What Is the Best Way to Learn Latin?

Eidolon (“an online journal for scholarly writing about Classics that isn’t formal scholarship”) presents a conversation between Eleanor Dickey, author of Learning Latin the Ancient Way: Latin Textbooks from the Ancient World, and Daniel Gallagher, who studied with Reginald Foster, author (with Daniel McCarthy) of Ossa Latinitatis Sola/The Mere Bones of Latin, “a Latin textbook using the legendary Vatican Latinist’s teaching methods.” The conversation is led by Michael Fontaine, Associate Professor of Classics at Cornell University. I have to say, Foster’s insistence on “total philological mastery” sounds off-putting to me, and I agree with Dickey when she says:

Reginald’s whole method is clearly a big-picture one when it comes to the range of texts used, and he’s emphatically against picking out easy stuff. The first reading sheet in his book is from Horace, an author so hard that I don’t think I’m up to reading him after 35 years of studying and teaching Latin. 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.

Much as I admire Reginald, in that respect the ancient method makes more sense to me. 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. 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 (but not too much) work — they enjoy it and learn from it. If you give them something too hard, they either do only a small amount or not even that, and they learn less.

But the whole discussion is thoughtful and interesting. Thanks, Trevor!

Comments

  1. Gallagher (very) recently left the Vatican to join the faculty at Cornell as a professor of practice: http://cornellsun.com/2017/01/30/cornell-classics-appointment-marks-tectonic-shift-for-the-study-of-classics/. A very weird and surprising move, both for him to want the job and for Cornell to offer it. By all accounts he is a gifted and effective pedagogue, but typically spoken Latin has been looked down upon within the discipline, even as a tool to recruit students to the language — interesting to see where this goes.

  2. To address the question posed by the title, how about:

    Skim the paradigms, start with Jerome, pronounce it like Italian, read as much as you can, and don’t waste time analyzing the grammar.

  3. Greg Pandatshang says

    @Ken,

    Forsaking the hard Cs of the new pronunciation? Barbarism!

  4. Did you ever try to sing Palestrina using the reformed pron? Barbarism!

    According to the article Gallagher speaks ecclesiastical Latin, presumably Italianate modo sedis apostolicae, and I assume that’s what he’s going to teach as well.

  5. Horace, an author so hard that I don’t think I’m up to reading him after 35 years of studying and teaching Latin.

    Is this some kind of weird classicist humblebrag, or what is this?

    I myself never learned any Latin to speak of because I learn languages by reading and I basically never found anything worth reading in Latin[1], but I flatter myself that if I put in 35 years of hard graft I’d probably have a decent working knowledge? (Not really a testable assertion at my current advanced age, of course.)

    [1] My tastes are not very classical, what can you do?

  6. “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 (but not too much) work — they enjoy it and learn from it.”

    Sounds like Krashen’s input hypothesis, which was very fashionable among second-language teachers in the 1970s and 80s: i+1 and all that.

  7. not leoboiko says

    @martinb …which is still “fashionable”, because it models exactly the experiences a lot of us had when learning foreign languages as adults. It’s how I learned English, for example; from videogames to comp-sci textbooks to Tolkien to Melville, with nary a single grammar lesson or flashcard, ever.

    (The only thing I’d add is a crash-course, 1-day deliberate study of the target language’s phonology and orthography, right at the start; and yes, please don’t pronounce Latin as if it was Italian.)

  8. ə de vivre says

    The best language teachers I’ve had always struck a virtuous Aristotelian mean between giving contextual chunks to memorize, more mechanical grammar lessons about how those chunks worked, and maybe most importantly a love of whatever the language is being used to express—my favourite Turkish class involved lengthy discussions about the gender politics of the soap opera Aşk-ı Memnu. Although I’m sure where that virtuous mean lies varies according to the taste and background of any given adult learner.

    I’ve probably mentioned it before, but verb paradigms go back to the 18th century BC in Sumerian-Akkadian tablets. It’s unclear, though, if they were used for learning or seen as an end in themselves. “List out every permutation of thing X” seems to be a robust form of knowledge production in the Old Babylonian period. There’s a mathematical text, for example, that goes through every permutation of a word-problem about man-hours and ditch-digging to give algebraic problems that wouldn’t have been solvable at the time. So it may be that these verb paradigms were a display of the scribe’s erudition rather than drills for beginners. It’s more probable that students learned from another perennial method, excerpts of texts arranged to show certain grammatical features, which have survived as “proverb collections”.

  9. Unless, of course, you’re Italian.

  10. Even if you are Italian. Otherwise you fall into the error of all modern Greeks except Nick Nicholas, namely that Greek has always been pronounced the same way right back to Homer.

    But it’s true that Italians are more likely to get the geminated consonants right. Besides, Sardinians are Italians nowadays, and their language “imitates Latin as monkeys imitate men”, as Dante rather rudely said.

  11. Well, of course you can learn the restored pronunciation even if you are Italian, but I’m not going to waste my precious time telling them to do so. See: Greeks.

  12. I use the Italian pronunciation for all authors from Augustine forward. And when singing, of course. Otherwise it’s Erasmus all the way.

  13. The only concession I make to non-classical pronunciation in singing Latin is [v] for /w/, as I admit that the wobble-wobble wonsonant is not esthetic when sung. But I do not admit that there is anything wrong with [k] and [g] before front vowels, and indeed consider them modestly superior to their variously palatalized equivalents.

    I don’t know of any Erasmian pronunciation of Latin, only Greek.

  14. Thanks. I had the impression Erasmus had reestablished the classical pronunciation of both languages.

  15. The argument has been made that Erasmus was taken in by a practical joke.

    The only thing worse than the Erasmian pronunciation of Greek is the reconstructed pronunciation with tones.

    Students with virtually unlimited access to both recorded and live native speakers of Chinese need months if not years to get the tones right, and many of them never succeed. Yet there are people out there who have convinced themselves that they are pronouncing the tones in ancient Greek correctly based on little more than the shapes of the diacritics chosen to represent the tones.

  16. >I use the Italian pronunciation for all authors from Augustine forward. And when singing, of course. Otherwise it’s Erasmus all the way.

    I try to use the Latin accent of the author or main subject – for anything on the Severan dynasty, a Punic lisp; for Cato, I follow his intentionally archaic pronunciations, many an S becoming an R, and I render it in an acerbic drone.

    Yeah, right. Sure I do.

    But Rodger, are you serious that you let your Latin accent change with your material?

  17. On the topic of pre-modern language learning tools, there is a Chinese textbook written in Korea around 1300 which has survived. Maybe “textbook” is too strong a word, since it just contains dialogues written in Chinese characters and nothing else. Presumably you would get a native speaker to read it out loud for you. Are there many other examples of ancient or medieval language learning tools that have survived?

  18. not leoboiko says

    I found another Eidolon article about learning Latin, from a teacher who uses a linguistics-aware, reading-based method: Teaching Latin to Humans.

    I’m about to unexpectedly immigrate to Germany, and I’ll apply similar methods to German myself up. Ask me in a few months whether or not it’s working 🙂

  19. I don’t think anyone thinks the acute was rising tone and the grave was falling tone, as in Pinyin. Rather, the circumflex represents high pitch on the first mora of a stressed bimoraic syllable, and the acute represents high pitch on the second or only mora of a stressed syllable. Grave represents the lack of high pitch, and in some MSS was also written on all unaccented syllables. Even if it was the case that the prominent syllables had low pitch rather than high, as in Swiss German, at least we understand the structure of the system.

  20. not leoboiko says

    But Rodger, are you serious that you let your Latin accent change with your material?

    There is, as far as I know, no widely accepted, de-facto-standard reconstructed pronunciation of Classical Japanese (there are several linguistic reconstructions, and the most recent ones tend to be compatible with each other, but so far most classicists in the literary side seem to just use Modern Japanese pronunciation, as of course do the Japanese themselves). I favor reconstruction, natürlich, but I often find myself at a loss when reading classical texts; generally I use my own understanding of early Heian pronunciation, for convenience, because that’s the one which was settled down into the orthography; but it postdates a lot of drastic phonetic changes from Old Japanese, and antedates a few important changes which affected most of the extant literature. (The window of time when the orthography was decently phonological was quite small, and most important works fall outside it.)

    I think a good argument may be made to use at least two reconstructed pronunciations, one for Old (which has a substantially distinct grammar/vocab anyway) and one for Middle onwards. Perhaps we can defend a single reconstruction for Middle on the grounds that Classical writing is demonstrably archaizing (as in Latin, they kept writing certain morphological forms long after they were gone from speech); they wrote imitating Early Middle works, so why not pronounce it as Early Middle people would? An archaic, “original” pronunciation certainly seems congruent with the aesthetic/literary values of most Classical writers, at least.

  21. are you serious that you let your Latin accent change with your material?

    I don’t do Latin much, but I certainly change my pronunciation of Greek depending on the era. (I may have mentioned before that this knack allowed me to give a tour de force performance as Dionysus in The Bacchae, in which I used reconstructed ancient pronunciation with everyone but an actress who was going to Greece for a semester and thus was using modern pronunciation, which I used with her for collegiality and continuity. It was fun!)

  22. David Eddyshaw says

    I was (in hindsight) fortunate in Latin learning. I started in school in Glasgow when I was ten years old, where the language was taught in the good old-fashioned way, i.e. by physical pain (corporal punishment was by no means used only for disciplinary purposes and was virtually inescapable even for angelic little boys such as I was.) When we moved to England (land of the soft) I was exposed to the Cambridge Latin project, which was light on morphology-learning and left many of my classmates adrift. What it was very good on was interesting texts, both composed and eventually authentic. Catullus is ideal for irreverent teenage boys. My previous Glaswegian experience left me able to recognise flexions without pausing to remember them, so unlike many in my class I was able to read fluently enough to actually enjoy it.

    I’m not sure if any viable modern strategy for teaching Latin arises from my experience, however …

    Pain is the key …

  23. Ah, that brings back memories of Brother Auger at St. Mary’s International in Tokyo, who used to chase unsatisfactory students around the classroom with a ruler, reducing to tears hulking lads who could have broken him in two with a single twitch of a bicep. Good times!

