The hornet's nest of grammatical complexity

[bradrn]

In any case, I note that we’ve drifted off the original topic a bit. We started off by disputing whether ‘complexity’ and ‘ambiguity’ are separate things, and we still haven’t resolved that.

I'm not sure I agree with the claim that English is more prone to ambiguity. Yes, we have a whole thread of lots of amusing examples. But-- we're English speakers, we enjoy that and we read enough stuff that we can note down fun examples. If it was really the case that every other sentence of English was ambiguous, the examples wouldn't be so remarkable.

Now, Sanskrit is highly morphologically marked: complex verbs, a full case system, gender, etc. So you or Moose would expect it to lack ambiguity, right? Yet there's an epic poem, Dvisandhana, which cleverly uses synonyms to simultaneously retell the Mahabharata and the Ramayana.

Or take Chinese, where poets love to play with understatement and ambiguity. E.g. Wáng Wéi's most famous poem:

空山不见人
empty mountain not see person
但闻人语响
however hear person words/speak sound
返景入深林
return brightness/view/situation enter deep/thick forest

You could write a book on how to translate these 15 syllables-- indeed, Eliot Weinberger and Octavio Paz wrote such a book. Is Chinese the summit of ambiguity? But you could also write this:

有一座山，空无一人。 我们看不到任何人，但我们听到了声音。 黄昏的光芒穿透了森林深处，再次照耀在森林地面上的青苔上。
There is a mountain that is empty. We couldn't see anyone, but we heard voices. The light of dusk penetrated the depths of the forest and shone again on the moss on the forest floor.

Of course, it is entirely possible to be ambiguous in Sanskrit, or concrete in Mandarin. I’m not disputing this. All languages allow people to be as concrete or as ambiguous as they like. However, languages certainly have a ‘preferred’ level of ambiguity in natural discourse.

For instance, I don’t know Chinese, but I have read a bit about Lao recently, and natural Lao sentences tend to be very ambiguous indeed, if you don’t know the context (Enfield 2008, in The Tai-Kadai Languages ed. Diller, Edmonson & Luo):

man²

3SG

bang³

block.from.view

hùan

house

He’s blocked from view by the house / He’s blocking the house from view

phuak⁴

group

juu¹

be.at

nam²

accompany

thaang²

road

ka⁰

FOC.PCL

qaw³

take

Thoseᵢ along the road, [theyⱼ] took ∅ᵢ / Those along the road took [them/it]

tamluat⁵/mak¹

police

dêj²

like

phu⁰-saaw³

PCL-girls

tòòn³

time

nan⁴

DEM.NONPROX

Policeᵢ, [theyⱼ] liked [themᵢ] you know, the girlsⱼ back then / Police liked [them] you know, the girls back then

Compared to Lao, English is much less ambiguous, precisely because its sentences are more complex: most of the time, it requires that you mark definiteness, grammatical relations (in three separate ways!), tense and aspect, amongst other categories. But spontaneous English sentences are still not as complex as German, which requires even more marking. And German is yet less complex than Japhug or Navajo.

---

On reflection, it seems like we’ve been equivocating between two different notions of complexity. Firstly, there’s the things each language requires marking: ‘very little’ in the case of Lao, ‘definiteness, tense etc.’ for English, ‘definiteness, gender, tense etc.’ for German, and so on. This is basically a measure of how much is needed in order to be at least vaguely comprehensible to the people around you (This is this post has been discussing so far.) Then there’s the things needed to sound like a 100% native speaker: things like adverb placement and idioms in English, SVC components in Kalam, and so on. You’re still understandable if you miss these out, but people will look at you strangely until you master them. (This is what I discussed in some of my previous posts.) Then there would be two different notions of complexity.

Except it’s even worse than that, because we don’t really have anything like this neat binary division. We can make a case that article choice is a ‘required category’ in a way that adverb placement isn’t, but ESL speakers leave out their articles all the time and they’re still pretty easy to understand. On the other hand, it does seem like the presence of definiteness marking should make a language ‘more complex’ in some way. Perhaps it would be best to talk about complexity in terms of a learning curve: for language L you need to know X% of language features in order to be Y% comprehensible.

[zompist]

On reflection, it seems like we’ve been equivocating between two different notions of complexity. Firstly, there’s the things each language requires marking: ‘very little’ in the case of Lao, ‘definiteness, tense etc.’ for English, ‘definiteness, gender, tense etc.’ for German, and so on. This is basically a measure of how much is needed in order to be at least vaguely comprehensible to the people around you (This is this post has been discussing so far.)

