symmetry
S S
Article
Space, Matter and Interactions in a Quantum Early Universe.
Part II: Superalgebras and Vertex Algebras
Piero Truini 1,2, Alessio Marrani 3,*
, Michael Rios 2 and Klee Irwin 2






Citation: Truini, P.; Marrani, A.; Rios,
M.; Irwin, K. Space, Matter and
Interactions in a Quantum Early
Universe. Part II: Superalgebras and
Vertex Algebras. Symmetry 2021, 13,
2289. https://doi.org/10.3390/
sym13122289
Academic Editors: Lucrezia Ravera
and Maxim Yu. Khlopov
Received: 13 September 2021
Accepted: 25 November 2021
Published: 1 December 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
1
Istituto Nazionale Fisica Nucleare (INFN), Sezione di Genova, Via Dodecaneso 33, I-16146 Genova, Italy;
piero.truini@ge.infn.it
2 Quantum Gravity Research (QGR), 101 S. Topanga Canyon Rd., Los Angeles, CA 90290, USA;
mrios@dyonicatech.com (M.R.); Klee@quantumgravityresearch.org (K.I.)
3 Centro Studi e Ricerche Enrico Fermi, Via Panisperna 89A, I-00184 Roma, Italy
* Correspondence: jazzphyzz@gmail.com
Abstract: In our investigation on quantum gravity, we introduce an infinite dimensional complex Lie
algebra gu that extends e9. It is defined through a symmetric Cartan matrix of a rank 12 Borcherds
algebra. We turn gu into a Lie superalgebra sgu with no superpartners, in order to comply with
the Pauli exclusion principle. There is a natural action of the Poincaré group on sgu, which is an
automorphism in the massive sector. We introduce a mechanism for scattering that includes decays
as particular resonant scattering. Finally, we complete the model by merging the local sgu into a
vertex-type algebra.
Keywords: Kac-Moody algebras; Borcherds algebras; early Universe cosmology; vertex algebras
1. Introduction
This is the second of two papers—see also [1]—describing an algebraic model of
quantum gravity.
Our guiding idea was to exploit the most fundamental principles of quantum mechan-
ics and general relativity as we believe they should apply in the extreme conditions of a
hot dense universe in its early stages.
We started from two very intuitive physical principles:
•
there is no classical observable to be quantized: one has to think directly in terms of
quantum objects and states of the system;
•
there is no spacetime geometry to start with.
As to the first point, the model departs from the conventional view that quantum
gravity ought to be realized as the quantization of gravity with its renormalization, with
the four fundamental forces unifying at the Planck scale. We are at the Plank scale: the
dynamics only depends on the quantum charges of the constituents and they are indeed
symmetrical. Gravity is identified with the way spacetime is created and evolves. It is
quantum because spacetime is created via quantum interactions, producing curvature, and
quantum expansion, as they both occur with complex probability amplitudes.
Although everything looks symmetrical, diversity comes from quantum theory itself,
and from the initial conditions.
As to the second point, the absence of spacetime to start with leaves us with the
sole interactions that we assume to be tree-like, as physical observations at very high
temperature suggest. The basic blocks involve therefore three objects, and this has led us
to consider algebraic models, with the algebra product playing the role of the building
blocks of all interactions. A mechanism for the quantum creation of spacetime suggested
the inclusion of momenta within the charges (roots) of the algebra, thus achieving charge
and energy-momentum conservation as well. The evolution of the universe, its quantum
interactions and quantum expansion from a chosen initial state of a finite set of generators,
Symmetry 2021, 13, 2289. https://doi.org/10.3390/sym13122289
https://www.mdpi.com/journal/symmetry
Symmetry 2021, 13, 2289
2 of 14
can be turned into an algorithm in which all physical quantities are, in principle, calculable.
We would need a huge powerful computer to do that, but this is not the point of our two
papers. Our point is: trying to think outside the box by going back to the basic physical
principles, and by building a consistent model that can accommodate the physics we know.
In this respect, we would like to emphasize that our main object is a superalgebra.
Again we were led to it by fundamental physics, the Pauli principle in particular. The way
we define it does not leave room for superpartners. Superparticles have not been observed
and we can describe all the physical interactions without any need of supersymmetry for
renormalization purposes: our theory is finite.
It also turns out, by construction, that spacetime in our model is discrete, simply
because, if we start, at say t = 0, with a finite set of generators, the number of interactions
will stay finite at t = 1, 2, ... and spacetime will be created in finite chunks, with certain
probability amplitudes, see [1]. What we have denoted here by t is an order parameter
for the succession of interactions, and is interpreted as universal time. Nevertheless we
show that there is a natural action of the Poincaré group in our model. This is crucial for its
relationship with the concept of spin and with relativity. Since the model has an intrinsic
notion of universal time, how do we compare future observers measurements? Where
is relativity of time and space? If a particle has a certain intrinsic momentum included
in a root of the algebra, how do we describe an interaction in the rest frame of a massive
particle? These are crucial questions coming from the physics we know and experiment.
To answer these questions we show that a local action of the Poincaré group is defined and
has the right properties as a Wigner representation, even though spacetime is discrete.
As a final remark on this preliminary conceptual outline, let us focus on the interac-
tions. Their building blocks involve three objects and are mathematically described by
the multiplication law of the algebra. A generator in the algebra is related to a particle,
with certain charges coming from the algebra roots, but it is also related to a quantum field,
since new generators are produced by multiplication in the algebra: a generator expands
in spacetime with complex quantum amplitudes but locally interacts, and disappears
and contributes to the creation of new generators. This local action can be considered as
a vertex, made of generators obeying the rules of an algebra. There is no vacuum since
space points exist only where generators are. We see therefore that we do have an algebra
at the core of the model, that we denote sgu, but the expansion of spacetime embeds it
into a larger picture: that of a vertex-type algebra describing quantum interactions and a
quantum generated spacetime.
We believe the above considerations are plausible and strongly based on fundamental
physics. Their concrete, calculable realization is what we have achieved with our model,
which has no claim other than being physically consistent and mathematically rigorous.
In the first paper we have described the basic principles of our model and we have
investigated the mathematical structures that may suit our purpose. In particular, we have
focused on rank-12 infinite dimensional Kac-Moody, [2], and Borcherds algebras, [3,4], and
we have given physical and mathematical reasons why the latter are preferable.
In our model for the expansion of quantum early Universe, [1,5], the need for an
infinite dimensional Lie algebra stems from the unlimited number of possible 4-momenta,
but at each fixed cosmological time the number of generators and roots involved is finite.
There is a known algorithm of Lie algebra theory that allows to determine the structure
constants among a finite number of generators of a Borcherds algebra [3,4]. Let us grade
the commutators by levels, by saying that the commutators involving n simple roots have
level n− 1. A consistent set of structure constants is calculable level by level, and once the
structure constants are calculated at level n, they will not be affected by the calculation
at any level m > n. There are computer programs that apply this algorithm and give the
explicit structure constants level by level, see for instance the package LieRing of GAP,
developed by S. Cicalò and W. A. de Graaf [6].
However, for the sake of simplicity, in [1] we have chosen to deal with a simpler
Lie algebra, gu, that extends e8 and e9. In the present paper, we will start investigating
Symmetry 2021, 13, 2289
3 of 14
a physical model for quantum gravity based on this particular rank-12 algebra gu. We
will start by focussing on local aspects of the algebraic model: in Section 2, we recall gu,
which is then turned into a Lie superalgebra sgu in Section 3. Section 4 will then discuss
interactions, scattering processes and decays, whereas the role of the Poincaré group is
analyzed in Section 5. Finally, in Sections 6 and 7 we will define the quantum states, and
then we will merge the algebra sgu into a vertex-type algebra, representing the quantum
early Universe with its expanding spacetime.
2. The Lie Algebra gu
We start and consider B+, the Lie subalgebra of the rank-12 Borcherds algebra B12
introduced in [1] and generated by the Chevalley generators corresponding to positive
roots. A further simplification will then give rise to gu, the Lie algebra that acts locally on
the quantum state of the Universe [1].
We recall from [1] that the generalized Cartan matrix for the Borcherds algebra B12,
with simple roots denoted by α−1, α0′′ , α0′ , α0, ..., α8, is

−1 −1 −1 −1
0
0
0
0
0
0
0
0
−1
0 −1 −1
0
0
0
0
0
0
0
0
−1 −1
0 −1
0
0
0
0
0
0
0
0
−1 −1 −1
2 −1
0
0
0
0
0
0
0
0
0
0 −1
2 −1
0
0
0
0
0
0
0
0
0
0 −1
2 −1
0
0
0
0
0
0
0
0
0
0 −1
2 −1
0
0
0
0
0
0
0
0
0
0 −1
2 −1
0
0
0
0
0
0
0
0
0
0 −1
2 −1 −1
0
0
0
0
0
0
0
0
0 −1
2
0
0
0
0
0
0
0
0
0
0 −1
0
2 −1
0
0
0
0
0
0
0
0
0
0 −1
2

(1)
By defining
δ := α0 + 2α1 + 3α2 + 4α3 + 5α4 + 6α5 + 3α6 + 4α7 + 2α8,
(2)
the 4-momentum vector can be written as
p := Epα-1 + px(α0′′ − α-1) + py(α0′ − α-1) + pz(δ− α-1).
(3)
Then, we restrict to the subalgebra B+ of B12, namely to positive roots r = ∑I λiαi,
I := {−1, 0′′, 0′, 0, ..., 8}, with λi ∈ N∪ {0}. Consequently, the 4-momentum (3) becomes
p = (Ep, px, py, pz) = (λ−1 + λ0′′ + λ0′ + λ0,λ0′′,λ0′,λ0)
(4)
with λ−1,λ0′′ ,λ0′ ,λ0 ≥ 0, implying
m2 := −p2 ≥ 0,
(5)
namely p either lightlike or timelike. In particular (i, j ∈ {−1, 0′′, 0′, 0}),
p2 = −
(
λ2−1 + 2λ−1 ∑i
6=−1 λi +∑i
6=j, i,j
6=−1 λiλj
)

= 0
if λ−1 = 0 and at most one λi
6= 0, i
6= −1,
= −1
if λ−1 = 1 and all λi = 0, i
6= −1,
6 −2 otherwise.
(6)
As in [1], we write a root r = ∑I λiαi as
r = α+ p,
(7)
Symmetry 2021, 13, 2289
4 of 14
with
Φ8 3 α = (λ1 − 2λ0)α1 + (λ2 − 3λ0)α2 + (λ3 − 4λ0)α3
+(λ4 − 5λ0)α4 + (λ5 − 6λ0)α5 + (λ6 − 3λ0)α6
+(λ7 − 4λ0)α7 + (λ8 − 2λ0)α8,
(8)
and p given by (4).
Remark 1. Notice that the mass of a particle cannot be arbitrary small, since there is a lower limit,
m ≥ 1.
Hence, we extend the possible values of the 4-momentum p := (E,~p) by including
those with opposite 3-momentump̃ = (E,−~p), as explained in [1], so that
p = (Ep, px, py, pz) , Ep ∈ N , px, py, pz ∈ Z , p2 ≤ 0.
(9)
The algebra gu extends the 1 + 1-dimensional toy model based on e9 discussed in [1];
it is defined as the algebra generated by xαp and xα+p, such that p2 ≤ 0, satisfying the
following commutation relations:
[
xαp1 , x
β
p2
]
= 0,
[
xαp1 , xβ+p2
]
= (α, β)xβ+p1+p2 ,
[
xα+p1 , xβ+p2
]
=

0,
if α+ β /∈ Φ8 ∪ {0};
ε(α, β)xα+β+p1+p2 ,
if α+ β ∈ Φ8;
−xαp1+p2 ,
if α+ β = 0,
(10)
where (·, ·) is the Euclidean scalar product in R8, the function ε : Φ8 ×Φ8 → {−1, 1} is the
asymmetry function [1,2,7], and
x−α
p = −xαp,
[
xαp1 , xβ+p2
]
= −
[
xβ+p2 , x
α
p1
]
,
(11)
in order to have an antisymmetric algebra.
Moreover, for consistency, we require that
xα+β
p = xαp + x
β
p .
(12)
Notice that p21, p
2
2 6 0 implies (p1 + p2)
2 6 0.
Remark 2. Notice also that p21, p
2
2 < 0 imply (p1 + p2)
2 < 0. Thus, there is a subalgebra g+u of
gu with the same commutation relations (10), but with generators xα+p and xαp such that p2 < 0
(only massive particles).
Proposition 1. The algebra gu with relations (10)–(12) is an infinite-dimensional Lie algebra.
The proof is in Appendix A.
The algebra gu has a natural 2-grading inherited by that of e8, due to the decomposition
into the subalgebra d8 and its Weyl spinor, [1]. The generators xα+p are fermionic (resp.
bosonic) if α is fermionic (resp. bosonic), whereas the generators xαp are bosonic, due to the
commutation relations (10).
3. The Lie Superalgebra sgu
In order to turn the Lie algebra gu into a Lie superalgebra, we exploit the Grassmann
envelope G(gu) of gu,
G(gu) := gu0 ⊗G0 + gu1 ⊗G1,
(13)
Symmetry 2021, 13, 2289
5 of 14
where gu0 is the boson subalgebra of gu, gu1 its fermionic part, and G0, G1 are the even, odd
parts of a Grassmann algebra with infinitely many generators. More precisely, we map
each generator X of gu to the generator X⊗ ex of G(gu), where ex is even if X is bosonic,
odd if X is fermionic, and ex
6= ey if X
6= Y. Then the graded Jacobi identity is satisfied, [8],
and one obtains, by linearity, a Lie superalgebra, that we denote by sgu.
Let us show this straightforward calculation explicitly.
Let X, Y, Z be generators of gu of degree i, j, k ∈ {0, 1} respectively. We remind,
from [1], that the generators xαp have degree 0, whereas the generators xα+p have degree
bαc = 0 if α is bosonic and degree bαc = 1 if α is fermionic.
Let [X, Y] still denote the product of X, Y in gu and X⊗ ex ◦Y⊗ ey the corresponding
product in sgu. Then:
X⊗ ex ◦Y⊗ ey = [X, Y]⊗ exey = −[Y, X]⊗ exey = −(−1)ij[Y, X]⊗ eyex
= −(−1)ijY⊗ ey ◦ X⊗ ex
(14)
and the graded Jacobi identity is satisfied:
(−1)ik((X⊗ ex ◦Y⊗ ey) ◦ Z⊗ ez) + (−1)jk((Z⊗ ez ◦ X⊗ ex) ◦Y⊗ ey)+
(−1)ij((Y⊗ ey ◦ Z⊗ ez) ◦ X⊗ ex)
= (−1)ik[[X, Y], Z]⊗ exeyez + (−1)jk[[Z, X], Y]⊗ ezexey+
(−1)ij[[Y, Z], X]⊗ eyezex
=
(
(−1)ik[[X, Y], Z] + (−1)jk(−1)k(i+j)[[Z, X], Y]+
(−1)ij(−1)i(j+k)[[Y, Z], X]
)
⊗ exeyez
= (−1)ik J⊗ exeyez = 0
(15)
where J = 0 is the Jacobi identity for gu.
Remark 3. The product in the Lie superalgebra sgu is effectively the same as in the Lie algebra gu
but its symmetry property is crucial for the elements of the universal enveloping algebra, that appear
point by point in the model for the expanding Universe. The universal enveloping algebra is indeed
the tensor algebra 1 ⊕ sgu ⊕ sgu ⊗ sgu ⊕ ... modulo the relations x⊗y− (−1)ijy⊗ x = x ◦ y for
all x, y ∈ sgu, embedded in the tensor algebra, of degree i, j respectively. In particular this makes the
fermions comply with the Pauli exclusion principle: x⊗ x = 0, for x fermionic, whereas the same
relation is trivial, 0 = 0, if x is bosonic.
Remark 4. The Lie superalgebra sgu does not involve superpartners. The elements are exactly
the same as those of the algebra gu. The importance we attribute to this algebra is solely due to the
fulfillment of the Pauli exclusion principle.
Remark 5. We also notice that the use of the Grassmann envelope produces zero divisors in the
algebra whenever the same fermionic root, with the same momentum, is in two interacting
particles. This is the precise mathematical statement we need in order to apply the previous two
remarks to our model. The Pauli principle is therefore fulfilled.
4. Interaction Graphs
As mentioned in the Introduction of [1], the interactions have a tree structure whose
building blocks involve only three particles, and they are expressed by the product in the
underlying algebra. The scattering amplitudes are proportional, up to normalization, to the
structure constants of the related products. An ordering of the roots has to be a priori set, so
that the commutator between two generators is taken according to that order. Quantum
interference is obviously independent from the ordering choice.
Symmetry 2021, 13, 2289
6 of 14
In this section, we set up a correspondence between graphs and products in the algebra
gu spanned by the generators
{xαp , xα+p ; α ∈ Φ8 , p = (E,~p), p2 ≤ 0},
(16)
and the procedure can then be trivially extended to sgu by Remark 3.
We include the decays among the possible scatterings as resonance interactions, a well
known and studied phenomenon in many physical processes, as we now explain. Suppose
that two particles, one with charge α ∈ Φ8 ∪ {0} and momentum p1, the other with
charge β ∈ Φ8 ∪ {0} and momentum p2, are present at the same space point and are
such that α− β ∈ Φ8 ∪ {0} and E1 > E2; then, a decay occurs, with a certain amplitude,
producing the outgoing particles of charges (α− β), β with momenta (p1 − p2) and p2
respectively, whereas the particle with charge β ∈ Φ8 ∪ {0} and momentum p2 shifts in
space according to the expansion rule, see [1]. The amplitude for the decay is proportional,
up to normalization, to the structure constant of the commutator between the outgoing
particles (we will comment on this viewpoint on the decays at the end of this section).
The possible situations for an elementary interaction are depicted in the following
graphs, given by Figures 1–3 (the resonant particle is also shown in case of a decay). In the
graphs we use wiggly lines for the neutral particles xαp and straight lines for the charged
particles xα+p. Red lines indicate outgoing particles and blue lines incoming ones.
We would like to stress that the orientation of the graphs is not significant; these are not
Feynmann diagrams, although they resemble them: only the distinction between incoming
and outgoing particles matters; it complies with 4-momentum and charge conservation.
according to that order. Quantum interference is obviously independent from the
ordering choice.
In this section, we set up a correspondence between graphs and products in the
algebra gu spanned by the generators
{xαp , xα+p ; α ∈ Φ8 , p = (E, ~p), p2 ≤ 0},
(4.1)
and the procedure can then be trivially extended to sgu by Remark 3.1.
We include the decays among the possible scatterings as resonance interactions, a
well known and studied phenomenon in many physical processes, as we now explain.
Suppose that two particles, one with charge α ∈ Φ8 ∪ {0} and momentum p1,
the other with charge β ∈ Φ8 ∪ {0} and momentum p2, are present at the same
space point and are such that α − β ∈ Φ8 ∪ {0} and E1 > E2; then, a decay
occurs, with a certain amplitude, producing the outgoing particles of charges (α−β),
β with momenta (p1 − p2) and p2 respectively, whereas the particle with charge
β ∈ Φ8∪{0} and momentum p2 shifts in space according to the expansion rule, see [1].
The amplitude for the decay is proportional, up o ormalization, to the structure
constant of the commutator between the outgoing particles (we will comment on
this viewpoint on the decays at the end of this section).
The possible situations for an elementary interaction are depicted in the following
graphs, Figs. 1 ÷ 3 (the resonant particle is also shown in case of a decay). In
the graphs we use wiggly lines for the neutral particles xαp and straight lines for the
charged particles xα+p. Red lines indicate outgoing particles and blue lines incoming
ones.
We would like to stress that the orientation of the graphs is not significant; thes
are not Feynmann diagrams, although they resemble them: only the distinction
between incoming and outgoing particles matters; it complies with 4-momentum
and charge conservation.
xβ+p2
xαp1
xβ+p1+p2
(a)
xβ+p1+p2
xβ+p2
xαp1
xβ+p2
shifted
(b)
Figure 1: [xαp1 , xβ+p2 ] = (α, β)xβ+p1+p2 ; (a): x
α absorption by xβ; (b): x
α emission
by xβ (similarly for x
α
p1 and xβ+p2 interchanged).
A particular case represented by Fig. 3 is the interaction among gluons.
7
Figure 1. [xαp1 , xβ+p2 ] = (α, β)xβ+p1+p2 ; (a): x
α absorption by xβ; (b): xα emission by xβ (similarly for
xαp1 and xβ+p2 interchanged).
xα+p1
x−α+p2
xαp1+p2
(a)
xαp1+p2
xα+p1
x−α+p2
x−α+p2
shifted
(b)
Figure 2: [xα+p1 , x−α+p2 ] = −xαp1+p2 ; (a): xα-x−α annihilation; (b): pair creation
(similarly for xα+p1 and x−α+p2 interchanged).
Figure 2. [xα+p1 , x−α+p2 ] = −xαp1+p2 ; (a): xα − x−α annihilation; (b): pair creation (similarly for xα+p1
and x−α+p2 interchanged).
Symmetry 2021, 13, 2289
7 of 14
xα+p1
xβ+p2
xα+β+p1+p2
(a)
xα+β+p1+p2
xα+p1
xβ+p2
xβ+p2
shifted
(b)
Figure 3: [xα+p1 , xβ+p2 ] = ε(α, β)xα+β+p1+p2 ; (a): xα-xβ scattering; (b): xα+β decay
into xα and xβ (similarly for xα+p1 and xβ+p2 interchanged).
8
Figure 3. [xα+p1 , xβ+p2 ] = ε(α, β)xα+β+p1+p2 ; (a): xα − xβ scattering; (b): xα+β decay into xα and xβ
(similarly for xα+p1 and xβ+p2 interchanged).
A particular case represented by Figure 3 is the interaction among gluons.
Notice that for each interaction as in (a) of Figures 1–3 there is an amplitude for a shift
of the two particles without interaction. This allows for an interaction as in (b) of the same
Figure at a later time.
These graphs represent the building blocks of the interactions and make the model
effective and calculable, being the amplitudes for each process and for the space expansion
well determined. We stress again that the overall picture is that of particles associated to
generators that interact while they expand in the same fashion as the wave function spreads
out in standard quantum mechanics. The expansion occurs however in a discrete space at
discrete time intervals. The graphs in the figures of this section show what happens locally
to the component of the expanded particle in a point where it is located at a certain instant
of the universal time, [1], with a certain amplitude.
We recall from reference [1] that our concept of a geometrical point in space reverses
that of locality: a point is where an interaction occurs. The initial set of generators are all
allowed to interact with each other, with a certain amplitude proportional up to normaliza-
tion to the structure constants of the algebra, at what we call time 0 of the universal clock.
This is equivalent to saying that at time 0 all particles are in the same point. The outcome of
the first interactions, plus the creation of space, which is a consequence of the momentum
part of the root associated with each generator, leads to a second set of interactions, and so
on. What we call universal time is this order parameter of the interactions.
5. The Poincaré Group
We refer to Section 2.3 of our previous paper [1], in particular we denote by ρ1, ρ2 the
roots k5 − k6 and k5 + k6 respectively.
We have a complex a1 ⊕ a1 Lie algebra M generated by x±ρ1 , x±ρ2 and the corre-
sponding Cartan generators hρ1 , hρ2 .
The spin subalgebra su(2)spin ∈ M is the compact form of the subalgebra with
generators R+ := xρ1 + xρ2 , R
− := x−ρ1 + x−ρ2 and HR :=
1
2 (hρ1 + hρ2), namely su(2)
spin
is generated by R+ + R−, i(R+ − R−) and iHR.
We denote by w the Pauli-Lubanski vector and we classify the gu generators xα+p or
xαp with respect to m2 = −p2 and w2, the two Casimir invariants of the Poincaré group. We
Symmetry 2021, 13, 2289
8 of 14
use the shorthand notation k := ±k1 ± k2 ± k3 ± k4 ± k7 ± k8, ke (resp. ko) when k has an
even (resp. odd) number of + signs.
(−p2, w2) generator
(m2, 0)
xα+p, xαp | α = ±ki ± k j, i, j /∈ {5, 6}, −p2 = m2
(m2, 34 m
2) xα+p, xαp | α = 12 (ko ± (k5 − k6)), −p2 = m2
(m2, 34 m
2) xα+p, xαp | α = 12 (ke ± (k5 + k6)), −p2 = m2
(m2, 2m2)
xα+p, xαp | α = ki ± k5 or α = ki ± k6, −p2 = m2
(0, 0)
all generators xα+p or xαp such that p2 = 0
(17)
Let geu be the extension of gu that includes all timelike and lightlike momenta (not
necessarily with integer component), and let g+e
u
the subalgebra geu of massive particles,
namely the extension of the subalgebra g+u introduced in Remark 2. We regard the following
proposition as fundamental for the relativistic behavior of our model.
Proposition 2. There is a natural action of the Poincaré group P : geu → geu. Let P = (Λ, a) be
an element of P, where Λ is a Lorentz transformation and a a translation.
The action extends by linearity the following action on the generators xα+p and xαp of geu.
1.
If p2 < 0, fix a transformation Λp such that Λp(m, 0, 0, 0) = p and let W(Λ, p) :=
Λ−1ΛpΛΛp be the Wigner rotation induced by Λ
P(xα+p) = eia·Λpead(R)xα+Λp ,P(xαp) = eia·Λpead(R)xαΛp
(18)
where ad(R) is the adjoint action of the generator R ∈ su(2)spin of the Wigner rotation
W(Λ, p).
2.
If p2 = 0 and w2 = 0 the action reduces to
P(xα+p) = eia·Λpeiθ(Λ)λxα+Λp , P(xαp) = eia·Λpeiθ(Λ)λxαΛp
(19)
where λ = 0,± 12 ,±1 is the helicity of α and θ is the angle of the SO(2) rotation along the
direction of ~p, analogous to the Wigner rotation of the massive case.
The Poincaré group is a subgroup of the automorphism group of g+e
u .
Proof. The action P on each generator with a certain mass and spin/helicity acts as the
irreducible induced representation, introduced by Wigner, [9].
We only need to prove that it is an automorphism of g+e
u , namely that P is non-singular
and preserves the Lie product (10). Part of the proof is similar to the classical one, see
Lemma 4.3.1 in [10].
The fact that P is non-singular comes from the obvious existence of its inverse trans-
formation. We are left with the proof that P([X, Y]) = [P(X),P(Y)].
Let us consider in particular P(xα+p) in (18). Since R is an e8 generator then ad(R) is
nilpotent, namely ad(R)r = 0 for some r and
ead(R) = 1 + ad(R) +
ad(R)2
2!
+ ... +
ad(R)r−1
(r− 1)!
(20)
We have
1
s!
ad(R)s[x, y] =
1
s! ∑
s
i=0 (
s
i)[ad(R)
ix, ad(R)s−iy]
= ∑
i,j
i+j=s
1
i! j!
[
ad(R)ix, ad(R)jy
]
(21)
Symmetry 2021, 13, 2289
9 of 14
and also that ad(R)t = 0 for t ≥ r implies
∑
i,j
1
i! j!
[
ad(R)ix, ad(R)jy
]
= 0 if i + j ≥ r
(22)
Let α+ β ∈ Φ8 and p21, p22 < 0. We get:
P([xα+p1 , xβ+p2 ]) = P(ε(α, β)xα+β+p1+p2) = eia·Λ(p1+p2)ead(R)[xα+Λp1 , xβ+Λp2 ]
= eia·Λ(p1+p2) ∑s≥0 ∑
i,j
i+j=s
1
i! j!
[
ad(R)ixα+Λp1 , ad(R)
jxβ+Λp2
]
= eia·Λp1+Λp2 ∑i≥0 ∑j≥0
1
i! j!
[
ad(R)ixα+Λp1 , ad(R)
jxβ+Λp2
]
= [P(xα+p1),P(xβ+p2)]
(23)
Similarly for the other commutators in (10).
The action P can be easily extended to sgu by acting accordingly on the Grassmann
variable in order to get the variable associated to the transformed generators of gu.
6. Initial Quantum State
The initial quantum state of our model of the expanding early Universe is an element
of the universal enveloping algebra Usgu of sgu, namely an element of the tensor algebra
built on the generators of sgu modulo the relations defining the product in the algebra itself.
The initial generators are all in pairs with opposite helicity and opposite 3-momentum, [1],
and have a phase or amplitude associated to each of them as a complex coefficient. The
interactions and expansions starting from the initial state are such that locally the quantum
state is an element of the universal enveloping algebra. Interference plays the crucial
role in the quantum behavior of the model, including repulsive versus attractive forces.
The quantum nature of gravity appears through the quantum nature of spacetime: at every
cosmological instant, a point in space has an amplitude which is the sum of the amplitudes
for particles to be at that point.
The initial state has the mean energy of the Universe concentrated on the generators
that interact with each other at t = 0. The choice of the initial state is crucial in determining
the likelihood for the existence of particles and of an eventual symmetry breaking. It is
beyond the scope of this paper to investigate this subject in depth; an algorithm based on
the algebra and the expansion rule that we have introduced can be the basis for computer
calculations, which should shed some light on the physical consequences of the choice of
the initial quantum state.
7. Vertex-Type Algebra and Gravitahedra
Space expansion leads to an enrichment of the algebra. The locality of interactions
suggests to embed the algebra in a vertex-type operator algebra, in which the generators of
sgu act as vertex operators on a discrete space that is being built up, step by step, by sgu
driven interactions.
The tree structure of the interactions allows for a description of scattering amplitudes
in terms of associahedra or permutahedra, [11–19], with structure constants attached to each
vertex; see Figure 4 for the interaction of four particles, producing the associahedron K4. A
vertex is interpreted as an interaction with universal time flowing from top to bottom in
the trees of Figure 4. However, if one includes the gravitational effect of space expansion,
one should describe the interactions through permutahedra Pn−1 rather than associahedra;
see Figure 5 for the interaction of four particles, producing the permutahedron P3.
The two trees in Figure 5b are different due to the spreading of particles in space,
because the same interactions occur at different times (represented by the horizontal lines).
A complete graphical description of the interactions, including the spacetime effects,
hence gravity, can be quite complicated and needs a deep study. A research program with
Symmetry 2021, 13, 2289
10 of 14
this goal has initiated, and the name gravitahedra has been coined for the polytopes that
will eventually, and hopefully, describe such interactions.
tudes in terms of associahedra or permutahedra, [11]-[18], with structure constants
attached to each vertex; see Fig. 4 for the interaction of four particles, producing
the associahedron K4. A vertex is interpreted as an interaction with universal time
flowing from top to bottom in the trees of Fig. 4. However, if one includes the
gravitational effect of space expansion, one should describe the interactions through
permutahedra Pn−1 rather thn associahedr; see Fig. 5 fr the interaction of four
particles, producing the permutahedron P3. The two trees in Fig. 5 (b) are different
Figure 4: Associahedron K4. Adjacent vertices (st)u→ s(tu), for sub-words s, t, u
due to the spreading of particles in space, because the same interactions occur at
different times (represented by the horizontal lines).
11
Figure 4. Associahedron K4. Adjacent vertices (st)u→ s(tu), for sub-words s, t, u.
Figure 5: Permutahedron P3. Interaction of 4 particles.
A complete graphical description of the interactions, including the spacetime
effects, hence gravity, can be quite complicated and needs a deep study. A research
program with this goal has initiated, and the name gravitahedra has been coined for
the polytopes that will eventually, and hopefully, describe such interactions.
The fact that locally the quantum state is an element of the universal enveloping
algebra means that we can assign to it labels q of space Q, which are triples of
rational numbers, due to the expansion by ~p/E, where E, px, py, pz are integers, [1].
