Forays into Mathematical Physics

by

Jonathan Hackett

A thesis
presented to the University of Waterloo
in fulfillment of the
thesis requirement for the degree of
Master of Science
in
Physics

Waterloo, Ontario, Canada, 2007

c Jonathan Hackett 2007

I hereby declare that I am the sole author of this thesis. This is a true copy of
the thesis, including any required final revisions, as accepted by my examiners.
I understand that my thesis may be made electronically available to the public.

ii

Abstract
Two different works in mathematical physics are presented:
A construction of conformal infinity in null and spatial directions is constructed
for the Rainbow-flat space-time corresponding to doubly special relativity. From
this construction a definition of asymptotic DSRness is put forward which is compatible with the correspondence principle of Rainbow gravity. Furthermore a result
equating asymptotically flat space-times with asymptotically DSR spacetimes is
presented.
An overview of microlocality in braided ribbon networks is presented. Following this, a series of definitions are presented to explore the concept of microlocality
and the topology of ribbon networks. Isolated substructure of ribbon networks are
introduced, and a theorem is proven that allows them to be relocated. This is followed by a demonstration of microlocal translations. Additionally, an investigation
into macrolocality and the implications of invariants in braided ribbon networks
are presented.

iii

Acknowledgements
I would like to thank Lee Smolin, Fotini Markopoulou-Kalamara, Sundance
Bilson-Thompson and Louis Kauffman for their guidance and discussions. I would
also like to thank both the University of Waterloo and the Perimeter Institute for
Theoretical Physics for their support of this work.

iv

Dedication
This is dedicated to my fiancee Sonia and to my family.

Contents
1 Introduction

2 Asymptotic Flatness in Rainbow Gravity[1]

2.1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2.2

Rainbow Gravity . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2.3

Conformal infinity in Rainbow Minkowski space-times . . . . . . . .

2.4

Asymptotically DSR space-times . . . . . . . . . . . . . . . . . . .

2.5

Correspondence between asymptotic flatness and dsr . . . . . . . .

2.6

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

10

3 Locality and Translations in Braided Ribbon Networks[2]

12

3.1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

12

3.2

Braided Ribbon Networks . . . . . . . . . . . . . . . . . . . . . . .

13

3.3

Topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

14

3.4

Aevol and Microlocal Translations . . . . . . . . . . . . . . . . . . .

16

3.4.1

Isolated Substructures . . . . . . . . . . . . . . . . . . . . .

17

3.4.2

Microlocal Translations . . . . . . . . . . . . . . . . . . . . .

20

3.5

Macro, Micro and Braided Locality . . . . . . . . . . . . . . . . . .

21

3.6

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

22

vi

List of Figures
3.1

Two trinions with a scored ribbon . . . . . . . . . . . . . . . . . . .

13

3.2

Twists and Braidings . . . . . . . . . . . . . . . . . . . . . . . . . .

13

3.3

Generators of Aevol . . . . . . . . . . . . . . . . . . . . . . . . . . .

14

3.4

Proposed form of the first generation . . . . . . . . . . . . . . . . .

14

3.5

Networks that can be differentiated by T2 . . . . . . . . . . . . . . .

15

3.6

Networks that can be differentiated by T3 . . . . . . . . . . . . . . .

16

3.7

Transforming a capped braid into an isolated substructure . . . . .

17

3.8

Examples of Terrain . . . . . . . . . . . . . . . . . . . . . . . . . .

18

3.9

Microlocal Translations . . . . . . . . . . . . . . . . . . . . . . . . .

20

3.10 -closer Translations . . . . . . . . . . . . . . . . . . . . . . . . . .

21

vii

It seems likely that this process of increasing abstraction will continue

in the future and that advance in physics is to be associated with a
continual modification and generalisation of the axioms at the base of
the mathematics rather than with a logical development of any one
mathematical scheme on a fixed foundation.
The theoretical worker in the future will therefore have to proceed in a
more indirect way. The most powerful method of advance that can be
suggested at present is to employ all the resources of pure mathematics
in attempts to perfect and generalise the mathematical formalism that
forms the existing basis of theoretical physics, and after each success
in this direction, to try to interpret the new mathematical features in
terms of physical entities...
- Paul A.M. Dirac

Chapter 1
Introduction
Presented here are two very different forays into mathematical physics. Both of
the papers were developed within a different mathematical framework, and both
of the frameworks represents a different attempt at making progress in theoretical
physics. As these frameworks fall into rather different fields and so to prevent a
comprehensive unified introduction to both of them would be impossible. Instead,
an overview of the principles of mathematical physics are presented to allow the
reader to see the underlying principles that led to the research.
The two quotes presented before the introduction are from [3] and present a
scheme under which the work presented within falls. Though Diracs scheme is
by no means accepted as the foundation of modern mathematical physics, it is an
exceptionally insightful description of the dilemma faced by modern mathematical
physicists. There is a temptation to use the strength of modern mathematics to
push the already developed physical theories to new limits. The works presented
here are not examples of such vehicles, they are examples more in tune with Diracs
vision of the field.
The first paper presents work done on Rainbow Gravity, an extension of Doubly
Special Relativity to a general theory. Though Rainbow Gravity does not immediately present itself as the latest frontiers of mathematics, it is based upon work
in Doubly Special relativity and provides an important comparison point for the
work done in the larger field. Meanwhile, Doubly Special Relativity and its related
works are using recent advances in mathematics - such as quantum group theory
and non-commutative geometry - to form the foundations of their theories.
The other paper presents work done in Braided Ribbon Quantum Gravity. This
paper follows Diracs description of mathematical physics: it introduces new mathematics, and then proceeds to relate this mathematical advancement to a physical
phenomenon, that of translations of particles.

