The Treaty of Nîmes (1704): fake news, propaganda and diplomacy during the War of the Spanish Succession

Abstract

‘TREATY: A Contract, a deal that is reached.[...] also refers to a negotiation & conclusion of peace, of confederation, marriage, capitulation’.¹

On 22 July 1702, the abbé du Chayla was brutally murdered by Protestant rebels known as ‘Camisards’ in the Cévennes. After nearly two decades of mounting persecution since the revocation of the Edict of Nantes in 1685, his death sparked what many historians have called the last French war of religion.² Far from a local conflict, the War of the Cévennes (1702–10) rapidly took an international dimension, as exiled Huguenots called upon the Anglo-Dutch alliance to support the rebels against Louis XIV in the midst of the War of the Spanish Succession (1701–14).³

The history of the Camisards’ revolt is well documented and has received growing attention from Anglophone historians in recent years.⁴ Nonetheless, uncertainty continues to surround precisely how and when this conflict ended. Although Languedoc saw ongoing violence until 1710, French authorities had negotiated peace with the rebels in May 1704. Contemporary sources provide discordant clues as to the outcomes of these negotiations but, according to Monahan, a ‘failure to communicate’ caused the conflict to resume until the last rebels were crushed.⁵

At the heart of this contention is the existence of a treaty, allegedly signed in Nîmes in May 1704, between the Camisard leader Jean Cavalier and Maréchal de Villars, Louis XIV’s envoy to Languedoc. The text was published by Cavalier in 1725 and presented as a partial restoration of the Edict of Nantes in Languedoc. In the aftermath of a religious war, both parties were likely to distort the reality of this little-known episode in French history. For the stakes were indeed very high: if authentic, the ‘Treaty of Nîmes’ would have constituted a severe blow to the Sun King while France was fighting most of Europe. If fabricated, however, the document may be regarded as a late piece of Protestant propaganda destined to rehabilitate Cavalier’s reputation and legacy.

The early modern period saw an explosion of printed news as production costs declined and literacy levels increased. By 1700, printers and journalists had built networks of informants to break news, beat competition and secure profits. Information circulated faster than ever across national and linguistic borders, sometimes unverified, and increasingly needed to be managed.⁶ This was paramount in times of international conflict, when courts suspected dissenters of colluding with the enemy. As Feyel, Rétat and others have demonstrated, French-language gazettes printed abroad circulated widely in France by 1700 and competed with Renaudot’s Gazette as part of a ‘double market of information’.⁷ Versailles’ efforts to conceal its negotiations with the Camisards fuelled a news race abroad based on leaks, hearsay and rumour. This lack of reliability against efficient royal censorship and propaganda has left its mark on historians, who have overlooked the significance of such an agreement. Yet, as shall be seen, although the document itself was indeed fake and only in fact reconstructed the terms negotiated between Cavalier and Villars for public knowledge, at least seven different versions circulated in Europe between 1704 and 1725. Such widespread interest in a treaty that never was should command our attention.

This article dissects the ‘Treaty of Nîmes’ in the light of recently discovered sources to determine the true circumstances that ended the Camisards’ rebellion. After briefly summarizing the fortnight of negotiations that led to Cavalier’s surrender, it considers the terms of the treaty against the backdrop of contemporary correspondences, both Protestant and Catholic, foreign and domestic, and reconstructs the genesis of that fake document based on newspaper reports and fragments collected from European and North American archives. I demonstrate, contrary to the existing historiography, that the ‘Treaty of Nîmes’ was contemporary to the Camisards’ revolt and that it circulated between the main protagonists of the War of the Spanish Succession. For this reason, it is argued that fake documents also have a historical value of their own right. Indeed, this ‘treaty’ illustrates, on the one hand, how the ‘stratégie de cabinet’, whereby military decisions were increasingly taken at Court rather than on the battlefield, also provided opportunities for conceited generals such as Villars to pursue their individual interests. The distance separating Nîmes from Versailles allowed him to take risky decisions and manipulate the information reported to Louis to his own advantage, forcing Europe’s most powerful monarch to rely entirely on his general’s brinksmanship to end the rebellion and regain control over the Cévennes.⁸ On the other hand, the ‘Treaty of Nîmes’ offers some insights into the communication networks that enabled its circulation and supported a transnational sense of Protestant solidarity in the post-Westphalian era.

I

On 29 March 1704, after nearly two years of religious violence, Louis XIV sent Maréchal Claude-Louis-Hector de Villars (1653–1734), who had been leading the Franco-Bavarian army with Elector Max Emanuel against Imperial forces, to the Cévennes. The fact that Louis solicited one of his highest-ranking officers to contain a peasant revolt in the midst of the War of the Spanish Succession speaks volumes about the severity of the situation. Not only had the Bavarian campaign been failing for months, but Savoy and Portugal had also recently joined the Grand Alliance, while French royal finances were showing frailty. As Louis told Villars: ‘You will be doing me an important favour if you are able to stop a revolt which may become very dangerous, especially at a time when, being at war against the whole of Europe, it is quite embarrassing to have one in the heart of the kingdom.’⁹ Languedoc was the largest Protestant stronghold then, with a long history of heresies and revolts behind it. Nowhere else in France was the crisis of trust between Catholics and Protestants more acute than in the Cévennes.¹⁰ The dragonnades and the Revocation had rekindled religious tensions, killing thousands of Protestants who refused to convert. Time was also pressing for Versailles as the Anglo-Dutch coalition prepared financial and military support to the Camisards.

