[davekinkead]

While I think programming can be a good way to teach general thinking skills like perspective, clarity, precision, and abstraction, I'm dubious about the claim that it is the best way to do so (§2.1).

Perspective - does learning multiple programming languages engender more awareness of different perspectives than say, learning multiple foreign natural languages?

Abstraction - does learning programming languages develop better use of conceptual abstraction than say, mathematics?

Precision and clarity - does learning programming languages develop better precision and clarity than studying philosophy and logic?

These are all empirical questions and no empiric data is provided. The lack of support in the argument for the conclusion given also hints that maybe learning programming isn't _the best_ way to learn clarity & precision after all.

Where I thinking learning to program stands out as superior is that it teaches a skill that is currently in high demand - programming.

[TeMPOraL]

Well, I think practical experience in programming can be[0] unparalleled in teaching one precision and clarity of thought. One could think that any hard science or engineering discipline would do, but the truth is, programming is unique - because as they say, you can't bullshit the compiler. You have to get the details perfectly right at all levels, or else your program either won't compile or won't work. In math, it's easy to just not realize your reasoning doesn't make sense. In science, it's easy to bullshit yourself with data. In engineering, well... you can't argue with nature, but your feedback loop is still much slower than in software.

I'd definitely say programming will teach you precision and clarity way better than logic and philosophy, because it's grounded in reality. If you want to build something cool and/or get paid, you have to actually think things through; the compiler won't be impressed by your long paper loaded with smart-sounding words.

Also, a side benefit here is that some time spent thinking about representing real world in programs lets one discover just how imprecise, messy and ad hoc everything is - doubly so if it involves humans. I'm sure I wouldn't internalize ideas like "map != territory" as well as I did if not for spending time thinking about modelling real world concepts in a computer program, and in the process realizing the "territory" they taught me in school was just a map (and a pretty shitty one).

> Abstraction - does learning programming languages develop better use of conceptual abstraction than say, mathematics?

I don't know. I haven't done much mathematics at the level when it gets interesting. I know algebra feels to me to require very similar kind of thinking as a lot of programming work.

--

[0] - I say "can be", not "is", because I've seen plenty of people who write code for a living, and yet their thoughts are as precise and reasoning as subtle as the Tsar Bomba. Somehow, some people don't seem to realize that the cognitive discipline they use when programming can, and should be applied to "normal life" too.

[GoToRO]

I think it works best with programming because you write your ideas in code and the compiler will quickly tell you that your idea is wrong, and you can't be mad at it because it is not a person. Now if a colleague tells you that you have no clue of what you are talking about... :)

But I agree that to reach a higher level, programming is not enough.

[seanmcdirmid]

It is difficult to gather empirical data on human behavior and efficiency, so I wouldn't ding Rinard points for that (as a systems PL researcher from MIT, he surely knows how to measure numbers). You'll find the lack of useful empirical data for most any pedagogic technique (what we can measure is performance in simple tasks, by this doesn't scale up).

Math and philosophy are hella abstract, while learning through programming at least has the learner building something concrete that can be more gratifying. When maintaining the student's interest is much of the battle, it might not matter as programming is not better than the other approaches, it only needs to be good enough.

[ujal]

"[…] Students can work with them in an interactive setting that provides immediate feedback and requires completely precise thinking." --

Math and philosophy provide the latter but not the former. I guess the tight feedback loop programming provides guides us by reducing the solution space.

[b0rsuk]

logic - not necessarily. But while math and most other sciences sprouted from philosophy, don't forget philosophy includes includes silly topics like "Is truth the same as beauty, and beauty the same as truth", or the nature of God. Personally I think these are bullshit topics. Philosophy definitely has capability for spending too much time on vague, subjective and poorly defined subjects. In contrast, there can be no ambiguity in a computer program.

Natural languages - in almost all cases they are used to describe the same concepts. Note books are are easy to translate to any language, even if you have to substitute lamb for guinea pigs in Bible. Everything else can be substituted almost word for word. Languages with substantially different syntax, like Japanese which uses reverse Polish notation (aka postfix), are rare. Verbs go to the end of the sentence in Japanese, that's why movies with Japanese often end up with characters screaming final bits of sentences. Without the end of a sentence, a Japanese can't be 100% sure what it's about. Now go ahead and translate Ruby to SQL, Rust to Javascript to Rust, Haskell to Python, HTML to vimscript, Logo/Processing to Bash, or Perl to anything else. I would say learning several languages with different paradigms and usage domains will be more useful for perspective.

I would say the value of programming is that it teaches being pragmatic, and using the right tool for the job. Math doesn't have to and often doesn't serve any practical purpose. Another big thing is it teaches problem solving and persistence.

[Singletoned]

> don't forget philosophy includes includes silly topics like "Is truth the same as beauty, and beauty the same as truth", or the nature of God.

That's like saying science includes silly topics like homeopathy and "conquering the seven signs of aging". Just because people claim they are doing science, doesn't mean that they are. The same applies to philosophy.

[mannykannot]

There is a middle way here. Currently-active fields in philosophy sometimes (usually?) lack bounded questions and precise answers, yet are not silly (ethics, for example.)

I don't think philosophy is about finding answers, the discussion itself is its purpose. Consequently, there is always the risk of generating closed, self-referential discussions in which the subject of the discourse is the discourse itself (post-structuralism vs. structuralism?) It also seems to be an accepted practice to hang an elaborate argument on nothing more than a personal opinion, as if doing so turned that opinion into an objective fact (I am thinking specifically of some of the arguments over consciousness.) For these reasons, I think you have to pick and choose from the offerings of philosophy if your goal is to teach general thinking skills.

[harrumph]

>Just because people claim they are doing science, doesn't mean that they are. The same applies to philosophy.

No True Scotsman is not the way to go here. It never is, actually.

[Chris2048]

There are standards of science, where are the standards of philosophy?

[JBiserkov]

>Perspective - does learning multiple programming languages engender more awareness of different perspectives than say, learning multiple foreign natural languages?

Honestly - I don't know. But I know from experience that learning different programming paradigms (structured, OOP, functional, logic, relational) is something very different from "Learning to write Fortran" in C, C++, Java, Python and Ruby. That learning Scheme without continuations or macros leaves most students with the impression that "[Scheme] can do everything C can [meaning iteration and if], but in a more weird way, so there is no use in this language whatsoever".

[naasking]

> Perspective - does learning multiple programming languages engender more awareness of different perspectives than say, learning multiple foreign natural languages?

Different perspectives on problem solving, absolutely. There is nothing comparable in natural languages to the paradigm shift required when first grokking logic programming or term rewriting.

> Abstraction - does learning programming languages develop better use of conceptual abstraction than say, mathematics?

You can think about the theorems involved in writing a todo list or balancing your chequebook, but to actually do these things you have to think programmatically. That makes programming far more applicable to most people's lives.

> Precision and clarity - does learning programming languages develop better precision and clarity than studying philosophy and logic?

Absolutely. Philosophy is muddied by the imprecision of natural language. You can't lie or confuse your computer, it's a strict task master.

[nicolapede]

Fascinating questions.

I am an applied mathematicians by educational background, while I am working more and more as a developer in my day-to-day job.

The answer that my personal experience gives to these questions is that programming has helped me a lot with precision and clarity. For me it probably boils down to the fact that a compiler is much more unforgiving than my own brain.

[Chris2048]

Math is over-abstract. At least programming often requires a concrete implementation. Mathematical expressions can easily have non-constructivity.