Question: Let us begin with the central theme — the relationship between truth and evidence. In your work, you have convincingly shown that truth is never a given, fixed entity; rather, it gradually emerges from traces, signs, and partial inferences. In your view, what is the role of truth in the work of the historian and, by analogy or contrast, in that of the social scientist?

Carlo Ginzburg: Truth, let’s say, is a point of arrival. The discovery of truth is a trajectory that starts with questions. I wrote an essay entitled “Our Words, and Theirs: A Reflection on the Historian’s Craft, Today” (Ginzburg, 2012a), republished in La lettera uccide (Ginzburg, 2021a). In it, I reworked Kenneth Pike’s (1954) etic-emic distinction in the light of the writings of Marc Bloch and, in part, Arnaldo Momigliano. I thus came to the conclusion that historians start from “anachronistic questions” which, however, they can reformulate through the analysis of empirical documentation, moving from etic questions (the observer’s point of view) to the reformulation of the same through emic documentation (the actor’s point of view). This interaction is present in the work of the historian, and truth is the endpoint of the trajectory that follows.

Truth is provisional, it is falsifiable, as Karl Popper taught us. This seems obvious to me and should be part of a sort of shared tradition. I mentioned anachronistic questions, and in that article, I quoted March Bloch, who in the late 1930s asked himself: Is it legitimate for historians to use the term “social class” for the Middle Ages? In an era when the term did not have that meaning? Marc Bloch (1953) posed this type of problem, which he then developed in his posthumous reflections entitled The Historian’s Craft, in a series of reflections that I found enlightening. In an essay on “Microhistory and World History”, I returned to the point and reformulated it after trying to trace the notion of thought experiment from Hobbes onward (Ginzburg, 2021b). I thus highlighted a remarkable chain whose links go from Hobbes to Vico, Marx, Labriola, Croce, Gentile, and finally to Robin George Collingwood’s (1993) concept of “re-enactment”.¹ I wrote that the path to truth is long and that the notion of mental experiment served to support my argument that all history is “comparative history”, i.e., I revisited Croce’s (1921) notion that all history is contemporary history, an idea that can be understood in two distinct senses. The more common interpretation holds that the historian inevitably writes from the standpoint of his own time. Yet this view tends to overlook the risk of anachronism — a concern that led me to engage with the etic–emic dichotomy. I have argued that every historical narrative is, in fact, a form of “comparative history”, insofar as it entails a dialogue between the context from which the historian poses questions and the contexts from which answers emerge, shaped by the categories of historical actors. This interplay between the etic and emic levels, to which I referred earlier, is marked by a sequence of thought experiments — an aspect of inquiry shared by both the natural and the social sciences. It was precisely on this point, I believe, that for the first time I found myself diverging from Marc Bloch.

Q: You have frequently engaged with the field of law in your reflections. In your view, what are the main similarities and differences between the ways in which law and history approach the notion of evidence? Why do you argue in your book The Judge and the Historian: Marginal Notes on a Late-Twentieth-Century Miscarriage of Justice (Ginzburg, 1999b) that Marc Bloch’s distinction between understanding and explaining needs to be revised somewhat? One might think that understanding is also judging. Even those who have to judge must first understand.

CG: At the beginning of my career, I worked on judicial records — specifically, the trials conducted by the Inquisition — and I encountered an unexpected situation that prompted extensive reflection. Initially, for reasons I later discussed in the afterword to the new Italian edition of I Benandanti (Ginzburg, 2020a), published by Adelphi,² there was a desire to know and recover the voices and attitudes of the victims: and this was also linked to my personal experience during the war, insofar as I retrospectively understood how I became a Jewish child through persecution. Later, when I had already written The Cheese and the Worms: The Cosmos of a Sixteenth-Century Miller (Ginzburg, 1980a), I realized that alongside this emotional contiguity with the victims, I had also an intellectual contiguity with the inquisitors, as I argued in an essay, first published in English (Ginzburg, 2013c), later included in the book Il filo e le tracce: Vero falso finto (Ginzburg, 2006a), on “The Inquisitor as Anthropologist”. This essay had some unexpected repercussions. I mentioned this when I republished it in Italian (Ginzburg, 2006b). I was in Moscow for a conference when I received a phone call saying, “We would like to hold a public discussion about your essay”. The essay had only been published in English up to that point, and the phone call was in English. I asked, “Who is this?” It was the Memorial Association.³ At that time, Memorial was working to protect human rights during the war in Chechnya. I said yes, I was very flattered, but I also wanted to understand what they wanted to talk about. They gave me a hint, and then the point was developed during the discussion, which was a public discussion. Their idea was that the “oblique” reading of the inquisition processes that I had proposed in my books could be used to read the trials connected to Stalin’s purges.

