Vowels

I’ve spent the last two weeks buried in research into vowels. That might sound boring to y’all, but I am a linguist by hobby and an English teacher by trade, and so vowels are right in my wheelhouse. This post might run a little bit long, so here’s the tl;dr version: teaching English vowels to Georgians is hard not because English vowels are inherently difficult, but because people on both sides have screwed it up so badly for so long.

Georgian supposedly has exactly five vowels and no diphthongs, although opinions vary on the matter and I personally am of the opinion that not nearly enough research has been done to state anything conclusively. It seems to me that Georgian has a much richer inventory of vowels than linguists give it credit for – that there are indeed five phonemic (meaning-carrying) vowels, but that realization of those vowels in different environments brings the total up quite a bit – perhaps as high as seventeen! That’s a huge gulf to fill, so it falls upon me to explain why Georgian has more than three times as many vowels as Georgians commonly believe.

The example that sort of started me off on this line of research is the Georgian o, written ო. According to most books on Georgian, this letter makes the sound found in English “saw” or “horse”. However, I hear the vowel as the “o” in “open”, except before r. Georgian “ori” (meaning two) might sort of rhyme with “story”, but Georgian “Tako” (a nickname for Tamara) definitely sounds exactly like the English (er, Spanish) word “taco”. The o in “taco” and the o in “story” generally do not sound the same, and in fact have different symbols in most phonetic transcriptions.

Another example is the famous Georgian reader “Ai ia” (აი ია), which is a book many Georgian children use when first learning their alphabet. Although basically every source claims that Georgian has no diphthongs – that is, no vowels that start in one place and end in another – I hear Georgian “ai” as a diphthong; in fact, to me it is indistinguishable from the English word “I”. If I were to follow the supposed rules of Georgian pronunciation, “ai” would have to be pronounced as two syllables, but in practice it is most certainly not. Other examples of Georgian diphthongs are oi (like in boy or noise) and ei (like in weight or cake), and I’d argue for au (like in house or now) and possibly some others.

Unfortunately, I can’t prove these claims without doing some serious science – recording Georgian speakers, creating spectrograms, plotting formants, etc. etc. I’m sure some of my readers will take my word for it, and I’m also sure that many of my Georgian readers will disagree with what I have said here. There are several reasons for this: one, native speakers generally don’t hear allophones as different. If I tell you that the a in “man” and the a in “mat” are different, will you believe me? Probably depends on where you are from and how much you know about English. If I tell you that the p in pot and the p in spot are different, will you believe me? They are – “pot” has an aspirated p and spot has an unaspirated p. Native English speakers don’t hear the difference because we don’t make a distinction between the sounds, but speakers of certain other languages (Korean, for instance) can easily hear the difference.

Additionally, the ideal spoken Georgian doesn’t blend vowels into diphthongs, and so if you listen, for instance, to a highly educated Georgian speaker reading poetry, a lot of the diphthongs will separate. Most Georgian speakers, most of the time, don’t speak in this literary register, but most educated Georgians (aka those who might be reading a foreign-language blog) like to think they do. Georgians react to informal Georgian the way Americans react to informal English – I’ve been told “exla” is not Georgian in exactly the same manner that I’ve been told that “ain’t” ain’t a word, but in practice “exla” is probably even more common in spoken Georgian than “ain’t” is in spoken English.

Part of the reason this interests me is that I just want to promote a correct understanding of Georgian to others who are learning the language. Another part is that I want to bridge the gap in the other direction – to promote a correct understanding of the phonology of Georgian and English so that Georgians can better learn English.

This brings me around to part two of my research: vowels in English and how they have been described and represented in academic and pedagogical literature. Modern sources like to use IPA, because IPA represents sounds with some degree of objectivity, and I admit I also like to use IPA for this reason. However, English has traditionally used a historical system that’s been grandfathered in from before the Great Vowel Shift – in other words, we describe our vowels using terminology that is outdated by at least four hundred years. Perhaps surprisingly, I have found this system to be not entirely without merit.

