From Pluralism to Autocracy: De-Constitutionalization of the Debate over Religion and State in Kaïs Saied’s Tunisia

Abstract

Existing scholarship has analyzed Tunisia’s 2022 Constitution primarily as an instrument of hyper-presidentialism, democratic breakdown, and authoritarian restoration. Bringing together the literature on authoritarian constitutionalism and religion–state relations, this article suggests that constitutional identity clauses constitute an additional, largely overlooked dimension of Tunisia’s neo-authoritarian transformation. Through a comparative analysis of the relevant religion-related provisions in the 2014 and 2022 Constitutions, it shows how the 2022 Constitution withdrew explicit constitutional recognition from competing Islamist and non-Islamist claims concerning Islam and civil statehood that had been accommodated in the 2014 Constitution, while entrusting the state alone with realizing the purposes of Islam. These changes in Kaïs Saied’s Constitution, the article argues, de-constitutionalized the debate over religion and state that the earlier Constitution had represented and mediated through a pluralist compromise framework, and concentrated authority over Islam’s public meaning within a state-centered authoritarian order.

1. Introduction

In the years following the 2010–11 popular uprising, Tunisia came to be regarded as the exceptional success story of the “Arab Spring.” As democratic transitions elsewhere in the region faltered or collapsed, it increasingly appeared to be the only Arab country in which the fall of an authoritarian ruler had produced a genuine democratic transition (Murphy 2013, p. 231; Pickard 2014, p. 259; Stepan and Linz 2013, pp. 18, 22–24). The decisive rupture came on 25 July 2021, the sixty-fourth anniversary of the proclamation of the Tunisian Republic, when President Kaïs Saied, who had himself come to power through democratic elections in 2019, invoked the exceptional-measures clause of the 2014 Constitution to dismiss the prime minister, suspend parliament, lift parliamentary immunity, and concentrate executive authority in his own hands (Ridge 2022; Souilmi 2023). What Saied initially presented as a temporary response to political paralysis, economic crisis, and popular disillusionment soon became the starting point for a far-reaching transformation of the constitutional order (Tamburini 2023). In September 2021, Saied issued Decree-Law 117, suspending key parts of the 2014 Constitution and allowing him to govern by decree. In March 2022, he dissolved parliament; and on 25 July 2022, a new constitution was approved by referendum with 94.6 percent support, albeit amid low turnout and an opposition boycott. The 2022 Constitution marked the constitutional consolidation of Tunisia’s post-uprising authoritarian turn, replacing the pluralist and semi-presidential framework of 2014 with a “hyper-presidential” order (Khan 2022, p. 16).

Existing scholarship and legal commentary have generally read the 2022 Constitution as the legal consolidation of democratic backsliding under Saied, authoritarian restoration, and the end of Tunisian exceptionalism (Limam 2023, pp. 8–10, 20–22; Gabsi and Sarihan 2025, pp. 20–21). Scholars have focused primarily on the institutional mechanisms through which the 2022 Constitution translated Saied’s “self-coup” into “a unilateral ‘refounding’ of Tunisia’s political system” around the presidency, including the subordination of the government, the weakening of parliamentary oversight, the redesign of independent bodies under presidential authority, the reduction in judicial and constitutional checks, the facilitation of exceptional powers, the rollback of decentralization, and the elimination of mechanisms of presidential accountability (Khaddar 2023, pp. 1327, 1332). So pronounced was this concentration of presidential authority that Desrues and Gobe conceptualize Saied’s rule as a “drift toward Caesarism” (Desrues and Gobe 2026, pp. 204–5, 212–15).

Less attention, however, has been paid to the way this authoritarian logic also operated through the 2022 Constitution’s treatment of religion-related identity clauses. To address this gap, the article brings together two bodies of scholarship that are often treated separately: the literature on authoritarian constitutionalism, which examines constitutions as instruments for organizing and concentrating power (Landau 2013; Ginsburg and Simpser 2014; Tushnet 2015), and the literature on religion-related constitutional provisions, which analyzes how constitutions establish, privilege, separate, or regulate religion within the state (Brown 2009; Temperman 2010; Fox 2015, pp. 201–30; Bâli and Lerner 2017; Bernard-Maugiron 2021). Reading these studies together makes it possible to see the treatment of identity clauses not as merely symbolic provisions, but as constitutional devices that may also redistribute authority and thus participate in a broader authoritarian logic. The article therefore argues that the 2022 Constitution de-constitutionalized the debate over religion and state at the heart of Tunisia’s 2014 constitutional compromise by dismantling the framework through which competing claims about Islam and civil statehood had been incorporated and negotiated, and by reinforcing the state’s exclusive authority over Islam’s public meaning.

