The discussion to follow was part of my master's degree dissertation, entitled "For a discussion of the theoretical contributions about cognition: the emergence of the metaphysics of subjectivity in the a priori historical of the XVII century." In that dissertation, I tried to rebuild a historical panorama to give a more concrete sense to notions such as evidence, mind, representation, sensation, sign, subject, which are almost always used when we thought about cognitive processes. In my opinion, this was an important goal because these notions sometimes give rise to some endless polemical issues, justly because we use them as if their meanings could be disembodied from specific human discursive practices in which they are linked to each other and acquire intelligibility (1). For reconstruction of such an historical panorama to give a more concrete sense to these notions, I drew upon mainly in some works of Michael Foucault ( Les Motes et les Choses ) and Ian Hacking ( The Emergence of Probability ). In these two books, both authors have work methodologies very similar, which can be called archeology and historical meta-epistemology, or even historical ontology.

The investigative perspectives employed by Foucault and Hacking in books such as Les Motes and les Choses and The Emergence of Probability , could be said to consist in a research of historical modifications in the space of possibilities of thought organization. We are to understand this space as an historical a priori : a region of dispersion of discourses, texts, enunciations (spoken or written), graphics, lists, schemas in a given context of human practices. In this region, the archeologist and the historical meta-epistemologist draw multiple and often-discontinuous displays and irruptions, which are masked by hurried reflections under the shape of explanative unities.

The aim of the present discussion is to present investigative perspectives such as archeology, historical meta-epistemology and historical ontology as research approaches that avoid transcendental postulations, and also withdraw from some problems that immobilize, in the past and today, much of the research done in Philosophy and in the Social Human Sciences. On the one hand, these problems have to do with the foundationalist imperative of certain very traditional epistemological perspectives that, proposing general theories of cognition and rationality, try desperately to abruptly break, at some or other point and once for all, the dynamics of our discursive practices. The label could apply to rationalists, empiricists, logical positivists, puritan popperians, etc. On the other hand, these problems have to do with the naïve posture of certain constructivist perspectives that, presenting themselves as antithesis of those very traditional epistemological perspectives, try desperately to empty the dynamics of our discursive practices, and look for explanations only in terms of power relations between given social subjects (2).

However, I would like to remember that in the Preface of his most recent book, Historical Ontology, Hacking says that

People sometimes take me to advocate the right methodology for philosophy in our times. Nothing could be farther from the case. There are many more ways for a philosopher to use history than I can imagine, and Foucault is an almost endless source of inspiration for people whose interest and abilities are different from mine (2002, p. v).

Hacking says another things like that also in the end of the first chapter of Historical Ontology : "I would never advocate any program as anything more than one way . Some epistemology - of a totally ahistorical sort - seems to me fascinating and certainly important." (Hacking, 2002, p. 25). In the light of these passages of Hacking himself, the present analysis could appear a little pretentious, a suspicious pamphlet! But I'm not saying that historical ontology or archeology, as I understand them, are necessarily the best and only one possible way to do philosophy in our times. What I'm saying is that historical ontology or archeology are ways to do philosophy in our times that do not suffer from certain compulsions and naiveties that are characteristic of other well known ways to do philosophy (and also to think in the Social and Human Sciences) we have at hand. But historical ontology or archeology are not necessarily immune to other diseases. Maybe if historical ontology or archeology were the leading work methodologies in my quarters, I would be here defending another perspectives that could overrun them. This is a kind of strategy used by an author that I think Hacking also admires and I will refers in this text only here: Paul Feyerabend.


(1) An instance of these endless polemical issues is the discussion around the concept of intentionality in AI debates: is the intentionality of our statements something present in the heads of the individual human beings, in such a way that no computer, no matter how much it dominate one of our languages, can be said to really understand our statements? There are people who would answer yes to that question, people who would align themselves with Searle (see his famous Chinese room argument). There are other people who would answer no to that question, people who would align themselves with Turing and his followers (see Douglas Hofstadter and his hypothesis of an artificial intelligence emerging as an epiphenomenon from a basis of computational units: Metamagical Themas: questing for the essence of mind and pattern , New York, Basic Books, 1985, pp. 636-646). It could be said that all the controversy can be understood as a legitimate insoluble sticking point that emerges from one determinate historical configuration of our discourses, practices and institutions. (I borrowed the expression 'sticking point' from Hacking 1999b, but the reader should be aware that in that book Hacking have used this expression in another - related - sense. He uses that expression referring to philosophical controversies about science, not to science controversies.) Maybe, I would also be inclined to add that, in my opinion, the people who answer negatively to the question stated above give a more plausible answer in view of that historical configuration, but this do not means that I believe the others are really wronged in giving a positive answer (in the light of the same historical configuration, which is the only Achilles' heel. oh no! I mean, the only Archimedean point we have for these meditations).

(2) As is stressed by Hacking (2004, p. 3) and others (Ruse, 1995), Foucault emphasizes in his later books that we should try to understand power not as something exerted upon subjects, but as anonymous arrangements in which we participate constituting ourselves. The constructivist perspectives that I'm criticizing here do not incorporate this lesson of the later Foucault, but keep talking confusedly about 'genealogy', as is common at least in some post-graduate circles in the Social and Human Sciences in Brazil. I have to apologize for not giving any concrete instance, but this kind of talking is really so confuse that it seldom goes to the print.

*Este trabalho foi apresentado no simpósio "Husserl and the historical a priori of the sciences", o qual ocorreu no Instituto Max Planck, em Berlim, de 1 a 3 de julho de 2004.