Concordance Tool (KWIC)

How to Use the Concordance Tool

1. Paste your text. A novel chapter, a set of interview transcripts, a collection of abstracts, or your own draft.
2. Type a word or phrase. Add an asterisk for any run of letters: analy* finds analyse, analysis, and analytical, and *ness finds every word ending in -ness.
3. Set the context and the order. Choose how many words to show on each side, from 1 to 15, and sort the lines by the word to the left, the word to the right, or the form that matched.
4. Read the lines and the collocates. The collocates table counts the words inside the context window, left and right separately. Download the CSV to keep working in a spreadsheet.

Reading a Concordance

A concordance turns a text on its side. Instead of reading sentence by sentence, you read every occurrence of one word at once, aligned so that the patterns around it stand out. Sorting is what makes those patterns visible: sort by the word to the right and every memory consolidation gathers in one place; sort by the word to the left and working memory separates from long-term memory.

The hit count says how often the word occurs; the rate per 1,000 words lets you compare texts of different lengths, since 8 hits in a 75-word passage is far denser than 8 in a 7,500-word chapter. Collocates are counted inside the same window as the context, so a wide window picks up looser associations and a narrow one finds fixed phrases.

Worked Example

Choose Try a sample: a 75-word passage on memory with the search set to memory. There are 8 hits — 106.7 per 1,000 words. With 3 words of context and the lines sorted by the word to the right, the concordance reads:

The two suffers lines sit side by side, which exposes the passage's contrast: working memory suffers first, long-term memory later. In the collocates table, working and long-term each appear twice and only on the left, and suffers twice and only on the right — the three phrases the passage is built on.

Uses in Research and Writing

• Corpus linguistics. Check how a word is actually used — the verbs it takes, the prepositions after it, the adjectives in front of it — before you make a claim about its meaning.
• Literary analysis. Trace a motif such as light, water, or home through a novel or a poet's work, and quote each occurrence with its context.
• Qualitative research. Find every mention of a theme across interview transcripts, with enough surrounding text to code it.
• Revision. See how often you lean on your own favorite words — significant, clearly, however — and whether the sentences around them earn them.

Search Rules and Limits

• Searches ignore case unless you choose Match case, and Whole words only stops art from matching inside start.
• A phrase matches across line breaks and extra spaces, but not across punctuation: working memory does not match working, memory.
• The context window counts words split at spaces and punctuation, with apostrophes and hyphens kept inside a word, so long-term is one word.
• There is no lemmatization: study does not find studied. A wildcard such as stud* finds both, along with student.
• Collocates are raw counts within the window, not association scores such as mutual information or log-likelihood.
• The page shows the first 500 lines; the CSV download includes every match.

Frequently Asked Questions

What is a concordance?

A concordance lists every occurrence of a word in a text together with the words around it. Hand-made concordances of the Bible and of Shakespeare took years to compile; a concordancer builds one instantly, so you can study how a word behaves rather than how often it appears.

What does KWIC stand for?

Keyword in context. The format — each occurrence on its own line, centered on the keyword with a fixed amount of context either side — was popularized by H. P. Luhn's keyword-in-context indexes at IBM around 1960 and is still the standard concordance view.

What are collocates?

Collocates are words that keep company with the keyword — strong tea, heavy rain. This tool counts how often each word appears within the context window, left and right separately. It does not compute association statistics, so frequent words can rank high simply because they are frequent; leave function words out to see the content words.

How is this different from AntConc or Sketch Engine?

Dedicated corpus tools handle large corpora made of many files, lemmatize words, tag parts of speech, and compute association statistics. This concordancer is for quick work on a text you paste: nothing to install, nothing uploaded, and results as you type.

Can I search for more than one word at a time?

One word or phrase per search. To gather a family of word forms, use a wildcard — organi* finds organize, organise, organization, and organism — then sort by the matched form to see each one separately.

How do I open the CSV in Excel or Google Sheets?

Download it and open the file directly, or import it as UTF-8 text. Every line becomes a row with its number, left context, keyword, and right context; quotation marks and commas inside the text are escaped so the columns stay aligned.

Is my text uploaded or stored?

No. The search runs in your browser, and the text is never sent to a server.