Version française
Home     About     Download     Resources     Contact us    

This site is updated infrequently. For up-to-date information, please visit the new OCaml website at

Browse thread
Coinductive semantics
[ Home ] [ Index: by date | by threads ]
[ Search: ]

[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
Date: 2006-01-18 (14:22)
From: skaller <skaller@u...>
Subject: Re: [Caml-list] Coinductive semantics
On Wed, 2006-01-18 at 13:58 +0100, Hendrik Tews wrote:
> skaller <> writes:
>    > Nobody is interested in final coalgebras in Set^op. 
>    Why not? This is really the key point of misunderstanding
>    I think. I'm not disputing your claim, I'm asking why not?
>    Perhaps they should be?
> Coalgebras in Set^op are for all intents and purposes identical
> to algebras in Set. If you want to study them, study them as
> algebras in Set. You will see nothing new if you look at these
> objects as coalgebras in Set^op. That's what duality means.
> Looking at an object through a mirror you see precisely what you
> can see looking at the object itself.

Perhaps my analysis is naive. But consider a simpler case
of products and sums. They're dual concepts, are they not?

In Ocaml we have representations of both, each can be used
with reasonable utility -- there is a degree of symmetry,
associated with the duality. It feels good!

Contrast to C, which has products, but the union construction
isn't a sum. And the many other 'popular' languages with 
this weakness.

Sometimes it seems looking in the mirror is good. 
It's what we want. We don't want something new!

>    >  Go out, read the papers on
>    > the Co-Birkhoff theorem! 
>    That's a pretty big ask of someone who isn't a
>    category theorist isn't it? Most mathematicians
>    can't understand category theory .. and I'm just
>    an ordinary programmer :)
> Well, you could try. I guess, that already the introductions
> contain enough information for what you are interested in: the
> duality of the Birkhoff and the Co-Birkhoff theorem. In any case,
> if you don't even try, your speculations about the contents of
> these papers remain wild guesses.

I often do try.. but seemed like a good idea to read Adameck first:

Still this is quite heavy going for me.

Incidentally .. if you look in Wikipedia for 'coalgebra' you may
be a bit disappointed.

John Skaller <skaller at users dot sf dot net>
Felix, successor to C++: