The Wetware of Grammar

 

Neurobiological Constraints on Minimalist Syntax

For more than half a century, formal syntactic theory and cognitive neuroscience advanced under largely incommensurable explanatory regimes. Within generative grammar, most prominently in the Minimalist Program of Noam Chomsky, language was modeled as an abstract, law-governed computational system, minimally characterized by structure-building operations such as Merge, feature valuation, and cyclic transfer via phases. Neuroscience, by contrast, pursued the biological substrate of language without a principled bridge to these formally articulated constructs, often treating syntax as an emergent behavioral epiphenomenon rather than a computational reality.

What defines the contemporary landscape is neither convergence nor contradiction in any simplistic sense but a more intellectually demanding configuration: partial commensurability under irreducible representational mismatch. Empirical neuroscience increasingly corroborates the existence of hierarchical, structure-sensitive computations in the human brain. Yet it simultaneously resists any clean isomorphism between neural dynamics and the derivational architecture assumed in formal syntax.

The central theoretical question is no longer whether syntactic structure is “in the brain.” It is, more precisely:

How can a formally discrete, recursive, and feature-driven combinatorial system be implemented by a noisy, time-bound, resource-constrained biological architecture without collapsing into either reductionism or metaphor?

1.1 Hierarchical Structure as a Neurocognitive Constraint, Not a Neural Copy

Hierarchy vs. Linear Order

A durable convergence across neuroimaging and electrophysiological research is the sensitivity of human language comprehension to hierarchical structure rather than linear adjacency. Across paradigms employing syntactically well-formed but semantically attenuated stimuli, most notably “Jabberwocky” constructions, the brain reliably tracks increasing structural complexity even in the near absence of lexical semantics.

$$[TP\; The\; vackish\; glim\; [vP\; snerled\; the\; flis]]$$

Such stimuli do not merely degrade semantic content; they isolate a more fundamental property of language processing: the persistence of syntactic organization as an independent computational constraint.

Neurocognitive data consistently implicate a distributed fronto-temporal network, most reliably including the left inferior frontal gyrus (IFG) and posterior superior temporal cortex, in responding to hierarchical load. However, the theoretically decisive point is methodological: these activations do not warrant the inference of explicit tree-like encodings in the representational sense assumed by formal syntax.

The empirically defensible claim is considerably more restrained and more powerful:

Core Constraint: The brain is sensitive to hierarchical dependencies, implemented through distributed predictive, integrative, and memory-dependent computations rather than discrete symbolic tree construction.

In this framing, “hierarchy” is not a neural object; it is a computational constraint realized across interacting neural populations.

Merge and Combinatorial Binding

Within the Minimalist architecture, Merge functions as the irreducible combinatorial operation generating hierarchical structure from simpler lexical elements. Its theoretical appeal lies in its radical parsimony: a single recursive mechanism replaces the proliferating rule systems of earlier generative models.

Neuroscientific investigations of combinatorial language processing confirm a robust distinction between structured and unstructured linguistic input at early stages of cortical response. Fronto-temporal circuits systematically differentiate between compositional phrases and non-integrated lexical sequences.

However, a critical constraint emerges from the empirical literature: no known neural signature uniquely instantiates Merge as a discrete, isolable operation. Instead, the evidence converges on a set of partially overlapping mechanisms:

Graded compositional integration: structure emerges as a continuum of binding strength rather than binary combinatorial steps

Predictive binding dynamics: cortical systems pre-activate likely syntactic continuations under uncertainty

Working-memory chunking: linguistic input is packaged into temporally stable representational units for downstream integration

The consequence is theoretically significant:

Merge retains its status as a formally indispensable explanatory primitive, but its neurobiological realization is best understood as distributed and emergent rather than localized and event-discrete.

Phase Structure and Temporal Dynamics

Minimalist Phase Theory posits that syntactic derivations proceed via cyclic domains, most notably vP and CP, which are transferred to interface systems upon completion of local computational phases. This cyclic architecture is intended to regulate derivational opacity and computational economy.

Converging evidence from neurophysiology suggests that speech perception and language comprehension unfold across multiple nested temporal scales, reflected in oscillatory brain dynamics.

[CORTICAL OSCILLATORY ARCHITECTURE]

Gamma / Theta (~4 Hz)  → syllabic parsing
Delta (~2 Hz)          → phrasal integration
Low Delta (~1 Hz)      → sentence-level structuring (CP-scale)

These temporal hierarchies are not merely descriptive conveniences; they reflect a fundamental property of cortical computation: multi-scale temporal integration under predictive sampling constraints.

Yet the theoretical temptation to equate oscillatory bands with syntactic phases must be resisted. The mapping is, at best, analogical and not identity-preserving.

Therefore, the most defensible formulation is:

Neural oscillations instantiate a general-purpose mechanism for hierarchical temporal segmentation, which language recruits but does not uniquely determine.

1.2 Predictive Processing and the Revision of Derivational Architecture

From Bottom-Up Derivation to Predictive Construction

Classical derivational models often present syntax as a bottom-up procedure: lexical items are selected and incrementally merged into larger structures until a complete syntactic object is formed.

However, electrophysiological evidence, particularly ELAN and P600 signatures, compels a substantial revision of this perspective. During real-time comprehension, the brain does not passively await structural completion. It actively constructs anticipatory models of incoming linguistic structure.

The processing architecture can be characterized as follows:

anticipatory generation of syntactic structure

continuous Bayesian-style updating under incoming evidence

rapid reanalysis upon violation of structural expectations (P600 effects)

This yields a fundamental distinction that must be stated with theoretical precision:

$$\text{Formal Derivational Space (Possibility)}\mathrm{\ne }\text{Neurocognitive Processing Trajectory (Prediction under uncertainty)}$$

Formal syntax characterizes what structures are generable. Neurobiology characterizes how such structures are probabilistically traversed in real time.

The two are not in conflict, but neither are they equivalent.

Modularity Reconsidered

Early minimalist interpretations often inherited a strong modular view of syntax: a “narrow syntax” insulated from semantic, pragmatic, and encyclopedic systems until interface transfer.

Contemporary neurocognitive evidence renders such insulation increasingly implausible.

Language processing networks exhibit rapid and temporally fine-grained interaction with systems supporting:

pragmatic inference and contextual modulation

discourse-level coherence tracking

Theory-of-Mind computations (notably right temporoparietal junction networks)

Crucially, these interactions occur within the same temporal window as syntactic parsing itself, not as post-derivational adjustments.

A more empirically adequate formulation is:

Syntax is not encapsulated; it is computationally specialized within a dynamically interactive architecture.

Modularity, in its modern neuroscientific interpretation, is best understood as gradient specialization constrained by network topology, rather than absolute informational isolation.

1.3 Constraints on Variation: From Computational Uniformity to Lexical Differentiation

If the neurocognitive substrate supporting language is broadly conserved across Homo sapiens, then grammatical variation cannot plausibly be attributed to differences in core computational machinery. Instead, it must arise within constrained representational loci.

This perspective aligns broadly with the Borer–Chomsky Conjecture, which situates cross-linguistic variation primarily in the lexicon and functional feature inventories rather than in syntactic rule systems.

On this view, the architecture of language can be schematized as follows:

Component	Theoretical Status	Neurocognitive Interpretation
Combinatorial system (Merge-like operations)	Invariant across languages	Supported by conserved fronto-temporal circuitry
Syntactic variation	Derived from feature specifications	Differences in lexical retrieval and feature valuation dynamics
Surface structural diversity	Interface-dependent realization	Variable mapping under performance constraints

Yet a critical caveat is essential: neuroscience does not directly verify this decomposition. It provides constraints on architectural plausibility, not a direct adjudication of formal linguistic hypotheses.

Constraint Without Reduction

The interface between syntactic theory and neuroscience is best understood not as a mapping of identities, but as a system of mutual constraint across incompatible representational domains.

Formal syntax specifies the space of possible hierarchical computations

Neuroscience constrains how such computations can be realized under biological pressure

Cognitive architecture mediates between formal possibility and real-time feasibility through predictive, resource-limited processing

From this perspective, grammar is not “located” in neural tissue as a structural replica. It is approximated under constraint, instantiated through distributed systems that simulate formal combinatorics within the limits of biological implementation.

