CMH MATH PATH — LESSON 7: NETWORK THEORY

The Hidden Web of Society — Why Connections Matter More Than Individuals

PART 1: THE CORE IDEA — THE WHOLE IS GREATER THAN THE SUM OF ITS PARTS

The Friendship Paradox and Why It Matters

Here’s a strange mathematical truth about your social life:

Your friends have more friends than you do.

This isn’t about popularity — it’s a mathematical certainty in any social network.

Try it: List your 5 closest friends. Count their friends (social media connections work). Average them. The number will almost certainly be higher than your own friend count.

Why? Because popular people (with many friends) appear in many people’s friend lists, skewing the average. You’re more likely to be friends with popular people than unpopular ones.

CMH Insight: This explains why people consistently think “everyone is going to that protest but me” or “everyone knows about this news but me.” Our local view of the network is biased!

PART 2: NETWORK ANATOMY — NODES AND EDGES

The Building Blocks of All Networks

Every network has just two components:

1. Nodes (Vertices): The “things”

  • People, countries, organizations, computers, genes
  • In CMH: Individuals, cities, institutions, ideologies

2. Edges (Links): The “connections”

  • Friendships, trade routes, alliances, communication lines
  • In CMH: Social ties, economic exchanges, diplomatic relations, information flows

Visualizing a Tiny Social Network:

    Alice
     / \
    /   \
  Bob — Carol
    \   /
     \ /
    David

This simple picture contains profound information:

  • Alice knows Bob and Carol
  • Bob knows Alice, Carol, and David (most connected!)
  • David only knows Bob (least connected)
  • Carol is the “bridge” between clusters

PART 3: NETWORK SHAPES — THREE PATTERNS THAT CHANGE EVERYTHING

Type 1: Random Network (The Party Model)

  • Everyone connects randomly
  • Like a cocktail party where you talk to random people
  • Real example: Early internet (websites linking randomly)
  • Key property: Most people have similar number of connections

Type 2: Scale-Free Network (The “Rich Get Richer” Model)

  • Most have few connections, few have MANY (hubs!)
  • Newcomers prefer connecting to already-popular nodes
  • Real examples: Social media (few influencers), air travel (major hubs), citations (few papers cited often)
  • Key property: Power-law degree distribution (heavy tail!)

Type 3: Small-World Network (The “Six Degrees” Model)

  • Highly clustered locally (friends know each other)
  • But short paths connect everyone globally
  • Real examples: Social networks, neural networks, power grids
  • Key property: High clustering + short average path length

Visual Comparison:

Random:       •••••••      (no pattern)
Scale-Free:   ••••••       (one big hub in center)
               \ | /
                •••
Small-World:  •••  •••     (clusters with bridges)
                | |
               •••••

PART 4: KEY NETWORK METRICS — MEASURING SOCIETY’S STRUCTURE

1. Degree Centrality: How Popular?

  • Count of direct connections
  • CMH meaning: Social influence, exposure to information
  • Formula: Degree(i) = Number of edges from node i

Example: In protest movements, high-degree individuals recruit more people.

2. Betweenness Centrality: The Bridge Maker

  • How many shortest paths go through this node?
  • CMH meaning: Gatekeepers, information brokers, bottleneck points
  • Formula: Betweenness(i) = Σ (paths through i)/(all paths) for all pairs

Example: In smuggling networks, the person connecting two groups has high betweenness.

3. Clustering Coefficient: How “Cliquey”?

  • Do your friends know each other?
  • CMH meaning: Community cohesion, echo chambers
  • Formula: CC(i) = (Actual links among i’s friends)/(Possible links)

Example: High clustering = insular communities where rumors spread fast internally but not externally.

4. Average Path Length: “Six Degrees”

  • Average shortest distance between any two nodes
  • CMH meaning: How quickly information/disease/revolt can spread

Famous finding: Average path length on Facebook = 4.7 (recently 3.5)! “Three and a half degrees of separation.”

PART 5: NETWORK DIFFUSION — HOW THINGS SPREAD

The Epidemic Models (Same Math for Ideas, Diseases, and Revolts!)

SIR Model (Epidemics):

  • S = Susceptible (haven’t heard idea)
  • I = Infectious (spreading idea)
  • R = Recovered (heard it, stopped spreading)

SI Model (Fads/Panics):

  • S = Susceptible (not panicked)
  • I = Infected (panicked and spreading panic)

Threshold Model (Revolutions):

  • Person joins protest when X% of friends have joined
  • Creates cascades — once critical mass joins, everyone follows!

Visualizing a Protest Cascade:

Day 1:   A protests alone
         ↓ tells B, C
Day 2:   A, B, C protest
         ↓ each tell 2 friends
Day 3:   A,B,C,D,E,F,G protest (critical mass!)
         ↓ everyone's friends see >50% protesting
Day 4:   WHOLE NETWORK protests!

Key insight: Network structure determines whether protest grows or fizzles!

