Author Topic: Is language really rule-based? [Details inside]  (Read 11084 times)

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1581
  • Country: us
    • English
Re: Is language really rule-based? [Details inside]
« Reply #15 on: January 05, 2014, 05:07:31 PM »
Quote
Wait, where did God come into the picture? and why would the rules have "had to emerge from God"?
Guijarro was, I believe, referring to the (historically popular/canonical) idea that ultimate order comes from God, given that God created the universe, so all principles should eventually be resolved to follow from Godly laws. No one in this thread has supported that, though. He was arguing against it and that therefore any of that type is also incorrect.
Personally I feel that if we replace "God" with "mathematics" (or something else) roughly that idea can apply-- there is something governing the properties of the universe and we can as scientists seek to understand it.
(In the end, it doesn't really matter what that is-- call it God, mathematics, random chance, whatever-- it'll still have the same external properties-- that is, the scientific data we collect and analyze.)


Quote
Also, this turn in the discussion reminds me of the relativity of wrong by Isaac Asimov, which views theories as not "right" or "wrong" but simply more or less wrong, or more or less complete.
Make it gradient or probabilistic if you'd like. I'm fine with that. But it's still roughly the same as (i the same family as) "right or wrong" as opposed to "we just don't know" or "it doesn't matter" or "it depends".




Edit: a very interesting article. I just skimmed the beginning-- I'll see if I have time to look through the rest later. It may address some of the controversial points earlier in this thread.

Quote
It follows that the one thing we can say about our modern "knowledge" is that it is wrong.
I agree!
But this necessarily implies that it would be possible to (in some way or other, with perhaps infinite knowledge) know the truth.
Science is likely the unending struggle to know partial truths-- to always be wrong, but strive to be less wrong. That's fine.
But what would very much bother me would be a situation in which we're not even wrong! That's what Guijarro seems to be suggesting.
I'm uncomfortable not being right or wrong. If I'm wrong, then I can seek to understand more and approach being right (even if I never reach it). But if I am simply not right or wrong, I don't know what to do next; truly, to me, I'm no longer aware of the point of science.
If (to borrow an example from the article) believing that the earth is flat and believing that the earth is round are equally wrong, then surely I am in the wrong field. I should instead be a sailor enjoying the sun, not worrying about these questions, for they have no answers, regardless of whether I, as just one scientist, could hope to discover them.



So in short:
I desperately hope that I am wrong.
I know that I am not (yet) right. And if I'm not wrong, then what is left?
« Last Edit: January 05, 2014, 05:15:32 PM by djr33 »
Welcome to Linguist Forum! If you have any questions, please ask.

Offline MalFet

  • Global Moderator
  • Serious Linguist
  • *****
  • Posts: 282
  • Country: us
Re: Is language really rule-based? [Details inside]
« Reply #16 on: January 05, 2014, 05:51:42 PM »
How can science not be a game if it isn't possible to achieve knowledge?

Who has suggested that it's not possible to achieve knowledge? Certainly not me, and I suspect not Guijarro either (though I'll let him speak for himself). The light that passes through a glass does not somehow become imaginary.

Simple questions:

1. Is there a fact of the matter? Is there, upon sufficient knowledge, an indisputable correct answer to (well-formed) questions? [Are Chomsky's theories necessarily either right or wrong, rather than just an arbitrary perspective, a "glass"?]

2. If not, what is the goal of science?

There is most certainly a fact of the matter, and the part of your question that I bolded is far and away the most frequent pitfall in its pursuit. I'm certainly no physicist, but (at least as far as I understand it) the fact of the matter seems to have something to do with the organized distribution of energy in space/time. Everything else is pattern recognition. Critically, and this is where your strawman comes apart, this does not suggest that patterns are somehow fake.

You could probably summarize the last few thousand years of human thought as an attempt to figure out which ontologies of patterning lead to well-formed questions. A "rule" exists only with respect to a mechanism of governance. This might be a brain, a social convention, a property of the material universe, or any number of other things. But, I don't suspect many people believe that language is ordered by rules that somehow stand autonomous in the universe.

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1581
  • Country: us
    • English
Re: Is language really rule-based? [Details inside]
« Reply #17 on: January 05, 2014, 05:58:45 PM »
Quote
The light that passes through a glass does not somehow become imaginary.
But if the light through one glass means that Chomsky is right, and the light through another glass means that he is wrong, then what's the point of either?
If a theory can simply be right by looking at it with the right glass, then I fail to understand the significance or how, for that matter, it is "right" at all.

Or, was the point merely that Chomsky is correct in one domain but that there are other domains? (That there are many wrong answers to Chomsky's domain, and that his is right, but that there are other domains with other right and wrong answers too.)