In Lyon, Villars met Jacques-Jacob de Rossel, Baron d’Aigaliers, a respected Huguenot loyalist at the Court who would become key to the negotiations. The intendant of Languedoc, Nicolas Lamoignon de Basville (1648–1724), welcomed them in Beaucaire and all three men finally reached Nîmes together on 22 April 1704.¹¹ Villars’ arrival in Languedoc was an immediate game changer. His predecessor Maréchal de Montrevel had failed to crush the rebellion by force, so Villars immediately announced his intention to resolve the conflict peacefully. This new, more diplomatic approach contrasted sharply with that of Basville, for whom negotiating with royal subjects debased the monarchy and discredited the king. But Villars was determined to capitalize on the hopes and culture of the local population: ‘They all want to expect that my arrival will bring a happy change and, prophecies being to their liking, they brought me one of Nostradamus’ Centuries which assures that a general entering Languedoc through Beaucaire will end all the sufferings of the Province.’¹² At the time, Villars thought that he was ‘dealing with madmen’, but later confessed in his memoirs: ‘I let them believe, as this could not harm my operations’.¹³ He immediately sent d’Aigaliers to Uzès to establish contact with Jean Cavalier (1681–1740), a former baker’s apprentice and leader of the Camisards. Not content with Villars’ trust in a Protestant nobleman, Basville sent his own agent, Isaiah Lacombe—Cavalier’s former employer—to approach the notorious rebel. Thus, instead of collaborating as they were expected to, Villars and Basville favoured two different approaches and competed to bring Cavalier and his troops to surrender.¹⁴

The conflict was beginning to turn in Versailles’ favour. By late April the Camisards were severely weakened. They had suffered heavy losses a few days earlier at the battle of Nages; stocks of supplies, food and ammunition were running low as the long-promised Anglo-Dutch aid had yet to come; denunciations were financially rewarded and dozens of Camisards were turning themselves in daily, adding pressure on Cavalier.¹⁵ On 30 April, Cavalier wrote to Villars for the first time:

MONSEIGNEUR, having heard that you were not informed of our request […] I wanted to take up my pen again to exhort you to accept this request for the good & prosperity of the Kingdom, which is the freedom of our conscience & the release of the prisoners & many galley slaves who suffer unfairly for defending the truth.¹⁶

It took several days for Villars and Basville’s respective agents to locate the Camisards in the Cévennes mountains and open negotiations with Cavalier. This is where contemporaneous sources increasingly diverge and the chronology of events becomes less clear. We can establish for certain that Lacombe was the first to reach Cavalier, on 11 May. He persuaded Cavalier to submit his requests in writing to Villars via the new maréchal de camp in Alès, the Marquis de La Lande.¹⁷ The next day, 12 May, Cavalier and 200 Camisards met La Lande outside Alès by the Pont d’Avènes, who only brought twenty dragoons and released Cavalier’s younger brother as a goodwill gesture. Despite many eyewitnesses, their conversation remained confidential. The two men talked alone for three hours. Their body language suggested that the tension was rising until La Lande presented some paper and ink and persuaded Cavalier to sign a written agreement. According to l’abbé L’Ouvreleul, a Catholic chronicler of the rebellion:

Cavalier replied that he did not need money & that he only begged him to obtain from the King permits to go to Geneva with eight of his friends, & those of his Troop willing to follow him, a permission to sell their property, & the freedom of their imprisoned or exiled relatives. Mr. de la Lande responded favourably to all these articles.¹⁸

A truce was declared with immediate effect and both parties withdrew to their quarters. The next morning, 13 May, La Lande presented the agreement to Villars in Nîmes, who in turn dispatched his nephew, Monsieur de Saint-Pierre, to Versailles for approval.¹⁹

Yet the Pont d’Avènes agreement thwarted d’Aigaliers’ plans. For months he had been working at the Court as a loyalist Huguenot, hoping to lead a French regiment abroad in return for his diplomatic services. D’Aigaliers thus met with Cavalier in Alès on 13 May and convinced him of his foolishness in conditioning his surrender to La Lande. He drafted a letter of apology begging the King’s forgiveness and surrendered without conditions for the Camisard to sign.²⁰ Villars immediately rejoiced at Cavalier’s volte-face in a letter to Michel Chamillart, Louis XIV’s minister of war:

Yesterday evening, Mr. d’Aigaliers returned, having spent the whole day with Cavalier. He brought me back the letter that you will find attached in the original. I beg you, Sir, to send it back to me; for since he signed it in the presence of his troop, it would be good to be able to show it back to him. You will find it much stronger than the first letter [i.e. the Pont d’Avènes agreement] because this one makes no conditions. It is not in Cavalier’s hand, & I recognized that of Mr. d’Aigaliers, who did not deny to me that he had helped to compose it; but the fact is that it was read again before Cavalier & his troop & signed by him.²¹

Having secured two signed submissions from Cavalier in two days, the French authorities naturally downplayed the Pont d’Avènes agreement and brandished the second letter without conditions as their victory over the rebels. This was reflected in the press: the Gazette de Lyon, for example, only reported Cavalier’s unconditional surrender, as did French allies.²² The Swedish press merely announced a truce without furnishing further details.²³ Similarly, La Gaceta de Madrid did not report on the Pont d’Avènes articles, but instead published Cavalier’s unconditional capitulation dictated by d’Aigaliers on 13 May.²⁴ Its coverage of ‘la Guerra de los Fanaticos’ reflects the confidentially of the negotiation process and its exploitation by the parties involved. Whereas its sources from Paris merely emphasized an unconditional surrender, those in Bayonne signalled an accommodation—‘el ajuste’—whereby the Camisards negotiated permission to go abroad, sell their estates and form a regiment to serve Louis, which the Sun King was expected to approve.²⁵

Foreign gazettes were increasingly circulating in France by the early 1700s. Dutch publications in particular provided lots of information, albeit biased, on French domestic affairs and competed against Renaudot’s Gazette, Versailles’ official voice, within a ‘double market of information’. Provincial reprints of the Gazette multiplied after the Revocation and the Nine Years’ War—Montpellier began in 1689—in an attempt to diffuse French propaganda in its Protestant strongholds.²⁶ Despite French efforts to publicize Cavalier’s unconditional capitulation, the terms of the Pont d’Avènes agreement rapidly leaked in London, where the Marquis de Miremont had been coordinating the Anglo-Dutch coalition’s military and financial support for the Camisards.²⁷ An anonymous copy of a letter written in Nîmes on 14 May announced:

Here is the Substance of the Treaty that is being reported and that is not yet known for sure. [I] A General Amnesty, [II] the freedom for Cavallier, to leave the Kingdom with 400 men from his Troop [III] the release of Exiles & prisoners made so since the Beginning of the War and [IV] the Return of confiscated property, & other articles that are not known. [Cavalier] did ask for freedom of Conscience, but it is not expected that the King will grant it.²⁸

This first letter from the Camisards to their contacts in England, likely via Miremont’s agent on the continent David Flotard, built momentum for a treaty among Protestant nations, the terms of which were published simultaneously in French, Dutch and German in June, indicating a highly organized and coordinated effort.²⁹ However, it was The London Gazette that first reported on the Pont d’Avènes agreement:

Nismes, May 17: [Cavalier] was at last prevailed upon to enter upon a Treaty for accommodating Matters, and had an Interview 6 or 7 days ago with Monsieur de la Lande, Marshal de Camp, wherein they agreed upon several Articles for that purpose, which are sent to Court to be ratified.³⁰

