In this post, I’ll cover the vowel chart. The IPA divides up sounds based on their articulations, and vowels and consonants have fundamentally different kinds of articulation. In particular, consonants are sounds produced with obstruction in the vocal tract, while vowels are sound produced without any obstruction.
Personally, I have always found vowels harder to explain than consonants. This is partly because vowels can vary tremendously between English dialects, so it can be difficult to find example words that will make sense to everyone. I’m a speaker of what you could call Standard Canadian English, so all my examples will be drawn from my dialect. If some pairs of words don’t work at all for you, tell me in the comments, and I’ll try to update the article to reflect more dialects. There’s also just my personal perception of vowels – I find their phonetics are very “squishy” and hard to pin down, whereas consonants are a little more “firm” and reliable. But that’s just me.
Vowels are organized into a little chart that is quite a different shape from the consonant chart. This is sometimes referred to as the vowel quadrant. It is also common to come across descriptions of a “vowel space”, in reference to the vowels of a particular language.
This chart is intended to roughly represent the physical space inside your mouth just like the consonant chart. The left of the chart represents the portion of your mouth closer to the lips, and the right side is the back of the mouth. The top of the chart is the roof of the mouth and the bottom of the chart is the jaw.
Because vowels and consonants are articulated in different ways, they cannot be described with the same features, and hence their charts are not organized the same way. Vowels, by definition, lack any obstruction in the oral tract, so it is impossible to describe them in terms of place or manner of articulation.
Vowels are always voiced by default, so voicing is not represented on the chart either. There are languages with voiceless vowels, but these are always predictable variants of normal voiced vowels. For example, in Japanese, vowels are predictably devoiced between voiceless obstruents. There are no pairs of words that differ only by the voicing of the vowel.
The three main vowel features are height, backness, and roundedness. Height and backness refer to tongue position. Roundedness refers to lip shape. When describing a vowel, the convention is to put information in that order as well. For example, the sound that [i] represents is usually a “high front unrounded vowels”, not “front unrounded high vowel”. Accordingly, I’ll describe each of these feature in more detail, in that order.
Vowels on the top row of the IPA chart are called “high vowels”. Vowels in the bottom row are called “low vowels”. The vowels in the middle rows are called “mid vowels”. I know that’s not what the IPA chart actually says, but the terms “high” and “low” super common and you should know them. They refer to the height of your tongue, relative to resting state. If you are just sitting still, not talking or chewing, your tongue will be in the mid position. To make a high vowel, you need to move your tongue up toward the roof the mouth – exactly which part of your tongue moves depends on which vowel. To make a low vowel, you tongue needs to move down, which is usually accompanied by jaw lowering. Differences in vowel height appear to be universal in human languages. Even the language with the smallest known vowel inventories, Ubykh, with possibly only 2 vowels, still had a height distinction.
The IPA calls high vowels “close” vowels, and low vowels are “open” vowels. I won’t use these terms, but to remember them, it’s useful to imagine your mouth as a tube with air flowing through: it is more closed up with a high vowel, when your tongue is near the roof of the mouth, and it is more wide open with low vowels when you lower your jaw.
English has four high vowels, represented as /i,ɪ,u,ʊ/. They are in the words leak, lick, Luke, look. Hooray for English spelling. I’ll come back to the difference between /i,ɪ/ and /u,ʊ/ towards the end of this post. English also has two low vowels: /ɑ/ and /æ/ in pot and pat.
A note about the vowel /ɑ/. This is, according to the IPA chart, the correct symbol for the vowel in English words like father, lot, pot, etc. However, linguists writing about English commonly use /a/ instead, even though this is technically a different vowel. This is pretty much just because it’s just easier to type /a/, since to use /ɑ/ you need to go into special symbol menus. Phoneticians tend to be a bit more careful about which symbol they select.
