Comment by quadrifoliate | Hacker News Reader

Comment by quadrifoliate | Hacker News Reader

Parent Full thread

> With so many, where do you draw the line?

The whole issue of language v/s dialect is a complex one; but I have adjusted to the one in this popular article [1]: if you speak a language and can enter into a conversation with a speaker of a variation on that language relatively easily, it qualifies as a dialect.

> A Mandarin-speaker can no more “adjust” to Cantonese than a Swede could “adjust” to German.

Most of the well-known Indian languages actually fail this test. Someone who has grown up speaking Hindi cannot reliably converse with someone in Bangla, Sambalpuri, or Tamil. If anything, the word "dialect" is overused regarding India – Bhojpuri [2] was often mentioned as a dialect of Hindi; but as a fluent Hindi speaker I cannot claim to understand or have a fluid conversation with someone speaking Bhojpuri. I am glad that it is at least sometimes referred to as a language now.

----------------------------------------------------

[1] https://www.theatlantic.com/international/archive/2016/01/di...

[2] https://en.wikipedia.org/wiki/Bhojpuri_language

That would make Hindi and Urdu dialects of each other. I'm not saying this is necessarily right or wrong; it's a landmine I don't want to touch :)

It is true Urdu is just a colonial imposition, it is synthetically derived from Hindi by insisting that the script be Persian.

Tarek Fatah makes a point about why Urdu isn't a distinct language from Hindi.

Yiddish is written in both hebrew letters and latin letters, but is one language. Language speakers don't care too much about the script when evolving their language; they use what they have.

I think Hindi/Urdu is quite a bit like Serbian/Croatian. The preferred script followed the religious divide, but (if I understand right) both sort-of standardised quite similar dialects out of a range of possibilities. And then later they were politically divided, and there were some efforts to move the official vocabulary away from each other's.

IIRC Sanskrit has also been written in dozens of scripts, there's a perfect one-to-one transliteration into most (maybe all) the north-indian scripts, although they have extra characters for their own sounds.

Since all indian scripts besides the Urdu Nastaliq were originally developed to write sanskrit, the transliteration is indeed perfect!

Why is it a landmine? I am an Urdu speaker, and totally fine with realizing that Urdu is the Persianised standard register of the Hindustani language, and Hindi is the standardised and Sanskritised register of the Hindustani language.

There is some issue with priority... modern Shudh Hindi (the standard literary dialect) only dates from the 19th century when it was essentially invented by a number of Hindu intellectuals mostly in Benares. Urdu is older, along with the lovely Hindi dialect Braj Bhasha (both dating from the mid 16th century onward). The earliest Hindavi writing is Kabir, from the the 13th century. Also, Panjabi is a complex relation, since it's also the same language, filtered through the gilded pen of Guru Nanak.

Urdu is definitely a dialect of Hindi. Speakers of Hindi, and of Urdu, can understand most of what is said and can have a conversation for most part. They may not understand 100% of the other language, especially when words that are closer to their Sanskrit or Persian origin are used and have not become part of the colloquial other.

It's more appropriate to think of Hindi (Rajbhashya/Khariboli) & Urdu as different registers of the same language.

For example, a lawyer & a rapper might both compose their works in English, while adopting a register & jargon geared towards their respective audiences. Their works might very well be largely incomprehensible to each others target audiences.

Bhojpuri, Maithili, Haryanwi, Braj, Bundelkhandi, dialects from Rajasthan, Himachal and so and so forth, and I'll say even Khari boli are all dialects of Hindi. We may not be able to have a fluent conversation, but for most part we'll understand the context and will be able to speak. In fact Hindi dialects change every 100km or so I'll say in the Hindi belt.

Punjabi, Gujrati, Marathi and others (picking from areas neighboring the Hindi belt) are not dialects, they are languages, even though from that list Gujrati and Marathi use Devnagri script (or very close to it).

hmm American here - but I vaguely recall.. Gurmukhi is a variation on Punjabi, used by Sikhs for religious purposes, but Urdu is a 'persianized' version of Hindi (?)

Checking wikipedia, it mentions a "Punjabi diaspora family of languages." These distinctions seem understandable and interesting to me, of European descent and non-political.

thanks yumraj, for the many posts here today on this, it feels important at some basic level.

I enjoyed reading "In Search of the Cradle of Civilization" long ago.. and studied personally with one of the authors, in California.

https://en.wikipedia.org/wiki/Subhash_Kak#In_Search_of_the_C...

Gurmukhi is the script used to write Punjabi, at least in India. It was created by a Sikh Guru. Link: https://en.m.wikipedia.org/wiki/Gurmukhi

Interestingly, Pakistan has a state called Punjab where Punjabi is the main language, but I’m not sure what script they use.

Regarding Hindi and Urdu, I found the wikipedia page to at least be informative. Link: https://en.m.wikipedia.org/wiki/Hindi%E2%80%93Urdu_controver...

Pakistani Punjabi uses the Shahmukhi script, which is based on the Perso-Arabic script.

Thanks. So, Punjabi must be one of those rare languages that is written in two entirely different scripts by roughly equal population sizes.

Are there any other examples of such a language?

There are, with many more examples historically.

Konkani in the Devanagari script is an official language of India & the state language of Goa - while Christian Konkanis (a sizable proportion of the population) prefer to write it in the Latin script.

Kashmiri Pundits used the Sharda script, while Muslim Kashmiris wrote it in the Persian script.

Historically, the picture is more complicated. People were familiar with multiple scripts, and shifted across them based on context and jargon. For example, when the British took Sindh from the Nawab, they found the natives writing the same language in multiple scripts. The administration of the erswhile Nawab worked in Persian, while merchants used Khudabadi, a landa script related to Gurmukhi. They found that women in the home preferred the Devanagari script, as did the priestly class. Since the British were most concerned with the needs of the Persianate clerks, they mandated that Sindhi only in the Persian script would be taught in schools. You will find similar situations across India, with people adopting scripts geared towards the audience they were writing to, rather than an exclusive formalized script unique to a language. Literate Medieval Indian would be fluent in multiple languages and scripts, and combined them in interesting ways.

On a tangent, I recall an argument by Bhagat Singh that the different languages of India should adopt a common script to foster unity - Devanagari in his opinion.

Italians can understand Spanish quite well and vice versa. Similar for Portuguese. That doesn't make them dialects of Latin.

I can understand some Punjabi, and given context might be able to guess Gujrati, Marathi, Bangla due to their origin in Sanskrit, but I did call them separate languages.

The ones I mentioned as dialects are indeed dialects, at least per commonly accepted standards.

Gujarati has its own script, fairly similar to Devanagari but without the horizontal bar above letters and a few more changes. Marathi uses Devanagari.

I believe both Urdu and Hindi came from Hindustani which was derived from Khari Boli.

On the issue of scripts, the matter was more fluid some generations ago. Different people would adopt different scripts in different circumstances, and it was the British push for standardization that settled the matter.

For example, when the British took Sindh, they found that officials of the erstwhile Nawab administration used the Persian script, while merchants used the Khudabadi landa script. Women and priests, however, largely used the Devanagari script. The British adopted the Persian script for Sindhi since the clerks they absorbed used it.

A similar story played out with Marathi. From the time of the Marathas till as late as the 1920s, Marathi was written formally in the Modi (ie, twisted) script. The British pushed for the adoption of the Devanagari script, since it would simplify logistics at printing presses.

More likely, people speaking different dialects will adopt a standard register to communicate with each other. For example, I have seen Himachali speakers from neighbouring districts communicate with each other exclusively using Khariboli Hindi, rather than their native Himachali dialects.

This is probably true of most large, diverse languages. Each dialect has drifted sufficiently far from each other than while it is possible to communicate with other dialects, it's most convenient to use the shared formalized standard register. So while a Malvani speaker can largely comprehend the dialect of somebody from Gadchiroli, more often both speakers will switch to the standard Marathi register they learnt in school.

That definition does not work with Czech and Slovak.

Why? Aren't they rather close, and more of a language family than two distinct languages?

They are distinct languages from the same language family, which are understood by speakers of the other language, but not spoken. Czechs understand Slovaks and vice versa, but both are talking in their own languages, and speaking the other language can't be done properly without studying it.

At least with Hindi and Tamil, you can be sure because they come from different language families.

True. This may sound obvious to you, but I am not so sure about its obviousness in general in the West.

A lot of western-educated people have asked me “So Hindi is the main language, and there are a lot of different dialects throughout the country, right?” – and I have no idea where to begin explaining that it's nothing like that, so I go with the time-tested “Well, it's slightly different...”

> and I have no idea where to begin explaining that it's nothing like that

(I have studied Sanskrit and Hindi in school for multiple years, and can understand a good amount of Marathi and Tamil.)

