Complete Pali-English Dictionary for Development of Interlinear Suttas

Dear @ElissaJ,

I don’t know your current status regarding this topic, but at one point the text linked below changed my perspective on learning Pali.

http://pali.nibbanam.com/immerse.htm

Metta and Smiles,
Mirco

1 Like

Thanks, I agree completely, you have to learn a language intensively otherwise you just waste time.

1 Like

hi adjahn sujato…i have a background in linguistics and recently greek during a christian phase. go to bswa in perth. surely given the script is phonological (far less messed up than english it would not be incredibly difficult with good coding (beyond me (i just know a bit about what is possible) but not beyond a jedi nlp coder…). to tag and parse the text automatically. made possible by phonological text. i know u have been to stanford but i dont know if u have talked to these guys.

professor christopher manning is their adjahn brahm…

they have all the software in python… and its open source

u could link it up with the roots in the pali society dictionary or any other after tagging automatically. pali dictionary seems very out of date but at least a good start. (i had a poke around in that)…

in recovery and newly sober atm but looking forward to helping out if i can get my s… together…looks awesome.

going from pali to english to another language is never going to work as well as adding in a good interlinear. i know u r working on some software to help other translators…may have already done it.

sorry about previous comment if wrong…but doesnt that the pali script is phonolgical mean its easier than english to automatically tag and parse (working out the right sense/phrase etc an entirely different matter and not possible atm…) english is a nightmare bcos not phonolgical like pali or italian.

these guys r the big cohunas… professor manning wrote one of the first statistical nlp texts.

python software is there…perhaps u would do a proportion then bootstrap… i

anyway as i said the stanford nlp group would b the people to ask.

Hey Tim,

Yes, it’s phonetic (like all Indic and south Asian languages), which solves a range of problems, but then there are a range of other problems. Pali has a rich and complex morphology, which means that the same word can appear in vastly different forms. Think, “Did he go?” “Yes he already went.” But on an industrial scale.

Then there is the problem that the corpus is relatively small compared to English, with fairly poor resources in terms of adequate dictionaries and the like.

Nonetheless, we are proceeding with this issue using ML approaches. A number of folks are looking into it and we should have some initial results soon. I am sure traditional NLP can also be useful, but the thing is simply finding someone to do the work.

do u know these guy?

looks like they already have something dont know how far they have got but some of the f measures look pretty good and even 0.5 looks like a good start.

now onward to learn some pali. is there a good set of graded lectures on youtube?

is there a grammar that is more technical and analytical with all the paradigms and an up to date analysis of morphosyntax.

i dont know python but i do know some r. a long way to go.

aware of this issue (have hons degree in linguistics)
…but i havent had a good enough look at pali…do know a good greek scholar who said he can understand sanskrit fairly easily and the case system looks the same. don’t know if u can use any of those existing tagged and stemmed corpora to help with that. anyway i dont know enough yet and i need to poke around more. this is so awesome. do have some ancient greek (well lame koine from greek at a seminary in christian phase) and was surprised to see the case system is virtually identical in pali. i think i rambled about this below but the first stage would be picking of the affixes and simple verbs like go/went (high frequency)… lower frequency words are far more likely to be regular (a universal i think)… anyway mostly overexcited. i think i have found a goal. i cant work fulltime (autism, bipolar ) . as i said un the other post the dictionary problem and finding the right sense for context and the semantics to difficult but tagging and stemming have got to b easier than english

Thankfully the relatively small corpus makes doing it by hand possible:

https://digitalpalidictionary.github.io/

This monk isn’t finished with the whole Tipiṭaka yet, but he is a large portion of the way there!

1 Like