The Diagonal Paradox can be extended in principle to any curve in 2D. For example, a circle of radius 1 has a circumference of 2pi, but if I apply my system of sine curves to the circumference I find that as they converge uniformly to the existing circumference, their lengths tend to infinity. Hence I am staring at what appears to be the simple circle I began with, but I now have one with infinite circumference, and hence infinite area.
Thus infinity is everywhere in plane geometry where it shouldn't be. — jgill
Hah! In the spirit of the infinite fractal coastline paradox. Nice paper.
I consider the Paradox an aberration that results from collapsing one dimension to a lower dimension in certain circumstances and insignificant although bizarre. But Wolfram claims that this crops up in Feynman diagrams. It goes to the very nature of lines and points. — jgill
No, I think it is significant and general. To relate it to my own interests – the Cosmos – there is this same virtual vs actual wrangle over renormalisation in quantum theory. Either renormalisation is some horrible kluge to be eliminated by better maths ... or in fact finitude and its dichotomous cut-offs have to be brought into the maths of the infinite somehow.
What I would point out again is how complex numbers may have their Penrosian "magic" as they speak simultaneously to the symmetries or conservation principles of rotation and translation. The incommensurability arise at that point – the foundational distinction between spinning at a spot and moving away in a straight line.
A 0D point could be spinning, it could be moving. It has no context, so we can't yet tell. It is technically vague. No symmetries have been broken and so no symmetries have been revealed.
But as soon as we imagine anything happens, the foundational symmetry breaking is the Noether symmetries that close the world for Newtonian mechanics.
Newton starts the world already in motion. The first derivative. The zeroeth doesn't really exist even at a Gallielean level of relativity. And this Newtonian world is defined by its twin inertial freedoms – to rotate and translate. One keeps things anchored at locales. The other sends them moving and so carving out the global largeness of space.
This would be why you find that converging to a limit has this fractal coastline property. Rotation and translation want to be different. They exist by being incommensurate.
It is the old squaring the circle story. Pi is irrational as curves and straight lines are at root incommensurate. They have to be that to co-exist in the same world and so be the symmetry-breaking making that world.
Complex numbers then speak to how the two directions of free action – one closed and cyclic, the other open and expansive – can be united under a unit 1 symmetry description. They explain how the symmetry is selectively broken in a way that is special to a 3D continuum. You get the chirality and commutative order arising as something that materially matters – unique in that it can produce those local knots, or Newtonian particles, that can't unknot, in the style modelled by twistor theory or fibre bundles..
So it is not a mistake resulting from the simple collapse of a higher dimensionality into a lower one. It is about a complex collapse – a reduction of dimensionality to a 3D continuum that then sets you up in a world where the number of spin directions finally match the number of momentum directions.
The collapse produces the internal gauge symmetry where rotation and translation become the new thing – a breaking of the symmetry achieved in 3D. You can get a cosmos founded on Newtonian mechanics with the Noether dichotomy to close it as a world, make it safe for local knots of energy that can't come untied, just endlessly shuffled about like bumps in a rug.
In short, you can collapse dimensionality to SO(3) and discover it then spits out SU(2). The magic happens. Commutative order becomes a thing. Time is born. Space is anchored in a way that can be described symmetrically by spherical or Cartesian coordinates. Etc, etc.
:grin:
And what of the cut off issue? I'm thinking this gives an argument in terms of the Planck scale representing a dimensional ratio.
The problem with infinities and infinitesimals is this urge to give them a concrete value. Or if not that, then they are processes without bounds. And neither answer is truly very satisfying in the light of physical reality.
The point about the Universe is that it is a story of fractal dimensionality or scalefree growth. The universe exists as a log/log story of cooling and spreading. This is both open and closed in some sense. And I've explained that in terms of it being an inversion at a deeper level.
It is going from very hot to very cold by going from very small to very large. Something is increasing as something is reducing. And if we look at a light ray, we see it is the same thing from opposite perspectives. Light waves – as your simplest sine wave or helix (the helix making the rotation~translation deal explicit) – are both stretched to their maximum possible extent and also redshifted to their maximum possible extent. They become as big as the cosmic event horizon and as cold as absolute zero.
So while we do like to measure all this using yardsticks like rulers and thermometers, it is essentially a dimensionless ratio. The Planck triad does not stand for some actual independently measurable number. It stands for a Platonic ratio - just like e scales unit 1 growth or pi scales unit 1 curvature. The cut off becomes simply the point where local spin and global translation first measurably come apart. And this is a qualitative distinction, not a quantifiable one.
The same goes for the Heat Death where everything arrives at its other end where the Planck triad are inverted - 1/planck - to give the cut off marking the other end of the effective 3D continuum. (Again, inflation and dark energy are unsolved aspects of this view.)
And this is why I argue for the numberline in these same terms - as the infinitesmal and the infinite as each other's reciprocal measure. A way to have the small and the large being both unboundedly open and yet also finitely closed.
The missing bit is that the numberline is based on naked spatial intuition. Peirce and Brouwer were trying to bring in time and energy (the two being complementary under quantum uncertainty) as the way to achieve the trick of an "open closure". Sure you can count forever. But that then takes time an energy.
You have two actions in opposition. This means that you can only expand until you run out of energy. And you can only contract until you run out of space.
Adding this to the numberline conception builds in the self-limiting finitude that the conventional spatial version lacks.
As I've said, Turing computation has run into a similar story of physical constraint. Infinity only reaches so far in a world where the gravity of being a hunk of circuits curling up into a black hole at a certain critical energy density – a cut off point.
So rather than either/or when it comes to taking sides on virtual infinity vs actual infinity, I've been pushing "both" in the dichotomistic sense for this reason. Openness and closure are what must emerge as themselves something that co-arise from the firstness of Peircean vagueness and become a combined continuity within the thirdness of Peircean generality.
A tricky business.