Writing with Sound

I sense that there is a change happening in my writing, because there is a change in how I write. I used to write, like most people, silently. My eyes and fingers worked together to lay down words on the page.

When you compose with your eyes, you read over what you’ve written with your internal voice. You supply all of the missing elements of speech: tone, emphasis, pause, and all the other things that add texture and life to the words on the page.

When I used to write with my eyes, I would become so familiar with the way I read a piece of text in my head that I could not imagine it any other way. I never considered that those invisible auditory elements would not be immediately evident to any other reader. At least, not until I came back to a piece of writing a few days later, when I had forgotten the finer details, the shape and flow of each individual sentence. Then I would inevitably puzzle over a phrase or paragraph until I realized how I had meant it to be read. I found these pitfalls lurking in the writing of others, too—places where the emphasis or tone was important but not apparent, waiting like rocks in tall grass to trip up unsuspecting readers.

Now I write with VoiceOver. It is Apple’s main accessibility feature, which gives immediate audio feedback for every character, word and sentence that I write, and reads me paragraphs so I can remind myself of their flow and argument. It is a multi-sensory experience, since I still see the words appearing on my screen, sometimes zoomed in close enough to read them but most often not. I used to find the constant wash of letters and words distracting and intrusive, but in the few weeks since I’ve started writing this way exclusively, it has become natural and I feel adrift when it is turned off.

When VoiceOver reads every word, line, sentence, and paragraph, it supplies its own intonation, emphasis, and rhythm. Computer voices today are not the abrasive robot voices of the eighties and nineties. They are getting closer and closer to the sound of natural speech. No one would mistake them for human, but developers are focused on improving their realism and fluency in reading long passages. So the computer adds pauses, shifts its tone, and inflects words up or down based on how it interprets the context. All those choices I used to make in composing, usually without even thinking about it, are now made by the computer. The computer becomes a controlling voice in my writing, because if it reads something awkwardly, I often change the text.

Of course, I don’t always cede to the computer. Sometimes it is obviously wrong (like when it pronounces a homonym inappropriately, like “read” as red instead of reed) or its interpretation is clumsily rule-bound (like when it thinks “Mr.” is the end of a sentence). But if the computer’s reading is misleading, ambiguous, or simply awkward, I will often rewrite it. I just can’t bear to hear the voice stumble over the same passage dozens of times as I write and edit, so I alter it to accommodate the computer’s flow.

The big question is, how will this affect my writing? Will it become more stilted or robotic as I adapt to a computerized interpretation of natural language? I hope not. My hope is that the instant feedback will remind me to consider the auditory dimensions of language more carefully than I did before. VoiceOver, imperfect as it is, is an audience that is present and interactive at every stage of my writing process, letting me know what I’ve written and at least one way it can be read. If it makes a mistake, there’s a chance another reader would also have made that mistake. VoiceOver was developed by people, after all, and those people defined the interpretive choices it makes.

VoiceOver is changing my writing, but how remains to be seen. I guess I’ll just have to wait for you, my human readers, to let me know.