June 23, 2025 • 55 mins
William Annis has come back to the show to talk to us about subordinate clauses! This will be a broad overview of the topic, which we may dive deeper on in future episodes. Links and resources: Lingweenie: https://lingweenie.org/conlang/ Wikipedia on balancing and deranking: https://en.wikipedia.org/wiki/Balancing_and_deranking Dixon, Robert M. W., and Alexandra Y. Aikhenvald. Semantics of Clause... Read more »
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(01:00:00):
They tend to be overlooked, because if you
see what you think is a dependent clause floating
about on its own, it invites the interpretationthat the speaker's made an error.
If you're really focused on the rulesor-- and we'll get to this in a moment--
if you're focused on sentences inisolation from the broader conversation.
[MUSIC PLAYING]
(01:00:33):
Welcome to Conlangery, the podcast
about constructed languages andthe people who create them.
I'm George Corley.
And with me, the next town over,is William Annis.
After a long time.
Yes, yes.
He used to be down the road a ways,but I have moved.
So I'm not in Madison anymore.
(01:00:54):
I'm not going to tell you exactly where I am, though.
But this is going to be something
I haven't actually donesince we started video
episodes, but it'skind of a return to form.
I've been doing a lot of interviews of people.
And William used to bejust the regular co-host of
(01:01:19):
Conlangery, if you werehere for audio episodes.
And he's coming back to talkabout a linguistics topic again.
And hopefully, we can do-- I'll bedoing more interviews in the future,
but I'll also come back and do a little bit more of this,
(01:01:42):
because this is the thing that peoplereally liked about the audio podcast.
And I think it's good to introduce,
now that we have video folks on YouTube as well, to there.
But William, thank you for coming on.
(01:02:03):
It's been so long.
There's fancy new technology.
I had to sully my computer with Chrome,but I'm like, for George, I can do it.
Yes, I don't know why-- nobodydevelops these video platforms for Firefox.
I don't know.
I wish they would.
It's just Chrome has too muchpenetration, and they just go for that.
(01:02:25):
And they could get a bunch of different browsers with that.
But anyway, yes, we are here.
You can see William for once.
And I have a face for radio.
I will see if people who arelongtime audio listeners who
(01:02:51):
have not seen William's face-- Iwonder what your reaction is going to be.
Because any time I see someone I'mused to hearing in audio see their face,
it's like, that face--the face doesn't match
the voice, like almostevery single time.
Why is his head so shiny?
(01:03:11):
Well, hopefully I can adjust with color correction there.
OK, so anyway, what are we talking about today, William?
We're talking about subordination.
This was an idea I had a long time ago.
And then it took us many months to finally get to this.
There was a little COVID in there for me.
And in the course of making my usual slew of notes,
(01:03:37):
my opinion is there are a few of thesetopics that deserve their own show.
We've already done one on relative clauses.
But I think we could do an entireepisode on temporal clauses.
Absolutely, we could do anentire show on conditional clauses.
If we haven't already, we could probablydo it again and other sorts of things.
So today, we'll do atop-level overview of...
(01:04:00):
.. whatever it is subordination means.
Different kinds of meanings and semantics
those can have, different ways of forming those.
I have a lot on my favorite thing,insubordination, which sounds fun.
I like the idea of insubordinate grammar, which
is the term that people have settledon for an underdescribed feature where
(01:04:23):
subordinate clauses cut loose fromthe main clause and go off on their own.
And that's super interesting.
So that's what we'll cover today.
And then some of the things we can deep dive
into in later episodes,I think, would be
better, rather than tryto cover them all here.
Or we'll end up withanother one of our two-hour episodes,
which isasking a lot of people.
(01:04:44):
Yeah, we don't do that anymore.
That was early on.
Before we knew better.
OK, so what is a subordinateclause, basically?
We'll start with a simple definition.
And that's just when a sentence is crammed
(01:05:06):
inside another sentence, either as an argument,
like a subject or direct object, or a modifier.
And those modifier ones usually get called adverbial clauses.
And that's different from coordination,
where sentences just follow each other like ducks in a row.
"And," "or," and "but" are normally
the standard coordinating conjunctions and clauses.
(01:05:29):
But obviously, immediately, theoretical issues arise.
And it's not clear to me anymore that speaking of subordination
is useful as a distinct thing fromother kinds of clause combining.
(01:05:50):
So my favorite heroes, Dixon and Aikhenvald,
have a book called "The Semantics ofClause Linking" that covers all of them.
And that probably makes sense, because there is more overlap.
"And" can do surprisingly subordinating-likethings in many languages, which we're not
going to talk about too much today.
(01:06:11):
So the distinction isn't always maybeas clear-cut as we might think.
In some cases, it absolutely is, but...
And sometimes I'll talk about embedded clauses
versus subordination, or dependentclauses versus subordination.
These are all terms used todescribe basically the same thing.
(01:06:31):
The big thing is, normally, a subordinateclause has been modified in some way such
that it cannot normally go wandering off on its own.
Right.
So if you think about just English examples,relative clauses do this in English,
(01:06:52):
where, like, "I saw the man who," I don'tknow, "I saw the man who watched the movie."
I don't know.
That's a bad example.
But anyway, but "who watched the movie,"like that-- there's no subject pronoun.
(01:07:12):
The "who" is like a linking thingthat signals it's a relative clause.
And so that "who watched the movie," Imean, you could turn it into a question,
but it's not really an independent clause anymore.
It has to be attached.
There are sometimes--
(01:07:36):
Tests for this can be tricky in Englishbecause we murdered our subjunctive.
Yeah, yeah.
And there's also times where the subordinating, like,
complementizer-type thing is, like, deleted,
like a "that" is deleted, where I thinkthat something-- "I think that it was blue."
(01:08:01):
You say, "I think it was blue."
So in English, a lot of times, it's fuzzier.
In a lot of other languages, you would
have a different verb form or something else that's signaling.
But we'll get into all of those strategies.
Yeah, it can be even-- in some languages,it is extremely blunt and obvious.
In other languages, it's trickier.
(01:08:21):
I think Mohawk, I have a noteon this, where a subordinate
clause is marked offentirely by intonation.
[LAUGHTER]
So nothing, no conjunctions, nothing else.
And the main point here is that, in general,
subordinate clauses have to be fiddled with some way
to make them work within the main clause grammatically.
(01:08:43):
And we'll talk about different kinds of fiddling shortly.
And normally, parts of the subordinate clause
are not allowed to go wanderingabout other parts of the main clause.
But with relative clauses, you have
a pivot word, which is part of both clauses, in effect.
(01:09:07):
But in general, the two clauses are kept separate.
And something has been done to the subordinate clause
to make it fit in with the rest of the sentence.
It's a separate constituent, as people say.
It is its own thing.
It's a sentence embedded into a sentence.
And it has to be kept unified, generally.
(01:09:27):
Yeah.
So when it comes to different strategiesfor subordination, there's a lot of--
there's so many different kinds of terminology.
One set of terms that I'm familiar with,
so I will tend to use, is a balancedverb versus a deranked verb.
(01:09:50):
A balanced verb is a kind of verbclause that can be used independently.
And a deranked clause or verb cannot be used independently.
And there are lots of ways to derank a verb.
Could be a completely different conjugation type.
Could be a change in mood.
Could be a converb, participles, nominalization, extra goo.
(01:10:14):
Like jumping back into form, talking about ancient Greek.
Some kinds of ancient Greek subordinate clauses
require both a mood change and a particle in the clause.
Oh, OK.
So you have to-- yeah.
And we have links to cover all of these things.
(01:10:35):
We just have the Wikipedia link will coverthe basics of balancing versus deranking.
But yeah, so you can have a situation.
So it's both a particle and a mood change in ancient Greek.
Yeah, for some clauses.
For some clauses.
(01:10:56):
Some clauses, just it's a mood change.
Yeah.
I think a lot of times subjunctive--subjunctive is like a broad, ill-defined term.
But often, the subjunctive is used in this way
to mark that this clause is dependent on the other clause.
(01:11:19):
And that is also kind of deranking.
Yeah, yeah.
So the point about balanced here is,
especially in some kinds of temporal clauses in English,
like before and after, the sentencejust looks like a normal sentence.
"Before, you've got to talk to the doctor, do this."
That is a balanced clause using a conjunction
(01:11:42):
versus a deranked, which, "if I were to fall asleep..."
So the last vestiges ofthe English subjunctive
sometimes occur inconditional clauses.
And I mean, could you also consider
infinitive clauses in here?
Because we have that.
Well, right.
So an infinitive, I consider a kindof nominalization most of the time.
(01:12:07):
So yes, absolutely infinitives are part of that.
So things like infinitives, thingsthat are clearly deranked in English
are things like infinitives, gerunds, or any--
gerunds, participles, anything withthe -ing form, those kind of things.
Yeah, "Sleeping during a thunderstorm is very difficult."
(01:12:30):
So that's a deranked, "Sleeping during a thunderstorm."
It's a nominalized clause.
Yeah, yeah, yeah.
And I think that's an important thing to say,
is like, nominalization is a really easy pathway to this.
Because very often, these subordinateclauses are taking an argument position.
(01:12:51):
They are taking the same place as a noun.
