According to the theory of natural pedagogy, humans have a set of cognitive adaptations specialized for transmitting and receiving knowledge through teaching; young children can acquire generalizable knowledge from ostensive signals even in a single interaction, and adults also actively teach young children. In this article, we critically examine the theory and argue that ostensive signals do not always allow children to learn generalizable knowledge more efficiently, and that the empirical evidence provided in favor of the theory of natural pedagogy (...) does not defend the theory as presented, nor does it support a weakened version of the theory. We argue that these problems arise because the theory of natural pedagogy is grounded in a misguided assumption, namely that learning about the world and learning about people are two distinct and independent processes. If, on the other hand, we see the processes as interrelated, then we have a better explanation for the empirical evidence. (shrink)