ANALYTICAL FRAMEWORK I: FROM TEXT TO MEANING

Comparative constitutional lawyers have long examined the travel of constitutional ideas across national borders,Footnote ⁴ and international relations scholars have also begun to study the diffusion of international constitutional norms (Skach Reference Skach2005; Wiener Reference Wiener2014). Yet, much of this work has focused on the migration of ideas between judges and courts during moments of constitutional adjudication (Perju Reference Perju2003, Reference Perju2012). Emerging scholarship on the travel of constitutional models at the constitutional drafting stage has tended to approach the issue from a strictly textual perspective, focusing on particular clauses that travel or the appearance of international law provisions in domestic constitutional settings (Goderis and Versteeg Reference Goderis and Versteeg2014; Law Reference Law2016). We focus, by contrast, on politically inflected changes in the meaning of imported constitutional provisions both during and after their initial adoption.

As noted above, Alan Watson and Pierre Legrand disagreed about the ways in which laws travel, with Watson’s focus on direct textual transplants being criticized in terms of what Legrand called the “cultural embeddedness” of law.Footnote ⁵ Legrand insisted that textual transfers typically “fail” owing to intersubjective breaks at the level of legal meaning. Watson and Legrand were interested in private, rather than public or constitutional, law. But, more recently, Goldbach, Brake, and Katzenstein (Reference Goldbach, Brake and Katzenstein2013, 146, 142n2) have noted that, while “public law is [considered] … more resistant [than private law] to being transplanted,” constitutional law has nevertheless emerged as “one of the most important areas of public-law legal transplants.” This observation, however, is buried deep in a footnote; Goldbach, Brake, and Katzenstein do not focus on constitutional law themselves (see also Langer Reference Langer2004). Indeed, even when sociolegal scholars have shifted their focus beyond textual “transplantation” to consider the practice of “translation,” they have rarely discussed the politically inflected translation of imported constitutional provisions.Footnote ⁶

We address this gap by highlighting the ways in which transnational constitutional borrowing and translation are complex processes closely tied to local political considerations. In doing so, we move beyond the work of legal-diffusion scholars like David John Frank, Bayliss J. Camp, and Steven A. Boutcher (Reference Frank, Camp and Steven2010), noting that, as a consequence of these political considerations, the global diffusion of specific constitutional provisions is often divorced from strong forms of convergence at the level of legal meaning or practice.Footnote ⁷ In fact, we argue that domestic political actors often have clear motivations for invoking or importing foreign constitutional elements, even if they have no intention of faithfully transplanting the prior (“foreign”) meaning of those provisions within their own political context (Scheppele Reference Scheppele2003).Footnote ⁸

In this sense, our work returns to that of sociolegal scholars like Grattet and Edelman, even though these scholars do not focus on constitutional law themselves (Edelman and Suchman Reference Edelman and Suchman1997, 499; Grattet and Jenness Reference Grattet and Jenness2005, 893). Edelman, for instance, examines patterns of employer compliance with American civil rights law, noting that employers rarely accept or reject such laws but, instead, creatively or discursively adjust them (Edelman et al. Reference Edelman, Krieger, Eliason, Albiston and Mellema2011). Grattet, similarly, examines how the meaning of a particular statute is often transformed via discretionary patterns of decision making within whole “fields” of social, organizational, legal, and political action (Phillips and Grattet Reference Phillips and Grattet2000; Grattet and Jenness Reference Grattet and Jenness2005). In fact, with Maximo Langer (Reference Langer2004), both Edelman and Grattet show how the practical meaning of a legal text is closely tied to work undertaken by deeply contextualized legal “translators.”

In the realm of comparative constitutional law, the question of borrowing is often framed by references to global convergence or isomorphism—as demonstrated, for instance, in the “world polity” tradition or in Günter Frankenberg’s “IKEA” theory of isomorphism (Frankenberg Reference Frankenberg2010; Beck, Drori, and Meyer Reference Beck, Drori and Meyer2012; see also Waters Reference Waters2007; Law Reference Law2008; Gardbaum Reference Gardbaum2013). Frankenberg’s model of constitutional migration and isomorphism, however, includes a middle stage of transfer in which constitutional concepts move from a particular local context to a global “IKEA center” before the idea (now “globalized”) is adopted in its new location. By contrast, our study does not track patterns of global isomorphism via directly transferable meanings. We highlight, instead, the power of local political dynamics in shaping conceptual transfers themselves. In fact the question with which Frankenberg ends his article (how do constitutional transfers of meaning happen?) is the question with which we begin.

In this context, our work returns, once again, to law-and-society scholarship on the relationship between sub-constitutional legal texts and situated forms of legal meaning. Specifically, we show how forms of legal ambiguity growing out of what Ryken Grattet and Valerie Jenness (Reference Grattet and Jenness2005) call a “surplus” of legal meaning (where the same text lends itself to several different meanings all at once) interact with particular social, institutional, and political contexts such that ostensibly common textual elements map onto very different forms of legal practice. Grattet and Jenness confine their attention to domestic contexts in which a surplus of meaning allows ordinary civil and criminal laws to be actively and substantively reinterpreted.Footnote ⁹ We highlight the political underpinnings of foreign constitutional provisions, focusing on the politically sensitive issue of religion-state relations.

ANALYTICAL FRAMEWORK II: INTRODUCING EMPTY SIGNIFIERS

Combining the work of sociolegal scholars like Edelman, Grattet, Jenness, and Langer with that of Laclau, we stress the ways in which borrowed constitutional provisions, models, and ideas are also translated to reflect the explicitly political concerns of those who borrow them. Specifically, we argue that those seeking to grasp the dynamics of constitutional meaning must appreciate the politically malleable gap that semioticians often illuminate between a linguistic “signifier” and what it actually “signifies” (that is, between a given term and its meaning).

Social theorists like Ferdinand de Saussure (Reference De Saussure1959) and Claude Levi-Strauss (Reference Lévi-Strauss1963) have noted that the link between a particular text/word/sound, on the one hand, and its meaning, on the other, is deeply contextualized within a network of conceptual distinctions prevailing in a particular time and place. Highlighting forms of political contingency in particular, Laclau’s (Reference Laclau1996) notion of “empty signifiers” (alongside so-called “privileged” or “floating” signifiers) ties these insights to our own focus on the politically inflected meaning of borrowed constitutional provisions. Laclau’s concepts, we argue, help us clarify the ways in which “politics matter” during episodes of constitutional borrowing.

In Laclau’s theory, the meaning of a “privileged” signifier is relatively stable or, at least, widely regarded as stable. Liberal constitutionalists, for instance, often treat a term like “religion” as a privileged signifier when they assume that its meaning is broadly uncontested (thus overlooking the ways in which colonial or postcolonial regimes struggle to locate the complexity of “customary” spiritual practices within the conceptual domain of “religion”).Footnote ¹⁰ Those writing about global constitutional isomorphism often rely on similar assumptions regarding important constitutional terms like “secularism.” Across multiple contexts, they treat the meaning of such terms as plain.

Laclau’s “floating” signifiers are terms that appear detached from the broader conceptual networks that might otherwise endow them with clear meanings: their conceptual location, and, thus, their meaning, are seen as irredeemably vague. Critical legal scholars scrutinizing British colonial encounters with Hinduism, for instance, often describe references to “religion” in this context as a floating signifier given their unstable relationship with ancillary concepts like doctrine, clergy, or conscience.Footnote ¹¹ Similarly, scholars writing about “constitutional incommensurability” often treat textual references to “religion” or “secularism” as floating signifiers because their ambiguous meanings are thought to render the stable transplantation of such concepts difficult or even impossible.

In this article, we draw on the work of Ernesto Laclau to highlight an intermediate position, focusing on what he calls “empty” signifiers or terms whose meaning is neither stable nor irredeemably vague but, rather, politically contested (see also Reference GallieGallie 1955–56). We pay special attention to the ways in which historically and politically situated agents, drawing attention to a particular term (say, “secularism”), wrestle with its multiple meanings and, then, actively press for one—for example, French laïcité rather than Indian equal treatment—as an explicitly political act. As Laclau (Reference Laclau1996, 44) points out, highlighting the importance of interpretive agents in contexts framed not merely by organizational discretion (as discussed by Edelman and Grattet and Jenness) but also by explicit forms of political intervention during key historical junctures, the meaning of an empty signifier “only exists in the various forms in which it is actually realized”—that is, the forms in which political actors produce it.

In this sense, the meaning of the borrowed constitutional concepts we examine is neither stable nor untraceable. Meanings are, instead, deliberately decontextualized and, then, actively recontextualized through the deliberate interventions of domestic political actors. This is, in effect, what Langer (Reference Langer2004, 33) describes as the “translation” of legal ideas owing, in part, to a political struggle “between different actors and groups within the target legal system over the meaning of the translated [text].” We examine this process of translation or “re-signification” in four cases involving the borrowing of constitutional provisions concerning religion-state relations. We demonstrate how borrowed texts, clauses, concepts, or models are intentionally emptied of their prior content (“decontextualized”) and, then, cast as “empty signifiers” before being filled up again by politically situated agents pressing for new forms of constitutional meaning (“recontextualized”).

Before proceeding, it is important to note that drafters may borrow a foreign model or text they actively contest, engaging in the forms of re-signification we associate with empty signifiers, but then still choose to “re-signify” the borrowed language, provision, or norm with forms of content that are left deliberately vague (to preserve the possibility of more than one interpretation later on). This process involves two overlapping strategies at once: the adoption of an empty signifier and, then, the embrace of an incremental approach to determining its re-signification in the future (Bâli and Lerner Reference Bâli and Lerner2017). Still, the empty signifiers we examine are never completely meaningless; their meanings are simply produced within a particular context, with evidence of this playing out in competing assertions of constitutional meaning (for example, “inclusionary” versus “exclusionary” religion-state relations) in a particular time and place.

FOUR PATTERNS OF BORROWING, EMPTY SIGNIFIERS, AND RELIGION-STATE RELATIONS: CASE SELECTION

In what follows, we examine four different ways in which constitutional actors, having borrowed or invoked foreign provisions, go on to remake the practical meaning of those provisions in particular contexts. Through selected cases, we highlight specific modes of politicized borrowing focused on religion-state relations. In choosing our four cases, we have sought to challenge a number of common assumptions about transnational constitutional borrowing in the context of religion-state relations. For example, it is common to expect patterns of transnational borrowing in territories formerly governed by external powers to originate in the traditions or practices of former rulers (Britain vis-à-vis Pakistan and Egypt; France vis-à-vis Morocco and Egypt; the Ottoman Empire vis-à-vis Israel and Egypt; and so on). The cases we have selected, however, reframe this conventional expectation. In Pakistan, for instance, we provide an account of borrowing rooted not in the constitutional traditions of Britain but, rather, in global networks of anti-British (“home-rule”) activism extending from Ireland to India. As such, our cases partially decolonize the study of constitutions and the practice of constitutional borrowing, reaching beyond the colonial and postcolonial frameworks underpinning numerous studies of sociolegal translation to include colonial and noncolonial (here, mostly non-Western) sources of constitutional borrowing.

Beyond borrowed texts, it is also common to expect deliberate borrowing from secular republics (for example, India, France, and Turkey) or famously pluralist regimes like the post-Tanzimat Ottoman Empire—itself influenced by post-Enlightenment models borrowed from Europe—to facilitate constitutional orders supporting and legitimating religious, sectarian, doctrinal, or ideological “inclusion.” But, again, our cases have been chosen to challenge this assumption. In each case, politically charged translations of constitutional elements from India (in Pakistan), from France (in Morocco), from Turkey (in Egypt), and from the Ottoman Empire (in Israel) push in the opposite direction. Rooted in political processes combining religious pluralism with domestic struggles to frame an independent constitutional identity, constitutional borrowing from secular contexts in these cases facilitated the practical “exclusion” of religious others.

This exclusionary trend might be expected to vary across different regime types, with civilian-democratic regimes adopting a more “inclusive” approach than military regimes, authoritarian regimes, or monarchies. Yet, again, our cases refute this assumption. In what follows, an “exclusionary” reading of imported constitutional provisions concerning religion-state relations cannot be traced to any particular regime type: the same exclusionary reading emerges in all four cases. In Pakistan and Israel, constitutional borrowing overlaps with popular bottom-up translations that feed exclusionary forms of majoritarian religious nationalism (Pakistani “Muslim” nationalism excluding a heterodox religious group known as the Ahmadiyya; Israeli “Jewish” nationalism marginalizing non-Jews and non-Orthodox Jews). In monarchical Morocco and authoritarian Egypt, the same exclusionary outcome emerges via top-down forms of authoritarian state formation (Morocco’s Muslim king borrows from secular France whilst restricting electoral democracy; Egypt’s military references a “Turkish” constitutional model prevailing under the elected regime of Turkey’s Islamist Justice and Development Party [AKP] whilst constraining the electoral prospects of Egypt’s Muslim Brotherhood).

