Things we can learn from an OD construction Adele E. Goldberg Predication of a role introduced by a context-evoked frame. Poset: the initial Np satands in a salient partially-ordered set relation to som eentity or entities already evoked in the discourse model. (see also Ward 1988; Ward and Prince 1991): IS-A-MEMBER-OF, IS-PART-OF, IS-A-SUBTYPE-OF, IS_AN_ATTRIBUTE-OF, IS-EQUAL-TO. Intro From LM’s manuscript: Should iterated events (a box repeatedly falling down) be treated the same as habitual events (brushing one’s teeth every morning)? Langacker makes a distinction (1996) GNOMIC sentences (Oil flots on water: can be paraphrased with conditinal sentence: if something counts as oil, it will float on substance that counts as water. Krifka et al. 1995: conflate habitual and gnomic sentences under GENERIC) “You have highway signs saying speed limit, uh, eighty, eighty-five, whatever would be the appropriate number.” “You have a bunch of people trying to do ninety” Ellen F. Prince. 1997. On the Functions of Left-Dislocation I English Discourse. In Akio Kamio (ed) Directions in Functional Linguistics.117-143. 29a. [Speaker has been married four times before the present husband] “the first time was in 1968, just to get out of my dad’s house,” she says. “Second guy, I just met him and didn’t have anything else to do.[LD] Didn’t work out…Third and fourth times were business partners. We got married for business reasons.” (Philadelphia Inquirer 4-J 73/88) 29b. I was just raised an old hillbilly and I’ll die one. Radio, it’s sitting up there, but I can’t hear too good. Don’t have a television. (Terkel 1974: 40). Fronted NP stands in a poset (partially ordered set) relation. Prince: “replacing the initial NP with an N’ in actually occurring ‘Simplifying’ LeftDislocations produces infelicity, if not ungrammaticality, as seen in (30)” (pg 135) 30a. It’s supposed to be such a great deal. #Guy, when he came over and asked me if I wanted a route, he made it sound so great. Seven dollars a week for hardly any work. Two problems: unanchored and punctual past: Anchored but punctual: odd: [It’s supposed to be such a great deal. #Nice kid like him, when he came over and aske dme if I wanted a route, he made it sound so great. He made it sound so great. Seven dollars a week for hardly any work.] Unanchored, non-punctual: odd: [It’s supposed to be such a great deal. # Guy, he comes over and asks if you want a route. He made it sound so great. Seven dollars a week for hardly any work.] Anchored and non-punctual: OK: [It’s supposed to be such a great deal. Nice kid like him, he comes over and asks if you want a route. He made it sound so great. Seven dollars a week for hardly any work.] BUT 29 a is punctual!! 30b. My sister got stabbed. She died. Two of my sisters were living together on 18th Street. They had gone to bed, and this man, their girlfriend’s husband, came in. He started fussingwith my sister and she started to scream. #Landlady, she went up, and he laid her out. So sister went to get a wash cloth to put on her… [I don’t want to manage an apartment building any more. People in apartments have been getting stabbed. #Landlady, she intervenes and she gets laid out. So my sister is going to open a new business with me.] cites Ward (1988): “topicalization” (Laura: is a focus construction): Customer W: ‘H recommends the cheesburger’ C: ‘Cheeseburger it is then. Prince (pg 136): “This suggests that further study of possible relations between poset inferences and initial N’ might prove fruitful. ********************* Lexical subjects in Switchboard corpus: 2,858 Argument focus: I was the only one who did not catch a single fish. My daughter caught fish, his daughter caught fish, he caught fish. *Daughter caught fish, son caught fish, I caught fish. SF: Doctor came out. (ok) Examples of lexical subjects from Switchboard, cited by Francis, Gregory and Michaelis (ms). Context: Conversation about drug testing We, that’s been an, a an issue, uh,in our company even though we don’t have the random or even regular drug screening. In fact, they’l have these little parties,and people will just get, I mean I’ve my brother lives where I work, and I have many a time called him to come get me, you know. *Brother lives where I work. Context: conversation about children “he sent him to kindergarten. As soon as he went there, the teacher took one look at him and he threw up again.” As soon as he went there, ?eacher took one look at him and he threw up again. The, uh, Governor, you know, has been trying to decide whether he’s going to commute it or not. ??Governor, you know, has been trying to decide whether he’s going to commute it or not. And you get up there and the prosecuting attorney presents his evidence. (definite because of bridging inferece (Clark and Haviland 1977) And you get up there and prosecuting attorney presents his evidence. We should and especially, if anything, be cutting taxes right now because of the recession and at the same time, the budget he sent to Congress has tax and fee increases, so, uh, I know the politicians, uh, aren’t straightforward. “As lexical NPs, they denote NEW referents. As subjects, they denote RECOVERABLE referents” (Francis, Gregory and Michaelis pg 13) Look at Steedman? 1984 Lng for topic intonation.