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

[Creyeditor]

Very short answer: no, they are different.

I think you are asking about the articulatory phonetics of slack voice. The thing is, our vocal cords don't just do voicing or not. There are several anatomically possible positions. Slack voice is between two of these idealized positions: spread vocal cords (as in aspirated sounds or /h/) and the 'sweet spot' where maximal vibration occurs (as in voiced sounds. (Yet a different idealized position is a glottal closure, as in glottal stops.)
Voicing, however, also has a temporal dimension (at least for plosives). What is the temporal relation between vibrating vocal cords and stop closure. This is often called voicing onset time. A voiced sound will have a negative VOT (voicing starts before the closure), a voiceless sound will have a VOT close to zero (voicing starts as the closure end). Half-voiced plosives are in between in that they might have a very small negative VOT.
I hope this helps.

[Salmoneus]

I agree with all that.

The other thing I'd add, though, is that any time someone starts from a point like "English stops have [this detailed subphonemic feature of realisation]", they're likely to be getting into some murk.

English is a hugely varied language; there are many dialects and sociolects, and even one speaker pronounces words differently in different phonological contexts, and in different pragmatic or sociological contexts too.

Take final voiceless stops, as in "tap". That final stop will be, by different speakers or in different contexts, pronounced aspirated or unaspirated. Or unreleased. Or glottalised (pre-glottalised, post-glottalised...). Or with devoicing of the latter stages of the preceding vowel. Or shortening (or non-lengthening) of that vowel. Or replacement by a glottal stop. Or, of course, it could be an affricate, or just a fricative! And those variations all occur just within England, let alone the broader anglosphere! I've no doubt that if you measure fine details of voice onset timing and degree of openness of the vocal folds, you'll find considerable variation.

[Creyeditor]

Very true. German 'devoiced voiced plosives' (notated the same as English half-voiced and Javanese slack voiced plosives) have a small negative VOT, a VOT close to zero, or a small-ish positive VOT, depending on speaker and place of articulation (and probably many other things). And all of these statements of course already ignore some variation by averaging VOT across several repetitions and excluding rare outliers where other stuff happens (e.g. no full stop closure, heavy breathing, gulping, etc.).

[KaiTheHomoSapien]

What is the exact process by which Sabellic derived /f/ from /ns/? /ns/ --> /f/ seems a strange sound change. Are there are examples of this sound change outside of Sabellic?

[WeepingElf]

What is the exact process by which Sabellic derived /f/ from /ns/? /ns/ --> /f/ seems a strange sound change. Are there are examples of this sound change outside of Sabellic?

I don't know, but some scholars reconstruct the PIE accusative plural as *-ms, which could have gone to -f via *-mf. But in Sabellic, both *bh and *dh have become f, which doesn't really help, though. However, *-ns > *-nþ > *-þ > -f is also a plausible pathway.

[Khemehekis]

Today I was doing P-words for my Landau Core Vocabulary with the excellent English-Irish dictionary at www.focloir.ie. While I was looking up "parrot", I came across an interesting idiom on https://www.focloir.ie/en/dictionary/ei/parrot:

"my mouth feels like the bottom of a parrot's cage"

As an American, I've never heard this expression, and figure it is probably unique to the British Isles, or at least the Commonwealth countries. It's a colorful and curious saying, though, so I'd like to know: What does it mean?

[Salmoneus]

I've never heard the expression, but to help you out next time you have a similar confusion, here's three suggestions:

- try working out what it might mean by imagining that what it said was true. What do you imagine the bottom of a parrot cage tastes like? Would it be a good taste, or a bad taste?

- try using www.google.com or a similar website. A quick search shows a variant on this expression being used by American author Stephen King, and somebody else explaining its meaning.

- the site you're talking about is actually a dictionary, so these phrases are translated there. Maybe try looking at those translations and seeing what they mean? If you don't know some of the words, the dictionary actually has translations for them too.

In this case, your dictionary gives you the translations tá drochbhlas in mo bhéal and tá boladh bréan in mo bhéal. Needless to say, "tá... i mo bhéal" means "there is ... in my mouth". And the dictionary a translation for both 'drochbhlas' (not that you'll need one, since it's obviously just droch+blas, where 'droch-' is a prefix meaning bad, and 'blas' means flavour or taste'). And, yes, it means a bad taste. And your dictionary actually confirms this by using drochbhlas a bheith i do bhéal as an example phrase. And, likewise, "boladh" is a smell, and "bréan" means foul or putrid, and it actually confirms this by using "boladh bréan" as an example phrase, meaning 'evil smell'.

So, it means to have a bad taste or scent in your mouth. Which is, coincidentally what you would have if your mouth did actually taste like the bottom of a parrot cage. Or the bottom of a birdcage, or the newspaper from the bottom of a birdcage, etc, as google suggests as variants of the expression.

[Keenir]

I've never heard the expression, but to help you out next time you have a similar confusion, here's three suggestions:

Given that it was probably a regional idiom, I don't see the downside to asking here, where pragmatics and expressions are part of what are dealt with, as parts of languages.

[Omzinesý]

Q
If verbs have classes that trigger agreement, what agrees with verb class? Does such even happen?

[Creyeditor]

Could you give a concrete example of what you mean?

[Salmoneus]

Presumably it would be the things in the verb phrase, whatever you put there - adverbs, TAM particles etc. Person markers seem plausible, maybe. Maybe obects - eg differential object marking triggered by lexically by the verb?

I'd recommend thinking about the origins of such marking.

In general it's not something you'd expect to happen often, because it's pointless. Nominal agreement happens a lot for a good reason: it tells you which noun the adjective goes with. Agreement with the verb makes less sense because that sort of ambiguity hardly ever arises anyway.

[Arayaz]

May I ask, how is ergativity lost in a language?

[eldin raigmore]

Q
If verbs have classes that trigger agreement, what agrees with verb class? Does such even happen?

I’ve seen lit wherein it is simply called “verb class”; however, most uses of the phrase “verb class” do not refer to the concordial verb-class you’re talking about.

I wrote to Rachel Nordlinger about it.
See these threads:
http://cbbforum.com/viewtopic.php?p=29 ... s#p292785
http://cbbforum.com/viewtopic.php?p=29 ... s#p292800

Spoiler:

http://cbbforum.com/viewtopic.php?p=29 ... s#p292804
http://cbbforum.com/viewtopic.php?p=29 ... s#p292808
http://cbbforum.com/viewtopic.php?p=29 ... s#p292882

http://cbbforum.com/viewtopic.php?p=29 ... s#p292894
http://cbbforum.com/viewtopic.php?p=29 ... s#p293140

(You might want to look at the last-listed one first! Then the others in the order listed, or any order you want!)

I’ve called it “verbal gender”. That parallels ideas like “nominal aspect” (e.g. are you at Kyiv or in Kyiv?) and “nominal tense” (e.g. spouse, ex-spouse, spouse-to-be). On CONLANG-L some linguist said she liked my “nominal aspect” idea; I don’t know if Professor Nordlinger liked my “verbal gender” term, but at least she understood it.

I hope that helps, and isn’t TL;DR!

—————

Edit: P.S.: You might want to read http://cbbforum.com/viewtopic.php?p=30 ... s#p304938 as well. It has only two posts!

[Creyeditor]

May I ask, how is ergativity lost in a language?

Three options come to mind:

• The language could lose core case marking alltogether (maybe through phonological erosion), such that ergative case and absolutive case become indistinguishable.
• The language could go through a phase of split-ergativity. Maybe new tenses are derived from existing constructions that happen to have accusative alignment. Later on, the original ergative construction just falls out of use and you get an language with accusative alignment.
• The language could go through a phase with optional ergative marking (attested in some Papuan languages). The frequency of ergative marked nouns just becomes gradually lower until its zero. Probably this is again influenced by the grammatical context, too.

[Arayaz]

May I ask, how is ergativity lost in a language?

Three options come to mind:

• The language could lose core case marking alltogether (maybe through phonological erosion), such that ergative case and absolutive case become indistinguishable.
• The language could go through a phase of split-ergativity. Maybe new tenses are derived from existing constructions that happen to have accusative alignment. Later on, the original ergative construction just falls out of use and you get an language with accusative alignment.
• The language could go through a phase with optional ergative marking (attested in some Papuan languages). The frequency of ergative marked nouns just becomes gradually lower until its zero. Probably this is again influenced by the grammatical context, too.

Thank you!

[Salmoneus]

May I ask, how is ergativity lost in a language?

Three options come to mind:

• The language could lose core case marking alltogether (maybe through phonological erosion), such that ergative case and absolutive case become indistinguishable.
• The language could go through a phase of split-ergativity. Maybe new tenses are derived from existing constructions that happen to have accusative alignment. Later on, the original ergative construction just falls out of use and you get an language with accusative alignment.
• The language could go through a phase with optional ergative marking (attested in some Papuan languages). The frequency of ergative marked nouns just becomes gradually lower until its zero. Probably this is again influenced by the grammatical context, too.

It could also be that the language allows direct objects to be elided when deduced from context. "Did he eat the cabbage?" - "He ate" (with no need to repeat "the cabbage" in the answer since it's obvious from context what you're saying he ate).

Once you allow that, it can easily be reinterpreted as zero-marked valency-reduction: bivalent "He ate [it]" (object elided) reanalysed as univalent "He ate" (no object).

At that point, an ergative language effectively has two classes of univalent verb: original univalents, for which the subject is zero-marked; and derived univalents, for which the subject is still in the ergative. This is a form of lexical split-ergativity, in other words, so technically I'm just repeating what you said, but perhaps coming at it from a different direction, since this approach is just driven by pragmatics and analogy rather than semantically-based rival and entirely independent constructions, which I think is what we might normally think of when we talk about split-ergativity and the ergative construction falling out of use.

From there, it's easy to spread the ergative marking of the derived univalents to (perhaps in stages) the original univalents - particularly if, due to earlier derivational processes, there are derived/original doublets that are homophonous or close to it. At that point, your "ergative case" has become a nominative case (albeit a marked one, as in Proto-Indo-European).

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

An alternative route: differential object marking moves some objects out of the absolutive and into a new case we can call the accusative (perhaps originally a dative, or partitive? Or both, why not, get some new noun cases in there!). This new case is then levelled to all objects.

This then creates a tripartite system, with marked ergative and accusative in divalent verbs, while univalent verb arguments retain a distint unmarked absolutive.

This tripartite system is cross-linguistically probably unstable, since it's rare - the absolutive is not 'contrastive' with either divalent case, since they can't appear in the same clauses (provided there isn't zero-marked valency reduction), and therefore is superfluous. Either it can be reinterpreted as the agent case and replace the ergative, or (perhaps more likely?) the ergative can be reinterpreted as a general agent case and replace the absolutive. Since neither spread results in any loss of information or creation of ambiguity, it should happen fairly easily.

I wonder whether (with the additional wrinkle of split ergativity / differential subject marking) this is what happened in PIE?

[Omzinesý]

Q
If verbs have classes that trigger agreement, what agrees with verb class? Does such even happen?

I’ve seen lit wherein it is simply called “verb class”; however, most uses of the phrase “verb class” do not refer to the concordial verb-class you’re talking about.

I wrote to Rachel Nordlinger about it.
See these threads:
http://cbbforum.com/viewtopic.php?p=29 ... s#p292785
http://cbbforum.com/viewtopic.php?p=29 ... s#p292800

Spoiler:

http://cbbforum.com/viewtopic.php?p=29 ... s#p292804
http://cbbforum.com/viewtopic.php?p=29 ... s#p292808
http://cbbforum.com/viewtopic.php?p=29 ... s#p292882

http://cbbforum.com/viewtopic.php?p=29 ... s#p292894
http://cbbforum.com/viewtopic.php?p=29 ... s#p293140

(You might want to look at the last-listed one first! Then the others in the order listed, or any order you want!)

I’ve called it “verbal gender”. That parallels ideas like “nominal aspect” (e.g. are you at Kyiv or in Kyiv?) and “nominal tense” (e.g. spouse, ex-spouse, spouse-to-be). On CONLANG-L some linguist said she liked my “nominal aspect” idea; I don’t know if Professor Nordlinger liked my “verbal gender” term, but at least she understood it.

I hope that helps, and isn’t TL;DR!

—————

Edit: P.S.: You might want to read http://cbbforum.com/viewtopic.php?p=30 ... s#p304938 as well. It has only two posts!

Thank you. I will read them and ask more if I still have questions. I had some memory that the topic had been discussed before but I didn't find anything.

[Arayaz]

Is there any natlang that historically replaced ambitransitive verbs' intransitive forms with the agent instead of the patient?
That sentence was word salad. Allow me to rephrase:

Assume an ergative language. A typical verb might be pyār, the lexical meaning of which is "bite." With one argument, it would have only the patient: jo pyār "the bone was bitten." With two arguments, it's the patient and then the agent, with the agent marked as ergative: jo kāvwa-y pyār "the dog bit the bone."

Now, let's say ergativity gets lost in some of the tenses. Perhaps the inflected future jo kāvway pyārghi is replaced with a construction that has both arguments in the absolutive: mēr jo pyārghi kāvway. Then maybe word-final diphthongs simplify, so the ergative marking is gone too.

Is it possible that then, jo pyār could be reanalyzed as having jo as the agent? (Which doesn't make much sense in this context; assume a sentient bone.) I ask because I want to get rid of ergativity in Hánásö.

While I was typing this, I was struck by inspiration and don't need to kill it this way, but I'd still like to know.

[Salmoneus]

This is unrelated to ergativity.

Univalent verbs derived from bivalent verbs can have either argument as their sole argument; and bivalent verbs can be derived from univalent verbs regardless of the semantic role of the sole argument. Cross-linguistically, that is; obviously individual languages can have specific rules of their own.

So English is nom-acc, and yet the bivalent verb "boil" (I boil water) has a univalent counterpart with which the sole argument has the same role as the patient of the bivalent verb (the water boils). Even though it has other verbs where the opposite is the case (I eat potatoes > I eat, not *the potatoes eat).

And some verbs can go either way. For instance, from the divalent "open", English gets two univalent verbs, one of which 'retains' the noun in the agent role of the divalent (the youngest player opened)(think of card games, for instance, and also some lectures/debates/etc) and one which retains the noun that's in the patient role for the divalent (the door opened).

--------

Ergativity means that the subject of an univalent verb takes the same case as the patient of a divalent verb (for a significant number of verbs, in a significant number of contexts).

It doesn't inherently have anything to do with derivational processes!

--------

It is true that there's also a tendency for verbs not to elide the unmarked argument, but that again is a separate thing and not universally the case. Old English, for instance, allowed unmarked agents to be elided, and that's actually not uncommon cross-linguistically. Other languages don't allow any elision at all.

Bear in mind that although in English the two processes are often superficially the same (to the extent that they can be hard to tell apart in some cases), argument elision is not necessarily the same as valency-reducing derivation cross-linguistically (or even arguably in English in some instances). Elision is driven by information structure (it might only be allowed when the elided object is clear from context, or only when it's entirely indefinite, for instance), whereas derivation is a lexical process. [I suspect that's a simplification in both directions, but it'll do for now]

[KaiTheHomoSapien]

Interesting. My conlang Arculese largely makes this valency distinction through the use of the active vs. middle voice:

Nardan soxti - he opens the door (ACTIVE)
Narda soxto - the door opens (MIDDLE)
Narda soxtor - the door is opened (by me). (PASSIVE)

"The player opens" equivalent would use the active voice here since the subject is an agent. The middle is inherently univalent.