Writing: Collaborative experiment: Speech tags · 3:36am Feb 12th, 2014
"I'm curious about speech tags," Bad Horse whinnied.
"Speech tags?" fimfiction asked.
"Those things that come before or after a quoted statement that indicate who said it and how they said it," Bad Horse answered. "Only sometimes they're not there. Like this:"
"Oh. Those things. They're supposed to be bad."
"Right," Bad Horse said, settling back into his armchair and producing a pipe. "But back around 1600, when people started writing novels, they were required." He lit the pipe and puffed on it professorially.
"Why?" fimfiction asked. "I don't always need them."
"Grammar," Bad Horse said.
"Oh." fimfiction thought about this for a moment. "But, wait--a quoted sentence is still a sentence. Grammatically, it's no different than if it weren't in quotes."
"No," Bad Horse said, pointing the stem of his pipe at fimfiction. "A sentence is the narrator telling you something that happened in the story world. A sentence in quotes is a character claiming something happened. They're semantically distinct. If you write a computer program to parse English and convert it to logical propositions, you'll find it matters a great deal whether something is in quotation marks or not, and who said it. The fact that some 19th-century amateur grammarian said they could both be represented by 'S' in his stupidly-oversimplified grammar doesn't change that."
"And yet," fimfiction said, "to me it sounds perfectly acceptable to alternate quoted sentences with no speech tags."
"Really?"
"Yes."
"And what does that prove?"
"Well, that speech tags aren't necessary. You are a descriptivist, I believe?" Here fimfiction peered suspiciously over its collective eyeglasses at the horse.
"Of course, of course. But I'm also something of a"--he coughed apologetically--"an aesthetic objectivist."
"No!" fimfiction gasped.
"Yes," the horse confessed in a low voice. "I like to believe that some things are better than others."
"Really." fimfiction glanced furtively about the surrounding internet, as if afraid of being seen with the horse. "Well," it went on in a firm, respectable, no-nonsense kind of voice, "I don't see what all this has to do with my OC / Twilight ship."
"It has everything to do with your fic," the horse said. "If stylistic rules are self-fulfilling prophecies that twist our perception, then every rule of style we know must be held up to the light and re-examined, post-haste!"
fimfiction coughed. "If," it said. "That's a lot of contention to hang on an 'if'."
"And that," the horse said, "is why we should look into speech tags. If we can find when the elimination of speech tags began," the horse went on, ignoring fimfiction's reservations, "and when grammars began treating quotations like any other sentences, we can see which came first. If the elimination of speech tags came first, and the grammarians followed after, why, then, they were merely describing what they saw, and the elimination of speech tags was the work of authors and readers. But if"--here the horse leaned forward conspiratorially--"if the grammar came first, then our aesthetic judgement followed the prescription. That would prove that our sense of style can be reversed entirely by an arbitrary and mistaken rule, without our even being aware of it!"
"I daresay that would be a death blow to your objective aesthetics."
"Not a death blow," the horse replied. "But a serious complication. It would make it damnably difficult to tell whether even the most unanimous writing advice were objectively good advice, or merely a cultural artifact."
"So what do you plan to do about it?"
"Well, for a start, we can look at novels from different years," the horse said, "count the number of speech tags, graph them by year, and see when they disappeared."
"That sounds like a lot of work."
"Indeed," the horse agreed. "And that's where you come in."
Here are some different types of speech tags. I give multiple examples of types when they have different variations.
1a. "That was fast!" Tom said.
1a. "That was fast!" Tom said swiftly.
1a. "That," Tom said, "was fast. I barely saw it."
1a. Tom said, "That was fast."
1a. Pangloss made answer in these terms: "O my dear Candide..."
1a. "Seek you any here?" asked Heyward; "I trust you are no messenger of evil tidings?"
1b. "That was fast!" Tom gushed / sputtered / exclaimed / ejaculated / shouted / speculated / offered. [Use of a tag other than "said", "continued", "went on", "added", "replied", "asked", "answered".]
1b. He mentioned the Count de Vereza. "He is certainly of all others the man most deserving the lady Julia." [I'm counting this as type 1 because "He mentioned" obviously refers to the quotation.]
18th and 19th century books often made a paragraph break just after "said" and before the quotation, like this:
She inquired of the youth who rode by her side:
"Are such specters frequent in the woods, Heyward?"
Count that one as a type 1b.
2a. "That was fast," Tom said. "I barely saw it." [A tagged quote followed immediately by an untagged quote.]
2b. "That was fast!" Tom exclaimed. "I barely saw it."
3. Tom looked up. "That was fast." [No tag, but next to an action whose actor we can assume said the quoted statement.]
3. "That was fast." Tom looked up at the sky.
4. "That was fast." [No tag; no adjacent sentence to indicate who spoke.]
I'd like to look at a bunch of novels and short stories in English (or translated into English), spread out at least over the time period 1700-present, and from each take the first 30 quoted statements and classify them as type 1a, 1b, 2a, 2b, 3, or 4. Don't count quotations that have an implied or indistinct speaker like this:
In his reading he came upon courtships and cartels, where he often found passages like "the reason of the unreason with which my reason is afflicted so weakens my reason that with reason I murmur at your beauty;" or again, "the high heavens, that of your divinity divinely fortify you with the stars, render you deserving of the desert your greatness deserves."
or
The host fancied he called him Castellan because he took him for a "worthy of Castile," though he was in fact an Andalusian.
Then summarize like so:
Year Translated Author Title 1a 1b 2a 2b 3 4
1605 1885 Miguel_de Cervantes Don_Quixote 20 9 1 0 0 0
1884 n/a Mark_Twain Adv._of_Huck_Finn 15 0 0 0 0 15
2003 n/a Jhumpa_Lahiri The_Namesake 15 3 6 1 2 3
I put underscores in the names because it makes it easier for me to copy-and-paste into Excel.
If you'll help by doing this for one or more books or stories, and post your results here, I'll compile them and we'll see if there's a pattern.
The data so far:
Jesus fuck it's about time I was vindicated on this. Sorry, Bad Horse, but I'm gonna stand against you on your objectivism...
Crowd-sourced literary analysis? Interesting. Not sure if I'll get in on it, though it's certainly relevant to my interests...
Well, maybe I'll dust off my copy of The Count of Monte Christo. Still, by no means should anyone read that as calling dibs.
Jesus Pope Fucking Christ, what the hell passes through this horse's mind in the wee hours of the night?
Tab doesn't work in comments apparently.
Year...Translated..Author.............Title............1a..1b..2a..2b..3..4
1942..1982...........Albert Camus..L’Étranger..18..8....0....0....4..0
Some I'm not completely sure about:
1a? …told me, “… (2x)
1a? …said to me/him, “… (2x)
1a? I said to the caretaker, “… (1x)
1a? I’d intervened to say, “… (1x)
1b? I interrupted him, “… (1x)
3? He gave the order into the telephone, lowering his voice, “… (1x)1825305 I don't see it as a two-sided issue. I expect that many stylistic rules are arbitrary conventions. The dislike of adverbs, for instance. I have a theory that most rules of writing style are just rules to "not do things the way our enemies do them". So when the modernists rose to power around 1930, they looked at 19th century literature with its adverbs, long sentences, flowery language, long introductions and languid pacing, and said, "You must eliminate adverbs, eliminate purpose prose, write short punchy sentences, get immediately to the point of the story, and keep the pace moving fast!" Similarly, the rules of literary fiction are largely not to do anything that might get your story confused with genre fiction.
So I expect that the grammar might have come first with speech tags. But I'm still objectivist about it--I think using speech tags is objectively right, because it makes the story objectively easier to understand. Believing in objective aesthetics doesn't preclude me from saying the rules of style may have shifted due to some arbitrary social convention. Then I just say that society's wrong.
And this isn't an airtight experiment, because it might be that the grammars just didn't mention sentences in quotation marks, ever, and at some point someone leapt to the conclusion that that meant that sentences in quotation marks didn't need speech tags. We probably have no way of knowing when that happened. It might even turn out that grammars have handled the issue pretty much the same way since before there were novels.
