When people discuss accents and dialects, they usually talk about consonants, vowels, diphthongs and the like. I know I do. But what about the musicality of an accent? We all seem to have a unique rhythm to our speech, a kind of linguistic “time signature,” if you will. Is there some way of quantifying this? And does an accent’s rhythm mean anything?
As a matter of fact, there is a way of measuring speech rhythm, although it’s an imperfect science. Linguists refer to this as timing (or to use a more scientific term, Isochrony). And when it comes to the “timing” of accents, there is a lot we can learn through measuring these differences.
Linguists divide language rhythm into two categories*:
1.) Syllable-timing. This means that each syllable in a language has roughly the same duration.
2.) Stress-timing. This means that
syllables which are stressed are longer in duration than those which aren’t stressed. [Ed. note: I have tried for hours to come up with a way of putting this into my own words and this definition never comes out right. So (sigh) I’m going to resort to quoting Wikipedia: “In a stress-timed language, syllables may last different amounts of time, but there is perceived to be a fairly constant amount of time (on average) between consecutive stressed syllables.”]
For example, compare Spanish, a syllable-timed language, with English, a stress-timed language. Spanish Yo Hablo español muy bien tends to have equal duration for each syllable:
Yo – ha – blo – es – pa – nol – muy – bien
Whereas in English,
stressed syllables are longer than unstressed ones, so “I speak English really well” is more like:
I speak EENG-lish really WEEELLL.
While all native English accents are basically stress-timed, for English accents with some “foreign” influence, the difference between stressed and unstressed syllables may be less pronounced. A study at North Carolina State University, for example, found that speakers of Jamaican English (influenced by West African languages) and Chicano English (influenced by Spanish) fall into a kind of “middle zone” in terms of timing. They aren’t as syllable-timed as Spanish, but they are less stress-timed than more “standard” varieties of English. To put that into plain English, these accents maintain a bit of a “foreign” rhythm: each syllable is weighed more equally than it would be for most English accents.
To see what I’m talking about, compare the accent of Bob Marley … with the accent of, mmm let’s say, Margaret Thatcher? Contrast the more steady rhythm of the former to the LONG-short-short-short-LONG-short-short pattern of the latter.
So what can we learn from findings like these? One thing they illustrate is how these “foreign-influenced” accents can change over time. In the same Duke study from above, the researchers compared modern-day speakers of African American Vernacular English with the recordings of ex-slaves before in 19th-Century. Their findings:
This result suggests that a difference in rhythm may have existed between earlier [African American English] … and that AAE has become more stress-timed over the years.
So, then, African Americans, who once spoke with rhythmic patterns closer to African languages, later adopted the timing patterns of more mainstream varieties of English. Perhaps we can use this to gauge how far a dialect or accent has diverged from its foreign origins?
This raises interesting questions about how other languages play a part in shaping English accents. Which features of another language remain part of an accent long after the foreign language has “left the room?” And which of these features are lost?
*I’m excluding the third division, mora timing, a considerably more complex phenomenon that I won’t go into here.
**Similarly, linguists Eivind Torgerson & Anita Szakay, in their study of Multicultural London English, found that people with this accent have more syllable-timing than other Londoners. Apparently this feature spreads!
Whereas in English, stressed syllables are longer than unstressed ones
I don’t think this is correct, at least in my accent (near-RP now Californicated). Stress in English is usually realized as increased loudness and/or as intonational peaks and troughs, not necessarily as extra length.
Stress-timing means that stressed syllables usually occur at regular intervals. When I say your sample sentence, it comes out like this (with each vertical bar representing an approximately equal amount of time):
| I | speak | ENG | lish | RE-al | ly | WELL
There is an alternation of stressed and unstresed syllables, except for RE-al-ly, where there are two unstressed syllables in a row. I end up shortening the first two the first two syllables have to be shortened to fit into the rhythm. So I actually end up shortening a stressed syllable, as well as an unstressed syllable, to get the stress timing.
In my speech, length in vowels is sometimes phonemic (“Mary” is longer than “Merry”; “caught” is longer than “cot”; “balm” is longer than “bum”).
There is a whole heck of a lot of controversy about this. I tend to go with the accepted notion of “stress timing” because stress timed languages like English or Portuguese (European) have a lot of vowel reduction. The controversy comes into play because between stressed and unstressed vowels that are NOT reduced, research found that there is limited difference in duration. At least from my understanding.
Do you have a source for the claim that “the accepted stressed notion of “stress timing” states that “stressed syllables are longer than unstressed ones”?
All the sources I have seen claim that stress timing means that stressed syllables occur at approximately equal intervals, which is not the same thing!
You know what? Crud. Looking over the original post, that definition I provide really does say something different. Sorry, I’m rather new to isochrony! My confusion derives from the fact that the original definition of “stress timing” appears to be wrong, so I’m a little perplexed about it’s current usage.
The implication of the stress timing seems to be that stressed syllables are longer in duration than unstressed, because the point is essentially that if you said “Jim and Dave” and “Jim and then it’s Dave” the words between “Jim” and “Dave” would take up the same duration regardless of how many of them there are. Furthermore, most explanations of stress-timing I’ve read also claim that such languages or accents exhibit tremendous variation in syllable duration.
I’m a bit confused, though, because (a.) later research apparently disproves this theory, but (b.) the idea that some languages/accents have greater or lesser differences in duration between syllables is demonstrable. I’ll stop here, before I get in over my head. But I’m curious to know how “stress timing” is used currently, since it’s original definition seems to have been rejected!
In teacher training sessions, I use three blind mice to illustrate. The first line “Three blind mice” has one syllable per beat. The you have “they run” two syllables in one beat. You eventually get to 4 syllables in one beat, ” with a carving”.