Welcome to December, everybody. On one hand I hope it does not fly by as quickly as November, but on the other hand I am ready to be done with 2020. Bring it on, or don’t. It’s whatever.
Today’s article will be an exercise in concise writing for me. What I’m prattling on about is a lesson that takes most editors years to learn and I have to knock it out in 2500 words or fewer. I am ever trying to keep things brief. You have things to do, I have things to do. Actually, my thing to do is write Shelf Life. But surely you have things to do. We’ll go with that.
Can you define what perfection means for whatever you’re doing? In my experience it’s almost always an arbitrary measurement. You can do a mathematical problem to a certain level of perfection—your solution is right or it’s wrong. What about computer programming? The code works or it doesn’t, sure. When the program runs without crashing, does that mean the code is perfect?
What about a book? You can’t write a perfect book—no two people agree on what constitutes a perfect book. You could write a book with no grammatical errors, I guess. Would that mean it was perfect? Would it mean that in the cast of characters, no one ever speaks with grammatical errors? In that case, would the absence of grammatical errors mean the book was perfect? Or imperfect?
Outside the constraints of money and time, almost any level of quality is achievable—including perfection. But if you don’t have infinite money and infinite time, then you can’t get to perfect consistently. This is just the law of diminishing returns. As you take your standard of quality higher and higher in pursuit of perfection, the amount of money and time it takes to achieve that standard rises disproportionately. Everyone decides at some point—consciously or subconsciously—to stop. “This is good enough, I have to call it done.” Even Leonardo Da Vinci had to put down the paintbrush eventually and stopped working on the Mona Lisa. Right before he got to her eyebrows, evidently.
Practically speaking, when you tell someone that the only acceptable outcome of a process or product is perfection, then you have not given them any meaningful or actionable information about the standard of quality you expect. And if all errors are equal to you—all of them unacceptable—then you don’t know how to classify them by type and magnitude.
I can make a perfectly fine editor out of most anyone who wants to learn. I think I’ve trained between ten and twenty brand-new fresh grads in my day. You can invest someone with the knowledge and skill to look at text, recognize errors, and then make and execute an action plan to resolve them. Any good editor can do those things.
A great editor is one who exercises judgment on the fly to determine which errors must be fixed and which ones can be let go. What separates editorial wheat from chaff is not the ability to eliminate errors. It’s the ability to conduct a high-velocity cost-benefit analysis to determine when an error can be allowed through. If your mandate is: “No errors ever. Any error, at any time, is unacceptable,” then your editors will never learn that. You might build a stable of decent editors over time but you’ll never cultivate a single great one.
Let me tell you a story. I was working in textbook production and I got a call from the head of our editorial group, fuming. One of our biggest titles had come out, just hit shelves, and someone already found a typo. Prepare yourself for what lies ahead, because it is a game-breaking bug: Feedback . . . was spelled . . . feed back. (Yes, really.) She was correct, that’s an error. Someone missed it in proofing and the book got out the door that way. I don’t disagree that it would have been better had that not happened. She let me know that it was completely unacceptable to let a major, big-budget title print with an error in it.
The book in question was made from a 340,000-word manuscript. What does that have to do with it? For one thing, one three-hundred-forty-thousandth of your breakfast cereal is rat feces and the FDA says that’s a-okay. One out of every million measles vaccinations we administer causes a severe reaction, but we’re fine with that. The frozen blackberries you get in the grocery store have to have more than three insect larvae per 500 grams to be considered contaminated. So more than three larvae per one-pound bag. Yum.
The publishing company had a service-level agreement in place that had been approved by the highest levels of the organization. The acceptable error rate was set at .001%. For every million words we processed, there might be up to ten typos. For a company that processes hundreds of millions of words a year? Thousands of typos.
Is that too many typos? If you can’t live with a .001% rate of error in text, then you should probably just go ahead and cease existing because your food isn’t that safe, your medical care isn’t that safe—your car isn’t even that safe. Look into the US Motor Vehicle Safety Code and see how often the word “risk” is preceded by the word “reasonable” if you don’t enjoy sleeping soundly at night.