Quote
There is most certainly a fact of the matter, and the part of your question that I bolded is far and away the most frequent pitfall in its pursuit. I'm certainly no physicist, but (at least as far as I understand it) the fact of the matter seems to have something to do with the organized distribution of energy in space/time. Everything else is pattern recognition. Critically, and this is where your strawman comes apart, this does not suggest that patterns are somehow fake.
Then why study linguistics? We're just listing patterns?

I suppose I'm confused at this point then. I asked whether there are rules in linguistics/language. If there are not, then what are we studying? If there are, then why are we discussing physics?

Quote
You could probably summarize the last few thousand years of human thought as an attempt to figure out which ontologies of patterning lead to well-formed questions. A "rule" exists only with respect to a mechanism of governance. This might be a brain, a social convention, a property of the material universe, or any number of other things. But, I don't suspect many people believe that language is ordered by rules that somehow stand autonomous in the universe.
I didn't mean to suggest that. But is there not a comfortable middle ground?
Welcome to Linguist Forum! If you have any questions, please ask.

Offline MalFet

  • Global Moderator
  • Serious Linguist
  • *****
  • Posts: 282
  • Country: us
Re: Is language really rule-based? [Details inside]
« Reply #18 on: January 05, 2014, 06:04:07 PM »
But if the light through one glass means that Chomsky is right, and the light through another glass means that he is wrong, then what's the point of either?

Thankfully, neither the world nor science work this way. Jkpate's last post contains two people famous for their ability to explain why.
« Last Edit: January 05, 2014, 06:23:25 PM by MalFet »

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1581
  • Country: us
    • English
Re: Is language really rule-based? [Details inside]
« Reply #19 on: January 05, 2014, 06:13:13 PM »
Then I'm at a complete loss for understanding this statement:
Quote from: Guijarro
I do think that Chomsky's glass is very convenient, but others might prefer other colours.
I thought I understood it a page ago, but maybe I did not.
« Last Edit: January 05, 2014, 06:25:00 PM by djr33 »
Welcome to Linguist Forum! If you have any questions, please ask.

Offline MalFet

  • Global Moderator
  • Serious Linguist
  • *****
  • Posts: 282
  • Country: us
Re: Is language really rule-based? [Details inside]
« Reply #20 on: January 05, 2014, 07:16:34 PM »
Then I'm at a complete loss for understanding this statement:
Quote from: Guijarro
I do think that Chomsky's glass is very convenient, but others might prefer other colours.
I thought I understood it a page ago, but maybe I did not.

I certainly won't presume to speak for Guijarro's meanings, but I don't see any particular conflict. Let me ask this: which do you think is right, biology or chemistry? And if your answer is "it doesn't work like that", why do you think knowledge is impossible and question whether questions can be asked? ;)

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1581
  • Country: us
    • English
Re: Is language really rule-based? [Details inside]
« Reply #21 on: January 05, 2014, 07:26:39 PM »
Well, simple enough then I guess.


But... you're naming fields of study rather than theories, so there is a slight difference. Guijarro was talking about Chomsky's theories, not linguistics (as opposed to chemistry).
Linguistics isn't right or wrong, because it is a field. But Chomsky's theories are right or wrong.
Guijarro's post suggests that with the right glass, Chomsky's theories are right. Does it not?


Therefore, a more well-formed version of your question would be:
Which is right: creationism or evolution?
Or: Minimalism or Construction Grammar?

For those, I believe the question is well formed.
« Last Edit: January 05, 2014, 07:28:23 PM by djr33 »
Welcome to Linguist Forum! If you have any questions, please ask.

Offline jkpate

  • Forum Regulars
  • Linguist
  • *
  • Posts: 130
  • Country: us
    • American English
    • jkpate.net
Re: Is language really rule-based? [Details inside]
« Reply #22 on: January 05, 2014, 08:06:36 PM »
But... you're naming fields of study rather than theories, so there is a slight difference. Guijarro was talking about Chomsky's theories, not linguistics (as opposed to chemistry).

Do fields of study have their own independent existence? Or do we simply say that a theory falls in a field when it seeks to answer questions that practitioners of the field find interesting?
All models are wrong, but some are useful - George E P Box

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1581
  • Country: us
    • English
Re: Is language really rule-based? [Details inside]
« Reply #23 on: January 05, 2014, 08:13:03 PM »
Quote
Do fields of study have their own independent existence?
Without getting too deep about it, yes. Biology is the study of life forms, astronomy is the study of space [and stars], and linguistics is the study of language. Within each there are certainly leading theories, such as evolution within biology-- evolution can be wrong as a theory, but studying life cannot be because it's not the sort of thing that can be wrong.
(Of course we could otherwise debate which fields of study are most productive, interesting or otherwise relevant. But that has nothing to do with "right" or "wrong" in a scientific sense.)

