I argue that the two-tier model represents a clear step forward, resolving debates and disentangling complexity in our understanding of human life history strategies. I relate the model to findings from my previous work, as well as to embodied capital and market integration, providing relevant schematics. I then discuss the timing of exposure to mortality cues and the age of the deceased, discussing what the two-tier model does and does not add to these debates.

The relationship between local extrinsic mortality rates and the timing of life-history milestones can be better understood when examined through food environments rather than by population “types,” like small-scale societies or high-income countries. By mapping observed life-history variation onto food environments, which mediate the fundamental relationship between mortality and fertility, the explanatory value of a 2-tiered model becomes less compelling.

Ellis, Reid, and Kramer advance a two-tier model of extrinsic mortality and life history strategies, emphasizing that extrinsic mortality can promote higher reproduction. The article implies this second tier operates via conscious, volitional processes. I present evidence that extrinsic mortality events are not merely environmental cues but also bereavement-inducing events that can define histories principally by disrupting, not facilitating, life strategies.

The two-tiered model proposed by Ellis et al. offers a compelling framework for understanding how extrinsic mortality (EM) sources shape life history strategies, including fertility. This commentary evaluates this model in light of modern fertility trends, identifies gaps in its current formulation, and proposes modifications to enhance its explanatory power in modern contexts.

In this commentary, I highlight the challenge of incorporating the harshness–unpredictability framework into the two-tiered model. While the effects of energetic stress (first tier), which are mostly mediated through biological pathways, can be decomposed into harsh (high mean) and unpredictable (high variance) components, ambient cues to EM (second tier) blur this distinction because they often share the same psychological mediators.

Life history strategies are linked to reproduction. Fast life history strategies place a particular burden on women biologically, socially, and politically. From the point of view of women’s rights and gender equality, slower life history strategies are the more optimal solution. In the article, I show how feminism has an important role to play in slowing down life history strategies.

While I applaud the effort that went into their article, I believe Ellis and colleagues’ two-tiered theory of life history has three problems: (1) relying on a straw man argument; (2) limitations in their data; and (3) alternate theoretical explanations for their correlations. Failure to attend to these issues resulted in a more confident presentation of their conclusions than is warranted.

The two-tiered life history model remains inadequate in addressing how social factors (particularly cultural norms and religious beliefs) interact with biological determinants to shape life history strategies. This limitation significantly constrains the model’s explanatory power when applied to certain historical demographic phenomena. We recommend introducing a third explanatory framework (the cultural-institutional tier) to augment the two-tiered model.

The two-tiered life history (LH) model proposes that different sources of extrinsic mortality (EM) have opposing effects that regulate development toward both slower and faster LH traits. Although the 53 commentators generally endorsed the two-tiered model and empirical conclusions of the target article, the devil is in the details. Some commentators challenged model assumptions (e.g., the mechanistic basis of the two-tiered model; whether the model is genetically confounded; the relative importance of child versus adult mortality). Other commentators proposed extensions/modifications of model concepts and hypotheses (e.g., incorporating density-dependent regulation; use of formal models to generate and test hypotheses; connection to the internal predictive adaptive response [PAR] model). In this reply, we review and address these challenges and proposed extensions/modifications. We hope that this iterative process advances our understanding of the complexity of EM, its opposing tiers, and their dualistic effects–both hierarchical and countervailing–on variation in human life histories.

Recent research on peripheral vision has led to a paradigm-shifting conclusion: vision science as a field must rethink visual attention. This article reviews the evidence for a crisis in attention science and examines supposedly attentional phenomena to ask which point to additional capacity limits. Based on the resulting list of critical phenomena, and what they have in common, I propose an alternative way to think about capacity limits and the underlying mechanisms.

How do our goals continually impact perceptual processing? The answer could arise from a computational specification of perception in terms of visual tasks, or perhaps several mechanisms operating over specific contexts. Here, we suggest an alternative: adaptive computation, a new algorithmic account of attention that rations the general resource of perceptual computations according to their impact on decision making.

Rosenholtz proposes to replace the concept of attention with peripheral vision models, such as the Texture Tiling Model (TTM). Here, we show that the TTM fails in many psychophysical studies due to its local, single-stage, feedforward, and low-level processing. Given that both attention and peripheral vision are unsettled fields, we argue that replacing one with the other is unwarranted.

Visual Attention in Crisis provides the reader with an alternative way to think about the visual attention phenomena–often interpretable in terms of perceptual processes and peripheral vision. We urge an extension of these considerations to developmental science. Infancy research underpins the foundations of mature attentional mechanisms. It may offer a critical test for evolving perceptual limits on attention.

I agree that, in some literature, the term “attention” is ill-defined and the constraints of peripheral vision have been overlooked. However, I disagree with omitting the empirical studies that have emphasized the importance of defining, operationalizing, and manipulating different types of attention and have investigated their effects on perception as a function of eccentricity and polar angle in peripheral vision.

This commentary refutes Rosenholtz’s claim that visual attention lacks conceptual validity. We contend that attention remains important for elucidating capacity-limited perceptual processing and explaining phenomenological experience. Alternative frameworks centered on tasks and decision boundaries fail to account for perceptual effects that attentional mechanisms can capture. Thus, preserving attention as a theoretical construct is important, providing interpretive frameworks for empirical investigations.

Rosenholtz describes “attention theory” as a scientific paradigm in crisis, in the Kuhnian sense. But is attention a theory? Here, I question this premise. Although there can be theories of attentional phenomena, attention is not a theory. Rather, like memory and emotion, attention is a psychological concept that captures a broad class of phenomena, requiring multiple mechanistic explanations.

Rosenholtz addresses the crisis of proliferating mechanisms for visual attention by redefining the concept in terms of (a) the known limitations of peripheral vision and (b) a proper assessment of task complexity. We argue that abandoning the see → decide → act pipeline model and the myth of a centralized gate or resource eliminates this crisis. In our view, “attention” describes an outcome—the consequence of multiple constraints on perception and action—rather than a reified cause.

Rosenholtz (2024) dismisses attentional capture, arguing that brief distractions (20–40 ms) are insignificant or intentional. However, we argue that distractions are never intentional nor negligible, and studying them is crucial both theoretically and for real-world applications.

Extensive research using the attentional blink phenomenon illustrates, through behavioural, modelling and cognitive neuroscience approaches, that distinct selection and attention capacity limits exist. Crucially, these effects cannot reflect peripheral visual processes nor distinct task operations across conditions controlling for issues raised by Rosenholtz. Moving away from attention and selection concepts hinder rather than facilitate a mechanistic understanding of vision.

We suggest that the crisis in visual attention is caused by problems prevalent in many other areas of psychology: low-power experiments and questionable research practices. We demonstrate that many attention experiments use sample sizes that are several orders of magnitude too small and argue that it is unreasonable to expect theoretical clarity based on such unreliable empirical data.