I had created a post called "Species as Grammars" in May of 2024.
I worked with Gemini AI to enhance that post and the results are posted below.
In a nutshell, AI knows more than us and can take our half-baked, weak ideas and enhance them by supplying conceptual connections to work that has already been done in that domain.
Posted here as well.
Species as Grammars: A Non-Reductionist, Biosemiotic Framework for Life and Intelligence
Abstract
We propose a non-reductionist framework for theoretical biology that treats Thought (Intelligence) and Extension (Matter) as irreducible, co-extensive attributes of the Universe, following the monistic philosophy of Baruch Spinoza. Abandoning the reductionist quest to derive mind from brute matter—a dead end analogous to the scholastic search for the essence of motion—we model biological species as generative formal grammars. Drawing on André Joyal’s Theory of Combinatorial Species, biosemiotics, the Chomsky Hierarchy, and the Chomsky–Schützenberger equivalence theorem, we demonstrate that organisms function as natural Turing machines whose genetic, epigenetic, and behavioral structures constitute formal languages. Within this architecture, evolution operates as an algebraic meta-grammar that composes new species-specific syntactic spaces from conserved molecular alphabets, while natural selection acts as the pragmatic evaluator testing semantic viability against physical reality.
1. Introduction: Beyond the Reductionist Deadlock
Under the influence of the monistic philosophy of Baruch Spinoza, we arrive at a foundational epistemological premise: we cannot fundamentally define "Thought," "Intelligence," "Life," or "Matter" ("Extension" in Medieval and Spinozist philosophy). Attempting to explain the origin of Life or Intellect purely in terms of brute Matter—as contemporary reductionist science often tries to do—is as barren as the Medieval Scholastics' search for the metaphysical origin of Motion.
Empirical physics only truly began when Galileo and his successors shifted focus from asking what motion is to formulating a quantitative, mathematical description of how motion behaves—whether a ball rolling down an inclined plane or a planet orbiting a celestial body.
The key conceptual shift, therefore, is to treat Thought or Intelligence as a qualitatively irreducible feature of the Universe—inseparably intertwined with Matter without being reducible to it—and to stop trying to derive mind from non-mind. One immediate implication of this perspective is the introduction of Language and Grammar into biology across every organizational scale, from molecular sequences to entire species:
- Biosemiotics and Molecular Linguistics: Recent empirical and theoretical work already points in this direction. Research on fungal electrical spiking activity suggests that fungi exhibit a lexicon of distinct electrical "words" for intercellular communication (Adamatzky, 2022). At the molecular level, Hubert Yockey’s Information Theory, Evolution, and the Origin of Life (2005) establishes how information theory governs molecular biology, while Bernd-Olaf Küppers (The Language of Living Matter, 2022, building on Manfred Eigen’s work on self-organization) argues that the boundary between physics and biology is marked by the emergence of semantic information—sequences that carry functional meaning relative to a biological context rather than mere statistical entropy. Furthermore, work bridging information theory with linguistic structure (e.g., Gibson et al., 2019) shows how functional complexity leads directly to mathematical descriptions of possible languages as solutions to constrained optimization problems.
2. Thought, Social Fields, and Cognitive Grammars
What, then, is a Language? For human beings, we use language to model the world, make judgments, and communicate those judgments to others. While some argue that non-linguistic thought exists, entertaining the premise that Thought, Language, and Intellect are coterminous (or deeply overlapping) yields a powerful insight: because verbal language is a social construct, a significant portion of individual intelligence emerges from participating in a communicating collective.
This mirrors mean-field theory in physics, where a single particle experiences an effective force generated by the collective behavior of all other particles in the system—an idea echoed in cognitive science through distributed cognition (Hutchins, 1995) and cultural evolution. Framed this way, comparing the cognitive capacities of different human or animal societies—asking, for instance, "Was Rome more 'intelligent' than Samoan tribes?"—ceases to be a vague philosophical question and becomes a tractable scientific inquiry. Rather than ranking societies on a single scalar axis, we can evaluate the expressive power and domain-specialization of their respective cultural-linguistic grammars against the specific constrained optimization problems they faced (e.g., imperial logistics and codified law versus oceanic wayfinding and ecological stewardship).
