Clear grammar builds trust with your audience
Because I am an English PhD, people often love to share with me their grammar horror stories. I can laugh at a lot of these because grammar mistakes can create funny pictures in people’s heads. I used to teach a class for Employers Council called Brush Up English (I still teach classes through Employers Council, but this specific class is no longer available through Employers Council; I do have grammar seminars, though) and the exercises on misplaced and dangling modifiers always make me laugh:
Piled up next to the sink, I washed the dirty dishes.
Do you get it? The grammar says that “I” am “piled up next to the sink.” 😆
This grammar is unclear, creates the wrong picture in the other person’s head, and may create some confusion.
***
SIDEBAR ON AI WRITING TOOLS AND GRAMMAR
Interestingly, Grammarly is unaware of the dangling modifier in that sentence as I write this. While it is offering me 16 other ways that I could “improve” my writing, it has no problem with “Piled up next to the sink, I”. Reason #1349 that AI writing tools are useless.
Ok, now I just took a 10-min side bar to converse with ChatGPT on the subject. It’s an idiot. While it could identify the problem with the sentence:
Yes, the sentence "Piled up next to the sink, I washed the dirty dishes." has a grammatical issue known as a dangling participle.
It couldn’t revise the sentence correctly. Here’s its first attempt at revision:
The dirty dishes piled up next to the sink while I washed them.
Which doesn’t make any sense. And another one:
Piled up next to the sink, the dirty dishes awaited washing.
Which is just weird. Dishes don’t “wait” because they are inanimate. Do they also await being eaten on? (Grammarly is now reprimanding me for ending that question with a preposition–which is not a real rule! And its suggested solution is absurd.)
Here’s another one from ChatGPT:
Next to the sink, I washed the dirty dishes.
Which creates a new misplaced modifier that–woohoo!–Grammarly *does* recognize! Except it marked it under “Clarity” and said that was a “premium suggestion” rather than just fixing the grammar.
Take away from this side bar: artificial intelligence is entirely artificial and has very little intelligence. Do not trust it.
***
HOW CLARITY IS INTEGRAL TO GRAMMAR
Back to clear grammar and its value. The utility of grammar is that the order of words in a language is the underlying structure that allows us to understand each other.
Words in the learned and expected order = meaning.
Words in an order that is unexpected and violates the “rules” = confusion.
For example, if I say to you “I need to go to the store.” You can understand that idea. You may not care about it, but you can follow what I am saying.
But if I put the same words in a different order and your ability to figure out my meaning is dramatically impaired: Store to I the to need go.
Yes, you know the words. Yes, you can guess at the meaning. But you are going to have to work really hard to figure it out.
Now, when my acquaintances send me mistakes that unintentionally create funny pictures in our minds, we laugh. Not to make fun of but because of the bizarre picture. Here’s a funny example from the grammarfail thread on reddit:
The original poster (OP) pointed out that in the last sentence on the ad “with” is very different from “without.” See how the meaning changes entirely because of one small error? But OP didn’t mention that “strudy” isn’t a word or that “protecting your sample integrity” is pretty hilarious out of context–you don’t have integrity, you are just testing it out! Errors in spelling, word choice, and word order! These are real problems!
CLEAR GRAMMAR IS NOT THE SAME AS STRICTLY-FOLLOWING-THE-RULES GRAMMAR
Why am I mentioning this? Because some of my acquaintances are dogged prescriptivists, which is a secret way of saying that they care about THE RULES more than they care about use or meaning and take pleasure in correcting people’s unintentional and perfectly clear errors for the sake of embarrassing them.
For example, many people in my family take great pleasure in correcting other people’s use of “I” and “me”. “And after we get dinner, you and me can go to the movies!” Sure, “me” is technically the wrong word in that sentence. But English allows “you” to be used in multiple ways: You and I are working together on this project. The project will be edited by you and me.
“You” stays the same, but the word for the self (I/me) changes depending on its position in the sentence. This shifting means that when we speak, we often put “me” where we should use “I” and vice versa.
The thing is, both “I” and “me” refer to the same person, so which word we use doesn’t affect the clarity or the meaning of the sentence. Correcting the grammar in this case–especially in speech–only serves to create a power difference between the person who knows the rule and the person who doesn’t.
Correcting the way people speak–if they aren’t asking for your input–is rarely appropriate. Speaking has nonverbal signals to support meaning. If you are confused, you can ask them right then and there. You can also tell more information about the person’s cultural background from the way that they speak which should allow you to determine whether correcting their grammar is supportive or classist (fyi: it’s usually classist).
Additionally, many people who correct other’s speech are not grammar experts and therefore get the corrections wrong. For example, if you said “The remote control is between him and me” one of these overcorrectors might automatically remind you that it is “him and I” or worse “he and I” and they would be 100% wrong and you would have been 100% right in the original.
