The speech tag is among the most utilitarian features in a piece of writing. Those little phrases connected to a fragment of dialogue—”he said,” “she asked,” “said the barkeep”—tend to serve exactly one purpose: making it clear who is speaking. They are as close to purely structural text that one will find in a story or an essay; they may be integrated into the main body of the text, but functionally they are more like the name labels in a script or an interview. They may assist the reader in comprehending and interpreting the text, but they don’t exactly contribute to its meaning.
A general rule of good writing is that the speech tags should not call attention to themselves, as they are boring almost by design. Oftentimes beginning writers, realizing that speech tags sound kind of dull, will attempt to spruce them up by using fancy synonyms for “said” or “asked,” or by appending needless modifiers to them. In doing so, they end up drawing the reader’s focus away from the dialogue, away from the important part of the sentence. And this assumes that speech tags are necessary in the first place. After all, if the voices of the characters are sufficiently distinct, the reader can suss out the speaker from the dialogue itself. (I’m certain that half the reason Ernest Hemingway’s “Hills Like White Elephants” is a classroom staple is how well it demonstrates this fact.)
Still, I’m not happy to just take a general rule and leave it at that. In much the same way that I’ve thought about how one can effectively use inanimate objects as narrators, I’ve wondered if there are ways to use speech tags creatively. One possibility that comes to mind is to vary the placement of the speech tag relative to the dialogue. It’s standard to close the sentence with the speech tag, but one can move it closer to the middle or beginning of the quoted text to suggest, say, a pause in the speaker’s delivery or an emphasis on an unexpected word. One may even do so just to give the reader a place to mentally breathe in a long passage.
This technique is definitely useful, if for no other than that it varies the rhythm of a conversation on the page, but I’m not sure I’d call it a creative use of speech tags, per se. When the speech tag interrupts dialogue, the content of the tag doesn’t really matter, only it’s presence. One could insert a brief action or a bit of description and obtain a similar effect. No, I’m looking for something stronger: a way to make a phrase like “I said” interesting in itself.
A speech tag doesn’t give the writer much meaning to work with, granted. But it does provide the reader with one indisputable piece of information: someone is speaking. If the act of speaking is thematically important to a piece of writing, then a writer could use speech tags, could use the constant repetition of “said” and such words, to underscore that theme. In fact, I’ve come across two short stories that do precisely that: “Fat” by Raymond Carver and “Say I won’t be there” by Ali Smith. Both stories revolve around characters who are struggling to communicate something, and both use an excess of speech tags to highlight that struggle.
“Fat,” which was included in Raymond Carver’s 1976 collection Will You Please Be Quiet, Please?, tells the story of the narrator’s encounter with a fat customer at the restaurant where she works as a waitress, and of the effect that the encounter has on her. When I say that it tells the story, I mean it is explicitly about the telling of it. It begins thusly:
I am sitting over coffee and cigarettes at my friend Rita’s and I am telling her about it.
Here is what I tell her.
It begins with two sentences, each its own paragraph, devoted to what at first seems like throat-clearing. In fact, Carver is setting up the story’s fixation on the act of speaking. This becomes apparent when the narrator first approaches the fat man to take his order:
Good evening, I say. May I serve you, I say?
Rita, he was big. I mean big.
Good evening, he says. Hello. Yes, he says. I think we’re ready to order now, he says.
Carver does not merely include those utilitarian speech tags; he uses way more than is necessary to get the point across. It’s not as though the reader will lose track of who is speaking mid-paragraph, so they must be serving some other purpose.
From what I see, these extra speech tags accomplish two things. First, they replicate the experience of orally telling a story with lots of dialogue. Unlike in print, the audience for an oral story does not have punctuation marks or paragraph breaks to clearly delineate whose dialogue is whose, so a good storyteller will need to remind the listener of who is saying what at more frequent intervals. Second, the repetition of “I say” and “he says” convinces the reader to pay special attention to the dialogue, even as we might be tempted to skim past all these pleasantries. The narrator certainly finds this dialogue interesting, noting that the fat man “has this way of speaking—strange, don’t you know.” Without all those speech tags slowing us down, we might not notice the fat man’s peculiar use of the royal “we,” which he ends up using consistently throughout the story.
Of course, if we pay close attention to the dialogue—and by extension, to the language the narrator uses throughout, for this whole story is being spoken—we get the sense that the narrator is not saying as much as they would like to. She has clearly been affected by that night with the fat man: by his speech, by his huge fingers, by the comments others make about him. But she seems unable to clearly vocalize just how she’s been moved. Instead she must resort to vague statements like, “Now that’s part of it. I think that is really part of it.” One senses that her only choice is to recap the entire evening, in the hopes that the feeling will shine through. Indeed, moments like the fat man’s last line come tantalizing close to an epiphany (“If we had our choice, no. But there is no choice”).
