Kat,
I knew it was.

Alex,
You didn't name one, I guess you can't. I at least have never met any person in many discussion who would have supported such a statement nor have I read such a notion in one of the many books I have read about such topics.

I mean that macro evolution can not be proven by repeatable experiment because you would have to repeat the history of the planet, which is not possible (Alex/Mousethief)

That's an interesting statement for all its implications on what constitutes scientific evidence and how theories are tested on many more fields than just evolution. It deserves a refutation for I think that it implies wrong ideas about how science works or should work.

The method of enquiry that has been termed science has in its history more often than not been applied to situations and cases that are in the strong sense unrepeatable. For instance, when I study human memory, all these humans in my lab are different and even the same person might be different (mood, attention, inner states) from one test period to the next and even from one test question to the next. Why do I think I'm allowed nevertheless to generalise across my findings? One simple reason is it works. That is, I can make successful predictions about future performance, about influences of other factors on performance. If my predictions wouldn't work, I couldn't convince any other scientist by armchair arguments about the validity of my research. As long as my predictions are sound, any arguments from philosophers don't count for most scientists I know.

In the physical sciences there are also many instances in which complete repeatability isn't possible. Think of forensic evidence which is considered admissible as proof. Never two bodies are completely alike (we wouldn't want to experiment by killing a volunteer) in terms of length of exposure, ambient temperature or weather. Nevertheless the rules in that branch of science work so good that this evidence is given greater weight in court than most other evidence.

Now let's consider science as applied to nonrepeatable (most of them in our past, but some like a meteorite coming dangerously close might be in the future) events. Let's skip for a moment evolution (for I do not have the impression that you can discuss that topic without falling below your usually high level of arguing). Let's talk about geophysics, cosmology, paleoanthropology, astronomy, and many others. They all make theories about unrepeatable events (in our timescale) like the continental drift, like the origin of planets, like the formation of rocks and many more. They test these ideas in repeatable experiments. When their predictions fail consistently they know that something is wrong in their theories and will either adapt them or discard them. When they have tested different theories with different predictions and one of them has 'won', they might even say for short that this theory has been proved though they know that it only has not been disproved yet and is only able at this moment in time to make better predictions about experimental results than competing theories. If you have a theory about let's say continental drift which makes correct predictions to all tests any scientist has thought of yet, you'll consider this the best fitting hypothesis about what has happened at a time when no human has watched. Yes, and scientists prefer these theories to e.g. Rudolf Steiner's (founder of anthroposophy) musings from occult sources about the early continents. One of the best reasons for this pattern of preference is the much larger predictive power of the scientific theories.

Science has repeatability of experiments as a prerequisite for proof (or better: failure of disproof) of theories and not as you seem to imply repeatability of events. If you really demand complete repeatability of events you'll have to throw out much more than just evolution.

As for evolution, it is the best fitting hypothesis (several of them competing, actually) at this moment and in my eyes much better and much more testable than accounts from several thousand year old folktales.

Wolfgang