Some stuff on writing dialogue. This is long, but there is a TLDR at the end.
Natural Is Relative
There is a very wide range of speaking types. Some people use ‘like’, like, every other word. Some people make formal speeches that the entire world watches. Some people have speaking disorders, like stuttering or motor issues that affect the mouth, or they’re missing mouth structures like teeth or a tongue. Some people are speaking in a foreign language that they aren’t 100% fluent in, that has tricky grammatical rules, that has its own dialects with different rules, or that has sounds that the speaker can’t pronounce. Some people are robots. Lots of different types of people, all with different ways of speaking.
And the same people might use different speech in different circumstances. A loudmouth with informal speech might be very formal and respectful when they’re talking with their boss. Someone who makes speeches for a living might relax when they’re out with their friends. If someone is tired, highly emotional, or drunk, they might get less formal than usual. Some might refuse to change their speech patterns no matter what.
But the point is, it’s perfectly natural for some people to be stiff and formal, or for some to be that way in some situations. But if everyone speaks the exact same way and does so in all situations, that’s where things start getting unnatural.
People of all types will often experience issues, like not being able to think of a word, mispronouncing a word, forgetting what they were going to say and stuttering, etc. These are called dysfluencies in the transcription/captioning world, or at least the corner of it that I was in.
Perfectly natural dialogue (based on something I’ve heard before) might look like this:
“So, um, you see, uh, we’re looking at the, um, the, um, the… the wall c***. I mean! The wall clock. (snickers) C***. So, uh. Yeah. The clock. On the wall. The wall-- wall clock. We’re looking at the wall clock.”
It’s not very fun to read. If you’re writing fiction, or copying speech down, you’d probably cut down on the disfluencies (stuttering, filler words, repetition, and so on) unless they’re important. In terms of copying, that is perfectly acceptable, unless the person who requested it wants every single word recorded faithfully. Our brains even cut these things out as well, unless our conscious mind brings attention to them and makes the brain think they’re important.
Incidentally, news articles that quote people will sometimes adjust the filtering based on how much they like the person or their position. If they want to discredit the person or make them look bad, they’ll keep in more disfluencies, while if they want to make them look good, they’ll go heavy on the filtering to make their speech look smooth and refined. A very subtle way of altering people’s opinions on things.
Anyway, the long and short of it is that most dialogue in fiction is unrealistic to some degree, and that’s not necessarily a bad thing. But it also means that you can make your dialogue instantly more realistic by adding in a few dysfluencies. Sprinkle in filler words, have the characters forget words or what they’re going to say, have them mispronounce words, make their sentences incomplete, and so on. Additionally, they can interrupt each other, speak over each other, mishear each other, and not listen to each other.
As for how much of this you should do, I’d suggest doing it relative to the narration. Natural is relative, and dialogue in a book is relative to the narration. For first person or a very close third, the narration and the viewpoint character should be exactly the same as the baseline. After all, first person is that character talking to the reader, more or less, and very close third persons are similar. For less-close third person, the baseline for dialogue can be a notch or two more dysfluent than the narration. From that baseline, you can reduce them for more formal situations and dialogue and increase them for less formal.
Adding emphasis to certain words can change how a sentence reads by making it the focus of the sentence.
“There’s a fly in my soup! How disgusting!”
“There’s a fly in my soup! I wanted it in the salad!”
This is especially useful when you’re using emotive descriptions.
“I said,” I snarled, “that I wanted meat in my soup, not flies.”
To figure out where to use emphasis, say the words. In your head or out loud, whichever works best for you. If the sentence changes meaning when you use emphasis or extra emotion on one word, add in emphasis on that word. If not, leave it out.
Dialogue As Foreign Language
When you’re talking about someone who is in the process of learning the language, typically there are three main differences from a native speaker: their speech works on the logic of their first language, they overcorrect, or they are too formal.
If you think about the stereotypical Russian accent, one example would be “get in car” rather than “get in the car”. This is because Russian doesn’t have the definite article; that is a literal translation of what the sentence would be in Russian. Similarly, if you’re working fast food and a Spanish-speaker wants to get it to go, they will often say “for to go”, because that’s a literal translation of what you would say in Spanish.
Alternatively, they might never have learned certain sounds. The “th” sound (dental fricative) is actually only found in a few languages, so most people who didn’t grow up with those languages wouldn’t know how to pronounce it. This is why so many people say ‘de’ instead of ‘the’. Similarly, Hindi is hard to learn for English speakers because most of us can’t distinguish between aspirated and unaspirated sounds, and Mandarin and Cantonese are difficult because we’re not used to the tones.
