1. Lawrence Hunter, in his review of Gene Rochlin's Trapped in the Net (Princeton U Press, 1997), notes many examples of how expertly designed information processing and decision system nonetheless fail, often with dramatic and terrible consequences: e.g., programmed trading leading to stock market crashes and the mistaken destruction of an Iranian airline by the USS Vincennes. Consider Hunter's observation: "Highly automated rapid-response systems that depend on real-time interactions between sophisticated computers and expert human operators may have an inherently high probability of error in situations not explicitly accounted for in their design."

What issues and principles of cognitive science does this observation evoke? How do computationalism and flexibility interact? What design issues are raised by the need to respond to new situations? What does this say about the knowability and completeness assumptions of cognitive science? Is cognitive science compromised by real-time computational mind?

2. The Navy just named a new destroyer the USS Hopper, after Admiral Grace Murray Hopper, a computer pioneer. She is responsible for two major ideas in computing:

a. the idea of "bug," which derives from her finding a moth caught in a switch contact and hence disrupting computation. Are all bugs hardware bugs? Are there bugs at different levels? What is the relation between software bugs and hardware bugs (supervenience and functionalism)?

b. programming languages composed of commands rather than mathematical code -- What are the "level issues" raised by this? What is the (dis)advantage of having a computational level that is independent of the hardware level? Is psychology about the mind's assembly language, or is this neuroscience?

3. The Clinton administration has proposed restrictions on the development and sale of encryption technology that could not permit immediate decoding by law enforcement. All proposals for decoding have some kind of electronic "trap door," either built into the chip (the Clipper Chip) or, now, third-party decoding -- that is, the establishment of a set of certified agents armed with the mathematical keys to decode all encryption technology under court order. Consider this problem: the language of thought is Nature's encryption technology. Does it have the same problem of potential inaccessibility to others? That is, can the language of thought always be decoded, intrinsically (mental clipper chip) or extrinsically (third party)? Here's another way to think about it: is the language of thought private or individualistic? If it is genetically given, does this make it different? Do you need a public standard of truth to decode the language of thought?