Because I Said So
By Ray Rhamey | September 15, 2011 |
Dialogue tags. In my book I call them “a game writers shouldn’t play.” I was talking about the use of verbs other than “said” to explain dialogue. You know:
”Please don’t do that,” he articulated.
“What?” she interrogated.
I’m sure you immediately saw the silliness of tags like this that try to explain dialogue by telling the reader instead of showing.
But there’s another problem with that example
I’ll wager that you didn’t even notice that the dialogue tag came after the speech. No doubt you’ve read hundreds and hundreds of lines of dialogue with this structure, a structure that I think is essentially bassackwards in what needs to happen to deliver dialogue to a reader.
I believe that where the dialogue tag goes is an important aspect of the stimulus-response sequence necessary for a reader to get what’s happening in the smoothest, most clear way.
So, in my novels, when a character utters a bit of dialogue and a tag is needed to show who is speaking, I put it at the front of the sentence.
Farnsworth said, “Where do you think the monster is hiding?”
My reasoning is simple; the sooner you know who is saying those words, the better. Why make the reader backtrack to know who said what?
That my approach differs from that of many writers was brought home to me by a beta reader of the next novel I’m publishing, The Summer Boy, when she wrote in her notes that “The frequent use of dialog tags before dialog was very distracting, even irritating before I eventually got used to them.”
I want to add that other readers did not mention this, mostly because the narrative is apparently pretty darned gripping. One wrote to me, “The story is alive. I kept reading even when my eyes were closing at night.” Love readers like that.
I agree that if there is speech after speech after speech that begins with “said” always coming first, it would be an irritating pattern. That’s why I try to avoid dialogue tags as much as possible by using action beats and to get rid of them altogether.
Farnsworth’s voice came from under the couch in a whispery hiss that ended with a sob. “Where do you think the monster is hiding?”
Nonetheless, I think the reader is better served when they know who is talking before the words come.
Delivering the sound of dialogue
I also think that the placement of how a character says a line simply has to come before the speech. Otherwise, when you read the words, you’re not “hearing” how it is said and, if you were so inclined, you would have to backtrack to re-hear the line, which no one is going to do. For instance, here’s a line from a manuscript that was submitted to me:
”Hellooooo.” The voice sounded ancient, an old lady maybe.
Besides the “telling” aspect of this, I believe that it takes a step back from what I think is the goal of strong narrative—to deliver the character’s experience into the reader’s mind.
So, in this example, it should at the minimum read in the following way so you can begin to get the sound the character heard:
The woman’s voice sounded ancient, maybe an old lady. “Hellooooo.”
But this fails to deliver, doesn’t it? What does “sounded ancient” really mean? Do you hear anything when you read those words? You have to interpret. You have to call up what an old voice might sound like and then fill in the blanks.
I don’t think we should make readers work that hard. The narrative, including description, should flow seamlessly into the reader’s mind with the experience of the sound, not a report of what it was like. For example, in this case:
The woman’s voice rasped and quivered like an old lady’s. “Hellooooo.”
”Said with”
I frequently see “he said with . . .” construction that creates the same backwards arrangement of the stimuli the reader needs to get the scene, the emotion, the reality, the experience. For example, another bit of dialogue from a manuscript:
”Ants,” he said with fear in his voice.
When you read “Ants,” you had no idea that there was fear in this character, did you? You learned that later. If the writer wants the reader to get that sense of fear, then go with something like this:
His eyes widened at the line of black specks marching under the door, and his voice broke when he said, “Ants.”
As far as I’m concerned, that comes a lot closer to delivering the emotion, and does it without using the word “fear” to tell the reader what the character is experiencing.
Am I wrong in thinking that it’s not nearly as effective to make a reader play mental catch-up to get the speaker, sound, and feeling of dialogue by putting tags and action beats after the words rather than before?
For what it’s worth.
Ray Rhamey is the author of five novels and one writing craft book, Mastering the Craft of Compelling Storytelling. He’s also an editor of book-length fiction and designs book covers and interiors for Indie authors and small presses. His website, crrreative.com, offers an a la carte menu of creative services for writers and publishers. Learn more about Ray’s books at rayrhamey.com.
