4 ms·

When I had to write Polytonic Greek, I had to write my own vim plugin that would transliterate on demand. For instance, while in insert mode h''als<alt-g> turns

When I had to write Polytonic Greek, I had to write my own vim plugin that would transliterate on demand. For instance, while in insert mode h''als<alt-g> turns into ἃλς. It's okay to laugh at the ancient Greek and its words that look similar to us, but we should not forget that the words with different accents sounded differently, so it was probably not confusing to native speakers. Moreover, every language has its homophonies (the English "Dick" is not just a first name), and many near-homophonies for strangers (Japanese untrained ears can't distinguish "r" from "l", and I can't identify "o" from "a" in most Russian words). A nitpick about the article: a paragraph about the other main diacritics would have given more context. The breathing marks chang the pronunciation as much as the accents did. And since Roman languages are not tonal languages, they kept the breathing as a "h" but dropped the tones.

oytis 5y ago
> and I can't identify "o" from "a" in most Russian words Unaccented "o" and "a" are exact homophones, Russian children are trained in school to memorize the spelling in order to differentiate between the two.
- bloak 5y ago
  They may well be exact homophones in some/all dialects of Russian, but I'm not sure that children having to memorise spellings is evidence for that. Don't both English and German, for example, have almost-homophones that children typically learn to distinguish by learning the spelling rather than by hearing the subtle difference in pronunciation? If you look at the spelling mistakes made by 12-16-year-olds in England and Germany, don't you find a significant proportion of them are caused by the child having incorrectly heard how people pronounce the word? I suppose it's an interesting phenomenon that some (but not all?) languages have subtle phonetic differences that even native speakers cannot always easily distinguish.
  - yorwba 5y ago
    > If you look at the spelling mistakes made by 12-16-year-olds in England and Germany, don't you find a significant proportion of them are caused by the child having incorrectly heard how people pronounce the word? I don't know about England, but I think in Germany it's more common that children represent the pronunciation accurately, but are "corrected" by adults who want to enforce the standard orthography. E.g. take the example on page 2 of http://www.edu.lmu.de/lbp/personen/wiss_ma/eckerlein/seminare/sse_stufen.pdf http://www.edu.lmu.de/lbp/personen/wiss_ma/eckerlein/seminar... : LOITETOTSUFARENiSTLBÖD. In Standard German, that would be "Leute totzufahren ist blöd." Except for switching L and B, the original isn't wrong about the pronunciation: <eu> represents [ɔʏ] and <z> represents [ts], so using <oi> and <ts> instead is perfectly reasonable.
    - tremon 5y ago
      It doesn't seem perfectly reasonable to me. Using "totsufaren" instead of the correct spelling because the 't' sound is already in 'tot' destroys consistency with hinzufahren, auszufahren. And I'm not convinced that 'oi' and 'eu' represent the same sound in German, since 'eu' is a diphthong and always sounds long, whereas 'oi' is a short vowel followed by a glide: [oᶦ].
    - bloak 5y ago
      > I don't know about England, but I think in Germany it's more common that children represent the pronunciation accurately, but are "corrected" by adults who want to enforce the standard orthography. Represent the pronunciation accurately, or the children's pronunciation accurately? It's a bit more interesting in English because a common mistake in English is to pronounce a word as it's spelt. For example, the T in often is supposed to be silent ... OK, that's me being prescriptive, but well-educated older speakers hear younger people pronounce the T in often and they disapprove. I suppose parents and teachers have different incentives here: parents want their children to resemble them, make a good impression, and sound educated; teachers want the children in their care to get good marks in written tests and exams. Knowing how to spell diarrhoea is less important nowadays in real life, because of computers. If you don't want to look stupid you still have to know it's guerrilla marketing rather than gorilla marketing, though. Computer will tell you how many Rs in guerrilla.
  - oytis 5y ago
    Quite the opposite, there are some dialects (e.g. Vologda dialect) where they are pronounced differently, there is a term for that, it's called "о́канье". But the norm is that a native speaker can't differentiate between the two unless he/she has other clues apart from the sound of it. I can imagine you can detect slight unconscious difference that an educated native speaker introduces by means of computer analysis of sound samples etc., but it's not a phonetically significant difference for a human.
- Grustaf 5y ago
  Well, you can very often deduce which one it is from cognate words, or general language patterns, so it's not just rote learning. e.g. this random headline from Pravda today: "Байден намерен обсудить поведение России на встрече с Путиным" NAMEREN - obviously a, since NA is a common prefix OBSUDIT' - obviously o, since OB is a common prefix POVEDENIE - obviously o, since PO is a common prefix ROSSII - hopefully you can spell the name of your country... or if you don't know how to spell OPRAVDAT' (justify) which has its stress on the ultimate syllable, you just need to recall the name of the glorious newspaper above, PRAVDA, where the stress is on the first A. or GOROD is easy to spell because -oro- is a very common cluster.
AntigoneJournal 5y ago
Many thanks for this. Yes, a fuller article would make mention of 'crasis' - where two successive words are melded into one utterance - which inevitably affected pitch and accentuation. As to breathings, their affect on pitch (as opposed to pronunciation) was null: the rough was a mere aspirate, and the 'smooth' was in effect the absence of any aspirate. Latin, as a fellow Indo-European language, had inherited initial aspirates, although these were fewer in number than in Greek, where rho and upsilon always had initial aspiration, and where Indo-European s- had become an aspirate (compare Lat. serpo with Gk herpo, or Lat. sal with your Greek hals.)
Grustaf 5y ago
> It's okay to laugh at the ancient Greek and its words that look similar to us I never laughed at this, it's not strange that there are words or word forms where the only difference is a single phonetic feature, or indeed where there is no difference at all. As you say that is the case in all languages. And regarding accents in Greek, I don't recall encountering any case where the position of the accent was the crux. Which isn't really surprising given that the texts didn't have accents originally, so the writers couldn't rely on that alone.
- Mediterraneo10 5y ago
  In Classical Greek, it isn't just the position of the accent that matters but the kind of accent there. You are right that the Greeks themselves did not need the accents notated, but students of Greek today often rely on the accent to tell e.g. the eta or eta+nu words apart.
  - Grustaf 5y ago
    Sure, I took 6 full-time terms of Classical Greek so I'm familiar with the different accents. But my Greek studies took place 20 years ago so I do wonder what you mean by eta and eta + nu words? Are these different subtypes of the third declension?
    - Mediterraneo10 5y ago
      Eta alone can mean ‘the’, ‘she who’, ‘either/or’, ‘indeed’, ‘than’, or ‘I was’, with the accent helping to narrow it down. Eta+nu is ‘her who’, ‘he/she/it was’, etc. Also, for example, tau+iota is either an interrogative or an indefinite pronoun, and it is the accent which tells you which without having to scan further through the sentence.
      - Grustaf 5y ago
        Ok, I thought you were talking about word endings. Well sure, there are minimal pairs, but most of the time it’s immediately clear from the context, and we have to keep in mind that the accents were added much later by modern philologists, so they have de facto been deduced from the context. But sure, there goldilocks zone where it’s not obvious to the lay reader, but the scholars are pretty certain what it should say, and then it helps. I wouldn’t say it’s a major struggle when reading Greek though.
eloisius 5y ago
Even accounting for tonally differentiated words, Mandarin Chinese is rife with exact homophones. Shì can be “is,” “try,” “thing,” “kind/style” and many more. It amazes me how we are able to pattern match so well provided a little context.
- dwheeler 5y ago
  Indeed. For an extreme example, "Lion-Eating Poet in the Stone Den" is a short poem in Classical Chinese in which every word is pronounced shi when read in present-day Standard Mandarin, with only the tones differing: https://en.m.wikipedia.org/wiki/Lion-Eating_Poet_in_the_Stone_Den https://en.m.wikipedia.org/wiki/Lion-Eating_Poet_in_the_Ston...

