What did ancient languages sound like? | Hacker News Reader

What did ancient languages sound like?

antigonejournal.com

197 pointsmewithoutyou149 comments

My favorite linguistic/sound story is this:

In 1878, a student called Ferdinand de Saussure proposed that PIE had only one vowel sound "e", and that all other vowel sounds in PIE languages came that single sound. How? He proposed that all other vowels came through a consonant "colouring" the original vowel "e". For example, the usage of e + w in words gave birth to the sound "u".

But this doesn't explain root words like "pas" that don't contain the core vowel "e" or any evidence of a rogue sound like "w" colouring it.

Here I will quote from the book: "Saussure conjured up a daring hypothesis. Once upon a time, he argued, verbs like pas did indeed have a rogue sound after the core vowel e, and this sound was responsible for 'colouring' the vowel from e to a. But at a later stage, and after it had wreaked all this havoc, the rogue sound disappeared from the scene. If one were to use the symbol X to represent this elusive rogue sound, then Saussure's claim was that pas originated as a regular root peXs, but that some time in prehistory X coloured the vowel e into an a, so peXs became paXs. Much later on (but still well before the earliest records), the rogue X was itself worn away because of [some other changes]. Schematically, then, the whole development can be represented as peXs -> paXs -> pas."

This remained a cool theory, outside mainstream linguistics. Then after Saussure died, archaeologists dug up some tablets and the oldest PIE language Hittite was deciphered. And there appearing in the very place where he had predicted it to be, the rogue consonant was found written down in clay.

Source: around page 109 from "The Unfolding of Language" by Guy Deutscher.

> In 1878, a student called Ferdinand de Saussure proposed that PIE had only one vowel sound "e", and that all other vowel sounds in PIE languages came that single sound.

That's not in perfect agreement with wiki:

> The beginnings of the [laryngeal] theory were proposed by Ferdinand de Saussure in 1879, in an article chiefly demonstrating that a and o were separate phonemes in PIE.

And: > In the course of his analysis, Saussure proposed that what had then been reconstructed as long vowels ā and ō, alternating with ǝ, was an ordinary type of PIE ablaut. That is, it was an alternation between e grade and zero grade like in "regular" ablaut (further explanations below), but followed by a previously unidentified element. This element accounted for both the changed vowel colour and the lengthening (short e becoming long ā or ō).

The book says it's a simplified version -- the paper he put out contains hundred+ pages. I'll check out wiki & sources because the book is quite specific about the single vowel thing.

That's not quite right. I might start to ramble but my last session was thankfully interrupted and I am not repeating it now.

The catch is that so-called laryngeals where they reflect in Hittite with Cuniform Characters like Ha and similar only in front of a word and only for two of three or more, meaning there is still no direct evidence of * paXs, or peh2s-, ph2es- depending on voewl length, modulo accentuation. Ah yeah, there are mostly three now.

The last point, about Ancient Greek read in the "modern" pronunciation (with Φ, Χ, Θ read as fricatives rather than aspirated stops) reminds me of how Mandarin, which is how Chinese poetry is often read nowadays, is not great for representing the sounds of the poetry as written, since it's lost all final stops (-k, -t, -p) in closed syllables.

Other topolects such as Cantonese preserved them, which makes it a better correspondence if you care about hearing something closer to the Middle Chinese that Tang poetry was composed in.

How do we know so much about Middle Chinese pronunciation in the present day? Check out https://www.youtube.com/watch?v=ME8wzyR6pO8 for a layperson's introduction.

Depends on when the poetry was written. Classical Chinese was in use for over two thousand years, with the written Mandarin-based vernacular Chinese only replacing it in the last century. So later poetry (e.g. during the Qing) may not have been written with Middle Chinese pronunciation in mind. For example, the somewhat (in)famous poem, 'Lion-Eating Poet in the Stone Den' consisting entirely of homophones in modern standard Mandarin was written in the 1930s, so the poem was certainly written with Mandarin pronunciation in mind.

It's hard to overstate the intellectual puzzle this is!

Classical Chinese scholarship had a developed body of proto-linguistics, including rime spelling, a system for phonetic transcription and analysis, which transcribed sounds using Chinese characters purely phonetically. It was used to indicate the pronunciation of characters, as well as foreign words in bilingual dictionaries and so on: https://en.wikipedia.org/wiki/Fanqie

The problem is, it's mostly self-referential. As the Wikipedia example gives, we don't precisely know how 東 was pronounced in 1000 AD from a fanqie dictionary entry. It would tell us something like that 東 (dong1 in Mandarin or dung1 in Cantonese today) was pronounced in the author's time like the initial consonant in 德 (de2, dak) with a closing syllable pronounced like 紅 (hong2, hung4). And of course, a dictionary from a thousand years before that, or even just from a speaker on the other side of China, might give a different entry...

If you collect all the known fanqie pronunciations given in all the eras from all the probable sister languages, and especially when you start comparing with our knowledge of other languages (Japanese, Sanskrit, etc.) you can start to piece together the puzzle and get a rather good understanding of Chinese historical sound change.

Chinese scholars back then were aware, on some level, of sound change, especially later in the Ming and Qing dynasties, but there was no general theory to address it, and the apparent contradictions in the ancient poetry and the old rime dictionaries were a source of a lot of perplexity.

It's very interesting that at the same time written Chinese is terrible for pronunciation, even an absolutely beginner (that's not too young a child) must surely be more aware of etymology the concept. Perhaps I am biased, but to me it is a writing system that schemes "history!" unceasingly.

It would make sense then if there was much earlier linguistics / folk etymologist, etc. in Chinese history. Whereas I didn't think think anything like modern linguistics got going until the enlightenment creation of the Classics' modern form.

Just because they were writing in Literary Chinese doesn't mean pronunciation was retained, or retained very well. Consider Mideval Latin and Church Latin "italianate" pronunciation.

