What were the inquisitions?

Heresy was understood as a public crime rather than a private error, which the jurists likened to treason, and the tribunals existed to detect it, extract confession, and impose penance or punishment. Most sentences were not death: fines, public humiliation, penitential garments, pilgrimage, imprisonment, exile. Those condemned to die were handed to the secular authorities, who carried out the sentence.

The Spanish tribunal was set up by Ferdinand and Isabella under a papal bull of 1478; the king nominated its grand inquisitor and the pope confirmed him. Its first decades were directed at converts from Judaism and their descendants, many from families forced into baptism in 1391 or converted around the expulsion of 1492. Converts from Islam, baptised by force in the early sixteenth century, became its targets later, with trials peaking in the second half of that century. Neither group’s conversion was trusted, and testing it was the tribunal’s first business. It went on to try many people for other things: heretical propositions, blasphemy, bigamy, witchcraft. In one study of 35,000 of its trials, charges of Judaism, Islam and Lutheranism make up barely half.

Torture was regulated rather than unlimited: it could be used only in trials for major heresy, and a confession given under it counted, in theory, only if the accused did not recant afterwards. That is what the rules said, and the rules are not the same as the practice.

The Roman tribunal is the one that did not end. Paul III set up a commission of six cardinals to watch over matters of faith in July 1542; it was renamed the Holy Office in 1908, the Congregation for the Doctrine of the Faith in 1965, and it is a dicastery of the same curia today.

The Spanish Inquisition kept summaries of its cases, and a database built largely from them covers more than 67,000 trials between 1480 and 1820. Counting works at all because each tribunal sent Madrid an annual summary of every case it heard, so those survived centrally rather than in local archives. The summaries run from about 1540, coverage peaks at over seventy per cent between 1560 and 1650, and the years outside that window are pieced together from a wide range of secondary studies. Extrapolating puts the whole life of the institution near 136,000 trials.

How many of those ended in execution is a much smaller number. Llorente, once general secretary of the Holy Office, put the executed at nearly 32,000. Two later figures each cover part of the period. Johnson and Koyama, cited in a 2019 economic history, put the converts executed before 1525 at about 2,000, with some 15,000 reconciled to the church. A table drawn from Parker (1982) gives executions in 3 per cent of the 49,092 trials held between 1540 and 1700.

A 2021 study assembled a dataset of trials by municipality, covering all of Spain and three centuries, and used it for something else entirely: measuring what persecution did to the places where it was heaviest.

It found that towns where the Inquisition was more active have markedly lower economic activity today, and lower trust and educational attainment. Whatever the death toll, the institution's effects outlasted it by nearly two centuries, which is a different kind of harm and a measurable one.

The revisionist case is that the inquisitions were less murderous and less omnipresent than the legend, and that much of the imagery belongs to the black legend of Spain rather than to the records. One statement of it is Henry Kamen’s survey, subtitled A Historical Revision, which its publisher describes as rebutting the myths and exaggerations that have distorted the subject.

All of that is defensible and none of it is an acquittal. A tribunal that imprisons on secret denunciation, tortures under rules, and sends a smaller number of people to be burned is not thereby a mild institution.

The honest statement is that the corrected numbers are the numbers, and the corrected numbers describe centuries of judicial coercion.

Archives were destroyed, dispersed and partly lost. Most of the full trial records of the Goa tribunal are gone, and its historians work from the case summaries that survived instead. In Spain the coverage is uneven rather than absent: Valladolid's documentation is particularly scarce, Seville, Cordoba and Granada are poor. Totals are extrapolations from what is left.

Where the records are dense, the picture is detailed enough to reconstruct individual cases; where they are thin, the figures carry wide uncertainties, and a total quoted without them says more than the records do.

In 2000 the Catholic Church held a liturgy of repentance in which John Paul II asked forgiveness for, among other things, violence in the service of truth, and a theological commission published a study of what it means for a church to ask pardon for its own past. The study recalls that he had already asked historians, at a 1998 symposium on the Inquisition, for "a reconstruction, as precise as possible" before any judgement.

That document distinguishes "the holiness of the Church from holiness in the Church", and confesses the sins of the church’s members "without obscuring this holiness". It records unease from those who feared that admitting faults would look like giving in to hostile accusers.

Whether an institution can repent, and what that would consist of beyond a statement, is the part that does not resolve into a fact.