It truly is "An Immense World": what Ed Yong's review of sensory biology can teach us about animal welfare
- Team AWE
- Aug 4
- 6 min read
Assessing animal welfare often, perhaps unavoidably, starts from our own point of view.
We might look at a space and ask whether we would want to live there. Ed Yong's An Immense World offers a gentle challenge to that habit. It's a tour through the sensory worlds of other species, and read with a welfare hat on, it becomes an invitation to reconsider how we think about what animals actually need.
We don't often review books on here, but this one was too important and brilliantly written to pass by. Here are four ideas from the book, and what they might mean for how we assess and improve welfare.
1. Different windows onto the same garden
Yong leans on the biologist Jakob von Uexküll's concept of the umwelt: that every animal inhabits its own private perceptual world, built from whichever senses matter most to it. Uexküll's own metaphor is architectural.
Picture a garden surrounded by a house full of windows, each one a different shape, a different height, a different tint of glass. A tick, a bat, an elephant, and a human are all looking at the same garden, but each window admits only a narrow slice of it, filtered through senses tuned to that animal's particular needs.
This matters for welfare because it means two enclosures can look identical to us and be entirely different environments to their residents. A "naturalistic" habitat designed around what we can see tells us very little about what the resident can smell, feel, or sense in ways we don't have a window for at all.
2. Smell as a social network
Yong's example of a dog on a walk is a good case to consider. To us, a carcass or a poo is unpleasant; something to steer the dog away from. To the dog, it is closer to an Instagram feed or a noticeboard: who passed through, how long ago, whether they were healthy, whether they were in season.
The same logic runs through the animal kingdom in all sorts of forms. Snakes and some lizards flick a forked tongue through the air not to taste in the way we understand tasting, but to sample scent molecules at two points simultaneously, allowing the brain to work out the direction a trail is coming from, and, in some cases, information about the health or reproductive status of the animal that left it.
For welfare, the implication is straightforward: removing scent-marking opportunities, sterilising enclosures, or cleaning too thoroughly and too often may strip out an entire information channel that matters enormously to the animal. Research on laboratory mice bears this out directly: a full change of bedding removes the odours that maintain a group's social hierarchy, and this has been shown to spike aggression between males, whereas leaving some used nesting material in place settles things back down (Van Loo et al. 2000).
This is the thinking behind "sniff walks": letting a dog set the pace and stop wherever her nose tells her to, rather than ours. And preserving informational cues by allowing scents to stay in the environment longer might be a reason for zoos to scale back their cleaning schedules (while still maintaining a safe environment of course).
3. Colour without our rules
Dogs and horses are dichromats. They have only two functioning cone cell types, sensitive to yellow-green and blue-violet light, rather than the three most humans carry. Practically, this means their world is built from blues, yellows, and greys, with reds and greens collapsing into similar, muted tones. Yong describes how a bright green toy, held up to a dog, stimulates both of her cone types almost equally; the signals effectively cancel out, and what she perceives is closer to white or grey than the vivid green we see.
Yong takes this further with a thought experiment: if horse racing had been designed around equine vision rather than ours, the markers, rails, and jump poles would be fluorescent yellow, white, or bright blue, since these are the colours that actually stand out against grass and track from a horse's dichromatic viewpoint. Instead, we've built the sport around what looks vivid to us.
As Yong puts it: “Since eyes define nature's palette, an animal's palette tells you whose eyes it is trying to catch." The colours an animal can see (or that a species has evolved to display) are not incidental. Enrichment items, target-training tools, and enclosure markers chosen for their appeal to human eyes may be functionally invisible or unremarkable to the animal they're meant for.
4. Considering the whole sensory landscape
Human activity is steadily narrowing the umwelten of wild animals too. Streetlights and artificial light at night disrupt species that navigate or forage by starlight or moonlight. Traffic and industrial noise mask the calls birds use to find mates or the low-frequency rumbles elephants use to coordinate over long distances. Underwater, shipping noise invades the acoustic spaces of whales and fish. None of this shows up as an obvious, or intentional, injury, but it removes information animals rely on just as surely as blocking a window would.
Consider, for instance, seasonal light shows now hosted at a number of UK zoos, often set up around or near the animal exhibits themselves. Bright, moving illuminations and amplified music are built for a human sensory experience, but for animals living immediately alongside them, this is functionally an imposed change to the light and sound environment they inhabit every night for weeks at a time. A species that navigates by dim, stable light cues, or that relies on quiet to rest or communicate, may find the visitors' magical evening to be a fairly significant disruption to its own umwelt.
Thankfully, we're seeing more research looking into the effects of such out-of-hours zoo events. For example, in this study the impacts of an evening light show events was measured on the gorillas, and as a result of the findings (less resting, potentially more Abnormal Repetitive Behaviours) the zoo closed these areas off for future events. But even these types of studies can go further to consider the appropriate sensory impacts relevant to different species.
Wherever possible, we can 'ask' animals what they choose to include or limit within their environments. When researchers gave rats a choice of soundscape, the animals preferred silence over any kind of radio sound, whether music or talk, and preferred the radio over white noise (Krohn et al. 2011). Furthermore, rodents can hear well into the ultrasonic range, far beyond human hearing, which means everyday sounds we don't register at all, like a running tap or a squeaky chair, can register to them as a constant, low-grade source of stress that we have no way of noticing ourselves.
From umwelt to enrichment: a sensory framework for animal welfare
Put these ideas together and they suggest a useful practical shift for welfare considerations: instead of designing enrichment around what looks stimulating to us, we could deconstruct an animal's needs by primary sensory domain, and build outward from there.
We should also consider an animal's umwelt when exploring welfare indicators within assessments, making sure to include environmental measures that give us meaningful information about how those senses are able to function.
A rough framework, adapted from the sensory hierarchy Yong focuses on across the book, might include the following considerations for animals that primarily use the bolded senses:
Smell and taste — provide scent trails, marking opportunities, chemical cues from other individuals, limit cleaning procedures, artificial perfumes and smells
Light — extra consideration of light intensity, day length, UV where relevant, visual cues
Colour — ensure visible spectrum matched to the species' actual cone types, not ours
Sound — consider frequency ranges the animal can hear, including infrasound or ultrasound where applicable, limit visitor noise
Electroreception — consider electric currents and wiring near enclosures, relevant for species like sharks, rays, some fish, and some amphibians and monotremes
Magnetoreception — relevant for migratory species and some invertebrates
Echolocation — integrate echolocation into enrichment (e.g. food release following echolocation beam), relevant for bats and toothed whales
Vibration — consdier substrate-borne cues, seismic information, whisker contact
Touch and contact — ensure that texture and temperature is considered across substrates, physical contact with conspecifics or objects will likely be important
A final thought: there is no single window
An Immense World resists a tidy conclusion, and that seems right. There is no universal, species-neutral way to measure a good environment, because there is no single umwelt that all animals share.
In our own lives, we generally know which sensations make an experience better or worse, and we usually have some control over them, or know they will only be temporary.
At present, animals in our care are rarely able to control when a sensory experience enters their environment. They can't opt out of a lighting schedule, a cleaning routine, or a Christmas light show, and they have no way of knowing it will end. This current asymmetry is exactly why the sensory audit Yong's book invites us to do matters: the responsibility for noticing what an environment sounds, smells, feels, and looks like from the animal's window. It should also prompt us to explore how we can give the animals genuine choice and control over as many of these elements as possible.
Comments