Excellence Perspective

The integration of Minimalist syntax with cognitive neuroscience does not diminish the autonomy of formal theory. On the contrary, it refines its epistemic status.

The strongest defensible position today is neither eliminativist nor reductive, but a form of constraint-based realism:

Syntactic theory remains formally autonomous as a theory of structure; neuroscience determines the boundaries of its biological realizability.

In this light, the “wetware of grammar” is not a neural transcription of syntactic theory. It is the biophysical medium that makes hierarchical computation possible at all, while inevitably reshaping it under the pressures of time, noise, and connectivity.

What emerges is not the end of formal grammar but its maturation under biological constraint.

Bibliography

Binder, J. R., Frost, J. A., Hammeke, T. A., Bellgowan, P. S., Springer, J. A., Kaufman, J. N., & Possing, E. T. (2000). Human temporal lobe activation by speech and nonspeech sounds. Cerebral cortex, 10(5), 512-528.

Bornkessel-Schlesewsky, I., & Schlesewsky, M. (2013). Reconciling time, space and function: A new dorsal–ventral stream model of sentence comprehension. Brain and language, 125(1), 60-76.

Coolen T, Wens V, Vander Ghinst M, Mary A, Bourguignon M, Naeije G, Peigneux P, Sadeghi N, Goldman S and De Tiège X (2020) Frequency-Dependent Intrinsic Electrophysiological Functional Architecture of the Human Verbal Language Network. Front. Integr. Neurosci. 14:27. doi: 10.3389/fnint.2020.00027

Coopmans, C. W., De Hoop, H., Tezcan, F., Hagoort, P., & Martin, A. E. (2025). Language-specific neural dynamics extend syntax into the time domain. PLoS biology, 23(1), e3002968.

Fitch, W. T., & Martins, M. D. (2014). Hierarchical processing in music, language, and action: Lashley revisited. Annals of the New York Academy of Sciences, 1316(1), 87-104.

Fedorenko, E., & Thompson-Schill, S. L. (2014). Reworking the language network. Trends in cognitive sciences, 18(3), 120-126.

Friederici, A. D., Chomsky, N., Berwick, R. C., Moro, A., & Bolhuis, J. J. (2017). Language, mind and brain. Nature Human Behaviour, 1(10), 713-722.

Friedrich, R., & Friederici, A. D. (2009). Mathematical logic in the human brain: syntax. PLoS One, 4(5), e5599.

Friederici, A. D. (2011). The brain basis of language processing: from structure to function. Physiological Reviews.

Friederici, A. D. (2017). Evolution of the neural language network. Psychonomic bulletin & review, 24(1), 41-47.

Friederici, A. D. (2017). Neurobiology of syntax as the core of human language. Biolinguistics, 11, 325-338.

Friederici A. D. (2011). The brain basis of language processing: from structure to function. Physiological Reviews, 91(4), 1357–1392. https://doi.org/10.1152/physrev.00006.2011

Giraud, A. L., & Poeppel, D. (2012). Cortical oscillations and speech processing: emerging computational principles and operations. Nature Neuroscience, 15(4), 511-517.

Grodzinsky, Y., & Friederici, A. D. (2006). Neuroimaging of syntax and syntactic processing. Current opinion in neurobiology, 16(2), 240-246.

Grodzinsky, Y., Pieperhoff, P., & Thompson, C. (2021). Stable brain loci for the processing of complex syntax: A review of the current neuroimaging evidence. Cortex, 142, 252-271.

Gutstein, D. B., Iorga, M., Stocks, J., Gill, N., Sleeman, J., Gefen, T. D., Los, M., Nelson, C., Geula, C., Weintraub, S., Parrish, T., Mesulam, M. M., & Barbieri, E. (2026). Identifying neuropathologic disease in primary progressive aphasia using narrative speech. Alzheimer's & dementia : the journal of the Alzheimer's Association, 22(3), e71294. https://doi.org/10.1002/alz.71294

Hagoort, P., Brown, C. M., & Osterhout, L. (1999). The neurocognition of syntactic processing. In The Neurocognition of Language (pp. 273-317). Oxford University Press.