The article proceeds by first elaborating the theoretical framework for understanding religion-related identity clauses not only as symbolic statements about the character of the state, but also as constitutional devices through which authority over contested meanings may be distributed or concentrated. The analysis then turns to a comparative reading of the 2022 Constitution’s religion-related identity clauses against the 2014 constitutional settlement: the removal of Article 1, the inherited state-religion clause that had identified Islam as the religion of Tunisia; the abolition of Article 2, the civil-state clause that had framed Islam’s constitutional presence within a civic and non-theocratic order; and Article 5, which fills the space left by the deletion of Articles 1 and 2 with a state-centered mandate to realize the “purposes of Islam.” The discussion of Article 5 also considers related provisions concerning the mufti, the requirement that the president be Muslim, and Islamic education. These sections show how the 2022 Constitution removed provisions that had borne the distinct imprint of the Islamist and non-Islamist forces that shaped the 2014 constitutional settlement, and shifted constitutional authority over Islam decisively toward the state.

2. Authoritarian Constitutionalism, Religion-Related Identity Clauses, and the Authority to Define Islam

Constitutions are no longer understood exclusively as liberal instruments for limiting power and entrenching rights. A growing literature on authoritarian constitutionalism has shown that constitutional texts may operate in authoritarian and autocratizing contexts in ways that cannot be reduced either to liberal constraint or to mere façade. Tushnet, for example, urges scholars to “pluralize the idea of constitutionalism” by considering “forms of constitutionalism other than liberal constitutionalism” (Tushnet 2015, pp. 394–96). Ginsburg and Simpser argue that constitutions in authoritarian settings may organize rule, coordinate actors, and “structure authoritative discourse” (Ginsburg and Simpser 2014, pp. 1–2, 14). Landau’s concept of “abusive constitutionalism” identifies a related mechanism: the use of constitutional amendment or replacement to erode the democratic order while preserving the formal language of constitutional legality (Landau 2013, pp. 189, 195). These approaches shift attention from whether constitutions constrain power to how they organize authority and structure the terms of constitutional discourse. This perspective is particularly useful for analyzing what Lührmann and Lindberg call the “third wave of autocratization,” a contemporary pattern of gradual democratic erosion carried out through formally legal or constitutional mechanisms (Lührmann and Lindberg 2019, pp. 1095–96, 1103–4).

Scholars have identified several constitutional processes through which authoritarian power is constructed and consolidated. Scheppele’s concept of “autocratic legalism” focuses on elected leaders who combine electoral mandates with constitutional and legal change in pursuit of an illiberal project, using constitutional change as their “preferred vehicle for achieving the unified domination of all of the institutions of state” (Scheppele 2018, p. 562). Huq and Ginsburg’s notion of “constitutional retrogression” highlights the cumulative character of the process: changes that may be innocuous or even defensible in isolation can, through their cumulative, interactive effect, produce a qualitative transformation of the constitutional order (Huq and Ginsburg 2018, p. 97). The resulting order may entrench a new distribution of power and weaken alternative sources of political and economic power (Albertus and Menaldo 2013, pp. 54–58). Frankenberg adds a distinct emphasis on mediation, identifying the “neutralization of intermediary institutions and organizations” as a characteristic operation of authoritarian constitutionalism (Frankenberg 2020, p. 243). Yet this literature has paid less attention to the ways in which identity clauses themselves may serve as instruments of authoritarian consolidation.

