Handing Over Delivery

I teach seven courses this semester at three schools. Filming and editing haven’t fit into that really at all, and I’ve stopped pretending they ever will. The paid HeyGen account, with a digital twin trained on video of me, is waiting on my bank payment to clear because I accidentally paid with ACH instead of my debit card. Until it does, Shaye’s first real video, a seven-minute introduction to the channel, is in script form.

Classical rhetoric splits the making of a speech into five canons: invention, arrangement, style, memory, and delivery. Delivery was the voice and the body: how loud, how fast, where the hands went. James Porter (2009) calls delivery and memory the two neglected canons. Once delivery got tied to speeches and the speaker’s body, writing teachers mostly let it go, and by the twentieth century composition hardly taught it at all (p. 207). It’s not lost on me, a writing teacher, that I just handed delivery to a machine after academia for the last century trained us to skip that canon.

Five Canons, 43 Seconds

The invention process of the video was mine. I dictated a very long brain dump, and Claude and I distilled it, over and over, down to 43 seconds. Arrangement was mostly Claude’s, with my edits. Style—the word choice—came out of that same back-and-forth. Memory went away entirely, since Shaye doesn’t memorize anything, and writing itself largely replaces memory anyway. That last part is an old worry.

Porter was writing about avatars seventeen years ago. Making one in Second Life, he says, is creating “a bodily representation of myself,” one that may or may not match the person who made it (p. 212). Shaye was modeled to look and sound like me, and he pretty much does. But he talks with more energy than I do, and honestly, he’s a better presenter. When he gets excited, he opens his mouth a little wider than I do, too (and I can’t decide if that’s good or bad).

Porter ties a speaker’s face, gestures, and manner of speaking to the speaker’s ethos (p. 212). If Shaye presents better than I do, some of the trust a viewer gives him comes from a body that’s only mostly mine. And better by whose standard? Antonio Byrd argues that language models learned “correct” English from decades of teaching that treated other Englishes as mistakes, and that it shows in what they write (2023, p. 137). He’s writing about text, not faces. But I wonder what taught Shaye that a good presenter opens his mouth that little bit wider.

Which Modes, for Whom, and Why

My students make multimodal projects: slide decks, web pages, videos. If one of them sent an avatar to present, here’s what I’d want for them to have nailed: to know their content, to have thought it through a lot, and to be able to answer questions about the modes they chose and why.

Hopefully most of my fellow adjunct professors who still teach with an ounce of process pedagogy would agree that delivery isn’t always the most important test or even the easiest to grade itself: Invention or ideas, along with the thinking behind the choices made, is what often gets graded. Of course, at the end of the day, the product has to succeed in reaching its intended audience and purpose for it to be successful. For those who prefer the real me over Shaye’s delivery style, I say you’re better off with Shaye most of the time, trust me.

Process Before Product

References

Byrd, A. (2023). Truth-telling: Critical inquiries on LLMs and the corpus texts that train them. Composition Studies, 51(1), 135–142.

Plato. (2008). Phaedrus (B. Jowett, Trans.). Project Gutenberg. https://www.gutenberg.org/ebooks/1636 (Original work published ca. 370 B.C.E.)

Porter, J. E. (2009). Recovering delivery for digital rhetoric. Computers and Composition, 26(4), 207–224. https://doi.org/10.1016/j.compcom.2009.09.004