Today I’d like to spend a little time on something kind of technical: phonology. Phonology （音韻論） is a branch of linguistics that focuses on the sounds of a language and how they’re converted into grammatical units. It’s distinct from phonetics （音声学）, which studies the production and decoding of sounds from a physical, psychological, biological, and acoustic point of view. In short, phonetics studies how we make and receive sounds in language, and phonology studies how we use them to express meaning. Now, phonology is a huge subject, and I’m just scraping the surface of it myself, but I’d like to share some of those surface-level revelations.
The basic unit of speech sounds is the phoneme （音素）. Phonemes form a single sound concept in the speaker’s mind, like /s/, /k/, or /m/. However, that is not to say that a phoneme is a single sound in reality! In fact, we often use multiple sounds to express the same phoneme. These sounds we call allophones （異音）. Take the words “cool” and “language”, for example. Say them out loud, and pay attention to where your tongue is in your mouth. Is the /l/ sound in both words exactly the same? For the vast majority of native English speakers, those two /l/ sounds are actually entirely different. “Cool” ends with a sound called a “dark l”, expressed [ɫ] in IPA. “Language”, on the other hand, starts with a “clear l”, expressed [l] in IPA. Conceptually, they’re the same sound to a native speaker, the same phoneme, but physically speaking, they’re different sounds, or allophones. For Japanese speakers, you can examine the difference in how you pronounce the ン sounds in 原因, 本物, and 簡単.
Phonemes then join together into syllables （音節）. Languages differ on how many phonemes can make up a syllable and their arrangements, but there are some common patterns. The middle part of the syllable, which is usually a vowel, is called the nucleus. Sounds before the nucleus are the onset, and sounds after the nucleus are the coda. In the word “clamp”, for instance, /kl/ is the onset, /æ/ is the nucleus, and /mp/ is the coda. English is quite free in what it will allow in a syllable compared to many other languages. In English, you always need a vowel in the nucleus, but the onset and coda are optional. Moreover, English allows multiple sounds in the onset and coda, as shown in “clamp” above. Contrast this with the much more constrained syllable rules of Japanese. In Japanese, the nucleus (as in all languages) must be present, but codas are completely non-existent. Only three kinds of syllables are allowed in Japanese: a single vowel, like /a,e,u,e,o/, a consonant onset plus nucleus, like /ka,ki,ku,ke,ko/, or a nasal nucleus alone, /n/. Interestingly, there seems to be no language that allows codas but doesn’t allow onsets. If a language will allow “ack”, it will always also allow “ka”—at least, of the languages documented so far.
If you’re a language learner, do yourself a favor and delve as deep as you can into understanding the sound rules of your target language. Nothing will do more for your pronunciation, in my opinion, than getting a firm grip on which sounds are in your target language, and how to produce them. And if you’re interested in phonetics and phonology, you might want to pick up Introducing Phonetics and Phonology, 3rd Ed. I’m reading through it for a class at the moment, and finding it pretty accessible.