Syntactic Analysis Quiz 1

some functional (closed) categories of English

1) prepositions (abbreviated P). Prepositions appear before nouns (or more precisely, noun phrases - Prepositions of English (P): to, from, under, over, with, by, at, above, before, after, through, near, on, off, for, in, into, of, during, across, without, since, until. 2) The class of determiners (D) is a little broader. - It contains a number of subcategories including articles, quantifiers, numerals, deictics, and possessive pronouns. - Determiners appear at the very beginning of English noun phrases. - Articles: the, a, an - Deictic articles: this, that, these, those, yon - Quantifiers: every, some, many, most, few, all, each, any, less, fewer, no - (Cardinal) numerals: one, two, three, four, etc. - Possessive pronouns:7my, your, his, her, its, our, their - Some wh-question words: which, whose" 3) Conjunctions (Conj) are words that connect two or more phrases together on an equal level - and, or, nor, neither...nor, either...or 4) The class of complementizers (C) also connects structures together, but they embed one clause inside of another instead of keeping them on an equal level: 5) Tense categories of English (T) - the category T consists of auxiliaries, modals, and the non-finite tense marker. In the older syntactic literature, the category T is sometimes called Infl (inflection) or Aux (Auxiliary). We'll use the more modern T. - Auxiliaries: have/has/had, am/is/are/was/were, do/does/did Modals: will, would, shall, should, can, could, may, might, must Non-finite tense marker: to 6) There is one special category containing only one word: not, which we'll call negation (Neg). There are other categories that express negation (e.g., the determiners no, any, and the noun none). We'll reserve the category Neg for the word not, however.

binding

A binds B if and only if A c-commands B and A and B are coindexed. A is the binder, B is the bindee. - Binding is a special kind of co-indexation. It is co-indexation that happens when one of the two NPs c-commands the other. Notice that co-indexation alone does not constitute binding. Binding requires both co-indexation and c-command. - it turns out that these different semantic types of NPs can only appear in certain syntactic positions, which are defined using the structural relations we developed in the last chapter. - Anaphors, R-expressions, and pronouns can only appear in specific parts of the sentence. - For example, an anaphor may not appear in the subject position of sentence: *Herself bopped Heidi on the head with a zucchini. - The theory of the syntactic restrictions on where these different NP types can appear in a sentence is called binding theory

clauses

A clause consists of a subject NP and a VP. The label we use for clause is TP (which stands for tense phrase) - TPs can also include other items, including (unsurprisingly) elements of the category T, such as modal verbs and auxiliary verbs like those in a) Cedric might crash the longboat. b) Gustaf has crashed the semi-truck. - Note that the T in the TP is optional. TP → NP (T) VP" Clauses don't always have to stand on their own. There are times when one clause is embedded inside another [TP Shawn said [TP he decked the janitor]]. - the clause he decked the janitor lies inside the larger main clause. - Often embedded clauses are introduced by a complementizer like that or if: [TP Shawna said [CP [C that] [TP she decked the janitor]]]. - We need a special rule to introduce complementizers (C): CP → (C) TP - For the moment we will assume that all embedded clauses are CPs, whether or not they have a complementizer. - Embedded clauses appear in a variety of positions. - the embedded clause appears in essentially the same slot as the direct object. - Embedded clauses can also appear in subject position: [TP [CP That she emailed the janitor] worried Amy]. - Because of this we are going to have to modify our TP and VP rules to allow embedded clauses. - Syntacticians use curly brackets {} to indicate a choice. So {NP/CP} means that you are allowed either an NP or a CP but not both. The modification to the TP rule is relatively straightforward. We simply allow the choice between an NP and a CP in the initial NP:

locality constraint

A constraint on the grammar, such that two syntactic entities must be "local" or near to one another - Notice that the difference between a sentence like (19) and a sentence like (12a) is that in the ungrammatical (19) the anaphor is in an embedded clause. - The anaphor seems to need to find its antecedent in the same clause. This is called a locality constraint. - The anaphor's antecedent must be near it or "local" in some way. - The syntactic space in which an anaphor must find its antecedent is called a binding domain. - For the moment let's just assume that the binding domain is the clause (TP).

terminal node

A node that dominates nothing. (A node that is not a mother).

Non-terminal Node

A node that dominates something (A node that is a mother).

