What question would you ask if you were in my position?

What an amazing question! Hmmm. I am going to take my cue from this clever questioner and flip things around again. Let me ask you, Ling 401 Syntacticians, the following:

What part of linguistics mystifies you the most? What fact or claim or assumption or theoretical decision leaves you totally baffled (because it is too complicated, or too bizarre, or too whatever)?

Does syntactic theory differ depending upon the language under discussion?

The details are likely to appear different, yes. Languages differ in numerous ways. They divide up their lexicons into different categories, and they employ different rules for combining elements of those categories into larger constituents. Emmon Bach, an emeritus professor in linguistics here at UMass, has an accessible essay on the degree to which languages can vary and what the variation means for linguistic theory:

Bach, Emmon. 2002. Structure and Texture: Toward an Understanding of Real Languages. Paper presented at the Western States Conference on Linguistics 2002, University of British Columbia.

Part of what Bach does is marvel at the surprising variation in the nature of the world's languages. But he is well aware that the languages of this earth have an astounding amount in common. The more one looks cross-linguistically, the more one finds the same patterns recurring.

What this means for our purposes in this class is the following: by studying English and its dialects, we will gain tools and techniques that will prove useful for studying other languages, even those that appear to differ wildly from English.

Noam Chomsky often says that a Martian observing people on Earth would conclude that there was just one human language. He says this on the grounds that human languages have much in common. The philosopher John Searle once replied that the Martian might think this at first, but if he stuck around and tried to talk with humans, he'd see that their languages differ quite radically. I feel that Searle is correct, but I would add that if the Martian hung around long enough to learn some theoretical linguistics, he would return to something like his first hypothesis.

Is it possible to write grammars that make learning foreign languages easy?

This would be great! I am not sure whether this is possible. It is something to strive for, though --- perhaps another application for linguistics.

If the class were geared towards creating a prescriptive grammar rather than an 'actual use' grammar, how would it differ? How would our methods be different?

The grammar formalism that we will develop in this class could be used to write prescriptive grammars. As long as the prescriptive claims in question make sense, it is at least conceivable that a real natural language could work according to them. So it is probably a virtue of our approach that we could write a grammar of, say, English-as-William-Safire-says-it-should-be.

The serious differences would be methodological, though. Consider the following sentence, which I used in class on the first day:

(T)

Toni Morrison's genius enables her to create novels that arise from and express the injustices African Americans have endured.

This sentence is condemned by some prescriptivists, because they maintain (P):

(P)

A noun phrase in genitive form (possessive) cannot be the antecedent for a pronoun.

Example (T) violates (P), because Toni Morrison's, a genitive, is the antecedent for her.

No one in the class knew about (P). And sentences like (T) are extremely common, as the prestigious linguist Arnold Zwicky has shown.

As theoretical linguists, we should react to Zwicky's evidence by concluding that hypothesis (P) is false. It is incorrect for English. Progress! Now we know not to write grammars that contain (P)!

But the prescriptivist doesn't work that way. The prescriptivist points to examples like (T) (and all those that Arnold Zwicky cites) and says "Yeah, this is just the kind of stuff I was talking about. People shouldn't say such things. They are incorrect." We see that the prescriptivist's claims cannot be falsified --- they seem even to thrive on counterexamples. But this is just to say that they are not scientific claims. In this class, we are in the business of making testable claims.

Can a knowledge of syntactic theory help one to understand how languages are related historically? Will we do any historical linguistics in this course?

Let me answer this indirectly at first.

We will spend quite a bit of time comparing different dialects of English. We will also branch out a bit to look at other Germanic languages. Why? Because if our grammar formalism is a good one, we should be able to describe the slight differences among these languages by making slight changes to our grammars.

So we'll find connections among languages that are spread out over the globe. What about languages that are spread out over time? Well, from our perspective, a temporal spread is basically the same as a geographic spread. The task is the same: to describe concisely the similarities and differences among related languages. We can ask how German is like Modern English just as we can ask how Old English is like Modern English. We can ask how New England English is like Appalachian English just as we can ask how 1904 English is like 2004 English.