There’s an acceptable error rate for everything on the planet. You have to figure out the acceptable error rate for whatever it is you’re doing before you can assess your quality and efficiency. If you don’t do that, then:
You’re just guessing at how many errors is really okay, because you know that zero isn’t achievable but you said zero anyway. So is it one? Five? One hundred?
You’re not planning for errors, so you probably don’t have an action plan in place to deal with the errors that inevitably crop up.
You don’t know which errors mean you need to reevaluate your process to prevent future errors and which indicate a need for more human training and which might not be actionable at all.
If you want to do whatever it is you do, make whatever it is you make, with a high and consistent level of quality, then you need to understand your own expectations. Make a reasonable service-level agreement with yourself for whatever it is you’re doing. Otherwise you’re just stumbling around in the dark.
Types of Errors
I break errors down into two types overall:
Those that indicate a failure or weakness of a person.
Those that indicate a failure or weakness of a process.
It’s critical to be able to figure out which kind is which. And, by the way, if your employees have been told that any level of error is unacceptable, they will do whatever they can to blame errors on process, making it difficult or impossible for you to tell whether the employees need training or the process needs improvement.
A critical consequence of the perfection mentality is the belief that every error must be a sign that process needs to be revamped. The refusal to accept that sometimes ad hoc errors get made, one-off mistakes with no preventable cause, leads to constant refinements of the process in question until it becomes so laborious that it doesn’t make sense anymore. Take this cautionary tale: The Lament of the World-Class Waiter. I will call him Lonny, in honor of the most dedicated and professional server I ever encountered.
Lonny is a career waiter and he takes his job seriously. As part of his skillset, he’s cultivated an excellent memory for orders so he does not write them down. He works five days a week at a busy chain restaurant, serving approximately 30 tables per shift. If there are on average four customers to a table, that’s 120 customers per shift; 600 customers per week; 30,000 customers a year if he gets a two-week vacation. Lonny is excellent at his job and his memory has a .01% error rate. Of the 30,000 orders he takes in a year, he’ll forget 3 of them between the table and the kitchen.
Now, if you’re one of those three customers and Lonny forgot your order on the way to the kitchen—forgot to put it in at all or forgot what you said and put in the wrong dish—you might ask to speak with his manager. You might tell the manager, “I noticed that he doesn’t write the orders down. Maybe a solution to this is to write the orders down so it doesn’t happen again?”
That is the solution a lot of process-oriented organizations want to take. The process must have broken down somewhere, because good servers don’t just make mistakes. Now Lonny is told he must write orders down to prevent future errors. Lonny’s efficiency decreases. He can’t turn the tables in his section as quickly, so over time he will serve fewer customers. He’ll earn fewer tips, and his employer will experience a loss as well. Although 29,997 customers per year were highly satisfied with Lonny’s service, he made a mistake in a corporate environment that doesn’t accept the reality of mistakes.
Did the process revamp meaningfully improve anything? Are things better? For whom are they better?
A specific and real example of this is a company I worked for that has the following rule: No book may be released to the printer until the project manager has personally done a search of the final PDF for the word “pubic.” Why? Because once, twenty years ago, a public policy textbook shipped with an embarrassing typo in a prominent place. Has this new process—which has been applied to hundreds of books each year for twenty years—ever turned up the same typo again? To my knowledge, it has not. But every time a different easily confused word causes a typo, it gets added to this list. And after twenty years, the team of project managers collectively spends weeks worth of their person-hours each year searching for words from the list. For nothing.
When you encounter an error, ask yourself:
Was there a breakdown in our process somewhere? If yes:
Would adjusting the process in a way that would have prevented this error lead to preventing more of the same error in future? If yes:
What will it cost us over time to make this process change, and is that worth the savings we’ll get from eliminating the errors? If yes, if the savings outweigh the cost, then proceed with a process adjustment.
If the answer to number one, above, was “no,” then ask yourself:
Could the person with whom this error originated have done better if we gave them more help in some way?
In some way, I mean: Is this person overworked due to staff shortages? Do they need more training? Is their workday too short to complete all of their tasks, and if so, can you approve overtime or lighten their workload? If the answer to the question about giving them more help is “no,” then maybe the error was just a one-off mistake. Humans make mistakes. If whatever you do can’t be done by an army of robots, then you need to live with that.