Moreover, verbal language is complemented by chemical, tactile, gestural, and musical languages. If epigenetic inheritance holds across behavioral and physiological domains, many of these non-verbal grammars can be acquired, stabilized, and transmitted across generations alongside genomic structures.
3. From Umwelt to Functorial Grammars
Aristotle defined Man as the talking animal, yet every species is defined by its genomic, genetic, epigenetic, and behavioral transformations. To make this precise:
- A Language ($L$) is a set of symbols (nucleotides, epigenetic marks, neural spikes, gestures, words).
- A Grammar ($G$) is a formal set of rules for generating syntactically valid structures from those symbols.
In Category Theory—specifically within André Joyal’s Theory of Combinatorial Species (1981)—a generative grammar or structure type can be expressed as an analytic functor (an infinite power series) acting on a base set $L$:
$$G = L + L \times L + L \times L \times L + L \times L \times L \times L + \dots$$
Viewed through this lens, every species is its own generative grammar: a Dog is its Language; a Human is its Language. This formalizes Jakob von Uexküll’s biosemiotic concept of the Umwelt (1934)—the species-specific perceptual and operational world in which an organism interprets physical stimuli as meaningful signs. Because humans and dogs can partially communicate, their respective grammars must share a non-empty intersection, a structural overlap mirrored by their shared evolutionary and physiological homologies. Extending this logic implies the existence of a Universal Biological Grammar on Earth governing organismal construction, intra-species signaling, and inter-species communication, with Evolution acting as a higher-order meta-grammar that composes new species-grammars from conserved modular blocks.
4. Connections to Physics
Far from being a mere metaphor borrowed from linguistics, the idea that physical reality and biological systems share a common syntactic and computational architecture aligns directly with modern developments in statistical mechanics, quantum information theory, and theoretical physics:
- Constructor Theory as Physical Grammar: David Deutsch and Chiara Marletto’s Constructor Theory (2013, 2015) reformulates fundamental physics not in terms of trajectories and initial conditions, but through the grammar of transformations—strictly defining which physical tasks are possible versus impossible. Living organisms act as universal constructors instantiating specific informational grammars.
- Generating Functions and Statistical Mechanical Partition Functions: Joyal's combinatorial generating functions ($G(x) = \sum_{n=0}^{\infty} |G[n]| \frac{x^n}{n!}$) share an exact mathematical isomorphism with the canonical partition function in statistical mechanics ($Z = \sum_{s} e^{-E(s)/k_B T}$), translating local combinatorial rules into global macro-states.
- Feynman Diagrams as Algebraic Grammars: In quantum field theory, particle interactions form a rigorous visual and algebraic grammar where scattering amplitudes are expressed as perturbative Dyson power series identical in structure to Joyal's analytic functors.
- Thermodynamic Information and Landauer’s Principle: Erasing or processing logical bits incurs a minimum heat dissipation ($E = k_B T \ln 2$; Landauer, 1961). Every biological syntactic operation (DNA proofreading, signaling cascades) operates as a non-equilibrium thermodynamic computation.
- The Renormalization Group and Multi-Scale Grammars: Just as Renormalization Group (RG) equations coarse-grain microscopic interactions into macroscopic physical laws (Wilson, 1971), biological grammars map nucleotide syntax into cellular, tissue, and ecological structures across spatial scales.
- Spontaneous Symmetry Breaking as Grammatical Constraint: Phase transitions restrict high-symmetry physical state spaces into ordered ground states (crystals, superfluids). Similarly, biological evolution enforces strict syntactic constraints (such as homochirality and the universal genetic code) to stabilize low-entropy living structures.
