> I've heard of this problem, but I've never really gotten why the solution in lambda calculus is to rename variables in lower scopes with overlapping names so that everything has a globally unique name. The way that programming languages usually solve this is by keeping a stack of symbol tables for scopes, right? Why is that never presented as a possible solution?
That's a perfectly fine way to implement this-or-that language (and is hence why many/most programming languages do it). However, it's not a valid solution in lambda calculus, since lambda calculus only has variables, lambdas or applications; there's literally no way to write down "a stack of symbol tables", or whathaveyou (note that you can implement such a thing using lambda calculus, but that would be more like an interpreter for some other language; "running" that implementation would still require a way to beta-reduce the "host" language in a capture-avoiding way)
That's a perfectly fine way to implement this-or-that language (and is hence why many/most programming languages do it). However, it's not a valid solution in lambda calculus, since lambda calculus only has variables, lambdas or applications; there's literally no way to write down "a stack of symbol tables", or whathaveyou (note that you can implement such a thing using lambda calculus, but that would be more like an interpreter for some other language; "running" that implementation would still require a way to beta-reduce the "host" language in a capture-avoiding way)