FOR SCIENCE!
Year Translated Author Title 1a 1b 2a 2b 3 4
1759 late 18th c. Voltaire Candide, ou l'Optimisme 25 0 0 0 0 5
Some dialogue did not use quotation marks, like this:
I didn't count these.
Variety is the spice of life.
As to your supposition, I think you might be looking at this wrong. I'd narrow down the search by looking for the first published article to decry the use of speech tags. From there I'd look at stories published in the years preceding it. So, if the first article was published in 1743, I'd look stories written in 1733 to 1743. Thus giving us more concise data points.
Hmm. Definitely a type 1 if that's a comma or a colon after "voice". I'd probably still count that as 1b if it's a period, because it's elaborating on "gave the order".
1825512
1b it is then. I'll adjust it accordingly.
1825502 Interesting! Were those type 4 quotations in paragraphs by themselves? I should probably require that for type 4.
1825521
Yes.
There was one other odd case, where explicit attribution was given across a paragraph break:
I counted this as 1a.
Year Translated Author Title 1a 1b *
1749 N/A Henry Fielding The History of Tom Jones, a Foundling 15 11 4
The asterisks are
1. a tag with a speaker, but no speaking verb: A third, "Ay, this comes of her learning."
2. "concluded with saying," followed by the quoted speech
3. "adding" (I'm not sure if it's close enough to "continued" to warrant being placed in 1a.)
4. "tells"
One instance of 1a was ". . . says she;" followed by the quoted speech, but there were other instances of use of a semicolon as a comma.
One 1b
That... was almost my actual objection.
I am not interpreting the examples in 1b as an exhaustive list- "added," for example, I'm putting in 1a instead of 1b.
Here is one, with more to follow when you write a happy Big Mac/Twilight ship:
1897--Bram Stoker--Dracula--21-3-4-0-1-1
Year Translated Author Title 1a 1b 2a 2b 3 4
1899 Joseph Conrad Heart of Darkness 21 4 0 0 5 0
By the way, if you want to help and can't think of something to look at:
http://www.gutenberg.org/ebooks/
1825495
1825305
Question: do either of you believe that style is what makes a story good or not?
This is graduate thesis level stuff here, and you're just throwing it out to us. You're nothing if not ambitious, Mr. Horse.
I am afraid I don't have much of a classical library, and what I do have is more often audiobook than anything else (which makes the whole process hard enough to not bother with), but I will look through what I have in the next few days. I look forward to seeing how the numbers stack up.
1825619 Can you put the zeroes in there? I can't tell if the 5 goes under type 3 or type 4.
1825653 I don't think I can even say what's style and what isn't. Like, Hopi Coyote stories have a style to them, but part of the style is a particular sense of humor, and an attitude toward Coyote, and toward life. It's pervasive. You couldn't take a story by Hemingway or Bukowski and rewrite it in Hopi Coyote-story style.
Take Cormac McCarthy. On the one hand you have his little-s style: He refuses to use quotation marks or apostrophes, uses commas where he should use periods and nothing where he should use commas, refuses to join "and" clauses with commas, doesn't use speech tags even when there's no way to tell who's talking, and doesn't care whether his pronouns resolve to the correct nouns. Basically his entire surface style is a juvenile rebellion against grammar, stupid shit that makes his stories harder to read, but also makes it obvious that this is Literature, not a Louis L'Amour western. For instance:
But on the other hand, his big-S style is the way he lingers over the landscape, and takes his time describing small things, details about the clouds, the birds, a horse, or a stream. He goes on in this simple way, then suddenly lets loose with a long run-on sentence like this:
Exactly where and when he chooses to do that is part of his Style. Saying the man stood "like a man come to the end of something" is part of his Style. I hate his style but love his Style.
I care a lot less about little-s style than most readers.
1825653
I most certainly do. How good your actual sentences and descriptions are or aren't is almost all I look for--I don't think I would ever enjoy a good idea despite bad prose, and good prose can sometimes make a bad idea good.
