Greetings
        I would like to present an alternate point of view to the notion that bees and other insects behave like "little robots." This idea has come about as researchers have attempted to understand bees and other animal behaviors using "models". The concept boils down to: if we can create a model that predicts what the organism will do, then we will have understood it.
        The computer model or the robot model is most attractive because we know that computers and robots are capable of extremely complex "behaviors" and *we understand them* -- because we built them. Therefore, if animal behavior can be shown to be mechanistic, we can understand that.
        One easy example where models fail is predicting the weather. Why do they fail? Because weather is a dynamic and essentially unpredictable system. Animal behaviour follows certain patterns, like weather, but it too is dynamic and changeable. To try to reduce it to mechanics or mathematics, is to oversimplify.
        An example in bees is swarming. In a given apiary, where colonies are very similar and are exposed to the same environmental conditions, some hives swarm and some don't. Each colony seems to evaluate the situation and make a decision to swarm or not swarm.
        Of course, it is not known whether there is a "spirit of the hive" which guides the activities of the colony, or if decisions are the result of some sort of "voting". Just as intelligence seems to be more than the sum of neural networks and enzymes, etc.,-- I believe that a colony has an intelligence that is more than just the collective mechanistic response of thirty thousand individual robots. But even if it is just "voting -- we now know only too well that given the same input, different colonies could come up with different results!

Peter Borst