According to https://en.wikipedia.org/wiki/Lion-Eating_Poet_in_the_Stone_..., the poem was written by a linguist, so I don't think it it is representative of how most later writers of Literary Chinese thought pronunciation ought to work, but rather a reductio ad absurdum argument again writing in dead languages without using an alphabet.

Why say 'topolect', and not just 'language'?

I realize topolect is a translation of an official chinese term, but as far as I'm aware we're not Chinese officials. We're speaking English, and we can be brave enough to call mutually unintelligible languages what they are - languages.

And before anyone gives me the 'language is a dialect with an army and a navy' - yes it's pithy but there's so many counter examples.

>We're speaking English, and we can be brave enough to call mutually unintelligible languages what they are - languages.

I'm not super familiar with the specifics of Mandarin, so I could be wrong here, but it's quite common to come across English words that have fallen out of common use by native speakers but are adopted by non-native speakers as a vehicle for expressing a concept that exists in their native language but doesn't have a precise translation into English.

Especially since they're talking about the evolution of Chinese pronunciation, "topolect", while jargon-y, seems to be a far more precise vehicle for expressing their sentiment.

Topolect was specifically invented in 1991 by Victor Mair as a translation of 方言 (fangyan) to get around the whole language/dialect bombshell when it comes to Chinese.

https://en.wiktionary.org/wiki/topolect

http://sino-platonic.org/complete/spp029_chinese_dialect.pdf

It's a way to take 方言 and decouple it from linguistic terms like language and dialects so a proper study and classification of the Chinese language family and its various languages and dialects can take place without interference from Chinese politics. When someone points out that Cantonese, Hokkien, and Mandarin are 方言 then linguists can nod politely, agree that they are in fact "topolects", then continue classifying them as Sinitic languages within the Sino-Tibetan language family each with their own collection of dialects.

Would be interesting to know if the study of Arabic has similar issues because from a layman's cursory knowledge Arabic has similar diversity when it comes to mutual intelligibility.

Thanks for the links and explanations. It seems political definitions of what is an isn't a "language" are more political than I realised.

On a side note, I actually found it a little bit ironic that Afrikaans was listed as a dialect of Dutch in his paper.

Many South Africans would disagree with that classification quite strongly, especially since Afrikaans has its own literature and regional dialects.

In turn, there's a movement to classify the coloured dialect of Afrikaans as its own language (Kaaps), in fear that it will be swallowed up by the more standard "white" form and erode their cultural identity.

Bravery doesn't enter into it -- I used the word because as linguists, we want to refer pragmatically to "whatever Chinese varieties actually are", rather than entering into endless debates about what is a language and what is a dialect.

By the way, I'm a linguist (and a native Cantonese speaker) who is acutely aware of the dialect/language issues surrounding our topolect, so while I sympathise and thank you for recognising that Cantonese would in other times and places be considered a language, using "topolect" allows us to focus on the topic at hand.

'Topolect' (an English word coined in America by an English-speaking linguist who was describing Chinese languages) has the advantage of being precise and preserving the relationship between Mandarin and Cantonese that the writer was specifically referring to, but that aside, might I suggest that courage is best seen and not heard.

'Topolect' is just a translation of 方言, a word that was previously translated as 'dialect'.

It's a weak, flimsy word for people who know better, but still don't have the courage to say 'language'.

I'm sorry, I did try to graciously point out that you had taken your patently noble defiance of the Chinese Communist Party and turned it into something that reflected poorly on your humility and empathy for others. Might I suggest writing the CCP a letter about the matter rather than speaking inappropriately about 'courage' and lecturing others?

Beg pardon. It's a made-up Greek word, comprised of "topos" and "lexis": "place" and "word" (or "saying") respectively.

It's actually very similar to "topology" as a construction - "lexis" comes from "logos".

Plays of Shakespeare is similar much of the rhymes and inside jokes are lost now due to modern pronunciation of old words.

http://originalpronunciation.com/GBR/Home

Similarly, in the Prologue to the Canterbury Tales, a young squire is described as conducting himself "in hope to stonden in his lady grace" (in the hopes of standing in his lady's grace).

Some take this to be a bawdy anatomical pun, expressing the young man's hopes of 'standing' in his lady's 'grass' — a distinction lost nowadays that the pronunciation of "grace" and "grass" have grown so far apart.

(Line 88 of Chaucer's General Prologue, https://chaucer.fas.harvard.edu/pages/general-prologue-0)

One of the more neglected and somewhat mind bendy aspects of Latin pronunciation is how often -m was not pronounced. Afaik the typical -um ending was a nasalized vowel with no /m/.

There was an article that made HN a while ago about censor being spelled cesor in an inscription. Probably because /n/, like /m/, was serving to nasalize the vowel before and was not itself pronounced.

It also starts to hint at how the modern romance languages dropped all those case suffices. Eg. if bonus and bonum were both pronounced /bonu/.

Portuguese, especially Brazilian Portuguese, retained this nasalization.

Portuguese has a lot of interesting stuff going on with nasal consonants. It always strikes me that they go for writing <m> in places other romance languages chose <n>. This also hints that over the history of latin languages, nasal consonants may have been kinda flexible.

I used to look at those topics and think in terms of what is "more like Latin", but at one point I realized... All of them are probably accurately capturing some form of the history of Latin, and there isn't right or wrong.

For words ending in "L" as well ... "som" (sound), "Brasil" (Brazil), the last consonants are dropped. Words ending in R like "lar" (home) often get a gutteralized or dropped R.

Portuguese did not retain the Latin nasal vowels, it was an independent development after the Latin nasal vowels had been lost.

Learning about this made Romanian, with all its words ending in -u, not seem strange suddenly!

Interestingly, spoken Catalan and spoken Portuguese also use -u as ending, while -o is written.

And further, latin nominative -us or -um has an -o- in some inflected forms. Example: bonorum, "of the good ones"

The article that mentioned cesor for censor also pointed out that in places where classical latin has -us or -um, early latin inscriptions might have -os, -om, or -o.