Chapter 2
Asymptotic Flatness in Rainbow
Gravity[1]
2.1

Introduction

The idea of a fundamental length scale has emerged in multiple approaches to quantum gravity. This length typically identified as the Planck length lp , is expected
to be the scale at which quantum gravitational corrections to our present theories
would be required. However, the idea that when probing below the Planck length
we will require new physics to describe the resultant phenomena is in direct contradiction with special relativity. How can we reconcile the idea of a fundamental
length scale when special relativity allows lengths to contract? This apparent paradox of quantum gravity is what gave the impetus behind the original work in doubly
special relativity (DSR) [4, 5, 6, 7]. Doubly special relativity fixes the planck length
and attempts to occupy the position of a flat space-time limit of quantum gravity.
Recent work on Doubly special relativity has been spurred on by current experiments that could provide a fertile testing ground for its results. Experiments
such as GLAST (The Gamma Ray Large Area Telescope) and AUGER (a cosmic
ray detector array) [8, 9] provide the opportunity to test the GZK cutoff and the
constancy of the speed of light, both of which are subjects which DSR is capable
of making predictions for. The other reason for the increased excitement in DSR
is the possibility of it providing increased insight into quantum gravity. The ingredients that lead to DSR are not dependent upon any particular attempt towards a
quantum theory of gravity and in fact are based solely upon attempting to combine
the ideas of special relativity and a fundamental length scale. Due to the simplicity
of its construction, it is possible that DSR could provide not only hints into the
structure of space-time in a complete theory of quantum gravity, but that it could
place restrictions upon one as well.
The picture of space-time resulting from DSR is in some ways still an open
question. Some approaches to DSR (particularly those that involve Hopf algebras)
3

have resulted in a picture of space-time which is non-commutative, these approaches

though interesting have yet to present any physical predictions, primarily due to the
difficulty in construction large-scale behaviour from non-commutative spacetimes.
Fortunately there is an alternative to non-commutative geometry as the arena in
which the space-time of DSR is understood, by attempting to extend DSR to general
relativity Magueijo and Smolin put forward Rainbow gravity. In Rainbow gravity,
the need for non-commutative geometry is avoided by introducing a metric which
runs with respect to the energy at which it is probed. Not only does this approach
avoid the need for non-commutative geometry to describe the space-time of DSR,
it allows DSR to be extended to a theory of gravity much like general relativity.
The intent of this paper is to explore the implications of Rainbow gravity further.
Of particular interest is the asymptotic behaviour of the theory as this is an area
where other approaches to DSR have not yet produced results. To this end we
will investigate the ideas of conformal infinity and asymptotic DSRness (in analogy
with asymptotically flat space-times in General relativity) in Rainbow gravity.

2.2

Rainbow Gravity

Rainbow gravity [10] is an attempt at construction an extension of DSR into a

general relativity framework which has at its foundation the proposal that the
geometry of a space-time runs with the energy scale at which the geometry is
being probed. The implication of this is that the metric of a space-time becomes
energy dependant. Fortunately a natural proposal for the manner in which the
metric could vary with respect to energy emerged from previous work on DSR [11],
which is the form that we shall use. In this sense we shall refer to as a rainbow
metric gab (E) the family of fundamental forms parameterized by the energy of the
particle probing it.
Rainbow gravity is governed by two principles, the correspondence principle and
the modified equivalence principle. These two principles are stated as follows [10]
Correspondence principle
In the limit of low energies relative to the Planck Energy, standard
general relativity is recovered. That is for any rainbow metric gab (E)
corresponding to a standard metric gab the following limit holds true.
lim gab (E) = gab

E
0
Epl

(2.1)

Modified Equivalance principle

Given a region of space-time with a radius of curvature R such that
1
R >> Epl

(2.2)

then freely falling observers measure particles and fields with energies E
observe the laws of physics to be the same as modified special relativity
to first order in R1 so long as:
1
<< E << Epl
R

(2.3)

where Epl is the Planck energy. Thus they can consider themselves to
be inertial observers in a rainbow flat space-time (to first order in R1 )
and use a family of energy dependent orthonormal frames locally given
by
E
)
e0
Epl
E
ei = g 1 (
)
ei
Epl

e0 = f 1 (

(2.4)
(2.5)

with a metric
g(E) = ab ea eb

(2.6)

We shall assume the existence of these two functions f (E) and g(E) which are
strictly greater than zero for small values of E; the small range of restriction is
to allow for the possibility that at significantly greater energies the geometry of
space-time could take on a significantly different character. The implications of
this assumption will be explored further in the context of asymptotic flatness.
One way to satisfy these principles is to require that the rainbow metric for any
space-time actually be a family of metrics given by energy-dependent orthonormal
frame fields - as presented above - which must satisfy a Rainbow Einstein equation
G (E) = 8G(E)T (E) + g (E)

(2.7)

where Newtons constant and the cosmological constant are now allowed to vary
with the energy so long as they obey the correspondence principle.
This is the form of Rainbow gravity which will be used to study the idea of
asymptotic DSRness in the following sections. It should be noted that conformal
mappings of rainbow gravity space-times pointwise with respect to the energy (at
specific energies instead of treating the energy as a dimension) are possible due to
the similarity between the Rainbow Einstein Equations and the original Einstein
equations. All Rainbow metrics are actually solutions to Einsteins equations in a
mathematical sense (treating the functions solely as mathematical concepts, instead
of allowing them to correspond to physical quantities) with the caveat that for
energies where G(E) varies from Newtons constant that the equation is slightly
modified, but not in a manner which would impact the behaviour of solutions under
conformal mappings, nor their compactifications.

2.3

Conformal infinity in Rainbow Minkowski spacetimes

In order to understand the asymptotic behaviour of space-times in Rainbow gravity

it is useful to have a consistent manner in which to evaluate quantities infinitely
far away in null and spatial directions. Additionally in order to be able to ascribe
the title of asymptotically DSR to a space-time we need to be able to identify
the behaviour of the flat DSR space-time at these asymptotic locations. To do
this we shall extend the ideas of I and 0 (the concepts of null and spatial infinity) from asymptotic flatness to DSR, proceeding in a similar manner to previous
demonstrations of these completions of minkowski space-time. [12]
We begin by considering the components of the metric of the flat DSR spacetime
(we shall call this the deformed Minkowski space-time or Rainbow space-time) gab
dr2
r2
dt2
+ 2
+ 2
d2
ds = 2
f (E) g (E) g (E)
2

(2.8)

Where d2 is the angular component of the spatial directions. By setting this equal
to zero (and likewise setting the angular component to zero) we are able to identify
the speed of light as a function of the energy:
c=

g(E)
dr
=
dt
f (E)

(2.9)

It is interesting to note that this will allow c to be non-constant. This allows us to

construct (energy dependent) null co-ordinates given by
f (E)
r
g(E)
f (E)
u=t
r
g(E)
v =t+

and change the metric accordingly to:



1
1
2
2
2
ds = 2
dvdu + (v u) d
f (E)
4

(2.10)
(2.11)

(2.12)

At this point we see that the only difference between this metric and the minkowski
space-time metric in null co-ordinates is a factor of f 2 1(E) . This means that we can
perform a conformal mapping of this space-time into a restriction of the Einstein
static universe by using a conformal factor of
2 =

f 2 (E) (1

4
+ v(E)2 ) (1 + u(E)2 )

(2.13)

such that the new metric gab is related to the old by

gab = 2 gab
6

(2.14)

The only restrictions being that

1
f 2 (E)

6= 0

(2.15)

and that
g(E) 6= 0

(2.16)

for all E. This mapping is made clear by choosing new co-ordinates of:
T (E) = tan1 (v(E)) + tan1 (u(E))
R(E) = tan1 (v(E)) tan1 (u(E))

(2.17)
(2.18)

The ranges of the new co-ordinates being

< T (E) + R(E) <
< T (E) R(E) <
0 <= R

(2.19)
(2.20)
(2.21)

and the components of the new metric being

2 = dT 2 + dR2 + sin2 (R)d2
ds

(2.22)

From here we are able to extend the original space-time to the boundary of the larger
space-time to yield an identification of the infinity of the deformed minkowski
space-time as follows:
Future Null infinity(I + ) is identified with T (E) = R(E) for
0<R<
Past Null infinity(I ) is identified with T (E) = + R(E) for 0 <
R<
Spatial infinity(0 ) is identified with R(E) = , T (E) = 0.
We therefore now have an identification of conformal infinity of the deformed
minkowski space-time with two reasonably physical restrictions given by equations
2.15 and 2.16. This allows us to examine asymptotic properties, and additionally
examine the concept of asymptotic DSRness in curved space-times.

2.4

Asymptotically DSR space-times

We now wish to build on the identification of I + , I , and 0 for deformed Minkowski

space-time by using them to define an asymptotically DSR space-time. To do this
we shall rely on the definitions of asymptotically flat space-times [12, 13, 14] and
expand them to incorporate the running metric of Rainbow gravity.
There is a concern however that as Rainbow gravity requires two as of yet
unknown functions - f (E) and g(E) - that we cannot assume that our definitions
of I + , I , and 0 hold for all values of E in the deformed minkowski space.
We therefore shall define an interval by
7

 is the interval (0, ), where is the smallest value greater than zero
such that when EEpl = , f (E) or g(E) fails to satisfy the restrictions
given by equations 2.15 and 2.16.
We will only address the concept of asymptotic DSRness where EEpl  as
our concepts of conformal infinity are ill defined outside of this interval. This
corresponds to the fact that the functions f (E) or g(E) tend towards zero as the
upper energy bound of DSR is reached. Such a restriction on the domain in which
we can define asymptotic structures is natural given the intent of Rainbow Gravity.
Within the interval , we shall require that for a space-time to be considered
asymptotically DSR at spatial infinity it must satisfy the following requirements
(analogous to those of asymptotic flatness conditions [13, 14])
Definition 1 A rainbow spacetime (M ,gab (E)) is considered asymptotically
DSR at spatial infinity if there exists a set of space-times defined by the parameter
(E),
E (M
gab (E)) where each space-time is smooth everywhere except at a point
0
is C 1 and gab is C 0 , and that there exists an imbedding of M (E)
(E) where M
satisfying:
into its respective M
(req. 1 ) The union of the closures of the causal future and causal past
(E), i.e.
of 0 (E) is equal to the complement of M (E) in M
M
J+ (i0 (E)) J (0 (E)) = M

(2.23)

that is C 2 at 0 (E) and

(req. 2 ) There exists a function on M
smooth everywhere else satisfying:
gab = 2 gab
(0 ) = 0
a (0 ) = 0

a
b (0 ) = 0

for all values of

E
Epl

(2.24)
(2.25)
(2.26)
(2.27)

 .

The motivation behind this definition is the desire for rainbow gravity to be
consistent at each value of E. Given any single value of E within all standard
rules of general relativity should apply and therefore the requirement for a spacetime
to be asymptotically DSR should be that it be able to be mapped to the deformed
minkowski space in a manner corresponding to the manner in which asymptotically
flat spacetimes are mapped to minkowski space through the Einstein static universe.
For a rainbow spacetime to be asymptotically DSR however, it must satisfy this
requirement at all energies within the interval however as we desire a definition
which is dependent upon the spacetime, not upon the specific energy at which it is
being probed. It should be noted that this requires that only spacetimes which are
asymptotically flat in the low energy limit can be asymptotically DSR.
8

Expanding this approach, we are able to expand the standard definition of an

Asymptotically flat and Empty spacetime as follows:
Definition 2 A rainbow spacetime (M ,gab (E)) is considered asymptotically
empty and DSR at null and spatial infinity if it is asymptotically DSR at spatial
infinity (as defined above) and satisfies
(req. 1 ) On the union of boundaries of the causal future and causal
a 6= 0 on the same.
past of 0 (E), = 0 and excepting 0 (E),
(req. 2 ) There exists a neighbourhood N (E) of the union of bound (E) such that
aries of the causal future and causal past of 0 (E) in M
(N (E), gab (E)) is strongly causal and time orientable, and in the in , Rab (E) = 0 (where
tersection of N (E) and the image of M (E) in M
Rab (E) corresponds to the original physical metric gab (E))
(req. 3 ) The map of null directions at 0 (E) into the space of inegral
a on I + (E) and I (E) is a diffeomorphism.
curves of na = gab
(req. 4 ) For a smooth function, , on the complement of 0 (E) in
with > 0 on the union of the image of M (E) in M (E)
with
M (E)
+

4
a
a ( n ) = 0 on the union of I + (E)
I (E) and I (E) which satisfies
and I (E) the vector field 1 na is complete on the union of I + (E) and
I (E).

for all values of

E
Epl

 .

These two definitions provide a means of identifying whether those concepts

that are defined through asymptotic behaviour (such as the energy of an isolated
system) carry over to Rainbow gravity in a consistent manner.

2.5

Correspondence between asymptotic flatness

and dsr

It is a natural question to inquire into whether an asymptotically flat space-time

will - when extended to rainbow gravity - correspond to an asymptotically dsr
space-time. This would seem to be likely, however it requires a proof in order to
justify this inclination.
Theorem 1 Given any Rainbow gravity metric of the form
ds2 =

A2 (t, r, ) 2 B 2 (t, r, ) 2 r2 C 2 (t, r, ) 2

dt +
dr +
d
f 2 (E)
g 2 (E)
g 2 (E)

(2.28)

where is shorthand for the solid angle of the two angular co-ordinates.
If the metric is asymptotically flat in the limit as EEpl goes to zero, the
rainbow gravity metric is asymptotically flat within .
9

Proof As the metric is in a form with a radial spatial co-ordinate we

can therefore find the speed of light by setting the distance and the
change in angular components to zero. We therefore get
dr
g(E)A(t, r, )
=
dt
f (E)B(t, r, )

(2.29)

as the speed of light, and can therefore choose Energy dependent null
co-ordinates u(E) and v(E) in the form of:
f (E)B(t, r, )
r
g(E)A(t, r, )
f (E)B(t, r, )
u=t
r
g(E)A(t, r, )
v =t+

(2.30)
(2.31)

This changes the metric to the form:

ds2 =

A2 (t, r, )
C 2 (t, r, )
dvdu
+
(v u)2 d2
f 2 (E)
4B 2 (t, r, )f 2 (E)

(2.32)

This is however just f 2 1(E) times the null metric for the original asymptotically flat space-time but with energy dependant null co-ordinates.
The fact that the co-ordinates are energy dependent is of no concern as
the definition of asymptotic DSRness is pointwise with respect to energy
and for each energy these co-ordinates are fixed. Likewise the functions
A(t, r, ),B(t, r, ), and C(t, r, ) are still constant with respect to energy as they were constant with respect to the original co-ordinates.
The construction of the conformal factor is identical to that used in
the derivation of conformal infinity in DSR in section 2.3. We therefore
find that we can construct a conformal factor 2 (E, t, r, ) from the
asymptotically flat space-times factor (denoted by 2 (t, r, )) by
2 (E, t, r, ) =

1
f 2 (E)

2 (t, r, )

(2.33)

which satisfies the requirements for the space to be asymptotically DSR.

2.6

Conclusion

The process of conformally mapping DSR into the Einstein static universe allowed
for a natural recreation of the definitions of the conformal infinities and allowed
the definitions for asymptotic DSR behaviour to follow from natural requirements.
These definitions, though highly stringent in their point-wise nature are put forward
as minimal (at least given the current knowledge of the functions f (E) and g(E))
should Rainbow gravity undergo further refinement it could be possible that these
10

restrictions could be relaxed based upon the running of the metric being well
behaved.
Though it is suspected that in general all asymptotically flat metrics will correspond to asymptotically DSR rainbow metrics, only a small result in that direction
is presented herein with further work on extending the class of metrics for which
this is true still in progress.

11

Chapter 3
Locality and Translations in
Braided Ribbon Networks[2]
3.1

Introduction

In the last century, there have been repeated discoveries of underlying structure.
Moving from macroscopic objects, to atoms, to components of the nuclei, to quarks,
it has been demonstrated repeatedly that the differences between supposedly fundamental particles are, in fact, merely consequences of the composite structure of
underlying reality. It only seems a natural progression that such an approach of
looking for underlying structure be used to explain the particles of the standard
model. Attempts towards this end, dubbed preon models,[16, 17, 18, 19, 20] met
with many obstacles, but still there was something deeper that presented itself as
a difficulty. The difficulty is that, as such a process does not have an end, we can
continue to suppose that below the currently understood structure is another set
of more fundamental particles. This idea quickly becomes unappealing at a philosophical level, or even a practical level, as the question then becomes What could
make it end?. The idea that the preons would be as fundamental as possible, such
as those in [22], provides a way of achieving the desired end. One way to achieve
this end is to suggest that the preons be composed of structure within space-time.
This suggestion gains further appeal by its convergence with recent approaches to
quantum gravity.
Such a preon model was recently proposed in [22] and then extended to the
idea of quantum gravity in [23]. The idea of having a composite model of particle
physics that is based upon topology in quantum gravity is appealing. The most
obvious basis for its appeal is that such a theory may be viewed as progress towards
a grand unified theory.
I shall investigate some features of this model and the topology of the structures
that it introduces. Based on this I will discuss the evolution algebra of this theory,
and demonstrate that translations of the large scale structures are a feature of the
theory.
12

3.2

Braided Ribbon Networks

The theory of braided ribbon networks [23] is concerned with two-dimensional surfaces in a compact 3-manifold. These surfaces are composed of the unions of trinions - intersections of three ribbons - and are scored to divide the surface into
clearly demarcated trinions (fig.3.1).

Figure 3.1: Two trinions with a scored ribbon

We allow the ribbons to be braided through the punctures in the surface, and
we also allow the ribbons to be twisted by multiples of 2 (fig.3.2). This network
evolves under Aevol , the algebra generated by the elements A1 , A2 and A3 (fig.3.3).
By viewing the trinions as nodes, we can consider the surfaces to be graphs. The
theory is then similar to loop quantum gravity in its structure, though with some
additional allowances for the labelings of the graph. We also note that a graph
can be changed to a ribbon graph by framing the edges of the graph: turning the
one-dimensional edges into two-dimensional surfaces.

Figure 3.2: Twists and Braidings

The reduced link of a graph is taken by treating each edge of the ribbons to
be a strand and then excluding unlinked unknotted strands. A subsystem is then
defined as a section of the graph where its reduced link does not intersect the rest
of the graph.
The first generation of the standard model is then proposed to be generated by
placing 2 twists on the strands of the two crossing capped braid of three ribbons
(fig.3.4), subject to the restrictions that all the twistings on a braid must be in the
same direction.
13

Figure 3.3: Generators of Aevol

Figure 3.4: Proposed form of the first generation

3.3

Topology

In order to properly discuss the idea of a translation we must first discuss the topology with respect to which the translations shall occur. Ribbon networks present
difficulties in this regard as there are several distinct classes of topologies. We begin
by considering the topology inherent in the idea of neighbours from a graph theory
perspective.
The Microlocal Metric Space
Consider a ribbon graph consisting of N nodes and M ribbons having
some braiding and twisting content. We construct a new metrical space
as follows: let X be the set of trinions within the ribbon graph. We

shall take each trinion x within X as a node in a pseudograph, and

construct edges for this pseudograph in the natural way: by making
an edge between two nodes if their respective trinions share a scored
ribbon. This is the reverse of the framing process that can be used to
construct a ribbon network.
The Microlocal Distance Function
Considering the set of all possible paths between two nodes
on the pseudograph, the distance between the nodes is the
minimum number of edges in any such path. This satisfies
the four requirements for a distance function: that it is nonnegative for any choice of two nodes, that it is strictly positive
14

for any two non-identical nodes, that it is reflexive and that

it satisfies the triangle inequality. This metric is equivalent
to the standard metric of graph theory.
Thus, the set X of nodes, along with the microlocal distance function,
create a metric space and, therefore, have a standard topology T1 defined
by the open balls given by the microlocal distance function on X. T1 is
thus the induced graph topology of the graph .
The microlocal topology surprisingly contains very little information about the
structure of the ribbon network. We should therefore consider topologies that
contain information about the braidings and twists of the ribbons.
Ribbon Topology
The Ribbon Topology is defined to be the topology corresponding to
taking the ribbon network as a bounded two dimensional surface with
a Euclidean metric. The Euclidean metric, together with the bounded
space of the ribbons, then becomes a metric space. Again, the open balls
generate the topology T2 . In contrast to T1 , T2 is able to differentiate
between graphs like those in figure 3.5.

Figure 3.5: Networks that can be differentiated by T2

Braided Ribbon Topology

The Braided Ribbon Topology is defined to be any topology of the ribbon
network that includes the braiding and the twisting of the ribbons. We
shall call this topology T3 . As the twisting is invisible to anything
living directly on the ribbons, this topology has to appeal to the higher
space that the structure is embedded within. T3 would then be able to
differentiate between graphs like those in figure 3.6. In the same way
that we have referred to the microlocal character of objects, we shall
refer to the braided-local character of objects.
If we consider T1 , T2 and T3 to be topologies on the original network , we can
see that each successive topology is finer than the last, with T3 being the finest.

15

Figure 3.6: Networks that can be differentiated by T3

3.4

Aevol and Microlocal Translations

We shall now demonstrate that there are indeed translations of braided-local structures with respect to the microlocal distance function. Also, we shall demonstrate
that even when using a more general notion of microlocal distance we can nonetheless demonstrate situations where braided-local structures have undergone a translation. These translations are generated by Aevol .
We shall first introduce a series of definitions and then prove a result using
them.
Ribbon Connected
Two nodes a and b are Ribbon Connected if there exists a sequence of
N + 1 nodes xn such that x1 = a, xN +1 = b and for each n the trinion
with node xn and the trinion with node xn+1 share a scored ribbon.

This is equivalent to the nodes being connected in the graph .

Connected Ribbon Network
A Connected Ribbon Network is a set of nodes X, such that all nodes
in X are ribbon connected to all other nodes in X.
Edge Segments
Consider the edges of a ribbon graph as a metric space E onto itself.
This space is essentially a collection of 1-d spaces which can be mapped
to the unit circle with the distance between two points being the minimum angle between the points on the unit circle. An Edge Segment is
then any connected subset of E with a non-empty interior. We consider
only sets of non-empty interior to avoid singleton sets that can produce
difficulties in later considerations.
Edge Connected
Two edge segments a and b are Edge Connected if they are connected
in the metric space E.

16

3.4.1

Isolated Substructures

In order to demonstrate translations within ribbon networks we must first define a

special class of elements within ribbon networks.
Isolated Substructure
An Isolated Substructure is a ribbon connected set of nodes where a
closed surface can be placed around it with exactly one ribbon intersecting the surface. We call this ribbon the Isolated Substructures
tether.
It should be understood that isolated substructures are not the same as subsystems as defined by [23]. This is readily apparent by considering the form of of
the reduced link of an isolated substructure.
It is interesting to note that, though the definition of an isolated substructure
appears to be restrictive at first glance, there are a significant number of structures
that can be packed up into the form of an isolated substructure. For instance,
all of the example definitions of particles from [23] can be changed into isolated
substructures through the use of exchange moves from Aevol as shown in figure 3.7.

Figure 3.7: Transforming a capped braid into an isolated substructure

Replaceable Edge Segments

An edge segment is Replaceable if it can be unambiguously replaced
by an isolated substructures tether. Specifically, a Replaceable Edge
Segment cannot be the edge of a node as this would cause four valent
nodes - such nodes are prohibited by our construction.
These definitions together allow us to consider the dynamics of isolated substructures under the generators of Aevol . We can consider a graph to be composed of
a set of isolated substructures attached to replaceable edge segments of a second
graph . As we do not require that all such isolated substructures be so removed,
this procedure can be done without ambiguity.
17

Theorem
Given a finite closed network with two edge connected replaceable
edge segments a and b, there exists a sequence of generators of Aevol
such that a graph a - composed of with an isolated substructure A
tethered to a - evolves to b , where b is composed of the same graph
but with A now tethered to b.
Proof
We shall proceed by induction on the number of nodes between a and
b, say N . As a and b are edge connected, the node created by A being
tethered to a is ribbon connected to the two nodes that are at either
side of b. We shall label these nodes x0 (for the node created by A at
a) through xN +1 in such a way that each xj shares a single ribbon with
xj+1 . The nodes on either side of b are then labeled xN and xN +1 .
Before we perform this induction, we need to show the ability to move an
isolated substructure through intermediate topological structures that
are not composed of nodes. These are comprised of three categories:
knots, twists and braidings. Examples of each of these is shown in
figure 3.8. As isolated substructures only have a single connection to
the outside network, we can move it past this terrain through the
following procedures.

Figure 3.8: Examples of Terrain

For each knot the isolated substructure is pulled through the knot by
stretching out the knot until the substructure can pass through it. As
the substructure is unconnected except through its tether, this leaves the
network unchanged other than the reversal of the position of the knot
and the substructure.
For each twist (consisting of a rotation by ), the isolated substructure
can run along the edge of the twist. Alternatively, one can view the
18

procedure as deforming the network itself by twisting the segment with

the isolated substructure in a manner that undoes the twist on the one
side and create a twist on the other.
For each braiding, we deform the ribbon of the braid by sliding it over
the isolated substructure to the other side. This is reminiscent of the
Reidemeister move of the second kind.
Our lemma is that we can move an isolated substructure tethered to
a node R so that it is tethered with no intermediate terrain between
it and some edge segment t (which is not a component of a piece of
terrain) that connects the node R to its nearest neighbours and is
edge connected to the edge which the isolated substructure is tethered
to. This is proven by induction on the number of elements of terrain
between R and t and the use of the above prescriptions. A consequence
of this is that the same method can be used for a node S which has
microlocal distance 1 to R. This ability shall be used heavily in our
proof.
Now, returning to the proof, we shall first prove the case of N = 1.
We apply the above lemma to move the isolated substructure through
any intermediate terrain between x0 and x1 , giving us A tethered to a
new node x00 (we shall use primes to denote nodes that have undergone
some change) that is immediately adjacent to x1 with no intermediate
terrain. We then perform an exchange move from Aevol on the node x00
and x1 to move x00 onto the edge on the other side of x01 . We can then
again use the above lemma (in its more general case) to move x00 to its
final resting place at b.
Now we shall assume that the case of N 1 nodes is correct and prove
the case of N nodes. The prescription for this is analogous to the N = 1
case. Given that there is a method for moving past N 1 nodes (by
inductive hypothesis), we shall use that method to change the situation
to a single intermediate node, and then invoke the method of the N = 1
case to bypass the final node.
The preceding gives the inductive argument and completes the proof.
To demonstrate translations we will need a further tool. We therefore consider
also the following lemma:
Lemma Translations Through an Isolated Substructures
Given an isolated substructure A that has been moved to the edge of
the tether of another substructure B, it is possible to translate A to the
opposite edge of the tether of B.
Proof
Due to the above theorem, it only remains to show that the two edges of
19

the tether of an isolated substructure are edge connected. Proceeding

by contradiction, we assume that they are not edge connected. As
we see that an edge of the network enters the isolated substructure
and does not exit, there must be some terminus of the edge within
the isolated substructure. However, such a situation is impossible, as
the edges of a ribbon network must form closed links or terminate at
some boundary (which we have not introduced into the theory of ribbon
networks). We therefore have a contradiction. Thusly we see that if an
isolated substructure can be moved to the edge of a tether, by the above
theorem, it can be moved to the other edge.

3.4.2

Microlocal Translations

The application of the theorem is straightforward and results in the ability to

demonstrate translations under the microlocal distance function. For instance, it is
possible to construct a sequence of moves of Aevol such that figure 3.9a evolves to
figure 3.9b. Under the microlocal distance function, the isolated substructure A is
now less distant from the isolated substructure C (measuring the distance between
substructures from the node at which they are tethered).

(a)

(b)

Figure 3.9: Microlocal Translations

Even applying a more restrictive definition of distances, we can demonstrate
translations in some form. Consider the following definition of closeness:
-closer
An isolated substructure A is said to be -closer to an isolated substructure B than it is to another substructure C, if for all paths along
the ribbons of the network, leaving the node at which A is tethered
and intersecting the node at which C is tethered the path intersects the
node at which B is tethered.
By expanding our definitions slightly to allow us to consider isolated substructures with identical structure to be treated equally, we can show that it is possible to
evoke a translation. Specifically, it is possible to take a situation where a substructure A is -closer to substructures of type B than to those of type C and to apply
a series of moves of Aevol such that the reverse is true afterwards. For instance,
consider figure 3.10a and figure 3.10b. Thus we see that we have translations even
under stringent requirements, thereby concluding our result.
20

(a)

(b)

Figure 3.10: -closer Translations

3.5

Macro, Micro and Braided Locality

The braided-local structure of the braid network is characterized by the reduced

link of the structure. The reduced link of a braid network can be shown to be
invariant under the generators of Aevol , by applying the definition of the reduced
link to the graphical representations of the generators. As a result, it is clear that
the braided-local content of a braid network is invariant under the generators of
Aevol . This invariance is a double edged sword. On one hand, it allows us to assign
some meaning to these invariant structures, as was done by the authors of [23].
On the other hand, it means that there is no way that these microlocal moves
can provide any form of dynamics in this content. As a result, I suggest that, to
construct a theory of quantum gravity containing particle physics from a ribbon
network, it is necessary to consider the existence of a second evolution algebra,
which I shall call Abraid .
In [24] and [25], the concept of macrolocality in networks is put forward as the
locality derived from the classical metric that would arise for a network with a spacetime as its classical limit. In [24], the authors then remind us that there is no need
for macrolocality to be coincidental to microlocality. It seems to be a consequence
of the ideas of [23] to suggest that, though microlocality and macrolocality are not
necessarily coincidental, the braided-local content - the invariants that we associate
with particles - should be part of the bridge of the gap between the two. I therefore
suggest that Abraid could be the bridge between microlocality and macrolocality.
For future consideration, I outline some general possibilities of Abraid . Regardless, it should be noted that any such algebra that could provide macrolocal dynamics, particle interactions included, would need to alter the reduced link of the
network if the identifications in [23] are to be considered seriously.
Nearly Microlocal Algebra
A candidate based upon the assumption that any move within the second evolution algebra should be as close to being microlocal as possible
is called a Nearly Microlocal Algebra. This could be completed by introducing moves involving next to nearest neighbor nodes. This suggestion
corresponds to the idea that there is a degree of coincidence between
21

microlocality and macrolocality (again, we should remember that such

a coincidence is not needed).[24]
Braid Algebra
An algebra based upon moves that alter the braiding content of the
network in ways that are roughly equivalent to elements of the standard
braid group is referred to as a Braid Algebra. Also, it can contain moves
that allow the composition of multiple braided isolated substructures.
Anti-Microlocal Algebra
An algebra premised upon the idea that microlocality should be dual
or completely unrelated to macrolocality is called an Anti-Microlocal
Algebra. Such an algebra can be constructed from a set of moves that
act upon the reduced links of a graph. Such moves could be realized
through the following algorithm:
Take the reduced link of the graph and apply a move
that composes or interacts parts of the reduced link (whether
through cutting and repairing links, or through allowing links
that correspond in some manner to annihilate each other).
Then take the new reduced link and equate it with a superposition of all graphs x 0. That any such graph x 0 should
exist should be provable by a generalization of the theorem
that allows the construction of a closed braid that corresponds
to any link. [26]
It is possible that a stronger candidate would draw upon multiple such programs.

3.6

Conclusion

The above results give rise to several key points. First, the results are restricted to
isolated substructures, without which it is impossible to bypass the terrain within
the network. Second, the definitions in the previous section may not necessarily
apply if labels are introduced to the network. Despite these restrictions, the result
remains promising and integral to attempts to attempts to develop Ribbon networks
into a theory of quantum gravity with matter. The primary candidates for the
fundamental particles within such a theory are all examples of systems that can
be made into isolated substructures. Indeed the form of the fundamental particles
was the motivation for demonstrating translations.
The demonstration of these translations provides great promise in further developing this model into a theory that involves particle dynamics. However several
key obstacles remain. As discussed in section 3.5, without adding more structure,
in the form of a a second evolution algebra (or at the least, expanding the original
22

evolution algebra), it is impossible to have any particle interactions. Indeed, even

the case that one might expect to be easiest to demonstrate - that of particle and
anti-particle annihilating one another - is impossible without some modification.
Developing candidates for Abraid remains the subject of ongoing work.

23

Bibliography
[1] J. Hackett, Class. Quant. Grav. 23, 3833 (2006) [arXiv:gr-qc/0509103].
[2] J. Hackett, arXiv:hep-th/0702198.
[3] P. Dirac Quantised Singularities in the Electromagnetic Field, Proceedings of
the Royal Society of London. Series A, Vol. 133, No. 821. (Sept. 1, 1931),
pp.60-72.
[4] J. Kowalski-Glikman, Introduction to Doubly Special Relativity,
th/0405273, (2004) submitted to Lecture Notes in Physics.

hep-

[5] G. Amelino-Camelia Status of Relativity with observer-independent length and

velocity scales in New Developements in Fundamental Interaction Theories:
37th Karpacz Winterschool of Theoretical Physics ed J. Lukierski, (2001)
(American Institute of Physics) gr-qc/0106004.
[6] G. Amelino-Camelia Testable scenario for Relativity with minimum-length
Phys.Lett. B510 (2001) 255-263, hep-th/0012238.
[7] G. Amelino-Camelia Relativity in Space-times with short-distance structure
governed by an observer-independent (Planckian) length scale Int.J.Mod.Phys.
D11 (2002) 35-60, gr-qc/0012051.
[8] A. de Angelis, Glast, the Gamma-ray Large Area Space Telescope, in New
worlds in Astroparticle physics, Proceedings of the Third International Workshop ed. A. Mouro et al, 2001 (World Scientific), astro-ph/0009271.
[9] N. Hayashida, K. Honda, N. Inoue et. al. Updated Agasa event list above 4109
eV Astrophys. J. 522 (1999) 255, astro-ph/0008102.
[10] L. Smolin and J. Magueijo, Gravitys Rainbow Class.Quant.Grav. 21 (2004)
1725-1736, gr-qc/0305055.
[11] J. Magueijo and L. Smolin, Lorentz invariance with an invariant energy scale,
Phys.Rev.Lett. 88 (2002), hep-th/0112090.
[12] R. Wald General Relativity, 1984. (Chicago, The university of Chicago Press)

24

[13] A. Ashtekar and R.O. Hansen A unified Treatment of Null and Spatial Infinity in General Relatvity. I. Universal Structure, Asymptotic Symmetries, and
Conserved Quantities at Spatial Infinity J. Math. Phys. 19 1542-1566, 1978
[14] A. Ashtekar Asymptotic Structure of the Gravitational Field at Spatial Infinity
in General Relativity and Gravitation vol. 2 ed. A. Held, 1980. (New York:
Plenum).
[15] L. Smolin and J. Magueijo, Generalized Lorentz invariance with an invariant
energy scale Phys.Rev. D 67 (2003) 044017. hep-th/0207085
[16] J. C. Pati and A. Salam, Lepton Number As The Fourth Color, Phys. Rev.
D 10, 275 (1974).
[17] H. Terazawa and K. Akama, Dynamical Subquark Model Of Pregauge And
Pregeometric Interactions, Phys. Lett. B 96, 276 (1980).
[18] H. Harari, A Schematic Model Of Quarks And Leptons, Phys. Lett. B 86,
83 (1979).
[19] M. A. Shupe, A Composite Model Of Leptons And Quarks, Phys. Lett. B
86, 87 (1979).
[20] H. Harari and N. Seiberg, A Dynamical Theory For The Rishon Model,
Phys. Lett. B 98, 269 (1981).
[21] S. O. Bilson-Thompson, A topological model of composite preons, arXiv:hepph/0503213.
[22] S. O. Bilson-Thompson, A topological model of composite preons arXiv:hepph/0503213.
[23] S. O. Bilson-Thompson, F. Markopoulou and L. Smolin, Quantum gravity
and the standard model arXiv:hep-th/0603022.
[24] F. Markopoulou and L. Smolin, Disordered locality in loop quantum gravity
states, arXiv:gr-qc/0702044.
[25] F. Markopoulou, Towards gravity from the quantum, arXiv:hep-th/0604120.
Expanded version of contribution to book Towards Quantum Gravity. Edited
by D. Oriti. Cambridge University Press, 2006.
[26] J.W. Alexander A lemma on systems o knotted curves Proc. Nat. Acad.
Science USA 9 (1923), 93-95.

25