Eleanor Dickey shows how Greek speakers learned a foreign language.

What was it like to learn Latin during the Roman empire? We often think that it must have been easy, because Roman children grew up speaking the language. But, as the empire expanded, many Greek speakers became Roman citizens, and some of them learned Latin. How did these people learn? Did they have textbooks, grammars, or dictionaries? Did they take formal classes? Did they read Cicero, Caesar and Virgil? What were their teachers like? Did they do prose composition? And how can we know any of this?

To take the last question first, one way we know about ancient Latin learning is that a set of ancient Latin textbooks has survived; they are known as the Colloquia of the Hermeneumata Pseudodositheana. These textbooks are bilingual, written in narrow columns with the Latin and Greek matching line for line. The idea is that a beginner would be able to study the Latin and work out exactly which bit meant what by using the translation as a glossary. Here is one of the prefaces to the Colloquia, with the original students’ native language, Greek, replaced by English to give a sense of how the line-for-line translation originally worked:

This image has an empty alt attribute; its file name is picture1.jpgGiving language students a running translation seems odd to us, but that is because we take for granted that texts will have spaces between the words. The ancients usually did not divide words (nor use punctuation, nor distinguish capital from lower-case letters), and therefore language learners needed a lot of help at the early stages. Until they had learned enough vocabulary to enable them to recognize where the words divided, ancient learners could not even use a dictionary. Because although Latin-Greek dictionaries did exist, you cannot look up a word in a dictionary unless you know where it begins and ends. So in fact the ancient experience of the passage just quoted (Colloquium Celtis 1–2) was more like this:This image has an empty alt attribute; its file name is picture2.jpg

Parts of these Colloquia tell the story of a boy who goes to school and learns Latin. These stories are wonderfully helpful for us, because they show exactly what learners did. For example, here are parts of the description of school in the Colloquium Celtis (19–39):

"I entered the school and said, ‘Hello, teacher!’, and the teacher greeted me in return. He handed me the book-stand and ordered me to read five columns of writing to him, and I read accurately and fluently. Then I passed the book-stand to another student and went to the teacher’s assistant. I greeted him and the other students, and they greeted me in return. I sat down in my regular place on the bench, ... and when I was seated the boy who carried my books passed me my writing tablets, case of styluses, ruler, counting board, and counters. ... First I erased the writing on the tablets and ruled lines following the model, and then I wrote. I showed my writing to the teacher, and he praised me because I had written well. I read aloud what I had written, with pauses in the right places. I recited. ... I was given an assignment to do. An unfamiliar work was explained to me. I was given a passage, and along with some other students I read it immediately; others read the same passage after preparing it carefully.

The little children practise in front of the teacher’s assistant: language textbooks and syllables, the inflection of the verb, the whole grammar book, conversation, the cases of nouns, the genders of nouns, their numbers, compositional status, and inflections, words in alphabetical order, letters, vowels, and consonants. They pronounce their readings with the right pauses and intonations. Then they go through lists of nouns classified by subject, and the eight parts of speech.

Then there is silence. The more advanced students go up to the teacher; they read a reading about the Iliad and another about the Odyssey. They are assigned a passage, a rhetorical exercise, a history, a comedy, stories, an explanation of the causes of the Trojan war, the basis for a recitation, the speeches of Cicero, Virgil, Persius, Lucan, Statius, Terence, Sallust, Theocritus, Thucydides, Demosthenes, Hippocrates, Xenophon, and the Cynics. Then they all go back to their seats, and each one does the work assigned to him: one writes and another learns a speech. When they have done their work they recite in order, each according to his ability. If someone has recited well, he is praised; if badly, he is punished. We are dismissed around the seventh hour."

But unfortunately, there is a complication. This cannot really be an authentic description of children learning Latin in school, because Greek speakers did not learn Latin as children. Their initial education was entirely in Greek, and only around university age did some of them decide to learn Latin. Latin speakers, by contrast, did learn Greek in school. In fact the Colloquia were originally composed for Latin-speaking children, to help them learn Greek, and were later adapted for use by Greek-speaking adults learning Latin. In the passage just quoted you can see chunks of material that must go back to the Latin-speaking children (the list of Greek authors and the references to a school and to little children) interspersed with material that must have been added during the adaptation for Greek speakers (the list of Latin authors and some of the grammatical topics in the second paragraph: although all those topics were probably studied by language learners in antiquity, it is unlikely that little children covered more than one or two of them in a single lesson).

But how do we know that Greek speakers did not learn Latin as children? Fortunately, we have some narrative sources in addition to the Colloquia, including the writings of Libanius and other authors who tell us about the education system in the eastern Roman empire. We also have fragments of exercises from Egyptian schools and (separately) from Egyptian Latin classes; these were written on papyrus, wooden tablets, and ostraca (pieces of broken pottery) and have been preserved by the dry conditions of the Egyptian desert until being found by archaeologists.