So if you have a good distinction betweennouns and verbs in your language, then you
can stick some kind ofa nominalizer onto the verb,
and use that foryour subordinate clauses.
But you don't have to.
This is always-- always, this is options foryou to do, because languages are diverse.
(01:13:15):
They do all kinds of different things.
Yeah.
I'm trying to find a fun example I saw.
But I can't find it.
We've talked about conjunctions, many kinds of those.
I will confess, a lot of the time, myapproach is very boring for temporal clauses,
(01:13:38):
when they could be a lot more interestingthan I have been making them, typically.
Conjunctions, nominalization-- and nominalization
doesn't just have to mean doingsomething to turn the verb into a noun.
Like, we have the classic--we'll go back to another standby--
Mandarin, de shihou (的时候),for when, where you're just
(01:14:01):
taking a clause, and then a possessive orattributive marker, and the word for time.
So "When I saw my father," whatever.
Right.
In that case, it'ssort of-- so shihou.
So in that case, shihou, "time", is like the noun there.
(01:14:25):
And then you're adding a relative clause
for what the actual subordinate clause is there.
Whereas in English, we have more of a--so "When something, something happened,"
(01:14:47):
then that's sort of amore like-- just like a lot of
our other clauses with therelative pronoun and such.
Yeah.
So that's another strategy, is you
(01:15:08):
could come up with one kind of subordinating structure
and reuse it to make other subordinating structures.
Correct, yeah.
Yeah.
I couldn't find an online resource for this,
but Hixkaryana , which is famous for its default word order,
(01:15:28):
also has an enormous number of nominalization types.
Which at first, you're like,what a weird derivational system.
But then you realize it's almost entirely structural.
It's for grammar purposes.
It's not for creating new kinds of words.
Right, right.
So you're nominalizing verbs just to get that--
(01:15:53):
Right, so an example one is a manner nominalization.
The way you sneeze is not a special kind of word
you normally use, except to shove it into a clause.
The way he talks annoys me.
So a special nominalization just for manner.
Where English is-- in some sense, Englishis doing the same kind of nominalization
(01:16:15):
with a relative clause, the way in which he talks.
Yeah, the same kind of thing.
So one fun thing I just wanted to note--
I found this little paper hereon the semantics of perception.
And this looks like paper, like apresentation notes, rather than a full paper.
(01:16:37):
It's talking about verbs of perception
and the kinds of subordinate clauses they can take.
"I see that it is raining."
"I heard that he went to the store."
That sort of thing.
There's a fun thing in Spanish that differences
between vision and hearing can lead to grammatical changes.
(01:16:57):
So visual perception mainly involves object perception,
while auditory perception aligns more with event.
So in Spanish, this affects pronominal agreement
with perception verbs.
Visual perception tends to favor object agreement,
while auditory perceptionoften triggers event
agreement, which means itwill always be in the singular.
(01:17:18):
So if you look on--
this is on page 3 of this paper I linked to, George.
Where is that?
Which paper is that?
It's the one under nominalization,so above the big table.
OK.
So this one.
I already opened that.
So we have a document thatI have also shared-- this one?
(01:17:43):
OK.
So page 3.
OK.
So we have-- Sothe example is 8-9.
"Se veían desfilar ese día por las ..."
OK.
OK.
Oh, OK.
So you have-- so the use of seeing versushearing, it changes the meaning here.
(01:18:07):
OK, that's interesting.
Right, because this sort of impersonal expression,
one can hear dogs barking, you'd think that it would be plural.
But because it's an event, and the verb is hear versus see,
it goes into the singular.
So there's subtle kinds of semantics related to, again,
(01:18:28):
the whole business of verbs of perceptionand how they work with everything.
But certainly, withsubordination is really interesting.
And again, probably we can do a deep dive on that later.
Yeah, that's-- I mean, but that's really important.
Because a lot of times, like the big thing--I know that you're not totally into it,
(01:18:52):
but a big thing that people do is historical stuff.
And they're looking for sources for grammaticalization.
Absolutely.
And this can be a source for a distinctionthat gets grammaticalized later.
Absolutely, yeah.
And it's just important to understand
that we're thinking, especially English speakers,
(01:19:15):
we're mostly thinkingin terms of conjugations
and maybe conjugationswith some mood dances.
But nominalization is super common all overthe place to do a lot of these functions.
And it's very important toremember that that noun phrase
has to be worked into therest of the clause somehow.
In many Australianlanguages, these subordinate
(01:19:35):
clauses take allkinds of case marking.
To add additional subtleties, the process,
this very exuberant process of case marking subordinate
clauses, is so strong that it has had a strong impact
on the historical development of a bunch of these languages.
So that's something to keepin mind if you're going with
(01:19:58):
a more nominalizing modefor a bunch of these things.
All right, so nominalization, conjunctions,intonation, as I said, mohawk.
Just like one clause, it happens at aslightly higher pitch, and then it drops.
(01:20:21):
We have another paper on clause combining
in languages of North America.
So here we have some differences between--so we have complements versus adjuncts.
That's a distinction you need to think about,
(01:20:43):
like a complement of a verb, versusan adjunct, a modifier kind of thing.
Sometimes they'll use different strategies for subordination.
Sometimes they'll be similar.
It's just going to be language dependent.
(01:21:03):
We have a bunch of papers tolook at here that you can check into.
Let's talk about, what is clause chaining?
Right, I'm not sure where clausechaining belongs in all this discussion.
(01:21:24):
But as I was saying earlier, that the distinction
between subordinate and coordinating clauses
is maybe not as strict as we'resuggesting while we're talking here.
So we talked about this when we had Matt Pearson on to talk
about converbs.
He has something called medial verbs in Okuna.
(01:21:46):
And these are verb forms that cannot exist on their own.
But in any normal translation, you would just use "and."
They're a verb form used for sequencing.
I did A, and then B, and then C, and then finally, I did X.
And depending on the structure of the word order default
(01:22:09):
in the language, you might save your fullyinflected verb for the end or the front.
So medial verbs are usually where you get a main verb,
and then these other subsequent verb forms that just say--
that basically mean "and then, and then, and then."
(01:22:29):
Whereas in languages with converbs,
they might use the general converbs as the sequencing
in addition to other sorts of more thingsthat we might think is more dependent.
Yeah, and just to refresh people, converbsare like a particular kind of verb form.
Yeah.
That we covered in the audioarchives-- I can link to that.
(01:22:52):
But those are-- oh, I mean, they're
sort of similar to a participle, similar to other things.
Often what they do is they encode
some sort of connective meaning between two verb phrases.
(01:23:15):
Well, many-- you can have alanguage, you can have one
or two converb forms,which mean all sorts of things.
Or you can get languages, especiallysome from the Caucasus, or my favorite,
Coptic, where they mean allsorts of very, very specific things.
And you have lots of converbs.
And they cover a lot of this.
In my opinion, many kinds ofconverbs qualify as subordinate clauses.
(01:23:38):
Right.
They can have adverbial meanings.
They can also have just likeconjunctive meanings attached to it.
And that gets into-- combinedwith this clause chaining thing,
that gets into what you're sayingabout the boundary being fuzzy.
Because if you have a chain of verbs where only one of them
(01:24:00):
has full inflection, but semantically it's not really
like embedded meanings, then the question is,
is that conjoined sentences, or is it a subordination?
Because grammatically, it's kind of looking like subordination.
(01:24:23):
But semantically, it's like, oh,these are just conjoined clauses.
So that can be an interesting way to play with that.
Right.
And that's why it's fuzzy.
These things can have multiple meanings.
And as I said, "and" can do weird things in English as well.
Right, right.
So popping back a bit for strategies for subordination,
(01:24:46):
we've talked about conjunctions and nominalization.
Or another thing you could do is have
a different conjugation paradigm for different clause types.
That can be as simple as the standard Indo-European,
like, oh, we'll use the subjunctive.
But you can have other languages like Coptic,
which has many pagesof paradigm forms, which
(01:25:10):
encode specificsubordinate clause types.
And there are other languages that have separate paradigms
that go beyond just having a different mood or something.
Right, right, right.
I've talked a bit about nominalizations thatcan take different kinds of case marking.
(01:25:33):
Right.
Another kind of search,you can-- go ahead.
Let's talk about this, because with case marking,
are we talking about they take an argument role,
and then they get ergative case, or accusative case,
or whatever, case marking to fit the argument that they're in?
(01:25:55):
Yes.
Or locatives are very commonly used for temporal.
OK.
Yeah, that makes sense.
Right, where you smacka dative onto an infinitive,
or a {unclear}, or somekind of verbal noun.
And you get that sort of temporal thing.
That's interesting.
You can do the same sort of thing
by sticking an adposition phrase there, too.
(01:26:20):
Yeah, yeah.
Yeah.
And in addition to different kinds of case marking,
the clause itself might have its normal case marking
capabilities mangled.
- So if you have- Right.
And again, we go back to ancient Greek.
If you have a clause where, because of its dependency,
it is using an infinitive for the verb, right?
(01:26:40):
It's been deranked.
It's an infinitive.
The subject has to be expressed with the accusative.
Right.
OK.
And in English, we do the same.
Like, "he was talking," "I saw him talking."
Yeah.
Right?
He is talking, it has to be the him.