(Christine Bartels) In this paper, I will focus on the case of determinerless noun phrases in initial position in English, an unusual pattern that has received scant attention, with few exceptions (most notably, Thrasher 1974, Cote 1996).Certain instances of this pattern are highly conventionalized: Do sentences typically involve recurrent, generic or stative predications? Main predication treated as if it is to be expected. Cannot add parenthetical between subject and Pred. Wife's always out of town. #Wife, after each surgery, is out of town. Check's in the mail. Sky's the limit. Receipt's in the bag. Can have PF interpretation… These cases might lead one to imagine that the existing cases are all idioms. The following examples attest to the fact that the pattern is in fact productive: Wife's out of town. Car's in the shop Copier's on the blink again. Printer's out of paper. Ice cream's in the freezer. Sun set early today. Old man's out of town. Foolish child though he was... (Fillmore 19XX) (LD construction) Elevator's broken. Boy's a real linguist! (are exclamatives SF?) Bus was late. (said as an explanation in context) #Bus ran into a stop sign. (Ok if uttered by Eye-Ore) Garbage collector came already. Radio said he was planning to run. Newspaper boy's at the door . % \# Husband's out of town. Let's go drink some beer (UIUC students liked this). TYPES Time expressions: ``How long you been here, anyhow?' ` Little over five weeks'" I said.. pg 60 I'll be back around ten. Couple of weeks ago,… Long time ago.. Little after' pg 103 KID… MODIFIERS often essential (to anchor N’?) % Good thing we didnt' run into anyone we know Last person I expected to meet was John Wife wants to go to the mountains this year. pg 5 %% (The) Damn dogs are taking over the city. % (A) Man your age shouldn't be working that hard. %% (The) Last man to see him alive was John. % (A) Couple of weeks ago I decided to stop smoking. % (The) Evening paper says he'll tough it out. %% (A) Funny thing happened on the way over here. It is argued below that these cases involve a grammaticalized construction which I will refer to as the Omitted Determiner (OD) construction. By construction, I intend the following: C is a CONSTRUCTION iff C is a pairing of form and function such that some aspect of the form or some aspect of the function is not strictly predictable from C's component parts or from other constructions already recognized to exist. That these expressions require a special construction is clear from the fact that singular count nouns in English do not generally appear without determiners and yet the nouns check, receipt, wife, car, copier, etc. all do appear in bare form in the examples above. The fact that nouns can appear in bare form in certain contexts requires that they be licensed by a minor pattern in the language, a special construction.We will see below that the construction has both syntactic and discourse constraints, and that in fact the form and meaning of the construction are closely related. The explanation of the relationship between form and meaning requires reference to a theory of paradigmatic opposition (see also Moore and Ackerman 1999,Lambrecht 1994,2000). Syntactic Constraint Determiners cannot be ommitted from just any noun in a sentence. For example, none of the following with determiners ommitted from non-initial position are acceptable: *I met wife today. *I saw a book on copier.1 *According to wife, ... Examples like these led Thrasher (1974) to propose that only determinres in sentence initial position are candidates for deletion. However, initial position is not quite the correct constraint (see also Thrasher 1974:89).A determiner is omitted from each of the following in a non-initial position: He always looks great. Hair's perfect, suit is tailored and shoes are newly waxed. [List intonation helps here, no?] Oy vey! Nine days out of ten, printer's out of ink around here! Last night, evening paper said she'll run for office. %[what about?: %\enumsentence{Wife and kids getting used to life in Ann Arbor? pg 83 \\ %*Are wife and kids getting used to life in Ann Arbor An initial generalization that covers both the facts in \ex{-1 and \ex{-0 is that determiners can only be ommitted from subject position. Whether or not the subject is sentence initial is in large part irrelevant. However, the following examples falsify this generalization, since all involve bare nouns that are not syntactic subjects. ``Kid like you' he said, peering across at me from the driver's seat, 'smart, good looking, good personality, you can make a million bucks here'" pg 54 There is a different minor pattern in English which involves bare nouns after prepositions to school, in bed, at work. The semantic constraints involved in these cases are quite distinct so that a distinct construction is clearly involved. 1 Husband and Wife, they got to work these things out themselves. You'll see. ```Smart, good-looking kid like you,' he said 'you can do better than this." pg 61 No poset relation, is there? “Simplifying” function, no?? So why are Prince’s examples 30 bad? Because events are punctual?? Each of these examples is an instance of the left detachment construction, and the bare NP appears in the leftward detached phrase. Thus it seems that the OD construction is licensed in subject or a left-dislocation position. For the time being, we will leave the statement of the generalization in these disjunctive terms. In section XX below, we return to refine the syntactic constraint, and argue that it is in fact highly motivated once the relevant discourse constraint is fully understood. Discourse Constraint It is clear that there is some kind of discourse constraint on this construction. If we compare the examples in \ex{1, provided by Thrasher (1974), with the same examples in the contexts in \ex{2, we find a distinct contrast:\footnote{I use the "\#" symbol here and below to indicate infelititous in the given context. "*" is reserved for examples which cannot be rescued by a change in context. Bill. (Your) Wife's on the phone. Got to go. (My/The) Kids are waiting in the car. % Bill's not coming in. (His) Son just called to say he has the flu. (Thrasher 1974: 83) Bill, is your family still on their trip? \\ \# (Your) Wife's on the phone again. % Bill has three children. \# (His) Son just called to say he has the flu. A: How are baby Greta and Ollie ? \\ B: They're fine. \# (The) Kids are actually waiting in the car. How are the children? Oh, kids are fine. Before trying to characterize the relevant discourse constraint, it is worth