Talk:Theta criterion
![]() | This article is rated B-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
|
![]() | This article was the subject of an educational assignment in Fall 2014. Further details were available on the "Education Program:University of British Columbia/Linguistics (Fall 2014)" page, which is now unavailable on the wiki. |
Untitled
[edit]RM Dechaine (talk) 05:50, 1 September 2014 (UTC)
I've added a few more links. If you feel the strong need to re-introduce the orphan tag, don't try too hard not to. Trigaranus (talk) 20:53, 27 January 2008 (UTC)
Peer review
[edit]Dear C3: Please leave your comment here. Thanks!Liu, Y. 90 (talk) 20:53, 2 November 2014 (UTC)
Kathyhht
[edit]In general, your article is solid and well-organized. You have provided good references and examples. I like how you use the bracket reference within the text. Here are some tiny suggestions. For the cross-linguistic generalizations section, you have provided three different languages which are great. I would suggest you to put a brief introduction before you talk about the examples of different languages in depth. Something about why did you choose to talk about those three languages but not some other languages? --Kathyhht (talk) 01:51, 3 November 2014 (UTC)
Argentum93
[edit]Like Kathy said, the article is generally well formed, and has detailed explanations supported by cited examples. I like how you used a box for the one-sentence definition for your topic, alone with three examples (of the same sentence) to explain the main concepts. But it would be even better to have some pictures (or tree diagrams) as visual guides. On the other hand, I feel that example sentences might look better under your explanations for them (maybe it's just me but.. I was confused at first glance, and thought that the examples belonged to the descriptions above them, but they were actually for the ones below them). --Argentum 93 (talk) 00:11, 4 November 2014 (UTC)
Sanjen50
[edit]I like the article in general. It is very well written and easy to understand with many examples given. I also liked how you put a lot of links that we can go to if we want more information on certain things, and you used the table to look very neat and to organize your content very well. Just small things though, there is some type errors. For example, the first sentence of the intro, there are two 'the's, so you might wanna fix that. Also, in the first sentence of the second paragraph, you guys use 'C' without saying what it stands for. I'm guessing that stands for complementizer and that is quite obvious for us, but you might want to use the full word at first and then use C for that. Also, I feel like the intro is too long... maybe some contents, especially some examples you have provided can be put under other categories? But again, in general, very well-formed and easy to read article. --Sanjen50 (talk) 01:15, 4 November 2014 (UTC)
Another thing I have found is that you guys use Theta role, theta role (without capitalization), and θ-role back and forth. I think it would be more consistent if you guys keep using just one term just to look clearer? But this is not a major issue. The citation is very well done! Again, the ideas seem all supported by some kind of literature and have references. I believe this article is a really good example for our own group to improve more.--Sanjen50 (talk) 06:19, 4 November 2014 (UTC)
Cnchia
[edit]Hi! Your group has done a good job linking terms to other Wikipedia pages, so that it’s really easy for the reader to obtain more information on a concept! The content is really good throughout and does provide a strong basis for your article. As for some suggestions, it would make the article a lot more concise if the references were linked with numbers at the end of the article using the Cite error: There are <ref>
tags on this page without content in them (see the help page). feature. Another possible suggestion would be to maybe consider writing out “theta” instead of using the symbol, just because theta criterion already uses alpha and beta symbols and by writing out theta, it may make the article a bit easier to read for the general public. Also, in respect to the layout, you may want to consider breaking up the first section by possibly adding a subsection to add clarity. Lastly, I really like the whole section of cross-linguistic generalizations and the language examples- really good progress so far and good job on providing proper citations for the examples.Cnchia (talk) 05:24, 4 November 2014 (UTC)
Danachos: intro & citations
[edit]Overall, you present a robust article. However, the introduction and page-layout could use some work.
Your introduction has two primary areas of concern for me: length and format. You put quite a bit of information where you should be summarising the entry in an easy-to-read/digest way. My recommendation is to gut the introduction and introduce instead a history subsection as well as an explanatory subsection, that way you can leave out the technical stuff from the introduction, effectively doing an ELI5 (explain like I'm 5) and getting into it below. The history would see much of your work with Chomsky and how the field began to look at theta criterion/roles over time. The explanatory subsection, titled something like "Function of theta criterion" or something, would then effectively take most of what is written in the intro and allow you to dig into it a bit more as well as be more clear in your explanation. For example, your usage of "position β" is unclear and should be better illustrated by either a diagram or further explanation. In terms of formatting, you have several, large hurdles. Firstly, please clean up your citations. The usage of citations (Bob, pg 137) should be minimally used if ever in Wikipedia. Please refer to the "ref" tag and do not write out citations under the "References" section. To do this, you do this: Example sentence is this (please look in the "Edit" section of the talk page to view this).[1]. After that, the references will automatically be referenced at the end of the article. To understand what I mean further, please refer to our page. Moreover, to cite multiple times the same source, you do this:[2], and then later on you just do this.[2] Please note that "cite news" can also be "cite book" or "cite web", more information can be found here.
