Identical or substantially similar content available at multiple URLs. Duplication alone does not establish a spam violation.
Duplicate content means the same or substantially similar main information is available at more than one URL. It can arise naturally through tracking parameters, alternate versions or publishing systems. The existence of duplicates alone does not establish a spam violation. The first task is to understand why the versions exist and whether they serve different purposes.
Search engines may group similar versions and select a representative canonical URL. Your preferred version is not necessarily the one selected. That makes consistent signals useful when several addresses genuinely represent the same content, but a canonical instruction should not be used to declare unrelated pages equivalent.
Duplication and low usefulness are different questions. Two addresses can contain an excellent duplicate guide, while a single unique page can still offer little value. Likewise, sharing a header, navigation or standard product specification does not mean every page is interchangeable. Compare the main content and intended task before deciding to merge or redirect anything.
Takeaway: Investigate the cause and purpose of duplicates before deleting pages. Preserve versions that serve a legitimate distinct need and manage genuine alternate URLs consistently.
Example
A craft store’s email software adds campaign parameters to a tutorial URL. The clean and tagged addresses show the same instructions. That differs from two tutorials teaching different techniques, even if both reuse the same materials list and site layout.
What to check
Check the content at each address, the declared canonical and Google’s selected canonical where available. If pages are consolidated, map them to a genuinely relevant destination and update links you control. Do not redirect distinct articles to a generic category merely because their titles share words.
For more detail, see Google canonicalization explanation.