- Wheeler’s "It from Bit" and Spinozist Monism: John Archibald Wheeler’s doctrine (1990) posits that every physical entity derives its ultimate significance from binary informational distinctions, providing a direct physical bridge to Spinoza’s dual-aspect monism where Matter (Extension) and Mind (Thought) are complementary readouts of the same informational substrate.
5. Syntax, Semantics, and Natural Selection: The Pragmatic Evaluator
A persistent challenge when mapping linguistics onto biology is the distinction between syntax (the structural rules governing symbol combination) and semantics (meaning, utility, and reference).
- The Syntactic Level: At the molecular scale, nucleotides ($\text{A}, \text{C}, \text{G}, \text{T}$) form an alphabet. Transcription and translation operate as syntactic rules that string these characters into codons, genes, and proteins. Similarly, fungal electrical spiking activity groups ionic action potentials into temporal "words" and "sentences" whose length distributions match human linguistic laws.
- The Semantic Trap: In human language, a sentence can be syntactically immaculate ("Colorless green ideas sleep furiously") yet semantically meaningless. In biology, a mutation can produce a syntactically valid protein sequence that folds correctly according to thermodynamic rules yet fails entirely because it cannot bind its molecular target or survive cellular stress.
- Natural Selection as the Pragmatic Evaluator: Natural selection does not write the grammar; rather, it acts as the pragmatic evaluator. It tests whether a generated grammatical sentence (an organism, a metabolic pathway, a behavioral routine) is functionally coherent within the physical and ecological constraints of its environment. Life is a continuous problem-solving computation where successful semantic alignment yields survival, and syntactic-semantic mismatch results in extinction.
6. The Chomsky Hierarchy Across Biological Organization
To formalize biological complexity, the four nested classes of formal grammars in the Chomsky Hierarchy (Chomsky, 1956; Searls, 2002) map directly onto distinct levels of molecular and cellular organization:
| Chomsky Level | Automaton Equivalent | Biological Mechanism | Structural / Functional Example |
|---|---|---|---|
| Type 3: Regular | Finite-State Automaton | Local linear sequence motifs & binding sites | TATA boxes, Shine-Dalgarno sequences, restriction enzyme recognition sites, and simple protein transmembrane helices. |
| Type 2: Context-Free | Pushdown Automaton | Nested base-pairing & hierarchical folding | RNA secondary structures (stem-loops, hairpins, cloverleaf tRNA) where nested complementary base pairs ($A-U, G-C$) require stack memory. |
| Type 1: Context-Sensitive | Linear-Bounded Automaton | Crossing dependencies & epigenetic chromatin state | RNA pseudoknots, allosteric protein-DNA interactions, enhancer-promoter looping, and context-dependent alternative splicing. |
| Type 0: Unrestricted | Universal Turing Machine | Whole-cell gene regulatory & metabolic networks | Self-modifying chromatin remodeling, embryonic morphogenesis, and adaptive cellular signal transduction. |
7. Species as Natural Turing Machines: Cellular Computation Without a CPU
In theoretical computer science, the Chomsky–Schützenberger / Turing equivalence establishes that Unrestricted Grammars are computationally equivalent to Turing Machines—the theoretical foundation of universal problem-solving. Because every living species is continuously solving the problem of survival and adaptation, its underlying biological grammar operates as a distributed computational engine.
Unlike a conventional von Neumann computer—which strictly separates passive memory (the tape) from an active central processing unit (the read/write head)—a single-celled organism functions as a "Natural Automaton" without a centralized CPU:
- Symbols and Memory Registers: Concentrations of transcription factors, metabolites, and epigenetic marks serve as formal symbols on the cellular "tape." Bistable double-negative feedback loops (genetic toggle switches) implement single-bit flip-flop memory registers that persist across cell divisions (Gardner et al., 2000).
- Logic Gates via Allostery: Cooperative binding of transcription factors to promoter regions (modeled by Hill functions) generates sigmoidal switches that execute Boolean AND, OR, and NOT gates.
- Turing Universality via Chemical Reaction Networks (CRNs): Any system capable of simulating a Register Machine with at least two unbounded counters (increment, decrement, and zero-test) is Turing-universal. Soloveichik, Winfree, and colleagues (2008) proved that mass-action and stochastic CRNs are Turing-complete: enzymatic synthesis increments molecular copy-number counters, degradation decrements them, and high-affinity competitive titration executes a zero-check.
- Self-Modifying Hardware: In living cells, the hardware and software are co-extensive. The cell’s membrane, cytoskeleton, and chromatin remodeling complexes simultaneously act as the tape and the read/write head—computing by physically restructuring their own logical circuitry.
8. Evolutionary Transitions as Algebraic Meta-Grammar Operators
Major evolutionary transitions—such as the origin of chromosomes, endosymbiosis (eukaryogenesis), multicellularity, and eusociality (Maynard Smith & Szathmáry, 1995)—can be formalized as algebraic operators acting on species-grammars within Joyal’s category of combinatorial species:
- Endosymbiosis as Species Product ($F \cdot G$): When two distinct prokaryotic grammars ($F$ and $G$, such as an archaeal host and an alphaproteobacterial proto-mitochondrion) merge into a chimeric eukaryote, the resulting grammar is a partitioned product $(F \cdot G)[U] = \sum_{U = U_1 \sqcup U_2} F[U_1] \times G[U_2]$, pooling two independent metabolic alphabets into a unified syntactic engine.
- Multicellularity and Hierarchical Differentiation as Species Composition ($F \circ G$): Multicellular morphogenesis substitutes differentiated cellular sub-grammars ($G_1, G_2, \dots, G_k$) into the spatial slots of a higher-order tissue or body-plan grammar ($F$): $$(F \circ G)[U] = \sum_{\pi \in \text{Part}(U)} F[\pi] \times \prod_{B \in \pi} G[B]$$ Here, gene regulatory networks act as context-sensitive parsers that silence or activate subsets of the shared genomic alphabet depending on spatial positional cues (morphogen gradients).
- Symbiogenesis and Horizontal Gene Transfer as Grammar Union ($F + G$) and Rewrite Rules: Horizontal gene transfer inserts modular syntactic subroutines (operons, antibiotic resistance cassettes) directly across distinct species grammars, accelerating evolutionary search without requiring de novo grammatical synthesis.
9. Empirical Research Program: Measuring Inter-Species Semiotic Overlap
To transition the Species-as-Grammars thesis from philosophy to testable science, we can measure the shared intersection between human linguistic/gestural grammars ($G_H$) and canine perceptual-behavioral grammars ($G_D$).
Rather than testing whether a dog "understands English" in an anthropomorphic sense, we treat interaction as a channel-coding problem between two distinct Umwelten. We decompose human communicative acts into orthogonal channels:
- Lexical/Phonemic Channel ($L_{\text{lex}}$): Arbitrary phonetic strings (e.g., "sit", "ball", "kitchen").
- Prosodic/Acoustic Channel ($L_{\text{pros}}$): Pitch contour, amplitude modulation, and rhythm (high rising pitch for arousal/play vs. low falling pitch for inhibition).
- Kinesic/Spatial Channel ($L_{\text{kin}}$): Gaze direction, pointing geometry, and postural orientation.
Experimental Protocol & Quantitative Metrics:
- Factorial Decoupling Trials: Present domestic dogs (Canis lupus familiaris) with systematically crossed stimuli—for example, pairing a reward-associated lexical token ("walk") with an inhibitory prosodic contour, or pairing a neutral nonsense word with a referential pointing gesture—while recording behavioral latency, choice accuracy, and gaze-tracking trajectories.
