In the early 1600s, alchemist Jan Baptist van Helmont planted a willow sapling in 200 pounds of soil. After five years of watering, the tree weighed 169 pounds—while the soil lost just 2 ounces. His conclusion? Water transformed into wood. He missed the truth: carbon drawn from the air. This flawed reasoning epitomizes pre-mechanistic science—a world where magnets lost power if rubbed with garlic, and basil allegedly bred scorpions.

Article illustration 5

The Triad of Scientific Maturity: Entities, Properties, Rules

Mature sciences transcend observation by defining three core elements:
1. Entities: Fundamental components (e.g., protons in physics, nucleotides in genetics)
2. Properties: Their measurable characteristics (e.g., atomic mass, nucleotide sequences)
3. Rules: Interaction principles (e.g., chemical bonding, evolutionary selection)

Chemistry’s maturity emerged when it replaced ‘essences’ with Mendeleev’s periodic table—118 elements reducible to protons, neutrons, electrons. Biology followed by shifting from classifying species to analyzing genetic machinery:

“Two ‘breast cancers’ may need different treatments, while tumors in different organs might share therapy—if they harbor identical mutations. Location is impressionistic; genetics is mechanistic.”

Biology’s Unfinished Revolution

Despite breakthroughs, biology struggles with exceptions:
- Cell types lack a finite taxonomy like the periodic table
- Central Dogma (DNA→RNA→protein) buckles under reverse transcriptase
- Tools like CRISPR reveal mechanisms but expose deeper complexity

Article illustration 1

The path from observation to mechanism requires sharper tools—like telescopes for Galileo or gene sequencers for modern oncologists.

Psychology’s Alchemy Problem

Psychology remains stuck in its pre-mechanistic phase:
- Abstract constructs (‘grit,’ ‘extraversion’) create tautologies:

Q: Why is Sarah outgoing? 
A: High extraversion. 
Q: What is extraversion? 
A: Whatever this scale measures.

- The Sims and neural networks hint at entity-based models but lack biological fidelity
- Control systems (e.g., thermostat-like feedback loops) offer promise for modeling drives

Why Tech Holds the Key

  • Dwarf Fortress’ drunk cats: Unintended vodka-poisoned felines emerged from nested entity interactions (alcohol + paw-licking code). Real-world systems exhibit similar emergent chaos.
  • Neural networks: Simplify cognition to neurons/weights but risk conflating emulation with explanation.
  • Flyball governors: Hierarchical control systems that automate tasks (cruise control, HVAC) mirror biological regulation.

For AI and computational biology, the lesson is clear: Mechanistic maturity isn’t about complexity—it’s about identifying irreducible components. Until psychology finds its ‘base pairs,’ it will keep naming phenomena without explaining them. The alchemists wore funny hats; modern scientists write Python. But without entities, properties, and rules? They’re just changing the dress code.

Source: Slime Mold Time Mold. What Makes a Mature Science. Asimov Press (2025).