reviewing some general facts about information structure. \subsection{Background on Information Structure % % %In this section, I would like to propose a further constraint on this %construction, having to do with the information structure of the %utterance conveyed. I adopt the following definition of information structure from Lambrecht (1994:5): \begin{quote Information structure: that component of sentence grammar in which propositions as conceptual representations of states of affairs are paired with lexicogrammatical structures in accordance with the mental states of interlocutors who use and interpret these structures as units of informaiton in a given discourse. \end{quote Information structure as used here is equivalent to Prince's ``informationpackaging" (Prince 1981, see also Vallduv\'i 1992). \subsection{Sentence Focus The contexts in the unacceptable examples in \ex{0 set up the subject referents as the topics of the discourse. The relevant subject in example \ex{0a, wife is already mentioned in the previous context; son in \ex{0b is likewise evoked by the previous mention of children. Finally kids in \ex{0c refers to Greta and Ollie just mentioned. In the original examples in \ex{-1 in contrast, each of the contexts encourages a particular interpretation in which the subject referent is not topical. Instead, the whole sentence expresses new information. Lambrecht (1994) dubs this type of expression, ``sentence focus" in contrast to topic/comment structure (or ``predicate focus"). The difference between predicate focus and sentence focus can be observed in the stress pattern in two syntactically identical sentential patterns in English. What happened to Bill? Bill was CONVICTED. (predicate focus) What happened? BILL was convicted. (sentence focus) An example of a predicate focus construction appears in \ex{-1. Bill is already evoked as the discourse topic, and the only new information is contained in the predicate phrase. As Lambrecht (1994) notes, predicate focus, or topic/comment structure is the most frequent, and unmarked way of packaging information. In \ex{0b, on the other hand, the entire sentence conveys new information (hence the term, ``sentence focus"). Bill has not been evoked by the question. The pragmatic distinction is marked formally in English by placing the a sentence accent on the subject NP instead of at the end of the predicate, the unmarked accent position (see Lambrecht and Polinsky 1997 for discussion). %The primary difference between %SF and PF are distinguished by TOPICALITY PRESUPPOSITION: %assumed status of a referent as a topic of current interest in the discourse. %One way to indicate SF information structure in English is by the stress pattern. %For example, in answer to the question, ``Why didn't she come to work today?" %one might respond with either of the examples in \ex{1: % %% Her SON is responsible (AF) % Her son had an ACCIDENT (PF) % Her SON is sick (SF) %In \ex{0a, we have an expression with predicate focus. To be more precise, Lambrecht (1994) offers the following definition: \begin{quote SENTENCE FOCUS CONSTRUCTION: the Sentence Focus construction is formally marked as expressing a pragmatically structured proposition in which both the subj and predicate are in focus. The focus domain is the sentence minus any topical non-subject arguments. \end{quote This type of information structure has alternatively been referred to as "news sentence" "neutral desctription" "all new utterance" "thetic sentence" "event-reporting sentence." If we look at the examples of determiner omission cited initially, we find that they each require a sentence-focus interpretation. For example, \ex{-5?a, repeated below, might be uttered, for example, as an explanation of why someone was eating alone at a restaurant. The fact that the man's wife was out of town serves to set the scene or explain the current situation. \enumsentence{Wife's out of town. The same sentence cannot be used if the wife was just mentioned in the discourse (cf. example \ex{-6>??. Similarly, example \ex{0c is intended to draw the listener's attention to the wife's specific actions and thus would normally be used when the wife was topical in the discourse; we see that the possessive determiner cannot be ommitted in this context (\ex{0b). \# Wife stood up suddenly. (His wife stood up suddenly.) As the contrasts in \ex{-6 and \ex{-5 show, while certain types of content lend themselves more or less easily to sentence focus interpretations, it is ultimately the contexts, not the contents which determine the interpretation and hence the acceptability of these sentences. %Compare \ex{1a with \ex{1b; % A: Why are you walking to work? \\ %B: Car's in the shop. % Where is your Honda? \\ %\# Car's in the shop. (with “Oh” this is ok!) The role of discourse context becomes even more clear when we consider more carefully which types of sentence focus expressions license the OD construction. That is, the term sentence-focus serves as a cover term to distinguish expressions from predicate focus. In fact it covers two distinct cases (Sasse 1987; Lambrecht 1994). The communicative function of sentence-focus constructions is either 1) to introduce a new discourse referent (``entity central" or ``presentational") or 2) to announce a new event involving a non-topical referent (``event central"). This is diagrammed below: \medskip \begin{tabular{ccc & sentence-focus & \\ &\hspace{.1in& \\ entity central & & event central \end{tabular \medskip \subsubsection{Presentational/Entity central % *Man appeared. % *Baby was born. Presentational expressions can generally appear in the existential there construction or the inversion construction.