InboxWednesday – Dialogue Construction

Hello and welcome to InboxWednesday, where I use the carnival claw machine on my inbox to pull out a question and answer it for public consumption. If you have a question you’d like to see answered (I should point out that if you submit a question, you also get an answer in email, I don’t just leave you hanging), you can email me.

Today’s question comes from Mark, who asks:

Hey John, got a question for you. I’m writing a lot of dialogue, and don’t want to keep saying, “Bob says,” before every quoted thing. What are some things I can do, and how can I keep it interesting, not just for the reader, but for me as I keep writing?

There’s actually a name for this issue, and it’s constructed repetition. Normally it’s a speaking concept where a person references the same phrase to establish it as a buzzword or theme for their discussion. Partner that with a physical gesture, and you’re looking at one of the tenets of neurolinguistic programming.

But since Mark isn’t asking me for a recipe on how to rally loads of people into action or organize disparate minds into a frenzy, we’ll just look at the constructed repetition that causes boredom.

To get into this, we’re going to have to first talk abstractly about what it means to read. I’ll spare you the load of science I know, and condense this down to one idea — reading is about pattern recognition, where the patterns are shapes and lines that have a sound associated with them. We call them letters, and groups of letters are called words. We see the same words over and over again, and we intuit and then understand their meaning.

This is how we know that a “dog” refers to a domesticated pet, and not that thing with a peel you eat as part of a balanced breakfast. This collective understanding is called context.

Pattern recognition works because our input system (eyes, braille, hieroglyphs, cuneiform, etc) processes the same symbols over and over again, and we derive meaning from them. (We call this reading.)

The tricky bit about reading is that it’s also about sameness in those patterns. We see the same letters forming the same words repeatedly, and we get lulled into less careful processing. We realize that all the paragraphs are going to start with “So”, and then we skip that word, because we’re used to seeing it, and we’ve built this momentary expectation that it’s going to be there, so skip ahead and get to the new bits.

When we get to dialogue, we need to focus, because not only can that dialogue convey new information (I mean stuff we didn’t already know, not specifically plot-stuff), it’s an opportunity for us get insight or access to the character doing the talking.

The two words Bob says are called a dialogue tag, because they tell us who is expressing themselves and how. The who is most often a proper noun, though it can be a pronoun, and the how is a combination of the verb, but also any adverbs hanging out nearby. Remember that verb choice carries with it an expectation and a context, so that we can distinguish saying and yelling by different volumes, for instance.

Dialogue tags can be found in three places: pre-line, mid-line, and post-line. Let’s take a look at each of them. And it’s important to think about a dialogue tag as two+ words functioning as one item. It’s not just the “said” or “Bob”, it’s their combination that we’re talking about here.

Bob opined, “Why can’t anyone pay attention?”

The sentence opens with a tag, using the comma to indicate a pause, and the quotation marks to trigger a switch from exposition/narration to character. All dialogue tags mark and make obvious this transition.

By setting this on its own line, as you should always be doing, you’re making a paragraph.

Giving Bob two more things to say, we can create this example:

Bob opined, “Why am I an example?”
Here is a sentence about something happening in the room.
Bob huffed, “I’m not sure I like being this example.”
Something else happened in the room.
Bob spoke, “I’m glad this example is over.”

Remember that Brady Bunch complaint about “Marcia, Marcia, Marcia” ? That’s what you get with a proper noun + verb dialog tag as a sentence or paragraph starter. It sounds whiny. It reads monotonously. It’s not striking visually. It seems safe and simple.

Is it wrong? No. It’s sort of like setting the cruise control at one mile under the limit then sticking to the center lane – you won’t get a ticket, but you’ll only aggravate the people around you.

There are uses for it, sure, but to have it dominate the text is to get a bit formulaic. Here’s another point about pre-line tags – where do you want me to focus. You’ve name-checked Bob, but to what idea within this line of letters am I supposed to be rapt? Should I care who’s talking and how they’re doing it, or am I supposed to care about what they’re saying?

“This,” Bob said, “isn’t much better.”

Don’t let the ‘mid’ fool you, the tag doesn’t need to be precisely 50% of the way in the line to have the effect I’m about to describe. It could be one word (most common is one to three), it could be a whole sentence, then a tag, then another heap of words. So what’s wrong with this?

First, it makes the text look like a pair of bubbles or wings, with the tag between them. My second writing professor referred to them as “swimmies on a toddler”, meaning those blood pressure cuff inflatable pontoons you lash to a kid’s biceps before putting them in water.

Second, and more critical to me, you’re breaking up the line of dialogue just to tell me who said it. It’s a sometimes unnecessary pause. If the context is clear enough, I should be able to figure out who’s saying what, particularly if you’ve only got the two characters and they’re going back and forth. Bob, then Alice, then back to Bob … even if you don’t name check a character with a “Holy dicksnot, Alice” or “By Zeke’s flowing beard, Bob!” So long as each line of dialogue gets its own line, trust me to be smart enough to know who’s saying what to whom.

Again, this has its uses, particularly if your swimmies run long and technical (I’m looking at you, SF/F manuscripts redolent with technocrunch), or if you adding a bit of weight to the moment the dialogue speaking creates. As in:

Now,” Bob loaded the shotgun, “this is where I stop being your monkey.”

“I think I really love Janice from the examples in the draft,” Bob said.

Here the dialogue tag is hanging out at the end of whatever’s been said, and that’s often vestigial. Again, context should help the reader figure out, so sometimes we don’t need any tags.

The tag at the end can render the whole line in a kind of decay, where all the good stuff happens (long) before the period, which means I don’t need to be paying complete attention at the end of the sentence, since I’m only there to see what the character says and move on.

So What’s The Fix?

The fix is pretty straightforward – mix and match. A combo of pre-and mid-. Throw in a few post-. And stop relying on the same verb. There’s nothing wrong with “said” but there’s nothing wrong with changing up the word while still being specific as to what’s happening.


Loads of words exist, and use them. Place tags where you like and need to, but also be willing to not have any. Remember, you’ve got revisions and editing and beta readers to tell you if something does or doesn’t need a tag.

Break the repetition. Make a conscious effort to try and not go with that first instinct with “said.” You might just surprise yourself.


Mark, I hope that answers your question. It’s a kinda crunchy one, and I’m hoping the examples streamlined it a bit. Let me know how it goes.

I’ll see you all back here Friday for more bloggity goodness.

Happy writing.

Leave a Reply