oytis 5y ago

> and I can't identify "o" from "a" in most Russian words Unaccented "o" and "a" are exact homophones, Russian children are trained in school to memorize the spelling in order to differentiate between the two.

bloak 5y ago
They may well be exact homophones in some/all dialects of Russian, but I'm not sure that children having to memorise spellings is evidence for that. Don't both English and German, for example, have almost-homophones that children typically learn to distinguish by learning the spelling rather than by hearing the subtle difference in pronunciation? If you look at the spelling mistakes made by 12-16-year-olds in England and Germany, don't you find a significant proportion of them are caused by the child having incorrectly heard how people pronounce the word? I suppose it's an interesting phenomenon that some (but not all?) languages have subtle phonetic differences that even native speakers cannot always easily distinguish.
- yorwba 5y ago
  > If you look at the spelling mistakes made by 12-16-year-olds in England and Germany, don't you find a significant proportion of them are caused by the child having incorrectly heard how people pronounce the word? I don't know about England, but I think in Germany it's more common that children represent the pronunciation accurately, but are "corrected" by adults who want to enforce the standard orthography. E.g. take the example on page 2 of http://www.edu.lmu.de/lbp/personen/wiss_ma/eckerlein/seminare/sse_stufen.pdf http://www.edu.lmu.de/lbp/personen/wiss_ma/eckerlein/seminar... : LOITETOTSUFARENiSTLBÖD. In Standard German, that would be "Leute totzufahren ist blöd." Except for switching L and B, the original isn't wrong about the pronunciation: <eu> represents [ɔʏ] and <z> represents [ts], so using <oi> and <ts> instead is perfectly reasonable.
  - tremon 5y ago
    It doesn't seem perfectly reasonable to me. Using "totsufaren" instead of the correct spelling because the 't' sound is already in 'tot' destroys consistency with hinzufahren, auszufahren. And I'm not convinced that 'oi' and 'eu' represent the same sound in German, since 'eu' is a diphthong and always sounds long, whereas 'oi' is a short vowel followed by a glide: [oᶦ].
  - bloak 5y ago
    > I don't know about England, but I think in Germany it's more common that children represent the pronunciation accurately, but are "corrected" by adults who want to enforce the standard orthography. Represent the pronunciation accurately, or the children's pronunciation accurately? It's a bit more interesting in English because a common mistake in English is to pronounce a word as it's spelt. For example, the T in often is supposed to be silent ... OK, that's me being prescriptive, but well-educated older speakers hear younger people pronounce the T in often and they disapprove. I suppose parents and teachers have different incentives here: parents want their children to resemble them, make a good impression, and sound educated; teachers want the children in their care to get good marks in written tests and exams. Knowing how to spell diarrhoea is less important nowadays in real life, because of computers. If you don't want to look stupid you still have to know it's guerrilla marketing rather than gorilla marketing, though. Computer will tell you how many Rs in guerrilla.
- oytis 5y ago
  Quite the opposite, there are some dialects (e.g. Vologda dialect) where they are pronounced differently, there is a term for that, it's called "о́канье". But the norm is that a native speaker can't differentiate between the two unless he/she has other clues apart from the sound of it. I can imagine you can detect slight unconscious difference that an educated native speaker introduces by means of computer analysis of sound samples etc., but it's not a phonetically significant difference for a human.
Grustaf 5y ago
Well, you can very often deduce which one it is from cognate words, or general language patterns, so it's not just rote learning. e.g. this random headline from Pravda today: "Байден намерен обсудить поведение России на встрече с Путиным" NAMEREN - obviously a, since NA is a common prefix OBSUDIT' - obviously o, since OB is a common prefix POVEDENIE - obviously o, since PO is a common prefix ROSSII - hopefully you can spell the name of your country... or if you don't know how to spell OPRAVDAT' (justify) which has its stress on the ultimate syllable, you just need to recall the name of the glorious newspaper above, PRAVDA, where the stress is on the first A. or GOROD is easy to spell because -oro- is a very common cluster.

AntigoneJournal 5y ago

Many thanks for this. Yes, a fuller article would make mention of 'crasis' - where two successive words are melded into one utterance - which inevitably affected pitch and accentuation. As to breathings, their affect on pitch (as opposed to pronunciation) was null: the rough was a mere aspirate, and the 'smooth' was in effect the absence of any aspirate. Latin, as a fellow Indo-European language, had inherited initial aspirates, although these were fewer in number than in Greek, where rho and upsilon always had initial aspiration, and where Indo-European s- had become an aspirate (compare Lat. serpo with Gk herpo, or Lat. sal with your Greek hals.)

Grustaf 5y ago

> It's okay to laugh at the ancient Greek and its words that look similar to us I never laughed at this, it's not strange that there are words or word forms where the only difference is a single phonetic feature, or indeed where there is no difference at all. As you say that is the case in all languages. And regarding accents in Greek, I don't recall encountering any case where the position of the accent was the crux. Which isn't really surprising given that the texts didn't have accents originally, so the writers couldn't rely on that alone.

Mediterraneo10 5y ago
In Classical Greek, it isn't just the position of the accent that matters but the kind of accent there. You are right that the Greeks themselves did not need the accents notated, but students of Greek today often rely on the accent to tell e.g. the eta or eta+nu words apart.
- Grustaf 5y ago
  Sure, I took 6 full-time terms of Classical Greek so I'm familiar with the different accents. But my Greek studies took place 20 years ago so I do wonder what you mean by eta and eta + nu words? Are these different subtypes of the third declension?
  - Mediterraneo10 5y ago
    Eta alone can mean ‘the’, ‘she who’, ‘either/or’, ‘indeed’, ‘than’, or ‘I was’, with the accent helping to narrow it down. Eta+nu is ‘her who’, ‘he/she/it was’, etc. Also, for example, tau+iota is either an interrogative or an indefinite pronoun, and it is the accent which tells you which without having to scan further through the sentence.
    - Grustaf 5y ago
      Ok, I thought you were talking about word endings. Well sure, there are minimal pairs, but most of the time it’s immediately clear from the context, and we have to keep in mind that the accents were added much later by modern philologists, so they have de facto been deduced from the context. But sure, there goldilocks zone where it’s not obvious to the lay reader, but the scholars are pretty certain what it should say, and then it helps. I wouldn’t say it’s a major struggle when reading Greek though.

eloisius 5y ago

Even accounting for tonally differentiated words, Mandarin Chinese is rife with exact homophones. Shì can be “is,” “try,” “thing,” “kind/style” and many more. It amazes me how we are able to pattern match so well provided a little context.

dwheeler 5y ago
Indeed. For an extreme example, "Lion-Eating Poet in the Stone Den" is a short poem in Classical Chinese in which every word is pronounced shi when read in present-day Standard Mandarin, with only the tones differing: https://en.m.wikipedia.org/wiki/Lion-Eating_Poet_in_the_Stone_Den https://en.m.wikipedia.org/wiki/Lion-Eating_Poet_in_the_Ston...