Writers-Whats the best way to stucture dialog?

I’m working on a novel and have occasionally passed around chapters I’m more or less done with to friends, family and a writing group I attend. Usually the biggest thing that gets pointed out is that my dialog should have breaks in between speakers, whereas I usually tend to integrate dialogue into a paragraph.

For example:

George said “Hello”. Anna responded “How are you doing?” George said “I’m doing well”.

That’s not an actual example from my novel but it does show how I tend to structure dialogue. Apparently, though, I’m supposed to be writing it like this:

George said “Hello”.
Anna responded “How are you doing?”
George said “I’m doing well”.

What is the rule for writing dialogue is what I’m asking? I never knew there was one in this regard, just that dialogue should serve a purpose

“New dialogue is always its own paragraph, unless you have a very compelling reason to have two characters speaking in the same paragraph,” Alice said.

“Why should it always be a new paragraph?” Bob asked. “Is there a novel police force that’ll throw you in jail?”

“No, there’s not. I mean, nobody will enforce this rule. But when you’re writing, your first concern should always be clarity. Which way is easier for the reader?”

“Why does that matter?”

“Because it’s unbelievably obnoxious to see all the dialogue mashed together. Not to mention confusing. Don’t ever give a reader a reason to put your work aside. It’s pretty self-defeating.”

Bob cocked his brow. “But it’s my book. Shouldn’t I be allowed to do whatever I want?”

“Yes, of course you can. Again, there’s no novel jail. But if you want others to read your book and take you seriously, you’ll have an understanding of convention. And in fiction, the convention is to start a paragraph for each new speaker.”

“Even if it’s just one word?”




You can break any rule if you’re a genius.

If you have to ask, you’re not a genius.

Therefore, don’t break the rules. Your reader will thank you. This applies to all the unwritten rules of writing. They’ve evolved over centuries because it’s hard enough to get lost in somebody else’s words without putting up additional artificial barriers.

If you don’t know what a rule is, pick up a book and look. You’ll see that 99.9% of fiction is written with the rule that each person starts a new paragraph. (If you haven’t read enough to internal that, you might want to go back and see what other conventions you’re missing.)

There are a couple of good reasons for giving each speaker a new paragraph. In the first place, it makes it really clear when you’ve switched speakers. If you mash it all together in one paragraph, the reader can easily become confused about who is speaking. And it also prevents having quotation marks butting up against each other which creates a big jumble of punctuation. For example:

“I wouldn’t do it,” The Hamster King said. “It will confuse the reader.” “Confuse the reader? How?” “By making it unclear where one person’s speech ends and the next begins.” “Oh,” HPL said, “I see what you mean.”


“I wouldn’t do it,” The Hamster King said. “It will confuse the reader.”

“Confuse the reader? How?”

“By making it unclear where one person’s speech ends and the next begins.”

“Oh,” HPL said. “I see what you mean.”

If you follow convention, it makes it easier to read. The thing I noticed about the OP, in both examples, is every change of speaker was explicit. That struck me as awkward, and I didn’t like it. I wonder if that’s something you need to do because you don’t break the dialogue up.

Agreed. Many dialogue tags are not necessary and are a wordy distraction besides.

What’s more, I’d say that even WHEN dialogue tags are used, nine times out of ten they should follow the dialogue, not precede it as in the OP. Using preceding tags really murders any sense of “flow” you might have for the dialogue.

A little trick that can help make things read a little more naturally is to replace preceding dialogue tags with small bits of character action:

John’s eyes glanced out the window. “The weather’s looking better.”
“What a boring thing to say,” she replied.

It’s not something you want to overuse, but can be a good way to avoid overusing bland tags.

I half agree with you.

I’d hate to see this:

Ari asked, “What’s that?”

but I don’t mind if they occasionally mix action in with preceding dialogue tags like this:

Ari’s eyes widen in surprise before he asked, “What’s that?”

I think it’s less awkward than doing it that way than telling us his reaction to surprise after he spoke.

Another rule you’re breaking is that, in American usage, periods always go inside the quotation marks.

It might be a good idea for you to read a good book on the conventions of writing. That, and/or read a bunch of novels and really pay attention to those mechanical kinds of things. Anytime you do things differently from the commonly accepted style, you run the risk of confusing or distracting your readers.

It’s been said in here, but to clarify, when you start a new paragraph with each speaker switch, you don’t even have to throw in any words to say that someone else is speaking.

Jill walked into the meeting room, strolling over to where Jack sat eating his lunch. “Whatcha doin in here?”

“Just grabbing some quiet time for lunch,” Jack said, “everyone bugs me with stuff to do when I’m eating at my desk during lunch.”

“You know, you could go eat outside in the courtyard.”

“Yeah, but I forgot my jacket this morning, and it’s a bit chilly outside. Do you need the room?”

Jill shrugged, “We have a meeting in 20 minutes, as long as you could clean up before then, it shouldn’t be a problem.”

“Cool, see you after work for coffee?”

“You betcha.”

