrates this well — Reginald is more interested in showing students how the Latin language functions and less in having them memorize forms. The forms are absorbed and digested gradually, precisely by doing chreiai day after day.

I believe that Reginald relies on chreiai precisely because, for a host of reasons, he is convinced that the thinking process most conducive to learning Latin today is much different than it was 2,000 years ago. Kids generally are not listening to Latin around the house or on the street. In your book, you do a wonderful job of showing that there never really was a single way of learning Latin, even in the ancient world. The method of learning Latin depended on — among other factors — where and into what social class you were born into, the purpose for which you were learning Latin, and the resources available to you. I think that your scholarly reckoning of the diversity of pedagogical approaches in the ancient world is an extremely important point, because it justifies Reginald’s life-long effort to develop an unorthodox system that allows a contemporary, usually adult learner to master the language in the shortest amount of time possible.

I should point out that, at least in the classroom, Reginald’s chreiai do in fact descend into farce. I have inherited this farcical tendency myself as a teacher. I often ask my Christian students to become pagans for a moment by having them put a sentence like this from Augustine …

et quis locus est in me quo veniat in me deus meus, quo deus veniat in me, deus qui fecit caelum et terram? (Confessions, 1.2.2).
And what is this place in me where my God comes into me; where God comes into me — the God who made heaven and earth?


et qui loci sunt in nobis quibus veniant in nobis di nostri, quibus di veniant in nobis, di qui fecerunt caelos et terras?
And what are these places in us where our gods come into us; where gods come into us, the gods who made the heavens and the earths?

Reginald calls this a total “reversal”: i.e., everything in the singular into the plural, and everything in the plural into the singular. As you suggest, whatever farce it falls into today is equally as useful as it was back then.

So in answer to your first question, Eleanor, Reginald does indeed know about the chreiai, but I think he incorporates them into a system not so much based on the memorization of forms as the recognition of them and, subsequently, the mastery of them precisely by taking a single five-word sentence from Cicero and, as Reginald often says, “finishing the language with it” or “exhausting it so that there is nothing more to do.”

Your second question is based on the accurate observation of how what Reginald understands by “total philological mastery” is quite different from what most intensive courses present as such. His presentation of morphology is indeed spread out over at least a two-year period (generally corresponding to his Experiences I and III). More importantly, after that two-year period, it never goes away. Reginald is constantly choosing passages for his students that guarantee that. His odium for a crash-course approach to forms is evident in the fact that, within an eight-hundred-some-page tome, the reader is hard pressed to find a single table, chart, or paradigm. My point is this: by “total philological mastery,” Reginald does not mean memorizing forms, but taking any entry in your dictionary — regardless of whether you’ve looked it up, got it from your teacher, read it off your ludus domesticus, or retrieved it from your memory bank — and doing anything you want with that word. In this sense, Ossa is truly a handbook. Lewis and Short and the entire body of Latin literature is the textbook.

ED: It’s awe-inspiring to think of the entire body of Latin literature as one’s textbook, but I see what you mean: Reginald’s whole method is clearly a big-picture one when it comes to the range of texts used (not to suggest that he wouldn’t pay attention to detail within a text, of course!), and emphatically against picking out easy stuff. The first reading sheet in his book is from Horace, an author so hard that after 35 years of studying and teaching Latin I don’t think I’m up to reading Horace yet. In this respect Reginald’s method is certainly different from that of the ancients, who believed in starting beginners off with something nice and simple that they could master easily. And much as I admire Reginald, in that respect the ancient method makes more sense to me, because it allows students to practice more. Realistically, students learn not from what teachers say, but from what they do themselves: it is the direct encounter between student’s brain and Latin text that really causes learning, and all we teachers can do is facilitate that encounter. I see that facilitation as a fine art. If you give students a task that is just challenging enough to be fun but not so challenging as to be discouraging — for example, a text that they can actually read by putting in some work but not too much — they do a lot of it, enjoy it, and learn from it. Whereas if you give them something too hard, they either do only a small amount or not even that, with the result that they learn less.

