In the last part of this series, I discussed how to create a consonant inventory. So in this part, I will be covering vowels.

One major misconception about vowels is that English has five vowels. Actually, written English has five letters that represent vowels (sometimes ‘y’ can as well), but phonetically there are more than five. The reason why there are only five vowel letters is that Latin had a five-vowel system.

Below, I have listed the vowels present in Received Pronunciation to show how many vowels are really present in English. Vowels can vary heavily among different varieties of English.

	Front	Central	Back
Close	iː		uː
Near-close	ɪ		ʊ
Close-mid	e
Mid		ə
Open-mid		ɜː	ɔː
Near-open	æ	ɐ
Open			ɑː,ɒ

So what is vowel? I remember asking this to my English teacher once. He never really gave me a straight answer, saying something along the lines of being a core component of a syllable. This is partially true, as vowels tend to be in the nucleus of a syllable. However, phonemes that are generally considered to be consonants can also be found in the nucleus, these are called syllabic consonants. The ‘m’ in the word ‘rhythm’ could be considered one of these.

Another way to define a vowel is by its phonetic features. Vowels are continuous sounds where at least some of the air is released through the mouth, along the centre of the tongue. Unlike consonants, vowels have no significant points of closure above the glottis. However, the consonants [j] and [w] fit this definition and may occur outside the nucleus of a syllable. These are called semivowels and are the consonant equivalents of the vowels [i] and [u] respectively.

Cardinal Vowels

The IPA symbols featured in the chart above are called cardinal vowels. Again, I recommend that you take a look at this interactive IPA chart to get a feel for how these sound. These are not vowels found in any particular language but are more like reference points. The IPA symbols for cardinal vowels are used to approximate the true value of a vowel in a broad IPA transcription. Certain diacritic marks can be used to more precisely transcribe vowels.

Vowels have three basic variables: height, backness and roundness. Height ranges from closed (high) to open (low). Closed vowels are equivalent to approximants. There is also backness, which ranges from front to back. Front vowels are equivalent to palatal consonants and back vowels are equivalent to velar consonants. Finally, vowels can be articulated with either rounded or unrounded lips. Rounded vowels are the equivalent of labialised consonants.

Basic Vowel Systems

The first step in creating a vowel system is to divide up vowel space. Latin divides its vowel space into five, while other languages, such as classical Arabic, have as low as three basic vowels. I recommend choosing a number between three and five.

Usually, languages make distinctions between height, backness and rounding. However, some languages only distinguish between height. An extreme example of this is Ubykh, a north-west Caucasian language which has only two vowels: /ə/ and /a/. However, Ubykh has eighty-four consonants, which a lot of them involve palatalisation or labialisation. This suggests that Ubykh’s vowels may have used to have distinctions in backness and roundness, but those features were absorbed into the consonants and then lost in the vowels.

Modifying vowels

The next step is to add additional vowels by modifying vowels from the basic set you have chosen.

Rounding

One way to add more vowels is by rounding or unrounding existing vowels. Front vowels are usually unrounded, while back vowels are usually rounded. A good way to add extra vowels is to round or unround an existing vowel. For example, French has a series of unrounded front vowels: /i/, /e/ and /ɛ/; each one of these has a rounded equivalent: /y/, /ø/, /œ/.

Length

A lot of languages make a distinction between long and short vowels. Sometimes there is just a distinction in length, like in Finnish. But sometimes there is also a slight difference in quality. For example, Latin’s short vowels are more open and central in quality than their long equivalents.

Nasalisation

If the velum is lowered during the articulation of a vowel, then some air will be released through the nose. This is called nasalisation. Ordinary vowels may become nasal vowels if they are next to a nasal consonant. The nasal consonant then has to be lost or denasalised for the nasal vowels to become phonemic. French has four nasal vowels: /ɛ̃/, /œ̃/, /ɑ̃/, and /ɔ̃/. Note that all of these vowels are relatively open. This is because nasalisation may cause vowels to become more open.

R-coloured Vowels

R-coloured vowels are vowels that are modified in a way that gives them an R-like acoustic quality. There are two ways to do this: retroflexion and rhotacisation.

Retroflexion is when a vowel is articulated with the tip of the tongue curled back. This is more common with open vowels.

Rhotacisation is a more complex process but produces a similar effect. The main difference from retroflexion is that the root of the tongue is pulled back.

Phonation

In the last part, I mentioned voiced and voiceless consonants. This is known as phonation, and it applies to vowels too. Vowels are usually voiced, but more rarely they can be unvoiced. There are other kinds of phonation, such as breathy voice and creaky voice.

Phonation depends on how open the glottis is during articulation. Voiceless phonemes have an open glottis. Voiced phonemes have the glottis closed tight enough so there is maximum vibration. If the glottis is slightly more open then it will produce a breathy voice. If the glottis is slightly tighter, then it will produce a creaky voice.

Diphthongs

Diphthongs are when two vowels occur in one syllable. These are articulated by moving the tongue from one position to another during vowel articulation. Single vowels are contrastively called monophthongs.

Pitch and volume may change across the duration of the diphthong. So they can be classified into two types based on which vowel is stronger. Falling diphthongs have a stronger first vowel, while diphthongs with a stronger second vowel are called rising diphthongs. If the weaker vowel is a close vowel, then it may it interpreted as the equivalent semi-vowel.

Another way to classify diphthongs is by how they move through vowel space. In opening diphthongs, the final vowel is more open than the first vowel. In closing diphthongs, the final vowel is more closed than the first vowel. Height-harmonic diphthongs are when both vowels have the same height, but these are rarer than closing and opening diphthongs. There are also centring diphthongs, which the final vowel is more central than the first one.

When choosing diphthongs, the classifications mentioned above can be used as a set of restrictions. For example, in received pronunciation, there are only closing and centring vowels.

The components of diphthongs don’t need to be present as monophthongs. General American English has the diphthongs /eɪ/ and /oʊ/, but /e/ and /o/ are absent as monophthongs.

More than two vowels can occur in the same syllable. If three vowels occur in the same syllable, then it is called a triphthong. Triphthongs occur in Received Pronunciation, for example in the words hour /aʊə/ and fire /faɪə/.

Vowel Harmony

Finally, I will explain an interesting phenomenon involving vowels: vowel harmony. Languages with vowel harmony have restrictions on which vowels can occur near each other.

For example, Finnish sorts its vowels into three groups: front (/y/,/ø/ and /æ/), back (/u/, /o/ and /ɑ/) and neutral (/i/ and /e/). A vowel from the front group and a vowel from the back group cannot occur in the same word, but a vowel from the neutral group can occur in the same word with a vowel from any group. Any front or back vowels in a suffix will change to its equivalent in the group of the previous vowel. The exceptions to this are compounds and recent loan words.

Afterword

The main resources I have used and recommend are J.C Catford’s ‘A Practical Introduction to Phonetics’; and Bernard Comrie’s ‘The World’s Major Languages’.

So far, I have discussed phonetic features that can be analysed as segments of a word or syllable. However, next time I will begin to look at features that belong to a whole syllable or word: suprasegmentals.