You can get a feel for the high distinction by trying the vowels [i] and [æ], which are in the English words beet and bat respectively (assuming you speak a Canadian-like dialect). Just say [iiiæææiiiæææ]. Pay attention to two things: (1) how your jaw lowers for [æ], (2) how close your tongue gets to your palate for [i].
English also has mid-vowels. The IPA vowel [e] appears in the word bay, the vowel [o] in boat, for example. The tongue and jaw position is, well, mid-way between high and low. Mid-vowels are a little harder to get a feel for. This is because low and high-vowels have limits imposed by your physiology. Your lowest vowel is limited by how low your jaw goes. Your highest vowel is limited by contact with the palate; once you make contact it’s a consonant, not a vowel any more. Mid-vowels occupy a space in between, and if you go too high or too low out of that zone, you end up producing a different vowel.
Actually, in my dialect, these vowels are always followed by a glide, so really my mid-vowels are /ej/ and /ow/. This is a very noticeable difference between Canadian and Irish accents: Irish truly has /e/ and /o/ without the following glides. If for some reason you have a speaker of Canadian English and a speaker of Irish English together, ask them to say “oh no potato boat” and “hey they say she’s great” and you can really hear the difference. Then report back here.
I should also point out that official IPA chart distinguishes two degree of “mid-ness”. This is not necessarily how every linguist does things. It may depend on the language. For instance, descriptions of English often use just a single mid-vowel level. If you are reading this blog and taking a linguistics class at the same time, I recommend that you use whatever method is in your textbook, or go clarify with your instructor.
You may be wondering why the vowel chart is not a square. Why are the front vowels on a diagonal? This is intended to represent articulatory facts. Your jaw swings on a hinge to lower down, it doesn’t drop down like a lift, so the vowel space isn’t a square, and it’s better represented with this little slant for the front vowels. Of course, the whole thing is an abstraction anyway, because it’s not like the top of your mouth is a straight line either. You may also come across different shapes when reading about specific languages. For example, the 3-vowel system /i,u,a/ is extremely common, and is sometimes represented as something more like a triangle.
That covers height, now on to backness. Vowels on the right-hand side of the chart are called back vowels. Those on the left-hand side are called “front” vowels. Those in the middle are called “central” vowels (not “mid” because that’s for height distinctions). You can get a feel for this difference with the English vowels [æ] and [ɑ], which appear in the words bat and bought. Just say [æææɑɑɑæææɑɑɑ]. Notice how your tongue has to be more retracted for [ɑ] than for [æ].
My dialect of English has 5 front vowels /i,ɪ,e,ɛ,æ/ in beat, bit, bait, bet, bat. There are also five back vowels /u,ʊ,o,ɑ/ in mood, mud, mode, mod (as in “moderator”).
Some of these symbols have names. The symbols /ɛ/ is called “epsilon”. The symbol /ʊ/ is “hooked-u”. The symbol /æ/ is sometimes called “ash”. Also, the symbol /æ/ is intended to be drawn as a single character, not as a back-to-back /ae/. The IPA is supposed to be a one-to-one system: one sound = one symbol. Writing /ae/ risks confusing the reader into thinking that there are two vowel sounds, not one. You can draw it as one character by starting with the top-left of the hook in the ‘a’, looping around into the ‘e’, then looping around back without taking your pencil off the paper.
Now for a different experiment, alternate the high vowels [i] and [u] in read and rude. Just say [iiiuuuuiiiuuu]. You will notice two differences: (1) your tongue has to retract somewhat for [u], because it’s a back vowel and (2) your lips become rounded.
Roundedness is the third major vowel feature and it describes your lips and not your tongue. On the IPA chart, rounding is indicated by where a vowel is placed. When two vowels appear in the same location on a chart, the vowel on the left is unrounded, the vowel on the right is rounded. The front and central vowels that appear alone are unrounded. The back vowel that appears alone is rounded.