Again, makes it fit into normal English structure better.
(01:27:05):
Right, right.
"I saw he talking," is not--
Yeah, "I saw he talking," that is not the way we do it.
Yeah.
But that is interesting, like, what
happens when you have a pronoun that's in between the two.
(01:27:27):
Obviously, that happens more oftenwith stuff like relative clauses and such.
But it's-- and we can do a whole thing on relative clauses,
because there's so many differentlayers of how those get worked together.
(01:27:50):
But-- We did a relativeclause show a long time ago,
like more than 10 years ago.
Maybe we can revisit it.
Revisiting some of them, nowthat we're both older and know more.
Yeah, I could re-listen to it and see.
I think something that unites all of this stuff
is one thing you need to do when you're doing subordination
(01:28:17):
is you have to think about what role isthis sentence taking in the matrix sentence.
Because-- so you talk about nominalization.
Well, why does nominalization-- why isnominalization used for subordination?
(01:28:37):
Because very often, it is taking theplace of a-- what would normally be a noun.
Either it is taking an argument role,
or it is in some oblique role, object of a preposition,
of an adposition, or a locative, orbenefactive, something like that.
(01:28:58):
But you have to keep in mind that you also
have to consider when a sentence becomessomething akin to an adjective or an adverb.
Especially adverbs, yeah.
Yeah, adverbs all the time.
That's temporal constructions.
You're adding an adverb all the time.
(01:29:21):
Which, I mean, that can be still related.
You could still end upusing a nominalization
with an adpositionor something like that,
or a nominalization with case marking to get that meaning.
Or you could go with something completely different.
But I think when you're doing your different flavors
(01:29:44):
of subordination, that's the thing to keep in mind,
is what role is this sentence playing in the matrix sentence?
Yep.
I'm going to start with an abstract thing, which, George,
I've not put in the notes here, but which I think about a lot.
So sorry about that.
And then we'll move on to the types.
And that is your different kinds of--
sometimes these clauses are called complement clauses.
(01:30:06):
And many of them have a broad distinction
between where the subordinate clause is either a fact,
it could be an activityindicating an extension
in time, or it could beindicating a potential.
Fact-type complements,fact-type subordinate
(01:30:27):
clauses are most likelyto not be deranked.
They're most likely to be balanced.
But they-- or maybe a mood change, but that's about it.
Activity types are very often nominalized.
And then finally, the potential type definitely
is going to have most likely to be reduced tense and aspect
(01:30:50):
marking, reduced pronominal marking,most often using a different mood.
Right, right, right.
And so certain kinds of verb types
are more often going to take one of these or the other.
Attention verbs or perception verbsusually take an activity complement.
"I see him walking to the store."
(01:31:13):
Or they might take facts.
Verbs of thinking take either fact or activity.
Decisions take fact or potential.
Verbs of liking or preferring typically take activities.
So all of these have-- it'sworth thinking about these,
because most languages aregoing to have five, six, seven
(01:31:37):
of these different types, possibly with some subtypes.
You do not need every kind of subordination strategy
that exists in your conlang, nor do you need two.
There will be differencesthat all have to do both the semantics
and thestructure of your language.
Right, right, right.
Any time we're going over all these different options,
(01:32:03):
there is a tendency for a lotof conlangers to do what we
call the kitchen sink andjust throw everything in, right?
But it's useful to know all of your optionsand then decide what you're going to pick.
And it may be the case that choosing one option
(01:32:27):
leads you to have different other varieties.
Maybe you end up with just one strategy that dominates,
like you have one nominalizerthat gets used for everything, right?
Especially if you can have nouns modifying nouns,
(01:32:48):
then you could do your relative clause withthe same thing, all that kind of thing.
But maybe if you go with a converb strategy
and then you decide to make a few different converb forms,
then you're like, OK, well,I've got-- now I've got lots
of different subordinationstrategies going all at once.
And also, you know what?
(01:33:13):
It's always an option to just stick asentence in a sentence in the relevant place.
Just like you mentionedthe Mohawk example, it's only
distinguished on intonation,that you can just do that.
And some languages' relative clauses to an English speaker
look bonkers, becauseit's like, you've just shoved
that sentence inside anotherone with no over-marking.
(01:33:36):
Like, in-place relatives can be tricky.
Or even just not even just stick that sentence in the place,
but also have the noun in therelative clause sentence impeded.
And it's even more wacky-looking to us.
It's a perfectly natural thing to do,but it looks strange to English speakers.
(01:34:00):
OK.
And then my last on the list was converbs,
but we've already talked about those, so we'll skip over.
So let's talk about some commontypes of subordinate conjunctions.
And I've got a listhere, just like cause, like
because, reason, orcause, reason, result, place.
I see where he's leaving.
(01:34:22):
Report, I said that someone said that.
Manner, which could be like a real manner,or a hypothetical, sort of like an as-if.
And then one that's a possible consequence,which you don't use much in English, lest.
Lest.
For fear that.
Yeah.
Some languages have things like this and use them a lot more.
Like, it seems like the Australianlanguages really like their lests.
(01:34:46):
Oh, really?
OK, that's interesting.
Yeah.
But it's not like you sound goofy
if you say that in English.
Yeah.
Chinese really likes resultative clauses.
In Chinese, it's kind of-- if it'sthe same subject, it's like the verbs
(01:35:06):
kind of form a pseudo-compound.
And so that's an interesting strategy that that uses.
And then it also messes with like aspect as well.
Yeah, it does funny things with aspect marking.
A lot is going on with those.
And there's very weird things, like thewhole example of kill-retreat, meaning--
(01:35:36):
do you know that example?
I do not know kill-retreat.
That's great.
OK, so there's an example that you hear a lot in Chinese.
他们杀退了敌人 (tāmen shātuì le dìrén)
So they kill-retreat enemies.
What that means is they were going to killthe enemies and the enemy retreated, right?
(01:36:01):
So kill, which is normally a telic verb,normally has an endpoint, becomes atelic
because it's not realized that you killedall the enemies because they retreated.
So that's-- telicity is another episode that we covered.
(01:36:23):
Yeah, we talked about the telicityone, didn't we, these cancelable--
Yeah, I think we may have covered that example.
But you can end up with weird semantics
when you add a different sort of subordination in there
because they are very good atcanceling things and adding weird--
(01:36:54):
canceling what you expected to be an entailment and stuff
like that, yeah.
So cause, reason, result, place, report, manner--
you have a whole section here on specifically temporal clauses.
Again, all of this-- and that'sjust the tiniest bit of stuff I read.
(01:37:16):
So we can go over this to getpeople thinking about some stuff.
But I really want to do a whole temporal clause show
because there's so much that goes on with these things.
Yeah, so you can have-- soyou're saying, OK, so the past--
so I just want to think about six temporal conjunctions
(01:37:39):
in English that distinguish between past--this is in comparison to the main clause.
It could be past, co-temporal,like at the same time, or the future.
And it could refer to either a point or a duration.
So past point is after, past length is since.
(01:37:59):
Same time point is when, same time duration is while.
Future point is before, future duration is until.
Right.
And we do happen to have a nice regular chart
(01:38:20):
for that in English.
I'm assuming that in many languages that is not the case.
Boy, if you can get it-- Ihave no non-copyrighted
references talkingabout this subject.
But the two or three pages you need
for a full chart of Koreansubordinate clause types is amazing.
(01:38:45):
It's huge.
And there are many temporal things thatall look like they're translated the same.
They can't possibly all be the same,but I've not had a chance to dig into it.
But I will try to find good references for that if I can,
if we can do a temporal subordinate clause episode,
because it's huge, many things, much morerich than most languages I'm used to seeing.
(01:39:08):
So let's talk about how they can overlap and things like that.
Right.
So again, many kinds of subordinate clauses
are polysemous with other kinds of subordinate clauses.
So in English, and in German, and in many languages,
when clauses, which looktemporal, can also be conditional.
(01:39:36):
"When it rains, get an umbrella."
That's not a temporal consideration.
That's a conditional.
While clauses can evolve into conditionalsor clauses of concession, like "although".
And while clauses are one of those
(01:39:58):
that locative case marking is a really popular thing.
After is much less likely to be polysemous with other things.
Before is often deranked and usuallyinvolves negation in complicated ways.
Have we ever talked about expletive negation?
Oh, honestly, I don't know.
(01:40:19):
It's been so long.
We've been doing this for so long.
Expletive negation, it doesn't really exist in English.
But again, it does occur in Greek.
Like, "I was afraid she died," in Greek was, "I was afraid--"
it looks like, "I was afraid she didn't die."
But that's coming from, we believe, a kind of optative.
(01:40:42):
"I was afraid, may she not be dead."
Right, so that's expletive negation.
And there are certain kinds of verbs where that happens.
And the before constructions are often
a situation where you get expletivenegation because of the semantics.
Right, right, right.
And you talk about-- so in English,obviously, we're mentioning these mostly
(01:41:09):
come from prepositions and one interrogative pronoun.
But we have-- all of the strategies
apply to temporal as well as any other.
You could use case marking.
You could use-- I mean, anywhere you coulduse adpositions, you can use case marking.
(01:41:32):
You can use converbs.
You can use any of those strategies.
Or again, just stick a sentence in a sentence.
