It's an error made by almost all developmental psychologists, so we shouldn't be too hard on Winston when he occasionally makes it too. The error is the assumption that what a child learns in his home environment is automatically carried along with him to other settings. This assumption is built into most theories of personality development. For example, there are researchers who believe that a child's attachment to his mother in infancy sets the pattern for all his later relationships. If his mother gave him all the love and attention he desired, he'll do well in life because he has learnt to trust people.
But babies don't work that way. A baby is wise enough to understand, almost from birth, that people differ. The fact that his mother treats him well doesn't lead him to expect that his sister or the babysitter will also do so. How other people will act towards him is something he will have to find out for himself, person by person. Researchers have discovered that the babies of mothers suffering from postnatal depression tend to act in a sombre, subdued fashion in the presence of their mothers. But around other familiar caregivers, these babies act quite normally—much more lively and cheerful. Just because Mummy is depressed doesn't mean everyone is depressed. Just because Mummy lets me get away with murder doesn't mean I can act that way in school.
Thanks to Swarup for the pointer.