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

A forum for discussing linguistics or just languages in general.
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

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

Post by sangi39 »

Khemehekis wrote: 04 May 2022 03:27 Is the onio- in the word "oniomania" cognate with the English word "own"?
Digging through Wiktionary suggests not. "Own" goes back to *h₂eyḱ- while the "onos" in "onio-" goes back to PIE *wósn̥ (from the root *wes-)
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
Khemehekis
mongolian
mongolian
Posts: 3883
Joined: 14 Aug 2010 09:36
Location: California über alles

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

Post by Khemehekis »

sangi39 wrote: 04 May 2022 09:27 Digging through Wiktionary suggests not. "Own" goes back to *h₂eyḱ- while the "onos" in "onio-" goes back to PIE *wósn̥ (from the root *wes-)
Oh, all right. I wondered. Seems like a false cognate!
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 86,336 words and counting

31,416: The number of the conlanging beast!
User avatar
Nel Fie
cuneiform
cuneiform
Posts: 139
Joined: 23 May 2022 15:18

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

Post by Nel Fie »

Are there any differences or patterns as to how tonal languages place tones in relationship to the speaker's natural "rest" pitch?

I've been told by a speaker that Mandarin uses the voice's rest pitch for the lowest parts in its tones (amongst much other useful information). But do all tonal languages do it that way? Or are there tonal languages that use the speaker's rest pitch for a mid tone - going beneath it for a low tone, for example?
:deu: Native (Swabian) | :fra: Native (Belgian) | :eng: Fluent | :rus: Beginner
DeviantArt | YouTube | Tumblr
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

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

Post by Creyeditor »

I think this question can be formulated in term of markedness. Is the low tone always the unmarked tone? The answer is: no, in some languages the mid tone is unmarked.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

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

Post by eldin raigmore »

Creyeditor wrote: 26 May 2022 13:09 I think this question can be formulated in term of markedness. Is the low tone always the unmarked tone? The answer is: no, in some languages the mid tone is unmarked.
But, do languages with only two tones, namely a low level tone and a high level tone, always make the low one the unmarked one?
I’m pretty sure the overwhelming majority do, but I could be wrong about nearly anything.
User avatar
Nel Fie
cuneiform
cuneiform
Posts: 139
Joined: 23 May 2022 15:18

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

Post by Nel Fie »

Creyeditor wrote: 26 May 2022 13:09 I think this question can be formulated in term of markedness. Is the low tone always the unmarked tone? The answer is: no, in some languages the mid tone is unmarked.
Good point. It would certainly make sense, thank you.
Last edited by Nel Fie on 26 May 2022 19:59, edited 1 time in total.
:deu: Native (Swabian) | :fra: Native (Belgian) | :eng: Fluent | :rus: Beginner
DeviantArt | YouTube | Tumblr
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

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

Post by Creyeditor »

Right, so I think googling[ "unmarked high" tone phonetics] might give you examples of languages that have been argued to have an unmarked high tone.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

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

Post by Salmoneus »

Creyeditor wrote: 26 May 2022 19:58 Right, so I think googling[ "unmarked high" tone phonetics] might give you examples of languages that have been argued to have an unmarked high tone.
I haven't googled this, but I know there are even languages in which stress is marked by a lower tone, so I'd have thought it almost certain that regular low tone is sometimes considered marked.
User avatar
Nel Fie
cuneiform
cuneiform
Posts: 139
Joined: 23 May 2022 15:18

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

Post by Nel Fie »

Salmoneus wrote: 26 May 2022 20:49
Creyeditor wrote: 26 May 2022 19:58 Right, so I think googling[ "unmarked high" tone phonetics] might give you examples of languages that have been argued to have an unmarked high tone.
I haven't googled this, but I know there are even languages in which stress is marked by a lower tone, so I'd have thought it almost certain that regular low tone is sometimes considered marked.
First off all, apologies for my last reply, it ended up a bit nonsensical due to me getting a few wires crossed in the process of writing it.

