Proto-Italic reconstruction

kizolk

Civis Illustris

  • Civis Illustris

Location:
Bourgogne, France
I wasn't sure where to post this, but I came across a nice reading of the Pater Noster in Proto-Italic:


The description of the video doesn't give much detail, but it is based on a reconstruction by Reuben Pitts, a linguist specialized in the languages of the Italic peninsula of the first millenium BCE, and it is read by Raphael Turrigiano, an American... language-lover? I don't really know his credentials, but I know he's interested in the phonology of ancient (and modern) languages.

I wonder how much I would understand of it if I didn't know the Pater Noster. In any case I find it beautiful.
 

kizolk

Civis Illustris

  • Civis Illustris

Location:
Bourgogne, France
Mr. Turrigiano himself :D Great to have you here. I didn't know much about your background but I think it's safe to say I've seen most of your YT content: Ecolinguist, Liga Romanica (probably the best channel for Romance geeks like me!), your appearances with Luke and your own videos, all very good stuff ^^

I imagine the process of reconstructing such a language is incredibly hard, and also tentative. Simply put: to what extent would you say this particular reconstruction is scientifically sound? To take a specific example, a few of those phonemes aren't present in Latin; I guess they come from PIE but on what grounds can you say they were present in proto-Italic, when they could have been lost earlier?
 

RaphaelT

New Member

Thank you very much! It's always great to hear that people enjoy the nerdy shenanigans we get up to haha. That's a really good question! So this particular reconstruction is overall very tentative, though the fact that it's such a simple text means it's not unlikely that a real proto Italic speaker would have been able to more or less understand what's being said. The most tentative thing is the syntax - that's basically just Latin with a handful of adjustments based on what we know about proto Italic, e.g. that there was a productive locative case. For a more complex text this would become more of a problem, but this is so simple that there's little reason to believe that a very Latin-y syntax would be wrong, as opposed to just a bit odd.

The next most tentative element of this reconstruction is the vocabulary - almost all of these words certainly existed in proto Italic since they are almost all attested in all of the Italic languages with more or less the same meaning. However, there may have been other terms that were more idiomatic but then fell out of use, much like how Latin 'equus' is completely gone in Romance (though the feminine 'equa' has some surviving descendents). To this end the reconstruction includes a couple of words not attested in Italic but rather sourced from PIE. This gives the impression to a modern listener of some divergence in vocab but is of course just an educated guess. Off the top of my head, 'venscā' is such a word which might have just not existed in Proto Italic or might have had a different meaning.

Finally there's the phonology, and this I would say is the least tentative element. Almost all of the phonological differences between Proto Italic and Latin are pretty clearly attested in either the oldest stages of Latin or in other Italic languages. Word initial stress, bilabial /f/, diphthongs for classical monophthongs, hard final nasal consonants, a variety of fricatives descending from PIE breathy stops, complex consonant clusters, etc. are all either present in the written record, or are necessary to explain divergence in the Italic languages.

For instance, how do we know F was bilabial? Developments in Faliscan (confusion of F and H) as well as in Old Spanish indicate that this pronunciation existed even in the classical period, and in the oldest latin text this sound is written VH /hw/ (you'll see it transcribed FH but the F in this case is a digamma representing /w/). This is a really good approximation of what a bilabial fricative sounds like if you don't have a dedicated letter for it.

Another example - how can we be sure about the word 'louðeros' which in Latin becomes 'līber'. Well, let's compare Italic forms:

Old Latin: loeber
Faliscan: loifero(s)
Oscan: loufris (genitive)
Venetic: louderobos (dative plural)

And now some Indo European cognates:

same formation:
Greek: eleútheros

Same root:
Icelandic: lýður
Russian: ljudi
Sanskrit: rodhati

Now let's go phoneme by phoneme:

/l/ - all through Italic and IE aside from Sanskrit which regularly has /r/ here.

/ou/ - this is generally reconstructed as /ew/ in Indo European (as in Greek) which then becomes ou in Italic. /oi/ seems to be restricted to Latino-Faliscan for this particular word, generally they share /ou/ which goes to ū in classical Latin. That this is attested in Oscan and Venetic, distinct branches of Italic, makes this the best candidate for the sound in Proto Italic.

/ð/ - this can't have been /b~v~f/ already because it is /d/ in venetic. It can't have been /d/ in Proto Italic because then it wouldn't have become b~v~f in Latin, Faliscan or Oscan. Other Indo European cognates make it clear that this sound was originally dh as in Sanskrit, which became a fricative in Italic. Thus the only candidate is /ð/.

