Subtitles section Play video Print subtitles So let’s talk about sounds. At first, you might think that writing down all the sounds you hear the way that you hear them should be pretty straightforward. But if you're listening to me now, you've probably struggled with enough crazy English spelling to know that's not true. So is there a way to accurately transcribe any word from any language anywhere ever? There sure is. I’m Moti Lieberman, and this is The Ling Space. Let’s just face it: English spelling is nuts. The same letters can represent all kinds of different sounds: so, ‘c’ can be [s] like “Cecil” or [k] like Carlos, ‘g’ can sound like [g] in “dog’ or [dʒ] in “angel”, and don’t even get me started about ‘ough’ - [ow], [u],[ʌf], [ɔ], [aw]. As if that wasn’t enough, the same sound can be represented by a bunch of different spellings: so, the [k] sound can be written like “key”, “ache”, “call”, “luck”, or even “box”. We also have silent letters, like the p in “pneumonia”, and sounds that we pronounce but can’t be bothered to write, like the [j] in “cute”. So English is a total failure for transparency - you can’t look at something and know for sure how you should say it. But it gets even worse when we try to represent non-English sounds with English spelling. We’ve tried lots of systems over the years, but you always end up with this inexact science. “dayjah voo” doesn’t actually sound that much like “déjà vu”. Or look at Pinyin, the system of transcribing Mandarin into English. If you look at “cuiruo,” it probably won't make you think you should say 脆弱. There may be rules, but it’s not transparent exactly what they are from looking at it, and even if you know them, you don’t really get exactly what it should sound like. And this is true for pretty much every writing system in the world. Even languages with tons of sounds will still have some sounds that they just won’t use, like the [y] that's tricky in French. And it goes both ways, too: Saying “furusato”, the Japanese word for hometown, doesn’t really capture the sounds that you hear in 故郷 (ふるさと), And on the other hand マフラー does not really get at “muffler”. So what do we do? Accurately transcribing the sounds of language is really important to linguists! It allows us to puzzle out phonemes and allophones, or map out different dialects, or capture the tantalizing glimpses of a rare or dying language. We need a way to make sure that everyone looking at the same alphabet knows exactly what those letters represent. That's the only way we can really talk about what language sounds like. And luckily, we’re not doomed to use approximations forever, when we're trying to capture all the sounds in all the languages. There’s a writing system that aims to do just that. It’s called the International Phonetic Alphabet, or IPA. It was first proposed in the late 19th century by the International Phonetic Association, and it's been revised a bunch of times since then, as we discover new sounds in the world’s languages. And we’ve needed those revisions - there were whole classes of sounds that weren’t on the chart at first because phoneticians weren’t really that concerned with capturing sounds outside European languages. So, just take a look at click sounds, like [!] or [ǀ]. The early IPA didn’t have room on the chart for sounds that weren’t made with air flowing out from the lungs up through the vocal tract. But we need those! We have to have them to describe a number of different languages from southern Africa, like Zulu or Xhosa. But they didn’t make it onto the chart for the first 45 years that it existed. By now, though, we’re pretty solid: the last change was in 2005, and it only added one sound. So what’s the IPA made of? Well, it basically comes in two parts: a vowel chart, and a consonant chart. Together they make up the roster of possible language sounds that humans can make with their vocal tracts. A lot of the symbols in the IPA should look pretty familiar to English speakers. So, [t] is written as ‘t’, [o] as ‘o’, etc. But the most important thing to remember about the IPA is that you get one symbol per sound. So you always know what sound you’re going to get for any given letter. Since the goal is transparency, and since English spelling is so confusing, the similarities stop pretty quickly. So for example, the sound ‘sh’ as in “shadow” is written in English with an s-h, but in the IPA with an [ʃ]. You also get [ð] for the ‘th’ in “them”, and [ŋ] for the ng sound at the end of “tongue”. And sometimes, letters from English are associated with different sounds in the IPA, like using “j” for [j] or “y” for that pesky French “u”. This means the problems we saw earlier with dealing with different languages just disappear. We can transcribe English “deja vu” as this and French “déjà vu” as that