How Many Words Do We Know? Practical Estimates of Vocabulary Size Dependent on Word Definition, the Degree of Language Input and the Participant’s Age

Researchers dealing with quantitative aspects of language are often asked how many words a typical native speaker knows. The question is raised not only by lay people but also by colleagues from various disciplines related to language processing, development, acquisition, and education. The answer usually starts with a deep sigh, followed by the explanation that the number depends on how a word is defined. As a result, in the literature one finds estimates going from less than 10 thousand to over 200 thousand (see below). In this paper, we try to give practical answers for American English depending on the definition of a word, the language input an individual is exposed to, and the age of the individual.

Terminology

In this text, we will need a number of terms related to words. For the readers’ ease we summarize them here.

Word Types vs. Word Tokens

Word types refer to different word forms observed in a corpus; tokens refer to the total number of words in a corpus. If the corpus consists of the sentence “The cat on the roof meowed helplessly: meow meeooow mee-ee-ooow,” then it has nine word types (the, cat, on, roof, meowed, helplessly, meow, meeooow, and mee-ee-ooow) and 10 word tokens (given that the word type “the” was observed twice). Somewhat surprisingly, in some word counts the words “The” and “the” are considered as two different word types because of the capital letter in the first token of “The.” If such practice is followed, the number of word types reported nearly doubles.

Alphabetical Word Type

This is a word type consisting only of letters. In the example above mee-ee-ooow would be deleted. The cleaning to get to alphabetical word types in addition involves eliminating the distinction between uppercase letters and lowercase letters. So, the words “GREAT” and “great” are the same alphabetical type.

Lemma

Uninflected word from which all inflected words are derived. In most analyses is limited to alphabetical word types that are seen by the English community as existing words (e.g., they are mentioned in a dictionary or a group of people on the web use them with a consistent meaning). In general, lemmas exclude proper nouns (names of people, places, …). Lemmatization also involves correcting spelling errors and standardizing spelling variants. In the small corpus example we are using, there are six lemmas (the, cat, on, roof, meow, and helplessly).

Word Family

A group of lemmas that are morphologically related form a word family. The various members are nearly always derivations of a base lemma or compounds made with base lemmas. In our small example corpus the lemmas “the, cat, on, roof, and meow” are all base lemmas of different families, but the lemma “helplessly” can be simplified to “help.”

Below we will see what the various definitions of “words known” mean for vocabulary size estimates. First, we discuss how many words there are to learn (in English).