Linguistics thread

[vijayjohn]

Sure! There are lots of Sanskrit loanwords like this. Also anything ending with -garh.

Could you name some? I still can't think of any.

Sorry, I couldn't do it earlier because I had to go for work soon after I saw your reply.

सौरभ
अविरुद्ध
अविरोध
गत्यवरोध
निषेध
निषिद्ध
आरूढ़
मेघारूढ़
अस्वारूढ़
अगूढ़
दृढ़ायुद्ध
सहोढ़
अमोघ
निदाघ

And some that aren't (entirely) tatsams:

आसाढ़
पेलढ़
रक्तकोढ़
सीयावढ़
नरसिंघ

[dEhiN]

Actually when I first learned about pitch accent vs tone in linguistics, I thought it interesting because music essentially uses those 2 words as synonyms. In fact, I can't think of any example right now where in a musical setting I would only be able to use one and not the other.

Interesting, I didn't know "pitch accent" was a term used in music as well.

No sorry, I should clarify. I meant the terms "pitch" and "tone".

[vijayjohn]

OK, "pitch" is completely different from "pitch accent" even in a linguistic context.

[dEhiN]

OK, "pitch" is completely different from "pitch accent" even in a linguistic context.

Oh, I didn't know pitch was used in linguistics. What's the difference then between pitch and tone in linguistics?

[vijayjohn]

OK, "pitch" is completely different from "pitch accent" even in a linguistic context.

Oh, I didn't know pitch was used in linguistics. What's the difference then between pitch and tone in linguistics?

Well, "pitch" probably means the same thing in linguistics as it does in music: an acoustic property correlated with frequency (i.e. if the frequency of sound waves in a particular audio sample is relatively high, our ears will probably perceive it as a sound with a relatively high pitch). It's "tone" that has a specialized meaning in linguistics, referring to pitch contrasts being relevant at the phonemic level in specific languages. I would think music was all about salient pitch variation whereas language is not necessarily, so in the context of music, this wouldn't seem to be a very useful distinction to make, and it would make sense that "pitch" and "tone" would be used interchangeably.

[mōdgethanc]

My understanding is that tone can appear on any syllable whereas pitch accent can only appear on one, much like how stress works in most languages. And that syllable need not be stressed because it's possible to have a language with pitch accent but no phonemic stress (Japanese, for example). In this case, the pitch functionally is the stress.

But every language with pitch accent seems to work in a somewhat different way so there is basically no clear definition but rather it's a catch-all term for a system that's somewhere in between having phonemic stress and being a tonal language.

Interesting, I didn't know "pitch accent" was a term used in music as well.

It's not, at least not that I know of (and I've played music for over a decade). "Pitch" is a term, and "accent" is a term, but "accent" is more like stress in linguistics.

[Vlürch]

Finnish doesn't even have phonemic stress, so just trying to get what all the pitch accent shit in any language is about makes my head hurt.

in the context of music, this wouldn't seem to be a very useful distinction to make, and it would make sense that "pitch" and "tone" would be used interchangeably.

Well, technically, there could be situations where they aren't interchangeable, but it wouldn't really make sense to extend those rare situations to all musical contexts. When speaking of the "tone" of guitars or bass or whatever, it can refer to the quality of the sound rather than the pitch; for example, you could have a "low/thick" tone without referring to the actual tuning of the guitar or the notes played, even if most of the time a "thick" or "heavy" or whatever "tone" goes hand in hand with a lower tuning and thus a lower pitch. The same applies to vocals, too, although I don't think anyone would generally describe any kind of singing as having a "low tone" unless the pitch was low...?

When it comes to stuff like death metal growls, though, that's a whole another world of random words describing random shit that makes no sense half of the time (I mean, "toilet growl" makes perfect sense, but actually describing what a toilet growl is... well... yeah) and even more ridiculously, some of the terms used by electronic music artists and especially the general "community" are the stuff headaches are made of; in reference to synths or whatnot, "tone" can refer to the central pitch of a sound while "pitch" would refer to the detuning/harmonising/whatever pitch that varies, without implying any relation to the actual notes in a song, and tons of other terms are thrown around in ways I have no clue about.

[dEhiN]

A few days ago, I read this article entitled Google's AI Has Reinvented the Master Language. Apart from the attention-grabbing headline, the content of the article seems quite exciting. Basically Google created a new engine for Google Translate, the Google Neural Machine Translation system. It seems to go beyond NLP, since according to Google's tests and demonstrations, this new engine relatively successfully handled a language pair that it had no prior data for. What do you all think?

[vijayjohn]

It seems to go beyond NLP, since according to Google's tests and demonstrations, this new engine relatively successfully handled a language pair that it had no prior data for. What do you all think?

That's not "beyond NLP"; it is NLP and precisely the sort of thing NLP people try to do, so you can guess how unimpressed I am. Neural networks go beyond the usual NLP approach but are most definitely still an NLP approach.

