Syntax - a multi-perspective introduction

Ceresz · Post by **Ceresz** » Sun Jan 15, 2006 10:53 am

Question:

I am a begginer when it comes to syntax, so I was wondering if this is right?

Space Dracula · Post by **Space Dracula** » Sun Jan 15, 2006 12:45 pm

finlay wrote:

Spack, two of your trees aren't loading.

At least one of my trees are hosted by Miekko there, whose server is his computer, and is thus prone to issues.

Also, everyone's tree analyses for homework are correct, which means that you're clever and I explain things better than I think.

Also, adjectives are usually ignored in LFG in trees, because most LFGers think there's nothing interesting to be found there; rather, how the adjectives relate to the words functionally is of more interest. For example, adjectives in Warlpiri, as shown before, can be completely disconnected from their heads and still make sense. What use, then, are trees (and remember LFG trees are without movement) in describing it?

Nonetheless, adjectives in English in LFG are usually analyzed this way:

Nuntar · Post by **Nuntar** » Sun Jan 15, 2006 12:50 pm

Space Dracula wrote:Nonetheless, adjectives in English in LFG are usually analyzed this way:

Then how do you account for the difference between "the little old man" (which contains "old man" as a sub-unit) and "the little, old man" (which doesn't)?

Space Dracula · Post by **Space Dracula** » Sun Jan 15, 2006 1:49 pm

Nuntar wrote:
Space Dracula wrote:Nonetheless, adjectives in English in LFG are usually analyzed this way:

Then how do you account for the difference between "the little old man" (which contains "old man" as a sub-unit) and "the little, old man" (which doesn't)?

I said usually. There I would personally put "old man" one level below "little", or even consider "old man" as a single word. But that's in using LFG.

Eventually in this thread I'm giong to present my own ideas, that they may be brutally criticized, and I'd handle that problem in a much different way.

Miekko · Post by **Miekko** » Sun Jan 15, 2006 3:08 pm

That difference is accountable for like this:
[DP [D the] [NP [N' [AP [A little]] [N' [AP [A old]] [N man]]
[DP [D the] [NP [N' [AP [A' A[ little] [A' [A old]]]] [N man]]
, ehm, roughly, at least.

Miekko · Post by **Miekko** » Sun Jan 15, 2006 3:10 pm

skurai wrote:Question:

I am a begginer when it comes to syntax, so I was wondering if this is right?

We haven't described very exactly how to deal with adjective attributes, yet, but if you replaced the NPs's branches with a big triangles, it'd be entirely correct.

finlay · Post by **finlay** » Mon Jan 16, 2006 8:20 am

Miekko wrote:
skurai wrote:Question:

I am a begginer when it comes to syntax, so I was wondering if this is right?

We haven't described very exactly how to deal with adjective attributes, yet, but if you replaced the NPs's branches with a big triangles, it'd be entirely correct.

The "Aux" shouldn't be in there. There is no auxiliary for this.

Also, the correct sentence is "The quick brown fox jumps over the lazy dog"; if you don't use "jumps" there's no "s" in the sentence.

Echobeats · Post by **Echobeats** » Mon Jan 16, 2006 11:08 am

finlay wrote:The "Aux" shouldn't be in there. There is no auxiliary for this.

There is if you're an X-bar-ist. It's just that it moves down to V and suffixes itself to the verb (affix lowering).

Ceresz · Post by **Ceresz** » Mon Jan 16, 2006 2:33 pm

finlay wrote:
Miekko wrote:
skurai wrote:Question:

I am a begginer when it comes to syntax, so I was wondering if this is right?

We haven't described very exactly how to deal with adjective attributes, yet, but if you replaced the NPs's branches with a big triangles, it'd be entirely correct.
The "Aux" shouldn't be in there. There is no auxiliary for this.

Also, the correct sentence is "The quick brown fox jumps over the lazy dog"; if you don't use "jumps" there's no "s" in the sentence.

Thanks, I wasn't sure about the sentence so I just made something up

Miekko · Post by **Miekko** » Mon Jan 16, 2006 3:15 pm

Echobeats wrote:
finlay wrote:The "Aux" shouldn't be in there. There is no auxiliary for this.
There is if you're an X-bar-ist. It's just that it moves down to V and suffixes itself to the verb (affix lowering).

At my current level of understanding, I'd be inclined to believe that the tense actually is outside the VP, and therefore is an AUX. However, orthodox X-bar would require the AUX to form it's own phrase around the VP, a phrase often called the TP (TAM-Phrase).

So, there should be a branch [S [NP ...] [TP [T' [T AUX]] [VP ...]

