Linguistics General

it is not?

I think it's a weird sentence, to say the least. Goldberg uses it to argue that constructions hold semantic content, but I find this to be more of a jocular construction than anything. People may use this to be funny, but wordplay isn't really grammatical. I understand what it means, but it doesn't strike me as particularly good English.

meaning is use, and no internal state can be described are two very important insights. then again i do come from philosophy....so yes it's interesting to us!

bro, do you even Colorless green ideas sleep furiously?

The great majority of linguists are mentalists, so we have a very different view of language than Wittgenstein and philosophers generally. We do not accept the thesis that meaning is use, as don't many cognitive scientists nowadays. That view seems somewhat dated, but to be fair to philosophers, it should be noted that linguists don't really offer any alternative theory of meaning.

While the syntax of 'Colorless green ideas sleep furiously' is fine, there is a theta-role violation in 'He sneezed the napkin off the table'. That is, 'sneeze' is typically an intransitive verb. Its use in that sentence is very irregular, and I contend that this is more wordplay than anything.

is there anything that says sneeze can not have a transitive version though?

No, nothing does, but this makes your lexicon more complex. This would imply that your lexicon must include a version of 'sneeze' which does not assign a theme theta-role and one that does. Construction grammar seeks to remedy this by assuming that semantic information is given by syntactic constructions. I'm not sure I agree with that, and as I mentioned previously, I find the example with sneezing somewhat unusual. My guess would be that the capacity to interpret 'He sneezed the napkin off the table' is likely more meta-linguistic than linguistic proper, although this is no more than a stipulation on my part. I have no solid grounds to make this assertion.

If you want a serious introduction to current minimalist syntax, look at David Adger's Core Syntax. For semantics, Heim and Kratzer's Semantics in Generative Grammar. Those two books will get you to about the level of a first year grad student.

Besides that, for culture anyone who wants to get into linguistics should read Syntactic Structures. It's a short little book that basically invented the field of modern linguistics.

You should start off with Saussure's 'General Course on Linguistics'.

I don't agree with that. While Saussure's work is of invaluable from a historical perspective, reading structuralist work in linguistics will not give you access to much of contemporary work in the field. Chomsky is without question the main author anyone should read if they want to know what most linguists do. Of course, transformational syntax is not accepted by everyone, but it remains the dominant theory and should serve as an introduction for all who wish to learn about linguistics.

Just a note: the search engine gives quite a lot of uses of "sneezes his way out of" (another transitive use of the verb).

>semantic information is given by syntactic constructions

I think you mean constructions, not syntactic constructions. Construction grammars usually try to incorporate syntax, semantic and phonological layers into one system and not separate them like mainstream generative grammars. The sneeze sentence is an extension of the elbowing someone something example and trying to figure out why they work.

>I think you mean constructions, not syntactic constructions.
Probably, I've not read much on construction grammar, being very much trained in transformational grammar myself. I find the idea interesting, but I remain unconvinced of the example given about sneezing. Have you had the chance to study construction grammar in much detail yourself?

Yeah, actually. Peter Culicover was in my dissertation committee, so I'm pretty familiar with his and Jackendoff's Simpler Syntax approach.

They both use the Sandy elbowed me the book across the table as a an example of a construction.

What are your thoughts on the divide people like Goldberg present between their work and that of generative grammarians? Do you think the theories are necessarily opposed? Are there ares where they converge? If I'm not mistaken, Jackendoff himself has tended to place himself at the intersection of generative and cognitive linguistics, so I imagine there are people looking to reconcile the various approaches.

I think it is very near-sighted, on both sides. One of the main lesson I learned from Culicover is not to be dogmatic and to be aware of what other people are doing that might be right or at least valuable. My own position is that linguistic research that is data driven and backed with experimental support tends to converge on their results, regardless of the school of thought you subscribe to, which ends up being mostly notational considerations.

If we want linguistics to be more like a science, we should take a page from the hard sciences and allow ourselves to be convinced by other people, even if we don't like them.

I agree with you. It's very important, especially in a young field such as contemporary linguistics, to resist becoming doctrinaires. One way of doing this is to be open to different theoretical frameworks. I doubt anybody would want to drench linguistics in bad blood the way it was in the 70s, when generative and interpretive semantics were at war with one another.

Language is the cornerstone of society and it is very fun to say this because no one knows why and find it frustrating.

I think one of the main causes behind the divide within the field is that Chomsky is such a humongous tool bag. He has generated enough animosity throughout the years that, in such as small field, it has created this sort of tribal structure.

What can I use linguistics for? I want to major in it when I go to big school

Are those balls?

You can use linguistics to do research linguistics and create more linguistics.

no. those are testicles.

