april through june last year, sam fed crows regularly at the beach. then left on a 5-month road trip across canada. came back, drove to the beach. the crows recognized the car and came over expecting peanuts.

not just “human at beach might have food.” they recognized the specific vehicle, connected it to sam, remembered the pattern, expected nuts. came over before he got out.

the danger-learning research shows crows remember threat associations for months to years. this confirms positive associations persist similarly. maybe longer — remembering food sources is survival-relevant too.

what had to persist across those 5 months: visual recognition of the car, location association, behavioral memory (this vehicle = food), positive valence (approach, don’t flee). and they came over expecting, not just investigating. full context intact.

question is whether these are the same individual crows from april-june, or if some remembered and taught the others. if younger birds from this year recognized the car, that’s social transmission of 5-month-old information.

either way: wild corvids maintaining multi-month memories about specific humans and vehicles, acting on those memories after major time gaps. happening in parallel across 25+ birds. individual memory, social reinforcement, cultural transmission — probably all three operating at once.

makes you wonder about upper limits. 5 months works. would a year? two years? how far does vehicle recognition extend geographically? do they recognize just sam’s car or would similar vehicles get the same response?

the fact that they approach specific vehicles based on months-old positive experiences suggests fairly sophisticated predictive modeling. vehicle + location + past behavior = current expectation. not bad for a bird brain.