and bam! Now we can tell them apart and reproduce them more accurately, and do all kinds of cool linguistic science to them. And because it’s international, linguists from anywhere in the world should be able to share their data with people from anywhere else, no matter what their writing system is. All they have to do is learn the IPA. So, great, with matchups between one sound and one symbol, you can express a whole lot about human language. But what if you want to put more information in there? What if it’s important to you to make sure you accurately include all the allophones and variation in what you hear? After all, just knowing what those basic sounds are can be useful for some analysis, but maybe you don't know what all those phonemes are yet. Or maybe you want to describe different dialects of the same language. Sometimes, you really need that extra detail. So, we can use the IPA to do more or less specific transcription. Let’s run this through with an example, say, a name like “Tamika.” Now, there are many different things that make up a word like “Tamika”. But to start off with, maybe all you want are those basic sounds, the phonemes. And, you need to add stress, too, which is the emphasized syllable. So, just the phonemes plus stress make up what we call “broad transcription”. Oh, and by the way, that flipped e is called a schwa,and it’s the secret key to English pronunciation. It’s just this little sound that turns up almost everywhere when we have a vowel that doesn't have emphasis on it. But maybe if you were listening closely, you heard that puff of air that came with the first sound in Tamika. And, maybe you want to include that. Maybe you also noticed that the first schwa is more nasal than the other one, and you want to mark that down too. If you include all the bits about pronunciation that are predictable based on what's going on elsewhere in the word - so, the allophones - you end up with something known as “narrow transcription”. Narrow transcription is crammed full of information, which is great, but you probably don’t want to work with it all the time, especially when that extra information isn’t relevant, because it can get pretty bulky. But beyond how awesome the IPA is for transcription, even the consonant and vowel charts themselves are great. The way they’re put together actually gives you information about the sounds in them, too. They’re kind of like the Periodic Table of the Elements that way! So if you look at the consonant chart, the rows are different ways humans can push air through their vocal tract to make sounds, from more closed to more free air flow. And the columns are the different places in the mouth that you can make those sounds - so, lips, like p and b, teeth like th, the soft palate of your mouth like k, and so on. We’ll come back to the different categories of speech sounds in a later episode, but for now, just looking at the chart makes it clear how much we divide up the space inside our mouths - how just a little difference can make a big meaning change, in some language somewhere. And if you thought the consonant chart was cool, the vowel chart is even cooler.This weird trapezoid is a rectangular oddity for a reason: it represents the inside of your mouth, and all the different ways you can make vowels. So an [i] has your tongue high up, near the front of your mouth, while an [ɑ] has it lower down and towards the back. It’s amazing! Just looking at the chart tells you where you should be pronouncing the sound, even if it isn’t always as easy to make it as it looks. Learning to spot and pronounce all the different speech sounds of the world might seem like a daunting task, but if you work it out then you can hear anything anywhere. And that’s a reward worthy of the challenge. So we’ve reached the end of the Ling Space for this week. If you were transcribing my speech sounds correctly, you learned that the IPA was invented to capture all the variation in all the languages in all the world; that you can transcribe broadly, with just phonemes, or narrowly, with all of the allophones; and that the consonant and vowel charts strive to actually depict what is going on in your face when you're making the different sounds of your language. The Ling Space is produced by me, Moti Lieberman. It’s directed by Adèle-Élise Prévost, and it's written by both of us. Our production assistant is Georges Coulombe, our music and sound design is by Shane Turner, and our graphics team is atelierMUSE. We’re down in the comments below, or you can bring the discussion back over to our website, where we have some extra material on this topic. Check us out on Tumblr, Twitter, and Facebook, and if you want to keep expanding your own personal Ling Space, please subscribe. And we’ll see you next Wednesday. Hasta luego!