(changed forum) Is Hebrew a Conlang?

A forum for all topics related to constructed languages
marcege
hieroglyphic
hieroglyphic
Posts: 45
Joined: 16 Jul 2018 22:19

(changed forum) Is Hebrew a Conlang?

Post by marcege »

Hi, just changing forums for this.

Wondering if I can get some feedback.

There are some aspects of the Hebrew language that tickled me to think that the Hebrew might itself be a constructed language. I wrote it up as a hypothesis that could potentially be tested, or maybe there is no great way to test it.

Draft / not final, I don't have all the references in / etc: https://docs.google.com/document/d/e/2P ... xjAnC7/pub

Is there a known way to tell if a particular language is a conlang vs. natural? Particular features; unusual distribution in phoneme space; something like that?

Criticism or other approaches much appreciated!


thanks,
-Marc
User avatar
Ahzoh
mongolian
mongolian
Posts: 4191
Joined: 20 Oct 2013 02:57
Location: Canada

Re: (changed forum) Is Hebrew a Conlang?

Post by Ahzoh »

It's not really constructed because it still uses Biblical Hebrew as a base in lexicon and grammar and took things from Yiddish to fill in the gaps. It's also got native speakers (and has for a while now), which definitively makes it a natural language. Biblical Hebrew itself is definitely a natural language.
Particular features; unusual distribution in phoneme space; something like that?
Generally not, except oligosynthesis as a morphological feature which is unattested.
Image Śād Warḫallun (Vrkhazhian) [ WIKI | CWS ]
marcege
hieroglyphic
hieroglyphic
Posts: 45
Joined: 16 Jul 2018 22:19

Re: (changed forum) Is Hebrew a Conlang?

Post by marcege »

Hi Ahzoh,

Thank you for your reply. My question was about Biblical Hebrew.

Did you happen to see the link I shared? One of the questions is whether a natural language is likely to have so many single letter words.

https://docs.google.com/document/d/e/2P ... xjAnC7/pub

Other questions too. Thanks, much appreciated!

Marc
User avatar
Ahzoh
mongolian
mongolian
Posts: 4191
Joined: 20 Oct 2013 02:57
Location: Canada

Re: (changed forum) Is Hebrew a Conlang?

Post by Ahzoh »

Well you talked of Gematria, but it's not unusual for languages to use letters of the writing system to indicate numerals. Glagolitic and Greek did it.

Biblical Hebrew has very clear origins from older languages (e.g. a descendant of Proto-Canaanite which is itself a descendant of Proto-Semitic) that can be traced back through time linguistically. And it really is the case that any language, no matter how strange or implausible its features may seem, could be a natural language if it fulfills the sole requirement of having native speakers.
One of the questions is whether a natural language is likely to have so many single letter words.
Yes. Lots of languages do. Especially if said languages are written in writing systems that represent only consonants or entire syllables.
Image Śād Warḫallun (Vrkhazhian) [ WIKI | CWS ]
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: (changed forum) Is Hebrew a Conlang?

Post by sangi39 »

First, on the note of "pangrams", it would seem reasonable to assume that any sufficiently large text will likely contain all of the letters within a language's alphabet at some at some point. Have you compared your premise that "faithful records written down of existing oral stories do not happen to include the development" of pangram sentences with other written versions of oral pieces, such as the Odyssey, Iliad, Serbian oral poetry, the Poetic Edda, etc.?

As to prefixes, what about Russian? At least /v/, /s/, /k/, /o/, and /u/, off the top of my head, are single phoneme prepositional morphemes, and then you can go even further with the "economy of affixes" in fusional languages where, in Spanish as an example, a single phoneme can represent the person, number and gender of the subject as well as the tense and aspect of the verb. I don't see how "economy of morphemes" can be an indicator of a language being a conlang.

On to the triconsonantal roots of the Hebrew language, this is a feature found in other related Semitic languages, and there are examples of roots in Hebrew which have as few as two and up to four or five "root" letters, which aren't borrowed from other languages (although I can't think of any off the top of my head). Proto-Indo-European is similar, for example, in having an abundance of roots which consist of a CVC syllable, with the vowel changes depending on the number and case of the noun or the tense of the verb, i.e. ablaut. The predominance of triconsonantal roots, at least in my opinion. Take a look at Old Chinese, which is reconstructed most often as being either entirely monosyllabic or sesquisyllabic at most, and at the very least transitioning into a wholly isolating morphology, a trait which is shared with modern languages like Khmer and Vietnamese. How is "Tất cả mọi người sinh ra đều được tự do và bình đẳng về nhân phẩm và quyền lợi" any more unnatural?

Numbers: I'd like to see your source for claiming that in Ancient Egyptian the order of the numbers doesn't matter. In every text I've seen, numbers are presented in decreasing orders of 10, with the only difference in order being the result of changes in the direction of writing across the entire text, e.g. in texts written left to right, higher powers of 10 appear on the left, lower ones on the right, and similarly in texts written right to left, higher powers of 10 appear instead on the right. I've also never heard of this claim for Hebrew except in the expressed instance that gematria is being used. Gematria is more properly a system of numerology, similar to Greek Isopsephy, Arabic abjad numerals, which are distinct from standard Arabic numerals. The assigning of numerical values to Hebrew letters is related to but distinct from this. As far as I know, while saying that גִּימֵל in (some forms of) Gematria is 3+10+40+30 = 83 is reasonable, it is not reasonable to say that this actually represents the number 83, which would instead be פג and never גפ (the same thing, as far as I understand it, is true in the Greek system, i.e. while you can assign a word a numerical value using a similar system, actual numbers are written systematically and uniquely).

37 and 73 are "star numbers" precisely because they form this pattern. However, given all of the verses throughout the entire Torah, are there any others? Are there any other verses which, through Gematria, have values of 1, 13, 37, 73, 121, 181, 253, 337, 433, 541, 661, etc., or are Gematria values more evenly distributed? Why is the tetragrammaton broken into two stars, presumably taken from a division of the value 26 divided into two 13s, which are star numbers, rather than a single value of 26? Are there any other geometrical patterns which appear. For example, 37 is also a centred hexagonal number, so are there any more of them? What about square numbers? Cube numbers? Triangle numbers? Do these appear relatively evenly, or does one pattern appear more often, and what might the chances of this be?

"It seems unlikely that anyone could master an existing natural language well enough to write a text with the level of wordplay and letterplay exhibited in Torah. Try in English to find multiple sentences and words that tell a story that also all exhibit the same geometric motif when letters are interpreted as numerals, and make sure at least one salient sentence (first, last, key concept) shows letter-level symmetry when curled into some shape". Yeah, you probably could. People have been trying this for centuries, the Torah included, even when written in other languages. Again, all you have to do is find the right word, phrase or sentence, and then find enough similar examples, and you're sorted. And given that different systems of numerology assign values differently, I wouldn't be surprised if this sort of thing appears everywhere in the Qur'an or the Odyssey. You have to show that your results are statistically significant and can't be interpreted any other way.

And that's about as far as I can go, I think. The idea that the author of the Torah is only one person really is up for debate, and the idea that there were multiple authors is the current consensus, especially amongst those who don't believe it was written by Moses (whether the direct word of God or not).

As for Hebrew in general being constructed, especially Biblical Hebrew, it really doesn't seem likely. The relationship it shares with languages like Akkadian, Arabic and Amharic are fairly well understood, as far as I can tell, and it does seem the Hebrew, like these other languages, derived from an ancestral Proto-Semitic language. Maybe it is possible that which ever person or people wrote up the final copy of the Torah threw in some interesting numerological nuggets, but to extend that to "Hebrew is a constructed language" is a bit of a stretch given evidence to the contrary.

You say, for example, that "it is hard to imagine how it could come to be that this is just a phonetic representation of speech that people were actually using", but then if that's the case, then Hebrew would have been constructed while the Torah was being written, and then an entire group of several thousand people would have had to have learned it in order to understand the Torah, rather than just, say, sticking with the religion and language they already had, and you're saying a single author managed to do this, when even in the modern day languages like Esperanto struggle to gain a foothold within any culture anywhere, let alone take it over?
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
marcege
hieroglyphic
hieroglyphic
Posts: 45
Joined: 16 Jul 2018 22:19

Re: (changed forum) Is Hebrew a Conlang?

Post by marcege »

Ahzoh wrote: 17 Jul 2018 02:23 Well you talked of Gematria, but it's not unusual for languages to use letters of the writing system to indicate numerals. Glagolitic and Greek did it.

Biblical Hebrew has very clear origins from older languages (e.g. a descendant of Proto-Canaanite which is itself a descendant of Proto-Semitic) that can be traced back through time linguistically. And it really is the case that any language, no matter how strange or implausible its features may seem, could be a natural language if it fulfills the sole requirement of having native speakers.
One of the questions is whether a natural language is likely to have so many single letter words.
Yes. Lots of languages do. Especially if said languages are written in writing systems that represent only consonants or entire syllables.
Ahzoh, thanks, that is very helpful. I am still learning how this community uses certain vocabulary. If a child grows up speaking Esperanto, sure, it is a natural language, and also it is a conlang. I am trying to pin down the origin of the Hebrew Torah text, and yes, today Hebrew is a natural language, simply because it has native speakers.

Can you point me to the immediately antecedent languages to the Torah text, so I can look at how different (and similar) the grammar and vocabulary is? It would make sense to me if some of the Torah words are invented, and others borrow convenient sounds and strings from other languages, like I understand (with my limited understanding) Esperanto does.

Thanks again!
Marc
marcege
hieroglyphic
hieroglyphic
Posts: 45
Joined: 16 Jul 2018 22:19

Re: (changed forum) Is Hebrew a Conlang?

Post by marcege »

sangi39 wrote: 17 Jul 2018 03:07 First, on the note of "pangrams", it would seem reasonable to assume that any sufficiently large text will likely contain all of the letters within a language's alphabet at some at some point. Have you compared your premise that "faithful records written down of existing oral stories do not happen to include the development" of pangram sentences with other written versions of oral pieces, such as the Odyssey, Iliad, Serbian oral poetry, the Poetic Edda, etc.?

As to prefixes, what about Russian? At least /v/, /s/, /k/, /o/, and /u/, off the top of my head, are single phoneme prepositional morphemes, and then you can go even further with the "economy of affixes" in fusional languages where, in Spanish as an example, a single phoneme can represent the person, number and gender of the subject as well as the tense and aspect of the verb. I don't see how "economy of morphemes" can be an indicator of a language being a conlang.

On to the triconsonantal roots of the Hebrew language, this is a feature found in other related Semitic languages, and there are examples of roots in Hebrew which have as few as two and up to four or five "root" letters, which aren't borrowed from other languages (although I can't think of any off the top of my head). Proto-Indo-European is similar, for example, in having an abundance of roots which consist of a CVC syllable, with the vowel changes depending on the number and case of the noun or the tense of the verb, i.e. ablaut. The predominance of triconsonantal roots, at least in my opinion. Take a look at Old Chinese, which is reconstructed most often as being either entirely monosyllabic or sesquisyllabic at most, and at the very least transitioning into a wholly isolating morphology, a trait which is shared with modern languages like Khmer and Vietnamese. How is "Tất cả mọi người sinh ra đều được tự do và bình đẳng về nhân phẩm và quyền lợi" any more unnatural?

