老師出一個題目，大家踴躍討論。 Compare the alternative structure in (1) with those in (2). Use verb subcategorization and thematic structures to explain the ungrammaticality in (2b) and (2d)
(1) a. I planted roses in the garden. b. I planted the garden with roses. c. I loaded hay onto the cart. d. I loaded the cart with hay.
(2) a. I poured water into the pot. b. *I poured the pot with water. c. I fill the pot with water. d. *I fill water in the pot.
plant,verb[_NP PP] load,verb[_NP PP]or load,verb[_NP] pour,verb[_NP PP] fill,verb[_NP PP] Based on the above, all the sentences in (1)and(2)have the same pattern of argument choosing. Yet sentences(2b)and(2d) are ungrammatical in thematic structures: *(2b)I poured the pot with water. [+liquid] [-liquid] [+variable] [-variable] *(2d)I fill water in the pot. [-liquid][+liquid] [+space] [-space] [+concrete][-concrete]
The subcategorizations of the four verbs are as follows: plant, verb [__NP] load, verb [__NP] pour, verb [__NP PP] fill, verb [__NP PP]
I think what differs most between the two groups is that the verbs in (i)take only two arguments --the subject and the direct object -- and the prepositional phrases are only adjuncts. The verbs in (2), on the contrary, take three arguments, a subject, a direct object, and an indirect object in the form of a PP.
Therefore, in (1), although the verbs are permitted to take either the object or the place as its direct objects and the thematic role of the direct object may vary, it won't violate the principle of the thematic structure: S --> agent, O --> theme(roses), or location (the garden), or goal (the cart), PP --> not taken into consideration because it's not an argument
As for (2), I agree with hsin-yu that because "pour" and "fill" both s-select their direct objects with the semantic property [+liquid] and [+container], respectively, the head noun in the PP cannot serve to be the direct object. If so, it will violate the thematic hierarchy.
I poured water into the pot. theme location *I poured the pot with water. location theme
The verbs in the group(1)only need two complement NPs as their arguments by subcategorizing as follows: plant, verb, [___ NP] load, verb, [___ NP] Note that the PP is described as a predicate despite varying its internal NP with thetaless grammatically.
Conversely, the verbs in the group(2) need three ones as their arguments with subcategorization as follows: pour, verb, [___ NP PP] fill, verb, [___ NP PP] Based on semantic hierarchy, the verbs, pour and fill ,constuct in 2a & 2c as below: Agent {pour} theme {goal} {fill} {instrument} 2b. *I poured the pot with water. {goal} {theme} 2d. *I fill water in the pot. {instrument} {theme}
Otherwise,2b and 2d also violate the semantic restriction as below: * 2b.I poured the pot with water. [+liquid] [-liquid] * 2d.I fill water in the pot. [-liquid][+liquid] [+space] [-space]
Subcategorization plant, V ﹝____ NP﹞ load, V ﹝____ NP﹞ pour, V ﹝____ NP PP﹞ fill, V ﹝____ NP PP﹞
In the first structure, the verb plant and load take only one complement. The prepositional phrases serve only as modifier. The verb, pour s-select an object whose semantic feature must be ﹝+liquid﹞. Therefore, since the object pot has the semantic feature of [-liquid], it cannot co-occur with the verb pour. On the other hand, the verb fill s-select an object whose semantic feature must be [+container]. Therefore, the ungrammaticality of 2d results from the unacceptable combination of the verb fill, with an object whose semantic feature is [-container].
Four verbs as the followed: plant, verb [__NP] load, verb [__NP] pour, verb [__NP PP] fill, verb [__NP PP]
First structure, there have two arguments and phrases are Adjunct
According to the hsin-yu points of view, the "pour" and "fill" both in s-select represent “pour” [+liquid] and “fill” [+container], the other way, but in the 2b[-liquid], 2d[-container]
I poured water into the pot. Agent theme location *I poured the pot with water. Agent location theme
Both 'plant' and 'load' in (1) take two arguments, which can be schematically represented as plant, verb [___ NP] load, verb [___ NP]
Following this, the PP, "in the garden" in 1a and the PP "with hay" in 1d are "adjuncts" on the ground that omission of these phrases will not cause ungrammaticality. 1a. I planted roses. 1d. I loaded the cart. Problem arises when we consider 1b. and 1c. The omission of the sentence final PP leads to ungrammaticality. 1b. *I planted the garden. 1c. *I loaded hay.
Nonetheless, we do not need to posit another argument structures for 1b and 1c. The reason is: Argument selection is based on the core meaning of the lexical item and necessarily manifests itself in the deep structure. Hence, 1b can be seen as derived from 1a and 1c from 1d via a transformation which moves an peripheral adjunct closer to the verb.
In (2), the subcategorization for "pour" and "fill" is as follows: pour, verb, [___ NP PP] fill, verb, [___ NP PP] The ungrammaticality of 2b. and 2d. may be accounted by the fact taht the trasnsformation, which is used to derive 1b from 1a and 1c from 1d, does not apply to this two verbs. The function of the transformation is to move PP adjuncts closer to the verb while the PPs "into the pot" and "with water" are complements, not adjuncts.
-- ps1. i think.. the ungrammaticality should be accounted for on the basis of the comparison; instead of treating the "ungrammaticality" and "comparison" as unrelated. 'cause if it's the case, why not just separate the q into two qs? (or i am mistaken?) anyway, i suppose differences between (1) and (2) is of paramount importance to a satisfactory explanation of the ungrammaticality of 2b and 2d. -- ps2. one of the biggest problems i ran into while attempting to tackle this exercise is that--> i don't know how many arguments the four verbs take. I did checked a dictionary but no definite answers were found. apparently, lexicographers' opinions do not converge. ex. plant NP plant NP PP pour NP PP pour NP fill NP PP pour NP usw.. ----
I have some questions to ask Brian :) 1. [the verbs are permitted to take either the object or the place as its direct objects ] in other words, you'll rate I planted the garden. I loaded the hay. as grammatical?
2. [I poured water into the pot. Agent theme location *I poured the pot with water. Agent location theme] then how do we account for I planted the garden with roses. Agent-Location-Theme I loaded the cart with hay. Agent-location-theme
I think the distinction between DS and SS, where transformation also plays a role, should be drawn to shed light on this problem? (or am i mistaken?) If not, then we can only recognise a. as well-formed but dismiss b. as unacceptable.
a.I give a dog to Mary. Agent-Theme-Goal b. I give Mary a dog. Agent-Goal-Theme
3. I planted flowers. -->"Plant" s-selects sth. [+live] among other features.
followiing this vein, then the ungrammaticality of c should be predicted, but it turns out acceptable.how do we account for this? c.I planted the garden with roses. the last qestion is, if s-selection is responsible for the ungrammaticality of 2b and 2d, then why do we bother to compare 1 and 2. or the "comparison" and "explanation" Qs are just unrelated? See PS1. :) if you got time, do say sth. btw, thank you so much for telling me that this board was opened. So nice of you~ best,
To Jarvis: 1) Sorry that i think i didn’t put my point that clear. i basically agree with your argument. Transformation is the factor causing the problem, but why i talk about thematic roles is ‘cause it’s one way i want to confirm which is DS and which is SS. As you say, “plant” s-selects sth [+flora], and “load” [+container] to be their objects, and my point is that we can accordingly get to know whose thematic role should be “theme,” or in other words, which is the appropriate direct object in DS, because it’s actually the one what the verb denotes directly has sth to do with or has effect on. After this, even though the adjuncts of DS may move to the position preceding the direct objects due to some transformational process, their thematic role will remain the same and won’t violate the hierarchy since we don’t care about adjuncts, the PPs. As for the grammaticality of “I planted the garden.” and “I loaded the hay.”, it is really not my inference. ( the former sounds OK to me, but the latter does sound odd) Any way, i think it won’t influence our analysis. :)
2) As i said previously, we don’t have to bother discussing the thematic hierarchy since the verbs in the two sentences are dyadic. As for a. I give a dog to Mary. Agent-Theme-Goal b. I give Mary a dog. Agent-Goal-Theme
i guess the hierarchy applies only to DS, since it’s one method to specify DS.
3) Truly, transformation is one way to explain sentences in (1), but the question is why it does not trigger in (2). (Transformation does take place in sentences whose verbs are triadic, say “give.”) And my opinion is that these verbs, “pour” and “fill,” strictly s-select their direct objects.
About your ps~
i think you’re right, and that’s why i said the crucial difference of (1) and (2) lies on the valence of the verbs of each group. Comparison is very important for the issue of detecting reasons for “ungrammaticalities.”
3) Truly, transformation is one way to explain sentences in (1), but the question is why it does not trigger in (2). (Transformation does take place in sentences whose verbs are triadic, say “give.”) And my opinion is that these verbs, “pour” and “fill,” strictly s-select their direct objects.
--> you know what~ this exercise, if not all the syntactic exercise, is really difficult for me.. sometimes i just don't know how to start and even having started, how to proceed. although the idea of transformation popped into my mind at the first glimpse, i questioned the validity of it in no time. We can proove nothing and get nothing done, if we bring ourselves in line with some previously-established conventions, in this case, a syntactic framework, the theoretical framework imposed by the question. if there's no belief then nothing could ever be done. So, the first thing i had to do was persuade myself that there was an answer to the question. The ensuing work was to try to come up with an answer, which would be at least deemed as acceptable under the theoretical framework in question, syntax in this case.