Next, whilst it seems like it would be proper to quote something like that as per Wikipedia's guidelines, in actuality, you do not use a quotation template unless the quotation is exceptionally long (the same formatting rules apply when writing a paper: anything over four lines and you indent it). Moreover, for an introduction, you would never use a quotation box. You should write out the quotation like this, "as if it were any normal quotation",[3] and if the sentence continues like this, you just put the citation in the centre (reminder: read source material).
Finally, I really enjoy what you did for your cross-linguistic analysis, however you need to fix how you cite there as well. You do not put a citation in a quote box, you put it as a "ref" tag just before it appears. Under Malagasy, you have "This is shown in the following examples:". You would put the "ref" tag where I put the "R": "This is shown in the following examples:R" To help illustrate what I said here, I will direct you to several other pages that I believe have done a bang-up job. First, here is the Alcubierre Drive. Please pay attention to how it introduces the subject in an easy-to-digest way, goes on to explain the history in a brief but effective manner and finally then dives into the subject itself using technical terms. Think of Theta Critereon this way: people who have never heard of a theta role/thematic role are coming to read this page. You want to limit the amount of technical talk in the intro and bring it up when people are actually reading-reading the page. Here, another few are Quantum gravity, Muggle quidditch, Gender binary and Pitch (music) (I chose these because these are most likely subjects that you are not immediately experts of so that you can see how a Wikipedia article is to those without intimate understanding of the material). Each of these do an excellent job at first introducing the subject in a way that allows even those who have no knowledge of the subject access to it. This is what you should do: introduce theta criteria, go over the history is a brief and effective way and then explain its function. Then, and only then, you can go onto things like cross-linguistic input and special cases.
Final note: for the references, please check the "Edit" page of the "Talk" page as well as scrolling down to the bottom of the live Talk page. --Danachos (talk) 08:16, 4 November 2014 (UTC)
References
Response to Peer Review
[edit]To Kathyhht
Great suggestion! The reason we chose those three languages is simply that they belong to different families and typologies, we thought this would give the data a wide variety of diversity. We took your advice into consideration and have added a sentence explaining this at the beginning of the data section.
To Argentum93
Thank you so much for reporting this! We thought your advice was very helpful and would benefit the overall look and cohesiveness of our article. We have reversed the order of some examples and explanatory paragraphs. Now every example has at least one explanatory sentence before it. Is it easier to read now?
To Sanjen
Thank you for your suggestions Sajen. We have done a thorough edit for all potential grammatical errors like the repetition of the word “the” that you previously pointed out in the opening sentence. As far as explaining what the “C” meant, we have decided to explain certain syntactic elements and terms in our article for those readers that will not be able to understand all the terms we are using. We have only done this to a certain extent however, we decided that because of the instructions for this Wikipedia article that state that it should be written in a language that appeals to a reader who has previous basic knowledge on the topic, we are not going to go into detail and explain every little symbol or new term we are introducing. This would make the article too long, and would take away from the actual content we are trying to introduce.
To Cnchia
Thank you for your comments. What we found really helpful from your critique was the part about the referencing and about theta in terms of writing the word “theta” out vs. using the symbol. In terms of referencing, we noticed some obscurity in terms of being uniform. This was because in the earlier stages of our article contained different citation formats and this could be confusing as it was for you. We have standardized our referencing style so that the reader will hopefully be clear about them. In terms of the symbol “θ” and the word “theta”, we have seen both forms used in the literatures but we agree with you on that “theta” might be easier for the general public. Right now we have unified all “θ” to the lower cased “theta” except for the definition because that’s a direct quote. This keeps uniformity and will prevent confusion.
We’ve broken up the first section into a shorter introduction and an application section illustrating this criterion with specific examples.