That said, I searched and found what seems like a good paper on the topic: Markedness and the Phonological Typology of Two-Height Tone Systems, by Larry M. Hyman

It certainly confirms that in a two-tone system either tone can be marked, although marking of the high tone seems more common. Beyond that, I'm out of my depth in terms of theory here, and it'll probably take a while and a few rereadings to fully digest and understand the paper itself. However, based on what I can gather, it seems that markedness and production of tone are not inherently tied, so the initial question might be back to square one.
eldin raigmore wrote: 26 May 2022 16:36 But, do languages with only two tones, namely a low level tone and a high level tone, always make the low one the unmarked one?
I’m pretty sure the overwhelming majority do, but I could be wrong about nearly anything.
As per my understanding of Creyeditor, Salmoneus and the paper linked above, either tone can be marked in a two tone system. It seems almost by necessity in some cases (unless I'm completely misunderstanding the terminology) as Hyman writes of languages with a low tone and a neutral tone. In such a case, I'd assume only the low tone could be marked - and I think Hyman makes that very point in the conclusion?
:deu: Native (Swabian) | :fra: Native (Belgian) | :eng: Fluent | :rus: Beginner
DeviantArt | YouTube | Tumblr
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

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

Post by eldin raigmore »

How did the name “Vandal” evolve into the name “al-Andalus” ?
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

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

Post by Salmoneus »

... you're asking how /wandalus/ became /andalus/? Well, I suspect the route was the dropping of /w/.
User avatar
Nel Fie
cuneiform
cuneiform
Posts: 139
Joined: 23 May 2022 15:18

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

Post by Nel Fie »

Is it even certain that the two are etymologically linked? What I could find suggest that it's more of a "best guess" that happened due to lack of actual information.

Wikipedia has an article on it, but the papers in reference 4 and 7 might be of more specific interest: https://en.wikipedia.org/wiki/Name_of_Andalusia
:deu: Native (Swabian) | :fra: Native (Belgian) | :eng: Fluent | :rus: Beginner
DeviantArt | YouTube | Tumblr
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

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

Post by eldin raigmore »

Salmoneus wrote: 30 May 2022 13:32 ... you're asking how /wandalus/ became /andalus/? Well, I suspect the route was the dropping of /w/.
I am on another forum where a group are discussing this question. One of them suggested the same thing Salmoneus suggests above. Someone else doubted it.


Nel Fie wrote: 30 May 2022 18:56 Is it even certain that the two are etymologically linked? What I could find suggest that it's more of a "best guess" that happened due to lack of actual information.

Wikipedia has an article on it, but the papers in reference 4 and 7 might be of more specific interest: https://en.wikipedia.org/wiki/Name_of_Andalusia
On that other forum no-one seems to doubt that Vandal and Andalus are etymologically related.
But there’s some question about, in which language the sound-change occurred, or in case there was more than one step, in which languages the various steps occurred.



…. …. …. …. …. …. …. …. …. ….

Thank you both for your responses!
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

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

Post by Salmoneus »

Factually, it's clear that that sound change did occur (assuming that is the etymology). Because that's the only difference between the two forms.

Specifically, the sound change occurred in the process of borrowing from Romance (where the /w/ is retained) into Arabic (where so far as I know it's never existed). Or possibly in Berber, if that was an intermediary.

It's a sound change so trivial - the loss of a weak initial semivowel in the process of borrowing a placename between radically different languages - that it doesn't really need any explanation. However, Wikipedia's suggestion of a parsing confusion (Berber 'wandalus', a plain borrowing, being mistakenly assumed by Arabs to actually be 'w-andalus', using the Berber definite article) seems extremely plausible.
User avatar
Nel Fie
cuneiform
cuneiform
Posts: 139
Joined: 23 May 2022 15:18

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

Post by Nel Fie »

So, I've been reading up on polysynthetic languages. I'm going to keep reading, but if I don't choose to ask questions at some point, I probably never will.

1) Are there any natural polysynthetic languages where the polysynthesis is not driven by polypersonal agreement and other markings on verbs?

2) Are there any polysynthetic natlangs that allow for the nominalization of verbs? And if yes, what happens to the agreement and other markings - are they preserved on the nominalized verb, or do they get removed?

One point of curiositiy here is whether such a language would then have to express the same idea in a more analytic way - if it even has the option to do so. Of course, I'd imagine both questions become kind of difficult to answer for languages where there is no clear noun-verb distinction.
:deu: Native (Swabian) | :fra: Native (Belgian) | :eng: Fluent | :rus: Beginner
DeviantArt | YouTube | Tumblr
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

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

Post by Salmoneus »

Nel Fie wrote: 12 Jun 2022 10:37 So, I've been reading up on polysynthetic languages. I'm going to keep reading, but if I don't choose to ask questions at some point, I probably never will.

