It is a bedrock principle of patent law that an inventor need not understand how or why an invention works. The patent statute simply requires that the inventor explain how to make and use the invention. But explaining how to make and use something without understanding how or why it works yields patents with uninformative disclosures. Their teaching function is limited; one who wants to understand or figure out the underlying scientific principles must turn elsewhere. This limited disclosure rule does not align with the norms of science and tends to make patent documents a less robust form of technical literature. To address this problem, this Article proposes a two-tiered disclosure paradigm. While compliance with the extant statutory disclosure requirements would still be sufficient to obtain a patent, the inventor could opt to provide a mechanistic disclosure—one that describes how and why the invention works. Providing a mechanistic disclosure would have several upsides for the inventor, improve patent (examination) quality, enrich the public storehouse of technical knowledge, and promote broader goals of the patent system.