Posts tagged phonetics

This post is entirely prompted by being in France, but I think there’s something really cool about nasal vowels. To make a nasal vowel start by making a ‘normal’ or ‘oral’ vowel. Then lower your velum (it’s the smooshy bit at the back of your mouth that opens and closes the air flow to your nose). It feels a bit weird if you’re not used to it, but you’ll know if you’re doing it right because your voice will sound nasal (big surprise there!).

French has a few nasal vowels, they’re tough to get your mouth (and nose) around at first if you’re an English speaker! There are lots of languages with nasalised vowels; in a sample on WALS 64 out of 244 languages had nasal vowels as well as oral vowels. There are no languages that I know of with only nasal vowels and no oral vowel.

The funny thing is that English doeshave nasal vowels. They often occur before a nasal consonant (m, n, or ng). So if you listen carefully to someone saying ‘tuna’ or ‘piano’ in a natural way it’s likely they’ll nasalise the vowel. The thing is that we’re not trained to hear this difference, because it doesn’t change the word, so we don’t ‘hear’ it!

When we speak we make sounds (unless the language is a sign language). We make these sounds by having air flow out of our lungs, setting out vocal folds vibrating and then constrict the airflow to get different sounds. This is the case in English, as with a majority of languages. This is calledegressivesound - but it’s not the only way people make sounds in some languages.

There are a small number of languages where people can make words by breathing inwards instead of out. This is referred to as ingressive sound (because the air is going in). Think about a gasp of astonishment in English, or try it out - just don’t hyperventilate!

Scandinavian languages like Swedish are famous for having ingressive sounds. They’re most frequently used when you’re listening to someone speak and say ‘yeah’ to agree with them. It’s a fact about Swedish that I’d completely forgotten until a cousin of mine used it. It’s really distinctive to hear the breathy in-rush of air. The Wikipedia page on ingressives has an example from Norwegian if you don’t have any Swedish cousins to hand.

Ingressive breathing is mainly used for short utterances, and often as ‘backchannelling’ when someone else is speaking. Robert Eklund has done a survey on ingressive sounds, including animals (for example, leopards).

I thought that since I was in Warsaw I would make this week’s TwwEh Poland-appropriate. The 24 hours since I got here have been rather hard on my brain; three years of Nepali have squeezed out most of my Polish, and it’s rather embarrassing when Nepali words come out instead of Polish ones. 

One thing I’ve always loved about Polish is that it has a regular stress system, which makes it a little easier on the learner. Stress is when a part of a word sounds more emphasised than the rest. Polish stress is almost always on second last syllable of a word (which is also known as the penultimate syllable). There are only a small number of examples, such as words from Latin or English, for example one I always remember is universitet (‘university’) where the stress is on the second syllable in the word.

Not all languages have stress, for example tone languages are generally considered to not have stress as well. Of languages that have stress, the location of the stress varies between languages. In a survey of 502 languages in WALS, Rob Goedemans and Harry van der Hulst found that 220 were like English and had no fixed stress pattern, and the other 282 were spread across different fixed patterns, for example the first, second or last syllable.

Stress is notoriously hard to hear for non-native speakers, most of the time when we talk quickly you don’t really hear the stress, but a native speaker always instinctively knows where it is. Take this English-learners’ website with its seemingly random collection of rules that learners of English are subjected to, and that’s not even all of them. This is one of those times I’m rather glad to be a native speaker of English!

For me one of the things about working with a language I’m not a native speaker of is getting my tongue and my brain around the sound system. It seems obvious to say, but dDifferent languages have different sound systems, that’s one of the reasons they sound so different. Sure, most languages might have a vowel that kind of sounds like ‘a’ but there’s every chance it’s a little bit different.

We lock into these sounds very early on in our language learning, some research shows that we begin to pick up on some of it by listening to our mothers voices before we’re even born. And we get very good very quickly at making the sounds in our own language. It’s how French speaking children have no trouble with those vowels that always get me, or why English speakers can say ‘th’ with ease

So we tune into the differences that are found in our own language which means we stop paying attention to distinctions that can be found in other languages. this is why I can’t hear the difference in Polish between three different sounds that all sound like ‘sh’ to me, or how some English speakers can’t roll their

When it comes to Yolmo, the worst sounds for me all sound like ‘ta’ or ‘da’… while in English we only perceive two sounds there’s a lot more happening in Yolmo.

Firstly, in English we only have ‘t’ and ‘d’ - which is a voicing distinction. Try saying each with your hand on your throat and you’ll notice that it vibrates more when you say ‘da’ than ‘ta’. In Yolmo, there is the voiced ‘d’ sound and the unvoiced ‘t’ sound, and also an aspirated ‘t’ sound (which gets written with a h next to it) - this third sound is a ‘t’ but with more air coming out with the sound. So we have three sounds ‘d’, ‘t’ and ‘th’ where English only has two.

Secondly, Yolmo has retroflexes, which English doesn’t. A retroflex sound is where you curl your tongue up so the bottom bit touches the roof of your mouth. It’s quite a common sound for this corner of the world - Nepali has it and it gives Hindi some of the rhotic (r-sounding) quality associated with it. There’s a retroflex for each of the above sounds, which I’ll write them in capitals. so there’s ‘T’, ‘Th’ and ‘D’.

Thirdly, Yolmo has tone. Tone is a feature of many languages - if you’ve learned Chinese or Vietnamese you’ve come across it. Basically the vowel sounds higher or lower. Fortunately Yolmo only has two tones, high and low. And, thanks to some complex historical reasons that I shan’t bore you with, aspirated sounds only have high tone, and voiced sounds only have low tone, but unvoiced and unaspirated sounds such as ‘t’ or ‘T’ can take either. So there’s another distinction that English speakers don’t make.

Finally, there is a vowel length distinction that’s very hard to hear. While there’s ‘a’ there is also ‘aa’. English has long and short vowels too, but there is also a quality different, the throat area tends to be more tense for short vowel and lax for long vowel. But in Yolmo it’s only length and can be quite hard to hear. So that doubles the amount of options that there were.

So in English we have ‘ta’ and ‘da’ - which means that we can hear the different between ‘tag’ and ‘dag’. InYolmo , instead of just a two way difference, all those variables give us sixteen different sounds. That means that Yolmo speakers hear the difference between ‘tag’ (with low tone), ‘tag’ (with high tone), ‘Tag’ (with low tone), ‘Tag’ (with high tone), ‘dag’, ‘Dag’ ‘thag’, ‘Thag’, ‘taag’ (high tone), ‘taag’ (low tone), ‘Taag’ (high tone), ‘Taag’ (low tone), ‘daag’, ‘Daag’, ‘thaag’ and ‘Thaag’.

It doesn’t mean that they are all necessarily words in Yolmo, it just means that they can hear the difference between them. While my ear has been getting better over the last few years these things still stump me sometimes!

[Lauren’s currently in off in the hills of Nepal for field work. This post originally appeared on Lozguistics.]

I’m not sure how this phonological feature has so far escaped a TwwEh post, especially considering it’s something that I deal with every day when working on Yolmo!

English speakers use pitch, but we use it at the ‘supra-segmental’ level, for things such as making questions but raising the pitch at the end of a sentence. Tone is where you use pitch quality of to make a lexical or grammatical distinction.

Tone is something that English and most Indo-European language speakers think of as exotic because it’s not a prominent feature of their language family, but tone is relatively common world-wide (and in Indo-European there’s some tonal distinctions in Swedish and Norwegian). Of the 527 languages surveyed on WALS 220 have tone.

So what you end up with are words that have exactly the same consonants and vowels but the change in pitch makes completely different words. To give you an example from Mandarin Chinese because it’s one of the most well known tone languages, and because it has an impressive-looking five way contrast:

1. mā “mom”
2. má “hemp”
3. mǎ “horse”
4. mà “scold”
5. ma (an interrogative particle)

The only thing to distinguish these words is the way the register rises, falls or stays level.

But not all tone languages are the same, there are lots of different ways tone can be expressed. Some languages like Chinese use the pitch height, but more importantly it also uses the contour of the pitch as well - it’s relatively easy to hear the rise or fall of these tones in context. Other languages, like Yolmo, make a different between high and low tone by the height of the pitch. This is fine, but often means that in one context a high tone might be lower than a low tone in another context which can make it hard to tune it. For some of these languages the high or low tone might correlate with something like creaky voice or certain initial consonants to make identification easier. For example, Yolmo low tone correlates with breathy voice and if a stop is affricated the tone will always be high.

Some languages have tone marked on the initial syllable of a word, while other languages mark tone on each syllable. The number of tonal distinctions also vary - as you saw above Mandarin has five tones while Yolmo only has two. Pitch accent tend to have more tone; Cantonese has seven tones, and there are possibly languages with more.

On a geeky level I like tone because of its ability to create more information density. Even if you only have a two tone system you automatically double the number of single syllable items you have. A friend once told me he read a paper that there is a higher density of people with ‘perfect pitch’ in languages with pitch heigh, but I didn’t get a reference. If anyone knows this study, do share!

A couple of weeks ago we came across this article on The Age website citing the findings of Dean Frenkel, a ‘speech analyst’, going to town on the speech of a couple of Victorian politicians. We were pretty incensed at the time, but decided to let it slide because here at Superlinguo we’re lovers not haters. When someone saying they’re a ‘speech expert’ is so rampantly misinformed it’s hard for a blog post to not appear to be a complete flaming. But then the crew over at Crikey’s Fully (sic) published this lengthy discussion of an interview Frenkel did on ABC radio and we realised this wasn’t a one-off case of dressing opinion up as a factual ‘expert analysis’, but that this guy was a serial offender.

You can read about the radio interview at length over at Fully (sic), so we thought we’d set the record straight on The Age article. Firstly, we should say we think this might be a matter of terminology. Frenkel’s a public speaking coach and that’s fine, public speaking is tough and any help people can get is cool. Also, we’re down with people having personal opinions on things, it’s a free world and we don’t much mind if you share your objective opinions with anyone. But please, don’t use the word ‘speech’ expert when you mean ‘public speaking’ and dress up opinion as fact with some pseudo-science and go publishing it in a newspaper.