Quote
Or do we simply say that a theory falls in a field when it seeks to answer questions that practitioners of the field find interesting?
Sure. But you're mixing subfields and theories. Syntax is not a theory, it's a subfield. So Syntax is not right or wrong. But a certain theory such as Government and Binding is wrong.

Likewise, we could distinguish theories as claims and subfields as topics-- evolution is arguably not "right or wrong" as a topic of study (religions often study and then disagree with such theories), but it is "right or wrong" as a theory.

I don't think any of that is controversial, but I might have phrased it ambiguously above.




To vastly simplify all of this:
Questions are not right or wrong.
Answers are right or wrong.
[assuming anything is "right or wrong"]


Therefore what Chomsky studies is not "right or wrong", but how he attempts to explain it with a theory is "right or wrong".
The relevant question, then, is which level the "glass" colors.

I believe that we can choose the questions but that some "fact of the matter" actually determines the answer, which we might or might not find or recognize.
« Last Edit: January 05, 2014, 08:16:16 PM by djr33 »
Welcome to Linguist Forum! If you have any questions, please ask.

Offline MalFet

  • Global Moderator
  • Serious Linguist
  • *****
  • Posts: 282
  • Country: us
Re: Is language really rule-based? [Details inside]
« Reply #24 on: January 05, 2014, 08:28:38 PM »
To riff on a very old joke in academia---
Q: What's the difference between a field of study and a theory?
A: 50 million dollars worth of endowments.

You said yesterday in another thread that you don't have much experience outside of g/Generativism. I understand the demands of training, but if you want to pursue the "big questions" (as you seem to suggest that you do), it is going to be very important for you to rectify that. Right now, the way you're framing the tensions in the discipline doesn't really fit with the reality of things.

In short, what characterizes generative linguistics (in contrast to other theories of language) is not chiefly a set of falsifiable, fact-driven claims. Sure, there's a bit of that, but the little of it that exists is not even particularly important to the paradigm as a whole.

Rather, what makes generativism generativism is a tier of abstraction and a scope of attention. Chomsky didn't overthrow Skinner and Bloomfield by disproving their claims but instead by arguing -- persuasively -- that they were asking the wrong questions in the first place. He convinced the world that they had misconstrued the problem so badly that it no longer even mattered whether their answers were correct. The questions themselves were so wrong that even the "right" answers were parochial.

To this end, the "theory" in "generative theory" does not mean "hypothesis". It refers instead to a tier of analysis and a framework of objects. This is very, very explicit as early as Syntactic Structures and was central to Aspects. Chomsky did not *refute* Behaviorism or American Structuralism...rather, he persuaded us that there was something else more useful to think about than stimuli and paradigms. He provided, in other words, a different glass to look through, and people liked the new refractions they saw so much that they largely lost interest in ones they knew before.

And, in good turn, Chomsky's critics do the exact same thing to him. If you spent the next six months reading, say, Dell Hymes, you'd have a hard time filling even one sheet of A4 paper with points of factual disagreement. Instead, you'd find vast and deep divisions about what is interesting about language.
« Last Edit: January 05, 2014, 08:30:25 PM by MalFet »

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1581
  • Country: us
    • English
Re: Is language really rule-based? [Details inside]
« Reply #25 on: January 05, 2014, 09:01:41 PM »
But, you see, now we are actually getting somewhere.

Chomsky has questions, which are judged by appeal.
Chomsky has answers, which are judged by science.



So then it seems that I should define a question, then ask about rules again. I will do that:

In my opinion, the questions of Linguistics are the following:
1. What linguistic behavior exists? [=the data]
2. What makes this behavior possible?

There are many subquestions assumed in that, such as how competence allows performance and so on.

Further, I'd argue that (1) is both obvious and uninteresting-- useful and very entertaining, but unimportant scientifically as an ultimate goal. I like descriptivism, but if that were the whole purpose I might as well go back to my other hobby of film making-- make some documentaries about people talking. Done.

That leaves us with (2) as, I fully defend, the central question for all linguists. And with that simple question, the rest of the field should follow. Certainly we can have different interpretations-- some would be interested in how linguistic behavior changes over time (historical linguistics), others in how linguistic behavior functions socially (sociolinguistics), and others (like me) are fascinating by the simple question: how do we use language to convey meaning, how do our minds do that, what is language as a system?


To argue by counterexample, there are many things that do not work like language and many organisms that do not use language-- somehow humans are not starfish, beetles or hedgehogs, and somehow language is not the same as gravity, cars or chess.

Up to that point, I would assume we can all agree.

So, perhaps just out of personal preference, I would like to ask the following question:
Does the way that language works depend on rules being part of that system? Is language what it is, and are human minds what they are because describable rules exist?


In the end... I think it's reasonable to ask if rules exist, within the domain of asking how language is possible.




---

Now, perhaps we should just move on from that. Generally the consensus seems to be that if we look at language with the right "glass" then there are rules to describe, or at least patterns.