But it is interesting--
I just want tointerject quickly, with temporal clauses,
there might be iconicity related tothe ordering of the clauses themselves.
So for example, even in a languagethat has very strict head final--
(01:41:56):
and so that normally means your dependent clauses come first--
things like result clauses mightstill come last for iconic reasons.
Right, right, right.
Because the result comes afterbecause that's what happens naturally.
And so the grammar gets changed to move the result clause,
- for example, to the end where normally--- Another thing with that is a result--
(01:42:17):
if you have any kind of broader, either historically
or synchronically in your language,you have any kind of broader-- what is it?
Verb sequences?
What is the term that I'm looking for?
Serial verbs?
(01:42:38):
Serial verb constructions, yes.
If you have any kind of serial verb constructions,
you're very, very likely to have that iconicity happen.
Oh, yeah, yeah, yeah.
Because serial verbconstructions are just like that.
They tend to have that iconicity of temporal order.
That's why Chinese has the resultativeimmediately after the matrix verb.
(01:43:06):
And they also have lots of cases of prepositional phrases
that come from verbs coming before the verb.
It all kind of leads intothat-- my theory, at least.
So that's all super interesting.
(01:43:29):
But my question is, with this ordering,
does that ever happen with past versus future clauses?
Oh.
I didn't see any of that in my reading.
But that would be worth looking to pay attention
to see if there's something like that in the future.
That would be super interesting.
It's kind of--
(01:43:51):
So I did seea paper specifically on the
conceptual metaphors of timeas it relates to subordination.
But I didn't dig into it heavily.
I should have, probably, because that would really
imply that, like, where is the futurein relation to you in your language?
Is it in front of you, or is it behind you?
(01:44:11):
Ooh, we should look at just Aymara conjugations.
I mean, I'm just saying the position of that in the sentence--
because obviously, in English, we can justplace the temporal clause wherever we want.
But it doesn't seem toreally-- not perfectly, but there's
several different slots inthe sense where it can go.
Because adverbs in English just go float on the sentence.
(01:44:35):
But there doesn't seem to be any of that.
Like, before or after will just occur anywhere.
And it's usually not tied to the actual temporal order of events.
But it would be interesting if we hada language that was militant about,
(01:44:57):
like, if it's a before clause, it has to happen.
It has to be placed before the matrix clause.
And if it's an after clause, ithas to be after the matrix clause.
I don't know if that would ever actually happen.
But it would be interesting if wefound a language that's like that.
Yeah, I don't know if there'd be--
it is interesting thatit's normally result
clauses that come infor this sort of dislocation,
(01:45:17):
where cause and effect seems to matter more than time.
Right, right, right.
One last thing that I think,since we're coming close
to time, I want you toget to your favorite thing.
Insubordination.
Insubordination, yes.
I love it so much.
So as I said, there's a recent entire book on it,
(01:45:40):
because it's only a recently investigated phenomenon.
It had all sorts of terrible names.
And now, finally, everyone's converging on insubordination.
I just think it's fun.
They tend to be overlooked, because if you
see what you think is a dependent clause floating
about on its own, it invites the interpretationthat the speaker's made an error.
If you're really focused on the rulesor-- and we'll get to this in a moment--
(01:46:02):
if you're focused on sentences inisolation from the broader conversation.
Oh, OK.
So insubordination is generally agreed
to evolve from full sentences, where a stereotyped main clause
got dropped, or a situation where a clause is implicitly
part of what the other speaker has just said.
So the other speaker has said something
that we're going to pretend is the main clause.
(01:46:24):
And the response-- Isa subordinate clause.
Yes.
So this is like a discourse thing.
Yes.
Conversation analysis, the discourse stuff.
Which a lot of conlangs just don't get to.
So it's important to think about this.
And if you think about how it's used,
it's really clearly related to discourse and pragmatics a lot.
(01:46:47):
So politeness rituals.
"If you could fill out this form, please."
Uh-huh.
Well, then what?
We leave it implied.
This is a normal way in English.
We use insubordination with conditionals to do politeness.
But we also can do it to use threats.
"If you touch my car!"
(01:47:09):
And then, let peopleimagine the rest of that sentence.
Right.
So yes, requests, threats, warnings, politeness rituals.
Also kinds of evaluation.
We don't say this much in English anymore,
but "that you dare tosay that" has an I can't
(01:47:32):
believe, that you dareto say that in front of it.
So evaluation.
So it's all much more emotional,pragmatic, and politeness things.
Conditionals seem to be a super common source for these.
Right.
But not the only.
(01:47:52):
The examples you gave are mostly conditionals, right?
Yeah, in English, they're conditionals.
Oh, and there are other ways youcan use result clauses and other things.
Let me look at Dahlstrom here, because Ihave some examples that I wanted to see.
So here's a fun one.
So Dahlstrom, again, now people are talkingabout insubordination a little bit more.
(01:48:12):
Where in Plains Cree, you can have anindependent question, "Are you hungry?"
And you can also have an "Are you hungry?"with what they call a conjunct inflection.
So a dependent form.
The dependence, the insubordinateversion, is used in a context where you're
(01:48:39):
trying to understand why they're doing something.
Like if someone'srunning to the fridge and
rummaging around, you'duse the subordinate one.
Whereas if you were just like, if someone was visiting,
and you're like, hey, do you want some food,
then you would just use the normal independent one.
So that's a very subtle contextual distinction.
So insubordination doesn't haveto be like these big things either.
(01:49:04):
So I just wanted to talk about that.
There's an entire literature on thisnow that if you search on insubordination,
you will find papers now, an entire book, as I said.
Right, right, right.
And we're going to have links in the show notes.
And if you're on YouTube, in the description to the papers
(01:49:26):
that we are referring to here.
And probably citations for books andstuff that we can't necessarily link to.
OK.
So that's really great.
Are there any other things that you
want to cover before we sign off?
(01:49:46):
Like any particular things?
No, nothing new.
It's just that there are many more--again, as is always the case with typology,
there's more things than probably mostof us have thought about available to us.
Well, it's definitely been helpful to me,and I hope it'll be helpful to listeners.
Because I need to dosome more interesting
(01:50:07):
subordination than I didwith the last language.
Which I think the last language, Idid something with relative clauses.
But then with other subordinations, I waslike, just stick a sentence in a sentence.
And that works.
You don't want to do that every time.
Yeah.
I've been casual about temporal subordinationin pretty much all of my languages.
(01:50:31):
And there's just so much more that could go on there.
I do like using adpositions for temporal clauses.
And having the thoughts about, OK, what spatial metaphor
am I going to use tochoose the adposition
(01:50:52):
that goes with thetemporal clause, right?
That is important.
You've got to work out your conceptual metaphor of time
before you start coming up with grammar for it.
Yeah, you can thinkabout, OK, is time in front
of you or behind you orabove you or below you
or all those things, or however that works out.
(01:51:13):
But again, there's lots of options out there.
Go read the materials that we're goingto link and take a look at those things.
And we'll try to do a few more episodes on some of these.
And I think we probablywill do a whole one
(01:51:34):
on temporal clausessometime in the future.
Yeah.
And conditionals, too, I think, if we've not done it before.
I'm going to-- probably we'll have interviews
and some other guests.
I think I will be moving on and doing
more of these linguistics episodes in the future.
(01:51:55):
So stay tuned for that.
Getting back into that either linguistics episodes
or like conlanging techniques kind of things canbe something I come back to a little bit.
But thank you, William, for being on.
(01:52:15):
Thanks for having me.
And we are-- do you have any final wisdom?
Nope, no wisdom this time.
You can check out William at Linguini.
(01:52:37):
Yes, Lingweenie.org.
Yeah, the Lingweenie blog.
I can -- Acta Lingweenie
I can link to that.
He is also the guy who madethe Conlanger's Thesaurus,
which a lot of peopleout there apparently use.
I'm glad they do.
I'm glad.
Yeah.
(01:52:58):
And the word generator that I use every single day,
lexifer, yes, which you've seen on my streams.
This is the guy who made that one.
But thank you, William, for being on.
And I hope to have you on again to talk,
(01:53:19):
to have another deepdive on another big
linguistics topic, moretypology sharing and things.
And with that, I'm going to say happy conlanging.
Ciao.
OK.
Thank you for listening to Conlangery.
You can find archives and show notes at conlangery.com.
(01:53:41):
And if you'd like to support the show, thebest way to do that is to go to our Patreon
at patreon.com/conlangery.
With just $1, you can get access to exclusivepolls for my Tongues and Runes series.
And also, for $5, you can get everypodcast episode early and ad-free.
(01:54:02):
And now to thank our patrons.
Mintaka, Kenan Kigunda,Conor Stewart Roe, Jesse,
Kaye, Alex Rossell-Hayes,Vyren Patrick, Tabby,
Alexis Hugelmann, Sylvia Sotomayor, Grammar Antifa,
Grakka Grunk, Sigourney Hunter,iloivar Jaana Mentoleum,
(01:54:25):
Niklas Norblad, Anthony Docimo, Jake Penny,
Artifexian, MilesWronkovich, Daniel Quigley,
Paul Roser, our table,Langworm, Nai, and Horn Bory.
Thank you.
Conlangery's theme music is by Null Device.
