When we put together our mental dictionaries, we don't stock up our neural space with words; instead, we use morphemes, the smallest pairings between sound and meaning. But just pairing up sound and meaning doesn't mean that the same sounds always emerge from our mouths for the meanings we intend. Instead, some morphemes get pronounced in different ways, depending on the environments in which they're pronounced: what other morphemes they attach to, and what sounds they're near. This is a process known as allomorphy.
Allomorphy is very similar to allophony. But while in allophony, the rules that change sounds apply whenever the sound finds itself in the right environment, for allomorphy, the rules are only enforced for that particular morpheme. We look at some examples of allomorphy, like with the English plural. Sometimes allomorphy is reflected in English spelling, like with the in- you find in indirect or improper, but much of the time, it's not, so you have to pay attention to your ears.
Allomorphy can change just one part of one sound, like causing a previous voiced consonant like the plural -s you get in dogs [dɑgz] to become voiceless, like in rocks [ɹɑks], or change all of the sounds out completely, as with cases of suppletion. But all of these changes are driven by rules that are learned with each morpheme. Knowing which version of the morpheme is the one to apply the rules to is a matter of finding the underlying morpheme. And for that, we look at which morphemes show up in the most different environments, as illustrated with our examples with the English in- and the German plural.
Even though morphemes may change in their pronunciations, they still have that core pairing of a sound to a meaning. We just need to know the different varieties of sounds a given meaning hooks into for a language.
In the episode above, we saw some pretty extreme examples of sound change within roots — the chunkier parts of words that smaller things like prefixes and suffixes attach to. Suppletion was just one example, where every single sound inside a root was either replaced or erased, like in the English examples below.
(1) “go” becomes “went”
(2) “be” becomes “is”, “are”, “was”, and “were”
But we can find intermediate examples, too, where the alternation happens somewhere inside the root, without leaving behind something completely unrecognizable. And a good place to look for these kinds of examples is in irregular morphology.
Regular morphology refers to processes by which words are changed to reflect (for example) the presence of information about number and tense, in a way that's fairly predictable and rule-governed. That is, in a way that could apply to an entirely new word nobody’s even heard of before. So, if I invent the word “gud”, tell you that it’s a verb, and then ask you to plug it into each of the sentences below:
(3) Ken ___ regularly.
(4) Mitsuru ___ yesterday.
Chances are, as an English speaker, that you’ll say something like “guds” in (3) and “gudded” in (4), having applied regular morphology to place the verb in the present or past tense.
Irregular morphology is, as its name suggests, less reliable. So these are things like “ate” instead of “eated” as the past tense form of “eat”, or “mice” instead of “mouses” as the plural form of “mouse”. English speakers know this, but only through memorization; every root with one or more irregular forms is another exception to the rule (in fact, because this is morpheme-specific, the choice between regular and irregular forms can even help us tell the difference between two meanings: "I have many mice" probably refers to my possessing lots of small, furry rodents, whereas “I have lots of mouses” is going to pick out the computing device).
In general, the manipulation of vowels in a root to indicate grammatical information is known as ablaut. Sometimes, ablaut occurs on its own, as in (5) and (6) below, as just part of how the morpheme functions. There's no specific reason for goose to become geese; otherwise, we wouldn't be able to get the verb version of goose to show up with the third person singular -s, like in Akihiko gooses his diet with lots of extra protein.
(5) “sing” becomes “sang” or “sung”
(6) “goose” becomes “geese”
Other times, both regular and irregular morphology occur together, as in (7), where the regular past tense morpheme -t triggers a change in the vowel we find inside the root. English doesn't like having that many segments together at the end of a word: [sli:pt], with a long [i] and two consonants, overloads that part of the word. This is a pattern that applies to other words with this profile, too, like creep or leap. Also, notice that the past-tense suffix undergoes some predictable allomorphy: it's a voiceless [t] rather than a voiced [d] because of the voiceless [p] at the end.
(7) “sleep” becomes “slept”
But changing the quality of the vowels in a word isn’t the only way to express this kind of information in a root. We can also shift the stress around to signal a change in both meaning and syntactic category. In (8), with the accent marking which vowel carries the main stress in the word, we get either a verb, meaning something like to take account, or a noun which represents the result of the verb, the account itself.
(8) “recórd” becomes “récord”
Finally, we can leave the vowels alone and play around with consonants. Depending on whether we voice the “s” in a word like “use” or “house”, we get either a verb or its corresponding noun:
(9) [juz] becomes [jus]
(10) [hauz] becomes [haus]
Like suppletion, ablaut, and the other kinds of irregular morphology, the effects of shifting stress and voicing consonants aren't captured very well just with rules. For the word “políce”, both the verb and noun get pronounced with the same stress pattern, and voicing the “s” in “bus” turns it into an altogether buzzier word. So instead, these kinds of changes must be represented within our lexicon on an individual basis. Luckily, when it comes to language, our memory’s pretty good.
So how about it? What do you all think? Let us know below, and we’ll be happy to talk with you about morphemes, the ways they change, and how we can track them. There’s a lot of interesting stuff to say, and we want to hear what interests you!