How to present dialogue spoken with text-to-speech software?

paradoxikay

everything happens so much
Registered
Joined
Apr 10, 2018
Messages
42
Reaction score
7
Location
Los Angeles, but not the cool part
I have a character who doesn't speak, and instead uses an iPad with text-to-speech software. Most of the time their dialogue sounds natural, but when they're stressed (or messing around) they start dropping punctuation and using chat speak... which doesn't come through the text-to-speech process well. Sometimes they sarcastically use emojis, which really don't come through well.

Does their dialogue belong in quotation marks or italics? And how do I present the dialogue that doesn't sound anything like natural speech?

For example, say they type something like:

lol sure u are šŸ¤”
yr not even trying rn

What they mean is:

They laughed derisively, rubbing their chin in mock contemplation. "Sure you are." A pause. "You're not even trying right now."

But text-to-speech would turn it into something like:

"Lol sure you are thinking face. Y-R not even trying R-N."

Do I write the dialogue as it was typed, or as it was "spoken" by the iPad? How do I do this without annoying the reader? (I mean, I guess one answer is "have Matt type full, grammatically correct sentences at all times", but I feel like I'd lose a lot of potential nuance that way. They don't have tone of voice to demonstrate that they're being short with someone or that they're super upset, after all.)
 

Bufty

Where have the last ten years gone?
Kind Benefactor
Super Member
Registered
Joined
May 9, 2005
Messages
16,768
Reaction score
4,663
Location
Scotland
I obviously don't know your story or who your intended audience are.

That said - No idea, really, but I would prefer to read normal dialogue (with careful word choice and sentence construction) with the knowledge, from the start, that it was effected through text-to-speech.

The emojis would help and I don't think they would be intrusive at all. Indeed, if there was an emoji I would know it was the text-to speak.

I suspect reader-friendliness would be the main issue with the pure phonetic or visible route.

Emotion can be covered via the POV - no?

Good luck. Others may well have had experience of this issue.

Kindest.

Bufty :Hug2:
 
Last edited:

Frenzy3

Registered
Joined
Apr 8, 2018
Messages
30
Reaction score
1
Just the same as regular dialogue, maybe add VC for voice control or TTS. I.E. "Hello. TTC"
 

Maryn

Sees All
Staff member
Super Moderator
Moderator
Kind Benefactor
Super Member
Registered
Joined
Feb 12, 2005
Messages
55,447
Reaction score
25,470
Location
Snow Cave
The Chicago Manual of Style, used by many fiction publishers even though it's not intended for fiction, has rules for everything, including synthesized speech. I don't have it at hand, but as I remember it, such speech appears in italic font, but inside non-italicized quotation marks, showing it to be speech.

Allow me to note that if you select one style, whether it's that or something else, and are 100% consistent with it, that should be fine. Publishing houses may have their own style manuals they follow and would change it to match no matter how the author chose to present it.
 

Hbooks

Super Member
Registered
Joined
Jun 18, 2011
Messages
558
Reaction score
72
I would italicize, and type it the first way, as the character would have entered it, leaving it to the reader to interpret just like a text message. You work into the text the first time that this is how the character communicates. If the publisher changes it later, so be it. You can put the action tags separately from the italicized speech. Whose pov is it from?

TTS can get confusing the way it garbles things it doesn't understand. Probably your first priority should be clarity. If the reader has to decode:

"Lol sure you are thinking face. Y-R not even trying R-N."

it's going to get frustrating quickly.
 

paradoxikay

everything happens so much
Registered
Joined
Apr 10, 2018
Messages
42
Reaction score
7
Location
Los Angeles, but not the cool part
The POV character meets Matt early in the story and quickly gets used to them using the iPad as their voice... that's the best way to establish the concept, actually, isn't it? Have her describe how it's synthesized speech the first time, because it catches her off guard, and then the reader gets used to it with her. And that way I can write the dialogue as typed, and Tristan's momentary confusion ("I'd never heard anyone say an emoji before, but Matt did just that") clarifies exactly what it ends up sounding like?