Numbers: I'd like to see your source for claiming that in Ancient Egyptian the order of the numbers doesn't matter. In every text I've seen, numbers are presented in decreasing orders of 10, with the only difference in order being the result of changes in the direction of writing across the entire text, e.g. in texts written left to right, higher powers of 10 appear on the left, lower ones on the right, and similarly in texts written right to left, higher powers of 10 appear instead on the right. I've also never heard of this claim for Hebrew except in the expressed instance that gematria is being used. Gematria is more properly a system of numerology, similar to Greek Isopsephy, Arabic abjad numerals, which are distinct from standard Arabic numerals. The assigning of numerical values to Hebrew letters is related to but distinct from this. As far as I know, while saying that גִּימֵל in (some forms of) Gematria is 3+10+40+30 = 83 is reasonable, it is not reasonable to say that this actually represents the number 83, which would instead be פג and never גפ (the same thing, as far as I understand it, is true in the Greek system, i.e. while you can assign a word a numerical value using a similar system, actual numbers are written systematically and uniquely).

37 and 73 are "star numbers" precisely because they form this pattern. However, given all of the verses throughout the entire Torah, are there any others? Are there any other verses which, through Gematria, have values of 1, 13, 37, 73, 121, 181, 253, 337, 433, 541, 661, etc., or are Gematria values more evenly distributed? Why is the tetragrammaton broken into two stars, presumably taken from a division of the value 26 divided into two 13s, which are star numbers, rather than a single value of 26? Are there any other geometrical patterns which appear. For example, 37 is also a centred hexagonal number, so are there any more of them? What about square numbers? Cube numbers? Triangle numbers? Do these appear relatively evenly, or does one pattern appear more often, and what might the chances of this be?

"It seems unlikely that anyone could master an existing natural language well enough to write a text with the level of wordplay and letterplay exhibited in Torah. Try in English to find multiple sentences and words that tell a story that also all exhibit the same geometric motif when letters are interpreted as numerals, and make sure at least one salient sentence (first, last, key concept) shows letter-level symmetry when curled into some shape". Yeah, you probably could. People have been trying this for centuries, the Torah included, even when written in other languages. Again, all you have to do is find the right word, phrase or sentence, and then find enough similar examples, and you're sorted. And given that different systems of numerology assign values differently, I wouldn't be surprised if this sort of thing appears everywhere in the Qur'an or the Odyssey. You have to show that your results are statistically significant and can't be interpreted any other way.

And that's about as far as I can go, I think. The idea that the author of the Torah is only one person really is up for debate, and the idea that there were multiple authors is the current consensus, especially amongst those who don't believe it was written by Moses (whether the direct word of God or not).

As for Hebrew in general being constructed, especially Biblical Hebrew, it really doesn't seem likely. The relationship it shares with languages like Akkadian, Arabic and Amharic are fairly well understood, as far as I can tell, and it does seem the Hebrew, like these other languages, derived from an ancestral Proto-Semitic language. Maybe it is possible that which ever person or people wrote up the final copy of the Torah threw in some interesting numerological nuggets, but to extend that to "Hebrew is a constructed language" is a bit of a stretch given evidence to the contrary.

You say, for example, that "it is hard to imagine how it could come to be that this is just a phonetic representation of speech that people were actually using", but then if that's the case, then Hebrew would have been constructed while the Torah was being written, and then an entire group of several thousand people would have had to have learned it in order to understand the Torah, rather than just, say, sticking with the religion and language they already had, and you're saying a single author managed to do this, when even in the modern day languages like Esperanto struggle to gain a foothold within any culture anywhere, let alone take it over?
Hi, thanks, these are all exactly the sort of questions I've been working through. Your feedback and thought is very much appreciated.

Just to touch on one of your (many) useful and thoughtful points. It's possible the Torah Hebrew did not take much of a foothold as a spoken language, and the text was used and interpreted by a smaller class of literate interpreters from its inception, like it was through the middle ages and up until the founding of modern Israel.

It would be hard to throw in a numerological nugget as a first sentence, if most of the words in that sentence would have needed to be constructed and then are also used frequently throughout the test of the text. Like, can you pick up a random book and add a numerological nugget to give it a new, meaningful start, and then over-write vocabulary in the rest of the book to make it consistent? Well, I guess someone could, but that would be an Interesting Thing and would represent a constructed (if small) aspect of the language.

I still don't understand how any language could have a non-normal distribution of root word lengths (i.e., 3-letter-roots) when words are spelled phonetically - I would suspect any language like that of having forced spoken words into a formula for some reason, and I might even think of that new, formulaic intentionally constrained vocabulary as a conlang.

Lots of other points to pick up - again, very much appreciated.

Marc
Last edited by marcege on 17 Jul 2018 04:55, edited 1 time in total.
User avatar
Ahzoh
mongolian
mongolian
Posts: 4191
Joined: 20 Oct 2013 02:57
Location: Canada

Re: (changed forum) Is Hebrew a Conlang?

