(Written by ChatGPT as a summary of a dialogue we had on this subject)
Where is human knowledge to be found?
The question sounds simple, but if we take it seriously it is not at all easy to answer. We can point to libraries, universities, encyclopaedias, books, scholarly journals, archives, and today an endless quantity of resources available online. But none of these places properly contains human knowledge. They contain fragments of knowledge, organized according to different and often incompatible criteria.
The problem becomes particularly evident in humanistic knowledge.
To know what sociology thinks, for example, it is not enough to “study sociology.” Which sociology? Which sociologist? Which school? Which textbook? One textbook presents a selection of authors and problems; another presents a different one. One author considers certain phenomena fundamental, while another considers them secondary or irrelevant. One theory interprets society in a certain way; another interprets it differently.
The reader who wants to form an understanding of society must therefore choose. But on what basis?
Here the fundamental problem appears: in order to choose well among the various offerings of knowledge, one would already need to possess a criterion by which to evaluate them. But in order to build that criterion, one would need to know the various offerings and compare them. The beginner is thus caught in a circle: he must know what to read before he can know what is worth reading.
The university solves this problem only in a practical and institutional way. A professor sets a syllabus, selects certain authors, excludes others, and establishes an order of reading. But this does not mean that he has discovered the objective organization of knowledge. He has simply adopted a particular organization.
The same holds for a book. A book may be excellent, rigorous, well documented and deeply instructive. But that does not make it “knowledge.” It is a construction of its author. The author decides what to include and what to leave out, what to place in the foreground and what to relegate to the background, which concepts to use, which authors to discuss and which to ignore, and which relations to establish among the various phenomena.
A book is therefore a narrative of knowledge, not necessarily an objective map of it.
This problem also concerns the disciplines that possess rigorous methods for reconstructing the past. History, philology and archaeology can establish non-arbitrary knowledge: a document contains certain words, an artefact has certain characteristics, an event took place in a certain period, a building belongs to a certain era. In these cases the task consists mainly in recording and reconstructing facts and forms on the basis of the available evidence.
The situation changes when we move from recording to explaining.
To say that A happened before B is one thing. To say that A caused B is another. To say that A was the principal cause of B is yet another.
To establish a causal relation, one has to select, out of an enormous number of conditions, those held to be relevant. One has to determine which are necessary, which sufficient, which secondary. One has to decide which economic, political, cultural, psychological or environmental factors are to be taken into account. It is at this level that interpretation acquires a decisive role.
The same holds for description. No historian can describe everything that happened. A description is necessarily a selection. And every selection implies a criterion.
We can therefore distinguish at least three levels.
The first is the fact, or the record: something happened, a document contains something, an artefact has certain characteristics.
The second is inference: from the available facts we derive something that is not directly observed.
The third is interpretation: we organize the facts according to certain categories, determining which are important, what relations hold among them, and what overall meaning they take on.
The further we move from the first level, the greater the risk of arbitrariness becomes.
This does not mean that every interpretation is equally good. One interpretation may be better documented than another, more coherent, broader, better able to account for the data. But what is often missing is a universally shared criterion that would allow these differences to be turned into a cumulative system of knowledge comparable to the one that exists in the formal disciplines and the natural sciences.
And it is precisely here that the fundamental difference emerges.
Mathematics and logic are not empirical sciences, but they belong to a domain in which there is an extremely high degree of intersubjective agreement about procedures and results. The natural sciences, though traversed by controversies and theoretical changes, likewise possess a far stronger cumulative structure: their practitioners share a great quantity of results, methods, definitions and verification procedures.
In the human and social sciences the situation is very different.
Knowledge is not lacking. What is chiefly lacking is an overall organization of it, and shared criteria for establishing how its parts should be connected.
The result is paradoxical. Humanity possesses an immense quantity of humanistic knowledge, but it does not necessarily possess humanistic knowledge as an organized system.
It possesses thousands of books about society, but it does not possess a universally shared answer to the question: what does one need to know in order to understand society?
It possesses thousands of interpretations of history, but it does not necessarily possess a shared criterion for establishing how those interpretations should be compared and integrated.
It possesses a boundless quantity of psychological, sociological, economic, anthropological and philosophical theories, but it has no general map allowing it to establish which parts of those theories are well established, which controversial, which incompatible, which complementary, and which by now superseded.
The problem, then, is not merely the fragmentation of knowledge. It is its lack of organization.
And the growth of knowledge does not automatically solve the problem. On the contrary, it may aggravate it. Every new book, every new theory, every new piece of research adds one more fragment to the available mass. If no system exists capable of integrating the new fragments with the earlier ones, a quantitative increase in knowledge may produce an increase in disorientation.
The internet has made this phenomenon particularly evident. Today almost anyone can find information on almost any subject. But finding information does not mean finding knowledge. A search produces results, not necessarily a structure. The problem is no longer merely access to information, but its selection, verification, ranking and composition.
Anyone who wants to learn something in the humanities is therefore often forced to build his own knowledge for himself. He reads a book, attends a lecture, comes across an article, encounters an author, follows a hunch, discovers an alternative theory. In the end he assembles the fragments he happened to run into.
But why those fragments?
They might be important or irrelevant. They might be true or false. They might represent a central position or a marginal one. They might be compatible or incompatible with other fragments. They might have been chosen because they are fundamental, or simply because a platform’s algorithm displayed them.
Personal knowledge thus risks becoming a contingent composition of chance encounters.
A paradoxical situation arises: the individual does not necessarily choose what he knows; often he knows what he happened to run into.
And he then tends to organize what he has encountered according to whatever connections he manages to construct. But those connections may be as arbitrary as the fragments he started from.
The consequence is that, in the absence of a good general organization of knowledge, an individual’s intellectual formation depends enormously on chance: on the teachers he happened to meet, the books he happened to find, the people he listened to, the bibliographies he consulted, the interests he developed by accident, the algorithms that select what appears on his screen.
This may also explain why many people give up the search for knowledge.
Not necessarily because they are uninterested in knowing. They may simply realize, more or less consciously, that the search is interminable and that there is no clearly guaranteed path. Faced with an infinite supply of information and no reliable map, it becomes reasonable to rely on an authority, to adopt a ready-made narrative, or to give up altogether.
The problem of humanistic culture, then, is not merely that different opinions exist. Diversity of opinion is inevitable. The problem is that we often lack a sufficiently shared system for locating those opinions in relation to what we know.
What is missing is a map of knowledge.
Such a map should not simply report what each author maintains. It should show the relations among the various claims: which are documented facts, which inferences, which hypotheses, which interpretations; which propositions are widely shared, which contested; which theories contradict one another, which can be integrated; which items of knowledge are fundamental and which depend on others; which questions are settled and which remain open.
A map of this kind would be something very different from an encyclopaedia or a textbook. An encyclopaedia tells us many things. A textbook organizes many things. A genuine map of knowledge would have to tell us how the things we know stand in relation to one another, and how far we are entitled to consider them true.
Perhaps this is the great gap in humanistic culture: not a scarcity of knowledge, but the absence of an overall organization of it that is sufficiently objective and sufficiently shared.
We have produced an enormous quantity of knowledge. We have produced an even greater quantity of interpretations of that knowledge. We have produced whole libraries of narratives.
But we have not yet produced, to a comparable degree, a system in which knowledge can be found as knowledge.
And as long as such a system does not exist, whoever wants to know the humanistic world must largely build it for himself.
It is an intellectually miserable situation — not because human beings have produced too little knowledge, but for the opposite reason: they have produced a great deal of knowledge without managing to organize it adequately.
Translated from the Italian original: “Sulla miserabile disorganizzazione del sapere umanistico“
