As instance, the sentence
let it be is pronounced /lɛt it bi:/ .
And in general, the
e seems to me pronounced just arbitrarily.
Does it come from Germanic languages? During language evolution, did different spellings converged to one single way to write?
As instance, the sentence
In Middle English (and even earlier), English scribes had a problem: they had a lot of vowels to represent, and not very many vowel letters to write them with. There were seven long monophthongs
/iː eː ɛː aː ɔː oː uː/ and five short monophthongs
/i e a o u/, and only six vowel letters in the Latin alphabet (A E I O U Y).
To solve this, they used multiple letters to write the long monophthongs—after all, they're pronounced longer, so it makes sense to use more than one letter. The convention became to spell these seven monophthongs as I_E, EE, EA or E_E, A_E, OA or O_E, OO, and OU. (The underscore here represents a single consonant. Spelling
/raːt/ as RATE looks a bit odd, but it came from earlier
/ratə/, thanks to a process that deleted certain final vowels and lengthened the vowels before them.) The short vowels were written simply I E A O U.
Then a major sound change happened near the start of the Modern English period: the Great Vowel Shift, which rearranged all the long vowels.
Around the same time, the length distinction was lost, replaced with a quality distinction: formerly "short" vowels were pronounced a bit closer to
[ə]. Unfortunately, these two changes happened after the introduction of the printing press, when the orthography was becoming standardized. So the ways of writing the vowels didn't change as their pronunciations did.
The result is that "short" and "long" vowels (no longer distinguished by length) are written in strange and unintuitive—but mostly consistent!—ways in Modern English.
This, as with so many pronunciations of English vowels that are counterintuitive for speakers of many other languages is due to the Great Vowel Shift (see the diagram below, from Wikimedia).
Time progresses as we go up the chart, from 1400 at the bottom to a reasonably modern pronunciation at the top. Along the bottom we see sample words, and at various stages in the chart we see IPA transcriptions of the pronunciation of the vowel in those words. Note that some of the short vowels that remained largely unchanged have been omitted.
As we can see, around 1400 the vowels were (for the most part) pronounced pretty similarly to how they were written, but over time some of diphthongs monophthongised into long vowels, and most of the long vowels (including some of the former diphthongs) broke into new diphthongs.
The long /e:/ in Middle English "been" (which gave Modern English "be") did not break, but instead rose to Modern English /i:/ filling the gap left after Middle English /i:/ broke into the diphthong /əɪ̯/ (giving today's /aɪ̯/ in "time").
Meanwhile the Middle English short /e/ in "let" remained relatively unchanged, with maybe just a slight lowering from [e] to [ɛ].
The Great Vowel Shift is why "short" and "long" vowel pairs in Modern English no longer obviously correspond. The "short" vowels haven't changed much, but the "long" vowels have (mostly) broken into diphthongs and moved around.
Pronunciation of is also not as arbitrary as it may seem for someone used to a language with a more transparent orthography (such as Spanish). In 2000 Mark Rosenfelder came up with a set of 56 rules that predicts the correct pronunciation of English words in General American ~85% of the time (more information available here).
Predicting the pronunciation of is simpler than that though. Avoiding digraphs like , , and following /r/'s (which vary more between dialects and are a little more complicated) and setting aside vowel reduction (which is a whole nother kettle of fish), the spelling is pretty good at telling you whether the vowel is "long" or "short".
If the is at the end of a word and there is another vowel earlier in the word (e.g. "mate"), it is almost always silent, and serves to clarify the pronunciation of an earlier segment.
If the stressed is in the final syllable, followed by a consonant, and without a silent at the end (e.g. "bet"), it is "short". If followed by a single consonant and a silent (e.g. "Pete") or word-finally (e.g. "be") it is "long".
Elsewhere, if followed by two or more consonants (e.g. the first in "execute", noting that represents the cluster /ks/), it's probably "short", otherwise (e.g. "redo") it's probably "long".
Modern English "be" derives from Middle English [be:n]. Here is the entry for Middle English. There was a historical sound change where [e:] changed to [i:], [æ:] changed to [e:] and so on – the Great Vowel Shift. This is why English spelling is so peculiar, given pronunciation, from the continental perspective, and why we call the letter "a" [e:], and "e" is [i:].
The etymological source of "be" goes back to Germanic and ultimately Indo-European *bhuH (reconstruction uncertain).