In this ambitious treatise on human life history variation, the authors first provide a useful review of empirical tests of life history theory predictions. The inconsistencies that they uncover motivate, as a resolution to these paradoxes, a verbal two-tier model wherein the first tier captures environmental influences affecting resource availability, and the second tier identifies the impacts of external mortality risks. The scope of analysis is admirable, given that different life history traits are usually analysed separately, despite the continuous calls for a more holistic treatment (Lawson & Borgerhoff Mulder, Reference Lawson and Borgerhoff Mulder2016; Frankenhuis & Nettle, Reference Frankenhuis and Nettle2020). This is an impressive and useful piece of scholarship, and a good example of how new theoretical frameworks can be built up from countervailing empirical observations. We do, however, have concerns – first with the added value of a hierarchically structured model and second with the absence of formal modelling.

There are many ways to divide up the triggers affecting life history decisions. They can be physiologically versus culturally mediated (Colleran, Reference Colleran2024). They can be based on current conditions versus expected future conditions (which can be complicated by the covariation between the two (Hayward, Rigby & Lummaa, Reference Hayward, Rigby and Lummaa2016)). They may be based on information about future environments versus current conditions that enduringly alter the somatic state of an individual (Nettle & Bateson, Reference Nettle and Bateson2015), or on internal versus external predicted adaptive responses (Nettle, Frankenhuis, & Rickard, Reference Nettle, Frankenhuis and Rickard2013). Such triggers may provoke non-adaptive responses, as has been seen in wild female baboons that show no evidence of accelerated reproductive schedules as an adaptive response to early-life adversity (Weibel et al., Reference Weibel, Tung, Alberts and Archie2020). Quite why and how the two-tier framework proposed here improves on these earlier categorisations of predictors is unclear to us, leading us directly to our second concern.

Lamenting the lack of a formal model is a somewhat cheap objection, especially without offering one ourselves, but the purely verbal nature of a two-tiered hierarchical model for explaining complex life history trade-offs inevitably raises questions about what would happen under formal analysis – what different predictions would this model make when compared to previous approaches to the analysis of the various life history trade-offs that an organism faces? Formally specifying complex developmental processes (e.g., emergence of sensitive periods (Panchanathan & Frankenhuis, Reference Panchanathan and Frankenhuis2016)) can help clarify the relationship between different conditions, triggers, and cues as causes of life history variability. It also allows us to think clearly about evolution in populations rather than isolated individuals, as demonstrated by de Vries, Galipaud, and Kokko (Reference de Vries, Galipaud and Kokko2023), who show that when mortality risk declines (and populations start to grow at a faster rate), there will be pressure to reproduce earlier (counter to standard theory). In short, the advantages and disadvantages of distinct approaches to studying life history variation do need formal attention, and posing a new framework without doing so is risky (Jones & Promislow, Reference Jones and Promislow2021).

So, without formal modelling to help sort things out, why are we somewhat sceptical of the hierarchical relationship being more informative than a model based on interactions that would allow greater flexibility in the relationship between these two variables? Using age at first birth as an example, the few empirical studies that analyse interactions between resource availability and environmental uncertainty (Low et al., Reference Low, Hazel, Parker and Welch2008; Uggla & Mace, Reference Uggla and Mace2016; Tan et al.; Reference Tan, Li, Meltzer, Chin, Tan, Lim, Neuberg and van Vugt2022) show inconsistent patterns with respect to the direction of the interactions. They also fail to provide any evidence that one predictor has any priority, either in time or in importance, over the other. In short, we agree with the authors on the need to account for energetic resources and extrinsic mortality risk together to better understand the reproductive strategies of women, but we are not convinced that construing this relationship as hierarchical is helpful.

The authors rightly highlight the extreme flexibility in human reproductive strategies (see Burger, Lee & Sear, Reference Burger, Lee and Sear2024 for a deeper review). We propose that a more explicit focus on understanding the impact of environmental uncertainty at different stages of the human life cycle can help explain the decisions humans take. Such approaches require longitudinal data and are data-intensive (e.g., Minocher, Borgerhoff Mulder, & Ross, Reference Minocher, Borgerhoff Mulder and Ross2024) but are open to agent-based modelling (e.g., Weitzel et al., Reference Weitzel, Wilson, Spake, Schaffnit, Lynch, Sear and Sosis2024). A step forward would be to understand how the effects of resource availability and external mortality risk on reproductive maturation and decision-making differ across ages. Theoretical work, including the two-tier model, focuses on the differences between early and current life conditions (Lummaa & Clutton-Brock, Reference Lummaa and Clutton-Brock2002; Nettle & Bateson, Reference Nettle and Bateson2015). However, longitudinal empirical analyses usually show that theoretical assumptions do not hold when finer granularity regarding development is accounted for (see Newmeyer et al., Reference Newmyer, McAllister, Alam and Shenk2025 for age-specific mortality risk and fertility behaviour). Hence, the gap in understanding how resource availability and external risks relate to the reproductive strategies requires theoretical work that aligns with the developmental granularity observed in longitudinal empirical studies.

In conclusion, the authors aim to move forward our understanding of the relationship between environmental uncertainty and age at sexual maturity, first birth, and total number of children, suggesting a two-tier verbal model to do so. However, their framework currently rests on the assumption that energetic resources and external mortality risk have a hierarchical relationship. Probing this relationship more carefully, ideally with formal modelling and through a developmental lens, should be the next step in examining how life history traits evolve and interact through life.