In what follows, the case of Egyptian borrowing vis-à-vis “the Turkish model” also shows how what we call foreign “borrowing” (combining references to foreign constitutional elements with domestic political debates that alter the meaning of those elements) is not confined to a narrow account of direct textual transfers. Like the rejected (“negative”) constitutional models studied by Kim Scheppele (Reference Scheppele2003), we argue that cases of foreign invocation without any transfer of constitutional text also illuminate an important feature of transnational constitutional borrowing (see also Salam Reference Salam2018, 3).Footnote ¹² Egypt’s public debate regarding religion-state relations, for instance, was repeatedly marked by rival invocations of “the Turkish model,” often by actors who selected very different aspects of that model to signal the (domestic) legitimacy of their own opposing political positions. The fact that Egyptian drafters later opted to avoid any direct textual transfers from Turkey does not diminish the power of the Turkish model as a borrowed constitutional “empty signifier.” On the contrary, the Turkish model was borrowed and politically re-signified in ways that directly shaped the meaning of religion-state relations in Egypt’s constitutional debates.

The four patterns of constitutional transfer and re-signification illustrated by our case studies can be summarized, briefly, as follows.

CONSTITUTIONAL PROVISIONS AS EMPTY SIGNIFIERS: RELIGIOUS FREEDOM FROM IRELAND VIA INDIA TO PAKISTAN

The most direct form of constitutional borrowing involves constitutional drafters importing explicit textual provisions from other countries. In 1922, Ireland’s provisional government published a book entitled Select Constitutions of the World (1922) with constitutions from the United States of Mexico, the Russian Socialist Federal Soviet Republic, and sixteen other countries to inform its constitutional drafters. Reflecting transnational networks rooted in anti-colonial activism, the book was republished in 1934 to assist India’s Constituent Assembly (Rao Reference Rao1934). This book was also combined with a second volume entitled Constitutions of Eastern Countries and used by the Constituent Assembly of Pakistan (Ahmad Reference Ahmad1951). Within Pakistan, these compendia set the stage for direct forms of borrowing—and, later, refashioning—foreign constitutional models (Nelson Reference Nelson2019a).

Archival records from India and Pakistan reveal extensive borrowing focused on religion-state relations and, specifically, enumerated and enforceable rights pertaining to religious freedom. Judicial and parliamentary records, in turn, reveal the ways in which grassroots violence perpetrated by conservative religious groups like the Majlis-e-Ahrar-e-Islam (Association of “Free” Muslims) informed mainstream constitutional debates in ways that later reconfigured prevailing forms of interpretation vis-à-vis Pakistan’s borrowed provisions. Briefly, grassroots violence inspired majoritarian notions of “Muslim nationalism” in which key state actors recast borrowed religious freedom provisions in an increasingly exclusionary mold.

Between 1922 and 1937, constitutional reformers in Ireland built on post-First World War constitutions in Eastern Europe—specifically, Poland’s “March” Constitution (1921)—to balance the religious freedom of individuals, on the one hand, with the freedom of majority and minority groups to “manage” their religious affairs, on the other (Keogh Reference Keogh2007, 154, 389). These provisions later traveled to India via transnational anti-colonial activists like Annie Besant (Reference Besant1926, 212), who lifted key provisions directly from the Constitution of the Irish Free State (1922) in her proto-constitutional “Commonwealth of India Bill.” This bill directly informed Article 25 of India’s Constitution (1950) concerning individual religious freedom (Austin Reference Austin1966, 54–55). India’s Constituent Assembly (1946–49) also adopted a provision concerning the rights of religious groups (Article 26) from Ireland’s second Constitution (1937).

The interim constitution unveiled by Pakistan’s first Constituent Assembly (1947–54) clearly revealed the influence of both Ireland and India.Footnote ¹³ In Article 8 of the first Irish Constitution (1922) and Article 44-2(1) of the second Irish Constitution (1937), it is noted that, with reference to individuals, “the free profession and practice of religion are, subject to public order, … guaranteed to every citizen.” India began with the phrase “subject to public order” before noting that all persons enjoy “the right freely to profess, practice, and propagate religion” (Article 25). And, finally, Pakistan followed suit, explaining that “the right to profess, practice, and propagate religion are guaranteed subject to public order” (Article 10).

The second Irish Constitution of 1937 also stipulated that “[e]very religious denomination shall have the right to manage its own affairs” and to “own, acquire, and administer property, movable and immovable, … for religious and charitable purposes’ (Article 44-2(5)). India’s Constitution guaranteed that “every religious denomination … shall have the right … to manage its own affairs in matters of religion … and to administer [its] property in accordance with law” (Article 26). Pakistan again followed suit, noting that “every religious denomination … shall enjoy freedom in the management of its religious affairs including … the acquisition of movable and immovable property for that purpose” (Article 11).

Hindu members within Pakistan’s Constituent Assembly tried to delay a debate regarding fundamental rights (including religious freedom) until the Assembly’s Committee on Minority Rights had issued its final report. But their colleague Abdulla al-Mahmood deflected their concerns, stressing that Pakistan’s efforts to protect the rights of minority groups were already explicitly drawn from Hindu-majority India. “Clause 10 of our Fundamental Rights,” he noted, provided “the same thing but on a little wider scale than what has been provided in Clause 35 [sic: 25]” of the Indian Constitution.Footnote ¹⁴

Inclusive efforts to accommodate Pakistan’s Hindus via borrowing from India, however, were not unlimited, as Pakistani Hindus failed to persuade their Constituent Assembly colleagues to reject two claims pushing toward a more exclusive, majoritarian, Muslim constitutional identity. The first claim emerged from right-wing religious activists affiliated with a group known as the Majlis-e-Ahrar-e-Islam as well as lay Muslim ideologues associated with Pakistan’s Jama’at-e-Islami (Party of Islam). These activists insisted that, because Pakistan sought to construct an Islamic democratic constitution, a “Muslim” head of state was essential. The second claim emerged from rival politicians within the ruling Muslim League, insisting that non-Muslims should be relegated to a separate “non-Muslim” electorate to prevent their electoral intervention as kingmakers in competitive races between Muslims.Footnote ¹⁵ On their own, these provisions sought to demarcate a special place for “Muslims” in Pakistan’s Muslim-majority “Islamic democracy,” but, over time, they also set the stage for dramatic changes in the meaning of Pakistan’s imported Irish-cum-Indian constitutional provisions.

