I must admit I find that style of describing things way too confusing. I have a much easier time understanding things if they are presented directly (one paragraph problem description, one paragraph solution). For example I understand the wikipedia description much better: <a href="http://en.wikipedia.org/wiki/Y_combinator" rel="nofollow">http://en.wikipedia.org/wiki/Y_combinator</a><p><i>Problem</i>: how to write a recursive function if the function doesn't have a name I could reference it by.<p><i>Solution</i>: since the result of the Y combinator applied to a function is the same as the function applied to the Y combinator applied to a function, we can use that instead (in brief Yg = g(Yg)).<p>However, it is interesting to me that maybe some people have an easier time with "dumb" explanations. On the other hand, maybe many people think they are bad at maths because all they ever got were dumb explanations.<p>The problem with the dumb explanation: it is far too long. All the stuff that is only decoration needs to be loaded in the brain and distracts from understanding the real issue.<p>Also, why explain Y combinator and recursion in one go? Would be easier to explain them separately, two smaller chunks, easier to load into the brain and easier to digest.<p>Sorry that I can't be more enthusiastic about that paper. It is certainly praiseworthy if someone at least tries to make such things understandable.<p>I have the same gripe with "Goedel-Escher-Bach" - it took thousands of pages (or something) just to explain Goedel's incompletenes theorem, which can actually be described in two pages. The unsuspecting reader might not even realize that GEB is about that theorem. (I simplify - it has been decades since I read it, but I think really Goedel's theorem was the gist of it). It was an entertaining read, but still.