We spoke about this: a conversation that affected me deeply. The notion of a certain, albeit partial, contiguity with the inquisitors plays an important role in The Judge and the Historian, the book I mentioned earlier. I should say that this work is definitely unique within my career, insofar as it is the only book I have written with the explicit intention of producing a practical effect. In other words, it was conceived as a response — one might say a reaction — to the trial of my friend Adriano Sofri and his comrades, militants in Lotta Continua.⁴

Adriano Sofri was accused of being the instigator and intellectual author of the murder of Luigi Calabresi. I was, and remain, absolutely convinced of Sofri’s innocence; however, as I wrote in the introduction, I did not wish to assert my personal convictions, my friendship, or any subjective stance. My aim was quite different. I sought to engage objectively with the documentation of the trial held in Milan, hoping that my analysis might contribute to the appeal trial. The idea for the book originated in a letter published in La Repubblica shortly after the verdict in the first trial, in which the proceedings were compared to a witch trial. Among the signatories was Adriano Prosperi, a long-time friend of both mine and Sofri’s. I obtained the transcripts of the first trial — approximately three thousand typed pages — and immersed myself in them.

The outcome of The Judge and the Historian was paradoxical. Although the Court of Cassation annulled the initial verdict, a new trial followed, and Sofri was condemned to twenty-two years of prison. After nine years, having risked dying due to a serious illness, he was sent to house arrest.

From a practical standpoint, my book was a complete failure. Yet paradoxically, it achieved an unexpected and wide resonance, as demonstrated by its numerous translations: into English, French, Dutch, German, Spanish, Japanese, a “pirate” Greek edition, and, most recently and to my astonishment, a Russian translation. For the latter, I wrote a special preface addressed to Russian readers. The fact that the book continues to circulate and has now found new life decades later strikes me as remarkable.

In the book, I emphasized that the pursuit of truth through evidence marks a zone of overlap between the work of the judge and that of the historian. The overlap is only partial, of course, since their aims ultimately diverge: the judge must pronounce judgment; the historian does not. On this point, I recalled Marc Bloch’s (1932) famous admonition: “Robespierrists! Anti-Robespierrists! For pity’s sake, simply tell us what Robespierre was” (p. 118).

Insisting on the importance of evidence in the work of the historian is not trivial. Let me recount a personal experience. In 1990, Hayden White, the author of well-known works on historical methodology, gave a lecture at UCLA in which he presented his neo-skeptical thesis that there is no rigorous difference between fictional narratives and historical narratives, since they both rely upon the use of rhetoric. I was in the audience (I had been teaching at UCLA for a year) and I intervened. White’s arguments seemed to me to be untenable and dangerous because they prevented the refutation of the thesis of so-called “negationists”, such as Robert Faurisson, according to whom the extermination of the Jews never happened. A civil but bitter clash ensued.

The evening after the conference, Saul Friedländer, a great historian of the Shoah, my dear friend as well as colleague, told me that a conference should be organized to assess the impact of positions such as Hayden White’s on research into the extermination of the Jews. Friedländer did just that, organizing a conference that resulted in a volume entitled Probing the Limits of Representation: Nazism and the “Final Solution” (Friedländer, 1992). It includes an essay of mine, later published in Italian, entitled “Just One Witness”, dedicated to Primo Levi (Ginzburg, 1992; 2023c). In that essay, I continued my polemic with White, according to whom the thesis of the deniers was “morally offensive and intellectually disconcerting”, but that “the distinction between a lie or an error and a misinterpretation can be more difficult to draw when we are dealing with historical events less widely documented than the Holocaust”. In this regard, I argued that the formula “unus testis, nullus testis”⁵ does not apply to the historian.

