At the beginning of this textbook, we defined morphology as the structure of words. Words are also important units manipulated by the syntax. But what is a word? It turns out that the definition of word is trickier than it may at first seem, especially when we consider the different kinds of words we see in different languages.

One reason wordhood is difficult to define is that, even within the same language, we use word to mean lots of different things.

Consider the following situations. What is meant by word in each of them?

To sort out the differences between these different uses of word, linguists use the technical terms lexeme, word token, and word form.

A lexeme is a word in the abstract sense, including all of the different inflectional forms that a word can take. In example (1c), what the teacher means is that the passive voice is constructed using the lexeme BE, but it doesn’t have to be be itself, it can be one of its other forms. In (1d), they are estimating that English includes approximately 1 million lexemes. Lexemes are usually rendered in all caps.

A word form is a a particular form of a lexeme. In (1a), you are arguing about whether the lexeme FURNITURE can appear with plural inflection, or, in other words, whether FURNITURE has a plural word form. In (1c), your teacher has taught you that the passive voice must have one of the word forms of the lexeme BE, including was, is, and been.

Technically, a lexeme is a set of all the inflected word forms associated with the root word. However, it is cumbersome to list all of the words in the set each time we refer to it. The word form we use to refer to a lexeme is typically called the citation form. It is the citation form that will typically be listed in the dictionary. Languages differ as to which word form is used as the citation form, depending on convention and what makes sense given the morphological properties of the language. For example, in English, we usually use the non-finite form of verbs and the singular form of nouns as the citation form.

The collection of all the word forms in a lexeme is called the paradigm for that lexeme. In other words, a paradigm is the set of all the inflected forms of a word. On the other hand, words that are related by derivation each belong to a different lexeme. A set of words related by derivation is called a word family.

A word token is a particular occurrence of a word. In example (1b), the teacher wants you to write a paper that has 500 word tokens. Your paper will likely use the word the several times, but you will count each instance as a separate word (unlike if you were counting lexemes).

Let’s look at an example together. Below, you will see the entry for buy from the Canadian Oxford Dictionary (Barber 2005).

The entry for buy includes at least two homonymous lexemes, the verb BUY and the noun BUY. These two words are related and belong to the same word family, but must be considered different lexemes because their relationship is derivational rather than inflectional, as the noun is formed by conversion from the verb. We might also wonder if the five different definitions of the verb buy should also correspond to different homophonous lexemes, or whether they should all be considered different uses of the same lexeme. I will save the finer points of this debate for your semantics class, but we can tell that this lexicographer, at least, decided they were one lexeme (or else they would have listed buy₁ verb and buy₂ verb, etc.), likely because the senses listed here all seem to be related to the same core sense by means of extension and metaphor. The citation form for both lexemes is buy.

The entry for the verb BUY has three word forms listed in addition to its citation form at the beginning of the entry: buys, buying, and bought. The entry for the noun BUY does not have any additional word forms listed, likely because the only other word form, the plural buys, is completely regular.

Overall, this dictionary entry includes 10 word tokens of different forms of buy. There is the headword, which begins the dictionary entry. The second, third, and fourth word tokens are the examples of the different word forms, buys, buying, and bought. The remaining six tokens are found in the examples, such as in money can’t buy happiness.

Check yourself!

References and further resources

For a general audience

🧠 Merriam-Webster. How many words are there in English? https://www.merriam-webster.com/help/faq-how-many-english-words.

Sources for examples

Barber, Katherine. 2005. The Canadian Oxford Dictionary. 2nd ed. Oxford: Oxford University Press.