Easy doesn’t have to mean inauthentic, since there is plenty of fairly easy Latin out there (especially for people who take a big-picture view of what counts as Latin literature); indeed some of it can be found in Reginald’s book, in the reading sheets for later Experiences. But he makes clear that he had a deliberate policy of choosing the readings at random, to give as it were an unbiased sample of what is actually out there in the world of Latin, rather than trying to find something particularly suitable for a given class. Which leads me to wonder: since texts that are easy are also part of what is out there, what is wrong with choosing easy texts for the beginners’ first year?

There is definitely something that I do not understand about Reginald’s method, namely what the students are actually doing. I’m sure they must be doing something, otherwise they wouldn’t learn, but they can’t actually be reading Horace in their first year (they could be listening to Reginald explain Horace, but that’s not the same thing). The key must be in those ludi domestici you mention, which are not in this book but are promised for volume 4 (this being volume 1 of a projected 5-part work). So can you tell me anything about these ludi, particularly ones that a student might do at a very early stage of learning Latin?

DG: Eleanor, you have definitely indicated one of the most unique and baffling aspects of Reginald’s pedagogy: namely, that from day one everything in the Latin literature is “fair game” for students. Your specific example of Horace is a good example.

The point Reginald wishes to emphasize by this virtually simultaneous presentation of the Latin language is that it is all one language. Despite the differences of style and vocabulary between authors and eras, a student should be able to master and deal with the entire language rather than with single author, time period, or style. And a student should start dealing with all these authors from the first encounter.

But it naturally brings up the question, “what are the students doing?”, especially in those first days? It’s difficult to explain exactly what they are doing until one experiences Reginald directly in the classroom, but basically he is asking beginning students to recognize, understand, and manipulate only those forms they have learned and nothing more.

So, let’s presume the students have had a couple of weeks of class with Reginald. What would they be able to do with the first two lines of Horace’s Epistula 1.8?

Celso gaudere et bene rem gerere Albinovano
Musa rogata refer, comiti scribaeque Neronis.
O Muse having-been-invoked, bring to Celsus Albinovanus, comrade and secretary of Nero, (greetings that he) be happy and do well!

Well, I admit not much. But by means of these lines, Reginald will illustrate in class the principles he presents in Encounter 1. The word order, at least to students whose native language is English, is strange. The subject Musa and the main verb refer are in the second line. “The position of the words,” as Reginald puts it on page 3, “is not fixed.” To further illustrate the point, Reginald would show the students that the words Celso and Albinovano go together, even though they are on opposite ends of the first line (he expects students to “ooh” and “ahh” at this point even though the fascination wears off over time). He would then point out that the only way you know that the two words go together is because you “know your vocabulary” (principle 8 on page 4) and you know your endings (principle 3 on page 3). In other words, you know (or in the case of the beginner, you will eventually learn) how the word works precisely by seeing how it appears in the dictionary, and then you will see (or you will learn) that both Celso and Albinovano are in the dative case (principle 4 on page 3).

Reginald would then immediately point out that not all words ending in –o go together, and that words can go together even though they don’t have the same endings, as with comiti and scribae (which, of course, also go with Celso and Albinovano). All of this can only be known by (1) knowing the dictionary, and (2) knowing which endings indicate which functions. After years teachers of Latin take all of this granted, but they cannot take it for granted that their students will take it for granted.

Reginald would also ask beginning students where you find gaudere and gerere in the dictionary. Of course, we will find them under gaudeo, gaudere and gero, gerere (and not under gaudere and gerere). A beginning student will already have learned that these forms are infinitives (Encounter 8) and would translate them appropriately. A student would only have to complete Encounter 3 to know that “-que” at the end of scribaeque means “and”, and of course the same student would immediately recognize et in the previous line.

From Encounter 2, the beginning student will know that if musa comes from musa, musae in the dictionary, then the object form is musam. Students who have had eight Encounters will also know how to use the perfect stem system of gero, gerere, gessi, gestus, -a, -um. So he or she could say the following sentences in Latin after just eight encounters:

I carried the muse: gessi musam.
The muse carried the muse: musam gessit musa.
We carried the muse: gessimus musam.
You carried the muse: musam gessisti.