  24. Stephen C. Carlson says

    I use the Italian pronunciation for all authors from Augustine forward. And when singing, of course. Otherwise it’s Erasmus all the way.
    Same here, except I start that with Tertullian. For NT Greek, I’ve been moving to reconstructed Koine.

  25. marie-lucie says

    JC: I never studied Greek (my Latin teacher having told my mother that I lacked the required imagination), so I did not learn the non-French meaning of the various “accents”. Among some of the Native American languages I have had an interest in is Takelma, a language formerly spoken in Oregon which does have the relevant accents, at least as understood by the great linguist Edward Sapir, who no doubt had a knowledge of Greek. He studied this fiercely difficult language with the last speaker before the first world war. In his grammar of the language Sapir does give instructions for how to interpret the symbols, transcribing the examples into musical notation. In spite of being very familiar with that notation I have always found it extremely frustrating to try to read aloud the linguistic material with my best approximation of the relevant accents. As a linguist, I feel that I should be able not only to do so, but also to try to determine the conditions governing the use of this or that accent. So I am going to copy your comment and keep it in a prominent place where I can keep referring to it while working on the language:

    the circumflex represents high pitch on the first mora of a stressed bimoraic syllable, and the acute represents high pitch on the second or only mora of a stressed syllable. Grave represents the lack of high pitch, and in some MSS was also written on all unaccented syllables. Even if it was the case that the prominent syllables had low pitch rather than high, …., at least we understand the structure of the system.

    Yet something else for me to work on!

  26. But Rodger, are you serious that you let your Latin accent change with your material?

    I seem late to the discussion of my own remark, but yes, I do. I’d feel it as silly to read Virgil or Horace in Church Latin (what would happen to the meter?) as to read the Vulgate or sing Palestrina (not that I can actually do that competently) in reconstructed Latin (what would happen to the meter?). (Remember that Jerome, forced to advert to pronunciation, writes “Scibboleth,” showing that he already spoke a Latin that was becoming Italian. Or so my Vulgate reads–I seem to recall a previous discussion that disputed this.)

  27. Sister Mary of the Cross broke my little finger. My tears were dismissed as self-indulgence which was sinful and so it was never treated (I was six); the tip of the finger is still at a 45 degree angle 50 years later. Still, I came top in Latin so can’t fault her teaching methods.

  28. That’s the Clementine Vulgate of 1592, which from the Council of Trent until 1979 was the Church’s official Vulgate. The verse reads: “interrogabant eum: Dic ergo Scibboleth, quod interpretatur Spica. Qui respondebat: Sibboleth: eadem littera spicam exprimere non valens. Statimque apprehensum jugulabant in ipso Jordanis transitu. Et ceciderunt in illo tempore de Ephraim quadraginta duo millia.”

    But in echt Jerome it’s “Interrogabant eum: dic ergo “sebboleth”, quod interpretatur “spica”. Qui respondebat “tebboleth” eadem littera spicam exprimere non valens. Statimque adprehensum iugulabant in ipso Iordanis transitu. Et ceciderunt in illo tempore de Ephraim quadraginta duo milia” (punctuation added). So Jerome doesn’t try to represent the Hebrew sounds directly, but instead cleverly constructs an analogue based on a Latin phonemic distinction between fricative and stop rather than between shibilant and sibilant.

  29. verb paradigms go back to the 18th century BC in Sumerian-Akkadian tablets

    Could that be the starting point of modern structuralism (including Chomsky), which sees language as a grand theoretical structure rather than a means of communication?

    Recently I met someone who studied Basque for many years. He told me that the verb conjugations of Basque are immense, but only a small proportion of the forms are actually used. Not knowing Basque I’m unable to judge, but the mathematician’s delight in setting the whole system out in glorious, regular detail certainly sounds familiar.

    Incidentally, I’ve never studied Latin. I’m not sure it’s something I regret since a habit of forcing all languages into the mould of Latin is apparently one of the sins of Western linguistics. But one day in my dotage I might give it a try to ward Alzheimer’s off.

  30. Trond Engen says

    Even if it was the case that the prominent syllables had low pitch rather than high, as in Swiss German, at least we understand the structure of the system.

    Also the case in Central Scandinavian. Western and Northern Norwegian use a high pitch, Central and Eastern a low pitch for stressed syllables. Nobody cares except we think the others sound sing-songy. I’m not sure about the distribution in Swedish.

  31. Are there many other examples of ancient or medieval language learning tools that have survived?

    https://earlytibet.com/2014/05/12/silk-road-phrasebooks/

  32. David Marjanović says

    Is this some kind of weird classicist humblebrag, or what is this?

    Perhaps it means she can’t read it really fluently and instead has to use the school method of looking for the verb, then the subject…

    Of course this stuff was never intended to be read fast. The reader was supposed to revel in the complex art and savor it slowly to appreciate every detail of the author’s ingenuity.

