Cats are not humans, but we design things like they are, for our convenience. What are the design patterns for cat-architecture? One missing design pattern: progressive concealment, for cat ledges, flaps & window boxes.
I suggest that cats have innate sensory preferences that existing cat-architecture (like ‘cat window boxes’) is blind to, and so fails to accommodate: while driven to monitor the outside world, they are highly sensitive to risk and personal exposure, and want to constantly adjust how much they can see or hear, or be seen or heard.
This essay proposes non-anthropocentric principles for cat-friendly architecture flowing from a cat’s-eye-view. Thoughtful cat-itecture enriches cats’ environments with: options of gradation, prioritizing soundscapes over sight-lines, and simplicity of use.
Current cat enclosures like window boxes are all-or-nothing designs which, while good for ventilation or simple construction, expose cats to extremes of exposure at the cost of control over their visibility or the intensity of sound/sight. Applying these principles could improve cat window boxes through features like sound baffling, opaque retreats, and clear vantages that balance seclusion and stimulation.
By looking through a cat’s eyes and listening, we can create spaces with use cases that reflect how cats want to use them.
To help my increasingly-elderly cat get to his cat-flap, I installed a ramp of cinderblocks on the inside & outside. The ramp worked well, and had the unexpected benefit of giving him more options in approaching the cat-flap beyond just ‘on’ or ‘off’, so I could watch how he would approach the cat-flap over the course of the day. What I noticed was something I’d suspected for a while: he acted like he had a ‘novelty’ or ‘risk compensation’ drive akin to temperature or hunger or thirst, where he craved a certain slowly-changing optimal degree of risk/stimulation, and would adjust his position to maintain his position on the inverted U-curve.
This outside-observation drive would be consistent with other observations about useful cat toys simulating the hunting kill-chain & knocking things over to test playing-dead: long-term monitoring of the outside while as invisible as possible is clearly useful preparation for hunting, as the cat can learn about prey movement patterns, what animals pass through when, what rivals or threats may be about, and so on. (We can answer Bradshaw’s question about what his cat thinks watching the outside for endless hours by noting that this information all expires, because the environment is non-stationary & changing, and so it can never become permanently unsatisfying any more than drinking water to quench thirst can become unsatisfying.)
Risk Compensation
![“I had to mow my lawn, but my cat loves chilling in it, so I left a little patch for him.”](/doc/cat/psychology/2023-11-04-jacobjanerka-cattryingtohideinapatchofgrassinthemiddleofamownlawn.jpg)
What is the cat’s-eye view of how best to watch the outside? I interpret this as risk compensation: there is a certain amount of stress or eustress he is willing to endure at any moment to watch the outside, which is an intrinsically-rewarding behavioral drive for a cat, but it still has diminishing returns (just like eating or playing or sleeping); so as the risk changes outside/inside, or the internal reward diminishes, or his overall energy level changes, he will adjust his position to match. The environment offers both exposure to danger, and protection, and a cat will make good use of fortifications for cover, like corners, edges, ledges, elevated surfaces, and especially love boxes to hide in (even illusory boxes!).
Back & forth. Like a cat hovering at the door deciding whether to go out1, the setpoint changes over the course of the day—highest in the crepuscular periods of dawn/dusk (when cats tend to be highly active), and then decreasing in between, depending on the activity outside.
A sudden unusual event (the arrival of strange humans or large animals like turkey vultures, bald eagles, or deer; yard work; airplanes; bad weather) would spike the risk, and he would retreat step by step. Long quiet periods would see the opposite, with advancing, culminating in leaving the outside ramp behind and boldly stalking the wilderness. Are there strangers outside? Concerning—retreat one step. His owner goes outside for a computer break, thereby scaring away enemies & making outside safer? Sitting around would be boring, so advance two steps.2 Did some extremely loud and unusual noise just happen, like a riding lawnmower starting up? Extremely frightening—retreat 3 steps! It’s been quiet & boring, and he can’t hear any birds from atop his ‘cat condo’?3 Try to get a better view of everything, so advance one step… And so on.
Sight Vs Sound
In terms of preferred position, the order appears to be:
-
underneath the bed in the dark bedroom down the hall (in extremis):
not seen nor able to see anything, but able to hear some things in the living room
-
inside the darkness of the ‘cat condo’ box:
able to see the living room but no one able to see him (due to black fur)
-
on top of the cat condo box:
able to see & be seen, and hear inside
-
on the middle of the cat ramp leading to an open cat-flap; or laying on the window ledge when the shades are down; or sitting on the desk near me:
able to hear the outside but neither see nor be seen
-
at the top of the cat ramp but not on the ledge:
able to hear and just barely see over the edge out the cat flap to the outside but almost invisible to the outside
-
on the cat-flap ledge inside, able to hear & see much of the outside and vice-versa, but with much still out of view4
-
outside the cat flap, laying or sitting on the middle ramp level, body partially shielded from view from some angles, but otherwise able to see & hear everything
-
outside the cat flap, on the top of the ramp, highly visible
-
walking or stalking around in general, away from the house
This risk compensation view is almost common sense (of course an ambush predator like a cat would have an acute sense of its own visibility, and being visible would make it anxious), but what’s interesting to me here is the multiple sensory modalities.
