Knowledge management has changed the way we look at knowledge in the current economy; it is a key factor in an enterprise's success or failure. In contrast to what we as engineers typically love, KM puts people first organizational issues second, and technology third. Seriously considering these issues to produce a successful KM system leads to at least three requirements. We must: encourage employees to participate; integrate KM with current organizational practice; and provide the natural tools such that people can easily recognize the benefits, align with current organizational practices, and use the system. The natural choice of substance for such a KM system is human language, and the required tools are based on human language understanding. As we all know, comprehensive human language understanding is out of reach for the foreseeable future. Nevertheless, although the knowledge system's substance is language and complete human language understanding is out of reach, the system need not be restricted to text nor the tools restricted to a keyword-based search. Human language technology can and should do better. The article elaborates on the interaction between human language technology and KM to achieve this goal.