I’m of a similar mindset. I try to mix up the placement of dialogue tags (before, in a break, or after the dialogue), but when multiple people are present and it is important for the reader to know who is speaking, I typically put the tag beforehand to avoid confusion.
Also like you, I tend to use action beats where possible. This has an added, hidden side benefit. There are only so many action beats you can use in all situations (characters glancing at each other, people fidgeting, pursing/chewing lips, etc.) If you find yourself forced to use the same action beats frequently, it is a good indication that you have a talking heads scene and you need to create more physical activity.
I really relate to this. A lot of times I catch myself tracking my eyes back to re-read dialogue now that I know who said it. Even though that can happen in the same sentence, it doesn’t read smoothly.
Thank you for sharing these thoughts. I frequently cannot finish a book because I can’t follow “late tags” or there are none.
It’s also extremely helpful if you write in multiple POVs to put the tags first because it helps the reader keep track of who is speaking, unless that person has already been identified by the text, then no tag is necessary.
I am glad you made this point. I am a new author, and the belief is, you start with the dialogue, then end with who said it. I have some long dialogue paragraphs, and like you believe it would be better to start with who is saying what.
Interesting perspective, though I disagree with you. I don’t think it’s a problem to read who’s speaking after the dialogue. It might be irritating to some people, but speaking only for myself, as long as the story has engaged me, somehow my eyes “catch up” and I don’t feel as if I have to backtrack. In fact, I think the first examples you’ve given above sound just fine. I wonder if readers notice this, or mostly writers notice it? To each his own, I suppose.
I do, however, appreciate your take on it. :)
As a reader, I share Rhamey’s opinion about the reader’s need to know who’s speaking before reading the words. But as a writer, I wasn’t always able to put this into practice. Sometimes, the rhythm of the dialog required me to place the tag at the end of the sentence.
Your question was: “Am I wrong in thinking that it’s not nearly as effective to make a reader play mental catch-up to get the speaker, sound, and feeling of dialogue by putting tags and action beats after the words rather than before?
My answer is: Yes.
Which is what it seems you are hearing from your beta readers. Despite your logic, and it is logical, people only have the patience to read things that fit how their mind is arranged.
Here is a point: we say , “Puerto Rico” (noun, adj) but “New York” (Adj, noun). In your opinion, should it be York New, as this takes out the mental exercise of rereading the adjective, once you know which city is referred to? This is how most of the world speaks, so why not?
Imagine how quickly T-shirts would sell with ‘I love York New’
Barb, it was just one beta reader who expressed that view. Out of the scores of readers of this novel and others that use the same approach, none have brought it up.
Nonetheless, I opened it back up and went said-hunting. In actuality, I have found very few instances in which I did as I say in the post above.
But I did find one spot where short pieces of dialogue from three characters in a row started that way, and there it was a little clunky. I changed it. I suspect it was that one that set the reader off.
Doesn’t seem to me that “York New” has anything to do with what I’m talking about as that is a place name. Nor do I think or suggest that an adjective modifying a noun should follow the noun. For one thing, they are intimately connected, one after the other. In a line of dialogue, it could be 20 words before you find out who the speaker is. I don’t think the usages are parallel.
I have no clue what the “right” answer is supposed to be, but I prefer to read dialogue where the tags, if used, follow the words. It replicates the natural world’s sequence of cause and effect, which my brain appreciates.
Also, there’s an opportunity to create microtension where I’ve read the words, but don’t necessarily know who spoke them, nor the emotion behind them. Used sparingly, I find the micro-second delay an interesting technique.
In the end, I’ve understood this to be a stylistic matter. I’ve seen advice that whatever a writer chooses, they probably want to be consistent in where they place the tags, said consistency rendering them less visible to the reader. Aaand of course, a writer might make their choice, and have an editor with a different preference. :)
I kind of disagree, but without any logical argument behind it. When dialogue tags are in the front of sentences, I find them to be like speed bumps. When there are several in a row, it sounds more like someone is reciting a story in actual conversation (i.e., …and then he was like, “Go away,” and then I was like, “No.”). Obviously, this is personal preference. I think I’m just really used to tags being at the end of dialogue. At the same time, when there are multiple characters speaking, I find myself reviewing dialogue to remember who said what. I don’t know if placement of tags would affect this; I just may need to pay closer attention while reading.