-eros is pretty much directly attested. The -er ending happens in Classical Latin due to loss of vowels between r and s, so basically ros > rz > rr > r. This is how you get lots of Latin words that end in r, e.g. ter < terr < terz < trz < tris, or ager < agerr < agerz < agrz < agros (the rr stage is attested)

So basically, pretty much every word in this reconstruction would have been pronounced almost exactly as we've reconstructed it at the phonological level, though there could be some phonetic details that aren't reconstructable.

So to sum up, I'm very certain most of this text would be understood if read aloud to an Italic speaker from 1500 BCE, but some words would probably be misunderstood and it would sound very strange in all likelyhood, especially in terms of syntax and wording but also potentially due to some missing phonological details.
 

kizolk

Civis Illustris

  • Civis Illustris

Location:
Bourgogne, France
Thanks for your great answer! Fascinating. I mean I understand some of the reconstruction techniques involved, but at this level of detail it's an art.

Keep up the great work! :)
 
Last edited:

Glabrigausapes

Philistine

  • Civis Illustris

Location:
Milwaukee
Not that I find it unlikely, but could I ask, @RaphaelT, what is the justification for reconstructing medial /h/ as velar fricative (if in fact that's what your /γ/ represents)? Are you maybe relying on analogous lenition in Romance?
 

RaphaelT

New Member

Not that I find it unlikely, but could I ask, @RaphaelT, what is the justification for reconstructing medial /h/ as velar fricative (if in fact that's what your /γ/ represents)? Are you maybe relying on analogous lenition in Romance?
Good question! So /h/ in Latin comes from PIE *gʰ. Just like how *dʰ becomes ð and *bʰ becomes β. These sounds all had voiceless allophones word initially, and these voiceless allophones subsequently merged such that there was only ɸ, which is what becomes Latin /f/. This either happened in late Proto Italic, or was an areal feature that spread across the Italic languages, e.g. proto Italic *ɸakjō (whence Latin faciō) came from earlier *θakjō, and it's actually a cognate to the English word 'do'.

Word medially however the situation is different. In all Italic languages ɣ ends up as ɦ (contrary to how most people pronounce it, H was probably never voiceless word medially given that it eventually disappears and it comes from a voiced sound). ð and β are a bit more complex - they partially merge in Latin (thus liber and ruber rather than lider and ruder, but medius rather than mebius) and completely merge in the rest of Italic. The difference is that in standard Latin they become stops in this position, whereas in the rest of Italic they remain (probably voiced) fricatives, and there is dialectic evidence of this in Latin as well (rufus, for instance, is from the same root as ruber).

So in short, PIE *gʰ must have at some point become ɣ, which became voiceless [h] word initially and voiced [ɦ] word internally. This shift may well have been completed before proto Italic split apart, so it's sort of an arbitrary decision whether to pronounce [ɦ] or [ɣ] as I did. It's also quite likely that both realizations coexisted for a while - a good example of this is the modern Slavic dialect continuum. Proto Slavic *g has the following realizations:

[g], as in standard Russian, all of south Slavic, Polish
[ɣ] as in standard Belarusian, dialectal Russian, dialectal Ukranian
[ɦ] as in standard Ukranian, dialectal Russian, dialectal Belarusian, Czech and Slovak

However, we also see that sometimes in standard Russian *g has ended up as [v] - this means that it was probably [ɣ] at some point, but then variably hardened to [g] or shifted to [v].

I hope that answers the question haha.
 

AndreaMiglio

New Member

Thank you very much! It's always great to hear that people enjoy the nerdy shenanigans we get up to haha. That's a really good question! So this particular reconstruction is overall very tentative, though the fact that it's such a simple text means it's not unlikely that a real proto Italic speaker would have been able to more or less understand what's being said. The most tentative thing is the syntax - that's basically just Latin with a handful of adjustments based on what we know about proto Italic, e.g. that there was a productive locative case. For a more complex text this would become more of a problem, but this is so simple that there's little reason to believe that a very Latin-y syntax would be wrong, as opposed to just a bit odd.

The next most tentative element of this reconstruction is the vocabulary - almost all of these words certainly existed in proto Italic since they are almost all attested in all of the Italic languages with more or less the same meaning. However, there may have been other terms that were more idiomatic but then fell out of use, much like how Latin 'equus' is completely gone in Romance (though the feminine 'equa' has some surviving descendents). To this end the reconstruction includes a couple of words not attested in Italic but rather sourced from PIE. This gives the impression to a modern listener of some divergence in vocab but is of course just an educated guess. Off the top of my head, 'venscā' is such a word which might have just not existed in Proto Italic or might have had a different meaning.

Finally there's the phonology, and this I would say is the least tentative element. Almost all of the phonological differences between Proto Italic and Latin are pretty clearly attested in either the oldest stages of Latin or in other Italic languages. Word initial stress, bilabial /f/, diphthongs for classical monophthongs, hard final nasal consonants, a variety of fricatives descending from PIE breathy stops, complex consonant clusters, etc. are all either present in the written record, or are necessary to explain divergence in the Italic languages.