(01:54:48):
Conlangery is distributed under a Creative Commons Attribution
Non-Commercial Share Alike 4.0 international license.
[MUSIC PLAYING]
[MUSIC PLAYING]
[BLANK_AUDIO]
They tend to be overlooked, because if you
see what you think is a dependent clause floating
about on its own, it invites the interpretationthat the speaker's made an error.
If you're really focused on the rulesor-- and we'll get to this in a moment--
if you're focused on sentences inisolation from the broader conversation.
[MUSIC PLAYING]
(01:00:33):
Welcome to Conlangery, the podcast
about constructed languages andthe people who create them.
I'm George Corley.
And with me, the next town over,is William Annis.
After a long time.
Yes, yes.
He used to be down the road a ways,but I have moved.
So I'm not in Madison anymore.
(01:00:54):
I'm not going to tell you exactly where I am, though.
But this is going to be something
I haven't actually donesince we started video
episodes, but it'skind of a return to form.
I've been doing a lot of interviews of people.
And William used to bejust the regular co-host of
(01:01:19):
Conlangery, if you werehere for audio episodes.
And he's coming back to talkabout a linguistics topic again.
And hopefully, we can do-- I'll bedoing more interviews in the future,
but I'll also come back and do a little bit more of this,
(01:01:42):
because this is the thing that peoplereally liked about the audio podcast.
And I think it's good to introduce,
now that we have video folks on YouTube as well, to there.
But William, thank you for coming on.
(01:02:03):
It's been so long.
There's fancy new technology.
I had to sully my computer with Chrome,but I'm like, for George, I can do it.
Yes, I don't know why-- nobodydevelops these video platforms for Firefox.
I don't know.
I wish they would.
It's just Chrome has too muchpenetration, and they just go for that.
(01:02:25):
And they could get a bunch of different browsers with that.
But anyway, yes, we are here.
You can see William for once.
And I have a face for radio.
I will see if people who arelongtime audio listeners who
(01:02:51):
have not seen William's face-- Iwonder what your reaction is going to be.
Because any time I see someone I'mused to hearing in audio see their face,
it's like, that face--the face doesn't match
the voice, like almostevery single time.
Why is his head so shiny?
(01:03:11):
Well, hopefully I can adjust with color correction there.
OK, so anyway, what are we talking about today, William?
We're talking about subordination.
This was an idea I had a long time ago.
And then it took us many months to finally get to this.
There was a little COVID in there for me.
And in the course of making my usual slew of notes,
(01:03:37):
my opinion is there are a few of thesetopics that deserve their own show.
We've already done one on relative clauses.
But I think we could do an entireepisode on temporal clauses.
Absolutely, we could do anentire show on conditional clauses.
If we haven't already, we could probablydo it again and other sorts of things.
So today, we'll do atop-level overview of...
(01:04:00):
.. whatever it is subordination means.
Different kinds of meanings and semantics
those can have, different ways of forming those.
I have a lot on my favorite thing,insubordination, which sounds fun.
I like the idea of insubordinate grammar, which
is the term that people have settledon for an underdescribed feature where
(01:04:23):
subordinate clauses cut loose fromthe main clause and go off on their own.
And that's super interesting.
So that's what we'll cover today.
And then some of the things we can deep dive
into in later episodes,I think, would be
better, rather than tryto cover them all here.
Or we'll end up withanother one of our two-hour episodes,
which isasking a lot of people.
(01:04:44):
Yeah, we don't do that anymore.
That was early on.
Before we knew better.
OK, so what is a subordinateclause, basically?
We'll start with a simple definition.
And that's just when a sentence is crammed
(01:05:06):
inside another sentence, either as an argument,
like a subject or direct object, or a modifier.
And those modifier ones usually get called adverbial clauses.
And that's different from coordination,
where sentences just follow each other like ducks in a row.
"And," "or," and "but" are normally
the standard coordinating conjunctions and clauses.
(01:05:29):
But obviously, immediately, theoretical issues arise.
And it's not clear to me anymore that speaking of subordination
is useful as a distinct thing fromother kinds of clause combining.
(01:05:50):
So my favorite heroes, Dixon and Aikhenvald,
have a book called "The Semantics ofClause Linking" that covers all of them.
And that probably makes sense, because there is more overlap.
"And" can do surprisingly subordinating-likethings in many languages, which we're not
going to talk about too much today.
(01:06:11):
So the distinction isn't always maybeas clear-cut as we might think.
In some cases, it absolutely is, but...
And sometimes I'll talk about embedded clauses
versus subordination, or dependentclauses versus subordination.
These are all terms used todescribe basically the same thing.
(01:06:31):
The big thing is, normally, a subordinateclause has been modified in some way such
that it cannot normally go wandering off on its own.
Right.
So if you think about just English examples,relative clauses do this in English,
(01:06:52):
where, like, "I saw the man who," I don'tknow, "I saw the man who watched the movie."
I don't know.
That's a bad example.
But anyway, but "who watched the movie,"like that-- there's no subject pronoun.
(01:07:12):
The "who" is like a linking thingthat signals it's a relative clause.
And so that "who watched the movie," Imean, you could turn it into a question,
but it's not really an independent clause anymore.
It has to be attached.
There are sometimes--
(01:07:36):
Tests for this can be tricky in Englishbecause we murdered our subjunctive.
Yeah, yeah.
And there's also times where the subordinating, like,
complementizer-type thing is, like, deleted,
like a "that" is deleted, where I thinkthat something-- "I think that it was blue."
(01:08:01):
You say, "I think it was blue."
So in English, a lot of times, it's fuzzier.
In a lot of other languages, you would
have a different verb form or something else that's signaling.
But we'll get into all of those strategies.
Yeah, it can be even-- in some languages,it is extremely blunt and obvious.
In other languages, it's trickier.
(01:08:21):
I think Mohawk, I have a noteon this, where a subordinate
clause is marked offentirely by intonation.
[LAUGHTER]
So nothing, no conjunctions, nothing else.
And the main point here is that, in general,
subordinate clauses have to be fiddled with some way
to make them work within the main clause grammatically.
(01:08:43):
And we'll talk about different kinds of fiddling shortly.
And normally, parts of the subordinate clause
are not allowed to go wanderingabout other parts of the main clause.
But with relative clauses, you have
a pivot word, which is part of both clauses, in effect.
(01:09:07):
But in general, the two clauses are kept separate.
And something has been done to the subordinate clause
to make it fit in with the rest of the sentence.
It's a separate constituent, as people say.
It is its own thing.
It's a sentence embedded into a sentence.
And it has to be kept unified, generally.
(01:09:27):
Yeah.
So when it comes to different strategiesfor subordination, there's a lot of--
there's so many different kinds of terminology.
One set of terms that I'm familiar with,
so I will tend to use, is a balancedverb versus a deranked verb.
(01:09:50):
A balanced verb is a kind of verbclause that can be used independently.
And a deranked clause or verb cannot be used independently.
And there are lots of ways to derank a verb.
Could be a completely different conjugation type.
Could be a change in mood.
Could be a converb, participles, nominalization, extra goo.
(01:10:14):
Like jumping back into form, talking about ancient Greek.
Some kinds of ancient Greek subordinate clauses
require both a mood change and a particle in the clause.
Oh, OK.
So you have to-- yeah.
And we have links to cover all of these things.
(01:10:35):
We just have the Wikipedia link will coverthe basics of balancing versus deranking.
But yeah, so you can have a situation.
So it's both a particle and a mood change in ancient Greek.
Yeah, for some clauses.
For some clauses.
(01:10:56):
Some clauses, just it's a mood change.
Yeah.
I think a lot of times subjunctive--subjunctive is like a broad, ill-defined term.
But often, the subjunctive is used in this way
to mark that this clause is dependent on the other clause.
(01:11:19):
And that is also kind of deranking.
Yeah, yeah.
So the point about balanced here is,
especially in some kinds of temporal clauses in English,
like before and after, the sentencejust looks like a normal sentence.
"Before, you've got to talk to the doctor, do this."
That is a balanced clause using a conjunction
(01:11:42):
versus a deranked, which, "if I were to fall asleep..."
So the last vestiges ofthe English subjunctive
sometimes occur inconditional clauses.
And I mean, could you also consider
infinitive clauses in here?
Because we have that.
Well, right.
So an infinitive, I consider a kindof nominalization most of the time.
(01:12:07):
So yes, absolutely infinitives are part of that.
So things like infinitives, thingsthat are clearly deranked in English
are things like infinitives, gerunds, or any--
gerunds, participles, anything withthe -ing form, those kind of things.
Yeah, "Sleeping during a thunderstorm is very difficult."
(01:12:30):
So that's a deranked, "Sleeping during a thunderstorm."
It's a nominalized clause.
Yeah, yeah, yeah.
And I think that's an important thing to say,
is like, nominalization is a really easy pathway to this.
Because very often, these subordinateclauses are taking an argument position.
(01:12:51):
They are taking the same place as a noun.
So if you have a good distinction betweennouns and verbs in your language, then you
can stick some kind ofa nominalizer onto the verb,
and use that foryour subordinate clauses.
But you don't have to.
This is always-- always, this is options foryou to do, because languages are diverse.
(01:13:15):
They do all kinds of different things.
