The recent proposal of a sister-group relationship between the Neoavian grebes (Podicipedidae) and flamingos (Phoenicopteridae) is chronicled, and morphological evidence claimed to be supportive of the grouping is examined. The hypothesis arose from an exiguous amalgam of molecular inferences, advanced in part by a pervasive, unsupported superiority conferred upon sequence data, and adopted by several societal committees on avian classification. Morphological characters marshalled specifically to support the hypothesis were found to be erroneous, and associated phylogenetic analyses, where given, were ambiguous. A combined analysis of large data sets for morphology and RAG-1 sequences found flamingos and storks to be sister groups but with reduced support. This example illustrates problems attending the synthesis of contradictory evidence and evaluation of unprecedented hypotheses, and reveals the informality by which revisions are adopted. Procedures for rational synthesis of evidence are needed for progress during this challenging but promising period of diversified phylogenetics, without which disputes will be dominated increasingly by polarized, intransigent prejudice regarding methods and data.© The Willi Hennig Society 2010.