Syntax
Syntax and Grammar (Mair and Leech)
3. Sentences and Their Parts (Leech)
3.1 Prologue: parts of speech
It can be difficult to define what a noun, verb, adjective and adverb are. For example, it is difficult to see anything in common between all the ‘things’ to which nouns can refer, except the fact that nouns can refer to them. However, it is true that the most typical or central members of the class refer to people, things and substances, these are often called concrete nouns.
Similarly with verbs, the definition of a ‘doing word’ applies naturally to went, looked, won, give, but does not so easily apply to are, hates, understand. We could say that a verb can denote ‘states’ as well as ‘actions’, but difficulty is partly that words like state and action are themselves vague in meaning.
Such definitions (e.g. ‘a noun is a naming word’) fail to keep the parts of speech apart, e.g. compare hates in Dracula hates werewolves with hatred in Dracula’s hatred of werewolves. One’s status as a noun and the other’s as a verb cannot be due to the meaning of these words.
We cannot always rely on meaning in defining word classes. This is evidenced by Carroll’s poem Jabberwocky in which we can tell the word class of the nonsense words even though we do not know their meaning (’Twas brillig, and the slithy toves, Did gyre and gimble in the wabe…)
We classify each word on the basis of its form and its position.
Semantic definitions are, then, fallible and dispensable - but not useless. Defining grammatical terms is like defining many other words; we can easily identify features of the most typical members of a class, e.g. a ‘prototype’ bird vs. a penguin which is less ‘birdy’. The same idea of prototype categories with fuzzy edges applies to grammar. The typical nouns are those which refer to people, animals and things.
This concept of a ‘fuzzy’ category applies not just to meaning, but also to formal aspects of definition, e.g. a typical noun has a plural in -s and a typical verb has a past tense in -ed, but there are exceptions to these rules (e.g. sunshine, oxen, won, ate).
3.2 The hierarchy of units
The sentence is the largest unit of language that we are concerned within grammar. A sentences is composed of smaller units, clauses, phrases, words.
Clauses are the major units of which sentences are composed; a sentence may consist of one or more clauses. Jack Sprat could eat no fat is a sentence, but can occur as a clause in a larger unit, e.g. [Jack Sprat could eat no fat] and his wife could eat no lean.
Phrases are units intermediate between clause and word: (My Uncle Olaf) (was munching) (his peach) (with relish). Like words, phrases belong to a number of different classes; (My Uncle Olaf) is a noun phrase, (was munching) is a verb phrase, (his peach) is another noun phrase and (with relish) is a prepositional phrase.
It is very important to notice that we use ‘high’ and ‘low’ in a special way in this hierarchy. It means that a unit of the higher rank consists of one or more of the units of the next lower rank.
3.3 Grammatical notations
For both clarity and brevity, we need a way of representing grammatical structure on paper. There are two different graphic notations: bracketing and tree diagrams.
We have already used a simple set of bracketing conventions. Sentences are marked with an initial capital letter and a final full stop. Clauses are enclosed in square brackets [ ]. Phrases are enclosed in round brackets ( ). Words are separated by spaces. It we need to separate the morphemes in words, we use a hyphen.
e.g. [(Our land-lady) (keep-s) (a stuff-ed moose) (in her attic)].
Bracketing is easy to use, but does not give a very clear visual picture of the relation between constituents. For this, when we want to, we can replace the brackets by a tree diagram.
The symbols Cl (clause), etc. are here used as labels for nodes on the tree so that all units of the same rank appear at the same level of the tree. Each branch of the tree represents a relation of ‘containing’.
3.4 Using tests
Expansion tests: In They’re playing we should expand ‘re into are, and so make it clear that it is a separate verb, belong to the phrase ‘re playing, rather than They. We can also add other words to show that the word is acting as a phrase, e.g. (Uncle Olaf) (has munched) (his peaches) (very contentedly).
Substitution tests: Sometimes by substituting a word sequence for a word we can see that the word is actually behaving as a phrase, e.g. They in They are playing can be replaced with Their team, proving that it is a phrase.
Subtraction tests: The opposite of an expansion test is a subtraction test, i.e. omitting some part of a construction, e.g. we could recognize Dear old Uncle Olaf as a noun phrase because we could subtract the initial three words and end up with a noun.
