It's a tasty topic.
True, if language learners interpret 'would' as the tail-end of a temporal ordering of predicates, which they do initially, and would have cause to continue to do -according to your stance - given the 'traditional' (as well as modern) definition, would, past tense of will.
Originally Posted by DBP
But why limit the sematic scope of of the term 'tense'? That's what I don't understand. Surely there's another way to interpret 'tense'. What about, 'would', pst tense of 'will', 'tense' in the sense of a temporal ordering of utterances?
'would' is defined as a past tense form, that's a given, but the term past tense, especially when it applies to auxiliary 'would', doesn't refer to the temporal ordering of predicates, which is where we agree. Evaluating reported speech, which is where we find modal-auxiliary 'would' is based on "polarity", as I am sure you would also agree. Given root/epistemic modal evaluations, modal-auxiliary evaluation time (ET) is understood as coinciding with the actual utterance time (UT). In other words, Max reports her root-modal evaluation 'would' on the basis of the actual utterance time; i.e., when Max said the words:
Max: "She said she would help." (UT= now; ET = then, hence 'would')
In short, the term "tense" refers to time: event time ('said'), and utterance time ('would'), so use the term 'past tense' to describe modal-auxiliary 'would', even 'could', 'might', and so on.
If you're interested, here's a wee bit of related background for you:
When modals [can/could] are used to convey the root-modal senses of ability and permission, they participate in a semantically viable present/past tense alternation, just like normal verbs.
a. Carl can't move his arm (ability at UT)
b. Carl couldn't move his arm (ability at a past time).
At the utterance time (UT), it is not possible for Carl to (habitually) move his arm. In (9b), could functions as a past-tense form of can in (9a); at some time prior to the utterance time, it was not possible for Carl to move his arm.
When could is used epistemically in simple sentences, it cannot have a past-tense interpretation:
a. Jack's wife couldn't be very rich.
It is not possible that Jack's wife is very rich.
*It was not possible that Jack's wife was very rich.
The speaker reports his or her epistemic modal evaluation holding at the actual utterance time. To force a past-tense reading of (10b) it is necessary to construe could as a root modal involving ability or permission.
When the modal is understood to have root-modal force, it is interpreted as though it falls under the semantic scope of the past tense, but when it is understood to have epistemic modal force, the past tense is interpreted as though it were a (non-finite) perfect occurring in the complement of a present-tense epistemic modal; that is, the modal evaluation time must be understood to coincide with the actual utterance time.
In Stowell (1995), I suggested that the so-called present and past morphemes in English are not actually present and past tenses per se (where tenses are understood as temporal ordering predicates) but rather polarity markers on time-denoting heads designating a particular scope relation with a higher (true) past tense.
Tense and Modals
, Tim Stowell