This is why i appeal to transformations, which in fact i have little faith in, to account for the ungrammaticality of 2. I know nothing about how to rate "pour" and "fill" as verbs that "strictly" select their arguments with [+X] let alone justifying that "plant" and "load" are more lenient argument selection. Consider this:
--all the sentences in 1 are grammatical but some of those in 2 are not due to the fact that the inherent meaning of the verbs in question are lenient and those in 2 are strict. --
this argument appeals to me and in fact,(you know i am more interested in semantics and philosophy than syntax) but a. I haven't learned enough about semantics and philosophy to approach this exercise from a semantics-based or philosophical perspective.
b. I'd rather be as 'syntactic' as possible when facing a syntactic exercise.
Thus, my argument is that the transformation does not apply to 2 'cause they take three arguments. The number of arguments required by each of the verbs were justified by grammaticality tests given in my first response to the exercise. The transformation here need not explain movements of other kinds(ex. ditransitive verbs) just like we'll never expect active-passive transformations to account for Wh- questions.
I do think the meaning of the verbs in this exercise may be sufficient for the ungrammaticality. The problem is, we need more reasoning or theories in support for that. Maybe you'll come up with some and share with us soon. :)
In my opinion, if observing from the categorization part, it only shows that different verbs will choose vary arguments which are specific to the verb. However; only observing from this part is not strong enough to provide why (2)b and (2)d sentences are ungrammatical because the arguments are all NPs. The subcategorization will show in the picture & below. picture & (1)a. plant, verb,【___NP】 b. plant, verb,【___NP PP】 c. load, verb, 【___NP PP】 d. load, verb, 【___NP】 (2)a. pour, verb,【___NP PP】 c. fill, verb,【___NP PP】 Researching from thematic structures, giving as the below list $, it clearly shows that particular verbs will take different semantic roles. Besides, theta roles will make a connection with the syntactic structure. list$ (1)a. agent, theme, location b. agent, theme, instrument (source) c.agent, theme, goal d.agent, theme, instrument
In short, comparing (2)b to (2)a,(2)c to (2)d ,and (1)a to d, it is an evidence that a particular verb will choose specific theta roles that is combined with syntactic structure to be its components. Further more, the arguments a verb needs can't make a move easily because of changing the original meaning when in the grammatical sentence.
hi Frank~ your explanation is both cool and original(maybe more like 'bold' in a sense). I like it very much:)
//////// PS. you may wanna skip the passages enclosed by the double lines below. They may not be as germane to the exercise as expected. //////// ============ You treated all the sentences in the exercise as "different" and it follows, most naturally, that the use of transformation, whose purpose is to account for the relationship between sentences with "identical" meaning, would be absurd(or at least inappropriate) here.
There are advantages to recognizing the subtle differences in meaning, and I just thought of two. 1. it reflects how we use lg. from a pragmatic point of view, we'll realise the distribution of sentences(eg. 1a & qb) may be significantly distinct. It's all about information packaging, which reflects what how the speaker would like to guide his or her interlocuters' attention to. 2. cognitive economy in terms of memory. If the sentences(eg. 1a & 1b) are thoroughly "identical," then in the process of evolution(in a linguistic sense), one of them should've been selected as representive of the others and the others discarded. Otherwise, the redudant realizations of the same meaning may do nothing but tax one's memory capacity.
Transformations are postulated with a hope to approximate "obervational adequacy(in Chomsky's words)." The need of such processes stems from the suspicion that sentences with "identical" meanning must be related and to store them separetely in the brain is uneconomical. However, problems arise when we take a closer look at the word "identical." Should we really employ a dichotic view? Are we always able to ignore the possibility that the relationship between sentences may be evaluated along a continuum with "identical" and "unrelated" being the two ends? In my understanding, convention overrides many other considerations as we talk about transformation, or syntax(at least the situation holds true for me. I'm slow in doing binary judges. No offences~) for example, why are we to treat active-passive sentences as related but others not so(eg. 1 & 2 below)? eg. 1. give me a hand. 2. i need help.
Sorry for all the digression above. ========== Let's get back to the exercise stop wandering around. I think the act to equate the preposition "with" with an "instrument" marker may lead to trouble. consider these: 1. I went to the movie with Jack. 2. I went to Taipei by train. Is "Jack" an intrument? Can't "train" be seen as "instrument," with the help which we achieve sth.? An one-to-one correspondence of "with" to "instrument" is hence called into question.
To the Jarvis yesterday, You beat me = = In fact, when i saw the question for the first time, the first things that came to my mind are transformation (you know, it's so popular these days, with the affix -er) and s-selection. However, even though transformation seems to work in sentences in (i), i just couldn't come up with a sensible reason why it isn't applied in (ii), and that's why i turn to the semantics-based ways, s-selection and theta roles.= =
"Thus, my argument is that the transformation does not apply to 2 'cause they take three arguments."
1) sorry i'm still obstinate that i don't understand why the tansformation, in your opinion, does not apply to those which take three arguments as in (ii), but it does in the verb "give." Do you only refer to "this sort" of transformation in question?
