The argument works by subtraction, and that is the most interesting thing about it. Researchers have catalogued more than 140 gestures in eastern chimpanzees and more than 120 in mountain gorillas [6], and the same movement can carry different meanings depending on the situation and on the history between the two animals [7]. Add planning, coordinated hunting, tool use, care for injured companions and the management of alliances [8], and the functional case for speech gets thin. Hobaiter says as much: decades of comparative work suggest you do not need language to learn from others, to organise foraging, to pass on tool and song traditions, or to navigate social politics, which makes it look as though language did not evolve for anything very useful [9]. What is left over, on this account, is the part of the day that fire created.
Take the upper bound literally and four extra hours a day is on the order of 1,460 hours a year [2], a block of social time with no daylight in it. The paper's case for why that block favoured voice over hand rests on two measured properties rather than on inference about behaviour. The first is geometry: preliminary measurements show a campfire lighting the faces nearest it while leaving people further off, or across the flames, hard to see [12], and the unsteady brightness makes gestures harder to follow [13]. The second is timing. A flicker peak at 2.7 cycles per second [15] is about 370 milliseconds per brighten-and-dim cycle [1], which is slow enough to sit alongside speech rhythm rather than inside the range where flicker becomes visual noise. Firelight also carries substantial infrared [17], and the light softens as the fire burns down, with embers suited to winding down for the night [22].
The thing this does not tell you is whether entrainment does any work. Visual neurons can lock to flickering light, and the authors raise alignment of attention and behaviour among listeners as a question rather than a result [16]. Nor do the numbers describe a Pleistocene fire; they describe fires the authors measured, and the descriptions are labelled preliminary [12]. The negative observation at the centre of the theory - that no other ape has been seen turning events into a story [11] - is an absence in a record built by people who were mostly looking for something else.
What raises this above an origin story is that the six closing predictions are not about the past [18]. Whether stories differ in pace and structure from ordinary conversation is a corpus question, and the cheapest of the three named tests to run. Whether speech beats gesture for comprehension under flickering light is straightforward psychophysics; the control it needs is steady illumination at the same average brightness, so that flicker is the only variable separating the conditions. Whether storyteller and audience synchronise more tightly is the hardest and the most interesting, because it is the step where the physics is supposed to become sociality. Hobaiter frames the ambition as escaping after-dinner speculation and asking whether we are the storytelling ape [19], and Dominy notes that firelight is a well studied chemical reaction except for the properties that hold attention [20]. On present evidence I would expect the intelligibility prediction to hold and the synchrony prediction to be where the hypothesis is decided.
