Os limites do consenso
Sobre uma das famosas contribuições de Robert Aumann, e suas implicações para o debate (entre pessoas racionais, entenda-se), escrevi esta crónica para o Diário Económico de hoje. Passo a arquivar estes escritos num novo caderno de aponTaMentos.
posted by Tiago Mendes at
2:43 da tarde
7 Comments:
ve o comment que pus no DE. A razao nao se funda a si propria (isso e um mito do iluminismo racionalista) mas e ela mesma dependente de algo que nao e nem racional nem irracional: uma visao de mundo dentro da qual decorre uma argumentacao...Lembra-te do Wittgenstein
By
Joao Galamba, at 11:36 da tarde
Deixo-te aqui um pequeno excerto sobre o fundacionalismo (o teu argumento depende da verdade do fundacionalismo)
Traditional "foundationalism" was the view that knowledge could be started, or started again, from nothing by finding pieces of certain and infallible knowledge, the "foundation," upon which all other knowledge could be constructed. The classic attempt to do this occurred with René Descartes (1596-1650), who believed that if he could conceive anything "clearly and distinctly," he then could rely upon it as being true and build the rest of knowledge on it. This became the pattern with the Rationalists, who based their systems on what they saw as self-evident first principles of demonstration, as these had originally been conceived by Aristotle. Something counted as "self-evident" if one knew it was true simply by understanding it. The Rationalist project, however, suffered from the disability that "self-evidence" was a subjective claim of certainty, which meant that different Rationalists could regard different things as being self-evident, with no way to rationally resolve the dispute. Hence, making such claims, the systems of Descartes, Spinoza, Leibniz, etc. produced very different results.
A similar foundationalism, however, occurred with the Empiricist opponents of Rationalism. The Empiricists also claimed a few things to be self-evidently true (Hume even put geometry in that category), but mostly they regarded experience as providing foundational pieces of knowledge. Statements about experience were not self-evident in terms of being understood, but they could be grasped as intuitively true as part of empirical observation. This kind of foundationalism came to a kind of grief much like the Rationalistic self-evident truths, since it turned out that disagreements and apparent mistakes could occur even in the course of direct empirical observation: The "foundational" pieces of knowledge were neither certain nor infallible. A similar difficulty would afflict any kind of "Intuitionism," which would not necessarily regard any truths as self-evident, but could regard pieces of knowledge based on "intuitions," whether empirical, rational, or otherwise, as foundational.
That no item of knowledge could be regarded as infallible or incorrigible, i.e. all knowledge can be mistaken and can be improved, has been taken as decisive disproof of foundationalism. As far as it goes, this is an inescapable conclusion. Whether there was anything insightful about foundationalism, however, must be determined once we see what the alternative has turned out to be. A certain alternative, indeed, has become all but dominant, not so much in philosophy, though it is powerful there, but in the popularized epistemology that we find in English departments and other areas of intellectual life that are liable to seize upon the "latest thing" as, indeed, self-evidently true.
If foundationalism is the thesis that we can construct knowledge with absolute certainty starting from nothing, then the denial of this can give us various possible theses: (1) knowledge cannot be constructed, (2) there is no absolute certainty, and (3) knowledge cannot be started from nothing. The first thesis gives us the idea of "deconstruction" to describe and symbolize the failure of foundationalistic projects. The second thesis, that we cannot have absolute certainty, is now accepted by all but everyone outside a few Aristotelians. But the third thesis is the best clue to an alternative theory: If knowledge cannot be started from nothing, what does it start with? With previous knowledge, of course. But what is to count as previous knowledge? Why, just whatever it was that we thought we knew before whatever happened that changed our minds. And if there is no certainty to knowledge, and no permanent, fixed system can be constructed, then the new knowledge will be what we think we know until something else happens to change our minds again.
This process ends up being described by the "hermeneutic cycle." "Hermeneutics," from Greek hermêneuô, "to interpret or translate" (from the messenger of the gods, Hermes), is the theory and practice of interpretation, originally the interpretation of texts, especially religious texts. The "hermeneutic cycle" is the process by which we return to a text, or to the world, and derive a new interpretation -- perhaps a new interpretation every time, or a new one for every interpreter. It is clear that this happens all the time. We can understand a book, a movie, etc. a little differently each time we read or see it.
This was serious business in the Middle Ages, when differing interpretations of scripture could produce heresies, schisms, persecutions, wars, etc. Some issues that emerged were the questions of who had the authority to interpret scriptures and of whether interpretation was something that anybody could do or if it required particular abilities or inspiration. Thus, one difference between Orthodox Islam and Shi'ism was the view in the former that, originally, virtually any Moslem could interpret the Qur'ân and the Traditions, while the latter held that the proper interpretation could only be given by one possessing the divine spark that descended from 'Alî, the son-in-law of the Prophet Muhammad. The Catholic Church is famous for the doctrinal authority of the Pope, but earlier Christianity invested doctrinal authority in Church Councils.
