Programming (explanations for the non-programmers, mostly)

Ok, that explanation makes sense, thank you.

@Adm You pretty much lost me, you´re talking programmer again.

Also in my defense i mentioned before in this thread that I absolutely suck at math. (I actually suspect dyscalculia, but have no idea how to find out)

@Mentarnes Thanks. Sorry no non rude way to give this info, but you answered a question i didn´t ask because i already knew the answer.

 

Trying my hand at answering, more to try to think through my own understanding than to offer actual response. Hopefully someone will look at this answer and tell us both where im full of shit/just straight up wrong.

You're telling the computer you have a thing called 'a'. Ok, the computer says, what type of thing is it? The computer knows about numbers, characters, boolean algebra values (true or false), and possibly some other stuff idk. 'Its a number' you tell the computer, and now it knows if you tell it a=b, you mean 'the number value of variable b' and not say 'the actual letter b'.

Boolean algebra. 'A' can be either true or false. If a is true, then b. If b is true then c. If a is false, d. If c, then a. If c, then not d.

http://en.m.wikipedia.org/wiki/Boolean_algebra

 

I did mention earlier that I only knew a bit of programming...
also a tiny dog has claimed one arm and i cant chicken peck decent explanations out fast enough XD

 

@seebs Yep, makes sense. Primitives are things which are not made up of other things.
@Mentarnes Ah yes, pets. They get in the way....

I hope I´m not being anoying, was just having fun playing question and explanation science.
Results so far: I got several answers I understand, one answer I did not understand, and two answers to questions I did not ask. Now I kinda wish I had a bunch of people in another field I know nothing about to try that on.

 

Though note that they're not saying that people don't learn; they're saying that the people with a consistent mental model at the beginning (even if wrong) still have a consistent mental model at the end (possibly/probably now right). It's the people without a consistent mental model who still don't at the end.

 

Yeah, that's exactly why I'm questioning how effective that test is. Now that I know that = in this context does not mean = in the familiar context, the problem's a cinch. And it took all of ten minutes to explain that the last statement actually meant "the value of thing a is now thing b." Even given that I'm quick at those things, it's definitely something that should have improved over the course of a couple weeks.

 

All the 'meaning' is assigned by us, not the computer, which is conceptually a really simple device; it can do math, logic, comparisons, move values to different places, and jump around. (It's a pretty complicated device in the interests of doing a huge fuckton of these simple things per second).

 

@Lissiel: Meaningless in that the computer has no idea what integers, a, = or 20 are. The statement means something to us, but to the computer it's just symbols.
The model I tend to use is Searle's Chinese Room analogy. Picture a room with no doors or windows, only a tiny slot just barely big enough to slip a piece of paper through. Inside the room is a man and a big book and a stack of blank paper and a pen. The man doesn't know any Chinese. Every once in a while, a piece of paper with Chinese characters written on it gets slipped in through the slot. He looks at the characters, matches them up to symbols in the book.
The book is not a dictionary. It's a set of instructions. All it does is say things like "if you see this symbol and that symbol and not this other symbol, write this symbol on the roll of paper." And when he gets to the end of the characters and the end of the instructions, he sends the bits of paper that he wrote on back out the slot. The man does no thinking; he just looks things up in the book and follows the instructions.
And the person on the other side of the door, who can read Chinese and has no idea what the room contains, thinks that whoever is in the room must be very wise indeed.
That's a computer. It does no thinking; it just follows instructions. Meaning is assigned by the people writing and reading the code and examining the results, but the computer only sees meaningless symbols and arbitrary instructions.

 

One thing I'm pretty sure on is that classic CS instruction fails a pretty damn large subset of those who try it. Whether it's because their minds just can't do it, or because the teaching's bad, I'm not sure. Probably some degree of both.

And damn, yes, there were plenty of CS students in my class who Never Got It. Some of whom graduated.

 

I'm kind of the converse; I don't think the Chinese Room is a good thought-experiment for a difference between human and computer cognition because I'm not convinced that humans definitely do think in a way that is distinguishable from the Chinese Room type setup. It's just several levels of emergent phenomena on top of heavily hardware-specific code.
The point of the Chinese Room is that the man has no idea what he's doing, but there's more to the room than the man; there's the book. The Chinese Room is the entire man-book-pen-paper-slot system, and if there is meaning to be had, it's in the book. The man is just a tool for the book to use. But the analogy has the computer being the man, while the book is a compiler/interpreter created by a human being, so it is allowed to have meaning imparted by its author. The man is still clueless.

 

Do you think there may be social or other biases in the way programming is taught that could account for that? Sort of like how you can teach geometry as abstract numbers and rules or as actual visible/tangible shapes moving through space, and different people will understand it better one way vs the other?

 

As an argument for anything in particular I agree that the Chinese Room is a bad argument. As a model I find it useful in terms of breaking apart the "instructions for how to do a thing" and "object that does a thing" very firmly. In particular, it makes a point that the I/O language is not the internal language of the processing unit, so any semantics of the I/O language are not necessarily translated to semantics of the internal language, and that the object that does the thing doesn't necessarily have to understand the semantics of the I/O language in order to do things in the I/O language. Maybe not naturally, maybe not adaptably, maybe only in certain subtopolects of Chinese that can't be considered complete languages, but it can do things.
I think this is an important distinction to make, between the instructions and the instructee and what they need to understand. The instructor can understand all she wants. The instructee follows orders.

 

Fun facts: you can fool google with genus-species names. I know this because of reasons.

geology geology geology