The coalition was fully aware of Louis XIV’s desperate need to end the rebellion and send his troops abroad against Anglo-Dutch forces.³¹ It was in this context of urgency that, three weeks after his arrival, Villars finally brought Cavalier to his negotiation table in Nîmes on 16 May 1704. The Camisard appeared magnificently dressed on a Spanish horse followed by some twenty men, sparking a wave of hysteria across the city. The meeting was held outside the city walls in the Recollect convent in the presence of Basville, La Lande and the governor of Nîmes, M. de Sandricourt. According to Villars’ memoirs, the discussions remained cordial, save for Basville’s angry outbursts at the sight of a baker’s boy negotiating with a Maréchal de France. Cavalier nevertheless surrendered and was promised a pension and a colonelcy in return. The ceasefire was prolonged; Cavalier’s troop were to stay in Calvisson at the king’s expenses and enjoy the free, public exercise of their faith until 1 June, while waiting for orders from Versailles.³²

II

What was agreed to at the Recollects meeting on 16 May 1704? The exact terms of these private negotiations caused much confusion at the time. Unsurprisingly, the memoirs of the two main protagonists differ substantially on this episode. Villars only acknowledged oral negotiations, whereas Cavalier claimed that he had signed a peace treaty restoring freedom of conscience throughout Languedoc and releasing Protestant prisoners and galley slaves.³³ Both memoirs were respectively published some thirty and twenty years after the events, both lack precision and suffer from a propensity to revisionism.

Since history, as we know, is written by the victors, little credit was given to Cavalier’s version during his lifetime and even less by modern historians. Early Catholic chroniclers of the War of the Cévennes remain vague about the content of the negotiations and certainly did not mention the existence of a treaty.³⁴ The day after the Recollect meeting (17 May), Basville announced triumphantly to the bishop of Alès:

Cavalier, leader of the rebels, has submitted entirely and without conditions. He came yesterday to this city [...] to place himself in the hands of Mr. the Marshal de Villars without any condition other than to beg for the king’s mercy; [...] Mr. the Marshal de Villars sent a second letter to receive the king’s orders on the fate of those who have surrendered in full.³⁵

Even Antoine Court, the great Huguenot minister and historian of the Camisard rebellion, dismissed the treaty’s existence, arguing instead for a late forgery used to restore Cavalier’s legacy.³⁶ Except for Bosc who, in his monumental history of the revolt, suggested that the so-called treaty may simply have been a draft; modern historians have largely concluded the same. Joutard discussed the peace negotiations in passing; Chabrol, Frey and Frey only allowed for vague promises made by Villars to Cavalier, while McCullough does not mention the possibility of a treaty at all.³⁷ Among those who investigated the issue, Rolland ruled that ‘No correspondence, no other Memoir than that of Cavalier mentions the signature of such a treaty, which one cannot imagine for one second that it could have been signed by Basville or by Villars’.³⁸ Monahan likewise dismissed the treaty as a ‘mythical document’, a ‘fantastic invention […] necessary to cover up [Cavalier’s] failure’.³⁹

The contentious treaty in question was published in Cavalier’s memoirs in 1725. The fact that it appeared two decades after the events and in English, most likely translated by the Huguenot minister Pierre Gally de Gaujac, has seriously compromised its authenticity,⁴⁰ for Cavalier’s memoirs were originally compiled in French around 1708–13. This manuscript is kept in the Walloon collection in Leiden University Library (LUL) under the reference AW2 1459. It is only one of three surviving manuscript copies of Cavalier’s memoirs, of which the original appears to be lost.⁴¹ The treaty published in Cavalier’s English memoirs remains nonetheless identical on every point to this French manuscript, except for its date: 17 [instead of 8] May 1704.⁴² The nine-day difference between the French manuscript and its English publication is too striking to be a mistake. It suggests the treaty’s date was edited retrospectively, possibly with some purpose in mind.

So where was this English version of the treaty copied from? Presumably from the standalone manuscript Traité also held at LUL under the shelfmark AW2 1458, which may be the original. It consists of eight articles each followed by the note ‘accordé’ or ‘refusé’, and lists Villars, Basville, La Lande, Cavalier and Billard as signatories. Although similar in appearance, these two manuscript versions differ slightly in wording and, more surprisingly, AW2 1458 bears yet another date—11 May 1704.⁴³ Since both versions are in French, but in a different hand, the treaty’s content may have been edited in preparation for Cavalier’s memoirs or else the latter was simply not based on the former, as previously assumed.

Our examination of the Leiden manuscripts could almost end here if it was not for the discovery of a third manuscript copy of the ‘Treaty of Nîmes’ in the Blenheim Papers at the British Library (BL).⁴⁴ This copy is also written in French, but in a different hand from the Leiden manuscripts; it shares the same date as AW2 1458—11 May 1704—and all five aforementioned signatories. It does not contain eight articles, however, but eleven and is clearly presented as a copy in the margin. The original might still be AW2 1458, but the different wording and the breaking down of article 6 into four new articles suggests that an earlier version may even have existed elsewhere.

We can establish for certain that the BL copy was circulating by 1707, that is before Cavalier wrote his memoirs. Indeed, it was published verbatim in La Guerre d’Espagne, de Bavière & de Flandres, a largely fictitious memoir ascribed to former musketeer-turned-writer Gatien de Courtilz de Sandras under the imprint of the equally fictitious Pierre Marteau.⁴⁵La Guerre d’Espagne was translated into English in 1708, but only listed nine out of the eleven articles of the treaty that Villars had allegedly granted to Cavalier in the BL manuscript.⁴⁶ The following year (1709), another English translation of the treaty was published in its entirety in The Life and History of Lewis XIV. Although identical in content, this version varies slightly from that in both La Guerre d’Espagne’s English translation and Cavalier’s memoirs in 1725, indicating that the original manuscript circulated between different hands at the time.⁴⁷

Regardless of its actual origin, the ‘Treaty of Nîmes’ generated considerable interest during the War of the Spanish Succession. At least seven different versions of this fake document have now been identified between 1704 and 1725. Three survive in manuscript form—two in Leiden and one in London—while another three were published in French and English between 1707 and 1709, making Cavalier’s memoirs only the fourth published version. To these must be added multiple re-editions and new translations published afterwards.⁴⁸ Despite their early circulation during the Camisards’ revolt, none of the three manuscript copies identified above can be considered as the original because they all present the signatures in the same hand. These seven versions are either dated 8, 11 or 17 May 1704, none of which matches the Recollect meeting on 16 May.

