Publication status: Preprint has been submitted for publication in journal

Theory of Dimensional Randomness

Lucas Resende

https://doi.org/10.1590/SciELOPreprints.5282

Submitted on: 2022-12-18

Posted on: 2023-01-10 (version 3)
(YYYY-MM-DD)

Powered by TCPDF (www.tcpdf.org)

 Springer Nature 2021 LATEX template
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness

Lucas Lopes Resende1,1*
1* Programa
de Pós Graduação em Modelagem de Sistemas
Complexos, Escola de Artes Ciências e Humanidades -
EACH/USP, Rua Arlindo Béttio, 1000, São Paulo, 03828-000,
São Paulo, Brazil.

Corresponding author(s). E-mail(s): [email protected]

orcid.org/0000-0002-9078-6291;

Abstract
There is still no scientific consensus on the existence of objective ran-
domness. Understanding it is especially important for physics, as the
probabilistic nature of quantum mechanics is believed to be irreducible.
This theory aims to explain how the existence of true randomness is pos-
sible for the referential of a non-omniscient observer. The theory is built
on a toy model in two thought experiments. We propose the concept
of invisible and inaccessible laws as explanation for the occurrence of
unpredictability and states that it represents the physical reification of
Gödel’s incompleteness theorem. The article proposes that the geometry
of space-time is able to explain several physical phenomena as a sub-
set of true random events. Among these are the event horizon in black
holes, the hypersensitivity to initial conditions, and the Heisenberg’s
uncertainty principle. It is also able to refute the ideas of determin-
ism in classical physics as well as to explain why deterministic Turing
Machines are only able to execute pseudo-random number generators.

Keywords: True Randomness, Heisenberg’s Uncertainty Principle, Black

Holes, Chaotic Systems, String Theory

1 Introduction
For this paper, two definitions of randomness [1] are considered: the definition
of subjective randomness (or pseudorandomness) by Demókritos and that of

1
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

2 Theory of Dimensional Randomness

objective randomness (or true randomness) by Epikuros. The first is random-

ness associated with ignorance of causation and a characteristic related to the
lack of human acumen to determine the phenomena that influence the outcome
of an event. The second definition, on the other hand, concerns randomness by
causelessness, in which, even accurately measuring all initial conditions that
influence an event, it is impossible to predict precisely the outcome. A pseudo-
random event can become deterministic if the user is given more information,
however, for a truly random event, there is no extra information that makes
the event deterministic. The discovery of irreducibly random events is prob-
ably one of the most significant discoveries of the 20th century [2], but the
existence of objective randomness is a hypothesis not yet confirmed by science
[3]. The topic still involves much controversy, as it is very difficult to differ-
entiate between a random or pseudo-random process given the complexity in
proving the exhaustion of all possibilities that would influence the analysis of
an event. In the field of quantum mechanics, the Copenhagen interpretation
states that the collapse of the wave function is explained by the definitions
of true randomness [4]. However, some important scientists disagree with this
interpretation and philosophically take determinism as a principle, bealiving
that quantum mechanics is an incomplete science that lacks a more funda-
mental theory. Because of this, the theory of hidden variables was formulated,
which is based on realism and aims to describe how unobserved phenomena
would lie behind the probabilistic nature of quantum mechanics. One of the
biggest proponents of the Hidden Variables Theory was Albert Einstein (1879
- 1955). Through the EPR paradox [5], Einstein, Podolsky & Rosen tried to
reduce to absurdity quantum mechanics by claiming that quantum entangle-
ment implied the exchange of information violating the locality proposed by
Special Relativity [6]. However, Bell demonstrated, via Bell’s Inequality [7],
that Hidden Variable Theories that assume both realism and locality are not
consistent with the results of quantum mechanics. The Bell’s CSHS test [8]
is a way to test Bell’s inequality and invalidate local realism. Based on these
results, the Copenhagen interpretation gained strength and became widely
accepted in the scientific community. However, the Theory of Dimensional
Randomness is not a theory of hidden variables or a theory that proves that
the probabilistic nature of quantum mechanics is inherited from dimensional
randomness. But it is a geometric theory that proves dimensional randomness
as a subset of events inside the set of true random events. Dimensional Ran-
domness is a nomenclature proposed by this work for the uncertainty derived
from the existence of invisible and inaccessible dimensions. From the results of
our theory, interpretations can be created to describe objective randomness in
several areas such as Statistics, Philosophy, Astronomy, Quantum Mechanics
and Complex Systems. A numerical series is random if the smallest algorithm
capable of expressing it to a computer has the same number of bits of infor-
mation as the series itself [9]. For the validation of our theory, we take the
randomness described by our modeling and apply the distribution to an algo-
rithmic analysis of randomness in order to check whether they fit the definition
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 3

of randomness as described by Shannon’s Information Theory [10] and Kol-

mogorov’s complexity [11]. The non-comprehensibility of the sequences points
out that the future of the systems is not computable by a deterministic tur-
ing machine, that is, the smallest algorithm capable of processing the future
must receive all the information from the system, making it random from an
algorithmic point of view.

2 Thought Experiments
The Theory of Dimensional Randomness is formulated in two Thought
Experiments on four abstract systems, not strictly concordant with the full
understanding of physical laws. The systems are built as toy models, so that
it can further a concise explanation to the occurrence of dimensional random-
ness as a particularization of true randomness. The basic premise we assume
for the modeling of Thought Experiments is realism and the prohibition of
mechanisms that create direct uncertainty or random behavior in any observed
object. After developing the concepts and properties of the theory, we will
check whether these extend validity to real physical systems.

2.1 Thought Experiment I

In the Thought Experiment we model a deterministic system, named s1, which
will be used to generalize the concepts of the theory.

2.1.1 Description of s1
In the foundations of s1, there are 8 rules which are described below:
A The system s1 is modeled over a two-dimensional vector space.
B The system s1 allows for circles and squares.
C In the initial configuration of the system s1, there are 5 circles. The location
of these is determined by the polar coordinate system. Take the two-
dimensional plane and a set of angular coordinates ca of 60°, 120°, 180°,
240°, and 300° for each of the circles, find the radial coordinate cr by means
of the expression cr = ca/60 .
D The circlespattract each other according to an intensity vector described by
v = 6400/ (x1 − x2)2 + (y1 − y2)2 , in which (x1, y1) and (x2, y2) repre-
sent, respectively, the Cartesian coordinates of circles 1 and 2 in the plane,
and the circles move, at each instant of time, a distance directly propor-
tional to the resulting vector v: s α v where s describes the displacement of
a circle. Squares also attract each other in the same way as circles.
E If eight circles simultaneously reach a proximity threshold defined by a
constant p, the eight circles are transformed into a square.
F The squares and circles
p repel each other with intensity described by the
vector v = −6400/ (x1 − x2)2 + (y1 − y2)2 , in which (x1, y1) and (x2, y2)
represent, respectively, the Cartesian coordinates of the square and circle
in the plane. The circles move at each instant of time, a distance directly
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

4 Theory of Dimensional Randomness

proportional to the resultant vector v: sc α v in which sc describes the

displacement of a circle. Squares move at every instant of time, a distance
directly proportional to the resultant vector v/8: sq α v/8, in which sq
describes the displacement of a square.
G Between discrete time instants, represented by a constant, a new circle is
generated at the mean coordinate of the existing circles at s1.
H Rules A through G are the most fundamental laws of the system s1 in the
sense that there is no more fundamental rule that explains why they exist.
The existence of the fundamental rules C,D,E,F and G are independent,
that is, one rule could vary, without changing the other.

2.1.2 Inferences based in the simulation of s1

Based on the rules A-H, a computer simulation of s1 was made and we recorded
11 frames in the figure 1:

Fig. 1 s1 evolution
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 5

Frame 1 of figure 1 shows the s1 initial configuration with 5 green circles as

described by rule C. The black arrows illustrate the attraction vectors between
the circles. In the second frame, respecting rule G, observe the creation of a
circle, numbered 6, at the mid-coordinate of the circles present in the system.
In frame 3, circles 3 and 6 reach a limit of proximity in which it is difficult
to visually differentiate the position of each circle. Between frames 4 and 9 in
figure 1, rules D and G continue to exert an influence on the dynamics of the
system. Especially in frame 10, the system reached a degree of specialization
that allowed the manifestation of rule E for the creation of a square. In the
next frame, it is already possible to observe how circle 5 changes its trajectory
given the repulsion between circles and squares determined by rule F. The goal
of the Thought Experiment in s1 is to demonstrate, by means of sentential
calculus, how some phenomena are unpredictable for a fixed observer. To this
end, we include an observer that will follow the evolution of the system s1 for
a discrete duration of time. The properties of this observer are:
I In principle, the observer does not know the A-H rules of the system s1.
J The observer has access to all coordinates in space, however is limited to
watch the passage of time between the instants represented by frame 1 and
4 of figure 1.
K The observer can take measurements of any accessible variable in the system
s1 with infinite precision.
L The observer has an intellect capable of inferring, when observed, one or
more equations to perfectly describe any occurrence in the system s1.
M The observer witnesses the passage of time just as we do, being able to
freeze frames 1 to 11 with absolute precision.
In this Thought Experiment, we wish to analyze whether the observer has
the ability to perfectly predict the position of circle 5 in frame 11 of figure
1. To this end, we define a language of sentential calculus L composed of the
following propositional variables: L = {A, B, C, D, E, F, G, vA, vB, vC, vD,
vE, vF, vG}. The validity of sentential variables A, B, C, D, E, F, G means
that the rules A, B, C, D, E, F, G of s1 are known by the observer. On the
other hand, the sentential variables vA, vB, vC, vD, vE, vF, vG indicate,
respectively, that the corresponding rules A, B, C, D, E, F, G were acted upon
between the frames accessible for the observer. According to the properties of
s1 and of the observer, we postulate the inferential basis in 3.
The base statements 1-7 mean that if the rules A-G act at an accessible
instant, the observer deduces them due to the L property. Sentences 8-14 say
that if the rules have not manifested at s1, the observer does not have the
ability to deduce them, since the rule H of the system states that rules A,B,C ...
G are at the most fundamental level of definition and are independent of C and
G, so that knowing one rule does not contribute to understanding the other.
Sentences 15-19 are findings based on the constraint defined by the observer
property J and the observed events between frames 1 and 4 of the simulation.
For proving unpredictability for the observer’s referential, it is checked whether
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

6 Theory of Dimensional Randomness

Table 1 Inferential basis.

Number Statement Explanation
1 vA → A L property from observer
2 vB → B L property from observer
3 vC → C L property from observer
4 vD → D L property from observer
5 vE → E L property from observer
6 vF → F L property from observer
7 vG → G L property from observer
8 ¬vA → ¬A H rule from s1
9 ¬vB → ¬B H rule from s1
10 ¬vC → ¬C H rule from s1
11 ¬vD → ¬D H rule from s1
12 ¬vE → ¬E H rule from s1
13 ¬vF → ¬F H rule from s1
14 ¬vG → ¬G H rule from s1
15 vC J property from observer
16 vD J property from observer
17 vG J property from observer
18 ¬vE J property from observer
19 ¬vF J property from observer

the observer is able to predict exactly the position of circle 5 in frame 11 of

figure 1. As pre-requirement, we postulate the following statement:

Table 2 Prediction postulate.

Number Statement Explanation
20 i ∩ (C ∩ D ∩ E ∩ F ∩ G) ⇐⇒ p5 postulate

Statement 20 says that the observer predicts with absolute accuracy the
position of circle 5 (p5) if and only if he determines with infinite accuracy the
variables of the initial conditions (i) and knows all the rules C,D,E ... G that,
in some of the frames 1 to 11 of figure 1, directly influenced the position of
circle 5. By means of deductive mechanisms, the base sentences is developed
in order to prove the validity of p5:

Table 3 Inferences.

Number Statement Explanation

21 i K property from observer
22 C Modus ponens, statements 3 ∧ 15 ∴ C
23 D Modus ponens, statements 4 ∧ 16 ∴ D
24 G Modus ponens, statements 7 ∧ 17 ∴ G
25 ¬E Modus ponens, ¬vE → ¬E, ¬vE ∴ ¬ E
26 ¬F Modus ponens, ¬vF → ¬F, ¬vF ∴ ¬ F
27 ¬[i ∧ (C ∧ D ∧ E ∧ F ∧ G)] ¬F ∴ ¬[i ∧ (C ∧ D ∧ E ∧ F ∧ G)]
28 p5 → [i ∧ (C ∧ D ∧ E ∧ F ∧ G)] Biconditional elimination of 20
29 ¬p5 Modus Tollens, statements 27 ∧ 28 ∴ ¬ p5
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 7

Variable p5 is false since one of the requirements for p5 to be true is the

simultaneous validity of E and F, but they have been respectively invalidated
according to the union of the sentences 12, 18 and 13, 19 present in the basis.
Because of this, there is unpredictability of the position of circle 5 from the
observer’s point of view. The existence of unpredictability, in the context of
this Thought Experiment, is attributed to the specialization dynamics of the
system. At s1, note that the fundamental rules have always existed in the logic
of the system in an abstract way, but not all rules can immediately manifest
themselves on the material plane of s1. For example, in order to form a square,
it is necessary that there be eight circles simultaneously nearby, but in the ini-
tial configuration, there were only five circles and duly spaced at a distance
greater than p. Such conditions are physical limitations that act as impedi-
ments to the formation of a square in the primitive stages of s1. Consequently,
the knowledge accessible to the observer is not complete, since the properties
inherent to squares cannot be studied by observation, inference and validation
as required by the scientific method. Rules that cannot be studied at a given
instant of time or location in space are named as invisible and inaccessible
laws. They are a set of rules that potentially act on a system, but that by irre-
ducible physical limitation, prevent inference and validation of their properties
by a non-omniscient observer. The invisible attribute indicates that the law
is not observable in a given referential of time, and the inaccessible attribute
states that the law is not inferable and validable under the same referential.
The system s1 is modeled so that the rules E and F were invisible and inac-
cessible to any limited observer between the instants represented by frames 1
and 4 of the evolution of the system. Invisible and inaccessible laws represent
the physical reification of Gödel’s incompleteness theorem [12]. The existen-
tial cause of invisible and inaccessible laws for the referential of the observer
is the invisibility and inaccessibility of the time dimension of s1 because if
he could see through the time dimension, it would be possible to observe the
information of all the frames of the system, and therefore vE and vF would be
satisfiable. Taking, vE and vE → E; vF and vF → F, E and F can be deduced
as true. Judging i, C, D, E, F, G to be simultaneously true, p5 also is true.
However, It is not possible to generalize the results of Thought Experiment I
to describe the existence of invisible and inaccessible laws or unpredictability
in a real system, because the system s1 is a toy model and does not respect,
with fidelity, physical laws. Rule G, for example, violates the principle of con-
servation of energy, since by creating new circles the total energy of the system
increases. But this violation does not invalidate the inner results of the exper-
iment for two reasons: 1st. The inference of unpredictability is not dependent
on the increase of energy in the system, because for the rule of creating squares
to manifest itself two prerequisites are necessary: a minimum number of circles
in the system and proximity less than p between eight simultaneously nearby
circles. The mere lack of proximity between the circles is sufficient to act as
an impediment to the manifestation of G. 2nd. If the concept of energy was
not specified in the fundamental rules of s1 this concept does not exist in s1
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

8 Theory of Dimensional Randomness

which does not make it inconsistent according to its own rules. Real physical
systems are not used in experiment I as it is not possible to guarantee that
we are omniscient to describe nature. Section 3 delves deeper to describe this
requirement for the thought experiment. Thought Experiment I demonstrated
that Laplace’s Demon, proposed by Pierre Simon Laplace (1749-1827), is not
valid for all moments of the evolution of s1 because the existence of invisible
and inaccessible laws limits the intellect of any observer. To assert the validity
of Laplace’s Demon for any fundamentally deterministic system, it is necessary
to prove that there are no invisible and inaccessible laws that affect the future
states of the system. In the next step, Thought Experiment II is proposed and
the manifestation of invisible and inaccessible laws describes unpredictability
that fits a probability distribution.

2.2 Thought Experiment II

Thought Experiment II is an extension of Thought Experiment I. In this
new experiment, the goal is to continue the exploration of the unpredictabil-
ity observed in fundamentally deterministic systems. The focus is to explore
how unpredictability can fit a probability distribution. Thought Experiment
I showed that the position of circle 5 in frame 11 of figure 1 is unpredictable
for an observer bounded in the referential represented by frames 1 and 4 of
figure 1. However, an observer’s prediction for the position of circle 5 can be
adjusted if the observer follows the evolution of the system. The temporal
evolution of s1 makes frames 5 to 11 visible and increases the scope of knowl-
edge accessible to the observer, making it possible to resume the deterministic
description for such a system if the observer’s referential is extended. In this
new experiment, we model a fundamentally deterministic system architecture
such that the observer is forced to use probability distributions to describe
certain phenomena.

2.2.1 Description of s2, s3 and s4

We define three parallel systems named s2, s3 and s4. All systems are abstract
and described at the most fundamental level by deterministic rules not strictly
in accordance with the understanding of physical laws.
The system s2 is described by the following rules:
A The system s2 is modeled over a two-dimensional vector space.
B The system s2 allows for circles and squares.
C In the initial configuration of the system s2 there are 8 circles. The location
of these is determined by the polar coordinate system. Take the two-
dimensional plane and a set of angular coordinates ca of 45º, 90º, 135º, 180º,
225º, 270º, 315º, and 360º for each of the circles, find the radial coordinate
cr by means of the expression cr = ca / 45º .
D The circles attract each other
p according to an intensity vector described
by the equation v = 6400/ (x1 − x2)2 + (y1 − y2)2 , in which (x1,y1) and
(x2,y2) represent, respectively, the Cartesian coordinates of circles 1 and
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 9

2 in the plane, and the circles move at every instant of time a distance
directly proportional to the resulting vector v: s α v, where s describes the
displacement of a circle. Squares also attract each other in the same way as
circles.
E If 8 circles simultaneously reach a proximity threshold defined by a constant
p, the 8 circles are transformed into a square.
F The squares and circles
p repel each other with intensity described by the
vector v = −6400/ (x1 − x2)2 + (y1 − y2)2 , in which (x1,y1) and (x2,y2)
represent, respectively, the Cartesian coordinates of the square and the
circle in the plane. The circles move at every instant of time, a distance
directly proportional to the resultant vector v: sc α v, in which sc describes
the displacement of a circle. Squares move at every instant of time, a dis-
tance directly proportional to the resultant vector v/8: sq α v8, in which
sq describes the displacement of a square.
G Between discrete time instants represented by t, a new circle is generated
at the coordinate representing the simple arithmetic mean between the
coordinates of the circles and squares existing in the systems s2 and s3.
H Between discrete instants of time represented by t, the oldest square in the
system s2 is erased if there are more than 2 squares in the system.
I The variables and initial conditions of the system s2 are invisible and inac-
cessible to any observer external to s2 with the exception of communication
with s4.
The system s3 is described by the following rules:
A The system s3 is modeled over a two-dimensional vector space.
B The system s3 allows for circles and squares.
C In the initial configuration of the system s3 there are 16 circles. The location
of these is determined by the polar coordinate system. Take the two-
dimensional plane and a set of angular coordinates ca of 22.5º, 45º, 67.5º
...360º for each of the circles, find the radial coordinate cr by means of the
expression cr = ca / 22.5º . It is defined that those closest to the arbitrary
center are older, in chronological order, compared to the others.
D The circles repel each other p according to an intensity vector described by
the equation v = −6400/ (x1 − x2)2 + (y1 − y2)2 , in which (x1,y1) and
(x2,y2) represent, respectively, the Cartesian coordinates of circles 1 and
2 in the plane, and the circles move at every instant of time a distance
directly proportional to the resultant vector v: s α v, in which s describes
the displacement of a circle. Squares also repel each other in the same way
as circles.
E If 8 circles simultaneously reach a proximity threshold defined by a constant
p, the 8 circles are transformed into a square.
F The squares and p circles attract each other with intensity described by the
vector v = 6400/ (x1 − x2)2 + (y1 − y2)2 , in which (x1,y1) and (x2,y2)
represent, respectively, the Cartesian coordinates of the square and the
circle in the plane. The circles move at every instant of time, a distance
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

10 Theory of Dimensional Randomness

directly proportional to the resultant vector v: sc α v, in which sc describes

the displacement of a circle. Squares move at every instant of time, a dis-
tance directly proportional to the resultant vector v/8: sq α v8, in which
sq describes the displacement of a square.
G Between discrete instants of time represented by t, a new circle is generated
at the coordinate representing the simple arithmetic mean between the
coordinates of the circles and squares existing in the systems s2 and s3.
H Between discrete instants of time represented by t, the oldest circle in the
system s3 is erased.
I The variables and initial conditions of the system s3 are invisible and inac-
cessible to any observer external to s3 with the exception of communication
with s4.
The system s4 is described as follows:
A The system s4 receives the mean coordinate of the circles and squares in s2.
B The system s4 receives the mean coordinate of the circles and squares in s3.
C The system s4 calculates the simple arithmetic mean between averages of
the coordinates of s2 and s3 and sends the result to s2 and s3.
D The system s4 does not allow for observers.
The characteristics of the observer are described below:
A The observer inhabits s3 but is external to s2 and s4.
B The observer can take measurements of any accessible variable in the system
s3 with infinite precision.
C The observer has an intellect capable of inferring one or more equations to
perfectly describe any occurrence of s3 with the exception of invisible and
inaccessible laws.
D The observer is initially unaware of the fundamental rules of s2, s3 and s4.
E The observer witnesses the passage of time just as we do.
According to the property A of the observer and property I of the system
s2, it is deduces that the observer has physical limitations to deterministically
infer the rule G of the system s3 since it presents direct dependence on vari-
ables of s2 and, according to property I of s2, is invisible and inaccessible to
any observer external to the system. Because of this physical limitation, the
rule G of s3 is an Invisible and Inaccessible Law to the observer referential.
Based on the definitions presented for the 3 systems, it was implemented, in
a computer simulation, the execution of s2, s3 and s4. The coordinates of the
generated circles in s3 were registered for further analyses. It is evaluated how
the interference of Invisible and Inaccessible Law G from s3 influences the
predictions of an observer for the positions of circles. Three numerical experi-
ments were carried out with the observer in order to verify the validity of the
following sentences:
1. The accessible knowledge of s3 does not allow the observer to predict with
absolute certainty the future position of the circles in the system.
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 11

2. The observer can use pattern recognition methods that remove uncertainty
from the future position of the circles.
3. The possible inaccuracy in observer predictions recorded by experiments
does not stem from floating point inaccuracy of our experiments.
Below, it is described the criteria for modeling the three experiments: In
numerical experiment 1, disregarding the existence of invisible and inacces-
sible laws, it was used all the knowledge of systems s2, s3 and s4 to make
predictions of the positions of the circles in space and time t + 1 of the sys-
tem. The execution of this experiment, although not feasible from the point
of view of the observer, aims to guarantee the validity of sentence 3 and the
validity of the conclusions reached in numerical experiments 2 and 3. Next,
numerical experiment 2 is performed. In this experiment, there is an observer
capable of using only the knowledge of visible and accessible laws to make pre-
dictions. It is known that rule G of system s3 is invisible and inaccessible to
any observer, however, for the other rules, there is no restriction that prevents
knowledge of the rule. Based on property C, the observer is able to deduce the
laws A,B,C,D,E,F and H of the system s3 and use them to make predictions
of the circle position in the instant t + 1. Numerical experiment 2 aims to
validate sentences 1 and 2. In numerical experiment 3, as in numerical exper-
iment 2, it was considered that the observer is able to use only the knowledge
of visible and accessible laws to make predictions, but we added the ability to
make assumptions about what would be the eventual invisible and inaccessible
laws of the system. Possibly, the assumptions increase the assertiveness of the
predictor method used. Experiment 3 re-evaluated sentences 1 and 2. Numer-
ical experiments 1, 2 and 3 were performed on the same computer and under
parity conditions of comparison.

2.2.2 Execution of numerical experiments in the system s3

In this section, we present the results of the application of the numerical
experiments proposed in the mental experiment II. The first is the applica-
tion Numerical Experiment 1, which applies a predictor that is capable of
using the complete knowledge of the systems s2,s3 and s4. Considering a set
of predictions made in the instant t before the appearance of a new circle,
we recorded 75500 deviations of the expected and observed circle position in
the time t + 1 of s3. In 75232 observations or 99.65% of the cases, there was
no deviation and the difference between the actual value and the forecast was
exactly equal to zero. In 268 observations or 0.35% of the cases, there was a
deviation between prediction and observation. However, we do not attribute
the existence of deviation in these observations to floating-point imprecision.
The explanation for the existence of these deviations is a limitation of the pre-
diction method that is taken in parallel with the execution of the system s2,
s3 and s4. In cases where there was a deviation, the prediction considered, at
time t, an old value of the mean that was present in the s4 system, but which,
according to property A, received between times t and t + 1 an update of the
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

12 Theory of Dimensional Randomness

mean of known coordinates of the system s2 and that caused the predictions
to deviate from the real value at t + 1. In general, the prediction method that
considers the use of all knowledge presented an average of the sum of deviations
(x,y) of 0.00018 units of distance of s3 and 0.00017 of variance. In order for
any recorded inaccuracy not to be attributed to the limitation of the numerical
experiment, it is necessary that the deviations are significantly distant from
those recorded in numerical experiment 1. We proceed with the application
of numerical experiment 2 in which the observer is capable of using only the
accessible knowledge to make predictions. We recorded, for the experiment,
500 thousand observations. Figure 2 shows the dispersion of deviations from
the predictor method used. In 99.34%, there was a deviation between the pre-
diction and the actual value. Compared with the reference predictor method,
the difference between the number of perfect predictions is 98.9% in favor of
the method used in numerical experiment 1. The average of deviations (x,y)
added in this new experiment was 0.04881, an increase of 271 times in relation
to that recorded in numerical experiment 1. Applying the Wilcoxon Signed
Rank Test, we verified that the p-value is equal to 0, asserting that there are
significant differences between the means.

Fig. 2 Deviations of observer predictions in system simulation s2, s3 and s4.

Numerical experiment 2 validates sentence 1 and confirms that the use of

knowledge of s3 accessible to the observer does not allow him to predict with
absolute certainty the future position of the circles in the system.
In numerical experiment 2, all accessible knowledge to the observer is used
to make predictions. However, we totally disregard assumptions that could be
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 13

made about what the invisible and inaccessible law G of s3 would be. First,
let’s assume that the observer perfectly supposes the existence of s2 and also
predicts that the average coordinate of s2 is influencing the coordinates of the
circles created in s3. This assumption, although strictly correct, cannot be con-
sidered valid for this experiment, as it does not contribute to the improvement
of the prediction method taken by the observer. After assuming the existence
of s2, it would then be necessary to assume the initial conditions of the exper-
iment, which include the position and number of circles in s2, the attraction
between the circles, all the constants that influence the attraction and every-
thing related to system s2. While this is not a valid assumption, we evaluate
the invisible and inaccessible law G to estimate what assumption the observer
could make to lessen the impact of ignoring G:
G: Between discrete instants of time represented by t, a new circle is generated
at the coordinate representing the simple arithmetic mean between the
coordinates of the circles and squares existing in the systems s2 and s3.
In the excerpt above, we verify that the invisible and inaccessible law G is
dependent on a constant t and on the coordinates of the circles and squares
in s3, information that is accessible to the observer. Based on this finding, we
infer that, to make the predictions, the observer adds rule G2 to the set of
known rules A-H:
G2: Between discrete instants of time represented by t, a new circle is generated
at the coordinate representing the simple arithmetic mean between the
coordinates of the circles and squares existing in the system s3.
We proceed with the execution of numerical experiment 3 which will use the
accessible knowledge for the observer plus the assumption of G2. For this new
experiment, we recorded 500 thousands observed deviations for the predictions
and displayed, in figure 3, a heat map that represents the deviations after the
adjustment assuming G2:
Visually, we observe that the deviations concentrated in figure 3 are lower
than those recorded in figure 2. We verified that, in 99.9% of the cases, there
was a difference between the predicted value and the actual value, a differ-
ence of 99.55% compared to the one recorded in numerical experiment 1. The
average of the deviations (x,y) added in this new experiment was 0.05142, an
increase of 300 times in relation to that registered in experiment 1. Although
visually the results give the impression that the new method is a better pre-
dictor, the mean value of the deviations recorded in this new experiment was
numerically higher than that recorded in experiment 2. To explain this appar-
ent contradiction, we need to consider the distribution of values by interval. In
numerical experiment 2, the range that contains 90% of the x-component data
is approximately [-0.0327;0.0327]; the 99% confidence interval is [-0.526;0.526]
and the 99.9% confidence interval is [-0.867;0.867]. After adjusting G2 in exper-
iment 3, we found that the interval that contains 90% of the observations is
smaller compared to before [-0.0205;0.0205], however the intervals of 99% and
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

14 Theory of Dimensional Randomness

Fig. 3 Deviations of observer predictions in the simulation of s2, s3 and s4 after the
assumption of G2.

99.9% are wider and with respective values [-0.664;0,664] and [-1.187;1.187].
According to the data, we found that the method that assumes G2 has a higher
probability of approximating any prediction to the real value. In about 70% of
the cases, the magnitude of the deviation in units of distance of s3 is smaller
compared to the method that neglects G2, however, when the method that
assumes G2 deviates significantly from the real value, the deviations are more
accentuated compared to the previous method. Because of this, we did not see
improvements in the mean deviation of the new predictor. Using the Wilcoxon
Signed Rank Test, we calculated that the p-value is equal to 0 and that there
are significant differences between the prediction methods, with the previous
method being 6% more accurate than this one. These results show that the
assumption that could be made did not bring the observer closer to a deter-
ministic description and reinforces the validity of sentence 1 that the use of
accessible knowledge in the system does not allow predicting the future posi-
tion of the circles. It is still necessary to judge the validity of sentence 2 which
says that the observer may use a pattern recognition method that removes
the uncertainty of the future position of the circles. To evaluate the previous
sentence, we need to resort to definitions and numerical tests of algorithmic
randomness. A numerical series is random if the smallest algorithm capable of
expressing it to a computer has the same number of bits of information as the
series itself [9]. Chaitin’s statement can be measured by Kolmogorov Complex-
ity. The Kolmogorov complexity K(x) of x is the length of a shorter string y
such that x can be computed from y by a Fixed Universal Turing Machine [13].
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 15

Table 4 Mutual Information between X1, X2 ... X12

MI X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 X11 X12

X1 .063 .018 .000 .000 .000 .000 .000 .001 .000 .000 .000
X2 .065 .076 .007 .001 .000 .001 .001 .000 .001 .000 .001
X3 .024 .078 .016 .001 .000 .001 .000 .000 .000 .000 .000
X4 .010 .011 .017 .001 .002 .001 .002 .000 .002 .001 .002
X5 .010 .007 .004 .002 .002 .001 .002 .002 .003 .003 .002
X6 .011 .007 .003 .002 .000 .001 .002 .000 .000 .002 .001
X7 .011 .008 .002 .001 .000 .000 .000 .000 .002 .003 .002
X8 .011 .007 .002 .002 .003 .002 .000 .001 .001 .002 .001
X9 .011 .008 .002 .002 .000 .001 .003 .001 .000 .000 .000
X10 .011 .008 .002 .001 .001 .001 .003 .003 .000 .001 .003
X11 .011 .008 .001 .000 .003 .001 .001 .000 .002 .001 .000
X12 .011 .009 .003 .004 .002 .000 .002 .002 .000 .001 .002

However, K(x) is non computable [14]. Kolmogorov Complexity was extended

to satisfy the strong invariance property and is related to the mutual informa-
tion metric which satisfies the same property [15]. A robust test is proposed
to check whether a sample distribution is random [16]. The test consists of
taking a perfect random generator as a comparison measure. Then it is pro-
posed to take the Kolmogorov-Smirnov test [17] to verify the hypothesis that
a sample from the perfect generator and the tested sequence obey the same
distribution. To perform the randomness test, it is taken the first 12 decimal
places of the observed deviations in the numerical experiment 2 and is con-
sidered that each decimal place comes from a random variable Xn, in which
n represents the position of the digit in the decimal places of the continuous
number. The metric of mutual information is evaluated to check which digits
are dependent. The variables Xn that have some degree of mutual dependence
were disregarded from the randomness test because there is no guarantee that
they respect the property of strong invariance, which must be satisfied when
considering Kolmogorov complexity. The table 4 shows the calculated mutual
information values between pairs of Xn.
The mutual information value means that by knowing one variable, you
can be certain of a percentage of digits of information from the other variable.
If the value of mutual information is 0, there is independence between the
variables. Notice in table 4 that the values of dependence between pairs of
variables Xn are generally low, the values are bounded between 0.000 and
0.076. We establish a criterion to interpret which values suggest dependence
and independence between the variables. To this end, the chi-square test of
independence [18] is taken at 95% confidence level. The values are recorded
in table 5. The Bonferroni Correction [19] was applied to the p-values. In the
cases where the letter D appears, the test result allowed us to reject the null
hypothesis and confirm the dependence between the variables. In the cases
represented by the letter I, the test did not reject the null hypothesis and it is
not possible to detect dependence between the two variables.
Notice, in table 5, that between the variables X1 and X5, the test detected
that there is dependence, while, beyond the variables X5, no dependence was
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

16 Theory of Dimensional Randomness

Table 5 Chi-square test of independence between X1, X2 ... X12

RV X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 X11 X12
X1 D D D I I I I I I I I
X2 D D D D I I I I I I I
X3 D D D D I I I I I I I
X4 D D D I I I I I I I I
X5 I D D I I I I I I I I
X6 I I I I I I I I I I I
X7 I I I I I I I I I I I
X8 I I I I I I I I I I I
X9 I I I I I I I I I I I
X10 I I I I I I I I I I I
X11 I I I I I I I I I I I
X12 I I I I I I I I I I I

Table 6 Metrics of Kolmogorov-Smirnov Test between randint and X1, X2 ... X12
RV D Statistic p-value adjusted p-value
X1 0.852253 0 0
X2 0.69383 0 0
X3 0.38587 0 0
X4 0.10114 0 0
X5 0.01003 <<0.01 <<0.01
X6 0.00074 0.999 1
X7 0.00208 0.232 1
X8 0.00099 0.968 1
X9 0.00150 0.629 1
X10 0.00219 0.183 1
X11 0.00228 0.148 1
X12 0.00169 0.471 1

detected. Based on this, we proceed to apply the randomness test proposed by

Shen (2020) on the variables X1, X2 ... X12. The random module of python
[20] is taken as a reference generator. Then, 7 samples of 500000 independent
and uniformly distributed integers are generated in the interval between 0 and
9 by means of the randint function. To each of these samples the Kolmogorov-
Smirnov test is applied to verify, at 95% confidence, that the sampling of
values of Xn and the sampling of values of the random module follows the
same distribution. For the variables X5, X6 ... X12, the test failed to reject
the null hypothesis and suggests that Xn is distributed like a discrete uniform
distribution in the interval [0,9]. See the metrics of Kolmogorov-Smirnov test
in the table 6:
Based on the results of table 4 and 6, the combination of variables X6, X7
... X12 did not fail any of the tests. We ensure that knowledge of a variable
Xn does not decrease the uncertainty of a variable Xn+c, where n ≥ 6 . Ran-
domness test includes the evaluation of serial correlation metric. The serial
correlation measures the relationship between a variable and a lagged ver-
sion of itself. It is a mathematical tool to detect periodicity in the data. The
serial correlation measure has already been used for randomness testing, with
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 17

Table 7 Maximum serial correlation varying the h parameter in the range [1,500000]

RV max(R1,R2,. . . ,R50250) RV max(R1,R2,. . . ,R50250)

X1 0.270789 X7 0.007012
X2 0.506885 X8 0.005890
X3 0.472550 X9 0.006338
X4 0.158905 X10 0.005800
X5 0.005646 X11 0.005709
X6 0.006055 X12 0.006359

emphasis on a serial correlation permutation test proposed by Wald & Wol-

fowitz [21]. The Rh serial correlation is measured for the RV X1, X2 . . . X12.
In table 7, It is shown the maximum Rh for the RV varying the lag parameter
h in the range of 1 to 500000. We evaluate the maximum Rh because it repre-
sents the worst case of periodicity observed between any two digits in the lag
interval [1,500000].

Fig. 4 Probability density function of max(R1,R2,...,R500000) for reference random

generator.

To interpret the values of table 7, it’s shown in figure 4 the values dis-
tribution of max(R1,R2,. . . ,R500000) built from running 20000 independent
samples of 500000 observations of the random reference generator. The esti-
mated 99% confidence interval is [0.00509; 0.00706] and can be seen in the
vertical lines of figure 4, the variables X1, X2 . . . X4 failed the serial cor-
relation test because the values are significantly distant from the confidence
interval. Considering the random variables X5, X7, X8 . . . X12, the maximum
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

18 Theory of Dimensional Randomness

values of Rh are inside the 99% confidence interval. Based on these metrics,
the variables X5, X7, X8 . . . X12 did not present regularities that significantly
differentiate them from a random reference generator in the analyzed interval.
The metrics point out that starting from the sixth decimal place, there is no
evidence of patterns in the variables and the system behaves like a real ran-
dom system. It is not ensured that a pattern may not occur in a future state
of the system. But while the pattern does not appear, the observer cannot use
a pattern recognition algorithm to take a deterministic description of s3, so
there is a moment in which there is unpredictability in the system. Actually,
we expect that a pattern may never occur, since the experiment is inspired by
Brownian Motion [22]. The distribution in figure 2 is extremely relevant to this
experiment, because even if it is possible for us to explain the causes for the
deviations of predictions, due to the invisibility and inaccessibility of G for the
observer’s referential in s3, there is not an existential cause for the imprecision
observed. For any non-omniscient observer in s3, the complete description of
G is inaccessible, with Gobs being the only possible description for G:
Gobs: Between discrete instants of time represented by t, a new circle is generated
at the coordinate representing the simple arithmetic mean of the circles and
squares present in s3 . To the value , add a random offset (x,y) sampled
from f3(x,y), in which f3(x,y) represents the probability density function of
the deviation from .
Because Gobs contain a probability function, there are some deterministic
systems in which randomness is irreducible for its description. The definition
of objective randomness describes a random event that is associated with the
absence of cause. Under an observer’s frame of reference, objective random-
ness is described as a relationship between the space-time topology and the
physical laws of a system. The actuation of invisible and inaccessible laws fits
the definitions of objective randomness, because it is not possible, under the
observer’s referential, to distinguish the class of events that have a fundamen-
tally unknown cause from the class of events that are fundamentally absent
of cause. In section 3, we provide a geometric interpretation that deepens the
referential view of objective randomness.
Code Availability: The python code used to run the Thought Experiments
I and II with the instructions to replicate the experiments are available on
Github1 .

3 Geometric Interpretation of Objective

Randomness
Thought Experiments I and II is generalized to a geometric interpretation
which will be extended to describe true randomness in real physical systems.
Figure 5 exhibits the block-universe representation of Thought Experiment
I. The toy model used this representation instead of Einstein’s theory of 4D

1
https://github.com/LucasLopesSI/ThoughtExperiments
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 19

Fig. 5 s1 geometry in Thought Experiment I.

space-time because the block universe provides extra capabilities for the 2D
observer to describe the universe. The first, represented by the J property of
the observer, is the ability to measure the state of all space variables at a given
instant. But according to the Special Theory of Relativity [6] that capability
is not feasible because the light speed c is an upper limit for the propaga-
tion of information. The second, represented by M property of the observer,
is the capability to despise the time dilation proposed by the Special Theory
of Relativity. We want the observer to have the least physical limitations to
describe the universe deterministically, being any eventual physical limitation
only inherited by our proposed toy model. In figure 5, the two spatial dimen-
sions x and y represent the coordinates at which the material entities of s1
can take at a given instant. The time dimension t, on its turn, describes the
stack of spatial frames of the system, and it is possible to describe the position
of any object in space-time assuming that there is visibility and accessibility
of the time dimension. Two observers are also represented in the figure. The
first observer, represented by a white eye, is limited to 2 dimensions and is
therefore dimensionally inferior to the second. The lower observer inhabits s1
and is able to measure the variables present in the system at a given instant
with infinite precision. The second observer, represented by the black eye, is
dimensionally superior to the first since it can visualize the space-time of s1
in 3 dimensions. In figure 5, the dimensions visible and accessible to the lower
observer are represented by the blue color and the invisible and inaccessible
dimensions are transparent. Unpredictability for an observer occurs when he
encounters limitations that prevent the visibility and accessibility of a dimen-
sion. In this sense, he does not have available the set of information that would
allow him to reconstruct the rules of the system.
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

20 Theory of Dimensional Randomness

Fig. 6 Visibility and accessibility of the dimensions after specialization of s1

The limitation to reconstruct the rules of s1 in Thought Experiment I is

the invisibility and inaccessibility of the time dimension for the lower observer.
Hypothetically possessed with all knowledge of the fundamental laws of s1, we
reconstructed the information present in all dimensions of the space-time of s1
and became the 3D observer of figure 2. It was possible to work with the Closed-
world assumption [23] and prove that, at certain instants, there is not enough
specialization of s1 so that the lower observer can reconstruct the fundamental
knowledge that we achieved as the 3D observer of s1. All the fundamental rules
that could not be deduced or inferred due to invisibility and inaccessibility of
the time dimension were called invisible and inaccessible laws and, when they
act on s1, they force the observer to deal with irreducible unpredictability. In
Thought Experiment I, it was not possible to derive a probability distribution
that describes unpredictability. This is because the invisible and inaccessible
dimensions are connected to the lower observer through the time component,
and from the specialization of the system, the invisible and inaccessible law
acts. For a sufficiently intelligent observer, it is possible to deduce the law
that was previously invisible and inaccessible to avoid unpredictability. Figure
6 exemplifies a case where s1 is specialized to the point that there are no
invisible and inaccessible laws in the evolution of the system. In the example,
the lower observer is located in a frame that allows all dimensions to be visible
and accessible, represented by blue color. We as the higher observers of this
experiment can say that the lower observer’s ability to make predictions is
equal to ours. However, he is not able to know this information, because since
he is embedded in the system, he must assume the Open-World assumption
[24] and cannot exclude the possibility of invisible and inaccessible laws acting
in the future.
Thought Experiment II is designed in order to explain how unpredictability
fits a probability distribution. In this new experiment, a system architecture
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 21

Fig. 7 Geometry of s2,s3 and s4 in Thought Experiment II.

Fig. 8 General mechanism for the existence of objective randomness according to Theory
of Dimensional Randomness.

with extra spatial dimensions is proposed using the parallelism of s2 and s3.
Moreover, there is an additional system s4 that is responsible for the com-
munication between s2 and s3. A two-dimensional observer is embedded in
the space-time of s3. We prove that to describe the position of circles in s3,
the lower observer must use a probability distribution. Figure 6 shows the
architecture of the Thought Experiment II:
Figure 8 describes the mechanism that was able to give rise to objective
randomness in thought experiments I and II. Initially, there is the existence
of an observer and one or more physical limitations. In thought experiment I,
the physical limitations for the observer is the possibility of living only in the
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

22 Theory of Dimensional Randomness

present and for a limited duration between frames 1 and 4. Furthermore, the
system was not specialized enough for the observer to theoretically or exper-
imentally predict the repulsion rule between squares and circles. In thought
experiment II, the observer is only able to live in the s3 system, whereas the
systems s2 and s4 are parallel to its existence and exchange information in
a way that the observer cannot observe or infer. Physical limitations vary
according to the type of system in which the observer is inserted. Using real
systems, we will address some physical limitations in section 5 of this article.
The observer’s association with irreducible physical limitations is the cause for
the existence of invisible and inaccessible dimensions. And the laws that act in
these dimensions, inherit the invisible and inaccessible attributes, being nec-
essarily described by probability theory. Thought experiments I and II were
modeled on hypothetical and non-physical systems, because in this way we
guarantee omniscience of the events that happen in the system. As superior
observers, the Closed-World Assumption is used to distinguish the class of
events with a fundamentally unknown cause from the class of events with an
unknown cause because of the observer’s lack of acumen. Using real physi-
cal systems, we must assume the Open-World Assumptions since we are 3D
observers immersed in a 4D universe, then the distinction between the two
classes of events becomes fuzzy.

4 Theorem of Dimensional Randomness

Based on the insights and experimental results of thought experiments I and
II, we propose the conjecture of dimensional randomness.

Conjecture of Dimensional Randomness:

Dimensional randomness comes from the existence of invisible and inacces-
sible dimensions in space-time.

Hypothesis 1:
There is a dimensionally random probability distribution Fx (x) if 1.1 or
1.2 hypothesis are true:

n
f (x) = ∃Fx (x), if [(Rt1,n t2,m t1,n t1,n
obs ∧ Rreal | t1 + n < t2 + m) ∨ (Robs ̸= Rreal )]
(1)

Hypothesis 1.1:
The Minkowski space denotes that the number of dimensions (t1+n) in
observable space is irreducibly smaller than the number of dimensions in
real space (t2+m). The dimensions of real space that do not participate in
observable space are the invisible and inaccessible dimensions.

Rt1,n t2,m
obs ∧ Rreal | t1 + n < t2 + m (2)
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 23

Proof of the conjecture for hypothesis 1.1:

Take the following equivalence for the observer:

f (x, y, ..., z) ≡ f (x, y, z, ..., z1)

f (x, y, ..., z) ≡ f (x, y, z, ..., z2)
(3)
...
f (x, y, ..., z) ≡ f (x, y, z, ..., zn)

The equivalence 3 is true from the observer’s point of view because real
space has more degrees of freedom compared to observable space. In the observ-
able space, the invisible and inaccessible dimensions mathematically contract
to be represented only by the dimensions (x,y ... z).

f (x, y, z, ..., z1) = ω1

f (x, y, z, ..., z2) = ω2
(4)
...
f (x, y, z, ..., zn) = ωn

Let’s assume that the variations in the parameters (x, y, z...z1), (x, y, z...z2)
and (x, y, z...zn) produce strictly different solutions, so that:

f (x, y, z, ..., z1) ̸= f (x, y, z, ..., z2) ̸= ... ̸= f (x, y, z, ..., zn) (5)

4 presents different outputs ω1 , ω2 , ω3 under the variation of the input

parameters (z1, z2. . . zn) represented by the invisible and inaccessible dimen-
sions. The statements in 3 and 5 may appear contradictory, but they actually
aren’t. The equivalence in 3 is only valid for the inferior observer perspective
and can only be written by a superior observer that sees simultaneously the
real space-time Rt2,m t1,n
real and the inferior observer from Robs . But by the per-
spective of the inferior observer, 3 is fundamentally unknown, and by the full
perspective of Rt2,m
real , 5 is strictly correct and 3 does not make sense.
From sentence 3, we know that f(x,y,z ... z1), f(x,y,z ... z2) . . . f(x,y,z ...
zn) are equivalent states of f(x,y,z). Replacing the terms of 4 by its equivalent
term in 3, we have:

f (x, y, ..., z) = ω1
f (x, y, ..., z) = ω2
(6)
...
f (x, y, ..., z) = ωn
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

24 Theory of Dimensional Randomness

6 shows that the inferior observer will see the output ω1 , ω2 , ..., ωn for the
event f(x,y, ..., z). The function definition says that an element of the domain
A cannot be related to more than one element of the codomain B. Applying
this definition, we conclude that:

ω1 = ω2 = ... = ωn (7)
However, assuming f(x,y, ..., z) as a function we get to a contradiction
between sentences 5 and 7. To maintain the validity of the assumption 5, we
must consider that f(x,y, ... ,z) is not a function, and an element of domain A
may have more than one correspondence in co-domain B. We therefore have:

f (x, y, ..., z) = ω1
f (x, y, ..., z) = ω2
... (8)
f (x, y, ..., z) = ωn ,
ω1 ̸= ω2 ̸= ... ̸= ωn

The definition of random event says that the same initial conditions (x,y,
... ,z) can originate different output states ω1 ̸= ω2 ̸= ... ̸= ωn . We note that
f(x,y, ... ,z) respects this definition, thus representing a random event Fx (x)
under the observer’s frame limited to the space Rt1,n obs .

Theorem 1.1:
An observable space that is dimensionally inferior to the true space gives
rise to dimensional random events described by Fx (x).

[Rt1,n t2,m
obs ∧ Rreal | t1 + n < t2 + m] =⇒ Fx (x) (9)

Hypothesis 1.2:

Rt1,n t1,n
obs ̸= Rreal (10)
Hypothesis 1.2 says that there is a true space which has the same numbers
of dimensions as the observable space, but which are strictly different.

Proof of the conjecture for hypothesis 1.2:

Rt1,n t1,n t1,n

obs ̸= Rreal ≡ ∃[α ∈ Rreal ∧ α ∈
t1,n
/ Robs ] (11)
11 says that the difference between the spaces is equivalent to defining
that there is an α that belongs to the real space, but does not belong to the
observable space. The equivalence can be made, as the observable space cannot
be larger than the true space. As the spaces are different, the only remaining
alternative is Rt1,n t1,n
obs ⊂ Rreal
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 25

¬f ∨ ¬α =⇒ ¬f (α) (12)
We postulate that to predict the outcome of an event determined by a
hypothetical function f (α) it is necessary to know the function f and the input
parameter α.

¬f (α), 11 ∧ 12 (13)
13 states that f (α) is not true because 11 states that α does not belong to
the observable space.

α ≈ α1 + ϵ, α ∈ Rt1,n t1,n t1,n

real ∧ α1 ∈ Robs ∧ ϵ ∈ Robs (14)
But, the α may be approximated by an α1 ∈ Rt1,n
obs plus an estimated error ϵ.

ϵ ∼ Fx (x) (15)
The ϵ is distributed like a probability density function Fx (x).

Theorem 1.2:

Rt1,n t1,n
obs ̸= Rreal =⇒ f (α1 + Fx (x)) (16)
A real space that has its dimensions partially observable gives rise to an
error distributed like Fx (x).

5 Applications of the Theory of Dimensional

Randomness
This section extends the application of the Theory of Dimensional Randomness
to explain phenomena and constraints observed in subfields of Mathematics,
Physics, and Computing.

5.1 Why do conventional computers use pseudo-random

number generators (PRNG)?
According to the Theory of Dimensional Randomness, the requirement for the
creation of an objectively random function is the invisibility and inaccessibil-
ity of one or more dimensions in space for the referential of a non-omniscient
observer and associated with the operation of at least one Invisible and Inac-
cessible Law. The first obstacle to creating objective randomness in modern
computers occurs because the architecture and models of computation are
specified. In this way, computers are implemented based on models controlled
by humans, making us able to use the closed world assumption to deny the
existence of objective randomness. We even understand the most fundamen-
tal level of a computer’s information and, as a result, become an omniscient
observer of the computations that take place. However, in physics, the process
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

26 Theory of Dimensional Randomness

is opposite. Theories and predictions are modeled so that they fit the observa-
tions of the real world. The open world assumption must be taken, implying
that ignorance of a certain phenomenon is not sufficient for the denial of it.
The second drawback in using conventional computation to generate objective
randomness occurs because the computations produced by a machine are in
the same spatial dimensions that are inhabited and accessible to us. Because
of this, there is no capability to create physical constraints on the operation
of invisible and inaccessible laws. To deal with this problem, we strive to cre-
ate invisibility of the random root, and complexification of the function used
for random number generation, however invisibility is a weaker property com-
pared to inaccessibility, allowing computations of a pseudo-random function
to be reduced to deterministic computations by an observer that breaks the
invisibility of the process.

5.2 Decay of Deterministic Equations

The operation of invisible and inaccessible laws that are not considered when
writing an equation adds randomness to the future predictions of a system. The
evolution of this phenomenon can occur in a smooth manner, so that predic-
tions lose accuracy over time. We call this process the Decay of Deterministic
Equations, which is defined as the transition from a deterministic equation
to a random function. The transition occurs from the manifestation of invisi-
ble and inaccessible laws in the observational results of the system. Figure 9
represents the decay process for a deterministic equation.
In figure 9, the upper left image demonstrates the predictions made by a
deterministic equation. Notice that there is a collapse of the function at a well-
defined peak. This is because the equation behaves deterministically at a given
time, and there is certainty of an outcome. As new invisible and inaccessible
laws manifest in the system, the quality of the predictions are reduced, and
can be observed through the emergence of a more curvilinear and smoother
mountain in the plane. The complete decay of a deterministic equation occurs
after the predictions made are comparable to predictions based on a uniform
distribution over the sample space. You can see the result of the complete decay
of a deterministic equation in the lower right corner of the figure. Take the
system s1 formed only by circles and squares and an equation which aims to
describe the trajectories of the circles, but which considers only the attraction
between the circles. Suppose that this system has a large number of circles and
that at a hypothetical instant the first emergence of a square is observed from
the fusion of 8 circles. Taking that the repulsion between squares and circles
is weak over long distances to the point of being negligible, we can state that
the initially written equation can describe with high accuracy the trajectory of
circles that are not close to squares. However, as new squares are added to the
system, the equation is less assertive in describing the trajectories of circles.
If the predictions of the equation are as bad as uniformly random estimations
in the sample space, the equation is said to have undergone complete decay to
a random function. To avoid the process of decay and aging of the equation,
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 27

Fig. 9 Smooth decay of deterministic equations.

updates and revisions are necessary according to the periodicity of the invisible
and inaccessible laws acting on the system.

5.3 Objective Random and Physical Phenomena

In subsections 5.3.1 to 5.3.4, we describe 4 examples of physical phenom-
ena that fit the definitions of objective randomness under the definitions of
theorems 1.1 and 1.2 provided in section 4 of this paper.

5.3.1 Event Horizon

The speed of light in a vacuum is a constant c equal to 299,792,458 m/s
and invariant to the referential [6]. It is not possible to exceed the speed
of light, because it represents the speed of propagation of space-time itself.
Schwarzschild provided the first description of what a black hole would be [25]
according to the solutions of Einstein’s equations of General Theory of Rela-
tivity [26]. Finkelstein interpreted black holes as a region of space in which it
is not possible to escape, because if the mass of a given body is compressed
to a radius less than the Schwarzschild radius, in order for bodies entering the
Schwarzschild radius to escape the gravitational influence of the black hole,
they must have an escape velocity greater than the speed of light in vacuum
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

28 Theory of Dimensional Randomness

[27]. However, it is not even possible for light to exceed this speed. Thus,
the information contained in a black hole is invisible to observers outside it.
The study of events within the event horizon are speculative. Generalizing the
results of the field equation of the theory of General Relativity, it is described
that inside a black hole there is a gravitational singularity, a point in space-
time at which the mass and curvature of space-time are infinite. However, these
results cannot be experimentally proven because black holes cannot be accessed
in order to compare or validate the results directly or indirectly. Unless black
holes propagate the internal content to the outer environment, the definition
proposed by this article for objective randomness is still valid. New scientific
discoveries may demonstrate that information inside a black hole can be prop-
agated to the external world, for example, the existence of an Einstein-Rosen
bridge at the event horizon. But, black holes lose mass through Hawking Radia-
tion and during this process convert a pure state into a thermal state, losing the
information contained in the process [28]. Taking current physical definitions,
black holes are described as invisible and inaccessible regions of the spatial
dimensions. More Formally, we can state that ∃[α ∈ Rt1,n real ∧ α ∈/ Rt1,n
obs ] is true
given the definition of Black Holes. In this sense, within the event horizon,
invisible and inaccessible laws can act only under the specific circumstances
that black holes provide, so that the knowledge formulated on the observations
of the visible and accessible part is not generalizable, suffering, in this way, ran-
dom decay of the predictions made by deterministic equations. The results of
the generalization of general relativity for the description of black holes present
indications of random decay, as several physical quantities cease to make sense
and assume infinite values. According to the current description of physics and
using theorem 1.2, we demonstrate that the event horizon of black holes gives
rise to dimensional random event. It is believed that a quantum-gravity the-
ory explains the events at the event horizon of black holes, but the existence
of such a theory that unifies the four fundamental forces does not rule out the
hypothesis that, exclusively, at the event horizon a fifth, sixth or nth force may
act and that it is invisible and inaccessible to our referential. The definition
of Objective Randomness remains valid unless new scientific facts prove the
possibility of breaking the invisibility and inaccessibility of the event.

5.3.2 Butterfly Effect

The flapping of a butterfly’s wings can cause a tornado in Texas [29]. The
interpretation for this phenomenon is the sensitivity to the initial conditions
measured for a system, in which small inaccuracies can trigger chaotic effects
that are amplified through the propagation of system states. The analogy,
dubbed the butterfly effect, was used to explain why weather forecasts devi-
ated after a day or two to degrees of uncertainty beyond what was expected.
Theoretically, to make predictions with high assertiveness, it is necessary to
collect data with arbitrarily defined precision, in addition to complete knowl-
edge of the laws that govern the dynamical system. The level of precision for
the measurements of an experiment is limited according to the quality and
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 29

advancement of technological instruments, which makes the unpredictability

observed for chaotic systems conform to the definition of Subjective Random-
ness. The measurement capability of the instruments applies to finite values.
For measuring variables belonging to the set of irrational numbers I there are
limitations, because, by definition, irrational numbers cannot be represented
as fractions of two integers. It’s proved that the decimal expansion of irra-
tional numbers is infinite and non-periodic [30]. Because any measurement has
a finite decimal expansion, any variable in which its value belong to the set
of irrational numbers cannot be precisely measured. Formally, ∀α ∈ I, then
α ∈ Rt1,n / Rt1,n
real , but α ∈
t1,n t1,n
obs . Given 11, we deduce that Robs ̸= Rreal . By theorem
1.2, the prediction of events in which the initial conditions are described by
irrational numbers is dimensionally random.

5.3.3 Heisenberg’s Uncertainty Principle

The Heisenberg’s Uncertainty Principle [31] defines that it is impossible to
simultaneously determine the position and momentum of a particle with
arbitrary precision for both variables according to the expression:

∆p.∆x ≥ h/(4π) (17)

Where p represents the uncertainty of the position x the momentum uncer-
tainty and h the planck constant. The mathematical inequality expressed above
indicates that if the uncertainty of one of the variables in the pair (p, x)
decreases, the other variable in the pair must increase its value for the expres-
sion to hold true. The Heisenberg uncertainty principle represents a physical
limitation that prevents one from being able to measure with absolute cer-
tainty, simultaneously, the momentum and the position of a particle. Due to
this limitation, we point out that the reconstruction of particle information in
the 3 dimensions of space is partial, making these dimensions partially invisi-
ble and inaccessible. We rewrite the The Heisenberg’s uncertainty principle in
the following terms:
(
p ∈ Rt1,n
obs =⇒ x ∈ / Rt1,n
obs
f (x) = t1,n (18)
x ∈ Robs =⇒ p ∈ / Rt1,n
obs

Considering 18, we deduce ∃[α ∈ Rt1,n / Rt1,n

real ∧ α ∈ obs ]. Applying theorem 1.2,
the Heisenberg’s Uncertainty principle is deduced to be a case of dimensional
randomness. Such a limitation renders all potential physical laws that simulta-
neously use the position and momentum variables of a particle as invisible and
inaccessible, making it necessary to use probability distributions to describe
them.

5.3.4 String Theory

String theory says that the fundamental constituent of matter are one-
dimensional objects that would replace the point-particle model on which
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

30 Theory of Dimensional Randomness

the standard model of particles and the quantum field theory are currently
built [32]. But, a stable and consistent version of String Theory requires extra
spatial dimensions in space-time. M-Theory [33] describes the space-time as
eleven dimensional R1,10
real . The explanation for why we do not experience the
7 additional dimensions is that those dimensions are compactified. Formally,
the observable space-time is R1,3 obs , but, in String Theory, the real space-time
is R1,10
real . The universe description provided by String Theory is compliant
t1,n t2,m
with the expression [Robs ∧ Rreal | t1 + n < t2 + m]. Applying theorem 1.1, if
string theory is right, we expect that dimensional randomness would occur in
a R1,3
obs space-time.

6 Conclusions
Through this work, we distinguish between objective and subjective random-
ness based on the type of limitation encountered. Subjective randomness is a
consequence of limitations inherent to the incompetence of an observer. On
the other hand, objective randomness is a consequence of limitations imposed
by the universe for the prediction of certain events, no matter how competent
an observer may be. Dimensional randomness is subset of objective random-
ness, because it is not possible, under the observer’s referential, to distinguish
the class of events that have a fundamentally unknown cause from the class
of events that are fundamentally absent of cause. The Thought Experiments
were of paramount importance for these definitions, because they allowed us
to use the closed world assumption and discriminate subjective limitations of
objective limitations imposed by the systems s1 ,s2 ,s3 and s4. According to
the definitions of the theory, it was possible to describe objective random-
ness for events beyond quantum mechanics, finding applications in astronomy,
computing, and complex systems. The Theory of Dimensional Randomness is
also important for computing, as understanding the mechanisms for generating
objective randomness is an extremely relevant topic for information security,
such as in the field of quantum cryptography. Understanding objective ran-
domness is important beyond orthodox, empirical science. Applications of the
Theory of Dimensional Randomness can also be extended to branches of meta-
physics, such as in the quest to understand the existence of omniscience, free
will and fate. Next steps for the theory aim to generalize the concept and
application of invisible and inaccessible dimensions beyond the spatial and
temporal definition, finding irreducible limitations beyond physics. Also the
generalization of the Theory of Dimensional Randomness to a General Theory
of randomness that explains completely the set of true random events. The
relationship between String Theory, Bell’s theorem and Dimensional Random-
ness would also need to be better analyzed if the results of this theory are
extended to explain the probabilistic nature of quantum mechanics. Finally,
we believe that the theory already delimits the boundaries of empirical sci-
ence in that it is able to differentiate between the fine line of accessible and
inaccessible knowledge for the description of randomness.
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 31

7 Conflicts of interest
The author declares that there is not conflict of interest for the publication of
this article.

References
[1] Silveira, J. (2001). Tipos de Aleatoriedade.. Retrieved December 7, 2022,
from http://www.mat.ufrgs.br/ portosil/probab3.html

[2] Zeilinger, A. (2005). The message of the quantum. Nature 438, 743.
https://doi.org/10.1038/438743a

[3] Schmidhuber, J. (2006). Don’t forget randomness is still just a hypoth-

esis. Nature, 439(7075), 392-392.

[4] Bohr, N. (1928). The quantum postulate and the recent development of
atomic theory. (Vol. 3). Printed in Great Britain by R. ands R. Clarke,
Limited.

[5] Einstein, A., Podolsky, B. and Rosen, N. (1935). Can quantum-

mechanical description of physical reality be considered complete? Physical
review, 47(10), 777.

[6] Einstein, A. (1905). On the electrodynamics of moving bodies. Annalen

der physik, 17(10), 891-921.

[7] Bell, J. S. (1964). On the Einstein Podolsky Rosen paradox.Physics

Physique Fizika, 1(3), 195.

[8] Clauser, J. F., Horne, M. A., Shimony, A. and Holt, R. A. (1969).

Proposed experiment to test local hidden-variable theories. Physical review
letters, 23(15), 880.

[9] CHAITIN, Gregory J. (1975). Randomness and mathematical proof.

Scientific American, v. 232, n. 5, p. 47-53

[10] Shannon, C. E. (1948). A mathematical theory of communication. The

Bell system technical journal, 27(3), 379-423.

[11] Watanabe, O. (1992). Kolmogorov complexity and computational com-

plexity. Berlin: Springer.

[12] Gödel, K. (1931). Über formal unentscheidbare Sätze der Principia

Mathematica und verwandter Systeme I. Monatshefte für mathematik und
physik, 38(1), 173-198.
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

32 Theory of Dimensional Randomness

[13] Vitányi, P. M. (2020). How incomputable is Kolmogorov complexity?

Entropy, 22(4), 408.

[14] Zvonkin, A.K. and Levin, L.A. (1970). The complexity of finite objects
and the development of the concepts of information and randomness by
means of the theory of algorithms. Russ. Math. Surv. 1970, 25, 83-124.

[15] Levin, L. A. (1984). Randomness conservation inequalities; informa-

tion and independence in mathematical theories. Information and Control,
61(1), 15-37.

[16] Shen, A. (2020). Randomness tests: theory and practice. In Fields of

Logic and Computation III (pp. 258-290). Springer, Cham.

[17] Massey Jr, F. J. (1951). The Kolmogorov-Smirnov test for goodness of

fit. Journal of the American statistical Association, 46(253), 68-78.

[18] McHugh, M. L. (2013). The chi-square test of independence. Biochemia

medica, 23(2), 143-149.

[19] Napierala, M. A. (2012). What is the Bonferroni correction?. Aaos

Now, 40-41.

[20] Random - Generate pseudo-random numbers (2022). Retrieved August

13, 2022, from https://docs.python.org/3/library/random.html

[21] Wald, A. and Wolfowitz, J. (1943). An exact test for randomness

in the non-parametric case based on serial correlation. The Annals of
Mathematical Statistics, 14(4), 378-388.

[22] Uhlenbeck, G. E. and Ornstein, L. S. (1931). On the theory of the

Brownian motion. Physical review, 36(5), 823.

[23] Bossu, G. and Siegel, P. (1985). Saturation, nonmonotonic reasoning

and the closed-world assumption. Artificial Intelligence, 25(1), 13-63.

[24] Drummond, N. and Shearer, R. (2006). The open world assumption.

In eSI Workshop: The Closed World of Databases meets the Open World
of the Semantic Web (Vol. 15, p. 1).

[25] Schwarzschild, K. (1916). Über das Gravitationsfeld einer Kugel aus

inkompressibler Flüssigkeit nach der Einsteinschen Theorie.. Sitzungs-
berichte der königlich preußischen Akademie der Wissenschaften zu
Berlin, 424-434.

[26] Einstein, A. (1922). The general theory of relativity. In The Meaning of

Relativity (pp. 54-75). Springer, Dordrecht.
 Springer Nature 2021 LAT X template
E
ciELO Preprints - This document is a preprint and its current status is available at: https://doi.org/10.1590/SciELOPreprints.52

Theory of Dimensional Randomness 33

[27] Finkelstein, D. (1958). Past-future asymmetry of the gravitational field

of a point particle. Physical Review, 110(4), 965.

[28] Hawking, S. (1975). Particle creation by black holes. Commun. Math.

Phys. 43, 199-220.

[29] Lorenz, E. (2000). The butterfly effect. World Scientific Series on

Nonlinear Science Series A, 39, 91-94.

[30] Weisstein, Eric W. Irrational Number. From MathWorld–A Wolfram

Web Resource. https://mathworld.wolfram.com/IrrationalNumber.html

[31] Heisenberg, W. (1927). Über den anschaulichen Inhalt der quantenthe-

oretischen Kinematik und Mechanik. Z Phys 43: 172-198.

[32] Blumenhagen, R., Lüst, D. and Theisen, S. (2013). Basic concepts

of string theory. (Vol. 17, pp. 21-23). Berlin: Springer.

[33] Witten, E. (1995). String theory dynamics in various dimensions.

Nuclear Physics B, 443(1-2), 85-126.
 This preprint was submitted under the following conditions:

The authors declare that they are aware that they are solely responsible for the content of the preprint and
that the deposit in SciELO Preprints does not mean any commitment on the part of SciELO, except its
preservation and dissemination.

The authors declare that the necessary Terms of Free and Informed Consent of participants or patients in
the research were obtained and are described in the manuscript, when applicable.

The authors declare that the preparation of the manuscript followed the ethical norms of scientific
communication.

The authors declare that the data, applications, and other content underlying the manuscript are
referenced.

The deposited manuscript is in PDF format.

The authors declare that the research that originated the manuscript followed good ethical practices and
that the necessary approvals from research ethics committees, when applicable, are described in the
manuscript.

The authors declare that once a manuscript is posted on the SciELO Preprints server, it can only be taken
down on request to the SciELO Preprints server Editorial Secretariat, who will post a retraction notice in its
place.

The authors agree that the approved manuscript will be made available under a Creative Commons CC-BY
license.

The submitting author declares that the contributions of all authors and conflict of interest statement are
included explicitly and in specific sections of the manuscript.

The authors declare that the manuscript was not deposited and/or previously made available on another
preprint server or published by a journal.

If the manuscript is being reviewed or being prepared for publishing but not yet published by a journal, the
authors declare that they have received authorization from the journal to make this deposit.

The submitting author declares that all authors of the manuscript agree with the submission to SciELO
Preprints.

Powered by TCPDF (www.tcpdf.org)