Sometimes i feel like when i hear speakers of modern romance languages talk really fast, unstressed o and u seem to get closer to one another. Maybe you could call it a continuum of "roughly that vowel".

>> And further, latin nominative -us or -um has an -o- in some inflected forms. Example: bonorum, "of the good ones"

And also, famously, "Babaorum".

Edit: oh, which btw does what you say, it's pronounced more with an "o" sound in French. Like all the "-um"'s. Otherwise "Petibonum" wouldn't work.

Of similar note, NativLang on Youtube has a fascinating series of narrated videos on how we know what various ancient languages sound like [1] (you'll have to forgive the cheesy graphics). I particularly liked the one on ancient Egyptian[2].

[1] https://www.youtube.com/playlist?list=PLc4s09N3L2h3d_c5Z5d0J...

[2] https://www.youtube.com/watch?v=J-K5OjAkiEA&list=PLc4s09N3L2...

The Egyptian example sounds like a Mitch Hedberg routine.

https://www.youtube.com/watch?v=_6_BpvhdIuU

I would be remiss if I didn't post Tocharian King of the Hill - https://www.youtube.com/watch?v=fNe6nke1HFk

This is amazing!

If anyone found this interesting, I highly recommend the audiobook/lecture "Language Families of the World" by John McWhorter. The title makes it sound dry, but he's a very entertaining lecturer. It's one of those works that - if you have never been exposed to the subject - is a window to an previously unnoticed detail in everyday life. Similar to when one learns about an architectural flourish and from that point forward sees it on buildings everywhere.

lovely article! another one of those letters that weirdly fascinates me is Beta, in modern Greek it's pronounced as a V whereas when we read Ancient Greek we usually pronounce it as B.

There was an excerpt somewhere where one Ancient Greek author (unfortunately I have forgotten who exactly) described the sound sheep were making, and he did so writing the following:

Sheep make the following sound: βαί, βαί

which could either be Bae, Bae, or Vae, Vae... (whereas nowadays we would probably use an 'm' to describe the sound, as in Mae mae) - so while the sound is described, it doesn't really provide a definite answer since both seem somewhat possible. But maybe the author of this article could help us out here!

Serious question: is it possible that the sounds made by sheep would have changed over time?

Different sheep make different sounds today. Mostly it’s just yelling. I think our perception of it insists on a leading consonant that just isn’t there.

The sheep/goats sometimes start yelling before opening their mouths, which seems to usually sound like maaaaaa, but can sometimes sound more like baaaaa, gaaaaa, waaaa, yaaaa. When they open their mouths before starting the yell there is no leading consonant sound.

Don’t think I have ever heard an initial 'v' sound though.

I mean, honestly, when we're talking about sheep vocalizations, the difference between ma/ba/va/wa is something more of a matter of perception than sensation...

You can't really generalize Ancient Greek, since in Attic, beta kept the /b/ sound the longest, whereas in many non-Attic languages it had already been /v/ for a long time.

In current Spanish the standard onomatopaeia for a sheep's sound is "beee". The b is pronounced like the English b, and the e is like the e in "test".

In English it is "baa", with the "aa" pronounced as the vowel in "trap". This is interesting in a couple of ways; the most immediately obvious is that it does not conform to ordinary English spelling.

But if you look a little closer, it turns out that it can't conform to ordinary English spelling. There is no conventional way to spell a syllable-final /æ/, because the phonology of English doesn't permit such a syllable to exist at all. /æ/ is a checked vowel in English -- except for this one onomatopoetic word.

(In the interest of completeness, it's probably worth noting that while the spelling is often simply "baa", it can be lengthened, and the pronunciation is nearly always several syllables long, ba-a-a-a, with the syllables separated by glottal stops.)

I'd love to see a source for this, never heard this before. In modern Greek it'd be "μπε μπε". With the μ + π giving the b sound.

However your example is interesting because sheep make an "eee" sound like in "men" and not an "i" sound like in "flight", which I think is kind of how the erasmian pronunciation says it should be.

In modern Greek "αί" is one vowel sound "e" like in "men". I really don't like the erasmian pronunciation of ancient Greek - it seems strange that vowels should change so drastically across time when for example Spanish and modern Greek have exactly the same vowels.

The fact is, languages change like this all the time. And sometimes not all the speakers go along with changes, for various reasons, and then the development splits. However, not always does this result in a new language (science uses cross-understandability as a criterion, outside of science it's... more complicated)

Just look at English: the current ortography is so weird because it got frozen before the language underwent a drastic change (the Great Vowel shift).

German changed radically too: during the High German consonant shift, many of the differences that set it apart from English and Dutch happened: t got changed to s, p got changed to pf and so on.

Vowel shifts happened too: the Early New German diphtongization changed many long vowel sounds. For example:

* "latin" to "latein" (the "i" in Latin is like "ee" in English "see", the "ei" like English "eye"),

* "niuw" (almost like English "new") to "neu" ("eu" like "oy" in English "boy"), or

* "mus" to "maus" ("u" as "oo" in English "good", "maus" exactl% like the English word "mouse").

The last example is especially interesting because the Great Vowel shift and the Early New High German diphtongization started almost a continent apart and 800 years after the languages split, but had the same result.

I think you are remembering a fragment by Cratinus, as quoted here: https://resgerendae.wordpress.com/2012/02/05/weird-and-wonde...

My recollection is that Erasmus was the one who argued, based on that verse, that the letter eta should be pronounced as a long "e" rather than "i" as in "police". I can't seem to find a reference, however.

You mean “What did ancient European languages sound like” perhaps? I would submit that Samskruta (NOT the anclicised Sanskrit) sounds exactly like how it did in ancient times. How come there is no discussion of that at all?

The article is about Indo-European languages, that includes 'indic, iranian, and slavic'. You can write a blog about Samskruta.

Samskruta is an Indo-European language.