- Mutual Information & Channel Capacity ($I(X; Y)$): Compute the Shannon mutual information between human signal features ($X$) and canine behavioral states ($Y$). High mutual information in $L_{\text{pros}}$ and $L_{\text{kin}}$ across untrained dogs quantifies the innate, evolutionarily conserved core of $G_H \cap G_D$ (shared mammalian affective and attentional syntax).
- Cross-Species Latent Space Alignment: Using contrastive representation learning, map high-dimensional acoustic/kinesic embeddings of human commands alongside canine vocalizations, tail kinematics, and awake-fMRI neural activation patterns (building on Berns et al., 2012). Structural isomorphism between the two latent spaces provides a direct geometric measure of grammatical overlap.
10. Mathematical Appendix: Protein Folding Grammar as a Combinatorial Species
To formalize how a biological grammar generates valid molecular structures from a finite alphabet, we apply André Joyal’s Theory of Combinatorial Species (Joyal, 1981; Bergeron et al., 1998).
10.1 Functorial Definition and Analytic Functors
Let $\mathbb{B}$ be the category of finite sets and bijections. A combinatorial species is a functor $F : \mathbb{B} \to \mathbb{B}$. For any finite set $U$ of labels (e.g., sequence positions $\{1, 2, \dots, n\}$), $F[U]$ is the finite set of all valid biological structures (such as secondary/tertiary fold topologies) constructible on $U$. Relabeling positions via a bijection $\sigma : U \to V$ induces a canonical transport of structure $F[\sigma] : F[U] \to F[V]$, ensuring the structural grammar is intrinsic and independent of arbitrary indexing.
Associating $F$ with its exponential generating function (EGF) yields:
$$F(x) = \sum_{n=0}^{\infty} |F[n]| \frac{x^n}{n!}$$
Extending this from sets to an alphabet-weighted set $L$ (where $L = \{\text{A}, \text{C}, \text{D}, \dots, \text{Y}\}$ is the 20-letter amino acid alphabet) defines $F$ as an analytic functor:
$$F(L) = \sum_{n=0}^{\infty} F[n] \times_{\mathfrak{S}_n} L^n$$
where $L^n$ is the set of all primary sequences of length $n$, and the quotient by the symmetric group $\mathfrak{S}_n$ accounts for structural symmetries.
10.2 Recursive Grammar for Nested Contact Topologies
Consider a simplified species $P$ representing valid nested intramolecular contact topologies (such as $\beta$-hairpins and antiparallel $\beta$-sheets) along a sequence of residues. Let $X$ be the singleton species representing a single unpaired amino acid residue ($X(x) = x$), and let $C$ be the species of a paired hydrogen-bonded contact bridge ($C(x) = x^2$). A recursive combinatorial grammar for nested fold architectures is defined by the species equation:
$$P = 1 + X \cdot P + C \cdot P \cdot P$$
where $1$ denotes the empty chain. Passing to the generating function $P(x)$ yields the algebraic equation:
$$P(x) = 1 + x P(x) + x^2 P(x)^2$$
Solving for $P(x)$ gives the closed-form analytic generating function:
$$P(x) = \frac{1 - x - \sqrt{1 - 2x - 3x^2}}{2x^2} = 1 + x + 2x^2 + 4x^3 + 9x^4 + 21x^5 + \dots$$
The coefficients ($1, 1, 2, 4, 9, 21, \dots$) are the Motzkin numbers, which exactly enumerate all syntactically valid non-crossing secondary contact topologies of chain length $n$.
Finally, introducing a semantic weight functor $W : P(L) \to \mathbb{R}_{>0}$ driven by the Boltzmann factor $W(s, \tau) = e^{-\Delta G(s, \tau) / k_B T}$ assigns each syntactically valid fold $\tau \in P[n]$ on primary sequence $s \in L^n$ its thermodynamic free-energy stability, bridging formal combinatorial syntax with physical selection.
11. Conclusion
By uniting Spinoza’s dual-aspect monism with formal language theory, category theory, theoretical physics, and biosemiotics, Species-as-Grammars offers a non-reductive framework for studying life: not as inert matter mysteriously animated by mind, but as self-computing, evolving languages written into the fabric of nature.
12. References
- Adamatzky, A. (2022). Language of fungi derived from their electrical spiking activity. Royal Society Open Science, 9(4), 211926. https://doi.org/10.1098/rsos.211926
- Aristotle (c. 350 BCE / 1998). Politics (Book I, 1253a). Trans. C. D. C. Reeve. Hackett Publishing.
- Bergeron, F., Labelle, G., & Leroux, P. (1998). Combinatorial Species and Tree-like Structures. Cambridge University Press.
- Berns, G. S., Brooks, A. M., & Spivak, M. (2012). Functional MRI in awake unrestrained dogs. PLoS ONE, 7(5), e38027. https://doi.org/10.1371/journal.pone.0038027
- Chomsky, N. (1956). Three models for the description of language. IRE Transactions on Information Theory, 2(3), 113–124. https://doi.org/10.1109/TIT.1956.1056813
- Chomsky, N., & Schützenberger, M. P. (1963). The algebraic theory of context-free languages. In Computer Programming and Formal Systems (pp. 118–161). North-Holland.
- Deutsch, D. (2013). Constructor theory. Synthese, 190(18), 4331–4359. https://doi.org/10.1007/s11229-013-0279-z
- Deutsch, D., & Marletto, C. (2015). Constructor theory of information. Proceedings of the Royal Society A, 471(2174), 20140540. https://doi.org/10.1098/rspa.2014.0540
- Eigen, M. (1971). Selforganization of matter and the evolution of biological macromolecules. Naturwissenschaften, 58(10), 465–523. https://doi.org/10.1007/BF00623322
- Gardner, T. S., Cantor, C. R., & Collins, J. J. (2000). Construction of a genetic toggle switch in Escherichia coli. Nature, 403(6767), 339–342. https://doi.org/10.1038/35002131
- Gibson, E., Futrell, R., Piantadosi, S. P., Dautriche, I., Mahowald, K., Bergen, L., & Levy, R. (2019). How efficiency shapes human language. Trends in Cognitive Sciences, 23(5), 389–407. https://doi.org/10.1016/j.tics.2019.02.003
- Hutchins, E. (1995). Cognition in the Wild. MIT Press.
- Joyal, A. (1981). Une théorie combinatoire des espèces formalisées. Advances in Mathematics, 42(1), 1–82. https://doi.org/10.1016/0001-8708(81)90052-9
- Küppers, B.-O. (2022). The Language of Living Matter: How Molecules Acquire Meaning. Springer. https://doi.org/10.1007/978-3-030-89488-7
- Landauer, R. (1961). Irreversibility and heat generation in the computing process. IBM Journal of Research and Development, 5(3), 183–191. https://doi.org/10.1147/rd.53.0183
- Maynard Smith, J., & Szathmáry, E. (1995). The Major Transitions in Evolution. Oxford University Press.
- Searls, D. B. (2002). The language of genes. Nature, 420(6912), 211–217. https://doi.org/10.1038/nature01255
- Soloveichik, D., Cook, M., Winfree, E., & Bruck, J. (2008). Computation with finite stochastic chemical reaction networks. Natural Computing, 7(4), 615–633. https://doi.org/10.1007/s11047-008-9067-y
- Spinoza, B. (1677 / 1996). Ethics. Trans. E. Curley. Penguin Classics.
- von Uexküll, J. (1934 / 2010). A Foray into the Worlds of Animals and Humans: With a Theory of Meaning. Trans. J. D. O'Neil. University of Minnesota Press.
- Wheeler, J. A. (1990). Information, physics, quantum: The search for links. In W. H. Zurek (Ed.), Complexity, Entropy, and the Physics of Information (pp. 3–28). Addison