In this century, interest in hermeneutics grew steadily in Continental philosophy without much notice in the Anglo-American world, until Thomas Kuhn published The Structure of Scientific Revolutions in 1962. As the (empiricist) foundationalism of Logical Positivism was tottering, Kuhn provided a powerful alternative vision: Scientific knowledge changed, not through confrontation with the hard facts, but by a social struggle between contending interpretations of intrinsically ambiguous evidence. As Logical Positivism actually then did collapse, the full force of the alternative burst through with Feyeraband, Habermas, Derrida, Foucault, etc. etc. While The Structure of Scientific Revolutions hardly mentioned truth as a concern of science, many of the new hermeneuticists positively rejected the possibility of any kind of objective truth. All was interpretation. "Reality" is only accessible to us in terms of how we understand and interpret it. Thus, if there is no "reality" to be independently compared with our knowledge, all we can do is oppose one interpretation to another, and each of these is ultimately going to be as well motivated by the "facts" as any other. There are no foundational pieces of knowledge.
By
Joao Galamba, at 11:40 da tarde
João: acho que percebo o teu ponto, e discordo, por uma razão simples. O argumento do Aumanna é "lógico-processual" e não relativo à "verdade". Ou seja, não é necessário que os "prior beliefs" estejam certos ou errados. A questão é que se duas pessoas partilharem esses prior, e se depois "engagarem" numa discussão, não podem discordar.
Há uma questão muito séria que tem a ver com a forma como as pessoas formam esses priors. Aí sim entra a tua crítica. Mas isso não abala o resultado que eu procurei trocar por miúdos. Neste ponto, acho que consigo concordar com a tua visão wittgensteiniana da (não) realidade, embora não seja um incondicional como tu (ou não tanto).
Há outros papers interessantes sobre estas questões epistémicas que quero referir daqui a uns dia, quando tiver tempo.
PS1: ainda não está online o teu comment, amanhã logo verei e actualizarei aqui se necessário.
PS2: we miss reading your thoughts, my friend. Bem sei que o tempo é escasso...
By
Tiago Mendes, at 12:13 da manhã
Esses priors nao sao beliefs (tipo uma lista de assumptions). Isso ainda e a mente racionalista a trabalhar. E um background holistico que constitui a nossa forma de estar no mundo. Nao e formalizavel nem representavel. Tanto o Wittgenstein como o Heidegger falam disto. O Heidegger chama-lhe Being-in-the-world which is that on the basis of which we can have beliefs and so it is not a belief or a set of beliefs itself...
E voltamos ao mesmo de sempre...
By
Joao Galamba, at 12:33 da manhã
Mas o artigo esta interessante. E pa nao tenho tido tempo para escrever no blog. So escrevi para bater no JM...
Abracos,
Joao
By
Joao Galamba, at 12:35 da manhã
Tambem deixei uma pequena provocacao no sinedrio sobre a questoa do agnosticismo...
By
Joao Galamba, at 12:36 da manhã
João, eu percebo o que dizes, mas acho que tu não percebes bem o resultado do Aumann, sinceramente. Ele não fala de "conhecer o mundo", "the world outhere". Quando tu escreves (no DE):
"A lei da contradicao nao chega para determinar o conteudo substantivo de uma linguagem, logo duas pessoas podem nunca chegar a acordo sem que uma esteja necessariamente em contradicao."
Isto revela que não percebeste o resultado. É claro que as pessoas podem discordar. O ponto do Aumanna é que quando elas tenham "priors" iguais, e as suas opiniões "a posteriori" sejam mutuamente partilhadas, não podem discordar. Agora, esses priors não tèm necessariamente que ser "racionalistas". Podem ser idênticos "being-in-the-world", ou outra coisa qualquer. O que interessa é que são uma certa "visão do mundo" à prior, que é partilhada. Os mesmos valores, o que quer que lhe queiras chamar.
"Lembra-te que se estiveres a discutir com uma pessoa que tenha conceitos diferentes dos teus (em que o significado de determinados conceitos e diferente do teu) a possibilidade de acordo e limitada."
Claro que sim, João. O meu ponto era que, neste caso, elas deviam "concordar em discordar". Isto é, que as diferenças "a prior" deviam ser mutuamente entendidas de modo a que cada um pudesse "adivinhar" e "concordar" com a opinião "a psoteriori" do outro, depois de discutida a evidencia e partilhadas as ideias.
"Discutir com Aristoteles biologia moderna implicaria uma mudanca na concepcao do mundo que ele tinha. Essa mudanca nao seria apenas via argumento, mas algo mais forte: implicaria que ele se transformasse e adquirisse uma nova linguagem..."
Sinceramente, o problema da linguagem e da perspectiva do Wittgenstein, não entra aqui, pelo menos não de forma importante.
Há um problema mais subtil nesta questão toda e que tem a ver com o 2º parágrafo do artigo que escrevi:
"Are people Bayesian?" Isto é, será que o "Bayesian updating" é constitutivo da racionalidade? Esta questáo é complexa e algo técnica e prefiro não me alongar, até por questões de tempo.
Mas sõ para terminar: o probelma não é as pessoas discordarem. Bolas, a 3ª frase do 1º parágrafo diz isso mesmo. O problema é a inconsistência das opiniões.
By
Tiago Mendes, at 12:04 da tarde
Enviar um comentário
<< Home