Religion and Diplomacy: The ‘Clash of Civilizations’ as Historical Libel

Abstract

The ‘Clash of Civilizations’ hypothesis suggested that global politics would revert to inter-civilizational (inter-religious) conflict with the end of the Cold War. Conceptual and empirical refutations followed, but the idea that pre-Cold War inter-polity interaction was generally characterized by such conflict was not addressed. We consider this a possible historical libel. First, we briefly review the position of major faith traditions toward making and keeping agreements with those of other faith traditions. Most forms of agreements are sanctioned, and there is inconsistent and minimal support for duplicity. Second, using the MATRS database of multilateral treaties, we identify 79 sovereign entities active between 1750 and 1900 (when multilateral treaties were numerous and official state religions were prominent), link states to their official religions, and analyze the pattern of 385 multilateral treaties’ signings. We conclude that there is no tendency among states with one official religion to avoid entering into treaties with those of other official religions. The ‘Clash of Civilizations’ hypothesis is a historical myth.

1. Introduction

In this work, we consider the suggestion that civilizational differences, defined primarily in the form of religious distinctions, increased conflict and hampered inter-polity cooperation broadly and historically. We address this question because we are concerned with the possible historical libel inherent in the post-Cold War ‘Clash of Civilizations’ hypothesis (Huntington 1993).

As the Cold War ended, the emerging and uncertain global system was a subject of significant concern. The search for a model of how the world might work was taken up by scholars from a variety of theoretical and methodological perspectives. The most prominent vision offered concerned a projected ‘Clash of Civilizations’. Shorn of ideological debate over ownership of the means of production, and relieved of a generally bipolar structure based loosely on that debate, the argument was that the global system would revert to an older historical pattern of hatred and conflict based on ‘civilizational’ differences. These historical differences have always generated, and would now continue to generate, intractable divisions among humanity.

The ‘Clash of Civilizations’ hypothesis generated a great deal of attention. It came from authoritative scholars at important universities, it was alliterative, and it predicted whom we would fear and hate. It played to the biases of some of the most traditional elements in US politics. Coverage in newspapers and textbooks helped reinforce this vision. Secretary of State Colin Powell (2001–2005) used the threat of a ‘Clash of Civilizations’ as a political bludgeon to clear away opponents to various forms of military intervention by suggesting that in the absence of certain forced changes, we would be adding fuel to a more serious ‘clash’ (Inglehart and Norris 2003).

However, the ‘Clash of Civilizations’ hypothesis is theoretically and empirically suspect. Ethnopolitical differences had long been an important element in political violence, but those that crossed ‘civilizational’ lines were neither dominant nor increasing with the USSR’s disappearance. Ted Robert Gurr, in his 1994 inaugural address as president of the International Studies Association, led a long line of both empirical and theoretical criticisms (Gurr 1994; and also see Gleditsch and Rudolfsen 2015; Inglehart and Norris 2003; Russett et al. 2000; Sen 2006; Shahi 2017). Today it may be suggested that the hypothesis is being kept alive primarily by its critics.

The intellectual dismantling of the ‘Clash of Civilizations’ hypothesis nonetheless overlooked a possible historical libel inherent in the argument. The ‘Clash of Civilizations’ did not describe post-Cold War interaction, but does it describe pre-Cold War interaction? Many critics assumed or suggested that the hypothesis was not correct currently. We suggest it was not true historically, either. Certainly, there are spectacular examples of inter-civilizational animosity, but was it a general trend? Instead of focusing on active conflict, we formulate this question in a different manner. We ask, did such boundaries circumscribe peaceful and cooperative interaction?

We address this question in two ways. First, we ask whether the religious doctrines that underpinned the differences suggested in the ‘Clash of Civilizations’ hypothesis provide the right of adherents of a given religious tradition to interact with those of other religious traditions in bad faith. Do those of other traditions merit our fidelity? Do authoritative religious commentaries suggest that we may make agreements with others that we intend in advance to disregard, specifically because they are members of different communities? Second, we ask whether religious doctrine created barriers to actually making agreements. The foundation of international interaction rests with the conclusion of treaties. We ask whether states with different official faith traditions avoided entering into treaties with those of other faith traditions, as would certainly be the case if duplicity and conflict were to be expected.

On the question of religious doctrine, we note that the earliest arguments regarding the acceptability of others as treaty partners came to rest on the issue of whether counterparties held on to any religious beliefs. The logic was that any belief in a superior or supreme being would allow oaths to be taken in earnest (Bederman 2001). We recognize that our short-hand reviews of different faith traditions cannot offer definitive conclusions on whether the making and keeping of agreements with others would always be viewed as legitimate and would always be upheld. Instead, we suggest that the identification of serious and authoritative arguments from within a faith tradition regarding the legitimacy of such agreements would be noted by others as evidence of their basic acceptability as treaty partners. If such arguments were missing, or worse, if there is ample support for the idea that such agreements can be made in bad faith, then it is likely that interactions based on oaths or promises would likely be avoided. What we find is that sophisticated and authoritative voices exist in the major faith traditions that accept the legitimacy of making and keeping agreements with those of other faith traditions.

