Linguistic Miscellany Thread

[bradrn]

Okay, so que is some kind of relative pronoun, and you can have an invariant relativization marker in the gap strategy.

I wonder if the pronoun retention strategy would not be better (easier) in an IAL, even though that's uncommon. So you'd say something like "boy buy book, he [did something]" for "the boy who bought the book".

A terminological point: the pronoun retention strategy would be “boy [he buy book] do something”, retaining the pronoun in the relative clause. What you’re describing is a non-reduced relative clause.

[KathTheDragon]

And there's the matter of "whose", which is probably enough to count as non-gapping.

You mean, like el libro cuyo autor yo conozco ('the book whose author I know')?

Now that's interesting to know, and confounds the case somewhat.

no gapping, relative pronouns: the book the boy bought which

Is "which" a pronoun?

[Kuchigakatai]

Ah... this topic again.

I've mentioned this bit about Spanish and its gap strategy (as defined by WALS) a few times here on the ZBB (probably three), and how it's different from English, but it looks like everyone here has forgotten my usual explanation/rant (or maybe you guys never read it ). I think Salmoneus was the one that asked the first time.

I was working with an old IAL of mine, that I've never gotten far with, and decided to read about relative clauses on WALS. Apparently the gap strategy is by far the most common type. This is the type used by Japanese, so I thought this must be most common among SOV languages. But to my surprise it was more common in SVO langs. Spanish is listed as one of those languages. Could someone please show me how this works in Spanish, because I can't imagine how. (Note that the relativized noun must appear as the subject of the main clause.) Mandarin is supposed to do the same, so similar examples from Mandarin would also be welcome.

WALS defines the ‘gap strategy’ as occurring when ‘there is no overt case-marked reference to the head noun within the relative clause’ — that is, when the head noun is entirely omitted from the relative clause, leaving behind a gap. From those examples, it certainly looks like Mandarin and Spanish straightforwardly satisfy this.

If Spanish does, so does English.

You guys are getting too distracted by the relative pronoun alone, thinking the question is whether Spanish is like e.g. Latin (which has a very case-marked relative pronoun: nom. sg. quī/quae/quod, acc. sg. quem/quam/quod, dat. sg. cui, abl. sg. quō/quā/quō...). Look at the overall syntax. When the WALS' definition says it's "case marking", it's in the syntactician sense, hence why the definition says "within the [entire] relative clause", not just the pronoun.

Spanish has an ambiguity in the 3rd person that English lacks, thanks to English's general insistence on using subject pronouns if there's no NP subject:

la mujer que saludó
'the woman that said hi'
'the woman that [another person] said hi to'

el perro que mordió
'the dog that bit [him/her/it]'
'the dog that [another animal] bit'

Standard English always has (syntacticians') case marking within relative clauses, clearly distinguishing whether the head noun is the subject or a direct object of the subordinate verb (if it is transitive like saludar and morder are), but standard Spanish doesn't.

So yeah, under the WALS' definition, you can have gapping while still having a relative pronoun, as long as there's syntactic ambiguity kind of like that of Mandarin.

This is why I rant the WALS should really have explanations of why certain surprising choices have been made. Otherwise we get people misreading the short explanations given in the chapters. (And I do think the WALS contains mistakes, but that's a separate and natural concern.)

[Kuchigakatai]

Could someone please show me how this works in Spanish, because I can't imagine how. (Note that the relativized noun must appear as the subject of the main clause.) Mandarin is supposed to do the same, so similar examples from Mandarin would also be welcome.

For Mandarin:

問候的女人
wènhòu de nǚrén
greet REL woman
'the woman that said hi'
'the woman that I/you/he/she/we/they said hi to'

For the second translation, recall Mandarin is "pro-drop" (a term I dislike, is something really dropped? the POV is too English-y), in spite of having no subject agreement in verbs. So yeah, it can have ambiguity for the subject and object of the subordinated verb.

Relativized instruments are not marked in any special way. Here you just know the knife must be involved in the hero killing the monster somehow, and pragmatically fill it as an instrument. Consider that either yīngxióng 'hero' or 'guàishòu 'monster' or both could be absent ("the knife with which he/she killed it" ~ "the knife of the killing") I think this can also apply to other relativized roles, but less commonly.

