Graphs are focal hypergraphs: strict containment in higher-order interaction dynamics

Elkaïoum M. Moutuou · 2026

Read Graphs are focal hypergraphs: strict containment in higher-order interaction dynamics by Elkaïoum M. Moutuou (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-04215.txt.

Introduction

Elkaïoum M. Moutuou presents graphs as focal contact systems in higher-order motion. In this view, hypergraphs extend graph neighbourhoods by permitting arbitrary contact domains with no distinguished focal node. This view also clarifies how graph and hypergraph dynamic models relate. Although sufficiently general graph node functions can reproduce higher-order contact motion, that dynamic match does not entail a match of core contact forms.

Catalog record

CREATORElkaïoum M. Moutuou
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE40996
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.04215

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Graphs are focal hypergraphs: strict containment in higher-order interaction dynamics Elkaïoum M. Moutuou Abstract We introduce a taxonomy of interaction types and show that graphs are focal hypergraphs: every graph is canonically a focal hypergraph via its closed neighbourhood structure, and every graph dynamical model is a special case of the general hypergraph dynamical model. The central distinction is between focal interactions, in which the interaction domain is defined relative to a designated reference node, and non-focal interactions, in which all participants stand in equivalent structural relationship. Closed graph neighbourhoods are precisely focal hyperedges, so hyperedges generalise graph neighbourhoods by removing the focal constraint. This yields a strict three-level hierarchy: graph models ⊊ \subsetneq focal hypergraph models ⊊ \subsetneq general hypergraph models. Moreover, graph models do encode genuinely higher-order (many-body) interactions, in the sense that each node’s update function may depend jointly on all members of its closed neighbourhood, but they remain a strict special case of the hypergraph dynamical model, not equivalent to it. We further show that universal encodings such as bipartite factor graphs are neutral with respect to this hierarchy, and that the symmetry condition of the hypergraph dynamical model — often treated as an additional constraint relative to the graph model — is in fact the dynamical definition of a non-focal interaction. The taxonomy is grounded in concrete phenomena from physics, biology, ecology, and social systems, and yields a principle of representational alignment: the choice between graph and hypergraph models should be governed by the type of interaction, not by a blanket preference for one formalism over the other. 1 Introduction What does it mean for a group of entities to interact? In physics, biology, and the social sciences alike, the answer shapes the mathematical language one reaches for. When interactions are binary, graphs are the natural model. When interactions are intrinsically multi-body, the question of appropriate representation becomes more subtle. A growing literature on higher-order networks has advanced hypergraph-based models as the necessary language for interactions that cannot be reduced to pairwise relationships [ 1 , 2 , 3 , 4 , 5 ] . The central claim of this literature is often stated informally, with the distinction between graph and hypergraph models left implicit. Our aim here is to make it precise. The key, we argue, is to keep two levels of description carefully apart: • the structure of a network, which records which entities are connected; and • the dynamics that runs on it, which governs how each entity’s state evolves. A bridge between two landmasses exists as a structural fact independently of any traffic that crosses it. But the flow through a road junction is a function of all roads meeting there simultaneously — a different, and richer, unit of description. This distinction, spelled out formally, determines the correct way to compare graph and hypergraph models, and yields results that differ from accounts that conflate the two levels. The paper makes four main contributions. First, we introduce a taxonomy of interaction types that distinguishes structural relations, focal dynamical interactions (in which one node’s state responds to its neighbourhood), and non-focal symmetric interactions (in which a group of nodes participates equivalently with no designated reference). Second, we show that every graph is canonically a focal hypergraph ; that is, when a dynamical model runs on a graph, the natural unit of interaction is not the edge but the node’s closed neighbourhood, and graph neighbourhoods are precisely hyperedges equipped with a designated focal node . Third, and as a direct consequence, we establish that graph dynamical models are a strict special case of the general hypergraph dynamical model; they do already encode genuinely higher-order (many-body) coupling functions, but they are constrained to a geometry that non-focal hyperedge models are not. Fourth, we show that the symmetry condition of the hypergraph dynamical model, often treated as an additional restriction, is in fact the formal definition of the absence of a focal node, which dissolves an apparent conflict in the recent work by Peixoto et al. [ 6 ] . The practical upshot is a principle of representational alignment; namely, the choice between graph and hypergraph models should be governed by the type of interaction one is modelling, not by a blanket preference for one formalism over the other. Most interactions studied in the empirical higher-order networks literature are focal, and graph models are fully adequate for them. Genuinely non-focal interactions (irreducible three-body forces, higher-order ecological coexistence, correlated social equilibria, etc.) require hyperedge-based models. The paper is organised as follows. Section 2 introduces the taxonomy of interaction types. Sections 3 and 4 develop the formal definitions and examples for focal and non-focal hypergraphs. Section 5 analyses the two natural embeddings of graphs into hypergraph theory and their implications. Section 6 shows that every graph is canonically a focal hypergraph. Sections 7 and 8 develop the graph and hypergraph dynamical models and the strict containment hierarchy. Section 9 addresses universal encodings. Section 10 concludes. 2 A Taxonomy of interaction types Before developing the formal concepts, it is useful to distinguish three types of interaction that arise in complex systems and that make different demands on the representational language. The taxonomy is not a partition of phenomena (real systems typically involve all three types simultaneously) but a partition of the questions one can ask. 1. Structural–Topological (ST) interaction. A binary combinatorial relation between two entities, independent of any dynamical state. A bridge, a synaptic connection, a territorial boundary. These are pairwise by nature and are faithfully represented by graph edges. The existence of the bridge is a structural fact; it does not depend on whether any flow is present. 2. Focal–Dynamical (FD) interaction. A process in which the state of a node is updated as a function of the states of all nodes in its neighbourhood, with that node as the designated reference. The entire neighbourhood participates simultaneously, so the coupling is genuinely many-body, but the interaction domain is always defined relative to the focal node. The flow through a road junction depends on all connecting roads jointly; the firing of a neuron depends jointly on all its pre-synaptic inputs; a person’s opinion shifts as a function of their entire social neighbourhood. Graph dynamical models encode FD interactions exactly. 3. Non-Focal–Symmetric (NFS) interaction. A group interaction in which no member plays the role of reference node. All participants stand in equivalent structural relationship; the interaction belongs to the group jointly, not to any individual’s neighbourhood. A committee reaching a decision by unanimous consent, a reaction requiring three co-present reactants with no designated catalyst, a synergistic neural ensemble whose collective firing encodes a stimulus. NFS interactions require non-focal hyperedges for their faithful representation. The distinction between FD and NFS turns on a single question: does the interaction have a natural focal node? Graph neighbourhoods are the right language when the answer is yes; non-focal interaction models are required when it is no. Remark 2.1 . It is important to observe that the majority of social interactions appearing in the empirical higher-order networks literature can be interpreted as focal (Type 2). Social influence and opinion dynamics fit this type exactly: person i updates their belief as a function of their

Read the complete text