Linguistics 322

Intermediate Syntax

Preposing is a syntactic process in which a constituent moves to the beginning of a sentence. A preposing operator is what causes a constituent to move to the beginning of the sentence or the beginning of certain embedded clauses:

(1)   The first movie I don't like at all.
(2)   Beans Mary likes, peas John likes.

Preposing is discourse dependent or pragmatically dependent. For example (1) could occur in the following discourse:

(3)   A.   Here is a listing of two movies. Do you like either of them?
     B.   The first movie I don't like at all.  

The use of the definite determiner marks the discourse property. It forces the references to one of the two movies. The adjective first determines which of the two movies A dislikes. Note that an observer of this short dialogue might not know which the first movie is, but presumably this information is known to the speaker :


The is linked to them in preceding sentence. Them is linked to its antecedent movies in the sentence that precedes it.

In a pragmatic context with no previous discourse, A and B could see a marquis listing a double feature. Knowing that A also sees the marquis, B could utter (1). But B could not enter into a situation where the two movies are not somehow given knowledge and utter (1). For example, suppose B is a professor and he walks into the classroom. (1) would be an inappropriate initial utterance, except in a very unusual situation where the professor had promised his class that he would go see two movies that were shown is a specific sequence. This, of course, would be a pragmatic situation.

Let us now look at the syntax of (1). The basic proposition without any modifiers is

(5) .   [I like movie].

Let us leave lexical modifiers aside for the moment. The basic proposition occurs within the scope of tense:

(6).   [ -Past [I like movie ]].

(6) occurs in the scope of negation:

(7)   [ Neg [ -Past [ I like movie ]]].

And (7) occurs within the scope of the declarative mood:

(8)   [ Decl [ Neg [ -Past [ I like movie ]]]].

And (8) occurs within the scope of focus:

(9)   [ Focus [ Decl [ Neg [ -Past [ I like movie ]]]]].

There are different kinds of focus and focus can be implemented in different ways. Here we will restrict our discussion to focus marked by preposing. In (4) the speaker chooses to focus on the topic of the minidiscourse. He answers to question by bring which movie he does like into focus by preposing it.

Now, let us suppose that there is a syntactic constituent called FP (focus phrase). This a constituent which is not generally accepted by syntacticians. Focus indicates the speakers intention to a constituent in the preceding discourse. The speaker must target the focus. He is targeting movie. Let us say that the feature [+focus] lowers the targeted constituent: movie. In the syntax we will represent focus as F (FP), Decl as C (CP) (following current usage), Neg as NegP (growing acceptance), T (TP) often used in place of I (IP), and V (VP) (standard).

Targeting a position occurs in the presyntax, not in the syntax proper. Targeting holds between semantics and syntax in what we will call the propositional component. The green line represents targeting. If no other targeting takes place, the result is a syntactic structure. The level of the syntactic structure is called D-structure (formerly deep structure):


[+Foc] is an operator. The term operator is borrowed from propositional logical. In syntax it refers to a predicate that ha referential meaning, but it has no sense (in the Fregean meaning of the term). It is well known that operators tend to raise in syntax. The reason, which we propose here, is that once an operator is lowered prior to D-structure, it must raise back to its original position in order to satisfy its original function: here [+Foc] must take scope over CP. Thus, [+Foc] must raise back to F to satisfy this requirement. However, [+Foc] is now attached to NP--it may be adjoined to NP or it may be embedded in the structure of NP. Once this happens, the entire constituent must raise, not just the feature. This is a syntactic property:


Once, all raising has occurred, the resulting structure is called an S-structure. Note that the subject of the sentence originates as part of VP--adjoined to VP. This corresponds to some degree with the basic proposition. The basic proposition is akin to a small clause (see small.clauses). The NP must raise and adjoin to TP (see raising and Case).

The raising of a node leaves an empty copy of the node. The trace behaves like an anaphor in that its antecedent, the moved item, must c-command its trace--in the above structure movie c-commands its trace.

The lowering of [+Foc] and other features does not leave a formal trace, since the lowered position cannot c-command the upper position from which it started. Trace is a syntactic realization, not a propositional one.

We omitted the modifiers in the above structure to keep it simple. Movie is modified by the and first::

(12).   [I like [the first movie]].

The VP is modified by the PP at all:

(13).   [[I like [the first movie]] at all].

Note that at all may only occur in the scope of a negative.

(14) *I liked the movie at all.
(15) I didn't like the movie at all.

Negation licenses at all and other similar forms that require the presence of the negative.

The dominant VP occurs as a complement of tense [-Past] (see tense), and TP occurs as a complement of Neg (see negation). NegP in turn occurs as a complement of C (C=mood, see mood), and CP, here, occurs as a complement of F. The following structure includes the modifiers; The processes involving negation, tense, and the subject NP are now included here, in order to focus on focus movement:

A link is relation that holds between two nodes that share some property. In the above structure we find a movement link binding the first movie and the trace of it. The movement arrows in the above diagrams can be though of as a link. Each link as a head and a tail. The arrow represents the head of the link. A chain is one or more links that link nodes sharing a common property. In the first diagram (10), there is one link. This link is a complete chain in the structure. In diagram (11), there are two links. The two links form a chain. In this case, the head of the chain is the same node as the tail of the chain. In this case we have a closed chain. If the head and the tail of the chain are different nodes, then the chain is open. Syntactic operators must be closed to be properly formed.



To return to course outline Click here.