In older poetry, the words "eye" and "die" are often rhymed with words that today end in an "ee" sound. Were "eye" and "die" pronounced like "ee" and "dee," or were words like "charity" pronounced "charit-eye"?

by ReaderWalrus

Anyone with even a passing interest in poetry (like me!) will have seen at least a few examples. Three off the top of my head are "Death, be not proud" by John Donne (1633; rhymes eternally/die), "The Tyger" by William Blake (1794; rhymes symmetry/eye), and "Sonnet—To Science" by Edgar Allan Poe (1829; rhymes eyes/realities).

Which words changed pronunciation, over what period of time did it occur, and how can we tell? (I assume Poe uses it as a deliberate archaism, but does Blake as well?)

Cowtheduck

Historical phonology is not my area of expertise, so apologies if this answer is not as comprehensive as you might expect from a flaired user, but as a theatre lover I do have a passing interest in Original Pronunciation (English as it was pronounced in Shakespeare's day), so hopefully I can get the ball rolling (with the view that someone more knowledgeable would expound).

Almost certainly in Donne's time, "eye" was pronounced similarly to the present-day pronunciation, while "eternally" would have been closer to "eternal-lie". The Oxford Dictionary of Original Shakespearean Pronunciation (OUP 2016; edited by David Crystal, an expert on the historical phonology of Early Modern English), which compiles the reconstructed pronunciations of words c. 1600, lists the following entries:

eye - ǝɪ

eternally - ɪˈtɐ:ɹnɑˌlǝɪ

You don't have to know the International Phonetic Alphabet (IPA) to observe that they have the same ending! Though note that the present-day pronunciation is very slightly different: aɪ rather than ǝɪ for eye (I suppose a very rough and ready description would be that ǝɪ would be closer to how someone with a strong modern Irish accent would pronounce "eye".)

As for the reconstruction process of Elizabethan pronunciation, Crystal has this to say in the foreword to his dictionary:

A reconstruction exercise is by no means straightforward because the textual evidence is often difficult to interpret. A distinctive spelling may genuinely indicate how a word was pronounced, or it may be a typesetter’s error. Words at the ends of lines may point to a genuine rhyme (as in a sonnet) or may have a fortuitous connection... The possibility of eye-rhymes [words that end with the same letters but are pronounced differently, like Sean Bean] also needs to be considered.

Notwithstanding, Crystal observes that "there are a sufficient number of clear cases of spellings, rhymes, puns, and comments to warrant a reconstruction of the English sound system of Shakespeare’s day". For a more comprehensive review of the methodology of phonological reconstruction, he recommends Eric Dobson's English Pronunciation 1500–1700 (2nd edn, Clarendon Press 1968), which unfortunately I don't have access to. As such I'm afraid I can't answer the second part of your question, that being when the pronunciation changed to its current form. Hopefully someone more knowledgeable will chip in with a fuller answer.