I remember the mille-feuilles...

Geez, some really strong wind here.

Gusts and squalls.

Ha, there's @Kosmonaut right when I need him.

Hm, not really on-topic.

More like Linguistics, yet again.

Or actually, Teaching.

I don't know, he just uses the word "English" at one point in the beginning.

Otherwise it wouldn't be connected to English at all.

But I dunno.

I feel the same way. He ducks and dodges just enough...

I left a comment.

@Kos: I have a question, not sure whether you know a lot about this. Modern linguistics is all about function: whenever what looks like a noun functions as an adverb, ML will call it an adverb, period. However, most of the time the fact that it still looks like a noun means that it was once a noun in function as well, but that its use in a particular context has transformed a bit. When studying historical development of language, ML terminology might thus obscure the patterns.

Is it widely recognised that this is a disadvantage of ML, though I am not saying it matters a great deal? Or does ML have some great counter argument that utterly destroys what I have just said?

Can you explain the purpose of calling something one part of speech when it is actually another part of speech?

I don't quite follow the logic.

Well, consider the following sentence:

Considering your unhealthy skin tone, you should probably stop eating caviar, don't you think?

OK

ML would probably call "considering" a preposition, instead of a participle.

@Cerb, really? i'd still call it a participle

especially since adverbial phrases are one of the classical purposes of participles

I wouldn't have called it a preposition.

but it's good friday, time to get to liturgy. see you all on monday.

A good friday to you, sir

@JSB: Really?? This would be one of the few instances where I think ML terminology, if that's what it is, would be a great improvement: if it were a participle, it would have to have a head noun/pronoun, which would normally be the subject of the sentence, or perhaps some other word present in the sentence.

Not everyone is a heathen here. Nice.

@JSB: Okay have fun in church!

@Cerberus Do you feel that each word should have only one part of speech, no matter how it is used?

@Kos: Hmm I'd have to think about that... but I'd probably say that it should have one part of speech that describes it most adequately?

I don't really see much purpose in assigning it one part of speech.

What can you predict about the word, what can you do with that information?

Oh! No, of course not: I meant one choice should be made for each instance of the word.

The choice would be based on context.

Isn't that what we do in linguistics?

In the context of the sentence, how is it used?

In my sentence, "considering" could be a preposition; in a sentence like "considering the nature of gravity, I think we should drop it", it would be a participle.

@Kos: Yes, that is what you do; I was giving an example where the merits of modern linguistics shine, at least in my humble opinion.

@Cerberus Well, it would depend on the framework you are using.

Sorry, I don't think I read your first paragraph carefully...

Right, but there seems to be some kind of standard framework that people are using on this website, with lots of new terms and acronyms.

Were you saying that modern linguistics treats lexical categories as things that are not context sensitive?

On the contrary!

So what is the disadvantage that modern linguistics has?

I was saying they were going very far in considering context, so far as to obscure our view of historical processes, where context matters less and the diachronic course of a word/form through history matters more.

How do we obscure our view of historical processes?

For example, if I say "considering" is a preposition in that sentence, fine; but suppose its use as a non-preposition died out in the next 100 years. If our offspring is then stuck with "preposition", it will be harder for them to quickly see the connection between other participles and their preposition "considering".

But linguistics wouldn't label it only as a preposition.

But wouldn't linguists of the 22th century do so, if that were the only way it was still used?

Perhaps this is not a great example, because the "participle" usage is still very much alive.

Etymologies would still have all of the old uses of a word.

@Cerberus Take some actual preposition, like our favorite "despite" or something.

Consider the subjunctive in "be that as it may, you are still a witch". If we stop calling it a subjunctive, but start calling it a base form exclusively, it will be harder for people to analyse where that form came from and compare it with other subjunctives.

It would get silly after a while to retain a word's earliest lexical category as its current one

@Cerberus Sure, but nobody would do that — it is subjunctive mood.

@Kos: I am not saying it shouldn't be done; I am just saying it is a disadvantage. It may very well be worth it in many cases.

I just don't see a disadvantage.

For example, a lot of affixes start out as other parts of speech; the "hood" in "neighborhood" used to be a noun.

Yes. And because we consider it a mere affix, its origin has been obscured. I didn't even know it was once a noun.

But, what is the value of calling it a noun?

The value would be that it would be easier to compare its use in that word to its use elsewhere.

Another example, the word "til" is related to "Ziel", a noun meaning "goal"

I don't know if Dutch has that word

I know it only in German, I think not.

That is a good example: because I didn't know -til came from a noun, it never occurred to me that it could be related to Ziel, which it might have otherwise.

Doel?

Oh!

Yes!

That is probably related.

+1 For Reg.

