Does repeating a noun increase density?

It can change the ratio because each occurrence counts as a lexical token; it does not necessarily increase vocabulary diversity.

Are all verbs counted as lexical?

Not under every convention. Main verbs and auxiliaries may be treated differently, so the rule must be stated.

Is high density always harder to read?

No. Familiarity, sentence structure, organization, and reader knowledge also matter. Density alone does not determine reading difficulty.

AI Glossary / Natural language and speech / Beginner

Lexical density

Definition

Lexical density measures how much of a text consists of lexical, or content-bearing, material. A common version is the percentage of word tokens classified as lexical words. It describes one aspect of a text's linguistic composition, not its overall quality, factual accuracy, or the intelligence of its writer.

What it means

Content words typically include nouns, main verbs, adjectives, and many adverbs. Function words, such as articles and prepositions, mainly express grammatical relationships. Both kinds are important for meaning.
Under the word-based convention used here, lexical density equals lexical word tokens divided by all word tokens, multiplied by 100. Repeated content words count again because the measure uses occurrences, not unique vocabulary.
Other conventions measure lexical items per clause or draw category boundaries differently. A density figure is therefore incomplete without its counting rules and denominator. Scores calculated under different definitions should not be compared as if they were the same measurement.

How it works

Define what counts as a word token and which categories count as lexical. State how the analysis handles punctuation, contractions, multiword expressions, auxiliary verbs, and ambiguous categories.
Tokenize the text and identify the relevant word classes, either manually or with a suitable linguistic analysis tool. Check uncertain classifications because tagging errors affect the numerator.
Count lexical tokens and total word tokens. Calculate the ratio using the declared convention, retaining the raw counts so another person can verify it.
Interpret the result alongside text type, language, and audience. A short headline, a spoken conversation, and a detailed instruction can have different densities for legitimate communicative reasons.

Three practical examples

A language researcher compares lexical-token proportions across two collections analyzed with identical rules.
An editor uses density as one descriptive signal when examining whether a technical paragraph packs many content words into little space.
A text-analysis project checks how different tokenization or part-of-speech decisions change reported density before comparing writing samples.

Think of it this way

Think of counting how much of a packing list names actual items versus how much explains relationships between them. The proportion tells you something about composition, but not whether the list is complete or useful.
The limitation is that function words are not empty filler. Removing them can destroy grammatical relations, so maximizing the measured content-word proportion is not automatically a sensible writing goal.

Worked example

Consider the original sentence “The small robot quickly sorted the blue boxes.” Excluding punctuation, its eight word tokens are The, small, robot, quickly, sorted, the, blue, boxes. Counting the tokens explicitly makes the denominator easy to verify.
Using adjectives, nouns, the lexical adverb, and the main verb as content words gives small, robot, quickly, sorted, blue, boxes: six lexical tokens. Density is 6/8 × 100 = 75%.
Now write “The robot sorted the boxes.” There are five tokens and three lexical tokens: robot, sorted, boxes. Density is 3/5 × 100 = 60%.
The first sentence is denser under this convention and adds descriptive information. The calculation alone does not establish that it is better writing: whether color, speed, and size are relevant depends on the purpose.

Illustrative case study

In a fictional project, Maple Learning compares short instructional passages. One automated report assigns a higher density to a compact passage filled with specialist nouns, and an editor initially treats that as a quality improvement.
The team checks reader comprehension and discovers that the specialist vocabulary needs explanation. It retains density as a descriptive measure while evaluating whether readers can follow the instructions.
It also records the lexical categories used by its analysis tool. This invented case illustrates a measurement decision rather than claiming a research result or a universal relationship between density and comprehension.

Common misconceptions and mistakes

Do not confuse lexical density with lexical diversity. Density concerns the proportion of lexical material; diversity concerns variation in vocabulary.
Do not count unique content words when the chosen formula requires lexical tokens. Repeated occurrences belong in the numerator.
Do not mix lexical-items-per-clause values with lexical-token percentages. Their denominators and numerical scales differ.
Do not treat function words as meaningless or automatically undesirable.
Finally, do not use a single density threshold to rank all genres, languages, or writers. The counting convention and communicative setting affect what a comparison can support.

Business use

For a business, lexical density can contribute to a broader review of reports, instructions, or customer-facing material. It may help identify passages that deserve closer inspection, but it cannot explain every readability problem.
Keep the underlying token counts and analysis settings available. Compare similar kinds of text using the same language and counting rules, and inspect examples behind unusually high or low values.
Combine the measurement with practical questions: can readers identify the required action, understand unfamiliar terms, and retain the essential message? Those outcomes are more informative than improving a percentage in isolation.

Check your understanding

A text contains 20 word tokens, including 12 lexical tokens, but only eight distinct lexical words. What is its lexical-token density?

Answer: 12/20 × 100 = 60%. The eight distinct words belong to a different kind of vocabulary count and must not replace the token numerator in this formula.

Related terms

Token
Part-of-speech tagging
Function word
Lexical diversity
Type
Readability

Learn next

Annotate a short paragraph manually, stating your treatment of auxiliaries and adverbs. Calculate lexical density and retain the counts.
Then compare your annotation with a tool's output and explain any disagreements. As a second exercise, replace a repeated noun with a synonym and consider why vocabulary diversity may change while lexical density stays the same.

Sources

Victoria Johansson, Lexical diversity and lexical density in speech and writing: a developmental perspective, original study with explicit discussion of counting conventions: https://journals.lub.lu.se/LWPL/article/download/2273/1848/4927 The robot sentences and numerical exercises are original teaching examples; their counts are not results from Johansson's study. Independent source review: Johansson’s definition and category-convention discussion were read, including token denominators, lexical adverbs, language dependence and the density/diversity distinction.

Previous: Lexical choice Next: Lexical resource