I hadn't planned to do any historical linguistics in this course, but I will be on the lookout for topics now. In any event, as I said above in a similar context, this course should provide you with the tools you need for such investigations.

Is there any biological support for phrase-structure grammar?

This is a tough question, one that many linguists, cognitive scientists, and philosophers have thought about and debated contentiously. I won't try to impose my opinion on you. Instead, I'll just say that I hope phrase-structure rules, trees, and features can provide a complete, accurate description of how language works. I want this description to be accessible to brain scientists. If those scientists find that linguists' descriptions are useful to them, then I guess we can conclude that those descriptions are in some sense present in (or inferable from) our biology.

I disagree with the idea that some English sentences are perfectly grammatical just because they are used often by native speakers of English. For example, Who did you see at the party? is ungrammatical, but most English speakers use such forms. Are linguists okay with that?

The issue here is whether who is grammatical when it is used to question something that is a verbal or prepositional object. The prescribed form for such questions is whom. The question, then, is which of the following we should try to account for in our grammars:

(A)		Who did you see at the party?
(B)		Whom did you see at the party?

Is it true that most English speakers use forms like (A)? Google is helpful here. Here are the results of a Google search for two relevant strings:

(A')		"who did you see"		2,960 hits
(B')		"whom did you see"		888 hits

The Significance Test for Comparing Two Proportions can help us to determine whether this is a statistically significant difference. Set X1 to 2960, X2 to 888, and both N1 and N2 to 3307998701, which is the number of pages that Google currently has indexed (see the bottom of http://www.google.com). The result is a p value of 0, which strongly suggests that the difference is statistically significant: speakers prefer who. (For more on the method I used, see Philip Resnik's post on the topic at Language Log.)

It certainly could have been otherwise. English could have evolved to be a language in which forms like (A) are completely ungrammatical. German is basically such a language. But this is not the path that English took. In fact, the use of whom is in decline. We would end up quite far from our goal of describing the language if we pretended that this was not true. If we tried to say that forms like (A) ought not to be used, then we would have to give up our notion that this is a scientific enterprise.

At this stage of English, we should probably ensure that our grammars generate both (A) and (B), because this accurately reflects the facts of English in 2004. I can think of no other coherent standard for making such decisions.

What practical applications does syntax have?

Here is a list of areas for which a solid understanding of natural language syntax is helpful or indispensable:

• Teaching foreign languages
• Learning foreign languages
• The law (especially copyright law and Constitutional law)
• Copy editing
• Translation
• All kinds of writing and research
• Psychology (both research and counselling)

Can people who speak English as a second language rely on their intuitions to determine grammaticality?

I think that it depends highly on their level of competence. I have studied a lot of German. I speak it passably well, and it's no problem for me to read German, watch movies in German, and so forth. And I have intuitions. But they are so often incorrect that I don't trust them beyond very simple cases. Also, I often get confused and find myself using English intuitions for German sentences.

But if you speak a language every day, then you probably have trustable intuitions. The bottom line, though, is that one should never rely entirely on one's own judgments for anything trickier than very basic examples. It is best to ask at least a few speakers. So my advice to everyone is to get multiple opinions on examples before basing theoretical decisions on them.

Will it ever be the case that I can't participate in an experiment or complete an assignment because I am not a native speaker of English?

Absolutely not. If you ever feel that has happened, then you should contact me immediately so that we can work something out. If you feel that you can't trust your intuitions about English, for whatever reason (because it isn't your native language, because you are tired, because you have a theoretical stake in the outcome), you should ask other speakers. You can even ask me.

Are there any grammatical sentences that we can't find meanings for?

Yes. Amazing though it might seem, there are such sentences. Consider this weird one:

More people have been to Brooklyn than I have.

Meaningless, right? But it sounds okay. For other things like this, check out this page.

Jabberwocky is also usually said to consist of sentences that are grammatical but mean nothing to us.

