Adverbs, Events, and the Immanence of Meaning in the Brain

Unpacking Composition: Adverbs, Events, and the Immanence of Meaning in the Brain

Language is often described as if it were a structure built upon structure, syntax layered over semantics, rules projected onto meanings, and meanings stabilized into forms. Yet beneath this architectural imagination lies a more unsettling possibility: that composition is not the execution of structure but the emergence of relations under constraint. This process precedes and exceeds the grammars we use to describe it.

What neuroscience of language increasingly suggests is not the neural instantiation of linguistic theory but the dissolution of its presumed primacy.

1. Against the Illusion of Structural Correspondence

To ask whether the brain implements syntax is already to assume that syntax is the correct scale of description. But neural data resist this assumption. What is observed is not the clean unfolding of rule-based composition but fluctuating regimes of integration, some rapid, some delayed, some sensitive to conceptual density rather than structural configuration.

A neural signal that correlates with a syntactic operation does not confirm the existence of that operation in the brain. It only reveals that, under certain conditions, linguistic theory has carved a space that the brain occasionally passes through.

But passing through is not the same as inhabiting.

Thus, the methodological gesture of neurolinguistics shifts from mapping rules onto regions toward identifying the conditions under which meaning becomes combinable at all.

2. The Left Anterior Temporal Lobe as a Field of Conceptual Convergence

The left anterior temporal lobe (LATL) does not behave like a syntactic engine. It behaves more like a threshold of intelligibility, a region where fragments of conceptual content briefly stabilize into compositional coherence.

Yet this coherence is not governed by syntax. It is governed by transparency: the degree to which concepts can enter into immediate relation without semantic displacement.

A “straightforward” composition is not structurally simpler; it is ontologically lighter. It requires no detour through reinterpretation, no folding of reference, no redistribution of roles across an event structure. It simply combines.

But even “combination” is already too structured a word. What occurs is closer to a local intensification of compatibility between concepts, a momentary alignment within a semantic field.

The LATL, then, is not a place where syntax is computed. It is where meaning becomes briefly coherent enough to persist.

3. Events, Adverbs, and the Fragmentation of Predicate Space

Adverbs reveal something crucial: predicates are not uniform objects. They are stratified fields.

A manner adverb like slowly does not merely modify a verb; it folds into the temporal texture of the event itself. But participant-oriented adverbs, reluctantly and vividly, operate differently. They do not attach to events directly. They emerge from other ontological domains: agency, perception, result, and aftermath.

Thus, what appears syntactically as a uniform class (“adverbs”) is in fact a heterogeneous topology of interpretive routes.

Each adverb indexes a different mode of access into the event:

• Some enter through motion (manner)
• Some through subjectivity (agent state)
• Some through residue (result)

Composition, therefore, is not the combination of words. It is the negotiation between different ways of accessing an event.

The brain does not assemble structure; it resolves routes into a temporarily stable interpretation.

4. Composition Without Constituency

One of the most radical implications of recent findings is that composition does not require constituency in the classical sense.

Words do not need to form a syntactic unit to be integrated. What is required is not structural adjacency but semantic availability: the presence of roles, participants, and conceptual affordances that enable integration.

This suggests a deeper reversal: syntax does not generate compositionality; rather, compositionality selects among possible syntactic realizations.

The LATL does not wait for trees to complete. It operates in a pre-constituent space, where meaning is already negotiating its own combinability before formal structure stabilizes.

5. The Brain as a Selector of Straightforwardness

A recurring pattern emerges: the brain prefers what is directly composable.

But “directness” is not simplicity in grammatical terms. It is a property of semantic friction. Some combinations slide effortlessly into each other; others resist, requiring reinterpretation, contextual repair, or structural detour.

The LATL responds most strongly when friction is minimal, when concepts meet without transformation.

In this sense, it does not compute meaning. It filters complexity, allowing only certain configurations of conceptual alignment to crystallize early.

This is not computation in the classical sense. It is a selection under a temporal constraint.

6. From Representation to Immanence

The deeper shift implied by this research is philosophical rather than merely empirical.

Linguistic theory has long operated within a representational paradigm: meaning as structure, structure as rule, rule as computation. But neural evidence suggests a different ontology.

Meaning is not represented before composition. It is produced within the act of integration itself.

This aligns more closely with an immanent view of cognition: not a system that manipulates symbols according to pre-given rules, but a system that continuously generates local coherence from distributed constraints.

Composition, then, is not a function over representations. It is the emergence of representability.

7. The Event of Language

If we follow this trajectory seriously, language ceases to be a structure we possess and becomes an event we undergo.

Adverbs, verbs, and nouns are not stable categories assembled by a grammatical engine. They are vectors within a dynamic field of interpretation, constantly renegotiated as neural systems attempt to stabilize meaning under time pressure.

The brain does not “apply” syntax to thought. It navigates a continuous space of possible compositions, selecting those moments where meaning becomes momentarily stable enough to be experienced.

Language, in this sense, is not a system of rules.

It is a field of constrained becoming.

Toward a Post-Syntactic View of Composition

What emerges from this line of inquiry is not the rejection of linguistics, but its displacement into a different register.

Syntax remains useful, but not foundational. It describes patterns that emerge after a deeper process of semantic alignment has already taken place.

The left anterior temporal lobe, in this view, is not a syntactic processor but a site of conceptual crystallization. It marks the moment when meaning ceases to be diffuse and becomes temporarily graspable.

Yet this graspability is always conditional, always fragile, always dependent on the ease with which concepts can enter into relation.

Composition, therefore, is not the assembly of linguistic units.

It is the momentary stabilization of an otherwise fluid semantic field.

And in that instability, language reveals its most fundamental property: not structure, but becoming.

Inspired by: Songhee Kim & Liina Pylkkänen | Leipzig Lectures on Language — Combinatorics 2021