Post by Ahzoh »

Can you point me to the immediately antecedent languages to the Torah text, so I can look at how different (and similar) the grammar and vocabulary is?
https://en.wiktionary.org/wiki/Appendix ... itic_stems
https://en.wikipedia.org/wiki/Proto-Semitic_language
https://en.wikipedia.org/wiki/Northwest ... nd_changes
I still don't understand how any language could have a non-normal distribution of root word lengths (i.e., 3-letter-roots) when words are spelled phonetically - I would suspect any language like that of having forced spoken words into a formula for some reason, and I might even think of that new, formulatic vocabulary as a conlang.
Triconsonsonantalism occurs through a complex series of processes including vowel mutation (e.g. i-mutation is when a vowel /i/ causes preceding OR following vowels [only one or the other] to become raised [a -> e, e -> i] or fronted [o -> e, u -> i]), vowel syncopation (when a vowel is deleted in a word) caused by syllable stress, and intense regularization/analogizing (e.g. verbs that conjugated irregularly now start conjugating regularly).
Alot of these processes are found in Proto-Indo-European and its descendants, though they never regularized to the degree that Proto-Semitic and its descendants did.
Last edited by Ahzoh on 17 Jul 2018 04:43, edited 1 time in total.
Image Śād Warḫallun (Vrkhazhian) [ WIKI | CWS ]
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: (changed forum) Is Hebrew a Conlang?

Post by sangi39 »

marcege wrote: 17 Jul 2018 03:44 Just to touch on one of your (many) useful and thoughtful points. It's possible the Torah Hebrew did not take much of a foothold as a spoken language, and the text was used and interpreted by a smaller class of literate interpreters from its inception, like it was through the middle ages and up until the founding of modern Israel.
Well, you'd have to have evidence to back up the idea that the early Jewish peoples spoke something else at the time that Biblical Hebrew was being written and before the Babylonian exile when Aramaic became the vernacular (since Judeo-Aramaic, if I remember correctly, is marked by influence from Hebrew in contrast to earlier stages of Aramaic), the most likely candidates being Phoenician, Ammonite, Moabite, etc.


marcege wrote: 17 Jul 2018 03:44 It would be hard to throw in a numerological nugget as a first sentence, if most of the words in that sentence would have needed to be constructed are then are also used frequently throughout the test of the text. Like, can you pick up a random book and add a numerological nugget to give it a new, meaningful start, and then over-write vocabulary in the rest of the book to make it consistent? Well, I guess someone could, but that would be an Interesting Thing.
Well, just as a random example, let's try Hamlet. If I take the first three spoken lines of Hamlet, assigning each letter a value between 1 and 26, take the value of each line and then add up the sum of their divisors, the first three lines return an octagon, a pentagon, and a square respectively. Bernardo, who speaks the first line, also leads to an octagon following the same process, as does Polonius (who is the seventh character to speak in the play, the divisor-sum number of his first line being, 756, which contains 7 lots of the number 108, the sum of its divisors also being an octagonal number).

The name of the play “The Tragedy of Hamlet: Prince of Denmark”, put through the same process, also gives square, like the third line, which is “Long live the King!”, the main plot of the play being that one king has been killed, the next is eventually killed as well, with the next dying soon after. The name Hamlet itself gives 60 following this process while “Prince of Denmark” yields, 300, one “Hamlet” for each of the five scenes of the play, failing to be a square (360) because the "King is Dead!" at the end of scene 5, instead of going on the rule in a 6th scene.

You don't have to overwrite the vocabulary in the rest of the book to find this sort of thing, you just need to find the correct examples of patterns, and in sufficiently large bodies of text you will find those examples.

marcege wrote: 17 Jul 2018 03:44 I still don't understand how any language could have a non-normal distribution of root word lengths (i.e., 3-letter-roots) when words are spelled phonetically - I would suspect any language like that of having forced spoken words into a formula for some reason, and I might even think of that new, formulatic vocabulary as a conlang.
You haven't demonstrated how the triconsonantal root system of Hebrew, though, is non-normal, given that similar systems exist in related languages like Arabic and Amharic, and what would make those systems non-normal in the first place, e.g. why we should consider it less normal that the ablaut system of Indo-European languages or the regularity of purely monosyllabic root morphemes in South-East Asian languages. You also haven't demonstrated why spelling them phonetically would make a difference to them being normal or not.
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
marcege
hieroglyphic
hieroglyphic
Posts: 45
Joined: 16 Jul 2018 22:19

Re: (changed forum) Is Hebrew a Conlang?

Post by marcege »

Ahzoh wrote: 17 Jul 2018 04:41
Can you point me to the immediately antecedent languages to the Torah text, so I can look at how different (and similar) the grammar and vocabulary is?
https://en.wiktionary.org/wiki/Appendix ... itic_stems
https://en.wikipedia.org/wiki/Proto-Semitic_language
https://en.wikipedia.org/wiki/Northwest ... nd_changes
Thanks. From what you sent it says that Proto-Semitic is "hypothetical," so it sounds like it does not provide a clear way to check the size of any spelling or grammar gaps between Torah Hebrew and antecedent languages.
I still don't understand how any language could have a non-normal distribution of root word lengths (i.e., 3-letter-roots) when words are spelled phonetically - I would suspect any language like that of having forced spoken words into a formula for some reason, and I might even think of that new, formulatic vocabulary as a conlang.
Triconsonsonantalism occurs through a complex series of processes including vowel mutation (e.g. i-mutation is when a vowel /i/ causes preceding OR following vowels [only one or the other] to become raised [a -> e, e -> i] or fronted [o -> e, u -> i]), vowel syncopation (when a vowel is deleted in a word) caused by syllable stress, and intense regularization/analogizing (e.g. verbs that conjugated irregularly now start conjugating regularly).
Alot of these processes are found in Proto-Indo-European and its descendants, though they never regularized to the degree that Proto-Semitic and its descendants did.
[/quote]

Interesting. How is it that the processes you describe lead to words with three consonants, rather than, say, 2, or 4, or a range?

thanks,
-Marc
marcege
hieroglyphic
hieroglyphic
Posts: 45
Joined: 16 Jul 2018 22:19

Re: (changed forum) Is Hebrew a Conlang?

Post by marcege »

sangi39 wrote: 17 Jul 2018 04:42
marcege wrote: 17 Jul 2018 03:44 Just to touch on one of your (many) useful and thoughtful points. It's possible the Torah Hebrew did not take much of a foothold as a spoken language, and the text was used and interpreted by a smaller class of literate interpreters from its inception, like it was through the middle ages and up until the founding of modern Israel.
Well, you'd have to have evidence to back up the idea that the early Jewish peoples spoke something else at the time that Biblical Hebrew was being written and before the Babylonian exile when Aramaic became the vernacular (since Judeo-Aramaic, if I remember correctly, is marked by influence from Hebrew in contrast to earlier stages of Aramaic), the most likely candidates being Phoenician, Ammonite, Moabite, etc.
Sangi39, this is to say that even if Torah Hebrew is a conlang, it does not need to immediately become the lingua franca of a society - this is to consider the analogy you brought up regarding Esperanto in which the constructed language can exist within a society, only exerting as much influence on the rest of the spoken language as it happens to have. Not everyone would have to immediately speak the text language. You bring up a good point, if Torah Hebrew is constructed, then there may be evidence of some other language in use as well, besides Torah Hebrew. My limited knowledge about this is that linguistic artifacts from the time periods of the origin of Torah are sparse, and also that some do include other languages like the ones you mention.
sangi39 wrote: 17 Jul 2018 04:42
marcege wrote: 17 Jul 2018 03:44 It would be hard to throw in a numerological nugget as a first sentence, if most of the words in that sentence would have needed to be constructed are then are also used frequently throughout the test of the text. Like, can you pick up a random book and add a numerological nugget to give it a new, meaningful start, and then over-write vocabulary in the rest of the book to make it consistent? Well, I guess someone could, but that would be an Interesting Thing.
Well, just as a random example, let's try Hamlet. If I take the first three spoken lines of Hamlet, assigning each letter a value between 1 and 26, take the value of each line and then add up the sum of their divisors, the first three lines return an octagon, a pentagon, and a square respectively. Bernardo, who speaks the first line, also leads to an octagon following the same process, as does Polonius (who is the seventh character to speak in the play, the divisor-sum number of his first line being, 756, which contains 7 lots of the number 108, the sum of its divisors also being an octagonal number).

The name of the play “The Tragedy of Hamlet: Prince of Denmark”, put through the same process, also gives square, like the third line, which is “Long live the King!”, the main plot of the play being that one king has been killed, the next is eventually killed as well, with the next dying soon after. The name Hamlet itself gives 60 following this process while “Prince of Denmark” yields, 300, one “Hamlet” for each of the five scenes of the play, failing to be a square (360) because the "King is Dead!" at the end of scene 5, instead of going on the rule in a 6th scene.

You don't have to overwrite the vocabulary in the rest of the book to find this sort of thing, you just need to find the correct examples of patterns, and in sufficiently large bodies of text you will find those examples.
Interesting. Shakespeare is famous for wordplay and inventing words and phrases, I wouldn't necessarily put something like this past him. There are 2 issues here. 1. One is about intentionality. Detecting whether the pattern is intentional is partly a matter of how overwhelmingly improbable the pattern is. My take is that the Hamlet pattern is the sort one might find in any text - but feel free to push back here - but the fact that Psalm 34 is an alphabetical acrostic is beyond uncanny chance and was probably deliberate.

Here is a piece on acrostics in Shakespeare that tries to get at their intentionality using statistics. Unfortunately, none of Shakespeare's acrostics are as long or clear as an entire alphabet, and the answers get a bit murky.
https://digitalcommons.butler.edu/cgi/v ... t=wordways

2. The 2nd issue is about how to go about making a pattern intentionally. You had brought up adding numerological nuggets to an existing text and I said this would be hard to do by way of adding a new introduction with new words to an existing text. In the case of Hamlet, adding the patterns you mention after the fact would involve taking an existing play and changing the title, rewriting a few carefully chosen lines of script, and choosing the name "Hamlet" because its numerology works so nicely. In this case, you would also have to edit the name of the main character to be "Hamlet" throughout the rest of the play - that part is what I referred to as "overwriting." If the pattern you brought up were, say, 100 times more unlikely-sounding, or even 10-times more unlikely-sounding, I might start to wonder if the numerology were part of how Shakespeare chose the name "Hamlet."

sangi39 wrote: 17 Jul 2018 04:42
marcege wrote: 17 Jul 2018 03:44 I still don't understand how any language could have a non-normal distribution of root word lengths (i.e., 3-letter-roots) when words are spelled phonetically - I would suspect any language like that of having forced spoken words into a formula for some reason, and I might even think of that new, formulatic vocabulary as a conlang.
You haven't demonstrated how the triconsonantal root system of Hebrew, though, is non-normal, given that similar systems exist in related languages like Arabic and Amharic, and what would make those systems non-normal in the first place, e.g. why we should consider it less normal that the ablaut system of Indo-European languages or the regularity of purely monosyllabic root morphemes in South-East Asian languages. You also haven't demonstrated why spelling them phonetically would make a difference to them being normal or not.
Hi, when I said "normal," distribution I meant in terms of statistics. Here's another way to say this, and hopefully you or someone will understand why I have the question that I have. When people talk to each other, out loud, over generations, the sounds people use change - accents develop, sounds get dropped off of words, new sounds get added, etc. If you take a language that people have been speaking out loud to each other for generations and then go to write it down phonetically, some words might need one letter to represent the sound (like the English word "I,") other words will need 2 letters, 3 letters, 4 letters, 5 letter, 6 letters, 7 letters, 8 letters, etc. etc. The same is true for the number of consonants, or the number of vowels. It's just however many it takes,

Here is what I am hoping to learn, maybe in the context of conlang. How can it be the case that a language root system would have a strong, even if not exclusive, predominance of words with "3" letters in particular. This sounds like someone systematizing an unruly system. When speaking out loud and learning new language from parents and friends, over generations, people - I don't think, correct me if this is where I'm wrong - don't naturally keep words to a regularized number of sounds. Therefore, if a story is represented using words with a highly regularized number of letters, the story is not straightforwardly a phonetic record of preexisting oral stories. Does at least my question make sense?


