You are on page 1of 4

Cosmology in Topos Quantum Gravity

M. D. Sheppeard

September 25, 2008

The observation of dark matter, dark energy, and now a dark flow, in-
dicate serious problems with the standard ΛCDM paradigm. In spin foam
quantum gravity, the cosmological constant is typically related to deforma-
tion parameters of classical group symmetries. However, the preselection of
any symmetry imposes restrictions on the geometry of spacetime which one
does not expect at the fundamental level in quantum gravity.
Richer structures arising from category theory allow the definition of
observables in terms of measurement algebras that are independent of the
concept of group symmetry. From this very different point of view, the
deformation parameter of a Hopf algebra reappears as a cosmological scale
parameter, where discrete scales correspond to categorical dimension. One
expects dark energy to be eliminated in the classical limit, where the scale
parameter is driven to zero (q → 1). In this case, observations may be
explained by a decreasing speed of light cosmology, as shown by Louise
Riofrio [6]. As a dimensionless parameter, the speed of light is inverse to
the variable ~ ' logq deformation.
In other words, the domain of applicability of general relativity in cos-
mology is limited by its independent fixing of c and ~, which actually become
quantised. A holographic emergent 3-vector time (or background tempera-
ture) T is identified with an entropy increasing process dual to a contraction
of measurement space that occurs for high energy observables. It is impor-
tant to note that this concept of time is observer dependent and not tied to
fixed foliations of a classical reality. A brief outline of this approach is given
in the next section, followed by a summary of the related current research
focus in cosmology.

1 Topos Quantum Gravity
Let categorical dimension be denoted by a parameter N . A classical topos is
a category with N = 2 levels of arrows, namely objects and 1-arrows. The
motivating example of a topos is the category of sets, and the categorical
structure includes the N = 2 valued Boolean logic of set union and intersec-
tion. In [7], a quantum mechanical version of a topos was considered, based
on the category of vector spaces and the logic of spin.
This identification of truth values with dimension is not usually con-
sidered a fundamental aspect of the categorical axioms. However, in topos
quantum gravity, one is prepared to consider novel axioms that are more in
line with physical expectations. In particular, categories themselves are sec-
ondary to operads (one object multicategories) and their higher dimensional
generalisations [1].
How does this philosophy affect the axioms for a higher dimensional
topos? To begin with, one seeks N − 1 categories with N natural truth
values. But instead of focusing on semantics, one constructs computational
tools for cosmology using simple matrix algebras over finite fields of truth
A useful interpretation of such a matrix is as a Chu space. The vertical
and horizontal indices are not identical, but rather represent dual variables,
which we will think of as position and momenta, although not in any classical
sense. The crucial example is given by the n × n matrices of the quantum
Fourier transform [4] at N = 2. Always generated by cyclic permutations
(234 · · · n1) with entries 0 and 1, these matrices represent Chu spaces with
n points over the Boolean truth values of the category of sets. For each
spatial point, there is one paired momentum for which Axp equals 1.
The 2 × 2 Fourier transform is generated by the Pauli operator σX and
σX2 = 1. But we also need to consider the full set of Pauli matrices, which

require the five values 0, ±1, ±i. These three matrices correspond to a set
of mutually unbiased bases for a two dimensional Hilbert space, namely the
measurement of two possible spins in the X, Y and Z directions. (Note that
these spatial directions are not defined prior to the concept of complementary
observable, of which there are always n + 1 for prime dimensions n).
Similarly, one considers the observables for inertial rest mass as a type
of three outcome experiment. The Koide relations for rest masses [3] rely
on complex number Fourier circulant matrices, or at least matrices with en-
tries in the field Q + Qi, arising from an infinite number of modes on the
quantum circle. This complexity (in all senses of the word) is not surpris-
ing from a Machian point of view, where inertial mass does indeed depend

upon all scales. In fact, many aspects of stringy cosmology appear in this
framework, but the physical interpretation is very different. For example,
the higher dimensions of string theory become the dimensions of quantum
number spaces, on which the measurement algebras operate.
A dual indexed matrix of zeros and ones may be viewed as an adjacency
matrix for a bipartite graph. That is, one connects white (vertical index)
and black (horizontal index) nodes with bits of string when the matrix en-
try equals 1. The basic circulant (234 · · · n1) corresponds to a basic braid
diagram with n strands, which join the white nodes to the corresponding
black ones. If we insist in ordering the nodes on an n = 3 diagram ac-
cording to the cyclic permutation, turning the adjacency matrix into the
identity by cubing it, then the selected braid is of the form used to describe
the (massless) standard model particles in [2]. There is no diagram for the
Higgs boson. The unknotted boson diagrams may be obtained by cubing
the circulant braids, which can be interpreted as a composite nature for the
The standard model particles are more usefully described [3] by 6 × 6
idempotent operators, which incorporate both mass and spin. This ex-
plains the 6 strand nature of the full twisted Bilson-Thompson diagrams [2].
Charge arises from the embedding of two dimensional spin operators into
the three dimensional circulants.
Although this framework has led to clear progress in deriving particle
rest masses from first principles, the question of how to apply it usefully to
cosmology remains.

2 Directions in Cosmology
The main observational tests that could distinguish this approach from more
classical approaches to cosmology would exploit the novel scale dependence
of a so called decreasing speed of light [6]. Both the dark energy and dark
flow are easily explained in this way, the latter by allowing much more
massive black holes to exist in the early universe. Explanations invoking
matter outside the observable universe are unacceptable from this point of
Untested predictions include radical T duality effects, such as the corre-
lation between WMAP data and the local ecliptic. Unlike in string theory,
it is not a fixed Planck scale that determines the T duality transformation,
since ~ is seen as a cosmic scale parameter [5]. Thus topos gravity promotes
T duality to a map that wraps observable scales back upon themselves. This

was a prediction that has been at least partly vindicated by the recent dark
flow observations.
At a theoretical level, cosmology requires the use of n × n matrices for
all n, just as in stringy matrix models, where large n operators approximate
Riemann surfaces with ribbon graphs on n vertices. There is room for a lot
of development here in categorical ribbon graph combinatorics. My focus so
far has been from the particle physics end [7], but this does not consider all
values of n together. Note also that ribbon graph T duality is a well studied
area of mathematics, having led to remarkable theorems about moduli spaces
for Riemann surfaces. Mulase et al [8] describe a scale dependent T duality
between real and quaternionic twisted ribbon algebras.
The ribbon diagram approach to the combinatorics of Fourier operators
allows an easier mixing of dimensions, since dimension corresponds to the
number of strands. This correspondence also appears in the zeta value
algebras of Feynman diagrams, as studied by Kreimer, Connes and others,
which lends further support to this approach.

[1] M. Batanin. The combinatorics of iterated loop spaces.

[2] S. O. Bilson-Thompson. A topological model of composite preons.

[3] C. A. Brannen.

[4] M. A. Nielsen I. L. Chuang. Quantum computation and quantum

information. Cambridge, 2000.

[5] M. Pitkanen.

[6] L. Riofrio.

[7] M.D. Sheppeard. Gluon phenomenology and a linear topos. Ph D

thesis, University of Canterbury, 2007.

[8] M. Mulase A. Waldron. Duality of orthogonal and symplectic matrix

integrals and quaternionic Feynman graphs. Commun.Math.Phys.,
240:553–586, 2003.