The key issue did not involve the religious freedom of non-Muslim Hindus. Rather, it involved a religious minority known as the Ahmadiyya whose pattern of religious self-identification as “Muslim” was treated as a source of debate. Even before Pakistan’s interim constitution was unveiled in 1954, some of the activists demanding special Muslim rights and privileges (for example, a Muslim presidency and a separate Muslim electorate) insisted that Pakistan’s Ahmadiyya should be relegated to Pakistan’s “non-Muslim” electorate owing to their departure from the orthodox view that Mohammad was the final prophet of God. (Some Ahmadiyya recognize a late nineteenth-century religious reformer named Mirza Ghulam Ahmad as a “prophet.”) These right-wing religious activists were excluded from the Constituent Assembly because, in 1946, they did not participate in the provincial elections that underpinned the Constituent Assembly’s late colonial formation.

Many of these activists, however, resorted to riots and violent pogroms to advance their exclusionary views. A particularly violent stretch in 1952–53 culminated in Pakistan’s first martial law—a spate of military intervention that inspired later efforts within Pakistan’s Constituent Assembly to shore up the power of parliament (Binder Reference Binder1961, 259–96; Report of the Court of Inquiry 1954; G. W. Choudhry Reference Choudhry1963, 48).Footnote ¹⁶ This focus on parliament, however, was offset in 1958 by a coup and more than ten years of military dictatorship. Yet, shortly after the military announced national elections in 1970, Pakistan was plunged into a civil war. Culminating in the separation of East Pakistan and the formation of Bangladesh (1971), this civil war led to the promulgation of Pakistan’s current Constitution in 1973—a Constitution in which the restoration of parliamentary power figured prominently alongside a special focus on the ostensibly unifying features of Pakistani “Muslim” nationalism.

Pakistan’s Constitution (1973) retained all of the religious freedom articles initially borrowed from Ireland (via India). At the same time, however, these articles were paired with two constitutional adjustments clarifying the terms of Pakistani Muslim nationalism. The first emerged in the Constitution of 1973 itself, requiring each new president and prime minister to not only swear an oath that he or she was a Muslim but also to clarify that there could be “no Prophet after [Mohammad].” The second emerged one year later, in Pakistan’s second constitutional amendment, wherein Pakistan’s Ahmadiyya were constitutionally redefined as “non-Muslims.”Footnote ¹⁷ The logic underpinning these changes was simple: if “Muslims” were endowed with special rights and privileges, the state must be able to distinguish its Muslim citizens from its non-Muslim citizens (Qasmi Reference Qasmi2014, 195). To do this, parliament merely sought to clarify, via oaths and constitutional definitions, which groups should be legally defined as “non-Muslim.”

Even apart from this basic legal logic, however, the specific political calculations underpinning these constitutional adjustments were important. Immediately after the approval of Pakistan’s new Constitution in 1973, the same activists who had challenged the rights of the Ahmadiyya during the 1950 s reasserted themselves in another round of violent skirmishes (Qasmi Reference Qasmi2014, 175–77). Pakistan’s new Prime Minister Zulfiqar Ali Bhutto (1973–77) did not declare martial law. After ten years of military dictatorship, he turned to parliament and its power of constitutional amendment instead.Footnote ¹⁸ Within this turn to parliament, however, the same activists stressed what they saw as the concerns of Pakistan’s religious majority, claiming that, by identifying as “Muslims,” Pakistan’s Ahmadiyya actively diluted Muslim access to Muslims’ special constitutional rights. Bhutto’s parliamentary majority did not reject this argument; facing rival coalitions tied to right-wing religious parties in two out of Pakistan’s four provinces (Balochistan and the Northwest Frontier Province), they simply co-opted it. In effect, Bhutto agreed that Pakistan’s parliament was entitled to curtail, by way of constitutional amendments, what his attorney general described as an Ahmadi “threat” of religious “false belonging.”Footnote ¹⁹

Public officials facing yet another round of protests by the same right-wing activists later intervened to restrict the fundamental rights of Pakistan’s Ahmadiyya even further—this time under the military dictatorship of General Zia-ul-Haq (1977–88), who amended Pakistan’s Penal Code (section 298) to prohibit the country’s Ahmadiyya from using ostensibly “Muslim” words like masjid (mosque) or azaan (call to prayer), describing any Ahmadi use of such words as a dangerously provocative form of “encroachment” on the special religious “property” of Muslims as a group (Saeed Reference Saeed2011, 88). When this reading was taken up in a landmark Supreme Court case known as Zaheeruddin v the State (1993), the court referenced Pakistan’s borrowed constitutional provisions. In particular, summarizing the decision, Ahmad Mahmood Khan (Reference Khan2003, 228) notes that, according to the court, “Ahmadi religious practice[s], however peaceful, angered and offended the Sunni majority,” so to preserve public order (Article 20), “Pakistan would … need to control [them].”

As such, the state drew on Irish and Indian provisions protecting a right to religious freedom “subject to public order,” suggesting that formal legal restrictions on the fundamental rights of the Ahmadiyya were necessary to prevent any possibility of religious provocation, outrage, or “disorder” within the country’s Muslim majority. The specific language of Irish and Indian religious freedom clauses was, in effect, emptied of its prior legal content and used to support an increasingly exclusionary reading in which Pakistan’s religious-cum-political majority actively reconfigured the meaning of Pakistan’s borrowed constitutional provisions.

CONSTITUTIONAL INSTITUTIONS AS EMPTY SIGNIFIERS: FRENCH CONSTITUTIONAL STRUCTURES IN THE SERVICE OF ISLAMIST MONARCHY IN MOROCCO

Beyond this focus on the transfer of explicit constitutional clauses, a second type of borrowing in the realm of religion-state relations involves the unacknowledged adoption—and, then, adaptation—of foreign constitutional ideas regarding state institutions. Morocco’s transfer of constitutional structures from the Constitution of France (1958) offers one example. Moroccan political actors drew on institutional elements from a democratic and secular European republic to legitimate, even sacralize, their country’s Muslim monarchy.

Perhaps because of its nonelected monarchical head of state, or its habit of frequent top-down constitutional drafting, Morocco has rarely been the focus of comparative constitutional analysis.Footnote ²⁰ Yet, particularly when juxtaposed with other Arab constitutions, Morocco’s postcolonial documents (each approved by a popular referendum) show several noteworthy features—most obviously, a pattern of imbuing the country’s ruling monarchy with “Islamic” forms of legitimacy (the king is referred to by the caliphal title amir el-mu’minin or “commander of the faithful”)—even as each constitution also references minority cultural influences and a pluralized understanding of rights.

There are several accounts of this explicit effort to infuse Morocco’s leader with quasi-sacred status alongside a push to highlight the country’s religio-cultural pluralism (Waterbury Reference Waterbury1970; Geertz Reference Geertz1971; Tozy Reference Tozy1999; Mednicoff Reference Mednicoff, Banerjee, Backerra and Sarti2017). Yet, there is also a rarely noticed facet of Morocco’s constitutional origins that may help to connect these ostensibly divergent elements. Morocco’s constitutions contain similarities of both language and structure with the 1958 Constitution that established the French Fifth Republic. This is no coincidence given France’s colonial and postcolonial influence on Morocco and Hassan II’s close ties to France, French political culture, and French constitutional experts.

Within postcolonial Morocco, the evidence of actual constitutional borrowing from the country’s former French colonizer is strong, if circumstantial. A new king seeking to restore Moroccan political autonomy and establish his own authority after a vocal independence struggle would hardly be expected to admit that he had relied on French texts or advisors to compile his country’s first postcolonial constitution. Indeed, King Hassan II claimed that the Constitution came from “his own hand.”Footnote ²¹ Yet the Moroccan Constitution (1962) included provisions very similar to those from the 1958 Constitution in France. Specifically, the Moroccan document seemed to draw on French provisions for states of exception,Footnote ²² the ruler’s ability to call a popular referendum (going over the heads of his elected parliament),Footnote ²³ and, in general, a comparatively high level of authority associated with the head of state.

Prominent French political scientists who advised the Moroccan throne, such as Maurice Duverger, devoted time to arguing for the new Moroccan Constitution (Beling Reference Beling1964, 172).Footnote ²⁴ Another French jurist with strong ties to Morocco and Hassan II noted significant textual overlaps and similar chronological histories, both in the constitutions’ initial promulgation and in their subsequent revisions (Vedel Reference Vedel1993, 363–65). A contemporary database comparing major structural features of the world’s constitutions finds significant similarities between France and Morocco in their high level of executive authority and their (correspondingly) low levels of parliamentary or legislative autonomy.Footnote ²⁵

De Gaulle’s brand of executive-centered politics was both a model familiar to the strongly Francophile Hassan II and an obvious (Western) template for a government centered on a strong ruler. If codifying monarchical advantage was Hassan’s purpose, then finding a familiar model that carried a strong flavor of global political legitimacy while, at the same time, translating Western institutional forms into Morocco’s institutional context was a sensible strategy. In fact, decades after the promulgation of Morocco’s 1962 Constitution, Hassan II actually confessed that his Constitution was inspired by France’s Fifth Republic, which, he argued, was a corrective to centuries of French political confusion as to whether their executive or their legislature was supreme (Laurent Reference Laurent1993, 73–74). This also helps to explain the late king’s serial constitutionalism. While maintaining a strong executive, the Moroccan ruling cadre found that it could use constitutional redrafting to respond to social changes and demands for rights and pluralism in a manner that periodically fine-tuned the powers of a monarchical regime.

Early Moroccan constitutionalism was thus a case of direct, if deliberately unacknowledged, institutional borrowing from the 1958 French Fifth Republic’s basic law.Footnote ²⁶ Yet the king and his constitutional advisors claimed inspiration, and borrowed, from a highly secularized French society—one in which the strong chief executive was elected through a contested democratic process. When Hassan II confessed that his Constitution was inspired by the French Fifth Republic, he went on to note that, although he had authoritarian tendencies, he had grounded his constitutional politics in the popular will of the people. In short, he used specific institutional models from a democratic state’s Constitution to secure popular legitimacy for the underpinnings of an authoritarian monarchy.

In this context, the Moroccan system that borrowed from France’s Constitution did so in the explicit service of, and alongside detailed provisions for, a nonelected monarchy that established a state religion (Islam) from which it derived its popular legitimacy. Indeed, Morocco’s set of constitutions under Hassan II and, to only a slightly lesser extent, those introduced by his successor, Muhammed VI, clearly established the king as the hereditary embodiment and symbolic head of Islam. Accounts of Morocco’s constitutional identity and meaning stress these functions and procedures around the king’s codified status as the amir el-mu’minin (Benjelloun Reference Benjelloun2002). Indeed, while Hassan II was careful to deploy both secular and sacred legitimizing discourses, he stressed that his link to the population was one of “allegiance,” the Arabic term for which bey’a refers to a religio-political ceremony in which the country’s leaders pledge fealty to him as amir (Laurent Reference Laurent1993, 76).

Hassan II took language and specific institutional arrangements from France’s Constitution out of that constitution’s democratic and secular context in a deliberate effort to straddle the line between a European system and Morocco’s. Again, this was part of a broader political strategy that sought to maintain close ties to France while, at the same time, establishing Hassan, and the monarchy, as essential to Morocco because they fused a new nation-state with traditional forms of authority rooted in a religio-political lineage of ‘Alawi rulers. Hassan II’s father, Muhammed V, had emerged as the country’s preeminent political force and a major nationalist symbol during Morocco’s independence struggle. When Muhammad V died unexpectedly in 1961, his relatively unknown thirty-two-year-old son Hassan II felt compelled to block the influence of competing nationalist parties like Istiqlal (Independence) and the more radically socialist and republican Union Nationale des Forces Populaires by reinforcing the historical and specifically religious salience of the Moroccan monarchy.

The king’s consultation with prominent French constitutional scholars, and his appropriation of phrases and frameworks from the French Constitution, thus served a political purpose: demarcating Morocco under the monarchy as comparable to France—a “modern” state—whilst simultaneously pushing aside the secular Moroccan political parties jockeying for postcolonial power.Footnote ²⁷ In effect, the king’s re-signification of French constitutional structures allowed him to reference the trappings of European political legitimacy without simultaneously conceding to democratic or republican forms of power.

