ON THE CUTTING ROOM FLOOR

There are some passages of the book as we originally wrote it that we had to drop as we cut the book down to the planned size. We plan to deposit all of those passages here for anyone who would like to have read a little more on certain topics.


Chapter 3. Verbs, tense, aspect and mood

p. 41: (d) Dare

We observe in the text that auxiliary dare is grammatically very much like auxiliary need: both are modals occurring only in non-affirmative contexts and taking a bare infinitival complement. Two differences between dare and need are as follows:

[25']
i a. I wouldn't dare jump out. b. I wouldn't need ask for help.
ii a. I wouldn't dare to jump out. b. I wouldn't need to ask for help.


p. 46: §5.1, The present tense

In the text we list four uses of the present tense, labelled (a)–(d). There is a further use that could be added:

(e) Timeless use of the present tense

Sometimes the present tense is used where no particular time is relevant at all:

<RDW:49>     [35]

[34'] Hobbes has a rather ordinary, though somewhat pessimistic, view of human nature.

This is from a philosophy reference work. A preterite would have been perfectly acceptable, but the present tense is justified for this permanent, timeless claim because Hobbes' 17th-century writings still exist and anyone can check up on his view of human nature at any time.


p. 53: The progressive futurate

Examples [50a–b] illustrate the distinction between the simple (non-progressive0 and progressive futurate constructions. While the former conveys that there is an arrangement in place for me to see my broker, the latter could be used in a context where I simply intend to see my broker. This element of intention largely restricts the progressive futurate to situations involving human agency. Thus in the following pair, only the non-progressive [a] would normally be used:

<REG:50>     [50]

[50']
a. The sun sets at 5.20 tomorrow. b. *The sun is setting at 5.20 tomorrow.


Chapter 4. Clause structure, complements and adjuncts.

p. 67: §2.1, Distinctive syntactic properties of the subject in English

The first property we mention is that the basic position of the subject is before the V (and the whole VP). It is worth noting here that languages differ strikingly on where the subject goes. Russian has considerable freedom in subject position (in Russian nouns have a richer inflectional system than in English, so that the subject can usually be identified by the inflectional form of the head noun rather than by position). Irish typically has the subject after the verb; and in several languages of the Amazon basin, the NP before the V is the object and the one following the VP is the subject, so we get Object–Verb–Subject, i.e. the exact opposite of the English order.


Chapter 5: Nouns and noun phrases

p. 98: §7.2, Fused determiner-heads

Examples [43i] illustrate the impossibility of the functioning as a fused determiner-head: that is used instead. The examples given involve singular NPs; others showing the same contrast for plural NPs are as follows:

<REG:50>     [50]

[1]
i c. *The of you who prefer can take the bus.
d. Those of you who prefer can take the bus.


p. 100: §8, Pronouns

We begin this section by noting that pronouns differ from common and proper nouns by their inability to take determiners as dependent. It is worth adding that the expression Is it a he or a she? is not a counterexample: here he and she have been recategorised as common nouns, and have a different meaning from the pronouns: a he is a male, a she is a female. The same applies to phrases like I'll never find another you: this involves a special use of you as a common noun meaning "person with your qualities". Evidence for this comes from verb agreement: we would not say Another you are never going to come along; it would be Another you is never going to come along. The agreement is 3rd person, not 2nd.


p. 106: The nominative–accusative contrast of case

A personal pronoun functioning as predicative complement may be in nominative or accusative form, as illustrated in [55iv]. The nominative is a marker of formal style, and as we observe is found only in certain constructions, mainly with it as subject. As further illustration of this point, consider a context where one is describing a role-playing skit where a mother acts the part of her daughter and vice versa): we might say In the skit, I was her and she was me, but the version with nominatives In the skit, I was she and she was I) would sound impossibly pedantic to most people.


p. 110: new section, §10, Proper nouns and proper names

We had little space in the text to talk about proper nouns beyond saying that they characteristically function as the head of proper names — we accordingly add a further section here dealing with this subcategory of noun, and elaborating on the distinction between proper nouns and proper names.

The central cases of proper names are expressions that have been conventionally adopted as the name of a particular entity — a person, animal, place (such as a town, country, lake, river, mountain), institution, and so on.

The reason why we need to distinguish between proper nouns and proper names can be seen from the following examples, where underlining marks proper nouns and brackets enclose proper names:

[64]
i a. Have you ever been to [Rome]? b. Have you ever been to [the Colosseum]?
ii a. She was born in [Edinburgh]. b. She studied at [Edinburgh University].
iii a. She lives in [Canada]. b. She lives in [the United States].

Proper nouns, being a subclass of noun, are word-sized units, whereas in the most straightforward cases proper names are NPs. In the [a] examples of [64] the proper name NPs consist simply of a proper noun as head, whereas those in the [b] examples contain dependents as well as the head.

(There are often variant forms of proper names. The United States, is a shortened version of the United States of America, and can be further shortened to the informal the States.)

It is also important to note that proper nouns commonly function as head of NPs that are not proper names. In the following underlining again marks proper nouns, but the bracketed NP is not a proper name:

[65]
i We were listening to [some Beethoven].
ii The Gallery has acquired [another Rembrandt].
iii It's very different from[the Edinburgh I remember as a student].
iv Don't forget that [the Smiths] are coming round this evening.

These represent secondary uses of proper nouns, where the meaning is derivative from that of their primary use as heads of proper name NPs.

Proper nouns thus differ from common nouns in that their primary function is as head of a proper name. In this function they mostly occur without a determiner: proper name NPs are inherently definite, and the definiteness generally does not have to be explicitly marked. There are, however, some proper names containing the definite article as an integral part: this is so with the Colosseum and also with various names of rivers the Nile), seas the Mediterranean), deserts the Sahara), newspapers or periodicals the Times), and so on.


Chapter 6. Adjectives and adverbs

p. 116: commentary on [14a]

We make the point in the text that the ambiguity of [14a] (They are entertaining) is resolved in favour of the verbal reading if we add some colleagues as object: in They are entertaining some colleagues the word entertaining can only be a verb. Note, however, that without the determiner some which marks the beginning of the NP the ambiguity remains. Entertaining colleagues could be either verb + object or a predicative NP with colleagues as head and entertaining an attributive modifier indicating that the colleagues are amusing or interesting.


Chapter 7. Prepositions and preposition phrases


p. 133: (b) Inflection and gradability

We note in the text that prepositions are normally (though not invariably) non-gradable. This relates to the fact that they prototypically express relations in time and space, and these mostly aren't matters of degree. (Your keys may be on the table or may have fallen into the wastebin, but they can't be more on the table than your watch, or fall more into the wastebin than your watch did.) So usingmore with the most standard meanings of prepositions likeat orinto will almost always sound linguistically odd.


p. 135: commentary on [19ii]

Notice that while verbal and prepositionalowing both licence a PP complement withto to the drought in [19iia],to the bank in [iib]), only the verb also licences an NP complement, an object so much in [iib]). In this interesting case, the gerund-participle of a verb that does take an NP complement has evolved into a preposition that does not.


p. 137: Grammaticised uses of prepositions

A further point that can be made here is that prepositions bear some resemblance to the inflections of certain other languages, where the work prepositions do in [20] would be done by inflectional endings on nouns.


p. 138: §5, Preposition stranding — Prescriptive Grammar Note

We point out in the text that although stranded prepositions are commonly condemned by prescriptivists they have widely used in English for hundreds of years. Stranding is found in the writing of Shakespeare and on down the years through the works of essentially all authors of the past three centuries. The allegation that it is improper seems to have originated with a personal opinion of the essayist John Dryden in 1672. By 1800 it had been picked up by a number of influential grammar-book authors, and it somehow survived the subsequent 200 years without being discredited — even though the article on it in Fowler's famous Modern English Usage (1926) opens by calling it `a cherished superstition'.


p. 142: §7.1, Goal, source and location

The location PPs in [32] qualify as complements because of their close relation to the verb.

PPs expressing location (as opposed to goal and source) can, however, also function as adjunct, as in The dog was sleeping underneath my bed/in her bedroom.

A further point to add to this short section is that the references made to positions or directions in space that are made by PPs have one thing in common with predicative complements: they exhibit either subject or object orientation. The following may thus be compared with [25] in Ch. 4 (p. 76).

[32']
subject orientation object orientation
i a. Liz went to school. b. We took Liz to school.
ii a. The coffee is in the refrigerator. b. We keep the coffee in the refrigerator.

In [i], the location change happens to Liz in each case, but Liz is subject in [a] and object in [b]. Similarly, in [ii], the location being given is of the coffee in each case, but the coffee is subject in [a] and object in [b].


p. 144: §7.3, Particles

As the text makes clear, the key thing about particles syntactically has to do with constituent order: they are the only complements that can freely come between the verb and its direct object. In most cases, the particle can occur either in this position (as in [36iia]) or else after the object (as in [36ia]). In a few verb + particle combinations, however, the particle must precede the object, as [37'iii] illustrates.

[37']

particle + object object + particle
i a. She took off her hat. b. She took her hat off.
ii a. I held out my hand. b. I held my hand out.
iii a. They put up strong resistance. b. *They put strong resistance up.


