Extended senses of “to know”

I know how to drive a car or form the Latin subjunctive, but an embryo knows how to grow arms, build a trillion neuronal connections, etc. The desire or fear with the second sense of know is that it’s just a metaphor, and that there is nothing deserving to be called knowledge at all – but what do we replace it with? A mechanical account would do nicely, because it makes the action just a push or a pull of a structure already set down in advance. But this seems only to shift the goal posts – how would it know to pump energy into the structure now as opposed to at some other time? We usually seek to address this problem by saying the mechanism must be supplemented by determinism or randomness – i.e. the change is either absolutely necessary or happens for no reason at all (though the repeated instances of randomness can forms patterns of probability). We haven’t shifted the goal posts here, but we’ve done something worse: we’ve shifted from a causal account of something to a modal account of it. But modes aren’t causes: to tell me something happened of necessity or with mere probability doesn’t tell me anything about the disposition of the cause: both an intelligent and a non-intelligent being use necessary and probable modalities all the time. A detective isn’t any closer to solving his case if, after figuring out that the couple dies because they were suffocated by gas, that the suffocation happened necessarily. Solving the case requires whether this modal necessity has an intentional cause (like some clever assassin, or negligent handiman) or whether it was just bad luck (wrong place at the wrong time).

And so giving a modal account of a mechanism doesn’t give us any help in unpacking what we mean when we talk about non-conscious or merely physical systems “knowing” what to do. Our own experience is ambiguous, and testifies to a sort of knowledge that is pre-rational and automatic. We have to allow a sense to non-conscious, automatic motives, and these motives are incoherent unless based on real knowledge. It also makes no sense to base these motives on higher levels of rationality, since even the non-rational have them.



  1. Matthew McCormack said,

    March 4, 2014 at 6:19 pm

    Let me be devil’s advocate here: Does a rock ‘know’ to roll down a hill ? I don’t think we would say it does. It is a very simple system of rock, hill and gravity. However, when we get to a much, much more complex system that involves hundreds of thousands of molecules and multiple physical and chemical laws (each one as simple as the rock rolling down the hill), then due to the complexity of the result, our minds interpret the result as the embryo ‘knowing’. Our eyes and minds can not see and keep track of all the many simple chemical and physical interactions, and it appears to us as if the embryo ‘knows’ to develop. When we say the embryo ‘knows’, the word ‘know’ is descriptive of our experience of embryonic development. Our experience of seeing the embryo develop is as if the embryo ‘knows’ the course of development, because we do not see nor comprehend all the many, many simple ‘rock rolling down a hill’ steps involved.

    • March 5, 2014 at 9:13 am

      The conclusion you present an argument for is that acting by knowledge is a name for our ignorance of a process that occurs without knowledge at all. If applied fully, it would lead to Eliminative Materialism, and so suffer the usual reductio ad absurdum of an intentional rejection of the intentional. But the argument makes a shift from the ontological to the epistemological: it tries to replace a claim about nature with a claim about what we can or cannot know about it. This is significant because, even if we are ignorant of how the simple causes add up to the complex ones, we still know that the complex must be (ontologically) either mechanical or not. If mechanical, the argument of the post kicks in; if not, then it seems it agrees with the argument of the post! After all, the whole claim is that calling nature “mechanical” does not resolve the question of whether we can speak of it knowing what it is doing in a robust and not merely fanciful way.

      The mechanical account of nature, or the account given by laws, always presupposes a nature already arranged and so as relating to some things and not others. In scientific jargon, laws presuppose initial conditions. In your own example this moment is suppressed: why is the stone rolling down the hill? If it’s because I’m rolling it down to crush my enemies, then its motion is a clear instrument of an intention; if it’s just slowly eroded to the point where it broke free, then it is also a part of a larger system of total causes, to which the action must be properly attributed. But this system presupposed by law has its own structure, with each of the parts ordered to each other. The “know” question arises here all over again. WE can. of course, chase after this with another law ad infinitum, and I’m sympathetic to Leibniz’s claim that we can keep finding laws forever since nature is a sort of infinite machine. But an infinite machine is just as posterior to intelligence as a finite one.

%d bloggers like this: