I find it useful to put the same form of argument into a different setting and see if the argument still works.
I find a watch in a field, and I know (or discover) a man who makes watches. That works fine.
I find an odd mechanical device in a field. I don’t know its purpose, or how it is made, BUT I may be able to infer these things. I then go and discover a man who makes these devices.
An Inuit Eskimo finds a pocket watch in a snowfield (circa 1845). This person has never seen anything like it, and may not even have a modern concept of time measured in regular units shorter than days. The Eskimo infers that animal spirits made the pocket watch.
We can break this analogy fairly easily, exposing where the assumptions are made. This basic approach works for other arguments too.