Zompist Bboard Again

Pabappa wrote: ↑Fri Feb 14, 2020 1:39 pm The Malagasy wiktionary is one of the best overperformers relative to its language's number of speakers...

Surprisingly, that was basically all the work of a single person. I once got interested enough to look it up, and it turns out that this one guy worked for five years and eventually made a series of bots which translated other articles into Malagasy (which explains why there are so many Volapük verb forms, which were really easy to translate en masse) or imported them from online dictionaries. They tell the story in detail here.

Xephyr once said Pali struck him as a Sanskrit that sounded more like a normal language, with most of the unusual clusters smoothed out. Knowing nothing about either language, I happened to come across some Buddhist terms in both languages side by side, and I was reminded of that comment...

dṛṣṭi ~ diṭṭhi
smṛti ~ sati
saṃkalpa ~ sankappa
karman ~ kammanta

The Wikipedia article on Pali provides similar examples.

vṛkṣa ~ rukkha
duḥprajña ~ duppañña
niḥsattva ~ nissatta
nirvāṇa ~ nibbāna
īśvara ~ issara

Xwtek wrote: ↑Sun Feb 16, 2020 7:43 amAlso, I noticed that I used the modal particle evidentially. In sentence:

"The snakes should have coordinated with the city before doing orgy. Now, everyone is panicked"

I hesitated to use:

"The snakes should have been coordinating with the city before doing orgy. Now, everyone is panicked"

instead because the latter implies that the snake is evidently coordinating with the city unless anything weird happened instead of implying that it's better to do the action, but it's not done instead. I'm not sure about this, so which sentence do you think it's the better one?

No, the difference between "should have coordinated" and "should have been coordinating" is purely aspectual. The former is perfective and focuses on there being a "bound" (an end, a finishing limit) to the event. The latter focuses on the coordination efforts as a continuous process, likely taking several days.

"The snakes should have coordinated with the city before doing an orgy. Now, everyone is in a panic."

"Everyone is panicked" might be okay, but is not at all the normal way of saying this... "Panicked" is usually a participle used as an adverbial, I think, like, "Panicked, I immediately went home to check on my mother". Otherwise it's the past tense of "to panic": "She panicked when I told her about the problem".

(EDIT: Sorry, I didn’t see there was a new page, and posted this without realising that Ser had already answered. I just hope this answer is helpful too!)

Xwtek wrote: ↑Sun Feb 16, 2020 7:43 am Also, I noticed that I used the modal particle evidentially. In sentence:

"The snakes should have coordinated with the city before doing orgy. Now, everyone is panicked"

I hesitated to use:

"The snakes should have been coordinating with the city before doing orgy. Now, everyone is panicked"

instead because the latter implies that the snake is evidently coordinating with the city unless anything weird happened instead of implying that it's better to do the action, but it's not done instead. I'm not sure about this, so which sentence do you think it's the better one?

The first sentence is entirely correct (except you should put an indefinite article before orgy), and has no evidential meaning; instead, the modal particle should means that the event (i.e. the snake’s coordination with the city) is obligated to have taken place, but may not have been done. (Recall that an evidential marks the evidence which the speaker has for a proposition — visual, auditory, hearsay etc.) The second sentence is also grammatically correct, but feels less natural than the first one; the difference is purely between perfective aspect (sentence 1) and imperfective aspect (sentence 2). Since we’re not at all interested about the internal structure of the snake’s coordination with the city — we simply want to know that it should have happened — it is more natural to use the perfective aspect, which is devoid of internal structure.

Ser wrote: ↑Sun Feb 16, 2020 3:47 pm No, the difference between "should have coordinated" and "should have been coordinating" is purely aspectual. The former is perfective and focuses on there being a "bound" (an end, a finishing limit) to the event. The latter focuses on the coordination efforts as a continuous process, likely taking several days.

…

"Everyone is panicked" might be okay, but is not at all the normal way of saying this... "Panicked" is usually a participle used as an adverbial, I think, like, "Panicked, I immediately went home to check on my mother". Otherwise it's the past tense of "to panic": "She panicked when I told her about the problem".

Weird… I’m a native speaker (of Australian English, I think), and these are both different for me. I see both ‘should have coordinated’ and ‘should have been coordinating’ as having natural end points, since they’re both perfects. And I’m absolutely fine with ‘everyone is panicked’, but I have never heard ‘panicked’ being used as an adverb, and such a usage is incorrect for me.

[EDIT: and *I* wrote *this* before seeing bradrn's posts!]

"Now everyone is panicked" didn't strike me as odd when I read it, so I would say it's okay, though I'd prefer "now everyone is panicking." I personally wouldn't say "everyone is in a panic," which sounds weird and other-dialect-y (for reference, I speak AmE). Also, at least in my dialect one doesn't "do" an orgy, you "hold" or "have" one. So for me the most natural phrasing would be "The snakes should have coordinated with the city before holding the/an [either works, depending on the fuller context] orgy. Now everyone is panicking/panicked."

But yes, the distinction between "have coordinated" and "have been coordinating" is basically aspectual, there's no difference in evidentiality. I would say the sense which "have been coordinating" gives, in this instance, is that it emphasizes that the snakes should have been doing this all along (by suggesting that there would have been many steps involved in a proper coordination and that it would have been carried out over a period of time), and thereby highlights a bit more the snakes' irresponsibility. But it's subtle. [EDIT: Both "have coordinated" and "have been coordinating" are fine for me.]

EDIT: I also agree with bradrn that "panicked" as an adverb like that sounds very weird and is not something I use in speech; it sounds like one of those really contrived example sentences out of a language textbook or linguistics paper that no one would ever say in real life. (It is okay in literary usage, though, but strikes me as high-register and a bit archaic.)

Well, then it's just me that thinks Past Modal/must + have + been + Ving as inferential, instead of imperfective.

Also for some reason can + have is ungrammatical to me except when used as "You can have this food."

Xwtek wrote: ↑Sun Feb 16, 2020 5:31 pm Well, then it's just me that thinks Past Modal/must + have + been + Ving as inferential, instead of imperfective.

Yes, I think that’s just you. I think part of the issue is that many English modals can’t be used in the past tense, so the expression of a past modal meaning requires the use of the perfect.

Also for some reason can + have is ungrammatical to me except when used as "You can have this food."

Could you give an example?

Xwtek wrote: ↑Sun Feb 16, 2020 5:31 pmAlso for some reason can + have is ungrammatical to me except when used as "You can have this food."

I have a problem with can + perfect too:

The snakes should have coordinated with the city.
The snakes may have coordinated with the city.
The snakes could have coordinated with the city.
*The snakes can have coordinated with the city.

But this is fine:

The snakes can't have coordinated with the city.

Can + passive is OK (I can be persuaded).

Can + progressive seems tricky.

?I can be going to Vyat right now.
While you sand that down, I can be thinking about the next step.

I wish I'd run into this while writing my syntax book— it's pretty neat.

zompist wrote: ↑Sun Feb 16, 2020 6:40 pm

Xwtek wrote: ↑Sun Feb 16, 2020 5:31 pmAlso for some reason can + have is ungrammatical to me except when used as "You can have this food."

I have a problem with can + perfect too:

The snakes should have coordinated with the city.
The snakes may have coordinated with the city.
The snakes could have coordinated with the city.
*The snakes can have coordinated with the city.

But this is fine:

The snakes can't have coordinated with the city.

Can + passive is OK (I can be persuaded).

Can + progressive seems tricky.

?I can be going to Vyat right now.
While you sand that down, I can be thinking about the next step.

I wish I'd run into this while writing my syntax book— it's pretty neat.

Interesting! I didn’t think about those examples, but I have the same problem with these sentences as well.

zompist wrote: ↑Sun Feb 16, 2020 6:40 pm Can + passive is OK (I can be persuaded).

Can + progressive seems tricky.

?I can be going to Vyat right now.
While you sand that down, I can be thinking about the next step.

I wish I'd run into this while writing my syntax book— it's pretty neat.

"Can + progressive" and "can + prospective" do not really work for me (your examples sound ungrammatical to me), even though "can + passive" works for me.

Whimemsz wrote: ↑Sun Feb 16, 2020 4:31 pmAlso, at least in my dialect one doesn't "do" an orgy, you "hold" or "have" one.

Same. "Do an orgy" sounds as off to me as "do sex".

So according to Elbert, one interesting feature of Hawai'ian is that, when deictics are paired, the distal always comes first. For example, i kēlā me kēia lā "daily" (lit. "on that and this day"), kēlā mea kēia mea "everyone; whoever" (lit. "that one, this one"), i ʻō i ʻaneʻi "here and there" (lit. "yonder and here"). How common is this?

Linguoboy wrote: ↑Mon Feb 17, 2020 11:21 am Same. "Do an orgy" sounds as off to me as "do sex".

