Panini, Sanskrit Grammar and Linguistic

[Astronuc]

I found this of great interest. I stumbled across Panini's name while looking for the food of the same (more or less) name.

Pānini (पाणिनि) was an ancient Indian grammarian (c. 520–460 BC, but estimates range from the 7th to 4th centuries BC) who lived in Gandhara and is most famous for his grammar of Sanskrit, particularly for his formulation of the 3,959 rules of Sanskrit morphology.

Pānini's grammar of Sanskrit is highly systematised and technical. Inherent in its analytic approach are the concepts of the phoneme, the morpheme and the root, only recognized by Western linguists some two millennia later.

http://en.wikipedia.org/wiki/Panini

 

[neutrino]

My Sanskrit professor in college would jokingly say that certain things are arranged like in a matrix (he was a maths major as an undergrad). Of course, being a physics student, I didn't study anything beyond the basics.

If I may ask, what's the food that sounds like Panini? :)

 

[Astronuc]

Panini is the food spelled the same as Panini the Sanskrit grammarian. Well they are similar in ASCII text.

The food panini (plural of panino) is a sandwich - http://en.wikipedia.org/wiki/Panini_(sandwich)

 

[honestrosewater]

Inherent in its analytic approach are the concepts of the phoneme, the morpheme and the root, only recognized by Western linguists some two millennia later.

Perhaps they mean that linguists didn't recognize for millennia that those concepts were included in his work. Otherwise, that claim strikes me as highly suspect. Below is a little ongoing list that I keep (poorly) of some major linguists and such. According to it, the Alexandrian Dionysius Thrax developed a morphology in the 2nd century BC, and I can't imagine something being called a morphology if it didn't include at least the concepts of morpheme and root. (Plus, Plato, Aristotle, and the Stoics were sharp blokes, and I'm hesitant to believe that such basic concepts escaped them (and all of the ancient Greeks and Romans who studied language?), even if their language studies were more logic-and-rhetoric-oriented.) I could look into it if anyone is interested.

Another cool linguistics history tidbit: an Egyptian papyrus dated ca. 1700 BC contains medical descriptions of language disorders following brain injury.

___
Panini (Indian; 5th cent BC; Sanskrit grammar)
Plato
Aristotle
Stoics
Dionysius Thrax (Alexandrian; 2nd cent BC; first morphology)
Apollonius Dyscolus (Greek; 2nd cent AD; first syntax)
Priscian (Latin; 6th cent AD; grammarian)
Petrus Ramus (French; 16th cent; logician)
Jacob Grimm (German; 18-19th cent; historical linguistics, sound change)
Wilhelm von Humboldt (Prussian; 19th cent; Langauge as activity, expression of individual and culture, inner and outer form, language is not utterances but rule for generating them (unlimitedly))
Ferdinand de Saussure (Swiss; 19th cent; 'father' of structuralism?)
Otto Jespersen (Danish; 19-20th cent; )
Franz Boas (German/American; 20th cent; anthropologist/linguist, focus on how to describe Native American languages)
Edward Sapir (American; 20th cent; student of Boas; )
Leonard Bloomfield (American; 20th cent; behavioristic approach)
Benjamin Lee Whorf (American; 20th cent; student of Sapir; Sapir-Whorf hypothesis)
Zellig S. Harris (Russian/American; 20th cent;
Kenneth L Pike (American; 20th cent; tagmemics)
Sydney M Lamb (American; 20th cent; stratificational grammar)
Noam Chomsky (American; 20-21st cent; student of Harris; transformational grammar)

Roman Jakobson
Morris Halle
Gunnar Fant
Daniel Jones
Henry Sweet

Frege, Tarski, & co.

 

[Astronuc]

Inherent in its analytic approach are the concepts of the phoneme, the morpheme and the root, only recognized by Western linguists some two millennia later.

I wondered about this statement myself. The Wikipedia article seems to praise Panini, so I was wondering how objective it is.

On the other hand, 2 millenia from 500 BC is about 1500 CE, and so maybe the statement is valid. In the list given by HRW, there is a big gap between Priscian (Latin; 6th cent AD; grammarian) and Petrus Ramus (French; 16th cent; logician), the later coming along about 2 millenia after Panini. But were the concepts of "phoneme, the morpheme and the root" only realized after the 1500's in Europe?

And the more I think about it, what have we lost along the way from invasions and migrations? What has been lost in the way of great literature and thought? I wonder - what was lost in the burning of the great library in Alexandria?

