Date: Thu, 19 Aug 1999
Induction
by Thomas M. Miovas, Jr.
Note: I'm combining several posts on induction in this one reply.
><HB: Can anyone come up with an example of circular induction?>
Would a "circular induction" be one in which a directly
perceivable inductive truth (or one of which one is directly
aware) is attempted to be validated by means of referring to
something not directly perceivable (or one of which one is not
directly aware)?
As an example: There are those who seem to want to validate the
nature of consciousness by means of referring to specific
scientific knowledge, such as the nature of neurons (and their
components). Since we can only be aware of neurons by means of
our consciousness, and only after a long line of inductive
inquiry (requiring actively using our consciousness to study the
brain), this is an attempt to validate something we already have
direct access to -- our consciousness. Does that make it circular?
Another example would be in trying to validate sight by means of
the electromagnetic theory of light.
Concerning induction and propositions:
><HB: The problem is to give an understanding of all these
>different types of proposition:
>
>1. Subsumption: "Dogs are animals"
>2. Differentiation: "Dogs are not cats"
>3. Attribution(?): "Dogs are big"
>4. Action: "Dogs bark"
>5. Relationship(?): "Dogs are bigger than snails"
>
>It's particularly action that I have difficulty with.>
Perhaps the difficulty arises because you have accepted the idea
that in order to get the cause of something one has to go one
level down (re the previous thread on causality -- i.e. to know
the cause of water boiling, one has to know something about its
molecular structure).
I understand the Objectivist position on causality to mean that
it is perceptually self-evident, insofar as we perceive entities
acting. In other words, the validation that the nature of the
water is the cause of its boiling is that we perceive the water
acting -- i.e. boiling, when it is under the right
circumstances. In one inductive sweep, we perceive the entity
and we perceive it acting (for those entities we can perceive,
and when these are acting; the water in the pot being the entity
in this case.)
So, a proposition like "Dogs bark," is one in which the totality
of the inductions (the perceptions of a dogs barking) is
separated out so as to be integrated according to perceived
similarities -- entities to entities and actions to actions --
but can be retained as a whole, because it was perceived as a
whole (if we are talking about each incident of a dog barking).
However, there is no similarity between dog and barking (or any
other entity and its action), so we can't form a concept that
would be the (hypothetical) equivalent of: dogbarking.
I think some primitive tribes have languages that attempt to do
this, but their vocabulary becomes overwhelming and violates the
crow tremendously. If you can imagine a (hypothetical) single
concept for each of the following: dogbarking,
bigblondedogbarking, blackpoodlescrathing, redbirdflying, etc.
Several modern languages do something like this, such as some
words in English and cumbersome run-on words in German, but
these are really a combination of two (or more) concepts that
can be retained precisely because the root concepts(s) are
previously know and can be isolated out for analysis. Try
recasting the above (hypothetical) concepts as: gobinzig,
trubik, ospile, lulo etc. and a different and unrelated concept
for each entity and each of its actions, and you will see the
difficulty.
What a proposition such as "Dogs bark" does is to put back
together (to re-integrate) the one perceivable event into a
whole. Measurement omission comes into play because we don't
specify, in this simple proposition, the details about what type
of dog it is or what type of barking it is doing.
