Let’s do an experiment. You will need a photograph of a man who is not bald. The average human head has 100,000 hairs, and for the sake of this experiment assume that this man has exactly 100,000 hairs on his head. Imagine that you pluck one out. He now has 99,999 hairs on his head. Is he still not bald? Of course. There is no significant difference between having 100,000 hairs and having 99,999 hairs. Imagine that you pluck out another hair, so that the man now has 99,998 hairs on his head. Again, this is not a significant change, and the man in the picture is still not bald. Repeat this enough, and eventually you will imagine plucking out the last hair on the man’s head, leaving him with no hairs. If, when starting with a man who is not bald, plucking out one hair at a time never changed a man from being not–bald to bald, wouldn’t this mean that a man with no hairs on his head is not bald?
Houston, we have a problem.
Clearly, our conclusion is unacceptable. We cannot say that no man is bald because our experiment resulted in a man with no hairs on his head, who would be bald by definition. We cannot deny that not–bald men exist, as our experiment started with someone who we agreed was not bald. This implies that somewhere along the line the man went from being not–bald to being bald. When did the man change and where was the boundary between being bald and not bald; equivalently, how many hairs does a man need to have on his head so that, should one be plucked out, he will go from being not–bald to bald?
We cannot pick any random number, say 10,000, because this brings up issues of arbitrariness; why not 10,002 or 9998? We cannot defend our choice of 10,000 being the boundary because it is a round number, because 10,000 is a round number because of the numeral base we are using (base 10 in this case). The Ndom language of Kolopom Island counts using base six. In base six, 10,000 is 114,144, not at all a round number. Hence, defending 10,000 as the boundary between being bald and not–bald because it is a round number implies that a man’s state of being bald or not–bald is dependent on the language we are discussing him in. That idea is, of course, ridiculous. Finally, we cannot claim that there are three categories— not–bald, unsure, and bald— because this necessitates asking where the boundaries between not–bald and unsure, and between unsure and bald, are. This therefore reduces to the same problem with selecting a boundary as before.
The resolution to this problem is realizing that even though there is no clear boundary between being bald and being not–bald, there still is a difference between them. The lack of a clear boundary between the category bald men and the category not–bald men in no way means that those categories are not distinct. The key point is that a large number of successive differences in degree (plucking out one hair at a time) eventually amounted to a difference in kind (not–bald versus bald). And that is exactly what the difference between micro and macroevolution is. A large number of successive changes (microevolution) eventually accumulate, causing evolution into a new species (macroevolution).
So now, whenever you debate cdesign proponentists, you can explain this to them and ask them, if they still accept microevolution but deny macroevolution, why they deny the existence of bald men.