Great questions! Here’s at least one thought: perhaps heuristics are limited to cognitive processes. Very roughly, cognitive process is a process that takes mental representations as its inputs (and, more controversially, takes also mental representations as its outputs.) This would mean, for instance, that rocks couldn’t be rational, since they don’t have mental representations and hence don’t have cognitive processes. Dogs can be rational, and, as you suggested, might be less rational than humans in ways that the ecological rationality crowd can readily accommodate. For instance, one mark of this is that our cognitive processes are successful in a wider variety of environments, or are faster and more frugal for a wide variety of cognitive tasks.

Best,

KK

This seems rooted in how we interpret Schroedinger’s original thought experiment, proposing the idea that a state outside of simply dead vs. alive exists. Suppose we agree that this state exists, in which case we could model three states of being: alive, dead, and neither. Then “The cat is alive” would be false until the box is opened. Here, the three states can be described as A, D, and (~A & ~D). This of course throws out the window the logical idea that alive = not dead and vice versa.

If we model the three states as alive, dead, or both, “The cat is alive” and “The cat is dead” are both true until the box is opened. Now our three states are A, D, and (A & D).

Now to evaluate “The cat is not alive,” in the first model it may be dead or neither. In the second model, the only option in which A does not have to be true is D, so the cat is dead. While the box is closed, the statement “The cat is not alive” is then true in the first model, since ~A is part of the box-closed state (neither). In the second model, the statement is false since the cat must be alive when the box is closed. However, the statement “The cat is alive” is false in the first model and true in the second when the box is closed, so neither model includes both the statement and negation in the box-closed state. This makes sense since both of my models involve abstracting the idea of “alive” from “dead” and separating the two concepts. ]]>