Do linguists say that a sentence is grammatical whenever a native speaker's intuitions says it is grammatical?

Yes. But, for linguists, there is always a concern: Am I getting this speaker's true intuitions, or are these responses clouded by extra factors (lack of sleep, prescriptive pressure, distractions, a desire to please, etc.)? A lot of messy real-world stands in between us and our true intuitions. This is why it takes some practice to get good at seeing how one's own grammar works, and it is why corpora (e.g., the Internet), speaker surveys, and the like are invaluable tools --- they can serve to support, complicate, or refute claims based on intuitions.

Are my, your, his, her, their determiners?

Yes, I think we can safely say that they are, based on parallels like these:

(D1)		his	bicycle
(D2)		the	bicycle
(D3)		every	bicycle
(D4)		many	bicycles

These parallels are strong evidence that possessive pronouns at least have a lot in common with determiners. This in turn suggests that they are of the same category, at least broadly speaking. We can use 'determiner' to name this class.

Why is natural language syntax so full of exceptions, quirks, and oddities?

It is probably because languages are big, complicated things that are used by a wide variety of people for a wide variety of purposes over a very long and complex time period. Once one is able to look beyond these details, one finds a great deal of regularity. This seems typical of sciences like linguistics. Economics, psychology, and anthropology must also look beyond the apparent mess to see the overarching generalizations.

What is the difference between ungrammaticality and semantic incoherence?

I'll try to answer this with some examples. Let's first look at an example that is semantically incoherent but nonetheless grammatical:

(C)

Colorless green ideas sleep furiously.

Nothing can be both colorless and green. Thus, this is a sentence that is false just in virtue of its basic elements. No imaginable situation makes it true. But it seems grammatical. Syntactically, it is just like Colorful green toads croak furiously.

Check out this answer for even wilder cases of semantic incoherence despite (apparent) grammaticality.

We can also find cases that are semantically coherent but ungrammatical:

(I)

*Me are happy.

We can see exactly what this means. But it is nonetheless an ungrammatical sentence in Standard English, in virtue of the fact that the Standard English 1st person subject pronoun is I, and Standard English 1st person subject pronouns call for the verb form am in predicative structures of this form.

Do you think that speakers' intuitions about sentences are innate, or does it all stem from what we "learn"?

Well, we must be careful. It seems quite likely that we are born with some mental mechanism that permits us to speak, and in turn to have intuitions about, natural languages. Our ability to have intuitions is probably innate.

What is definitely not innate is for us to have the particular intuitions that we have. For instance, I have intuitions about English. But had I been born in Japan, I would have had intuitions about Japanese. Perhaps I would have had no intuitions about English at all. Nothing about our genetic make-up says that we will be able to speak only one, or just a select few, languages. Which languages we end up having intuitions about is something that definitely traces back to environmental factors.

This ability of ours to learn the native language of whatever part of the globe we're born into is a strong argument in favor of the idea that all natural languages share important properties. If they did not, then we could probably expect to find people born with the ability to speak a select few languages but unable to speak the language of their homeland. This seems not to happen, though.

Are the syntactic structures for English (say) different from those of pidgeon languages? Are they different from those of sign languages?

We can safely say that two languages are different if, and only if, they have different structures. So, to the extent that the pidgeon language is different from English, it will have different structures. That is as precise as I can be in for the general case.

The structures for sign language might be quite different from those of English, because the medium of expressions is so different. Issues like constituency and linear order might play out much differently in sign languages than they do in spoken languages.

Rest assured that linguists are hard at work studying sign languages to try to understand them better. You can hear about this from one of the top researchers in this field, Judy Kegl, who is giving a pair of talks here this month (February). The details can be found here.

Why do English speakers persist in teaching prescriptive grammar when so many professional linguists think that this is no the way to study language?

Well, there is a lot to be said for learning prescriptive grammar in some form. In truth, the vast majority of what they teach concerns the conventions for punctuation and spelling. Together, these represent a giant system of conventions that we all must learn in order to be considered competent writers. So, in this area, they are doing you a service by teaching you these things.