R-expression

An NP that gets its meaning by referring to an entity in the world Felicia wrote a fine paper on Zapotec. - If you heard this sentence said in the real world, the speaker is assuming that you know who Felicia is and that there is somebody called Felicia who is contextually relevant. - Although you may not have already known that she wrote a paper on Zapotec, this sentence informs you that there is some paper in the world that Felicia wrote, and it's about Zapotec. - It presupposes that there is a paper in the real world and that this paper is the meaning of the phrase a fine paper on Zapotec. - Both a fine paper on Zapotec and Felicia get their meaning by referring to objects in the world. - This kind of NP is called a referring expression (or R-expression) - The vast majority of NPs are R-expressions. But it is by no means the case that all NPs are R-expressions. Consider the case of the NP herself in the following sentence: Heidi bopped herself on the head with a zucchini. - In this sentence, Heidi is an R-expression and gets its meaning from the context, but herself must refer back to Heidi. - It cannot refer to Arthur, Miriam, or Andrea. - It must get its meaning from a previous word in the sentence (in this case Heidi). - This kind of NP, one that obligatorily gets its meaning from another NP in the sentence, is called an anaphor

C-command

C-command (informal): A node c-commands its sisters and all the daughters (and granddaughters, and great-granddaughters, etc.) of its sisters. C-command (formal): Node A c-commands node B if every node dominating A also dominates B and neither A nor B dominates the other. Symmetric C-command: A symmetrically c-commands B if A c-commands B and B c-commands A. Asymmetric C-command: A asymmetrically c-commands B if A c-commands B but B does not c-command A."

English phrase Structure Rules

CP -> (C) TP TP -> {NP/CP} (T) VP VP -> (AdvP+) V (NP) ({NP/CP})(AdvP+)(PP+)(AdvP+) NP -> (D) (AdjP+) N (PP+)(CP) PP -> P (NP) Adj -> P(AdvP)Adj Adv -> P(AdvP)Adv XP -> XP conj XP X -> X conj X

additional rules

CP → (C) TP TP → NP (T) VP XP → XP Conj XP X' → X' Conj X' X → X Conj X

constituent in terms of nodes

Constituent: A set of terminal nodes exhaustively dominated by a particular node. Constituent of: A is a constituent of B if and only if B dominates A. Immediate Constituent of: A is an immediate constituent of B if and only if B immediately dominates A."

adjectives

Derivational Suffixes: Adjectives often end in derivational endings such as -ing (the dancing cat), -ive (indicative), -able (readable), -al (traditional), -ate (intimate), -ish (childish), - some (tiresome), -(i)an (reptilian), -ful (wishful), -less (selfless), -ly (friendly). Inflectional Suffixes: Adjectives can be inflected into a comparative form using -er (alternately they follow the word more). They can also be inflected into their superlative form using -est (alternately they follow the word most). Adjectives are typically negated using the prefix un- (in its sense meaning "not", not in its sense meaning "undo"). Note that the following affixes have homophonous usage with other parts of speech: -ing, -er, -en, -ed, un-, -ly. Syntactic Distribution: Adjectives can appear between determiners such as the, a, these, etc. and nouns (the big peanut). They also can follow the auxiliary am/is/are/was/were/be/been/being (warning: this distribution overlaps with verbs). Frequently, adjectives can be modified by the adverb very (warning: this distribution overlaps with adverbs)."

nouns

Derivational Suffixes: In English, nouns often end in derivational endings such as -ment (basement), -ness (friendliness), -ity (sincerity), -ty (certainty), -(t)ion (devotion), -ation (expectation), -ist (specialist), -ant (attendant), -ery (shrubbery), -ee (employee), -ship (hardship), -aire (billionaire), -acy (advocacy), -let (piglet), -ling (underling), -hood (neighborhood), -ism (socialism), -ing (fencing). Inflectional Suffixes: Nouns in English don't show much inflection, but when pluralized can take suffixes such as -s (cats), -es (glasses), -en (oxen), -ren (children), -i (cacti), -a (addenda). - Note that the following endings have homophonous usage with other parts of speech: -ing, -s, 's, -er, -en, but also sometimes are found on nouns. Syntactic Distribution: Nouns often appear after determiners such as the, those, these (e.g., these peanuts), and can appear after adjectives (the big peanut). Nouns can also follow prepositions (in school). All of these conditions can happen together (in the big gymnasium). Nouns can appear as the subject of the sentence (we will define subject rigorously in a later chapter): The syntax paper was incomprehensible; or as the direct object: I read the syntaxpaper. Nouns can be negated by no (as opposed to not or un-): No apples were eaten. - One easy way to see if something is a noun is to see if you can replace it with another word that is clearly a noun. So if we want to see if the word people is a noun or not, we can substitute another word we know for sure to be a noun, e.g., John in I saw people running all over the place vs. I saw John running all over the place)."