The authoritarian implications of constitutional change may also be traced in provisions that appear, at first sight, to concern identity rather than institutional power. Constitutional provisions on religion are commonly examined as expressions of the polity’s foundational identity and as markers of the national or religious character of the state. Lerner describes the constitution as “the state’s charter of identity” and identifies religion–state relations among its foundational concerns (Lerner 2011, p. 4), while Hirschl emphasizes religion’s role in defining the boundaries of collective identity and the constitutional metanarrative of the polity (Hirschl 2010, pp. 3–5). In societies divided over the place of religion in public life, such provisions may also manage disagreement through ambiguity, ambivalence, or the deferral of contested decisions to future political institutions (Lerner 2010, pp. 70, 84–85; Lerner 2013, pp. 610–11, 617–19; Dixon and Ginsburg 2011, pp. 636–38). These approaches illuminate the identity-forming and compromise-managing functions of religion-related provisions but leave open a question central to authoritarian constitutionalism: how such provisions distribute authority over religion.

Religion-related constitutional provisions may therefore also perform an authority-allocating function. As Lombardi demonstrates with regard to Sharia guarantee clauses, their practical effect depends substantially on institutional design—most importantly, on “who has been given the authority to interpret Islam’s constraints” and the procedures through which that authority is exercised (Lombardi 2013, p. 616). Depending on their wording and institutional setting, religion-related provisions may thus structure relations among political, judicial, and religious institutions and determine where the authority to define, interpret, and implement religious norms is located. Hirschl further shows that the constitutionalization of religion may not merely recognize or elevate it, but may also bring religious institutions under state regulation and make the state a key participant in selecting and supervising official interpretive authorities (Hirschl 2010, pp. 13–14, 51, 102). In pluralist constitutional settlements, religion-related identity clauses may preserve space for competing actors to attach different meanings to religion, civil statehood, and state identity. In authoritarian or autocratizing contexts, by contrast, their redesign may narrow this space of contestation and concentrate interpretive authority within the state.

This framework guides the analysis of the changes introduced by Tunisia’s 2022 Constitution in relation to the 2014 constitutional settlement. Rather than revisiting the institutional features conventionally identified as authoritarian—such as the concentration of presidential power and the weakening of institutional checks—the article examines changes to a set of provisions not usually included in this category: religion-related identity clauses. Instead of treating these changes simply as indications that the Constitution had become more religious or more secular, the empirical question is how the removal or redesign of provisions related to Islam and civil statehood altered the distribution of constitutional recognition and interpretive authority. It is in this sense that the article develops the concept of the de-constitutionalization of the debate over religion and state: the displacement of a plural and mediated constitutional arena by a state-centered framework. The following sections apply this perspective to the relevant provisions of the 2014 and 2022 Constitutions.

3. Removing Article 1 (“Islam Is Its Religion”): Dismantling an Islamist Constitutional Anchor

Tunisia’s 2022 Constitution marks an unprecedented rupture in the country’s constitutional history: for the first time since independence, the phrase “Islam is its religion,” enshrined in Article 1 of both the 1959 and 2014 Constitutions, was omitted from the constitutional text (Tunisian Constitution 1959, 2014, 2022). Moreover, with the adoption of the 2022 Constitution, Tunisia became the first Arab state to remove a previously established state-religion clause from its postcolonial constitutional order. Shortly before the Constitution was submitted to referendum, President Kaïs Saied framed the anticipated omission not as a secularizing move but as a conceptual distinction: “The next constitution of Tunisia will not mention a state with Islam as its religion, but of belonging to an umma which has Islam as its religion. The umma and the state are two different things” (TRT ʿArabī 2022). Saied’s distinction suggests that Islam is not conceived as the source of state authority. Rather, the state appears as a distinct and prior institutional entity, governed by its own logic and not derived from religion as a constitutional source of legitimacy (BBC News Arabic 2020). To fully grasp the weight of these shifts, it is necessary first to trace what Article 1 meant, how it functioned across Tunisia’s constitutional history, and why its deliberate ambiguity became a defining feature of the post-independence constitutional settlement.

Article 1 of Tunisia’s 1959 constitution declared that “Tunisia is a free, independent, and sovereign state; Islam is its religion, Arabic is its language, and its system of government is a republic.” Most Arab states adopted some version of this state religion clause in their post-independence constitutions and have adhered to it ever since.1 Since constitutions are an imported practice that entered the Arab-Muslim world through colonial influence, state religion clauses were incorporated into Arab constitutions in part to adapt this foreign norm to the local environment and render it consonant with Islam (Ahmed and Ginsburg 2014, p. 619). In many cases, however, including the Tunisian one, the debate over the wording of this clause exposed deep internal divisions between groups holding different views on the role of religion in the state, and led to the adoption of deliberately ambiguous language that allowed competing parties to coexist with the same text and afforded the constitution a broad basis of legitimacy (Böckenförde 2016). This phenomenon is not unique to the Arab world, but a well-documented strategy in constitution-drafting processes, particularly in societies divided over the character of the state: the use of ambiguous or ambivalent formulas as part of what Lerner calls the drafters’ “incrementalist constitutional tool-box” (Lerner 2010, p. 70).

