Menary's objection here seems to me to be a bit compressed. The idea, as I've articulated it before, is that when Otto uses his notebook, there are by A&A's lights non-derived representations in Otto's brain. So Otto + notebook meets the A&A condition that cognition must involve non-derived representations.Weak cognitivism: cognition involves the processing of representations.Note that even the strong version is ambiguous, because it allows that there may be some underived representations that are processed, for example, in an extended process that also involves conventional representations (for example see Menary 2006 and Clark 2010). In which case there really is no case left to answer.
Strong cognitivism: all cognition involves the processing of representations with underived content.
So, I think that Clark and Menary are right to note that "involve" is too weak. But, the straightforward strengthening that gets at what we are about is to require that the vehicles of content bear non-derived content. But, the vehicles of content in Otto's notebook are paradigms bearers of derived content. So, the challenge remains.