\footnote{While presentational sentences can generally appear in these constructions, it is not the case that these constructions require the presentational conversational import. As Birner (1994) observes, for example, the inversion construction has the weaker requirement that the preposed phrase conveys more familiar information than the postposed logical subject. If the logical subject is newly presented in the discourse it can be assumed to be the least familiar information in the sentence and thus inversion is generally allowed. Consider the following examples: There (suddenly) appeared a wheel. In came his wife. At the door stood the garbage collector. Notice that the canonical paraphrases of these sentences cannot appear with omitted determiner: *Wheel (suddenly) appeared. *Wife came in. *Garbage collector stood at the door. Thus it seems that the ommitted determiner construction is incompatable with presentational information structure. At first it looks like the following are exceptions to this generalization since they involve the ommission of an indefinite determiner and thus superficially at least seem to introduce a new participant onto the scene. Crazy situation came up. Funny thing happened on the way over here. (Thrasher 1974) However, upon closer inspection it is clear from the content of these examples that they are actually introducing a new event. The ``crazy situation" or ``funny thing" both refer to events that will be described subsequently, despite the fact taht the events are encoded by NPs Notice the corresponding inversions and existential constructions are also unacceptable: *Up came a crazy situation. *There came up a crazy situation. *On the way over here happened a funny thing. *There happened a funny thing on the way over here. Thus it seems that the examples in \ex{-1 are not in fact entity central expressions, but rather are event central sentences. That is, the distinction is ultimately a semantic one whether an entity or an event is introduced into the discourse. Only ``event central" sentence-focus contexts license the OD construction. \section{Inferrables We have seen that the ommitted determiner construction requires the event central sentence-focus information structure. This generalization implies that the subject cannot be topical in the discourse, since when the subject is the topic, we have an instance of the predicate focus pattern. %For exmaple, wheel in the following example is %topical, since A introduces a question about it: %\enumsentence{A: Why did you put that wheel away? \\ %B: \# Wheel's smashed up. \\ %(B: The wheel's smashed up.) At the same time, we have seen that the subject cannot be newly presented on the scene. %\enumsentence{ \# (A) Wheel rolled down the street. %What role in the discourse can the subject argument play in %the ommitted determiner construction? %We can get a clue by considering a context in which B's response in \ex{-1 is %made acceptable: %\enumsentence{A: Why did you put that bike away? \\ %B: Wheel's smashed up. %In \ex{0, wheel is {\small INFERRABLE. The OD construction requires that the subject referent be distinct from both topic and new information: it must be inferrable; that is, the speaker assumes that the hearer can infer its existence from context, general world knowledge or semantic frames previously evoked (Prince 1981, Fillmore 1982, Lambrecht 1994, Byrne 1997).\footnote{ In fact it seems that all event central sentences require inferrable subjects. %The constraint that the subject must be inferrable appears to be a %general constraint on SF expressions. %This explains the unacceptability of the following: %\enumsentence{A: What happened? \\ %B: *MAN died. %Although the stress pattern and context favor a sentence focus interpretation, %man is not inferrable. Chafe (1987) defines an accessible discourse entity as one which is semi-active. A discourse referent may be accessible because it has been partially deactivated from an earlier state due to intervening discourse, because it is inferrable from some semantic frame or schema that has been evoked, (Chafe 1987), or because it is present in the text-external context (Lambrecht 1994: 99). %\subsubsection{Inferrables %Various types of inferences Clark (1977), Prince (1981a). %Prince: set to subset, set to number, number to set). % Check's in the mail. INFERRABLE % Sky's the limit. INFERRABLE % Receipt's in the bag. INFERRABLE %These cases might lead one to imagine that the existing cases are %all idioms. %However the following further examples attest to the fact that the %pattern is in fact productive: % Wife's out of town. INFERRABLE % Car's in the shop INFERRABLE % Copier's on the blink again. % Printer's out of paper. % Ice cream's in the freezer. % Sun set early today. % Old man's out of town. % Foolish child though he was... (Fillmore 19XX) % Elevator's broken. % Boy's a real linguist! (are exclamatives SF?) % Bus was late. (said as an explanation in context) % Garbage collector came already. % Radio said he was planning to run. % Newspaper boy's at the door . % \# Husband's out of town. Let's go drink some beer (UIUC students liked this). % % %x, {x e X %X, x e X %But not active (easily accessible, in the forefront of the hearer's consciousness) (and not topical) in all of the above. % Left detachment. or ``marked topic construction" Lambrecht (1994): Accented topic: ``[the leftward NP bears ] the primary function of announcing a new topic or of marking a shift from one topic to another" (1994:202). Gregory and Michaelis (1999) operationalize this definition and confirm it using corpus data. In general, left detached NPs are typically not brand new, but are instead accessible (Lambrecht 1994: 183). The detachment or marked topic construction can ...be defined pragmatically as a grammatical device used to promote a referent from...accessible to active status, from which point on it can be coded as a ... an unaccented pronominal." (Lambrecht 1994: 183) ``Kid like you' he said, peering across at me from the driver's seat, 'smart, good looking, good personality, you can make a million bucks here'" pg 54 Husband and Wife, they got to work these things out themselves. You'll see. ```Smart, good-looking kid like you,' he said 'you can do better than this." pg 61 \section{Relating form and discourse function In this section we will address the question of the relation between the syntactic form and discourse constraint.. that the relationship between the formal characteristic of the missing determiner and the discourse function of non-topicality is strongly motivated. Lambrecht (1994) and Lambrecht and Polinsky (1996) observe that sentences with the marked pragmatics of sentence focus tend to be formally marked as distinct from canonical predicate focus in motivated ways. \section{Misc >From Gregory Ward, Personal communication 5/11/98: Evaluative N's are ok: good idea! idea! \# square window! wonderful plan! ?a wonderful plan! happy day! ?a happy da Also %\subsection{Bare NPs % %Cross-linguistically, %determinerless nouns are often associated with noun incorporation when they %are in object position. Common kind of noun incorporation in Micronesian and Mayan languages. Noun forms a syntactic unit with V, and noun is not specific, does not refer and is not marked for definiteness or number (Mithun 1984). \subsubsection{Lambrecht and Polinsky Building on work by Lambrecht (1994) and Lambrecht and Polinsky (1996), this paper focuses on the relationship between form and meaning and the notion of paradigmatic contrast. It is argued that the distinction between ``predicate focus" or topic/comment information structure and other types of information structures is represented. Even within English there are various types of SF constructions, each varying in form. While this fact could lead one to the view that the particular form of a construction is unmotivated (cf. Prince 1996), this paper argues on the contrary that more fined grained distinctions motivate the differences in form. Prince suggests that each of these constructions serves to relocate a discourse-new entity from a syntactic position disfavored for discourse-new entities (namely, subject position) and create a separate processing unit for them. This rational in fact presupposes that the relationship between form and meaning is not an arbitrary one in the sense of being unmotivated or senseless. The form of the constructions is strongly motivated by its pragmatic function. $<$ My sister got stabbed in her bed$>$ \\ The landlady, she went up... (Left dislocation) There are a lot of Americans, they can't read. ("Run-on") There are a lot of Americans, who they can't read (resumptive pronoun RC) \section{Attested examples ``How long you been here, anyhow?' ` Little over five weeks'" I said.. pg 60 I'll be back around ten. Little after' pg 103 "' Truth is, I don't really care.'" pg 110 [means "frankly" notice tag would be "now do I?" `You told them about me?' 'Sure. Long time ago." pg 165 ```So I can help Earl crack that wood loose and load it. Stuff's frozen into the ground, most of it." pg 187. (SEPARATE MASS Ns from COUNT??? >From Switchboard corpus: Jury was unorganized. Population is a lot lighter up there. Peer group is the other way now instead of... Third question was how [mumble] Car is going fifty five. Phone's always ringing. Poor kid was in school all day. And once they found you guilty, you know, judge said, he's guilty, kill him. Woman goes in,.. And fact is, we're going to take a vacation this year. Well, thing is, it's a global market place these days. Smith and Wesson revolver isn't goin to do you much good. Uh, well, climate would be, you know... Place is forested with those concrete trees. \section{In a larger context \subsection{Initial Deletion Determiner ommision can be seen as part of a larger phenomena involving the dropping of initial elements (see also Thrasher 1974). That is, Thrasher (1974) suggests relating these cases to the more general phenomena of omitted entitites in sentence initial position. %Bolinger The Imperative in English (no date given): % like it? %% care to come along? % tell the differnce? Gotta go now See you next Tues Too bad about old charlie No need to get upset about it Been in Ann arbor long? Ever et a chance to use your Dogrib? Ever get to Japan, look me up. pg 5 Subjects (I, you, there, it and aux's, if, articles are dropped. \begin{enumerate % ommission from initial position As with other ommissions, only ``redundant" elements can delete (meaning must be recoverable) Contrastive determiners cannot be ommited. For example, \enumsentence{Hey Mike. *(YOUR) Turn now. (Thrasher 1974) grammatical, not lexical, information is deleted Dummy but not anaphoric it and there delete TOo bad you lost More than one way to skin a cat. % must occur in face to face encounter (or as if face to face) Hierarchy: \enumsentence{ aux $>$ dummy subj $>$ if $>$ subj pronouns $>$ articles $>$ possessive pronouns For some speakers, all of the possessive determiner deletion is implossible. ``the variation gives support to Ross' squishiness principle. In fact, there seems to be a complete heirarchy starting with auxilary verbs adn running to possessive pronouns." pg 88 used only in informal contexts ``Rapport deletion" (pg 99): \enumsentence{Got a match? is abrupt, even rude. But ``it is the custom of some people to use such language with strangers.. part of a behavior pattern that superficially treats the whole world as a special friend." pg 99 \end{enumerate \subsection{Discourse Deletion The present cases must be distinguished from ``discourse fragments": certain sentence fragments that require linguistic context to be acceptable (Morgan 1972). Morgan (1972) discourse fragments in 5 minutes: require previous discourse context. Present cases are distinguished from the following in NOT requiring special context: Must be broken Probably hasn't gotten home from work yet Long discussion in chapter 1. Hints at relation of htis type of deletion with that found in newspapers, telegrams, signs. pg 21-23. Main difference: place of