To Danachos
Thank you so much for giving us such a detailed feedback! Please know that it is taken with sincere gratitude. Below is a brief response to the issues you raised:
--We didn't introduce many historical developments in the introduction section because we only meant to give a very brief definition there. More detailed info is provided in the "Special cases and controversies" section. We haven't got time to order the theories chronologically yet but that is definitely our goal.
--Your suggestion about parsing out Chomsky’s formulation of the theta criterion (with [γ ...α...β…]) is well appreciated. We have replaced all Greek symbols with words are still trying to make it clearer.
--It's so considerate of you to give us so many examples and useful links! The reason why we didn't change our citation style is that we've consulted Jonathan in tech support meeting and Rose-Marie also suggested in her announcement that such citation style is ok as long as it is unified within our article. The resources you provided are very useful though! We'll refer to them when needed.
--So far the only direct quotation we used is the one-sentence definition in the introduction section, but we'll pay more attention to quotation if we add more in the future. We haven't changed the "quotation box" in the introduction yet because Jonathan didn't show disapproval of it in the tech support meeting. We'll double check it later. If it does matter we'll change it.
Liu, Y. 90 (talk) 08:11, 14 November 2014 (UTC)
die the death
[edit]Is this the same as "John buttered the sandwich with *(expensive) butter"? (Pustejovsky would call this a shadow argument---"parameters which are semantically incorporated into the lexical item. They can only be expressed by subtyping or discourse specification" p.63--64 in The Generative Lexicon.) --Kiwibird (talk) 15:08, 1 May 2009 (UTC)
Adverbial usage not needed as an argument
[edit]Wouldn't this argument be the same as an adverb? He died + adverbial modifier. How did he die? He died painfully, quickly, etc. Here he died - [the death befitting a Roman soldier].
Merge with theta role
[edit]Shouldn't this stub be merged with theta role? —Preceding unsigned comment added by 70.138.177.168 (talk) 14:35, 7 October 2010 (UTC)
C4 Milestone 2: Annotated Bibliography of Theta Criterion
[edit]1970s
[edit]Jackendoff (1972)[1] is one of the early endeavors to accommodate semantic information into syntax and utilize it to limit the possible grammars. Following Gruber’s semantic analyses in the 1960s, the book defined thematic relations as the arguments of semantic functions CAUSE (takes an individual and an event argument respectively) and CHANGE (takes an individual, an initial state, and a final state argument respectively). Such relations between nouns and verbs are determined by the lexical entry of verbs and thus can be put into the formal representations of verbs under the framework of Chomsky’s Syntactic Structures. In addition, the author suggested a thematic hierarchy in which the relative positions of different thematic relations are universal: Agent > Location, Source, Goal > Theme. Inchoate as the theory is, it provided the rough building material for θ-criterion as well as many other later works.Liu, Y. 90 (talk) 11:09, 2 October 2014 (UTC)
In the article “Stimulus as a Semantic Role”(1978), written by Edward L. BLansitt Jr., Blansitt uses the term semantic role, which refers to the term theta role. He then compares the use of roles such as experiencer, agent and patient from the perspectives of different researchers. “The nuclear role which co-occurs with experiencer in emotive experience predications has been called patient ([Wallace L.] Chafe), neutral ([John T.] Platt), instrument ([Charles J.] Fillmore), an goal ([Robert E.] Longacre) (Blansitt, 1978 ). Blansitt argues that the nuclear role co-occurring with emotive experiencer represents a separate semantic role that he refers to as stimulus. “The term ‘patient’ is used by Chafe (1970: 164) in a very broad sense: ‘Every sentence contains a patient or agent noun, if not both, unless the verb is in the exceptional ambient category’” (Blansitt, 1978). [1] It is important to know where the roles originated from to know how and why they came to the way we use the roles today. This work is biased from the Blansitt’s point of view, as at this time every author had their own way of determining semantic roles. I personally do not find this article particularly helpful for writing a Wikipedia article on Theta Criterion because there are may different perspectives all over the place in this article, which would confuse a reader who is trying to learn about Theta Criterion currently as opposed to in the 1970’s.Cadicksh92 (talk) 01:54, 2 October 2014 (UTC)[2]
1980s
[edit]In his landmark book Lectures on government and binding[3], Chomsky brought θ-criterion to the world as a subsystem of the universal principles. The pith is that each argument bears one and only one θ-role, and each θ-role is assigned to one and only one argument. Its most important function is to define well-formed LF and restrict syntactic movement, namely, movement must be initially from a θ-position in D-structure to a non θ-position in S-structure and hence to subject. In the modular structure of Government and Binding Theory, θ-criterion interacts closely with other subtheories, especially projection principle and Case theory. Focusing on methodology building, Chomsky’s discussion was highly abstract and open to different interpretations, but no radical change has been made to this concept in later literature.Liu, Y. 90 (talk) 11:14, 2 October 2014 (UTC)
