I would like to thank Callard, Holmes and Solms for taking the time to discuss my previous paper. ^{Reference Ramus1} I actually find little in their commentaries that is not already addressed in the target article. I will therefore focus on a few points.

According to Callard, my first point reflects ‘the ‘'classic manoeuvre” by those opposing Freud’, that ‘is to argue that if what he says is right, he stole it from somewhere else’. In my view, whether the manoeuvre is ‘classic’ matters little compared with whether it is well founded. Whether broadly accepted Freudian ideas about the existence of unconscious processing, unconscious motives, conflicts between desires, reason and society's constraints, etc. were Freud's ‘discovery’ or originated from earlier thinkers is simply an empirical matter that can be decided by checking the works of Janet, Galton, Charcot, Krafft-Ebing, Schopenhauer, Nietzsche and others. And whether his more original contributions to these ideas (e.g. the Oedipus complex) have any validity is also an empirical matter.

Holmes provides a nice illustration to my second point. The studies by Strathearn et al ^{Reference Strathearn, Fonagy, Amico and Montague2} and Coan et al ^{Reference Coan, Schaeffer and Davidson3} are perfectly well understood using mainstream psychological concepts such as attachment, which have nothing to do with psychoanalysis. It is indeed ironical that, although John Bowlby was trained as a psychoanalyst, he found psychoanalytical concepts so inadequate to explain his observations that he had to develop an entirely independent theoretical framework for attachment, based on up-to-date knowledge in ethology and the various areas of cognitive science, and that turned out to be rejected by the psychoanalytic community of the time. ^{Reference Bowlby4} It is of course perfectly fine for contemporary psychoanalysts to now admit the errors of their predecessors and embrace attachment theory. However, adding another layer of psychoanalytical concepts to an already functioning theory would really need to increase explanatory power in order to remain parsimonious. Merely finding ‘consistencies’, as Carhart-Harris et al ^{Reference Carhart-Harris, Mayberg, Malizia and Nutt5} attempt to do in their review, adds little. It is also fine, as Solms proposes, to attempt to ‘finish the job’ and test hypotheses inspired from Freud's writings. What matters is whether these hypotheses are better empirically supported than competing ones, not whether they seem ‘coherent’ or ‘intellectually satisfying’ to some.

Finally, I entirely agree with Holmes and Solms that the French psychoanalytically inspired treatment of autism does not by itself justify rejecting psychoanalysis as a whole. This was indeed not meant as a definitive condemnation, but rather as an illustration of the unfortunate side-effects of uncritical Freudism (and Lacanism, for that matter). It remains troubling, though, that despite neuropsychoanalysts' admirable ambitions, when one takes a worldwide perspective, psychoanalysis seems to be the main factor of resistance against evidence-based psychology and psychiatry.