Haueis and Colaço compellingly argue for integrating metabolic considerations into cognitive modeling. We suggest that current discussions suffer from a scale mismatch: metabolic constraints are typically framed at global or microscopic levels, whereas cognitive models may operate at a systems level. We propose mesoscale metabolic properties of functional neural networks as a more promising interface for constraining cognitive models.

Explicitly modeling metabolic constraints remains rare in cognitive neuroscience due to methodological, empirical, and conceptual challenges. Although metabolic limits already shape the biological data that many models are fitted to, direct measurements of energetic costs remain scarce and difficult to relate to model variables. We argue that recent experimental and computational advances may soon enable more explicit integration of metabolism into cognitive models.

We build upon Haueis and Colaço’s account in two ways. First, we argue that the backgrounding of metabolism was a pragmatic response to a data granularity problem faced by traditional cognitive modelers. Then, we suggest that equivalence demonstrations offer a promising pathway for incrementally introducing metabolic considerations into extant modeling practice.

Haueis and Colaço point out that mammalian brains are much more energy efficient than AI supercomputers because the former use orders of magnitude less energy. I argue this issue is more nuanced, and the “efficiency” is scale dependent. Contrary to popular belief, I show that the human brain uses much more energy per operation than the most powerful AI supercomputers.

We argue that under minimal thermodynamic and control-theoretic assumptions, metabolic constraints imply generative cognition. We formalize cognition as a resource-bounded controller of probability flows, where metabolic budgets define path-space divergence constraints. This reframes inference as a physically instantiated process governed by metabolic regulation, implying that progress in the cognitive sciences requires grounding cognitive models in nonequilibrium thermodynamics and dynamical systems.

Haueis and Colaço argue that motivational/computational models of effort should incorporate metabolic factors. But metabolic and motivational accounts predict the same behavioral patterns, making their contributions hard to distinguish empirically. Moreover, philosophy provides independent reasons favoring motivational accounts: effort is unified across domains and normatively significant in ways metabolic factors cannot capture.

Multiple realizability (MR) is not necessarily unclear nor does it purely operate at the computational level. To understand potential relationships between MR and other constraints, such as metabolic, we formalize possible meanings of function in cognitive science. We build on these to formalize MR, thus resolving its apparent vagaries. Importantly, MR formalisms meaningfully guide and constrain theory building.

Integrating metabolism into cognitive science is essential. Beyond a mere constraint, metabolism may be constitutive of cognition itself. This requires expanding metabolism beyond neuronal energy use to include glial contributions, compartmentalized and dynamic processes, and broader brain–body interactions (e.g., interoception). Capturing this complexity will require more refined, integrative models, potentially leveraging artificial intelligence to unify multiscale biological data.

H&C leave unanswered whether metabolic considerations (MCs) are merely one neglected type of biological detail that constrains cognitive models or a distinctive kind uniquely poised to do so. Their examples give little reason to believe the latter. If the former, a methodology built around MCs specifically – rather than neuroscientific facts generally – is unjustified unless embedded within resource-rational analysis.

Haueis and Colaço argue cognition must be modeled with metabolic constraints, challenging more abstract, computational views. This commentary supports their approach but identifies a key omission: multiscale dynamics. Cognition is fundamentally dynamic, emerging across coordinated timescales where metabolism is embedded, not merely constraining. Incorporating multiscale dynamics would strengthen their framework by aligning models with biologically realistic, time-dependent processes.

Constraining cognitive models based on the metabolic realities of the brain is a laudable goal. However, imposing strong constraints on nascent theories risks prematurely discarding potentially productive models. Furthermore, despite models of cognitive fatigue being an intuitive test case for Haueis and Colaço’s approach, the strictly medium-independent nature of non-resource models (e.g., opportunity cost models) highlights some limits to the proposed approach.

I offer further support for a metabolic turn in cognitive science and raise two questions for metabolic considerations in modeling: I first invite the authors to consider broader tensions between computational and metabolic models; I then discuss possible connections between the relevance attributed to metabolic considerations for modeling and ontological background commitments about cognitive systems.

Haueis and Colaço argue that metabolism is a generative principle for cognitive modelling. We extend this by bridging the “scaling gap” through functional PET (fPET) and network neuroscience. By measuring glucodynamics and the metabolic network of the brain, we can model how glucose metabolism enables network efficiency, temporal flexibility, information transfer, and ultimately the cognitive capacity of the human brain.

I agree with Haueis and Colaço that integrating metabolic considerations into cognitive modeling is important, given evolution by natural selection. However, I question their exclusive focus on neurocentric models, arguing that ecological models of neural function in embodied, situated contexts should be added to the space of possible models and that extraneural metabolic factors are important to consider as well.

Haueis and Colaço avoid referencing “neural activity,” instead of speaking “neural activities”. This subtlety speaks to a strength of their framework: Their metabolically oriented account helps bypass abstractions and stimulus–response assumptions endemic to traditional neuroscience paradigms. We highlight two advantages of this pivot to “neural activities,” and how they pave the way for a unified, metabolically grounded, systems neuroscience.

The human brain makes up 2% of body mass but consumes closer to 20% of the body’s energy. Nonetheless, it is more energy-efficient than most computers. Although these facts are well known, models of cognitive capacities and neural systems rarely address metabolic factors. Here, we argue that metabolic considerations should be integrated into cognitive models. Metabolic considerations can be used to evaluate existing and generate novel models. Our account of these considerations offers fresh insights into debates over cognitive models of mental effort, philosophical views of multiple realizability and medium independence, and the comparison between biological and artificial computational systems.

In the Target Article, we argued that “egalitarian societies” are real, but that outcome-oriented, society-level taxonomies have caused confusion and disagreement. After showing that inequalities exist in even the most egalitarian societies, we called for greater research emphasis on processes underlying egalitarianism under a range of ecological/economic conditions. Thirty commentaries discuss, extend or, in several cases, query elements of our thesis. In this response, we address each commentary in turn, focussing on four unifying themes: (1) gendered inequalities; (2) outcome vs process; (3) cultural norms and enculturation; and (4) the psychology of egalitarianism. We finish by reiterating our call to shift focus beyond outcome-oriented society-level taxonomies. In synthesising insights from a range of commentaries, we show that the motives and processes underlying egalitarianism in smaller-scale societies have far-reaching relevance to the study of political psychology and political organisation in all human societies, including contemporary nation-states.

This article argues that Duncan Stibbard–Hawkes and Christopher von Rueden’s article, “Egalitarianism is Not Equality,” makes an important point about egalitarianism, but it misidentifies its target. The problem is not outcome-versus-process definitions, as the authors suggest but perfectionist-versus-substantive definitions. The solution they propose qualifies as a good substantive definition.

The target article by Stibbard Hawkes and von Rueden (SHVR) provides useful terminological housekeeping, but its individualistic account of egalitarianism in small-scale societies risks overlooking important group-level functions. We argue that egalitarian norms are not merely by-products of individual strategies but actively sustain cooperation, coordination, and collective decision-making, enhancing group stability and collective intelligence. A full account of egalitarianism must integrate both individual-level processes and emergent group-level outcomes. Egalitarianism not only constrains inequality but also underwrites the cooperative capacities of small-scale societies.

Stibbard-Hawkes and von Rueden highlight consensus-based decision-making as a key process underlying egalitarianism. I argue that this terminology remains unnecessarily imprecise and can be potentially misleading. In line with the authors’ own argument, I underline the need to focus on deliberation instead and provide several reasons why this could further both the study of egalitarianism and that of deliberation.