5.3.1 The auxiliary as a dummy
One very interesting fact about aspectual auxiliaries is that each auxiliary is accompanied by a specific morpheme which is always realised on the verbal element which follows the auxiliary:
(156) | a | has seen the light |
b | has been seeing the light | |
c | is seeing the light | |
d | is being seen |
The fist two examples in (156) show that the auxiliary have is always followed by a verbal element in the ‘en’ form (though there is irregularity here and the morpheme is not always represented like this – see chapter 1). This element may be a main verb, as in (156a) or the auxiliary be as in (156b). The auxiliary be is followed by a verbal element in its ‘ing’ form and again this can either be a main verb (156c) or another auxiliary, as in (156d), where we have the passive auxiliary be. We have already seen that the passive morpheme is another instance of en. This attaches only to main verbs, a fact which follows from the analysis given above where the morpheme was treated as a light verb immediately above the thematic VP to which the main verb will move.
What is the nature of the two parts of each aspectual elements, the auxiliary and its associated morpheme? A classic analysis dating back to Chomsky (1957) is that the auxiliary and its morpheme are inserted into a structure as one element and then the morpheme is ‘hopped’ backwards onto the following verbal element:
(157) |
However some of the details of this analysis were never fully worked out. What is the lexical status of the auxiliary plus morpheme element, for example? If it is to be considered a single lexical item, how is it possible that a syntactic rule can break it apart? But if it is not a single element, what is the relationship between the two parts and how do we ensure that they are always inserted into a structure together?
A related issue concerns the meaning that aspectual elements bring to the sentence. Of the two elements, which is the meaningful one? There are at least three possibilities. Perhaps the most intuitive one is that the aspectual meaning is contributed by the auxiliary and the morpheme has no semantic input. However, it is possible that the meaning contribution is made by the morpheme and the auxiliary is meaningless, or that both elements have a contribution to make. One relevant observation is that the use of meaningless auxiliaries is not unheard of in English. The classic example is the auxiliary do which seems to have a variety of uses, mainly to do with providing an element to fulfil a purpose that the main verb is not suited for. For example:
(158) | a | did you see that? |
b | I didn’t see that | |
c | you DID see that! | |
d | you saw that, didn’t you? |
In these examples, the auxiliary do adds very little to the meaning of the sentence, apart from the fact that it carries tense. However, given that main verbs can do this, this is clearly not the main function of the auxiliary in these examples. Instead the auxiliary is used to do something that main verbs cannot do. In (158a), the auxiliary is moved to the front of the clause to form a question, in (158b) it is used to bear the contracted negative, in (158c) it bears stress in order to assert something that had previously been denied and in (158d) it is used to form a tag question, the main function of which is to lessen the force of a statement. As the following show, these are all things that we cannot use a main verb to do:
(159) | a | *saw you that? |
b | *you sawn’t that | |
c | you SAW that | |
d | *you saw that, sawn’t you? |
The ungrammaticality of most of the sentences in (159) shows that the main verb cannot be used in this way. That (159c) is not ungrammatical does not indicate that it is an exception, however, as this has a different meaning to (158c). In (159c), the emphasised verb is used to question or contradict a previous statement in terms of the content of the verb itself. So, for example, if someone claimed to have seen something that was invisible to others, (159c) might be an appropriate response. Importantly, (159c) could not be used to contradict someone’s claim that they did not see something, i.e. contradicting the truth of their statement. The point is, then, that the auxiliary do is inserted into a sentence to do something that is impossible for a main verb to do and hence it has a purely syntactic role rather than a semantic one. For this reason it is often called the dummy auxiliary. Perhaps its main function is to support the tense morpheme when, for whatever reason, this cannot appear on the main verb and hence the phenomenon is also called do-support.
Another instance of the use of a meaningless element might be the use of the copula be in examples such as the following:
(160) | a | Tim is tall |
b | Ferdinand is a fake | |
c | Gertrude is in the garden |
Considering the semantic relationships that exist in these sentences, we notice that they are established between the subject and the predicative element after the verb: Tim and tall, Ferdinand and a fake and Gertrude and in the garden. Again, the main function of the verb here appears to be to support the tense morpheme and it seems to make very little contribution of its own. Indeed, in circumstances where there is no need to express the tense morpheme, the verb is not used:
(161) | a | I consider [Tim tall] | = | I consider [that Tim is tall] |
b | I deem [Ferdinand a fake] | = | I deem [that Ferdinand is a fake] | |
c | I wanted [Gertrude in the garden] | = | I wanted [Gertrude to be in the garden] |
The bracketed part of the sentences on the left in (161) express the same predication relationships as those on the right and the only difference between the two is the expression of tense in the latter. Thus it looks as though the copula is used to support the tense morpheme when predication relationships are established between a subject and a non-verbal element. It is interesting that the verb be is used in this case, not the verb do as it is in cases of do-support. Even though both elements seem to contribute little to the interpretation of the sentence, it seems that their use is specialised to certain contexts: do for cases where the main verb fails to be able to support tense and be for cases where there is no verb present to support tense. I know of no explanation for this fact.
One more possible use of a meaningless verbal element which follows from an analysis developed above is the use of the auxiliary be in the passive. We analysed the passive construction as involving the replacement of an agentive light verb with a non-agentive one, realised as the passive morpheme. From this perspective then, the passive morpheme is the defining element of the passive construction. Of course, most passives also make use of the passive auxiliary be:
(162) | a | Sam was seen |
b | Harry was being hit | |
c | Barry was believed to have been murdered |
What is the function of the auxiliary in these sentences? Note that the auxiliary bears some morpheme: in (162a) the tense, in (162b) the ing associated with the progressive auxiliary and in (162c) the tense on the first passive auxiliary and the en associated with the perfective have on the second. In these examples, the main verb cannot bear these morphemes for the simple reason that it is already bearing the passive morpheme and it seems a basic principle of English morphology that no word can bear more than one overt inflectional morpheme:
(163) | a | *it seened/sawen | = | it was seen |
b | *she fallend/fellen | = | she had fallen | |
c | *Ron runninged/ranning | = | Ron was running |
Thus, again, we might say that the passive auxiliary is used to support a morpheme that the verb is unable to due to restrictions on the morphological structure of English.
Again note that the form of the auxiliary used is restricted to context: the auxiliary must be be in this case, not do. Support for this approach can be gained from observing contexts in which there is no other morpheme to be supported, in such cases the passive morpheme is not present:
(164) | a | I saw [the treaty signed] |
b | they heard [the charges read out] | |
c | we felt [the earth moved] |
In these cases, there is an ambiguity that must be checked for. For example, what was seen in (164a) could either be taken to be an object (the treaty that was signed) or an event (the treaty being signed). It is the latter interpretation that is relevant here as this clearly involves a predication-like structure that simply lacks tense, similar to the examples in (161). Of course, the important observation is that here we have a passive construction involving a passive morpheme, but no passive auxiliary. This indicates that the function of the passive auxiliary is to bear an inflection rather than to add any semantic content.
Given the similarity between the passive construction and those constructions involving aspectual elements, it seems likely that they should receive a similar analysis. From this perspective, it is the aspectual morpheme that carries the semantic content and the associated auxiliary is merely a dummy inserted to bear another morpheme that the verb is prevented from bearing by the aspectual morpheme itself.