On the question of hesitancy to interact with others, it is important to note that there are various reasons to suggest that co-religionists might conclude more treaties with one another and avoid those of other faith traditions. Shared languages and iconography should facilitate communication. Similar beliefs about the orientation of the deities toward duplicity or contravention of oaths should place negotiators in a more secure environment. Shared understandings of treaty language should reduce friction and suspicion. Individuals of similar religions should find it easier to seal important agreements with augmented social interaction. Similar dietary traditions would facilitate the use of banquets, toasts, and other public signals of amity, which are so important in early diplomatic interactions (Numelin 1950; Shannon 2008, pp. 81–102; Sharp 2009). Intermarriages could more readily establish the foundations for ‘kinship diplomacy’ both immediately and in later generations (Jones 1999; Meier [2000] 2004). Third-party guarantors might be more easily chosen. Religious invocations (used in European treaties up until 1879) could more easily be agreed upon (Phillipson 1916, p. 168). However, what we find is that between 1750 and 1899, there was no significant alteration in the tendency of countries with a given official faith to avoid entering into multilateral treaties with those of other official faiths. Indeed, multilateral treaty making is robustly heterogeneous, declining only as we approach the 20th century. We believe that the move toward homogenization is due to the declining number of countries with official faiths and our use of ‘secular’ as a faith category, which given the number of secular states interacting with one another, decreased the tendency for cross-’faith’ interaction.

An important concern may be raised regarding this work: whether we should expect either the ideas that underpin such consistency or other contextual variables, such as relative power resources, to remain the same over a century and a half. We know that relative power position played a role in the history of Islam’s relations with others (Sertkaya 2023), as well as on a broad range of issues regarding peace and conflict in Catholicism (Jose 2023). Keene (2012) speaks to the influence of changing ideas on treaty making in a global context. We offer a similar response to both. Our primary goal is to ask whether cross-civilizational participation in multilateral treaties, which might, for many reasons, be considered less likely than intra-civilizational agreements, presents us with any empirical differences. From this perspective, we are concerned with treaty signing, and not with variables that might explain differences in treaty-signing behavior. The fact that multilateral treaty-signing practice remains robustly heterogeneous across civilizations throughout the 150-year period in question suggests that neither it, nor ideas regarding obligation and international agreements or power contexts, present much of a barrier to a formal agreement. The ‘Clash of Civilizations’ hypothesis appears as poorly supported historically as it is in the more contemporary environment.

2. Religious Doctrines and Cross-Faith Treaty Making

We find little support from within the major faith communities for the idea that one should avoid or make disingenuous agreements with those of other convictions. This should not be surprising. Treaty making has been an important inter-polity process for nearly 5000 years, and was originally geared toward facilitating the making of agreements among highly diverse communities. Early treaties in western Asia were written in multiple local languages and read publicly on a set schedule; they would include a ‘deity list’ that included the most important divinities of each society involved and, most importantly, entreaties by leaders to bring destruction upon their own societies if they broke their agreements (Bederman 2001, p. 53). The challenge of making cooperative agreements across various boundaries is not new, and successful methods by which to carry this out are of impressive antiquity.

2.1. Judaism

Rules regarding respect for diplomacy and treaty making may be found in the Five Books of Moses and the works of the Hebrew prophets. In the Book of Genesis (especially 21: 25–34), treaties emerge in essentially the same form as domestic contracts (Rosenne 1958, p. 78). Such agreements are sealed with a religious oath, and the breaking of the oath was an act of sacrilege punishable in a severe manner.

There were also times when diplomacy and the making of peace were considered obligatory. While scripture dictates that seven historic tribal enemies could not be offered peace of any kind, all others had to be given a chance to make accommodations before any hostile actions could be pursued. If accepted, these peace agreements had to be respected (Deuteronomy 20). Rosenne considers these rules to be “the Biblical equivalent of the pacta servanda sunt rule … essential to the stability of any legal order” (Rosenne 1958, pp. 141–42). In all instances, a foreign people were to be respected if they accepted the basic foundations of civilization, referred to as the ‘Seven Commandments of the Children of Noah’, which included prohibitions against murder, adultery, robbery, eating the limb from a living animal, idolatry, blasphemy, and a positive injunction to establish courts of justice. If people followed those rules, then they were fit to enter into agreements that could not be abrogated even in instances of duplicity (Joshua 9), or for reasons of state. (As an example, Israel refused an advantageous treaty with Alexander the Great, as it would have abrogated a pre-existing agreement with Darius (Rosenne 1958, p. 142).) Diplomatic emissaries, including those of other faiths, are viewed as personally important in the work of Josephus ([93–94] 1998, XV, 5 para. 2), and are protected with a quasi-religious status.

In what is acknowledged to be the last great recapitulation of Jewish Law, the Code of Maimonides (1180), we find these rules still intact. Treaties of Peace had to be offered and respected among all civilized peoples. Kings were bound to follow these rules (Maimonides [1208] 1977, pp. 220–22).

