Skip to content

How many errors are acceptable in a text?

Errors appear everywhere, from grocery stores’ signage, to John Irving and Stephen King’s novels, to The LA Times. Because daily newspapers are on tight deadlines, seeing errors on their pages – while somewhat understandable – does annoy their more discerning readers.


On social media, typos are fairly common, as is “how’r u 2day?” and so on. While we may not correct ourselves on Twitter, we do represent a (personal) brand and/or an employer.


Errors are common, in part owing to the various ways and the speed in which we communicate. While the pressure to publish and the quantity of publications are increasing, the average publication quality may well be decreasing. Sometimes – say if a book that is guaranteed to sell comes out weeks before the movie based on it – there may be little incentive to proofread.


So, are errors acceptable in a paper, thesis or book? While there are many divergent views, my short answer is: as few as possible, preferably none. Errors influence others’ professional opinion of one – including examiners, colleagues, institutions, friends, parents, prospective employers and other readers.


If we submit something that is in some way erroneous, it may not be picked up in the review process. Academic publications are digitally archived and will remain visible for the world to see, for the rest of our life and thereafter.


Most e-book formats and print-on-demand services allow one to correct and update a file. With printed material, the risks associated with errors are too high – especially if you’ve paid for advance copies.


But there is no such a thing as a perfect article, thesis or book. Usually, the errors in these – even in professionally edited books from commercial publishers – include minor typos, the misplacement of an apostrophe or omission of a comma.

Technical books and manuals tend to undergo additional vetting, because a mistake can result in bad information. At the same time, technical publishers know that errata occur and often have a page listing issues and corrections.

Is there an acceptable percentage of errors?

According to some, a manuscript that that was hastily proofread by one proofreader may have one typo per 1,000 words – still too many. Some use the figure of three typos per 10,000 words. One error per page is considered an extremely high error rate.


Total errors over total pages is one formula; some prefer defects per million (DPM). Whether or not such thresholds exist is irrelevant, since we can never know whether or not we have met them. One reason for us to keep the number of known errors as low as possible is that we cannot count unknown errors. Thus, there is no way to know if one’s error rate is below a given threshold.


But nothing’s perfect, right? ProofreadNOW: Wrong. “It can be done, and it is done many times each day by careful writers. I have a copy of ‘Gone With The Wind’… that is over 500 pages long, and there is not a single error in the entire book.”


Do people care if there are mistakes in a published document? ProofreadNOW: “Not as many as in yesteryear… And that’s too bad! … Our bottom-line answer… is NO, there is no acceptable number of mistakes for any published document.”

Proofreading our own work is not enough

There are different standards in different publishing areas. The nature and purpose of your text may determine the number of reviewing layers. In some traditional publishing processes, at least three sets of eyes see a manuscript; still, they seldom catch everything.


Some self-published manuscripts undergo no editing or proofreading. One criticism of such work is that they do not meet professional standards of editorial quality. That is, there are many mistakes, including grammatical, punctuation and spelling errors as well as continuity and logic mistakes.


According to Scott Bury, as a writer and an editor,


the first rule… is this: you cannot effectively proofread your own writing. It’s so easy to make mistakes. Your fingers hit the wrong key, or Auto-correct gives you ‘ethylene’ when you wanted to type ‘Ethel.’ [] And no matter how many people read a manuscript before it’s published, somehow there are mistakes that slip through to the published edition, and then a reader will point it out… At some point, a large number of minor errors becomes frustrating. It shows that the author did not care enough about the reader’s experience to follow the process necessary to produce a good book: have it edited by a professional editor, proofread by a professional proofreader. Submit it to beta readers and reviewers, and make the effort to correct the errors. [] It costs money and it takes time, but as all our parents and grandparents told us, there are no short cuts when it comes to doing something well.


While unacceptable, errors seem to sneak past, sometimes in less obvious ways than typos. In academic publishing, errors are usually introduced into a manuscript mainly by an author or authors making changes after copyediting was done by a professional.


When a copyeditor must do more work on a text, they may miss mistakes, especially if there is a tight deadline; then, if the author and the publisher are not highly accountable for errors, more will appear on the page.


Attention to grammar, spelling, style, figures and tables may not be at the top of our to-do list when writing. Our research is complete, and we now need to write it up and create figures. After so much hard work, we may not feel like taking the time to get these right.


Reviewers or typesetters will correct this, surely? Careful figure legends, axis indications, units… the text makes clear what they are, right? Unfortunately not.


Although readers don’t judge a text solely by its errors, the single fastest way for an editor, reviewers and readers to lose confidence in your work, your results and you remains poor attention to detail.


Sloppy work disrespects an editor and reviewers, especially if their work is voluntary. They are not our spell-checkers. Also, it is widely assumed that an author who cares little about grammar, spelling and style is also imprecise in their research.


Causing a journal editor or a publisher’s production editor to work hard may also jeopardise one’s future assignments. Further, time is lost when a manuscript comes back with this single instruction: Have it edited by a professional.


Few people can afford to have a manuscript edited/proofread ad infinitum. One’s cost vs. error acceptability rate almost always comes down to how much one is prepared to spend on it and the quality of those you spend it on.