OK, that's certainly an interesting aspect of language... I'm not sure it has a set name... requiredness? mandatoriness?

I don't think it's a type of complexity... if nothing else, because it's conceptually very simple! If you have to indicate evidentiality in Quechua, or tense in Latin, or aspect in Mandarin, or plurality in English... well, you just do. It's not some sort of burden on the native speaker.

(Well, it can be if it has a lot of irregularity. But that's complexity in the ordinary sense-- lots of things to learn and memorize.)

I think it comes up in these discussions because if you're not used to a requirement, it's a difficulty for the language learner.

And as several people have pointed out, not having a required parameter is not necessarily "simple." E.g. English does not have required evidentiality... but that means that the concept has to be expressed via various constructions, or nuances added to other features, or various lexemes. Plus an English textbook might or might not think to cover any of these things. By contrast the required evidentials in Quechua are pretty simple to explain.

[Linguoboy]

Then there’s the things needed to sound like a 100% native speaker: things like adverb placement and idioms in English, SVC components in Kalam, and so on. You’re still understandable if you miss these out, but people will look at you strangely until you master them.

Just I can't agree with this statement. Grammatical scope is a stuff.

[bradrn]

On reflection, it seems like we’ve been equivocating between two different notions of complexity. Firstly, there’s the things each language requires marking: ‘very little’ in the case of Lao, ‘definiteness, tense etc.’ for English, ‘definiteness, gender, tense etc.’ for German, and so on. This is basically a measure of how much is needed in order to be at least vaguely comprehensible to the people around you (This is this post has been discussing so far.)

OK, that's certainly an interesting aspect of language... I'm not sure it has a set name... requiredness? mandatoriness?

I don't think it's a type of complexity... if nothing else, because it's conceptually very simple! If you have to indicate evidentiality in Quechua, or tense in Latin, or aspect in Mandarin, or plurality in English... well, you just do. It's not some sort of burden on the native speaker.

If I correctly understand what I’ve read, many people do in fact consider this a type of complexity. It certainly can cause problems even for native speakers: it’s hard to write a text which is ambiguous with plurality in English, or with tense in Latin, or with evidentiality in Quechua. (I recently read a book noted for how it never specified the protagonist’s gender. This would be trivial in Mandarin, and obviously contrived in Latin.) That being said, it certainly isn’t a particularly big burden most of the time, and I tend to agree with you in doubting its importance in evaluating complexity.

(Well, it can be if it has a lot of irregularity. But that's complexity in the ordinary sense-- lots of things to learn and memorize.)

I find it interesting that you say this: people like McWhorter tend to lump in morphophonology with all the other types of linguistic complexity. I feel it’s clearer to separate it out, as you do here.

And as several people have pointed out, not having a required parameter is not necessarily "simple." E.g. English does not have required evidentiality... but that means that the concept has to be expressed via various constructions, or nuances added to other features, or various lexemes. Plus an English textbook might or might not think to cover any of these things. By contrast the required evidentials in Quechua are pretty simple to explain.

If other people have already mentioned this, I must not have noticed. But now that I do think about it, this is a good point. Especially interesting is that English replacements for, say, evidentials tend to be set phrases, so they do undoubtedly contribute to complexity. (e.g. ‘I heard the bank was robbed yesterday’, starting with the standard construction for auditory or hearsay evidence, is more idiomatic than ‘I was told the bank was robbed yesterday’.) But on the other hand, how does that compare with being forced to mark your source of evidence with every single clause? There is still a sense in which English is simpler here, because you don’t have to know the standard evidential constructions in order to construct coherent sentences. And this comes back to the problem of dealing with several different kinds of linguistic complexity, like I was saying before.

Then there’s the things needed to sound like a 100% native speaker: things like adverb placement and idioms in English, SVC components in Kalam, and so on. You’re still understandable if you miss these out, but people will look at you strangely until you master them.

Just I can't agree with this statement. Grammatical scope is a stuff.

…and this is where we see the problem with trying to define any simple notion of ‘complexity’ or ‘grammaticality’, as I said literally one paragraph below that quote. Even adverb placement is heterogenous with respect to grammaticality. (‘Just I can’t agree’ is obviously ungrammatical; ‘Still I can’t agree’ less so.)