III

The fact that the Treaty of Nîmes never existed should not undermine its historical value. A close reading of its content and, more importantly, the inconsistencies between its seven versions, sheds considerable light on its genesis, the motivations behind it and the negotiation process more generally. The remaining part of this article therefore examines the terms and conditions allegedly agreed between Villars and Cavalier on 16 May and shows how the devil was quite literally in the detail.

The first article was by far the most important. It guarantees freedom of conscience for all Protestants across Languedoc, exempts them from attending Mass and, despite forbidding the construction of new temples, it allowed for public worship in the fields. This was of course extremely contentious and likely incorporated lessons learned from previous failed negotiations. Already in 1683, before the Revocation, Claude Brousson had led a civil disobedience movement with Huguenots worshipping on the ruins of their destroyed churches in response to the dragonnades, edicts and arrêts that increasingly restricted their rights.⁴⁹ A decade later, exiled ministers such as Elie Benoist and Pierre Jurieu had unsuccessfully lobbied the Anglo-Dutch coalition to force France to restore Protestant worship at the Peace of Rijswijk in 1697.⁵⁰ While this first article’s parameters proved more restricted, both in its nature and geographical scope, it would still have been perceived as a severe blow to Louis XIV. If freedom of conscience was indeed discussed at the Recollects, Villars certainly did not mention it to Versailles. Instead, he emphasized Cavalier’s unconditional surrender drafted by d’Aigaliers the day after the Pont d’Avènes agreement: ‘upon the promise I made to Cavalier, he came with his chief leaders to submit entirely to Your Majesty’s mercy, & to offer without condition, either to leave the Kingdom, or to go and atone for their fault by sacrificing himself in Your Majesty’s service.’⁵¹ Villars’ letter, written from Nîmes on 17 May, reached Versailles just three days later, a record delivery reflecting the urgency of the matter when couriers between Paris and Montpellier normally took five days at that time.⁵² The French officer and diarist Philippe de Courcillon, Marquis de Dangeau noted on 21 May: ‘The letter from Mr. de Villars which arrived yesterday [May 20] in the evening brought the news that Cavalier had come to find him, surrendering entirely to the mercy of the king and no longer demanding any conditions.’⁵³

Cavalier’s alleged silence on freedom of conscience seems hard to believe, for his meeting with Villars offered a once-in-a-lifetime opportunity to defend a cause for which so many of his coreligionists had died. Not only had Cavalier made freedom of conscience his prerequisite for the negotiations in his first letter to Villars on 30 April, but he had also renewed his demand to La Lande at the Pont d’Avènes, to which the latter answered that the king would advise. More importantly, Villars’ memoirs contradict his own correspondence at the time and reveal that Cavalier went much further in his demands at the Recollects: ‘[Cavalier] asked [...] permission to profess their religion publicly in designated places. I replied that this last request would never be permitted.’⁵⁴ Not content with this answer, Cavalier obtained a significant compromise when Villars then promised him:

you insist upon liberty of Conscience, the King is willing to grant it to you; you shall live as you think fit; he will also permit you to meet, and have Service in your own Way, but not to build Churches, you ought to be satisfied with this, having taken up Arms against his Majesty.⁵⁵

In considering the value of a baker’s boy’s testimony against that of a Maréchal de France, historians have always sided with the latter. Yet Villars’ last letter to Louis before meeting Cavalier, dated 13 May, corroborates the Camisard’s version on every point: ‘My view [is] to never allow them but to pray to God in themselves as they wish, without any appearance of an assembly. I also believe that they should not be forced into any exercise of the Catholic Religion.’⁵⁶ Three days before the Recollects meeting, Villars supported granting the rebels freedom of conscience and likely promised it to Cavalier without Louis’ permission. This is probably what infuriated Basville during their meeting and what almost compromised Cavalier’s surrender.⁵⁷ Indeed, Villars’ letter echoes almost exactly the terms of article 1 of the treaty, which respectively guarantees freedom of conscience, bans public worship and exempts Protestants from attending Mass. He promised freedom of conscience as a short-term pragmatic attempt to appease the Camisards long enough to lure Cavalier into leaving the country with his troops. Further, he allowed them to stay in Calvisson (18–28 May 1704) until their agreement was approved by Versailles and to worship publicly for the first time since the Revocation. Up to 12,000 Protestants, according to both Catholic and Protestant sources, prayed and sang psalms together, and Cavalier reportedly preached on the main square. Local priests lamented that many young prophetesses fell under ecstasy claiming to be possessed by the Holy Spirit.⁵⁸

This shocking spectacle only helped spread rumours that Villars had indeed restored the Edict of Nantes in Languedoc. The fact that Major de Bombelles promised freedom of conscience to the Camisards in Calvisson without his hierarchy’s permission did not help either.⁵⁹ Villars came under fire at court, where Louis Phélypeaux, Marquis de la Vrillière, Secretary of State for Protestant Affairs and close to Mme de Maintenon, publicly accused him of having conceded freedom of conscience to the rebels: ‘It is reported that I have granted indefinitely the free exercise of religion, and that I only owe to this condition the return of those who surrender’.⁶⁰ Villars categorically denied such allegations in his defence both to Chamillart and Louis XIV:

it would be a capital [crime] to promise freedom of conscience when the rebels who have surrendered never thought to demand it, & that I made clear that I would have hanged the first man daring to bring me a letter from the rebels that mentioned this & any other condition than to beg only His Majesty’s clemency.⁶¹

In his ordonnance dated 29 May, Villars further protested that the Camisards:

instead of following all the commitments they have made by requests they have signed, by letters they have written, & by promises they have given us themselves, (they) some disrupters & enemies of peace have thought only of creating false expectations of freedom in the minds of the people for the exercise of the so-called Reformed Religion, for which no proposal has ever been made.⁶²

Villars’ defence is further evidence of his duplicity. For whereas he claimed to rule Languedoc with an iron fist, freedom of conscience, as we have seen, remained indeed central throughout his negotiations with Cavalier.⁶³