adverbs

Derivational Suffixes: Many adverbs end in -ly: quickly, frequently, etc. Inflectional Suffixes: Adverbs generally don't take any inflectional suffixes. However, on rare occasions they can be used comparatively and follow the word more: She went more quickly than he did. Adverbs typically don't take the prefix un- unless the adjective they are derived from does first (e.g., unhelpfully from unhelpful, but *unquickly, *unquick). Syntactic Distribution: The syntactic distribution of adverbs is most easily described by stating where they can't appear. Adverbs can't appear between a determiner and a noun (*the quickly fox) or after the verb is and its variants.5 They can really appear pretty much anywhere else in the sentence, although typically they appear at either the beginning or the end of the clause/sentence. Frequently, like adjectives, they can be modified by the adverb very."

verbs

Derivational Suffixes: Verbs often end in derivational endings such as -ate (dissipate), and -ize/-ise (regularize). Inflectional Suffixes: In the past tense, verbs usually take an -ed or -t ending. In the present tense, third person singular (he, she, it), they usually take the -s ending. Verbs can also take an -ing ending in some aspectual constructions, (she was walking) and most take either an -en or an -ed suffix when they are passivized (more on passivization in later chapters): the ice cream was eaten. - Note that the following endings have homophonous usage with other parts of speech: -ing, -s, -en, -ed. Syntactic Distribution: Verbs can follow auxiliaries and modals such as will, have, having, had, has, am, be, been, being, is, are, were, was, would, can, could, shall, should, may, must, and the special non-finite marker to. Verbs follow subjects, and can follow adverbs such as often and frequently. Verbs can be negated with not (as opposed to no and un-4)."

replacement test

First, the smallest constituent is a single word, so it follows that if you can replace a group of words with a single word then we know that group forms a constituent a) The woman from NY flew only ultra-light planes. b) She flew only ultra-light planes. - There is one important caveat to the test of replacement: There are many cases in our rules of optional items (those things marked in parentheses like the AdjP in NP → (D) (AdjP+) N). - When we replace a string of words with a single word, how do we know that we aren't just leaving off the optional items? - To avoid this problem, we have to keep the meaning as closely related to the original as possible. This requires some judgment on your part. None of these tests is absolute or foolproof.

principle of modification (revised)

If a YP modifies some head X, then YP must be a sister to X or a projection of X (i.e., X' or XP).

the principle of modification

If an XP (that is, a phrase with some category X) modifies some head Y, then XP must be a sister to Y (i.e., a daughter of YP). informal: modificeert are always attached within the phrase they modify The adverb very modifies yellow, so it is part of the yellow AdjP by contrast, big doesn't modify yellow, it modifies book, so it is attached directly to the NP containing book.

stand alone test (sometimes called the sentence fragment test)

If the words can stand alone in response to a question, then they probably constitute a constituent. a) Paulette ate at a really fancy restaurant. b) Paulette ate at a really fancy restaurant. If we ask the question "What did Paulette do yesterday afternoon?" we can answer with the italicized group of words in (a), but not in (b): a) Ate at a really fancy restaurant. b) *Ate at. Neither of these responses is proper English in prescriptive terms, but you can easily tell that (a) is better than (b)."

The problem of traditional definitions

If you were taught any grammar in school, you may have been told that a noun is a "person, place, or thing", or that a verb is "an action, state, or state of being". - Alas, this is a very over-simplistic way to characterize various parts of speech. - It also isn't terribly scientific or accurate. - The first thing to notice about definitions like this is that they are based on semantic criteria. - It doesn't take much effort to find counterexamples to these semantic definitions. Consider the following: The destruction of the city bothered the Mongols. - The meaning of destruction is not a "person, place, or thing". It is an action. By semantic criteria, this word should be a verb. - But in fact, native speakers unanimously identify it as a noun. " - It thus seems difficult (if not impossible) to rigorously define the parts of speech based solely on semantic criteria. - This is made even clearer when we see that a word can change its part of speech depending upon where it appears in a sentence

why does binding theory matter to syntacticians

In addition to showing the relevance of structural relations, binding theory is an important tool for syntacticians. By using binding judgments we can actually probe the syntactic structure itself. Because it is defined over c-command, we can determine if two positions in the tree c-command each other. Consider the problem of VSO languages (see for example, the Irish problem set at the end of chapter 4.). These languages don't lend themselves easily to a structure with a VP because the verb and its object are separated by the subject.

generative grammar

In generative grammar, generalizations about structure are represented by rules. - These rules are said to "generate" the tree. So if we draw a tree a particular way, we need rules to generate that tree. - The rules we are going to consider in this chapter are called phrase structure rules (PSRs) because they generate the phrase structure tree of a sentence.