2) If this is the case, then is there a convincing reanson for this difference and phenomenon?
After these days and after reading everyone's arguments, i begin to think my opinion is rather weak and lacks a crucial aspect of standpoints. Maybe this question involves a multiple perspectives, and maybe our viewpoints are included there as two surfaces of one identical geometrical object.
By the way, don't use asterisks. It made me think they are unacceptable. Orz
That's always good to know the limitation of reasoning. I wanna present the thread of my argument once again and end further defending my argument on this exercise('cause at the present moment, i don't see how we can go any further~ maybe it's time to turn to our teacher). :) 1. the questions in the exercise have answers, and they are not supposed to come out of nowhere. ex. "because the verbs are different, so they behave differently."
2. to approach them in a syntactic framework --> thus i resort to transformations. --------
-yes, i do think this transformation is not generalizable to just any situations.
-nontheless, bear in mind that the soundness of a rule is not to be able to take care of everything, sentences in this case, but to deal with things with similar nature. I remember reading Chomsky argued against other non-Chomskyans like what i just wrote.(i just did my best to paraphrase it. or it's my interpretation after reading his argument) Thus, I think it's alright to recognise the limitation of the transformation, which I used at my responses. This very transformation apply not only to the sentences in the exercise, but also to the followings: -plant -laod 1. sprinkle 2. spread and so on, i can't come up with more off the top of my head. the sentences are taken from BNC. 1a. A89 321 Ladle on the bechamel, sprinkle the remaining cheese on top, 1b. ABB 412 Sprinkle the avocado slices with lemon juice, then arrange them with Mozzarella and tomato slices. 2a. ASE 2191 She spread mulberry jam in reckless bounty upon a slice of tea cake. 2b. ABB 1755 Spread the cake with cream and then top with the drained cherries. ---
An alternative way to go, is to focus on the inherent meaning of the two sets of verbs. As you may already notice, the verbs "plant" "load" "spread" "sprinkle" do have sth. in common. I don't know how to phrase it in a technical way, but i'll say, it'll be sth. like.. "to cause a particular container to have more X" "the X is not liquid" "..." the list goes on and on, and it can , in the end, distinguishes itself from the set of verbs in 2 (and possibly the rest of other verbs). I took semantics, but it was an introductory course. I think it's possible to solve the whole question with a semantic approach. But.. I, and maybe we, haven't learned that much yet..
thus, I resorted to transformation...
---
"Maybe this question involves a multiple perspectives"
ya, every question and problem is not as simple as we would like it to be. every phenomenon is a function of loads of variables and we'll never pin down all of them. Then what should we do? My solution is.. to be aware of what we are doing before proceeding. How to do sth. has much to do with the end of the "sth." eg. the end of doing exercise is to sharpen our writing skills and to help us do well in exams. i'm not to treat the questions in the way as i would my research topic. So... I took a simple and rule-like approach to it, which i think would be good for exams.
------ ==By the way, don't use asterisks. It made me think they are unacceptable. Orz==
=I carried over my msn habit over onto the wrong place. asterisks, are interpreted the other way round when it comes to a linguistics discussion board. Thanks for telling me that. :)
Sentence(2b) and (2d) are ungrammatical because “pour” and “fill” are verbs with the feature[+liquid], but ” the pot” is the NP with the feature[-liquid]. They are contrast semantically.
22 則留言:
老師出一個題目，大家踴躍討論。
Compare the alternative structure in (1) with those in (2). Use verb subcategorization and thematic structures to explain the ungrammaticality in (2b) and (2d)
(1)
a. I planted roses in the garden.
b. I planted the garden with roses.
c. I loaded hay onto the cart.
d. I loaded the cart with hay.
(2)
a. I poured water into the pot.
b. *I poured the pot with water.
c. I fill the pot with water.
d. *I fill water in the pot.
plant, verb, 【___NP, PP】
load, verb, 【___NP, PP】
pour, verb, 【___ NP, PP】
fill, verb, 【___NP, PP】
Four verbs have identical subcategorization, but the sentence (2b)and (2d) violate the selection restriction.
* I poured the pot with water.
﹝+liquid﹞ ﹝-liquid﹞
plant,verb[_NP PP]
load,verb[_NP PP]or load,verb[_NP]
pour,verb[_NP PP]
fill,verb[_NP PP]
Based on the above, all the sentences in (1)and(2)have the same pattern of argument choosing.
Yet sentences(2b)and(2d) are ungrammatical in thematic structures:
*(2b)I poured the pot with water.
[+liquid] [-liquid]
[+variable] [-variable]
*(2d)I fill water in the pot.
[-liquid][+liquid]
[+space] [-space]
[+concrete][-concrete]
The subcategorizations of the four verbs are as follows:
plant, verb [__NP]
load, verb [__NP]
pour, verb [__NP PP]
fill, verb [__NP PP]
I think what differs most between the two groups is that the verbs in (i)take only two arguments --the subject and the direct object -- and the prepositional phrases are only adjuncts. The verbs in (2), on the contrary, take three arguments, a subject, a direct object, and an indirect object in the form of a PP.
Therefore, in (1), although the verbs are permitted to take either the object or the place as its direct objects and the thematic role of the direct object may vary, it won't violate the principle of the thematic structure:
S --> agent, O --> theme(roses), or location (the garden), or goal (the cart), PP --> not taken into consideration because it's not an argument
As for (2), I agree with hsin-yu that because "pour" and "fill" both s-select their direct objects with the semantic property [+liquid] and [+container], respectively, the head noun in the PP cannot serve to be the direct object. If so, it will violate the thematic hierarchy.
I poured water into the pot.
theme location
*I poured the pot with water.
location theme
sorry, the last two sentences should be:
I poured water into the pot.
Agent theme location
*I poured the pot with water.
Agent location theme
The verbs in the group(1)only need two complement NPs as their arguments by subcategorizing as follows:
plant, verb, [___ NP]
load, verb, [___ NP]
Note that the PP is described as a predicate despite varying its internal NP with thetaless grammatically.
Conversely, the verbs in the group(2) need three ones as their arguments with subcategorization as follows:
pour, verb, [___ NP PP]
fill, verb, [___ NP PP]
Based on semantic hierarchy, the verbs, pour and fill ,constuct in 2a & 2c as below:
Agent {pour} theme {goal}
{fill} {instrument}
2b. *I poured the pot with water.
{goal} {theme}
2d. *I fill water in the pot.
{instrument} {theme}
Otherwise,2b and 2d also violate the semantic restriction as below:
* 2b.I poured the pot with water.
[+liquid] [-liquid]
* 2d.I fill water in the pot.
[-liquid][+liquid]
[+space] [-space]
Subcategorization
plant, V ﹝____ NP﹞
load, V ﹝____ NP﹞
pour, V ﹝____ NP PP﹞
fill, V ﹝____ NP PP﹞
In the first structure, the verb plant and load take only one complement. The prepositional phrases serve only as modifier. The verb, pour s-select an object whose semantic feature must be ﹝+liquid﹞. Therefore, since the object pot has the semantic feature of [-liquid], it cannot co-occur with the verb pour. On the other hand, the verb fill s-select an object whose semantic feature must be [+container]. Therefore, the ungrammaticality of 2d results from the unacceptable combination of the verb fill, with an object whose semantic feature is [-container].
Four verbs as the followed:
plant, verb [__NP]
load, verb [__NP]
pour, verb [__NP PP]
fill, verb [__NP PP]
First structure, there have two arguments
and phrases are Adjunct
According to the hsin-yu points of view, the "pour" and "fill" both in s-select represent “pour” [+liquid] and “fill” [+container], the other way,
but in the 2b[-liquid],
2d[-container]
I poured water into the pot.
Agent theme location
*I poured the pot with water.
Agent location theme
Both 'plant' and 'load' in (1) take two arguments, which can be schematically represented as
plant, verb [___ NP]
load, verb [___ NP]
Following this, the PP, "in the garden" in 1a and the PP "with hay" in 1d are "adjuncts" on the ground that omission of these phrases will not cause ungrammaticality.
1a. I planted roses.
1d. I loaded the cart.
Problem arises when we consider 1b. and 1c. The omission of the sentence final PP leads to ungrammaticality.
1b. *I planted the garden.
1c. *I loaded hay.
Nonetheless, we do not need to posit another argument structures for 1b and 1c. The reason is:
Argument selection is based on the core meaning of the lexical item and necessarily manifests itself in the deep structure.
Hence, 1b can be seen as derived from 1a and 1c from 1d via a transformation which moves an peripheral adjunct closer to the verb.
In (2), the subcategorization for "pour" and "fill" is as follows:
pour, verb, [___ NP PP]
fill, verb, [___ NP PP]
The ungrammaticality of 2b. and 2d. may be accounted by the fact taht the trasnsformation, which is used to derive 1b from 1a and 1c from 1d, does not apply to this two verbs. The function of the transformation is to move PP adjuncts closer to the verb while the PPs "into the pot" and "with water" are complements, not adjuncts.
--
ps1.
i think..
the ungrammaticality should be accounted for on the basis of the comparison; instead of treating the "ungrammaticality" and "comparison" as unrelated. 'cause if it's the case, why not just separate the q into two qs? (or i am mistaken?)
anyway, i suppose differences between (1) and (2) is of paramount importance to a satisfactory explanation of the ungrammaticality of 2b and 2d.
--
ps2.
one of the biggest problems i ran into while attempting to tackle this exercise is that--> i don't know how many arguments the four verbs take. I did checked a dictionary but no definite answers were found. apparently, lexicographers' opinions do not converge. ex. plant NP
plant NP PP
pour NP PP
pour NP
fill NP PP
pour NP usw..
