I'm wandering in the library linguistics section looking for books worth reading. It's more like the Internet than you might think. You start out following one link, the shelving number of the book from the catalogue, but you can accidentally get distracted by things like the children's story of The Whiting and the Flatfish, written in parallel text in Russian, English and the endangered Siberian language Ös. (I'd tell you, but I already forgot how the flatfish became flat. I guess it wasn't that good a story. The pictures were nice, though.) You pick up a book and the table of contents leads you to something interesting, so you read that chapter and then follow the bibliography to more on the same subject. Sometimes you pick up a book and discover that the author knows less than you on the subject, has an irritating writing style, or has dedicated pages to ranting about the shortcomings of some other author. Sometimes the links are dead, and there is no book with that call number at that location. Sometimes you pick up a promising-seeming book and when you open it discover it's typeset in non-proportional Courier. Seriously. This is late 1970s technology at work. You even get the problem of inane comments. Almost all the books I read have pencil marks in them underlining passages some reader found important. Some of the books are in languages I can't read. (Can't as in don't know how to as opposed to can't as in it would drive me insane, like the ones that are just typewritten manuscripts someone slapped a binding on).
I pick up a textbook by David Odden. Introducing Phonology. Cambridge University Press: 2005. I'm hoping to learn about the muscles and other anatomy underlying the interesting phonology course. I had no idea that linguistics would go so deeply into how language is produced and interpreted.
This book doesn't go where I was hoping, but I read it anyway. (I have to stop doing that). Sounds are variations in air pressure. The vowels are periodic waves defined by amplitude and frequency. Prominent frequencies are those that resonate and which those are is related to the shape of the vocal tract. We use our speech apparatus, tongue, lips, and velum to change the shape of the vocal tract and produce the frequencies we want. A longer tube gives a lower resonance frequency, which is why when you blow in a recorder it sounds all squeaky until you can figure out how to get your fingers over all the holes.
Sound is a continuous property; soundwaves show little in the way of breaks between letters. Spectrograms show more than is needed for language analysis. We need a way to represent speech such that speakers of a language agree it is the same word, when it is, but that shows differences that are important to distinguish between words in a language. An assumption in phonetics is that there are systematic limits on what constitutes a speech sound in a human language. My problem with linguistics textbooks is you have to know how the author talks when they say "the æ sound is like a in hat." There are a lot of different ways English speakers pronounce /hat/.
Vowels are classed by height (high-mid-low) and vowels at the high and mid positions can be tense or lax. Any vowel can be rounded or unrounded, and front, centre or back. It is common, but not necessary for back vowels to have lip rounding. So multiply that out and it makes five times three times two equals thirty possible vowel sounds. And then there are more multipliers like nasalization, length, stress, tone, phonation (i.e. creaky or breathy), and glides such as y and w sounds.
There are eleven places of articulation for consonants: bilabial, labiodental, dental, alveolar, alveopalatal, retroflex, palatal, velar, uvular, pharyngeal, laryngeal-glottal. A consonant may have a primary and secondary place of articulation. There is also a manner of articulation: stop, fricative, nasal, or affricate, and a consonant can be voiced or unvoiced.
The chapter discusses laryngeal consonants, comparing series like p b pʰ bʰ bʱ b̤ʱ p’ ɓ according to movement of the larynx, but without the anatomical detail I was looking for. This whole chapter is leading towards the International Phonetic Alphabet, which is all very useful, but not what I was looking for. My problem is that this is a book on phonology and what I need is a book on phonetics. Here's the difference. With my luck my next try will be a book on phonics. Or telephony. In courier extra faint. Or maybe that early-1970s pseudo-computer font.