[Bluej-discuss] bluej-discuss Digest, Vol 53, Issue 23

[Barry Brown]

On Nov 28, 2007, at 10:12 AM, TJ Evert wrote:
> What's more simple (to a 1st-year student):
>
>       a  = b + c
> or
>       a.set( plus(b, c) )

Taken out of context (as these expressions are), both are equally  
confusing. The first one is obvious to programmers: the sum of b and c  
are used to change the state of the "a" variable.

But to a student who has had only prior experience in mathematics, it  
is clearly an assertion that "a" is immutably equivalent to "the sum  
of b and c." Using the same prior knowledge, it follows, then, that  
the statement can be re-written thusly:

     b + c = a
or
     c = a - b
or
     a + 5 = b + 10 / 2 + c

or any of a multitude of other combinations. Raise your hand if your  
students have tried these.

Teaching a first programming language that uses pseudo-algebraic  
notation strongly favors students with prior experience in pseudo- 
algebraic programming and is confusing to students who may be strong  
in math, but otherwise lack experience in procedural programming. In  
other words, the students who succeed in CS1 are those who come to the  
class with pre-existing CS1-like knowledge. The others are left in the  
dust. Is this how to encourage students to consider CS as a major?

By using a non-pseudo-algebraic language (Lisp and Scheme come to  
mind; Smalltalk and ML, too), we "level the playing field," so to  
speak. Everyone is equally confused.

Those with prior procedural programming experience have to "un-learn"  
some of what they thought they knew about programming. By learning a  
different style of programming, they come to think more abstractly  
about problem solving. Win.

Those with strong mathematics skills are able to apply what they  
already know about functions, recursion, and induction without  
altering much (if any) of that knowledge. Win.

And those with weak mathematics skills sometimes make the best  
students because you get to start with a blank slate. They learn  
problem-solving skills that they can apply in their math and science  
classes. (Dr. Bloch will tell you that they also learn how to write a  
history paper.) Because they have to consciously convert from  
algebraic expression notation to non-algebraic Lisp or Smalltalk  
expressions, the silly precedence rules of algebra actually get  
reinforced. Without anything to unlearn, an intelligent group of these  
students can quickly find themselves at the top of the class. Win, win.

> Note - No joke: there's a growing minority of college instructors  
> that are pushing for LISP to be a 1st-year language!


Is that a bad thing?

> There are extremes to any position. My aim isn't a reactionay  
> throwback to assemly (or bit-coding Turing machines). The ultimate  
> questions is "Will this help a student who knows *nothing* about  
> programming?"
>
> However, the *real* $64,000 questions is "What prior experience does  
> a 1st year computer science student have?" Some algebra, experience  
> working with files, maybe a little bit of logic with Venn diagrams.  
> So (especially in the first semester) why not use what they're used  
> to: algebraic statments, Boolean algebra, control structures -  
> procedural programming!
>

I can tell you with certainty that the syntax-overladen control  
structures of Java/C/C++/Perl/PHP frustrate the majority of CS1  
students who have had no prior experience. By contrast, almost every  
one of my CS0 [not a typo] students can write a correct Scheme cond  
expression.

-B