Articles 2 and 3 are identical in all seven versions of the treaty, albeit with slightly different wording. They respectively guarantee the release of all prisoners and galley slaves arrested on religious grounds since the Revocation within six weeks and authorize exiled Huguenots to return to France provided that they swear allegiance to the king.⁶⁴ Regardless of the treaty’s authenticity, both requests ought to be credited to Cavalier, who had indeed insisted on these terms in his first letter to Villars on 30 April and again at the Pont d’Avènes on 12 May. By contrast, the return of Huguenot exiles was never put in writing and is nowhere discussed in official correspondence. Still, the Nîmes magistrate Charles-Joseph de La Baume, author of one of the most important memoirs written during the revolt, not only confirmed that Cavalier had asked for the return of exiled Huguenots, but also that ‘M. de la Lande led him to believe that these conditions would be accepted’.⁶⁵

Evidently La Lande went to great lengths to persuade Cavalier to sign his surrender at the Pont d’Avènes, so much so that, on 17 May, Villars felt compelled to justify to Versailles his own meeting with the Camisard at the Recollects:

It was vital that I speak to Cavalier & his people, primarily because the Marquis de la Lande believed that it was essential to allow them the sale of their goods & the freedom of the prisoners, and that with these two conditions, they could have negotiated infinitely, which was not acceptable. They did not even mention it to me. Cavalier has always been deeply respectful, relying on Your Majesty’s grace.⁶⁶

Villars was once again playing on words, this time using his diplomatic tact to appeal to Louis XIV’s clemency towards the rebels. As we have seen, the terms of articles 2 and 3 had been discussed before. Once Cavalier had taken the bait, Villars and Basville downplayed his requests and negotiated instead based on his second, unconditional surrender of 13 May.⁶⁷ This is reflected in the time lag of the correspondence between Nîmes and Versailles, distance remaining without a doubt the greatest barrier to the Sun King’s military absolutism. Two days after the Recollects meeting, on 18 May, Louis was only responding to the Pont d’Avènes conditions received the day before and approved Cavalier’s requests:

I see nothing in Cavallier’s declaration that cannot be granted to him & to those who have become mixed up with him in this revolt. You [Villars] can in my name give them whatever written assurances they desire that the lives of those who will want to remain in my Kingdom will be in complete safety. […] I am willing to grant pardon to those in the prisons […] I am also willing to allow those who will leave my Kingdom to sell their goods […] In the written statement you will give to Cavalier, you will force him to leave my Kingdom as soon as possible, & to leave in the first of next month with those willing to follow him.⁶⁸

Contrary to what French propaganda and modern historians have claimed, Cavalier’s surrender was anything but unconditional. The release of some Protestant prisoners as a result of these negotiations provides further evidence of an agreement. In his testimony given in London in January 1707, the ‘French Prophet’ Jean Cavalier of Sauve, alleged cousin of the homonymous Camisard leader, declared: ‘I was released from Perpignan prison (even though I had been sentenced to life imprisonment) by an article of the Treaty that Mr Cavalier had made’.⁶⁹

The remaining articles vary from one version to another and reflect the technical nature of the negotiations. Article 4 concerns the restoration of the parlements’ original powers and privileges—that is the chambres mi-parties de l’Édit—abolished in 1679.⁷⁰ This was a key demand for the rebels, for these courts consisted of an equal number of Catholic and Protestant judges, thereby guaranteeing Huguenots a fair trial as well as an opportunity for social mobility for their local elite.⁷¹ Unsurprisingly, this was allegedly denied by Villars in all versions, except for Cavalier’s memoirs, which state ‘the King will Advise’, perhaps as a late attempt by Cavalier to overstate his achievements at the Recollects. Likewise, the fifth article grants a tax exemption for Protestants both during and after the war, except again in Cavalier’s memoirs, where it was reportedly denied.⁷² Of all seven versions of the ‘Treaty of Nîmes’, article 5 is the only one to feature contradictory responses, which further confirms that Cavalier’s memoirs were composed after copies of the ‘treaty’ began circulating in Europe.

Article 6 reveals precious chronological clues about both the negotiations between Cavalier and Villars and the composition of the treaty. It allowed Cavalier to raise a regiment of 2,000 Camisards to serve Louis XIV in Portugal and granted a general amnesty to those who would surrender. All seven versions of the treaty are consistent on these terms, except for their presentation. The BL copy and its published translations all list these conditions not in one, but four distinct articles (6, 7, 8, 9). As established earlier, the BL copy was circulating between England and the Dutch Republic by 1707, but it was not the original draft. The question of a Camisard regiment serving the Sun King abroad is central here. Throughout the negotiations with La Lande and Villars (12–16 May), the prospect of a Camisard regiment was only discussed orally because Cavalier himself refused to put it in writing, possibly lest he be accused of sedition.⁷³ Moreover, his regiment was originally destined to serve in Germany and leave France via Geneva. The Camisards had first offered to form a regiment on 25 April—three days after Villars’ arrival in Languedoc—in return for their surrender. The Maréchal judged that ‘this would be very good’ at the time.⁷⁴ Chamillart’s response, dated 12 May, was unequivocal about Versailles’ position and feared that the rebels might join the Protestant coalition once in Germany: ‘The King did not like the proposition you [Villars] made to form a regiment of those Camisards who surrender to serve in Germany; those people still have a mind too full of their own rebellion to make good use of it, & we should not think about it’.⁷⁵ The same day, Cavalier renewed his offer to La Lande at the Pont d’Avènes. Villars was desperate to negotiate the Camisards’ departure from Languedoc, but also concerned about their loyalty. On 13 May, he wrote to Chamillart: ‘I must beg you, Sir, to send me various routes to send these people out of the Kingdom, whether they want to go to Geneva or to Strasbourg to reach Holland. I am convinced that it will not be impossible to commit them to our service’.⁷⁶ The next day, he added:

I beg you, Sir, to please honour me with His Majesty’s orders on this regiment, if it can be set up; for I would find it better still to have these people for us than against us; it is quite certain that they will increase our enemies’ battalions, the moment after they have left the Kingdom.⁷⁷

Catholic and Protestant sources concur that the formation of a Camisard regiment due to serve against Portugal was first discussed at the Recollects on 16 May.⁷⁸ Villars sought to drive the Camisards out of France and promptly realized that Spain was a much safer destination than Germany to prevent the Camisards from turning against their sovereign. The French were well aware of the Anglo-Dutch coalition’s planned military and naval expeditions to support the rebels in Languedoc. Sending a Camisard regiment to Protestant territories constituted a risk that Versailles could not afford to take. So, Louis reluctantly agreed to Villars’ proposition on 24 May:

I am not convinced, wherever it may be, by the prospect of a regiment at my service that incites in my subjects & Enemies an idea of revolt [...] All I can do is to agree that those who have decided to leave my Kingdom will go to Spain to serve in my troops against the Portuguese.⁷⁹

In addition to offering an insight into the European stakes of these negotiations, this sudden change in destination helps date the composition of the treaty. Since all seven versions include an article relating to service in Portugal, the agreement could not have been drafted before 16 May 1704 at the earliest. The fact that all manuscript copies antedate the Recollects meeting confirms, in light of article 6, that they were written retrospectively from memory, shortly after Cavalier departed from France.