[dEhiN]

It seems to go beyond NLP, since according to Google's tests and demonstrations, this new engine relatively successfully handled a language pair that it had no prior data for. What do you all think?

That's not "beyond NLP"; it is NLP and precisely the sort of thing NLP people try to do, so you can guess how unimpressed I am. Neural networks go beyond the usual NLP approach but are most definitely still an NLP approach.

Then I guess I don't know what exactly NLP is! I thought it basically meant using a whole bunch of textual data of a language pair in use, and drawing inferences from that data about what different words mean and such. So I figured that approach wouldn't work with a language pair that basically has no data for it. I guess this is where the limits of my understanding of linguistic principles come in!

[vijayjohn]

No, NLP is the sort of thing computational linguists and computer scientists try to do: plug their favorite algorithms into various kinds of problems and keep tweaking parameters until yay, they finally have a result they can publish in a paper! (As opposed to, you know, actually trying to solve the problem).

[md0]

Goes under the general field of Theory-less Science, which is a tendency that will doom us all.

[vijayjohn]

Goes under the general field of Theory-less Science, which is a tendency that will doom us all.

[dEhiN]

Quick question. I was having a debate with a friend about the use of "him and I" versus "he and I". (I know this is the Linguistics thread and perhaps my questions should be in the English subforum. But I am curious about this from an English linguistic perspective).

I was talking to a friend about another friend. So in my first sentence I reference the other friend by name. Then in my next sentence I said "him and I" because him is the oblique case and the phrase "him and I" is a nominal phrase that acts as the subject. To my thinking then, "him and I" is a perfectly acceptable use of current English.

But my friend interrupted me and say "no, it's 'he and I'", and that resulted in a mini-debate. My assumption is that she is going by older English grammar rules that were based on the old school belief that English grammar should be like Latin grammar. These same rules are the ones that say you should never have a split-infinitve in English, nor end a sentence with a preposition. My friend is only 5-6 years older than me, so I was surprised she was so adamant that the correct way is "he and I".

[linguoboy]

This was probably something beaten into them by a misguided grammar teacher.

If they know any Romance, you could try pointing out that in French the equivalent expression would be "lui et moi" and not *"il et je", but I'm not sure that would help much. This tends to be a classic instance of "a little learning".

If you want to research linguistic explanations, the term of art is "disjunctive (use of) pronouns".

[md0]

Huh, Greek requires both pronouns in the nominative case here. TITAIATIL

Any chance "me and he" is acceptable btw?

[linguoboy]

Huh, Greek requires both pronouns in the nominative case here. TITAIATIL

Any chance "me and he" is acceptable btw?

For some speakers, certainly, but I'm not sure how many. In addition to prohibitions on non-subject pronouns, most speakers also have thoroughly internalised stylistic rules which mark it "impolite" to put yourself first. So this construction violates not one but two widely-obeyed rules governing the formation of pronominal subject phrases. (I Googled "me and he" and most of the valid examples that popped up were drawn from rap lyrics.)

[dEhiN]

So apparently this thread has become a hot topic right now on the Skype UL group! Linguoboy, I did look up the disjunctive pronoun, but I think it's a shame they use French as an example. As someone on the Skype group pointed out, if we're not going to determine English grammar based on Latin grammar, we also shouldn't determine it based on French either.

Though someone else mentioned that none of the other Germanic languages allows a mixed case subjective nominal phrase.

I think perahps this mixed case is a situation of English changing, and so there are some that still cling to the older way of speaking and others that are embracing the change.

[linguoboy]

As someone on the Skype group pointed out, if we're not going to determine English grammar based on Latin grammar, we also shouldn't determine it based on French either.

I figured you'd get that objection. This misunderstands the purpose of an example. It's just showing that other languages--including those directly descended from Latin--don't have issues with this kind of construction, so why should we? I'm not proposing we base English grammar on anything other than English usage, which clearly allows this kind of construction or we wouldn't have to explicitly teach rules against it.

Though someone else mentioned that none of the other Germanic languages allows a mixed case subjective nominal phrase.

There are a lot of Germanic languages in the world. Are we sure that none of the others allow such a construction?

Even so, I'm not sure what difference would make. It would hardly be the only unique feature of English grammar.

I think perahps this mixed case is a situation of English changing, and so there are some that still cling to the older way of speaking and others that are embracing the change.

There doesn't have to be a change in process for there to be a competition between variants. Change also doesn't have to be unidirectional. Look at the spread of "...and I" in objective pronominal phrases. In origin, it's a hypercorrection, but it's since acquired a life of its own.

[vijayjohn]

There are a lot of Germanic languages in the world. Are we sure that none of the others allow such a construction?

I just found this from this Swedish website:

Vi har dock ett villkor för att genomföra köpet. Det är att vi, Charlottendals Mc, kör motorcykeln till den närmaste tull station för oss och vi, du och oss, tillsammans deklarerar motorcykeln.