A commonly held notion in the nineteenth century was that experience was composed of immutable “pure sensations.” This idea, which William James described as “as mythical an entity as the Jack of Spades” (James Reference James and Holt1890, p. 236), was abandoned early in the twentieth century. Firestone & Scholl's (F&S's) arguments follow from a similar anachronistic assumption that there exists a state of immutable pure perception.

The notion of pure perception is as untenable today as pure sensation was 100 years ago. Retinal ganglia cells terminate in at least a dozen different brain areas, and feedforward and feedback reciprocal neural connections are ubiquitous in the visual system. According to our contemporary understanding from neuroscience, the brain cannot be carved at its joints so as to isolate pure perception anatomically and functionally.

Perception is, by definition, a meaningful awareness of one's environment and one's perspective on it. Lacking access to cognition, pure perception would be devoid of meaning and consist instead of an absolute associative agnosia in which the conceptual recognition of environmental objects and events is entirely absent. F&S argue that pure perception can be carved out of the flow of mental experience, but they are vague about what its constituents would be. It cannot include conceptual entities, by their account, because concepts fall on the other side of the perception–cognition divide. Being conceptual, even familiar objects cannot serve to scale space in pure perception.

F&S's definition of pure perception, however, is even more restrictive; they place the meaning of perceived distance outside the purview of pure perception. In fact, to perceive distance, the angular units of visual information must be transformed into linear units appropriate for measuring extent. This transformation requires geometry and a ruler. The meaning of a distance therefore is specified by the magnitude that it subtends on a ruler.

Following Gibson (Reference Gibson1979), Proffitt and Linkenauger (Reference Proffitt, Linkenauger, Prinz, Beisert and Herwig2013; Proffitt Reference Proffitt2013) proposed that spatial layout is scaled by perceptual rulers derived from the affordances associated with intended actions. As was certainly the case for Gibson, we view affordances as being neither top-down nor cognitive. F&S, however, lump action influences on perception in with cognitive ones, and they provide no alternative account for the meaningful scaling of spatial layout.

The studies F&S critiqued are an unrepresentative sample of the relevant literature. For example, the first experiments that found an influence of wearing a backpack on people's perception of hill slant were conducted 20 years ago (e.g., Proffitt et al. Reference Proffitt, Bhalla, Gossweiler and Midgett1995). Since then, numerous studies from multiple labs have found evidence for bioenergetic scaling of perceived spatial layout using designs that avoid all of the methodological pitfalls emphasized in the target article. In those studies, researchers too approaches such as:

• The “biological backpack.” Taylor-Covill and Eves (Reference Taylor-Covill and Eves2016) used an individual differences design, devoid of experimental manipulations, to show that percent body fat – what they called a biological backpack – was positively correlated with the perceived slant of stairs. Moreover, by assessing participants over time, they found that changes in body fat, but not fat-free mass, were associated with changes in apparent slant.

• Bioenergetic scaling of walkable distances. Also employing an individual differences design, Zadra et al. (Reference Zadra, Weltman and Proffitt2016) found that physical fitness (measured as VO₂ max at blood lactate threshold) was inversely correlated with perceived distance.

• Energy expenditure correlation. White et al. (Reference White, Shockley and Riley2013) used a treadmill and a virtual environment viewed in a head-mounted display to decouple the relationships between energy expended while walking and the optically specified distance traversed. They found that increasing energy expenditure while keeping constant optically specified distance led to increased distance judgments.

Such findings show that extents and slants are scaled by the bioenergetic costs of walking and ascending. The theoretical generalization from the original backpack studies has thus been repeatedly upheld, but F&S claim that such results merely reflect experimenter demands. We find their evidence unconvincing, for the following reasons:

• The studies purporting to control for demand were themselves demand-ridden (e.g., Durgin et al. Reference Durgin, Klein, Spiegel, Strawser and Williams2012). Participants were warned not to respond the way others did, inflating hill slant estimates to please experimenters. Not surprisingly, those participants responded by giving low estimates of slant compared with the estimates of unwarned participants. Far from eliminating implicit demand, such instructions introduce explicit demand. The literature on “debiasing” (e.g., Schwarz Reference Schwarz, Borgida and Bargh2015) indicates that such warnings tend to bias, rather than debias, judgments.

• Participants in Durgin et al. (Reference Durgin, Baird, Greenburg, Russell, Shaughnessy and Waymouth2009) were told that the backpack contained equipment to measure muscle potential. An elaborate story ensured that participants were alerted to the irrelevance of their experience of the backpack for judging slant. Additionally, a noisy cooling fan on the backpack served as a constant reminder of its irrelevance. Rather than eliminating demand, the elaborate effort to make the backpack both salient and distinct ensured that its heaviness would be segmented from other bioennergetic cues when making slant estimates. The results seem foreordained by the procedures. The findings likely reflect not only the odd backpack manipulation, but also the fact that the “hill” was only a 2-meter-long ramp. This hill did not afford the opportunity to walk more than a step or two, rendering any bioenergetic costs of wearing a backpack irrelevant (Proffitt Reference Proffitt2009).

• To assess awareness, experimenters (e.g., Durgin et al. Reference Durgin, Baird, Greenburg, Russell, Shaughnessy and Waymouth2009) first asked participants to consider the backpack heaviness and then asked for their hypotheses. The literature on assessing awareness has long warned against such procedures (e.g., Bargh & Chartrand, Reference Bargh, Chartrand, Reis and Judd2000; Dulany Reference Dulany1962), because participants cannot reliably distinguish hypotheses elicited by being asked questions from hypotheses actively generated during the experiment. To avoid such contamination, recommended methods involve carefully designed funnel interviews (e.g., Dulany Reference Dulany1962), which were not used in this research.

• Other studies have found that replenishing depleted glucose can lower slant estimates (Schnall et al. Reference Schnall, Zadra and Proffitt2010), but F&S suggest this occurs not because changes in resources affect perception, but because added glucose empowers participants to resist experimenter demands. We are not aware of any data supporting glucose effects on susceptibility to demand. More important, such suggestions cannot account for the results of multiple new studies, only some of which we cited earlier, which are simultaneously immune to the criticisms of F&S and supportive of the perceptual effects they attack.