These are the kind of things students would be doing on their ludi domestici.

So, in short, in the earliest days of Latin with Reginald, students are “looking” for verbs, nouns, adjectives, etc. based on their ability to use a dictionary and recognize the forms they have learned. The stuff they don’t know is pointed out by Reginald to illustrate the principles they learned in Encounter 1.

Later, I’ll have a word to say how spoken Latin fits into all of this. But for now, Eleanore, I want to marvel at how you’ve collected a wonderful sampling of resources for learning Latin from the ancient world, and how they help to illustrate what is so unique about Latin in comparison to modern and other ancient languages. Particularly fascinating is the focus of grammars on those things that the audience doesn’t know, as evident in your presentation of Dositheus’s treatment of the ablative case (p. 88 and following). You point out that the audience would not have been so familiar with this case, which “from a Greek perspective was sometimes equivalent to a date and sometimes to a genitive.”

This would be similar to Reginald’s insistence that not call cases should be presented at the same time. Reginald finally gets around to presenting the ablative case in Encounter 27, because it will be the least familiar to native English speakers. But once he does introduce it, he floods the learner with copious examples to emphasize both the frequency of the case and the flexibility it has in expressing different albeit similar concepts.

Personally, I find Dositheus’ presentation of the ablative case one of the most delightful aspects of your book. To me, it illustrates the importance of recognizing “where students are coming from,” so to speak, and that is essential for good pedagogy, both in the ancient world and for today.

So, I’m wondering if you could say a few more words about what your research has revealed about the sensitivity of ancient pedagoges to their students and their prior linguistic backgrounds? I suspect there will be more interesting parallels to Reginald’s pedagogy.

ED: Thanks for that explanation of the ludi domestici! Yes, recognizing where students are coming from is an essential of good teaching, anywhere and any time — because how do you help someone get from A to B if you don’t know where A is? And in antiquity A was definitely not where it is today. Ancient Latin students were mostly coming from a knowledge of Ancient Greek, which means that they had no trouble with some aspects of Latin that modern students often find hard. Free word order and the use of cases, for example, were just taken for granted by ancient students, and the reason Dositheus concentrates on the ablative is that the other cases (nominative, genitive, dative, accusative, and vocative) also existed in Greek and therefore posed no conceptual difficulties for Greek-speaking students: he doesn’t bother to explain those cases at all. But ancient students had a lot of trouble with the Latin alphabet, something that their modern counterparts usually take for granted. They really struggled to learn the alphabet, and some just gave up and learned Latin in transliteration. So we find ancient copies of dictionaries, grammars, and even little learners’ dialogues with the Latin in Greek script, like this:

“βενε νως ακκιπιστι ετ ρεγαλιτερ, ουτ τιβι δεκετ.”
“νη κοιιδ βουλτις ικ δορμιρε, κουοδ σερω εστ;”
“ετ ιν οκ γρατιας αβημος.”
“You have entertained us well and royally, as befits you.”
“Do you want to sleep here, because it is late?”
“For this too we are grateful.”

That kind of thing would really not be helpful for a modern beginner! In theory transliteration might be helpful for a modern student of Ancient Greek, but we don’t use it: even though students sometimes find the Greek alphabet a challenge, no-one tries to learn ancient Greek in transliteration. That’s because if you learn a foreign language in transliteration, you end up being able to speak and understand it, but not read or write it. That kind of knowledge would be pointless nowadays with ancient Greek, but in antiquity an oral command of Latin without the ability to read or write would go a long way: ancient society depended less on writing than ours does, and many people were entirely illiterate. Probably the main goal of all ancient Latin learners was a good command of the spoken language, and literacy in Latin was a secondary goal even for those who worked on it at all. Which is another interesting point of comparision with Reginald’s teaching, isn’t it?