The acceptability and importance of treaty making and sanctity across religious boundaries are reinforced by the prophet Amos who wrote circa 750 BCE. Amos outlines the sins of various peoples, illustrating what is to be considered unacceptable practice. He is especially harsh on the people of Tyre for abrogating a treaty of friendship with their neighbors (Amos verse 9 in Polley 1989, pp. 78–79). In doing so, Amos is reinforcing an existing norm regarding the inviolability of diplomatic agreements. Barton considers the question of the abrogation of treaties of peace and friendship to be problematic under the accepted (if rather minimal) norms of international behavior across the entire ancient near east at the time (Barton 1980, pp. 51–61).

From the perspective of Judaism, you both can (with few exceptions), and in some cases, must make treaties with those of other faiths, so long as those with whom you dealt maintained the minimum values of most religious groups in that period. Once concluded, such treaties had to be respected.

2.2. Christianity’s Major Variations

Christian Europe is said to be the cradle of modern international law, and the work of the Protestant Hugo Grotius in the first years of the 17th century is usually identified as its genesis. However, Grotius did not write in a vacuum. Hermann Conring, who taught at the University of Helmstaedt in the 17th century, suggested that if Grotius “excelled in philosophy and produced the incomparable book, De Jure Belli Ac Pacis, he owed it to his reading of the Catholic jurisconsults” (Corning cited in Nys 1917, p. 98). These included Balthazaris Ayala and Francisco Suárez (Westlake 1912, p. i). Writing in 1582, Ayala acknowledges that the terms of a peace treaty must be obeyed, and that duplicity is not permissible (Westlake 1912, v2, p. 57). However, he also suggested there were broad exceptions when treaties could be concluded without any intent to hold to them. No obligations emerge in an agreement with unlawful enemies: pirates, brigands, or rebels. This is a traditional exclusion. However, Ayala then turns his back on his earlier argument and suggests that a treaty that included provisions that might be offensive to God or opposed to the interests of the state could be concluded in bad faith (Ayala in Westlake 1912, v2, pp. 57–68). The former has a specifically religious interpretation, though the latter could be applied equally to co-religionists. This is a particularly muddled defense of treaty making across religious lines, but it was Ayala’s argument in favor of fidelity that was extended by other scholars.

Writing 30 years later, Francisco Suárez suggests the following:

… treaties of peace and truces may be placed under this head [that is, under the law of nations, or ius gentium in the strict sense of the term]; not in so far as relates to the obligation to observe such treaties after they are made, since this obligation pertains rather to the natural law, but in so far as [offers of] such treaties should be heeded and not refused, when presented in due manner and for a reasonable cause.

(Suárez et al. 1944, p. 348)

His conclusion is particularly powerful since he defines treaties as falling under natural law and therefore relevant to everyone, without exception, as “natural law is truly and properly divine law, of which God is the Author” (Suárez et al. 1944, pp. 131–33, 198).

Grotius himself, hailed as the father of international law, raises some inconsistent arguments. Though in one place Grotius proposed that international law only emerges “among nations of the same class or family, united by the ties of similar origin, manners and religion” (cited in Janis 2004a, p. 134), he provides later an unequivocal statement regarding the legitimacy of treaties made across religious boundaries:

Concerning Leagues [security agreements], it is often disputed whether they may be lawfully made with those who are not of the true Religion, which is not to be doubted in respect to the Law of Nature…For the Right of making Alliances is common to all Men, and admits no Exception on the Account of Religion.

(Grotius 1715, Book II, p. 342, all capitalizations here and used subsequently are original)

Grotius then turns to the question of whether treaties between those of different religions can be abrogated without cause or prejudice. Among the Hebrews, he argues, treaties were readily made with ‘Others’ and were to be upheld without distinction. This is true both in the area of security and for “leagues of commerce” which Solomon made “according to the Wisdom GOD had given him” (Grotius 1715, Bk 2, p. 343). Grotius continues with this logic: “the Gospel has made no Alterations in this Respect [regarding leagues with ‘others’]; nay, it gives greater Encouragement to such Leagues, by Virtue of which, those who are not of the true Religion may be relieved in a just Cause; forasmuch as we are to do Good unto all Men … as what we are commanded and obliged to do” (Grotius 1715, Bk 2, p. 345). Grotius further strengthens this argument:

Nor must we conclude, that it is unlawful to make Treaties and Alliances with Pagans and Infidels, because we are not to put ourselves voluntarily under their Government, or to intermarry with them; for in both these Cases there is evidently more Danger of being exposed to the Temptation of renouncing the true Religion, or at least more Difficulty in maintaining the Profession of it ….

(Grotius 1715, Bk 2, p. 346)

Perhaps Grotius is most direct when he suggests, “For by the Example of GOD, who makes his Sun to arise on the just and unjust, and sends his Rain on the Wicked as well as the Righteous, we are taught to exclude no Man from the benefit of our Kindness” (Grotius 1715, Bk 2, p. 345).

