Wednesday, February 17, 2010

Build-Yer-Own Language (Part 3)

Part 1
Part 2

Now that you know how your language's syllables work and I haven't scared you off, you're going to want to make words. This is good, because words are awesome. Occasionally confusing, but awesome. The linguistic study of word formation is called morphology.

(I apologize in advance for all the bold font.)

Linguists distinguish between lexemes, or the core meanings of a chunk of sound (cat + cats = 1 lexeme, because they both refer to the four-legged, tailed animal that meows), and word forms, which are variations on lexemes. Cat and cats are different word forms because one is singular and one is plural, with an added -s. Having explained the difference between a lexeme and a word form, I'm now going to forget I did so and just use 'word' most of the time.

Word forms can be grouped into paradigms, or lists of all variations on a single lexeme. For instance, go, goes, going, gone, went is the paradigm for "to go". For the language maker, paradigms can be useful as memory aids, especially if there are complicated word-making processes in the language.

Each piece of sound that goes into a word form is called a morpheme. Morphemes also happen to be what we code into our brains. We don't code going and being and seeing and dancing individually. We code go, be, see, dance, and -ing, because we've all learned (somehow) that there's a rule for when -ing shows up, meaning that -ing is a separate entity.

There are two types of morphemes: root morphemes, or the ones that make up the basic meaning (cat, go), and affixes, which add extra meaning (-s, -ing). You're probably not familiar with the word affix, and this is fine—prefixes, suffixes, and infixes are the three kinds of affixes out there.

Wait, you don't know infix either? That's a chuck of sound that gets slipped into the middle of a word. Abso-freaking-lutely <-- like that.

Remember my last post in this series, where I talked about sound rules and how sometimes sounds get changed around to make syllables easier to say? They apply to morphemes too, especially affixes. Sometimes those sound rules intervene to make two or more allomorphs with an identical meaning. We've got some of these in English, of course*— cats, beds, and dishes have three different sounds that all mean 'plural'.

Ah yes, you say, this is all fascinating, but how do we use this to make words? There are several ways, and most languages use a combo or all of them.

  • derivation - making a new word by adding an affix with non-grammatical meaning: form --> formation, derive --> derivative, fiction --> fictional, absolute --> abso-freaking-lute-ly
  • inflection - making a new word by adding an affix with grammatical meaning (i.e., plurals, gender markings, tense, possession): cat --> cats, go --> going, mom --> mom's
  • compounding - putting two words together to make new meaning: dish + washer --> dishwasher, cat + burglar --> cat burglar 
  • reduplication - copying some or all of a word/morpheme as a way of changing meaning, sometimes with slight differences in sounds - politics --> politics-schmolitics, rolly --> rolly-polly, doggy --> doggy-woggy

You'll also find irregular paradigms, where nothing really has affixes but the words change anyway. Go, gone, went is an example, and so is they, them, their, theirs, and foot, feet.** Other languages change stress or vowel quality as a way of signaling meaning. Also note that you can do a bunch of these to a single word, multiple times, in any order: anti-dis-establish-ment-ar-ian-ism.

Your made-up language will most likely use derivation and inflection, so you'll need to come up with affixes for common meanings. Start with a syllable (or several) for the root, and then add a syllable or group of sounds, then make a note that what you added means "X meaning". That way you'll be consistent with other words. Derivational affixes can have just about any meaning you want, but inflectional ones are going to stick to a few categories: tense, mood, voice, aspect, person, number, gender, and case. (Look up any term there you don't recognize. I'm not going to bore you any more here.)

One more thing to bear in mind at this stage: type of language. This is important for a couple reasons. 1) The language type is going to determine what kinds of word formation processes you're going to need, and how many of the various types of affixes you'll have to come up with. 2) The language type you choose will have connotations to many readers, in that they'll see certain features and think of certain languages. You can exploit this by choosing sounds that will remind the readers even more of a given language.***

Types of Languages

  1. Isolating - Every word is a morpheme; every morpheme is a word. There won't be any affixes, but those meanings will still need to be there, as separate words or in the word order. Alternatively, those meanings could be "irrelevant" to the speakers, so that past/present/future is all the same verb form, or there's no concept of gender or number. There will still probably be a way to signal case (a.k.a. direction), because we'll need to know who hit who with what where. Compounds are still possible. Examples: Chinese, other South-East Asian languages.
  2. Agglutinative - Most if not all words are formed by adding lots of morphemes together. There's probably a morpheme (with allomorphs) for most kinds of inflection, and plenty for derivation too. Every morpheme will have a single meaning. The words will tend to be mid-length to longish. Examples: Japanese, Bantu, Turkish, Quenya, Klingon
  3. Fusional/Inflecting - Also uses lots of affixes, but the affixes are more likely to have multiple meanings, such as "continuous + past" or "male + plural + indirect object". There will be lots of paradigms and declension tables, not just for verbs, but also nouns, pronouns, and adjectives. Examples: Latin, Greek, German, Hindu
  4. Polysynthetic - Really big strings of morphemes, to the extent that one or two words can contain a sentence's worth of meaning. Subjects, objects, adjectives, adverbs, pronouns, spatial relationships, etc. are all joined together into a single word. Some polysynthetic languages keep nouns and verbs in separate words; others done. There's often a very loose distinction between what's a verb and what's a noun. Examples: Inuktitut, Cherokee, Ainu
  5. Semitic - Not technically a language type (they're fusional), but cool enough to deserve its own paragraph. This language family uses something called "transfixes" pretty heavily— strings of vowels that break up the consonant roots and hold meaning. Sometimes the meaning is derivational, sometimes inflectional. For instance, there's the Arabic kataba, katabnaa, yaktubu, naktubu, kaatib. k-t-b means "write". Examples: Hebrew, Arabic

Few if any languages are purely one type. They tend to mix things up a little. English is a mix of isolating and fusional, for instance. Languages also tend to steal borrow from each other, not just with sounds and new words, but also with word formation. They have to be in contact with each other for that to happen, though, like the bleeding between Elvish and Numenorean in Lord of the Rings.****

I'll be doing the whole Pamak morphology example thing next time around.

*We've got almost everything in English, somewhere.
** Let it be noted that there is a possessive -s in theirs, and that these paradigms were much more regular in Old English, a.k.a. 1000-1400 years ago.
*** Yes, that probably means research. Sorry.
**** Yes, it's there.

No comments: