Oi Valéria, oi Eduardo!
Muito obrigado por mostrarem interesse na entrevista...
O Danilo Lawvere me disse que assim que puder, vai trabalhar na edição do texto (inserir referências para as obras citadas, corrigir uma ou outra imprecisão).
Mas talvez ele demore um pouco agora, porque já estava assoberbado de trabalho e mãe dele, Fatima Fenaroli, faleceu há poucas semanas... Ele ficou sozinho com a tarefa de administrar a página, presumo:
Em 2017 eu era bem mais ignorante do que sou hoje, então, peço alguma condescendência.
Preciso apenas dizer algumas palavras como preâmbulo, para que as perguntas tenham sentido.
1. Lógica
A teoria de categorias tem uma relação muito sofisticada com a lógica... Fiz uma pergunta sobre essa relação, mas, surpreendentemente, o Lawvere desviou dela e disse que gostaria de ter tido a oportunidade de estudar mais o Aristóteles...
Qual o problema aqui e por que ele desviou do assunto?
Minha opinião: Goldblatt.
Goldblatt é um escritor e expositor muito talentoso. Ele escreveu uma obra chamada Topoi: A Categorial Analysis of Logic.
Neste livro, na opinião de Lawvere, aconteceu uma séria distorção do que seria de fato a topos theory. Essa distorção poderia levar um leitor desprevenido a não pegar o ponto, a se desviar do alvo, a sair com uma visão errada do que é a topos theory e de quais são as relações entre lógica e teoria de categorias.
Parte das reações do Lawvere a esse livro estão nos e-mails que ele enviou para a lista de categorias (análoga dessa nossa lista de lógica, mas de âmbito internacional e que está no ar desde o final da década de 1990).
Uma forma meio grosseira de tentar expor o nó é dizer, como o Peter Freyd, que "teoria de categorias é teoria de funtores". Ou seja, precisamos de funtores para trabalhar. Onde não existem funtores e transformações naturais, não existe teoria de categorias. O objeto básico são os funtores. As relações visadas são as transformações naturais, que são relações entre funtores.
Um funtor é um processo de interpretação, que transporta problemas de uma área para outra. Uma transformação natural é algo como um dicionário que me permite relacionar processos de interpretação.
Teoria de categorias é uma disciplina que estuda um tipo especial de relação entre processos de interpretação, relações que nos parecem naturais ou canônicas...
Se você escreve um livro sobre teoria de categorias onde os processos de interpretação e as traduções entre eles não são o objeto fundamental do estudo, então você não está fazendo teoria de categorias.
Então, Lawvere e seus seguidores mais próximos opinavam que o livro do Goldblatt é na verdade lógica e teoria de modelos clássicas com roupagem categorial, vocabulário categorial. A propósito, os funtores só aparecem na segunda metade do livro.
Só que essa abordagem do Goldblatt fez um enorme sucesso, pois de fato o livro é muito bem escrito. Por exemplo, ela influencia claramente o Peter Smith, cujo livro introdutório de teoria de categorias também só apresenta os funtores no final.
Perde-se algo com essa distorção.
Outra coisa: a teoria dos topos originalmente surge em estudos sobre física: "da física para a física", diz o Lawvere num ensaio sobre a origem da teoria dos topos. Ela não é a análise da lógica...
Mas a parte realmente técnica da crítica ao Goldblatt ficou a cargo do Philip J. Scott, coautor ao lado do Joachim Lambek daquele livro clássico "Introduction to Higher Order Categorical Logic". Este Philip Scott escreveu uma resenha do livro do Goldblatt apontando exatamente onde o enfoque enviesado leva a problemas. É uma resenha bem técnica.
Mas, então, afinal, onde entra a lógica?
Isto era o que eu gostaria que o Lawvere respondesse, mas ele evitou entrar no assunto.
Me parece que ele se sentia bastante incomodado com as confusões que existem sobre essa questão no senso comum...
Eu tenho uma opinião, que eu achei no Francisco Miró Quesada. Ele achava que a teoria dos topos, ao mostrar que em cada universo aparece uma lógica diferente ---isto é, clássica quando o topos é booleano; não clássica caso contrário (em geral multivalorada, sem terceiro excluído)--- teria confirmado a intuição de Hegel de que a lógica é conteúdo-dependente. Ao meu ver, confirma a opinião de Brouwer segundo a qual a matemática não depende da lógica; a lógica é que depende da matemática.
Miró Quesada opinava que a topos theory teria inaugurado uma grande revolução no âmbito da lógica e da filosofia da matemática. Ela nos levaria, por exemplo, a abandonar a ideia de "lógica subjacente". Isso estaria meio inadequado. O que realmente existe é matemática subjacente com lógicas adaptadas ao conteúdo, mas jamais uma lógica pairando acima dos conteúdos.
(O Miró Quesada diz tudo isso num artigo que saiu numa revista onde o Lawvere também publicou um artigo... Então, ele conhecia a discussão, mas não quis entrar nela...)
2. Teoria do Conhecimento
Filosofia da Matemática não é Matemática. História da Matemática não é Matemática. Crítica literária não é literatura.
Haveria alguma chance de epistemologia da matemática ser uma disciplina matemática e não apenas filosófica?
A ideia não é minha, é do próprio Lawvere e ele escreveu sobre isso várias vezes em diferentes momentos de sua vida.
O ponto alto dessas investigações parece ter ocorrido num momento em que havia um grupo de trabalho coordenado no Canadá por John Macnamara, mas ele faleceu em 1996, pouco depois de os primeiros frutos terem sido publicados... Aqui Lawvere mostra o quanto ele apreciava esse trabalho interdisciplinar e o quanto acreditava no trabalho em equipe. Na entrevista ele apenas diz: "me dediquei durante algum tempo sobre isso". Acho que sem o Macnamara ele desistiu de continuar.
3. A Matemática Conceitual
O livro "Conceptual Mathematics" é o registro de um experimento didático e também é o lançamento de um programa: ele contém um esboço de uma filosofia da matemática e o ensino é o laboratório. Várias perguntas giram em torno desse livro.
Bem, era isso.
Aí vão as perguntas e as respostas (desculpem o meu inglês do Google Tradutor):
-×-×-×-
--
Interview with F. William Lawvere. Conceptual Mathematics: Theory of Knowledge and Category Theory
1. There seems to be a fundamental idea underlying the structure of the book "Conceptual Mathematics": revealing or exploring the connection between teaching (in particular, teaching at a fundamental level or at the most elementary level possible) and the development of knowledge and scientific research...
What are, in general lines, the epistemological or philosophical ideas that guided the realization of this project?
LAWVERE: My late friend and collaborator Stephen Schanuel shared with me a belief that in the modern world, there needs to be a broader and deeper knowledge of mathematics. Both clarifications and mystifications are often accompanied by a mathematical background, but only with some knowledge of mathematics can someone begin to distinguish between the two. Our passion for teaching became known and a representative of a publishing company invited us to write a book on ‘Discrete Mathematics’. Steve replied ‘No, we will explain methods that apply to both, continuous and discrete mathematics’.
Important for us was the realization that new scientific developments are often responses to questions of students.
2. In what circumstances were the sessions (the classes) held? The classes took place in what year? For how long? In which school or institution? How were students selected?
Lawvere: THE STATE UNIVERSITY OF NEW YORK at Buffalo agreed to the unusual arrangement whereby two professors collaborated in teaching one course. The course was designed as a beginners’ course for second year undergraduates. The students were nurses, physicists, engineers, and psychologists, et cetera (who had no idea what they were getting into when they chose this course). My wife Fatima and my son Danilo also participated in it and contributed to the discussion, parts of which are recorded in the book. My former student Emilio Faro took accurate notes that became part of the basis for the book. My collaborator and I did not always discuss what we were going to say in the next lecture before delivering it, so there were some surprises, for example when Steve devised the intuitive account of Brower’s fixed point theorem and immediately presented it in his lecture.
The first publication of the record of the course was published by our own Buffalo Workshop Press in 1991, and an Italian translation, and improvement based on
Aurelio Carboni’s Milano elaboration, was published by Franco Muzzio in 1994.
3. The examples that appear in the book are very impressive, because they revalue how mathematical thinking materializes, that is, how mathematical thinking interprets real-world events. In particular, we would like to know how the ideas present in the example of lunch in the Chinese restaurant appeared: did a restaurant really exist or did it not exist? And the example of the museum director? How did these amazing ideas happen to you?
LAWVERE: We were coming up the stairs to go to the class and at the landing I said to my wife ‘I have nothing to say’. But then I thought of the Chinese Restaurant, where we used to go after the seminars in New York City, and of the mathematical use of stacks and fibers of maps to represent families. From these thoughts I produced the lecture for the day. (I later learned that this particular style of Chinese Restaurant is well known.)
The example of the Museum director was encouraged by the late John Macnamara, a psychologist whose later work explicitly took up the application of category theory.
(I published some papers in books dedicated to him.)
4. What part did each author have in the preparation of the book and in the execution of the classes? Or rather: what was the role of your friend Stephen Schanuel in this venture?
LAWVERE: The preparation of the book: I wrote a draft of a small part, Steve put in his ideas; Steve wrote a part, I put in my ideas. Fatima corrected and simplified. In the end you cannot always tell which parts where written by whom.
5. One aspect of this work that has gradually awakened the attention of philosophers is the connection with certain ideas of Hegel, such as the relation between the "subjective" and the "objective." What studies or what problems of science have led you to discover in these ideas a guide for the construction of research instruments?
LAWVERE: All problems of science that I have had the strength to study seem to lead me to these ideas. The works of Marx and Lenin made unique contributions to the on-going development of dialectical materialism. I still need to study more the work of Hegel, as well as its clarification by Lenin. Looking back I realize that I often worked in a dialectical manner, but the specific mathematical connection with ‘Aufhebung’ et cetera in the lattice of subtoposes of a topos came later.
Probably the Mathematician most interesting as a self-confessed dialectician is Hermann Grassmann, whom I began to analyze in the book by Gerd Schubring, published by the late Robert S. Cohen, recording a meeting on Insel Ruegen, in Germany in honor of the 150th anniversary of the work of Grassmann.
6. Does category theory provide any new light for the interpretation of non-classical logics in the context of the theory of knowledge? Should the connstruction of a theory of knowledge necessarily be based on non-classical logics?
LAWVERE: I am not sure what you mean by non-classical logic. I would very much like to learn more of Aristotle’s logic, wherein some scholars claim that already both Heyting and co-Heyting logic are considered, with Boolean logic being offered as a first level of study. The modern interpretation, made more explicit with the help of category theory, is that there is an important contrast between variable and constant, between cohesive and discrete. Of course, these aspects were long known, but we can now create more precise models of them, bringing new light.
7. There is an emphasis in the book "Conceptual Mathematics" in the construction of small subjective categories, which, once dominated, become operative as research instruments of the larger objective categories. How can this notion be generalized?
LAWVERE: One relevant tool is the notion of natural transformation, specifically between functors that naturally arise in the investigation. An important case of this is called ‘algebraic structure’ in my thesis which can be read in TAC Reprints, a part of the on-line journal ‘Theory and Application of Categories’. The concept had been explicit earlier in the special case of ‘cohomology operations’.
8. Session 6 of the book seems to contain the practically finished program of a theory of knowledge, in particular a theory of logical-mathematical knowledge. In what way, in your opinion, does Category Theory relate to General Theory of Knowledge? Is the Category Theory the Knowledge Theory itself of our time?
LAWVERE: These questions should be the focus of ongoing research, avoiding the temptation of facile answers. I have been thinking of these matters for some years but still the program is not nearly finished, needing contributions from other researchers.
9. In your opinion, should there be a reformulation of the curriculum of the higher courses of mathematics putting the theory of categories as a compulsory subject? Should it be taught before or after traditional courses in abstract algebra?
LAWVERE: This problem requires serious consideration. Abstract algebra and general topology can be seen to arise naturally in the dialectic of category theory, (see Appendix 1 of Conceptual Mathematics) and many find that such an ordering is congenial and effective.
It is difficult to make such a program ‘compulsory’, because many mathematicians believe that the way they first learned functional analysis, algebraic geometry, set theory, or logic, is the only way.
According to the late Pierre Gabriel, a mathematician from whom I learned much, it was the specialists in various branches of abstract algebra that most eagerly took up the categorical method. In my case, it was the 1959 actual working out of some results in the study of functional analysis that called for the more explicit use of categories (which I had seen before, but dismissed as too abstract).
10. What advice would you give to young math students who have just come into contact with the Theory of Categories?
LAWVERE: They should peruse such works as my thesis, Mac Lane and Moerdijk’s book on sheaves in geometry and logic,
Anders Kock’s book on Synthetic Differential Geometry,
Adamek, Rosicky, and Vitale on Algebra, et cetera,
in order to determine which of these works they would like to first study more systematically.
11. What projects are you currently working on?
LAWVERE: The consolidation of my several works on proof theory, which are still not properly integrated by theoretical computer scientists; the further refinement of my unified theory of finite and infinite-dimensional smooth dynamical systems, including a new realization of Newton’s definition of force as used in principle by mechanical engineers.
The End.
-x-x-x-