The article “Multiple Subcategorization and the Theta-Criterion,[4]” written by Ray Jackendoff encompasses the main argument that there is a correspondence between semantic argument positions and syntactic theta positions. The article showed how multiple argument positions are available in the cases of transitive verbs, transaction verbs and intransitive verbs. Ungrammatical cases were discussed which lead to show that conceptual semantics can resolve issues that could not be proven with the properties of syntactic structural positions. Finally, the article addressed that theta structure and argument structure are thought to be constraining on one another because the syntactic properties of a verb can make sentences ungrammatical, not the syntactic structure that has more than one post verbal complement. The significance of this article has to do with clarification of the general concept of Theta Criterion. It shines a light on a semantic way of addressing theta roles, oppose to the more traditional adopted view of predicate argument syntactic structure. The work exemplified through this article will add to the readers knowledge of how conceptual semantics plays a significant role in the theta criterion and can be a factor towards investigating issues in specific theta role examples. The work is written in a perspective that is slightly bias, because the author clearly states evidence trying to prove that certain issues regarding theta criterion and ungrammaticality cannot be investigated without the help of conceptual semantics. The author also clearly states that syntactic theory is missing out because it has a habit of only adopting traditional formalizations of syntactic predicate argument structure and does not explore the field of conceptual semantics. The bias in the work is clearly shown throughout the article in the language and the way the article is written. The bias is also demonstrated in the evidence that the author portrayed in the article, which only shows positive outcomes towards adopting conceptual semantic views.This article focusses on another way of explaining sentences that show ungrammaticality, different from the traditional syntactic view. It is interesting and intriguing for a reader to be able to understand the concept of theta criterion from another viewpoint(conceptual semantics), which is essentially one that is different. Angieghirra (talk) 00:05, 3 October 2014 (UTC)
1990s
[edit]In the article Thematic Roles and Syntactic Structure[5], Baker specified the earlier version of his influential Uniformity of Theta Assignment Hypothesis (UTAH) and accommodated it into Chomsky’s Minimalist Programme. He argues that in all languages, arguments of verbs can be classified into three categories of θ–roles that merge into different places of VP-shells: (i) An agent is the specifier of the higher VP (ii) A theme is the specifier of the lower VP. (iii) A goal, path or location is the complement of the lower VP. These simple and operative rules fixed down the absolute (instead of relative) positions of θ–roles in VP and made θ–criterion more falsifiable. The author reduced his bias when comparing his theory with others’ by using a wide range of languages and structures drawn from rich literature, which gives us a bird's-eye view of researches on θ–roles in the 1980s and 1990s. Liu, Y. 90 (talk) 11:50, 2 October 2014 (UTC)
In the Article” Trace Deletion, Theta-Roles, and Cognitive Strategies,[6]” a main argument that is discussed is regarding theta roles and how they are not universally agreed upon as being part of grammar. It discusses how they are only positions that hold a place and are not thematic labels. The article argues that these labels do not have syntactic meaning therefore have no reason to be encoded in syntactic data structures. The article then addresses Thematic Hierarchy Condition which is a principle that gives syntactic structure and constraints to theta roles because they have no grammatical and syntactical relevance alone. The value and significance of this article in respects to Theta Criterion is that it explores theta criterion beyond the basics of what an avid reader would previously know. The addition of the Thematic Hierarchy Condition gives additional insight that theta roles have constraints and conditions in terms of syntactical structure.The article is written with no obvious bias because it entails an experiment that involves the Thematic Hierarchy Condition and theta roles.The only form of bias in the article takes place towards the beginning where the author shows considerable preference towards proving the fact that theta roles have no syntactic power and that they are only labels that need additional constraints imposed on them.This article is interesting in terms of being able to visualize through experiment and examples the restrictions and constraints that theta roles experience, and how they have more significance than what an avid reader would know. This article does a very thorough job at showing how theta roles can be addressed at a syntactic level. Angieghirra (talk) 02:49, 3 October 2014 (UTC)