Accoding to the context, exactly. The city is panicked because of the snakes having sex. https://www.reddit.com/r/BrandNewSenten ... too_weird/

zompist wrote: ↑Sun Feb 16, 2020 6:40 pm Can + passive is OK (I can be persuaded).

To me, it's not only OK, it's also common and quite important. However, I'm a native speaker of Indonesian, where such construction is more common.

For example, this sentence is grammatical (aside of definite article, which I'm not sure):
(If you go there,) you can be killed by monsters there.
However, this sentence is doubful (aside of definite article, again):
?(If you go there,) monsters there can kill you.

The latter is more grammatical if you're describing a place, instead of cautioning a naive adventurer.

(Also, I heard someone saying Indonesian has some aspects of Obviative marking, but I can't get the citation)

I read a paper that says that a language spoken by millions of native speakers tends to lack object agreement and applicatives. But I forgot the exact paper. Could you tell me the exact paper?

Xwtek wrote: ↑Thu Feb 27, 2020 10:42 pm I read a paper that says that a language spoken by millions of native speakers tends to lack object agreement and applicatives. But I forgot the exact paper. Could you tell me the exact paper?

I’ve never heard of this claim, but it sounds suspicious to me. I imagine this would only be the case because by far the most widely spoken languages are either Indo-European, East Asian, Afroasiatic, Austronesian or Dravidian, all of which are families or areas which happen to lack these characteristics.

Xwtek wrote: ↑Thu Feb 27, 2020 10:42 pm I read a paper that says that a language spoken by millions of native speakers tends to lack object agreement and applicatives. But I forgot the exact paper. Could you tell me the exact paper?

Languages spoken by millions of people also tend to lack evidentiality, fourth person pronouns, and the phrase "isht taloa."

Moose-tache wrote: ↑Fri Feb 28, 2020 5:24 am

Xwtek wrote: ↑Thu Feb 27, 2020 10:42 pm I read a paper that says that a language spoken by millions of native speakers tends to lack object agreement and applicatives. But I forgot the exact paper. Could you tell me the exact paper?

Languages spoken by millions of people also tend to lack evidentiality, fourth person pronouns, and the phrase "isht taloa."

Well, the vast majority of languages lack these features, so it’s no surprise that the most widely-spoken languages also lack them. (This also applies to applicatives.) On the other hand, the majority of languages with verbal agreement do agree with objects (see my ergativity post for some WALS data), so it is at least reasonable to ask about that.

bradrn wrote: ↑Thu Feb 27, 2020 11:31 pm

Xwtek wrote: ↑Thu Feb 27, 2020 10:42 pm I read a paper that says that a language spoken by millions of native speakers tends to lack object agreement and applicatives. But I forgot the exact paper. Could you tell me the exact paper?

I’ve never heard of this claim, but it sounds suspicious to me. I imagine this would only be the case because by far the most widely spoken languages are either Indo-European, East Asian, Afroasiatic, Austronesian or Dravidian, all of which are families or areas which happen to lack these characteristics.

On object agreement, the claim seems like complete nonsense. To start with, it ignores Arabic, which has object agreement, as did the earlier imperial languages Akkadian and Aramaic. Spoken French arguably has a verbal complex that includes object agreement— /ʒnətevypa/ "I didn't see you". Of the imperial languages of the New World, Quechua and Nahuatl— and it's historical accident that tens of millions speak Spanish rather than those— both have object agreement.

Even if the original paper does some statistics on this, my guess is that they do so improperly, because languages are not independent variables. You can't apply techniques that apply to independent variables to a database where the variables are starkly, defiantly dependent. (E.g., as brad points out, a huge number of major languages are all I-E.)

Applicatives in the strict sense (using a dedicated morpheme) are probably just uncommon. Even here, Swahili has one. And a syntactic applicative is probably not uncommon— cf. English we walked a mile and a half, where a measure (usually expressed with "for") has been promoted to direct object.

It sounds a bit like arguments I associate with Trudgill and McWhorter, that relatively isolated languages with relatively few speakers are more likely to preserve certain kinds of complexity; or conversely that certain kinds of simplicity arise pretty much only when a language gets large numbers of adult learners. (I'm not endorsing the arguments.)

On applicatives, double object construction of the sort you have in English are widely analysed as involving a covert applicative, as are experiencer constructions in many languages. I think if you buy that sort of analysis they end up being pretty common.