Bakunin's log

[Bakunin]

A few months ago, I’ve started learning a variety of Lao spoken in North-Eastern Thailand, Isaan. It’s not a super serious undertaking but it would be nice to develop good comprehension and get access to the culture. Lao and Thai are already close, and Isaan has some influence from Thai so it’s even closer. I have one speaker from Khon Kaen province I’m working with and should have access to for a few more years.

After mainly working with recordings on an on-and-off fashion, I’ve spent last week to analyse the tones of that particular variety. Doing this, I learned a great deal. It was (and is) pretty exciting to add a new layer of understanding; there’s a lot of regularity in languages, and tones are no exception to this.

For Tai languages (which includes Lao and Thai), tones can usually be understand using a grid of 20 syllable categories; each syllable falls into one of these 20 boxes. Words in the same box have the same tone. Overall there are usually between 5 and 7 tones, so many of the boxes share the same tone, but the distribution and the contours of the tones vary from variety to variety.

Using word lists prepared for testing the 20 boxes, one can relatively easily determine the number of tones and their distribution. The following picture is the result of my analysis (plus Thai as a reference); I found a published paper with an identical result for a neighboring district and am reasonably confident in my analysis. Reading these papers, I’ve also realised that other Isaan varieties have different tone contours and distributions, and some of these features were confirmed by native speakers who are, of course, also aware of differences in tonal shape and distribution from place to place and can sometimes articulate them.

Screen Shot 2016-12-17 at 09.16.01.png

Tone contours between the two languages are different. None of the Isaan tones is really identical to any of the Thai tones. What I call “mid” for Isaan is slightly different from mid in Thai… mid in Thai is “flat”, mid in my Isaan variety can have a slightly falling then rising shape. Similarly, “high-rising” in Isaan is nothing like I know from Thai, and the two “high” tones are also quite different, especially for syllables ending in vowels.

Using further information encoded in the Thai writing system (which is fortunately quite conservative), I have constructed a little tone conversion table converting Thai tones into Isaan tones for native Tai words shared between the two languages. Nothing of that is revolutionary in any sense, but it was pretty exciting to work through this on my own, and it definitively helps me to understand the language much better.

Understanding the tones and the conversion from Thai tones to Isaan tones allows me to transcribe the language with correct tone marks. I’ve started experimenting with that and will probably continue for a bit to see how useful a “private transcription system” is.

The tone conversion (and other sound conversions… there are some regular vowel and consonant shifts, e.g., ch becomes s, ʉa becomes ia etc.) reminded me of my ongoing quest to learn Swiss German. Lacking a standard, it’s basically the same story: all phonological features exist in a continuum and vary from place to place. For Swiss-German, vowels are the most obvious feature which changes from place to place, for Isaan it’s probably tone contours. It’s a good idea to pick one specific variety and learn that in order to avoid a strange mix of dialects. In the case of Swiss-German, that’s obviously the place where I live, and in the case of Isaan, that’s now this one speaker and his community (and the larger region with millions of inhabitants but *not* all of the North-East). It doesn’t make sense to learn “Swiss-German” if that is meant to include all high-alemannic varieties spoken in Switzerland, and it doesn’t make sense to learn “Isaan” if that is meant to include all Lao varieties spoken in North-Eastern Thailand.

[Bakunin]

I’m slowly morphing from a regular forum member into a lurker …anyway, here’s an update.

Someone recently posted a link to the excellent open-access journal Language Documentation & Conservation. One of their recent volumes had a large section on working with tone languages, and in parallel I read a linguistics textbooks on tone (Moira Yip: Tone). Plenty of new information, and I learned a lot about language diversity and the complexities of tonal systems. Turns out that Thai and Lao are among the easiest tonal languages one can imagine. There’s all kind of weird stuff out there like floating tones, morphological tones, polysynthetic tone languages etc. I also learned a lot about the interplay of tone and intonation, tone sandhi, tone articulation and how to go about determining the tones of a new language.