英雄殺死怪獸的刀
yīngxióng shāsǐ guàishòu de dāo
hero kill monster REL knife
'the knife with which the hero killed the monster'

[zompist]

la mujer que saludó
'the woman that said hi'
'the woman that [another person] said hi to'

el perro que mordió
'the dog that bit [him/her/it]'
'the dog that [another animal] bit'

Those are good examples... I wasn't thinking of clauses with no overt argument. When there's an argument it does work more like English.

On the other hand... I just checked with my native speaker wife, and she doesn't like them. I asked her to translate (A) and she came up with (B):

A: We're looking for the wolf and the dog that it bit.
B: Estamos buscando al lobo y al perro al cual mordió.

I asked if she'd accept "Estamos buscando al lobo y al perro que mordió" and she said no, in part precisely because it's ambiguous.

IIRC WALS sometimes allows non-binary choices. I don't think they've correctly characterized either Spanish or English. "There are certain constructions that behave this way" is not the same as "This is how the language works." I understand that they're creating a massive database, but it's not like they're saving database size or anything. You can store which ones of several strategies are used in a single byte.

[Kuchigakatai]

Those are good examples... I wasn't thinking of clauses with no overt argument. When there's an argument it does work more like English.

On the other hand... I just checked with my native speaker wife, and she doesn't like them. I asked her to translate (A) and she came up with (B):

A: We're looking for the wolf and the dog that it bit.
B: Estamos buscando al lobo y al perro al cual mordió.

I asked if she'd accept "Estamos buscando al lobo y al perro que mordió" and she said no, in part precisely because it's ambiguous.

But the ambiguity is precisely the point! She might be thinking about stylistics there. Even stylistically, I suspect it might just be too little context, and she'd would accept "y al perro que mordió" if she was given the sentence within a story, say, if it is obvious the group is looking for the wolf and the dead dog after hearing the dog's owner crying.

IIRC WALS sometimes allows non-binary choices. I don't think they've correctly characterized either Spanish or English. "There are certain constructions that behave this way" is not the same as "This is how the language works." I understand that they're creating a massive database, but it's not like they're saving database size or anything. You can store which ones of several strategies are used in a single byte.

Yeah, I agree with that too. In the end, what Spanish has is not what is normally meant by gapping anyway, since all other roles are clearly distinguished via prepositions (al que, con el que, por el que, en frente del que...). What Mandarin has, with its simple bare relativizer, is the stereotype. And English sitting there next to Russian because of its syntax seems awkward. English and Spanish seem so similar to each other instead, in most respects.

[Qwynegold]

I've mentioned this bit about Spanish and its gap strategy (as defined by WALS) a few times here on the ZBB (probably three), and how it's different from English, but it looks like everyone here has forgotten my usual explanation/rant (or maybe you guys never read it )

Huh, I've never seen it. Thanks for the examples!

[bradrn]

So yeah, under the WALS' definition, you can have gapping while still having a relative pronoun, as long as there's syntactic ambiguity kind of like that of Mandarin.

So does this mean that English doesn’t use gapping? I always thought it did, but it doesn’t have any of that ambiguity.

[priscianic]

I'm confused why people seem to be assuming that Spanish que is a relative pronoun. Is the idea that there's a relative pronoun que that just so happens to be homophonous with complementizer que? Is there a principled reason why people are assuming that there are two different ques? The most parsimonious thing is to say is that que appears on all kinds of (finite) embedded clauses: complement clauses (decir que... ‘say that...’), temporal adjunct clauses (después de que... ‘before (of that)...’), clausal standards of comparison (más que... ‘more than...’), and also of course relative clauses.

Spanish and English are alike in that they have both gapped relative clauses: in English, with the general-purpose complementizer that (the cat that I saw Ø) or a null complementizer/nothing (the cat I saw Ø); in Spanish, with the general-purpose complementizer que (el gato que vi Ø ‘the cat that I.saw’). They are also alike in that they have a relative pronoun strategy as well, with a typical IE similarity between interrogative pronouns and relative pronouns: the child who I saw, el niño a quien vi ‘the child to whom I.saw’. Spanish additionally has what are known as "light-headed relatives", which involve a determiner on top of a relative clause: el que/cual ‘the that/which’.

(Of course, the distribution of the gapping vs. relative pronoun (vs. light-headed relative) strategies is another matter, and Spanish and English differ wrt the syntactic/semantic environments in which each strategy is chosen.)

When the WALS' definition says it's "case marking", it's in the syntactician sense, hence why the definition says "within the [entire] relative clause", not just the pronoun.

I'm firstly not sure what you mean by "syntactician's case marking". WALS is clear that when they say case-marking, they mean case-marking by some overt morpheme: "the position relativized is indicated inside the relative clause by means of a clause-initial pronominal element, and this pronominal element is case-marked (by case or by an adposition) to indicate the role of the head noun within the relative clause" (emphasis mine).

The reason why WALS insists on being case-marked "to indicate the role of the head noun within the relative clause" is because the relative head is in some sense "in two places at once": it plays a role in the relative clause, and it plays a role in the matrix clause, and you could in principle imagine the (overt, morphological) case-marking on the relative pronoun to either reflect the case the relative head would get in the relative clause, or the case the relative head would get in the relative clause. To illustrate the difference, they provide an example from MSA where the relativizer is inflected for the case that relative head gets in the matrix clause, and for them that's not a "relative pronoun strategy".

I'm not sure why WALS decided that English counts as having the relative pronoun strategy but Spanish counts as having the gapping strategy; both languages have both strategies. (I'm also not sure why they made the choices they did dividing up the different relativization strategies or why they think languages can be divided up like this when many languages make use of multiple strategies...)

[zompist]

I'm confused why people seem to be assuming that Spanish que is a relative pronoun. Is the idea that there's a relative pronoun que that just so happens to be homophonous with complementizer que?

No one has said there are two words que, but yeah, people have not been distinguishing between relative pronouns and complementizers.

OTOH, is it that clear what's a pronoun and what's not? E.g. in English we have

This is the key which/that launches the missiles.
This is the officer who/that holds the key.

Which/who/that all have the same surface syntactic position, and it's one suitable for either role. Are they Comps or NPs?

I tend to think "that" is a complementizer, and que too, because they clearly are elsewhere ("I think that the officer is loyal"). But that's not a very strong argument; homonyms do exist. (Including, um, "that", as in the last sentence!)

Maybe it's relevant that in English, we have an animacy distinction between who/which. Complementizers don't do that. (At least I think they don't. Just because I can't think of one offhand doesn't mean much. You can have gender on some pretty weird things.) We also have who/whom, but for a lot of speakers "whom" doesn't exist.

Mandarin de is much easier, since its syntactic behavior is entirely unlike a pronoun.

I'm not sure why WALS decided that English counts as having the relative pronoun strategy but Spanish counts as having the gapping strategy; both languages have both strategies. (I'm also not sure why they made the choices they did dividing up the different relativization strategies or why they think languages can be divided up like this when many languages make use of multiple strategies...)

Yes, I agree.

[Zju]

Are there languages, in which transitive verbs differ in morphology from intransitive ones, other than having markers for transitivity and/or direct object?
Tangentially related, but weird verbal morphology can be observed in Georgian, where some person-number combinations are expressed with prefixes, others with suffixes, and yet other with circumfixes.

[bradrn]

Are there languages, in which transitive verbs differ in morphology from intransitive ones, other than having markers for transitivity and/or direct object?

Two clear cases:

• In direct-inverse languages, only transitive verbs can (and occasionally must) take direct and inverse markers
• Passive, antipassive etc. are usually restricted to transitives

I’m sure there are many more examples, but these are the ones I thought of first.

Tangentially related, but weird verbal morphology can be observed in Georgian, where some person-number combinations are expressed with prefixes, others with suffixes, and yet other with circumfixes.

Interesting! I shall have to have a look at this.

[Kuchigakatai]

Are there languages, in which transitive verbs differ in morphology from intransitive ones, other than having markers for transitivity and/or direct object?