Chapter 9. Clause type: asking, exclaiming and directing

p. 166: Multiple interrogative phrases

The examples in [15] are clauses with more than one interrogative phrase, with only one of them occurring before the verb. There are constraints on which one of them occurs in this position. If the subject is an interrogative phrase (as in [i]), it begins the clause; if not, then if the object is an interrogative phrase it can be fronted (as in [ii]); otherwise some other interrogative-phrase complement may be fronted (as in When did you buy what?).

p. 167: which

Interrogative which is a determinative, and the text illustrates its use in determiner function: Which video shall we get? It should be borne in mind that it can also function as fused determiner-head, as in Which shall we get? (see Ch. 5, §7).


p. 168: §2.7, Echo questions

One point we would add to the discussion in the text is that the echo question features can be superimposed on any of the clause types. In [18] the stimuli are declaratives, while two other possibilities are illustrated in [18']:

[18']
stimulus echo question
i A: What did Tom say? B: What did who say? [open interrogative]
ii A: Give it to Pat. B: Give it to Pat? [imperative]

The echo question features, then, are irrelevant to clause type: B's echo questions in [18'] belong to the same clause type as A's stimulus. The type of speech act involved, though, is determined by the echo feature superimposed by B. For example, B's utterance Give it to Pat? in [iv], is a (closed) question if we're concerned with the speech act made, but an imperative if we're asking about clause type.

This is a further reason, of course, for keeping the clause type categories very clearly distinct from the speech act categories.


Chapter 10. Subordination and content clauses

p. 176: §3.2, Declaratives as complement

Our commentary on example [5v] analyses the content clause (that we just can't afford it) as complement of be in its specifying sense. Content clauses don't occur, however, as ascriptive predicative complements, so we don't get sentences like *Jill seems that she's enjoying herself; instead we have Jill seems as if she's enjoying herself, with the content clause functioning as the complement of as if.

It will also be noted that we begin this section by saying that declarative content clauses `mostly' function as complement of a verb, noun, adjective or preposition. The qualification `mostly' is needed because they are also occasionally found as adjunct. In this construction the subordinator that is obligatory, as in:

[5']
i What has happened, that you are looking so miserable?
ii `This is my party card', he said, holding it high, that all might see it.


p. 178: §4.1, Closed interrogatives and the subordinators whether and if

Subordinate closed interrogatives are normally marked as such by one or other of the subordinators whether and if, as illustrated in [9]. For the most part, whether and if are interchangeable, but there are some constructions where only whether is permitted. They include those where the subordinate clause precedes the verb of the matrix clause, as in:

<RED:12>     [11]

[9'] Whether/*If she'll change her mind remains to be seen.

Conversely, if tends to be preferred when the construction is reporting a request. We saw in Ch. 9, §4.4 that main clause interrogatives are often used to make requests, so that Will you chair the meeting, for example, will typically be interpreted as a request to do so, and the subordinate clause in [9'iib] will typically report such a request.


p. 179: §4.3, Interrogatives as complement

In general, open interrogative complements are licensed by the same items as closed ones, but there are a few exceptions:

[11']
closed interrogative open interrogative
i a. I doubt whether they'll help us. b. *I doubt why she told them.
ii a. *It's amazing whether they replied. b. It's amazing what they said.

Doubt and related lexemes accept only closed interrogatives — and note that the meaning here does not involve "the answer to the question", for [ia] simply means "think it unlikely that they'll help us. Conversely, amazing and similar lexemes accept only the open type.


Chapter 11. Relative clauses

p. 189: Antecedents allowed for supplementary relatives

We point out in the text that supplementary relatives allow a wider range of antecedents than integrated ones, including clauses and proper nouns without determiners. nevertheless, there are certain types of antecedent that supplementary relatives never have: NPs containing every, any or no such as everyone, every member, anything, any book, no one, no candidate, etc. Note, then, the contrast illustrated in [13']:

[13']
i No member who cares about our work should miss this moving documentary.
ii *No member, who cares about our work, should miss this moving documentary.

NPs like no one combine well with integrated relatives, as seen in [13'i], but they provide no basis for loosely attaching a secondary assertion about someone satisfying the condition x cares about our work: [13'ii] is not grammatical, and sounds incoherent (it says that we should all see the documentary; but who is this person who cares about our work?).


