This clearly happens, as evidenced in the different accents in different regions of the country. But it has also happened to entire countries as well, e.g. the Great Vowel Shift.
My question is why this happens, and why this would vary by region. Is it solely due to the influence of other languages? Or what else might be involved?
I’m under the impression that linguists are not in accord.
I once read a delightfully geeky science fiction time travel story in which some linguists went back in time to study it, and one team comes back almost incapable of communicating: “Hyulf! Vwer gut koht in deh shibt!”* – i.e., that the Great Vowel Shift was caused by some immediate wavelike phenomenon that went ripping through villages instantly changing people’s speech patterns on the spot!
[sub]* or something like that[/sub]
I’m sure there are many reasons. Fromm’s Law seems to work without any outside influences; it’s just that G>k seems to just happen because of how we talk.
Others are influenced by other languages or other words.
Among the possible explanations are:
laziness: What happened to the first r in February. Oh, you say it? I don’t. What happened to the w in what?
harmonization: Some people say “cot” for “caught”. It seems likely that the back c is attracting a further back vowel. The o in the old Germanic word foti for foot was fronted to foot in English and a similar vowel fuss in German presumably dragged forward by the f.
Error: nucular for nuclear. Errors by high mucky mucks tend to propogate.
Reading pronunciation: Traditionally, the t in often was not pronounced and my dictionary doesn’t even offer of-ten as an alternate. But in Canada, no one says of-fen (well I do, but I am an outlier) and a great many Americans do too.
None of these explain the great vowel shift though. I think that is pretty much of a mystery. At least I have never heard any explanation.
I don’t understand any of these explanations.
All of these are explanations of why it would be easier to pronounce things one way rather than the other. None of them seem to be explanations of what changed.
If there are reasons for why people subsequently came to pronounce things in a certain way, then the question is why they originally pronounced them the original way? Why wouldn’t the same factors have been present at the outset?
That’s why the only thing I can think of is the influence of other languages, since it’s determined by non-language-related historical events, like migration, travel, wars etc.
When I was young I rarely heard anyone say OFF-ten, now I hear it more OFF-ten then not.
Same thing for alms and palms and so on, I always hear the “L” now.
The public acceptance of rap music as mainstream pop will go a long way to changing acceptable pronunciation. Now let’s pwarty
They probably were but the ‘starting’ pronunciation was different.
That seems a bit circular. The whole question here is why things changed to begin with. Any explanation that relies on the starting pronunciation being different boils down to saying that some changes were caused by other changes, and leaves the original question unanswered.
A related question is: Given the following general tendencies
[ul][li] t –> ts[/li][li] ts –> s[/li][li] s –> z[/li][li] z –> r[/li][/ul]
why haven’t all the T’s in all the languages changed to R’s ? 
The change z –> r may be rarish, but it’s much more common than r –> z, and human language has had tens of thousands of years to approach equilibrium.
These are fascinating questions and the full answers must be complicated, but part of the answer is very simple: The tendency to simplify language (in part due to oral laziness) reduces the information content of human speech. Therefore those “lazy” changes must be countered by a need to make new informative distinctions.
It’s cuz the kids want to sound all cool. And some of it sticks.
Well, partly.
Origins of English: The Great Vowel Shift
Two other points from those articles. The effect wasn’t random. All the vowels changed in an identical way. And the change was very slow, taking centuries.
You see some of the same changes within the U.S. The huge in-migration from the North to the South after WWII has lessened the difference in Southern accents from one another and from those in the North.
When I was in grade school ('66-72), the teachers were high priestesses of the Funk and Wagnall’s. I remember having to look up plantain and kiln and victuals. All of these are easily mispronounced if one tries to pronounce them as they’re spelled. Schools don’t seem to drill with dictionaries much any more.
For those of you wondering, the last syllable in plantain is pronounced as in curtain, victuals is pronounced vittles, and the N in kiln was silent back then. I still have that dictionary.
In modern times I would say it was the influence of TV and film. In England we see a great deal of American film and TV and ‘Americanisms’ are commonplace these days. Not everything gets taken up though: one frequently hears and uses ‘guess’ instead of ‘think’ or ‘assume’, while ‘gotten’ is rarely heard other than in the phrase ‘ill gotten gains’…
This is a huge part of it in Australia - pretty much every young person (under 25 or so) that I’ve encountered lately pronounces the letter “Z” as “Zee” instead of Zed and many of them use American spellings for words (color instead of colour, flavor instead of flavour, etc) as well - all thanks to a combination of US TV shows, movies, the internet and default language settings on popular word processing programmes.
Here’s another thought that occurred to me.
I imagine there are some sounds and types of sounds that are harder to pronounce if you don’t have a lot of practice, but easier if you do. So I wonder if changes in lifestyle, due to technology or whatever, can themselves cause certain words to be used less often, and with that, become harder to pronounce, which could trigger changes in pronunciation.
One thing to which I attribute changes in pronunciation is what happens to words when the person saying them hasn’t heard them - or doesn’t realize they have. If there is no evident phonetic correspondence, you end up with people who pronounce chastity three different ways*, and who look funny at the foreigner that says “maybe we should ask a priest… what? It’s a virtue, they should know how to say it!”
But that is at the level of “personal error” so long as it’s only one. As it spreads (and then the question would be how and why), it becomes a dialectal variation and maybe even the new standard pronunciation.
Not directly relevant to the question, but that’s also sometimes true in the other direction in the US. Not Zed (that’s more likely to remind Americans of the line in the movie Pulp Fiction, ‘Zed’s dead’ 
) but various anglicisms are more common. It’s fed by internationalized mass media, which isn’t all in the direction of spreading American norms to the other English speaking countries, and mobility of people. It’s particularly common IME in the financial industry in NY, where lots of people often speak to Brits or spend time in London, ‘cheers’ as a phone sign off for example, never heard decades ago in the US.
Formal pronunciations have changed a huge amount since the 1940s. I found a famous radio announcer’s test for pronunciation. A large percentage, maybe even a majority, are no longer pronounced the way they were formally preferred back then.
SAHN-toh VOH-chay isn’t a typo? They actually throw in an N? The rest made sense, but that one is weird.
Notwithstanding all the above theories, I think the OP’s question is at the core of the puzzle that theologists, philosophers and later linguists have tried to solve since the beginning of civilisation: why are there different languages? The answer is both simple and complicated: because no one speaks his language exactly the same as another speaker of that language, although the differences may be too small to be observed. But over time and under influences mentioned (bottlenecks/founder effects, other languages, fashion) ways of speaking slowly grow apart. I assume you are all familiar with the Indo-European language family with common words as far apart as ‘quarto’ and the like in Roman languages and ‘four’/‘vier’ in Germanic languages which apparently share a common ancestor. (I’m not a linguist myself so please don’t quiz me on the Farsi or Urdu)
My point, in short, is that there is no why. Changes in pronunciation (and in exact meaning of words btw) is simply a characteristic of human languages. And, come to think of it, not human only: apparently there are species of songbird whose calls have regional peculiarities which are learnt, not genetic.