I didn't say it wasn't...

Are you saying Samskruta is not an Indic language?

No, I didn't say that.

A discussion of that would have to also include a discussion of why Lithuanian has also not changed much since the vocabulary is filled with Samskruta cognates that Lithuanian does NOT share with Slavic languages like Russian or Polish, some examples from Wikipedia:

- S: sunus, L: sunus (son)

- S: avis, L: avis (sheep)

- S: padas, L: padas (foot)

- S: viras, L: vyras (man, also compare Latin vir [pronounced weer])

- S: dhumas, L: dumas (smoke)

- S: antaras, L: antras (other, second)

- S: vṛkas, L: vilkas (wolf)

- S: yuñje, L: jungiu (I join)

- S: jántis, L: gentys (tribes, also compare Latin gentes [a/the people])

- S: dantas, L: dantis (also compare Latin dentes)

- S: naktis, L: naktis (also compare Latin noctes)

Lithuanian in general is considered one of the most conservative IE languages. There isn't really any clear reason that historical linguistics has produced for whether languages change quickly or slowly afaik.

My understanding was that the isolation of the Baltic speakers was a big contributor, though off the top of my head I wouldn't know the geographical obstacles, hence also why they were the last to Christianity in Europe

Knowing Russian, Lithuanian and bits of Polish I would also add that it's amazing how lithuanian is very similar to slavic languages in grammar structure but radically different phonetically.

I.e. a direct translation, word to word, works (with some style issues), but it's hard to recognize the sentence when pronounced.

This makes me wonder how sanscrit sounds...

Also, the language is quite complex, probably more complex than polish or russian, which is typical for archaic languages, afaik.

I find the rhythm/meter is very Romance-like but without identifying with any one specific Romance-speaking region, the phonetics vaguely remind me of Brazilian Portuguese.

Heh... Brazilian Portuguese is my favourite language I'll never learn :-)

No, it doesn't. End of discussion.

Could you elaborate please? How do you establish that?

I agree with the sentiment but ironically your transliteration of the name of that language reveals that it certainly has undergone at least some sound changes. Why do people in the north pronounce the vocalic "r" as "ri" while in the south they use "ru"? Why do people in the north drop word final schwa but in the south they add a nasal bilabial?

While it's very easy to pronounce Samskrta in a way similar to what was likely used in classical times, there have definitely been sound changes to an extent.

In classical recitation, with ghanapatha and other techniques, I would expect that pronunciation is strict, meters are accounted for, and things have remained relatively unchanged in the way the Vedas and Upanishads are recited. Unfortunately, there is little exposure of the rigorous training in the Western world.

I've listened to rgveda samhita patha of what I believe is the shakala shakha which is predominant in north india but even they seem to mispronounce the vocalic "r" as "ri".

It's more of a Pali/Prakrit influence than a mispronunciation.

Exactly. This transliteration, in eastern India would be 'Shongskrito'. Which is no more right or wrong compared to saying 'Sanskrit'.

Songuahkrito is not Samskruta but Bengali. So what’s your point?

Sure, and that's exactly why pronunciation has changed over the last few thousand years.

People are always going to speak how they want and eventually the 'correct' version isn't the majority. None of the many thousand people who have learnt Sanskrit in the eastern part of India would pronunce the name as 'Samskruta'. You could argue all day about how they are all wrong, that doesn't really change the fact.

While there is an element of truth to that, it's largely because the language was deliberately fossilized for religious reasons, while the "real" language that people actually used day to day continued to evolve around it.

How fossilized? While it is not used as an everyday language, there is a yet growing library of literature, and vigorous philosophical debate that happens in Samskruta. BTW, an unrelated rant, the other thing that grinds on me is that the Ganga river is ALWAYS referred as Ganges (another anglicization)! Ganga is easier to pronounce, so why did that get anglicized, I wonder?

Ganges is what the Hellenic Greeks called it. Most English words for Indian things come from Greek.

I think it is really hard for speakers of European languages to believe that a language could have the same pronunciation that it had centuries ago.

Obvious example:

Arabic from at least the 7th century AD:

As an Arabic speaker, I can totally read a text the way it was pronounced 14 centuries ago and also perfectly understand it.

You cannot say the same thing for European languages.

Probably the same for sanskrit.

> As an Arabic speaker, I can totally read a text the way it was pronounced 14 centuries ago and also perfectly understand it.

I may be able to do something similar in Spanish. I've read Emilianensian /Glosas/ in Navarrese-Aragonese and understand near the 95% of it. That being a close Romance to Castillian Spanish.

Scroll down and compare, you don't need to understand Spanish:

https://es.wikipedia.org/wiki/Glosas_Emilianenses

Also, I can perfectly read Don Quixote. My main issues would be the old Medieval slang with things that do not exist anymore or they changed a lot. But without the footer annotations, I had no issues at all to understand the 80% of it.

But, spelling it? Heck, it didn't change at all, our Spanish standards (Nebrija's grammar) date back to that era.

> As an Arabic speaker, I can totally read a text the way it was pronounced 14 centuries ago

You probably cannot. Every region with a tradition of reciting the Qur'an has their own way of pronouncing it. It would be a strange coincidence if your tradition in particular happened to keep the pronunciation unchanged while all the others didn't.

E.g. do you pronounce ﺽ as a pharyngealized voiced alveolar lateral fricative? That's the most likely reconstruction of the sound: https://en.wikipedia.org/wiki/%E1%B8%8C%C4%81d

>> You probably cannot.

Exactly my point.

You’re under the impression I cannot because of your limited knowledge of this topic: even if 1 letter was slightly different (which it is not), how does that discard the prononciation of 99% of the rest?

>> Every region with a tradition of reciting the Qur'an has their own way of pronouncing it

Plain wrong.

If you knew the topic, you would know that the various way to read the Holy Quran are so close that you could just say they are identical.

In my original post, I’m not even necessarily talking about the holy Quran.

