Watch out, I might only be playing games with words like "true" and "obvious" ;-)
Since we can make many useful statements about this universe the quicksand we are all standing in (if we indeed are) must be very slow indeed. So what if there are no absolute metrics. That doesn’t seem to matter all that much.
If I count my sheep on their way out of the pen and if I count them again on their way back into the pen I can be reasonably sure that if those numbers match, there are no more sheep on the pasture. I don’t have to look for missing sheep. This seems like a useful statement. I would call it true.
Sure, sometimes it doesn’t work. Tests with other people have shown that I sometimes tend to overlook a sheep. Further tests have shown that I sometimes misremember the number by the time the sheep are coming back. But other than that it seems very robust.
Why not call that statement true? Do postmodernists expect that this would only work here but not in other cultures? And if that were the case couldn’t you just test that, no need to throw your hands up in the air and give up? Do they expect that it might work on Earth but maybe not on Mars or Alpha Centauri? Testing that would also be no problem. It’s at least not impossible, no reason to give up.
I'm not sure post-modernism advocates any kind of 'giving up' -- otherwise no one would bother with all that effort of writing laboriously long and dauntingly opaque essays on why we should all just give up.
Why not call the statement true? Go ahead. The issue is whether we (can) classify truth as an essential quality of a statement, or as a description of a statement from a particular perspective.
To make an analogy that dawkins would surely crucify me for, it's like the difference between the notion of absolute time, and time as described by Einstein's relativity. From a post-modernist view, time/truth is relative, but being relative doesn't mean it doesn't matter, and just as with physical time, there are narrow frames of reference (day to day life on Earth) where the relativity of time/text/truth is... irrelevant.
God, it really is impossible to advocate for post-modernism without writing in that multiple meaning style.
How could perspective even begin to change whether or not counting sheep works?
I have no trouble imagining universes in which, for various reasons, counting sheep doesn’t work. But that’s not what you seem to allude to. In my imagined universes counting sheep wouldn’t work because of properties of those universes (different arithmetic, different physics, different results of natural selection).
Postmodernism seems to suggest that a mere change in brain state (i.e. change in perspective) can have an impact as to whether counting sheep works or not. That seems … unlikely to me.
Consider the similar task of counting humans. The American constitution originally considered Native Americans as equivalent to 3/5 of a person. We now count them as a full person. What changed other than perspective/brain state?
Even with sheep, consider the possibility that out of say 10 of the sheep you counted, one or all of those sheep are pregnant. Not only that, but you don't know which are or may be pregnant, or how many sheep they may be pregnant with. Does an unborn sheep even count as a full sheep? And yet you counted 10 sheep, and if counting sheep works without changing perspective, then there are 10 sheep regardless. I'm not sure I would agree with that.
If you change your definitions you will very likely end up with different results.
I, as someone who hates arguing about definitions, will be the first to agree with you that there exists no “true” definition of sheep or, for that matter, humans. Definitions are just lists of properties and conditional statements to which we assign certain words. There is nothing true and nothing magic about them. They help us communicating and that’s that.
My statement about sheep is useful (true?) only if all the concepts I use in my statement are properly defined. I can’t just start to count pregnant sheep differently because then my statement isn’t useful (true?) anymore. Counting pregnant sheep differently wouldn’t help me tell whether or not all sheep are back in the pen at the end of the day and that’s the whole point of the statement. I’m not making a moral statement about whether or not unborn sheep count. That has got nothing to do with my statement and is very much beside the point.
It’s not that surprising, actually. If you change a statement which has proven to make good predictions by redefining one of its concepts, you will likely break the statement. It won’t make good predictions any more.
Counting pregnant sheep differently wouldn’t help me tell whether or not all sheep are back in the pen at the end of the day and that’s the whole point of the statement. I’m not making a moral statement about whether or not unborn sheep count. That has got nothing to do with my statement and is very much beside the point.
Whether it's beside the point depends on your perspective :). I spent a few days working on an actual sheep farm in New Zealand, and whether or not your sheep are pregnant is actually a highly relevant fact. Whether or not a sheep that starts out pregnant is still pregnant at the end of the day is also highly relevant. I wasn't trying to say that unborn sheep have anything to do with 'morality', but they are very much not beside the point.
It’s not that surprising, actually. If you change a statement which has proven to make good predictions by redefining one of its concepts, you will likely break the statement. It won’t make good predictions any more.
This is actually one of the salient points of post-modernist thought, with the caveat that post-modernism argues that all statements are always changing, 'definitions' of concepts are always changing. Take our sheep debate for example -- you make statements that are useful to you, and then I respond by interpreting the statements so that they are useful to me, and vice versa.
Since we can make many useful statements about this universe the quicksand we are all standing in (if we indeed are) must be very slow indeed. So what if there are no absolute metrics. That doesn’t seem to matter all that much.
If I count my sheep on their way out of the pen and if I count them again on their way back into the pen I can be reasonably sure that if those numbers match, there are no more sheep on the pasture. I don’t have to look for missing sheep. This seems like a useful statement. I would call it true.
Sure, sometimes it doesn’t work. Tests with other people have shown that I sometimes tend to overlook a sheep. Further tests have shown that I sometimes misremember the number by the time the sheep are coming back. But other than that it seems very robust.
Why not call that statement true? Do postmodernists expect that this would only work here but not in other cultures? And if that were the case couldn’t you just test that, no need to throw your hands up in the air and give up? Do they expect that it might work on Earth but maybe not on Mars or Alpha Centauri? Testing that would also be no problem. It’s at least not impossible, no reason to give up.
Where’s the beef?
(The sheep story is blatantly stolen from this essay: http://yudkowsky.net/rational/the-simple-truth)