Smalltalk and Python

Alex Martelli aleaxit at
Sat Dec 16 09:35:59 CET 2000

"Neil Cerutti" <cerutti at> wrote in message
news:slrn93l1tf.5l.cerutti at
> Luckily, the objects we model in software are generally simpler
> and easier to define than a Vehicle class. In the real world,

I'd put it another way: the _models_ we build in our minds
(particularly for software-design use, but not only!) are
simpler than the 'real world' objects.  That's why they're
models: the map is not the territory.  It's the source of both
their usefulness and their limitation.

> things have "family relationships", (Wittgenstein?)--you can't

Wittgenstein is talking about _human language_ (which in his
terms covers human thoughts, and actions in as much as they
communicate, to another or to oneself): a word (a concept)
denotes a 'cloud' of entities (he might have called it a fuzzy set,
had the term existed) which are joined by a vague 'family

There is no implied assertion on any 'deep nature of reality'
(no needed noumena under those phenomena, if you will),
because "whatever we're talking about, we're only talking
about natural history of human beings" -- i.e., all of our
words/thoughts/concepts/models/actions/abstractions are
ALWAYS in the context of our biology/needs/desires/
abilities (that's in the _mature_ Wittgenstein; the Tractatus
of the 20-something Wittgenstein foreshadowed that key
realization but didn't truly and fully grasp it).

> always make an easy determination about what class a thing
> belongs to. In an ocean simulation, you might have a Fish class
> and it would be useful for the Dolphin class to inherit, but it
> IS-NOT a fish.

At roughly the same time as Wittgestein's, Alfred Korzybski was
busy building general semantics, and the law of non-identity.

The map-territory relationship was his conception, and more
fully stated: "The map is not the territory it represents, but, if
correct, it has a similar structure to the territory, which accounts
for its usefulness...." and a few pages later "it is enough to train
children to abandon the ‘is’ of identity", to build "consciousness
of abstracting".

The focus on _usefulness for our purposes_ of our models,
which need to be consciously built as such, as abstractions,
parallels the 'natural history of human beings' idea.

If this be an ontology/epistemology split, then, make the most
of it... it seems to me that most thinkers after W, K and other
contemporary analyzers of language (Sapir and Whorf also
come to mind:-) have more or less accepted this split, i.e,
focused on 'how we know what we know and what do we do
with it' more than on 'deep nature of reality' issues.

> Perhaps it makes sense to instantiate a Vehicle if you believe in
> Plato's idea of an ideal vehicle. All other vehicles will be
> different or imperfect in some way, but the ideal could exist and
> be instatiated--by God, er... by the programmer. ;-)

If you accept Plato's conceptions, the Ideal cannot be instantiated
in this world; there is a different, 'true' world, where _only_
Ideals lie.  Plotinus (and other Platonists before the Renaissance)
may be closer to this 'Ideal Instantiated' possibility.  But _this_
is indeed ontological more than epistemological, and lacks the
key distinction of maps/models/concepts (in our minds, for our
purposes) vs 'noumena'.


More information about the Python-list mailing list