As linguists, we are taught to view speech as one component of language, and to observe speech patterns as part of a complex system. It’s for good reason that as a discipline, linguistics has close ties to neuroscience, psychology, sociology and other humanities and sciences. It’s far too complex to start making assertions that link a person’s vocal pitch to their personality.

Frankel complains that Ted Baillieu, our current State Premier “has a voice that is at least half an octave higher than it should be, given his size.” If he’d done some basic research into human physiology he’d know that the pitch of your voice has nothing to do with the size of the person, but the size of their larynx and vocal folds (to go all academic citation on you see page 17 of John Laver’s Principles of Phonetics). A big person is more likely to have a big larynx but there isn’t a direct correlation. Also, Frenkel is basing this entirely on his subjective opinion, if he’d got in touch with a physician or a phonetician (someone who studies speech sounds) it wouldn’t be too hard to actually measure the pitch of Baillieu’s voice and compare it to other men of his stature. 

We’ve complained about this kind of journalism in a cross-posted blog piece on Fully (sic) before, because these kind of rampantly speculative discussions distract from people forming opinions of politicians based on policy, not personality.  

Again, we don’t want to be haters. Frenkel is clearly a language lover and word nerd, just like us. We just hope that in future, Dean Frenkel is labelled more accurately (by the man himself and by journalists covering his opinions), as a singer and public speaking coach. To call himself a ‘speech expert’ detracts from centuries of rigorous and objective work by speech pathologists and linguists.

A new Coldplay album, called Mylo Xyloto, has just been released into the world and its success (or not) is being watched as a kind of “healthcheck for the record industry”, apparently.

I reckon it could also be used as a check-in for the way English speakers deal with words beginning with “x”.

When I saw that the album was called Mylo Xyloto my mind immediately jumped to that formidable ‘x’ at the start of the second word. How the heck would I pronounce that? As an amateur broadcaster, I live in constant fear (well, mild anxiety really) of mispronouncing a band or album name on air due to serious cred loss, which may explain a bit of my reaction to this strange album title. There are very few words that start with ‘x’, the least of any letter. Ack. How do we know which sound/s corresponds to this word-initial ‘x’?

The given name Xavier, which is fairly common these days, and the word xylophone with their [z] or [s] sounds at the beginning (depending on your accent, really) are a couple of the rare examples that we English speakers have as a guide when it comes to pronouncing word-initial ‘x’.

But this letter occurs so often in written language later on in a word or at its end, we are more used to its corresponding sound being a [ks] sound, as in six or oxygen. We’re so used to this sound for ‘x’ that I still think that many of us instinctively go to say [ks], or even [eks], at the beginning of our utterance when faced with word-initial ‘x’. I’ve definitely heard people unfamiliar with the name pronounce Xavier as ‘Ecks-ay-vee-ah’ (NB. if that rough pronunciation guide looks weird to you, remember here in Australia we don’t pronounce our ‘r’ sounds).

To further confuse things, ‘x’ in a written word can also correlate to a [gz] sound, as in exhaust or exam, when the ‘x’ is followed by an accented syllable and a vowel sound.

I might even be a [kʃ] sound (or ‘ksh’) in words like crucifixion or transfixion, or a [ɡʒ] sound, as in luxury for many speakers of British English.

So, big thanks to you Coldplay for using ‘x’ at the start of a word in your album title and making life hard for us. Sheesh. Thank goodness this is just a rare occurrence where a group of stubbled musicians have decided to be a bit difficult, just for the heck of it. According to Chris Martin, “It is pronounced My-low Zy-letoe* … and even the lads admit the title doesn’t mean anything.” Righto then.

* or /ˈmaɪloʊ ˈzaɪlətoʊ/ for those who prefer the IPA

Tonight I noticed Leigh Sales applying a little effect to the final ‘t’ in her “7.30 Report” sign-off. It was a bit of the ol’ hyperaspiration - a fascinating thing.

This Fairfax article from 2007 uses Kath & Kim and their alter egos Prue and Trude as exemplars of this Australian English feature. (Excerpt shown below, with condemnation of the casual homophobia which sadly also features in the article.)

“Kath and Kim also hypercorrect themselves. Like Prue and Trude, they say their final “t”s funny from time to time. Pauline Hanson did it as well. You know, that kind of breathy sound? It’s called hyperaspiration, and it’s a classic case of hypercorrection. These women, real and fictional, are on the run from the broad Australian “t”, which sounds like a “d”, as in “budda” for “butter”, or gets dropped at the end of phrases. Some do it more flamboyantly than others. Prue and Trude say “great” as “grayshsh” and “beautiful” as “beyoushshiful”. And Kath says: “Yis, I am hoigh maintenance and, frankly, I enjoy ishsh.”

I find hypercorrection really interesting - it comes in different shapes and sizes, like when “yourself” is used in place of “you”, or when we try to pronounce foreign names. I think we probably all do it in certain situations (there’s a bit of Prue & Trude in all of us, perhaps?).