But Dositheus also did something else important in terms of paying attention to where his students were coming from: he recognized that they did not necessarily know the material that they were supposed to have learned before getting to his classes. In antiquity, just like today, there were certain basics that children were supposed to learn at school, and which one doesn’t normally teach at university for that very reason. But a good university Latin teacher knows that you had better check, before using a word like “noun” or “subject,” that students really know what those words mean. Dositheus was in exactly the same position: at the start of his Latin course he went over explanations of the parts of speech, for example, despite knowing that the students should have learned those in school. He also went over the different punctuation marks and their meanings, which indicates that he did not take it for granted that the students would know what periods (full stops) and commas were. Now I personally have never discussed punctuation with a Latin class, but looking at the work produced by some of my students, I now wonder: should I maybe just tell them what periods and commas are? Maybe there are teaching tips we can pick up from Dositheus despite the fact that his students had such different backgrounds from ours…

DG: Eleanor, I think you have touched on one of the most interesting parallels between ancient methods of teaching Latin and Reginald’s: i.e., the goal of helping students develop the skills they will find most useful. Whereas once the ability to speak Latin was the learner’s greatest asset, today it is literacy. Developing our students’ ability to understand and digest ancient texts is the reason most of us have dedicated our lives to teaching Latin.

So what, according to Reginald, is the place of spoken Latin today?

It is a common misunderstanding that Reginald aims to instill oral proficiency in his students or that he speaks to them exclusively in Latin during class. Neither is true. What he aims at is helping students to deal with any Latin text by any author from any time period. Anything that enhances this capacity should be used liberally by the teacher, including well-designed oral exercises and drills. These include translating newspaper and magazine articles, expressing one’s personal thoughts and desires in Latin, and chatting about the World Cup. Insofar as such activities are aimed at helping students to incorporate pure and genuine Latin usage, they are encouraged and enjoyed. “The issue” when it comes to spoken Latin, Reginald writes, “is not simply a matter of adding vocabulary, because we can always develop ways of speaking about space ships, automobiles, cell-phones, and the internet.” Rather, “the greater difficulty for us concerns how to deal with the language, its flow and natural structure” (p. 204, emphasis mine).

Within the sequence of the five “experiences” that make up Reginald’s curriculum, the Second is dedicated to spoken Latin. This Second Experience occupies a mere fraction of Reginald’s 800+page book. He describes the Second Experience as “an immediate introduction to living, spoken Latin with no notes and no commentaries, but just the use of the language” (Ossa, p. 203). This means that, like the other four experiences, the Second Experience is embedded in real, unadulterated Latin literature taken from every time period. Wheelock, quite frankly, is the antithesis of Reginald’s approach in that it takes classical loci and manipulates the grammar and word order therein to make them more digestible to a student who has had only a year or two of formal Latin training. Reginald simply asks, why not give them the real stuff? For only by giving them the real stuff will they have a genuine taste of the “flow and natural structure” of Latin. So by making students speak in that flow and natural structure, they will more naturally develop a taste for how the ancients themselves spoke.

To counter the Wheelock approach, Reginald has collected five hundred short sentences from Cicero’s letters: the famous “500” (they do not appear in Ossa, but perhaps they will in a subsequent volume). These passages are taken directly from the Ciceronian corpus with no alteration of word order, verb tense, or mood. After translating the original sentence, the student is invited — using “no notes and no commentaries” — to say whatever he or she wants in Latin.

So, from a sentence like duas a te accepi epistulas heri (“I received two letters from you yesterday,” Att. XIV.2.1, given on page 205), a student can derive:

Duas a te accipiam epistulas cras.
Duas a te acceperam epistulas nudiustertius.
Duae a te sunt a me acceptae litterae heri.
Dixit heri Cicero duas a te se accepisse epistulas nudiustertius.
Dixit heri Cicero duas a te se accepturum epistulas cras. (all on p. 205)
I shall receive two letters from you tomorrow.
I had received two letters from you the day before yesterday.
Two letters from you were received by me yesterday.
Cicero said yesterday that he had received two letters from you the day before.
Cicero said yesterday that he would receive two letters from you tomorrow.