Article 1 of the 1959 constitution was thus deliberately ambiguous, reflecting a debate that had been unfolding in Tunisia since the 1950s between conservative currents—centered around the venerable al-Zaytuna religious establishment—and progressive elites, many of whom had been educated in France, over the character of the postcolonial state. According to the first reading, favored by Islamist currents, the phrase “Islam is its religion” establishes Islam as the official religion of the state, carrying legal force and implying that legislation must conform to Islamic principles, even though the constitution made no explicit reference to the Sharia as a source of legislation. According to the second reading, preferred by civil and non-Islamist forces, the phrase was understood as an elliptical formulation requiring the reader to supply an implied subject (taqdīr): “Islam is the religion of Tunisia’s Muslim-majority society.” On this reading, Article 1 did not establish Islam as the official religion of the state, nor did it carry normative implications for the legal order (Zeghal 2016, pp. 109, 114, 124; Krishan 2022). This latter interpretation also aligned with the vision of Presidents Bourguiba and Ben Ali, who pursued modernization as a state priority and viewed a strong Islamic establishment as a potential impediment to their agenda (Dalmasso and Cavatorta 2013, p. 232). Similar compromise formulations were adopted in several other Arab states (Saʿd 2014). Over time, such clauses became constitutional taboos: rulers were reluctant to reopen them, lest their removal awaken destabilizing public debates (Ahmed and Ginsburg 2014, p. 637).

The 2014 Tunisian Constitution preserved this clause and reworked the logic of compromise, enabling both Islamist and non-Islamist actors to claim continuity with their respective visions (Moine 2014, p. 228). The post-revolutionary constitution maintained the ambiguity of Article 1 by preserving the 1959 wording, “Islam is its religion,” while adding that the article could not be amended, thereby turning it into an eternity clause (Suteu 2017). Yet this continuity was embedded in a broader architecture of mutual reassurance. Most importantly, the retention and entrenchment of Article 1 reassured Islamist and conservative actors that the revolution would not produce a secular state or sever Tunisia from its Islamic identity. This reassurance was reinforced by Article 6’s designation of the state as “guardian of religion” (Böckenförde 2016). At the same time, the new definition of Tunisia as a “civil state” in Article 2, together with the guarantees of freedom of conscience, the neutrality of mosques, and the prohibition of takfīr in Article 6, reassured non-Islamist, liberal, and progressive actors that Islam would not be transformed into a binding source of legislation or a tool of Islamist domination (Lavie 2021b; Gana and Sigillò 2023; McCarthy 2024). The 2014 Constitution thus did not resolve the competing meanings of Islam in the state; rather, it constitutionalized their coexistence by preserving textual imprints through which competing actors could locate their respective visions within the constitutional order (Maboudi 2020, p. 774).

Against this background, the abolition of Article 1 in the 2022 Constitution altered the function it had acquired as a textual site through which the identity debate was constitutionally mediated. Islam was no longer presented as a shared, if ambiguous, identity marker around which competing political forces could negotiate. The rupture was all the more striking because Article 1 of the 2014 Constitution had expressly placed itself beyond amendment. Read against the 2011–2014 constitution-writing process, which Marzouki describes as shaped by a sustained “commitment to compromise” rather than by the victory of one ideological camp over another, this protection can be understood as preserving one of the textual anchors of the post-revolutionary identity settlement (Marzouki 2017, p. 343). By overturning this formula, the 2022 Constitution dismantled one of the constitutional safeguards of that settlement and initiated a broader process of de-constitutionalizing a debate over religion and state that the 2014 Constitution had sought to contain within its text (Lavie 2021b).

The abolition of Article 1 thus did more than eliminate an inherited formula. It withdrew a constitutional reassurance that had enabled Islamist and conservative actors to locate their conception of Tunisia’s Islamic identity within the constitutional settlement. In this respect, the removal of a principal textual foothold for Islamist actors also reflected their political marginalization and the broader decline of political Islam following its rapid ascent after the 2011 uprising. At the same time, the change amounted to a formal-symbolic constitutional disestablishment, weakening one form of what Temperman describes as the state’s “positive identification with a single religion” (Temperman 2010, p. 49). Yet it did not remove Islam from the constitutional order. Rather, it eliminated an ambiguous identity clause that had accommodated competing readings of the relationship between Islam and the state. The concentration of authority over Islam’s public meaning did not result from this omission alone, but from the broader constitutional reconfiguration, above all Article 5, whose significance becomes clearer when read alongside the omission of Articles 1 and 2.

4. Removing Article 2 (“A Civil State”): Withdrawing Reassurance from Non-Islamist Actors

The omission of Article 2 from the 2022 Constitution complements, but should be analytically distinguished from, the omission of Article 1. Whereas Article 1 was Tunisia’s inherited state-religion clause, Article 2 was a post-revolutionary innovation that made Tunisia the first Arab state to define itself constitutionally as a “civil state.” It described Tunisia as “a civil state based on citizenship, the will of the people, and the supremacy of law” and, like Article 1, was expressly shielded from amendment under the 2014 Constitution. The pairing of these two unamendable provisions was central to the architecture of the 2014 identity settlement. While Article 1 reassured Islamist and conservative actors that Tunisia’s Islamic identity would remain constitutionally recognized, Article 2 reassured non-Islamist actors by giving their conception of Tunisia as a civic, non-theocratic polity explicit constitutional expression and by ensuring that Islam’s recognized place in the state’s identity would not transform it into a binding source of legislation (Lavie 2021a, pp. 461–64). Their joint omission in 2022 therefore removed the two textual guarantees through which the 2014 Constitution had accommodated the rival concerns of Islamist and non-Islamist actors. To understand the distinct significance of Article 2’s omission, it is necessary to examine what the “civil state” meant, how it became a shared formula across ideological camps, and what its removal withdrew from the post-revolutionary constitutional settlement.

The notion of the “civil state” (dawla madaniyya) initially carried a strong secularizing connotation, associated with the separation of religion from the state and with a political order grounded in modernism and citizenship rather than religious authority (El-Daghili 2016, p. 190). In recent decades, however, the term has been adopted by Islamist and non-Islamist currents alike, acquiring meanings that complicated the conventional binary between the secular state and the religious state (Lavie 2017a, 2017b). For centrist Islamist thinkers and movements, the “civil state” did not mean separation between state and religion, but rather a non-theocratic state: one headed by civilians rather than clerics or jurists, governed through modern institutions and procedures, and willing to adopt Western norms insofar as they did not contradict the principles of Sharia. In this reading, the civil state remained compatible with the understanding of Islam as both religion and state and as an overarching source of authority for public life (Bahlul 2018, pp. 334–36; Steuer et al. 2023). Non-Islamist currents, for their part, also tended to use the term in a moderated rather than strictly secularist sense. For them, the civil state did not necessarily require a complete separation of religion from the state, but it did require distancing religion and religious authorities from politics, legislation, and decision-making, while preserving Islam’s place in culture, education, history, and collective identity (Lavie and Shalaby 2021). This semantic flexibility enabled the concept to function in Tunisia as a compromise formula. As Bras notes, the “civil state” operated as both a “notion charnière” and a “notion coopérative”: a concept that Islamist and secular actors could endorse while continuing to attach divergent meanings to it (Bras 2016, p. 63).

In this respect, the acceptance of the “civil state” formula by Ennahda, Tunisia’s principal Islamist movement, went further than that of several other Islamist movements in the Arab world. This position drew on a longer ideological evolution, led in part by the movement’s leader, Rached Ghannouchi, that began in the 1980s, when Ennahda articulated commitments to consultation, pluralism, public freedoms, and opposition to authoritarian rule (Cavatorta and Merone 2015, pp. 30–31). This trajectory made it possible for Ennahda to accept the constitutionalization of the civil state in 2014, despite internal disagreements and earlier demands within the movement and among its supporters to constitutionalize Sharia (Netterstrøm 2015, p. 121). The contrast with Egypt is instructive. Although the Muslim Brotherhood and its Freedom and Justice Party also used the language of the civil state in their political discourse, they rejected attempts to define Egypt as a civil state in the post-Mubarak constitutional order (Lavie 2017a). The Tunisian case was therefore distinctive: the civil-state formula was accepted by the dominant Islamist party itself as part of a “strategic conciliation” with non-Islamist forces (Erdoğan 2020, pp. 62–63; McCarthy 2024, pp. 1065–67).

