Comment by tareqak | Hacker News Reader

Comment by tareqak | Hacker News Reader

The article title asks “Why cats won’t punish a stranger who harms you”, but the experiment that the article discusses has “helper”, “non-helper”, and passive actors. Maybe cats do not make the same association that dogs seem to do in this experiment that not helping is the same thing as harming?

The question of reacting to harm incurred by a human companion would be better answered if the action that the animals observe appeared to be harm such as a mock assault where the companion appears to be in pain as a result of the active participant e.g. the active participant appears to punch the companion in the belly without actually making contact such that the companion falls to the ground reeling in false pain. The Japanese study linked in the article discusses this possibility in its “Discussion” section.

> It is conceivable that the cats in this study did not understand the meaning or goal of the owners’ behavior (attempting to open the container to get the object inside). To our knowledge, cats’ ability to recognize others’ goals or intentions from their actions has not been investigated. But even if they did understand the owner’s goal or intention, they might have failed to detect the negative intention of the non-helpful actor (deliberately withholding assistance). Further studies would be valuable for testing alternative third-party evaluation situations and for clarifying whether cats recognize others’ goals or intentions. [0]

[0] http://animalbehaviorandcognition.org/uploads/journals/30/AB...

Agreed. I'm a big believer in animal intelligence and I've lived with both cats and dogs, and this setup seems to me to be difficult for cats to understand, too abstract and removed from cat's natural capacities. They made no effort to establish that the subject animal had any understanding of what was happening before varying the experimental parameters. Basically I would say this experiment is useless as it is; it's just not good science.

I don’t think it’s necessarily bad science. The null hypothesis is that cats would react the same as dogs, and the study suggests that’s not true. While this is not surprising to most people, it is still a more rigorous attempt to understand the differences between cats and dogs than simple intuition and anecdotal experience. I do think the experimenters should probably have done a follow-up study more attuned to cats’ natural abilities, but they likely feel the same way.

I agree. "Not helping" is a much harder cue to pick up than an actual aggression.

As the article mentions:

> " So we shouldn’t be too quick to conclude that our cats don’t care if people are mean to us. What’s more likely is that they just can’t tell."

Cat's social theory of mind is different from humans and dogs. Cats do understand 1-1 relations (between them and other people) but relations between other people are harder.

(On the other hand they know how to use the litterbox so...)

yes, cats are very curious and perceptive, constantly trying to figure out how the world works (and how to eat that bird outside the window!). just because they don’t perceive the human problem in this case doesn’t mean they don’t generally want to help.

cats are social, just not the same way dogs or humans are, and they form habits based on things they’ve figured out about the world (the litterbox being a good example, a hybrid instinct & learned social behavior). my cat has figured out that her human and her canine sister are harmless to her and she can relax and even play and cuddle with us. she’s also food motivated, so can be trained on that basis.

Reply on news.ycombinator.com