From Molecules to Meaning: Can Assembly Theory Scale to Cognitive Systems?

Series: Assembly Theory | Part: 6 of 9

In the first five parts of this series, we explored Lee Cronin's audacious proposal that complexity can be measured through assembly—that the difference between life and non-life isn't some mysterious spark, but the accumulation of constraints in construction history. We've seen how assembly index distinguishes biological molecules from random chemistry, how life chemistry becomes special through recursive construction, and even how assembly theory differs fundamentally from Shannon information.

But here's the question that demands asking: If assembly theory works for molecules, does it work for minds?

Can the same framework that detects life in mass spectrometry readings detect meaning in neural activation patterns? Can we measure the assembly index of an idea, a memory, a cultural practice? Is there something about the construction of cognitive structures that parallels the construction of complex molecules—and if so, what does that tell us about consciousness, culture, and the nature of meaning itself?

This isn't just speculative play. If assembly theory captures something fundamental about how complexity emerges through selection, it should apply wherever selection operates. And selection operates on ideas, behaviors, and meanings just as surely as it operates on molecules.

The Assembly of Concepts

Start with a simple observation: Not all thoughts are equally complex.

The concept "red" is simpler than the concept "infrared photography." "Mom" is simpler than "attachment theory." "Hungry" is simpler than "intermittent fasting for autophagy optimization." These aren't just differences in word length—they're differences in conceptual construction.

Assembly index for concepts would measure how many prior concepts must exist before a new concept becomes thinkable. "Red" requires only direct sensory experience—low assembly. "Infrared photography" requires: light, color, spectrum, camera, film, invisible, extension-beyond-visible. Each of these concepts has its own prerequisites. The construction path is long.

In this framing, consciousness in a preliterate human has lower assembly index than consciousness in a literate one. Not because one is "smarter," but because literacy adds layers of cognitive infrastructure—phonemes, graphemes, syntax, grammar, genre conventions. Each layer enables new constructions that were previously impossible.

This isn't metaphor. Andy Clark's extended mind thesis and the entire 4E cognition framework (see our 4E Cognition series) argue that cognitive scaffolding is real infrastructure. Writing systems don't just record thoughts—they enable thoughts that couldn't exist without them. Try thinking "Keynesian macroeconomic policy" without language. Try thinking "integral calculus" without mathematical notation.

The assembly index of a thought is the depth of its dependency tree.

Memory as Constraint Accumulation

If concepts have assembly indices, what about memories?

A memory isn't a recording. Neuroscience abandoned the "memory as videotape" model decades ago. Memory is reconstitutive—every retrieval reconstructs from distributed neural patterns, often altering what gets reconsolidated. From a free energy principle perspective (see Karl Friston's framework), memory is a generative model that minimizes surprise about the past.

But here's where it gets interesting: Some memories are harder to construct than others, not because they're older, but because they require more cognitive infrastructure to encode in the first place.

Episodic memory—memory of specific events—requires: self-representation, temporal ordering, spatial context, emotional tagging, causal narratives. A rat can remember where food is (spatial memory), but it can't remember "the time I found food in the maze on Tuesday when the lights were dim." That kind of memory has higher assembly requirements.

Autobiographical memory—the constructed narrative of your life—has even higher assembly. It requires episodic memory as substrate, plus: narrative coherence, identity continuity across time, theory of mind (to imagine yourself as others see you), cultural templates for what counts as a "life story."

Assembly theory suggests memory isn't one thing—it's a stack of increasingly complex construction processes. And crucially: Not all organisms have access to all layers of the stack.

This maps cleanly to what we know about memory development. Children don't have autobiographical memory before age 3-4, not because they don't experience things, but because the cognitive infrastructure for that kind of memory hasn't assembled yet. The hardware is there. The assembly pathway isn't complete.

Cultural memory—shared narratives passed through generations—has higher assembly still. It requires language, yes, but also: ritual practices, mnemonic devices, institutional support, transmission protocols. The Iliad survived for centuries as oral tradition because Greek culture assembled the infrastructure (bards, meter, formulaic phrases, cultural importance) to maintain it. When that infrastructure collapses, high-assembly memories disappear, even if individual humans still have functioning brains.

Memory persistence scales with assembly infrastructure.

Ideas as Selected Objects

Now we're ready for the provocative move: treating ideas as units under selection.

Richard Dawkins proposed "memes" in 1976—ideas that replicate through cultural transmission. But memetics mostly failed as a research program because it lacked formal grounding. Assembly theory might provide what was missing: a way to measure the construction complexity of cultural replicators.

High-assembly ideas require extensive cognitive and cultural infrastructure to transmit.

Consider: "Electrons orbit nuclei like planets orbit stars" (Bohr model) versus "Electrons exist as probability distributions in quantum superposition until measurement collapses the wavefunction" (Copenhagen interpretation). Both are models of atomic structure. But the second has vastly higher assembly index—it presupposes: probability theory, wave mechanics, the measurement problem, the observer effect.

The Bohr model dominated early quantum mechanics not because it was more correct (it wasn't), but because its assembly requirements were lower. It could fit into existing cognitive infrastructure (planetary orbits). The Copenhagen interpretation required constructing entirely new conceptual categories.

Ideas with lower assembly index spread faster, all else being equal. This is why conspiracy theories outcompete nuanced analysis on social media. "Powerful people coordinate to control outcomes" has low assembly—humans grasp social hierarchy and deception intuitively. "Complex systems exhibit emergent properties through multi-agent interaction without centralized coordination" has high assembly—it requires: emergence, non-linearity, distributed causation, systems thinking.

But—and this is crucial—all else is rarely equal. High-assembly ideas can dominate if the cultural infrastructure supports them. Academic institutions, peer review, mathematical training, specialized vocabulary—these are assembly enablers. They make certain high-complexity ideas transmissible within communities that maintain the infrastructure.

Scientific knowledge is high-assembly information preserved by institutional infrastructure.

When the infrastructure degrades (e.g., civilizational collapse), high-assembly knowledge disappears first. You lose calculus before arithmetic, genome sequencing before agriculture, antibiotics before wound cleaning. Not because the high-assembly knowledge is less "useful," but because it requires more constructed preconditions to maintain.

This is exactly what assembly theory predicts for molecules: complex molecules require complex production machinery. When the machinery disappears, so does the complexity.

The Assembly of Language Itself

Let's go deeper: What about the infrastructure that makes conceptual assembly possible at all?

Language has an assembly index.

This isn't controversial. Historical linguistics tracks exactly this—how languages accumulate complexity through historical processes. But assembly theory offers a precise framework: language complexity = the minimum number of construction steps required to generate the grammatical and semantic structures of the language.

Pidgins have low assembly. They emerge when populations with no shared language need to communicate—stripped-down grammatical structure, limited vocabulary, mostly present tense, minimal subordination. Creoles have higher assembly—they're pidgins that have acquired native speakers and developed full grammatical complexity across one generation. The infrastructure assembled rapidly because children's brains do the construction work.

Academic registers have extremely high assembly. "The post-structuralist deconstruction of hegemonic discourse reveals latent power dynamics" requires: abstract nouns, nominalization, embedded clauses, specialized vocabulary, theoretical frameworks. This sentence is linguistically possible only after centuries of intellectual infrastructure construction.

But here's the key insight: Language doesn't just transmit ideas—it is cognitive infrastructure. Having words changes what you can think. The Sapir-Whorf hypothesis (linguistic relativity) gets mocked for its strong form ("language determines thought"), but the weak form is clearly true: Language constrains and enables thought.

You can't think "The day after tomorrow" efficiently if your language lacks grammatical future tense (you'd construct it from other pieces, but with higher cognitive cost). You can't think "schadenfreude" efficiently if your language lacks that lexical item (you'd describe it, but the single-word version is faster and thus more likely to be used).

Every language is an assembly pathway for thoughts. Different languages assemble different thought-spaces. This isn't mystical—it's mechanical. The available linguistic building blocks determine which complex thoughts assemble easily and which require extensive construction.

And crucially: Languages evolve under selection pressure. Words that enable high-value constructions stick around. Grammatical structures that reduce cognitive load spread. Language is a selected cognitive technology, and its assembly index reflects the accumulated construction history of the culture that built it.

Cultural Assembly Pathways

Zoom out to culture as a whole.

Culture is accumulated assembly infrastructure for thought and behavior. Every cultural practice—ritual, tool use, social norm, artistic tradition—is a construction that makes other constructions possible.

Consider mathematics. You can't do calculus without algebra. You can't do algebra without arithmetic. You can't do arithmetic without number concepts. You can't have number concepts without... what, exactly? Language? Finger counting? Subitization (automatic recognition of small quantities)? The assembly pathway goes all the way down to neural substrates and embodied experience.

But here's what makes culture wild: The assembly pathways are path-dependent and contingent.

The Greeks developed geometry to high sophistication without algebra because their mathematical tradition assembled through geometric construction (literally drawing shapes). Islamic mathematicians developed algebra (al-jabr) in the 9th century, creating an entirely different assembly pathway for mathematical thought. When European mathematics imported algebra in the Renaissance, it unlocked new constructions (analytic geometry, calculus) that were extremely difficult in the Greek geometric framework.

Assembly order matters. Some mathematical truths are hard to reach from geometry but easy from algebra. The infrastructure shapes what gets discovered.

This applies to everything cultural. Science assembled differently in different places—Chinese astronomy, Islamic optics, European mechanics—not because of inherent capacity differences, but because the construction pathways diverged based on what infrastructure already existed.

And once a pathway is established, switching costs are high. The QWERTY keyboard is famously suboptimal—it was designed to slow typists down so mechanical typewriters wouldn't jam. But the installed infrastructure (billions of keyboards, muscle memory in millions of fingers, typing instruction) makes switching to better layouts nearly impossible. High-assembly infrastructure resists replacement even when superior alternatives exist.

This is precisely what we see with biological molecules in assembly theory. Once a complex molecule is deeply integrated into metabolic pathways, replacing it requires reconstructing huge swaths of the system. Evolution conserves high-assembly infrastructure.

Culture does the same thing. The year 2025 is completely arbitrary (based on contested dating of Jesus's birth), but the infrastructure for using it—every computer system, every legal document, every historical record—makes switching to a better calendar essentially impossible.

Cultural lock-in is assembly entrenchment.

Meaning as Assembled Coherence

Now we can ask the hard question: What is meaning?

AToM (A Theory of Meaning) proposes: M = C/T — meaning equals coherence over time (or tension). But what is coherence, mechanically?

Assembly theory offers a precise answer: Coherence is the degree to which a system's components assemble into integrated structures rather than remaining independent.

A molecule with high assembly index is coherent—its parts are constrained by construction history to fit together in specific ways. Random molecules are incoherent—their parts could recombine arbitrarily without violating any constraints.

A thought with high conceptual assembly is coherent—its components (sub-concepts) are tightly constrained by logical relationships, semantic dependencies, inferential connections. Word salad is incoherent—the parts don't assemble into anything.

A person with integrated memories, consistent values, and stable identity is coherent—the components of selfhood assemble into a unified system. Someone in acute psychosis is incoherent—thoughts, perceptions, and identity fragments don't assemble correctly.

A culture with shared language, coordinated practices, and transmitted knowledge is coherent—the components assemble into a functioning system. A collapsing civilization is incoherent—the assembly infrastructure breaks down and complexity disappears.

Meaning is what happens when assembly creates coherence that persists through time.

This isn't metaphorical. When you understand a sentence, you're assembling a cognitive structure from words. When the structure coheres (syntax fits, semantics makes sense, pragmatics is appropriate), you experience meaning. When it doesn't cohere ("Colorless green ideas sleep furiously"—grammatical but semantically incoherent), you experience meaninglessness.

When a scientific theory makes sense, it's because the concepts assemble into a coherent model. When it feels confused or ad hoc, the assembly is incomplete or contradictory.

When your life feels meaningful, it's because your experiences, values, and actions assemble into a coherent narrative. When you experience existential crisis, the assembly has failed—the pieces don't fit together anymore.

The geometry of meaning is the geometry of assembly pathways in high-dimensional cognitive space.

Selection for Cognitive Assembly

If meaning is assembled coherence, what selects for it?

At the molecular level, selection is straightforward: configurations that enable replication persist. The molecules that assemble correctly get copied. The ones that don't, disappear.

At the cognitive level, it's more complex but structurally similar: Thoughts that assemble coherently propagate. Thoughts that don't, fade.

This happens within individual minds—coherent thoughts get reinforced through use, connected to other thoughts, integrated into memory. Incoherent thoughts are forgotten or revised. Your brain is constantly selecting for coherence.

It happens between minds—ideas that assemble easily in listener brains get transmitted. Ideas that require too much construction work to understand get lost. Communication is selection pressure for assembleability.

It happens across cultures—knowledge systems that cohere (science, mathematics, accumulated practical wisdom) get preserved and elaborated. Systems that don't cohere (pre-scientific cosmologies, alchemy, phrenology) eventually collapse under their own contradictions.

Cultural evolution is selection on assembly pathways for meaning.

And just like molecular evolution, it's not selecting for "truth" in any absolute sense—it's selecting for what works in the current environment. Conspiracy theories cohere within communities that share certain background assumptions (distrust of institutions, pattern-seeking bias, us-vs-them framing). They're assembled from available cognitive materials. They're meaningful to their adherents, even if they're false.

Scientific theories cohere within communities that share different infrastructure (mathematical training, experimental standards, peer review norms). They're assembled from different cognitive materials. They're meaningful in a different way—not emotionally satisfying narrative, but predictive and explanatory power.

Different assembly environments select for different forms of coherence.

The question isn't "which is objectively correct?" The question is "which assembly infrastructure produces the kind of coherence we value?"—and that's a choice about what infrastructure to build and maintain.

The Hard Problem: Subjective Experience

But wait. There's a gap here that needs addressing.

Assembly theory explains construction. It explains why some structures are more complex than others. It explains selection and persistence. But does it explain experience?

When I think "the sunset is beautiful," assembly theory can analyze:

• The conceptual dependencies (sunset, beauty, aesthetic judgment)
• The neural assembly process (perception → categorization → valuation)
• The cultural construction (beauty standards, romantic associations with sunsets)
• The transmission dynamics (why this thought is easily shared)

But does it explain what it's like to have that experience? The phenomenology, the qualia, the felt sense?

This is the hard problem of consciousness—the explanatory gap between mechanism and experience. And here assembly theory faces the same challenge every physical theory faces: You can explain the structure without explaining the feeling.

Or can you?

Consider this move: What if experience is what assembly feels like from the inside?

When you understand a sentence, the experience of "getting it" is the phenomenology of successful cognitive assembly. When you're confused, the feeling of confusion is the phenomenology of assembly failure. When something is meaningful, the felt sense of meaningfulness is the experience of coherent assembly. When life feels meaningless, it's because the assembly pathways have broken down.

If consciousness is (as Friston and Levin suggest) what free energy minimization feels like, and if coherent assembly is a form of free energy minimization (prediction error reduction), then subjective experience might be the intrinsic aspect of assembly processes.

This would make experience and construction two sides of the same process—the outside (physical) and inside (phenomenal) of what's happening when systems assemble coherent structures.

Is this satisfying? That depends on whether you think the hard problem requires a dualist solution (physical stuff + separate experiential stuff) or whether you accept that experience and mechanism might be the same process viewed from different perspectives.

Assembly theory doesn't solve the hard problem, but it offers a framework where experience isn't an add-on to mechanism—it's what mechanism is, intrinsically, for the system doing the assembling.

Practical Implications: Engineering Meaning

If meaning has assembly structure, we can engineer it.

Education becomes assembly pathway design. You can't teach calculus before algebra not because of student capacity but because the assembly order matters. Curriculum is construction order. The question isn't "what should students know?" but "what assembly pathways maximize later construction possibilities?"

The current educational system often teaches high-assembly content (abstract theories) without ensuring the low-assembly prerequisites are solid (conceptual foundations, worked examples, embodied understanding). Students memorize without comprehending—they have the words but not the assembled cognitive structure. It's like trying to synthesize a complex molecule without the precursor compounds.

Therapy becomes coherence repair. When someone's life story doesn't assemble coherently (trauma fragments, conflicting values, identity confusion), the therapeutic work is reconstruction. Not "finding the truth" but "building a version that coheres"—that assembles into something livable. Narrative therapy explicitly does this. So does psychodynamic integration. So does somatic experiencing (reassembling body-mind connection).

Cultural preservation becomes infrastructure maintenance. You can't preserve high-assembly cultural practices (traditional music, craft techniques, ritual forms) by just documenting them. You need to maintain the assembly infrastructure—the apprenticeship systems, the community contexts, the prerequisite skills. When the infrastructure disappears, the practice dies, even if you have perfect written instructions.

Communication becomes assembly cost management. The reason technical writing is hard isn't vocabulary—it's ensuring the reader has the assembly prerequisites. Every unexplained concept increases assembly cost. Every missing inferential step makes coherence harder. Good writing manages the reader's construction process.

Meaning crisis becomes visible as assembly failure. When people report lives feeling meaningless, assembly theory predicts: The traditional assembly pathways (religion, stable careers, tight-knit communities) have broken down, but new pathways haven't fully assembled yet. The infrastructure for constructing coherent life narratives is in transition. No wonder meaning feels scarce.

