Ray Comfort once ventured to a mysterious pagan temple known only as a university with one goal in mind: like Darth Vader searching for the stolen plans, he would stick his microphone in the face of every sentient creature there until one of them gave him observable evidence of a change of kinds. When no one had any idea what he was talking about, he declared that the jig was up and evolution had finally been debunked. Ever since that day, creationists have continued to run this argument into the ground and no one still has any idea what they are talking about. The go to response is “well, what is a kind?” A valid question, but the creationist will inevitably spike the football and chide the stupid evolutionist for not knowing what a cat is, or something equally vague. What is a kind you ask? It is the most diabolical thing ever conceived by a creationist. William Jennings Bryan is probably in his grave right now smacking himself for not coming up with this shit. No one seems to be able to articulate what is wrong with the argument because there is nothing wrong with it. They have actually taken the way evolution is supposed to work and used it as an argument against evolution.
The first definition of a kind is a group of animals that they are willing to admit share common ancestry. This definition alone makes it something that cannot possibly change. Sometimes they say it is more at the family level of classification. In evolution, a family is nothing more than a group of animals who share a set of characteristics they inherited from a common ancestor who also had those characteristics. If a turtle evolves into a giraffe for some reason, it would still be a part of the turtle family, the only thing that has changed is what it means to be a turtle has become less specific because the family has become more diverse. There’s already a concept in taxonomy that describes this: it’s called a monophyletic group. Over the course of evolution, nothing ever leaves the groups that their ancestors were a part of. They only form new subgroups as the old groups become more and more diverse. It’s like Russian nesting dolls.
The second definition of a kind is a group of animals that look so much alike that not even creationists can deny they are related. This is because evolution happens one change at a time. Animals with a more recent common ancestor will have more in common because there has been less time for changes to accumulate between them. Animals with a more distant common ancestor will have less in common. That’s the entire basis of what Darwin was getting at, yet that’s where they’ve gotten the idea of micro-evolution and that’s where they’ve gotten the idea of kinds. In order to prove evolution, they demand to see less in common from animals that share a more recent common ancestor, the exact opposite of what evolution predicts. That’s where Ray’s sneaky request for observable evidence comes into play. Since you can’t observe when a group of animals has a more distant common ancestor, it falls into the realm of historical science and is thus on equal ground with the science of iron age tribal poetry.
You can try to explain all of this to creationists if you want, but they will probably just tell you to show them a fish giving birth to a monkey.