The Entropies

✨

This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Idea: One Size Does Not Fit All

Imagine you are trying to measure the "messiness" or "disorder" of a room. In science, this messiness is called Entropy.

For a long time, scientists thought there was only one ruler to measure this messiness. This ruler is called Shannon Entropy (or Gibbs Entropy in physics). It works perfectly for some situations, like a room where the temperature is controlled by a thermostat (you can add or remove heat freely).

However, the author of this paper, Roumen Tsekov, argues that this ruler is broken when you try to use it on a sealed, isolated room where no heat can enter or leave. In that specific case, the ruler gives you a nonsensical answer (like saying the room has "negative infinite" messiness).

The paper suggests we need two different rulers depending on the situation:

The Thermostat Ruler (Shannon/Gibbs): For systems that can swap heat with the outside world.
The Sealed Box Ruler (Boltzmann/Hartley): For systems that are completely isolated and have a fixed amount of energy.

The Two Scenarios: The Hotel vs. The Vault

To understand why we need two rulers, let's look at two different scenarios.

Scenario A: The Hotel Room (The Canonical Ensemble)

Imagine a hotel room where the air conditioning is on. The temperature is constant because the AC is constantly adjusting. If you leave a window open, heat flows in or out to keep things steady.

The Physics: This is a system exchanging heat with its surroundings.
The Ruler: Here, the standard Shannon Entropy works great. It's like a flexible measuring tape that stretches and shrinks to fit the situation.
The Result: It correctly predicts that things will naturally move toward a state of maximum comfort (equilibrium) and that the "messiness" will increase over time, just like the Second Law of Thermodynamics says.

Scenario B: The Sealed Vault (The Microcanonical Ensemble)

Now, imagine a super-strong, perfectly insulated vault. Once you lock the door, no energy can get in or out. The total energy inside is fixed.

The Physics: This is an isolated system.
The Problem: If you try to use the "Hotel Ruler" (Shannon Entropy) here, it breaks. Mathematically, it tries to calculate the messiness of a single, specific energy level and ends up dividing by zero, resulting in negative infinity. This is physically impossible.
The Solution: Tsekov argues that for this sealed vault, we must use Boltzmann's Entropy. Instead of looking at the probability of a specific state, we look at the volume of the entire space the system can occupy.
The Analogy: Imagine the vault is a giant ballroom.
- The Shannon Ruler tries to count how likely it is for a dancer to be in one specific spot. In a sealed vault, the dancer must be somewhere, but the math gets confused about which spot.
- The Boltzmann Ruler simply measures the size of the ballroom. It says, "The more space the dancers have to run around in, the higher the entropy." This works perfectly and gives a real, positive number.

The "Time Travel" Problem

The most critical part of the paper is about Time.

The Second Law of Thermodynamics says that in an isolated system, entropy (disorder) must always increase. This is what gives us the "Arrow of Time"—it's why an egg breaks but never un-breaks.

The Failure: When Tsekov applies the standard Shannon formula to an isolated system, the math shows that entropy never changes. It stays flat. It's like a movie where the characters are frozen in time. This contradicts reality; we know things do change and get messier.
The Fix: When he switches to the Boltzmann formula (measuring the volume of accessible states), the math shows that entropy does increase over time. This aligns with our real-world experience that time moves forward and systems evolve toward equilibrium.

The Metaphor:
Think of the Shannon formula as a camera that takes a perfect, frozen snapshot of a single moment. In an isolated system, that snapshot never changes, so time seems to stop.
The Boltzmann formula is like a wide-angle lens that sees the whole landscape of possibilities. As the system evolves, it explores more of that landscape, and the "width" of the landscape (entropy) grows, showing us that time is passing.

Why Does This Matter?

You might ask, "Why should I care about a math error in a physics paper?"

It Fixes the Foundation: If our basic definition of entropy is wrong for isolated systems, then our understanding of how the universe works (from stars to black holes) has a crack in it.
Black Holes: The paper mentions that this new way of thinking helps explain weird things about black holes, like how they might have "negative temperatures" (which sounds impossible but is actually a real concept in physics).
Beyond Physics: The concept of entropy is now used in AI, economics, and even sociology (to measure "social freedom"). If the math is wrong for isolated systems, we might be misinterpreting how information or economies evolve when they are cut off from the outside world.

The Takeaway

Roumen Tsekov is telling us: "Stop using the same ruler for every job."

If you are dealing with a system that can trade heat with the world, use the Shannon Entropy (the flexible tape measure). But if you are dealing with a system that is completely sealed off, you must use Boltzmann Entropy (the volume measure). Only by using the right tool can we correctly understand why time moves forward and how the universe settles down.

Based on the provided text, here is a detailed technical summary of the paper "The Entropies" by Roumen Tsekov.

1. Problem Statement

The paper identifies a fundamental discrepancy in contemporary statistical mechanics regarding the definition and application of entropy, specifically the conflict between the Gibbs-Shannon entropy ( $S_G$ ) and the Boltzmann entropy ( $S_B$ ).

The Core Issue: While the Gibbs-Shannon entropy ( $S_G = -k_B \int \rho \ln \rho \, d\Gamma$ ) successfully describes systems in the canonical ensemble (systems exchanging heat with a reservoir at constant temperature $T$ ), it fails to adequately represent the microcanonical ensemble (isolated systems with constant energy $E$ ).
The Consequence: When applied to an isolated system in equilibrium, the Gibbs-Shannon formula yields an infinitely negative entropy due to the Dirac delta function nature of the microcanonical probability density. More critically, the paper argues that $S_G$ is mathematically incapable of supporting a theoretical derivation of the Second Law of Thermodynamics for isolated systems, as it remains constant over time under Hamiltonian dynamics, contradicting the observed arrow of time.

2. Methodology

The author employs a comparative theoretical analysis of statistical mechanics frameworks, utilizing:

Probability Theory and Information Theory: Examining various entropy definitions (Shannon, Rényi, Tsallis, Hartley) and their mathematical properties.
Classical and Quantum Statistical Mechanics: Analyzing the behavior of probability densities ( $\rho$ ) in phase space for both ideal gases and real systems defined by Hamiltonian functions $H(p, q)$ .
Thermodynamic Derivations: Calculating conjugate thermodynamic parameters (pressure, chemical potential, temperature) via derivatives of different entropy functions to test their consistency with established thermodynamic relations.
Dynamical Systems Analysis: Utilizing the Liouville equation to determine the time evolution of entropy in isolated systems.

3. Key Contributions and Arguments

A. Critique of Gibbs-Shannon Entropy in Isolated Systems

The paper demonstrates that for an isolated system where energy $E$ is constant, the equilibrium probability density is $\rho_{eq} = \delta(E - H) / \Omega$ .

Substituting this into the Gibbs formula ( $S_G = -k_B \int \rho \ln \rho \, d\Gamma$ ) results in a term involving $\ln[\delta(0)]$ , leading to infinitely negative entropy, which is physically unobservable.
Furthermore, the time derivative of $S_G$ in an isolated system is calculated using the Liouville equation ( $\dot{\rho} = \{H, \rho\}$ ). The author shows that $\dot{S}_G = 0$ , meaning Gibbs entropy is conserved and does not increase over time. This directly contradicts the Second Law of Thermodynamics, which requires entropy to grow until equilibrium is reached.

B. Reaffirmation of Boltzmann Entropy for Isolated Systems

The author argues that the correct entropy for isolated systems is the Boltzmann entropy (often associated with the volume of phase space), defined as:
$S_B = k_B \ln \Phi$
Where $\Phi(E, V, N) = \int \theta(E - H) \, d\Gamma$ is the total number of accessible states (volume of phase space) bounded by energy $E$ , and $\theta$ is the Heaviside step function.

Thermodynamic Consistency: By differentiating $S_B$ , the author derives correct expressions for pressure ( $p$ ) and chemical potential ( $\mu$ ) that align with standard thermodynamic definitions.
Temperature Definition: The temperature is derived as $k_B T = \Phi / \Omega$ (where $\Omega = \partial \Phi / \partial E$ ), satisfying the exact thermodynamic relation $T = (\partial E / \partial S)_{V,N}$ .
Connection to Ideal Gas: For an ideal gas, this formulation reduces to the well-known Sackur-Tetrode equation, validating its physical correctness.

C. Distinction Between Ensembles

The paper clarifies that:

Canonical Ensemble (Constant $T$ ): The Gibbs-Shannon entropy is valid because the exponential Boltzmann distribution ( $\rho \propto e^{-\beta H}$ ) is the inverse function of the logarithm in the entropy definition, allowing for a consistent derivation of the Second Law (via non-equilibrium free energy minimization).
Microcanonical Ensemble (Constant $E$ ): The Gibbs-Shannon entropy fails. The system requires the Boltzmann entropy ( $S_B$ ), which is proportional to the Hartley function (logarithm of the number of microstates). The author notes that $S_B$ grows over time, consistent with the Second Law, whereas $S_G$ does not.

4. Results

Mathematical Proof of Failure: The paper provides a rigorous proof that $\dot{S}_G = 0$ for any functional of the form $S = \int g(\rho) d\Gamma$ in an isolated Hamiltonian system, proving that Shannon-type entropies cannot describe the thermodynamic arrow of time in isolation.
Resolution of the "Meaningless Constant": The paper refutes the textbook notion that $S_G$ and $S_B$ are equivalent up to a constant. In the microcanonical limit, the ratio $\Phi/\Omega$ implies a formal zero temperature if $S_G$ is used, highlighting a fundamental incompatibility rather than a simple offset.
Generalization: The author extends the discussion to modern physics, noting that the Bekenstein-Hawking entropy (proportional to black hole horizon area) and concepts of negative temperature in bounded spectra further challenge the universality of the Gibbs-Shannon framework.

5. Significance

Theoretical Correction: The paper challenges the prevailing view in many statistical mechanics textbooks that equates Gibbs entropy with Boltzmann entropy in all contexts. It argues for a strict separation: Gibbs entropy for open/isothermal systems and Boltzmann entropy for isolated systems.
Foundational Impact: By resolving the inability of Shannon entropy to derive the Second Law for isolated systems, the paper suggests that the "molecular chaos hypothesis" (Stosszahlansatz) and the factorization of the probability distribution are essential for entropy growth, a feature not captured by the linear Liouville evolution of the full density $\rho$ .
Interdisciplinary Relevance: The author highlights the importance of these distinctions for emerging fields such as quantum computing, artificial intelligence, and sociophysics, where the definition of entropy (information change) dictates the behavior of complex, non-equilibrium systems. The paper suggests that applying the wrong entropy definition (e.g., using Shannon entropy for isolated quantum states) could lead to incorrect physical predictions.

In conclusion, Roumen Tsekov's paper serves as a critical examination of the foundations of statistical mechanics, advocating for the Boltzmann entropy ( $S_B = k_B \ln \Phi$ ) as the physically correct measure for isolated systems, while acknowledging the utility of Shannon/Gibbs entropy only within the specific constraints of the canonical ensemble.