Dorothee Mölle emphasizes the importance of this constitutional settlement in enabling cooperation between secular and religious elites. In her reading, the 2014 Constitution helped Tunisia avoid the “secularism trap”: a destructive struggle in which secular and religious elites seek to neutralize one another rather than construct institutional arrangements for cooperation. The 2014 settlement thus gave constitutional expression to what David T. Buckley labels “benevolent secularism”—a non-exclusionary secular framework that permits religious actors to participate in public life while preventing the subordination of the state to religious authority. It also reflects what Alfred Stepan terms “twin tolerations”: the mutual acceptance whereby religious actors recognize constitutional authority and secular actors accept the legitimate participation of religiously grounded parties in public life (Stepan 2000; Buckley 2017; Mölle 2025). The simultaneous omission of Articles 1 and 2 in 2022 can therefore be read as a textual manifestation of the reversal of this settlement. It removed both the inherited Islamic identity clause and the civil-state clause that had bounded Islam’s constitutional presence within a civic, non-theocratic order, thereby undoing two central anchors of the post-2011 accommodation.

The omission of Article 2 was consistent with Saied’s prior rejection of the very concept of the civil state. As Bouassida shows, as early as 2019, before his election, Saied had declared that the notion of the civil state “does not exist and has no meaning” (Bouassida 2023, p. 31). In the 2022 Constitution, this rejection acquired constitutional form through the omission of Article 2. It was further reinforced by the replacement of Article 49 of the 2014 Constitution, which permitted limitations on rights only when necessary in a “civil and democratic state,” with Article 55’s reference to a “democratic system,” thereby removing the civil-state criterion from the constitutional framework governing limitations on rights. The broader significance of these changes, however, lay in the reconfiguration of the 2014 identity settlement. Just as the omission of Article 1 removed a textual foothold through which Islamist actors had secured recognition of Tunisia’s Islamic identity, the omission of Article 2 deprived non-Islamist and civil actors of the corresponding constitutional foothold through which they had secured recognition of their vision of Tunisia as a civic, non-theocratic polity. Read alongside Article 5, which entrusts the state with realizing the purposes of Islam, these omissions contributed to what Limam characterizes as a shift “from the state of civil values to the state of the nationalisation of religion” (Limam 2025, p. 35). The Section 5 therefore turns to Article 5, where the new logic of state-centered control over Islam’s constitutional meaning is most explicitly articulated.

5. Article 5 and the Authoritarian Monopoly over Islam

Article 5 of the 2022 Constitution can be read as a reformulation of the statist logic already present in Article 6 of the 2014 Constitution. Article 6 defined the state as the “guardian of religion,” but embedded that role within a broader compromise over religion and the state. Adopted within a Constitution produced through dialogue and broad consensus (Ferreiro Galguera 2026, p. 99), it coupled the guardianship formula with guarantees of freedom of belief and conscience and the free exercise of religious practice, as well as commitments to the neutrality of mosques and places of worship, the dissemination of moderation and tolerance, the protection of the sacred, and the prohibition of takfīr—the declaration that another Muslim is an unbeliever—and incitement to hatred and violence (Donker 2019, p. 501; Badar and Florijančič 2021, pp. 490–93; Valciukas and Al Majali 2021, pp. 5–7). This arrangement belonged to the same compromise package through which Ennahda abandoned demands to constitutionalize Sharia as a source of legislation and accepted Article 2’s definition of Tunisia as a civil state (Erdoğan 2020, p. 61).

