Those of us in the business regularly consult the Regret the Error Web site, which aggregates published corrections, to see what blunders our peers are fessing up to. Craig Silverman, the proprietor of the site, does an annual year-in-corrections roundup.

And, since 2004, he has also provided an annual roundup of reports of plagiarism and fabrication. These are, mind you, reported instances. As teachers and professors will likely concede, what gets caught appears to be a fraction of what is committed.

The range is impressive. Incidents occur at student papers, metropolitan dailies and national magazines. Columnists are well represented — perhaps they imagine that the rules don’t apply to them. People lift material from Wikipedia, from other periodicals, from Web sites, shoving it all under their own bylines.

No one is immune. In recent years, scandals of plagiarism and fabrication have blighted The New York Times, USA Today and The New Republic. Accusations of what was either plagiarism or extremely sloppy research practices have cast shadows on the work of historians Stephen Ambrose and Doris Kearns Goodwin. Careers at The Sun have been destroyed by evidence of plagiarism and fabrication.

It falls to editors — assigning editors and copy editors — to protect the integrity of the publication. Indeed, in recent years the instances of premeditated or accidental plagiarism that have been identified in-house at The Sun have been caught by the copy desk.

This, by the way, is one good reason to make sure that the copy desk has the staff and the time to edit, rather than merely process, the copy.

For those of you who teach or edit or have some supervisory responsibility over written material, I offer some commonplace tips on what to watch for.

Changes in diction: If the vocabulary of an otherwise amateurish student writer or cliche-ridden hack journalist should abruptly grow sophisticated, lifting is likelier than an infusion from the muse.

Changes in syntax: Same thing. If a writer who struggles to cobble together a noun and a verb suddenly masters the compound-complex sentence, with attendant Ciceronian participial ornaments, it’s time to start looking elsewhere.

Specialized information: Ask Howard Baker’s question from the Watergate hearings of beloved memory: What did he know, and when did he know it? Sudden access to biographical details, historical information, ecclesiastical terminology or scientific or medical expertise has to have come from somewhere. Demand an explanation of the source.

Dubious sources: Any article based on a single source is automatically suspect — how can you tell that the source wasn’t lying? Where’s the confirmation? Similarly, anything based on second- or third-hand sources demands scrutiny. In addition, readers are justifiably suspicious of anonymous sources. Even when anonymity has been granted for good reason, such as the source’s reasonable fear of physical or economic injury, the writer should be obliged to reveal the source to the assigning editor, acquire supporting information, and give the reader as much information as is prudent about the anonymous source’s credibility.

Improbabilities: When Jack Kelley filed his famous story with USA Today about seeing, in the aftermath of a bombing, human heads rolling down the street, their eyelids still blinking, it would have been a good thing for the paper if an editor had said, “What the hell?” and followed up. In journalism, as in investment offers, if it looks too good to be true. ...

Your job is to be skeptical, not gullible. Any writer’s work ought to stand up to questioning, particularly about sourcing. So ask the questions.

As it happens, the very ease of theft that the Internet provides also offers ease of detection. Use Lexis-Nexis or Google to find information on the subject that the suspect article covers. Do searches on distinctive and anomalous phrases. (Some colleges and universities employ specialized software and run term papers through it.) Check it out.

Follow up. The first question that must always be asked when a plagiarism is detected is this: Has he/she done this before? This has to be checked out, but it won’t be unless you, who have detected the misdeed, report it to someone in authority.

Don’t agonize over fear of appearing to be an informer. If the instance you identify is a first-time mistake made out of ignorance, you may save a colleague’s career. If it turns out to be one in a pattern of lies, then the career wasn’t worth saving.