Yeah.
I'm trying to find a fun example I saw.
But I can't find it.
We've talked about conjunctions, many kinds of those.
I will confess, a lot of the time, myapproach is very boring for temporal clauses,
(01:13:38):
when they could be a lot more interestingthan I have been making them, typically.
Conjunctions, nominalization-- and nominalization
doesn't just have to mean doingsomething to turn the verb into a noun.
Like, we have the classic--we'll go back to another standby--
Mandarin, de shihou (的时候),for when, where you're just
(01:14:01):
taking a clause, and then a possessive orattributive marker, and the word for time.
So "When I saw my father," whatever.
Right.
In that case, it'ssort of-- so shihou.
So in that case, shihou, "time", is like the noun there.
(01:14:25):
And then you're adding a relative clause
for what the actual subordinate clause is there.
Whereas in English, we have more of a--so "When something, something happened,"
(01:14:47):
then that's sort of amore like-- just like a lot of
our other clauses with therelative pronoun and such.
Yeah.
So that's another strategy, is you
(01:15:08):
could come up with one kind of subordinating structure
and reuse it to make other subordinating structures.
Correct, yeah.
Yeah.
I couldn't find an online resource for this,
but Hixkaryana , which is famous for its default word order,
(01:15:28):
also has an enormous number of nominalization types.
Which at first, you're like,what a weird derivational system.
But then you realize it's almost entirely structural.
It's for grammar purposes.
It's not for creating new kinds of words.
Right, right.
So you're nominalizing verbs just to get that--
(01:15:53):
Right, so an example one is a manner nominalization.
The way you sneeze is not a special kind of word
you normally use, except to shove it into a clause.
The way he talks annoys me.
So a special nominalization just for manner.
Where English is-- in some sense, Englishis doing the same kind of nominalization
(01:16:15):
with a relative clause, the way in which he talks.
Yeah, the same kind of thing.
So one fun thing I just wanted to note--
I found this little paper hereon the semantics of perception.
And this looks like paper, like apresentation notes, rather than a full paper.
(01:16:37):
It's talking about verbs of perception
and the kinds of subordinate clauses they can take.
"I see that it is raining."
"I heard that he went to the store."
That sort of thing.
There's a fun thing in Spanish that differences
between vision and hearing can lead to grammatical changes.
(01:16:57):
So visual perception mainly involves object perception,
while auditory perception aligns more with event.
So in Spanish, this affects pronominal agreement
with perception verbs.
Visual perception tends to favor object agreement,
while auditory perceptionoften triggers event
agreement, which means itwill always be in the singular.
(01:17:18):
So if you look on--
this is on page 3 of this paper I linked to, George.
Where is that?
Which paper is that?
It's the one under nominalization,so above the big table.
OK.
So this one.
I already opened that.
So we have a document thatI have also shared-- this one?
(01:17:43):
OK.
So page 3.
OK.
So we have-- Sothe example is 8-9.
"Se veían desfilar ese día por las ..."
OK.
OK.
Oh, OK.
So you have-- so the use of seeing versushearing, it changes the meaning here.
(01:18:07):
OK, that's interesting.
Right, because this sort of impersonal expression,
one can hear dogs barking, you'd think that it would be plural.
But because it's an event, and the verb is hear versus see,
it goes into the singular.
So there's subtle kinds of semantics related to, again,
(01:18:28):
the whole business of verbs of perceptionand how they work with everything.
But certainly, withsubordination is really interesting.
And again, probably we can do a deep dive on that later.
Yeah, that's-- I mean, but that's really important.
Because a lot of times, like the big thing--I know that you're not totally into it,
(01:18:52):
but a big thing that people do is historical stuff.
And they're looking for sources for grammaticalization.
Absolutely.
And this can be a source for a distinctionthat gets grammaticalized later.
Absolutely, yeah.
And it's just important to understand
that we're thinking, especially English speakers,
(01:19:15):
we're mostly thinkingin terms of conjugations
and maybe conjugationswith some mood dances.
But nominalization is super common all overthe place to do a lot of these functions.
And it's very important toremember that that noun phrase
has to be worked into therest of the clause somehow.
In many Australianlanguages, these subordinate
(01:19:35):
clauses take allkinds of case marking.
To add additional subtleties, the process,
this very exuberant process of case marking subordinate
clauses, is so strong that it has had a strong impact
on the historical development of a bunch of these languages.
So that's something to keepin mind if you're going with
(01:19:58):
a more nominalizing modefor a bunch of these things.
All right, so nominalization, conjunctions,intonation, as I said, mohawk.
Just like one clause, it happens at aslightly higher pitch, and then it drops.
(01:20:21):
We have another paper on clause combining
in languages of North America.
So here we have some differences between--so we have complements versus adjuncts.
That's a distinction you need to think about,
(01:20:43):
like a complement of a verb, versusan adjunct, a modifier kind of thing.
Sometimes they'll use different strategies for subordination.
Sometimes they'll be similar.
It's just going to be language dependent.
(01:21:03):
We have a bunch of papers tolook at here that you can check into.
Let's talk about, what is clause chaining?
Right, I'm not sure where clausechaining belongs in all this discussion.
(01:21:24):
But as I was saying earlier, that the distinction
between subordinate and coordinating clauses
is maybe not as strict as we'resuggesting while we're talking here.
So we talked about this when we had Matt Pearson on to talk
about converbs.
He has something called medial verbs in Okuna.
(01:21:46):
And these are verb forms that cannot exist on their own.
But in any normal translation, you would just use "and."
They're a verb form used for sequencing.
I did A, and then B, and then C, and then finally, I did X.
And depending on the structure of the word order default
(01:22:09):
in the language, you might save your fullyinflected verb for the end or the front.
So medial verbs are usually where you get a main verb,
and then these other subsequent verb forms that just say--
that basically mean "and then, and then, and then."
(01:22:29):
Whereas in languages with converbs,
they might use the general converbs as the sequencing
in addition to other sorts of more thingsthat we might think is more dependent.
Yeah, and just to refresh people, converbsare like a particular kind of verb form.
Yeah.
That we covered in the audioarchives-- I can link to that.
(01:22:52):
But those are-- oh, I mean, they're
sort of similar to a participle, similar to other things.
Often what they do is they encode
some sort of connective meaning between two verb phrases.
(01:23:15):
Well, many-- you can have alanguage, you can have one
or two converb forms,which mean all sorts of things.
Or you can get languages, especiallysome from the Caucasus, or my favorite,
Coptic, where they mean allsorts of very, very specific things.
And you have lots of converbs.
And they cover a lot of this.
In my opinion, many kinds ofconverbs qualify as subordinate clauses.
(01:23:38):
Right.
They can have adverbial meanings.
They can also have just likeconjunctive meanings attached to it.
And that gets into-- combinedwith this clause chaining thing,
that gets into what you're sayingabout the boundary being fuzzy.
Because if you have a chain of verbs where only one of them
(01:24:00):
has full inflection, but semantically it's not really
like embedded meanings, then the question is,
is that conjoined sentences, or is it a subordination?
Because grammatically, it's kind of looking like subordination.
(01:24:23):
But semantically, it's like, oh,these are just conjoined clauses.
So that can be an interesting way to play with that.
Right.
And that's why it's fuzzy.
These things can have multiple meanings.
And as I said, "and" can do weird things in English as well.
Right, right.
So popping back a bit for strategies for subordination,
(01:24:46):
we've talked about conjunctions and nominalization.
Or another thing you could do is have
a different conjugation paradigm for different clause types.
That can be as simple as the standard Indo-European,
like, oh, we'll use the subjunctive.
But you can have other languages like Coptic,
which has many pagesof paradigm forms, which
(01:25:10):
encode specificsubordinate clause types.
And there are other languages that have separate paradigms
that go beyond just having a different mood or something.
Right, right, right.
I've talked a bit about nominalizations thatcan take different kinds of case marking.
(01:25:33):
Right.
Another kind of search,you can-- go ahead.
Let's talk about this, because with case marking,
are we talking about they take an argument role,
and then they get ergative case, or accusative case,
or whatever, case marking to fit the argument that they're in?
(01:25:55):
Yes.
Or locatives are very commonly used for temporal.
OK.
Yeah, that makes sense.
Right, where you smacka dative onto an infinitive,
or a {unclear}, or somekind of verbal noun.
And you get that sort of temporal thing.
That's interesting.
You can do the same sort of thing
by sticking an adposition phrase there, too.
(01:26:20):
Yeah, yeah.
Yeah.
And in addition to different kinds of case marking,
the clause itself might have its normal case marking
capabilities mangled.
- So if you have- Right.
And again, we go back to ancient Greek.
If you have a clause where, because of its dependency,
it is using an infinitive for the verb, right?
(01:26:40):
It's been deranked.
It's an infinitive.
The subject has to be expressed with the accusative.
Right.
OK.
And in English, we do the same.
Like, "he was talking," "I saw him talking."
Yeah.
Right?
He is talking, it has to be the him.
Again, makes it fit into normal English structure better.
(01:27:05):
Right, right.
"I saw he talking," is not--
Yeah, "I saw he talking," that is not the way we do it.
Yeah.
But that is interesting, like, what
happens when you have a pronoun that's in between the two.