Hagoort P. (2013). MUC (Memory, Unification, Control) and beyond. Frontiers in Psychology, 4, 416. https://doi.org/10.3389/fpsyg.2013.00416

Hagoort, P. (2009). Reflections on the neurobiology of syntax. In Biological Foundations and Origin of Syntax (pp. 279-296). MIT Press.

Hagoort, P. (2017). The core and beyond in the language-ready brain. Neuroscience & Biobehavioral Reviews, 81, 194-204.

Heard, M., & Lee, Y. S. (2020). Shared neural resources of rhythm and syntax: An ALE meta-analysis. Neuropsychologia, 137, 107284.

Hickok, G., & Poeppel, D. (2007). The cortical organization of speech processing. Nature Reviews Neuroscience, 8(5), 393-402.

Hickok, G., & Poeppel, D. (2004). Dorsal and ventral streams: a framework for understanding aspects of the functional anatomy of language. Cognition, 92(1-2), 67–99. https://doi.org/10.1016/j.cognition.2003.10.011

Hu, J., Gauthier, J., Qian, P., Wilcox, E., & Levy, R. (2020, July). A systematic assessment of syntactic generalization in neural language models. In Proceedings of the 58th annual meeting of the Association for Computational Linguistics (pp. 1725-1744).

Keitel, A., & Gross, J. (2016). Individual human brain areas can be identified from their characteristic spectral activation fingerprints. PLoS biology, 14(6), e1002498.

Kemmerer, D. (2022). Cognitive neuroscience of language. Routledge.

Kuperberg, G. R. (2007). Neural mechanisms of language comprehension: Challenges to syntax. Brain Research, 1146, 23-49.

Lakatos, P., Karmos, G., Mehta, A. D., Ulbert, I., & Schroeder, C. E. (2008). Entrainment of neuronal oscillations as a mechanism of attentional selection. science, 320(5872), 110-113.

Maran, M., Friederici, A. D., & Zaccarella, E. (2022). Syntax through the looking glass: A review on two-word linguistic processing across behavioral, neuroimaging and neurostimulation studies. Neuroscience & Biobehavioral Reviews, 142, 104881.

Matchin, W., & Hickok, G. (2020). The Cortical Organization of Syntax. Cerebral cortex (New York, N.Y. : 1991), 30(3), 1481–1498. https://doi.org/10.1093/cercor/bhz180

Moro, A., Tettamanti, M., Perani, D., Donati, C., Cappa, S. F., & Fazio, F. (2001). Syntax and the brain: disentangling grammar by selective anomalies. Neuroimage, 13(1), 110-118.

Murphy, E. (2024). ROSE: A neurocomputational architecture for syntax. Journal of Neurolinguistics, 70, 101180.

Osterhout, L., Kim, A., & Kuperberg, G. R. (2012). The neurobiology of sentence comprehension. The Cambridge handbook of psycholinguistics, 365-389.

Patel, A. D. (2003). Language, music, syntax and the brain. Nature Neuroscience, 6(7), 674-681.

Petersson, K. M., & Hagoort, P. (2012). The neurobiology of syntax: beyond string sets. Philosophical Transactions of the Royal Society B: Biological Sciences, 367(1598), 1971-1983.

Petersson, K. M., Folia, V., & Hagoort, P. (2012). What artificial grammar learning reveals about the neurobiology of syntax. Brain and language, 120(2), 83-95.

Poeppel, D., & Assaneo, M. F. (2020). Speech rhythms and their neural foundations. Nature Reviews Neuroscience, 21(6), 322-334.

Zaccarella, E., & Friederici, A. D. (2017). The neurobiological nature of syntactic hierarchies. Neuroscience and biobehavioral reviews, 81(Pt B), 205–212. https://doi.org/10.1016/j.neubiorev.2016.07.038

Zaccarella, E., Meyer, L., Makuuchi, M., & Friederici, A. D. (2017). Building by syntax: The neural basis of minimal linguistic structures. Cerebral Cortex, 27(1), 411-421.