Movement tests: In Olaf savagely devoured his sixth peach, savagely is treated as a separate phrase rather than as part of a phrase savagely devoured because the word can be moved elsewhere in the clause without noticeably changing its meaning or function in the clause.
Since grammatical categories have fuzzy edges, one test is rarely enough; we often have to rely on a number of different tests in deciding which analysis is the correct or best one.
3.5 Form and function
Form classes: Phrases are subdivided into phrase classes such as noun phrase, verb phrase, adjective phrase, adverb phrase and prepositional phrase. One reason why we need to identify these is to explain the order in which elements of the sentence occur
Function classes: elements of the clause
Classes such as NP, VP, AjP are called form classes because the classification of phrases in this way depends on how the unit is composed of smaller units, or on how its form can vary. But it is also necessary to classify units into function classes (classify them according to how they are used to form larger units - this determines what positions it can fill and whether it is optional).
To exemplify the idea of function, we can limit our attention at this point to three elements, S, O, and P. We can think of them as ‘slots’ in the clause pattern.
P is the only element of a clause which is a verb phrase.
In English S typically comes before P, whereas O typically comes after P.
S typically denotes the actor of the action represented by P, whereas O typically denotes the receiver of the action.
S must normally be present whereas O is often not needed.
Function classes: elements of the phrase
Two main function of words in phrases: head (H) and modifier (M). In general, the head is the word which cannot be omitted from the phrases, whereas modifiers are optional. In verb phrases, the relation between the constituents is different and instead we use the terms auxiliary verb and main verb.
5. Phrases (Leech)
5.1 Classes of phrase
We will recognize six classes of phrase; of these, noun phrases (NP), adjective phrases (AjP) and adverb phrases (AvP) all have the same basic structure: ({M*} H {M*}).
The asterisk means that there can be one or more than one modifier (M), these phrases must have a head (H) but the modifiers are optional ({ }).
Pre-modifiers precede the head, while post-modifiers follow the head.
Prepositional phrases (PPs) and genitive phrases (GPs) can be though of as NPs with an extra particle or marker added to them.
PP: the bride (of (M) the (H) heir (MAj) apparent)
GP: ((M) the (H) heir (Maj) apparent-’s) bride
The difference is that the preposition is added to the front of the PP, whereas the genitive marker (’s) is added to the end of the GP.
The verb phrase (VP) is peculiar to itself, having a rather different structure from those of other phrases - this is explored in chapter 6.
5.2 Main and subordinate phrases
A main phrase is one which is a direct constituent of a clause while a subordinate phrase is one that is part of another phrase.
Subordinate phrases
We have to allow the possibility that units are not merely divisible into units of the next lower rank, but can contain as their elements units of the same, or even of a higher rank - this is the phenomenon of subordination and it allows us to make complex sentences.
e.g. (the bride (of the heir apparent)), where we have two sets of round brackets, one within the other, the inner brackets enclose a subordinate phrase.
Main phrase: […(Ph)…], Subordinate phrase: (…(Ph)…)
Subordinate phrases will always be directly or indirectly a part of a main phrase.
Once we have allowed the possibility of phrases within phrases, this can go on indefinitely:
(his reply (to my review (of his book (on gastronomy (in the Dark Ages))))).
Subordinate clauses
Subordinate clauses work on the same principle, this time subordination is indicated by a nesting of square brackets:
[Joel thinks [that Anna thinks [that he thinks [that she loves him]]]].
Taking the matter one stage further, we have to allow for the possibility of the subordination of one unit (say a clause) within a unit of lower rank.
(the rat [that ate (the malt [that lay (in the house [that Jack built])])])
The clause that Jack built here is called a relative clause and is part of an NP.
5.3 Noun phrases and related phrase classes
Like words, phrases can be classified partly by their external function and partly by their internal form; by form we mean the way the structure of the unit is made up of other, smaller constituents.
The noun phrase (NP)
(a) Function: In the clause, NPs are important and versatile elements: they can act as subject (S), as object (O), or as complement (C), some kinds of NPs can even act as adverbials (A).
Subordinate NPs act as modifiers in other NPs: (my friend (the professional driver)) - this construction in which one NP is defined by another is called apposition.