I can't stop laughing

IA, deep learning, neurolinguistics... Lots of bridges to other fields.

Any major IT company employs linguists.

If you want a job outside of academia, intelligence defense contractors and forensic linguistics pay linguists really well.

Cool to see someone else who worked with Culicover. He has a great perspective on the whole linguistics as science issue.
What camp would you say you're in now, even if it mostly amounts to just notational preference?

I do syntax/semantics interface so usually Heims & Kratzer, lambda calculus approach on the semantics side and some sort of mainstream generative grammar for the syntax.

I also do a lot of experimental stuff, so R and LaTek take up more of my time than theoretical notation. As long as it makes sense and most people get what I'm trying to say, I leave the structural pirouettes to other people.

Where do you work now (if you don't mind me asking)? I'm a grad student at MIT myself.

I'm a grad student in a really chomskyan program and it seems like the inside scoop is Chomsky's been working on a significant update to minimalism for the past year or so. You know anything about that?

I went to Lancaster and saw this all the time. A good portion of our module on evolutionary development was spent discussing the shtick Dan Everett got over the whole Piraha thing, and the 'danger' of deviating beliefs.

Can't say I have, but it's entirely possible.

To be fair though, Everett's fame also comes from the commotion he caused in generative linguistics. He certainly profited from the polemic.

>the 'danger of deviating beliefs
what does this mean, that non-mainstream opinions are dangerous?

I think he means that you'll be under heavy fire if you challenge fundamental assumptions made by mainstream linguistics.

That's not really a bad thing, is it? I mean, wouldn't the same thing happen if you challenged the mainstream in the hard sciences?

OSU. Spanish Dept.

I don't think it's a bad thing, and you're right that similar things happen in the hard sciences. However, exchanges between schools of thought in linguistics have typically been particularly vitriolic. I think every linguist should read Allen Harris' 'The Linguistics Wars' to see just how much bad blood there was between the different factions in the field.

Everett committed the ultimate sin: he was a SeaOrg member of Scientology (MIT PhD) and went off the reservation in a major way, claiming L Ron was wrong. He needed to be taught a lesson.

I seriously doubt it. Chomsky is not really in the loop of cutting edge minimalism anymore. It sounds more likely that Lasnik and others would be working on it, bouncing ideas off the old man.

>MIT PhD
I would be very surprised if this was true. I could have sworn he went to unicamp.

It's interesting, everyone I keep meeting from Spanish departments is heavily trained in linguistics. Is it common for Spanish departments to be so focused on linguistics?

He's been giving talks on it this spring at a couple places. One of the things it seems like he's working on is finding ways of making multidominance impossible. Because obviously the way everything's set up now it's vague enough to allow multidominance to be a natural possibility.

It is, or at least it was. Things are changing rapidly, and even linguistic sections in Spanish departments are being taken over by SJW. Theoretical linguistics is quickly being replaced by quantitative sociolinguistics approaches and applied linguistics, mostly centered on SLA and bilinguals.

That is what I mean. He is still capable of coming up with good insights, but actually working out the formalism is something he is probably not interested in anymore. If there is any sort of major revision of the minimalist approach, the brunt of the work will fall on others.

That really sucks. But I feel like it's partially linguists' own fault that that's happening, because they've done such a horrible job of disseminating results to the public. If you talk to a random layperson, almost everyone either hasn't got a clue what linguistics is, or thinks it's translation/interpretation, language learning, documenting exotic languages, etc. It's not that there's nothing to report either. So much of what we take for granted is totally unknown to most people, even something as seemingly obvious as the existence of syntactic constituents.

Well, there's nothing wrong with sociolinguistics and applied linguistics per se. I only hope that the SJW takeover doesn't implicate a rise in politically charged research, which is really the last thing research on language needs.

In the case of the current predicament of Hispanic Linguistics sections, the main reason is that we coexist with Culture sections, which have always been highly politicized, but are now on the warpath over the whole Trump, build the wall climate. They have a lot of power in the departments and they have no use for theoretical linguistics. They, on the other hand, love when linguists talk about how awesome Latin American dialects are, and how language reflects social struggle and how bilingual children are smarter than monolinguals. Hence the shift. For instance, I am currently working on a project that involves getting intuitions from Spanish-Swedish bilinguals. When I asked for funding from the department, they basically told me to go fuck myself, since I would not be making any socially related claims.

In our particular case, the enemy is at home.

That is the point. There is. The method is not the point. The main objective is to prove the superiority of poor, oppressed people over white imperialists.

The trouble will start when SJWs start infusing their research with "theory" i.e ideas from Frankfurt school or French deconstructionist continental philosophers that you see in literature, sociology and antropology departments. I don't study the subject but I have dabbled in via computational and mathematical linguistics and from there I've read quite a few linguistics papers (only on syntax).

From my small experience with the linguistics literature, theory doesn't seem to have infected the papers yet to a considerable degree, in fact most papers I read followed the scientific method. All this theory crap has infected so many decent subjects, even really prestigious security and international relations journals are completely infested with it, I really hope linguistics doesn't become the next victim. I wouldn't be surprised if it doesn't though because quite a few French philosophers like Deleuze have attempted to deal with the subject in the usual retarded way that continental philosophers do: grand meaningless and unfalsifiable statements couched in incomprehensible language etc. I think even Chomsky has written about how he was appalled at the scholarship of these types

Oh God, please no. After roughly sixty years of making linguistics as rigorous as possible, this is some terrible news.

relevant
youtube.com/watch?v=OjQA0e0UYzI

Learning languages has little to nothing to do with linguistics.

It's happening. If you want to piss one of these "quantitative sociolinguists" off, just mention how the use of masculine for mixed-gender plurals does not reflect any sociological bias, and that it does not affect society at all.

wut? This doesn't deserve an answer.

>Linguist master race of soft-science.

A Linguist who knows only 1 or 2 Languages is a Brainlet.

A Linguist deserve respect only once He mastered several languages.

Linguistics is the theory behind the pragmatic use, grammatical structure & historical evolution of Languages.

If you know less than 3 languages, this degree is useless, senpai. Philosophy tier.

But if you know how to apply it to several languages you can become a mastermind, worthy of respect, because Linguistic is a powerful tool & intellectual framework to master them.

...

Nobody cares about learning languages in linguistics.

Exactly. Monolingual linguists are frauds, just like biologists who don't have pets, or geologists who don't have rock collections!

So, I took like three linguistics classes in my undergrad but then stopped because I didn't have time to finish up the minor. I want to review and expand my knowledge. Where's a good place to start? I'd like to go over syntax again because I feel like I didn't take it seriously enough when I took a class on it and I've forgotten a good amount of theory behind it. I never took many classes beyond that, either. All I did was the intro class, syntax, and a typology course that was a little disorganized, disappointingly.

Also, my undergrad was in math, is jumping to linguistics for grad school a practical idea? Are there any topics in linguistics (besides those that cross over into computational linguistics) that have interesting applications of mathematical theory?

> Where's a good place to start? I'd like to go over syntax again because I feel like I didn't take it seriously enough when I took a class on it and I've forgotten a good amount of theory behind it
The textbook I used to learn basic syntax was Andrew Carnie's 'Syntax'. I would recommend reading through it.

>Also, my undergrad was in math, is jumping to linguistics for grad school a practical idea?
It depends what you mean by practical. Historically, many linguists have begun their studies in mathematics. It would not doubt give you an edge over other linguists.

>Are there any topics in linguistics (besides those that cross over into computational linguistics) that have interesting applications of mathematical theory?
There's a whole field called Mathematical Linguistics. Here's a book on it.
helsinki.fi/esslli/courses/readers/K54.pdf

As a linguist, I can say that mathematicians are an extremely welcomed addition to the roster of researchers on language. If the field interests you, then by all means join us.

Egō dhídhēmi-dh∂kjō ghawōd bhāmoi "indo european".

theoretical linguistics needs more people who have a background in math. you might want to look at "a formalization of minimalist syntax" by Collins & Stabler. it's a reasonably concise summary of some of the main components of minimalism. you might also want to look at categorial grammar. Morrill's "displacement calculus" is totally beyond me because of the algebra involved, but maybe you could understand it.

Biologists observe life. Geologists observe rocks. Ownership isn't equivalent to or required for observation.

Linguists observe languages. It makes sense for them to observe a few, so they can form abstract ideas of their own.

Certainly knowing a few languages is helpful, but one can do very good linguistic research while being monolingual. Moreover, you can always consult the work of others to gather data on other languages.

>Ownership isn't equivalent to or required for observation.
same goes for languages and linguists. you don't need to speak multiple languages to do linguistics.

Any linguistic theory of a generative bent will always undergenerate and anything else lacks explanatory power. Linguistics is at a stalemate.

I highly recommend Style: The study of Sociolinguistics for anyone looking to work in finance.

>Any linguistic theory of a generative bent will always undergenerate

How do you figure? If anything, generative grammar overgenerates by the observable data by a fair margin.

wut?

Then it's a bullshit major for tumblr dykes

I have a BSc in Neuroscience and am interested in Linguistics, would a masters be a good idea? GPA was pretty crap at right under a 3.0, where could I get in? (had lots of extracurriculars)

Just research programs that focus on neurolinguistics or psycholinguistics. Maryland is a pretty good one. It shouldn't be too hard, especially if you get in touch with one of the professors there and tell them what you are interested in.