Hayden White died a few years ago. By chance, I came across an interview in Il Manifesto in which someone asked him: what do you think of Ginzburg and your discussions? He replied that Carlo Ginzburg has a “biblical” notion of truth (a phrase I found incomprehensible), adding that my book The Cheese and the Worms is a work of imagination based on a document of a few lines (White, 2006). Now, the proceedings of the trials against Menocchio occupy 250 printed pages (see Del Col, 1996). It seems to me that Hayden White showed a clear disregard for the facts (even a document is a fact).

Certainly, the discussion with Hayden White was useful to me. When I was invited to Jerusalem to give some lectures, I realized that there are two rhetorical traditions. One dates back to Aristotle, then taken up by Quintilian and Valla, in which proof is central; another, deliberately and explicitly anti-Aristotelian, which we can trace back to Nietzsche, in which rhetoric is opposed to proof. The English edition, based on the lectures I gave in Jerusalem, is entitled History, Rhetoric and Proof (Ginzburg, 1999a); the edition that I later expanded in Italian and translated into various languages is called Rapporti di forza (Power Relations) (Ginzburg, 2000). The importance of the rhetorical tradition dating back to Aristotle is crucial (especially today), not only for historians.

Q: I return to the subject of trials because it is very interesting for us, i.e., sociologists have not made trials a subject of study. There is a school of thought that argues that it is not possible to understand a social phenomenon through law enforcement agencies, because the way documents are written reflects the institutions’ labeling of the phenomena themselves, often for purposes of domination and power. If this were true, however, we would know nothing about the Inquisition, which has been studied through the records and documents of the trials. So, as you rightly write, we need to read the documents, taking up Walter Benjamin’s (1990) teaching, namely to “brush history against the grain” (p. 257), even going against the intentions of those who produced them. This is a major gap in sociology, because many social phenomena are only accessible through judicial documents. Among other things, Italian documents are perhaps excessively voluminous, while in the United States, judicial records are very streamlined. In our case, there is a wealth of information, a wealth of material that helps you understand a lot. I return to the point: the study of court records and trials allows you to learn about phenomena that you would otherwise not be able to learn about. So, the question is this: can we say that, as a judge and historian, you have adopted a philological approach to reading court records?

CG: Undoubtedly. Here, the example of my father, Leone Ginzburg, is decisive.⁶ He taught Russian literature as a freelance lecturer in Turin. In 1934, when the oath of allegiance to the Fascist regime was extended to non-tenured lecturers, my father refused to comply, thereby bringing his academic career to an abrupt end. Shortly thereafter, he was arrested for his anti-fascist activities — he was a member of the clandestine group Giustizia e Libertà — and spent two years in prison. After his release, he collaborated with Giulio Einaudi in founding the Einaudi publishing house and developed a profound intellectual and personal friendship with Benedetto Croce, who frequently visited Turin. Croce asked my father to edit an edition of Giacomo Leopardi’s Canti for the Scrittori d’Italia series, which Croce himself had created and directed (Leopardi, 1938). What has always struck me, however, is that despite his close relationship with Croce, the person my father referred to as “his teacher” was Santorre Debenedetti — a distinguished philologist who, being Jewish, could not openly appear as editor of the Einaudi series “Nuova raccolta di classici italiani annotati”, which included Dante Alighieri’s Rime, edited by Gianfranco Contini.

Philology held a crucial place for my father — not only as a technical discipline but as a broader moral and political disposition. His judgment of those who had embraced Fascism out of convenience or coercion differed sharply from the severe and uncompromising standards he applied to himself (see L. Ginzburg, 2000). In this divergence, I perceive a form of human understanding that informed his historical sensibility — one grounded in distance. I have often thought that this distance is precisely the distance of the philologist. I completely agreed with the title under which an interview with me was published: “Mio padre Leone filologo della libertà” (“Leone, my father philologist of freedom”; Ginzburg, 2009). My father’s approach to texts and his political and moral convictions were both inspired by philology in a broad sense, as Giambattista Vico understood it: a general method applicable across different fields. Among my father’s books that I have in my library, there is a copy of the third, posthumous edition of Giambattista Vico’s Scienza Nuova, published in 1744. When he was in internal exile, in a village of Abruzzi, my father wrote repeatedly to Santorre Debenedetti discussing who should edit the new edition of Scienza Nuova (L. Ginzburg, 2004, pp. 222 n. 229, 231, 242, 243, 248, 249, 251, 252, 253). I am convinced that Vico’s broad notion of philology can contribute to defeating “fake news”.

Q: But in the case of the judge and the historian, to give an indication of a method that may also be useful to our readers, in what sense did you use philology?

CG: Let’s say, first of all, that I explicitly state that my conviction, my certainty about Adriano Sofri’s innocence, is not something I bring into play as evidence. I address those who have no opinion, or perhaps have a contrary opinion. And so, I explicitly distance myself from them and analyze the documents from a philological perspective. The etic-emic conceptualization I mentioned earlier helped me to formulate this distance, precisely because Kenneth Pike (1954) clearly separated the two perspectives. This has important implications for understanding the differences between empathy and distance. Empathy abolishes distance, that is, it presupposes an identification that projects the provincialism of our ego onto others, whereas it is precisely through distance that we can overcome the provincialism of our ego. So, let’s say that the rejection of empathy, on the one hand, and the contrast between empathy and philology, on the other, seem to me to be central and complementary elements.

Q: Let’s return to the circumstantial paradigm by bringing together some of the issues that the participants in the Symposium hosted here have pointed out to us. Traces and footprints are pieces of information released in conditions of low control and low surveillance. So, a first issue here is how to distinguish these pieces of information, these bases of knowledge that are then used to make inferences, from error and mistake. Since there is no intentionality, how is it possible to draw a distinction between the detail of the earlobe in that particular painting as sufficient evidence to attribute it to that artist, to take your example from Morelli, and the error, i.e., the fact that that detail is an error and therefore does not “speak” of that painter’s typical style of painting?

CG: I would respond with another question: Is the error a deliberate error or an involuntary error? If it is an unintentional error, we return to the distinction between “Unintentional revelations. Reading history against the grain”, which is the title of an essay I published in La lettera uccide (Ginzburg, 2021c). In reality, this term is used simultaneously, and perhaps independently, by Marc Bloch and Alessandro Manzoni. Unintentional revelations may or may not include an error or a trace. Here we are moving into a field which, shifting to the ethical plane, gives certainty to the historian, sociologist, or anthropologist, because there, let’s say, is something that the author of the text or gesture does not control. In fact, I am continuing to work on this topic: I have just given a lecture on it at Oxford.

Q: Let’s explore this point further, still on the subject of the indiciary paradigm. Some sociologists equate this paradigm with an abductive inferential strategy. Do you agree with this idea?

CG: At one point in my essay [“Clues: Roots of an Evidential Paradigm”], I cited Charles Pierce, the American pragmatist who codified abduction (Ginzburg, 2013c, n. 38). However, it seems to me that this similarity is quite generic, whereas the thing that becomes both more specific and provocative derives from the type of empirical documentation that is included in the abduction process. For example, the inclusion of Morelli alongside Freud. I do not deny the connection between the circumstantial paradigm and the abductive approach. However, I find it more interesting to start from empirical documentation than to draw more general conclusions in an “abductive” key.⁷

Q: The third issue we are interested in exploring is related to your answer and concerns the work of historians and sociologists in relation to semiotics. Do you often take “signs” into consideration?