These materials provide other information besides the age at which Greek speakers learned Latin. For example, the exercise fragments tell us which texts ancient Latin students commonly read. By far the most popular student texts were Virgil’s Aeneid and Cicero’s Catilinarians; Sallust and Terence were also occasionally used, but Caesar apparently not. Students never read isolated sentences, only connected passages, and they never translated sentences into Latin – but they did sometimes translate passages into Latin, usually fables. Their prose compositions are often pretty good, indicating that at least some ancient Latin students succeeded in learning Latin fairly well.

Yet learning Latin must have been pretty hard in antiquity. The exercise fragments also tell us that the texts, glossaries and other aids used by ancient students often contained mistakes. Now of course most copies of most texts contained mistakes in antiquity; that was the natural result of producing books by copying them out by hand. But an error that would not cause much trouble for a native speaker could easily make a sentence hopelessly indecipherable for a language learner. Moreover, many ancient Latin teachers were not native speakers of Latin but rather Greek speakers who had themselves learned Latin as a second language. Some of these probably knew Latin very well, just as today there are plenty of teachers who have acquired an excellent command of Latin by learning it in Latin classes. But some ancient Latin teachers apparently did not know Latin very well, or else they were not very diligent as teachers – for otherwise they would have corrected the mistakes in the teaching materials before passing them to the students. Of course, sometimes we find exercises where someone has indeed written in corrections. But often the corrections do not fix the problems; in fact sometimes they make matters worse, by changing things that are right rather than those that are wrong.

On the other hand, some of the ancient teaching materials were perfectly good. For example, we have a set of Latin noun paradigms produced for Greek speakers in late antique Egypt (P.Louvre inv. E 7332). One of these reads as follows:

This image has an empty alt attribute; its file name is picture3.jpg

This way of presenting a paradigm is basically similar to how we would present it, but there are some differences. The dative and ablative plural endings have been abbreviated, so that aequorib῾ is written instead of aequoribus. The order of the cases is different, following the normal ancient practice: ancient grammarians always used the order nominative, genitive, dative, accusative, vocative, ablative for Latin (and the same order, minus the ablative, for Greek). The cases are indicated by forms of hic haec hoc ‘this’ instead of by their names; there were Latin names for the cases, but people found it more efficient to write huius than genetivus singularis ‘genitive singular’. And perhaps most strikingly, not only the paradigm itself but also the heading and the labels for the individual forms are in Latin; the only Greek visible here is τὸ πέλαγος ‘the sea’, which translates aequor. Nowadays everything except the paradigm itself would be in the learner’s native language, but in antiquity grammars were always written in the language they described: a grammar of Greek was in Greek and a grammar of Latin was in Latin, even if the Greek grammar would be used by Latin-speaking students and the Latin grammar by Greek-speaking students. So a translation of that ancient paradigm into modern terms would be something like this:

This image has an empty alt attribute; its file name is picture4.jpg

But the biggest difference between the ancient Latin-learning experience and our own may actually have been the alphabet. Unlike modern English speakers, ancient Greek speakers needed to learn a new alphabet before they could read Latin. That learning was just as difficult for them as learning the Greek alphabet is for us; numerous finds from Egypt show students struggling with the Latin letters. Some appear to have given up entirely and tried to learn Latin in transliteration. The fragment below comes from a papyrus with Latin verb forms in transliteration (P. Strasb. inv. g 1175); only three forms of each verb are given. Can you work out which forms they are, and which verbs they are from?

νοκετ, νοκες, νοκεω, τονατ, τονας, τονω, ουιδετ, ουιδες, ουιδεω

The person using this papyrus must have been aiming for oral fluency in Latin without literacy – another difference from today’s students, most of whom aim for literacy without oral fluency. But only a minority of ancient students used transliterations. Just as today’s Latin students are a diverse group with very different reasons for learning the language, so in antiquity Greek speakers studied Latin for different purposes, from merchants hoping to do business with Romans to law students who needed to read Latin legal works. They all had a learning experience fundamentally different from our own, but with occasional delightful similarities.

Eleanor Dickey is Professor of Classics at the University of Reading. She is the first person ever to edit and translate the Colloquia and has made the most of that opportunity, publishing three different versions: Learning Latin the Ancient Way (Cambridge 2016) for Latin students, Stories of Daily Life (Cambridge 2017) for those who want only the translation, and The Colloquia of the Hermeneumata Pseudodositheana (Cambridge 2012–15) with full scholarly edition, translation, and commentary.