🔬 physics

The Impossibility of Cohesion Without Fragmentation

This paper presents a static structural theory demonstrating that social cohesion and fragmentation are complementary outputs of positional constraints, where relational collapse is structurally inevitable under plurality because maintaining cohesion requires unanimous compatibility while fragmentation requires only a single veto.

Original authors: Daisuke Hirota

Published 2026-03-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Daisuke Hirota

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ 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: You Can't Have Your Cake and Eat It Too (Structurally)

Imagine you are trying to organize a massive party for 200 people. You want everyone to get along (cohesion), but you also want to respect their individual differences (division).

This paper argues that you cannot have a perfectly united group without breaking some relationships. In fact, the very act of making a group "stick together" is what causes other parts of the group to fall apart.

The author, Daisuke Hirota, uses math to prove that division and cohesion are two sides of the same coin. You can't have one without the other if people have specific "requirements" for who they can be friends with.

The Core Concepts (The "How")

1. The "Position" of a Person

Think of every person in society as having a specific position. This isn't just where they live; it's their opinion, their job, their political party, or their religion.

The Analogy: Imagine everyone is wearing a colored badge. Some wear Red, some wear Blue, some wear Green.
The Rule: In this paper, a relationship (friendship, alliance, trade) can only exist if the two people's badges are "compatible."

2. The "Bifurcation Event" (The Moment of Decision)

Sometimes, a big event happens that forces everyone to pick a side or lock into a specific role.

The Analogy: Imagine a sudden storm forces everyone to run to a specific shelter. Or, imagine a company forces everyone to pick a specific team for a project.
The Paper calls this: A Bifurcation Event. It's the moment where everyone's "position" becomes fixed.

3. The "Compatibility Filter"

Once everyone is in their fixed position, a filter automatically checks: "Can Person A (Red Badge) talk to Person B (Blue Badge)?"

If the rules say Red and Blue don't mix, the relationship collapses. (This is Fragmentation).
If the rules say Red and Red get along, the relationship strengthens. (This is Cohesion).

The Two Main Discoveries

Discovery #1: The "Split and Stick" Effect (Theorem 10.2)

The paper proves that whenever you force people into specific positions, two things happen at the exact same time:

Some people get pushed apart. (The "Reds" and "Blues" stop talking).
Some people get pulled together. (The "Reds" and "Reds" start hugging).

The Analogy: Think of a magnet. When you bring a magnet close to a pile of iron filings:

The filings that are attracted clump together tightly (Cohesion).
The filings that are repelled fly away and scatter (Fragmentation).
You cannot have the clumping without the scattering. They are the same physical process.

Discovery #2: The "Total Unity" Trap (Theorem 11.3)

The paper asks: "Is there a way to make everyone get along without anyone getting pushed away?"

The answer is No, unless you do something extreme.

The Only Way to Avoid Splitting: You must force everyone to have the exact same position.
The Analogy: Imagine a school where the principal says, "Everyone must wear the exact same uniform, eat the exact same lunch, and think the exact same thoughts."
- In this case, no one is incompatible. Everyone gets along.
- But: This is called "Coercive Homogenization." It means you have destroyed all diversity to get unity.
- The Conclusion: If you want to keep people's differences (different opinions, different roles), you must accept that some relationships will break. You cannot have a diverse world where everyone is friends with everyone else.

Why Does This Matter? (Real World Examples)

The paper explains why our world feels so polarized right now, and why it's not necessarily a "failure" of society, but a structural necessity.

Political Polarization:
- When a society forces people to choose between "Party A" and "Party B" (a bifurcation event), people with similar views bond tightly (Cohesion).
- But, people with different views stop talking to each other entirely (Fragmentation).
- The paper says: This isn't a bug; it's a feature of how social rules work. You can't have strong political identities without losing some cross-party friendships.
Workplace Teams:
- If a company forces employees into strict departments (Engineering vs. Marketing), the engineers bond deeply with other engineers.
- But, the engineers might stop understanding or talking to the marketers. The "silos" form naturally because of the structural constraints.
Social Media Algorithms:
- Algorithms act as the "Bifurcation Event." They sort you into groups based on what you like.
- This makes your feed very cohesive (you see things you agree with).
- But it fragments your view of the world (you stop seeing things you disagree with).

The Takeaway in One Sentence

You cannot build a strong, unified group of people with different identities without simultaneously breaking the connections between those who are different; the stronger the unity within a group, the sharper the division from the outside.

The paper provides a mathematical proof that division and cohesion are not enemies fighting each other; they are partners dancing together. One cannot exist without the other.

1. Problem Statement

The paper addresses a fundamental paradox in social science: why do division (fragmentation) and cohesion often intensify simultaneously? Existing paradigms, such as utility-based models (e.g., network formation games where relations form based on positive payoffs) and probabilistic models (e.g., random graphs where links form with probability $p$ ), treat relation formation as a matter of degree, preference, or stochastic likelihood. These approaches implicitly assume that relations are always feasible and focus on their strength or probability.

Hirota argues that these models fail to address the structural boundary conditions under which a relation can exist at all. The core problem is to determine whether a relation is logically possible given the positions agents occupy, rather than whether they want to form it. The paper seeks to explain how the very act of fixing agents' positions (e.g., through voting, institutional assignment, or social sorting) structurally necessitates the simultaneous creation of cohesive subgroups and the collapse of other relations.

2. Methodology and Framework

The paper develops a static, axiomatic theory of relation maintenance based on minimal positional constraints. It abstracts away from psychological motivations, temporal dynamics, and quantitative attributes (like tie strength), focusing exclusively on logical compatibility.

Key Definitions and Axioms:

Position Space ( $L$ ): An abstract space where agents occupy specific positions ( $\ell_i$ ). Positions represent physical location, opinion, role, or affiliation.
Gain Axes ( $G$ ): Types of value or conditions realized through a relation (e.g., recognition, authority).
Minimum Condition Function ( $a$ ): A binary function $a(i, g)$ indicating if agent $i$ requires gain axis $g$ to maintain a relation. If $a(i, g)=1$ and the condition is not met, the agent structurally withdraws.
Compatibility Function ( $f_g$ ): A function $f_g: L \times L \to \{0, 1\}$ $f_{g} : L \times L \to {0, 1}$ determining if a gain axis $g$ $g$ is satisfied between two positions.
- Position-dependent: $f_g$ is not identically 1; compatibility depends on the specific positions of the agents.
- Existence-dependent: $f_g \equiv 1$ ; the condition is satisfied regardless of position.
Bifurcation Event ( $E$ ): An exogenous mapping $E: P \to L$ that fixes every agent to a unique position. This represents a "snapshot" of the system after a sorting event (e.g., an election or policy change).

Core Logic:

The theory posits that a relation $R_{ij}$ is maintained after a bifurcation event $E$ if and only if, for all position-dependent gain axes $g$ required by either agent, the compatibility function evaluates to 1:
$(a(i, g)=1 \lor a(j, g)=1) \implies f_g(E(i), E(j)) = 1$
If this condition fails for any required axis, the relation collapses.

3. Key Contributions

Reframing Relation Formation: Shifts the question from "Why do agents form relations?" (utility/behavior) to "Can agents form relations?" (structural feasibility).
Structural Asymmetry of Veto Power: Establishes that fragmentation is a unilateral veto (one agent's condition failure breaks the link), while cohesion requires bilateral alignment (both agents' conditions must be met).
The Structural Trade-off: Identifies that fragmentation and cohesion are not opposing forces but complementary outputs of the same structural mechanism.

4. Key Results and Theorems

Theorem 10.2: Structural Necessity of Fragmentation and Cohesion

This is the paper's central result. It states that under any non-degenerate positional constraint (where the compatibility function is not constant on the set of relevant edges), a bifurcation event necessarily produces both:

Fragmentation: Relations between agents with incompatible positions ( $f_g=0$ ) must collapse.
Cohesion: Relations between agents with compatible positions ( $f_g=1$ ) are structurally confirmed.
Conclusion: Division and cohesion are the "0-output" and "1-output" of the same compatibility function. You cannot have the confirmation of shared conditions (cohesion) without the simultaneous rejection of incompatible conditions (fragmentation).

Theorem 11.3: Conditional Impossibility of Universal Cohesion

This theorem establishes a conditional impossibility result regarding "g-forced cohesion" (cohesion without collapse).

Result: A gain axis $g$ forces no collapse among $g$ -shared relations if and only if the compatibility map is identically 1 on the set of relevant edges.
Implication: The only bifurcation events that guarantee no relational collapse for any possible initial set of relations are coercively homogenizing events (where all agents are forced into positions that are mutually compatible).
Corollary 11.4 (Structural Trade-off): In any non-homogenizing system, the existence of cohesive subgroups (where $f_g=1$ ) mathematically necessitates the existence of collapsed relations (where $f_g=0$ ) within the same network layer.

Corollary 12.1 & 12.2: The Inescapability of Out-Groups

The only way to achieve cohesion without fragmentation is if the gain axis is existence-dependent (universally compatible).
If a gain axis is position-dependent, the formation of any "in-group" (confirmed shared condition) structurally guarantees the existence of an "out-group" (collapsed relation).

5. Significance and Implications

Theoretical Impact: The paper provides a rigorous mathematical foundation for understanding polarization. It demonstrates that polarization is not merely a failure of communication or a result of extreme preferences, but a structural inevitability arising from the constraints of positional compatibility.
Policy and Institutional Design: It suggests that institutions aiming for universal cohesion without fragmentation must resort to "coercive homogenization" (forcing everyone into the same position), which is often undesirable or impossible in diverse societies. Conversely, accepting diversity (positional plurality) structurally guarantees some level of fragmentation.
Analytical Tool: The framework offers a common language for analyzing polarized networks, institutional sorting, and identity-based alignment, moving beyond probabilistic descriptions to deterministic structural constraints.

6. Conclusion

Hirota concludes that division and cohesion are intrinsically relative phenomena. They are not independent variables but are co-determined by the structural constraints of the position space. The "impossibility" in the title refers to the fact that universal cohesion without fragmentation is structurally impossible unless the system is forced into a state of total homogeneity. Any attempt to maintain distinct positions (plurality) inevitably generates both cohesive clusters and fractured links.