The final two articles appear consistent in form, though with slightly different wording every time. Article 7 addresses the restoration of Montpellier, Sète and Perpignan as Protestant strongholds guaranteeing the safety of the Reformed community and the free exercise of their faith. The Edict of Nantes had created 150 of these safe havens across France in 1598, but all had been abolished at the peace of Alès in 1629.⁸⁰ Cavalier’s request to restore them was denied. By contrast, Article 8 allegedly granted a seven-year tax exemption to all Cévenols whose houses had been destroyed during the war, a clause that appears consistent across all seven versions of the treaty.

IV

As he walked out of the Recollects, Cavalier rejoined his troops, who were eager to hear about his agreement with Villars. His response dealt a sudden blow to the Camisards, as one of them, Jacques Bonbonnoux, recalled:

As soon as he returned to the troop, we wanted to know what the articles of the treaty he had just agreed were; but, since he did not judge it appropriate to tell us, our principals urged him [...] so that he was forced to declare that red clothes were being made [for us] in Nimes, and that we were to be sent for Portugal.⁸¹

Cavalier’s visible unease further suggests that he could not produce a written agreement from his meeting with Villars. Instead, he had placed all his trust in the Maréchal and had to wait an excruciating ten days for the king’s response to Villars’ bold promises. Preserving the secrecy of the transactions was vital to Louis, who feared that the Anglo-Dutch coalition might propagandize a written agreement against him while he was at war against most of Europe. When the Camisard leader Elie Marion negotiated the exile of another group of rebels with Villars in 1705 and requested a signed agreement, the Maréchal answered ‘that nothing had been signed with Cavalier […], that the King’s promise sufficed’.⁸² In short, both Cavalier and Villars confirmed, through Bonbonnoux and Marion’s respective accounts, that no treaty was signed on 16 May 1704.

Still, conflicting reports and rumours added to the confusion. The five-day delays in the circulation of information between Languedoc and Versailles further complicated the matter when, on 18 May, still unaware of the Recollects meeting, Louis responded to Cavalier’s agreement with La Lande at the Pont d’Avènes: ‘If they ask for an amnesty or a written pardon, I will have one sent to them, but it seems to me that you cannot take too much care to avoid giving the public an act that may spread among foreigners & from which they would benefit.’⁸³ Louis was evidently willing to make unprecedented compromises and thus to follow Villars’ decisions to end the revolt as soon as possible. His letter did not reach Villars in Nîmes until several days later, by which time the negotiations with Cavalier were over and his troop were waiting in Calvisson in full compliance with their agreement.⁸⁴ Meanwhile, news of a truce was beginning to spread in foreign gazettes, as we saw earlier. Time was therefore pressing and, like Cavalier, Louis had also placed all his trust in Villars. Yet the Maréchal proved a tactful diplomat coaxing both of his interlocutors to his own advantage. A letter from a professor of the University of Montpellier to a friend at the Sorbonne even cast doubts as to what Villars actually reported to Louis about his meeting with Cavalier: ‘I do not know whether Mr. the Marshal made an accurate report of this meeting to His Majesty; but it is at least certain, that a few days later he received an order from the Court, to end this matter completely, to the satisfaction of Rolland & Cavalier.’⁸⁵ Louis’ response to the Recollects talks, dated 24 May, followed Villars’ recommendations to grant a general amnesty, the release of prisoners detained on religious grounds, the right to sell their properties and to leave France, but excluded ‘allowing them publicly in their error’ as he had done since the Revocation.⁸⁶ By the time it arrived, however, the Camisards had been worshipping in public in Calvisson for two weeks with Villars’ permission and unbeknownst to the king.

The confidentiality of the negotiations between Cavalier and Villars followed by the public worship in Calvisson could not but create confusion in the minds of the coalition. The Nouvelles extraordinaires in Leiden reproduced via Zurich the letter of a Camisard under Cavalier’s orders, written from Calvisson on 26 May, which largely echoes the main terms of the Treaty:

My Brother, [...] I wanted to wait for the outcome of the requests we had made for the restoration of the Protestant Religion in all of Languedoc & in all other Provinces, with the release of all the galley Slaves, & of all the Prisoners on account of Religion. This was granted to us by the Marshal de Villars & by Mr. the lntendant [Basville] with all their Council. But they said they wanted to write to the King, & ordered us to gather here, with the promise that they would give us His Majesty’s answer 10 days later. […] Our Brother Cavalier did not go to Nimes as you think he did; but it is true that he received Letters with proposals for Peace.⁸⁷

Three days later, on 29 May, the Duke of Marlborough wrote from Mainz to the Dutch Grand Pensionary Heinsius that the Camisards had capitulated and called for an immediate halt to the financial and military support for the rebels. By 4 June, the court of Savoy had received intelligence from its spies in Languedoc that ‘His most Christian Majesty [Louis XIV] has ratified what Marshal de Villars had promised to the Camisards’.⁸⁸ An anonymous letter written around the same time that I discovered in Berlin even proclaimed that the Camisards had recovered their freedom of worship and the right to rebuild their temples—most likely based on the public celebrations in Calvisson—and exhorted Huguenot exiles to return to France to rescue their brethren from prisons and galleys.⁸⁹

Despite the confusion created by such conflicting reports, the coalition promptly celebrated the Recollects meeting as a humiliation inflicted upon France. The English press praised Cavalier’s achievement as a symbol of the Sun King’s decline. Comparing the Recollects meeting to Aesop’s fable of the Lion forced into a truce after failing to catch a fly, the Monthly Journal for July 1704 exulted:

For my part, I am almost in an Extasie, as often as I think of it. How was it possible for so great a Monarch to Stoop so low, and that at the End of a long, glorious, and arbitrary Reign? […] When a Prince signs by force an Amnesty, he signs a Diminution of his Power, which is his dearest and most precious Jewel […] But King Lewis has done more than that, […] [he] has done [the Camisards] the honour to Treat, to Stipulate, and enter upon a Negotiation with them. […] The Most Christian King, who makes all Europe tremble, sends no less than a Marshal of France, to Treat in his Name with one of his Highlanders […] as Nature makes sometimes sad Tools with the finest Blood, so she makes often her Master-pieces out of Dirt. […] A Baker’s poor Prentice falls in the Way of Lewis the Great, and checks him in his full Career for Universal Monarchy.⁹⁰

In many ways, these words reflected the coalition’s growing momentum in the War of the Spanish Succession. It proved prophetic as the wind was about to turn in their favour. Indeed, having recalled Villars from Bavaria to end the Camisards’ revolt, France suffered a milestone defeat a month later at the battle of Blenheim in August 1704.

