Duplicate content is text that's identical or substantially the same as content that already exists, either elsewhere on the same site or on a different one entirely. It shows up more often by accident than by design: printer-friendly page versions, URL parameters creating multiple paths to the same page, scraped or syndicated articles, and boilerplate product descriptions copied straight from a manufacturer are all common, unintentional sources.
Google's own position is more nuanced than "duplicate content gets penalized." There's no blanket penalty for the most common, accidental cases; instead, search engines try to identify the canonical version and show that one, which mostly just means the other versions won't rank rather than that the site as a whole gets punished. The exception is content that's duplicated deliberately and manipulatively, entire articles scraped from other sites and republished wholesale, which can trigger real ranking consequences. Canonical tags, redirects, and simply writing original copy instead of reusing manufacturer or competitor text are the standard fixes.