Magnitudes of Error
In editing text, I break errors down into four types, presented here in descending order of importance.
Life-Threatening Errors
Sounds dramatic but I’ve seen a few life-threatening errors in my day. Let’s say the Greek lowercase letter μ (mu) doesn’t pull into your typesetter’s software correctly and the software substitutes the Roman lowercase letter m because that’s the next best thing. If that μ was part of a drug regimen you just changed micrograms to milligrams and somebody’s dead. Maybe a lot of somebodies. Depends on how fast you can act when you find out about the error.
Think it’s unique to medical publishing? I once edited a cookbook of baby food recipes. Essentially, the big book of one hundred different fruit-and-vegetable combinations to puree in your fancy blender. A recipe for six-to-nine-month-old infants called for honey as a sweetener. Clostridium botulinum in honey isn’t a big deal for adults and older kids but it’s extremely dangerous and sometimes deadly to babies under one year.
Errors that can endanger people always need to be fixed and can never slide, no matter how late in the game. Further, this type of error should always start a process evaluation, even if you believe it was truly a one-off mistake.
Errors of Fact
These, too, should always be fixed if they’re discovered in time. Don’t publish information that is factually wrong if you can possibly help it. Doing so will likely lead to a retraction later anyway, and you’ll end up pulling down an electronic publication or destroying and reprinting hard copies. When you’re attempting to repair an error of fact, be as transparent as you can. Don’t fix it and move forward as though the error never happened if there’s any chance that misinformation made it out into the world. Own your mistake and do what you can to ensure the correct information reaches anyone who may have been misinformed.
Errors of Language
Some things are not a matter of style but are always incorrect usage within the language. For instance, in English an -ly adverb is never hyphenated with whatever it’s modifying: a “hysterically laughing child,” for instance, never a “hysterically-laughing child.” Errors of language also include lack of clarity (when the lack does not rise to the level of a life-threatening error). In these cases, ask yourself: Will my average reader, looking at this, be confused and not know what it means? If the answer is yes and the error could truly cause misunderstanding of the content, then you should fix it before publication is final if you can. But if you can’t, this probably won’t rise to the level of destroying and reprinting.
Errors of Style
Finally, sometimes text is correct in the language but wrong according to the style manual you are using; for example, an omitted serial comma. These errors will almost never misinform or cause misunderstanding. They should only be fixed at the early stages of production when it is inexpensive and expedient to do so. I’ve been the editor who sends a final proof back for one more round of corrections at 4:58 PM over a comma. Don’t be that editor. You’re only adding value for yourself at that point—satisfying yourself that your work is perfect. You’re not adding any real value for your reader.
I hope today’s article has given you some food for thought regarding the pursuit of perfection. My spider sense is already tingling in anticipation of the messages I’ll be receiving from various among my perfectionist friends. When you’ve always valued something, it’s unpleasant to consider the possibility that it might not be as valuable as you believed. Give some hard thought to perfection, what it costs and what the pursuit of it pays you back. I don’t know a whole lot of people who confidently tell me it’s working out great for them.
In my opinion, very few things—very few—have absolutes. If you believe perfection is the only goal and any level of error is unacceptable, while you know full well that involving humans in any process means that you cannot achieve a completely nil error rate, then you probably don’t know what your real expectation of quality is. And if you don’t know that, you have no idea how to measure whether you are succeeding.
Make sure you don’t miss Thursday’s article, in which I’ll dip a toe in the librarianship pool to help you apply the extremely cool science of reader advisory to your writing. December is a terrible time to put any part of your body in the pool, but I will do it anyway—for you. Till Thursday, then. Stay warm.
If you have questions that you'd like to see answered in Shelf Life, ideas for topics that you'd like to explore, or feedback on the newsletter, please feel free to contact me. I would love to hear from you.
For more information about who I am, what I do, and, most important, what my dog looks like, please visit my website.
After you have read a few posts, if you find that you're enjoying Shelf Life, please recommend it to your word-oriented friends.