x- bar schema

NP → (D) N' N' → (AdjP) N' or N' (PP) N' → N (PP) VP → V' V' → V' (PP) or V' (AdvP) V' → V (NP) AdvP → Adv' Adv' → (AdvP) Adv' Adv' → Adv (PP) AdjP → Adj' Adj' → (AdvP) Adj' Adj' → Adj (PP) PP → P' P' → P' (PP) or (AdvP) P' P' → P (NP)"

(Proper) Dominance

Node A dominates node B if and only if A is higher up in the tree than B and if you can trace a branch from A to B going only downwards.

exhaustive dominance

Node A exhaustively dominates a set of terminal nodes {B, C, ..., D} provided it dominates all the members of the set (so that there is no member of the set that is not dominated by A) and there is no terminal node G dominated by A that is not a member of the set.

government

Node A governs node B if A c-commands B, and there is no node G such that G is c-commanded by A and G asymmetrically c-commands B. Phrase-government: If A is a phrase, then the categories that count for G in the above definition must also be phrases. - G and M don' count as interveners, even though they both are c-commanded by AP and they both c-command B. This is because they are not phrases- they are heads. Head-government: If A is a head (word), then the categories that count for G in the above definition must also be heads." - similarly, phrasal interveners don't count because they are phrases

Immediate Dominance

Node A immediately dominates node B if there is no intervening node G that is dominated by A, but dominates B. (In other words, A is the first node that dominates B.) - A node immediately dominates another if there is only one branch between them A is the Mother of B if A immediately dominates B. B is the Daughter of A if B is immediately dominated by A. Sisters: Two nodes that share the same mother.

Free

Not bound

The Special Case of Possessive Pronouns

Possessive pronouns are an especially tricky case. - They clearly function semantically like nouns. So for example, Susan's father might be the same person as her father, where her refers to Susan. - In chapter 5, you'll see cases where possessive pronouns behave like pronouns with respect to a phenomenon called binding. - But in other regards, possessive pronouns actually behave more like determiners: they are in complementary distribution with determiners (*the her book). - They appear at the beginning of the noun phrase. This gives possessive pronouns the flavor of a determiner. So are possessive pronouns a subcategory of noun or a subcategory of determiners? That's a really tricky question. - Once you learn about head movement in chapter 10, you might consider an analysis where they start out as nouns and become determiners via the mechanism of head movement. - But for the first part of this book, it's probably easier just to treat them as determiners, because they normally appear in the same syntactic positions as determiners.

The binding Principles

Principle A: An anaphor must be bound in its binding domain. Principle B: A pronoun must be free in its binding domain. - a pronoun cannot be bound by an antecedent that is a clause-mate (in the same immediate clause). You'll notice that this is exactly the opposite of where anaphors are allowed. Principle C: An R-expression must be free. - R- expressions receive their meaning from outside the sentence (i.e., from the context). That they don't get their meaning from another word in the sentence (via binding) is entirely expected.

proper names vs. common nouns

Proper names are nouns like Andrew Carnie. Common nouns are all other nouns. For the most part, proper names resist taking determiners: a) Andrew Carnie b) *the Andrew Carnie - There are some exceptions to this generalization. - For example, when referring to a family it's common to say the Smiths. - In other languages, proper names can take determiners.

do-so replacement

Replacement of a V' with do so. There is a similar process to one-replacement in the syntax of VPs. This is the process of do-so- (or did-so-) replacement. Consider first the VP in the following sentence, which has both an NP and a PP in it. I [eat beans with a fork]. - The rule we developed for VPs in chapter 3 generates a flat tree - In this tree, there is no constituent that groups together the V and NP and excludes the PP. - However, do-so-replacement targets exactly this unit: 16) I [eat beans] with a fork but Janet [does (so)] with a spoon. Let's formalize this rule as: 17) Do-so-replacement: Replace a V' with do so (or do or do so too or do too). For this to work we need the following rules: 18) VP → V' 19) V' → V' (PP) or V' (AdvP) 20) V' → V (NP)"

precedence

Sister Precedence: Node A sister-precedes node B if and only if both are immediately dominated by the same node, and A appears to the left of B. Precedence: Node A precedes node B if and only if neither A dominates B nor B dominates A and A or some node dominating A sister-precedes B or some node dominating B. No Crossing Branches Constraint: If node X precedes another node Y then X and all nodes dominated by X must precede Y and all nodes dominated by Y. Immediate Precedence: A immediately precedes B if there is no node G that follows A but precedes B.

complement

Sister to X, daughter of X'

subject, object, direct object, indirect object, oblique

Subject(preliminary): NP or CP daughter of TP. Object of Preposition(preliminary): NP daughter of PP. Direct Object: a. With verbs of type V[NP__NP], V[NP CP] and V[NP__ NP PP], the NP or CP daughter of VP. b. With verbs of type V[NP NP {NP/CP}], an NP or CP daughter of VP that is preceded by an NP daughter of VP. Indirect Object(preliminary): a. With verbs of type V[NP__ NP PP], the PP daughter of VP immediately preceded by an NP daughter of VP. b. With verbs of type V[NP NP {NP/CP}], the NP daughter of VP immediately preceded by V (i.e., the first NP daughter of VP). Oblique: any NP/PP in the sentence that is not a subject, direct object of a preposition, direct object, or indirect object.

