Do editors and proofreaders check facts?

Is fact-checking part of a copy-editor or proofreader’s role? If so, what sort of checking should they do, and how should they handle factual errors?

​Is there a rule about this?

People have all kinds of different understandings about whether editors do, and should, check facts. I’ve seen some apparently saying we should routinely check all facts, and others suggesting we shouldn’t check any. And there are plenty of other views in between.

As with many things in the editing world, there’s no universal rule. It all depends on the client or employer’s needs, the nature of the project, and how the editor or proofreader fits into the workflow.

​What we check

As well as checking the accuracy of statements made in the writing, ‘fact-checking’ includes verifying that names – of people, places, organisations, products and so on – are spelled and presented as they should be.

In conventional publishing, neither copy-editing nor proofreading is generally expected to include comprehensive fact-checking. The idea is that the author is mainly responsible for getting things right.

But we usually should look out for factual discrepancies within the material. This is a valuable part of what we do. These differences could be within the main text, or between that text and other items such as figures, tables and lists. For example, different chapters might say an event occurred on different dates, or the commentary on a graph may not tie in with the figures that the graph shows.

It’s also common practice to correct anything that we know is wrong, and to check anything that we suspect may be wrong. For an experienced editor or proofreader, there are certain things that will raise a red flag and prompt them to do some research, or at least ask a question.

​How we deal with factual issues

Here’s what I consider to be general best practice, although a different approach might sometimes be more sensible according to the circumstances.

When a copy-editor tries to clear something up, they may not be able to find an answer from a reliable source quickly. In that case, the best thing to do is to leave the text unchanged and just raise a query for the author.

If they do feel fairly confident that a change would be correct, they should make it and provide an explanatory note for the author. If possible, the note should include a link or reference to at least one source that backs up the decision. This way, the author gets a chance to check that they’re happy with the change, and either accept or reject it.

If they know something is definitely wrong: again, the best practice is to change it and provide a note for the author, citing one or more sources to confirm that the change is correct.

All of the above may also apply to proofreaders, depending on the project. If it has both a copy-editor and a proofreader (as in traditional publishing), it may be agreed that only one or the other will do any fact-checking. The copy-editor may well be expected to go further in checking things that they find suspicious: doing some research and suggesting how the solution should be worded. The proofreader may just need to flag the issue, clarifying what it is (rather than simply writing ‘Please check.’).

​Changing times

There is an old-school belief that copy-editors and proofreaders should keep fact-checking to a minimum, if they do any at all. This seems to be a legacy of the pre-web days, when it involved thumbing through reference books and maybe making a trip to the local library, and so it would have seriously slowed down their work.

Now, of course, vast amounts of information are available to us within seconds. That comes with its own dangers, though, as so many online sources aren’t necessarily reliable.

The rise of AI – with the risk of ‘hallucinations’ – has increased the value of fact-checking in some contexts: we may sometimes be checking text that’s been written with some help from AI tools. But it also means we need to be even more wary about the reliability of sources we might use for verifying facts, as they themselves may include information that can be traced back to AI.
​
Even with the web at our fingertips, fact-checking can be quite time-consuming when done with an appropriate level of care. It’s a very different skill from the core elements of editing and proofreading, which are about language, structure and presentation. So, I’d argue that we still should be cautious about agreeing to check all of the facts in a manuscript or document, and that clients shouldn’t just assume we’ll do so.

​Making things clear

With all this uncertainty about what (if anything) a client might expect their editor or proofreader to do in relation to fact-checking, it’s important to have an agreement on this right from the outset, before the schedule and terms are finalised.

I routinely mention my proposed stance on fact-checking when giving a quote for potential work, to help make sure the client won’t have unrealistic expectations. I don’t remember anyone asking me to go any further than this. If anyone does, though, I’ll gladly consider it.

So, if you’re in discussion with an editor or proofreader about using their services, it’s well worth making sure you have a common understanding of how they’ll approach fact-checking: in what circumstances should they do it, and how should they deal with potential changes? This can help to ensure they’ll protect the integrity of your writing to a suitable degree, while staying within the agreed schedule and budget.

​Further reading

My colleague Hazel Bird has some interesting thoughts on this topic, backed up by some research into how often clients set out their fact-checking needs up front (spoiler alert: not very often): ‘Are copyeditors responsible for fact-checking?’