What about the vast poetry written in Arabic in pre-Islamic times that anyone today can read the same way?

I will not get into the subject of whether I would understand it or not since you are still debating whether I would pronounce it the same way.

BTW, this is very different from : how does one pronounce Latin etc.

It seems our disagreement is mostly about the semantics of what it means to pronounce something the same way.

Because it is necessary to read the Qur'an properly, and in order for it not to change or deteriorate over time, old scientists wrote in books how to read each letter and its rules. It is written in detail where each letter will come out of the mouth and what features(like the sibilants, the stressed letters, the soft letters, the labiolinguals, and tremolo etc. ) it will have. For us in Turkey it is different ve more difficult but we learn it to read the Qur'an properly. You can listen a surah from the Quran from a Turkish or Arabic reciter there is no big difference.

> old scientists wrote in books how to read each letter and its rules. It is written in detail where each letter will come out of the mouth and what features(like the sibilants, the stressed letters, the soft letters, the labiolinguals, and tremolo etc. ) it will have

Indeed, that is why modern scientists believe that ﺽ must have represented some kind of lateral sound, like a lateral fricative (or maybe a lateral affricate or a stop with lateral release – the descriptions aren't detailed enough to distinguish between these options.) But most modern speakers nonetheless pronounce it as a plain stop or dental fricative instead of lateral.

Having a description of the pronunciation doesn't mean that someone reading the description will be able to follow it perfectly. Otherwise everyone learning a foreign language would be able to develop a native-like accent just by reading about the correct pronunciation.

Certainly vocabulary has been added to, but the whole point of the work of Pāṇini and others was to prevent the Vedas from drifting over time. That freezing of the core of the language was deliberate and artificial. (The same thing applies to the Arabic of the 7th century. It exists nearly unchanged at its core for religious reasons, and a version with various additions functions as a lingua franca and language of academic discourse - rather as Latin or Greek did in other cultures - in the Islamic sphere.)

And Paninian/Classical Sanskrit is not even Vedic Sanskrit!

Not only that, but in Vedic ceremonies you have a priest who is just there so no mistakes are made.

>The brahman was the reciter of hymns from the atharvaveda who was largely silent and observes the procedures and uses Atharvaveda mantras to 'heal' it when mistakes have been made.

It seems to me that you are proving my point.

That's how foreign names work. People hear the name used in another language, and then they adapt the pronunciation and/or spelling to their own language.

As a Finnish speaker, "Ganges" has the familiar form of a name borrowed from Greek or Latin, and it does not need further adaptation. "Ganga", on the other hand, sounds awkward in Finnish. Had ancient Finns been in contact with India, we would probably know the river as Kanka or Kankajoki ("Kanka river"). The latter in particular sounds like a proper Finnish river name.

That's if we can even all agree on how to pronounce "Ganges"! Should it rhyme with cheese or oranges; are the g's hard or soft? In en_GB the first g is hard and the second one soft and the word rhymes with cheese. Oh, and the a is short.

Place names are notoriously tricky. The town I live in is now called Yeovil. It has well over sixty documented spellings since around 500CE. It started off as Gifle - bend in the river. G -> Y is a common Old English/Saxon to modern English thing (-gard and -yard, -burg and -bury, dag or tag and day.) If you squint really hard you can get the rest of the word OK. It was Ival, Givelle, Evil and all sorts of spellings on the way to today. There was also a vowel shift en-route and a generally open approach to spelling.

if that'd peace your mind, we call it "Gang" in Russia. But I wouldn't be surprised that it came from greek too, most greek words lost their endings when loaned to Russian

Can you give some examples please? I'm Greek and I don't know any Russian.

In Greek we have adopted many Turkish words and we always stick an "-i" at the end. For example, "passoumi" (a kind of footwear, "passoum" in Turkish), "briki" (for making coffee, "brik" in Turkish), "ati" (horse, "at" in Turkish) etc. I think we've also taken some Slavic words and treated them this way, for example "boureki" (borek, dumpling), kokoretsi ("kokorec") and probably others I can't think of now.

it's tamil-ized as Kangai. Proper nouns are always absorbed according to the target language's own rules. Why quibble?

It's been a long time since I've done anything with Sanskrit so please forgive any misunderstandings, but didn't Vedic ळ become Classical ड, as in नीळ and नीड ?

> Vedic ळ

There is strong evidence that the ळ came from Substratum languages (Dravidian or Munda) and many Indo-Aryan languages still retain it e.g. Marathi, Gujarati.

Yes but it's still present in Vedic, literally in the second word of rgveda 1.1.1.

Vedas were not passed unmodified since they were originally written. Neither were the purānas.

The reason there's no discussion about Samskruta is that all the scholars of संस्कृतम are in India and most of the University departments lack the cultural and linguistic nuances that are needed to learn संस्कृतम. Unless Indians come and share their research with the world, there won't be any discussion on it.

I also think this is so because the prevalent narrative is post colonial in nature and the (English) Internet is came out of Anglosphere. This is not a criticism of the prevalent order, it's just the way things have evolved naturally post WW2.

It's संस्कृतम् not संस्कृतम

Another casualty of transliteration keyboards.

Lot of Sanskrit scholars are priests and their bias is shown a lot as well. They cry whenever something becomes slightly better offensive to 2000 year ago people.

No Sanskrit absolutely doesn't sound like it did in ancient the ancient times. Modern (north Indian) Sanskrit pronunciation uses Hindi/Urdu style of pronouncing vowels which was significantly influenced by Arabic and Persian.

Ancient Sanskrit pronunciation was essentially similar to what we call the Prakrit language, which was a more colloquial version and spoken by most people. Bengali, Marathi and Gujarati are a bit closer to that pronunciation, but of course they have changed a fair bit too.

> Modern (north Indian) Sanskrit pronunciation uses Hindi/Urdu style of pronouncing vowels which was significantly influenced by Arabic and Persian.