Within the Moroccan document, this recontextualization and re-signification of constitutional elements from France offered modern constitutional cover for direct monarchical links to religious history, effectively buttressing the political legitimacy of the king with rural and traditional Moroccans (Leveau Reference Leveau1976). Efforts to imbue the monarchy with quasi-sacred status also helped to justify the absence of any provisions for judicial review in the 1962 Constitution (and subsequent constitutions under Hassan II), thus precluding any judicial reinterpretation of the king’s constitutional push to balance Western modernity with quasi-Islamic traditionalism.

Morocco’s constitutional system attaches the legalized status of a quasi-sacred monarchy to the political institutions of France’s Fifth Republic. This link has persisted through multiple Moroccan constitutions as well as a gradual expansion of nonmonarchical institutional powers. What French drafters likely viewed as a set of institutional provisions concerning executive power serving the needs of a relatively centralized democratic secular order was carried over by Moroccan ruling elites as an “empty signifier” that was politically “re-signified” by a king and his advisors to amplify the authority of a nonelected monarchy imbued with quasi-religious power.

CONSTITUTIONAL “MODELS” AS EMPTY SIGNIFIERS: THE TURKISH MODEL IN POST-ARAB SPRING EGYPT

A third pattern of empty-signifier-based borrowing involves invocations of foreign constitutional models—without any explicit transfer of text—to mark out key positions in a domestic constitutional debate. After the Arab Spring uprisings of 2011, Egyptians invoked Turkey’s constitutional model as a guide for their domestic political transition, but they did not adopt any Turkish constitutional text. When the Turkish model was invoked by Arab publics demanding constitutional and political reform during and after the Arab Spring, the reference was generally linked to notions of sociopolitical “inclusiveness”—that is, the inclusion of Islamist political parties. With the governing Turkish AKP standing as a marker of this Islamist inclusion, such references regularly surfaced in the Egyptian press between the fall of 2011 and the end of 2012.Footnote ²⁸ Specifically, the success of the AKP in Turkey’s national and local elections (2002–11) created the impression that the Turkish republic’s history of assertive secularism had given way to a pluralist vision of democratic inclusion (Kuru Reference Kuru2013). In this sense, Turkey stood for a constitutional solution to one of the many problems faced by post-2011 Arab states—the problem of allowing more explicit Islamic political influence while, at the same time, striving for a stable democratic system. Reviewing the appeal of the Turkish model, one Cairo University professor cited the AKP’s ability to secure “democratic” stability under an “Islamic-oriented” ruling party (Nafaa Reference Nafaa, Tocci, Taşpınar, Barkey, i Lecha and Nafaa2011).

The notion that Turkey could serve as a model might have suggested that reformers would borrow directly from Turkey’s Constitution (1982). In fact, invocations of the Turkish model were offered by key figures directly involved in Egypt’s constitutional drafting process, including Mohamed Morsi, the Muslim Brotherhood politician who became Egypt’s first democratically elected president.Footnote ²⁹ When Morsi visited Turkey for an AKP convention in 2012, he explicitly highlighted his admiration for the Turkish model even as the constitution-drafting process led by his own party was still unfolding in Egypt.Footnote ³⁰ The specific meaning of the Turkish model, however, was slowly decoupled from any specific Turkish constitutional provisions; in fact, at one point, the idea that textual elements of the Turkish model could be transplanted or replicated was flatly rejected.Footnote ³¹ None of the key players in Egypt’s constitutional debate—neither drafters (like Morsi) nor pundits nor the general public—sought to adopt specific provisions of the Turkish Constitution. When the Turkish prime minister, Recep Tayyip Erdogan, traveled to Cairo in September 2011 and suggested that specific aspects of the Turkish model of secularism could be adopted in Egypt, the idea was rejected by members of the Muslim Brotherhood and other Egyptian officials who, in the past, had themselves invoked the merits of “the Turkish model.”Footnote ³²

Following the earliest invocations of the Turkish model in broader debates across the Arab world, Egyptian political figures expressed ambivalence about which aspects of the Turkish example might be worth emulating, particularly given the fact that Turkey’s actual Constitution had been written under military tutelage in 1982 and amended over a dozen times. Typically, the success of the Turkish model was judged in light of perceptions regarding its political outcomes and, specifically, its outcomes in the realm of religion-state relations: on the one hand, Islamists who admired the ability of Turkey’s secular state to accommodate Islamist political parties; on the other, a more explicitly authoritarian appreciation for the ways in which Turkey’s military guarded the secular character of the state. Recalling Langer’s (Reference Langer2004, 33) description of legal translation owing to a struggle “between different actors and groups within the target legal system,” which of these elements was emphasized depended on the domestic interests of the Egyptian political actors who sought to embrace the “model” (Ottaway and Brown Reference Ottaway and Brown2012).

While grassroots popular perceptions focused on the electoral successes of a moderately Islamist political party, non-Islamist elites tended to focus on the secular facets of Turkey’s political trajectory. Specifically, Egyptian military and business elites were less enamored of the post-2002 AKP model of Islamist inclusion than a pre-2002 model of Turkish politics characterized by a strong and centralized executive.Footnote ³³ Indeed, even as President Morsi was praising the model as an important reference point for Islamist inclusion, prominent Arab columnists described the Egyptian military as adopting the Turkish model “to check Morsi.”Footnote ³⁴ The Turkish model was, as such, an “empty signifier” with many possible meanings—indeed, a “surplus” of meaning. The specific content of its meaning, following Laclau, emerged as a function of the political preferences of those who intervened to invoke it.

Of the various ways in which Turkey was invoked, however, the one that ultimately had the most influence in Egypt was the least inclusionary interpretation of what Turkey’s system represented. Specifically, an understanding of the Turkish model as according a preeminent role to the military in civilian governance was embraced by the leadership of the Supreme Council of the Armed Forces during Egypt’s post-uprising transition.Footnote ³⁵ The draft Constitution adopted under the leadership of President Morsi in December 2012 did not reflect this version of the model, offering no formal tutelary role to the military in civilian governance. But, within six months, President Morsi was ousted in a military coup, and the Constitution drafted during Egypt’s brief democratic interlude was abrogated, then replaced (Brown and Dunne Reference Brown and Dunne2013). The new ruler, General Abdel Fattah el-Sisi, viewing Turkey as a major sponsor of the ousted Muslim Brotherhood, privileged a different version of the same Turkish “model” that radically curtailed Islamist political participation, giving the military a key role in defining the nature of Egyptian religion-state relations.Footnote ³⁶

As an “empty signifier,” the Turkish model came to signal the legitimacy of (contradictory) Egyptian positions for the purpose of domestic political debate. In fact, invocations of the Turkish model served as a type of shorthand for the political positions taken by an array of drafters making divergent appeals to mobilize their domestic audiences. In the end, the course of constitutional drafting in Egypt came to reflect the most exclusionary version of the Turkish model: a Constitution drafted in the wake of a military coup to regulate religion-state relations through top-down imposition and, crucially, the exclusion of Islamist parties.