2000s
[edit]In the article “Integrating Verbs, Situation Schemas, and Thematic Role Concepts” (2001), researchers Todd R. Ferretti, Ken McRae, and Andrea Hatherell examine four experiments with respect to verbs’ thematic roles. In experiments 1 and 2: single-word priming shows that verbs immediately activate knowledge of agents, in experiment 3: verbs prime features common to patients, and in 4: single-word priming shows that verbs immediately activate knowledge of agents, and a cross-modal sentence priming – which shows the activation of agent and patient is modulated by syntactic cues (Ferretti, Hatherell & McRae, 2001). The perspective these researchers take is that “using an integrative account of the processing of words, thematic roles, and schemas, we conclude that this type of detailed world knowledge is tied tightly to on-line thematic role assignment, and thus should be considered as part of thematic role knowledge” (Ferretti, Hatherell & McRae, 2001). This article will be helpful when looking at the examples of thematic roles used in different concepts for those with little knowledge about theta criterion. The article is written with a neutral point of view from examining experiments and basing judgments on findings. I believe this article will be helpful for examining different verbs’ thematic roles and comparing the findings for a full understanding of the concepts.[7] Cadicksh92 (talk) 03:03, 2 October 2014 (UTC)
In the article “Do we Do We Need a Distinction between Arguments and Adjuncts? Evidence from Psycholinguistic Studies of Comprehension (2008).” researchers Tutunjian & Boland aim to settle the inter-theory discrepancy of creating a distinction between arguments and adjuncts or not. Because the Theta Criterion relies on and applies only to arguments, the article will be useful in helping briefly describe the notion of an argument. It will also provide useful as a way to make clear how an argument differs from an adjunct, thus making the idea of the Theta Criterion more clear for those with little knowledge of the subject. The work is conducted with a neutral point of view and later adopts support of the distinction between the two above notions after examining psycholinguistic evidence. While this article will be used only for a quick and simple explanation, I think it is an important distinction to note. [8]Hannahjtalbot (talk) 18:43, 1 October 2014 (UTC)
2010-2014
[edit]This article brings up an important question of theta roles in regards to reflexives: namely if a reflexive and it’s antecedent co-refer, which of the two bears which theta role? Sportiche then continues on to describe how this lack of distinction can hinder the clarity of unaccusativity. Through the use of French examples, Sportiche creates a new test to clarify both the strength and unaccusativity of a construct. Sportiche writes with the perspective that reflexives in French are not accusative. This article may prove to be significant to our research on the Theta Criterion because it will provides us with at least groundwork for exploring a problem of assignment of theta roles in our article. Sportiche does note that her findings may not apply cross-linguistically (specifically to English), hinting at a possible shortcoming of her work. This article does focus on unaccusativity and only uses theta roles as groundwork, which may prove for this specific article to be not very useful in the long run. On the other hand, it does raise an interesting point about theta roles and reflexives that we might not have discovered without it. [9]Hannahjtalbot (talk) 06:05, 2 October 2014 (UTC)
With the case of copy raising, Asudeh and Toivonen tease apart the concepts of thematic roles and semantic roles, and perception verbs and copy raising verbs, for copy-raising involves control syntactically and raising semantically.[10] This article has thorough arguments, informal and formal, embellished with examples from two Germanic languages, English and Swedish. -C4gw (talk) 07:07, 3 October 2014 (UTC)
References
[edit]References
- ^ Jackendoff, Ray S. (1972). Semantic interpretation in generative grammar (2. print. ed.). Cambridge, Mass.: MIT. ISBN 9780262100137.
- ^ Blansitt, Edward L. (1978). "Stimulus as a Semantic Role". Valence, Semantic Case and Grammatical Relations. 1: 311–324. Retrieved September 23, 2014.
- ^ Chomsky, Noam (1981). Lectures on government and binding : the Pisa lectures (5. ed. ed.). Dordrecht: Foris. ISBN 9789070176136.
{{cite book}}
:|edition=
has extra text (help) - ^ Jackendoff, Ray (1985). "Multiple Subcategorization and the θ-Criterion: The Case of Climb". Natural Language & Linguistic Theory. 3 (3): 271-295. doi:10.1007/BF00154264.
- ^ Baker, Mark C. (1997). "Thematic Roles and Syntactic Structure". In Haegeman, Liliane (ed.). Elements of grammar : handbook in generative syntax (Student éd. ed.). Dordrecht [u.a.]: Kluwer. p. 73-137. ISBN 0792342976.