For Grotius, the only possible exception appears to be in cases where we might be asked to deal with those who profess no faith, since the offering of oaths rests at the center of treaty making. Janis concludes the following: “Grotius sought to fashion a law of nations that could appeal to and bind Catholics and various Protestants alike … Treaties were to be fulfilled and interpreted in good faith … Even between enemies …” (Janis 2004b, pp. 217, 219). Subsequent European scholars followed Grotius in these regards. Emerich de Vattel’s The Law of Nations argued that any religion that would suggest that you could not make treaties with those of other faiths was not legitimate (Vattel [1758] 1916, p. 162). Janis reviews key authors and debates in international law in Europe from the 17th through the 20th centuries, and none support the suggestion that treaties may not be made, or are legitimately abrogated, simply because they are made among individuals or between states of different religions (Janis 2004a).

2.3. Islam

The first 16 verses of Sura Nine of the Quran speak to the making and breaking of treaties. This is an important topic in a religious tradition famed for taking the position that the non-Islamic world exists in a perpetual state of war. Nonetheless, treaties enjoy a status similar to that found in Christian traditions. The making of treaties is covered in the Quran, and failure to follow its injunctions constitutes a serious religious infraction (Holt 1980, p. 75). Sura 9:4 requires that treaties made with non-believers who do not engage in duplicity or complicity with an enemy must be fulfilled to the end of their term. “Thus the principle pacta sunt servanda is … recognized by all Muslim jurist-theologians” (Khadduri 1955, p. 203). Sura 9:7 suggests that treaties should be concluded near a Mosque, reinforcing their sanctified nature.

Islamic treaty norms are also derived from the experience of Muhammad. An agreement is understood to exist when an authoritative proposal receives an authoritative acceptance. Written codification is unnecessary (Khadduri 1955, p. 203). Treaties with those of other faiths who wish to live in peace in the areas controlled by Islam may be granted in perpetuity. Those among the subject people who contravene the terms of such a treaty are to be viewed as individuals, and not as representatives of their communities. Individual actions may not be understood to abrogate the treaty (Khadduri 1955, pp. 213–15).

Treaties with those outside the control of Islam are to be of a limited term, following the model established in the Treaty of Hudaybiya signed by Muhammad in 637. If Islamic forces are strong, the term should be no greater than four months. If Islamic forces are weak, then treaties may be for no longer than ten years (Badr 2004; Holt 1980, p. 67). An end date need not be specified, but termination may not come without due notice, and safe conduct must be provided for those who are threatened by such action (Khadduri 1955, 1966). Finally, treaties themselves must not be disadvantageous (Khadduri 1955, p. 221). If a treaty is made that surrenders arms or is deemed particularly contrary to the interests of the community, it may be considered ‘irregular’, and notice of termination, along with safe passage for all non-combatants, must be given.

There are a number of special provisions regarding treaties that were re-considered by Muslim clerics over time. Among the earliest and more comprehensive treatments is offered by Shaybani in the eighth century. Much emphasis is placed on travel and trade, especially on slaves. An excellent sense of the acceptability and sanctity of treaties with non-Muslims may be gathered from this reported exchange:

I asked: If the Muslims entered into a peace treaty with some of the inhabitants of the territory of war and these were attacked by … would it be lawful for the Muslims to purchase any of the captives? He replied: They should not purchase any of them, and if they did I would order them to send them back … For a narrative has been related to us from the Apostle of God, in which he said: ‘The one lowest in status can bind others if he gives a pledge …’.

(cited in Khadduri 1966, p. 192)

Later practice confirms and extends treaty making with non-believers well beyond the proposed time limits. Holt points out that the ten-year maximum is extended to treaties of peace even in eras when Islam was strong, as evidenced in the 13th-century agreements between the Ottoman Empire and Europeans (Holt 1980, p. 68). The making and keeping of agreements are well illustrated by the complexity of Ottoman diplomatic interaction both within the empire and with Europe (Hurewitz [1961] 2004, p. 305). Similar extensions were evidenced in Muslim Spain. By the middle of the 13th century, we find a web of cross-cutting treaty relationships that “mark the beginning of a major multicultural society” in which agreements were artfully crafted with every expectation that they would be observed (Burns and Chevedden 1999, p. 17).

2.4. Asian Traditions

Treaty making emerged in various Asian contexts. In India, Hindu principles were used to construct the norms of an interstate political system of significant complexity. Nanda (1991) reviews trans-state relations in ancient India, especially in the fourth-century BCE writings attributed to Kautilya. Treaties played an important role in maintaining peace and developing stable systems among units of different capabilities. Modelski notes that there existed a rudimentary collective norm against breaking treaties or profiting from such an act (Modelski 1964, pp. 555–57). These norms transcended Hindu areas. “The separation of law and religion had occurred by Kautilya’s time; interstate relations in India and further India were based on principles of secularism regardless of religion, race or ethnicity,” but more importantly, “Alliances and treaties were common practice; the principle of pacta sunt servanda was universally recognized” (Nanda 1991, p. 58).

In China, early treaty relations mirrored the blood oaths that sealed alliances among the aristocracy (Britton [1935] 2004). Confucianism emerged in China between about 700 and 500 BCE during a time of vicious political competition. Wars were plentiful, but “General meetings of the princes for the purpose of forming or renewing treaties of alliance were of frequent occurrence,” the envoys involved held a “sacred” status and had to be treated with the utmost respect, and “such intercourse was always characterized by courtesy, and mostly by good faith” (Martin 1883, pp. 70–71). Confucians argued that treaties were to be respected, both among those sharing that tradition and with other groups of ethically minded individuals. Chen explains that those who generally “followed the Chinese rules of conduct (‘li’)” were to be respected, unless they chose to “act as barbarians by engaging in duplicity” (Chen 2004, p. 34). Special provisions and dispensations were to be made for those from abroad to give them a chance to illustrate their intentions and good behavior.

Japanese religious studies are hampered by the changing definition of ‘Shinto’ over time. Modern formulations appreciate Shinto as the indigenous religion of Japan, but the term applies to any set of folk beliefs, and there were many in Japan as there were in China, where Shinto was also a term used to designate various modes of animism (Kuroda 1981). When Buddhism entered Japan in the sixth century, it was readily accepted and care was taken to make it consistent with the beliefs of Shinto, and vice versa. One could easily claim fidelity to both (Larsen 2018). Buddhist belief and Shinto both generally recognize the sanctity of promises made. Nothing in either belief legitimizes the idea that agreements with ‘others’ need not be followed.

A more consolidated form of Shinto became the official religion of Japan with the Meiji Restoration (circa 1866) when recognition of the superior power of the west led to the adoption of select western technology and an attempt to strengthen Japanese political and military power. This included the codification and adoption of Shinto, viewed in nationalist terms, as an official faith. Along with Shinto, Bushido (an ethical system for the warrior) was also adopted in an official manner. Like Shinto, Bushido went through many iterations, but the restoration period saw efforts to codify it as a professional martial code. Among the elements of Bushido are honesty and honor. Agreements made had absolute precedence, and breaking one’s word could require atonement by ritual suicide (Nitobe 1908). While there is ample room for expediency on the battlefield, none of these various elements countenance making agreements without the intent to uphold them.

Though religious commentaries on the issue of making and keeping agreements with those of other communities are not unanimous, we find strong and authoritative voices across the various traditions that contend that agreements may be made in good faith and should be upheld.

3. Cross-Religious Cooperation in Practice: Making Treaties

In this section, we interrogate the tendency of states with official religions to enter into treaties with those holding other official faiths. There may be a number of reasons to suggest that joining into agreements with those of other official faiths might be problematic. It is possible that religious injunctions against breaking agreements with others would reduce their incidence if there was concern that fidelity would not be reciprocated. Domestically, official state religions institutionalized discrimination at many levels. Church organizations were wealthy, powerful, and politically active. Mistrust was difficult to overcome, given the continued insularity of communal groups. (It was not until 1871 that religious affiliations could no longer be used to restrict access to elite education in Britain, which was among the more enlightened states). Distant ‘Others’, such as the Muslim Turks, generated fear and hatred (Knobler 1996, p. 197). It would be reasonable to suggest that accommodation would be facilitated if groups shared an official religious orientation, and more difficult if they did not.

There are also reasons to also suggest that co-religionists might conclude more treaties with one another. Shared languages and iconography should facilitate communication. Similar beliefs about the orientation of the deities toward duplicity or contravention of oaths should place negotiators in a more comfortable environment. A shared understanding of the treaty process should reduce friction and suspicion. Individuals of similar religions should find it easier to seal important agreements with augmented social interaction. Similar dietary traditions would facilitate the use of banquets, toasts, and other public signals of amity, which are so important in early diplomatic interactions (Numelin 1950; Shannon 2008, pp. 81–102; Sharp 2009). Intermarriages could more readily establish the foundations for ‘kinship diplomacy’ both immediately and in later generations (Jones 1999; Meier [2000] 2004). Third-party guarantors might be more easily chosen. Religious invocations (used in European treaties up until 1879) could more easily be agreed upon (Phillipson 1916, p. 168).

In order to assess the tendency of states with official religions to enter into treaties with those holding other official religions, we consider signatory behavior in multilateral agreements. Multilateral agreements are the most important indicators of global cooperative interaction. They constitute new global norms by virtue of their size, hold considerable legitimacy, and are less often held in secret (Muldoon et al. 2011). We argue that treaties have become such a “taken for granted norm” that we no longer appreciate the need to interrogate the practice (Wiseman 2011, p. 3; Glas et al. 2018).

