Null auxiliaries

So far, all the clauses we have looked here have contained a TP projection headed by a finite auxiliary or infinitival to.

The obvious generalization suggested by this is that all clauses contain TP. An important question begged by this assumption, however, is how we are to analyze finite clauses which contain no overt auxiliary. In this connection, consider the construction illustrated in (17) below:

Both clauses here (viz. the he clause and the bracketed she clause) appear to be finite, since both have nominative subjects (he/she). If all finite clauses contain a TP projection headed by a finite T constituent, it follows that both clauses in (17) must be TPs containing a finite T. This is clearly true of the he clause, since this contains the finite modal auxiliary could; however, the she clause doesn’t seem to contain any finite auxiliary constituent, since have is an infinitive form in (17) (the corresponding finite form which would be required with a third-person subject like she being has). How can we analyze finite clauses as projections of a finite T constituent when clauses like that bracketed in (17) contain no finite auxiliary?

An intuitively plausible answer is to suppose that the string she have helped him in (17) is an elliptical (i.e. abbreviated) variant of she could have helped him, and that the T constituent could in the second clause undergoes a particular form of ellipsis called gapping. (Gapping is a grammatical operation by which the head of a phrase is given a null spellout – and so has its phonetic features deleted – when the same item occurs elsewhere within the sentence, and is so called because it leaves an apparent ‘gap’ in the phrase where the head would otherwise have been.) If so, the second clause will have the structure (18) below (where  marks an ellipsed counterpart of could, and have is treated as a non-finite AUX/Auxiliary heading an AUXP/Auxiliary Phrase – the rationale for AUXP):

The head T position of TP in a structure like (18) is filled by the ellipsed auxiliary . Although an ellipsed item loses its phonetic features, it retains its grammatical and semantic features, so that  in (18) is a silent counterpart of could. The null T analysis in (18) provides a principled account of three observations. Firstly, the bracketed clause in (17) is interpreted as an elliptical form of she could have helped him: this can be straightforwardly accounted for under the analysis in (18) since T contains a null counterpart of could. Secondly, the subject is in the nominative case form she: this can be attributed to the fact that the T position in (18) is filled by a ‘silent’ counterpart of the finite auxiliary could, so that (like other finite auxiliaries) it requires a nominative subject. Thirdly, the perfect auxiliary have is in the infinitive form: this is because  (being a null copy of could) has the same grammatical properties as could, and so (like could) requires a complement headed by a word (like have) in the infinitive form.

A further argument in support of the null T analysis in (18) comes from facts relating to cliticisation (a process by which one word attaches itself in a leech-like fashion to another). The perfect auxiliary have has a range of variant forms in the spoken language. When unstressed, it can lose its initial /h/ segment and have its vowel reduced to schwa /ə/, and so be pronounced as /əv/ e.g. in sentences such as You should have been there. (Because of is also pronounced /əv/ when unstressed, some people mistakenly write this as You should of been there – not you, of course!) However, when have is used with a pronominal subject ending in a vowel or diphthong (e.g. a pronoun like I/we/you/they), it can lose its vowel entirely and be contracted down to /v/; in this weak form, it is phonetically too insubstantial to survive as an independent word and encliticises onto (i.e. attaches to the end of) its subject, resulting in structures such as:

However, note that have cannot cliticise onto she in (20) below:

so that she’ve is not homophonous with the invented word sheeve. Why should cliticisation of have onto she be blocked here? Let’s suppose that have-cliticisation is subject to the following structural conditions:

The asymmetric c-command condition (21i) in effect requires the pronoun to be ‘higher up’ in the structure than have. (In the relevant technical sense, one constituent X asymmetrically c-commands another constituent Y if X c-commands Y, but Y does not c-command X.) The adjacency condition (21ii) requires have to be immediately adjacent to the pronoun which it cliticises to. (A descriptive detail which we set aside here is that (21) applies specifically to encliticisation of have: encliticisation of the ’s variant of has is subject to far less restrictive conditions on its use – but this will not be pursued here.)

To see how (21) works, consider the structure below:

Here, the pronoun they ends in a diphthong and so is the kind of pronoun that have can cliticise onto. The asymmetric c-command condition (21i) is met in that they c-commands have, but have does not c-command they. The adjacency condition (21ii) is also met in that there is no constituent intervening between they and have.

Since both its structural conditions are met, (21) correctly predicts that have can encliticise onto they, so deriving They’ve left. The kind of cliticisation involved here is essentially phonological (rather than syntactic), so that they and have remain separate words in the syntax, but are fused together in the PF component (i.e. the component responsible for determining Phonetic Form) once the structure generated (i.e. formed) by the syntax has been handed over to the PF component for morphological and phonological processing.

In the light of our discussion of have cliticisation, now consider why cliticisation of have onto she is not possible in (20) ∗He could have helped her or she’ve helped him. Under the null T analysis suggested above, the second clause in (20) contains a null variant of could and has the structure shown in (18) above, repeated as (23) below:

Although the asymmetric c-command condition (21i) is met in (23) in that she c-commands but is not c-commanded by have, the adjacency condition (21ii) is not met in that she is not immediately adjacent to have because the null auxiliary could intervenes between the two (in the sense that  c-commands have, and  is in turn c-commanded by she). Thus, the presence of the intervening null auxiliary   blocks cliticisation of have onto she in (23), thereby accounting for the ungrammaticality of (20) ∗He could have helped her or she’ve helped him.

Turning this conclusion on its head, we can say that the ungrammaticality of (20) provides us with empirical evidence that the bracketed clause in (17) contains a null counterpart of could intervening between she and have – as is claimed in the analysis in (23) above.

Our discussion so far in has suggested that some seemingly auxiliariless clauses are TPs headed by a T containing an auxiliary which (via ellipsis) is given a null phonetic spellout. A rather different kind of null-auxiliary structure is found in African American English (AAE), in sentences such as the following (from Labov 1969, p. 717):

In AAE, specific forms of the auxiliary be have null variants, so that we find null forms of are and is in contexts where Standard English (SE) would require the contracted forms ’s and ’re. Hence, in place of SE he’s getting crippled we find AAE he gettin cripple (with a null counterpart of ’s). Evidence in support of the assumption that AAE sentences like (24) incorporate a null variant of is comes from the fact that the missing auxiliary is may surface in a tag, as in sentences such as the following (where the sequence following the comma is the tag) (from Fasold 1980, p. 29):

In tag sentences, the auxiliary found in the tag is a copy of the auxiliary used in the main clause. This being so, it follows that the main gonna clause in (25) must contain a null variant of the progressive auxiliary is. In other words, the main clause in (25) must be a TP with the structure shown in skeletal form in (26) below ( indicating that the phonetic features of the auxiliary are not spelled out):

Interestingly, the form am (contracted to ’m) has no null counterpart in AAE, nor do the past-tense forms was/were. It would seem, therefore, that the only finite forms of BE which have a null counterpart in AAE are the specific auxiliary forms are and is. No less interestingly, Wolfram (1971, p. 149) reports that in non-standard Southern White American English the use of null auxiliaries is even more restricted, and that the only form of BE with a null counterpart is are; cf. the parallel observation by Fasold (1980: 30) that ‘There are many southern whites who delete only are.’