- ^ Grodzinsky, Yosef (1995). "Trace Deletion, θ-Roles, and Cognitive Strategies". Brain and Language. 51 (3): 469-497. doi:10.1006/brln.1995.1072.
- ^ Ferretti, Todd; McRae, Ken; Hatherell, Andrea (May 2001). "Integrating Verbs, Situation Schemas, and Thematic Role Concepts". Journal of Memory and Language. 44 (4): 516–547. doi:10.1006/jmla.2000.2728.
- ^ Tutunjian, D., & Boland, J. E. (2008). Do We Need a Distinction between Arguments and Adjuncts? Evidence from Psycholinguistic Studies of Comprehension. Language and Linguistic Compass, 2, 631-646. doi: 10.1111/j.1749-818X.2008.00071.x
- ^ Sportiche, D. (2014). Assessing Unaccusativity and Reflexivity: Using Focus Alternatives to Decide What Gets Which θ-Role. Linguistic Inquiry, 45, 305-321. http://dx.doi.org/10.1162/LING_a_00156
- ^ Asudeh, Ash; Toivonen, Ida (2012). "Copy raising and perception". Natural Language & Linguistic Theory. 30: 321-380. doi:10.1007/s11049-012-9168-2.
Technological Questions
[edit]- Are there any trends in terms of technological problems that other groups are having? Any pointers on those common problems?Hannahjtalbot (talk) 23:08, 27 October 2014 (UTC)
Here's a question: I uploaded a tree (a .jpg file), and wiki tells me that I haven't specified the copyright and licensing information. How am I supposed to do that? — Preceding unsigned comment added by C4gw (talk • contribs) 22:37, 29 October 2014 (UTC)
If we used a software to draw pictures, do we need to mention the name and developer of the software?Liu, Y. 90 (talk) 01:39, 4 December 2014 (UTC)
Proposed Outline
[edit]Introduction
[edit]- Definition
- Short Description/Links to: argument, Theta role
--Hannahjtalbot (talk) 23:35, 29 October 2014 (UTC) Do u think I should avoid using the symbols like α, β or C (this is not complementizer though, just a letter)? I copied them from Chomsky's book because I thought that would be accurate, but after C3's suggestion I guess it's better to replace them with words?Liu, Y. 90 (talk) 21:19, 12 November 2014 (UTC)
Theta-criterion applied
[edit]Theta grids
[edit]Data Section
[edit]Inuit
[edit]Inuit is an ergative language in the Eskimo-Aleut family that allows systematic noun incorporation into verbs in sentences (Bitter 1994) . However, the theta-criterion restricts the type of theta-roles of noun phrases that can incorporate into the verb. Such effect can be seen from example (2) below:
<AGENT> <THEME> [VERB] (2) Juuna alIa-mik ilinniartitsisu-siur-p-u-q |
In example (2), only the theme ilinniartitsisu 'teacher', but not the agent Juuna, can be incorporated into the verb root. Thus, it is impossible for this sentence to be interpreted as 'another teacher is looking for Juuna'. According to Baker (1997, p. 82) this difference is due to the verb obligatorily assigning the theta-role "theme" and "agent" to different positions with the former structurally closer to the verb.
Chinese
[edit]As a pro-drop language (Sportiche, Koopman & Stabler 2014, p. 244) , Chinese can drop subjects more freely than English. However, theta-criterion does hold in Chinese because without context predicates cannot have more or less arguments than their theta-roles. This can be seen from the Mandarin examples below:
<AGENT> <GOAL> <THEME> (3)a. tamen gei-le jingli *(yi-fen baogao). they give-LE manager a-CL report[b] 'They gave the manager *(a report).' <AGENT> <??> b. ta zou-le (*women). he walk-LE us 'He walked (*us).' (Huang, Li & Li 2009, p. 39) |
In (3a), the verb gei 'give'has three theta-roles to assign: Agent, Goal, and Theme. Thus there must be three arguments in the sentence to shoulder them respectively: tamen 'they', jingli 'manager' and yi-fen baogao 'a report'. If yi-fen baogao is omitted the sentence will be ungrammatical because the theta-role Theme cannot be assigned.