1825848
Hmm...so does this mean you don't care about what happens in a story, or what the characters do? That's apart from the style one writes with, or their quality of prose.
1825805 .
Ah, my brother is reading that book (I believe the passages were from Blood Meridian?), and really loves his writing. I've read a teensy bit and, despite how many "rules" it breaks, I find myself getting absorbed in it.
So. You have writing...and you have story telling. Would you say they're one and the same?
Also: "Clear and Simple as the Truth: Writing Classic Prose" by Francis-Noël Thomas & Mark Tuner states, "A style is defined by its conceptual [my emphasis] stand on truth, presentation, writer, reader, thought, language, and their relationships." They make the distinction that style is a set of ideas, not a set of skills or simple surface features. Technically, the book is about nonfiction writing, but it informed a lot of my current feelings regarding prose.
I only understand:
ABout 10% of what you write in these blogs, but that 10% always makes me think. All this "dialogue tag" stuff is firmly in the 90% that I don't see the point of, but I'll try to help out with some data.
Mike
Year Author Title 1a 1b 2a 2b 3 4
1833 Edward Bulwer-Lytton Godolphin 14 4 1 1 1 9
1846 Edgar Allan Poe "The Cask of Amontillado" 13 0 2 0 0 15
1826031
Yes. Poor prose has ruined several stories for me that I really wanted to like.
I don't think I have ever encountered a bad idea that was carried by good prose. I have, however, encountered plenty of stories that were elevated from "meh" to awesome because they were executed well. A story doesn't need to be groundbreaking in order to be cute, exciting, or touching.
True, style is more than just technical skills, but a lot of the author's soul DOES come out on the "surface" level, the level of descriptions and dialogue and sentence structures.
Okay. I might as well. I'm not as opposed to untagged speech as you seem to be, so long as A: the conversation has only two participants and B: it's short enough. It just so happens I have two temporally disparate works of literature on hand with me!
Year Translated Author Title 1a 1b 2a 2b 3 4
1880 1912 Fyodor Dostoevsky The Brothers Karamazov 17 5 1 2 3 2
1961 N/A Joseph Heller Catch-22 19 3 0 2 4 2
Bonus trivia: It took six chapters of The Brothers Karamazov for me to find thirty quotes for this experiment. Catch-22? Twelve pages.
Also, I got a little confused as to how to handle 2a and 2b. If there are two or more untagged quotes following a tagged quote in a single paragraph, does it just count as one instance of 2a or 2b? I assumed so.
There's something to that, but it might be too general to be useful.
Author, writer, & storyteller have different connotations, in decreasing order of pretentiousness and sophistication. A storyteller can be oral, because the exact words can be changed around--a storyteller is more improv, a performer, more interactive. A storyteller gives the impression about being less concerned with style, but real storytellers practice their stories over and over again. They may tell every story in the same style, or choose only stories that fit their style.
1825506 That's a good idea, but I'm still interested in the changes over centuries. What if it changed back and forth a couple of times?
So far I see no pattern.
Oh, and:
People have actually been writing things that a modern reader would recognize as a novel since at least the big "B.C./A.D." switchover two thousand years ago. Bryan Reardon, one of my professors when I was studying Latin and Greek, put together a collection of the few surviving "Greek novels," so I've read a couple of them in the original Greek and in translation.
The salient point to the discussion here, though, is that ancient written Greek had no punctuation marks--they didn't even put spaces between the words. So using a dialogue tag like "ait" was definitely required to tell the reader that what followed was the character speaking rather than the writer.
Mike
Interesting dilemma. But even you, in your imaginary quarrel with fimfiction's
, omitted speech tags in some part of your dialogue. Take this excerpt, for instance:
collectivehive mindWould it really add to the dialogue if you added speech tags when the horse and fimfic exchange one-word arguments? I believe it would cloud the narrative, perhaps destroying the fast pace that this particular part of the dialogue probably should have.
I was always under the assumptions that speech tags
are unnecessaryaren't welcome when it is perfectly clear who is speaking at the moment. Especially if all you added was (s)he said. The body language or more vivid descriptions (type 2 and 3, if I understand those correctly) obviously don't fall into this category, but wouldn't adding them to a one-word gun shots (like in the excerpt above) put the reader off?1826603
1826129
Well, they do spend the next fifty pages expanding on that single sentence, so yeah haha it's totally too general.
Anyway, thanks for answering my questions you two. I was going to ask more, but I need to retreat into my mind and think a lot more about these issues: writing and style, how much they contribute to the quality of a story, is there something deeper than the prose we can grapple with when addressing a story's woes, etc.
I'll add my two cents. I really think you're over thinking it Bad Horse. Not to be rude, but to me, when and what type of speech tag to use depends on what most things in writing and life depend on: context.
Evening. A little contribution:
Year; Translated. Title Author 1a..1b..2a..2b..3..4
1930; 1965. The Eye — Vladimir Nabokov. Whole novel sifted.
Entire novel: 1a: 66,1b: 16, 2a: 57, 2b: 7, 3: 18, 4: 7
Thirty (30) first points of data: 1a: 14 1b: 0 2a: 12 2b: 0 3: 2 4: 2
1774; 1779. The Sorrows of Young Werther — Johann Wolfgang von Goethe.
The thirty (30) first points of data included, sequestered, of course, with auxiliary five points of data.
No qoutationmarks to be found, but the few lines of prose, that clearly infers the use of qoutation marks, had them subjoined; their meaning and use remains intact.
I can hardly see the difference in adapting un-qouted mono-/dialog into qouted ones, especially pellucid specimen, where their utility is defined beoynd doubt. Even though I cutton on the reason that misconstruction, and other lapses, will threaten the integrity of the final data(statistical result; or what it's called): I thought that the data might interest you.
The sorrows of Young Werther isn't fun to read, it's down right horrible, for the reader, due to the structure of its prose (Yes, I know it's a corpus of letters). But that's another issue.
A feeling, I've got, tells me that you'll expunge this perticular set of data. Don't do me much, it was a fun exercise in either case.
Adequate. 1a: 10, 1b: 3, 2a: 11, 2b: 2, 3: 2, 4: 2.
Extended: 1a: 11, 2a: 15.
Going to add nothing exhilarating, and to add to the verbosity, about some notions I garnered whilst doing this. Mainly from the eye by Vladimir Nabokov.
Point four is for speed and only dual conversations, where the agents/speakers are clear to the reader (when doubt is close to nil) when one speaks and the other is silent. Nothing much but still.
Further, it hurries the narrative along to get into the 'meat' of it.
That's cheap writing, when the author feel she has to hurry the narrative along for something interesting to happens. Close to claiming that readers needs 'flashing' imagery to keep their attention fixed where it matters.
But not only that, as it also, for me at least, adds a speeding feeling of flow by scattering some readability.
The third one seems to act as variability and, again, to gain the affect of speed and, obviously, animate the characters for the reader.
Hm, I wonder how a narrative, written mainly in telling prose, would yield in terms of data? And vice versa.
1826733 I did that deliberately. Omitting speech tags sounds good to me, but I'm suspicious that I might just be conditioned to it.
1827813 My goal is to see whether use of them changed over time, to get an idea whether this aspect of style is just fashion.
What's the difference between the "adequate" and "extended" data?
1828117
The first: it's the number you've asked for.
It's a given that I've got the worst use of terminology, ever. I hope that the data is enough.
I could stalk around for novels in my misshapen and a towering book-labyrinth, if you wish. It ended being a clean and ordered library years ago. Aaaa... those were the days.
To stalk and use my primal senses to find those grand, flat, but unmoving, preys! Oh! It makes my heart almost beat!
*Turns around in the swivelling chair, with a coil of my arms I push myself towards the bed. The clattering of plastic and iron-rimmed wheels can be heard. Ducks, rummage beneath its murky realm. Aha! A two-handed sword of finest quality and temper! Expectation brimming in my eyes, I turn around, quick steps, and confidently hovers my hands just above the black keyboard and...*
I'm armed! Now the little, nasty critters will not get me this time! Silver! A-waaay!
