Morphology

← Course contents · Using LLMs in this course

What's morphology doing in a module on vocabulary?

In English, it's often the case that, even if you've never seen a word before, you can work its meaning out if you know something about morphology. For example, you've probably never seen the word decountrify, but you can probably guess what it might mean because you know de-, country, and -fy. It's the same for your students. If they know industry, along with some common suffixes, then they get industrialization "for free".

Fortunately, there aren't even that many common suffixes, and if your students already speak another European language, they may already be familiar with many of them.

There are, of course, others, but most of them apply only for a few words and aren't really worth studying.

The same is true of most non-word bases. For example, sta- means "to stand, set down, make or be firm," and it appears in quite a few words. But these are mostly obscure (e.g., assist, cost, stem), unhelpful, or both. For instance, it's hard to imagine how knowing the meaning of sta- would allow most students work out what any of the following words mean: constant, contrast, destination, distant, estate, instant, prostitute, restore, stage, system, and understand. For this reason, it's generally fine to ignore bases like thiss.

Inflectional & derivational morphology

The following video introduces the morphological concepts of inflectional and derivational suffixes. This is the standard story, so if the distinction between inflection and derivation is familiar to you, you may wish to skip it.

Word families, lemmas, types, and tokens

A word family includes a base word and all its inflectionally and derivationally related words. All the words related to wide in the previous video belong to a single word family, the wide family.

All the inflectionally related words belong to a single lemma. In the video, that's one lemma per column. For example, the lemma widen includes widen, widens, widened, and widening. There are five lemmas in the video: an adjective, an adverb, a verb, and two nouns. Generally, a lemma is what you look up in the dictionary. If you want to know about widest, you look under the lemma wide. If you look at widest, there may be nothing there, or you may be redirected to wide.

Each inflectional or derivational word form is a different type. In the wide column, there are three types: wide, wider, and widest. The wide family includes the 11 types shown in the video.

A token is the occurrence of a word in a text or utterance. If an article is 800 words long, it has 800 tokens. This is the normal way of counting words for school assignments and publishing.

The reason we need these different ideas of a word is that they're all useful in different ways.

If you want to know how many words a learner has learned, you're probably talking about word families. If we learned run, for example, and you know something about English then, runs, running and runners all come "for free"; we haven't really learned four words, we've learned one word family. Similarly, if the teacher used members of the run family in its various forms 12 times during the class and the students also read it 15 times, we don't want to say we've learned 27 words. So word families are helpful for thinking about our vocabulary size.
But if you want to know how many words somebody can read in a minute, that's tokens; if somebody says, "I can read 160 words per minute," it doesn't matter how many of those words are repeated or even related. And, again, when a teacher says, "write 500 words," they mean word tokens.
As we saw above, what you look up in a dictionary is usually a lemma. If a dictionary says it has 180,000 words, it means 180,000 lemmas. Inside each lemma (e.g., run), there are more types (e.g., running, ran, etc.), but the dictionary doesn't include those in the number of words covered. On the other hand, it counts and run and and runner as two words, even though they're in the same word family.
Finally, if want to know how many different forms are included in a lemma or word family, you're asking about types. For example, the lemma go has 5 types, but the lemma be has lots more types than any other verb (15 in fact!).

(BTW, If you're searching in COCA, you usually just enter a word (e.g., run), and this will give you just the type run (every word spelled R-U-N). But you can also enter [run] with square brackets, which returns the various types of the run lemma: run, running, ran, etc.)

This video very helpfully explains what we mean by "word" and goes into more depth about morphology.

← Course contents