Few questions in science are as controversial as the origins of knowledge. Whether knowledge (e.g., “objects are cohesive”) is partly innate has been debated for centuries. Here, we ask whether our difficulties with innate knowledge could be grounded in human cognition itself. In eight experiments, we compared reasoning about the innateness of traits that capture knowledge (cognitive traits) with noncognitive (sensorimotor and emotive) traits. Experiments 1–4 examined adult and infant traits; Experiment 5 presented detailed descriptions of published infant experiments. Results showed that people viewed cognitive traits as less likely to be innate in humans—the stronger the association with “thinking,” the lower the rating for “innateness.” Experiments 6–8 explored human, bird, and alien traits that were presented as innate. Participants, however, still considered cognitive traits as less likely to emerge spontaneously (i.e., be innate). These results show that people are selectively biased in reasoning about the origins of knowledge.