(3b) is the opposite type of violation where there are two arguments in the sentence ta 'he' and women 'us' but the verb zou 'walk' can only assign one theta-role Agent. This sentence crashes because the argument women 'us' cannot get a theta-role. [c]
Malagasy
[edit]Malagasy is a language in the Austronesian family (Guilfoyle, Hung & Travis 1992) . The theta-criterion in Malagasy restricts the relative position of verb and its arguments. As in English cases, regular transitive verbs in Malagasy take two arguments and assign one theta-role to each, consistent with the theta-criterion. For example, the verb sasa 'wash' has two theta-positions to assign: agent and theme. The verb matches agent to its specifier and theme to its complement. If a third noun phrase exists in the sentence, it must be an adjunct and linearly further to the right in the surface form of the sentence. When the verb raises to the head of IP and appears to the left of the whole phrase in the surface form, the noun phrases assigned agent and theme will appear to be adjacent to each other and one of them will be adjacent to the verb. This is shown in the (3a) below:
[VERB] <AGENT> <THEME> [ADJUNCT (Topic)] (3a) An-sasa-na (anasan')ny zazavavy ny lamba ny savony. XT-wash the girl the clothes the soap[d] 'The soap was washed (with) the clothes by the girl.' (Guilfoyle, Hung & Travis 1992, p. 381) |
In this sentence, ny zazavavy 'the girl' is the agent, ny lamba 'the clothes' is the theme, and ny savony 'the soap' is an adjunct. Because the theta-roles assigned in D-structure can be transmitted through trace (Chomsky 1981, p. 101), the noun phrases ny zazavavy and ny lamba still have proper theta-roles in the surface form and the sentence is grammatical.
Sometimes either the agent or theme will be moved to the topic position at the end of the sentence. In such cases, the other one will be directly adjacent to the verb. This is shown in examples (3b-c):
[VERB] <THEME> [ADJUNCT] <AGENT>(Topic) (3b) M-an-sasa (manasa)ny lamba amin'ny savony ny zazavavy. AT -wash the clothes with the soap the girl 'The girl washes the clothes with the soap.' [VERB] <AGENT> [ADJUNCT] <THEME>(Topic) (3c) Sasa-na (sasan') ny zazavavy amin'ny savony ny lamba. wash-TT the girl with the soap the clothes 'The clothes are washed with the soap by the girl.' (Guilfoyle, Hung & Travis 1992, p. 380) |
Again, the theta-roles are not lost after movement so these two sentences are still deemed grammatical. However, if an adjunct noun phrase intervenes between the verb and its agent and theme, the verb will have no way to assign the theta-roles properly and the sentence will consequently be ungrammatical. For example:
[VERB] [ADJUNCT] <THEME> <AGENT>(Topic) (3d) *Nanasa omaly ny lamba ny zazavavy. past-AT-wash yesterday the clothes the girl 'The girl washed the clothes yesterday.' [VERB] [ADJUNCT] <THEME> <AGENT>(Topic) (3e) *Nanasa ho an'ny ankizy ny lamba ny zazavavy. past-AT-wash for the children the clothes the girl 'The girl washed the clothes for the children.' (Guilfoyle, Hung & Travis 1992, p. 383) |
In these two sentences, although the agent ny zazavavy moved to the end of the sentence properly, there are adjuncts omaly 'yesterday 'and ho an'ny ankizy 'for the children' intervening between intended theme ny lamba and the verb. Because of the intervening adjuncts, the verb cannot assign a theta-role to the argument ny lamba, leaving a mismatch between the number of theta-roles and arguments. Thus, neither example 3d or 3e is grammatical (Guilfoyle, Hung & Travis 1992) .
Japanese
[edit]![](http://upload.wikimedia.org/wikipedia/commons/thumb/7/76/Japanese_example_before_movement.png/220px-Japanese_example_before_movement.png)
In Japanese, theta-criterion triggers affix raising (Kitagawa 1986) . For example, in a sentence with -ga, -ni, -o, and -sase like (4) below, sase 'CAUSE' first combines with tabe 'eat' to form a complex lexical item when it is introduced into the sentence, after which it raises to the sister node of the verb phrase [kodomo-ni [okasi-o [tabe]]] (the child eat sweets) to satisfy theta-criterion:
(4) Hanako-ga kodomo-ni okasi -o tabe-sase ru Hanako NOM child DAT sweets ACC eat-CAUSE PRES[e] 'Hanako let the child eat sweets' (Adopted from Kitagawa (1986, p. 69) ) |
![](http://upload.wikimedia.org/wikipedia/commons/thumb/4/48/The_position_where_%27sase%27_lands.png/220px-The_position_where_%27sase%27_lands.png)
The motivation behind such movement is that, both sase and tabe need to assign two theta-roles: agent & event and agent & theme, respectively. Theta-criterion thus forces sase to move to the sister node of the verb phrase [kodomo-ni [okasi-o [tabe]]] (the child eat sweets) and take Hanako-ga as subject in LF so that sase can assign "agent" to Hanako-ga and "event" to the verb phrase [kodomo-ni [okasi-o [tabe]]]. tabe, at the same time, assigns "agent" to kodomo 'child' and "theme" to okasi 'sweets'. Both of their theta-marking properties are satisfied and every argument gets a theta-role. If sase doesn't move in LF or land on other sites like a sister node of [okasi-o [tabe]] (eat sweets), the sentence will crash for the violation of theta-criterion.