This has to be the most far-fetched and absurd claim I’ve heard so far. Sanskrit has been orally transmitted since several millennia and there’s no change in its pronunciation and grammar since days of Panini. That robust oral tradition is well researched and proven to be fool proof.

That would make it the only language in the world to be passed down unmodified, which seems... unlikely.

I understand it comes across as jingoistic, but truly Sanskrit grammar and pronunciation hasn’t changed since Pānini. The word Sanskrit itself means “constructed” and there was a rigorous effort to prevent ambiguity.

Grammar hasn't. Pronunciation has, a lot.

You are almost completely wrong. How do you account for the precision of Panini as the previous commenter pointed out? How do you know to postulate that? There is NO other formalized grammar akin to Panini, so don’t judge Samskruta by other so called Indo European standards!

there are no recordings from panini's times and panini's texts pertain to grammar. they don't prescribe pronunciation - like precise position of tongue lips etc...

What are you talking about? Panini did detail the pronunciation of each character. And the actual pronunciation (i.e. accounting for things like assimilation) is then encoded in the writing via Sandhi.

Likewise, we have older information on pronunciation as well in the Pratishakhyas.

Ancient Indians cared a lot about accurate pronunciation, especially for mantras where the thinking was that without correct pronunciation, the mantra would not work. Picture the Harry Potter "wingardium leviosa" scene.

Modern don’t actually hew to the prescribed pronunciation for certain letters, but it’s certainly specified by Panini.

That’s actually the first thing you learn and it’s quite rigorous where to vocalize and how.

Yeah, and it differs based on where you're from!

It is NOT Sanskrit, which was foisted on the world by the british. Please be respectful and call it by it real name, given by Panini which literally codified and formalized it.

Maybe if more people learnt the language instead of crusading on about it's supposed heritage, it would have been the official language of India.

I mean, you could listen to samveda chants from different parts of India and see the pronunciations are very audibly different...

Your kind of hot take about Sanskrit is somewhat common.

I once went to a wedding involving academic linguists and Hindu prayers. A Hindu priest chanted a prayer in Sanskrit and then an academic linguist followed up by chanting the same prayer with his view of what it would have anciently sounded like. Notwithstanding the priest's own view of his Sanskrit as an ancient language, his pronunciation didn't match up with the academic linguist's view of the historical pronunciation.

Relatedly, I went to a spoken Latin event where I (an American) spoke Latin with an Italian man. His pronunciation and mine were wildly different, although we could understand each other. Neither was all that similar to the academic linguists' view of what our conversation would have sounded like thousands of years ago, though I guess his vowels were generally closer and my consonants were generally closer.

People retaining ancient languages for liturgical and scholarly uses are generally not that great at preventing the pronunciation from shifting, and I think that's true across the board. ("Atque memento, nulli adsunt Romanorum qui locutionem tuam corrigant." -- Henry Beard.)

"There are no more Romans to correct your speech"?

"And remember, there are no Romans around to correct your pronunciation."

Cheers. Yeah, I let the "remember" bit drop :)

I don't understand why an academic linguist is more qualified than a traditionally educated scholar in India, who has deep knowledge of the classics. I don't deny that there were likely some influences, but the care and precision with I was taught recitation is something I have not encountered in any other discipline. I am an engineer, but if I start from the Ashtadhyayi, I don't see how things could have changed much.

My understanding is that it's not clear to what extent Panini understood or attempted to explicitly treat articulatory phonetics. Addressing articulatory phonetics well in writing alone is still hard, and understanding it in detail was immensely hard before the creation of audio recording, spectrograms, and x-rays.

https://en.wikipedia.org/wiki/Shiva_Sutras#Arrangement

So while a present-day understanding of classical Sanskrit grammar could be derived fairly directly from his work, that's still not any guarantee that you would pronounce Sanskrit just as Panini would have.

A moderately imperfect analogy is that while Jews have been carefully learning to read, recite, and copy the Torah for thousands of years -- including reading commentaries, which, while nowhere near as systematic or thorough as Panini, elucidate lots of grammatical issues -- there still emerged huge, systematic regional differences in the pronunciation of particular Hebrew letters, even in liturgical use.

https://en.wikipedia.org/wiki/Ashkenazi_Hebrew

https://en.wikipedia.org/wiki/Sephardi_Hebrew

https://en.wikipedia.org/wiki/Mizrahi_Hebrew

https://en.wikipedia.org/wiki/Tiberian_Hebrew

among several others. The people using each pronunciation were often incredibly sophisticated in their religious scholarship and incredibly familiar with the underlying material, but they definitely didn't (and don't) pronounce it the same way.

Modern linguistics has been able to find evidence for comparing these pronunciation variants and also understanding their histories, and I believe something similar is true for Sanskrit pronunciation.

While it’s very true that we have a much better idea of how Sanskrit was pronounced millennia ago than we do for most languages, we can’t make absolute claims like yours.

No, that’s not true. there’s a bunch of varieties of Sanskrit, including Vedic Sanskrit, Classical Sanskrit, and various local dialects. Not all of these have been preserved in oral traditions.

Second, even within the preserved oral traditions, pronunciations vary across schools. For example, the “kru” part in your Samskruta is pronounced differently in different parts of India; some people do say “kri”

> How come there is no discussion of that at all?

Leaving aside the linguistic debate in the comments here, there’s often no fair or truthful or representative discussion of the ancient East’s legacy (or others) for reasons relating to imperialism, colonialism, and supremacist thinking. I don’t mean that in the explicit “we’re taking over your country way” but in the more subtle or implicit ways. Western world views and Eurocentric histories are everywhere, from academia to Wikipedia, and due credit or consideration isn’t given elsewhere.

The problem persists even in the very countries we are talking about, like India, since their earlier “original” Hindu culture was changed by invasions (whether from the Islamic Middle East or Christian Europe). My understanding is that when the British finally left India, their academic culture still persisted in the form of Western education that replaced earlier forms of education, textbooks sourced from Western publishers, taxpayer funding into this system, and the adoption of languages like Hindi and English as the official languages (rather than say Sanskrit or Tamil).