----
I have some questions to ask Brian :)
1.
[the verbs are permitted to take either the object or the place as its direct objects ]
in other words, you'll rate
I planted the garden.
I loaded the hay.
as grammatical?
2.
[I poured water into the pot.
Agent theme location
*I poured the pot with water.
Agent location theme]
then how do we account for
I planted the garden with roses.
Agent-Location-Theme
I loaded the cart with hay.
Agent-location-theme
I think the distinction between DS and SS, where transformation also plays a role, should be drawn to shed light on this problem? (or am i mistaken?)
If not, then we can only recognise
a. as well-formed but dismiss b. as unacceptable.
a.I give a dog to Mary.
Agent-Theme-Goal
b. I give Mary a dog.
Agent-Goal-Theme
3. I planted flowers.
-->"Plant" s-selects sth. [+live]
among other features.
followiing this vein, then the
ungrammaticality of c should be
predicted, but it turns out
acceptable.how do we account for
this?
c.I planted the garden with
roses.
the last qestion is, if
s-selection is responsible for
the ungrammaticality of 2b and
2d, then why do we bother to
compare 1 and 2. or the
"comparison" and "explanation"
Qs are just unrelated? See PS1.
:)
if you got time, do say sth.
btw, thank you so much for
telling me that this board was
opened. So nice of you~
best,
To Jarvis:
1) Sorry that i think i didn’t put my point that clear. i basically agree with your argument. Transformation is the factor causing the problem, but why i talk about thematic roles is ‘cause it’s one way i want to confirm which is DS and which is SS. As you say, “plant” s-selects sth [+flora], and “load” [+container] to be their objects, and my point is that we can accordingly get to know whose thematic role should be “theme,” or in other words, which is the appropriate direct object in DS, because it’s actually the one what the verb denotes directly has sth to do with or has effect on. After this, even though the adjuncts of DS may move to the position preceding the direct objects due to some transformational process, their thematic role will remain the same and won’t violate the hierarchy since we don’t care about adjuncts, the PPs.
As for the grammaticality of “I planted the garden.” and “I loaded the hay.”, it is really not my inference. ( the former sounds OK to me, but the latter does sound odd) Any way, i think it won’t influence our analysis. :)
2) As i said previously, we don’t have to bother discussing the thematic hierarchy since the verbs in the two sentences are dyadic. As for
a. I give a dog to Mary.
Agent-Theme-Goal
b. I give Mary a dog.
Agent-Goal-Theme
i guess the hierarchy applies only to DS, since it’s one method to specify DS.
3) Truly, transformation is one way to explain sentences in (1), but the question is why it does not trigger in (2). (Transformation does take place in sentences whose verbs are triadic, say “give.”) And my opinion is that these verbs, “pour” and “fill,” strictly s-select their direct objects.
About your ps~
i think you’re right, and that’s why i said the crucial difference of (1) and (2) lies on the valence of the verbs of each group. Comparison is very important for the issue of detecting reasons for “ungrammaticalities.”
3) Truly, transformation is one way to explain sentences in (1), but the question is why it does not trigger in (2). (Transformation does take place in sentences whose verbs are triadic, say “give.”) And my opinion is that these verbs, “pour” and “fill,” strictly s-select their direct objects.
-->
you know what~ this exercise, if not all the syntactic exercise, is
really difficult for me..
sometimes i just don't know how to
start and even having started, how
to proceed. although the idea of
transformation popped into my mind
at the first glimpse, i questioned
the validity of it in no time. We
can proove nothing and get nothing
done, if we bring ourselves in line
with some previously-established
conventions, in this case, a syntactic framework, the
theoretical framework imposed by the question. if there's no belief
then nothing could ever be done. So, the first thing i had to do was persuade myself that there was an answer to the question. The ensuing work was to try to come up with an answer, which would be at least deemed as acceptable under the theoretical framework in question, syntax in this case.
This is why i appeal to transformations, which in fact i
have little faith in, to account for the ungrammaticality of 2. I know nothing about how to rate "pour" and "fill" as verbs that "strictly" select their arguments with [+X] let alone justifying that "plant" and "load" are more lenient argument selection. Consider this:
--all the sentences in 1 are grammatical but some of those in 2 are not due to the fact that the inherent meaning of the verbs in question are lenient and those in 2 are strict.
--
this argument appeals to me and in fact,(you know i am more interested in semantics and philosophy than syntax) but
a. I haven't learned enough about
semantics and philosophy to
approach this exercise from a
semantics-based or philosophical
perspective.
b. I'd rather be as 'syntactic' as
possible when facing a syntactic
exercise.
Thus, my argument is that the transformation does not apply to 2 'cause they take three arguments.
