The author has maintained that among the things that cognition requires are: non-derived content, scientifically tractable and non-motley processes (Adams and Aizawa 2001; 2008a; 2008b), and the capacity to figure in agent-centered reasons that explain purposive behavior (Adams and Garrison 2003). So what will be discussed here is what someone who accepted these considerations about the mark of the cognitive would require for extended knowledge. Of course, cognition could extend without knowledge. Just as contemporary skeptics might be right (not that the present author thinks they are) and we might lack non-extended knowledge, even if cognition extends into the environment that alone wouldn’t mean that knowledge extends. Yet, if cognition were to extend, what else would be required for extended cognition to yield extended knowledge? Attention will also be given to Gricean thought and processes, procedural thought, mirroring, and we-intentions.