(L&N) Q&A Thread - Quick questions go here

[Arayaz]

Can a coda nasal consonant cause tone on the previous vowel?

Yes, but maybe not in the most direct way. I can think of three ways.

• First: moras. In some languages, syllables that end in nasals (in other languages in all sonorants or in any consonant) are counted as heavy. One theoretical model is that heavy syllables have two moras (or morae). In some languages, heavy syllables can bear contour tones consisting of two simple tones, i.e. one tone per mora. This is very frequent in tone languages of Africa and Middle America.
• Second: more general coda constraints. In some languages, the tone of a syllable depends on the kind of coda consonant either synchronically or diachronically. Think of checked tones in Sinitic languages or Tonogenisis in (South) East Asia.
• Third: via nasal vowels, as WeepingElf mentioned. I would guess that nasal voweks develop low tones because nasal vowels have their formants masked by antiformants. This is similar to murmured or breathy vowels which tend to develop low tones. But this is really just a guess.

All right, thank you very much!

[clawgrip]

I google translated 'I love you' into Japanese and it gave
愛してます。
Aishitemasu.

It apparently just means 'love'. Is there some politeness thing or something that hints thats 'I love you' instead of, say, 'He loves her'?

Japanese often doesn't use pronouns as it's kind of expected to be obvious from the context. Aishitemasu is a more casual form of aishiteimasu. Which are more polite compared to aisu/aisuru. All mean "to love". However, as far as love confessions go, you're probably going to hear more of aishite(i)ru. These are all very poetic, using the kanji 愛, being a Chinese loan word. More commonly, Japanese people would say daisuki 大好き (literally "big fondness") to say "I love you". Yes, without any pronouns as long as it's clear from the context.

For "he loves her", it depends on the rest of the conversation. Are we already talking about him? Then we could say "kanojo wo aishiteiru 彼女を愛している", literally "love her". However, if someone asks me "what do you think of your girlfriend?" and I'd reply with "kanojo wo aishiteiru", then it would mean "I love her". If you really need to specify, you could say "kare wa kanojo wo aishiteiru 彼は彼女を愛している" (this is literally "he loves her") or simply use the people's names. If you always put these pronouns in in every situation even when not needed, it would probably sound kind of stiff to a native speaker though.

Actually I'm just trying to understand anaphoric zero in Japanese. How does reference tracking go. "Inferred from the context" often means that we don't know.
But I understood from the answer that some politeness forms don't hint that it must be 1>2.

First of all, I just want to point out that "inferred from context" doesn't mean we often don't know, it means we generally do know... from context; otherwise, this kind of system would be untenable, and no language could employ it.

That being said, there are a few factors at play here.

Japanese rather aggressively avoids personal pronoun use in many situations, particularly second and third person. Depending on the given social situation, it could be considered pejorative or confrontational, or conversely, familiar (because familiar language in the wrong context becomes rude in Japanese). Every first person pronoun also forces you to choose what sort of relationship you want to convey to your interlocutor. As a result, Japanese speakers will instead often use personal names, titles, or nothing at all.

When nothing at all is used, it's up to situational or conversational context, or pragmatics.

A phrase like, "Feel sick, so going home," on its own is ambiguous, but in real conversation, it rarely would be.

If someone asked me, "Is there any update on Jeff?" or whatever, and I answered with this, conversational context would make it clear who was going home. Conversely, if I looked and sounded sick and just said that out of nowhere, situational context would make it clear.

There are also pragmatic concerns. The sentence "ai shite ru" mentioned above, will of course generally not be uttered out of context, but it's also most likely being said by someone expressing their feelings to someone, so in the vast majority of cases, it will be a first person subject and second person object.

When you get into humble and honorific language, potential subjects also become limited, because you or anyone perceived as being in your in-group compared to the person you're talking to can never be the subject of an honorific verb, and conversely, only you or a member of that in-group can be the subject of a humble verb. If a store clerk uses an honorific verb when talking to a customer, the customer is obviously the subject, so there is no ambiguity.

I hope this helps.

[Omzinesý]

When you get into humble and honorific language, potential subjects also become limited, because you or anyone perceived as being in your in-group compared to the person you're talking to can never be the subject of an honorific verb, and conversely, only you or a member of that in-group can be the subject of a humble verb. If a store clerk uses an honorific verb when talking to a customer, the customer is obviously the subject, so there is no ambiguity.

I hope this helps.

This is interesting!


Of course the scope of an anaphoric pronoun can we wider or smaller. If the first sentence was about N, it is natural to assume the second is about N, as well, if nothing else is said. (I sometimes see how reference tracking is lost in books translated badly into Finnish when he and she are just replaced with hän.) The (zero) anaphora can be first or second petson too, why not.

My understanding (no example, I just read that somewhere) is that you can though start a passage without a topic in Japanese. But probably that is just some discourse trick and not even meant to be understood as once.

[Creyeditor]

How does stress-ogenesis work? Does it work like tonogenesis or can stress just spontaneously appear in languages without word stress?

[Salmoneus]

How does stress-ogenesis work? Does it work like tonogenesis or can stress just spontaneously appear in languages without word stress?

I don't know.

I've always assumed that it was more that stress became more prominent, rather than just appearing out of nowhere - do any languages really have absolutely zero stress?

If stress were to become phonemic it would presumably also become prominent. So for instance if there's naturally non-prominent stress on the second mora, and then long vowels shorten leaving stress in place, you'd have prominent and phonemic stress.

I could also imagine concomitant features of stress (longer duration and higher (or sometimes lower) pitch) being reanalysed as stress, if existing stress had lost prominence?

[eldin raigmore]

How does stress-ogenesis work? Does it work like tonogenesis or can stress just spontaneously appear in languages without word stress?

IM(H?)O:
I don’t believe there are any natural languages, with two-or-more-syllable words, totally without ANY primary stress (on syllables in words), AND also totally without ANY rhythms (on syllables in words).
I remain unconvinced by any of the examples put forth by some professional linguisticians (including some famous and respected ones).
…
I think I saw an article claiming that one minor non-prestige DIALECT of some natlang had multi-syllable words but neither primary nor rhythmic stress. I don’t recall accepting nor rejecting their conclusions.
…
There do appear to be some several natlangs with only primary stress but no rhythm; and some several others with only rhythm but no primary stress.

….

Apparently (IMHO) the only ways a spoken natlang can totally avoid every kind of stress and rhythm, are:
1. Don’t have syllables.
2. Don’t have words.
3. Don’t have words with more than one syllable.

If that’s so, the term “stress-genesis” becomes a rather murky one.
There aren’t (in that case) any stress-less languages in which stress can be created.
Stress-creation would be either a part of language-creation; or a part of word-creation; or a part of syllable-creation.

….

If I’m wrong, it’s still a good question, but I’m obviously not someone who can answer it!

[qwed117]

How does stress-ogenesis work? Does it work like tonogenesis or can stress just spontaneously appear in languages without word stress?

Tone and stress are linked and there's that but I don't think there's a true stress-less tone-less language. In cases where there's no phonetic word level stress there's usually either phonological stress (eg. I think Indic languages, where 'stress' is realized as supershort syllables and collapse to minimal syllable) or prosodic stress (eg. French, ask a French speaker for more detail not me please)

For what it's worth, stress is not really a clearly defined feature as much as phonologicians and phoneticians want make it seem.

[VaptuantaDoi]

ask a French speaker for more detail not me please

I once did ask a French speaker and they insisted that French had initial stress.

[qwed117]

ask a French speaker for more detail not me please

I once did ask a French speaker and they insisted that French had initial stress.

I should clarify, a French speaker who insists that French only has prosodic stress

[LinguoFranco]

So, is RTR the same thing as - ATR?

[Arayaz]

So, is RTR the same thing as - ATR?

Yes?

[eldin raigmore]

So, is RTR the same thing as - ATR?

Yes?

I’m afraid the answers are really “maybe” and “sometimes” and “it depends”.

In vowels, Advanced Tongue Root is contrasted with Relaxed Tongue Root. (One can’t articulate a vowel with a Retracted Tongue Root.)
In consonants, Retracted Tongue Root is contrasted with Relaxed Tongue Root. (One can’t articulate a consonant with an Advanced Tongue Root.)

So it’s not safe to use those RTR abbreviations without more information, such as context, or more words, or at least more letters than just the initials!

Like so many other linguistic terms, there are more ideas to express than there are abbreviations or acronyms or initialisms to call them.

….

It’s a problem with medications and pharmacies as well!
When CVS texts me about refilling my ATO, I have to call and speak to a human being to find out whether they mean Atomoxetine or Atorvastatin.
Or if they text me about refilling my OXY, I don’t know whether their robot is asking about my Oxybutynin or my Oxycodone.

[Salmoneus]

So, is RTR the same thing as - ATR?

No, they're different things. -ATR is the absence of tongue root advancement, whereas +RTR is the presence of tongue root retraction. I'm not aware of any languages contrasting more than two tongue root positions phonemically, but different languages differ in the usual phonetic realisation of the binary contrast, and linguists working on different languages describe them differently, which may or may not be because the realisations differ in practice.

[both vowels and consonants can be either +ATR or +RTR]

[Creyeditor]

In general, what Eldin said.
For phonetics (or the phonetics-phonology interface) what Sal said.
For phonology what Arayaz said.
In phonological theory, some discussions of this are based on phonological markedness of binary features (a very much vaguely-defined concept). Whereas people are relatively sure for other vowel features about which feature value is more marked, people haven't really found conclusive evidence for eitger advanced tongue root or retracted tongue root as a marked feature value. It seems to vary on a language-specific base. Some people have concluded that languages use different features. Languages that use [±RTR] have [+RTR] as the marked value and languages that use [±ATR] have [+ATR] as the marked value. Under these assumptions [+ATR] could be phonetically identical to [-RTR] or it could not. They would refer to the same natural class of phonological segments, but you would expect differences in their phonological alternations and their phonotactic restrictions.

[Arayaz]

I think most people/books I know call /i/ advanced or non-retracted and /ɪ/ retracted or non-advanced. But remember, I make a lot of very bold claims from a position of what looks like robust knowledge, but where I do not have as much authority as I seem to present myself as having.

[eldin raigmore]

I think most people/books I know call /i/ advanced or non-retracted and /ɪ/ retracted or non-advanced. ….

But they’re not talking about the Tongue-ROOT.
They’re talking about the main body of the part of the tongue that is in the mouth.
/i/ is Front and /ɪ/ is Near-Front.

The “root of the tongue” is, roughly, the front wall of the pharynx.
With ATR you’re widening the pharynx; with Retracted TR you’re narrowing the pharynx.

….

I’m guessing most of those books weren’t originally written in English?

[Creyeditor]

Re: Stress-ogenesis
@Eldin, @qwed: I think I have seen good arguments that some languages have only phrasal prominence (as diagnosed by pitch and duration cues) which is not aligned to any particular position inside a word. Some of these languages (e.g. some varieties of Standard Indonesian) also do not show any indirect evidence for metrical prominence in their segmental alternations, IINM. I would argue that these languages do not have word stress.

@Sal: Vowel length is a good candidate; that makes sense. Tone of course is a good candidate. I was actually thinking if phrasal prominence good become word stress in some way.

[Omzinesý]

West Greenlandic stressing is also interesting.
It basically has prominence on every strong syllable (closed or long vowel). Words thus often have several prominent syllables. The concept of word is generally difficult in Greenlandic.

[eldin raigmore]

….
@Eldin, @qwed: I think I have seen good arguments that some languages have only phrasal prominence (as diagnosed by pitch and duration cues) which is not aligned to any particular position inside a word. Some of these languages (e.g. some varieties of Standard Indonesian) also do not show any indirect evidence for metrical prominence in their segmental alternations, IINM. I would argue that these languages do not have word stress. ….

Thanks for this!

I’ve also seen arguments; IM(NSH?)O the ones I’ve seen weren’t good enough.
Can you point me to URL, or otherwise indicate to me, where I can read some of the good arguments you mention?
Maybe they’ll include some that I’ve not read yet, and will find convincing.

I’d generally take, as probably reliable, your word for anything you say about Indonesian.
But I think you claim (only) some varieties of Standard Indonesian don’t have any stress or rhythm in syllables within polysyllabic words?
If so, I’m not sure they constitute a counter-example to my own opinion.
….
Another interesting difference between dialects and languages is in the timing.
My ‘lects of English are just about (stress-)foot-timed; pretty much every prosodic or metrical foot in each utterance is very close to the same length, time-wise, as every other stress-foot in the same word.

But in at least some varieties of British Stage English, (almost all) utterances are (almost perfectly) syllable-timed, with each syllable in the utterance taking (very close to) the same time to utter, as each other syllable in the same sentence. (Or, at least, I think I mean “sentence”.)
In such English, “stress” and “rhythm” are shown by pitch, rather than length.

There are and/or other languages or dialects —— St. Augustine’s Latin, for instance —— which are mora-timed.
Any bimoraic syllable takes exactly* twice as long to say as any monomoraic syllable.
*(At least, St. Augustine said that!)

[Creyeditor]

I was specifically thinking about a series of experiments by van Zanten and Goedemans (reported a.o. here: https://www.academia.edu/download/81343 ... 03ebc8.pdf) which seems to suggest that Standard Indonesian as spoken by Javanese native speakers has phrasal stress but no words stress. This holds for production and perception. This pdf is not behind a paywall, other publications on this by the same authors may very well be.
If your claim is that no set of varieties of a single language can only include varieties without any word stress (or equivalently that any set of varieties of a single language must include at least one variety with word stress) than Standard Indonesian as spoken by native speakers of Javanese (SIASBNSOJ) is of course not a counterexample. But still, if the variety of SIASBNSOJ would gain word stress in the future, could we call this stress-ogenesis?