INHERITING EMPTY SIGNIFIERS: THE OTTOMAN MILLETSYSTEM FROM INTERRELIGIOUS TO INTRARELIGIOUS REGULATION IN ISRAEL

Our fourth case study concerns the Ottoman millet system of legal pluralism in the area of religious family law as it moved from the Ottoman Empire through the British Mandate to the quasi-constitutional architecture of Israel. This example differs from the preceding modes of constitutional movement in two key respects. First, it represents a case of constitutional inheritance from a prior legal system (common in postcolonial state building) rather than any proactive search for constitutional “imports” from other countries. Second, Israel lacks a written constitution, so this example illuminates some of the ways in which foreign models of quasi-constitutional religion-state relations might be assimilated via ordinary legislation. Legislation might be considered more flexible than formal constitutional provisions, but the basic regulatory structure of legal pluralism in Israel, inherited from the millet system and adapted during the first few years of the state, remains largely unamended even after several decades. As such, we describe Israel’s regulation of marriage and divorce as “quasi-constitutional.”Footnote ³⁷ Since the religious monopoly on marriage and divorce that emerged from Israel’s inheritance of the Ottoman millet system has come to represent a core principle of religion-state relations in Israel, we use it to illuminate a pattern of constitutional assimilation and, then, constitutional re-signification.

The term “millet” emerged in the early nineteenth century to designate recognized non-Muslim communities living within the Ottoman Empire (Quataert Reference Quataert2005, 175–76). While contemporary legal and political scholars often regard the millet system as representing a highly pluralistic system of religious law, more recent historical research has emphasized the term’s versatility, which, already during Ottoman times, was influenced by centralizing reforms with a (Muslim) majoritarian flavor as well as external political pressures from Europe. Before the nineteenth century, different religious communities across the empire operated under distinct legal arrangements within the Ottoman state, with separate courts, judges, and legal principles. Following a set of mid-nineteenth-century Ottoman reforms known as the Tanzimat reforms, however, a more centralized structure gradually replaced the empire’s differentiated legal system, culminating in 1917 with the publication of an Ottoman Family Code that was partly meant to unify family regulations on a territorial (rather than a religious or communal) basis (Reference AgmonAgmon 2016–17).Footnote ³⁸ This new legislation, however, was never fully implemented due to the outbreak of the First World War and, thereafter, the collapse of the Ottoman Empire.

When the British occupied Palestine, they generally preserved the system of differentiated religious tribunals, with some exceptions.Footnote ³⁹ Whilst claiming mere continuation of the millet system, for instance, the British reversed emerging Ottoman reforms that strove for a more unified and territorially based family code. Whereas under Ottoman rule, Jews were free to appeal to Muslim courts, the British established an exclusive rabbinical court defined as the sole religious authority for all of the Jews in Palestine. Moreover, the British recognized the 1917 Ottoman Family Code, which was developed as a territorial state code, as a “Muslim Family Law” to be applied to Muslims alone, erasing the inclusive articles referring to Christians and Jews (Reference AgmonAgmon 2016–17, 17–18).

With independence in May 1948, Israel’s Jewish leadership faced questions about the extent to which Ottoman or British family law arrangements should be preserved or reformed. From the early stages of the debate, however, the dilemma was structured around intra-Jewish considerations, initiating a further stage of legal “re-signification” surrounding the state’s inherited millet model. Israel’s new political leadership never seriously doubted broad millet principles of legal pluralism, which allowed complete autonomy in the area of family law for all recognized religious communities. Still, intense debates took up the question of whether the Jewish rabbinical court established by the British mandate should retain its monopoly over family law for all Jewish citizens.

Ultimately, Israel’s “borrowing” of the Ottoman millet model and its later re-signification occurred in two stages. The first stage involved Israel’s first legislation, passed by the Provisional State Council in 1948, three days after independence, preserving the general contours of British Mandate law.Footnote ⁴⁰ The drafting of the bill had begun even before independence, and the debate concerned the extent to which Jewish law should be incorporated into state law (Radzyner Reference Radzyner2010). During the first meeting of the Provisional Council (one day after independence), a representative of the Communist party criticized the bill for its failure to separate the state from religion. But, within the Provisional Council, this claim represented a small minority.Footnote ⁴¹ The final version of the bill neither increased nor decreased the role of Jewish law compared with previous Mandate arrangements (Radzyner Reference Radzyner2010).

Yet the application of a millet-style system of legal pluralism in the context of a Jewish majority state created new bureaucratic problems. Under both Ottoman and British rule, rabbinical courts could apply Jewish law only to those who were officially registered as members of the Jewish community. Because the state of Israel ceased to maintain such official registration, the question of official membership in the Jewish religion became extremely controversial. New legislation was required, and, in 1953, five years after independence, the government proposed the Jurisdiction of Rabbinical Tribunals (Marriage and Divorce) Bill (1953), stirring intense political controversy.

During Knesset debates on the proposed bill, Jewish religious parties claimed that a unified system of marriage and divorce amongst Israeli Jews, dictated by Halachic rule, would be essential to preserve a unified identity among the Jewish people. Otherwise, they argued, religious Jews would not be able to marry the descendants of those who had married without a ceremony according to Halachic provisions.Footnote ⁴² In contrast, many in the majority secular/liberal camp argued against the bill’s failure to allow for civil or interfaith marriage.Footnote ⁴³ The Rabbinical Authority, it was argued, did not recognize gender equality and followed “medieval traditions.”Footnote ⁴⁴ Others opposed any infringement on “freedom of conscience,” arguing that the law would force nonreligious citizens to act against their secular worldview.”Footnote ⁴⁵ Eventually, however, and despite this criticism, a majority of the Knesset led by Prime Minister David Ben Gurion and the Mapai party voted for the Rabbinical Courts Act (1953). This 1953 law formalized a monopoly for the Orthodox rabbinate on personal status laws regarding Jewish marriage and divorce even as it limited the scope of rabbinical authority with regard to other aspects of family law.Footnote ⁴⁶

Whereas the millet system under Ottoman rule encouraged formal state recognition of diverse religious traditions and allowed various forms of religious self-governance, this inherited order was transmogrified under Israeli law into a centralizing mechanism of religious imposition and national homogenization for the state’s majority Jewish population. Rather than protecting religious minorities from potential abuse by the majority, the plurilegal system of religious autonomies in the realm of marriage and divorce was refashioned—in part—to limit a right of marriage for certain parts of the population (including interfaith couples) and to regulate intrareligious relations within the dominant (Jewish) community (Amir Reference Amir2016).

