Harmony consists in a mixture of opposites.
— Heraclitus
In the previous chapter, we performed a single, simple gesture: folding the number line at the point $x/2$. Nothing more.
This apparently elementary act revealed something decisive. When the observed interval is divided into two halves, prime numbers begin to play distinct roles, not because of their intrinsic nature, but because of their relative position with respect to the point of observation.
In the first half, primes still actively participate in the formation of composite numbers. In the second half, they no longer do.
This distinction was not imposed by any theory. It emerged directly from counting and folding.
The next natural step is not to interpret this difference, but to measure it.
If two complementary roles act simultaneously, it is legitimate to ask which one predominates at each stage of observation.
Let us call structuring primes those that lie in the first half of the observed interval. Because they are in $[1, x/2]$, their multiples still fall within $[1, x]$, and it is through them that composite numbers are formed.
Let us call stabilising primes those that lie in the second half, in $(x/2, x]$. These primes no longer generate new composites within the observed interval. They occupy positions not filled by the action of the structuring primes.
These two roles are not fixed identities. They are transient functions.
As the interval grows, a prime that stabilises today will, at a later stage, become a structuring prime. The sequence of primes is not static. It carries an internal dynamic of functional redistribution.
It is this dynamic — not the isolated existence of primes — that begins to reveal the system’s complexity.
To observe this dynamic, we require a simple, direct, and purely arithmetical instrument.
We define the function $\pi(x)$ as the exact count of primes less than or equal to $x$. Using it, we can quantify:
The difference between these two quantities defines a natural contrast functional:
\[\Delta_\pi(x) = \pi(x) − 2\, \pi(\lfloor \frac{x}{2} \rfloor).\]This functional does not predict, adjust, or smooth. It simply measures, at each value of $x$, which of the two roles predominates.
When $\Delta_\pi(x)$ is positive, stabilisation outweighs structure. When it is negative, structuring capacity dominates. When it vanishes, there is a momentary balance.
Nothing beyond this is claimed here.
When $\Delta_\pi(x)$ is computed along the number line, something unexpected occurs.
Its value neither grows smoothly nor decays monotonically. It oscillates.
These oscillations are not random. They follow an irregular yet persistent pattern that repeats as the scale increases. After the initial values, a clear predominance of the negative phase becomes visible — a signal that, on average, structure outweighs stabilisation.
This negative predominance is the arithmetical record of a familiar fact: primes become sparser as we advance. The first half of the interval always accumulates more structuring potential than the second half can counterbalance.
This behaviour can be described as an observational pulse: a measurable alternation between two complementary forces, derived exclusively from the exact counting of primes.
There is no statistics, no spectrum, and no chaos here in the technical sense. There is only a discrete signal, obtained without approximation.
At this point, abstraction must give way to visual evidence.
Notebook 02 (02_pi_and_delta_pi.ipynb) was designed for this purpose.
When you execute it, the graph of $\Delta_\pi(x)$ appears. Observe the nature of the signal: it is a discrete, step-like signal — dry and uncompromising. Observe how the blue forces (structure) and the orange forces (stabilisation) compete for dominance along the number line.
There are no smooth curves here. What you see is the raw record of each new prime entering the system and altering the balance between structuring and stabilising roles.
Increase the interval. Notice how the oscillations acquire a kind of “texture”. What initially appears as a counting artefact reveals itself as a persistent alternation.
This is the signal we call the pulse.
It is impossible not to notice a curious aspect of this construction.
The value 1 sustains the entire sequence, yet it does not participate directly in any of the relations measured by $\Delta_\pi(x)$. It neither structures nor stabilises.
The One acts as the phase reference of the system. It is the neutral element that makes the fold possible, but it is only from 2 onwards that functional asymmetry (structuring versus stabilising) becomes observable.
The first effective relation emerges with 2. With it, duplication begins. From it onwards, the fold produces observable effects.
The gaps created by the fold are measured in units of 1, but only numbers greater than 1 occupy structural positions within the system.
This fact is not interpreted here. It is merely recorded.
Methodological discipline is essential.
Up to this point:
We have constructed a simple functional and observed its behaviour.
What appears is not noise. It is a deterministic signal, born from the most elementary rule of arithmetic: the repetition of the One.
This chapter concludes the phase of direct measurement.
In the following chapters, this signal will be:
Nothing new will be added to the system. Only the manner of observation will change.
Everything that follows — including any spectral reading — depends on this initial, silent, unavoidable pulse, which arises when we fold the line and count with care.
The arithmetical function $\Delta_\pi(x) = \pi(x) - 2\,\pi(x/2)$ does not describe a distribution, but a functional imbalance. It renders visible the alternation between two complementary roles exercised by the primes along the observed interval.
Primes do not act homogeneously. In certain regimes, they structure the mesh of composite numbers; in others, they stabilise the density of the line. From this alternation a pulse emerges — not as noise, but as the expression of an internal dynamic.
The system is therefore not static. It redistributes functions without loss of coherence, and it is this redistribution — not isolated values — that sustains the observed structure.
| $\gets$ Previous Chapter | Contents | Next Chapter $\to$ |