Antecedent

The element that binds a pronoun, anaphor, or R-expression. When this element c-commands another coindexed NP, it is a binder of that NP.

parameterization

The idea that there is a fixed set of possibilities in terms of structure (such as the options in the X-bar framework), and people acquiring a language choose from among those possibilities.

syntactic distribution

The other kind of test we use for determining part of speech uses syntactic distribution. - Syntactic distribution refers to what other words appear near the word. - For example, nouns typically appear after determiners (articles) such as the, although they need not do so to be nouns. We can thus take appearance after the to be a test for noun- hood.

recursion

The possibility of loops in the phrase structure rules that allow infinitely long sentences, and explain the creativity of language - The TP rule has a VP under it. Similarly, the VP rule can take a CP under it, and the CP takes a TP. This means that the three rules can form a loop and repeat endlessly

coindexed

Two NPs that have the same index are said to be coindexed

compliments

We started off with the observation that there seemed to be more structure to our trees than that given by the basic phrase structure rules we developed in chapter 3. - In particular, we introduced the intermediate levels of structure called N', V', Adj', and P'. - The evidence for these comes from standard constituency tests like conjunction, and from processes like one-replacement and do-so-replacement. - We also saw that material on different levels of structure behaved differently. - Complements exhibit one set of behaviors and adjuncts a different set. to account for the data seen above in the introduction, let us revise our NP rules to add the intermediate structure: 9) NP → (D) N' 10) N' → (AdjP) N' or N' (PP) 11) N' → N (PP) These rules introduce a new character to our cast of nodes. This is the N' node. It plays the role of the intermediate constituent replaced by one above. " -Without the intermediate N' node, we would have no way of accounting for one-replacement or conjunction facts. With N', explaining these sentences is easy, since there is more structure in each phrase.

adjunct rule

X' -> X' (ZP) or X' -> (ZP) X' Since the adjunct rule takes an X'-level category and generates another X' category, it will always be higher in the tree than the output of the complement rule (which takes an X' and generates an X). - Since lines can't cross, this means that complements will always be lower in the tree than adjuncts and will always be closer to the head than adjuncts."

complement rule

X' -> X(WP) or X'. -> (WP)X -First observe that because the complement rule introduces the head (X), the complement PP will always be adjacent to the head. Or more particularly, it will always be closer to the head than an adjunct PP will be.

specifier rule

XP -> (YP) X' or XP -> X' (YP)

constituents

a group of words that function together as a unit - we are going to claim that the words are grouped into units (called constituents) and that these constituents are grouped into larger constituents, and so on until you get a sentence. - In order to capture these intuitions (the intuition that certain words are more closely connected than others, and the intuitions about relationships between words in the sentence), we need a more complex notion. - The notions we use to capture these intuitions are constituency and hierarchical structure. - The idea that the and student are closely related to one another is captured by the fact that we treat them as part of a bigger unit that contains them, but not other words. We have two different ways to represent this bigger unit. - One of them is to put square brackets around units: [the student] - The other is to represent the units with a group of lines in what is called a tree structure: - These bigger units are called constituents. - Constituency is the most important and basic notion in syntactic theory. Constituents form the backbone of the rest of this book. - They capture the intuitions mentioned above. - The "relatedness" is captured by membership in a constituent.

branch

a line connecting two parts of a tree

index

a subscript mark that indicates what an NP refers to - We use a special mechanism to indicate that two NPs refer to the same entity. - After each NP we write a subscript letter. If the NPs refer to the same entity, then they get the same letter. If they refer to different entities they get different letters. - Usually we start (as a matter of tradition) with the letter i and work our way down the alphabet. - These subscript letters are called indices or indexes (singular: index).

coordination test

also called conjunction - Coordinate structures are constituents linked by a conjunction like and or or. - Only constituents of the same syntactic category can be conjoined: a) [Anita] and [the man] went to the store. b) *Anita and very blue went to the store. - If you can coordinate a group of words with a similar group of words, then they form a constituent."