PART 6: HANDS-ON NETWORK LAB

Exercise 1: Map Your Social Network

Materials: Paper, pen, different colored pens

Step 1: Write your name in center
Step 2: Add 10-15 closest people (family, friends, coworkers)
Step 3: Draw connections between THEM (who knows whom?)
Step 4: Calculate:

  • Degree: Who has most connections?
  • Clustering: Are your friends friends with each other?
  • Betweenness: Who connects different groups?

Step 5: Color code:

  • Red = Family
  • Blue = Friends
  • Green = Work/School
  • Yellow = Other

Insight: Most people have highly clustered, multi-community networks!

Exercise 2: Simulate Rumor Spread

Setup: Make a 10-person network:

1—2—3—4—5
|     |
6—7—8—9—10

Rules:

  • Day 0: Person 1 knows rumor
  • Each day: Each person tells ALL connected friends
  • Record who knows rumor each day

Calculate:

  • How many days to reach everyone?
  • Which person was last to hear? Why?
  • Remove person 4 (the bridge!). Now how long to spread?

Key finding: Removing bridges isolates whole sections!

Exercise 3: Build a Scale-Free Network

The “Rich Get Richer” Algorithm:

  1. Start with 3 connected people
  2. Add new person one by one (up to 20)
  3. New person connects to existing people with probability proportional to their current connections

Example: If Alice has 10 friends and Bob has 2, new person is 5× more likely to connect to Alice!

Result: You’ll get few hubs with many connections, many with few.

CMH insight: This explains why inequality emerges naturally in social systems!

PART 7: CMH APPLICATIONS — REAL-WORLD NETWORK EFFECTS

Application 1: Protest Movement Success

Research finding: Successful protests have decentralized leadership networks.

Failed protest network:

          Leader
         /  |  \
        A   B   C
        (if leader arrested, movement dies)

Successful protest network:

    A — B — C
    | X | X |
    D — E — F
    (multiple leaders, resilient!)

CMH rule: For resilience, build modular, decentralized networks with multiple bridges.

Application 2: Economic Inequality

The Job Search Paradox: People find jobs through weak ties (acquaintances, not close friends).

Why? Your close friends know the same opportunities you do. Distant connections know DIFFERENT opportunities!

Network consequence: People with diverse, weak-tie networks get better jobs → economic advantage compounds!

Inequality mechanism:

  1. Poor communities → Dense, strong-tie networks (trust but few opportunities)
  2. Rich communities → Sparse, weak-tie networks (access to diverse opportunities)
  3. Result: Network structure reproduces inequality

Application 3: Innovation Spread

Diffusion of innovations follows network structure:

  • Early adopters: Well-connected to outside world
  • Early majority: Connected to early adopters
  • Late majority: Need many friends to have adopted
  • Laggards: Isolated, highly traditional

The “Chasm” problem: Many innovations fail between early adopters and early majority (different networks!)

CMH strategy: To spread innovation (or idea), target bridges between network clusters!

PART 8: SOCIAL NETWORK ANALYSIS IN HISTORY

Case Study 1: Paul Revere vs. William Dawes

Both rode warning of British invasion (1775). Why do we remember Revere but not Dawes?

Network analysis reveals:

  • Revere: Central in multiple networks (political, commercial, social)
  • Dawes: Peripheral in fewer networks

Result: Revere’s warning spread through more and faster paths!

Revere's Network:  Militia — Merchants — Politicians — Artisans
                    \         |          /            /
                     \        |         /            /
                      \       |        /            /
                    ALARM → FAST SPREAD TO MANY GROUPS

Dawes's Network:     Friends — Family — Neighbors
                              \       /
                               \     /
                                \   /
                              SLOWER, LIMITED SPREAD

CMH lesson: Influence depends on network position, not just personal qualities!

Case Study 2: The Fall of the Habsburg Empire

Historical puzzle: Why did multi-ethnic Habsburg Empire collapse while similar Swiss Confederation survived?

Network analysis of elite marriages:

  • Habsburgs: Centralized network (everyone marries Habsburgs)
  • Swiss: Decentralized network (multiple interconnecting families)

Result: Remove Habsburg hub → whole network fragments. Remove any Swiss family → network still connected.

CMH insight: Network resilience matters for regime survival!

PART 9: ETHICAL DANGERS — THE DARK SIDE OF NETWORK ANALYSIS

The Surveillance Problem

Network analysis can:

  1. Map dissent networks for repression
  2. Identify key influencers for targeting
  3. Predict cascades to preempt protests
  4. Isolate communities through network disruption

Historical example: COINTELPRO (FBI) used network analysis to disrupt civil rights movement by targeting bridges between groups.

The Filter Bubble Problem

Algorithmic recommendations create:

  1. Homophily: Similar people connect more
  2. Increased clustering: Less exposure to different views
  3. Echo chambers: Ideas amplify within, don’t escape
  4. Polarization: Different clusters develop incompatible realities

CMH concern: Healthy societies need weak ties between clusters! Algorithms destroy these.

The “Network Determinism” Fallacy

Dangerous idea: “Your network position determines your fate.”

Reality: Networks shape opportunities, but individuals have agency. People can:

  • Build new connections
  • Bridge divided groups
  • Change network structures

CMH ethical rule: Study networks to empower connection, not to determine destiny.

PART 10: YOUR NETWORK ANALYSIS FIELD KIT

Quick Diagnostic Questions

When studying any social system, ask:

  1. What are the nodes? (Individuals? Groups? Institutions?)
  2. What are the edges? (Friendship? Communication? Trade?)
  3. What’s the network shape? (Random? Scale-free? Small-world?)
  4. Who are the hubs? (Influencers, leaders, bottlenecks)
  5. Where are the bridges? (Between communities, cultures, classes)
  6. How resilient is it? (Remove a hub — does it fragment?)
  7. How would X spread? (Info, disease, innovation, revolt)

Network Intervention Strategies

To spread something (innovation, information):

  • Target hubs first (scale-free network)
  • Or target bridges between clusters (small-world network)

To stop something (disease, rumor, panic):

  • Block hubs (vaccinate superspreaders)
  • Or block bridges between clusters (containment)

To build resilience:

  • Create redundant connections
  • Develop multiple hubs (decentralize)
  • Foster weak ties between clusters

YOUR MATH HOMEWORK

Exercise 1: Map an Historical Network

Choose one:

  • The American Revolution: Map connections between Founders
  • The Renaissance: Map patron-artist connections
  • Scientific Revolution: Map citation network of key papers

For your chosen network:

  1. List 10-15 key nodes
  2. Research their connections
  3. Draw the network
  4. Identify: Hubs, Bridges, Isolates
  5. Hypothesize: How did network structure affect outcomes?

Exercise 2: The “Small World” Experiment

Test “six degrees of separation”:

  1. Choose a target person (celebrity, politician, overseas friend)
  2. Start with yourself (degree 0)
  3. List people you know personally (degree 1)
  4. For each, list who THEY know who might know target (degree 2)
  5. Find shortest path
  6. Calculate actual degrees

Most people find: Distant targets are 3-4 degrees away, not 6!

CMH insight: The world has shrunk since the original 1960s experiment (found 5.2-5.7 degrees).

Exercise 3: Simulate Network Fragmentation

Create a polarized society network:

  1. Draw 20 nodes (people)
  2. Color 10 blue, 10 red (two groups)
  3. Initially: Each person connects to 3 others (random)
  4. Rewiring rule: Each year, people drop connections to other group, add connections to same group

Run simulation for 5 “years”:

  • Calculate: Between-group connections over time
  • Measure: Average path length between random blue-red pair
  • Observe: When does network fracture into two components?

Real-world parallel: Social media algorithms that show “more like this” increase homophily!

Exercise 4: Design a “Healthy Society” Network

You’re designing a new city. What network structure promotes:

  1. Innovation spread?
  2. Disease containment?
  3. Social cohesion?
  4. Economic mobility?
  5. Political resilience?

Consider trade-offs:

  • Dense networks vs. sparse networks
  • Strong ties vs. weak ties
  • Centralized vs. decentralized
  • Homophilous vs. diverse

Draw your ideal city network and explain your choices.

KEY TAKEAWAYS:

  1. Networks matter — connections create emergent properties
  2. Three basic shapes: Random, Scale-free, Small-world
  3. Key metrics: Degree, Betweenness, Clustering, Path length
  4. Spread follows structure: Same math for disease, ideas, revolts
  5. Position is power: Hubs and bridges have disproportionate influence
  6. Homophily is natural but dangerous (creates echo chambers)
  7. Weak ties are powerful for opportunities and innovation
  8. Resilience requires redundancy and multiple hubs

The CMH Mantra:

“Linear regression asks: What traits predict behavior?
Network theory asks: Who do you know, and who do they know?
In social systems, your position in the web of relationships matters as much as your personal attributes.
Revolutions don’t spread because individuals decide independently; they cascade through networks of trust.
Innovation doesn’t diffuse through populations; it hops across bridges between communities.
Inequality isn’t just about wealth; it’s about access to different networks of opportunity.
To change society, you must understand its hidden architecture of connections.”

NEXT LESSON PREVIEW:

Lesson 8: Agent-Based Modeling — Simulating Societies from the Bottom Up

Where we’ll learn:

  • How to simulate thousands of individual agents following simple rules
  • How emergence creates complex patterns from simple interactions
  • The Schelling Segregation Model: How mild preferences create extreme segregation
  • SugarScape: How resource distribution shapes societies
  • How to combine networks + agents for realistic CMH simulations

Thought to Ponder: Networks give us the structure of connections. Time series gives us dynamics over time. Agent-based modeling lets us simulate how individual decisions percolate through network structures to create historical dynamics. This is the holy trinity of CMH: Micro-level rules + Meso-level networks + Macro-level time patterns. When we combine all three, we can finally simulate societies in computers — not to predict the future, but to understand the space of possible futures.

Remember: In CMH, we’re not just studying societies as they are. We’re learning how to re-wire them for better outcomes. Network theory shows us that changing a few key connections can transform entire systems. Remove the bridges between hostile groups → conflict. Add bridges between isolated communities → innovation spreads. The social fabric isn’t fate — it’s architecture, and we can be the architects.

Similar Posts