You can say this - I think it is true, not quite verified (partly heard from older relatives, partly deduction based on observing similarities in the mentioned languages' vocabularies):

Sanskrit came first - thousands of years ago. It is the language of the Vedas, 4 of India's most ancient scriptures. Many northern (including east and west) Indian languages are similar to Sanskrit and likely are descendants of it. Hindi, Gujarati, Marwari, Punjabi, Bengali, Oriya (not sure, but it sounds some like Bengali), Marathi, are some of these.

Languages of Indian states further east than Bengal, called the North Eastern states, like Assam, Nagaland, Manipur, etc. may not show a lot of Sanskrit influence because they may have Oriental influence instead, such as from SE Asia, China, etc. Except Assamese may have some Sanskrit influence, also not sure of Manipuri.

And of the southern states in the peninsula, Tamil has the least Sanskrit influence and is also said to be very ancient. The other three, Malayalam, Kannada and Telugu, have some Sanskrit influence.

Caveat: I am not a language expert.

Bengali is very much a descendent of Sanskrit. Assamese (and Oriya) is essentially a somewhat distant dialect of Bengali. Most languages of Asia that aren't part of the Chinese world (Chinese, Japanese, Korean, and modern Vietnamese) have tons of Sanskrit or Pali (or both) in them. Tamil has lots of sanskrit in it (good examples are naka meaning finger nail, singa meaning lion, ampala from ambara meaning sky) it just has less than other dravidian languages, presumbaly because of its much older literary tradition. Literaly Malayalam looks literary Tamil, with all the words converted to Sanskrit. It's kind of funny!

And Tamil have long tradition of resisting Sanskirt influence. With elaborate grammar rules dictating if you must use loan words when, where and how to use them. It is more true for literary setting and never true for day-today speaking. Also, Tamil words also made into Sanskrit, don't know what it means. Languages are basically for interchanging ideas. ¯\_(ツ)_/¯

I mean... Tamilians sort of dominated of dominated Sanskrit literature in 7th-10th centuries. Dandin, Kumarila, Shankara, Dharmakirti, Ramanujan and Madhva. Not to mention the Bhagavatapurana, which very quickly became probably the most important religious text in India! I wouldn't say Tamilians rejected sanskrit... they just kept the two seperate. There aren't very many tamilian words itself in sanskrit (mina meaning fish is the only one I can think of), but there are proto-dravidian loan words in Sanskrit.

Most of the people in that list were likely not Tamilian - Shankara was likely Malayali, Kumarila & Dharmakirti are just as likely to have been from Assam, and Madhva was Tuluva.

What about the Bhagavatapurana is Tamil?

It literally quotes the Azhvars... you're right about Malva, that was a mistake. I don't agree with a possible assamese origin of either Kumarila or Dharmakirti, and see no reason to doubt Taranatha, and in the 9th century, the distinction between Malayali and Tamil is academic.

Assamese is NOT a dialect of Bengali. It is older than Bengali and infact the first non-Sanskrit language to which Ramayana was translated to. The Assamese script itself is in a continuous development since 5th century.

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

The earliest extent translation of the Ramayana is the Tamil Kampan Ramayana from the late 12th century. The might have been a Khmer translation earlier. The Assamese Kotha Ramayana is about two hundred years later. The Jain Ardhamagghadi Ramayana is very old, perhaps older than the Valmikiramayana!

I stand corrected about Assamese's relationship with Bengali.

No worries. I meant in the Indic language family, not Dravidian. The Tamil version definitely preceded that.

> Bengali is very much a descendent of Sanskrit

Yes, that is nearly same as what I said above, except I said "likely", because I was not sure, not being a language expert. And that "likely" applied to the other languages I mentioned too, in that sentence.

Edit: Actually, I am (still anecdotally) pretty sure of it for at least Hindi and Marathi, because both of them have too many words which are the same as, or derived from Sanskrit, for it to be a mistake.

"The situation is a lot more nuanced than that."

"it's slightly different" confuses the issue.

It seems easy to say, "actually, no, India has multiple unrelated language groups as different from each other as English is from Hindi, and I don't just mean in isolated villages"

I usually explain it as being similar to the differences in languages across Europe. If you take portuguese at one end and russian at the other and start tracing across countries, you'll notice certain similarities and patterns but also encounter a lot of unique character especially in some areas. Spanish and portuguese people might be able to understand a little of each other like tamil and malayalam speakers and so on.

Reply on news.ycombinator.com