English does not have any pairs of vowels that contrast based on rounding. Rounding is predictable from backness: Back vowels are generally all rounded in English, whereas front vowels are not. In fact, rounded front vowels are relatively rare around the world, although they can be found in some major world languages. For example in French:
[ty] tu 2nd person singular subject pronoun (high front rounded vowel)
[ʒœn] jeune ‘young’ (mid front rounded vowel)
Beyond height, backness, and rounding, there is a fourth feature you’ll see on the chart, and one that I alluded to earlier and avoided because it’s a bit messy. What are we to make of the distinction between [i] and [ɪ]? They are both high, front, unrounded vowels. The difference between these vowels depends somewhat on which language is being described. English has both of these sounds. Here are some [i] vs. [ɪ] pairs: beat/bit, Pete/pit, seat/sit, eat/it. Traditionally in English, [i] is described as a “tense” vowel and [ɪ] is described as a “lax” vowel.
There are three lax vowels in English: [ɪ] in bit, [ɛ] in bet and [ʊ] in book. All others vowels are considered tense. What makes these sounds articulatorily distinct is more than I want to go into here. They do all share a particular phonotactic restriction though: there aren’t any English words that end with any of these sounds. Words like [bɪ], [samɪ], [kopɛ] or [trompʊ] don’t exist. (There is one possible exception to this, which is the exclamation spelled “meh”, which would probably be [mɛ] in the IPA.)
This distinction between these vowels is important in many Niger-Congo languages, but a different terminology is used. In these languages, the vowels [ɪ,ɛ,ʊ,ɔ] are often grouped together and called “retracted tongue root” vowels, or [-ATR] whereas the vowels [i,e,u,o] are considered “advanced tongue root” or [+ATR]. The particular patterning the vowels depends on the language.
Lastly, let’s consider the mid-central unrounded vowel /ə/, usually known as “schwa”. This is the most neutral vowel you can make, and it probably exists in all languages. Put your tongue in resting position, and then make a vowel. It’s the sound you might make if you were put on the spot and didn’t know the answer to the question. There’s a lame joke in linguistics that goes like this:
Phonetics professor: What is the unstressed mid-central unrounded vowel?
Phonetics professor: Exactly!
In English, and many other languages, vowels which are unstressed often become schwa. Consider the pair Canada and Canadian. In Canada, the stress is on the first vowel, and it would be transcribed [kænədə]. The suffix –ian changes the stress, and it moves to the second syllable, and the words is [kənediən]. Notice how the first and second vowels vary between schwa and something else, depending on the stress.
This pretty much covers the entire chart. Every vowel is some combination of high/low, back/front, round/unround, and maybe tense/lax. There aren’t quite as many distinctive feature to cover as with consonants. However, this doesn’t mean that’s all there is to say about vowels. What I’ve covered here are the basic “segmental” features of vowels. There are a great many things you can do with your vocal tract to vary the vowels beyond these four things, but all of the other things are do-able without fundamentally changing any of these segmental features.
Let’s take nasality as an example. All the vowels described on this chart are “oral” vowels, meaning that the velum (soft palate) is raised, blocking off the nose, and air flows only out the mouth. It is also possible to lower the velum and produce a nasal vowel, where air flows out the nose in addition to the mouth. In principle, any of the vowels on the IPA chart can be produced as a nasal vowel or as an oral vowel.
Nasality depends on the position of the velum, which doesn’t really interact with tongue position. The IPA doesn’t need a whole new set of symbols for nasal vowels. Instead, an accent mark is used, though the technical term is a “diacritic”. In this case specifically the tilde is used above a vowel. For example:
Oral vowels: /i,e,a/
Nasal vowels: /ĩ,ẽ,ã/.
Diacritics are used in cases where it would be confusing or unnecessary if completely new symbols were introduced. There’s no need for special nasal vowel symbols, since it would obscure an articulatory relationship between vowels. We can tell from looking at /i/ and /ĩ/ that they are similar vowels, and that’s very useful.
There are dozens of diacritics in the IPA, used for both consonants and vowels. I will not be covering them here, but they will be the subject of the next part in this series.