How long does it take for human languages to fork?

May 13, 2006 7:52 PM

linguists: i vaguely remember a story that i can't verify. the gist is this: two groups of people who spoke the same language became separated by some minor geographical obstacles, like hills, and remained so for a period of years (decades, i think). upon analysis some years later, their common tongue had forked noticeably into fairly different languages. is this true?

i want to say this is an american story, and more specifically appalachian -- that would make sense, since the course i'm recalling was in colloquial language and i was studying at the university of tennessee at knoxville. but i'm not sure if the story is factual or not. enough time has passed that i'm not sure if this is getting muddled in my memory with something else.

does this sound familiar to any of the linguists in the house?

thanks in advance. this has been tickling at me for some time.

posted by patricking to writing & language (17 comments total)

Your citing of the Appalachians suggests you're thinking of the Melungeons, but I don't think their history supports linguistic bifurcation.
posted by SPrintF at 8:37 PM on May 13, 2006

Are you sure this was a story about a specific historical event? I ask because it sounds to me exactly like the kind of hypothetical example given in historical linguistics classes to show how dialects develop.
posted by nomis at 9:03 PM on May 13, 2006

To address your question "how long does it take for human languages to fork?", this varies depending on the circumstances of the two groups (are they completely separate or do they still have some contact?), whether they have any writing system (which can allow them to keep records of older forms of the language) and even how you define "fork" - do you mean they end up with different accents? A few different words? Two mutually intelligible dialects of one language? Complete mutual unintelligibility?

Having said all that, it is generally accepted that the process of separating into anything more than two slightly different varieties would take centuries rather than decades.
posted by nomis at 9:12 PM on May 13, 2006

One of my linguistics professors liked to say that if you magically separated Americans, Canadians, Brits and Aussies for 200 years, you'd have four completely different languages.
posted by frogan at 9:20 PM on May 13, 2006

Of course, this is one of the reasons why Hindi and English aren't the same language. Your example is really hypothetical and not really specific. This sort of thing has been going on for a long time.

Although, with the advent of communication technology, I'd bet that it's less prevalent than before.
posted by taschenrechner at 9:30 PM on May 13, 2006

I'll give you a random example of Pittsburgh, PA.. it's largely separated from major surrounding areas, while still a major city, and evolved a dialect of its own. It's really just one of many as people have noted above.
posted by kcm at 9:35 PM on May 13, 2006

If there is any truth to the claim, it would have to have happened within a small population like a small Native American tribe.

Children are the main force in language evolution. They miss hear things, they try to rationalize what they speak, and they invent slang so that their subgroup has a solidifying secret language.

If a group of 100 people was split in two it would take only 2 generations for there to be strong differences in their speech.
posted by MonkeySaltedNuts at 9:40 PM on May 13, 2006

I don't have any academic reference, but a Korean friend told me that the languages of North and South Korea have diverged a lot since the Korean War and the ensuing break between the two societies. There may have been some north-south dialect differences already, but she thought the imposed (one assumes rather Orwellian) vocabulary and attitude difference in the north has been something of a pressure cooker for language change.
posted by zadcat at 10:00 PM on May 13, 2006

Language divergence through geographical isolation has happened a lot through out human history. It is one of the reasons that there are so many languages in areas of rough terrain. If there are mountains or other physical obstacles between groups of people it is much harder for them to remain in contact. Also, roads and commerce make a huge difference, as does colonization or strength of national identity.

Take a look at these language maps of New Guinea and Asia. Look at the terrain of the island shared by New Guinea and Indonesia. It is made of rough country, too rough even for roads and colonization to have made a dent in the number of languages. This area has the highest concentration of languages in the world. Also, notice the Asian map, the clusters of lots of languages correspond with mountainous regions or underdeveloped regions with thick forests or swamps. Notice how the distribution of languages changes over the border of a country with a prominent national identity like China. The number of languages is greatly diminished. (If you think this is interesting check out the World Atlas of Language Structures.)

The rate of decay depends on isolation and what you would call a "separate language". There is actually a lot of controversy about what constitutes a separate language. What is the maximum number of words two dialects must differ before becoming two separate languages? A number I have seen a few times is 20% (check out the first chapter of Language Death by David Crystal for a better handle on this). This is further complicated by divergence in accents and spelling (if a language has a writing system). Also, one must consider how much the grammar has diverged, but that is a bit too complicated for this exercise.

The only real number for divergence I have found is a loss rate of about 14% per millennium and a chance similarity rate of 5% (which sounds a bit high). With no writing system (writing slows change) I believe that the loss rate would be much higher, as much as 25%. So, How long does it take for human languages to fork? My answer is a little less than a millennium depending on conditions.

For a less half-assed calculation check out "The Logical Problem of Language Change"

PBS also did a series on language change in America.
posted by Alison at 10:46 PM on May 13, 2006 [1 favorite]

In college I read Albion's Seed which addresses how different dialects have grown in different regions of the east coast and southern US. If you listen to middle Virginia accents they sound vaguely British to this day, while here in the deep south our broad accents can make us sound uneducated even when we're not.

It follows the same hypothesis of people being separated for generations and their speech patterns becoming markedly different from each other.
posted by hollygoheavy at 5:10 AM on May 14, 2006

No. It takes a lot longer than that.
posted by languagehat at 7:16 AM on May 14, 2006

You might find some information by trolling the University of Pennsylvania. They've been studying the changes that have occured in the Philadelphia accent for the last thirty-odd years. While not geographically separate from other areas, the Philadelphia accent has changed dramatically and continually over that time.
posted by Spoonman at 7:20 AM on May 14, 2006

For some off the wall reason I connected this with "Snow Crash." I think Juanita said something like this on the Raft.
posted by Fejery at 8:23 AM on May 14, 2006

Zadcat -
Yeah, North / South Corean is differentish, partly because of preexisting regional dialects, partly because of the North's resistance to foreign words (in seoul, we say "computer" in pyeongyang, the have a Corean word instead), and partly because of the separation.

Very far from different languages, however.
posted by Joseph Gurl at 9:06 AM on May 14, 2006

okay, so it's sounding like i'm remembering a hypothetical situation. thanks for all the responses.

quickie to SPrintF: my mother does a lot of research on melungeon culture, so we have a lot of information about those folks lying around. no language bifurcation, you're right.
posted by patricking at 2:50 PM on May 14, 2006

What is the maximum number of words two dialects must differ before becoming two separate languages?

Sorry, that should read minimum number of words.
posted by Alison at 3:00 PM on May 14, 2006

Here's an example of a split: Dutch vs. Afrikaans. They are very similar, but sometimes use the same words to mean very different things. (The Afrikaans word for highschool translates as whore-school, rest-area translates as a place to wank! Two very funny ones we saw right in a row.)
posted by Goofyy at 5:00 AM on May 15, 2006