If someone learns that they’re making a mistake in a language, they may switch to overcorrecting themselves. For instance, a Russian who learns about the definite article might begin using it in places where an English speaker wouldn’t.
And people who learn a language from a book or a formal class from their own language will probably be more formal in speech. I learned Spanish. Part of that includes learning the word para. Puerto Ricans, however, use a cut verson, pa’, instead. If I went to Puerto Rico and used para, because that’s what I was taught was correct, my speech might sound overly formal. In English, “should have” is correct, but in speech, we cut off the word ‘have’, so we say the phrase more like “shoulda”. Someone who learned it from a book might not know the difference in pronunciation and they might teach others the same thing.
So, in order to write someone who isn’t 100% fluent, it helps to know at least a little about the grammar of their original language. If you don’t, then going with something that’s a bit stiff and formal will also work, provided their background allows for them to take those kinds of lessons.
Side note: please try not to go overboard with phonetic speech. It gets old really fast and it can be difficult to understand. If the point is that the narrator can’t understand it, they’ll probably learn to understand it after a few minutes so you can drop it. You can also just say that someone has a particular accent or that they’re from a particular region and readers can look up videos on it if they’re unfamiliar.
In general, the preferred way of talking about someone signing is the same way you’d talk about anyone else speaking. “In quotations, like this,” I sign, “but you can also use quotations with italics if you need to distinguish between speaking and signing.” Italics without quotations are also used by signing authors, but it can be a problem since italics are often used to express thoughts, so it means you need to tag every single line of dialogue to make it work. For tags, you can use say/said instead of using signed all the time.
Body language is important. When you’d describe a signing person’s tone, like saying that someone is raging, you’d describe how someone is signing by talking about what their arms are doing. Think about someone’s body language - especially hand or arm gestures - during different emotions and carry that over to signs. When someone is angry, their gestures become more aggressive and jerky, and so would their signs. Someone sad or depressed or tired might make halfhearted signs. You can also carry this over to showing personalities: a boisterous person takes up more space with large signs and a shy person signs in a way that’s closer to their body, smaller and more reserved.
Also, sign languages have dialects, as well. There are different ways of signing based on geographical location. There are also multiple sign languages; North Americans, Brits, and Australians can all speak English together, but they speak ASL, BSL, and Auslang, respectively, and can’t understand each other when signing.
I’ll leave any further advice - like on sign language grammar or specific community quirks - to people who actually speak a sign language, but here’s an article discussing the matter from an author who is Hard of Hearing.
One major problem is something called Talking Head Syndrome, in which the only text is the two people talking. It’s named that because it gives the impression that the only thing that exists in that moment are talking heads floating in a void. I am very guilty of this. Still, adding in surroundings and how they affect the conversation not only gets rid of Talking Head Syndrome, but it can also make the conversation more interesting and natural.
Sometimes, it helps to have the person doing something while talking. For instance, if they’re eating with a fork, they can wave it around while talking and stab it for emphasis. They can also interrupt the conversation to chat about the food, like asking someone to pass the salt. If they’re out walking, they might step on a rock, causing them to yelp and get distracted. Or, if they’re engrossed in the conversation, they might walk into something or someone else.
Similarly, other people can affect the conversation. If they’re in a park with kids, a kid might suddenly shriek, causing a few words to be missed in a phone dialogue, or the person has to stop and repeat themself. Or they pause to thank the waiter or give orders. Maybe they have to lower their voices or stop talking if someone else is nearby and they’re talking about a sensitive issue. An emotional outburst might be heard by other people, causing those people to become concerned or the speaker to become embarassed.
You can also use actions to show emotions rather than having people say things adverbally. Someone puts their hands on their hips, cringes, starts wiping their eyes, laughs, and so on.
- Have people vary their tones based on who they are and what they’re doing
- Add in filler words, stuttering, repetitions, interruptions, and so on, to about or slightly more than what the narration has
- Put emphasis on words if it changes the meaning or emotion of the sentence to what you want
- People learning your language will make mistakes that are based on their original language, or else they will be overly formal
- Treat sign language like another language, but put emphasis on body language rather than tone
- Think about what’s around the person or what they’re doing and how that can affect the conversation