(b) Structure: The structures of NPs are diverse but the chief elements are:
i. The head of an NP can be: a noun (the doll), a pronoun (it), or less usually an adjective (the absurd), an enumerator (all fifteen) or a genitive phrase (John’s), in which case there is usually a noun which semantically is understood to be the head, e.g. the greedy (people), all fifteen (people), John’s (house)
ii. The pre-modifiers of an NP can be: determiners (this morning), enumerators (two eggs), adjective (red shoes), nouns (London pubs), genitive phrases (someone else’s problems), adverbs (quite a noise), and some others.
iii. The postmodifiers of an NP can be: prepositional phrases (the best day of my life), relative clauses (a quality that I admire), various other types of modifier, including adverbs (the girl upstairs), adjectives (something nasty), noun phrases in apposition (the bandicoot, a tiny marsupial), etc. In postmodification there is no limit to the length of NPs.
Pronouns (pn) and determiners (d)
Pronouns and determiners are two closed word classes in the NP which have similar subdivisions
i. Pronouns function as H; there are personal pronouns (I, you, he), possessive pronouns (mine, ours, yours), demonstrative pronouns (this, that), and quantifier pronouns (general (all, some, any), compound (anyone, no one, everything), gradable (many, much, more)), Wh-pronouns.
ii. Determiners function as Ml there are articles (the, a/an), possessive determiners (my, our your), demonstrative determiners (this, that, these), and quantifier determiners (general (all, some, any, no) and gradable (many, much, more, most)), Wh-pronouns.
These two groups have many words in common but the rule is that those which function as heads are pronouns and those which function as modifiers are determiners.
This rule applies even to possessive pronouns. They can stand alone as the head of an NP: (This) (is) (mine). Possessive determiners are modifiers, which need a head to follow them: (This) (is) (my mouse).
Prepositional phrases (PP)
(a) Function: In the clause, PPs act as adverbials (A)
[(On Monday) (we) (arrived) (by train) (in Omsk)].
The adverbial PPs have various meanings. Thus in the three PPs are adverbials of time-when, of means, and of place, answering the questions When? How? Where? PPs are also common as subordinate phrases and often act as post-modifiers.
(b) Structure: PPs have exactly the same structures as NPs, except that they are introduced by a preposition; i.e. PP = p + NP. Normally prepositions are inseparable from the head that follow them. There are, however, circumstances in which a preposition can be separated, compare:
1. (Problems [(with which) (everyone) (must live)]
2. (Problems [which) (everyone) (must live) (with)])
In (1) with which is a PP, but in (2) the preposition has been left stranded at the end of the clause and the pronoun which which follows it in (1) has turned into a separate NP.
Another way in which prepositions can separate themselves from the following NP is in idioms like look at, look for, approve of, deal with, etc, compare:
1. Most monkeys live in trees
2. The hands indulge in wild gunplay
In (1) the PP clearly has adverbial function (Where do most monkeys live? In trees), but in (2) in seems more closely connected with indulge so that indulge in can almost be regarded as an idiomatic compound verb. So the structure of (1) is S P A and of (2) is S P p O.
Another type of idiom is that of phrasal verbs (make up, take off, hang about), but in these cases the second word is an adverb rather than a preposition.
The genitive phrase (GP)
(a) Function: GPs can function either as pre modifiers (M) or as head (H) in NPs, compare:
1. [(These) (are) (the technician’s) (tools)]
2. [(These tools) (are) (the technician’s)]
(b) Structure: GPs are just like NPs except that they end with the particle ’s. In the plural, after a noun with -s, the particle is not separately pronounced and is simply written as an apostrophe (‘).
5.4 The adjective phrase and the adverb phrase
Compared with NPs, AjPs, and AvPs tend to have a simple structure. Although potentially they consist of the same elements ({M*} H {M*}), in practice they often consist only of a head.
The adjective phrase (AjP)
(a) Function: In a clause, AjPs function as complement (C):[The weather has turned (milder)].
As subordinate phrases, AjPs can function as pre modifiers in NPs:
[The Mad Hatter took (a (very large) slice)].
But we only analyze the modifier as an AjP if the adjective has one or more modifiers. A single adjective without modifiers in front of a noun does not count as an AjP.
(b) Structure: The head of an adjective phrase is an adjective, which can be plain (big), comparative (bigger) or superlative (biggest). Pre-modifiers are mostly adverbs: typically, adverbs of degree. Some, especially very and too, can be repeated. Postmodifiers can be either adverbs (indeed, enough) or PPs. We also find certain kinds of clauses as postmodifiers (younger [than I thought], too hot [for me to drink], etc.
The adverb phrase (AvP)
(a) Function: AvPs function in the clause as adverbials (A).
(b) The head of an adverb phrase is an adverb (Av). Otherwise, the structure of AvPs is the same as that of AjPs.
5.5 The verb phrase (VP)
(a) Function: The VP is the pivotal element of the clause and always acts as predicator (P).
(b) Structure: The structure is of two elements: the main verb (Mv) and auxiliaries (Aux). The auxiliaries are options and precede the main verb. {Aux} {Aux} {Aux} {Aux} Mv, with any number of auxiliaries from zero up to four.
Auxiliaries come in four types: modal (Mod), perfect aspect (Perf), progressive aspect (Prog), and passive voice (Pass). Mod is always filled by one of the modal verbs, Perf is always filled by the primary verb have, and Prog is always filled by the primary verb be; similarly, the Pass position is always filled by be and the main verb position (Mv) can be filled either by any full verb (V) or by one of the primary verbs (be, have, do).
So the structure is: ({Mod} {Perf} {Prog} {Pass} Mv), always in this order.
Verb forms
In the table below, the verbs are expanded to include operator-verbs (modals and primary verbs).
We also make a distinction between tensed and tenseless forms. In the tensed verb phrase, the first word is always a tensed verb form.
The rules for combining verbs in the verb phrase can now be expressed:
Mod: modal + Vi
Perf: have + Ven
Prog: be + Ving
Pass: be + Ven
Mv: V
The initial parts of tensed VPs are tensed, whereas non-instal parts of any VP are always tenseless (Vi, Ving, Ven), so we can describe the structure as: +T {-T}{-T}{-T}{-T} where +T is tensed and -T is tenseless.
The ‘dummy operator’ do
The auxiliary do has a special role. Much like can, be, have, the particle not can simply be added after the first word of the VP (can’t, isn’t, hasn’t, doesn’t). So, while one cannot say “She sang not”, the same message can be conveyed using the verb do: “She did not sing” - this is the dummy operator do.
6. Clauses (Leech)
6.1 Elements of the clause
The five clause elements are illustrated in:
[S(Many people) P(are painting) O(their houses) C(white) A(these days)]. This example also demonstrates the typical ordering of these elements.
Predicator (P) and subject (S)
The predicator is the only element which is a verb phrase. The subject normally precedes the predicator an there is concord between them as regards number and person.
It is often possible to use a substitution test, substituting one of the subject pronouns for the phrase in subject position. There is a need to use several tests for identifying clause elements:
1. [A(In the box) P(are) S(six skulls)]
2. [S(The box) P(contains) A(six skulls)]
The criteria of position and meaning do not help to identify S in (1), but we can still rely on tests of concord and pronoun substitution.
Object (O) and complement (C)
The object is very closely tied to the predicator in terms of meaning, and typically denotes the person or thing most intimately affected by the action or state denoted by P. The complement can look superficially like an object (both can be NPs) but in terms of meaning it describes or characterizes the S or O. Objects and complements normally follow the P.
[S(They) P(‘ve elected) O(Harry) C(the next leader of the school)].
Adverbials (A)
Adverbials fill out the clause by adding extra circumstantial information of various kinds, ranging from time and location to the speaker’s attitude. Adverbials are the least closely integrated clause element in clause structure.
There is no fixed number of adverbials in a clause; in this they are rather like modifiers in the NP. Adverbials are generally much more mobile in the clause than the other elements. They are also optional in most clause types and can normally be omitted from the clause.
[She works at home very rarely] -> [She works at home] -> [She works].
6.2 Complex sentences
So far we have dealt with simple sentences consisting of a single main clause (MCl), but the majority of English sentences likely to be met in texts are complex sentences (sentences which contain additional clauses).
There are two ways in which additional clauses can occur in a complex sentences. Two or more clauses can be coordinated; that is, they can be linked as units of equal status:
[You support your team-mates] and [they’ll support you].
