Blog by Sumana Harihareswara, Changeset founder
If You Give A Speech You Care About, Post A Transcript
When I give a keynote address at a conference, I sometimes commission* and post a transcript or a near-transcript afterwards (example). And sometimes I do this for non-keynote speeches (example); this year I decided to get and post a transcript for the recent GitHub OCTO talk I gave, on what it would look like if open source were healthy.
If you give speeches that you're proud of, that you want people to think about and pass along so they can keep influencing people, you should do the same thing. Get transcripts up as soon as possible.
I often don't use slides, which makes presentation easier. When I do, I often include the transcript alongside each slide, the way Dustin Ingram and my spouse Leonard Richardson do.
And here's the sort of thing that enables:
My GitHub OCTO talk: so many people have informally said, upon seeing a link to the transcript of the hourlong talk, "oh thank goodness, a transcript." And one of the followup messages I've received says, "I just read your talk on healthier open source, it was terrific." before sharing a super cool related thing.
And I realized Cory Doctorow might be interested, and I emailed him with a link and a summary. He blogged and tweeted about it soon thereafter, which drew a bunch of readers/viewers. He's a really fast reader so I'm reasonably sure he read the talk rather than listening to it.
My WikiConference USA keynote from 2014, on "Hospitality, Jerks, and What I Learned": a writer for the weekly Wikipedia Signpost news outlet was able to use extensive excerpts in his coverage of my talk, which he loved. This talk is one of the things that Wikimedian colleagues still refer to and remember about me, and the fact that they can easily pass around and quote a transcript is part of why.
My code4lib 2014 keynote, "UX Is A Social Justice Issue": Here I did not post a true transcript, but I had written out my talk as an essay ahead of time so I could immediately post it online. So, several months later, when the editors of code4lib Journal were interested in turning it into a journal article, that was easy to do.
And that journal article has been quoted and cited by scholars! Some researchers cite me while discussing libraries and archives, which makes sense, since code4lib is a conference about the intersection of libraries/archives and tech -- examples include "User Experience Strategies for Every Library. Yes, Even Yours" and "Beyond ADA Compliance: The Library as a Place for All". It looks like Keshab Raj Acharya has cited it in a couple of papers on usability for empowerment and social justice, like here:
The last question in the sampling asks participants to briefly narrate both awkward and good experiences with the biotechnological equipment and devices at their workstation. Because user experience is a social justice and human right issue (Harihareswara, 2015), participants were asked to express their negative and positive experiences with the northern products being used in southern sociocultural and contextual settings.
Further afield, Kyle D. Trott cited it in an article on how to design instructional packets (more info). How neat! I've been able to give people a little useful pebble that they can easily reuse in their scholarly cairns, in ways I seriously did not expect.**
In "Documentation to the People: Building Empathy into Technical Documentation for Digital Archiving", L. Work and H.E. Kelly quote me:
.... Second, in the context of digital work, Harihareswara  contends that the larger technology field "systematically undervalues the jobs and roles that require empathy and has deeply gendered associations with hospitality and empathy." In digital archiving, then, it can be argued that we are prone to downplaying both the human element of our own work and the humanity of our users....
I contend!! But actually I do. I put substantial work into these speeches, and in them I contend in the marketplace of ideas. Making my arguments and observations legible, searchable, quotable, skimmable, and accessible*** means my impact is bigger and more durable. How many people have read Allison Parrish's excellent 2016 speech "Programming is Forgetting: Toward a New Hacker Ethic", and how many would have been influenced by it if it had remained video-only?
So: make or commission transcripts, and post them.
And: if there's a talk you love that you want more people to appreciate, maybe transcribe and post an excerpt.
* I pay Keffy R. M. Kehrli to transcribe my talks. I've also worked with Keffy to prepare a talk -- I was having trouble writing a talk, but I could talk about it, so I recorded myself talking about it, Keffy transcribed it, and I used that as the raw material for the talk script.
** And then there's a citation in "Design and Security Analysis of Improved Identity Management Protocol for 5G/IoT Networks", which, ridiculously, cites me (as footnote #8 in the excerpt below) for (as far as I can gather) mentioning PGP.
Yet it focuses on minimizing degradation in MO and maintenance expanse. It is done by constructing trusted-base with cross-certificate between service provider and MO, and implying principles of PGP (Pretty Good Privacy) [8,9] based on PKI (Public Key Infrastructure) [10,11] enables mutual dependence-base communication, and also able [sic] ID management to service providers.
This is nonsensical, and it makes me think less of the journal that accepted it.
*** The accessibility case for transcripts is so obvious that I wanted to focus on this entry on a different, I think underappreciated, argument.
In my opinion, it would be ideal for conferences to live-caption and provide sign language interpretation for all talks, to help Deaf attendees, people who speak English as a second language, people with ADHD, and so on. And the live captions could then at least be a first draft for a human-edited transcript, which, again ideally, would be commissioned and hosted by the conference alongside the video, with the speaker given the opportunity to add hyperlinks to relevant resources.