Part of the reason for the “taken for granted” status is the length of time we have used treaties to codify agreements among polities. The known history of treaties extends back at least 5000 years. The transition to written treaties took place when scribes began recording ritualized interactions among kings who met to establish their relative status (McCarthy 1968, p. 22). A fairly consistent treaty format emerged in western Asia by the third millennium BCE and generally included an introduction of the authors, a stylized history of previous relations, an enumeration of the current relationship, specific sets of stipulations, a set of deities said to serve as witnesses or guarantors, and a set of blessings to be afforded for compliance, and curses to be leveled should the agreement be broken. As an indicator of the age and continuity of this treaty form, Mendenhall traces the western Asian treaty structure from the Hittites to the New Kingdom Egyptians (circa 1500 BCE) and to the Hebrew Book of Deuteronomy (circa 7th century BCE), while Weeks argues that the Israelites’ application of the treaty form was simply an innovative use of an already well-established treaty-making practice (Mendenhall 1955; Weeks 2004, p. 179). Formal diplomacy resulting in treaties of various sorts emerged with the earliest communities and can be understood as constituting international relations (Jönsson and Hall 2005; Glas et al. 2018).

The historical record finds direct evidence of diplomatic interaction and bilateral and multilateral treaties emerging at about the same time. Cooper, Heine, and Thakur suggest that the earliest known diplomatic document is a copy of a letter from Mesopotamian Ebla to Amazi (some 600 miles away) that was inscribed on a cuneiform tablet in about 2500 BCE (Cooper et al. 2013, p. 4). The first explicitly multilateral treaty for which we have evidence was also drafted in about 2500 BCE when a third-party sovereign was listed as the guarantor of the disputed boundary’s official markers between Lagash and Umma (Bederman 2001, p. 139). Multilateralism is also inherent in any agreement that requires alliance partners to mobilize in case of violence. The old dictum that ‘my friend shall be the friend of my friend and the enemy of my enemy’ has been part of diplomacy and formal treaty making since it was noted in Hittite letters to Amarna, Egypt, in the 24th century BCE (Munn-Rankin 1956, p. 93).

To help assure compliance, important treaties were written in the vernacular, chiseled into a stele, deposited in temples, and read to the public on a pre-set schedule. Local leaders from all signatory polities were quoted as asking their deities to bring ruin upon their own societies if they ignored or contravened treaty provisions (Bederman 2001, p. 53). Changes were allowed, and official alterations were chiseled into the text. Unauthorized changes were treated as a serious crime, though they may appear to posterity as vandalism (Bederman 2001, p. 173).

4. Treaties as Evidence

To interrogate patterns of treaty making, we use the Multilateral Agreement and Treaty Record Set (MATRS), which includes nearly 7000 multilateral treaties signed between 1595 and 1995. Multilateral treaties are defined as any agreement signed by two or more parties where at least one of them is a sovereign entity, and the other/s may be sovereign or groups of sovereign actors (international organizations or alliances). Included in that data are the signatory states, dates of signature, and two substantive categorizations (one broader, such as ‘war and peace’ or ‘transportation’, and one more specific, such as ‘prisoner exchange’ or ‘diseases at sea’). The dataset remains incomplete. Full signatory information is available for only about 6000 of the 7000 treaties noted, and the farther one goes back, the fewer complete elements there are (for a discussion of this dataset, see Denemark and Hoffmann 2008).

We will look at treaties concluded between 1750 and 1900. By 1750 treaties had largely lost their status as the private property of sovereigns and were being placed in public archives. States often held official religions during this period, making it more difficult to justify agreements with ‘Others’. We end our inquiry in 1900 as official faiths were being constitutionally abandoned at the dawn of the twentieth century.

MATRS includes complete signatory information on 385 multilateral treaties signed between 1750 and 1900. This is just over 50% of all 744 treaties listed in the database for that period. Our concern is that the missing treaties might bias the result. The argument derived from the ‘Clash of Civilizations’ hypothesis is that those from different civilizations would be less likely to sign agreements, while those who join together in cooperative treaties would be more likely to be religiously homogeneous. If the treaties for which we have incomplete information had parties that were primarily homogeneous in terms of their official religions, our sample would be biased against the hypothesis. To guard against this, we look closely at the treaties for which there is incomplete information. A total of 184 of those treaties, 51% of those for which we lack information, were signed by groups concerned with transportation and communication, international organizations, as separate annexes to Peace Conferences, or by groups such as the Zollverein. The treaties concluded by those groups, and other similar groups for which there was complete information, are predominantly heterogeneous in terms of the official religions of their members. This means that half the ‘missing’ treaties are likely to be heterogeneous in terms of the official faiths of their signatories. We conclude that using the sample of 385 for which we have complete information is not biased against homogeneity. (Both the overall MATRS database and the Religion and Diplomacy database may be accessed at Figshare, https://figshare.com. MATRS is 10.6084/m9.figshare.21936447, and the Religion and Diplomacy database is 10.6084/m9.figshare.21936024. Both were created and posted by the authors and accessed 2 March 2023.)

