klonk wrote:If you are unhappy with the word "premise," you can say meta-premise instead.
That would be an improvement! Just so long as we agree that where the problem with supernatural explanations fits in is not with the initial hypothesis, but in what constitutes testability of a prediction.
Take this matter of repeatable experiments. This is a rock bottom requirement for science. The subtext is that the experiment is repeatable by anyone. Given suitable conditions it is also repeatable at any time. I should be able to devise a demonstration that you can repeat at your leisure and see that it works the same for you as for me.
But if you open my holy book, or most others, you find accounts of miracles. A miracle is by definition a one-off event, or something that happens only rarely, and perhaps not for everyone. (Not everyone is a miracle worker.) So that whole business is outside scientific investigation.
Repeatable experiments are nice, but not required. There are plenty of statistical methods that could be applied to infrequent events which are not controlled by an experimenter - some aspects of meteorology and most branches of economics and finance rely on non-experimental methods since researchers cannot control their respective inputs. So, sporadic observations can be dealt with if the data is well recorded.
There is a problem if the prediction is entirely unobservable and has no other observable consequences (eg trying to test if a person acting by some definition of morality actually got into heaven). However, we could test the likelihood of people being smitten down by lightning by hiring one person to stand in a field and curse God repeatedly while another in the same field sings his praises. Do you think we'd find a significant difference between the two or does God just move in ways too mysterious to fathom?
But note that it is possible to apply methodical, logical reasoning to accounts of the miraculous. It's something theologians do. It's not that logic doesn't work in this case, but clearly science won't. The worldview that says, truth claims are founded in repeatability, is what is in the way. If this is not an assumption, what is it, exactly?
Where repeatability comes in is that even with a statistical approach, some underlying causal framework needs to be present. In these cases, the supernatural bit is assumed to act so randomly that it cannot be statistically separated from any other random error term making the model unspecified. It is therefore very difficult to separate a miracle from an outlier ex post.
With miraculous accounts there is also a problem with endogeneity (ie going by self reported observations by people who have been brought up under a particular religion who are perhaps more likely to frame their experience in terms of that religion) so this sort of evidence is sketchy at best.
So also for the scientific rejection of unseen agencies, outside observation and measurement. Where science meets with those, the general response is we just don't know enough yet; we will find a perfectly natural explanation by and by.
That's very true. Some would say we just haven't modeled the situation well, others that it is due to some external agency. I think the trend has been to steadily nibble away at the things that are assigned to the latter, leaving it the more general issues of origin and purpose.
The rules you adopt going in, when you reason, seem to me implicit premises in the reasoning you are going to do. Thus my claim that using science to buttress a stance of atheism is essentially circular. You are merely reiterating something assumed from the beginning.
Personally I would say science buttresses agnosticism rather than atheism. There are certainly aspects of existence that do not lend themselves to providing evidence
Are you sure we disagree?
Perhaps in some small areas. I find details to quibble about with most things though
Why does man Kill? He kills for food.
And not only food: frequently there must be a beverage.