The same is true for the injunctions they make about grammar: your knowledge of them could be the deciding factor in your getting (or not getting) a job, for instance. This is just the way that our society works. You can think of your knowledge of prescriptive grammar as intellectual armor. If you know it, you are more likely to get through the professional world on two feet.

What is important to remember is that most of these prescriptive injunctions are arbitrary. They cannot be motivated on objective grounds that make sense. Why should it be possible for a possessive to be the antecedent to a pronoun (an example discussed here)? There is no rational reason why this condition should exist. It certainly will not make you write better. In fact, following it slavishly will probably lower the quality of your writing. One senses that these injunctions are actually testing to see whether you have read the right textbooks, talked with the right people, gone to the right schools, etc. This makes them elitist. The generalizations that we will adopt in this class are different. They will represent things that one knows simply in virtue of being a member of a particular speech community. Our objective, nonarbitrary standard is people's verbal behavior.

Are the exams going to be problems sets like the regular daily assignments, or will they be research papers?

The exams will have the same format as the regular assignments: mostly short answer problems, perhaps with a sprinkling of tree construction and lexical entry composition. Note that you are encouraged to work in groups on the regular assignments (turn in your own assignment, with the names of who you worked with at the top), but your exam work must be independent from start to finish.

You won't have to do outside research for the regular assignments or the exams. But you will have to think long and hard! Enjoy!

Is the final exam one of the two take-home exams?

Yes.

How should I go about studying for exams and digesting the course material?

I see two routes to an outstanding semester in this course:

First, ask lots of questions in class. Make sure that you always know what is going on. Ask questions if you don't know what's going on. Venture hypotheses that seem reasonable to you.

Second, write a lot. Do syntax. On assignments, make sure that you answer the question thoroughly, and then try to push your answer beyond what's required. The practice will serve you well, the effort will impress me, and your acumen for syntactic analysis will sharpen quickly.

How much work will the assignments typically involve?

I can't answer this in a general way. The amount of time you need for the assignments is going to depend heavily on the amount of effort you put in, your background, your level of interest in the problem, and your goals for the semester. I think there will be somewhat more written work in this class than in other courses because there will be relatively little reading.

Are the supplementary readings reserved anywhere?

The textbook by Andrew Radford is on reserve at DuBois. I didn't put it on reserve, but it has that status, and you are welcome to look at it there.

The Cambridge Grammar of the English Language is a reference book at DuBois. The call number is on the syllabus.

I will arrange to have the other two books (McCawley's textbook and the O'Grady et al. introduction to linguistics) put on reserve.

Will the class notes ever be posted ahead of time at the website?

Here's a snippet from a message I sent to a student in this class about just this issue:

Regarding when I post the notes: I will think about how best to do this, and my decision might depend on the nature of the material.

For instance, today I am going to hand out notes, but not until the end of the class period, because I want everyone to be focussed on the discussion. This delay tactic wouldn't do much good if people had downloaded the notes before class.

Another reason why I might often post the notes only right before, or sometime after, the class meeting, is that all the materials for this course are new and under development. Teaching this class is a chance for me to revisit and rethink foundational issues in syntax. It is refreshing for me, and it will make for a more rewarding semester for all of us. But it means that I can't post the notes way ahead of time. If I could, then it would mean that everything was preprepared, nothing unpredictable --- a snooze.

I have a double major, linguistics plus a foreign language. Do you think Ellen Woolford's class would benefit me more than this one, since Ellen plans to talk more about languages other than English?

Will we learn about how children acquire language?

Will this class be difficult if my background in linguistics is limited?

We are starting at the beginning. I am doing my best to presuppose very little about how linguistic analysis works, what terms linguists use, and the like. I hope that the new beginning is refreshing for experienced students, and I hope that it is encouraging to inexperienced ones.

If your experience is limited, I suggest that you ask lots of questions, so that you do not get left behind. An advantage of my decision to design the class as we go is that the students can play a big role in setting the pace.