    The only concession I make to non-classical pronunciation in singing Latin is [v] for /w/

    That’s non-classical only for a very narrow definition of classical, it seems. There’s an Umbrian city and its inhabitants whose names appeared in Latin as Mefanas and Meflanus, then as Mevania from Livy onwards. Presumably even earlier, Latin scaevus apparently gave rise to an Umbrian name which showed up in Etruscan as scefi(a).

    Source: p. 108 of: Jane Stuart-Smith: Phonetics and Philology – Sound Change in Italic. Oxford Linguistics. The pdf is lying around cyber-somewhere, there’s a link in a LH post from March or April of last year.

    Students with virtually unlimited access to both recorded and live native speakers of Chinese need months if not years to get the tones right

    That strikes me as an exaggeration.

    However, the closest analog is probably Central Franconian, not a full-blown tone language like Mandarin; and what I’ve read about that (e.g. the dialects of Cologne and Maastricht) is rather scary when it comes to the interaction of tone and intonation.

    Even if it was the case that the prominent syllables had low pitch rather than high, as in Swiss German

    It wasn’t! This guy said, according to the German Wikipedia, that their pitch was higher than that of unstressed ones by about a musical fifth.

  33. Mr. Carr eschewed all these barbarous methods of the Britanni and taught Latin exactly as he taught French, except with less conversation. (To be sure, all his discipuli were entirely volunteers.) After my time, he dropped dead with his second heart attack during a Latin class, thus defying the old saw that Latin teachers never die but just go on declining.

    And by classical I mean the reconstructed pronunciation I was taught.

  34. Students with virtually unlimited access to both recorded and live native speakers of Chinese need months if not years to get the tones right

    I think it is an exaggeration. You can pick Chinese tones up in a few days. The problem is wearing them in — like a pair of shoes, or indeed like any language that you’re learning. For most people, it takes time, effort, exposure, familiarity, and actual use to pronounce the sounds of a language naturally in speech.

  35. David Marjanović says

    Also the case in Central Scandinavian.

    And in Scotland and Ireland.

  36. David Eddyshaw says

    “Of course this stuff was never intended to be read fast. The reader was supposed to revel in the complex art and savor it slowly to appreciate every detail of the author’s ingenuity.”

    Profoundly true, at least for the vast majority of Latin authors on the traditional school syllabus; though some authors more than others. Tacitus’ prose must have been virtually incomprehensible to *native* Latin speakers who had not undergone an elite education.

    A concrete example is the ablative absolute, delight or bane of Latin syntax depending on your background. It basically doesn’t occur at all outside literary works, where it probably began largely as a device to get round the fact that Latin, unlike Greek, doesn’t have past-tense active participles.

    There’s also, from the modern barbarian schoolchild’s standpoint, the fact that the classical authors weren’t writing for schoolchildren at all (except maybe the ever-uplifting patriotic mythmaker Livy.) When I was at school I thought Tacitus was just an old misery, as well as a writer of entirely gratuitously convoluted Latin; fortunately I’d been taught well enough to want to return to him as an adult, when I (a) understood better what made him that way and (b) had learnt to appreciate his style instead of resenting it.

  37. Part of me wants to suggest rendering “sibboleth” in Latin as “xibboleth”, since the Greek letter ξ comes from samekh – except that the notion of [ks] as a mispronunciation of [s] surely makes no sense. According to the Septuagint: “Then they said to him, Say now Stachys; and he did not rightly pronounce it so.”

  38. Michael Hendry says

    David Eddyshaw’s 8:18pm comment reminds me of G. C. Lichtenberg on Tacitus:
    “A sure sign of a good book is that the older we grow the more we like it. A youth of 18 who wanted and above all could say what he felt would say of Tacitus something like the following: Tacitus is a difficult writer who knows how to depict character: and sometimes gives excellent descriptions, but he affects obscurity and often introduces into the narration of events remarks that are not very illuminating; you have to know a lot of Latin to understand him. At 25 perhaps, assuming he has in the interim done more than read, he will say: Tacitus is not the obscure writer I once took him for, but I have discovered that Latin is not the only thing you need to know to understand him — you have to bring a great deal with you yourself. And at 40, when he has come to know the world, he may perhaps say: Tacitus is one of the greatest writers who ever lived.”
    I blogged that 9+ years ago (here), but didn’t add much except the reference (second half of Sudelbücher E 197) and the German text.

  39. In Standard Danish (which I don’t think is usually considered Central Scandinavian) I would say stressed syllables without stød usually have lower pitch, around a musical third or so, while stressed syllables with stød have the same or slightly higher pitch, maybe a whole tone or so.

    Using Lower pitches for stressed syllables sounds like (the imitation of) Swedish or Bornholmsk, and higher pitches like (the imitation of) Jysk (Jutish).

  40. –Are there many other examples of ancient or medieval language learning tools that have survived?

    Not quite medieval, but very interesting:

    Tönnies Fenne’s Low German Manual of Spoken Russian, Pskov, 1607 AD

    I’ll quote his warning to learners of Russian: “Ahne Gotts gnade kan nemandtt de rusche sprake lehren.”