Also, if you want to pace the dialog, you can throw in the transitions anyways. Jill shrugging shows that there is a pause before her response instead of just a quick flowing reply. Once you introduce all the speakers, especially in two-person conversations, there’s no need to point out who is speaking. One person obviously starts speaking after the other one is finished.

When writing a long scene that consists mostly of dialogue, I aim for the following structure.

*“My feet hurt,” **Skald **said.

“Oh, you poor baby,” Viva replied. “What’s wrong?”

“It’s been a long day.”

“Go on.”

“My feet hurt because it’s been a long day,” Skald said.

“Huh? That doesn’t make any sense.”

“What do you mean?”

“The fact that it’s been a long day doesn’t have any direct connection with how your feet feel,” Viva said.

HPL came into the room just then. “What are y’all talking about?” he asked.

“I’m not sure,” Viva said. “Skald was telling me that his feet hurt, but his explanation doesn’t make much sense.”

“Sure it does,” Skald said.

“No it doesn’t,” Viva said. “Just because it’s been a long day doesn’t mean your feet should hurt.”

“I have to agree,” HPL said. “What exactly do you mean, Skaldie?”

“I mean that I’ve been wearing new shoes all day long, and while they didn’t bother me at first, eight hours later they do.”

“Well, why didn’t you just say that?” Viva said.

“Because my feet hurt and that is distracting me,” Skald said.*

My idea here is that each time a character speaks for the first time, there needs t be a tag. If only two characters are speaking, then, after those two initial tags, I note who is speaking on every third change. But if more than two characters are speaking, I tag each change, unless one character addresses another by name in asking a question and the next speaker the character so addressed. I find this to be minimally intrusive, yet clear. I also avoid minimize tags like he asked, and she replied; those are allowed a maximum of once per page. And I refuse anything more elaborate, like “I explained,” except for comic effect.

An exception to this if if I have two characters with sharply different speech patterns. If, say I were writing about a scene with Spike and Illyria from Angel, I might not bother with any tags after the initial two, because Illyria’s speech is so distinctive. But if it were Spike, Angel, and Wesley talking, I’d stick with the pattern described above.

ETA: If I began a paragraph by noting what a character is doing, I consider that the tag and don’t include the “X said” (though I may use a “he/she said.” And I DON’T do paragraphs that describe one character’s actions but have another person speaking, because it’s ambiguous. Also, if two characters in the dialogue are the same gender, the tag is always “X said,” never “he said” or “she said.” I never want my reader to have to stop to figure out who is speaking. I hate that when I’m reading.

Looks good to me. Reminds me a bit of Hemingway’s “Hills Like White Elephants,” which relies heavily on dialogue and uses only a tag now and then so you remember who’s speaking.

I think of Hemingway when I structure dialogue. Sometimes I am tempted to eliminate the dialogue entirely, but I rarely do. For instance, I’ve a story in the pipeline in which three of the four characters are from Memphis, and the fourth is from New York. Part of me thinks that, since the Knickerbocker never says y’all or similar regionalisms, while the Southerners always do, I could eliminate the tag for him every time he says you. But I don’t want the reader to have to infer that, as I don’t want to highlight the mechanisms of dialogue and characterizations.

ETA: Also, in the snipped I posted above, I would probably edit it so that dialogue always began a paragraph. It’d be

“What are y’all doing?” HPL said. He had just come into the room.

rather than what I wrote.

Or just ‘John glanced out the window’.

Skald: for me, that snippet has dialogue tags maybe just a bit too often. When I encounter them frequently with lines whose speaker I had already identified by context and alternation, they begin to feel repetitive. This is especially true when the tag comes at the end of a long-ish line; if I already know who’s speaking, the tag isn’t necessary, but if I have to get to the end before I know whose voice to read it in, something’s wrong!

A related issue that I don’t think has been brought up specifically yet but is worth watching out for is the use of description in dialogue tags when they follow the line being spoken. For example:

“I’m going to rape you and kill your family,” Bob said cheerfully.


“This is without a doubt the most joyous moment of my life,” Eve deadpanned.

In both cases the initial reading of the line may be at odds with the way it’s subsequently described. If you need to set the tone of a particular line, you need to do it before the reader has already read it, which may not be optimal if you have an aversion to putting a dialogue tag at the beginning of a paragraph.

Personally, for this and other reasons, I like to replace most dialogue tags with action beats. Not only do they provide a more visual way of describing character interaction, they’re a valuable tool for varying the pacing and intensity of exchanges. And, of course, they cut down on the number of “saids.” I like to keep the use of “said” to a minimum. Other tags, like “asked,” “replied,” “exclaimed,” and so on, should be used approximately never. :stuck_out_tongue:

Well, sure it does. I wrote it in about 30 seconds, after all. :cool:

When I’m rewriting, I edit out a lot of dialogue tags as unneccessary. In the fantasy novel I used to post about here, for instance, the three protagonists have very voices. One speaks very ungrammatically (what we’d call country hereabouts), one speaks very proper unless exhausted, and one talks rather like Thor, God of Thunder. I can get away with no tags at all if they’re the only characters talking.