Isaan
All of that is pretty relevant as I’m currently trying (well… I’m beyond trying I guess) to write the Lao variety a friend of mine from North-Eastern Thailand records for me. I have a pretty solid grip on the tonal system, but I’m still not entirely sure if there are 5 or 6 tones. I’m currently assuming 6 tones, but two of them seem to have a lot of overlap in their actual realisation (so that one can sound like the other and vice versa, depending on position in the phrase, whether stressed or spoken in isolation, speed etc.). I also don’t have a minimal pair yet for these two, another pointer that it’s maybe just one tone with various allotones which cover a rather wide range of contours. Fortunately, it doesn’t really matter as the writing system looks the same whether these are two distinct tones or just one.

I’m writing transcripts of picture recordings but also of other recordings on specific topics or words. It’s intense but also incredibly interesting. I encounter lots of words I hadn’t noticed before. Most of these new words (not familiar from Thai, that is) have Lao cognates but some are specific to this variety. I guess I’m learning a lot of Lao on the way, like @iguanamon has gotten familiar with French through his study of French-based creoles.

I’ve also developed an appreciation for the variability of Lao dialects in Thailand. It’s no good to learn an unwritten language by mixing different dialects together. In Isaan, it’s tones, in Swiss-German vowels which vary from place to place and guarantee total confusion if mixed freely. So I’lll stick to that one dialect for the moment, and I hope I can work with more speakers from this village in the future to get variety within this particular dialect.

There’s quite some material on the web on Isaan vocabulary, mostly in Thai, even a university dictionary project. However, none of these are precise with tones but rather approximate whatever variety the word is from with Thai tones. Completely useless for me as a primary source but still good to cross-check meanings.

Khmer
Ploughing along. I feel I’m over the first hump; I was able to speak (broken halting) Khmer with my teachers for several hours each day in PP last month without getting exhausted, and I had entire conversations entirely in Khmer with strangers. Khmer has taken a back seat recently but I’m still doing stuff on a regular basis. It looks like alternating bursts of Khmer and Isaan for a few days each instead of doing both at the same time.

Thai
I deleted a legacy Anki deck which still served me about 20-30 cloze deletion or recognition cards per day even though I hadn’t added anything in a long time (years?). Anki may be efficient but its also a huge pain. At the advanced stage, I don’t really see the point anyway… there are so many words I don’t know and don’t need to know because I don’t encounter them in my life, so I don’t need to see them in Anki either — and the others which matter to me will come again and again just by using the language. It was a great feeling to delete the app.

I’ve reached a plateau in Thai. To get to the next level I would have to invest a lot of time and effort and specifically work on certain skills. I’m enjoying Khmer and Isaan too much to do that, so Thai stays where it is. Of the three languages in focus, I’m still using Thai more than the others, so there’s little danger that things start moving backwards.

[lingua]

I for one look forward to your log entries since there is so little Thai here and I enjoy your perspective on the similar languages.

[Bakunin]

Isaan
I've written a post for a Thai learning blog describing how I'm going about learning Isaan. Instead of reposting the content here, here's a link. I really enjoy working on Isaan It's a big adventure, every new word I encounter (which is not cognate with Thai) gives rise to a little burst of excitement.

Swiss-German
Somehow it never occurred to me to look for something like this and I only found it by accident... there's a weekly hour-long podcast on Swiss-German published by SRF: Schnabelweid. It's a great find, I really enjoy listening to Swiss-German from regions I don't have much exposure to. There's such a beautiful diversity of sounds and vocabulary.

[iguanamon]

Well done, indeed, Bakunin! You continue to be an inspiration. What you have done with your drawings and constructing your own learning materials can serve as a blueprint for preservation of non-written languages. Not bad for a non-academic, non-linguist, amateur language-learner!

I had a thought, could generic videos of people doing things, i.e.: a person walking/sitting down/jumping/throwing a can away/turning on a fan etc., be used to similar effect with a voice over description in L2 and L1? Or, would it be too complex and difficult to do? What do you think?

[Bakunin]

Well done, indeed, Bakunin! You continue to be an inspiration. What you have done with your drawings and constructing your own learning materials can serve as a blueprint for preservation of non-written languages. Not bad for a non-academic, non-linguist, amateur language-learner!

I had a thought, could generic videos of people doing things, i.e.: a person walking/sitting down/jumping/throwing a can away/turning on a fan etc., be used to similar effect with a voice over description in L2 and L1? Or, would it be too complex and difficult to do? What do you think?

I don't know if I'm an inspiration but you most definitively are one! Your log is one of the most interesting on LLORG, and I always enjoy to read your updates. Most of the ideas I apply in my own learning were developed by field linguists! There’s a lot of creativity out there which, strangely enough, rarely crosses over into the language learning community.

I’ve seen people describing filming culturally meaningful practices (daily activities, work, ceremonies etc.) as a technique for field linguists. Native speakers would then describe what they see and provide further commentary. This can result in quite interesting recordings, I guess, which would be very useful for learners. Audio commentaries could be further complemented with written materials.

You seem to have a more generic set of videos in mind. I haven’t seen anything like this in the literature. It definitively takes much more time and effort to film than to illustrate, and it might be difficult to reach the required level of abstraction, but it could be quite effective and also fun for learners.

My own ideas are all based on recordings, though. I need to work a lot more on the basics, but I’ve already started to transcribe my first little story (a childhood memory) in order to keep moving ahead. There’s a huge difference between structured illustrations-based recordings and the language used in recounting a childhood experience, and it’s currently quite a challenge. But story-telling is definitively the next step. After that I was thinking of trying two-person conversations about topics related to food, agriculture/work, handicraft, festivals/religion and customs etc. Another area of interest is ‘life stories’ of people, again ideally as a conversation between two native speakers.

There’s an interesting technique to make such recordings more accessible, and that’s to provide native speaker audio commentary on the native speaker recording. This could help with clarifying fast or confusing passages, it could be employed to explain (and maybe translate) vocabulary, and, of course, to add more context. I’ve recently bought a field recorder supporting multiple inputs and have plans to try something in that direction later this year. Whatever comes out of it, it’s a lot of fun

[tuckamore]

Isaan
I've written a post for a Thai learning blog describing how I'm going about learning Isaan. Instead of reposting the content here, here's a link. I really enjoy working on Isaan It's a big adventure, every new word I encounter (which is not cognate with Thai) gives rise to a little burst of excitement..

I read your post on WLT yesterday, and hadn’t connected the you here (Bakunin) with the you over there (Andrej) until reading this log entry today. It seems obvious now, but I was slow to make this connection. Great post on WLT, by the way! I strongly feel, too, that: “Learning a tonal language without getting the tones right doesn’t work for me.” And, I struggled on this front.

I have also now spent an hour+ reading through your entire log. So much of what you have written about listening strikes a cord with me in starting out with Thai. I found I needed more listening than ever in Thai right at the get-go*. No course that I looked at worked for me. No course had enough audio. Tones were so foreign to me. I knew I could not learn tones from a book and the meagre audio that comes with beginner Thai resources. I think I would have enjoyed and profited from taking an approach like yours.

Also, thank you for making your recordings publicly available! They are a tremendous resource. Once I am at a place where I can get the general gist from your recordings, they will play a central role in my studies. I’m already thinking of ways I will want to utilize them.

*I probably need this in other languages, too, but it isn’t so obvious at the beginning.

