We don't know when plate tectonics began

Geologists have speculated when plate tectonics began since the dawn of that revolution, with recent estimates ranging from the last 20% of Earth history to within the first 5%. All such estimates rest on six types of evidence: preservation of modern plate-tectonic features, detrital zircon age spectra, trace element and isotope geochemistry, atmosphere–crust–mantle exchange, palaeomagnetism and model calculations. ‘Burke's Law’ suggests that the null hypothesis puts the onus on demonstrating when plate tectonics was not operating rather than requiring observation of plate-tectonic features in the geological record, as assemblages presumed characteristic of plate tectonics could be lost from the geological record (owing to preservation biases) or never existed (owing to secular changes). This issue has become salient as our community appears to be coalescing around the paradigm that plate tectonics began during the late Archean in the face of problematic evidence, in much the same way it has done repeatedly over the past century (e.g. rejecting continental drift but accepting geosyncline theory). I summarize evidence from the six evidentiary types and conclude that we do not know when plate tectonics initiated. Claiming we do forestalls the day when we might truly understand an event at the heart of Earth history and habitability.