5. Defining an Official Religion

To determine official faith, we look for formal constitutional provisions, forced church attendance, or state support for building the churches or paying the clergy of a single faith. Some of these practices disappeared after 1750, but states often found other ways to express a formal religious orientation. Non-conformists might be granted the right to worship, but not to meet in groups beyond the family. Even if groups were allowed to meet in special-purpose buildings, we define a state as having an official religion if such buildings were not allowed to be identified by sight or entered from the street. Finally, an official religion exists if a specific denomination must be professed to enter into the service of the state or public educational institutions. Even where constitutions officially forswear such practices (as they did early on in France and Russia, or in many US states into the 19th century), we deem an official religion to exist wherever de facto practices limit participation in public life through mechanisms such as requiring potential officeholders or students to take denominational oaths, or if we note the regular use of intimidation (including assassination) of non-conforming political aspirants. Using these criteria, we were able to identify the official religious orientations of 79 sovereign entities.

By this definition, we will need to treat Protestantism and Catholicism as separate civilizational entities as well. There is no question that the Protestants and Catholics in Europe fulfilled all of these definitional elements of separate civilizations. Further, the policies of those countries ruled by adherents of one or the other also evidenced myriad forms of discrimination or suppression against the other, and sometimes outright violence. The ‘CofC’ hypothesis did not distinguish between Christian sects initially, but Huntington offered an analysis separating the historical US identity, derived from Protestant roots, from the threat of “Hispanization,” of which the majority Catholic faith of immigrants was identified as an important factor (Huntington 2004). While there is a relatively unified Christian tradition regarding treaty making, Protestants and Catholics nonetheless evidenced and retained separate civilizational foundations that were prominent during the time period we are considering, and arguably afterward as well.

Official faith is not always clear-cut. Some states abandon requirements regarding membership in specific sects but retain rules regarding families of religions. For example, some came to accept officeholders of any Christian sect, but not others. This might be true even when constitutional provisions suggested otherwise, as was the case with the formal, but not de facto, acceptance of Muslims in Catherine the Great’s Russia. We identified these states as ‘mixed Christian’, denoting that the historical rivalry among Catholics and Protestants had been overcome, but not bias against non-Christians.

We also wish to avoid separating specific state churches (such as the Church of Sweden) from those that were identical in terms of creed, but took different names (such as the Church of Norway). Treating these as separate religions would not do justice to our desire to identify truly different faith traditions, so we present our data using both all religions with different names, and also those grouped by religious families. (Information on religions and religious families is derived from Barrett et al. 2001).

We also chose to treat secular states as their own unique religious category. Our concern is to find out whether states with different official religious orientations will enter into treaties with one another. A secular state shares a religious orientation with other secular states, and is different from states with an official traditional faith. Put differently, we do not discount secularity as a belief system.

6. Official Religions and Multilateral Treaties

We measure the distribution of religious orientations and treaty making in four ways. Figure 1 presents an average of the number of religions and religious families represented in each treaty signed between 1750 and 1900 in ten-year increments. The figure offers two important insights. First, none of the 15 decades considered in this analysis finds multilateral treaties with an average of fewer than two states with different official religions. Over the 150 years considered here, multilateral treaty making was never a religiously homogeneous practice. These data raise an interesting question. If multilateral treaties were primarily used to end wars, and the ‘CofC’ hypothesis was correct in suggesting that wars would be more likely across civilizational boundaries, then we might expect treaties to be heterogeneous. This would support, not detract from, the ‘CofC’ hypothesis. However, the treaties that emerged during the 15 decades in question were not dominated by treaties of peace that followed war. Treaties of ‘War and Peace’ (one of the categories MATRS adopts that was derived from UN designations) are dominant in the data only through 1819, after which they fall to below 50% of all treaties concluded. Our review of the treaties that fall into the ‘War and Peace’ relative to those in other categories is just below 23%. Overall, we note that treaties of ‘War and Peace’ were the largest contributor to the dataset across this entire period, but nonetheless constituted just under one-third of all treaties included. The idea that treaties were basically instruments used to end conflicts, and might therefore be expected to cross civilizational lines, is not accurate.

A second conclusion we may draw from Figure 1 is that the number of religions and religious families entering treaties tracks very closely. This suggests that the heterogeneous nature of multilateral treaty making is not an artifact of states signing treaties with those of faiths that differ in name only. Treaty signing heterogeneity is robust.

Figure 2 is sensitive to the size of treaties. We ask whether larger or smaller treaties appear more or less heterogeneous. We compare the average number of religions to treaty signatories. There are no significant differences based on the number of parties. Across 15 decades, we find the average number of official religions represented in multilateral treaties is greater than half the overall number of signatories. In four of the five decades between 1780 and 1829, they are within 10 percent or 15 percent of parity across treaty size. Multilateral treaty making is a religiously diverse activity throughout the period, regardless of the size of the signatory group.

We note that from about the 1830s onward there is a slight tendency for religious diversity to decline relative to the period from the 1770s through the 1820s. The tendency is not great, but is contrary to general perceptions of increasing secularization through the mid-19th century. We consider this in more detail below.