[Linguoboy]

…and this is where we see the problem with trying to define any simple notion of ‘complexity’ or ‘grammaticality’, as I said literally one paragraph below that quote. Even adverb placement is heterogenous with respect to grammaticality. (‘Just I can’t agree’ is obviously ungrammatical; ‘Still I can’t agree’ less so.)

I'ma disagree wi you again, dawg. "Just I can't agree" parses similar to "Only I can't agree" IMD.

I think the real issue isn't just that it's easy to produce ungrammatical or unidiomatic results, but that it's easy to produce results that are grammatical and even idiomatic but don't mean what the speaker intended. I can't tell you how many times I've corrected a sentence from an L2 speaker, guessed what they meant, and been wrong, because the interaction here between syntax and semantics is (dare I say it?) quite complex.

[Kuchigakatai]

The second example is actually a good one: "Still I can't agree" should be grammatical and idiomatic for most (North American?) native speakers, as they'd add a small intonation break or pause after "still", i.e. "Still, I can't agree", meaning something similar to "And yet, I can't agree". And of course, this doesn't mean the same as "I still can't agree", so a non-native writing "Still I can't agree" (while having the meaning of "I still can't agree" in mind) would convey an unintended meaning...

[Raphael]

The second example is actually a good one: "Still I can't agree" should be grammatical and idiomatic for most (North American?) native speakers, as they'd add a small intonation break or pause after "still", i.e. "Still, I can't agree", meaning something similar to "And yet, I can't agree". And of course, this doesn't mean the same as "I still can't agree", so a non-native writing "Still I can't agree" (while having the meaning of "I still can't agree" in mind) would convey an unintended meaning...

Now I'm reminded of the classic juxtaposition between "Woman, without her man, is nothing!" and "Woman! Without her, Man is nothing!".

[alynnidalar]

The second example is actually a good one: "Still I can't agree" should be grammatical and idiomatic for most (North American?) native speakers, as they'd add a small intonation break or pause after "still", i.e. "Still, I can't agree", meaning something similar to "And yet, I can't agree". And of course, this doesn't mean the same as "I still can't agree", so a non-native writing "Still I can't agree" (while having the meaning of "I still can't agree" in mind) would convey an unintended meaning...

"Still I can't agree" with no intonation break or comma seems perfectly grammatical to me, if a little poetic, and would absolutely mean "I still can't agree" when said in that way.

Found a couple examples on Google Books:

I said to her only the other day about Francisco, It's no good crying in front of them, it only puts them off. And he won't marry you, I said, no, not if you were the Virgin Mary dear he wouldn't. It's always this family business. But still I can't complain about Enrique. Apart from his family Enrique is one in a thousand.

I know we both had equal parts in everything, but still I can't fathom how someone I consider to be my best friend, my confidant, my love, and my future, can also be the person I'm fighting with

(you don't need to nitpick these examples, I'm sure you could read them either way! Just pointing out that it does work for me.)

[Travis B.]

The second example is actually a good one: "Still I can't agree" should be grammatical and idiomatic for most (North American?) native speakers, as they'd add a small intonation break or pause after "still", i.e. "Still, I can't agree", meaning something similar to "And yet, I can't agree". And of course, this doesn't mean the same as "I still can't agree", so a non-native writing "Still I can't agree" (while having the meaning of "I still can't agree" in mind) would convey an unintended meaning...

"Still I can't agree" with no intonation break or comma seems perfectly grammatical to me, if a little poetic, and would absolutely mean "I still can't agree" when said in that way.

Found a couple examples on Google Books:

I said to her only the other day about Francisco, It's no good crying in front of them, it only puts them off. And he won't marry you, I said, no, not if you were the Virgin Mary dear he wouldn't. It's always this family business. But still I can't complain about Enrique. Apart from his family Enrique is one in a thousand.

I know we both had equal parts in everything, but still I can't fathom how someone I consider to be my best friend, my confidant, my love, and my future, can also be the person I'm fighting with

(you don't need to nitpick these examples, I'm sure you could read them either way! Just pointing out that it does work for me.)

One nitpick there, though is that examples use "but still I can't agree" which is slightly different from "still I can't agree" and is perfectly kosher to me (whereas "still I can't agree" at the beginning of an isolated sentence is a tad poetic to me).