Between July and September 2017, a team from California Polytechnic State University buried speakers along roughly a third of a mile stretches of trail in the Boulder Open Space and Mountain Parks system, wired them to play recordings of birdsong, and waited to see whether hikers who walked past could tell the difference. Most, by the researchers’ own account, could not.
The study, published in Proceedings of the Royal Society B in December 2020, was led by Danielle Ferraro, then a biology graduate student at Cal Poly, with Clinton Francis, Zachariah D. Miller, Lauren Ferguson, Derrick Taff, Jesse Barber and Peter Newman as co-authors. The paper’s title calls the setup a “phantom birdsong chorus,” and that is the more precise term. It was not run at dawn. Data collection took place across the middle hours of the day, when hikers were actually out on the trails, and the recordings were built to simulate a denser, more varied mix of bird species than the trail would ordinarily produce at that time, not to recreate the specific acoustic conditions of first light. The “livelier dawn chorus” framing captures the spirit of the manipulation, the layering of many species’ songs into a chorus richer than what was really there, but the timing detail is worth keeping straight.
What the speakers actually did
The researchers alternated the trail sections between “on” blocks, when the hidden speakers played a mix of recorded songs from birds that could plausibly be heard in that habitat, and “off” blocks, when hikers heard only the ambient soundscape. The blocks rotated on a weekly basis across the two trail sections. Hikers were intercepted as they came off the trail and asked to complete a questionnaire about their experience, without being told in advance what the study was investigating.
Clinton Francis, who oversaw the project, told National Geographic that the deception was deliberate and that his team worried about it being discovered: “One of my graduate students was so nervous that an incredibly good eBirder was going to find a speaker and be outraged.” To the researchers’ knowledge, that never happened. The speakers went undetected for the length of the study, which is itself part of the finding: whatever effect the phantom chorus had did not depend on hikers consciously registering that something artificial was being piped into the landscape.
What hikers actually reported
The exposure window was short. Francis put it at seven to ten minutes; a later account of the study describes it as lasting “about 10 minutes.” In that window, hikers who walked through an “on” block reported higher restorative effects on their questionnaires than hikers who walked through the same stretch during an “off” block, on both trail sections. The effect was consistent across the two trails, but it worked through different routes. On one trail, hikers who heard more birdsong simply reported feeling better, without indicating that they thought more bird species were present. On the other trail, the improved sense of well-being was mediated by perceived biodiversity: hikers who heard the phantom chorus were more likely to believe more species lived along that stretch, and it was that belief, not the sound alone, that the researchers’ statistical model linked to the reported improvement.
This is one study, not settled consensus about birdsong and well-being generally. It is a controlled field experiment run on two trail sections in one Colorado park system over a single summer, using self-reported questionnaire data collected immediately after a short hike. The authors themselves frame it as adding to “a growing body of evidence linking mental health to nature experiences,” not as closing the question of why nature helps.
Why the perception mattered as much as the sound
The split between the two trails is the part of the paper that repays a second look. If the phantom chorus worked purely as a pleasant background noise, its effect should have looked the same on both trails: more birdsong, better mood, independent of whatever the hiker believed about the ecosystem around them. Instead, on one of the two trails, the effect ran through a hiker’s belief about biodiversity, meaning the sound worked partly by changing what people thought was true about the place they were standing in, not only by changing what they heard.
That distinction matters for how the result should be used. It is not evidence that ambient bird noise, piped in indiscriminately, produces a reliable calming effect on its own. It suggests, on this dataset, that the perception of a richer natural environment, formed on the basis of what a person can hear, is doing some of the work. Francis described the effect in blunter terms to National Geographic, calling it a “clearing of the mind,” and said he was “flabbergasted” that so short an exposure produced a measurable shift at all.
What the study does not show
The paper does not measure long-term outcomes. It says nothing about whether the restorative effect persists after a hiker leaves the trail, or whether repeated exposure compounds or fades. It relies on self-report, collected once, right after the walk, which is a real limitation of the design rather than a footnote: people are not always accurate narrators of their own mental states, particularly when asked to summarise them in a brief questionnaire under mild time pressure. And the study was carried out in a single protected area with a specific existing soundscape, hikers who chose to be there, and a demographic that is unlikely to represent hikers, or non-hikers, more broadly.
The authors’ own conclusion is narrower than the headline version of the finding. They argue that audition, the sense of hearing, is an underexamined route by which people register the biodiversity of natural environments and draw benefit from that environment, and that protected-area managers might treat noise pollution as something worth managing, not merely tolerating, if they want visitors to have a richer experience. That is a modest, specific claim about soundscapes and park management. It is not a claim that recorded birdsong is a reliable treatment for anxiety, or that any hiker exposed to more birdsong will feel calmer in a predictable, reproducible way.
The part that is easy to miss
What stands out about the design, more than the headline result, is what it required the researchers to give up. To test whether birdsong causes a change in how people feel, rather than merely correlating with it, they needed a controlled manipulation, which meant deceiving the people they were studying about what was actually happening around them. The alternative, telling hikers in advance that some of what they were hearing was recorded, would have compromised the very thing being measured. The paper’s contribution is less about birds than about method: proving a causal link between a single sense and a self-reported outcome in an uncontrolled outdoor setting is difficult, and this is one of the few times it has been attempted with this kind of rigour outside a laboratory.
Seven years on, the speakers are long gone from those trails. What remains is a dataset suggesting that a short burst of manufactured birdsong, undetected as manufactured, shifted how a group of strangers described their own state of mind on the way back to the car park.