Thanks again!
-Marc
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: (changed forum) Is Hebrew a Conlang?

Post by Salmoneus »

It's worth pointing out that it's commonplace - perhaps even the default - for languages to have relatively fixed word-shapes. CVCV is the most common, but I've seen CVCVC - isn't the basic Austronesian wordshape CVCVC? And some SE asian families tend toward C.CVC. The only odd thing about Semitic is that there's pervasive ablaut.

And is that weirder than Proto-Indo-European's obsessive stop series constraints? Other than a few weird former compounds and loanwords, PIE only allows roots of the form C(R)V(R)C (and that 'V' is almost always /e/), and the three series of stops have to follow strict rules: the D series may be paired with either of the other two series, but never with itself; the Dh and T series can each be paired with the D series, or with themselves, but never with the other series. Does that look more natural than "most words have three consonants in them"!?


(and yes, Proto-Semitic is 'theoretical'. Gravity is also 'theoretical'.)
User avatar
Ahzoh
mongolian
mongolian
Posts: 4191
Joined: 20 Oct 2013 02:57
Location: Canada

Re: (changed forum) Is Hebrew a Conlang?

Post by Ahzoh »

Thanks. From what you sent it says that Proto-Semitic is "hypothetical," so it sounds like it does not provide a clear way to check the size of any spelling or grammar gaps between Torah Hebrew and antecedent languages.
It's a reconstruction based on comparing words and grammar between the Semitic languages and so there are a set of sound and grammar changes showing how each descendant language might have diverged from it. You could compare Biblical Hebrew to any Semitic language and see the striking similarities in vocabulary and grammar (such as their verb prefixes and object pronoun clitics) with even closer resemblances to other Northwest Semitic languages like Aramaic and Ugaritic.

The roots of words in Semitic languages are well preserved enough you could even trace them back to the (also reconstructed) Proto-Afro-Asiatic language, which ties Semitic with 5 other language superfamilies in Africa and Asia.
Interesting. How is it that the processes you describe lead to words with three consonants, rather than, say, 2, or 4, or a range?
It doesn't but it just so happens that the ancestor of Proto-Semitic possibly had large numbers of words with only two syllables, which isn't at all uncommon in world languages, with some of them possibly having approximant consonants (like yod, waw, he, and ayin) added to them to fit the paradigms if they ended in a vowel. There could also have been roots derived from two-consonant roots that had derivational affixes added to them.
Image Śād Warḫallun (Vrkhazhian) [ WIKI | CWS ]
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: (changed forum) Is Hebrew a Conlang?

Post by Salmoneus »

In terms of why this would happen: because words are not random!

More specifically, the evolution of words is not purely random, but is subject to environmental constraints: like animals, words evolve to suit their habitat. People want words - particularly ones they use frequently - to be easy to say. Words are continually evolving to be 'easier' to say, both in absolute terms (common sound changes like lenition, haplology, assimilation, etc) and relative to the specific language (if you're used to having words with final consonants, it's not a big problem if a word has a final consonant, but if hardly any of your words have final consonants, the fact that a few of them do may be a 'difficulty' for you). This happens through regular processes throughout the language, which causes convergence.

For instance, say a language has CVCV, CVCVC, CVCVCV, CCVCV, CCVCVC, and CCVCVCV words. Quite a variety. Now let's have two universal rules throughout the language: final vowels are lost, and, because the second syllable is stressed, the first vowel is lost in words with more than one syllable. Now suddenly those wordshapes look like: CVC, CCVC, CCVC, CCVC, CCCVC, CCCVC. A third rule: in clusters of three consonant, the middle consonant is lost (because CCC clusters are hard to say). So now five of those six wordforms have a CCVC shape, and only one has a CVC shape. What's more, if there are three vowels, and the high vowels break when stressed (/i/ > /ji/, /u/ > /wu/), and then glides are deleted after clusters, two-thirds of those CVC words have also ended up as CCVC.

It also, of course, happens in an ad hoc manner, particularly when a long, unwieldy word becomes important. The word may, through overuse and awkward shape, have irregular or fast-forwarded soundchanges applied (the way a lot of nautical terminology has experienced: 'boatswain' becoming 'bosun', for instance). It may be abbreviated. Or it may be replaced by an easier synonym (creating one, if necessary, by repurposing an existing word). English is generally quite tolerant of long and/or unusual words. But consider modern forms of travel, for example. I no longer travel by locomotive, by automobile, by aeroplane, by omnibus, by electric street railway, by motor wagon, by velocipede, or by taximeter cabriolet; instead, I go by train (synonym), car (synonym), plane (abbreviation), bus (abbreviation), tram (synonym), lorry (synonym), bike (synonym, then irregular soundchange and/or abbreviation, no-one really knows what happened there), or either taxi or cab (both abbreviations, with sometimes different meanings).
marcege
hieroglyphic
hieroglyphic
Posts: 45
Joined: 16 Jul 2018 22:19

Re: (changed forum) Is Hebrew a Conlang?

Post by marcege »

Salmoneus wrote: 17 Jul 2018 14:24 In terms of why this would happen: because words are not random!

