The principal rationale asserted for this success criterion is that one or more speech recognition systems inappropriately forwards dictated text to Web applications when no control that accepts text input has the keyboard focus. This is an accessibility problem that merits a solution for users of speech recognition systems - especially those with disabilities.

The current proposal seeks to solve this problem by requiring every relevant Web application to include options whereby users can turn off or remap the single character key shortcuts used in the application. The cost of this proposal is borne by Web content (especially application) implementers, and by users of affected speech recognition technology, who must find and change an option in every application that uses a single character key shortcut (where the controls responding to the shortcut are not in focus). A personal needs and preferences profile could in theory be used to simplify the configuration process, but, to date, no interoperable Web standard is available to support this approach.

A more effective and better engineered solution would make use of accessibility APIs (e.g., at the operating system level) by speech recognition tools to ensure that spoken input is not processed as dictation unless a user interface control that accepts text input has the keyboard focus. Proponents of this success criterion have repeatedly asserted (without substantive technical arguments and without the involvement of the various speech technology vendors) that the acknowledged problem cannot be solved by speech recognition tools.

Given the difficulties summarized above with the rationale for this proposal, it should be deferred for further consideration in a forum that engages speech technology developers, ARIA and accessibility API developers, to investigate the available options for solving the underlying accessibility problem. That is, this proposal should not proceed to CR.