Although the Recollects meeting dealt a severe blow to Louis XIV’s image internationally, the reality on the ground was more complicated. Disappointed with the terms of Cavalier’s agreement, most Camisards refused to serve Louis in Spain against Portugal and only a minority followed Cavalier to Calvisson. Meanwhile, Tobie Rocayrol, another envoy of the Anglo-Dutch coalition after David Flotard, had arrived in Nîmes from Turin at the end of May and compromised further negotiations by promising imminent support from the coalition.⁹¹ According to Villars’ nephew, Rolland sought to follow Cavalier’s footsteps and had sent two envoys to negotiate with the Maréchal on 4 June. ‘This renew’d the Treaty’ and, as with Cavalier, the same ‘Articles were agreed upon’, including ‘a full Liberty of Conscience’, but four days later because of Rocayrol’s interference, ‘Villars declar’d that the Negotiation and Treaty were broke’.⁹²

Back in Versailles, Louis was forced to approve the terms of the Recollects meeting that Villars had submitted him. The Daily Courant reports ‘from the Paris Gazette a la main’ that:

The King has sent his last Orders to the M. de Villars to conclude the Accomodation with the Camisards; they are to have Passports to go out of the Kingdom, part by the way of Geneva, and part to the Rhine and so into Germany; and will be permitted to sell what they have, or to leave a Power to others to dispose of their Effects for them after they are gone. But the last Letters say, the chief of them are unwilling to leave the Kingdom, and demand only a Pardon and to be admitted to Serve in the King’s Armies.⁹³

The fragile mutual trust established in May was rapidly breached as it became clear that Louis would not honour Villars’ risky promise to restore freedom conscience after Cavalier’s departure from Languedoc. By mid-June the Camisards felt their leader had been lured into surrendering and vowed to the coalition to continue fighting:

As for us, we are determin’d to fight as long as God shall give us strength, and to the last Drop of our Blood; […] We have declar’d unanimously that the first among us who shall talk of Peace upon any other Terms than restoring the Protestant Religion, shall suffer Death without Mercy.⁹⁴

V

After the Recollects, Cavalier and his men stayed in Calvisson until 1 June 1704, while waiting for Louis’ orders before their departure to Spain. Their destination changed once again on 29 May, when Chamillart informed Villars that they would go to Neuf-Brisach on the Rhine and serve in Germany under Maréchal de Tallard.⁹⁵ Chamillart’s orders did not reach Languedoc until a week later, by which time the truce of Calvisson had ended and the Camisards faced a new wave of repression. Cavalier eventually left on June 23 with 100 men, heading towards Versailles.⁹⁶

Although he was appointed lieutenant–colonel in Mâcon on 9 July, mistrust continued to grow on his way to the Court.⁹⁷ The same day, Baron d’Aigaliers presented the Camisard with a new plan to serve under Prince Ferenc Rákóczi II (1676–1735) in Hungary against the Habsburgs. The Marquis de Bonnac, French ambassador to Sweden, based in Danzig and in charge of both Polish and Hungarian affairs, had negotiated an agreement with the Magyar regime to send the Camisards to Hungary, which Cavalier accepted.⁹⁸

Meanwhile, rumours had begun to circulate at Court to discredit Cavalier ahead of his arrival at Versailles. As is often the case, foreign observers offer a precious source of information in this context. The Swedish ambassador Daniel Cronström asked his correspondent on 5 June 1704:

On the matter of Cavalier, do you know, Sir, who he is? He is M. Tiquet’s man. This was not known when he was granted an amnesty and up until now there has been a willingness to keep the promise regarding the crime of rebellion, but no one knows what to do about the murder.⁹⁹

The name ‘Tiquet’ referred to a recent cause célèbre: in 1699, the wealthy and adulterous Angélique Carlier was beheaded after conspiring to assassinate her husband Claude Tiquet, an impoverished conseiller at the Parlement de Paris.¹⁰⁰ This tragic affair was well known in Versailles and the rumour was clearly intended to smear Cavalier ahead of his visit to the Court.

It was in this toxic context of rumours and suspicion that Cavalier met Chamillart in Versailles on 16 July to receive his colonelcy and 1,500l pension, but that he was also denied freedom of conscience for his coreligionists.¹⁰¹ Whether or not he really met Louis XIV, as Cavalier’s memoirs state, is another claim that historians have dismissed as an affront to the Sun King.¹⁰² But their interpretation is based on well-known, hostile sources such as Saint Simon, Mme Dunoyer and Voltaire.¹⁰³ An account of such an incongruous encounter was already published early in September 1704, in which Louis allegedly asked Cavalier to convert to Catholicism, to which the latter answered: ‘my Body & my goods are Your Majesty’s, but my Soul is GOD’s, who alone will dispose of it’.¹⁰⁴ French authorities denied of course that such a meeting between Europe’s most powerful ruler and a baker’s boy ever happened, but foreigners once again leaked Versailles’ best-kept secrets. Madame la Princesse Palatine wrote on 20 September 1704:

Only at Marly did I learn that Cavalier saw the king at Versailles and if Dibagnet, the caretaker of the Palais-Royal, had not by chance dined with him, I would not yet know it still. No one had been informed of this at Versailles. He is certainly a fanatic. I do not think Villars can overcome the camisards, he is too romantic for that…¹⁰⁵

And again on 10 June 1706:

The king would be very surprised if he were told that I know the entirety of the conversation he had with Cavalier … I find it very foolhardy to have told the king to his face where he had sourced his weapons and ammunition from. Villars, without doubt, committed to him more than the king had allowed him to promise…¹⁰⁶

Madame’s account counterbalances the Court’s secrecy and the natural bias of French sources. Not only does she corroborate Cavalier’s account on every point, but she also further substantiates suspicions that Villars promised the Camisard leader more than he could deliver without compromising himself before Louis.