1) Are there any natural polysynthetic languages where the polysynthesis is not driven by polypersonal agreement and other markings on verbs?

This is largely a question of definitions. "Polysynthesis" is often used to mean a particular style or feel of language characterised by polypersonalism, noun incorporation, and extensive affixation on the verb.

Alternatively, we can in theory define polysynthesis in abstract terms, as simply a high level of synthesis. But this is more problematic than it seems, because how do you measure synthesis? Is it the average morphemes per word? Is it the average morphemes on the most synthetic word in an average clause? Or is it the most morphemes that can exist on one word - or the most that does exist in an average conversation? In which case, how long a conversation? These questions matter, because many languages not traditionally seen as polysynthetic are nonetheless capable of extreme synthesis (as anyone who has talked to an over-pseudoantidisestablishmentarianist-hating theologian can testify!), of nouns and/or verbs; on the other hand, many traditionally 'polysynthetic' languages will have their average morpheme-per-word ratios dragged down by isolating nouns; and indeed, many traditionally polysynthetic languages will in practice produce a large percentage of sentences in which no word happens to have a high degree of synthesis (most newbie-scaring examples are artificially-concocted potential words, rather than words that will actually obligatorily be used repeatedly in any given conversation). And then of course what level of synthesis (however we define it?) is sufficiently high?

So depending on the exact definitions, a purely statistical definition could leave you with a vast number of polysynthetic languages of all natures, OR an extremely small number of polysynthetic languages that probably have a lot in common.

That said, there are certainly 'traditionally polysynthetic' languages that lack polypersonal agreement - or even any argument agreement of any sort. Iirc Athabaskan languages are often examples? Certainly a bunch of north american languages.

There are no traditionally polysynthetic languages that lack a vast number of markings on verbs, because that's traditionally how polysynthesis is defined. There are languages with quite a lot of synthesis on nouns and not much on verbs.

However, in general it's hard to get up to really 'polysynthetic' levels of affixation on nouns without getting a fair bit on verbs too, just because so many more categories are naturally marked on verbs. And also, if polysynthesis is measured by the synthesis of the most synthetic word in a clause, non-verb-y languages will lose out, because they will tend to spread their affixes over the many non-verbs, none of which will have as many affixes as the one verb of a highly verb-y language: a language that has three obligatory suffixes on every non-verbal word will have a high average synthesis level, but no word with as many morphemes as the verb in a language that has no suffixes on any non-verbal word but five potential suffixes on the verb. So, again, definitions.

That said, if you wanted to have a vey highly synthetic language in which that synthesis was mostly nominal, it would certainly be possible. Things that actually exist include:
- cases
- compound cases (eg one morpheme for location and another for how the noun's position relates to that location - so, "on" vs "in" plus "toward" vs "away from" would give you four compound cases)
- deictic markers on the noun (particularly distal vs proximal)
- clausal agreement (in which all or most words in a clause agree with a certain element - sometimes more than once per word!)
- overt noun class markers
- plurality
- semi-random 'thematic' morphemes to enable inflection
- suffixaufnahme (in which nouns with adjectival function agree with the case of their head noun as well as bearing their own case marking)
- nominal TAM marking
- possessed vs unpossessed markers
- agreement with possessor
- agreement with possessed
- agreement of actor nouns with implied patients (eg having "breaker [of plates]" have a different affix than "breaker [of human skulls]")
- possessive classifiers (showing the typical or actual function of the object for its possessor)
- numerical classifiers (additional morphemes indicating class when indications of number are present; note that numerical classification, possessive classification and conventional noun class can all be independent classification systems for the same noun)
- overt animacy marking
- definiteness marking
- specificity marking
- focus marking (potentially multiple focus types)
- topic marking
- obviation marking
- switch reference marking
- transitivity marking
- qualifiers
- demonstratives
- markers of degree or comparison
- verb incorporation
- noun incorporation for relative or deverbal nouns
- extensive compounding
- a vast array of derivational affixes
- fixation (in which an entire phrase or clause is re-interpreted as a reusable lexical item)
- possibly direct/inverse marking?
- a wide variety of clausal-adverbial and conjunctive things expressed by clitics that happen for reasons of word order to generally affix to a noun
- etc etc.

[the first seven of these are found in Northeast Caucasian, for instance.]