The number of arguments required by each of the verbs were justified by grammaticality tests given in my first response to the exercise. The transformation here need not explain movements of other kinds(ex. ditransitive verbs) just like we'll never expect active-passive transformations to account for Wh- questions.
I do think the meaning of the verbs in this exercise may be sufficient for the ungrammaticality. The problem is, we need more reasoning or theories in support for that. Maybe you'll come up with some and share with us soon. :)
good night~
to increase readability to my last
response..
*we "don't" bring us in line...
*more lenient "in" argument selection...
*are not, due to the fact that...
*to me (and in fact..
In my opinion, if observing from the categorization part, it only shows that different verbs will choose vary arguments which are specific to the verb. However; only observing from this part is not strong enough to provide why (2)b and (2)d sentences are ungrammatical because the arguments are all NPs. The subcategorization will show in the picture & below.
picture &
(1)a. plant, verb,【___NP】
b. plant, verb,【___NP PP】
c. load, verb, 【___NP PP】
d. load, verb, 【___NP】
(2)a. pour, verb,【___NP PP】
c. fill, verb,【___NP PP】
Researching from thematic structures, giving as the below list $, it clearly shows that particular verbs will take different semantic roles. Besides, theta roles will make a connection with the syntactic structure.
list$
(1)a. agent, theme, location
b. agent, theme, instrument
(source)
c.agent, theme, goal
d.agent, theme, instrument
(2)a.agent, theme, goal
*b.agent, location, instrument
c.agent, location, instrument
*d.agent, theme, location
In short, comparing (2)b to (2)a,(2)c to (2)d ,and (1)a to d, it is an evidence that a particular verb will choose specific theta roles that is combined with syntactic structure to be its components. Further more, the arguments a verb needs can't make a move easily because of changing the original meaning when in the grammatical sentence.
hi Frank~
your explanation is both cool and original(maybe more like 'bold' in a sense). I like it very much:)
////////
PS. you may wanna skip the passages
enclosed by the double lines below.
They may not be as germane to the
exercise as expected.
////////
============
You treated all the sentences in the exercise as "different" and it follows, most naturally, that the use of transformation, whose purpose is to account for the relationship between sentences with "identical" meaning, would be absurd(or at least inappropriate) here.
There are advantages to recognizing the subtle differences in meaning, and I just thought of two.
1. it reflects how we use lg. from
a pragmatic point of view, we'll
realise the distribution of
sentences(eg. 1a & qb) may be
significantly distinct. It's
all about information packaging,
which reflects what how the
speaker would like to guide his
or her interlocuters' attention
to.
2. cognitive economy in terms of
memory. If the sentences(eg.
1a & 1b) are thoroughly
"identical," then in the process
of evolution(in a linguistic
sense), one of them should've
been selected as representive
of the others and the others
discarded. Otherwise, the
redudant realizations of the
same meaning may do nothing but
tax one's memory capacity.
Transformations are postulated with a hope to approximate "obervational adequacy(in Chomsky's words)." The need of such processes stems from the suspicion that sentences with "identical" meanning must be
related and to store them separetely in the brain is uneconomical.
However, problems arise when we
take a closer look at the word "identical." Should we really
employ a dichotic view? Are we always able to ignore the possibility that the relationship between sentences may be evaluated along a continuum with "identical" and "unrelated" being the two ends?
In my understanding, convention overrides many other considerations as we talk about transformation, or syntax(at least the situation holds true for me. I'm slow in doing binary judges. No offences~) for example, why are
we to treat active-passive sentences as related but others not so(eg. 1 & 2 below)?
eg. 1. give me a hand.
2. i need help.
Sorry for all the digression above.
==========
Let's get back to the exercise stop wandering around.
I think the act to equate the preposition "with" with an "instrument" marker may lead to trouble. consider these:
1. I went to the movie with Jack.
2. I went to Taipei by train.
Is "Jack" an intrument?
Can't "train" be seen as "instrument," with the help which we achieve sth.?
An one-to-one correspondence of "with" to "instrument" is hence called into question.
*the exercise and stop wandering..
*with the help of which
To the Jarvis yesterday,
You beat me = =
In fact, when i saw the question for the first time, the first things that came to my mind are transformation (you know, it's so popular these days, with the affix -er) and s-selection. However, even though transformation seems to work in sentences in (i), i just couldn't come up with a sensible reason why it isn't applied in (ii), and that's why i turn to the semantics-based ways, s-selection and theta roles.= =
"Thus, my argument is that the transformation does not apply to 2 'cause they take three arguments."
1) sorry i'm still obstinate that i don't understand why the tansformation, in your opinion, does not apply to those which take three arguments as in (ii), but it
does in the verb "give." Do you only refer to "this sort" of transformation in question?
2) If this is the case, then is there a convincing reanson for this difference and phenomenon?
After these days and after reading everyone's arguments, i begin to think my opinion is rather weak and lacks a crucial aspect of standpoints. Maybe this question involves a multiple perspectives, and maybe our viewpoints are included there as two surfaces of one identical geometrical object.
