According to Steven Lubet, documentation freezes facts in time. For this reason, it is, he thinks, more reliable than other kinds of evidence. That makes it especially useful for fact checking. However, this is not a realistic way to read for evidence in documents. It ignores or downplays a crucial fact that inheres to all paperwork, which ethnographers have long heeded: documents do not report on facts; they render them. Ethnographically informed readings of paperwork that attend both to facts and to how they are rendered are, by contrast, realistic. In this contribution, I discuss why and illustrate with reference to two studies of policing and law in Thailand.