By Miranda Bell-Tilcock, Rachel Alsheikh, and Malte Willmes

Doing science is hard. Even in the best of times, it’s incredibly difficult, with many failures, mishaps, and disappointments along the road. More so than just smarts, perseverance, resilience, and teamwork are essential to seeing a project from initial field and lab studies to final conclusions.

If everything aligns, you may write a scientific paper to add to the vast library of science knowledge. Maybe it’s OK, then, that at the end of the road you let your guard down a little bit. Next thing you know, your final paper, in all its published glory…has a typo. Some typos are small, a missing comma somewhere (though remember, commas save lives. Friendly intent: Let’s eat, Grandma! vs. Cannibalistic: Let’s eat Grandma!) or a misspelled word that has little or no impact on the overall meaning of your text.

Recently, we published a paper detailing a diet-shift experiment that spanned multiple years with how isotopes can help identify off-channel habitat used by juvenile Chinook Salmon. This study used three stable isotopes (δ¹³C, δ¹⁵N, δ³⁴S) to characterize differences in food webs between off channel habitat and the adjacent river. We found differences in δ¹³C and δ³⁴S in the stomach contents between the two habitats with the off-channel having much lower values than the river. These differences were seen in the muscle tissue as well as the otolith for each fish. Using these two isotopes together, provides a toolset to quantify the role of off-channel habitats for fish. 

After years of work, we were excited to publish these data about food webs and isotopes but sometimes,  

sh#ft just happens: 

Can you spot our typo below?

That’s supposed to say “diet-shift”. With an f. That f is really crucial, isn’t it?

Getting a typo pointed out to you minutes after a paper goes live is embarrassing. While Miranda can find the funny factor in anything and immediately embraced the humor in the situation, this quote from Darwin’s famous letter to Charles Lyell better summarized Malte’s feelings:

“But I am very poorly today & very stupid & I hate everybody & everything. One lives only to make blunders.”― Charles Darwin, The Correspondence of Charles Darwin, Volume 9, 1861

Our typo had survived several months with multiple drafts that traveled through many people, including all of our coauthors, four reviewers, and the journal’s copyeditors. Although many edits were made to the paper and all numbers and figures were triple-checked, nobody caught this typo. Maybe because it was in a subheader, our brains skipped over it. We even read the paper out loud, and still the typo wasn’t caught until a friend pointed it out after publication.

We immediately reached out to the journal to inquire about getting the typo fixed. You would expect that correcting an innocent mistake, one that doesn’t change the results of a manuscript, would be a straightforward and easy process, but that’s not the case. Any significant changes to a paper require a formal correction process (for example, comments; or in serious cases, retraction and resubmission), but there is no standard method to fix a small, annoying (embarrassing) mistake. The journal only offered that a short, regretful comment could appear below the paper. This imperfect solution was not unique to this journal, in fact, it’s the same answer for most, and in many ways it is counter-intuitive. Since the typo didn’t constitute a significant error, we decided against making a formal correction. We were left with one option: leaving it as is. And then writing this blog post about it.

But why do we even make typos, and why do they seem common in scientific papers that are supposedly well-checked by many smart and dedicated people? Part of the answer comes from our familiarity with text we spend a lot of time on. Our brain loves to take shortcuts. By the time we’re looking at a final version, we know the text well, so well that our brain doesn’t actually read it anymore. It doesn’t need to. It already knows it.

A good example of our brain’s amazing ability to recognize text is that yuo cna raed tihs. Btu how cmoe we can raed wrods eevn wehn the letrtes are julbmed? Our brain generally doesn’t read words one letter at a time but rather processes them in their entirety all at once. Moreover, it skims sentences at high speed to infer their meaning while barely looking at the individual words that make them up. With our own writing, all of this happens even faster, because we learn to anticipate the words and sentences. To put it another way, the more familiar we are with a text and its context, the more efficiently we can read it — and the more easily we can miss a typo. Of course sloppy and fast writing can cause typos, but highly detailed and involved writing can lead to typos just as well.

When does this become a problem? Well, apart from creating some inappropriate headings, typos can cause all sorts of serious problems, like distracting from the message being communicated or undermining the authority of the work being presented. This gets exacerbated when we talk numbers. For instance, we all “know” that spinach is a good source of iron because of one misplaced decimal point (it’s got 2.5mg/100g, not 25mg/100g). And then there’s NASA Mariner 1, the first spacecraft sent to explore Venus, which lost steering capabilities shortly after launch due to a single missing subscript bar in its code.

Typos happen to even the most prominent scientists. For example, Peter Moyle told us that if you carefully read the key of the 1976 version of the benchmark reference Inland Fishes of California, LIVEBEARER FAMILY accidentally becomes LIVERBEARER FAMILY. And while no one ever pointed it out to him, that typo made him cautiously proof titles and subtitles ever since. 

So, in light of the fact that typos are extremely difficult to remedy in a published manuscript, the best strategy we have is to avoid them as much as possible during the writing and proofreading process. We’ve gathered some proven strategies to do just that. Note: most of the tips try to trick your brain into treating your writing as new and unfamiliar to stop your brain from assuming the text is correct and allowing a review from square one. Approaching your writing in a new way helps you notice things you never saw before, like that last typo.

• Use a trusted spell and grammar checker (ideally one not desensitized to curse words).
• Read your writing aloud to yourself, or have someone else or even a text-to-speech program read your writing to you. Don’t skip titles and headings!
• Change the arrangement. Switch up the order of the paragraphs, then read your paper that way.
• Change the appearance. Switch up the font and size of the text, the spacing between the lines, or the margins of the page. Brace yourself: some writers suggest proofreading in Comic Sans (psychologists have found that Comic Sans Italicized can increase reading comprehension).
• Proofread backwards. Start at the bottom of your paper and read one sentence at a time until you’re at the top. The meaning of the text gets completely lost, but the typos don’t have much room to hide.
• Get your colleagues, who haven’t seen countless drafts of your paper, or a friend, who is completely unfamiliar with your area of expertise, read it for clarity. 
• Google (or should we say Googol?) common misspellings. Have a checklist of words you personally misspell often or that spell checkers commonly confuse. NPR calls this an accuracy checklist and has a handy list of ideas to get you started.
• Triple-check numbers and all those critically important bits to minimize catastrophic typo potential.  Check decimal places.
• Let the text (and yourself) rest. Step away, sleep on it, and come back fresh.
• Print the paper out to read it through a different medium.
• Send the paper to your coauthors with the word “final” somewhere in your email, because you’ll inevitably spot an error the moment you hit send.
• Finally, don’t let the short deadline (many) journals give you to submit your final proof cause you to stress out and rush your last edits. Ask for more time and then take a deep breath. You’re nearly done, and it’ll be worth it.

We used many of these techniques, but nevertheless did not spot the typo staring us in the face. Shift happens. At the end of the day, we hope it doesn’t affect the science we presented. While in this case we take responsibility for the typo, other cases are completely outside of any writer’s control. Typos can even get introduced in typesetting during publication. Our final lesson is to be more forgiving of our typos and of others’, and to realize one misteak does not reflect the years of effort that go into a research paper. 

We would love to hear the best, or dare we say the worst, typo that slipped past you, and to learn what’s on your accuracy checklist. Comment below or share with us on Twitter using the hashtags #shifthappens and/or #typosci.

Further reading

Why typos are hard to catch

Why can we read words that are jumbled

MRC Cognition and Brain Sciences Unit