Article 5, by contrast, defines Tunisia as “part of the Islamic umma” and provides that “the state alone shall work, within a democratic system, to realize the purposes of true Islam in preserving life, honour, property, religion, and freedom.” As Zeghal notes, this enumeration echoes an older Tunisian constitutional vocabulary associated with the explanatory text of the 1857 Fundamental Pact (Zeghal 2024, p. 185). As noted earlier, Saied had already distinguished between the state and the umma, arguing that the new Constitution would refer to Tunisia’s belonging to an umma whose religion is Islam rather than to Islam as the religion of the state. The reference to the umma was not itself new: the 2014 preamble had already affirmed Tunisia’s cultural and civilizational affiliation to the Arab and Muslim umma. What changed in 2022 was its constitutional function. Article 5 moved the reference to the umma from the identity language of the preamble into an operative provision and coupled it with an exclusive mandate assigned to “the state alone” to realize the purposes of Islam. The provision therefore juxtaposes the transnational idea of the Islamic umma with the exclusive authority of the territorial state. This further supports reading the provision as the incorporation of Islamic universalism into a state-centered framework of authority.

The change combined continuity with rupture. The explicit language of guardianship disappeared, but its statist logic was intensified. Some elements of Article 6’s counterbalancing framework were retained elsewhere in the 2022 Constitution: Article 27 preserves freedom of belief and conscience, while Article 28 protects the free exercise of religious rites, albeit subject to public security. Other safeguards disappeared, however, including the neutrality of mosques and places of worship, the commitment to disseminating moderation and tolerance, and the prohibitions on takfīr and incitement to hatred and violence. A state role previously constrained by plural and civic safeguards was thus reformulated into an exclusive and active mandate to realize Islam’s purposes. Islam remained within the constitutional order, but its constitutional function shifted from an element of a negotiated framework in which competing claims were constitutionally mediated to an object of state definition and implementation.

The significance of this shift becomes clearer in the Constitution’s treatment of the “purposes of Islam” (maqāṣid al-Islām). The maqāṣid are a broad juristic concept through which Muslim scholars articulate the rationale, moral principles, and public interests underlying Islamic law. As Jasser Auda shows, modern interpretations range from general values such as justice, freedom, human dignity, and free will to explicitly Islamic purposes such as building a strong Muslim umma, preserving sound belief, purifying the soul, and reforming moral conduct (Auda 2008, pp. 1–8). In the 2014 preamble, the maqāṣid appeared within a broad statement of the Tunisian people’s Arab-Islamic identity and commitment to “the teachings of Islam and its purposes” (Kahlaoui 2015). The 2022 preamble replaced this language with the narrower notion of adherence to “the human dimensions of the Islamic religion,” which can be read as challenging the Islamist conception of Islam as an all-encompassing system (shumūliyyat al-Islām). Moreover, by specifying the purposes of Islam as the preservation of “life, honour, property, religion, and freedom,” and assigning the realization of these purposes to “the state alone,” Article 5 privileges and constitutionalizes one state-selected formulation of a contested Islamic concept while displacing competing interpretations from explicit constitutional recognition.

Existing interpretations of Article 5 diverge over whether it should be understood primarily as an intensification of Tunisia’s statist tradition or as a step toward constitutional Islamization. Zeghal reads the provision not as a complete break with Tunisia’s longer tradition of state custodianship over Islam, but as its deepening. Saied’s Constitution, she argues, “made it thicker,” while Article 5 “gave the state a full monopoly on the custodianship of Islam, hence strengthening it” (Zeghal 2024, p. 185). Other scholars emphasize the provision’s potentially theocratic implications. Nidhal Mekki argues that Article 5 places the Constitution within the realm of religion and may turn Islamic law into a “supra-constitutional standard” (Mekki 2022). Khan similarly reads it as a possible constitutionalization of Sharia, arguing that the shift from “Islam is the religion of the State” to “the State will work to achieve the objectives of Islam” creates an active constitutional obligation where the 2014 text had imposed none (Khan 2022, p. 15). Limam likewise interprets the 2022 Constitution as transforming Islam from a cultural and social descriptor into an operative determinant of legislative policy and judicial interpretation. Her analysis, however, also identifies the political logic connecting these seemingly divergent readings. In her words, “Islam is not propagated for its own sake; it is primarily disseminated to solidify the authority of the state by shaping the consciousness of the masses” (Limam 2025, p. 37). Whether Article 5 is understood as an intensification of state custodianship or as a possible vehicle of constitutional Islamization, both readings point to the same authoritarian dynamic: Islam is not restored as an autonomous source of authority but subordinated to a project of state consolidation.

