Although knowledge in a specific discipline resembles a tree, knowledge across disciplines resembles a forest — a latticework of facts and ideas with many crisscrossing connections.
Much of our progress now comes from finding these novel relationships between subjects. As I recently wrote, Claude Shannon did this when he paired Boolean alegra with circuitry design, which laid the foundation for modern computers.
And Shannon's cross-disciplinary breakthrough is not an anomaly. David Epstein in his book Range: Why Generalists Triumph in a Specialized World, writes, "Fifteen years after publication, studies that made multiple new knowledge combinations were way more likely to be in the top 1 percent of most-cited papers."
Even though those same papers tended to be published in less prestigous journals and take longer to garner attention, they ultimately became more influential.
Intellectual innovation, therefore, often comes from remixes of existing knowledge, rather than completely new discoveries.