This is a response (a reply? a riposte?) to Peter Griffiths (Int. J. Nurs. Stud., in press, doi:10.1016/j.ijnurstu.2004.11.004) that attempts to answer some of the charges, which he levels at our book 'Deconstructing Evidence-based Practice' (Routledge, London, 2004). It begins by countering Griffiths' mistaken assertion that, in deconstruction, 'anything goes'. It argues that Griffiths is wrong because he has literally mistaken the meanings of certain words; that is, he has taken them wrongly. His biggest mistake, on which all of his others rests, is to mistake the word 'deconstruction' to imply a form of extreme relativism in which there are no right or wrong readings. In this, he is simply wrong. He is wrong in his assumption that there are no wrong readings, and the fact (yes, fact) that he is wrong demonstrates that some readings can be wrong. In particular, he mistakes the word 'challenge' to mean 'deny', and the word 'authority' to mean 'legitimacy'. This is not simply our reading of what we took him to mean (which could, by our own argument, be mistaken). It is his reading and his writing, there on the page in black and white. And this misreading, this mistake, inevitably leads him to a wrong conclusion. Having clarified the small matter that, in deconstruction, anything does not go, and that deconstructionists are not constrained to accept everything that is written about them, we then attempt to point out some other mistakes in Griffiths' non-review. Most importantly, we reject Griffiths' accusation that postmodernism is a strategy to 'save us from thinking' and instead, with Lyotard, advocate it as an attempt 'to save the honour of thinking'.