Movement vs Parasitic Gaps:
Beatrice Santorini's introduction to syntax (above) assumes the idea of syntactic movement. a common idea in theoretical syntax is that constituents can only move to positions in the tree that "c-command" their trace (the place they moved from). movement should also be motivated both by a reason for the phrase to be in the moved position and also for why it started in its original position.
the c-command relation is also theorised to explain the grammaticality of certain "licensing" structures, for example negative words licensing NPIs or certain reflexive pronouns, the prepositional phrase "at home" can appear at the beginning (topicalised) or at the end (as a VP adjunct) of a sentence:
- I do not sleep at home
- at home I do not sleep
but the NPI "at all" necessarily occurs at the end of sentences where it can be c-commanded by its licenser "not":
- I do not sleep at all
- *at all I do not sleep
- *I sleep at all (at all lacks a licenser)
another example involves the licensing of "each other" by the DP "two men"
- two men were accused by the DA during each other's trial
- *the DA accused two men during each other's trials
the passive sentence is grammatical because "two men" c-commands "each other", but in the active sentence "two men" is too low in the tree to license "each other"
now to introduce the concept of parasitic gaps!
- [what book] did you review _ without actually reading _
this sentence from wikipedia seems to involve "what book" moving to the front of the sentence from two distinct locations. this should be impossible in a movement based analysis. But, if you instead take the view that "what book" is actually merged into the tree at the same place that it is pronounced, then you could simply explain it as "what book" licensing two coreferrential traces/gaps, just as a negative word is perfectly capable of licensing two NPIs as in "I don't find joy in [anything] [any more]"
Topic and Focus
Topic and focus are two parasemantic features (likely common to all languages). The meanings of the two can be seen more clearly in languages like japanese which have specific topic and focus particles, the topic particle "は" and the focus particles "も" (non-unique) and "が" (unique) (which is distinct from the nominative "が"). It is strongly preferrenced in languages to put topics early and focuses late in the linear order of a sentence. A surprising generalisation can be made that, at least most of the time, specifiers are topics and complements/adjuncts are focuses.
a topic is something that the listener is expected to know, it is a backgrounded element that sets the context for an utterance. The focus of a sentence is something that you assert in an utterance. when an element is focused, it's alternatives are brought to mind. Topics and Focuses inherently oppose one another, if an element of a sentence is topicalised, the rest of it is necessarily in focus, and vice versa.
arguments of verbs can be topicalised in english by moving them to the front of the sentence:
- I saw a cat last thursday - weakly topicalised "I" and weakly focused "last thursday", normal sentence
- last thursday, I saw a cat - "last thursday" topicalised, this sentence could occur in a conversation about last thursday, "last thursday" is known, but the seeing of the cat is pushed into focus by the topicalisation, it is asserted, it was unknown, and it contrasts with other things that could've/didn't happen last thursday
there are three ways of focusing in english that I will discuss:
1. Stress; you can stress a word to contrast it with alternatives:
- he will drink the coffee - someone else is drinking it
- he will drink the coffee - stressing an auxiliary (or "not") causes verum/polarity focus*
- he will drink the coffee - he isn't doing something else with the coffee
- he will drink the coffee - he's not drinking something else
*in many other languages this is expressed by stressing the lexical verb, leading to ambiguity (ily do-support)
2. Clefts; a specific weird structure in english (which also exists identically in chinese, but looks simpler given it allowing null subjects, lacking complementisers and WH-movement, and having a lexical copula)
- it's him who will drink the coffee
- (verum focus is not possible like this) (though maybe "??it is that he will drink the coffee" is comparable?)
