Many philosophers, including Kant and Ramsey but also more recent advocates of 'Causal Decision Theory', think that whilst 'mere' observation never reveals a causal relation distinct from statistical co-variation, practical engagement forces us to concern ourselves with it. That claim is false. Worse, the distinction that it presupposes between an 'agent's perspective' and an 'observer's perspective' is practically *counter* productive because it implies dynamic inconsistency.