2) Are there any polysynthetic natlangs that allow for the nominalization of verbs?
I think so.
And if yes, what happens to the agreement and other markings - are they preserved on the nominalized verb, or do they get removed?
It probably depends, as it does in non-polysynthetic languages.
User avatar
Nel Fie
cuneiform
cuneiform
Posts: 139
Joined: 23 May 2022 15:18

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

Post by Nel Fie »

Salmoneus wrote: 12 Jun 2022 17:44 This is largely a question of definitions. "Polysynthesis" is often used to mean a particular style or feel of language characterised by polypersonalism, noun incorporation, and extensive affixation on the verb.

Alternatively, [...]
Thank you for the in-depth reply! Well, if my question lacked precision as to what kind polysynthesis I meant, and if you managed to provide such a wide range of different definitions and interpretations, at least it confirms that the world of polysynthetic languages is broader and more varied than the descriptions I read make it seem.
Salmoneus wrote: 12 Jun 2022 17:44[...] indeed, many traditionally polysynthetic languages will in practice produce a large percentage of sentences in which no word happens to have a high degree of synthesis (most newbie-scaring examples are artificially-concocted potential words, rather than words that will actually obligatorily be used repeatedly in any given conversation)[...]
That's something I've seen suggested or implied quite a few times. Though whether you're writing about newbie learners, or newbie conlangers, I have no doubt some of them would be thrilled at the idea of building such monoliths, ha ha!
Salmoneus wrote: 12 Jun 2022 17:44
2) Are there any polysynthetic natlangs that allow for the nominalization of verbs?
I think so.
And if yes, what happens to the agreement and other markings - are they preserved on the nominalized verb, or do they get removed?
It probably depends, as it does in non-polysynthetic languages.
It's very interesting - the few languages I'm familiar with seem to drop everything, aside from noun incorporation. Admittedly they are all rather analytic to begin with, but the idea of a nominalized verb that nonetheless preserves information of TAM, agreement, etc... is inspiring.
:deu: Native (Swabian) | :fra: Native (Belgian) | :eng: Fluent | :rus: Beginner
DeviantArt | YouTube | Tumblr
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

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

Post by Salmoneus »

Nel Fie wrote: 13 Jun 2022 18:11
Salmoneus wrote: 12 Jun 2022 17:44 This is largely a question of definitions. "Polysynthesis" is often used to mean a particular style or feel of language characterised by polypersonalism, noun incorporation, and extensive affixation on the verb.

Alternatively, [...]
Thank you for the in-depth reply! Well, if my question lacked precision as to what kind polysynthesis I meant, and if you managed to provide such a wide range of different definitions and interpretations, at least it confirms that the world of polysynthetic languages is broader and more varied than the descriptions I read make it seem.
Well, again, definitions. The world of what we might call "traditionally polysynthetic" languages - polypersonalism, noun incorporation, bunch of other affixes (TAM, directionals, etc) - is relatively narrow. The world of what we might call "statistically polysynthetic" languages - scoring high on at least some measure of synthesis - is very broad and varied indeed, but also poorly-defined and sort of arbitrary.
Salmoneus wrote: 12 Jun 2022 17:44[...] indeed, many traditionally polysynthetic languages will in practice produce a large percentage of sentences in which no word happens to have a high degree of synthesis (most newbie-scaring examples are artificially-concocted potential words, rather than words that will actually obligatorily be used repeatedly in any given conversation)[...]
That's something I've seen suggested or implied quite a few times. Though whether you're writing about newbie learners, or newbie conlangers, I have no doubt some of them would be thrilled at the idea of building such monoliths, ha ha!
To be clear, it's probably true that such words do crop up a lot more in such languages than they would in, say, English. Just less than you might expect on the basis of some over-excited descriptions of what it theoretically possible in a certain language. Though I should say, I don't actually know anything in detail about any polysynthetic language, so I could well be wrong. But, as you say, it's something I've seen said quite a few times.
Salmoneus wrote: 12 Jun 2022 17:44
2) Are there any polysynthetic natlangs that allow for the nominalization of verbs?
I think so.
And if yes, what happens to the agreement and other markings - are they preserved on the nominalized verb, or do they get removed?
It probably depends, as it does in non-polysynthetic languages.
It's very interesting - the few languages I'm familiar with seem to drop everything, aside from noun incorporation. Admittedly they are all rather analytic to begin with, but the idea of a nominalized verb that nonetheless preserves information of TAM, agreement, etc... is inspiring.
Well, there again, we're back to definitions. What makes a verb nominalized? The more verbal stuff you preserve in the process, the less likely someone is to describe the result as a noun!