More specifically, the evolution of words is not purely random, but is subject to environmental constraints: like animals, words evolve to suit their habitat. People want words - particularly ones they use frequently - to be easy to say. Words are continually evolving to be 'easier' to say, both in absolute terms (common sound changes like lenition, haplology, assimilation, etc) and relative to the specific language (if you're used to having words with final consonants, it's not a big problem if a word has a final consonant, but if hardly any of your words have final consonants, the fact that a few of them do may be a 'difficulty' for you). This happens through regular processes throughout the language, which causes convergence.

For instance, say a language has CVCV, CVCVC, CVCVCV, CCVCV, CCVCVC, and CCVCVCV words. Quite a variety. Now let's have two universal rules throughout the language: final vowels are lost, and, because the second syllable is stressed, the first vowel is lost in words with more than one syllable. Now suddenly those wordshapes look like: CVC, CCVC, CCVC, CCVC, CCCVC, CCCVC. A third rule: in clusters of three consonant, the middle consonant is lost (because CCC clusters are hard to say). So now five of those six wordforms have a CCVC shape, and only one has a CVC shape. What's more, if there are three vowels, and the high vowels break when stressed (/i/ > /ji/, /u/ > /wu/), and then glides are deleted after clusters, two-thirds of those CVC words have also ended up as CCVC.

It also, of course, happens in an ad hoc manner, particularly when a long, unwieldy word becomes important. The word may, through overuse and awkward shape, have irregular or fast-forwarded soundchanges applied (the way a lot of nautical terminology has experienced: 'boatswain' becoming 'bosun', for instance). It may be abbreviated. Or it may be replaced by an easier synonym (creating one, if necessary, by repurposing an existing word). English is generally quite tolerant of long and/or unusual words. But consider modern forms of travel, for example. I no longer travel by locomotive, by automobile, by aeroplane, by omnibus, by electric street railway, by motor wagon, by velocipede, or by taximeter cabriolet; instead, I go by train (synonym), car (synonym), plane (abbreviation), bus (abbreviation), tram (synonym), lorry (synonym), bike (synonym, then irregular soundchange and/or abbreviation, no-one really knows what happened there), or either taxi or cab (both abbreviations, with sometimes different meanings).
Thanks. That's a clear explanation of how words might lose variety and converge towards similar forms. Does it also explain why single-consonant words would expand upwards to the identical form that longer words have expanded downwards towards? Would it be expected that all verbs, say, naturally converge to 3 letters in the root? For example, there is still variability in your list: car and cab each need 2 consonants to spell, plane and train need three, and kids these days definitely get around by scooter.

Do 3-letter-roots have too little variability to have been capturing what people who talk actually say to each other? Maybe I'm too stuck in English to have a feel for other options with less variability, but for me it's still feeling like far too little variability for expressing how humans use oral speech.

thanks again, still thinking!
Marc
User avatar
Ahzoh
mongolian
mongolian
Posts: 4191
Joined: 20 Oct 2013 02:57
Location: Canada

Re: (changed forum) Is Hebrew a Conlang?

Post by Ahzoh »

marcege wrote: 17 Jul 2018 16:02Does it also explain why single-consonant words would expand upwards to the identical form that longer words have expanded downwards towards?
Vowel breaking might, or compounding with other words.

Would it be expected that all verbs, say, naturally converge to 3 letters in the root?
No, they would not be all expected to become triconsonantal. Also it's important to understand that the idea of "triconstonal roots" is an abstraction and not a concrete thing. Speakers don't think of putting abstract collections of letters into patterns in the same way you don't think of the difference between foot and feet as you putting the root f-t into two separate patterns. If anything, you probably think of foot as the base form with feet as the changed form. It's the same way in triconsonantal languages.
Do 3-letter-roots have too little variability to have been capturing what people who talk actually say to each other? Maybe I'm too stuck in English to have a feel for other options with less variability, but for me it's still feeling like far too little variability for expressing how humans use oral speech.
Given a language like Arabic, which has 28 consonants, even if every word had roots consisting of 3 letters, you'd still have about 21,952 possible roots (28 x 28 x 28). But Arabic will never use this many. In fact, it only needs to use about 2000 roots and everything else taken care of through verb conjugation or noun declension. Most of those might be triconsonantal, large numbers of others might be biconsonantal, qudraconsonantal, or even quinticonsonantal.
Image Śād Warḫallun (Vrkhazhian) [ WIKI | CWS ]
marcege
hieroglyphic
hieroglyphic
Posts: 45
Joined: 16 Jul 2018 22:19

Re: (changed forum) Is Hebrew a Conlang?

Post by marcege »

Ahzoh wrote: 17 Jul 2018 18:55
marcege wrote: 17 Jul 2018 16:02
Would it be expected that all verbs, say, naturally converge to 3 letters in the root?
No, they would not be all expected to become triconsonantal. Also it's important to understand that the idea of "triconstonal roots" is an abstraction and not a concrete thing. Speakers don't think of putting abstract collections of letters into patterns in the same way you don't think of the difference between foot and feet as you putting the root f-t into two separate patterns. If anything, you probably think of foot as the base form with feet as the changed form. It's the same way in triconsonantal languages.
Ahzoh, thanks. If so, can we make the inference that if a text is based on abstract collections of letters that are organized into patterns, then the text does not represent speakers' speech?
Ahzoh wrote: 17 Jul 2018 18:55
Do 3-letter-roots have too little variability to have been capturing what people who talk actually say to each other? Maybe I'm too stuck in English to have a feel for other options with less variability, but for me it's still feeling like far too little variability for expressing how humans use oral speech.
Given a language like Arabic, which has 28 consonants, even if every word had roots consisting of 3 letters, you'd still have about 21,952 possible roots (28 x 28 x 28). But Arabic will never use this many. In fact, it only needs to use about 2000 roots and everything else taken care of through verb conjugation or noun declension. Most of those might be triconsonantal, large numbers of others might be biconsonantal, qudraconsonantal, or even quinticonsonantal.
Ahzoh, that makes sense, and I am definitely learning. Your comments are most appreciated.