The rules of language change over time. Sure, we have some historical precedent and debate around whether the rules reflect speech or writing and how quickly the rules should change and whatnot. Noah Webster of Webster’s dictionary advocated in the late 18th century/early 19th century that American English should spell “colour” as “color” because who needs that extra “u”? (The French. It was their word to begin with and they spelled it with a “u”. And British English kept that history. But Webster wanted a linguistic revolution to go along with the political one. Sigh.) But he also wanted American English to spell “island” as “iland”, which makes the exact same sense and appeared in the exact same version of his dictionary. And yet the editors and other Americans said, “No, Noah Webster, that is too far! “Island” must be spelled with a silent “s” because that’s the way it always has been done!”
So that’s why some of our words are spelled differently than British English, but not all of the words in American English have had their spelling Webster-ized. Because language is entirely embedded in culture. And there isn’t a single grammarian in all of human history who has managed to exert his authority over the entirety of a language (ok, there’s the guy who invented Esperanto, but that never really took off because he didn’t have a culture to go with it).
All of this is to say that the goal of clear grammar is not to go around flaunting our knowledge of the rules and telling people how to follow the rules correctly. End your sentences with prepositions! For that matter, start sentences with “and” or “but” or “which”! I do and have done in this blog post. Use “I” instead of “me” or “me” instead of “I” and don’t freak out about it since both those words point to the same human being.
On the other hand, make sure you are considering grammar carefully so that you are getting the right idea–the one that you intend–quickly and correctly into the head of the person that you are communicating with. (See, that was another sentence ending with a preposition.) The goal of clarity in grammar is to make our communication easier and more effective.
All of us make mistakes in our communication. When we speak, the person is there and can ask us for clarification.
When we write, they aren’t there with us and they have to guess at our meaning and intention. Or write us back asking “Huh?!” Which wastes everyone’s time.
Making sure we write as clearly as possible the first time should make us more efficient. The less mental substitution and guesswork a person has to do when reading our messages, the better.
And, correct grammar sends more signals than just “trust me; I know what I’m talking about.”
When we take the time to check our grammar before sending written messages, we demonstrate to the reader that we *care* about them and their perception of us.
Sending sloppy messages with errors suggests that we don’t care about the work the other person has to do to decipher the message and/or we don’t care about what they think of us.
Correct grammar is, rightly or wrongly, a sign of trust. When the grammar is correct, the reader tends to believe the content. When the grammar is incorrect, the reader tends to be suspicious of the content. Because correct grammar signals “I know and I care” where grammar errors signal “I don’t know and I don’t care.”
Which type of person would you rather do business with?
To check your grammar, make sure you are doing things that support good outcomes.
Most native speakers of a language have enough knowledge to check their own grammar simply by reading their work out loud. The trouble is that most professionals do not take the time to do this. Instead, most professionals write something up, glance over it, and send it out. When we glance over work we’ve just written, our brains take a shortcut and show us what we think we wrote. Not what we actually typed. Which means that we are highly likely to send out messages with errors in them if we just look it over before sending it out. And many of us trust our AI tools like spellcheck and Grammarly to underline things that could be wrong. But those tools are not as good as an actual human being. You know this because you also press 0 or say “representative” rather than go down automated customer service hell with any company.
So, if you are going to show the people you are writing to that you are both knowledgeable and caring, you need to implement better processes for proofreading your work.
Here are some suggestions:
#1 Wait 15 minutes between writing a message and proofreading it. I know that sounds like a long time, but I bet that you usually take coffee and bathroom breaks, right? All I’m suggesting is that you put that break right inbetween the writing and revising process instead of after it. It’s not more total time; it’s just a different order than you’re used to. But I promise that making this small tweak to your process will significantly improve the quality of the messages you send out.
#2 Use multiple senses. When you do review your work 15 minutes after writing it, read it out loud to yourself. And maybe even run your finger or a capped pen along the screen.
#3 Ask someone else to review it for you. And when they do, either you or they should read it out loud.
#4 Use AI tools to help you identify potential errors. But keep in mind that these AI tools have been programmed with THE RULES. They do not know when to use them or when to break them. They do not know that all rules have exceptions because they cannot be programmed with every possible exception. Also, they can’t check the validity of your content. What I mean is that they can’t make sure that the meeting is on 4/15 instead of 4/16 or that the fee is $1500 or $150. These tools don’t know anything. All they can do is direct your attention to potential problems. You have to check them.
You can make sure that your grammar is clear and mostly correct. All you need is the time and the tools to do it. And if taking that time and using those tools can help build trust with your reader, but not doing that will ruin that trust, isn’t it worth it to make sure your grammar is correct and helps ensure you are credible?