So far we've mostly concentrated on using affixes and other processes to create slight variations on the words that we already have. For example, "dogs" is just a variation of "dog" meaning that there is more than one of them, but they still both refer to a particular kind of four-legged animal. This is called inflection.
In contrast, words like "dogpile", "doggedly", and "hot dog" are obviously related to the word "dog", but they don't refer to the same thing. It's time to dive into derivation.
There are two major sub-fields of derivation; compounding and affixing. Compounding means putting whole words together to form a new word, as in the following examples:
- "roommate" (noun + noun)
- "bluebird" (adjective + noun)
- "babysit" (noun + verb)
Notice that in all these compound words, the second word determines the general kind of thing the compound refers to, and the first word makes it more specific. A roommate is a kind of mate, not a kind of room; a bluebird is a kind of bird, not a shade of the color blue; to babysit is to take care of (sit with) a baby (or other child), not to baby a sit (if that means anything). Because of this we say that compound words in English follow a modifier-head order: the modifier morpheme comes first, and then the head morpheme. Some other languages have head-modifier order. Some languages also allow more than two words to be combined.
When compounding is used to generate new words, care must be taken not to violate the phonotactics of the language. In the case where an illegal phoneme combination occurs, you will have to decide how to modify the word to follow the phonotactics.
Affixing is when you use affixes to change the meaning of a root in a more or less predictable way. Examples from English include:
- -er, -or
- doer, worker, contractor, enabler
- unable, unworthy, undo
- prefix, prepay, preauthorize
One of the major uses for derivational affixes is to change the type of word; turn a noun into a verb for example. In the example above you can see that the English affix -er takes a verb (such as "to work") and turns it into a noun ("worker"). There are all sorts of affixes for all the major transformations.
It's possible to apply more than one such transformation at a time; a noun like "concept" can become an adjective with the addition of "-al", "conceptual" can become a verb with "-ize", and "conceptualize" can come full circle to a noun with "-ation", conceptualization.
In some languages these means of deriving new words are more productive than in others. Languages like German and Esperanto give the individual speaker great freedom to coin new compounds and affixed words as needed, while languages like French are more restrictive, usually tending to use phrases for things German might coin a new compound for. English is somewhere in the middle of the spectrum. This may be partly a matter of culture as well as linguistic structure.