In particular, when I was taught English – using Phonics, and then phonetic respelling systems used in dictionaries – I was taught about the existence of “short” and “long” vowels. Short vowels are the sounds made by a, e, i, o, and u in cat, pet, sit, box, and cut. Long vowels are the sounds made by those representations in cake, Pete, bike, hope, and cute. There are many more vowels than this in English, some of which are also called “short” and “long” in different descriptions, and some of which aren’t (the schwa and reduced short i are often considered neither short nor long), but these are the most important and relevant ones for this discussion, for reasons with will become clear.

Long vowels may in fact be longer in duration than short vowels, but that’s not what distinguishes them in modern English. In modern English, long vowels are generally “higher” than short vowels, in that they are typically pronounced with the tongue higher up and closer to the roof of the mouth (short and long i are the exception; long i starts off lower than short i but glides over it). Long vowels may also be diphthongs, while short vowels are always monophthongs. Historically, long vowels were just short vowels that you said for a longer amount of time, but they underwent a lot of weird changes for unknown reasons, so now they vary a great deal in pitch from their ancestors; short vowels, on the other hand, seem not to have changed much during this time.

There are two main things that still “connect” English long and short vowels, that make the distinction relevant and, in my opinion, merit the teaching of these concepts even though the vowels involved no longer have the same sounds. One is orthography – that is to say, English spelling still generally spells long vowels using the corresponding short vowel plus a “silent e” or other helping vowel (cake, meet, time, boat). The other is a process called “laxing”, whereby a long vowel becomes short in certain circumstances – perhaps the most notable being “trisyllabic laxing”, where a word like “nation” (long a) becomes “national” (short a) when it becomes the third-to-last syllable. See also: “serene, serenity”, “divine, divinity”, “provoke, provocative”, etc. You also have words where ou becomes short u (“pronounce, pronunciation”) leading one to conclude that “pronounce” might once have been pronounced “pronoonce”. You also have “closed-syllable laxing” (“moon, month”, “keep, kept”, “consume, consumption”) and a variety of other, possibly related processes that all share these common relations between short and long vowels. So short and long vowels aren’t just connected by spelling, but often also by meaning – the a in “nation” and the a in “national” are clearly the “same” in a very intuitive way even though they are pronounced differently.

I bring up this whole sordid business because the designation of long vowels as “long” was apparently taken a bit too seriously by some Soviet scholar, who taught Georgians that English long vowels are distinguished from their shorter cousins by duration, rather than pitch. Thus moon will be pronounced “moooooooon” and sweet will be pronounced “sweeeeeet”.

Georgians have also been taught that English short i is the same as Georgian “ი” – but in fact Georgian “ი” is English long e, and while short i and long e are acoustically quite close, to an English speaker they sound worlds apart. They’re the difference between bin and bean, sin and seen, it and eat. I originally thought that Georgians mispronounced these English words because Georgian lacked a short i sound, but I have since discovered that there is a consistent misrepresentation of Georgian “ი” in English literature, equating the two, which means that Georgians might actually think that “it” is supposed to be pronounced like “ით” which sounds like “eat”.

Yes, it’s true that the vowel in “eat” is typically transcribed as having a markedly longer duration, but that is mitigated by the fact that English long vowels are typically shorter before unvoiced stops, like t – so the “e” in “bean” is somewhat longer than the “e” in “beat”, but this doesn’t really matter in terms of comprehension, whereas the difference between “beat” and “bit” matters a lot.

Instead of learning the correct distinction, Georgians have specifically been taught to distinguish these sounds by length: they say “eat” instead of “it” and “eeeeeeeeeeeeeet” instead of “eat.” This is extremely problematic from my perspective since my job is to teach pronunciation, and I’m not only trying to teach foreign sounds, but I’m trying to teach foreign sounds in the face of decades of scholarship that instructed my students, and their teachers, to pronounce those sounds incorrectly. I honestly don’t know how to surmount this problem, since Georgian speakers of English in general are not receptive to my corrections and have a highly over-inflated sense of the accuracy of their ideas about English. I don’t want to tear down my students’ – or my coteachers’ – confidence, and yet they have to be told, somehow, that basically everything they know about English vowels is wrong.

So that’s the bad news. The good news is that, because of the variations and diphthongs that I mentioned above, Georgians actually can produce most English vowels using only native sounds. Of the 23 vowel sounds and diphthongs that Oxford lists in its Advanced Learner’s Dictionary, Georgians can hit 17 almost exactly using sounds native to Georgian (18 if they speak Mingrelian or one of the other Kartvelian languages that maintains the schwa). They just need to be taught the correct correspondences.