deletion is more restricted in present case. T argues for actual deletion because reflexives can occur in fragment. ``the choice has been narrowed to deciding between generating subjectless fragments as part of the grammar of Enlgih or having them generated by a special fragment component tacked on to the end of it." pg 33 %Reasons for creating them as part of grammar: 1) they refer to grammatical elements Not all dummy subjects delete. deletion. Hedges, quantification are suggested to help ``it is also obvious that the totally unexplored are of conversational expectation is involved. ...if your guest commented that he was thirsty or really hot..": (there's) beer in the refridgerator. Avoidance of ambiguity: Deletion is blocked if remaining fragment looks more like a simple NP or an imperative than an S. pg 41: \enumsentence{*(I'll) try to leave as soon as possible Discussion of use of this deletion in actual texts here pg 99--. not clear how controlled it is). (Interesting, %\enumsentence{Use your phone? (from Dragnet from Jerry Morgan) %Jerry morgan: can't answer with ''no you can't" %These cases of the omission of initial elements is termed ``rapport deletion" %because of the informal register it conveys. %Thrasher notes several ways in which the phenomena are alike: %\begin{enumerate % %Initial deletion (``rapport deletion") is used %to indicate an informal register, i.e., less %social distance between speaker and adressee. % %\end{enumerate Thrasher futher includes the following examples: \enumsentence{ (A) Couple of weeks ago I decided to stop smoking However, this example involves a temporal adverb... These are sentence focus cases: subject argument looks less definite because it is not topical!! For some speakers, N needs only specific reference, not definite: Dog kept me up last night (hearer does not need to identify) \section{Lambrecht, Knud and Maria Polinsky. 1997. Typological Variation and Sentence-Focus Constructions. CLS 33. \begin{tabular{lcc Sentence Type & Arg in Focus & Predicate in Focus \\ Argument Focus & + & - \\ Predicate Focus & - & + \\ Sentence Focus + & + \\ \end{tabular SF(thetic) and PF(categorical) are distinguished by TOPICALITY PRESUPPOSITION: assumed status of a referent as a topic of current interest in the discourse. in AF: subj or some other constituent is in focus Why didn't she come to work today? Her SON is responsible (AF) Her son had an ACCIDENT (PF) Her SON is sick (SF) SENTENCE FOCUS CONSTRUCTION: Sentence construction formally marked as expressing a pragmatically structured proposition in which both the subj and predicate are in focus. The focus domain is the sentence minus any topical non-subject arguments. has been called "news sentence" "neutral desctription" "all new utterance" "thetic sentence" "event-reporting sentence" Given an SF construction, three questions arise: questions of FORM, questions of FUNCTION, *the question of the relationship between the two* (what this paper is about). Lambrecht and Polinsky suggest the following principle \begin{quote The Principle of Paradigmatic Contrast: SF constructions have the form they do because they are to be minimally distinct from corresponding PF constructions within a given language. \end{quote SF marking crucially involves DETOPICALIZATION of the nominal constituent involved. How can the subj argument be coded in such a way that it will not be interpreted as a topic? 1) cancel formal properties conventinoally associated with topics 2) code the SF subject with features conventinoally associated with FOCUS arguments (prosodic prominence, specific linear position, non-nominative case marking, lack of grammatical agreement). SF constructions tend to lack a formal opposition betwen subj and obj (tend not to have NP VP structure). They ignore distinction between SF and AF. PF: Sonny Bono DIED. SONNY BONO DIED. (requires iconic focus accent on predicate; allows co-occuring topic activation accent on subj) SF: SONNY BONO died. *TRUMAN DIED (can't have SF interpretation: lack of accent can be explained via reference to contrast with PF) Note examples resist separation of parts: *Black-as-sootbird *JOHNSON, after a short illness, died. [ Ways to neutralize Subj=Obj distinction crosslinguistically: English: prosodic inversion Italian: syntactic inversion: si \`e rotta la MACCHINA \\ ``The CAR broke down" (SF) Ho rotto la MACCHINA \\ ``I broke the car" (PF) Subject in SF in a has position and accent of focal obj in b. Russian allows both syntactic inversion and prosodic inversion: pticy POJUT (PF) \\ birds. NOM.PL sing.PRES.PL \\ ``The birds are SINGING pojut PTICY (SF) \\ sing.PRES.PL birds,NOM.PL \\ ``There are BIRDS singing PTICY pojut (SF) \\ PTICY pojut \\ ``The BIRDS are singing" Detopicalizatoin in SOV lngs: the SF subj may be placed in imediately preverbal (focus) position as in Jap and Kor; or the SF subj may be placed postverbally (Lat, Armenian, Chukchi). VSO: most verb initial languages use preverbal position for topic, so SVO order would have a PF interpretation. But if this is not possible as in Irish, SVO order can serve to mark SF as predicted. If SVO is used for topic comment, then the focus position available for objects is used (e.g. in Malagasy). In some languges, SF subj is morphologically incorporated into the verb (demonstrating the cohesion between subj and v) (Sasse on Boni): \enumsentence{a\'ddige\'ee~juudi \\father-may~died \\ "my father died" Since objects only are usually incorporated, the incorporation of the subject obeys their generalization. Suspended subj-verb agreement: Jespersen ``In Danish the verb was here put in the singular before a plural word...in Eng, there is hte same tendency to use there's before plurals..in Italian, too, one finds v'e\` instead of vi sono..." See chart on page 15 for cross-linguistic differences. Multiple Motivations: why determinor omission to mark SF in this case? Deleting the determiner makes the argument look less independent and therefore less topical. \section{Schmerling, Susan. 