Sequence of Tenses in English Grammar

In English grammar, the term sequence of tenses (SOT) refers to an agreement in tense between the verb phrase in a subordinate clause and the verb phrase in the main clause that accompanies it.

As observed by R.L. Trask, the sequence-of-tense rule (also known as backshifting) is "less rigid in English than in some other languages" (Dictionary of English Grammar, 2000). However, it is also true that the sequence-of-tense rule doesn't occur in all languages.

Examples and Observations

Geoffrey Leech: Most commonly [sequence of tenses] is a case of a past tense in a main clause being followed by a past tense in a subordinate clause. Compare:

(a) I assume [you are going to be late].
(present followed by present)
(b) I assumed [you were going to be late].
(past followed by past)

The interesting thing is that the past tense of the subordinate clause can easily refer to the present time, as in Hello! I didn't know you were here. In such cases, sequence of tenses overrules the normal meanings of past and present tenses.

R.L. Trask: [W]hile we can say Susie says that she is coming, if we put the first verb into the past tense, we normally put the second verb into the past tense as well, producing Susie said that she was coming. Here Susie said that she is coming is somewhat unnatural, though not strictly ungrammatical . . ..

Sequence-of-Tense Rule (Backshifting)

F.R. Palmer: [B]y the 'sequence of tense' rule, present tense forms change to past tense after a past tense verb of reporting. This applies to the modals as well as to full verbs:

'I am coming'
He said that he was coming
'He may be there'
She said that he might be there
'You may come in'
He said that I might come in
'I'll do it for you'
She said that she'd do it for me

Sequence of Tenses With Modals in Indirect Discourse

Paul Schachter: [A]lthough it is true that modals do not inflect for number, there is some evidence that they do inflect for tense. The evidence that I have in mind has to do with sequence-of-tense phenomena in indirect discourse. As is well known, it is generally possible to replace a present-tense verb by its past-tense counterpart in an indirect quotation after a past-tense verb. For example, the present-tense form of the main verb have that occurs in the direct quotation of (3a) may be replaced by the past-tense form had in an indirect quotation, as in (3b):

(3a) John said, 'Little pitchers have big ears.'
(3b) John said that little pitchers had big ears.

Note in particular that the quoted material in (3a) is a proverb learned as a fixed formula so that the change in this (otherwise) fixed formula attested in (3b) provides especially clear evidence for the application of a sequence-of-tense rule.

Now consider in this connection the following examples:

(4a) John said, 'Time will tell.'
(4b) John said that time would tell.
(5a) John said, 'Beggars can't be choosers.'
(5b) John said that beggars couldn't be choosers.
(6a) John asked, 'May I be excused?"
(6b) John asked if he might be excused.

As these examples show, it is possible to replace will by would, can by could, and may by might in an indirect quotation after a past-tense verb. Moreover, these examples, like those of (3), involve changes in fixed formulas (proverbs in (4) and (5), a social formula in (6)), and thus provide similarly clear evidence that the sequence-of-tense rule is involved. It, therefore, seems that the present-past distinction that is relevant to verbs, in general, is relevant to modals as well, with will, can, and may, for example, being classified as distinctively present forms and would, could, and might as distinctively past.