That is only under extended X-bar theory, though.

Echobeats · Post by **Echobeats** » Mon Jan 16, 2006 4:20 pm

Miekko wrote:
Echobeats wrote:
finlay wrote:The "Aux" shouldn't be in there. There is no auxiliary for this.
There is if you're an X-bar-ist. It's just that it moves down to V and suffixes itself to the verb (affix lowering).
At my current level of understanding, I'd be inclined to believe that the tense actually is outside the VP, and therefore is an AUX. However, orthodox X-bar would require the AUX to form it's own phrase around the VP, a phrase often called the TP (TAM-Phrase).

So, there should be a branch [S [NP ...] [TP [T' [T AUX]] [VP ...]

That is only under extended X-bar theory, though.

Actually, TP would be where S is. In X-bar theory, S was replaced by IP was replaced by TP.

Tim.

Miekko · Post by **Miekko** » Mon Jan 16, 2006 4:29 pm

I quite distinctly remember - and I have the notes to back it, at least - that I've been taught that TP does enclose the VP, but not the subject. OTOH, I can see a case be made for TP being further out, so ...

Salmoneus · Post by **Salmoneus** » Mon Jan 16, 2006 4:41 pm

If you can't cross lines, what do you do when the lines cross?

Echobeats · Post by **Echobeats** » Mon Jan 16, 2006 5:01 pm

Miekko wrote:I quite distinctly remember - and I have the notes to back it, at least - that I've been taught that TP does enclose the VP, but not the subject. OTOH, I can see a case be made for TP being further out, so ...

Subjects are generated in Spec-VP, and in many languages raises to Spec-TP. When it doesn't, you get VSO word order (if you have V-to-T raising).

Salmoneus wrote:If you can't cross lines, what do you do when the lines cross?

They don't cross. But you are allowed to move things around, if the movement is motivated.

Tim.

Junes · Post by **Junes** » Mon Jan 16, 2006 5:59 pm

Okay, I'm a bit confused now. In the particular Chomskyan system I'm somewhat familiar with (Government and Binding, Haegeman), you would have an IP (inflectional phrase) with the PAST in specifier-position. So is this TAM-phrase (Tense/Aspect/Mood) more or less equivalent?
And, if we step outside strict, binary X-bar for a while, then we could say that the PAST is actually at the AUX. Correct?

My problem with most of the traditional approaches to syntax I know of (granted: not that many) is that they ignore the fact that the information is coming in incrementally, rather than in whole sentences. That may well be due to the fact that I'm mostly interested in psycholinguistic stuff.
To be fair, some people have tried to apply the Chomskyan framework to actual processing. However, concepts like late closure and minimal attachment don't actually seem to be all that important in parsing. Also, it seems that semantic factors can play an early role in parsing, rather than it being based purely on some syntactic rules of thumb.

I will keep following this thread (as well as the very interesting cognitive linguistics thread); your work is appreciated

.

Echobeats · Post by **Echobeats** » Mon Jan 16, 2006 6:45 pm

Junes wrote:My problem with most of the traditional approaches to syntax I know of (granted: not that many) is that they ignore the fact that the information is coming in incrementally, rather than in whole sentences.

No they don't. Are you familiar with the term "to be garden-pathed"? If not, I'll explain.

Salmoneus · Post by **Salmoneus** » Mon Jan 16, 2006 6:51 pm

But what about when they DO cross? Surely you can't claim that they actually used a different sentence just to avoid having evidence that conflicts with the arbitrary 'straight lines are good' theory of linguistics?

Echobeats · Post by **Echobeats** » Mon Jan 16, 2006 7:27 pm

Salmoneus wrote:But what about when they DO cross? Surely you can't claim that they actually used a different sentence just to avoid having evidence that conflicts with the arbitrary 'straight lines are good' theory of linguistics?

Lines aren't present in people's speech. They're drawn by syntacticians. "You can't draw lines that cross" is one of the constraints of the theory.

Junes · Post by **Junes** » Tue Jan 17, 2006 2:41 am

Echobeats wrote:
Junes wrote:My problem with most of the traditional approaches to syntax I know of (granted: not that many) is that they ignore the fact that the information is coming in incrementally, rather than in whole sentences.
No they don't. Are you familiar with the term "to be garden-pathed"? If not, I'll explain.