So to answer your question in the broader sense beyond this article and discussion, there is no discussion of these things in part because of the power of Western spheres and in part because Eastern spheres themselves haven’t retained their own knowledge or promoted themselves adequately either domestically or internationally.

Afaik the original culture (take India) has been massively influenced by the English occupation and Mogul occupation.

Lots of the original way people thought, lived their culture and worshipped in their religion has changed.

Which ancient times? It was only frozen after Panini’s time, but there had been changes from original rigvedic pronunciation by then.

And your quibbling is indicative that there have been also been changes today as well - e.g. southern speakers tend to pronounce vowel r as “ru” while others say “ri” and at some point in time the pronunciation was likely more like American English r.

Vedic Samskrutham was pretty much different I guess.

And regular people didn't speak sanskrit in classical age. They spoke regional languages like Prakrit and Pali.

Vedic Samskrutam was a different language than Paninis classical Samskrutam. We need to recognize size that.

This has been an incredible discussion, thank you all for contributing. I hope my original point was made, which was that there is often bias when the western world talks about “ancient”, and I place the blame squarely on the imperialistic british for that. The other issue I want to point out is that there is a certain arrogance that the linguistic academia has as it relates to traditional scholars by regularly discounting their knowledge and assuming their own superiority. Not trying to change anyone’s beliefs, just an opinion I am expressing. I am going to drop out not for want of willingness to engage, but because this discussion is devolving into familiar patterns of debate. All I can say as a practitioner is that when I was taught, I was actually taught the “correct way” and the differences so I was always aware. There is an academic rigor in traditional teaching that Western academia seems to be unaware of.

If you want a real mind trip, check out the reconstructed pronunciation of Classical Chinese.

https://en.m.wikipedia.org/wiki/Old_Chinese_phonology

A video sample: https://www.youtube.com/watch?v=zyaFKnUumAM. I find myself quite liking the sound of it, in particular the poem which is recited — the writers of the Shijing (from which the poem is taken) paid a lot of attention to rhyme and assonance. Even so, Old Chinese phonology is really weird: pharyngealisation is decidedly odd even for Sino–Tibetan (though some Qiangic language have pharyngealised vowels), and labialised glottals are only slightly less so.

This is the reason I’m not fully convinced by Baxter and Sagart’s reconstruction of pharyngealized consonants everywhere: it’s just too weird. In languages like Arabic that do have pharyngealization, it doesn’t apply to every consonant like in the BS2014 reconstruction. Still, I don’t see a better way to explain the Type A / Type B distinction…

Well, Sino–Tibetan is one of the only places I would expect to see pharyngealisation in the first place. I can easily imagine a scenario in which vowels had a uvularisation or pharyngealisation contrast, which then was transphonologised onto the previous consonant. But I don’t know too much about Chinese languages, so until another explanation comes along I can only accept Baxter and Sagart’s conclusion.

Shijing is a collection of folksongs, so of course the words must sound very nice at the day, or the music behind them must be good. Jian Jia is in my opinion one of the best, which by the way rhymes well even in Mandarin.

But I don't like how it sounds in the video, and those weird sound bits (no idea what you call them) is not very friendly for singing either.

> But I don't like how it sounds in the video, and those weird sound bits (no idea what you call them) is not very friendly for singing either.

Could you describe in more detail which bits sound weird? Of course Old Chinese would sound very different to Mandarin, but I didn’t notice anything actually weird about it.

Perhaps this is the right thread to ask the question as to how people figured out what ancient Egyptian words sound.

From what I understand Rosetta Stone simply provided a dictionary of sorts, so I can understand how they know which words mean what. But how do they know that a certain combination of hieroglyphs translates into words pronounced as say “Ramses” or “Tutankhamen “ and so on.

https://youtu.be/5NeccdDpO5I

I found this to be a great source on the topic. Your question is a great one, and hints as to why Jean-Francois Champollion is hailed as the genius he was. The key insight was essentially the theory that the Coptic language spoken still as the liturgical language of Coptic Christians in Egypt was a modern form of the ancient Egyptian language encoded by Hieroglyphics. Champollion made a key insight attempting to decipher a cartouche (the hieroglyphic form used to write a name) that a symbol at the beginning of a name looked extremely similar to a sun. The Coptic word for sun was Ra. At the time, they’d already identified the symbol used to writes a ‘ss’ sound, and so Champollion was able to guess at the middle symbols spelling Mes, leading to the deciphering of Rameses. Note that the European world at large obviously already knew the name Rameses, which was key here.

I believe in general names were key to deciphering many of the sounds used in the language, coupled with pronunciations hinted at by Coptic.

I typed this all by memory so I may be off in some details but I believe I have the gist right.

The video looks great, appreciate the recommendation.

We know these names because the Rosetta Stone (and other dual-language inscriptions) allowed researchers to figure out which Greek letter the egyptian hieroglyphics mapped to. There is also the Coptic language, which is a modern descendent of Ancient Egyptian.

However, these pronunciations are full of issues and uncertainties - in part because, as the OP states, even figuring out how exactly Greek was pronounced is sort of a challenge, and because Ancient Egyptian did not denote most vowels, so some uncertainty is unavoidable.

For example, Egyptologists will usually transcribe the name we usually pronounce as Tutankhamen as "twt-ꜥnḫ-jmn". We can tell which phonemes are distinct and have some idea of how they must have been pronounced, but we can never truly know for sure.

So the answer is: The word probably wasn't pronounced as "Ramses", the name "Ramses" is an approximation of a Greek approximation of the actual name of the pharaoh, optimized to be easy to pronounce by English-speakers.

Makes sense, BTW I appreciate all the sibling replies as well.

So basically it seems they used the proper nouns to start and then built up on that. And also used Coptic to fill gaps..