For instance, many IE "infinitives" (eg Latin, Greek) inflect for TAM, and some (eg Portuguese) even inflect for person. This is a big part of why they're not called verbal nouns. But of course diachronically infinitives are just derived from old verbal nouns. In English, they can be governed by a preposition like a noun. In Old English, they could inflect for two cases (in very Old English and iirc Old Saxon, even for three - the usual dative and nominative/accusative but also a genitive). So why aren't infinitives nouns exactly? Likewise gerunds.

So really there's two different questions: first, can you use an inflected verb in a role normally reserved for nouns?; and second, will your local linguist admit that this is a noun, or will they insist that it's still a verb?
User avatar
Nel Fie
cuneiform
cuneiform
Posts: 139
Joined: 23 May 2022 15:18

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

Post by Nel Fie »

Salmoneus wrote: 14 Jun 2022 01:25 Well, again, definitions. The world of what we might call "traditionally polysynthetic" languages - polypersonalism, noun incorporation, bunch of other affixes (TAM, directionals, etc) - is relatively narrow. The world of what we might call "statistically polysynthetic" languages - scoring high on at least some measure of synthesis - is very broad and varied indeed, but also poorly-defined and sort of arbitrary.
Apoligies for misusing the term, ha ha! But it's good of you to provide the second one - I'm not sure it'll always pass without additional explanation, but I think "statistical polysynthetic" is probably a better descriptor for the kind of languages I - and possibly some of the material I've been reading - are thinking of.
Salmoneus wrote: 14 Jun 2022 01:25 To be clear, it's probably true that such words do crop up a lot more in such languages than they would in, say, English. Just less than you might expect on the basis of some over-excited descriptions of what it theoretically possible in a certain language. Though I should say, I don't actually know anything in detail about any polysynthetic language, so I could well be wrong. But, as you say, it's something I've seen said quite a few times.
Perhaps it is time I parse some texts in those languages then, to see what the average word length is.
Salmoneus wrote: 14 Jun 2022 01:25 Well, there again, we're back to definitions. What makes a verb nominalized? The more verbal stuff you preserve in the process, the less likely someone is to describe the result as a noun!
[...]
So really there's two different questions: first, can you use an inflected verb in a role normally reserved for nouns?; and second, will your local linguist admit that this is a noun, or will they insist that it's still a verb?
Good point. If the latter questions are rhetorical (apologies if they aren't), could it be said that there is a noun-verb gradient, then? And where one particular word falls on it depends on the language - perhaps even on case-by-case usage of that word in different phrase constructions - with additional provision for situations where the language doesn't distinguish between the two, and others where the word effectively "dual-classes"?
:deu: Native (Swabian) | :fra: Native (Belgian) | :eng: Fluent | :rus: Beginner
DeviantArt | YouTube | Tumblr
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

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

Post by Salmoneus »

Nel Fie wrote: 14 Jun 2022 18:03
Salmoneus wrote: 14 Jun 2022 01:25 Well, again, definitions. The world of what we might call "traditionally polysynthetic" languages - polypersonalism, noun incorporation, bunch of other affixes (TAM, directionals, etc) - is relatively narrow. The world of what we might call "statistically polysynthetic" languages - scoring high on at least some measure of synthesis - is very broad and varied indeed, but also poorly-defined and sort of arbitrary.
Apoligies for misusing the term, ha ha! But it's good of you to provide the second one - I'm not sure it'll always pass without additional explanation, but I think "statistical polysynthetic" is probably a better descriptor for the kind of languages I - and possibly some of the material I've been reading - are thinking of.
Oh, I don't think you're misusing the term! I actually think that what I've called the 'statistical' definition makes more sense given the surface meaning of the word ('lots of synthesis'), and the traditional definition closed linguists eyes to a lot of interesting phenomena. The only problem is that as linguists have (in some cases) become more sympathetic to this definition and looked at a wider scope of 'highly synthetic' languages, they've found it harder and harder to actually make any meaningful generalisations about them, or to robustly define them as a category of languages at all. Whereas the traditional definition is effectively hijacking what should be a broader term, and is weirdly parochial (why should we care so much about these specific languages? Why do they deserve to be treated as an entire 'type' of language on their own, while so many other weird and wonderful languages are lumped together?)... but on the other hand, people using that definition are at least using relatively well-defined concepts and talking about a group of languages that to some degree at least some generalisations can be made about.