Or, in Ad Atticum I.9, we have: Peto abs te, ut haec diligenter cures. From this, Reginald might ask students to derive sentences such as:

Peteris ut haec a te curentur.
Petebam ut haec diligenter curares.
Petebaris ut haec a te curarentur.

until the entire Latin language is “exhausted,” or everything that could be said in Latin using those words has been said.

So, in Reginald’s opinion, the ultimate purpose of spoken Latin is to improve Latin reading comprehension. Those who wish to learn how to order in a restaurant, talk about the weather, or make travel plans in Latin are doing nothing wrong, provided that the acquisition of such skills contributes to understanding Plautus, Cicero, Virgil, and Ovid.

I want to add, Eleanor, how interesting I found the papyrus containing Cicero’s first Catilinarian oration that you discuss and reproduce on p. 144 of your book. That’s precisely the kind of text that Reginald would have his students “manipulate” in every direction to hone their grammatical and stylistic skills. It is the perfect passage for a “ludus domesticus” at any level.

Also interesting is your comment that Dositheus paid close attention to what was lacking in his students’ knowledge of basic grammar. This relates to an aspect of Reginald’s method that has caused some controversy: his elimination of traditional nomenclature. This is a large can of worms, so I want to limit myself to one example. Reginald consistently refers to the accusative case as the “object form” precisely because many students have not learned what a direct object is. By replacing “accusative case” with “object form” he has essentially eliminated a layer of confusion. Although I admit that he may be adding a layer of confusion for those who do know some Latin and are used to referring to animam as the accusative case rather than the “object form”. In any case, his replacement of traditional nomenclature with his own is not without good reason.

Eleanor, I want to thank you for your book and for this interesting discussion. I wish it could go on! As I add Learning Latin the Ancient Way and Ossa to my toolbox and use them in the classroom, perhaps we can exchange ideas further on how to best help students unlock the treasure chest of Latin literature!

MF: As I reflect on this conversation, Eleanor and Dan, a famous remark Cicero makes in the Brutus keeps coming to mind:

nam ipsum Latine loqui est…in magna laude ponendum, sed non tam sua sponte quam quod est a plerisque neglectum: non enim tam praeclarum est scire Latine quam turpe nescire.
Just getting Latin right should be well regarded, but not so much because it’s intrinsically impressive as that most people can’t be bothered with it. The reality is that it’s not as impressive to know Latin as it is embarrassing to not know it.

We are beginning to investigate, reflect on, and debate pedagogy more and more. As we do, the truer it seems to be that there’s not just one way to learn Latin. The beauty of these two new books is that they show us two radically different approaches, and both of them clearly do work well. As we’ve seen, they complement each other in surprising ways. Anyone interested in learning or teaching Latin could learn a lot from both of them — and should do it right away, since, as Horace quips, dimidium facti, qui coepit, habet: “Once you’ve started, you’re halfway there.”

Eleanor Dickey is is professor of Classics at the University of Reading in England and a Fellow of the British Academy. She is the author of numerous books and articles on the Latin and Greek languages and how they were understood, used, and studied in antiquity, including Learning Latin the Ancient Way (Cambridge 2016). She has also spent decades teaching both languages to students in England, Canada, the United States, and the Netherlands.

Daniel Gallagher, who succeeded Reginald Foster in the Office of Latin Letters at the Vatican Secretariat of State, is the Ralph and Jeanne Kanders Associate Professor of the Practice in Latin at Cornell University. He also teaches at the Paideia Institute for Humanistic Study. In addition to helping students “experience” the Latin language according to Foster’s thought and system, he lectures in philosophy and has published numerous articles in metaphysics and aesthetics.

Michael Fontaine teaches Latin at Cornell. His reviews of Learning Latin the Ancient Way and Ossa Latinitatis Sola will soon appear in Gnomon and The Weekly Standard respectively.

Published by the Paideia Institute. You can read more about the journal, subscribe, and follow it on Facebook and Twitter.

The views expressed in this article are those of the author and do not reflect the views of the Paideia Institute.