Yeah, I am. Okay, maybe that was a bit overstated. Of course, research has been done on parsing. That's not the point. However, I think that at least in early Chomskyan linguistic analysis, human performance was not the basis for the theory. The theory was developed on the basis of the analysis of whole sentences, and then it was applied to parsing. For instance, in trying to find evidence for transformations. Not necessarily wrong, of course, given the absence of empirical data at the time.
I'm familiar with parsing models like the garden path model but I don't know to what amount they (or the empirical evidence) influence mainstream syntactic theory in linguistics (rather than psycholinguistics). Perhaps more than I thought?

finlay · Post by **finlay** » Wed Jan 18, 2006 3:23 pm

My syntax teacher finally taught us trees yesterday.

She does weird things with various phrases like instead of using just NPs she'll use "DPs", for "determiner phrases", which dominate NPs...

I challenged her about it yesterday and she said that it was the "correct" way and that some guy had written about it in 1987 or something, and collected evidence... apparently... such as how you can shorten "those pictures" to "those"...
I still think she's wrong; just wondering what your thoughts on this are...

Miekko · Post by **Miekko** » Wed Jan 18, 2006 4:03 pm

finlay wrote:My syntax teacher finally taught us trees yesterday.

She does weird things with various phrases like instead of using just NPs she'll use "DPs", for "determiner phrases", which dominate NPs... I challenged her about it yesterday and she said that it was the "correct" way and that some guy had written about it in 1987 or something, and collected evidence... apparently... such as how you can shorten "those pictures" to "those"...
I still think she's wrong; just wondering what your thoughts on this are...

That is the variety I've been taught, and I agree with it.

The same variety has CPs (complementizer phrases) to dominate the entire clause, and TPs to dominate the VP.

Echobeats · Post by **Echobeats** » Wed Jan 18, 2006 8:58 pm

Miekko wrote:
finlay wrote:My syntax teacher finally taught us trees yesterday.

She does weird things with various phrases like instead of using just NPs she'll use "DPs", for "determiner phrases", which dominate NPs... I challenged her about it yesterday and she said that it was the "correct" way and that some guy had written about it in 1987 or something, and collected evidence... apparently... such as how you can shorten "those pictures" to "those"...
I still think she's wrong; just wondering what your thoughts on this are...
That is the variety I've been taught, and I agree with it.

The same variety has CPs (complementizer phrases) to dominate the entire clause, and TPs to dominate the VP.

Even above main clauses?

I haven't yet got round to asking why Ds are no longer put in Spec-NP, but apparently there's some justification along the lines that Ds allow NPs to refer to things in the world as opposed to concepts. Hence the difference between "I shot the boar" and "I had boar for lunch". Though that still doesn't sound terribly convincing to me ? I may have got the details wrong. Or it may just be a weak theory, but I'm more inclined to think the former.

Please remind me to ask that ? I have a syntax supervision on Monday.

Yours, Tim.

finlay · Post by **finlay** » Wed Jan 18, 2006 10:51 pm

Miekko wrote:
finlay wrote:My syntax teacher finally taught us trees yesterday.

She does weird things with various phrases like instead of using just NPs she'll use "DPs", for "determiner phrases", which dominate NPs... I challenged her about it yesterday and she said that it was the "correct" way and that some guy had written about it in 1987 or something, and collected evidence... apparently... such as how you can shorten "those pictures" to "those"...
I still think she's wrong; just wondering what your thoughts on this are...
That is the variety I've been taught, and I agree with it.

The same variety has CPs (complementizer phrases) to dominate the entire clause, and TPs to dominate the VP.

Hmm.... yeah, I'm still not convinced.

But fair enough...
Part of it is of course that I've read several different descriptions of syntax and/or linguistics over the years, especially on the ZBB, before I came here, and not one of them has mentioned these DPs. In fact the photocopied notes from textbooks that we get given use NP for everything....
And then the other thing that got to me was that "John" is supposed to be a full DP instead of a straight NP, with ? for a D, which just makes no sense. (type in [DP [D ?] [NP John]] to the tree maker)

Tengado · Post by **Tengado** » Wed Jan 18, 2006 11:02 pm

Could someone post a link to a site that explains all the basic terminology - such as what a complementizer is - so that people with non-linguistics degree backgrounds can get the most out of it?

Miekko · Post by **Miekko** » Thu Jan 19, 2006 8:50 am

Tengado wrote:Could someone post a link to a site that explains all the basic terminology - such as what a complementizer is - so that people with non-linguistics degree backgrounds can get the most out of it?

I haven't found any such site; however, for now, knowing that a complementizer is a particle (or whatever) that introduces a clause or a subclause should be enough.

Some words we should explain:

mother node
A node in the tree that is above other nodes.

daughter node
a node that is below a mother node

sister node
a daughter node of the same mother node as another node.

these I should look up some more formal definitions of:
c-command

domination