Above is the deleted data section just as backup material. Feel free to use them if you like. The reason I deleted them is that none of the foreign languages I listed here revealed typical effects of theta-criterion. The Inuit and Malagasy data are more about UTAH than about theta-criterion because the negative data are not crushed for wrong numbers of arguments, but for the arguments' not in proper position. The Japanese data is fine but the so-called affix raising doesn't produce any difference in word order so it's unfalsifiable. The Chinese data is the only section that is really about theta-criterion, but it is pretty much the same as the English data so doesn't help deepen the understanding. Besides, most research on theta-criterion in Chinese is on resultative structures and pro instead of the simple clauses listed here. If we really want to use it we can put it in the "applied" section and don't need an independent "foreign" section for it. Liu, Y. 90 (talk) 02:08, 9 November 2014 (UTC)Liu, Y. 90 (talk) 05:45, 11 December 2014 (UTC)
Theory Section
[edit]optionality: some verbs can be intransitive, transitive, and ditransitive depending on context. e.g. verbs that take cognate objects (as in “He died a gruesome death.”).C4gw (talk) 23:35, 29 October 2014 (UTC)
control verbs: since big PRO is invisible, it seems on the surface a sentence like this violates theta criterion: John wants [PRO to be arrested t].
Can we move the raising example in PRO section to DP raising section? DP raising section can include raising to subject and raising to object. Liu, Y. 90 (talk) 20:36, 12 November 2014 (UTC)
movement: the surface form of DP-raising sentences seem to violate theta criterion, too
small pro (or subject drop): there seem to be no subject. e.g. Sta bene.<--Italian
If literature on pro is so scant maybe we can merge this section with big PRO?
Maybe trace deletion (Grodzinsky 1995)? Both Grace and Angie have read this, and it is widely cited, so why don't we write something about it?Liu, Y. 90 (talk) 04:19, 2 November 2014 (UTC)
Refuting Theory
[edit]-Pesetsky (1995) http://webcat2.library.ubc.ca/vwebv/holdingsInfo?bibId=1014333 <--out textbook recommended this but I haven't read it
-Hornstein (1999) https://www.msu.edu/course/lin/834/pdf/hornstein.pdf <--about PRO but abandoned theta criterion. cited by 1000+
Supporting Theory
[edit]DP Raising
[edit]example: (9) a. Jean wants Briani [ti to leave]. b. Jean persuaded Brian [PRO to leave]. (Carnie 2012, p. 429)
Example (9a) is a subject-to-object raising sentence; "Brian" raises to the object position of the verb wants. In contrast, (10b) is an object control sentence.(Carnie 2012, p. 430) The verb persuade has three theta-roles to assign: "agent" to Jean, "theme" to Brian, and "proposition" to the clause [PRO to leave]. There is no raising, but there is a PRO in the subject position of the embedded clause that takes the verb leave's only theta-role, "agent". Since Brian does not receive theta-role from leave, it only bears one theta-role, nor does PRO receive a second theta-role from persuade. Every argument only receives one theta-role, and every theta-role of the two predicates is assigned to only one argument. The sentence is thus grammatical.
other things we can talk about:
-Aspect: http://search.proquest.com/docview/58305547?pq-origsite=summon, http://psycnet.apa.org/psycinfo/1991-97168-000
-Light verbs: textbook chapter 13, http://www.jstor.org/stable/4178587
-A, N, P as the Prof suggestedLiu, Y. 90 (talk) 03:08, 31 October 2014 (UTC)
Cite error: There are <ref group=lower-alpha>
tags or {{efn}}
templates on this page, but the references will not show without a {{reflist|group=lower-alpha}}
template or {{notelist}}
template (see the help page).