The rest, as they say, is history and is well documented. Cavalier returned to Mâcon for nearly a month until his regiment was sent to the fortress of Neuf-Brisach on the Rhine in late August. Fearing his imminent imprisonment amid conflicting diplomatic signals and rumours, he escaped to Lausanne on 26 August, never to return to France again.¹⁰⁷ His grievances were made public in Geneva on 6 September: ‘Cavalier pretends that the French King having not set at liberty the Protestants on board the Galleys, nor taken care to perform several other Articles, which were part of his Treaty, he did not think himself obliged to perform that agreement.’¹⁰⁸ Another letter from Bern from the same day broke Cavalier’s silence over his capitulation for the first time since the Recollects:

[Cavalier] replied to them that he would never have come to an accommodation, if he had not had their word that his closest Relatives would be freed from their Shackles, together with all those of our Religion who groan in the Prisons & in the Galleys: And if he had not also been promised freedom of Conscience for them all, excepting, he says, Public Worship in the Temples.¹⁰⁹

In many ways, this letter paved the way for the ‘Treaty of Nîmes’. For, while Cavalier only mentions verbal promises made by La Lande and Villars, their content respectively corroborates articles 1 and 2 discussed above. Evidently Villars lured the Camisards into believing the truce of Calvisson restored freedom of conscience. His deceitfulness had become clear among the coalition by the summer. A satirical dialogue imagined between Rolland and Ravanel, two Camisards leaders who refused to surrender, captured the Protestant perspective on the negotiations: ‘be assur’d [Villars’] Orders from the King are to murder all, to promise any thing, and keep his word in nothing […] Too visible an Instance of his Treachery we’ve seen already, how well he kept Conditions with Revolted Cavalier.’¹¹⁰ Cavalier himself felt betrayed and considered returning to the Cévennes to resume the insurrection at the beginning of December.¹¹¹ At last in 1705, more ‘Camisards […] came out of France by virtue of the Treaty concluded by the Famous Cavalier, with the Marshal of Villars’ and went to the Canton of Zurich.¹¹² But this meagre gesture did not suffice to convince the remaining Camisards, who judged Cavalier harshly as a result. Elie Marion, who himself negotiated with Villars in October 1704, recalled years later:

Marshall Villars […] found means to deal with Cavaliere by some flattering Spirits who surpriz’d him and caus’d him to fall into the Snares which the Devil had laid for him by pride, he hearkened to the flattering and deceitfull Voice suffering himself to be blinded by the promises and hopes which did flatter his pride and Ambition and he yielded and did Capitulate and Laid down his Armes […] Cavaliere haveing thus withdrawn himself in May 1704 from among the Combatants The Enemy granted to him some of the Prisonners among the oppressed who with a small Number of those that had follow’d him were to be imploy’d for the Kings Service.¹¹³

Cavalier’s word and credibility suffered irremediably in all this confusion, deceit and disinformation. Perhaps the last word on the Recollects meeting and the true nature of Cavalier’s surrender should go to Villars himself, whose memoirs once again contradicted decades later what he had reported to Versailles at the time: ‘From the moment that Cavalier began to negotiate until the end, he always acted in good faith. Several conditions were agreed and rejected before an agreement was reached’.¹¹⁴

VI

The Treaty of Nîmes, as a legally binding diplomatic document between the Camisards and the French authorities, never existed. A close examination of contemporary sources against the backdrop of a well-established chronology leaves no room for the ratification of such a document in the middle of May 1704. Not only do none of the seven versions presented above match the date of the Recollects meeting, but Catholic and Protestant sources concur that Villars never signed a written agreement with Cavalier. Instead, this fake document turns out to be a mostly accurate reconstruction of the conditions agreed verbally between the two protagonists.

Nonetheless, the forgery of the ‘Treaty of Nîmes’ should not undermine its historical value. Autopsying a fake document in its multiple versions offers considerable insight into the motivations behind its genesis and the period as a whole. As I have shown, news of a treaty preceded the Recollects meeting and emerged from the Pont d’Avènes capitulation on 12 May. The actual text, as it survives in various forms, was most likely composed in 1705–06, possibly when Cavalier arrived in The Hague to prove what he had actually negotiated. This mythical document should therefore be read as a faithful reconstruction of the verbal agreement that brought Cavalier to surrender and precipitated the end of the last French war of religion. Each of its articles, including freedom of conscience, had indeed been agreed to orally by either La Lande or Villars. The contradictions between Villars’ letters and his later memoirs reveal his pragmatic handling of the negotiations and willingness to overstep the king’s authority to his own benefit.¹¹⁵ That is not to say that nothing was put in writing. Louis accepted the conditions to Cavalier’s surrender at Pont d’Avènes, which the fake treaty subsequently incorporated, and he also instructed Villars to give the Camisard written guarantees after the Recollects meeting.

The treaty’s publication in the European presses from 1707 suggests that it served a very different purpose than rehabilitating Cavalier’s reputation two decades after the events. Instead, it fuelled a wider propaganda campaign by Huguenot diplomats, such as the Marquis de Miremont and his agent David Flotard or pamphleteers such as Abel Boyer, promoting the Camisards’ revolt to the coalition as key to defeating Louis XIV.¹¹⁶ Cavalier had by then integrated the Dutch army, while other prominent Camisards such as Elie Marion and Abraham Mazel were prophesying in London to build momentum for a military expedition to the Cévennes. Since none of the extant manuscript copies of the ‘treaty’ are in Cavalier’s hand, a third-party involvement by exiled Huguenots would also explain the inconsistencies between the seven versions.

More generally, this contested episode illustrates the urgency of conflict resolution in times of slow communication against the active propaganda of transnational Protestant networks during the War of the Spanish Succession. It also highlights the varying degrees of trust in the spoken word between different social ranks of the Ancien Régime. In matters of diplomacy more than anything else, spoken words fly away, but written ones remain.

An earlier draft of this paper was presented at the SFHS annual conference in Washington DC in 2017 and at the University of St Andrews in 2018. I would like to thank the reviewers for their insightful comments. I am grateful to Jeroen Duindam, Michiel van Groesen, David van der Linden and David Thompson for their feedback on earlier drafts of this article.

Footnotes