part of speech

also known as syntactic category or word class - The most common parts of speech are nouns, verbs, adjectives, adverbs, and prepositions

pronoun

an NP that may (but need not) get its meaning from another NP in the sentence Alvina said that she played basketball. In this sentence, the word she can optionally refer to Alvina (i.e., the sentence can mean "Alvina said that Alvina played basketball") or it can refer to someone else (i.e., "Alvina said that Cassandra played basketball"). Typical pronouns include: he, she, it, I, you, me, we, they, us, him, her, them, his, her, your, my, our, their, and one.

anaphor

an NP that obligatorily gets its meaning from another NP in the sentence Typical anaphors are himself, herself, themselves, myself, yourself, ourselves, yourselves, and each other. There are actually (at least) two different kinds of anaphors. One type is the reflexive pronouns like herself, himself, and themselves. The other kind are called reciprocals, and include words like each other. For our purposes, we'll just treat this group like a single class, although there are minor differences between the distributions of reflexives and reciprocals.

intermediate projection

any projection that is neither the head nor the phrase (i.e., all the X' levels)

hierarchical structure

constituents in a sentence are embedded inside of other constituents - The sentence constituent (represented by the symbol TP) consists of two constituents: a subject noun phrase (NP) [the student] and a predicate phrase or verb phrase (VP) [loved her syntax assignments]. - The subject NP in turn contains a noun (N) student and a determiner (or article) (D) the. - Similarly the VP contains a verb (V), and an object NP [her syntax assignments]. - The object NP is further broken down into three bits: a determiner her, an adjective syntax, and a noun assignments. - As you can see this tree has constituents (each represented by the point where lines come together) that are inside other constituents. - This is hierarchical structure. Hierarchical constituent structure can also be represented with brackets. - Each pair of brackets ([ ]) represents a constituent. - We normally put the label of the constituent on the left member of the pair. - The bracketed diagram [TP[NP [D The] [N student]] [VP [V loved] [NP [D her] [AdjP[Adj syntax]] [N assignments]]]]."

general X-bar-theoretic rules

next, there was the observation that all trees have three levels of structure. They all have specifiers (weak evidence here), adjuncts, and complements. In response to this we proposed the following general X-bar-theoretic rules specifier rule: XP -> (YP) X' or XP -> X'(YP) adjunct rule: X' -> X'(ZP) or X' -> (ZP)X' complement rule: X' -> X(WP) or X' -> (WP) X these rules use variables to capture cross-categorical generalizations. In order to limit the power of these rules, and in order to capture differences between languages, we proposed that the options within these rules were parameterized. Speakers of languages select the appropriate option for their language.

one-replacement

replacement of an N' node with one. Similar evidence comes from conjunction: Calvin is [the [dean of humanities] and [director of social sciences]]. Give me [the [blue book] and [red binder]]. - We need these "intermediate" N' (pronounced "en-bar") categories to explain the items that are conjoined in these sentences." The flat structure seen in (2) is clearly inadequate and a more articulated structure is needed. This chapter is about these articulated trees. The theory that accounts for these is called X-bar theory."

adjunct

sister to X', daughter of X'

specifier

sister to X', daughter of XP

argument structure

subject predicate (traditional definitions) - The definition of subject isn't too bad (we'll refine it later though), but syntacticians use the term "predicate" entirely differently. - The syntactician's definition of predicate is based on the mathematical notion of a "relation". - The predicate defines the relation between the individuals being talked about and the real world - as well as among themselves. - The entities (which can be abstract) participating in the relation are called arguments. To see how this works, look at the following example: "Gwen hit the baseball." - There are two arguments in this example, Gwen and the baseball. - These are elements in the world that are participants in the action described by the sentence. - The predicate here is hit. Hit expresses a relation between the two arguments: more precisely, it indicates that the first argument (Gwen) is applying some force on the second argument (the baseball). - This may seem patently self-evident, but it's important to understand what is going on here on an abstract level. - This usage of the terms predicate and argument is identical to how they are used in formal logic. - We can speak about any particular predicate's argument structure. - This refers to the number of arguments that a particular predicate requires. Another name for argument structure is valency. - Take, for example, predicates that take only one argument (i.e., they have a valency of 1). - These are predicates like smile, arrive, sit, run, etc. - The property of transitivity refers to how many arguments follow the verb. - In predicates with a valency of 1, no arguments follow the verb (the single argument precedes the verb), so these predicates are said to be intransitive. - Predicates that take two obligatory arguments have a valency of 2; some examples are hit, love, see, kiss, admire, etc. These predicates are said to be transitive, because they have a single argument after the verb (the other argument precedes the verb). - Finally predicates that take three arguments have a valency of 3. Put and give are the best examples of this class. These predicates have two arguments after the verb so are said to be ditransitive. - In determining how many arguments a predicate has, we only consider the obligatory NPs and PPs. Optional ones are never counted in the list of arguments. Only obligatory elements are considered arguments. - Predicates not only restrict the number of arguments that appear with them, they also restrict the categories of those arguments. - A verb like ask can take either an NP or a clause (embedded sentence = CP) as a complement I asked [NP the question]. I asked [CP if you knew the answer]. - But a verb like hit can only take an NP complement: I hit [NP the ball]. *I hit [CP that you knew the answer]."