The solution isn't nostalgia (resurrect old assembly pathways that no longer fit current conditions) or nihilism (accept that meaning is impossible). It's construction: Build new assembly infrastructure. Create communities, practices, narratives, and institutions that enable high-complexity meaning to assemble in contemporary contexts.

If meaning is assembled, then meaning-making is a design problem.

Limitations and Open Questions

Assembly theory for cognition is speculative. It's not yet a research program with empirical results. So where are the gaps?

Measurement: We don't have a mass spectrometer for ideas. How do you actually compute the assembly index of a concept? What are the "building blocks" and "construction steps"? These need operational definitions before assembly theory becomes testable for cognitive systems.

Selection vs. drift: Molecular assembly assumes strong selection pressure—only configurations that work persist. But cultural evolution includes enormous amounts of drift. Fashion, slang, meme mutations—lots of cultural content spreads without functional value. How do we distinguish selected assembly from random drift?

Emergence: Complex molecules don't have novel properties beyond their components—they're just arranged differently. But consciousness seems to have emergent properties (subjective experience) that aren't obviously reducible to arrangement of parts. Does assembly theory handle emergence, or does it only handle construction?

Normativity: Assembly theory describes what is complex, but not what should be valued. A conspiracy theory can have high assembly (elaborate interlocking claims) without being true or beneficial. How do we distinguish high-assembly that increases coherence in valuable ways from high-assembly that's just elaborate delusion?

Individual variation: People construct different meanings from the same inputs. Two people read the same book—one finds it profound, one finds it nonsensical. Their assembly processes diverged. Can assembly theory account for why different cognitive systems assemble differently, or does it only describe the structure of what gets assembled?

These aren't fatal problems—they're research directions. If assembly theory for cognition is right, there should be answerable questions here.

Where Assembly Meets Free Energy

We're ready to connect this to the broader theoretical landscape.

Karl Friston's Free Energy Principle says living systems minimize surprise—they maintain themselves by predicting and controlling their sensory inputs. Assembly theory might be the flip side: Systems that minimize free energy accumulate assembly.

Here's the mechanism: Every time you reduce prediction error, you're reinforcing a cognitive structure (this worked, encode it). Over time, structures that consistently reduce surprise get assembled into increasingly complex predictive models. Your brain assembles higher-order concepts because they compress prediction—"dog" is a high-assembly concept that lets you predict a vast range of sensory inputs with a single label.

Free energy minimization drives assembly construction. The systems that survive are the ones that build high-assembly predictive models, because those models reduce surprise most efficiently.

This works for molecular systems: A cell assembles complex metabolic machinery because it reduces surprise about resource availability. It works for cognitive systems: A brain assembles complex conceptual hierarchies because they reduce surprise about sensory inputs. It works for cultural systems: A society assembles complex institutions because they reduce surprise about social coordination.

Assembly is what you get when free energy minimization operates over time.

The direction of our next article is clear: We need to explore where assembly theory meets free energy principle—two frameworks that might be describing the same deep principle about what it takes for complex systems to persist.

But before that, we have one more speculative extension: If assembly theory works for cognitive systems, does it work for detecting cognitive systems? Could we use assembly index to search for intelligence elsewhere in the universe?

This is Part 6 of the Assembly Theory series, exploring how selection and construction interact to produce complexity.

Previous: Selection as Constructor: Where Assembly Theory Meets Constructor Theory
Next: Detecting Life on Other Worlds: Assembly Theory's Cosmic Implications

Further Reading

• Cronin, L., & Walker, S. I. (2016). "Beyond prebiotic chemistry." Science, 352(6290), 1174-1175.
• Marshall, S. M., et al. (2021). "Identifying molecules as biosignatures with assembly theory and mass spectrometry." Nature Communications, 12, 3033.
• Clark, A., & Chalmers, D. (1998). "The Extended Mind." Analysis, 58(1), 7-19.
• Friston, K. (2010). "The free-energy principle: a unified brain theory?" Nature Reviews Neuroscience, 11, 127-138.
• Dawkins, R. (1976). The Selfish Gene. Oxford University Press.
• Boyd, R., & Richerson, P. J. (1985). Culture and the Evolutionary Process. University of Chicago Press.
• Henrich, J. (2015). The Secret of Our Success: How Culture Is Driving Human Evolution. Princeton University Press.