Other constitutional references to Islam point in the same direction. The requirement that the president be Muslim was retained in Article 88 of the 2022 Constitution, continuing a condition already found in Articles 38 and 40 of the 1959 Constitution and Article 74 of the 2014 Constitution. In the context of the 2022 Constitution’s broader concentration of power in the presidency, however, this requirement further links Islam to presidential authority. Similarly, Article 44 of the 2022 Constitution, echoing Article 39 of the 2014 Constitution, preserves the state’s educational role in rooting younger generations in their “Arab-Islamic identity” and national belonging. Here too, Islam is embedded in the state’s constitutionally defined role in transmitting identity to the population. The treatment of the Mufti points even more clearly to the same pattern. The Mufti appeared explicitly in Article 78 of the 2014 Constitution, which empowered the president to appoint and dismiss the Mufti of the Republic. The 2022 Constitution, by contrast, no longer accords the Mufti explicit constitutional status. The Mufti’s disappearance from the constitutional text thus exemplifies the weakening of intermediary institutions characteristic of authoritarian constitutionalism. Read alongside Article 5, it also points to the concentration of authority over Islam in the state.

The significance of these changes is therefore not merely symbolic. From the perspective advanced here, their longer-term importance lies in the narrowing of the arena in which competing claims over religion can acquire constitutional recognition. If sustained, this reconfiguration may make future contestation over Islam’s public role more dependent on political change at the center of the state than on a plural constitutional framework capable of mediating disagreement. The provision’s practical legal meaning, however, remains unsettled. Tunisian legal scholars have warned that the ambiguity of the maqāṣid may turn Islam into a normative reference for assessing legislation and rights, potentially placing previously secured freedoms under renewed religiously framed scrutiny (e.g., Klibi 2022). Other interpretations, by contrast, emphasize that the phrase “the state alone” assigns the realization of Islam’s purposes exclusively to the state, thereby establishing a state monopoly over their constitutional implementation (e.g., Bras 2022). These competing interpretations have not yet been resolved through constitutional adjudication, as the Constitutional Court envisaged by the 2022 Constitution has yet to become operational. The practical implications of Article 5 thus remain open, even as its significance for the constitutional organization of religion and state is already evident.

6. Conclusions

This article has argued that Kaïs Saied’s Constitution of 2022 de-constitutionalized the debate over religion and state that had been incorporated into Tunisia’s earlier constitutional order. While the 2014 Constitution gave competing understandings of Islam’s place in the state parallel constitutional expression and mediated their coexistence through an architecture of mutual reassurance built around Articles 1, 2, and 6, their redesign in 2022 replaced these compromise formulas not with constitutional silence, but with a different arrangement, assigning to the state alone the mandate to realize the purposes of Islam in Article 5, a state-centered vision closely associated with the president’s political credo. The transformation therefore concerned not merely Islam’s formal constitutional status, but a redistribution of constitutional voice and authority over its public meaning. It thinned the constitutional framework for plural identity negotiation while thickening the state’s substantive mandate in relation to Islam. Yet the purposes of Islam remain sufficiently flexible to retain a degree of constitutional permissiveness (Lerner 2016).

The Tunisian case therefore broadens the analytical scope of authoritarian constitutionalism. Religion-related identity clauses are not merely declaratory statements about the character of the polity; they may also structure political disagreement, distribute constitutional recognition, preserve space for competing interpretations, and sustain intermediary institutions. Their authoritarian redesign can consequently operate through subtraction and addition—not only by granting new powers to the state, but also by removing compromise formulas, rival claims, and institutional intermediaries from explicit constitutional recognition. Saied’s Constitution combines this logic with Tunisia’s longer étatist tradition of governing religion, while embedding it within a more centralized and hyper-presidential constitutional order.

The broader religious and social consequences of this reconfiguration are more difficult to predict. Constitutional texts alone cannot establish how Tunisian believers will understand Islam, whether autonomous Islamic movements will emerge or re-emerge, or how competing claims excluded from explicit constitutional recognition will be articulated outside the constitutional framework. Future research should therefore examine not only how Article 5 is translated into legislation, judicial interpretation, religious administration, and educational policy, but also whether religious, political, and social actors develop alternative arenas in which to contest Islam’s public meaning and the state’s claim to define it.