For instance, how do we know F was bilabial? Developments in Faliscan (confusion of F and H) as well as in Old Spanish indicate that this pronunciation existed even in the classical period, and in the oldest latin text this sound is written VH /hw/ (you'll see it transcribed FH but the F in this case is a digamma representing /w/). This is a really good approximation of what a bilabial fricative sounds like if you don't have a dedicated letter for it.

Another example - how can we be sure about the word 'louðeros' which in Latin becomes 'līber'. Well, let's compare Italic forms:

Old Latin: loeber
Faliscan: loifero(s)
Oscan: loufris (genitive)
Venetic: louderobos (dative plural)

And now some Indo European cognates:

same formation:
Greek: eleútheros

Same root:
Icelandic: lýður
Russian: ljudi
Sanskrit: rodhati

Now let's go phoneme by phoneme:

/l/ - all through Italic and IE aside from Sanskrit which regularly has /r/ here.

/ou/ - this is generally reconstructed as /ew/ in Indo European (as in Greek) which then becomes ou in Italic. /oi/ seems to be restricted to Latino-Faliscan for this particular word, generally they share /ou/ which goes to ū in classical Latin. That this is attested in Oscan and Venetic, distinct branches of Italic, makes this the best candidate for the sound in Proto Italic.

/ð/ - this can't have been /b~v~f/ already because it is /d/ in venetic. It can't have been /d/ in Proto Italic because then it wouldn't have become b~v~f in Latin, Faliscan or Oscan. Other Indo European cognates make it clear that this sound was originally dh as in Sanskrit, which became a fricative in Italic. Thus the only candidate is /ð/.

-eros is pretty much directly attested. The -er ending happens in Classical Latin due to loss of vowels between r and s, so basically ros > rz > rr > r. This is how you get lots of Latin words that end in r, e.g. ter < terr < terz < trz < tris, or ager < agerr < agerz < agrz < agros (the rr stage is attested)

So basically, pretty much every word in this reconstruction would have been pronounced almost exactly as we've reconstructed it at the phonological level, though there could be some phonetic details that aren't reconstructable.

So to sum up, I'm very certain most of this text would be understood if read aloud to an Italic speaker from 1500 BCE, but some words would probably be misunderstood and it would sound very strange in all likelyhood, especially in terms of syntax and wording but also potentially due to some missing phonological details.
Thank you kizolk for sharing this video!
Dear Rapahel, I just joined this forum because I have some questions regarding your Proto-Italic version of the Pater nostere:
1) I don't understand the meaning of the term "Rēins" and its purpose in the period. Whats does it mean?
2) As far as I know the term "χōdke" means "In this day", so it already includes the term "today". Is "djowe" still required to complete the sentence correctly?
 

RaphaelT

New Member

Thank you kizolk for sharing this video!
Dear Rapahel, I just joined this forum because I have some questions regarding your Proto-Italic version of the Pater nostere:
1) I don't understand the meaning of the term "Rēins" and its purpose in the period. Whats does it mean?
2) As far as I know the term "χōdke" means "In this day", so it already includes the term "today". Is "djowe" still required to complete the sentence correctly?
Ciao Andrea!

1) rēins is the accusative plural of reis, which became classical Latin rēs. So rēins dēɣabetās = literally "things owed." I'm actually not sure why Reuben didn't just use the neuter plural as in the classical original, but I presume he had a reasoning, and even if not it still makes sense.

2) No, xōdke is the equivalent of classical hōc, so it just means 'this'. djowe is day, so together they form 'on this day'. These words got smashed together to eventually form classical Latin hodiē.
 

RaphaelT

New Member

I love it, both the phonetics and the usage of the cases (and limited usage of prepositions), but he could have tried more to actually recite it... as in talk slower, so one can catch details. But awesome!
Heh, well I can't please everyone! In this case I followed my own aesthetic preference - I dislike the over recitation that is typical of the way people read ancient texts, and prefer to make it sound as natural as possible, especially since the point here is mainly to show what reconstructed Proto Italic sounded like - the content is largely incidental. But still, I appreciate the feedback! :)
 
 

Godmy

Sīmia Illūstris

  • Censor

Location:
Bohemia
Heh, well I can't please everyone! In this case I followed my own aesthetic preference - I dislike the over recitation that is typical of the way people read ancient texts, and prefer to make it sound as natural as possible, especially since the point here is mainly to show what reconstructed Proto Italic sounded like - the content is largely incidental. But still, I appreciate the feedback! :)
Ah, it is your work. Still, an incredibly good job!

Edit: I'm only now reading your additional posts in this thread! Very interesting...
 
Last edited:
Top