Or there may be one or more subordinate clauses (clauses which are grammatically subordinate because they are part of other clause. A subordinate clause is either (1) an element in a ‘higher-up’ clause or (2) a postmodifier in a phrase within a clause:
1. [[If you support you team-teams] (they) (‘ll support) (you)]
2. [(Players [who support their team-mates]) are supported in turn by them].
6.3 Tensed and tenseless clauses
So far we have dealt only with tensed verb phrases, containing a tensed verb, which is a verb showing tense and subject concord (for person and number). A tensed verb is either the operator (first auxiliary in the VP) or main verb (the only verb in the VP) if there’s no operator.
There are also many tenseless verb phrases containing no tensed verb. In a tenseless VP, all the verbs must be Vi, Ving or Ven (this can be any verb except modals which have no tenseless form).
The terms ‘tensed’ and ‘tenseless’ also apply to clauses. The clauses we have dealt with so far have been tensed clauses, in which P has been a tensed VP. But there are also many tenseless clauses, for example:
Tensed: [It would be best [if you (told) everyone]].
Tenseless: [The best thing would be [for you (to tell) everyone]].
The tensed subordinate clauses have tensed verbs as well as tenseless ones, while the tenseless subordinate clauses have only tenseless verbs.
6.4 Declarative, interrogative and imperative clauses
There are three major types of main clauses in English:
The declarative type is generally used to make statements, it is the most usual form of the interrogative type is mostly used to ask questions. The imperative type is most commonly used to give orders or make requests.
e.g. Jim will post these letters - Will Jim post these letters? - Post these letters.
There are two kinds of interrogatives:
1. Yes-no interrogative - asks for a yes/no answer. It is always the tensed operator that carries this contrast between yes and no: Jim will/won’t post these letters.
In forming yes-no interrogatives we place the operator in the prominent position before S: “Have you seen the latest blockbuster?” As with negation, if there is no operator in the declarative, then the corresponding interrogative needs dummy auxiliary do followed by an infinitive: He squeezed her hand -> Did he squeeze her hand?
In the interrogative, the verb phrase is split into two parts, rather as the verb phrase is split when an adverbial is shifted into it: Is he really making things so difficult?
2. Wh- interrogative - questions starting in what, when which, how, etc - open-ended questions. The operator is normally placed before S, and then, because the wh-word is the focus of attention, it is placed before the operator: [O(What) Vdid S(she) P(say)]? In those cases where the wh-word is in the S element, the rule that places the wh- word first is obeyed without changing the normal declarative order: [S(Who) P(ate) O(that sandwich)]?
The imperative
Unlike the declarative and interrogative, the imperative has a tenseless P and no S. We can think of it as being derived from a declarative by omission of S(you) and of the VP operator will/must:
[You) (will listen) to me] -> [Listen) to me].
In the imperative, the first verb in the VP is thus a Vi.
6.5 Active and passive clauses
The active voice is the basic, unmarked, most frequent form of the clause:
[S(She) P(has eaten) O(my porridge)].
The passive voice is the more marked form of the clause in which the S corresponds in meaning to an O of a corresponding active clause:
[S(My porridge) P(has been eaten)].
The S of the corresponding active becomes an optional A of agency in the passive, nearly always a PP marked with the preposition by:
[S(This porridge) P(had been eaten) (by Goldilocks).
6.6 More on clause structure
There are two kinds of object (O) and two kinds of complement (C).
A direct object (Od) is the most usual kind of object, and an indirect object (Oi), when it occurs, comes between the predicator and the direct object: S P Oi Od. Oi is normally optional and can often be replaced by an A element, a PP introduced by to or for, coming after the Od:
[S(You) P(can’t teach) Oi(an old dog) Od(new tricks)].
[S(You) P(can’t teach) Od(new tricks) A(to an old dog)].
Semantically, the Od generally refers to something directly affected by the action of the verb. The indirect object refers (generally) to a person who is indirectly affected by the action.
There is also a subject complement (Cs) and an object complement (Co). The Cs characterizes or describes the subject, whereas the object complement characterizes or describes the Od:
[S(Joe Walcott) P(was) Cs(a great boxer)].
[S(Everyone) P(considered) O(Joe Walcott) Co(a great boxer)].
The Cs normally follows P and the Co normally follows O. A clause normally contains only one complement, either Cs or Co.
Conjunctions and interjections can also occur in a clause:
ij: Ugh, it’s a coffee-cream again!, cj: And so it went on.