You’re correct, Jan, there really is no “right” answer. As usual, this is all terribly subjective.
I’m a psychology major, though, and I can’t help but apply what I know about the stimulus-response paradigm to what goes on a page.
Think of it this way: Pavolov’s dog doesn’t salivate before the bell rings, it’s after.
Either mode, in a fiction narrative, is of course correct. I just question which is the most effective in a reader’s mind. I guarantee you that I see in editing examples where putting the speaker last causes confusion, usually caused by what came just before the speech.
Tell you what–here and now I’ll offer to anyone reading this post a PDF version of the novel in question, The Summer Boy, so you can see if your brain just goes haywire when reading dialogue tags before the speech.
Clarity FTW, Ray. That’s for sure. :)
Oops. I should have included an email address for those who take me up on the PDF offer. ray at rayrhamey dot com
I agree that it can definitely be helpful to identify speakers prior to dialogue when it’s unclear who is speaking, but I absolutely disagree with you about describing the speech ahead the dialogue. Variation in this once in a while is fine, but I would find a novel very frustrating to read if dialogue was consistently introduced in this way.
In real life, we hear the voice AS it is speaking. On the page, however, simultaneous action is impossible. Actions are necessarily read sequentially. Even when we describe unanimous actions, they are still limited by language to be listed one at a time. In real life, it is impossible to know what someone’s voice is going to sound like until after they have opened their mouth. So when the narrator of a past- or present-tense narrative begins to regularly forecast things I am about to read, I am going to be pulled out of it.
No one would ever describe actions in this way:
“Brian stopped in front of the door. He was aghast at the state of the room on the other side. He turned the knob and pushed.”
It’s just not logical. The narrator can’t tell us what Brian sees before he sees it, so why should the narrator be able to tell us what he hears before he hears it?
“Brian turned to the woman. She spoke in a throaty whisper. “Hello Brian.”
Describing a voice as a throaty whisper is a summary–we have heard what was said and the overall impression of it was throaty. I think far and away, readers don’t mind going back once in a while if they accidentally misread dialogue once and a while. Backtracking is a totally normal part of reading to begin with. This kind of future-telling does happen once in a while, but does not come accross quite as naturally.
I think the ideal for longer pieces of dialogue is to pause and describe it somewhere in the middle.
I have to disagree as well. I’ve always followed Elmore Leonard’s rule: Dialogue tags should follow the quote and be as unobstrusive as possible. And always use, “said.” Readers are accustomed to receiving a story in a certain way. Putting the name first is jarring. Just one person’s opinion.
Hmm…As I read the examples, I was thinking of a few things. First, the old woman rasping—it would be hard to determine if that construction would bug me without seeing the sentence prior to it.
This is off the point, but in this example:
“His eyes widened at the line of black specks marching under the door, and his voice broke when he said, “Ants.” ”
That doesn’t necessarily denote fear to me–surprise maybe, but not necessarily fear. Which only goes to show what a task we have as writers in choosing the right words.
I understand the logic of what you’re saying and used judiciously, as with any other tags, it’d be fine. I think it would be as maddening as could be to read a steady stream of the “before” tags–at least until I got well used to it.
But I also don’t think readers who don’t write spend time thinking about such things. Personally, as a reader, I don’t consider it “work” to go back and re-read a sentence because first, it happens instantaneously and second, because if I’m re-reading it, it means I like what I’m reading.
For me, the bottom line is what conveys the story most clearly. Use it.
Interesting point, although generally I prefer dialogue tags after, since I’m used to it. I think in my case it doesn’t matter too much, as I read quickly and sort of see the dialogue and the tag at the same time, or at least close enough together for my brain to process them that way. Putting the tag first is more preferable if the dialogue is long, although at that point I prefer to put the tag near the beginning, where the speaker might pause for a moment. Or, as you said, eliminate the tag altogether and imply the speaker through mention of an action. That’s my favorite strategy, although it too can get annoying if somebody is fidgeting for every line of a long conversation. It all comes down to there being no hard-and-fast rules, and a combination of all these techniques will probably make for the best scene.