  41. TR: Scandinavian pitch

    A few years ago I took Swedish lessons from a Norwegian former colleague in preparation for attending a conference in Stockholm. Among all sorts of things I learned to say (and did my best to use in relevant contexts), the most useful sentence I learned was Tala Du engelska? ‘Do you speak English?’ which had a definitely lower pich on the stressed “en”. Similarly for svenska and a number of other stressed words. After speaking English for decades, I was surprised to encounter the ower pitch of the stressed vowel. The following year I switched to learning Norwegian before going to a conference in Oslo as you know. and found the same pitch difference in Norwegian. But after your comment I wonder if the Swedish I learned first was influenced by my teacher’s Norwegian native tongue? Perhaps I was not proficient enough in either language to notice the difference? (I have forgotten almost everything except how to ask the question).

  42. (Replying to two commenters above) My characterization of foreign students’ ability to master Chinese tones reflects my own first-hand observation, which is not to say that others who have similar experience may not characterize it differently. I’m not even sure that Bathrobe is disagreeing with me, because it takes months or years to achieve something that is not cringeworthy in actual use, and a lot of people never get there. If one’s argument for using a reconstructed pronunciation is an aesthetic one (e.g. the poetry doesn’t sound right otherwise), then the fact that many people cannot acquire a tolerable accent in a modern foreign language under far more favorable conditions is, I think, a significant objection.

  43. Tala Du engelska?

    It is TalaR Du engelska?, though. In the combinations r + t, d, n, l, s the r gets absorbed into the following consonant, resulting in a retroflex coloring, as in ord (=word).

  44. marie-lucie says

    THanks juha. I could not remember if it was tala or talar in this context, since the pronunciation is the same, so I chose the simpler version just in case.

  45. I’m almost jealous of these war-stories. Here I started on my own with a book literally-literally titled “Latin Made Simple.” And when I finally did get a chance to take Latin before college, my teacher was (again) a literal-literal former hippie with a passion for things like homemade yogurt. Clearly I missed out on a prime formative experience.

    As far as pronunciation regimes go, this has come up in a number of threads before, but I don’t think it’s possible to be a purist about this in a reasonable way: the circumstances of individual languages/orthographies just don’t lend themselves easily to treatment under a unified set of consistent principles.

    Take Greek: I also tend to use reconstructed Attic for pre-Christian texts, and a Modern Greek pronunciation for the NT onward, with some fitful (and ineffective) guilt that I should be observing the open/close distinction for epsilon/eta etc. of the “reconstructed Koine” for the early centuries AD. Reasonable enough, I think, and per above I’m clearly not the only one. But it’s not really very accurate, no? Maybe we can forgive ourselves for not throwing in digammas and what not with Homer because we’re stuck with an Athenian recension. But what about those dialects more contemporary with classical Attic that we have good epigraphic evidence for? How many of us take a stab at anything more than the eye-dialect the texts present us with? On top of which we have the problem of people not necessarily writing in the dialect they natively spoke. It’s a mess. And on the other end, even in later antiquity, do we suppose that all the scattered Greek Fathers spoke Koine the same way?

    This isn’t to say there aren’t hard facts to recommend the distinctions some of us make. I may read to myself in some franken-Greek of anachronistically combined features, and in an American accent to boot, but there’s still little doubt I’m closer on average to the way Homer was meant to be heard if I don’t spirantize like Modern Greek. But I think it’s important to remember that “more accurate” is still quite a bit less more accurate than we tend to think. And for all the focus on consonant and vowel quality, how careful are most Westerners about their quantities in Greek (or Latin)? Here in Japan the consonant inventory is woefully inadequate for any period of Greek, but any child can repeat and distinguish the vowels in ἄνθρωπος like a champ.

    And if our accuracy isn’t as accurate as we would like to think, not even for a comparatively well-understood language like Ancient Greek, case-by-case considerations of fluency and communicability might easily (or at least arguably) favor favoring existing reading practices over more scientific reconstructions. And so while it’s certainly inconsistent to worry about reconstructed Koine vowels on the one hand and recoil in shudders from the idea of romanizing Old Chinese 愛 ài as qˁəp-s or something à la Baxter-Sagart, on pragmatic grounds I think it can be justified. Even with Greek, after hearing a modern Greek so fluently read his Homer, it’s not always easy to prefer the plodding Attic of the classroom. There’s a lot to gain from discarding purism.

  46. Von Bladet of the Desert, NM says

    I’m almost jealous of these war-stories.

    It is by now clear that the Best Way To Learn Latin™ involves a brutal and ill-tempered monastic, and there was me thinking it unlikely I could regret not learning it any less.

    It is also clear that if pronunciation is part of the desired outcome it needs to be an intrinsic part of the pedagogy. (But chins up, one and all: I learned to speak tolerable Swedish tones back in the day without even being actually bodily harmed although perhaps unsportingly my teacher was a native speaker.)

    But I’m still in the dark about what exactly the deal is Horace.

  47. not leoboiko says

    @Ken: How would you compare foreign learner’s (in)ability to converse in fluent Mandarin with the proper tones, vs. the ability of scholarly students to pronounce wényán poetry with the tonal patterns in place?