We treat these simply as particles.
An NP may occur as a vocative (Voc), a phrase which identifies the person addresses:
It was pretty fresh, (dude).
Vocatives are optional, and mobile, and are therefore more like adverbials than any other type of constituent. Only the vocative can be omitted or moved to the end of the clause, e.g.
Voc(Jock), pay attention. vs. S(Jock) pays attention.
6.7 Clause patterns
When we strip away the optional elements of each clause, we are left with a nucleus which can be called its clause pattern, and which expresses the kernel of its meaning.
eg. But gee, Alice, you must be kidding me now, baby. [cj ij Voc S P O A Voc]. Once we throw away the optional elements, we are left with a nucleus [S P O] (you must be kidding me).
When illustrating clause patterns, it is convenient to use a main clause in the declarative form in the active voice and with the unmarked word order: The chef stuffed a chicken.
The hierarchy of the principal clause elements is: S P Od Oi C A, and using S as a reference point, the further down the hierarchy we move, the more likely the element is to be optional.
Passive clause patterns
It is a general rule that a clause pattern with an object can be changed into a passive clause pattern, with the same verb, in which this object functions as subject.
[(He) (sold) (me) (the car)] -> [(The car) (was sold) (me) {by him}].
The passive allows us to change the semantic focus of the clause and also to omit reference to the doer of the action if we want to.
6.8 The structure of tenseless clauses
Corresponding to the three tenseless forms of the verb (Vi, Ving, Ven) there are three types of clause where those forms occur as the first (or only) element of the predicator. There are the infinitive clause (Cli), ing-clause (Cling) and en-clause (Clen).
Cli: We need space [for (the kids) (to play) (volleyball).
Cling: [(Her) (being fired)] was an act of discrimination.
Clen: We had [(the brakes) (fixed)] last year.
There are one or two details, apart from the verb, which distinguish tenseless from tensed clauses. For example, in an infinitive clause as above, a conjunction for normally precedes the subject and the infinitive particle to normally precedes the VP. In ing-clauses, the subject is sometimes the possessive form, like her above. Apart from this, these tenseless clauses have almost the same structural possibilities as tensed clauses.
6.9 Parsing a simple sentence
We now have all the grammatical resources we need to parse a simple English sentence containing all word classes: But alas, the two ugly sisters had gone home without her. There is no single right way to parse a sentence, but generally one can work down the rank scale from the sentence and working down to the word of which it is composed.
4. Syntax I/general principles - the structure of the clause
Orientation
Words, phrases, clauses and sentences
Syntax is the traditional term used for the study of the rules which help us combine words into sentences. The relationship between morphology, syntax and grammar can be conceived as:
morphology = word formation + inflectional morphology
syntax = grammar - inflectional morphology
grammar = syntax + inflectional morphology
A clause can be regarded as the linguistics representation of a state of affairs, usually built around a verb as a nucleus and a varying number of constituents. With regard to their outer form, clauses are divided into three categories:
i. finite clauses: verb is marked for tense and person, Where are you going?
ii. non-finite clauses: built around an infinitive, gerund or participle, Made in England.
iii. verbless clauses: depending on form of the verb (or its presence or absence), Off with you!
With regard to their function, clauses are divided into:
i. main clauses: capable of standing on their own, I’m going home
ii. subordinate clauses which expand or otherwise modify a main clause or another subordinate clause, e.g. [I’m only putting up with this] because I don’t want to lose you as a friend
Non-finite and verbless clauses tend to occur in subordinate function (there are exceptions)
The first step in a grammatical analysis of a clause is to distinguish between the subject and the rest (one big verb phrase consisting of the verb and its objects).
The crucial importance of the phrase or constituent level of structural organization becomes apparent particularly in sentences which are ambiguous (which have two different interpretations depending on the analysis one adopts).
e.g. I bought him a book about medieval sculpture in Ireland could be (1) [[I] [[bought] [him] [a book about medieval sculpture in Ireland]]] or (2) [[I] [[bought] [him] [a book about medieval sculpture] [in Ireland]]]. In (1) in Ireland is part of the NP, while in (2) it is a PP.