- ??it's drinking he will do to the coffee - kinda weird, clefting is usually limited to arguments of the verb
- it's the coffee that he will drink
3. Focus particles/adverbs; adverbs, such as 'only' or 'also', can be used before a focused argument, or lexical verb to convey a specific type of focus to an argument of the verb:
- I eat [only bread] in france - "bread" in focus, if you know that I am eating something in france, then you know that that thing is bread
- I eat bread [only in france] - "in france" in focus, if you know that I am eating bread, then you know that I am in france
- I only eat bread in france - ambiguous in meaning between the previous two sentences, stressing either "bread" or "france" puts that one in focus and makes it unambiguous
'only' is a unique focus adverb, but 'also' is a non-unique one, similar examples (with non-unique meanings) can be created by replacing 'only' with 'also'
topic and focus has a similar impact to definiteness. Definite nouns (marked with "the") are nouns that the speaker of an unterance expects the listener to know and indefinite nouns (marked with "a") the opposite is true for. it is obviously more wierd and complicated than this, but this is broadly correct. in the following examples, it can be seen that the indefinite "a cat" can sit in either a later focus position, using the expletive "there" to appease english's no null subject constraint, or it can be raised to the topic position of most english subjects. However, the definite noun "the cat" is forced to be in the topic position due to it's definiteness:
- A cat is here
- There's a cat here
- The cat is here
- *There's the cat here
there is also a connection between topic and focus, and restrictive and nonrestrictive adjunction: if a group of travellers all took the correct path and got where they needed to go then you can say (1) "the travellers, who took the correct path, got where they needed to go", but if some travellers took a different path and thus did not make it, then you would say (2) "the travellers who took the correct path got where they needed to go". sentence (1) adds a set of travellers to the discourse and then asserts two things about them, that they took the correct path, and that they got where they needed to go. sentence (2) however, adds a larger set of travellers to the discourse, and then intersects that set with those who took the correct path, and then makes only one assertion about that restricted set, that they got where they needed to go. The difference is that the relative clause "who took the correct path" is in focus in (1), but is a topic in (2).
some languages like french, spanish, portuguese, and romanian can, on occasion, distinguish between restrictive and non-restrictive adjectives, but not in the way one would expect, given this analysis. restrictive (topic) adjectives will regularly follow the noun, whereas nonrestrictive (focus) adjectives will precede.
english clefts and the 是⋯的结构
When a sentence has a lexical verb in english, the verb can take certain adverbs as adjuncts before it:
- I will [VP definitely [V help] you]
- That is not [VP only [V doing] what you want]
- Beatrice has [VP effectively [V taught] you syntax]
but the word "is" works differently, it can stand seemingly on its own as the only verb in a clause, but acts like an auxiliary in that it undergoes subjects-auxiliary inversion in questions and adverbs are included afterwards. This placement of adverbs suggests that we would need another projection in english copular clauses like: "I am [VP also [V _] your friend]" where an unpronounced lexical copula is included to take the adverb "also" as an adjunct. One could imagine that this is the starting position of the copula before undergoing V to I movement, thought this would then be the only verb to do so in english. Analysing this without movement is also strange as there seems to be no reason why two projections are needed to perform copulation.
the sentence "the fact that he slept is a good thing" can also be expressed more simply as "that he slept is a good thing" which is perhaps a little unnatural and better expressed as "it is a good thing that he slept". I will analyse this sentence as such: 
([IP [DP^ it] [I' [I is] [VP [V' [V 0] [DP^ a good thing]] [CP^ that he slept]]]]) this is perhaps what the sentence "that he slept is a good thing" looked like before the subject of the VP raised to spec IP if you assume that verbal subjects begin as adjuncts. It is also possible that this tree started as "it is [VP that he slept [V _] a good thing] before the CP undergoes rightward movement due to its syntactic weight. This structure is how I analyse cleft sentences mentioned in the previous section
onto chinese. there are a variety of similar structures in chinese. 是 is the copula, but is lexical and acts just like other verbs. 的 is perhaps best analysed as a complementiser, and is almost always optional. Sentences such as "是[DP 咖啡][CP [IP 他喝][C（的）]]" and "[CP 他喝的]是[DP 咖啡]" appear similar in structure to english cleft sentences "it is coffee that he drank" and "what he drank was coffee" but there are sentences such as "我们是昨天到的" which are more confusing, but which I will analyse as clefts that have had the subject of the imbedded clause raised to the subject of the main clause; "[CP [DP 我们i] [C' [VP [V' [V 是] [XP 昨天]] [CP ti 到的]]]] [C _]" (one would need to assume that subjects sit in spec CP in chinese, both because the main verb 是 already has a subject in this sentence and likely could not take another, and because I am suggesting that subjects can be raised out of CPs which would requre them to be in their specifier position in order to be accessable)