So, considering the first question answered "yes" (with some footnotes), the next questions are:
1. What kinds of rules/patterns?
(See jkpate's post on page 1 for a good list of candidates.)
2. And do those rules/patterns result in an interesting scientific analysis? Or are they mere descriptions?


Personally, my answers at the moment (which may be wrong!) are:
There are no rules on linguistic form, just tendencies based on what is useful and constraints based on factors like linearization.
There are recurring patterns to the structure of the meanings these forms encode-- for example, I'm confident that modification is a real property of languages.


Does that seem reasonable? Any good arguments against that position?
Welcome to Linguist Forum! If you have any questions, please ask.

Offline MalFet

  • Global Moderator
  • Serious Linguist
  • *****
  • Posts: 282
  • Country: us
Re: Is language really rule-based? [Details inside]
« Reply #26 on: January 05, 2014, 09:07:38 PM »
But, you see, now we are actually getting somewhere.

Chomsky has questions, which are judged by appeal.
Chomsky has answers, which are judged by science.

I'm not sure where you're getting this, but it's certainly not anything I said (or at least nothing I meant to say), and it's not an accurate reflection of the situation.

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1581
  • Country: us
    • English
Re: Is language really rule-based? [Details inside]
« Reply #27 on: January 05, 2014, 09:10:22 PM »
Quote
I'm not sure where you're getting this, but it's certainly not anything I said (or at least nothing I meant to say), and it's not an accurate reflection of the situation.
Why not? Really?

You did say the following:
Quote
Chomsky didn't overthrow Skinner and Bloomfield by disproving their claims but instead by arguing -- persuasively -- that they were asking the wrong questions in the first place. He convinced the world that they had misconstrued the problem so badly that it no longer even mattered whether their answers were correct. The questions themselves were so wrong that even the "right" answers were parochial.
That leads to what I said, no?
Quote
Chomsky has questions, which are judged by appeal.

So, do you disagree about the second part, that scientific answers are judged by science?




--
Or, to be fair, perhaps you just disagree with my phrasing.
« Last Edit: January 05, 2014, 09:12:53 PM by djr33 »
Welcome to Linguist Forum! If you have any questions, please ask.

Offline MalFet

  • Global Moderator
  • Serious Linguist
  • *****
  • Posts: 282
  • Country: us
Re: Is language really rule-based? [Details inside]
« Reply #28 on: January 05, 2014, 09:49:47 PM »
That leads to what I said, no?

I don't see why it should. This distinction you're pitching between "appeal" and "science" is another false dichotomy. It just doesn't work like that, and unfortunately I don't know how else to explain it. If you'd like references to major works in other paradigms, I'd be happy to provide them.

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1581
  • Country: us
    • English
Re: Is language really rule-based? [Details inside]
« Reply #29 on: January 05, 2014, 10:01:43 PM »
Quote
This distinction you're pitching between "appeal" and "science" is another false dichotomy.
Appeal is subjective. For example, there is nothing technically wrong with every scientist in the world studying the mating behavior of gnats, but it's unlikely that would happen because people wouldn't want to do that-- funding wouldn't exist, interesting wouldn't exist, and eventually someone would decide to study something else.
That's completely unrelated to the scientific validity of a given theory about the mating behavior of gnats. It may have no appeal (or a lot), and either way it's still right or wrong, based on whether it's right or wrong.



In the end, we may simply disagree.


Let me make a point that was discussed in one of my syntax courses last summer at the LSA summer institute:
Science provides no roadmap to what a scientist should study or which theories should be tested. It does, however, provide a way to test whether a given theory is false.
In science, we have falsifiability as an axiom. We have no parallel axiom for measuring value.

Thus:
Quote
The questions themselves were so wrong that even the "right" answers were parochial.
Scientifically speaking, the "right" answers are the "right" answers, and that's that.
If they are not actually right because of some underlying problem in the theory, then those theories are false. That's not how you described it.
Either way, if an answer is right, then it is right. That's how science works.

The next and perhaps more important question is whether those answers are useful.




If this is what you intend to say, great. If not, then I'm happy to disagree. But I'd need a lot of convincing to change my beliefs on any of that. I'm not saying a roadmap isn't something I'd like, though!



Falsification is deductive and solvable, given a specific hypothesis.
"Does Data X falsify Theory Y?" -- can be solved by a computer, just add data!
Everything else is inductive and unsolvable (in a strict sense).
"What question would find the {best, next, important} theory? How would I falsify that theory?" -- cannot be solved by a computer or deterministically by a scientist.








--
Edit:
Quote
If you'd like references to major works in other paradigms, I'd be happy to provide them.
Ok, sure. What other paradigms?
« Last Edit: January 05, 2014, 10:17:38 PM by djr33 »
Welcome to Linguist Forum! If you have any questions, please ask.