    Because reading Homer out loud is a far cry from talking in fluent Ancient Greek. The pitches are right there, visible, staring you in the face; and what’s more, phonetic details aren’t as important as the bare-bones phonology. It seems to me to be no harder than reading IPA transcriptions, or musical notation for singers. And, as in Chinese poetry, it seems reasonable to argue that the melody is important to Greek poetry, and worth reproducing (assuming we have reasonably reliable reconstructions, defended on solid linguistic grounds).

  48. @not leboiko
    The rules for meter in classical Chinese poetry were based on a distinction between tones that has been effaced by sound changes since the Tang dynasty. If you read classical poetry in present-day Mandarin, you don’t get a good sense of the meter. It works better if you use certain of the dialects, though even where modern dialects preserve the distinctions, the actually pitch of the tones is no longer the same, so the subjective experience of the sound should be different. (Excuse me if you knew this already and I missed your point.) But not to quibble. If I understand your point correctly, you are saying that reciting poetry is different from speaking naturally, and if I can extrapolate from that, that the meter is your guide to pronunciation, and maybe that the meter overrides some of the more difficult things about pronouncing a tonal language, such as combining the tones of the individual words with the intonation of the entire sentence. This is certainly a fair point.

    I have not kept up with my reading on Greek prosody, and I have had Devine and Stephens’s “The Prosody of Greek Speech” on my bookshelf for a number of years waiting to be read. Maybe that will give me more respect for the state of the reconstruction of the Greek accents. My impression from earlier reading was that the contours of the pitch on individual words was thought to be fairly well understood, but that it was a mystery how the pitch in successive words might interact. Is that still the case?

    In Mandarin, two successive third tones produce a second and a third, but there are other cases as well — for example, if you have two successive fourth tones, the first one does not drop as low as the second one does, or if you have a second tone sandwiched between two first tones, there’s little perceptible rise in the middle syllable. In Taiwanese (a/k/a Hokkien), any syllable that precedes another syllable (as opposed to preceding a pause) will change in tone unless it is the last syllable of a noun. These tonal changes are not normally marked because they are automatic and predictable, but they apply to poetry as well as prose, and who knows whether something similar is lurking under those Greek tone marks. (Well, maybe everyone knows except me, because I haven’t kept up with my reading. But if anyone knows, it would be interesting to learn about it.)

  49. Pitch isn’t the same as tone. Pitch languages like Norwegian or Ancient Greek don’t usually show sandhi effects like Chinese, because the majority of syllables in pitch languages are unstressed and so don’t have contrastive pitch. Mandarin has a few neutral-tone syllables; most Sinitic languages have none. Shanghainese has pretty much become a pitch language.

  50. The best language to read Classical Chinese poetry in is Vietnamese. All 平 (bằng) tones are modal-to-breathy in phonation, while all 仄 (trắc) tones are creaky or tense in phonation. Best thing with Vietnamese is that unlike any modern Chinese language, the living tradition of folk verse in Vietnam follow the same 平-仄 scansion as classical regulated verse, and so the psychological apparatus to pick up the music of the poetry is already there.

  51. Neat, I didn’t know that.

  52. Jim (another one) says

    “Even if it was the case that the prominent syllables had low pitch rather than high, as in Swiss German, at least we understand the structure of the system.
    Also the case in Central Scandinavian. Western and Northern Norwegian use a high pitch, Central and Eastern a low pitch for stressed syllables. Nobody cares except we think the others sound sing-songy. I’m not sure about the distribution in Swedish.”

    Trond, I saw a paper a while ago about tones across Athapaskan that said all these languages have the same tones in all the same places and the apparent chaos is simply that there is a system of marked versus unmarked, or something like that, and in some languages the marked tones is realized as high, in some as low.

    “It is by now clear that the Best Way To Learn Latin™ involves a brutal and ill-tempered monastic, and there was me thinking it unlikely I could regret not learning it any less. ”

    The best way is to go into a monastery at six or seven and memorize the Psalter by reciting Psalms by simple imitation five times a day. It doesn’t take long to go through the whole thing and then the cycle resumes, without cease. Then one day, if you show promise, someone sits you down and starts writing out some lines for you, of a text you already know by heart, so it’s easy to associate letters with sounds. Then if you show real promise, you may be brought in to learn the structure and lexicon of what you memorized a few years before. To develop the skill of actually producing anything in Latin you start content classes such as theology.

    Reportedly this is also how little Aramaic-speaking Jewish kids learned how to read and write whatever Hebrew they could.

  53. Ah, that brings back memories of Brother Auger at St. Mary’s International in Tokyo, who used to chase unsatisfactory students around the classroom with a ruler, reducing to tears hulking lads who could have broken him in two with a single twitch of a bicep

    E.C. Baade would throttle the slow (in the nicest possible way.) Scudder was more the absent minded professor school of teaching. I’ve seen him rootle about for eyeglasses that were on top of his head. Classic!

    (And while we’re playing name drop and for what it’s worth, which is zilcho, this guy is a g-g-uncle.)