(01:27:27):
Obviously, that happens more oftenwith stuff like relative clauses and such.
But it's-- and we can do a whole thing on relative clauses,
because there's so many differentlayers of how those get worked together.
(01:27:50):
But-- We did a relativeclause show a long time ago,
like more than 10 years ago.
Maybe we can revisit it.
Revisiting some of them, nowthat we're both older and know more.
Yeah, I could re-listen to it and see.
I think something that unites all of this stuff
is one thing you need to do when you're doing subordination
(01:28:17):
is you have to think about what role isthis sentence taking in the matrix sentence.
Because-- so you talk about nominalization.
Well, why does nominalization-- why isnominalization used for subordination?
(01:28:37):
Because very often, it is taking theplace of a-- what would normally be a noun.
Either it is taking an argument role,
or it is in some oblique role, object of a preposition,
of an adposition, or a locative, orbenefactive, something like that.
(01:28:58):
But you have to keep in mind that you also
have to consider when a sentence becomessomething akin to an adjective or an adverb.
Especially adverbs, yeah.
Yeah, adverbs all the time.
That's temporal constructions.
You're adding an adverb all the time.
(01:29:21):
Which, I mean, that can be still related.
You could still end upusing a nominalization
with an adpositionor something like that,
or a nominalization with case marking to get that meaning.
Or you could go with something completely different.
But I think when you're doing your different flavors
(01:29:44):
of subordination, that's the thing to keep in mind,
is what role is this sentence playing in the matrix sentence?
Yep.
I'm going to start with an abstract thing, which, George,
I've not put in the notes here, but which I think about a lot.
So sorry about that.
And then we'll move on to the types.
And that is your different kinds of--
sometimes these clauses are called complement clauses.
(01:30:06):
And many of them have a broad distinction
between where the subordinate clause is either a fact,
it could be an activityindicating an extension
in time, or it could beindicating a potential.
Fact-type complements,fact-type subordinate
(01:30:27):
clauses are most likelyto not be deranked.
They're most likely to be balanced.
But they-- or maybe a mood change, but that's about it.
Activity types are very often nominalized.
And then finally, the potential type definitely
is going to have most likely to be reduced tense and aspect
(01:30:50):
marking, reduced pronominal marking,most often using a different mood.
Right, right, right.
And so certain kinds of verb types
are more often going to take one of these or the other.
Attention verbs or perception verbsusually take an activity complement.
"I see him walking to the store."
(01:31:13):
Or they might take facts.
Verbs of thinking take either fact or activity.
Decisions take fact or potential.
Verbs of liking or preferring typically take activities.
So all of these have-- it'sworth thinking about these,
because most languages aregoing to have five, six, seven
(01:31:37):
of these different types, possibly with some subtypes.
You do not need every kind of subordination strategy
that exists in your conlang, nor do you need two.
There will be differencesthat all have to do both the semantics
and thestructure of your language.
Right, right, right.
Any time we're going over all these different options,
(01:32:03):
there is a tendency for a lotof conlangers to do what we
call the kitchen sink andjust throw everything in, right?
But it's useful to know all of your optionsand then decide what you're going to pick.
And it may be the case that choosing one option
(01:32:27):
leads you to have different other varieties.
Maybe you end up with just one strategy that dominates,
like you have one nominalizerthat gets used for everything, right?
Especially if you can have nouns modifying nouns,
(01:32:48):
then you could do your relative clause withthe same thing, all that kind of thing.
But maybe if you go with a converb strategy
and then you decide to make a few different converb forms,
then you're like, OK, well,I've got-- now I've got lots
of different subordinationstrategies going all at once.
And also, you know what?
(01:33:13):
It's always an option to just stick asentence in a sentence in the relevant place.
Just like you mentionedthe Mohawk example, it's only
distinguished on intonation,that you can just do that.
And some languages' relative clauses to an English speaker
look bonkers, becauseit's like, you've just shoved
that sentence inside anotherone with no over-marking.
(01:33:36):
Like, in-place relatives can be tricky.
Or even just not even just stick that sentence in the place,
but also have the noun in therelative clause sentence impeded.
And it's even more wacky-looking to us.
It's a perfectly natural thing to do,but it looks strange to English speakers.
(01:34:00):
OK.
And then my last on the list was converbs,
but we've already talked about those, so we'll skip over.
So let's talk about some commontypes of subordinate conjunctions.
And I've got a listhere, just like cause, like
because, reason, orcause, reason, result, place.
I see where he's leaving.
(01:34:22):
Report, I said that someone said that.
Manner, which could be like a real manner,or a hypothetical, sort of like an as-if.
And then one that's a possible consequence,which you don't use much in English, lest.
Lest.
For fear that.
Yeah.
Some languages have things like this and use them a lot more.
Like, it seems like the Australianlanguages really like their lests.
(01:34:46):
Oh, really?
OK, that's interesting.
Yeah.
But it's not like you sound goofy
if you say that in English.
Yeah.
Chinese really likes resultative clauses.
In Chinese, it's kind of-- if it'sthe same subject, it's like the verbs
(01:35:06):
kind of form a pseudo-compound.
And so that's an interesting strategy that that uses.
And then it also messes with like aspect as well.
Yeah, it does funny things with aspect marking.
A lot is going on with those.
And there's very weird things, like thewhole example of kill-retreat, meaning--
(01:35:36):
do you know that example?
I do not know kill-retreat.
That's great.
OK, so there's an example that you hear a lot in Chinese.
他们杀退了敌人 (tāmen shātuì le dìrén)
So they kill-retreat enemies.
What that means is they were going to killthe enemies and the enemy retreated, right?
(01:36:01):
So kill, which is normally a telic verb,normally has an endpoint, becomes atelic
because it's not realized that you killedall the enemies because they retreated.
So that's-- telicity is another episode that we covered.
(01:36:23):
Yeah, we talked about the telicityone, didn't we, these cancelable--
Yeah, I think we may have covered that example.
But you can end up with weird semantics
when you add a different sort of subordination in there
because they are very good atcanceling things and adding weird--
(01:36:54):
canceling what you expected to be an entailment and stuff
like that, yeah.
So cause, reason, result, place, report, manner--
you have a whole section here on specifically temporal clauses.
Again, all of this-- and that'sjust the tiniest bit of stuff I read.
(01:37:16):
So we can go over this to getpeople thinking about some stuff.
But I really want to do a whole temporal clause show
because there's so much that goes on with these things.
Yeah, so you can have-- soyou're saying, OK, so the past--
so I just want to think about six temporal conjunctions
(01:37:39):
in English that distinguish between past--this is in comparison to the main clause.
It could be past, co-temporal,like at the same time, or the future.
And it could refer to either a point or a duration.
So past point is after, past length is since.
(01:37:59):
Same time point is when, same time duration is while.
Future point is before, future duration is until.
Right.
And we do happen to have a nice regular chart
(01:38:20):
for that in English.
I'm assuming that in many languages that is not the case.
Boy, if you can get it-- Ihave no non-copyrighted
references talkingabout this subject.
But the two or three pages you need
for a full chart of Koreansubordinate clause types is amazing.
(01:38:45):
It's huge.
And there are many temporal things thatall look like they're translated the same.
They can't possibly all be the same,but I've not had a chance to dig into it.
But I will try to find good references for that if I can,
if we can do a temporal subordinate clause episode,
because it's huge, many things, much morerich than most languages I'm used to seeing.
(01:39:08):
So let's talk about how they can overlap and things like that.
Right.
So again, many kinds of subordinate clauses
are polysemous with other kinds of subordinate clauses.
So in English, and in German, and in many languages,
when clauses, which looktemporal, can also be conditional.
(01:39:36):
"When it rains, get an umbrella."
That's not a temporal consideration.
That's a conditional.
While clauses can evolve into conditionalsor clauses of concession, like "although".
And while clauses are one of those
(01:39:58):
that locative case marking is a really popular thing.
After is much less likely to be polysemous with other things.
Before is often deranked and usuallyinvolves negation in complicated ways.
Have we ever talked about expletive negation?
Oh, honestly, I don't know.
(01:40:19):
It's been so long.
We've been doing this for so long.
Expletive negation, it doesn't really exist in English.
But again, it does occur in Greek.
Like, "I was afraid she died," in Greek was, "I was afraid--"
it looks like, "I was afraid she didn't die."
But that's coming from, we believe, a kind of optative.
(01:40:42):
"I was afraid, may she not be dead."
Right, so that's expletive negation.
And there are certain kinds of verbs where that happens.
And the before constructions are often
a situation where you get expletivenegation because of the semantics.
Right, right, right.
And you talk about-- so in English,obviously, we're mentioning these mostly
(01:41:09):
come from prepositions and one interrogative pronoun.
But we have-- all of the strategies
apply to temporal as well as any other.
You could use case marking.
You could use-- I mean, anywhere you coulduse adpositions, you can use case marking.
(01:41:32):
You can use converbs.
You can use any of those strategies.
Or again, just stick a sentence in a sentence.
But it is interesting--
I just want tointerject quickly, with temporal clauses,
there might be iconicity related tothe ordering of the clauses themselves.
So for example, even in a languagethat has very strict head final--
(01:41:56):
and so that normally means your dependent clauses come first--
things like result clauses mightstill come last for iconic reasons.