Well, as promised this post has run very long – and I’ve barely scratched the surface. I’ll close on that note, with a link to an amusing but stunningly comprehensive resource for information on vowels:

The Society for the Prevention of Cruelty to Vowels

About these ads
This entry was posted in Education, Linguistics. Bookmark the permalink.

7 Responses to Vowels

  1. I’d love to see future posts on all the topics you cover here (especially “exla” and other non-literary speech).

    You point out that Georgian does have many of English’s vowels, if we understand “have” to include Georgian vowel allophones in addition to the usual five vowel phonemes. This might be useful for Georgian linguists, since they are trained to listen for allophonic differences. Can it really help regular people? That’s like saying that learning the aspirated / unaspirated distinction in Thai should be easy for English speakers, since they have both if allophones are included. But they don’t have free access to both (the “t” in “stop” is hard to say in isolation), and, as you say, very few are aware that they make two different sounds.

    • panoptical says:

      Well, let me list some explicit ones and you tell me what you think:

      ა = o in hop
      ე = e in bet
      ი = ee in beet
      ო = o in hope
      უ = u in duke
      აი = i in bike
      ეი = a in cake
      ოი = oi in noise
      აუ = ou in house
      ია = ea in idea, ear (British)
      ეა = ai in fail, air (British) / ერ = air in air (American)
      ორ = or in horse

      The o in hope is realized as a different diphthong in different dialects, but the use of Georgian ო in long-o words would be totally comprehensible and unmarked to native English speakers. The u in duke is sometimes pronounced with a /j/, but not in my yod-dropping accent. The l- and r-colored vowels are possibly the toughest to pin down and the ones with the most variation in English, but those Georgian sounds should get quite close to them.

      This set leaves out a in cat, i in bit, u in put, u in cut, er in her, and schwa, all of which will have to be drilled a lot more – but that’s down to six sounds, which is not nearly as daunting as 16-20.

      In any case, teaching these relationships would certainly aid in pronunciation, especially over the erroneous set that is currently taught, viz:

      ა = u in but
      ი = i in bit
      ე = a in had, or in work, er in were
      იი = ee in sweet
      უუ = oo in moon

      • Ah, I thought you meant that Georgians already had allophonic access to short i, short a, and sounds like that. I agree completely with the examples you’ve given, and indeed have been using those vowel combinations when I transcribe English words into Georgian.

        • panoptical says:

          A weird thing I noticed today was that Georgians seemed to be producing a reduced /i/, close to a short i, when saying “jellyfish” – specifically, the “y” was produced as much shorter than it should have been and the contrast between my expectation and their production sort of evoked a short i.

          I have had a lot of success in getting my classes to render short a without too much difficulty, but I personally have to keep in mind that they do much better when I demonstrate actual short a (a in cat) as opposed to New York diphthong short a (a in man, often transcribed [eə]).

          Also, the younger classes do all the short vowels better, with definitely fits with a lot of the research on language acquisition and with my theory that learning incorrect Georgian-English mappings is more responsible for mistakes than a simple lack of vowel diversity in Georgian.

  2. Gio says:

    “ი” – “ea”, as in ‘eat’ is very good finding.

  3. Having recently taken a trip to the beach, it occurs to me that the time-length of English vowels can be important in some cases. It’s why when Georgians (and many others) try to say “beach” it comes out sounding like “bitch”: although they don’t have the “short i” sound, they say the “long e” as quickly as we say “short i”, when in fact we sometimes sit on the “long e” sound for a bit more time.

    • panoptical says:

      Vowel length can be important – and can distinguish between two words in some accents – but I think in this case something else is going on.

      For example, when Georgians say “ბიჭი”, does it sound to you like they’re saying “bitchy”? Would you ever mistake that word for “bitchy” in any of the Georgian accents you have encountered?

      No, I suspect that in the case of pronouncing “beach”, the Georgians who say it wrong have simply learned it wrong. They’re aiming for the wrong pronunciation, rather than aiming for the right one and falling short.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s