If people speak a language based on mostly triconsonantal roots, you are saying we would also expect a fair number of biconsonantal and quadraconsonantal roots to appear in the same language. People drop sounds sometimes in dialect and in the development of language, add sounds, etc., and there is no constraint in everyday language to stick to words that are "3" of anything in particular.

I'm struggling to understand how the list of verbs at the following link, which I think is a word count of words in the Torah, is so regular with respect to how many letters they use. That is to say, regular compared to my intuitions that roots, or words, or really any linguistic construct that is there to express the phonemes that people speak will need to be fairly variable, simply because people's speech is naturally somewhat variable. The list of verbs is also strangely regular compared to the list of nouns that follows.

http://www.rabbihorowitz.com/PYes/BB_Fi ... 10Page.pdf

Should we assume there is some error, and 2-letter and 4-letter terms are missing from the list of verbs?

Thanks again for your thoughts,
-marc
User avatar
Shemtov
runic
runic
Posts: 3283
Joined: 29 Apr 2013 04:06

Re: (changed forum) Is Hebrew a Conlang?

Post by Shemtov »

I'll avoid my own view as much as possible, but I think this raises three interesting questions:
1. To Religious Jews and some Christian Denoms, who believe that Hebrew is Adamic, can Religious Studies Academia and Anthropologists analyze their believe as that Bib. Hebrew is a Conlang and G-d is a conlanger? The same applies to Muslims and Quranic Arabic.
2. Is a literary language, confined to a small corpus, (The 24 of the Jewish Tanakh divided into 39 by the Protestent OT, plus the fragments of Sirach) have aspects of a conlang? Do we want to include Rabbinic Hebrew (Mishnaic and Post-Mishnaic) in the corpus and does that change the answer to this question?
3. Talking of Rabbinic Hebrew, are the forms of a language no longer spoken but still generating texts considered to have aspects of a Conlang? The same applies to the Scholarly Latin of the Middle Ages/Renaissance, Modern Vatican Latin, and possibly Al-FusHa.,
Many children make up, or begin to make up, imaginary languages. I have been at it since I could write.
-JRR Tolkien
Keenir
mayan
mayan
Posts: 2399
Joined: 22 May 2012 03:05

Re: (changed forum) Is Hebrew a Conlang?

Post by Keenir »

marcege wrote: 17 Jul 2018 03:44It would be hard to throw in a numerological nugget as a first sentence, if most of the words in that sentence would have needed to be constructed and then are also used frequently throughout the test of the text. Like, can you pick up a random book and add a numerological nugget to give it a new, meaningful start, and then over-write vocabulary in the rest of the book to make it consistent?
You're making an assumption:

a. that either all written languages have numerology, or only Biblical Hebrew has numerology

b. that all interpretations of a text are of equal age. Aside from being a natural history observation and an alphabet aide, what signifigance is "The quick brown fox jumped over the lazy dogs" ?
I still don't understand how any language could have a non-normal distribution of root word lengths (i.e., 3-letter-roots) when words are spelled phonetically -
define "normal".
marcege wrote: 17 Jul 2018 16:02Thanks. That's a clear explanation of how words might lose variety and converge towards similar forms. Does it also explain why single-consonant words would expand upwards to the identical form that longer words have expanded downwards towards? Would it be expected that all verbs, say, naturally converge to 3 letters in the root? For example, there is still variability in your list: car and cab each need 2 consonants to spell, plane and train need three, and kids these days definitely get around by scooter.
scooter is 3 consonants.

bicycle -> bike -> biking ...etc. people add to words the things they want to say.
Do 3-letter-roots have too little variability to have been capturing what people who talk actually say to each other?
wait...you were saying that Biblical Hebrew is a conlang that took over the language of a faith for over a millennia...and now you're saying you're not sure if such languages as Hebrew have enough variability to help people talk to each other? (you are aware that there are conversations within the Torah and Tanakh, right?)
marcege wrote: 17 Jul 2018 20:29If people speak a language based on mostly triconsonantal roots, you are saying we would also expect a fair number of biconsonantal and quadraconsonantal roots to appear in the same language. People drop sounds sometimes in dialect and in the development of language, add sounds, etc., and there is no constraint in everyday language to stick to words that are "3" of anything in particular.
not everyone writes casual speech...particularly when writing something important: even people who speak somethin' likes dis, will often write properly and counter to how they speak.
Last edited by Keenir on 18 Jul 2018 19:39, edited 1 time in total.
At work on Apaan: viewtopic.php?f=6&t=4799
User avatar
Shemtov
runic
runic
Posts: 3283
Joined: 29 Apr 2013 04:06

Re: (changed forum) Is Hebrew a Conlang?

Post by Shemtov »

Keenir wrote: 18 Jul 2018 19:24 the Torah and Tanakh

Sorry for being pedantic, but that phrase is redundant- the word "Tanakh" is an acronym, the Ta part standing for "Torah". Separating the Torah from the rest of the Tanakh, we are left with what is called "Nakh" but as most people here probably don't know the word, "Torah and the rest of Tanakh" would be more accurate.
Many children make up, or begin to make up, imaginary languages. I have been at it since I could write.
-JRR Tolkien
Locked