I do get the point about Greek approximation since just as an example in modern India, Alexander is referred to as Sikander. Similar, yet very different.

The Al- most likely got lost in Arabic due to being reanalyzed as a determiner instead of as part of the name.

For the Greeks, it even had an intelligible meaning

https://en.wiktionary.org/wiki/%E1%BC%80%CE%BB%CE%AD%CE%BE%C...

I think the a- in the al- in Alexander is supposed to be the alpha privative

https://en.wikipedia.org/wiki/Alpha_privative

like our "ward off" or something.

I don't know if that's right. It could be but it sounds wrong in that I've never heard of "έξω", on its own, without the possible privative, at least not as a verb (there is a modern noun "έξω" meaning "out").

Also, wikipedia says this about alpha privative:

>> It is derived from a Proto-Indo-European syllabic nasal n̥-, the zero ablaut grade of the negation ne, i.e. /n/ used as a vowel.

But the wiktionary article says this about αλέξω:

From Proto-Hellenic aléksō (“to protect”), from Proto-Indo-European h₂lek- (“to defend”) (...)

Note the difference between "n̥" and "h₂". I don't even know what those are, but they seem different enough that they may be ... different?

Alexander is pronounced as “İskender” in modern Turkish as well.

> So basically it seems they used the proper nouns to start and then built up on that. And also used Coptic to fill gaps..

In the case of Ramses specifically, there is also a piece of biblical evidence, the city recorded in Hebrew as something like "Raamses". https://en.wikipedia.org/wiki/Pi-Ramesses#Biblical_Ramesses

The two biblical cities of Pithom and Raamses are traditionally identified as the Egyptian "per-Atum" and "per-Ramesses", for some convention on spelling the Egyptian names.

Royal names in cartouches provided a clue. The Rosetta Stone provided the spelling of the name Ptolemy in Greek, Hieratic, and Egyptian hieroglyphs. Other cartouches with known values filled in a lot. One major problem is that Egyptian hieroglyphs don't really encode vowels. Another set of clues about pronunciation comes from the Amarna letters written in the 18th dynasty. They are written in Akkadian cuneiform (then the language of international diplomacy), a script that does include vowels. These letters contain Egyptian names and sometimes loan words.

> They are written in Akkadian cuneiform (then the language of international diplomacy), a script that does include vowels. These letters contain Egyptian names and sometimes loan words.

I was interested to learn that the name of Ra is attested in these letters! From that, we know the vowel -- it was /i/.

(The "a" in Ra is a consonant which is traditionally rendered as "a" by convention.)

Another angle I didn't see in the other comments was comparative linguistics. This came later and wasn't used in the initial deciphering, but it's a major part of the modern understanding.

Ancient Egyptian was related to other languages, and was part of the rather large still-existing Afroasiatic language family, which includes Semitic (Hebrew and Arabic and others), Berber (of north Africa), and many others. (Importantly, Akkadian is one of the first written languages aside from Ancient Egyptian itself, and it was a very early Semitic language, and not so dissimilar from Ancient Egyptian in some ways.)

While only reconstructions, this gives us, in a sense, a model of the ancestral language of Ancient Egyptian. This insight provides more data on possible phonetic values and the nature of that language. There is always a great deal of uncertainty though.

By comparison, Sumerian, of similar antiquity, was an isolate and we do not know any language related to it. We have many texts in Sumerian, and many parallel texts between Sumerian and Akkadian, as well as old Indo-European languages like early Persian. We kind of have the sketch of its sounds. But even then, we have less idea what Sumerian sounded like, or how its grammar works, or what its words mean, than we do Ancient Egyptian. And that is despite the fact that Ancient Egyptians didn't write their vowels while the Sumerians did! Having related languages you can compare to is valuable here.

I don't doubt that part of the pronunciation problem is the deeper meanings that creep in.

'CAESAR' sounds positively German and Legio VI Victrix has an effeminate sound. I'm afraid that Biggus Dickus had at least a little to do with attitudes towards proper Latin.

TLDR:

The difficult task of reproducing the pitch accents in Greek might be a step too far for most people, but consider the following line from Homer’s Iliad 1.345:

ὣς φάτο Πάτροκλος δὲ φίλῳ ἐπεπείθεθ᾽ ἑταίρῳ

It develops beautifully from sauntering labial alliteration (of p and ph sounds) to rapid-fire dental consonance (of th and t sounds), but if we pronounce Φ and Θ as fricatives here:

https://antigonejournal.com/wp-content/uploads/2021/07/Iliad...

rather than aspirated stops:

https://antigonejournal.com/wp-content/uploads/2021/07/Iliad...

then the distinction between Φ and Π and between Θ and Τ becomes so great that the sound effect is lost completely. Homer, for one, would have cared.

Yeah, that was beautiful to hear the difference!

Never mind the fricatives. From a modern Greek point of view it's the pronounciation of ἑταίρῳ as "eta-eroy" that is the hardest to stomach.

That's the Erasmain pronounciation, if I remember correctly: saying vowel dipthongs as two separate vowels. That really sounds awful and very much not like Greek.

For people who are interested in ancient Greek, Podium Arts' YouTube channel is filled with professional readings of Attic Greek works, like Apologia[0], Plato's Apology

[0]https://www.youtube.com/watch?v=k3tlBO3agFU

Ancient Chinese reconstruction:

https://www.youtube.com/watch?v=VIkQSuZbbt0

To my untrained ear this sounds almost like a mix of Hindi and Mandarin. Interesting.

On Spanish, Middle Age Spanish had different spellings for v and b, and c, s and z were spelt differently, all of them closer to Latin.

If this is interesting to you, you may also enjoy this ted talk on Indigenous languages, personally I found it very interesting:

A history of Indigenous languages -- and how to revitalize them: https://www.youtube.com/watch?v=BoRXbjGJrI0

Reply on news.ycombinator.com