Chapter 13. Non-finite clauses and clauses without verbs

p. 219: Auxiliary verbs

We note in the text that auxiliaries, when used as markers of tense, aspect, mood or voice, are catenative verbs, entering into the simple catenative construction. Recall from Ch. 3, §3.4 that not all auxiliaries are invariably used as markers of these categories. In particular, be and (for some speakers) have have other uses too, and in such cases they are not catenative verbs. Be, for example, is a catenative verb in [30'a], but not in [b]:

[30']
a. She is working hard. b. She is a hard worker.


Chapter 14. Coordination and more

p. 231: Asymmetric coordination

We point out in the text that asymmetric coordinations — those where the order of coordinates cannot be reversed without loss of acceptability or a change in interpretation — commonly carry implications beyond the basic additive or alternative meanings of the coordinators and and or. Some of these are illustrated in the examples of [19]. A further property associated with coordinations of this kind is that the usual `across-the-board' (p. 230) does not always apply. This is seen in such examples as:

[19']
i This is the homework which she [stayed up all night and eventually finished __/].
ii Which are the bills [that we need to pay __ today or we won't get the discount]?


p. 234: Joint coordination

We say in the text that the central cases of joint coordination are of NPs. But there are other possibilities, as illustrated by the contrast in:

[27']
distributive coordination joint coordination
a. I had some [red and blue] crayons. b. I wore a [red and blue] shirt.

The natural interpretation of [a] is that some crayons were red and some blue: the colour properties apply separately to the roses. In [b], by contrast, the shirt was partly red and partly blue, so the colour properties apply jointly to the shirt.


p. 235: §8.1, Expansion of coordinates by modifiers

It will be noted that all the examples in [29] involve distributive coordination. Joint coordination does not permit the expansion of one or more coordinates by a modifier, as illustrated by the sharp ungrammaticality of, for example, *Kim and Pat too are a good pair.


Chapter 15. Information packaging in the clause

p. 244: §2.4, Lexical restrictions

We illustrate in [13] various transitive verbs which, either generally or in certain senses, do not permit passivisation. There are also two or three verbs taking clausal rather than NP complements which occur only in passive clauses (again either generally or in certain senses). Two examples are the verb rumour and the verb say in its catenative use with an infinitival complement:

[13']
active passive
i a. *They rumour her to be engaged. b. She is rumoured to be engaged.
ii a. *Max said Sue to be very rich. b. Sue was said by Max to be very rich.

The passives here are thus further cases of non-canonical clauses with no direct canonical counterparts.


p. 246: §2.9 Adjectival passives

The examples in [20] illustrate the distinction between be-passives and adjectival passives at its sharpest, with the clauses describing events when interpreted as be-passives and states when interpreted as adjectival passives. It must be emphasised, however, that be-passives don't always describe events. Whether a be-passive has an event or state interpretation will depend on the semantic properties of the verb, and the passive will match the corresponding active:

[21']
active be-passive
i a. Kim wrote the letter. b. The letter was written by Kim.
ii a. Everyone despised him. b. He was despised by everyone.
iii a. Rioters surrounded our car. b. Our car was surrounded by rioters.

In all of these, the active and the be-passive describe the same situation. In [i] it's dynamic (an action); in [ii] it's static (a state); in [iii] it's ambiguous — either rioters took up positions all around our car (dynamic), or our car had rioters around it (static).

Where a be-passive has a static interpretation, the semantic difference between passive clause and adjectival passive is lost, but syntactically an example like [21'iib] can still be either a passive or a complex-intransitive:

Thus there are two different syntactic structures without a corresponding difference in meaning.


p. 248: §3.1, Subject extraposition

A number of verbs, such as appear, happen, seem, and turn (as used with out) occur in a construction resembling extraposition in that they have dummy it as subject and a content clause at the end. There is, however, no counterpart with the content clause in subject position:

[24']
i a. *That I underestimated you seems. b. It seems that I underestimated you.
ii a. *That this was false turned out. b. It turned out that this was false.


p. 250: Extended existentials

The construction with a hollow extension, illustrated in [29iv], is of interest because it is one of those that allows a personal pronoun as displaced subject. The point of interest is that the pronoun occurs in accusative case, not nominative: There's her to consider, not *There's she to consider. This is clear evidence that what we call displaced subjects are not actually subjects. Personal pronoun subjects of finite clauses have nominative case form.


p. 250–1: Constraints on the use of basic and existential constructions

Besides the two constraints described in the text (those concerning indefinite NPs and definite NPs), there is a third concerning quantified NPs. Displaced subjects generally can't be quantified by all, most, each, or every(as opposed to some or many):

[32']
i a. Every bike was in the bike rack. b. *There was every bike in the bike rack.
ii a. Most members are here now. b. *There are most members here now.
iii a. Many members are here now. b. There are many members here now.


1