constituency tests

tests that show that a group of words functions as a unit. There are four major constituency tests given here: movement, coordination, stand-alone, and replacement.

binding domain

the clause containing the NP (for our purposes) The definition we've given here for "binding domain" is clearly over-simplistic. For example, when there is an NP that contains an anaphor and an NP marked with's, that NP seems to function as a binding domain: i) Heidii believes any description of herselfi. ii) *Heidii believes Marthaj's description of herselfi. iii) Heidii believes Marthaj's description of herselfj. The literature on this is extensive and beyond the scope of this chapter. But you should be aware that the definition given here needs extensive revision; we will return to this in chapter 17.

node

the end of a branch A node with two or more branches below it is said to be branching; a node that has a single branch below it is said to be non-branching.

label

the name given to a node (e.g., N, NP, TP, etc)

root node

the node that dominates everything, but is dominated by nothing. (The node that is no node's daughter).

projection

the string of elements associated with a head that bear the same category as the head (N, N', N', N', NP, etc)

maximal projection

the topmost projection in a phrase (XP)

Head

the word that gives its category to the phrase

head

the word that gives its category to the phrase the head of the NP is the N, the head of a PP is the P, the head of the AdjP is Adj and the head of an AdvP is Adv.

syntactic trees and bracketed diagrams

these are means of representing constituency. They are generated by rules.

corefer

two NPs that are coindexed are said to corefer (refer to the same entity in the world)

are numerals of category D or adj?

- Cardinal numerals, in phrases like two books, seem to function like quantifiers like all or few at least in terms of their function as counting elements. In the case of one big book, the one can stand in place of an obligatory determiner (normally we'd require a determiner like a or the -- leaving off the determiner makes the phrase ungrammatical (*big book). This suggests, then, that numerals are of category D. - However, consider the case of these two books. In this phrase, the numeral functions more like an adjective, in that it appears between the deictic article these and the noun. - What category are numerals? That's a difficult question to answer, and one that requires more theoretical tools than I can give you in this book, at this stage in your learning at least. I tell my own students that for the purposes of doing assignments in this book either analysis is plausible, and possibly both situations exist.

a common mistake in binding

- Consider the sentence *Shei loves Maria. Which of the two NPs in this sentence is the antecedent? - Common sense might tell us that Maria is. But common sense is wrong. The antecedent here is she. This is because she c-commands Maria, and not vice versa. - One easy way to avoid this mistake is not to think in terms of antecedent and anaphor/pronoun, but in terms of binder and bindee. - The binder here is she because it is coindexed with Maria and c-commands Maria. - Maria is the thing being bound (the bindee). - Note that binding is typically an asymmetric relationship." *Herselfi bopped Heidii on the head with a zucchini. Even though the two NPs are coindexed, they do not form a binding relation, since the antecedent doesn't c-command the anaphor. You might think that Heidi binds herself, since the anaphor c-commands the antecedent.7 But notice that this is not the way binding is defined. Binding is not a symmetric relationship. The binder (or antecedent) must do the c-commanding of the bindee (anaphor or pronoun), not the reverse.

count vs. mass nouns

- Count nouns represent individual, "countable" elements. - "Mass nouns" usually can't be counted in the same way. For example sincerity and air are mass nouns. - There are two easy distributional tests to distinguish between mass and count nouns. Mass nouns take the quantifier much, while count nouns take many. - like plurals, mass nouns generally don't require a determiner but count nouns do ex: *I ate apple. I ate the apple.

lexical vs functional

- Lexical parts of speech provide the "content" of the sentence. - Nouns, verbs, adjectives, and adverbs are all lexical parts of speech. - Functional parts of speech, by contrast, provide the grammatical information. - Functional items are the "glue" that holds a sentence together. - One way to tell if a lexical item is functional or lexical is to see if it is left behind in "telegraphic speech" (that is, the way a telegram would be written; e.g., Brian bring computer! Disaster looms!). - Functional categories include determiners, prepositions, complementizers, conjunctions, negation, auxiliaries, and modals.

