1) Some vowels and consonants are more common than others. The most common vowels are /i/, /a/, and /u/*, since they're the most distinct, and no language has less than three vowels.Common consonants are trickier to pinpoint, since there are so many more of them than there are vowels. Most languages will select from the plosive, fricative, and approximant rows of the IPA chart, with /s/, /p/, /t/, and /k/ being common. /m/ and /n/ also show up frequently. Languages will often, but not always, have multiple sounds in the same area of the mouth, and will often match up voiced and voiceless sounds—/p/ and /b/, for instance.
2) Sonorance refers to the amplitude of a speech sound. Languages tend to group syllables according to a sonority hierarchy, so that quiet sounds (like /t/) are on the outside and loud sounds (like /a/ and /n/) are in the middle. Not every language has the hierarchy I linked to, however, so you can make up your own if you want. Just bear in mind that the hierarchy exists because of how sounds carry.
3) A minimal pair is a pair of words that have distinct meanings but only differ by one sound. They're a great way to figure out which sounds are coded to the language and which are interchangeable (and therefore don't get you killed if you mispronounce them). For instance, deaf and death are a minimal pair (with dead, debt, den, deck). "I am deaf" and "I am death" have radically different meanings. Your average xenolinguist is going to be interested in minimal pairs because of what the pairs say about the sounds of a language.
Since some languages distinguish between normal and nasalized vowels**, or the length of a vowel or consonant, or whether /p/ comes with a puff of air or not, there can be all kinds of minimal pairs and linguists get really good at distinguishing pronunciations.***
4) Not all sounds are distinctive. Sometimes you'll find Sound A only in one "environment" and Sound B everywhere else. For instance, you never find a nasal vowel before a non-nasal consonant in English. If I deliberately did that, it would sound strange, like I had an accent. These kinds of sounds are called allophones and are said to be in complimentary distribution.**** Every language has rules to determine which allophone happens when.
Rules For Changing (Non-Distinctive) Sounds
- As a general rule, two sounds side by side will try to becoming more similar. Two reasons for this: It's easier to pronounce sounds when they're in the same part of the mouth, and humans have a "bad habit" of prepping for sounds before we get around to making them. This is why we get nasalized vowels before nasal consonants, and why [mp], [nd], and [ngk]***** occur a lot of the time (same spot in the mouth for each sound). Any time you want to have a sound change rule in your language, write it out for reference.
- Some languages have rules that all vowels in a word must be made the same way as the first vowel (rounded, for instance), or that if the first consonant is made in the throat, all the consonants need some throatiness to them. These are harmony rules, and they're an extension of the assimilation rules in the first bullet.
- Sometimes you'll get rules where two adjacent sounds which are a lot alike being less similar, generally so that it's harder to get confused about which word's being said, or because the original sequence of sounds is a tongue twister.
- Other rules allow us to add in or drop out sounds/syllables to make words easier to say.
- All languages have a mix of these rule types, with several (at least) in each category. I'd suggest doing some deeper research into common changes before coming up with ones for your language (start here).
Want examples for all this?
Sonorance in this language (let's call it Pamak) is going to follow the chart I linked to. Why? Because it's what English uses and I want people to feel comfortable saying the words. If I wanted the language to feel more alien, I'd use a different hierarchy.
Minimal pairs in Pamak can involve any consonant or vowel (pa vs. ba, mifs vs. mafs). Pamak doesn't distinguish between nasal and non-nasal vowels, or puffs of air, or length of sounds.
For familiarity for readers, Pamak stress is iambic, like in Shakespeare: shall I comPARE thee TO a SUMmer's DAY….†
1. Vowels will be nasal before nasal consonants.
2. A nasal consonant will always be produced in the same spot as the consonant that comes after it, if there is a consonant.
3. No other consonant needs to match for place of articulation. (This isn't a rule in the linguistic sense, but I'm writing it to remind myself.)
4. Any consonant preceding a high vowel (/i/ or /u/) will be palatalized, meaning that a faint y (/j/) sound gets added. Think "cute" [kjut].
5. If two adjacent consonants are made in the same spot, but one is voiced and the other is voiceless, then the second consonant will change its voicing to match the first, unless the second consonant is a nasal. If the second consonant is a nasal, the first consonant doesn't change.
6. Unstressed vowels should be dropped if the word is 4+ syllables and dropping the vowel won't result in a non-syllable. Remember that Pamak syllables have one of the following structures: CV CVC CCVC CVCC CCVCC (repeat all with VV). By the time this rule, and rule 7, are applied, all the sound changes have happened.
7. If two adjacent syllables result in a very awkward consonant cluster such as xpxm, add /e/ in the middle of the cluster.
Rules 1, 3, and 4 won't be rules reflected in the writing system. They'll be spoken-words only. Rules 2, 5, 6, and 7 will be written and spoken. Examples:
- Rule 2: klen + pa --> klenta, vlan + klen --> vlangklen, san + mis --> sammis
- Rule 5: bim + fa --> bimva, tlant + bal --> tlantpal
- Rule 6: klen + mamf + ba + vlask --> klemmamvbavlask --> klmmamvbavlask, not *klmmamvbvlask
- Rule 7: vlaxt + xmal --> vlaxtexmal, not *vlaxtxmal
Get the idea? (For the record, these are examples of combinations. They may or may not be actual words. I'll get to that next time.)
One more thing before I finish: some languages use tones/pitches as a "speech sound", often with associated minimal pairs. Mandarin Chinese uses fives tones (rising, falling, level, dipping, neutral), for instance. So that's another thing to play around with.
Before starting with full-out word formation, it's a good idea to get the basics of the sound system laid out, so you have a better idea of what sounds should go where. That said, I tend to do some sound stuff, some word stuff, then sounds, then words, as I become aware of other rules I want or need.
See you next time!
* As always in this series of posts, letters in brackets will represent sounds in the phonetic chart from Part 1. The slashes and brackets are a linguist thing, in case you were wondering. You don't really need to know the difference for this overview.
** Nasalized vowels being those that generally come before nasal consonants, because the route through the nose is already open. English pat has a normal vowel; English pan has a nasal one.
*** Hence all the funky marks at the bottom of the IPA chart, and the four month course I took on phonetics.
**** I'm totally giving you these terms just so you can look them up if you want to.
***** /ng/ = the sound in thing
† Stress can also be trochaic (TIger, TIger, BURning BRIGHT…), or only occur on the first or last syllable, or be a strange mix of things which gets really complicated and is taught to advanced third-year students.