But, given Rita’s reaction to the story, it would appear that telling a story doesn’t mean conveying it. After the narrator mentions how she felt “terrifically fat” while her boyfriend Rudy raped her, she realizes that Rita has missed whatever point she was trying to make:
That’s a funny story, Rita says, but I can see she doesn’t know what to make of it.
I feel depressed. But I won’t go into it with her. I’ve already told her too much.
No matter how many times that narrator “says” something within her story, she cannot guarantee that her audience will understand. That goes not just for Rita, but also for the reader, for whom the last lines are famously enigmatic: “My life is going to change. I feel it.”
Whereas in “Fat,” the problem is that the audience doesn’t understand the narrator, in Ali Smith’s “Say I won’t be there” (collected in the 2015 book Public Library and Other Stories), the problem is that the audience doesn’t want to listen to her. As it happens, the story starts off in a similar fashion to “Fat,” with two one-sentence paragraphs that foreground the act of telling a story to an audience:
I had a dream, I say.
Don’t tell me about any dream right now, you say, I can’t listen to it right now.
And, much like Carver, Smith makes liberal use of mundane speech tags to reinforce the importance of speaking, although in her story’s case there’s a more palpable sense of frustration:
It’s not just any dream, it’s the recurring dream, I say. The one I’ve been having all year. I had it again. I keep having it.
Tautology, you say.
What? I say.
You just said the same thing four times over, you say. And I can’t hear about your dream right now. I’ve got work in a minute.
Something is obviously eating away at the narrator. They’re having trouble getting it out, though, so they just end up repeating themselves as a preface, saying that they’re want to say something. Not helping matters, they don’t have someone like Rita who is at least game for a story; the audience here is actively trying to shut down the discussion.
One thing you’ll have notice is that “Say I won’t be there” features not only a first-person narrator but also addressee, in this case the narrator’s romantic partner. The use of “you” in this context encourages the reader to place themselves in the perspective of this character, which is a rather conflicted position: both intimate (the reader is sole audience member) and confrontational (the reader-character resists being the audience). This tension, this need to negotiate between the desires of the reader and those of the reader-character, may explain why the story progresses the way it does, with the addressee insisting they don’t want to hear about this dream while they keep asking questions about it.
Over time, we get a rough notion of what the narrator’s dreams: she keeps hearing stories about how Dusty Springfield was being photographed in a nearby graveyard. The addressee finds this perplexing, because that sort of dream would seem to have much more relevance to their own life—their a fan of Springfield’s music and they work for a company that wants to repurpose a graveyard for commercial purposes. They playfully accuse the narrator of “filching [their] subconscious,” which brings the narrator back to an earlier period in their relationship, back when they made it a point of sharing dreams with each other. They’d write them down “because it’s really boring to have to sit and listen, in the morning when you’re hardly awake yourself, to a dream someone else has had.” In that moment, the narrator’s entire quest seems both hypocritical and hopeless.
As it turns out, the narrator is not the only one with a lot to say that they’ve been holding back. Later that day the addressee, inspired by their earlier conversation, sends the narrator a text message, an email, two voicemails, and a letter, all filled with fun facts about Dusty Springfield. It’s as though the addressee sees the earlier dream discussion as an excuse to share their interests with the narrator. Through that deluge of a trivia, though, it’s difficult to see what, if anything, the narrator takes from it—a role reversal from earlier on. This is why I find Smith decision to have “I say”/”you say” volley back and forth so crucial: it implies that the narrator and the addressee are in similar positions. We are reminded, constantly, that both characters are speaking, but never reassured that either is hearing.
From the examples of Carver and Smith, we can see that beyond structuring and stitching together dialogue, speech tags are an excellent tool for getting the reader to consider speaking as an action. Both “Fat” and “Say I won’t be there” are stories driven by and about the struggle to communicate; that struggle is where the majority of the conflict lies. Not all stories feature such conflicts, and even in those that do it may not be necessary or advisable to go to the extremes of Carver and Smith. But if you’re writing a passage of tense or uncertain dialogue, perhaps consider what those little words around the quotation marks might do for the scene.
What are your thoughts on speech tags? Are there other potential ways of using them for creative ends? Can you think of any stories which use them especially well, or poorly? Let me know in the comments! And if you want more advice on making the most of the little things in writing, you might want to check out my post on how Brave New World and Hiroshima use section breaks.