Most approaches to syntactic/grammatical description are based on a distinction between form (category) and function (the task a form performs in a grammatical context). Normally one linguistic form may have several different functions and a specific grammatical function within the sentence can be realized by many different form categories (e.g. to every student can work as a prepositional phrase, a complement of adjective, a modification of noun, etc. depending on the context of the rest of the sentence).
Examples of form categories as grammatical subjects:
Noun phrase (usual case): The holiday was a disaster.
Clause (common enough): That she was willing to forgive him is a little miracle.
Prepositional phrase (rare): From Stratford-upon-Avon to Luton Airport takes longer than the flight from London to Rome.
Demonstration/discussion
Basic strategies for the expression of grammatical relations
All languages use the same four fundamental strategies for the expression of grammatical relations, and the differences among them are due to the fact that the relative prominence of each strategy is variable across languages.
Grammatical relations can be expressed through: word order, inflection, function words (or free grammar morphemes), and intonation.
Word order
In modern English word order is one of the most important ways of coding grammatical relations in a sentence. Changes in the order of constituents are almost always associated with changes in meaning or lead to ungrammaticality.
Movement for emphasis in order to produce a balanced information structure in the clause is highly restricted in English (but is fairly common in German). We can sometimes front objects but only if we put an emphatic sentence accent on them.
Inflection
In the course of its history, English has shed most of its inflectional endings. Today we find inflection used to form the plural and genitive of nouns (books), the 3rd person singular (she plays) and past tense of verbs (I played) and the comparative (hotter) and superlative (hottest) forms of monosyllabic and some disyllabic adjectives.
Grammatical relationships expressed through inflection in German are expressed by grammatical function words (to,of,on) in English.
Function words
Function words are usually identical in form to prepositions, but differ in that they have lost their concrete spatial or temporal meanings and are used to express purely grammatical relations, e.g. dedicate something to someone - the original directional meaning is lost.
Intonation
Intonation plays a minor role both in English and German in coding grammatical structure.
Both languages have the option of turning a statement into a question through a change from falling to rising intonation, e.g. You are a student?
In English, intonation helps to distinguish between restrictive relative clauses and non-restrictive relative clauses, thus the students who were waiting in the lobby were sent home implies that there is a larger number of students and only a portion of them (the ones in the lobby) were sent home. However, the students, who were waiting in the lobby, were sent home means that a group of students, all of whom were waiting in the lobby, were sent home.
Typological classification of languages
Languages can be classified into diachronic “families” on the basis of their common origin and historical relationship. English turns out to be related to German or Dutch, more distantly to Latin or Albanian and even more distantly to the ancient Indian Sanskrit language.
Languages can also be classified from a synchronic point of view; such a typological classification is usually based on a limited number of parameters and allows for historically unrelated languages to end up in the same typological group, e.g. English may turn out to share some basic grammatical properties with Chinese.
The isolating type tends towards maximally simple words (where most words consist of one morpheme only), e.g Chinese
The agglutinating type may have long and complex words, but these are structured transparently and can be assembled and disassembled easily, e.g. Turkish
The inflectional (fusional) type has words which are complex and show signs of fusion (i.e. are much less transparent than the agglutinating type), Latin/Russian
The polysynthetic (incorporating) type has words which are extremely long and complex and very often encompass almost the entire information in a clause, Inuit
A simper two-way classification has been proposed into analytical and synthetic languages, the parameter of comparison being clause structure.
Analytical grammars code grammatical relations through word-order rules and free grammatical morpheme, whereas synthetic languages rely on inflection to do the job. Clearly, languages with a lot of inflected or fused words will end up in the synthetic type, whereas isolating languages will be highly analytical.
Problems and challenges
An obvious challenge to grammatical descriptions is the fact that there are hardly any rules which have not got numerous exceptions. No matter how much a rule may be refines to accommodate exceptions, these will always remain to some degree.
However, in natural human language a rule must be a tendency strong enough to give the speaker and listener orientation, but a margin of flexibility is desirable if only because it allows flexible responses to new communicative challenges.
5. Syntax II/the fundamentals of English grammar
Orientation
Parts of speech
Grammarians of ancient Greece and Rome left us a grammatical classification of words which has turned out to be surprisingly robust and is still at the heart of most modern taxonomies. They suggest a distinction between nouns, pronouns, adjectives, prepositions, verbs, adverbs, conjunctions and interjections. To this we might add the category of articles which is useful in English but was no really needed for Latin.
