Tense and aspect in corpora

The preference for lexical work in corpora is understandable given that corpora consist of words much more visibly than they consist of structures (syntagms). It therefore depends on your corpus being tagged and lemmatised, and on how your concordancer can search your corpus.

If it uses a corpus query language (CQL), it is quite straightforward to search for instances of the perfect because this consists of the verb HAVE (in any form) followed by a past participle: [lemma = have][tag = VVD]. Similarly, the continuous can be found by searching for any form of BE followed by an -ing form: [lemma= be][tag = V.G]. And space can be made in these queries for discontinuous forms, e.g. I'd HARDLY EVER tried it.

These are just two examples, and they can be searched because of the periphrastic nature of English, that is, English uses auxiliaries (words) rather than inflections to create aspect.

If our approach involves using corpus data with students to discover features of tense and aspect that they can apply in their own language production, CQL is clearly of considerable value. However, without such well-endowed corpora and concordancers, we can still look at the lexical environments of these forms.

One productive area is adverbs of time: a search for already in CLT's concordancer reveals a great deal of perfect aspect, mostly present perfect.

A search for never in the corpus of Presedential speeches returns 806 lines, with a great number of HAVE immediately before it, and also many modals. Following never, there is a great number of BE, including been in the present perfect passive chunk have never been.

At the bottom of this CLT interface, there is a useful box of left collocates:

to=19 would=11 not=8 might=7 we=7 can=6 cant=5 should=5 you=5 cannot=4 didnt=3 and=2 could=2 may=2 must=2 never=2 people=2

We can search for more specific time adverbials, such as tomorrow, which has very few instances of going to after it compared to will. Yesterday co-occurs much more frequently with be than with have. Students can observe the use of be and have as lexical verbs or as auxiliary verbs, and if the latter, which aspect is used in the company of yesterday.

This approach is clearly guided discovery, and while the above discussion is rich in terminology, sessions with students do not need to be. Neither do they need to be at a computer. Pages of concordances can be printed out with specific tasks attached to them. They could be pinned on the wall for students to wander around and complete the tasks and present their findings.