i have been trying to get my cat to stop sitting on top of our living room table. this is for two reasons: one, her cat litter paws are a little gross, and two, she is blind, so she often can accidentally knock things off the table. that said, she is a cat, and likely would try to knock things off the table anyway if she could see, so i am indexing more heavily on the paws

my primary way of accomplishing this is by annoying her whenever she is on the table, in an attempt to create a strong association between (on the table) and (someone is overstimulating me). this has recently started to work very well. but i've discovered something very troubling: i have caught her at 4am, when i go get a glass of water, sitting on the table and staring at nothing at all. when she realizes i am up, she scrambles to run off the table

i had hoped that the causal mapping she learned is between a negative stimulation (being poked) and a recently-arrived-to, current state of being (being on the table). i was wrong. what she has really learnt is something much more sophisticated: object permanence. even in her blindness, she has learned that i am not an omnipresent force but a localized one, and she has adjusted her table-sitting to occur outside my zone of enforcement

this is actually worse than no learning at all. no learning would mean a cat who sits on the table at random intervals, some of which i catch and some of which i don't. what i have instead is a cat who has arrived at the. pareto frontier of table-sitting: maximum table time, minimum poking. i am now considering whether the correct move is to introduce randomness into my own sleep schedule so she cannot predict my enforcement windows, but this seems like it will degrade my quality of life faster than it degrades hers. she does not have a job. she is a blind cat. she has nothing but time