[Bakunin]

Thanks for your comment, tuckamore! Actually, I don’t think tones are more difficult than other sound features but they are a bit unusual for native speakers of non-tonal languages at first. What seems to do the trick for me is a massive amount of listening without any production at first. I mix in some careful and contrastive listening, but mostly I’m going for a lot of exposure with a focus on meaning. After a while the tones feel more and more familiar and I start sounding out odd words and phrases here and there. At this point in the process, I will have a relatively solid mental model and can clearly hear how close I am. I believe that’s critical, otherwise I’d be groping in the dark. Now I would also start using one or two word responses in actual conversation should I get into one. It’s also the time to start chorusing if I’m in the mood (but not shadowing, that comes much later). After that, things develop naturally. Key for me is to develop the mental model first by listening and subconsciously processing a lot of comprehensible speech first. It helps that I’m not in a rush to get anywhere

I think the approach of listen and repeat is a bit flawed, in particular for first-time learners of tonal languages. How is the learner supposed to do it right without having had a chance to process the tonal shapes first? That takes months of exposure, not just three example words. The actual production of tones is pretty complex as they are embedded in general intonation patterns, usual have different shapes depending on surrounding sounds, then there's tone sandhi and always a certain range of acceptable allotones. No way this can be acquired consciously. The brain is perfectly capable of zooming in on the underlying structures by itself but it takes a bit of time.

Also have a look at what lingua is doing. From what I understand he/she's doing a lot of narrow listening/watching (cooking shows). That can be a great entry point as well. You want to get to a point where you do understand what’s going on. The magic of comprehensible input works only if it’s comprehensible to start with. I used to binge watch travel shows in Thai, it doesn’t have to be food

[lingua]

Actually I do find the tones difficult. Our Thai tutor has gone through all five tones on a few examples where the words are otherwise spelled the same (excluding the accent) and I can barely tell the difference most of the time. I also have noticed that on the one cooking show I watch where the hostess talks fast that I barely hear any tones out of her at all. Context is everything in Thai. I wonder if voice lessons (or even acting/singing) would help with tones?

I also appreciate your site and have listened to a few of them. Love the drawings.

Oh and Lingua is a she who loves to cook and eat.

[Bakunin]

Actually I do find the tones difficult. Our Thai tutor has gone through all five tones on a few examples where the words are otherwise spelled the same (excluding the accent) and I can barely tell the difference most of the time. I also have noticed that on the one cooking show I watch where the hostess talks fast that I barely hear any tones out of her at all. Context is everything in Thai. I wonder if voice lessons (or even acting/singing) would help with tones?

I also appreciate your site and have listened to a few of them. Love the drawings.

Oh and Lingua is a she who loves to cook and eat.

Thanks for your comment, lingua! I didn't say tones aren't difficult, just that I don't think they're more difficult than other phonological features Phonology is generally pretty difficult and takes a lot of exposure. What I don't understand is how a learner is supposed to get a good feeling for tones after hearing a few rounds of ga-ga-ga-ga-ga with all five tones, and then some listen-and-repeat. This seems to be a fundamentally flawed approach to me. For me, it takes a massive amount of listening to allow my brain to hone in on the underlying tonal shapes.

Native kids seem to acquire tones in order. I've recently read a study on Mandarin kids who seem to acquire certain tones earlier than others. Full tonal accuracy takes a while (and is achieved around age 3 if I remember correctly). The same has to apply to second language learners. Some tones come earlier, others come a bit later.

Another really important realization for me was to understand that tone is not static. How a tone is realized depends on many things, most importantly surrounding sounds and tones and the overall intonation pattern. There is no fixed tone (starting, ending) height and also no one shape - there's a lot of flexibility. I struggle to see how all of this could be aquired consciously.

Oh, and there's another thing about tone: Tones are almost always perceived in relation to each other. Even for native speaker it can be difficult to determine the tone of an isolated one-syllable word, especially if the language has level contours which only differ in relative height. What helps me when I'm trying to determine tones in Ton's Isaan is to have him contrast the word I'm curious about with other words I know the tone of. It's much easier to determine whether two tones are the same or different than to say a certain tone is this or that. Concretely, if I'm unsure about word X and think it could have the tone of A or maybe B, then I ask Ton to say A X A X, and then B X B X, and in 99% of cases it's totally clear to both of us whether X is like A or rather B. Maybe try this with your tutor, and don't worry about hearing the tone of a word spoken in isolation too much. I use the same approach with tricky vowels in Khmer, and you can also use it with other sound features, e.g., บ (b) versus ป (unaspirated p) versus พ (aspirated p))...