#notes; #onwards (currently reading)
The Trainable Cat (John Bradshaw + Sarah Ellis)
A book subtitled, “A practical guide to making life happier for you and your cat”
Cat Stuff
“Playing” is hunting
Cats’ cognition is not very sophisticated.
They aren’t really capable of planning
They won’t understand your intent behind punishments. Spraying them with water to keep them off the table will simply associate the negativity of being sprayed with you, not with being on the table
Rewards/consequences must immediately follow behaviours (ie. for conditioning)
Cats don’t care that much about you
...and as such, don’t care that much about getting your approval. So training will generally work better with treats than excitement, and punishments generally will not work—they don’t care that you’re displeased.
Cats are control freaks
Comfort comes from security. Places over people; run from danger; predictable routine; social stability.
One should allow them to approach stimuli on their own (reward exploration!)
To help with a new stimuli, can collect your cat’s scent (eg from facial glands) and apply to the new stimuli
As corroborated by previous bulletpoints, punishments “are always counterproductive” (p 18)
Conditioning
Reward desired behaviour (eg treats, play, touch); ignore unwated behaviour
Train very gradually; cats are sensitive
Stimulation: find a sweet spot. Understimulated (disinterested, grooming) → increase reward, remove distractions, switch activities. Overstimulated (meowing, can’t focus) → decrease reward, take a break, switch activities.
‘Mark’ behaviour, eg. by saying “good!". Once the positive association is made with the mark, this gives a flexible way to reinforce behaviours without requiring immediate rewards.
Train frequently and in short sessions. Mark the end of a session (eg, say “done” and begin packing up). Else, your cat may be frustrated when rewards do not continue to come with behaviour
Reward schedules:
When first training a new skill, reward it every time (continuous reinforcement)
Once it’s established, stop and switch to rewarding it intermittently, with varying reward levels and time between behaviour/mark and reward (intermittent reinforcement). Reasons:
A continually-reinforced skill is more quickly lost (subject to extinction), but an intermittently-reinforced skill lasts far longer. Think of slot machine that cashed out 100 times in a row and then stopped—one would quickly stop playing. Compare this to the addictive nature of how slot machines actually work.
A cat will be more engaged with an intermittently-reinforced behaviour. Think how one is excited to check on the shipping status of a package, but would only be mildly so if the progress and delivery were precisely guaranteed.
We can’t always reward behaviour (eg. hands are full)
If a cat expects reward every time, they may become frustrated any time rewards are not produced
Over-rewarding can lead to lethargy (if treats) or exhaustion (if play)
Example applications
Train a behaviour (eg coming when called) by rewarding it
Reward relaxation on a particular blanket → cat associates with relaxation → helps relax at vet, etc
Reward interaction with a scary stimulus (eg, vacuum cleaner) to re-train the negative association. Start very mild and slowly step up intensity → cat no longer scared during vacuuming
Many more possibilities; get funky with it!
Less-important things
“when a cat seems to have gone into a brief trance with its mouth gaping open, it has probably just come accross a scent mark left by another cat." — who knew??