Thanks Ray.
Thanks for giving me a new way of looking at things! Great post.
I’m gonna have to give a “nope” vote to this one as well. That being said, I sometimes put them in front (surely I do sometimes…? probably), and of course I leave them out when I can or use action instead…but I would certainly say that the majority of my tags are after the dialogue. Honestly, I think the majority of tags are after the quotes in most books. I’m a fan of mixing it up sometimes and putting them in front occasionally, but it changes the tone to my “ear”…so I think that needs to be considered.
One thing I will say is that when climb in bed to read in a few minutes I’m going to watch and see where they are in the book I’m reading…so you made me think, that’s for sure!
Thanks Ray, these are great dialogue examples.
I’ll use quickies like “said with a snarl” in my early drafts, because I’m in the flow of storytelling and need to quickly get the gist of the scene down.
In my second draft, I replace these placeholders with description or beats ahead of the dialogue.
In my third draft, if it’s the most elegant solution, I may use an adverb. I feel a bit naughty when I use adverbs, but I swear I do so sparingly.
Incredibly helpful post — printing as I comment. Great suggestions that I will be using a little later today as I edit away on a first draft. THANKS!
I avoid dialogue tags as much as possible. You really don’t need to say who is talking every time someone opens their mouth. If there are two people in a room and Jack says something then the person who responds has to be Jill. If it’s a long exchange then it does help to remind the reader who is talking but you can do that in other ways like having Jack scratch his chin in response to something Jill said before continuing.
I’ve also seen a couple of authors format the dialogue like a play:
JILL: Says something and…
JACK: I respond.
I’m surprised more people don’t do this. A large chunk of my last novel takes place in a chat room and so I present the chat log and pretty much leave it at that.
That this was a great post is evinced by the responses elicited. Nice work, Ray.
I will be reviewing my tag placements now that you have shined a light on the subject. For myself, I try to minimize tags and, if a character has been well established, what and how the line is delivered will oft times do the identification adequately.
You didn’t mention slipping a tag into the middle of the sentence. My guide is whatever serves the flow. And, this can be illuminated by reading it aloud.
I have seen dialogue written both ways and either way is fine with me. As long as there are not that many dialogue tags. Too many tags will bounce me out of the story.
I think what most people have issue with is that we are just used to seeing it a certain way.
I have often wondered if dialogue is written the way it is because someone started doing that a long time ago and it stuck. Maybe because it was different, or just easier to write. I’m always asking why with things.
Interesting, but I’m afraid I can’t agree with much of what you said.
I find dialogue tags after the dialogue to be fine. It doesn’t hinder my comprehension of a sentence, and I don’t feel the need to backtrack, as you mentioned. It’s a seamless process for me. I don’t see any problem with mixing it up every now and again. I do it once or twice in each ms, usually for long passages of dialogue. But to see that over and over again would be irritating and distracting, even if you did break it up with action beats. My mind isn’t accustomed to reading that way.
I also don’t see any problem with using “said with.” In fact, given the example you shared above, the first example was better and I had a clear picture of how he felt about the ants.
Everyone has his or her opinion, but this post was superb (for me) because I hate it when I’ve read an entire sentence only to find out that Bob wasn’t speaking but rather it was his friend Sue or something to that effect. I find that irritating. I’m now going to go completely through my current ms and see how I feel about rewriting some of my tags. Thank you SO much, Ray.
Patti
I doubt there is a “right” or “wrong,” but rather an “industry preference” that an editor is going to dictate, no matter how you originally write the dialogue.
For me personally, I prefer the tag at the end because it creates a better rhythm to my ear. Plus, I can usually tell who’s talking and the tag is really nothing more than confirmation that I’m still tracking the conversation.
I know I have a problem with tagging. I put them EVERYWHERE. I simply want the reader to know who is speaking.
(I’m working on it.)
But I do feel that sometimes I tag before merely to show the emotion the dialog is meant to be said with. Sometimes I find it more confusing to re-read due to the fact that the emotion I had originally read it with was completely off…but the tag came after.