It's about building mathematical models of language. Next time you shit your pants because SIRI isn't properly understanding your request to find a transgender South American escort whose into scat, be happy that linguists are working on improving that for you.

Maryland or NYU have good neurolinguistics grad programs.

Yes, it does overgenerate. But it undergenerates by failing to account for even small scale reanalyzation and diachrony. In order to be better able to describe language phenomena it needs to take time into account. Not necessarily language change per se, but the rapid, everyday differences in grammaticality judgements. In linguistic work on your native language you will encounter this often, like when a sentence seems completely wrong to begin with but you rationalize your way to grammaticality. Another similar but different phenomenon is creative use of language in general where use frequently breaks any established "grammar" under scrutiny but nevertheless doesn't raise eyebrows, meaning it's not strictly "performance" in the Chomskyan sense as general rulebreaking. On the other hand what I mean is not the traditional "competence" either as even this term failed to take into account temporality.

Linguistics isn't "learning languages" you dumbfuck.

I agree with you, except in the sense that I would not say that generative grammars undergenerate, more than they are inadequate. They mostly deal with syntax, not with change or probabilistic structures.

>the rapid, everyday differences in grammaticality judgements

I have never heard about this. In fact, grammaticality judgments are fairly robust synchronically. Where have you heard about this ?

>master race of soft science
I don't think linguistics are 'soft' science that much.

One could make a decent argument that all other than physics, chemistry and biology are soft sciences but the reality is that any field of science out there is just field of science. Any of them can be made hard or easy.

Thing with physics, for example, is that the tools are already quite advanced and any major can do some small research about how ball goes here and there quite well. On the other hand such small experiments don't have much to give to the field and the majority do just that, the soft science, since that's all they can and know.

Does linguistics have more soft scientists? Yes. Does that make it soft or worthless? No.

Sure, I never claimed that linguistics being a soft science made it worthless. That would be pretty strange given that it is my area of specialization.

I guess it depends on what level of grammaticality you deal with. In work on information structure for instance, the data quickly become slippery. If you work on it based on personal intuitions, then grammaticality (maybe acceptability is better) will change during the course of your work. Language informants even don't always agree with themselves.

Another, clearer, example would be the example earlier in the thread:
>He sneezed the napkin off the table.

Transitivizing of intransitive verbs are done all the time, and to call it ungrammatical is grammar nazism and way too restrictive as a grammar theory.

degruyter.com/view/j/tlir.2010.27.issue-1/tlir.2010.001/tlir.2010.001.xml

We can call it "grammaticality sliding", where exposure causes the syntactician to accept weirder sentences.

Any fennougrists here?

That is different than saying there is a rapid change in acceptability. The notion of making an intransitive verb transitive is one that exists already. The only innovation is the use of sneeze.

Whenever I claim that a structure is unacceptable, there is always some sort of experimental research behind it, and I focus on the structure, not the particular use of one verb or another. That quick change in acceptability can be attributed to lexical frequencies or priming, two areas where, admittedly, generative syntax does not perform well.

I suppose one of the main cop outs of generative grammar is its reliance on the notions of competence and performance, as it has for the most part claimed to focus on the former and not the latter. While I do think there is some merit to this distinction, I do think it is somewhat abused and all too often used to escape arguments pointing to the shortcomings of generative grammar, the same way pragmatics was used as the waste bin of semantics. If we take the semantics/pragmatics case seriously, we now see that there are important interactions between semantics and pragmatics, and we might expect such interactions between competence and performance as well.

Is that someone who studies Finnic languages or something?

you guys should know that even if this phenomenon is real, there's no reason to believe it has an impact on the reliability traditional acceptability judgments.
sprouse.uconn.edu/papers/SSA.comparison.pdf

>"this holds only for the most statistically powerful of the formal tasks, the forced-choice task, as both magnitude estimation and Likert scale tasks yielded lower convergence
rates"

But, no, I agree with you. The problem arises to a greater degree in the syntax/pragmatics/semantics interfaces where more context is required to see the sentence as grammatical. Therefore it's nice that pragmatics are no longer a wastebasket.

As for generative grammar undergenerating, the undergeneration is partly a result of it's historic, but hopefully about to stop, failure to account for phenomena only licensed in deep context, like different ways of information structure handling and how it can surface in the syntax.

You can check the other tasks yourself, they're not much different

I think Sprouse is talking about how the task used to determine acceptability is mostly irrelevant, not that the effects that cause a difference in acceptability are not there.