The basic concepts and main ideas aren't that hard: certainly no harder than understanding high-school level mathematics. Ever see what Alan Kay was able to teach children to do in the 70s, with an intuitive enough development environment and programming language.
Being able to write straightforward little algorithms to solve data-munging tasks is more and more a matter of basic literacy, especially for people like lawyers, or scientists, who need to sort and sift through masses of data, find the interesting patterns, and explain it to the world.
There is no easy way to do this at the moment. Not if you want to ignore details, just write something and have it work. The best I can do at the moment is guess if something is possible to do, guess the amount of work and tell someone (who’s an expert) what to do.
You are probably right about scientists, though. It’s a pain to see social scientists working with the mess that is called SPSS when r is so cool. I don’t know why that is but when toying around with r I feel for the first time that I am able to actually do something useful with what I know about programming.
Granted, he's a smart guy, and obviously he should really spend a few weeks or months learning the ins and outs if he wants to get especially sophisticated, but seriously, these things are essential for anyone with any kind of research or curation job ... and frankly for anyone: so much of our society runs on searching and sorting through masses of data that it’s important to understand just to know what’s going on, even for those uninterested in answering their own questions.
Just as it’s hard to be a doctor and a lawyer at the same time, so is it hard to be a programmer and a lawyer at the same time. Even if you only know a bit about medicine. Even if you only know a bit about programming. That’s not to say you can’t do it, but it’s not easy and many won’t do it.
It’s frustrating: I (not being a programmer) spent a few hundread hours of my life understanding programming and programming. Heck, I can explain bisection search or bubble sort to you. I can write cute little (clumsy) programs that do stuff (useless stuff, that is). But that’s not enough to save me hours in Excel. The most frustrating part about this is that I know what kind of automation is theoretically possible but I can’t actually do it. That knowledge is so often so unbearable. But I don’t have the time to invest a few days worth of work into understanding how to go from theoretical to practical, not for this one specific problem.
But that’s just how it is. You can’t be an expert at everything.