1973. Subjectless Sentences and the Notion of Surface structure. CLS 9 577-586 Seems like the class always wakes up five minutes before the bell rings. Guess I should have been more careful Going to lunch? differ from fragments in that they require no preceeding discourse. Perception verbs: Looks like an accident Sounds like another ghost Smells like hot metal Tastes like almonds Feels like real silk. Modals: \enumsentence{Must/might/could/may be an accident up ahead `` Wastebasket class": Happens that way all the time *Happens that way close at ten Turns out you can't do that in Texas Seems you can't do that in Texas. (Almost no analysis here) ``there is some elusive element of spontaneity and implusiveness involved in uttering them." pg 583 Guess I should be going *Guesses he should be going *think I should be going Think I'l have anohter cup of coffee Wish I hadn't done that Gotta be going *Hafta be going. ??Got a lot of nerve. Got a lot of nerve, doesn't he?/I'd say/eh? \section{References Birner, Betty J. 1994. Information Status and word order: An analysis of English Inversion. Language 70 2 . 233-259. Byrne, Bill. 1998. University of California, Dissertation. Gregory, Michelle L. and Laura A. Mich aiUse s. 1999. Topicalization and LeftDislocation: Using Computational Methods to Analyze an Opposition. LSA presentation and handout. Gregory, Michelle L. and Laura A Michaelis. 1999. Topicalization and LeftDislocation: Lambrecht Lambrecht, Knud and Maria Polinsky. 1997. Typological Variation and Sentence-Focus Constructions. CLS 33. Schmerling, Susan. Surface structure. 1973. Subjectless Sentences and the Notion of CLS 9 577-586. Thrasher, Randolph Hallett Jr. 1974. Shouldn't Ignore These Strings: A Study of Conversational Deletion. University of Michigan dissertation. \section{MISC Recognizing the discourse dimension to this construction allows us to explain another seemingly quirky fact about the construction first observed by Thrasher: CHECK JUDGMENTS: *? Boy is at the door. (A) Boy who says he's a student of yours is at the door. (The) Boy who called earlier is at the door. pg 37 [In these cases the modification makes it clear that speaker is INTRODUCING the boy] Modification of N sometimes encourages deletion: Notice further that Who are you rooming with? \\ Boy I went to high school with [ok? Why do you have to go home? Son's sick. seems topical...] Permission statements or reports of permission allow my,your to delete: \enumsentence{(my your *his) wife can go home tomorrow \end{document >From green Mon Sep 8 17:16:41 1997 Received: from bach.cogsci.uiuc.edu.uiuc.edu (bach.cogsci.uiuc.edu) by lees.cogsci.uiuc.edu with SMTP id AA05855 (5.67b/IDA-1.5 for <adele@bach.cogsci.uiuc.edu>); Mon, 8 Sep 1997 17:16:40 0500 From: Georgia Green <green> Message-Id: <199709082216.AA05855@lees.cogsci.uiuc.edu> Subject: Re: car's in the shop To: adele@cogsci.uiuc.edu (Adele Goldberg) Date: Mon, 8 Sep 1997 17:14:56 -0500 (CDT) Cc: green (Georgia Green), morgan (Jerry Morgan) In-Reply-To: <199709082035.AA05285@lees.cogsci.uiuc.edu> from "Adele Goldberg" at Sep 8, 97 03:35:52 pm X-Mailer: ELM [version 2.4 PL25] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1126 Status: RO This is a little vague, but it'll give you an idea of where to look. Susan Schmerling presented a paper on English sentences with finite verb but no subject NP (somebody named Cole from Penn now has a whole diss. on the pragmatics of this) at a CLS around 1971. John Lawler drew attention to the determiner cases, if she didn't cover them also, and may have had a CLS paper himself the following year on that topic. In any case his student Randy Thrasher wrote a dissertation at the University of Michigan (1973) which takes up the issues in detail. As I recall the constraint, determiners, NPs, and auxiliaries can be left off as if serially from the left end of the sentence, subject to the condition that stuff relating to the NP has to refer to the speaker in a declarative sentence and to the addressee in an interrogative sentence. Georgia >From adele Fri Mar 27 12:26:09 1998 Received: from grimm.cogsci.uiuc.edu by lees.cogsci.uiuc.edu with SMTP id AA26993 (5.67b/IDA-1.5 for <adele@cogsci.uiuc.edu>); Fri, 27 Mar 1998 12:26:07 -0600 Date: Fri, 27 Mar 1998 12:26:07 -0600 From: Adele Goldberg <adele> Message-Id: <199803271826.AA26993@lees.cogsci.uiuc.edu> To: adele@cogsci.uiuc.edu, michaeli@spot.colorado.edu Subject: Re: LD Status: RO Oh, thanks, yes hte LD file would be great. Your paper with Michelle sounds good. Isn't LD used to establish a topic. I remember Knud calling it the marked topic construction or something like that. *that, I don't know about it, is a great example! I don't think you can get ommitted determiners is topicalization, insofar as it involves a real topic. If you happen to come across a case, though, please let me know! Prince said something silly like several different constructions have the "same" discourse constraint (to avoid new things in subject position), therefore the mapping of form and meaning is arbitrary. A complete non sequitor, but it was interesting that the three constructions (and this one also) have similar but NOT the SAME discourse constraints. I'm hoping to motivate the form of this along the lines that Knud and Knud and Masha suggested: a difference from PF is indicated by a minimal (and also motivated) difference from standard PF form. Bare NPs are generally used when Ns are associated with the predicate (as in N incorporation), and do not stand as independently referential. It's still kind of vague in my mind, but that's the basic intuition. Hopefully M will be able to procreate with just a little technological help! But are you spending time thinking of possible donors? That is a kind of interesting topic... LOve, Adele >From michaeli@spot.colorado.edu Mon Apr 27 15:23:59 