CG: Yes, although I must say that very often semiotics moves in a perspective that, in my opinion, is a bit tautological, while what I find interesting is to question oneself starting from “signs” and, if possible, to be taken by surprise. Here, there is a convergence, which exists in Italian and not in other languages, between “caso” and “caso”, “case” and “chance”. I have worked extensively on this idea. I wrote an essay entitled “Conversing with Orion” (Ginzburg, 2001; 2005a) and then I wrote an essay in English entitled “Latitude, Slaves, and the Bible: An Experiment in Microhistory” (2005b), which is the result of a case study carried out using the digital resources offered by the Web. In reality, this approach has older roots. In the early 1970s, I began teaching at the Facoltà di Lettere in Bologna; Adriano Prosperi was teaching at the Facoltà di Magistero. We had been together at the Normale, and so we decided to hold a seminar on a 16th-century religious text entitled Trattato utilissimo del Beneficio di Cristo (A Very Useful Treatise on the Benefit of Christ), published in Venice in 1543, copies of which had been destroyed except for one. The text was rediscovered and reprinted in the nineteenth century. We decided to bring together the students from our two classes and held a seminar, which resulted in an article we dedicated to Auguste Dupin, Edgar Allan Poe’s detective, who discovers a truth before everyone’s eyes. A well-known church historian immediately understood the allusion and was irritated. We subsequently wrote a book called Giochi di pazienza (Puzzles), published in 1975, in which we tried to explain to young people how to do research, describing the path to discovery, the mistakes we had made, and so on (Ginzburg & Prosperi, 2020). It was a truly unusual book that, at the time, received a single, very critical review from a former Jesuit, if I’m not mistaken. Then it was reconsidered; almost 50 years later, a Spanish translation was published in Mexico (Ginzburg & Prosperi, 2019), followed by a Portuguese translation in Brazil (Ginzburg & Prosperi, 2022). An English translation will soon be published by Seagull, the Indian publishing house distributed by The University of Chicago Press (Ginzburg & Prosperi, 2026). The anomaly of this book lies in the idea of analyzing the way in which prejudices and assumptions filter the relationship of the historian, sociologist, or anthropologist with empirical documentation or informants, and, at the same time, can be corrected or controlled. Adriano Prosperi and I worked here in Bologna, writing this book literally with four hands, although I occasionally remember pieces he wrote or pieces I wrote. In the Archiginnasio Municipal Library,⁸ there was a magnificent card catalog, which has now been replaced by an electronic catalog. We used to compare the text we were working on to a tree in a forest, made up of Italian religious texts from the 16th century. We asked ourselves: was it possible to reconstruct this forest in order to characterize the anomaly of the Trattato utilissimo del Beneficio di Cristo? We started to search in the card catalog, checking names such as Marco, Giovanni, Benedetto, and so on: members of religious orders who could have written a text comparable to the Beneficio di Cristo. Our strategy aimed to collect and explore what, in digital terms, is called “noise”, which traditional search for online information tries to avoid (when you look for the address of a restaurant or the title of a book, you try to get as focused and accurate a result as possible). When I started teaching in Los Angeles in the late 1980s, I continued with the UCLA library’s digital catalog what we had done in Bologna with the Archiginnasio card catalog. It was a game: you ask for a word (for example, “bizarre”) and see what comes up. An unexpected title can lead to unexpected conclusions in a search.

Q: One of the authors participating in this symposium wonders whether the introduction of artificial intelligence could lead to a combination of the indiciary paradigm with quantitative analysis.

CG: I believe so. What I have just described does indeed go in that direction. I have given a number of lectures at Cambridge and Oxford and was struck by the fact that the discussion of artificial intelligence is a constant there. I am completely ignorant of this subject, but I have noticed the contiguity you mention; I also had the impression that, for the moment, artificial intelligence is probably not as intelligent as it seems. This does not exclude the possibility that it may become so, and probably will. Even in this case, however, interaction with humans will probably (hopefully?) be unavoidable and may lead to unpredictable directions. Much will depend on how these developments allow scholars to still be “taken by surprise”: not to look for an answer, but to try to be surprised by unexpected questions. In this case, it is possible and probably plausible that artificial intelligence, at a level much more sophisticated than the current one, could go in this direction.

