Do you mean that we can do logic habitually, without really concentrating, and that this is what you call "mechanical"? So, for example, the fact that we can argue "if x then y... not y... therefore not x" quite habitually, without really thinking too hard, it can therefore be done by a mechanism. Or perhaps you mean that it can be mechanical, because I have removed the propositions themselves and replaced them with the letters "x" and "y"? You use the word "inevitability" for both mechanical and logical results, but that does not mean that mechanisms and logic are the same thing.
Why do mechanisms and logic have to be "the same thing" before we can accept that mechanisms can do logic? "Wings" and "flight" are not the same thing but planes and birds can still fly.
When I say logic has inevitability I mean that if you are doing logic correctly (which is to say if you are doing logic at all) then your actions are predictable and inevitable. Like a forced move in chess. Obviously a machine could work in this way.
It is not clear that a machine - whether a lever or a computer - employs logic whatsoever. Causality among material things is not the source of logic, nor may logic be reduced to causality -- that at least is my contention.
So one thing can only be a source of another thing if it is the same kind of thing, is this your contention? And the same similarity must exist if one thing is to be reduced to another thing. But this just rules out all emergent or supervenient properties. In the other thread you mention I quoted the following:
"We can say that a high level phenomenon is weakly emergent with respect to a low-level domain when the high-level phenomenon arises from the low-level domain, but truths concerning that phenomenon are unexpected given the principles governing the low-level domain."
In this sense logic would be a weakly emergent phenomenon of certain configurations of the physical world.
In your view it would seem that brains and computers may causally implement illogic just as well as logic; in fact, we humans seem to do illogic much more efficiently than logic. If causality is the source of logic, then logic is no better than illogic.
This doesn't follow. Logic is usually better than illogic for creatures like humans as it better aids our survival. Or helps us satisfy desires that originally evolved to aid our survival.
Again, if the designer does correct logic and therefore the machine does logic... but the designer makes a mistake and the computer does not make the mistake, how is that not special pleading? The machine does what it is designed to do, whether logical or illogical.
I don't understand your point here. In as far as the human is doing logic he is also merely doing what he is "designed" to do by his training. Note, if he is doing correct logic then he is not exercising "free will" either (he is not free to infer that 1 + 1 = 3 while still being correct). Of course there may be rather more than just "logic" involved in programming a computer. But you are not arguing that human creativity or emotions make us different from machines so you can't really use that argument.
If the human restricts himself to doing logic there is an exact parallel between what he is doing and what a computer doing a similar task would be doing. Tiredness may cause him to reason incorrectly just as a fluctuation in voltage may cause a computer logic circuit to malfunction. The computer malfunctions in the sense that it is not correctly implementing its program. The programmer malfunctions in that he is not correctly implementing the specification he has been given.
So what is the ultimate arbiter of correctness? Who says that the specification is correct? Perhaps the writer of the specification was handed a requirement from the marketing department and simply (mechanically, logically) translated this into a set of technical requirements. Perhaps marketing pieced together their requirements by asking customers what they wanted. At each stage correctness is defined as compliance with the requirements set by the previous stage.
Let us suppose that there is some ultimate source for correctness. I think it is clear that this could not be an act of "logical inference". It would be the place where the buck has to stop, it couldn't be the kind of forced decision that a correct logical inference always is, a decision made inevitable by the very premises. If it was we could then question the correctness of the premises and so on, ad infinitum.
Of course I don't think there is or ever could be an absolute source of this sort for correctness or truth. In the case of the programmer the ultimate arbiter of what the machine ought to do, against which errors can be judged, is the profits of the company, driven by the need for everyone involved to make money and the customer to have certain problems solved (perhaps to make
their businesses more profitable by better satisfying the needs of their consumers). We all want our material and other needs taken care of because we are evolved creatures who have survived this long precisely because we give a high priority to such things.
The buck doesn't actually stop anywhere. Evolved creatures have no justification for wanting the things they want. They don't want truth, they just want what they want. "Error" is what is prejudicial to survival. We see the natural world through the filter of evolutionary competition - we only see the survivors which means we see a world populated by those creatures who shun error. This includes humans.