The vertex-type algebra is therefore the algebra Usgu(Q), whose relations have been
extended in order to include the commutation of elements with different space-labels.
8 Conclusion
In the pair of papers given by [1] and the present paper, we have presented an in-
trinsically quantum and relativistic theory of the creation of spacetime starting from
a quantum state as cosmological boundary condition, which we conceive to play a
key role in any fundamental theory of Quantum Gravity. We have discussed the
general framework of a workable model, based on a rank-12 infinite dimensional Lie
superalgebra, which can be applied to the quantum era of the first cosmic evolution.
Our model can accommodate the degrees of freedom of the particles we know, with-
out superpartners, namely spin-12 fermions and spin-0 and spin-1 bosons obeying the
proper statistics.
The seed of quantum gravity has to be searched at the big bang era. In the
algebraic realm there is no need to have the whole set of generators at the big bang,
since a proper choice of a (finite) number of them can produce all generators in the
algebra. The use of Serre’s theorem, for instance, states this fact in the proper and
elegant mathematical language in the case of a Lie algebra. The same is true for
infinite dimensional Kac-Moody and Borcherds algebras.
In our model a proper finite set of generators is suitable to represent the initial
Figure 5. (a): whose; (b): renders a magnification, shows the permutahedron P3, pertaining to the
interaction of 4 particles.
The fact that locally the quantum state is an element of the universal enveloping
algebra means that we can assign to it labels q of space Q, which are triples of rational
numbers, due to the expansion by ~p/E, where E, px, py, pz are integers, [1]. The vertex-type
algebra is therefore the algebra Usgu(Q), whose relations have been extended in order to
include the commutation of elements with different space-labels.
8. Conclusions
In the pair of papers given by [1] and the present paper, we have presented an
intrinsically quantum and relativistic theory of the creation of spacetime starting from a
quantum state as cosmological boundary condition, which we conceive to play a key role in
any fundamental theory of Quantum Gravity. We have discussed the general framework of
a workable model, based on a rank-12 infinite dimensional Lie superalgebra, which can be
applied to the quantum era of the first cosmic evolution. Our model can accommodate the
degrees of freedom of the particles we know, without superpartners, namely spin- 12 fermions
and spin-0 and spin-1 bosons obeying the proper statistics.
The seed of quantum gravity has to be searched at the big bang era. In the algebraic
realm there is no need to have the whole set of generators at the big bang, since a proper
choicof a (finite) number of thm can pr duce all generators i
the algebra. The use
of Serre’s theore
, for
stance, states this fact in the proper and legant mathematical
language n the case of a Lie algebra. The same is true for infinite dimensional Kac-Moody
and Borcherds algebras.
In our model a proper finite set of generators is suitable to represent the initial state,
but as soon as time starts flowing the state necessarily becomes a fully entangled pure
Symmetry 2021, 13, 2289
11 of 14
state—one may start with few generators of a much simpler algebra than ours, to deduce
this fact from our rules. It is intuitive to say that the state stays this way, a pure entangled
state, until the first observations, supposedly occurring after the universe has cooled down
a lot and undergone a large expansion. The measurement by an apparatus that entangles
with the observed object, changes the pure state of the universe: namely it disentangles a
very tiny part from it and eventually, after subsequent observations, turns that pure state
into a mixture. Quasi-classical phenomena may thus appear in a very small part of the
universe—locally, we may say. The density matrix of the universe changes a little tiny bit,
some information is lost and the entropy increases.
The quantum nature of gravity is intrinsically unobservable, because observation
implies the destruction of the entanglement and the collapse of the wavefunction.
This is the reason why objects that are expected to be intrinsically quantum, such as
black holes or the rapid expansion characterizing inflation in early Universe, can effectively
be described by (semi)classical structures, such as a Riemannian metric and a potential
with flat directions, respectively.
Thus, the deal in Quantum Gravity is the following: the intrinsically quantum and
relativistic description of an intrinsically unobservable regime should be made consistent
with the existence of a macroscopic observer, and thus of a (semi)classical observational
symmetry, emerging in the thermodynamical/macroscopic limit in which the entanglement
becomes irrelevant. Our model, by exploiting Occam’s razor, tackles this crucial issue of
Quantum gravity, providing an elegant solution, which can be regarded as a “third way”,
alternative to both supersting/M- theory and loop quantum gravity: indeed, the Poincaré
group emerges from both the “spin” sector (e8) and the kinematical sector (complementary
of e8 in gu) of the Lie superalgebra sgu. Besides the absence of superpartners and the
implementation of the Pauli exclusion principle, the emergence of the Poincaré group is
a crucial feature of our model. We should stress that, of course, the Poincaré group can
be defined only in the thermodynamical limit in which the observer can be consistently
decoupled from the evolutive dynamics of the Universe, given in toto by sgu. Especially in
an early Universe, the back-reaction of the observer on the object of the observation should
be relevant, and thus the abstraction of a decoupled and distinct observer is not totally
consistent during the early stages of the Universe.
Many physical properties have still to be verified and/or fulfilled, like the proton
decay, the confinement of quarks, the attractive nature of gravity on the large scale. The
general framework of the model leaves however a great freedom of choice, and this is to
be regarded as a benefit for those who believe this is a promising approach and wish to
explore it.
There is much left for future work, to start with the definition of a particular quantum
initial state allowing to perform some preliminary computer calculations that may give
an idea of how the model effectively works. In particular, the density matrix, von Neu-
mann entropy, mean energy, scattering amplitudes can be explicitly calculated according to
our model.
We end this series of two papers by recapitulating what we consider the main physical
features of our approach:
(I)
spacetime is the outcome of the interactions driven by an infinite-dimensional Lie
superalgebra sgu; it is discrete, finite and expanding;
(II)
the algebra sgu incorporates 4-momentum and charge conservation; it involves
fermions and bosons, with fermions fulfilling the Pauli exclusion principle;
(III) sgu is a Lie superalgebra without any supersymmetry forcing the existence of super-
partners for the particles of the Standard Model;
(IV) every particle has positive energy and it is either timelike or lightlike;
(V)
the initial state is an element of the universal enveloping algebra of sgu;
(VI) the interactions are local, and the whole algebraic structure is a vertex-type algebra,
due to a mechanism for the expansion of space (in fact, an expansion of matter
and radiation);
Symmetry 2021, 13, 2289
12 of 14
(VII) the emerging spacetime inherits the quantum nature of the interactions, hence Quan-
tum Gravity is an expression for quantum spacetime—in particular, there is no
spin-2 particle;
(VIII) the Poincaré group has a natural action on the local algebra;
(IX) once an initial state is fixed, the model can be viewed as an algorithm for explicit
computer calculations of physical quantities, like scattering amplitudes, density
matrix, partition function, mean energy, von Neumann entropy, etc.
Author Contributions: Conceptualization, P.T., A.M., M.R. and K.I.; methodology, P.T. and A.M.;
formal analysis, P.T., A.M. and M.R.; investigation, P.T.; writing—original draft preparation, P.T.;
writing—review and editing, P.T., A.M. and M.R.; supervision, K.I. All authors have read and agreed
to the published version of the manuscript.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.
Appendix A
We prove Proposition 1.
The algebra gu with relations (10)–(12) is obviously infinite dimensional, and its
product is antisymmetric. We only need to prove that it fulfills the Jacobi identity.
Throughout the proof we strongly rely on the following standard results, see [1] and
Refs. therein.
Proposition A1. For each α, β ∈ Φ8 the scalar product (α, β) ∈ {±2,±1, 0}; α+ β ( respectively
α− β) is a root if and only if (α, β) = −1 (respectively +1); if both α+ β and α− β are not in
Φ8 ∪ {0} then (α, β) = 0.
For α, β ∈ Φ8 if α+ β is a root then α− β is not a root.
Proposition A2. The asymmetry function ε satisfies, for α, β,γ ∈ L:
(i)
ε(α+ β,γ) =
ε(α,γ)ε(β,γ)
(ii)
ε(α, β+ γ) =
ε(α, β)ε(α,γ)
(iii)
ε(α, α) = (−1) 12 (α,α) ⇒ ε(α, α) = −1 if α ∈ Φ8
(iv)
ε(α, β)ε(β, α) = (−1)(α,β) ⇒ ε(α, β) = −ε(β, α) if α, β, α+ β ∈ Φ8
(v)
ε(0, β) =
ε(α, 0) = 1
(vi)
ε(−α, β) =
ε(α, β)−1 = ε(α, β)
(vii)
ε(α,−β) =
ε(α, β)−1 = ε(α, β)
By linearity it is sufficient to prove that the Jacobi identity holds for the generators of
the algebra. For each triple of generators X, Y, Z we write
J1 := [[X, Y], Z] ,
J2 := [[Z, X], Y] ,
J3 := [[Y, Z], X]
(A1)
We want to prove that J := J1 + J2 + J3 = 0.
For p
6= 0 we call the generators xαp of type 0 and xα+p of type 1.
We consider the various cases.
(a) At least one of X, Y, Z is of type-0
(a1)
If X, Y, Z are all of the type-0 then Jacobi holds trivially.
(a2)
If X = xαp1 , Y = x
β
p2 are of type 0 and Z = xγ+p3 is of type 1 then J1 = 0,
J2 = (α,γ)(β,γ)xγ+p1+p2+p3 and J3 = −(α,γ)(β,γ)xγ+p1+p2+p3 , hence J = 0.
(a3)
If X = xαp1 is of type 0 and Y = xβ+p2 , Z = xγ+p3 are of type 1, then J1 =
(α, β)[xβ+p1+p2 , xγ+p3 ], J2 = (α,γ)[xβ+p2 , xγ+p1+p3 ] and J3 = −[xαp1 , [xβ+p2 ,
xγ+p3 ]]. We have 3 cases:
(a3.i) β+ γ
6∈ Φ8 ∪ {0} then J1 = J2 = J3 = 0;
Symmetry 2021, 13, 2289
13 of 14
(a3.ii) β+ γ ∈ Φ8 then J3 = −(α, β+ γ)ε(β,γ)xβ+γ+p1+p2+p3 = −(J1 + J2);
(a3.iii) β+ γ = 0 then J1 = −(α, β)xβp1+p2+p3 , J2 = (α, β)x
β
p1+p2+p3 and J3 = 0,
hence J = 0.
(b) None of X, Y, Z is of type-0. Let X = xα+p1 , Y = xβ+p2 , Z = xγ+p3 be all of type 1. For
any two roots of Φ8, say α, β without loss of generality, we have three cases:
(b1)
α+ β
6∈ Φ8 ∪ {0}:
(b1.i)
if both α+ γ, β+ γ
6∈ Φ8 ∪ {0} then J = 0 trivially;
(b1.ii)
if β + γ
6∈ Φ8 ∪ {0} and α + γ ∈ Φ8 ∪ {0} then J1 = J3 = 0. Since
both (α, β), (β,γ) ∈ {0, 1, 2} then (α+ γ, β) ≥ 0 hence if α+ γ ∈ Φ8,
then α+ β+ γ
6∈ Φ8 ∪ {0} and J2 = 0. On the other hand if α = −γ
then J2 = [xαp1+p3 , xβ+p2 ] = (α, β)xβ+p1+p2+p3 . But (β,γ) = −(β, α)
and (α, β), (β,γ) ∈ {0, 1, 2} imply (α, β) = 0 hence J = 0;
(b1.iii) if β+ γ ∈ Φ8 and α+ γ ∈ Φ8 then J2 = ε(γ, α)[xα+γ+p1+p3 , xβ+p2 ] and
J3 = ε(β,γ)[xβ+γ+p2+p3 , xα+p1 ]. If α+ β+γ
6∈ Φ8 ∪{0} then J2 = J3 =
0 hence J = 0. If α+ β+ γ ∈ Φ8 then J2 + J3 = ε(γ, α)(ε(γ, β)ε(α, β) +
ε(β,γ)ε(β, α))xα+β+γ+p1+p2+p3 . Since 2 = (α+ β+ γ, α+ β+ γ) = 6+
2(α, β) + 2(β,γ) + 2(α,γ) = 2 + 2(α, β), we get (α, β) = 0 and, from
Proposition A2, ε(α, β) = ε(β, α) and ε(γ, β) = −ε(β,γ), implying
J2 + J3 = 0 and J = 0. Finally if α+ β+ γ = 0 then (α, β) = (α,−α−
γ) = −2 + 1 = −1 and α + β would be a root, contradicting the
hypothesis.
(b1.iv) if β + γ ∈ Φ8 and α + γ = 0 then J2 = (α, β)xβ+p1+p2+p3 and J3 =
ε(β, α)ε(β− α, α)xβ+p1+p2+p3 = −xβ+p1+p2+p3 . But (α, β) = −(γ, β) =
1 hence J2 + J3 = 0 and J = 0.