The biggest problem historically was that a lot of people using the traditional definition didn't seem to realise that it wasn't the same as the 'statistical' definition. That is, people thought that all languages that were highly synthetic worked a certain way, and they don't. But once you realise they don't, it makes the whole concept of "polysynthetic language" as a category become of dubious usefulness. So some people seem to have reacted by retreating to their original definitions, and no-true-scotsmaning all other highly synthetic languages, so that they can keep talking about the languages that interest them, while others have accepted that the concept should be expanded, but as a result don't see much reason to keep using it...

At least, that's my impession, as a layman!

To be clear, though, I think there's a lot of fascinating synthesis out there, beyond the limits of the 'traditionally polysynthetic' languages! And we should be thinking less of distinct categories and more of continua - which means that even someone making a traditionally polysynthetic language could still take interesting things from less-traditionally highly synthetic languages...
Well, there again, we're back to definitions. What makes a verb nominalized? The more verbal stuff you preserve in the process, the less likely someone is to describe the result as a noun!
[...]
So really there's two different questions: first, can you use an inflected verb in a role normally reserved for nouns?; and second, will your local linguist admit that this is a noun, or will they insist that it's still a verb?
Good point. If the latter questions are rhetorical (apologies if they aren't), could it be said that there is a noun-verb gradient, then? And where one particular word falls on it depends on the language - perhaps even on case-by-case usage of that word in different phrase constructions - with additional provision for situations where the language doesn't distinguish between the two, and others where the word effectively "dual-classes"?
[/quote]

Absolutely there is a noun-verb gradient, in my opinion. We can see this even in English:
He finds cats
He was finding cats
This is my finding stick
He was a-finding cats
He likes to find cats
Him finding Mrs Malmesbury's cat saved the day!
In finding cats, he found mammals
I was in the middle of finding my keys when you distracted me!
The court's finding of legal liability was disappointing
What a great find!
Finding is better than keeping
My findings are as follows...
His discovery of sodium led to many complications

The first of these is clearly verbal; the last, clearly nominal. But where exactly you draw the line in between is not really clearcut. Traditionally, we say that these middle sentences variously involve a true participle ("finding"), a possibly-a-participle used as a periphrastic verbal construction ("finding"), a gerund ("finding"), an abstract noun ("finding"), a verbal noun ("finding"), another verbal noun ("find"), and an infinitive ("find"). And of these, the infinitive is said to be verbal, the gerund and participles are sometimes said to be verbal but sometimes only partly verbal, and the various nouns are said to not be verbal. But obviously these distinctions are not, prima facie, synchronically obvious. And in some cases they are not well-founded (no pun intended) diachronically either. They come from the combination of recognising slight differences in possible usage with our external knowledge of how certain other languages (chiefly Latin!) work. And the distinction between a gerund and a verbal noun in particular is extremely tenuous in actual spoken English [is the choice between 'his finding' and 'him finding' really as seismic a shift as moving from a noun to a verb!?]


-----

More to the point, though, there's actually THREE continua at play here.

First, there's the continuum of functions. In a given sentence, a given word can seem to have a more or less verbal function.
Then there's the continuum of forms. A form of a word (like 'finding') that appears in verbal functions may be called a verbal form - but if it also appears in nominal functions, should it be called a mixed noun-verb form, primarily a verb form, or primarily a noun form? Or should we say that there are two underlying forms, with different functions, that are just identical on the surface?
And then there's the continuum of roots. If a root produces both verbal and nominal forms, should we call it primarily a verbal or primarily a nominal root - or is it two identical roots?


EDIT: just to give an example from one of my own conlangs: Rawàng Ata doesn't explicitly set out to have no distinction between nouns and verbs. But there is some ambiguity around possession, because inalienable possession on a noun looks exactly the same as regular verbal agreement on a verb...

By roots here I don't even mean etymological roots, I mean lexical items themselves.

All languages have clearly nominal and verbal functions and I think they all have clearly nominal and verbal forms - though they may also have mid-continuum functions and forms. But there are some languages for which it has been argued that they have no distinction between nominal and verbal roots - that is, that any word can as easily (if not as frequently) be used verbally as nominally.
Post Reply