By the way, don't use asterisks. It made me think they are unacceptable. Orz
That's always good to know the limitation of reasoning. I wanna present the thread of my argument once again and end further defending my argument on this exercise('cause at the present moment, i don't see how we can go any further~ maybe it's time to turn to our teacher). :)
1. the questions in the exercise
have answers, and they are not
supposed to come out of nowhere.
ex. "because the verbs are
different, so they behave differently."
2. to approach them in a syntactic
framework
--> thus i resort to transformations.
--------
-yes, i do think this transformation is not generalizable to just any situations.
-nontheless, bear in mind that the
soundness of a rule is not to be
able to take care of everything,
sentences in this case, but to
deal with things with similar
nature. I remember reading
Chomsky argued against other
non-Chomskyans like what i just
wrote.(i just did my best to paraphrase it. or it's my interpretation after reading his argument)
Thus, I think it's alright to
recognise the limitation of the
transformation, which I used at
my responses. This very transformation apply not only to the sentences in the exercise, but
also to the followings:
-plant
-laod
1. sprinkle
2. spread
and so on, i can't come up with more off the top of my head.
the sentences are taken from
BNC.
1a.
A89 321 Ladle on the bechamel, sprinkle the remaining cheese on top,
1b.
ABB 412 Sprinkle the avocado slices with lemon juice, then arrange them with Mozzarella and tomato slices.
2a.
ASE 2191 She spread mulberry jam in reckless bounty upon a slice of tea cake.
2b.
ABB 1755 Spread the cake with cream and then top with the drained cherries.
---
An alternative way to go, is to focus on the inherent meaning of the two sets of verbs. As you may already notice, the verbs "plant" "load" "spread" "sprinkle" do have sth. in common. I don't know how to phrase it in a technical way, but i'll say, it'll be sth. like..
"to cause a particular container to have more X"
"the X is not liquid"
"..."
the list goes on and on, and it can
, in the end, distinguishes itself
from the set of verbs in 2 (and possibly the rest of other verbs). I took semantics, but it was an introductory course. I think it's possible to solve the whole question with a semantic approach. But.. I, and maybe we, haven't learned that much yet..
thus, I resorted to transformation...
---
"Maybe this question involves a multiple perspectives"
ya, every question and problem is not as simple as we would like it to be. every phenomenon is a function of loads of variables and we'll never pin down all of them. Then what should we do? My solution is.. to be aware of what we are doing before proceeding. How to do sth. has much to do with the end of the "sth."
eg. the end of doing exercise is
to sharpen our writing skills and
to help us do well in exams. i'm
not to treat the questions in the way as i would my research topic.
So... I took a simple and rule-like
approach to it, which i think would be good for exams.
------
==By the way, don't use asterisks. It made me think they are unacceptable. Orz==
=I carried over my msn habit over
onto the wrong place. asterisks,
are interpreted the other way round when it comes to a linguistics discussion board. Thanks for telling me that. :)
good day,
Jarvis
Sentence(2b) and (2d) are ungrammatical because “pour” and “fill” are verbs with the feature[+liquid], but ” the pot” is the NP with the feature[-liquid].
They are contrast semantically.
請問用X-bar theory畫數有啥規則嗎?
像之前PS-rule就有他的規則,Vp→V(NP)...
但是X-bar好像遇到什麼字,就用什麼詞組帶進去,有規則嗎?像是在PS rule 裡面,Np裡面就不會出現VP...但在X bar裡面好像就有.
我感覺X-bar好抽象,為什麼X-bar就比PS rule 好?可以給我X bar簡單的概念,和如何畫數的技巧嗎?
謝謝~~~
我感覺X-bar好抽象,為什麼X-bar就比PS rule 好?
---
上課中記得老師有講過!
最一開始的方式 是線性的分析(此種分析假設同一條線 僅能對應一種語義!!) 但發現這樣的
方法 無法解釋句子的歧義
ex. I saw the girl with a telescope.
棄線性而改用ps rule，因為它是二維的結構。同一個線性結構可能有不同的二維結構，因此，上述的句子即獲得解釋。
然而，ps rule把論元與非論元，皆一律置於XP之下，因為無法分辦兩者*論元與非論元)之間的差別。此刻，X-Bar就出現了。它與PS-rule的差別，就再於區辦論元和非論元的能力。
----
至於x-bar的畫法，我只知道最基礎的，因為有點缺課的情形.. 首先，你可以先弄清楚X-bar的template(樣版).
1 XP--> X' Specifier
2 X'--> (Adjunct) X' (Adjunct) 擇一
3 X'--> X Complement
依情況對2.做數量上的增減..
x-bar是理論 理論的存在是想對現象做解釋。所以，畫出來的圖，只要不違背理論內部的規範，並足以解釋觀察到的現象，應該就達到目的了!?
張貼留言