Year Author Title 1a— 1b— 2a— 2b— 3— 4
1818 Mary Shelley Frankenstein 14— 8— 1— 0— 1— 6
1847 Charlotte Bronte Jane Eyre 10— 4— 3— 4— 0— 9
1852 Harriet Beecher Stowe Uncle Tom's Cabin 19— 0— 1— 1— 3— 6
1868 Louisa May Alcott Little Women 11— 9— 2— 1— 1— 6
1887 Arthur Conan Doyle A Study in Scarlet 10— 0— 8— 2— 1— 9
1938 P.G. Wodehouse Code of the Woosters 0— 3— 3— 0— 3— 21
1959 Walter M. Miller Jr. A Canticle for Leibowitz 6— 11— 0— 3— 5— 5
1961 Robert Heinlein Stranger in a Strange Land 3— 1— 0— 3— 7— 16
1965 Frank Herbert Dune 7— 1— 6— 0— 11— 5
1969 Kurt Vonnegut Slaughterhouse-Five 13— 1— 6— 0— 1— 9
1851— N/A— Herman Melville— Moby Dick— 30: Quotations followed by an em-dash and the name of the person or work quoted.
1828109
I wouldn't call it "fashion":
I'd call it "adapting to new technologies" or maybe even "the law of unintended consequences." Once someone invents quotation marks and paragraph spacing, you don't need the speech tags all the time. Just like the way people have stopped wearing wrist watches because their cell phones tell them the time now....
Mike
1825495
>I think using speech tags is objectively right, because it makes the story objectively easier to understand.
Do you consider third person pronouns to be "objectively wrong"?
So do sentences like "This is an important plot point" and "This made Harry feel sad" and "Desultory means lacking enthusiasm". Is ease of understanding necessarily a desirable quality? (See also: Ulysses)
1829465 That's a perfectly valid point. What I really mean is that speech tags make it easier for computer programs to parse the story. That's a different, simpler, syntactic level of understanding.
1829248 Why would I? But I do consider pronouns that don't resolve unambiguously to be wrong. See my reply just above to Prof. Whooves.
1829041 Thanks! That's great!
Re. Moby Dick, you were reading the quotations in the prologue. The rest of the book is (more) normal. That raises a point:
should probably be 1a, because it clearly associates a person with the quotation.
1828238 So what is "extended"? Does that mean what you found when you kept on going after the first 30?
1829681
That I don't think is arguable for the kind of parsing you probably want a computer to do, but do speech tags always make it easier for a human to understand? Maybe in the sense where story is a sequence of objective events, but it seems to break down when the story is a first person perspective within the sequence of events.
When I'm talking to someone in person, I don't (always) think "he said" after he's done speaking. In those cases, speech tags are things that are just thrown on top of the actual story so that the reader can understand the plot. It's intrusive, but maybe it's necessary or difficult to do without. I think of it more as a kludge than a story element.
1830184 I don't think of it as a story element, but a style element that writers argue about a lot.
1830228
Right, I'm getting lost separating "telling a story" from "telling a story". Carry on.
1829681
Maybe it's just:
That I'm a humanities major, but don't see the value in making stories easier for computer programs to parse. Is it an exercise in trying to get a computer to work the same way that a human brain does? And how is this different from trying to make a pebble work the same way an apple seed does?
Mike, Wondering
1829716
Yes.
1830277 Writing a program that can parse English reveals how English really works, particularly when you then try to use those parses to reason with. This, I think, is a better guide to what its grammar really is than the intuitions of grammarians. Because I'm trying to separate those elements of syntax and grammar that made English work from those which are mere accidental fashion, a good criterion is that things that help the parsing are functional, while things that don't, aren't.
Quick question first, which I'm surprised nobody else has asked:
What makes this a "1a" when it's identical to your 2a example?
1831920 The semi-colon. Type 2 has a sentence entirely within quotes.