According to the hypothesis that the universe is discrete, the elementary particles are structures built around world points with a given symmetry. Let us recall that the cosmic bits of a world point are all connected to each other through ferromagnetic interactions [1] . Then space-time disappears inside a world point which therefore seems punctual from an experimental point of view. According to [3] the size of a world point would be of the order of.

Our approach suggests that a particle could be structured along a three levels organization

1) A particle is built around a special world point (whose size is), called its seed. The internal interactions matrix of the seed G_P must transform according to gauge symmetries and the symmetry properties of the seed determine the symmetry of the whole particle.

2) The seed modifies the polarization states of the neighbouring world points over a range. This region of influence is called the core of the particle and the system made of a seed and its core constitutes a bare particle.

3) Finally the bare particle is surrounded by a cloud of gauge particles whose nature and properties are determined by the symmetry of the seed according to the gauge symmetry mechanisms that have been discussed in [3] . The system made of a bare particle and its cloud of virtual particles constitutes a dressed particle.

The permutation group S₄ of four objects, a finite group, plays a prominent role in our approach as already stressed in [1] - [3] . Physics, indeed, must be invariant under any permutation of the four axes used to describe the internal states of world point i since no specific orientation of axes can be defined inside a world point. The permutation group S₄ of four objects has 24 elements distributed into 5 classes. The table of characters of S₄ may be found in [1] . The internal interactions matrices G_P of the seed must commute with any four dimensional representation of group S₄. Therefore it must transform according to direct sums of irreducible representations of S₄. There are two types of irreducible representations. On the one hand and, that are insensitive to mirror operations that is to an odd number of permutations. On the other and are sensitive to mirror operations. Finally does not take the mirror operations into account. We see below that the particles imply the representations, and and the antiparticles the representations, and. The problem of finding four-dimensional matrices G_P as direct sums of, and representations has the following solutions.

a) 4 = 1 + 3

b) 4 = 2 + 2

c) 4 = 1 + 1 + 2

d) 4 = 1 + 1 + 1 + 1

The matrices G_B that transform along to solution a), that is, define the seeds of boson particles. The matrices G_F that transform along to solution b), that is, define the seeds of fermion particles (see [2] ).

Let us recall that the state of a physical system containing a particle P obeys the following eigenvalue equation

,

The mass of P is given by.

This way of introducing particle masses seems to go against the usual conclusion that all particles must have zero masses due to the dichotomy of the universe between a right and a left universe. The well known argument goes as follows:

The mass term in the Lagrangian of a fermion field writes

.

We consider the right handed term. We define a projection operator with where are the four Dirac matrices. The projection operator transforms a general state into a right handed state

.

Since one also has

,

and thus

.

Likewise. All other terms, such as, mix the chiralities and are therefore forbidden. Consequently there are no more mass terms left in the Lagrangian.

If there is no universe dichotomy this derivation no longer holds and the particles may have non zero-masses.

The universe dichotomy is introduced to account for the experimental observation that the neutrinos are left handed. We show in Section 4.1 that this property of neutrinos is, in reality, a consequence of the model and no Higgs mechanism is needed to give a mass to particles. It must be stressed, however, that a Higgs scalar field, identified with the world point polarization state, is still necessary simply to give an existence to particles.

The states of an isolated seed, associated with a particle P, are obtained by minimizing the Lagrangian

,

under the constraint. The states of the seed thus satisfy the following eigenvalue equation

.

The mass of the seed is given by

.

By using a first order perturbation expansion we find that the eigenvalue of a bare particle associated with a seed P is simply proportional to the eigenvalue (here and in the following sections s or seed will be used for seed, d or dressed for dressed, bare for bare and finally vb or virtual bosons for virtual bosons).

Proof:

A bare particle, we have seen in Section 2.1, is comprised of a seed and its core. Let i be the label of world points belonging to the core. The vacuum states of world points belonging to the core are modified by the state of the seed. To first order they become. Similarly the states modifications in the core induce a reorganization of the cosmic bits of world points i, and, therefore of the internal interaction matrices. G_i becomes.

The eigenvalue of the bare particle then writes

.

There is no orientation correlations between and. Therefore, the cross terms vanish and we obtain

.

The Lagrangian operator of a particle, that is of a coupled pair of world points, one of bosonic type, the other of fermionic type, must keep the symmetry properties of the two components taken separately so as to preserve the organization of the particles of the Standard Model. This compels the Lagrangian to be a 16-dimensional operator given by

is the four dimensional unit matrix. The coupling parameter is determined by the organization of interactions between the two world points that constitute a particle, more precisely by the relative numbers and of ferromagnetic and antiferromagnetic interactions that link the two points. For example. We have no indication as regards an evaluation of the coupling parameter except that the values and are forbidden, because the two world points that form the particle would not be coupled and because the coupling would be so strong that the two world points would be in the same state. A value could be a reasonable guess but we will see that the results are, in actual fact, weakly dependent.

The Lagrangian (4) has four eigenvalues. The first is associated with the bare down quark

The second is associated with the bare up quark

That is

and

Since the masses of bare particles are proportional to the masses of their seeds one has

For leptons, however, no calculation of bare particles has been carried out. It is then necessary to add the effects of virtual bosons. Then to calculate their dressed masses one must write

for the neutrino and

for the electron. Explicitly one has

and

The definition of the bare mass implies that. If one has a problem since the mass of the bare particle becomes imaginary. If one wants to keep the minimization principle, and to give nevertheless a physical meaning to negative eigenvalues, one is forced to interpret the negative eigenvalues as zero mass (or almost zero mass) particles. To make this point more precise we consider the set of states defined by. This set forms a surface (a manifold) in the internal space of the world points that generates the bare particle P. During the minimization process the state evolves so as to minimize under the constraint. The trajectory of possibly crosses at a point. Let us then consider a state. One has

.

has a physical meaning if that is if. Crossing, however, induces a sign change of. If has a physical meaning in one side of it loses this meaning on the other side. To a trajectory of state, however, there corresponds a mirror trajectory of the anti-state. On this trajectory the sign of is changed. Therefore while crossing, a state, attracted by, keeps a physical meaning if it transforms into its anti-state which is also attracted by. This mechanism reminds the Zitterbewegung process, an oscillation between a particle and its antiparticle. The trajectory of therefore stops on, or remains close to,. Since the state corresponds to zero, or almost zero mass particles.

The mass of a dressed particle is given by. If while is positive, the mass of the dressed particle keeps a physical meaning which, in the framework of the present model, seems to be the case for the electron (but not for the neutrino). Otherwise the electron would have an imaginary mass which is possibly the reason that makes impossible a calculation of the electron bare mass.

There is no direct measurement of quark bare masses because it is not possible to observe separate quarks due to their confinement properties. The value of separate bare quark masses can only be obtained through various theoretical calculations that usually lead to different numerical values. The results for quarks bare masses are rather scattered. For example it has been proposed [7] that

and

Calculations, carried out on lattices by C. Davies et al. [8] (see also [9] ) give values for the bare masses of light quarks that, the authors claim, are much more accurate than the previously published data. They find

and

which gives.

The ratio between bare quark masses is better determined than their explicit values and this is the only value that one can calculate in the present theory without appealing to free parameters. There are two parameters in eq. (9): c and. The constant c, the dimensionless speed of light, may be determined through considerations relating to electroweak interactions given in contribution [3] . Explicitly where is the Weinberg angle. Then. With the experimental value this yields. From the above formula (Equation (9)) one has

With the value (that is) one finds, that fits the Davies calculation within a 1% error.

The range of values given by Davies calculation is and corresponds to the range

The average value of Davies is obtained for close to the proposed value.

With the values adopted for c and one has and. Note that must be positive and large enough for to be positive. Since (the neutrino is neutral) one has.The mass of the neutrino therefore vanishes or, at most, remains very close to zero. In our interpretation the state of a neutrino associated with negative eigenvalues oscillates from one side of the manifold to the other side and from the particle to its antiparticle. The mass of a neutrino is the mass it has in the positive side of. Since the particle must never go far away from its mass is very small.

The discovery (by Wu) that the parity symmetry is completely broken by weak interactions and that the electronic neutrino is left handed compelled the physicists to assume that the universe is, in actual fact, made of two different sorts of universes: a left-handed universe L and a right-handed universe R. This hypothesis, we have seen, leads to the conclusion that all fermions are zero-mass particles. The eigenvalue associated with neutrinos is negative and the state of a neutrino must stay on or remain in the close vicinity of the manifold, a surface on which the neutrino state can move freely until the Lagrangian is minimized, that is until the neutrino state is parallel to a relevant eigenvector of G_F. G_F can be written as

with

. The eigenstates of G_F and those of are identical. has two eigenvalues. The eigenvalue is associated with the neutrino and the eigenvalue is associated with the electron. Let us write the eigenstate of a neutrino as

.

For to be an eigenstate of it is necessary for and to be eigenstates of with the same eigenvalue, (corresponding to neutrinos), and and must be identical to one another within a phase factor whence

.

The Lagrangian of a neutrino then writes

that is

is minimum for: and the eigenstate of a neutrino writes

is an eigenvector of matrix with eigenvalue −1 since

The eigenvalue associated with, an eigenvector of, is negative and the neutrino is, therefore, left-handed. In our approach the neutrino’s left handed chirality is a consequence of the general properties of the universe which does not need to be shared into left handed, L, and right handed, R, universes. The same argu- ment does not hold for the electronic state because and the electron state trajectory does not cross the manifold. Then the electron state is an eigenstate corresponding to the eigenvalue which has no defined chirality.

The particles masses can be computed without appealing to a Higgs mechanism as we have seen. We have already said that the Higgs field, identified with the polarization amplitude of world point i, is still necessary, not to give a mass to particles, but simply to give them an existence. The Higgs field is also necessary to make sure that the photon mass is strictly zero as found by the GWS approach of electroweak interactions [3] .

The model can however yield a value for the mass of Higgs particles if any. The polarization mini- mizes the Landau expression (see [1] )

,

where

.

The minimum of F is given by. We expand F to second order around this minimum

,

that is

.

In other respect one has and therefore

.

We now give to c its physical value. With, and one finds

,

a value that is far out of reach even for the most modern colliders.

The model presented here also provides a very simple explanation of Regge trajectories [10] . We have seen that the eigenvalue of a particle is related to the mass of the particle by the relation:

.

Let us gather fermions. The stability of the cluster is generally weak and the cluster, called a resonance, decays rapidly. The spins of the individual fermions adds up to a maximum spin. Since the links between the individual fermions are weak their eigenvalues approximately add up and

that is

.

Finally

.

This linear dependence of the spin j along the square of the mass is called a Regge trajectory.

The parity symmetry P is broken by weak interactions as we have seen. On can wonder whether the time inversion symmetry T, or the charge conjugation symmetry C can also be broken. We do not know the answer but a very general property, the CPT theorem, a product of these three invariance shows that the product CPT must be always verified.

The vector

,

represents the state of a particle. The state of the associated anti-particle is obtained from through a mirror transformation that is a sign change of one of its components but, as discussed above, all signs have to be simultaneously changed. Whence

,

C is the charge conjugation operator.

In other respects the time reversal operator T changes the sign of the time component of

that is

.

Finally the operation of reversing the direction of one axis, say the direction, amounts to a reflection of space (a mirror symmetry operation) in any number of space dimensions, and in three space dimensions it is equivalent to reflecting all space coordinates, because one can add an additional rotation of 180 degrees in the x-y plane to complete the transformation. Let P be the operator that reflects all space coordinates. P amounts to changing the signs of the three spatial components, and.

,

that is

.

We see that

.

a simple and straightforward derivation of the CPT theorem.

The electric charge Q_f of a given fermion, lepton or quark, is derived from the hypercharge Y_f through the formula

,

where is the isospin of the particle.

The three generators of are the three Pauli matrices

.

The interaction Lagrangian between leptons and their gauge field writes [3]

.

The quantities are gauge fields. To make this expression more symmetric, the 2-dimensional identity matrix

is added (and substracted) to the list of Pauli matrices (as in GWS theory)

.

The amplitude of the hypercharge is defined by the coefficient of the generator of group U(1) namely the coefficient of the unit matrix and, therefore,. One chooses. The isospin of the electron is. Therefore its electric charge is (that is). The isospin of the neutrino is. Therefore its electric charge is, that is the neutrino is neutral.

Let us now turn towards the electric charges of quarks. The eight generators of SU(3) are the eight Gell-Mann matrices

.

To make the expression of the Lagrangian more symmetric the 3-dimensional identity matrix

is added (and substracted) to the list of Gell-Mann matrices

.

In order to compare the electric charges of leptons and quarks it is necessary to establish an endomorphism between the set of the eight Gell-Mann matrices and the set of three Pauli matrices. This endomorphism is given by

in the main term and in the last term whence

.

One observes that. One chooses. The isospin of the up quark is. Therefore the electric charge of the up quark is

(that is).

The isospin of the down quark is. Therefore its electric charge is

(that is).

As already argued, the symmetry properties of bare particles are determined by the matrix. The particles are created in high energy collisions experiments. In this type of experiments much energy is put down inside the core of a world point to such an amount that the value of cosmic temperature inside the core could be severely modified and, thereby, could modify its geometrical properties. We propose to look for the origin of families of fermions in these cosmic temperature perturbations.

We have seen, in [1] , that the dimensionality d of space-time is given by, where J is the binary interaction between the cosmic bits and b is a measure of cosmic bits disorder that is of cosmic temperature. The spatial dimension is. A decrease of the cosmic temperature results in an increase of the dimensionality d. When decreases, the dimensionality of the core of the point of collision undergoes a series of changes.

1):. The core of a world point is very “hot” and dimensionless.

2):. The core has one time dimension and no space dimensions.

3):. The core has one space dimension and one time dimension.

4):. The core has two space dimensions and one time dimension.

5):. This is the case for the ordinary space-time. The core has then three space dimensions and one time dimension.

The three domains (3, 4 and 5) of cosmic temperatures would correspond to the three families of particles. The domain (5) would be the domain of the electron family, the domain (4) that of the muon family and the domain (3) that of the tau family. There is no room for any other family. There are three families of fermions because there are three space dimensions in our space-time.

The idea that the three-fold multiplicity of families of particles may be explained by a dimensional re-organiza- tion of space-time in the core of a particle is supported by the following analysis of the CKM (Cabbibo, Kobayashi and Maskawa) matrix.

The CKM matrix V describes the amplitudes of transition probabilities between the set of three quarks with isospin −1/2, namely u, c and t, and the set of three quarks d, s and b, with isospin +1/2 through weak interaction. The CKM matrix V writes

.

The experimental values of the elements of the CKM matrix are given below in Equation (15).

Here we strive to find the CKM matrix by using a very simple geometrical argument. The probability of a transition of quark into quark is assumed to be proportional to the overlap between quark and quark, that is by the relative number of world points that are common to and.

The perturbation in the vicinity of a particle seed that is in the core, in a d-dimensional space behaves as

with., the dimension of bare particles, that is the size of the core, is a standard of length. Let be the range of a d-dimensional bare particle. This size is the distance where

For 3-dimensional quarks (u and d) one has. Then and therefore. The number of world points gathered to build a 3-dimensional bare particle is

For 2-dimensional quarks (s and c) one has. Then

and, therefore,. The number of world points in a 2-dimensional bare particle is

.

Finally for 1-dimensional quarks (t and b) one has. Then and, therefore,. The number of world points gathered to build a 1-dimensional bare particle is

.

According to our hypothesis, the probability of transition between particle and particle is

.

That is

and

.

The matrix W then writes

.

Since one must have the probabilities are normalized

.

Finally the transition probability amplitude is given by. To find the range parameter we use an optimal mean square deviation procedure. The quantity to be minimized is

.

One finds. Then the computed CKM matrix writes:

,

to be compared with the experimental matrix [10]

The agreement is not that satisfactory, especially for the transitions between the quarks of muon family and those of tau family. This is not surprising given the crudeness of the calculations. One may consider, however, that the general trend is in the right direction and that the approach brings a valuable support to the granular structure of space-time and to the thermodynamic interpretation of the nature of fermions families. With the value the size of a bare particle is

,

perhaps not completely out of the reach of realizable colliders.