movement test

- Movement is our third test of constituency. - If you can move a group of words around in the sentence, then they form a constituent - because you can move them as a unit. 1) Clefting involves putting a string of words between It was (or It is) and a "that" at the beginning of the sentence. It was [a brand new car] that she bought. (from She bought a brand new car) 2) Preposing (also called pseudoclefting) involves putting the string of words before a "is/are what" or "is/are who" at the front of the sentence. [Big bowls of beans] are what I like. (from I like big bowls of beans 3) Passive involves putting the object in the subject position, the subject in a "by phrase" (after the word by) and changing the verb form (for example from kiss to was kissed). [The big boy] was kissed by [the slobbering dog]. (from The slobbering dog kissed the big boy) Again, the movement test is only reliable when you keep the meaning roughly the same as in the original sentence.

endocentricity generalization

- Next we observed that our rules were failing to capture several generalizations about the data. - First was the endocentricity generalization: all NPs have an N head, all AdjPs an Adj head, etc. There is no rule like *NP V Adj. "

Open vs. Closed Parts of Speech

- Some parts of speech allow you to add neologisms (new words). - New words may be coined at any time, if they are open class (e.g., fax, internet, grody). By contrast there are some parts of speech that don't allow new forms. - Parts of speech that allow new members are said to be open class. Those that don't (or where coinages are very rare) are closed class. - All of the cases that we've looked at so far have been open class parts of speech.

distributional criteria

- The criteria we use for determining part of speech then aren't based on the meanings of the word, but on its distribution. We will use two kinds of distributional tests for determining part of speech: morphological distribution and syntactic distribution.

subcategories of verbs

- There are really two major ways in which we can divide up verbs into subcategories. - One is along the lines of tense/finiteness (i.e., whether the verb is left, leaves, (will) leave or (to) leave). - The other way to divvy up verbs is in terms of the number of noun phrases (NPs) and prepositional phrases (PPs) or clauses (CPs) they require. This second kind of division is known as argument structure.

when constituency tests fail

- Unfortunately, sometimes it is the case that constituency tests give false results (which is one of the reasons we haven't spent much time on them in this text). - Consider the case of the subject of a sentence and its verb. These do not form a constituent. - However, under certain circumstances you can conjoin a subject and verb to the exclusion of the object: i) Bruce loved and Kelly hated phonology class. - Sentence (i) seems to indicate that the verb and subject form a constituent, which they don't. - As you will see in later chapters, it turns out that things can move around in sentences or be deleted. - This means that sometimes the constituency is obscured by other factors. - For this reason, to be sure that a test is working correctly you have to apply more than one test to a given structure. Always perform at least two different tests to check constituency, as one alone may give you a false result.

coordination

- combines two adjectives into a single modifier, combines two NPs, combines two VPs, two sentences and two PPs. - Coordination seems to be able to join together two identical categories and create a new identical category out of them.

Adjectives and Adverbs: Part of the Same Category?

- the only major distinction between them is syntactic: Adjectives appear inside NPs, while adverbs appear elsewhere. This kind of phenomenon is called complementary distribution. (Where you get an adjective vs. an adverb is entirely predictable.) - When two elements are in complementary distribution in linguistics, we normally think of them as variants of the same basic category. - We might extend this analysis to parts of speech: There is one "supercategory" labeled "A" that has two subcategories in it (allo-parts-of-speech, if you will): Adj and Adv. - In this book we'll stick with the traditional Adj and Adv categories, simply because they are familiar to most people. - But you should keep in mind that the category A (including both adjectives and adverbs) might provide a better analysis and might be better motivated scientifically.

morphological distribution

- this refers to the kinds of affixes (prefixes and suffixes) and other morphology that appear on a word. Let's consider two different types of affixes. - First, we have affixes that make words out of other words. We call these affixes derivational morphemes. + These suffixes usually result in a different part of speech from the word they attach to. + For example, if we take the word distribute we can add the derivational suffix -(t)ion and we get the noun distribution. The -(t)ion affix thus creates nouns. Any word ending in -(t)ion is a noun. This is an example of a morphological distribution. + A similar example is found with the affix -al, which creates adjectives. If we take distribution, and add -al to it, we get the adjective distributional. + The -al ending is a test for being an adjective. Derivational affixes make a word a particular category; by contrast inflectional morphemes don't make a word into a particular category, but instead only attach to certain categories. - Take for example the superlative suffix -est. This affix only attaches to words that are already adjectives: big, biggest (cf. dog, *doggest). - Because they are sensitive to what category they attach to, inflectional suffixes can also serve as a test for determining part of speech category.