I think knowing how the word actually works synchronically is useful to everyone, but this historical info is useful mostly just for linguists and for people who want to get a better appreciation of language

People who look at diachronic change are aware of these shifts in lexical category.

@Kos: Well I was considering it from the perspective of linguistics.

But only for historical linguists like me.

I think for a syntactician looking at modern English, it doesn't really give them any helpful information.

But it also becomes a bit harder for you, I'd think?

Okay, well, do you understand what I mean by a disadvantage? I wasn't even saying it was significant.

Maybe I don't know what you mean by disadvantage.

Something bad that the other method doesn't have.

It also has great advantages, of course.

I just don't see anything bad about it.

Nothing is hidden.

@RegDwight, @Kosmonaut: would the diagramming question be better for meta.EL&U?

It is still less accessible.

Hi Mitch

@Mitch It should really go in linguistics.SE, if only it existed...

Isn't there a button diamonds can press?

Yes.

Oh!

Then they should.

I am pressing buttons all the time, but Linguistics still doesn't go live.

As RegDwight said, it is a bit of dodge and weave here.

You are apparently unable to press its buttons then.

but I only know of the Reed-Kellog method ever being used for English.

@Cerberus The info about till can be found in the OED or anything similar.

Yeah, someone actually suggested a tag for all linguistics related questions that we close.

But that would be a meta tag, so that's not an option.

which should then be immediate questions once Ling.SE goes beta.

@Mitch I have never heard of Chomskyan syntax or any other types of trees being taught in any English class in the US.

@Mitch Well, I kind of try to maintain a list.

@Kos: Yes but that still gives one less of an easy overview than if we'd been taught to call the -til part a noun—which I am not proposing we should, by the way: this just for the sake of argument.

Well, I guess the chomskyan syntax trees would have been a question ('is it taught at all in secondary school in the US?')

@Cerberus I think the disadvantage of doing that would be so incredibly huge, linguistics and language education would be a mess.

@Mitch: We are certainly not taught such trees outside university in Holland. Would have been cool though.

Nearly all of the prepositions used to be nouns and whatnot.

@Kos: Agreed; but that doesn't mean the opposite doesn't have its (perhaps small) downsides too.

@Mitch the thing is, I think we've closed quite a few teaching-related questions, even if they specifically focused on English.

Cause it's kind of more about politics and culture and whatnot.

The meta argument would then be 'where else to ask such a question?' which of course has its own problems.

Yeah, that's always hard.

I mean, we had quite a few questions that got closed but would be on-topic for Writers now.

Old stuff can't be migrated, because the OP is usually no longer around to accept, etc.

@Kosmonaut: Nearly all prepositions? In English (even in, to, of, at, by, on)? In most languages?

@Cerberus But, as far as I can see, the only downside is that you wouldn't know what part of speech the word was in its earliest known existence. So you wouldn't know that till was a noun in Proto-Indo-European, unless you look it up.

@Kos: OK if you will agree that it is an infinitesimal downside, I am happy.

@Kosmonaut: and in some sense you don't really -know- such things about PIE.

@Mitch True, it's reconstructed.

@Mitch: Most theories say something similar about prepositions in Ancient Greek: they were all once adverbs, and can still be so used on occasion.

@Mitch Not necessarily the most basic ones, but prepositions often do come about from other parts of speech, yes.

Well "often" in the sense of thousands of years, gradually :)

The same applies to inflection.

In a way the distinction between inflectional and synthetic languages is blurred.

Yes, that's true.

Therefore I propose that we call all words of all the old lexical categories simply "roots".

That obscures historical developments the least.

Or perhaps "word" would be nice all-replacing lexical category.

Call words "words"?

No, too inexact. Better call them all utterances.

Next thing you know, someone will submit that there's no such thing as "words".

'utterances' always sounds like it's referring to grunting to me.

Exactly!

@RegDwight Some people do say things like that.

I know.

Just listen to some Russians talking. Can you make out a single "word"?

Just some grunting, really.

But seriously: the downside I just described becomes more serious when modern linguists change the old lexical categories more liberally, as when they call "until death do us part" simply an adverb.

Aha, so that's where the wind comes from.

This is a bit more extreme than what I have observed in nature, but it might one day come to pass.

I have in fact observed their calling prepositional phrases (what's wrong with that term, by the way) simply adverbs, like "in the water".

You mean in English?

Yes!

@Cerberus Linguists don't do that.

Cause there are enough languages where "in the water" would be a single grunt.

Maybe overly aggressive dictionaries?

@Kosmonaut No true Scotsman...

@Kos: Okay well I have heard people say that who also used lots of frightening new terms and especially acronyms.

@RegDwight I have never heard of anyone doing that.