About the SA Blog Network

Can Geography Shape the Way We Speak?

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint

View from Devil's Path east of Indian Head Mountain in Catskill Mountains. | Image, CC by Miguel Vieira. Click for license and information.

There’s a pretty neat series of maps featured in Business Insider making the rounds on various social channels. They tell us about ourselves, showing how Americans speak differently from each other. My favorite is slide 17 because I have only ever owned sneakers and was absolutely confused when someone recently suggested that my concept of sneakers was really a generalization of “athletic shoes.” (I still think they’re sneakers. All of them.) It’s a treasure trove of linguistic diversity revealing which everyday words in my lexicon vary (e.g., garage sale) and which seem fairly common (e.g., subway) to the overall American experience.

The maps are the work of NC State Ph. D. student Joshua Katz, who drew on a linguistic survey on the ways Americans pronounce words—you can view the original set of maps here. Regional variations in language are well-studied. Every speaker can be identified by region, social class and gender. Accents are incredibly revealing, which is why some people take great pains to hide theirs even while others use it to weave an identity. These identities reach beyond personal definition to explicitly include regional and social histories and cultural nuances.

However while accents can be hidden or faked, the sounds that we’re able to make may not be so readily manipulated. A recent study published in PLOS One shares evidence that geography may play a part in shaping these sounds. Anthropologist Caleb Everett analyzed 567 language locations and found a commonality that crossed dialectical boundaries and language families: languages with ejective phonemes tend to occur at higher elevations throughout the world.

What are ejectives? They’re a basic unit of sound—a type of phoneme—which when combined with other phonemes create words. Ejectives are unusual in that they’re non-pulmonic. They’re produced by the closing of the vocal cords. It is the vibration of the vocal cords that allow us to meld phonemes together to generate words. These sounds are essentially voiceless. They’re hard to describe: it’s almost as though you’re trying to make the sound of a consonant while holding your breath. They’re really better understood when you can hear them for yourself:

Fifteen percent of the habitable space in the world is at altitudes higher than 1500 m (approximately 1 mile) above sea level; these spaces are home to ten percent of the world’s population. Everett divided the 567 languages he tapped for his study into two groups: those having ejectives (92) and those without (475). Among those languages with ejective phonemes, sixty-two percent are found at higher altitudes. The remaining languages with ejective components are disbursed, but the greatest concentrations are seen within 500 km (approximately 300 miles) of these linguistic peaks.

Figure 1. Plot of the locations of the languages in the sample. Dark circles represent languages with ejectives, clear circles represent those without ejectives. Clusters of languages with ejectives are highlighted with white rectangles. doi:10.1371/journal.pone.0065275.g001

Everett believes the conditions at higher altitudes may encourage the production of ejective phonemes. These utterances require that the vocal cords are closed and raised. Everett holds that this is easier to accomplish at higher altitudes where atmospheric pressure is lower, which means that air pressure in the mouth and lungs is lower so it may be easier to force the vocal cords closed. Everett also proposes that the higher incidence of ejective phonemes at higher altitudes may represent a biological adaptation. With ever word uttered, we’re generally exhaling during some portion of the utterance. This exhalation also releases water vapor, which Everett informs us is not a trivial matter. Apparently, we lose up 400 ml of water vapor though exhalations. At higher altitudes, this can lead to dehydration and severe forms of altitude sickness. Because ejective phonemes require the vocal cords to the closed, they’re also not drawing on air contained within the lungs and therefore releasing water vapor—in this way, ejective phonemes could be a biological adaptation to an extreme living condition.

However, not everyone is quite sold on Everett’s analysis. Over at Language Log, a blog run by the Research for Cognitive Science at the University of Pennsylvania, Dr. Mark Liberman raises some concerns over the statistical work done in this paper cautioning that this may be a causal effect. Liberman’s primary concern appear to be that data can be molded to tell any story, but acknowledges that there is independent support for Everett’s findings.

Given the amount of attention Katz’s maps have received, we are definitely interested in exploring our linguistic differences. In the case of languages that are in isolated, as those with ejective phonemes, they will likely attract a fair amount of attention as we work to better understand ourselves and our differences. Do you have a accent or dialect story to share? Or are you an ejective speaker? How has your locality influences how you speak—both physiologically and acoustically?

Krystal D'Costa About the Author: Krystal D'Costa is an anthropologist working in digital media in New York City. You can follow AiP on Facebook. Follow on Twitter @krystaldcosta.

The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions

Comments 3 Comments

Add Comment
  1. 1. AntonioMario 6:40 pm 06/19/2013

    Hi Krystal,

    Very interesting article.

    There’s one thing that puzzles me. I live in Sao Paulo state’s countryside and people (here and other parts of the countryside in other states) have that characteristic ‘r’ sound, one that is spoken with your tongue curled up more than usual and for longer than a city dweller say.

    Now, that sound is essentially the same as the one that you hear in the Western US. Listen to America’s ‘A horse with no name’ (yes, I’m 60+… ;-) ) when the lead singer pronounces ‘… the *air* was full of sound…’; that ‘air’ has exactly the ‘r’ sound I’m talking about. Some of the Eagles’ lyrics show that too. So two completely (geographically speaking) separated areas but same accent. Hmmm.

    So, that whole thing got me thinking whether people that live in rural areas, at least where Indo-european languages are spoken (or is that just in the Americas?), even across different countries, have sounds that might be similar and that that has got to do with their environment, rural in this case.

    I’d like to know what you think. Just coincidence?


    Antonio Mario

    Link to this
  2. 2. Krystal D'Costa in reply to Krystal D'Costa 12:27 pm 06/21/2013

    HI Antonio,

    You raise an interesting question, and unfortunately, it’s not one I can answer definitively. You’ve cited mostly songs for this similarity, and I think there is a tendency in music to exaggerate certain vowels and consonants to make them clearer (I’m drawing on my middle-school choral training there). So in this case, it could be a coincidence. However, there is something to be said for the spread of languages themselves and the overall Westernization of language wherein people adopt Western modes of speech as a means of better identifying with a dominant culture, which in this case would the US as it’s in the same hemisphere. The immediate counter to this argument is that people would tend toward speech as it’s represented in media (movies, music, etc) and that would lead to a generalization of how things should sound and not link to a specific region.

    There’s still a lot of work to be done in terms of how environmental factors may shape speech–so regional weather, regional flora and fauna, and many contextual details need to be considered.

    This will be an interesting realm of work to follow.


    Link to this
  3. 3. AntonioMario 2:37 pm 06/24/2013

    Hi Krystal,

    Thanks for your insights.

    You raise the argument for a possible influence of a dominant folk culture, as that of the US pop music. I tend to agree with you that there would indeed be counter-arguments to that.

    Our local, countryside ‘r’, like in ‘porta’ (door, in Portuguese) in which country people here put an emphasis on that ‘r’, really is a recognizable accent of old, predating say the American influence, for example, which picked up only after the 40′s or so.

    As you say, an interesting topic of study, no doubt.

    Look forward to your future articles. I don’t have FB, only Google+, unfortunately, but will keep an eye on the SciAm newsletters.

    Thanks and best regards,

    Antonio Mario Magalhaes

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article