A human concerned about varying levels of privacy in architecture would focus almost entirely on vision, and how to progressively block it, with sound being an afterthought. (See eg. “open plan” offices.) But cats have acute hearing, and he appears to often be content to close his eyes and quasi-nap on the ramp or behind the window drapes, ‘keeping an eye’ on me & the outside by instead listening.5 So for him, the optimal locations will vary sound as much as sight. (I suspect that many cat owners have noticed the risk compensation part, but not had cat-itecture with enough steps which would let them notice the sensory difference.)
Cat Window Boxes
Default Design
![Imagequilt (3×7 grid) of 21 cat window boxes selected from Google Images search for “cat window box”, showing general vulnerability of design in exposing cats to the outside.](/doc/cat/psychology/2023-11-03-gwern-googleimages-catwindowbox-imagequilt.png)
The bird cage. Human vision-centrism carries over to cat architecture, or ‘cat-itecture’ one might say. When people build those ‘outside boxes’ or ‘cat window boxes’ (also called ‘solarium’, ‘catio’, ‘cat balcony’) for their cats to sit in or make window flaps, invariably the design tends to be focused on vision & physical exposure to the outside.
A typical cat window box might be a rectangle or cube, with a wood floor and then made of plexiglas/window screen, an open window frame on the back side, and visible on all sides with mesh/holes for ventilation/sound, and have a portal/cat-flap: the cat is either all the way in, and able to see & hear everything and vice-versa, or not. There are no steps where visibility is minimized to a slit, or where it is sound-only. (Sometimes, one is reminded of old-fashioned zoos, or the dangling cages-in-midair in so much fiction where the hero and/or heroine will be locked by the villain, or of window prostitution; hardly a relaxing situation!6)
It is true that many cats love their cat window box and spend hours in them, but that doesn’t mean that the design is optimal: “hunger is the best spice”. (What would it look like if a design were not optimal? It’s not like cats can say that they would prefer a different design, when even most humans struggle to articulate their dissatisfaction with something or quietly put up with extremely broken designs.)
From the risk compensation view, the typical cat flap or cat window box is a crude half-measure missing many steps; like all too much modern human architecture, it appears designed to create anxiety. A cat must be at a fairly high level to want that level of exposure, and then all the lower levels are pessimized, leaving the cat bored.
Cat Design Patterns
How should a cat window box be designed, from a cat’s point of view rather than a human’s vision-centric point of view? (Or to put it another way: If cats are autistic, “what cat window box would Temple Grandin design?”)
What should be the design patterns of cat-itecture?
-
Gradation: we would want it to cover as many steps as possible.
-
Sight vs Sound: the steps should separately vary vision and sound exposure: if they co-vary unnecessarily, then the steps will be coarser.
-
Sight, then Sound: they should follow a vision-then-sound approach, as sound is the safest & least-stressful modality at each step.
-
Simple Use: the steps should be static, intrinsically part of the construction, implemented mostly by moving short distances.
It would be unreasonable to expect a cat to have to manipulate a set of controls to raise or lower blinds, or something like that—unless one could make a case for the mechanism being cheap, highly robust, with instant feedback, and something a cat could easily train itself to do. Anything more complex is doomed.
A mechanically-simple design that might work would be something like a single slider that the cat might push/pull by accident, and it would learn to use it because it would immediately observe the increase/decrease in visibility.
Improved Cat Window Box
So one design might be a double box, with the boxes designed to have multiple ways of mutually obscuring the outside & inside:
-
The outer or ‘second’ box is clear Plexiglas on the top and sides, with finely-pierced ventilation holes towards the top, but opaque on the bottom & in a band around the sides starting at the bottom, with a Plexiglas cat-flap to the outside (if it is an indoor-outdoor cat). This box is always outside.
This provides 3 steps: (1) the cat can exit it for the outside; (2) a cat can sit or stand in this, and be fully visible with complete sight & sound; and (3) it lays down, surrounded by the opaque band, and becomes invisible to anyone looking at it from a ground angle, able to see the sky while also able to lift its head just enough to look over the band into the yard.
-
The inner or ‘first’ box is connected to the outer box with a round narrow open porthole; it has some but fewer ventilation/sound holes, and it is all opaque7, except perhaps for a sliding segment placed at roughly eye level for a crouching or sitting cat which removes the opacity for a narrow slit. This box may be outside with the first box ‘next to’ it side-by-side, or it may be inside with the first box then the outside half, split by the window.
This provides another 2 steps: (1) the cat can sit or crouch with its head sticking through the porthole, granting it greater invisibility against both ground & sky but compromising vision (sound remaining the same); (2) slit closed, laying down, completely invisible but also zero vision, sound-only.
If we add in a slider, we get a 3rd step: (3) the cat can withdraw and pull/push the slider to ‘open’, and crouch looking through it (as if it were an embrasure), now even more invisible at the cost of more vision.
Past that, the cat retreats into the house proper, where the next step is presumably sitting at the window, unable to hear anything but seeing a decent amount and being somewhat visible to the outside. (Then it can retreat to furniture like cat condos or chairs or sofas or under Christmas trees, and then to a bedroom and then under a bed and so on.)
This improves over the typical cat window box by giving the cat fine-grained control over its exposure and sensory overload, while not taking up much more space than a typical cat window box does.
With more engineering complexity, one would add further subdivisions to the hierarchy (eg. perhaps the outer box could be divided in half vertically with two floors, and the upper floor be open completely?), but I think this design sketch demonstrates that if we think in terms of steps and a vision-then-sound hierarchy, we would approach the design of cat-architecture differently.
-
Itself an example of risk compensation: when I experiment with opening a door for a cat, I notice that most seem to go to the threshold to watch the outside intently, and then at last, almost violently choose to either go outside or retreat deep into the house (possibly into another room entirely).
It is an arrogant & fearless cat indeed which will simply sit there, having decided not to go out after all, and calmly watch you close the door.↩︎
-
Many cat owners have noticed that cats will appear to be ignoring them, but ‘just happen’ to keep wandering around near the humans, or napping in places which let them watch everything from a safe distance (in stark contrast to dogs who insist on getting into the thick of things, regardless of how wise that would be).
When my cats follows me around outside, it amuses me to test how far they will lag behind and then carefully break the sight-line and see how they catch up.
At one point strolling around the corner, noticing my cat had lagged unusually far behind, I silently climbed up a convenient tree about 1.5 meters, and waited. Within a minute or two, he had come trotting along, head swiveling back & forth, and halting near the tree in confusion—presumably my smell had stopped at that point, and yet, I was nowhere in sight! Finally, he looked up and was visibly startled to see me—and then turned away to ostentatiously sniff some acorns on the ground. Surely he hadn’t been looking for me (baka!), he was just interested in the acorns in that particular spot, that was all.
These sorts of stories serve to illustrate a common theme of cat-human psychology research: cats are more alert to their environment, and bonded with humans, than humans think, because they attempt to conceal it or they express it in ways which are not as blatant anthropomorphic as more domesticated animals like dogs.↩︎
-
If there aren’t enough birds in general, bird seed is cheap in bulk & highly effective. The step-wise concealment cat behavior is even more striking when one uses bird seed to frequently bring a large number of birds/squirrels extremely close to the cat flap.↩︎
-
My cat-flap is currently just an open window. It was a window mesh with a pivoting door before I built the ramps, but it broke and I had to remove it, and have not yet installed a more robust cat-flap.
While it was working, I noticed my cat would often deliberately lay against it and prop it open with his paw (letting in mosquitoes)—but without the ramp providing extra steps, I was never sure if he was doing that just to have somewhere convenient to prop his paw, or if it was about visibility. Less ambiguously (and much more amusingly), when birds were nearby outside, he would jump up to the desk and begin stalking towards the cat-flap—sometimes not completing the step across the gap, freezing, and simply remaining stretched out half-on half-off. Presumably he froze when he felt he had become too visible through the cat-flap to the unsuspecting prey.
(This also serves as an example of the benefits of the cat gait implementing ‘direct registering’—moving the hindpaw to exactly where the forepaw was—because once the forepaw has been correctly placed across a gap, the cat doesn’t need to think about the step any further. So for a cat, stepping across a gap is no harder than stepping anywhere else.)↩︎
-
What about smell? Do we need to take into account a cat’s other major sense? I think probably not.
It is true that cats have an excellent sense of smell, but smell doesn’t seem to play a major role in how I’ve watched cats watch the outside: they do not inhale deeply or show Flehmen responses. It is usually sounds or motion which triggers alertness. When they smell, it seems like more of a generic “what does the outside smell like today” sort of smell (eg.).
I suspect that smell is, for all its virtues, too physically close & temporally low-resolution to be useful for passive observation. A cat has to go outside and get close to things to smell them—and then they intensely smell things. But not otherwise.
This is good because controlling smell independently of sight & sound, while maintaining ventilation & safety, would be quite a challenge!↩︎
-
One horrific design puts a litter box in the transparent cat window box. Their ad copy boasts that “The Katio features a window on the outside of the box so your cat can get a nice view of the world while pooping, much like man has lusted after for ages…”
Aside from violating all advice on how to avoid litter box aversion, this raises questions about the author’s understanding of cat and human psychology.↩︎
-
This opacity may need to be as minimal as possible: cat window boxes exposed to the sun may get uncomfortably, not to mention dangerously, hot.↩︎