  54. David Eddyshaw says

    “It is by now clear that the Best Way To Learn Latin™ involves a brutal and ill-tempered monastic, and there was me thinking it unlikely I could regret not learning it any less.”

    Presbyterian, in my case. We had Presbyterians and Jews. Catholics went to some other school … I suppose. I didn’t know any.

    “But I’m still in the dark about what exactly the deal is Horace.”

    In some ways I think Horace is the touchstone of whether one has a full appreciation of what Latin can do. Whereas our ideas of literary greatness place a high value on content and on originality, Horace is mostly writing on pretty mundane and well-worn themes. What makes him great is the beauty of the form of his verse, along with his amazing virtuosity in making complex Greek meters work in Latin. Our culture underrates that sort of achievement.

    As far as simply understanding the sense of his words goes, I don’t think Horace is all that difficult. It’s having enough of a feel for the language that you can really appreciate his artistry which is the difficult bit.

    Video meliora proboque; deteriora sequor. As for me, I prefer Lucretius to Virgil and Catullus to Horace. A cultured Roman would immediately and rightly peg me as having a mere Silver Age sensibility.

  55. Von Bladet of the Desert, NM says

    Thanks, David!

  56. J.W. Brewer says

    Re Br. Auger, that’s why you should have made your parents send you to the mellow hippies at ASIJ instead. (Although I realize I have no idea if they taught Latin in the grades more advanced than what I’d reached by the time we moved back to the States. Us elementary school kids at ASIJ just got an hour a day of Nihongo plus I taught myself Cyrillic out of some reference book in the library because I was weird like that.)

  57. Re Br. Auger, that’s why you should have made your parents send you to the mellow hippies at ASIJ instead.

    Is that the one that used to be in Quonset huts that were torn down to create the Olympic Village in ’64? If so, that’s where I spent fifth and sixth grades (not sure about fourth), until my parents decided I wasn’t getting enough of an education and sent me to St. Mary’s. Which was great from an educational point of view and a disaster from a social-development point of view (no girls).

  58. J.W. Brewer says

    I guess by the time I got to Tokyo in ’73, the new post-Olympics campus (ok, actually wiki says “The current campus in Chofu was opened in 1963”) was sufficiently old (from a child’s perspective) we didn’t understand it as new or even new-ish but treated it as though it had been there from time immemorial.

  59. David Marjanović says

    I didn’t know that about Vietnamese, and I’ve read some detailed explanations of Middle Chinese tonogenesis!

    Trond, I saw a paper a while ago about tones across Athapaskan that said all these languages have the same tones in all the same places and the apparent chaos is simply that there is a system of marked versus unmarked, or something like that, and in some languages the marked tones is realized as high, in some as low.

    I thought the standard explanation in this case was that Proto-Athabaskan was full of glottal stops, which turned into high tone in most languages (as expected) but into low tone in others (which would make sense if we assume a preceding shift to [h])?

  60. Minus273: So the most faithfully Classical Chinese-like language is Vietnamese rather than any variety of (Modern) Chinese? I didn’t know that either, but I am not surprised: it is very similar to the reality that in many ways Finnish is far more faithful to Proto-Germanic than any Germanic language/dialect spoken today: have a look at page 25 of this dissertation I ran into recently:

    https://helda.helsinki.fi/bitstream/handle/10138/135714/bidragti.pdf?sequence=1

  61. Trond Engen says

    dainichi: Using Lower pitches for stressed syllables sounds like (the imitation of) Swedish or Bornholmsk, and higher pitches like (the imitation of) Jysk (Jutish).

    As I never get tired of claiming, Jysk is West Scandinavian.

    marie-lucie: Perhaps I was not proficient enough in either language to notice the difference?

    Not at all. Swedish is mostly low-tone. What I meant to say is, it’s not completely low-tone, and I have not been able to learn the geographic distribution of high-tone dialects.

    David M.: I thought the standard explanation in this case was that Proto-Athabaskan was full of glottal stops, which turned into high tone in most languages (as expected) but into low tone in others (which would make sense if we assume a preceding shift to [h])?

    Note how (most of) Danish uses stød instead of pitch to mark stressed syllables.

  62. David Marjanović says

    Sure, tonogenesis can go backwards. But having lots and lots of glottal stops is the ancestral condition in the Athabaskan case, retained by the group’s closest relative, Eyak.

  63. @Trond, I thought it was only Vestjysk that was West Scandinavian.

    Danish uses stress to mark stressed syllables, stød originally marked syncopated bisyllables in contrast to original monosyllables (I think, it could be the other way around), like one of the tones in Norwegian and Swedish, and there are still many minimal pairs even though the distribution has changed in the 1000 years since then. A fun example is the surname Møller with stød and the occupation name møller (‘miller’) without, probably the result of dialectal differences.

    But it’s true that stressed syllables with stød tend to have a higher tone than surrounding unstressed syllables, and lower without.

    Googling Vestjysk, I just learned of a phenomenon kalled klusilspring (‘stop jump’) where the glottal gesture of stød is strengthened to a whole extra syllable: PG *erþō >> VJ jokwer (!).