1998 Received: from spot.Colorado.EDU by lees.cogsci.uiuc.edu with SMTP id AA07433 (5.67b/IDA-1.5 for <adele@cogsci.uiuc.edu>); Mon, 27 Apr 1998 15:23:25 -0500 Received: from localhost (michaeli@localhost) by spot.Colorado.EDU (8.8.5/8.8.4/CNS-4.1p) with SMTP id OAA28460 for <adele@cogsci.uiuc.edu>; Mon, 27 Apr 1998 14:19:22 0600 (MDT) Date: Mon, 27 Apr 1998 14:19:22 -0600 (MDT) From: "Laura A. Michaelis" <michaeli@spot.colorado.edu> To: Adele Goldberg <adele@cogsci.uiuc.edu> Subject: hey! Message-Id: <Pine.GSO.3.96.980427134043.420D-100000@spot.colorado.edu> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Status: RO Sorry I wasn't able to call you back this weekend, by coincidence I had been planning to call you, actually. But on Sunday I was hit with a cold and 'female trouble', which is somewhat better today. Did I tell you the good news on on procreation attempts? No, no result yet but M recently was able to have the doctor analyze a sample of his output and the report was very encouraging--he's got something like half the adult population of the US in there. So, the doctor says that he doesn't think that there will be any problem with fertility. We might have to do AI (the other kind), but at least we can use M's chromosomes! On other topics. Things are going well here. There have been many events and parties celebrating my big news, and this Saturday Lise is having a party to honor me and Dan--Dan because he will also stay; Berkeley didn't want him after all! I am pretty optimistic that I will get all my needs met in negotiations, e.g., to come up for tenure in 3 years and even to get money to hire an RA for 2 years. On another front entirely, there is a short discussion in an article by EP on the functions of left dislocation in which she discusses 'initial N'' LDs, like I was raised an old hillbilly and I'll die one. RADIO, it's sitting up there, but I can't hear too good. Don't have a television. (from Terkel) She uses the determinerless leftward N's to argue that 'posset' denoting LD (basically, contrastive LD) differs from other functions of LD syntactically, in that not all functions of LD support this determinerless NP in leftward position. E.g., 'simplifying LD', whose function is to get a discourse-new referent out of a position disfavored for such referents doesn't allow a determinerless Leftward NP: GOOD: Some guy, he objected, so we shut up. BAD: *Guy, he objected, so we shut up. She also cites Ward as discussing determinerless instances with Topicalization, where the example given is 'Cheeseburger it is, then' in answer to a waiter's suggestion, but this example is pretty clearly focus fronting and not topicalization at all! I wish I could give you a fuller reference to this paper, but all I know right now is that it is in a volume edited by Akio Kamio, and that its title is "On the Functions of Left Dislocation in English Discourse'. Did I tell you that my student Michelle and I will give an anti-Prince paper at this discourse conference in Madison in July? We basically argue that the functions Prince gives for LD don't allow us to predict anything about the accessibility status of the leftward NP denotatum, which is strongly upperbounded at the givenness status generally associated with definites. Topicalization, which Prince says is more marked than LD, is actually less marked with regard to constraints on the fronted NP. I think we can present our findings as being compatible with hers, though. The main thing we feel we have to show is that in discourse pragmatics one needs to use a lot of conversational data, and not just pull 20 examples out of one's own conversations with friends. The students are really getting me interested in a lot of stuff that I wouldn't ordinarily take on. The other somewhat brave thing for me is a paper on Basque with my student Phyllis (whom Alan King knew). It's going to be a bit controversial, but the title is already planned: Basque as a Verb-Initial Lg. Sounds like McCawley's whacky paper on English. The main point I want to make is that arguments, in the form of heavily inflected auxs, follow the verb in this language, and anything preverbal is actually a detached, i.e., fronted, topic or focus. I was really influenced by Aissen's paper on topic and focus in Mayan, which has this very behavior. And on yet another front. Bob and Giulia may be coming your way. Bob has a postdoc offer in Psych from Murphy, whose first name I don't know, and Giulia was thinking of possibly transfering there. Would it be OK if she called you to discuss this? If you'd prefer to be called at work, could I give her your work number? Her going there would certainly be our loss and your gain. She is extremely smart, is typologically knowledgeable, knows stats very well, is very big on construction grammar, and is in fact right now running an experiment designed to elicit reading facilitation effects of syntactic priming, to show that we really do have evidence of syntactic priming at the comprehension level. She could work with you or Kay Bock, depending I guess on which department she applies to. Anyway, let me know what you think. (She's also a great Italian consultant.) And on the penultimate front, my mom moved on Saturday, assisted by the very preg Voz and my brother, who flew out. Apparently, her new place is pretty smashing. She seems happy, which is great, although Voz says that the old place will need a lot of work prior to selling. Speaking of which, when will you move into your new house? How is everyone, including a? How's her vocab? Any 3-place predicates yet? Looks like Bowerman's kids had some at around a's age. When will you come to visit? Is Ali still planning the talk at NIST at some point? Hope so, cause I don't know when else I will see you. What are your summer plans? We will be in Germany and Italy (visiting Giulia's parents--the BMW crashers) from mid June to mid July, but otherwise here. Love, L