Players of video with synchronised audio offer a mode that speaks the available captions aloud, unless the captions cannot be determined programmatically, as with bitmap subtitles.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.