Skinner and the Mousewrap

Mousewrap -, an interesting interface design experiment by Alex Frank, included this amusing idea, the Mousewrap, to ‘train’ users not to click any more “through physical pain”.

It did make me think: is the use of anti-sit spikes on window sills, ledges, and so on, or anti-climb spikes on walls, intended primarily as a Skinnerian operant conditioning method (punishment – i.e. getting spiked – leading to decrease in the behaviour), or as a perceived affordance method (we see that it looks uncomfortable to sit down, so we don’t do it)? How do deterrents like this actually work?

It might seem a subtle difference, and in practice it probably doesn’t matter; it’s probably a bit of both, in fact. Most people will be discouraged by seeing the spikes, and for the few who aren’t, they’ll learn after getting spiked.

But on what level do anti-pigeon spikes work? Do pigeons perceive the lack of ‘comfort’ affordance? Or do they try and perch and only then ‘learn’? How similar does the spike (or whatever) have to be to others the animal has seen? Do animals (and humans) only learn to perceive affordances (or the lack of them) after having been through the operant conditioning process previously – and then generalising from that experience to all spikes?

What’s the accepted psychological wisdom on this?

Some spikes in Windsor, Poundbury, Chiswick and Dalston, UK.


  1. Steven Fusty outlined a typology of spaces in which he would describe these as prickly spaces — spaces which deter or prevent occupation.

    Greg Smithsimon writes, “Prickly space technology has the advantages of being easily retrofitted onto existing spaces, and of representing an unambiguous message of control to potential users.”

    But are they always so unambiguous?

    Influence or control can be exercised by social or physical means. Consider the relative merits of a “no loitering” sign, regular police/security patrols to move loiterers on and the simple absence of anywhere (comfortable) to sit.

    In my neighbourhood, there are large areas of the footway with raised, irregular cobblestones that are clearly designed to say “don’t walk here” and to channel pedestrians away from their natural desire lines onto routes that the designers considered to be safer. Yet they seem to be largely ineffective, the stones neither reading unambiguously as “no walking” nor sufficiently prickly to prevent walking rather than to make it more difficult (and ironically, more risky).

    I suspect the intention of these cobbles is observed more often by locals who are familiar with the area and will also have learned the “correct” routes to take rather than visitors to whom their intention is unclear and for whom the alternatives are less obvious. A study of who walks on the cobbles, who doesn’t, and why, might be enlightening.

    Other examples spring to mind: cattle grids and electrified fences on farms.

  2. Peter

    >electrified fences on farms

    Not sure about cattle grids but I can definitely say that electric fences are a learned obstacle. On our farm, when I was growing up, when we wanted to train a new horse that an electric fence should be avoided, we would place food on one side and the horse on the other. A nasty, but memorable, shock resulted when the horse went for the food, and from then on it took only a single strand of wire to keep half a ton of horse under control – it didn’t even need to be switched on all the time.

  3. rjh

    Cattle grids are learned, and also it turns out are visual. Painted grids will also work once cattle are trained. But painted grids also lead to unlearning, when accidental movement reveals the grid to be fake.

    Pigeon spikes turn out to be discomfort based rather than pain based. The effective ones are not sharp. They are flexible. The pigeon cannot roost properly. They try to hold onto the spike, but it falls over under their weight. There is no position where they can roost without a spike interfering with wings or feet because they are too densely placed to avoid.

  4. tom

    Arbitrary associations (like wires and shocks, which I’m sure seem pretty arbitrary to a horse) have to be learnt, but all animals also come equiped with some some stimuli which they are drawn to and some which they are predisposed to avoid (eg monkeys and snakes, rats and bright lights).

    Humans are different from other animals in the extent that they can use mental simulation to provide pseudo-feedback for themselves (“I wonder what would happen if I tried to climb over that razor wire?”)

Comments are closed.