In Figure 3, we consider coercion or tokenism. A treaty with three Catholic and three Protestant states includes parties with two religions. A treaty with five Catholic but only one Protestant state also includes parties with two religions, but our perceptions of their diversity are not alike. To capture this phenomenon, we measured the proportion of parties in the largest bloc by dividing the number of signatories representing the most common religion/religious family by the total number of signatories for each treaty. A measure of 0.5 represents a perfectly balanced representation. A measure of 0.7 suggests that at least 30% of treaty signatories are from numerically subordinate groups. Figure 3 shows that in only one-third of the decades considered is there a level as high as 0.7. We see more homogeneity among religious families, though less than half of our decades rise above the 0.6 level, and only two are greater than 0.7.

We see neither tokenism nor the type of lopsided plurality that would cause us to question the true nature of heterogeneity when considering official faiths (though we note a bit more homogeneity in terms of religious families). Counterintuitively, homogeneity among parties to multilateral treaties appears to increase as we approach the twentieth century.

While religious heterogeneity in multilateral treaty making is the norm, an interesting puzzle emerges. There is a slow upward trend in homogeneity from 1829 onward. We believe this is an artifact of growing secularity. We take secularity to be a religious orientation, and therefore a significant increase in the number of secular parties signing treaties may appear as a decrease in heterogeneity. Figure 4 tracks the rise of secular signatories both in terms of numerical averages (on the left y-axis) and as a proportion of all signatories (on the right y-axis, and note the lines do not match levels on the right side, which are denoted separately). The proportion of officially secular parties increases quickly from zero through the first three decades of our time period (through 1779) to 29 percent by 1799. From 1799 begins a slower increase, reaching one-third of all signatory states between 1840 and 1849 (just as our measures begin to show declining heterogeneity), and ends at 39 percent in 1899. This level of increase on the part of any of the faith traditions included in this study would necessarily increase the level of homogeneity among treaty signers. It is this significant increase in officially secular states that explains the growing homogeneity we see as we approach the 20th century.

7. Conclusions

Our major faith traditions have historically offered authoritative voices in support of the legitimacy of making and keeping agreements with others. While these are not the only voices that might be heard on such issues, and practice may sometimes differ, it is significant that this position exists. In its absence, it is unlikely that early treaty making would have become a widespread practice among societies with vastly different ideas about authority and appropriate forms of interaction. The traditional treaty-making form that emerged in the third millennium BCE was dependent upon polities having and respecting a faith tradition that accorded value to oaths. That treaty form remained remarkably stable over the years as a result.

The consistency of cross-faith treaty making led us to consider the ‘Clash of Civilizations’ hypothesis, which makes the claim that civilizations harboring different beliefs and ethics (often defined in terms of different religions) will create rifts across which we are likely to see conflict. Treaties, based primarily on elements of trust, would be vulnerable to the existence of such rifts. To see if such conflict-prone rifts are evident, we undertake an analysis of multilateral treaty making, the most transparent and important form of cross-polity agreements.

Identifying the 150-year period from 1750 to 1899 as relevant, given the prevalence of official religions among its polities, and the transparency of the multilateral agreements made, we consider the 385 treaties signed by 79 sovereign entities. We suggest that if conflicts or duplicity were more readily expected across civilizational/religious boundaries, there would be a tendency for treaties to be made among co-religionists and to be avoided across civilizational lines. We find instead that cross-faith treaties are the norm, and never the exception, across the entire 150-year period. This is true in large and small treaties, it takes into account the chance that some ‘token’ member of another official faith might be included in treaties that are homogeneous otherwise, and it considers official religions both individually and in terms of religious ‘families’ that share all critical orientations but may have different formal names (i.e., The Church of Norway and The Church of Sweden).

We note an anomaly. We see strongly heterogeneous treaty making from the very beginning of the period, but homogeneity increases as we approach the 20th century. We attribute this to the growing number of officially secular states. There were no such states during the first three decades of our analysis, but their number grew to 39% of all treaty-signing states by the end of the period. That large of an increase in any ‘faith tradition’ would militate in favor of greater homogeneity. Since we treated ‘secularity’ as a faith tradition (assuming that secular states might well view one another with greater levels of trust based on their similar orientations), we believe the increase in homogeneity is due to that change.

While we consider the foundations of different faith traditions and their orientation toward making and keeping agreements with others, we do not delve deeply into other ideational or contextual variables. It is doubtlessly true that different ideas and environments can have an impact on how we view and make agreements with others, but in the context of the simple existence of faith traditions that value oaths, those differences may be less fundamental than expected.

We conclude that the ‘Clash of Civilizations’ hypothesis is not supported by the evidence. Quite the opposite proves to be the case. Global-level cooperation has long been designed to deal with the challenges of cross-faith agreements, and evidence from the period from 1750 to 1899 shows no tendency for such differences to play a role in keeping polities with different civilizational backgrounds from engaging in mutually agreed-upon cooperative interaction. The ‘CofC’ hypothesis may have been a popular polemical position that sought to identify our enemies and mobilize a domestic following in the wake of the Cold War, but it is untrue not only in the contemporary era, but for a significant historical period as well.