  64. David Marjanović says

    Wikipedia says, in a place I can’t find right now, that there are Danish dialects that have generally turned stød into /k/, and some of those turn it into [t͡ʃ] before front vowels! From word length to affricate in a thousand years or less – go sit in a corner and be ashamed, Armenian erku.

    At the same time, there are also dialects in southern Denmark that never developed stød and retain Scandinavian word-tones.

  65. David, go to dialekt.ku.dk/dialektkort/ and select overlay 9. If you want to listen there are 4 juicy examples between 1:30 and 1:45 here (uk < ud, titj < tid, swenneritj < svineri). These are all in Auslaut though, word-internally there is ‘normal’ stød even after high vowels — that might be specific to this dialect of course..

    Anyway it seems that these affricates are conditioned on the frontness of the stressed vowel; most of Vestjysk seems to have velar stops in that position, even if they add a j-glide after front vowels.

  66. Etienne: The Sinitic component in Vietnamese, in most aspects, is like any other Chinese dialect: some archaisms, some strange innovations (the Great Consonant Shift : t > ɗ, ts, s > t, tʂh > ʂ > s, tɕh > ɕ > s, et cetera), some strange innovations reflecting deep archaisms (the chóngniǔ distinction in Middle Chinese between pjen and pien is reflected in Sino-Vietnamese by pjen > *tsien > tiên), some spelling pronunciations, some character readings carried on from earlier, “native” Chinese borrowings.

    The suitability of reading Classical Chinese poetry in Vietnamese concerns one archaism in particular: due to its southerly location on the periphery of the Chinese Sprachbund, Vietnamese has not completely forgotten the phonation origins of the tonogenesis, so, I believe, still retains what has made the literati invent the regulated meter in the first place.

  67. David Marjanović says

    Thanks, Lars! The 3rd or 4th time I managed to read along and noticed. What’s up with the [w], do you think that’s a retention?

  68. Latin is alive and well in specialised technical fields. For example, in law, the traditional English pronunciation is still used in Latin phrases: eg. ratio decidendi (reɪʃio disaidendai) = reason for deciding or the principle of a legal decision; sub judice (sʌb dʒudisi:) = before the court. There is no way that these phrases would be spoken in the reconstructed classical pronunciation by anyone in the legal profession.

  69. Michael Hendry says

    zyxt:
    One of the chapters in A. P. Herbert’s Uncommon Law features a young lawyer who tries to use the restored pronunciation of Latin in court, pronouncing prima facie ‘pree-mah fah-kee-aye’ and so on, and the judge’s hostile reaction. (Date: 1930s, I think.) Very amusing, like the rest of the book.

  70. @David, look at map 11 — that does look like a classical retention pattern with /v/ spreading from East Danish, but it is usually said that most features of Jysk are innovations. I don’t have any sources with me so I can’t get closer than that.

  71. David Marjanović says

    Geographically, that does look very much like retention!

  72. There is no way that these phrases would be spoken in the reconstructed classical pronunciation by anyone in the legal profession.

    But per se and pro se are partial exceptions.

    pronouncing prima facie ‘pree-mah fah-kee-aye’ and so on

    Ah yes, “The Dead Pronunciation”. And here is Mark Liberman being sandbagged by “The Reasonable Man.” But I confess that my very favorite (and I have read both books as well as the uncollected pieces on the web) is “Port to Port”, which addresses the burning question: when a car on a flooded roadway in England, properly keeping to the left, meets a rowboat on the same flooded roadway, properly keeping to the right, which is to prevail, the law of the land or the law of the sea?

  73. I find it interesting to see how the methods for teaching Greek in ancienty are mirrored in my experience learning Mandarin Chinese in the 21th century. When I studied Mandarin as a foreign student in China, in the beginning our teacher would make us read a practise sentence from the book and exchange just one word (usually from a box with alternatives, though). I find the idea similar to the chreias described in the article. Another similarity is the transcription: I know of no one who would learn Mandarin Chinese in the original characters without using some transcription (pinyin, bopomofo).

  74. David Marjanović says

    I finally clicked on the link. The ancient Greek transcription of Latin is fascinating – too bad it’s not mentioned how old exactly it is…

  75. John Cowan says: There is no way that these phrases would be spoken in the reconstructed classical pronunciation by anyone in the legal profession. But per se and pro se are partial exceptions.

    I’ve only heard them pronounced as [pɜː seɪ] and [pɹoʊ seɪ]. Wouldn’t a reconstructed classical pronunciation be something like [per se] and [pro se]?

  76. I imagine that’s why he said partial exceptions.

  77. A fully English pronunciation of se would be /si/.

  78. I notice a tendency to write “per say.”

  79. @Trond Engen:

    >Note how (most of) Danish uses stød instead of pitch to mark stressed syllables.

    If we regard stød as a phonemic concept, I would guess that pitch is one of the most important phonetic cues for it (at least in my dialect, which is pretty close to Standard). Stressed syllables without stød need to be marked as well; here pitch would be an even more important cue (on top of syllable length, stress-time rhythm etc., of course).

Trackbacks

  1. […] Hat talks about the best ways to learn […]

Speak Your Mind

*