(b1.v) If β + γ ∈ Φ8 and α + γ
6∈ Φ8 ∪ {0} then J2 = 0 and (β, α) ≥ 0,
(γ, α) ≥ 0 imply (β+ γ, α) ≥ 0 hence β+ γ+ α
6∈ Φ8 ∪ {0} therefore
J3 = ε(β,γ)[xβ+γ+p2+p3 , xα+p1 ] = 0 and J = 0.
(b1.vi) If β+ γ = 0 and α+ γ ∈ Φ8 then
J2 = −ε(α, β)ε(α− β, β)xα+p1+p2+p3 = xα+p1+p2+p3 and
J3 = −[xβp2+p3 , xα+p1 ] = −xα+p1+p2+p3 , being (α, β) = −(α,γ) = 1,
implying J = 0.
(b1.vii) If β+ γ = 0 and α+ γ = 0 then J2 = [x
β
p1+p3 , xβ+p2 ] = 2xβ+p1+p2+p3
and J3 = −[xβp2+p3 , xβ+p1 ] = −2xβ+p1+p2+p3 and J = 0.
(b1.viii)If β+ γ = 0 and α+ γ
6∈ Φ8 ∪ {0} then J2 = 0; (α, β) = 0 since (α, β) ≥
0 and (α,γ) = −(α, β) ≥ 0, therefore J3 = −[xβp2+p3 , xα+p1 ] = 0 and
J = 0.
From now on α+ β, α+ γ, β+ γ ∈ Φ8 ∪ {0}.
(b2)
α+ β ∈ Φ8:
(b2.i)
If α + γ, β + γ ∈ Φ8 then (α + β + γ, α + β + γ) = 0 hence α + β +
γ = 0. Then J1 = −ε(α, β)xα+β
p1+p2+p3 , J2 = ε(α+ β, α)x
β
p1+p2+p3 , J3 =
ε(β, α+ β)xαp1+p2+p3 . Since ε(α+ β, α) = ε(β, α+ β) = ε(α, β), x
α+β
p =
xαp + x
β
p , see (12), we get J = 0.
(b2.ii)
If α+ γ ∈ Φ8 and β+ γ = 0 then α− β ∈ Φ8 which is impossible.
(b2.iii) If α+ γ = 0 and β+ γ ∈ Φ8 then β− α ∈ Φ8 which is impossible.
(b2.iv) If α+ γ = 0 and β+ γ = 0 then α = β which is impossible.
(b3)
α+ β = 0:
(b3.i)
If α+ γ ∈ Φ8 then β+ γ = −α+ γ
6∈ Φ8; we can only have β+ γ = 0
implying α = γ, that contradicts α+ γ ∈ Φ8.
Symmetry 2021, 13, 2289
14 of 14
(b3.ii)
If β+ γ ∈ Φ8 then α+ γ = −β+ γ
6∈ Φ8; we can only have α+ γ = 0
implying −α = β = γ that contradicts β+ γ ∈ Φ8.
(b3.iii) If both α+ γ = 0 and β+ γ = 0 then α = β which contradicts α+ β =
0.
This ends the proof.

1.
Truini, P.; Marrani, A.; Rios, M.; Irwin, K. Space, Matter and Interactions in a Quantum Early Universe. Part I: Kac-Moody and
Borcherds Algebras. arXiv 2020, arXiv:2012.10227.
2.
Kac, V.G. Infinite Dimensional Lie Algebras; Third Edition Reprinted with Corrections; Cambridge University Press: Cambridge, UK,
1995.
3.
Borcherds, R.E. Generalized Kac-Moody algebras. J. Algebra 1988, 115, 501–512. [CrossRef]
4.
Borcherds, R.E. Monstrous moonshine and monstrous Lie superalgebras. Invent. Math. 1992, 109, 405–444. [CrossRef]
5.
Truini, P. Vertex operators for an expanding universe. In Proceedings of the Symmetries and Order: Algebraic Methods in Many
Body Systems, Francesco Iachello, Yale, 5–6 October 2018; Volume 2150, p. 040005.
6.
The GAP Group. GAP—Groups, Algorithms, and Programming, Version 4.10.2. 2019. Available online: https://www.gap-system.
org (assessed on 12 September 2021).
7.
de Graaf, W.A. Lie Algebras: Theory and Algorithms; North-Holland Mathematical Library 56; Elsevier: Amsterdam, The Netherlands,
2000.
8.
Shestakov, I.P. Prime Malcev Superalgebras. Mat. Sb. 1993, 74, 101–110. [CrossRef]
9. Wigner, E.P. On unitary representations of the inhomogeneous Lorentz group. Ann. Math. 1939, 40, 149. [CrossRef]
10. Carter, R.W. Simple Groups of Lie Type; Wiley-Interscience: New York, NY, USA, 1989.
11. Sheppeard, M.D. Gluon Phenomenology and a Linear Topos. Ph.D. Thesis, University of Canterbury, Christchurch, New Zealand,
2007.
12. Arkani-Hamed, N.; Trnka, J. The Amplituhedron. J. High Energy Phys. 2014, 10, 030. [CrossRef]
13. Mizera, S. Combinatorics and topology of Kawai-Lewellen-Tye relations. J. High Energy Phys. 2017, 08, 097. [CrossRef]
14. Arkani-Hamed, N.; Bai, Y.; Lam, T.J. Positive Geometries and Canonical Forms. J. High Energy Phys. 2017, 11, 039. [CrossRef]
15. Stasheff, J.D. Homotopy associativity of H-spaces. I. Trans. Am. Math. Soc. 1963, 108, 275–292.
16. Stasheff, J.D. Homotopy associativity of H-spaces. II. Trans. Am. Math. Soc. 1963, 108, 293–312.
17. Stasheff, J.D. From operads to physically inspired theories. In Operads: Proceedings of Renaissance Conferences (Hartford, Connecti-
cut/Luminy, France, 4–5 March 1995); American Mathematical Soc.: Providence, RI, USA, 1997; Volume 202, p. 53.
18. Tonks, A. Relating the associahedron and the permutohedron. In Operads: Proceedings of Renaissance Conferences (Hartford,
Connecticut/Luminy, France, 4–5 March 1995); American Mathematical Soc.: Providence, RI, USA, 1997; Volume 202, pp. 33–36.
19. Loday, J.L. Realization of the Stasheff polytope. Arch. Math. 2004, 83, 267. [CrossRef]