Just a thought.
I’ve made it my mission to eliminate dialogue tags whenever I can. I use action beats instead and feel it flows much easier. I can’t say I’ve read anything where the ‘saids’ come before the dialogue, but I think I would notice it.
Great post!
IMO, clarity, euphony, rhythm, and pacing all determine where the tag goes, or if a tag is used at all. Having it always in front of the dialogue would be an annoyance to me. It would be like reading a script. Most readers take in whole sentences at once anyway, so it’s easy enough to see who said something just at a glance.
Also, the example of the description of the voice appearing before the actual words doesn’t work for me. Yes, it would seem logical, but in fact, as soon as I read “The woman’s voice rasped and quivered…” I’d stop and wonder “What woman’s voice? What did I miss?” For me, in most cases it works better to get the description afterward. Maybe not always, though. It depends on what’s around it.
And that said…one of my little pet peeves is when writers get action out of order in the narrative, but the reaction before the action. For example: “Susie uttered a blood-curdling scream as the ghost appeared in the window.” Well, no. The ghost appeared, then Susie screamed.
I use both tag constructs, before and after. In the middle sometimes too. It’s all in how it flows when I read it aloud. That’s key for me in determining the placement. Also, ending with the dialogue instead of the tag sometimes delivers a stronger punch. It depends on the context, rhythm, pacing, and the tension I’m trying to build.
But it seems most of my editing of dialogue consists of replacing tags with beats.
Revealing too much for the reader may kill the reading experience, since throughout the novel characters usually develop their own voice in the head of the reader. In many of Dostoevsky’s novels there are long dialogues with only ‘said Raskolnikov’ at the end. For me, when reading Dosty, Raskolnikov has his own voice and not a pre-programmed one. If a character is well developed, he will have his own voice and thus adding rasps and groans becomes redundant since the text before the dialogue often describes the characters physical or psychological state.
Just sharing my thoughts on your advice. It would probably best not to listen to me since English is not my native language and I’ve never written a novel :-)
I agree that tags should be used as little as possible. That said, I don’t think it’s a problem to have a tag at the end of a relatively short sentence. Most people’s brains are able to process what was said and who was speaking almost simultaneously. The difficulty arises when the dialogue goes on for several lines before the speaker is identified.
As always, variety is key. I think I would find it irritating if every paragraph of dialogue began with a “setup” tag, but exclusive use of ant of the other possibilities would bug me just as much.
Ray!
I read a blog post on this very topic a few weeks ago then immediately did some editing then worried maybe it was just his pet-peeve.
Whew!
Thanks for all your help!
https://thewritepractice.com/a-critical-dont-for-writing-dialogue/
Interesting post! I got here through the Sterling Editing link.
I have to say, though, that I’m in the Elmore Leonard camp of placing tags after the dialogue. Not only because of the forecasting problem pointed out by others in the comments, which messes with cause and effect, but because I don’t think there is any cognitive load that needs to be lightened.
Readers are not actually re-reading the sentence to reconstruct it, but are reading the entire sentence in a block, THEN assigning meaning.
The simplest example of this is punctuation. “That is my mother.” reads differently than “That is my mother?” or “That is my mother!” But did you notice your brain backtracking? Probably you felt no confusion at all.
That’s because we don’t read character by character, or even word by word, like a computer. We don’t even read sequentially at a physiological level — the fovea of the eye is jumping around, fixating on some words and jumping to other words along the line, then back. The brain assembles meaning from all those various flashes of detail. In some ways, silent reading is more like viewing a painting than hearing a text spoken aloud.
The key for writers, I think, is to keep the tag (and the ending punctuation) within the mental “buffer” of the reader. If the tag comes much too late (after three lines of text, it’s revealed that the dialog is spoken by the dog), then you trigger actual re-reading. (On the other hand, this is a useful tool for comedy.)
The good news is that the brains of readers are flexible, and you can train them to read your sentences, even if they’re constructed differently than other writers’. Hell, you can go full Yoda on them, and they’ll still follow you, if the story’s good enough. But it seems to me that that retraining is adding to the cognitive load, not lessening it.