Why would the secular/socialist leadership of the Mapai party support such religious legislation, which conflicted with their personal commitment to religious freedom and entrenched an exclusivist re-signification of the millet system?Footnote ⁴⁷ Archival materials reveal that the 1953 law was part of a coalitional agreement within the Knesset between the Mapai majority and various religious parties.Footnote ⁴⁸ The two sides agreed to pass the Rabbinical Courts Act hand in hand with another bill establishing a mandatory national service for religious women (in lieu of military service). Another consideration involved efforts to avoid conflicts with the more religious parts of world Jewry that might be expected to immigrate to Israel.Footnote ⁴⁹ The records of the Mapai party and government meetings also reveal that many party members viewed the 1953 Rabbinical Courts Act as a temporary compromise; they did not intend to entrench religious personal status regulations as a quasi-constitutional regulation for decades to come. Leading members of the Mapai party even explicitly raised the option of future reforms, including a civil marriage bill.Footnote ⁵⁰

Nevertheless, the bill played a central role in national debates regarding Israel’s “constitutional identity” (Jacobsohn Reference Jacobsohn2010). In the eyes of many, the rabbinical monopoly on marriage and divorce was designed to promote a homogenization of Jewish identity, on the one hand, while differentiating non-Jewish identities, on the other (Shafir and Peled Reference Shafir and Peled2002; Amir Reference Amir2014; Triger Reference Triger2014). As such, the meaning of Israel’s inherited millet system was actively translated and narrowed—from a policy protecting diversity between religious groups to a centralized mechanism of religious homogenization targeting the members of Israel’s Jewish majority.

CONCLUSION: POLITICS DRIVES TEXTUAL MEANING

These four cases suggest the complex ways in which foreign constitutional provisions, institutions, and concepts are borrowed, invoked, or self-consciously inherited during constitution-drafting processes and then actively engaged during politically charged moments of constitutional contestation, reinterpretation, and translation. Our main argument is that, well beyond the borrowing of constitutional “text,” the politics surrounding borrowed constitutional elements often shapes the meaning of those ideas and, therein, the significance of borrowing itself.

Constitutional borrowing, in this sense, is a political process that leaves room for local actors to introduce normative slippage, institutional adaptation, and new meaning. We do not see these new and narrower meanings as cases of “unsuccessful” borrowing.Footnote ⁵¹ Instead, we see them as contested constitutional and political choices. As such, our focus on empty signifiers moves beyond the forms of organizational decision making that feature in recent accounts of domestic legal translation (Edelman Reference Edelman1992; Grattet and Jenness Reference Grattet and Jenness2005) to examine explicitly transnational contexts of constitutional borrowing underpinned by explicitly political decision making (Laclau Reference Laclau1996).

Of course, not all cases of borrowing rely on empty signifiers. Where the meaning of a borrowed text or model is clearly bound by its prior usage (as in South African constitutional references to provisions of the International Covenant on Economic and Social Rights [1966]), it makes little sense to move beyond the notion of stable “privileged” signifiers to highlight the intervening influence of politically inflected “empty” signifiers (Reference YoungYoung 2008–9). As noted above, there is also a distinction to be drawn between the contested meanings we associate with empty signifiers and instances in which the meaning of a given text or norm is left deliberately “floating” or vague, as in persistent Indian debates regarding the meaning of a constitutional principle like “secularism” (Sen Reference Sen2010) or the refusal, by Léopold Sengohr and later politicians and judges in Senegal, to define a constitutional term like “laïc” (Diagne Reference Diagne, Bâli and Lerner2017).

In all four of the cases we examine, domestic political dynamics explain the shifting meaning of constitutional empty signifiers. In Pakistan, domestic legislators and judges reshaped the meaning of both Irish and Indian constitutional provisions regarding religious freedom in ways that became more closely tied to forms of parliamentary majoritarianism and a cynical effort to bind Pakistan’s “Muslim” community together via that community’s juxtaposition vis-à-vis the country’s heterodox Ahmadi minority. In Morocco, domestic political elites underpinned the re-signification of French constitutional provisions in ways closely tied to the codification of quasi-religious monarchical advantage against domestic political challengers. In Egypt, the political underpinnings of the so-called “Turkish model” harbored a surplus of meaning that pointed in two directions at once—toward bottom-up forms of Islamist political inclusion as well as top-down forms of secular authoritarianism—with the latter coming to prevail via the military coup of General Abdel Fattah el-Sisi after 2013. In Israel, core political dynamics underpinned a shift in the meaning of a quasi-constitutional personal law system inherited from the Ottoman Empire via the British Mandate. This system was reframed by Jewish political leaders to advance their goals regarding both interreligious and intrareligious relations in the realm of marriage and divorce. With reference to borrowed constitutional provisions in the realm of religion-state relations, further research will undoubtedly help to isolate the mechanisms that drive politically inflected patterns of “re-signification,” not only in the direction of religious exclusion but also, at least potentially, in the direction of greater inclusion.

Centering our analysis on the active importers of constitutional concepts, rather than exporters,Footnote ⁵² our study challenges trends highlighting the degree to which traveling constitutional texts might point to global constitutional isomorphism or convergence. Constitutional borrowing, we argue, is closely tied to the idiosyncratic political circumstances surrounding the deliberate choices of importers—to provide political signals, borrow symbolic capital, or invoke patterns of domestic constitutional legitimacy—without any associated commitment to reproducing the “original” meaning of a particular constitutional text, institution, or idea. Grasping the subtleties of local political contestation, we argue, is essential for those with an interest in illuminating the conceptual modulations that underpin the meaning of borrowed constitutional texts.