What the Woods Are Said to Repair

AAnyone who has spent a morning proof-reading knows the condition the psychological literature calls directed attention fatigue: the machinery that holds one task in mind while suppressing everything else wears down, and errors follow. Attention restoration theory, set out in the late 1980s, proposes a remedy. It separates directed attention, which is effortful and finite, from involuntary attention, which is captured by things interesting enough to hold the mind without any effort at all: moving water, drifting cloud, a bird crossing a gap in the trees. Natural settings, on this account, are unusually rich in such material and unusually poor in the demands that drain the directed system, so time spent in them lets the depleted faculty recover. Four ingredients are commonly listed: a sense of being away, gentle fascination, enough coherence for the place to feel like a world of its own, and a fit between the setting and what the visitor wants from it. The claim is a precise one, and precise claims can be tested.

BThe most cited field test was run by Marta Vidal at the Halstead Institute in 2004. Sixty-eight undergraduates completed a demanding proof-reading task, then walked for fifty minutes either along a wooded path beside the Ellerby reservoir or through a commercial street of the same length, before repeating a backwards digit-span test. The wooded group's scores rose by twelve per cent, the street group's by two. Vidal was careful about what the result did not show. Both groups reported feeling better afterwards, and by almost identical margins, so the attention gain could not simply be read off from mood. She was equally frank about the study's weakness: the two routes were unalike in traffic noise, in temperature and in the number of decisions a walker had to make, and no single one of these could be isolated by the design she had used. Effects of the size she recorded, around half a standard deviation, have since become the usual benchmark for outdoor work.

CA rival reading of the same evidence came from Daniel Okonkwo, whose papers of 2011 do not dispute that people score better after time outdoors. What he disputes is the mechanism. On his account restoration is a by-product of lowered physiological arousal: green settings are quiet, cool and undemanding, stress hormones fall, and performance on almost any cognitive test improves as a consequence. Nothing in this requires a separate faculty of directed attention, and nothing in it is special to nature, since a warm bath or an empty corridor might serve as well. The two accounts agree about most of the findings, but they part company over a single prediction. If Okonkwo is right, the benefit should disappear once mood and arousal are held steady; if the older theory is right, some of it should survive that treatment.

DThe prediction was put to work by Ines Brandt in 2016. Two hundred and forty volunteers viewed either woodland or urban scenes on screens in a laboratory whose temperature, lighting and noise never varied, and self-reported mood was recorded throughout. The two sets of images had been matched in advance for brightness, colour range and visual complexity, so that neither could be dismissed as merely the easier thing to look at. In the subgroup whose mood ratings did not move at all, attention scores still improved after the woodland scenes, which the older theory had predicted and the arousal account had not. The gain was small: an effect size of 0.21, against the 0.55 typical of walks taken out of doors. Brandt drew the modest conclusion available to her, namely that a part of the effect outlasts the removal of feeling while most of it apparently does not, and that screens remain a weak substitute for places.

EThe sharpest criticism, however, is aimed not at either theory but at the evidence beneath both. Reviewing ninety-six restoration experiments in 2021, Lin Feng found that sixty-one had measured attention with a single test administered twice, a design in which improvement may arise from mere familiarity with the test rather than from anything the participant did in between. Only fourteen of the ninety-six had lodged an analysis plan before any data were collected, and among those fourteen the average effect was half the size reported by the rest. The largest effect anywhere in her sample, she added, had been obtained from eleven students in a single afternoon. Feng did not conclude that restoration is an illusion. Her report goes no further than the observation that the field's headline figures come disproportionately from its least constrained studies.

FNone of this has slowed the theory's career outside the laboratory. It appears in hospital design, in school grounds policy, and in the guidance now issued by several city authorities recommending a weekly dose of twenty minutes in a park. That figure has no experimental basis whatever: it can be traced to the booking slots of one laboratory, whose sessions happened to run for twenty minutes, and no study has yet mapped the curve that would tell a planner where the returns begin to flatten. The prudent position is the one Brandt and Feng converge on. There is something here worth building for, and the case for planting a courtyard does not wait on a ruling about which faculty is being repaired; but a number repeated often enough begins to look like a measurement, and this one is not.