If in the documentation we have collected, there is nothing unexpected, then there is something wrong with our research trajectory. I remember starting to study witchcraft trials with a series of assumptions and prejudices typical of a left-wing student in the late 1950s, who had drawn attention to the culture of the subordinate classes from reading Antonio Gramsci’s (2011) Prison Notebooks. Delio Cantimori, a great historian who was my mentor, told me, “Go to Modena, go to the State Archives and look for the Inquisition trials which are preserved there”. I went to Modena and came across a trial against a peasant woman, Chiara Signorini, who had been accused of witchcraft in the early 1500s. The specific charge was that she had cast a spell on the mistress who had driven her off the farm where she worked with her husband. This seemed to confirm my naive hypothesis that witchcraft could be regarded as an elementary form of class struggle. I still remember the disappointment I felt when I thought I had found this confirmation. It must be, I thought, an uninteresting hypothesis if a confirmation comes so quickly. In reality, other elements emerge from the trial. When the inquisitor asks her, “But why did you do this?”, Chiara replies: “The Madonna appeared to me and told me, ‘Chiara, do this’. The Madonna was beautiful, ruddy-cheeked”. I still remember my amazement when reading this description because, between the lines, something emerged that was part of peasant culture, that the inquisitor had not anticipated. The essay I published (the first piece I ever published) was entitled “Witchcraft and Popular Piety” (Ginzburg, 1961; 2013e). Chiara was subjected to a series of questions and then tortured. In many cases, she followed the inquisitor’s suggestions: for example, she admitted that it was not the Virgin Mary who had appeared to her but the devil, but she denied ever having gone to a witches’ sabbath.

At the end of the essay, I wrote that this trial, despite its specificity, could perhaps have a paradigmatic value. When I reread this sentence a few years ago, I thought that the term “paradigmatic” echoed Thomas Kuhn’s (2012) The Structure of Scientific Revolutions, a book that had enormous importance for me, as it did for everyone who read it. In fact, my essay was published in 1961, a year before the first edition of Kuhn’s book. I used the term paradigm in one of the many meanings used by Kuhn, namely “exemplary”. But what surprised me was that the case I had focused on was an anomaly that I considered exemplary. The idea of studying “exemplary anomalies” is something that has accompanied me throughout my life since then.

Q: I wanted to briefly return to the topic discussed in the book Puzzles (Ginzburg & Prosperi, 2026) to conclude the argument. How can it be useful to think of the “biases” of the scholar as part of an assumption that is not necessarily linked to an ideological element? Or do all biases have an ideological matrix?

CG: Not necessarily. We approach a topic through filters that can shape our analysis, leading us to find what we are looking for. As I never tire of repeating, the idea of “finding what we are looking for, period” is not enough. The problem is being caught off guard, finding something unexpected. The philological method helps historians, sociologists, and anthropologists to break out of an anachronistic and/or ethnocentric perspective. It also helps them to combat the thesis, popularized above all by Hayden White, that there is no strict boundary between fictional narrative and historical or sociological narrative.

In a lecture in honor of Natalie Zemon Davis, which I presented in Budapest, I argued that the spread of ideas attributable to postmodern neo-skepticism has weakened resistance to fake news (Ginzburg, 2023a). A few years ago, I was in Lima, where I was invited to participate in a conference on “post-truth”. I accepted immediately because I detest this notion; in my contribution, I described it as an “old new story” (Ginzburg, 2020b). The idea of using lies for political purposes has ancient roots, but technology is transforming the project in a decisive way.

When I arrived at UCLA, I was amazed to see that the most curious and intelligent students were seduced by these postmodern ideas that “facts do not exist” and that there are only narratives. I have said many times (echoing a page by Gramsci, on a completely different subject) that if one were to react to postmodern positions with a war of position, that is, with trenches, one would have to argue that historical narrative is historical narrative, while fictional narrative is fictional narrative, and so on. With a war of movement, on the other hand, one must enter the opponent’s camp to wrest their weapons from them: for example, as I said before, by remembering that there are two rhetorical traditions, and that only one of them allows us to recognize the crucial role played by evidence. This also helps us to understand what it means to “read history against the grain”. If the production of historical evidence, especially voluntary evidence, largely reflects the relations of production and, more generally, the balance of power within a society, reading history against the grain means — as Walter Benjamin (1990) pointed out in his Theses on History — looking for traces of oppression in that evidence. To do this, it is necessary to reconstruct, on the one hand, how historical testimonies were produced; on the other, their effects, which are often unintended or unforeseen.