Skip to content

Instantly share code, notes, and snippets.

@ericTsiliacos
Last active November 14, 2017 06:30
Show Gist options
  • Save ericTsiliacos/ef6562a46f30c7bfa6dadd5c5f5d580a to your computer and use it in GitHub Desktop.
Save ericTsiliacos/ef6562a46f30c7bfa6dadd5c5f5d580a to your computer and use it in GitHub Desktop.
Understanding
-- "I agree with you that it is important once and
-- for all to examine all of our assumptions in order
-- to establish something solid." - Letter to Foucher by Leibniz
-- Break down propositions into defintions (recusively) until one arrives at identities who are called such
-- because their defintions are necessary truths (or non-contradiction).
-- Question(s): Are the natures or essences of things necessary truths? Is it provable?
-- Using the "for all" qualifity, one can arrive at the essence of a thing?
-- But there lies a contradictions: How would one to include an object into
-- the set of elements one is trying to describe without having already
-- known it belonged in the set? From whence does this ability come from?
-- Are we merely abstracting out the differences between objects?
-- For depending on which two object one compares, or the ordering one chooses
-- one might arrive at a different abstraction.
-- It could be thought of as the uses of objects that bring them together
-- and for a particular purpose that then an abstraction is discovered.
-- In nature, does one observe many objects of various kinds,
-- as Scipture points to the nature or form of created things -
-- "according to its [or their] kind".
-- "There are two things that should be distinguished in every argument, namely, form and subject matter."
-- - Samples of the Numerical Characterisic
-- http://yogsototh.github.io/Category-Theory-Presentation/#slide-46
-- https://people.mpi-sws.org/~dreyer/tor/papers/wadler.pdf
-- Sum types (or): Closed universe
-- Product types (and)
-- Exponential types (functions)
-- Recursive types (recusive functions)
-- Necessary truths and contingent truths
-- Types are structural
id :: x -> x
id x = x
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment