You can watch the rise of modals in English if you survey old texts and actually jot down data. This has been done. Numerous times. And oft published. What's varied over time isn't the data but the analysis hung onto the data. There's a huge literature on actualization that often delves into English modality.
You also get tense agreement in English out of the rise of the modals. "I knew they would go" = "I knew they were wanting to go", the last bit of which was interpreted as future. Same for may/might (might = past tense of may), can/could. Tense concord creates hell for English language learners--why do you sudden screw with the verb instead of saying, "I knew they will go." Every other language I know just uses "will": "Sabia que iran", savait qu'ils iront, ja znal, chto poidut, vedel jsem ze pojdut, etc. (Span/French/Russian/Czech, the same's true for Serbian, Italian, Polish ... but I only really know Romance and Slavic).
Modals are common. Slavic--also Indoeuropean--winds up with some forms based on esti 'be' or imeti 'have' or even khocheti 'want'. Bg shchu 'I want' is one way of saying "I will." Ukrainian has two imperfective futures, otrymatyme 'he will receive' and 'bude trymaty' "he will receive", but they mean slightly different things (Ukr, like other Slavic languages, distinguishes broadly between verbs that are perfective and have outcomes and imperfectives where no outcome is claimed, which is different from saying there is no outcome). One's more dependent on what he wants and does, the other has a bit more of a modal flavor (look up deontic and epistemic). Notice I've just shifted what "modal" means a bit, from simple tense/aspect marker to modality in the sense of "should" or "ought". (Which are, of course, the past tenses of "shall" and "owe" ... "must" didn't have a good past tense, and that's a nice instance of suppletion.)
Modals are often transient. They're on the way from highly inflected languages (which tend not to have them because they have other ways of saying the same thing) and isolating languages, if anybody still uses that term, which have no morphology. Tibetan vs Chinese, for instance.
Modals are often unnecessary. The future tense is a funny sort of beast. A lot of languages don't have it, a lot didn't have it--at least as far as a formally identifiable thing. You infer it. Like in English; before what, 1100 AD or so English didn't have a clear future tense. You used periphrasis to, well, work around the idea.
Modality is also broader than in English. Check out "evidentials" for fun. In German, evidentiality can be shown by a tense shift; Bulgarian has something like that, where you deny responsibility for something's truth value by shifting it one "tense" back in time. Some nothern South American languages have truly astonishingly complex evidentials, stipulating whether you know something because you heard it, saw it, was told it ... If you include this under "modality" then you probably want to peel off some tense/aspect markers as "auxiliaries" to keep your analysis focused.
= new reply since forum marked as read