Parts of speech come in two major classes:
Auto-semantic words are those which have a fairly precise meaning by themselves, these are nouns, verbs, adjectives and adverbs.
Synsemantic words express very general concepts such as spatial, temporal or causal relations and acquire specific meanings only in conjunction with auto semantic words, these include a small number of adverbs as well as articles, pronouns, prepositions, and conjunctions.
The weakness of this classification is that some categories may be messy, e.g. pronouns - some may function as a pronoun, but can also display adjectival or adverbial properties.
New types of part-of-speech categories have been suggested in recent linguistics, e.g. determiners (a(n), the, all, both, each, some many, etc.)
Phrases
Phrases/constituents combine into clauses. The most common types of phrases are noun phrases (NP), prepositional phrases (PP), verb phrases (VP), adjective phrases (AdjP) and adverb phrases (AdvP). The phrases are need after they heads and include the heads and their modifiers.
A verb phrase typically consists of a main verb, the auxiliaries which modify it, and those phrases or constituents other than the subject which are necessary to form a basic clause, e.g. the roof [collapsed] last night OR I [put the alarm clock on the shelf] last night (‘last night’ is not bracketed because it can be dropped without making the sentence structurally incomplete).
Phrases, like words, are form-categories, and their function only becomes apparent in the context of the clause.
The seven basic clause patterns
There is an immense variety of different clause structures, but at closer inspection a relatively simple underlying mechanism is revealed. All simple sentences in English correspond to one of the seven basic clause patterns illustrated above.
The patterns differ with regard to the number and type of constituents which are needed in addition to the subject to form a grammatical sentence.
1 (SP): We left, Tom snores, The children are sleeping.
2 (SPCs):. Sally is a student, Sally is intelligent, Sally became very interested in sports.
3 (SPA): The film lasts two hours, I ran down a steep hill, We are going home.
4 (SPO): My sister is reading the paper, Arsenal beat Manchester United.
5 (SPOO): Who gave you this book?, The guide explained everything to us.
6 (SPOCo): You make me nervous, Everybody considers Sally the best choice.
7 (SPOA): He put the book on the shelf, We couldn’t get the piano through the narrow door.
Demonstration/discussion
It is possible to assign any given clause that we might find in a text to one of the seven basic clause types if we take into account a few simple expansion strategies which provide the link between the simple basic system and the multiplicity of structures encountered in “live” data:
Addition of optional adverbials indicating circumstances in the clause
Unlike the obligatory adverbials in the SPA and SPOA patterns, there are optional adverbials - (A) - which can be added to any of the basic clause types without any formal restrictions as long as such an addition makes sense.
e.g. Peter yawned (a lot) (at the meeting) (last night).
Expansion of constituents through pre- and postmodification
Constituents can also be made more complex and longer internally through elaborate pre- and postmodification.
e.g. (In a recent paper), we demonstrated a modified sodium dodecyl sulphate polyacrylamide gel electrophoresis method for visualization of factor VIII heavy chain polypeptides.
This is a case of (A)SPO.
Clause embedding and clause combining
It is also possible to replace any constituent phrase except the verbal predicate itself by a clause.
e.g. I cannot believe this -> I cannot believe that you never noticed that he was in love with you
Here the object is expanded into an object clause.
If there is a hierarchical relationship between the clauses of a sentence we speak of a complex sentence. If we coordinate independent clauses (and/but/or), we produce a compound sentence.
Problems and challenges
One fragment of the grammar in which present-day English develops extreme degrees of complexity is non-finite clauses, that is dependent clauses organized around non-finite (tenseless) forms of the verb. In many cases, the use of such non-finite clauses tends to blur the boundary between superordinate and subordinate clauses.
e.g. [People in the region expect [that [the new high-speed railway] [will boost [the local economy]]]] vs. People in the region [the local economy to be boosted [by the new high-speed railway]].
Similar analytical problems are posed by superficially similar structures.
e.g. She is looking for an au-pair speaking Portuguese - seems to be a relative clause.
She found the au-pair reading her diary - seems to a complement of the object or adverbial.
She insisted on the au-pair staying in all evenings of the week - seems to function as one unit in replacement and question tests and is therefore best regarded as an object.
Therefore, though these seems similar on the surface, they cannot be treated in the same way.