There are several hallmarks of science. One of them is that scientific theories make predictions about future observations. Another is that if these observations cannot be explained by the theory, a new theory has to be developed. Science history is full of such flip-flopping. Probably the most well-known example is Einstein's theory of relativity replacing Newtonian mechanics. Another a lot less spectacular example is a very personal, cherished hypothesis of my own. When I first started studying operant conditioning, I was convinced the biological mechanism would be the same as the one for classical (Pavlovian) learning. The beauty of a unified theory was just too attractive. However, over the years, my experiments kept proving me wrong and last year we published a study which showed that there actually are two distinct molecular learning mechanisms. It is precisely because these results run counter to my own presuppositions, that I trust the data, that I'm more confident than I usually am, that I haven't subconsciously made an error in my experimental design or evaluation just to make the outcome fit my expectations. Scientists should be (and mostly are, I would hope) especially cautious and more suspicious and critical than usual whenever their results fit their hypothesis.

To make this more clear: what can cost politicians their career is one of the fundamentals of science: if the experiments challenge current opinion, scientists gladly change their mind. Multiple times if this is what the data requires. In that way, the ideal scientist is an opportunist, following the data wherever it goes. Konrad Lorenz once famously wrote: "It is a good morning exercise for a research scientist to discard a pet hypothesis every day before breakfast. It keeps him young."

This is in stark contrast to what creationists think about their "theory": they already know the answer (magic man done it) and are now picking observations that suit their agenda. Changing your mind is totally out of the question. Of course this is nothing new, but it really struck me in a very direct way when I saw the interview Michael Shermer did with Georgie Purdom (from Panda's Thumb). Unfortunately, the video is no longer available (I wonder why). In it, young earth creationist Purdom goes on and on trying to explain how she already knows that her view of the Bible is true and how she thinks this is the same as scientists believe evolution is true. Of course, what she calls "presuppositionalism" is a red herring. In contrast to her, scientists change their mind all the time. Sometimes it takes quite a while until some "presuppositions" fall, e.g. 200 years for "Newtonianism" (sounds like "Darwinism" doesn't it?) or about ten years for my own pet working hypothesis. Sometimes it only takes a single experiment, though.

As is true for any scientific theory, there are observations which can be predicted which would make scientists change their mind about evolution. Steven Gould's precambrian rabbit is of course such an observation. Importantly, an observation that the theory in question currently cannot explain, yet, is clearly NOT such a case. For instance, we currently do not understand how gravity really works (which is why we have the most complicated experiment in all human history: the LHC). This ignorance does not mean we should now all become adherents of intelligent falling. There are a few observations which would change my mind about evolution.

• The precambrian rabbit
• The discovery of a species whose individuals transfer fitness benefits to other individuals at fitness costs to themselves and without return fitness benefits (hint: cooperation is not such a case, but close). It's very difficult to unambiguously show this, but should in principle be possible. Not surprisingly, many evolutionary biologists are working on cases where this possibility could be considered, because it would change their mind about evolution.
  
• The discovery of speciation without a trace in the DNA. Speciation events can be traced to events in the DNA (mutations, translocations, duplications, inversions, etc.). Observing a speciation event in the lab where, say, a bacterium changed into a fish without any trace of the molecular changes in the DNA would constitute such a case. Coincdentally, this is the case creationists sometimes cite as the one they would need to see to accept evolution ("but it's still a bacterium!"), which is of course ludicrous. But such an observation would make me change my mind about evolution.
  
• A mechanism by which creation works. Of course, "magic man done it" is not an explanation. A scientific theory explains how something happens in order to derive predictions for future observations. Discovering the biochemical processes by which clay turns into a human or how the not really all that intelligent designer changes around the DNA (e.g., did he/she/it use PCR?), would make me change my mind about creationism.
• I'm already beginning to change my mind about some of the more detailed mechanisms of evolution. I think the link between epigenetic phenomena such as genomic imprinting and more permanent DNA changes in the germline provide for an interesting alternative to more or less random DNA rearrangements.
  
• I'm sure there are more points, if one thinks longer and harder than I currently have time for?
  

So, creationists (and I don't really care what denomination of creationism you subscribe to), what would change your mind? Would anything change your mind?