Right, right, right.
Because the result comes afterbecause that's what happens naturally.
And so the grammar gets changed to move the result clause,
- for example, to the end where normally--- Another thing with that is a result--
(01:42:17):
if you have any kind of broader, either historically
or synchronically in your language,you have any kind of broader-- what is it?
Verb sequences?
What is the term that I'm looking for?
Serial verbs?
(01:42:38):
Serial verb constructions, yes.
If you have any kind of serial verb constructions,
you're very, very likely to have that iconicity happen.
Oh, yeah, yeah, yeah.
Because serial verbconstructions are just like that.
They tend to have that iconicity of temporal order.
That's why Chinese has the resultativeimmediately after the matrix verb.
(01:43:06):
And they also have lots of cases of prepositional phrases
that come from verbs coming before the verb.
It all kind of leads intothat-- my theory, at least.
So that's all super interesting.
(01:43:29):
But my question is, with this ordering,
does that ever happen with past versus future clauses?
Oh.
I didn't see any of that in my reading.
But that would be worth looking to pay attention
to see if there's something like that in the future.
That would be super interesting.
It's kind of--
(01:43:51):
So I did seea paper specifically on the
conceptual metaphors of timeas it relates to subordination.
But I didn't dig into it heavily.
I should have, probably, because that would really
imply that, like, where is the futurein relation to you in your language?
Is it in front of you, or is it behind you?
(01:44:11):
Ooh, we should look at just Aymara conjugations.
I mean, I'm just saying the position of that in the sentence--
because obviously, in English, we can justplace the temporal clause wherever we want.
But it doesn't seem toreally-- not perfectly, but there's
several different slots inthe sense where it can go.
Because adverbs in English just go float on the sentence.
(01:44:35):
But there doesn't seem to be any of that.
Like, before or after will just occur anywhere.
And it's usually not tied to the actual temporal order of events.
But it would be interesting if we hada language that was militant about,
(01:44:57):
like, if it's a before clause, it has to happen.
It has to be placed before the matrix clause.
And if it's an after clause, ithas to be after the matrix clause.
I don't know if that would ever actually happen.
But it would be interesting if wefound a language that's like that.
Yeah, I don't know if there'd be--
it is interesting thatit's normally result
clauses that come infor this sort of dislocation,
(01:45:17):
where cause and effect seems to matter more than time.
Right, right, right.
One last thing that I think,since we're coming close
to time, I want you toget to your favorite thing.
Insubordination.
Insubordination, yes.
I love it so much.
So as I said, there's a recent entire book on it,
(01:45:40):
because it's only a recently investigated phenomenon.
It had all sorts of terrible names.
And now, finally, everyone's converging on insubordination.
I just think it's fun.
They tend to be overlooked, because if you
see what you think is a dependent clause floating
about on its own, it invites the interpretationthat the speaker's made an error.
If you're really focused on the rulesor-- and we'll get to this in a moment--
(01:46:02):
if you're focused on sentences inisolation from the broader conversation.
Oh, OK.
So insubordination is generally agreed
to evolve from full sentences, where a stereotyped main clause
got dropped, or a situation where a clause is implicitly
part of what the other speaker has just said.
So the other speaker has said something
that we're going to pretend is the main clause.
(01:46:24):
And the response-- Isa subordinate clause.
Yes.
So this is like a discourse thing.
Yes.
Conversation analysis, the discourse stuff.
Which a lot of conlangs just don't get to.
So it's important to think about this.
And if you think about how it's used,
it's really clearly related to discourse and pragmatics a lot.
(01:46:47):
So politeness rituals.
"If you could fill out this form, please."
Uh-huh.
Well, then what?
We leave it implied.
This is a normal way in English.
We use insubordination with conditionals to do politeness.
But we also can do it to use threats.
"If you touch my car!"
(01:47:09):
And then, let peopleimagine the rest of that sentence.
Right.
So yes, requests, threats, warnings, politeness rituals.
Also kinds of evaluation.
We don't say this much in English anymore,
but "that you dare tosay that" has an I can't
(01:47:32):
believe, that you dareto say that in front of it.
So evaluation.
So it's all much more emotional,pragmatic, and politeness things.
Conditionals seem to be a super common source for these.
Right.
But not the only.
(01:47:52):
The examples you gave are mostly conditionals, right?
Yeah, in English, they're conditionals.
Oh, and there are other ways youcan use result clauses and other things.
Let me look at Dahlstrom here, because Ihave some examples that I wanted to see.
So here's a fun one.
So Dahlstrom, again, now people are talkingabout insubordination a little bit more.
(01:48:12):
Where in Plains Cree, you can have anindependent question, "Are you hungry?"
And you can also have an "Are you hungry?"with what they call a conjunct inflection.
So a dependent form.
The dependence, the insubordinateversion, is used in a context where you're
(01:48:39):
trying to understand why they're doing something.
Like if someone'srunning to the fridge and
rummaging around, you'duse the subordinate one.
Whereas if you were just like, if someone was visiting,
and you're like, hey, do you want some food,
then you would just use the normal independent one.
So that's a very subtle contextual distinction.
So insubordination doesn't haveto be like these big things either.
(01:49:04):
So I just wanted to talk about that.
There's an entire literature on thisnow that if you search on insubordination,
you will find papers now, an entire book, as I said.
Right, right, right.
And we're going to have links in the show notes.
And if you're on YouTube, in the description to the papers
(01:49:26):
that we are referring to here.
And probably citations for books andstuff that we can't necessarily link to.
OK.
So that's really great.
Are there any other things that you
want to cover before we sign off?
(01:49:46):
Like any particular things?
No, nothing new.
It's just that there are many more--again, as is always the case with typology,
there's more things than probably mostof us have thought about available to us.
Well, it's definitely been helpful to me,and I hope it'll be helpful to listeners.
Because I need to dosome more interesting
(01:50:07):
subordination than I didwith the last language.
Which I think the last language, Idid something with relative clauses.
But then with other subordinations, I waslike, just stick a sentence in a sentence.
And that works.
You don't want to do that every time.
Yeah.
I've been casual about temporal subordinationin pretty much all of my languages.
(01:50:31):
And there's just so much more that could go on there.
I do like using adpositions for temporal clauses.
And having the thoughts about, OK, what spatial metaphor
am I going to use tochoose the adposition
(01:50:52):
that goes with thetemporal clause, right?
That is important.
You've got to work out your conceptual metaphor of time
before you start coming up with grammar for it.
Yeah, you can thinkabout, OK, is time in front
of you or behind you orabove you or below you
or all those things, or however that works out.
(01:51:13):
But again, there's lots of options out there.
Go read the materials that we're goingto link and take a look at those things.
And we'll try to do a few more episodes on some of these.
And I think we probablywill do a whole one
(01:51:34):
on temporal clausessometime in the future.
Yeah.
And conditionals, too, I think, if we've not done it before.
I'm going to-- probably we'll have interviews
and some other guests.
I think I will be moving on and doing
more of these linguistics episodes in the future.
(01:51:55):
So stay tuned for that.
Getting back into that either linguistics episodes
or like conlanging techniques kind of things canbe something I come back to a little bit.
But thank you, William, for being on.
(01:52:15):
Thanks for having me.
And we are-- do you have any final wisdom?
Nope, no wisdom this time.
You can check out William at Linguini.
(01:52:37):
Yes, Lingweenie.org.
Yeah, the Lingweenie blog.
I can -- Acta Lingweenie
I can link to that.
He is also the guy who madethe Conlanger's Thesaurus,
which a lot of peopleout there apparently use.
I'm glad they do.
I'm glad.
Yeah.
(01:52:58):
And the word generator that I use every single day,
lexifer, yes, which you've seen on my streams.
This is the guy who made that one.
But thank you, William, for being on.
And I hope to have you on again to talk,
(01:53:19):
to have another deepdive on another big
linguistics topic, moretypology sharing and things.
And with that, I'm going to say happy conlanging.
Ciao.
OK.
Thank you for listening to Conlangery.
You can find archives and show notes at conlangery.com.
(01:53:41):
And if you'd like to support the show, thebest way to do that is to go to our Patreon
at patreon.com/conlangery.
With just $1, you can get access to exclusivepolls for my Tongues and Runes series.
And also, for $5, you can get everypodcast episode early and ad-free.
(01:54:02):
And now to thank our patrons.
Mintaka, Kenan Kigunda,Conor Stewart Roe, Jesse,
Kaye, Alex Rossell-Hayes,Vyren Patrick, Tabby,
Alexis Hugelmann, Sylvia Sotomayor, Grammar Antifa,
Grakka Grunk, Sigourney Hunter,iloivar Jaana Mentoleum,
(01:54:25):
Niklas Norblad, Anthony Docimo, Jake Penny,
Artifexian, MilesWronkovich, Daniel Quigley,
Paul Roser, our table,Langworm, Nai, and Horn Bory.
Thank you.
Conlangery's theme music is by Null Device.
(01:54:48):
Conlangery is distributed under a Creative Commons Attribution
Non-Commercial Share Alike 4.0 international license.
[MUSIC PLAYING]
[MUSIC PLAYING]
[BLANK_AUDIO]
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.