A Woman, Facing Away
The video opens with her back to the camera. She rocks at a steady pace, humming under her breath. Her fingers glide along a keyboard without pressing a single key. She twists a drawer knob back and forth, again and again. At one point she slaps a sheet of paper against a window, and later presses her face into an open book, as if trying to read it through her skin. No subtitles, no music, no one explaining what any of it means.
Watching this cold, the urge to look away isn’t a strange reaction. Most of us have simply never learned to read this as anything other than distress.
Why Would Anyone Post This?
Who records themselves like this on purpose, then uploads it for strangers? To eyes trained on “normal” behavior, footage like this looks like a private moment that should have stayed private, maybe even a malfunction. There are really only two explanations: either this is raw footage of distress with no context attached, or we’re watching something and simply lack the vocabulary to read it.
The clip is titled “In My Language.” That alone hints that more thought went into it than the opening footage lets on.
The Translation
Partway through, a title card reads “A Translation,” and the whole video changes direction. A synthesized voice, produced through text-to-speech software, starts narrating, explaining each gesture from the first half one by one. The rocking. The humming. The paper against the glass. Every one of them, she says, was a form of sensory engagement, a way of staying in ongoing exchange with the space around her. Touching, tasting, smelling: together they form a “constant conversation,” and that conversation is her “native language.”
The woman is Amanda Baggs, 26 years old at the time, a non-speaking autistic woman who would later go by the name Mel Baggs. She posted the video to YouTube in January 2007. Nothing about the footage itself changed between the first half and the second. What changed was our access to what it meant.
Whose Silence Is the Problem?
The argument underneath the video is sharper than “I have an inner life, please be kind.” Baggs is challenging an assumption most communication research never bothers to question: that spoken or typed language is the only real language, and anything else counts as a deficit waiting to be fixed.
Then she turns it around. Her inability to speak is treated as her failure, something to diagnose and correct. Meanwhile, everyone else’s inability to understand her way of communicating is treated as unremarkable, nobody’s problem at all. That double standard, over who has to do the adapting, is the real argument sitting inside a video that opens looking like nothing more than a few minutes of repetitive motion.
One Video, Reconsidered
The clip did not stay small. Within a month, according to a CBC interview from that period, it had already been watched nearly 700,000 times. By the end of February of that year, CNN ran a story on her, and Baggs went on to guest-blog for Anderson Cooper’s show, whose medical correspondent, Dr. Sanjay Gupta, later visited her at her home in Vermont. Wired covered the video too, framing it as part of a broader rethink of autistic cognition. Eventually the view count passed 1 million, and Baggs went on to speak at disability conferences and work alongside autism researchers at MIT.
That’s a real ripple effect, and it’s worth being precise about its shape. No single study got overturned because of this video, and no named scientist is on record saying it changed their mind. What actually happened is that one piece of media reached a huge audience and pulled journalists, advocates, and researchers into the same conversation. That’s smaller, and more honest, than calling it a settled scientific reversal, and the difference is worth keeping straight.
What We Still Don’t Know
Some parts of the story were never filled in. How Baggs actually composed the narration, how much was written ahead of time versus assembled on the spot, exactly how the text-to-speech process worked behind the scenes: none of that made it into the record.
What’s left is a bigger kind of not-knowing. Most of us have never had to question what counts as language, because we’ve never needed anything other than the standard kind. The next time someone’s behavior looks like noise with no signal, it might be worth asking whether the signal is there and you just don’t have the receiver for it yet.