Well... Mandarin monosyllabic transitive verbs often take a monosyllabic generic object, or need to be used in an extended form (where the extra morpheme is a synonym, or reduplication, or some sort of further direction or resultative morphemes...) to become intransitive. Classic example would be 吃 chī 'eat sth' and 吃飯 chī-fàn 'eat' (where fàn = 'rice' as a generic object for any food). I'd also mention 笑 xiào 'to laugh at sth' and 笑起來 xiàoqǐlái 'to laugh' (where -qǐlái is 'up and coming', not really meaning much here except vaguely inchoative).

But note this isn't necessary if the intransitive verb is given weight via an adverbial or a following result clause or even just a negator... 不吃 bù chī (gloss: "not eat") 'sb isn't eating, didn't eat' is alright for the intransitive usage, and so is 吃得很快 chī de hěn kuài ("eat ADV very fast") 'eat fast'. You could also say e.g. 大笑 dà xiào 'to laugh hard'. The real reason why a generic object like 飯 fàn 'rice' is needed with 吃 chī 'to eat' has more to do with prosody than transitivity, because once you remove the direct object, the monosyllabic main verb is felt to not be heavy enough to end a sentence. Aren't non-European languages just weird and fun?

[Rounin Ryuuji]

Aren't non-European languages just weird and fun?

Yes.

[priscianic]

I'm confused why people seem to be assuming that Spanish que is a relative pronoun. Is the idea that there's a relative pronoun que that just so happens to be homophonous with complementizer que?

OTOH, is it that clear what's a pronoun and what's not? E.g. in English we have

This is the key which/that launches the missiles.
This is the officer who/that holds the key.

Which/who/that all have the same surface syntactic position, and it's one suitable for either role. Are they Comps or NPs?

I tend to think "that" is a complementizer, and que too, because they clearly are elsewhere ("I think that the officer is loyal"). But that's not a very strong argument; homonyms do exist. (Including, um, "that", as in the last sentence!)

Maybe it's relevant that in English, we have an animacy distinction between who/which. Complementizers don't do that. (At least I think they don't. Just because I can't think of one offhand doesn't mean much. You can have gender on some pretty weird things.) We also have who/whom, but for a lot of speakers "whom" doesn't exist.

This is an important question! You're entirely right that it's not entirely obvious what complementizery things and relative pronouny things are, whether a given element in a given language is a complementizery thing or a relative pronouny thing, etc.

As far as I understand things, the standard assumptions/analysis is as I laid out: that and que are complementizers, and who, which, whose etc. and quien, cuyo, cual, etc. are relative pronouns. However, people have gone different ways about this. Even just for English, people have argued that that is a relative pronoun, even arguing that when it appears in complement clauses, it's still a relative pronoun (Kayne 2010; "Why Isn't This a Complementizer" has to be one of my favorite paper titles of all time). Conversely, people have argued that who and which are actually complementizers that agree in animacy with the relative head (Pesetsky and Torrego 2006).

Sidenote: complementizer agreement, in various different guises (e.g. what exactly is the complementizer agreeing with? what features is it agreeing with?), is a thing in several typologically-distinct languages, e.g. Flemish (De Vogelaer, Neuckermans, and Vanden Wyngaerd 2002) Lubukusu (Diercks 2013), and Nez Perce (Deal 2015). So Pesetsky and Torrego's (2006) proposal isn't completely out of left field from a typological perspective (whether this analysis of putative "relative pronouns" in English is right is, of course, an independent matter)

[Richard W]

Are there languages, in which transitive verbs differ in morphology from intransitive ones, other than having markers for transitivity and/or direct object?

Your average European tends to allow past participles of intransitive verbs an active meaning; probably the extreme consequence is the avoir/être distinction in perfects as in French, though the 'avoir' construction tends to win out.

[Man in Space]

(Kayne 2010; "Why Isn't This a Complementizer" has to be one of my favorite paper titles of all time)

I misread "Kayne" as "Kanye"…

No one man should have all that power
Relative clauses that go on for hours
Reading the papers, get older and wiser
Why isn't "this" a complementizer

[Raphael]

Newbie question: is there any historical relation between German Loch "a hole" and Scottish Gaelic loch "a type of lake" (that is, arguably, "a big hole in the ground with water in it"), or is that just a standard chance resemblance?

[Ares Land]

Apparently they're not.

German Loch is on the other hand a cognate of English lock.

[Raphael]

Thank you!