 

[selfAdjoint]

A really fine book on the languages of the world is Empires of the Word, by Nicholas Ostler. It discusses Panini's work and its role in the subsequent language history of the Indian subcontinent and Southeast Asia. Panini did not just do a grammar in the sense of Priscian or what you get in modern language classes. He analytically codified just about every aspect of spoken communication, tone, gestures, rhetorical devices, everything, so that people of future ages could reconstruct the correct recitation of the Vedic scriptures even if the living tradition of master to student might have lapsed. I'll bet that somewhere in Panini is something equivalent not only to matrices but to regular expressions.

 

[honestrosewater]

On the other hand, 2 millenia from 500 BC is about 1500 CE, and so maybe the statement is valid. In the list given by HRW, there is a big gap between Priscian (Latin; 6th cent AD; grammarian) and Petrus Ramus (French; 16th cent; logician), the later coming along about 2 millenia after Panini.

Well, that list isn't complete in any way, but, yeah, that is a big gap. It does correspond roughly to the Middle ('Dark') Ages (so-called for a reason), so a dearth is perhaps to be expected. You might find some work on language having been done by the so-called medieval philosophers. They did work on logic during that time.

But were the concepts of "phoneme, the morpheme and the root" only realized after the 1500's in Europe?

The phonemes of a language are roughly just the 'sounds' of the language, e.g., /b, p, t, d, f, v, s, z/ are some of English's phonemes. The concept of phoneme is contrasted with the concept of phone. Phones are basic speech sounds (i.e., speech as acoustic signals), while phonemes are basic speech sounds as used in a particular language. The two concepts are useful because not only do different languages use different phones, they can also use the same phones in different ways. For example, English doesn't make a phonemic distinction between the sounds underlined in (1).

(1) a. bat [bæt]
(1) b. batter [bæɾɹ̩]

In English, the phones [t] and [ɾ] are variants of the same phoneme, /t/. The conditions under which, or environments in which, these phonetic variants occur is predictable, with /t/ occurring as the plosive [t] in environments like (1a) and as the flap [ɾ] in environments like (1b). Since these two phones don't occur in the same environments, English can't use them to distinguish between words. (I don't mean to imply a one-way cause-effect relationship there though.) However, in Spanish, [t] and [ɾ] do occur in the same environments, as in (2). (I'm stealing this example from a book -- I don't kow how to spell (2a).)

(2) a. ?? [pita] meaning 'century plant'
(2) b. pira [piɾa] 'funeral pyre'

In Spanish, [t] and [ɾ] are used to distnguish between words and are each phonemes, /t/ and /ɾ/.

As another quick example, Spanish doesn't distinguish between English's dare and there.

A morpheme is basically just an atomic unit of meaning, and a root is the 'base word' to which you add other pieces or make changes. For example, in

(3) unwillingly = un-will-ing-ly

each of those segments represents a morpheme. Will is the root -- it is the only morpheme that can show up on its own, so we say that will is a free morpheme, while the other 3 are bound.

Anywho, they are basic concepts that I imagine are likely to fall out pretty early on in a close examination of language, though the phone vs. phoneme distiction might not be as obvious (it might take a more scientific, cross-linguistic study to notice -- though spelling vs. pronunciation might bring it out).

I wonder - what was lost in the burning of the great library in Alexandria?

Yep, very sad.

 

[honestrosewater]

I'll bet that somewhere in Panini is something equivalent not only to matrices but to regular expressions.

I don't know about his work yet, but this is from the Wikipedia article:

Pāṇini uses metarules, transformations, and recursions with such sophistication that his grammar has the computing power equivalent to a Turing machine. In this sense Panini may be considered the father of computing machines. His work was the forerunner to modern formal language theory, and a precursor to computing. Paninian grammars have also been devised for non-Sanskrit languages. The Backus-Naur form (Panini-Backus form) or BNF grammars used to describe modern programming languages have significant similarities to Pāṇini's grammar rules.

I'm growing into view that speakers use natural languages to model their worlds, which include both physical and abstract objects, in a way similar to how mathematicians and physicists use formal languages and mathematical structures to model their worlds. The latter group and their ilk are just more organized and careful -- and restrict themselves to relatively well-behaved worlds. So finding familiar mathematical structures in natural languages wouldn't be surprising. Formal languages are derived from natural languages by humans, after all -- and even if that didn't come into play, I suppose it would be rather like, say, the same types of visual systems evolving repeatedly and separately. I'm still growing into this though. I wonder how useful (or accurate) it is to view math and science as just refining what already exists in/as natural language.