To probe this, we randomly sampled 1 million word pairs (4 days processing on 32 cores), to get a strong statistical sampling of the connected core of English. 1,525,522 headwordsWe built a semantic network of 1.5 million English terms by casting a wider net than traditional resources. This doesn’t affect puzzles—which start from common words—but reveals an interesting property of the semantic network. Five Data SourcesThe Linguabase integrates five complementary knowledge sources into a unified semantic network. Understanding Our BiasesEvery semantic network encodes particular worldviews about which words relate to each other and how strongly they connect.