1For some time now, Turkey has been a textbook case in the rapidly growing political science literature on contemporary forms of authoritarianism. President Erdoğan is invariably included in the shortlist of “strongmen” along with, say, Vladimir Putin of Russia, Victor Orban of Hungary, and Narendra Modi of India, who are often cited en masse as a way of referring to the global rise of authoritarian politics in the hands of elected leaders. The phenomenon is multifaceted, and so is the literature. Drawing on analytical frameworks developed in different traditions of scholarship, political scientists explore current Turkish politics under various rubrics, including “competitive authoritarianism,” “authoritarian neoliberalism,” and “populism,” to name a few.

2Admittedly, these labels and related approaches shed light on certain aspects of Turkey’s political transformation. Without denying their respective merits, however, I want to suggest that they do not sufficiently capture the constitutional matrix of the change under consideration. By this, I do not simply refer to the constitutional amendments of 2017, which created a super-presidency or “Turkish style presidentialism,” as its defenders have sometimes coined it. Equally or perhaps more important, is the fact that the new regime was born as a “dual state” in the context of an unbridled state of emergency declared in July 2016.

3This paper explores the recent (post-2015) phase in Turkey’s autocratization process through a conceptual perspective based on Ernst Fraenkel’s classical study, The Dual State. Originally published in 1941, Fraenkel’s work offers a penetrating analysis of the relationship between law and politics in the early years (1933-1938) of the Third Reich. More specifically, it analyzes the concurrent existence of a “normative state” (Normenstaat) that somehow observes its own laws and a “prerogative state” (Maßnahmenstaat) acting arbitrarily and ad hoc on political decisions unrestrained by any legal norm. Despite its immediate engagement with the historical context of the Third Reich, Fraenkel’s work contains a set of crucial insights regarding the institutional logic and patterns of permanent state(s) of emergency, which are of significance to the study of contemporary forms of authoritarianism in general, and the transformation of Turkey’s political regime in particular.

4To make a preliminary case for the thesis that Turkey’s political regime has evolved into a dual state, I begin with a selective survey of the recent works considering contemporary forms of authoritarianism. The following section introduces Fraenkel’s original conception of the dual state, and offers an interpretive outline regarding its analytical value for contemporary studies of autocratization. Portraying the post-election period in 2015 as a critical juncture in the formation of the dual state in Turkey, I then explore how a sweeping wave of securitization, along with the extensive abuse of emergency powers created an extralegal sphere and fostered a comprehensive institutional transformation towards a dual state. The article’s final section discusses what contemporary transformations have occurred in the broader context of Turkish constitutional history.

Situating Turkey’s authoritarian drift

5The contemporary rise of authoritarian politics could be perceived as so globally pervasive that it seems plausible to describe it as a “new wave of autocratization” (Lührmann, Lindberg 2019). The growing body of work sparked by the erosion of constitutional democracies and the rise of new authoritarian regimes seems to highlight three characteristic features of contemporary autocratization processes.

6Firstly, the incremental modus operandi of autocratization has shown constitutional democracies’ gradual erosion at the hands of elected leaders instead of penultimate death through momentous events such as military coups (Urbinati 2014; Levitsky, Ziblatt 2018; Rodríguez-Garavito, Gomez 2018). Scholars already have at their disposal a rich and ever-expanding vocabulary of descriptive terms that draw attention to this incremental process of autocratization. Terms such as “backsliding,” “decline,” “setback,” “retreat,” “recession,” “de-democratization,” “autocratization,” et cetera, are widely in circulation, each with their own nuances (Cassani, Tomini 2020).

7Another point that stands out in the extant scholarship is the use and abuse of law, increasingly constitutional law, as a key factor in contemporary processes of autocratization. More often than not, deformative changes that substantially undermine constitutional democracies and serve to entrench autocratic modes of rule take place under legal disguise through seemingly constitutional means, including plebiscitary practices of constitution-making and constitutional amendment(s). On this register, too, there is a set of new concepts, such as “rule by law” (Ginsburg, Moustafa 2012), “abusive constitutionalism” (Landau 2013), “autocratic legalism” (Corrales 2015; Scheppele 2018), “populist constitutionalism” (Blokker 2018), and “strategic legalism” (Yılmaz 2020), each of which uniquely draws attention to the instrumentalization of legal and constitutional forms in the service of autocratic politics (see also Huq, Ginsburg 2018).

8Lastly, there is also broad agreement regarding the hybridity of political forms fostered by contemporary autocratization processes. Such regimes blend a range of formal democratic institutions (most notably, elections) with practices of autocratic governance, thereby creating hybrid systems categorized as “competitive authoritarianism” (Levitsky, Way 2010) or “electoral authoritarianism” (Schedler 2013), i. e. systems in which “elections are real but not fair,” and it becomes increasingly difficult to replace incumbents even though they are not immune to being challenged on a political sphere.

9As various studies have shown, the Turkish case overlaps in many ways with the preceding threefold picture (Özbudun 2015; Akkoyunlu, Öktem 2016; Esen, Gümüşçü 2016; Somer 2016; Çalışkan 2018; Yılmaz 2020). Yet, it also displays certain traits that cannot necessarily be easily categorized. While Turkey’s autocratization process unfolded incrementally under the rule of AKP, it was also punctuated by a drastic state of emergency from July 2016 to July 2018, which served as a springboard for a rapidly imposed regime change. The new, so-called “presidential” system was born in the context of an unbridled emergency rule through the abuse of emergency powers, which are “dictatorial” powers in the classic sense of the term (Arato 2014; Schmitt 2014). Further, the aforementioned use and abuse of law have played a crucial role in Turkey’s autocratization process, but so has extralegality, to be discussed shortly. Law has not only been abused but also set aside when seemingly expedient to do so.

10Harkening back to before the state of emergency, Turkey was in a process of “deconstitutionalization” fostered by President Erdoğan and the ruling AKP through deliberate and systematic breaches of the constitution (Kaboğlu 2013, Gözler 2016). From 2015 onwards, the process of deconstitutionalization, interlocked with a sweeping wave of securitization, culminated in a nationwide state of emergency that institutionally regulated extralegality through executive decrees exempt from judicial review (Kaygusuz, Aydın 2020). Therefore, it is imperative to explore this type of institutional hybridity arising from Turkey’s autocratization from a constitutional point of view, paying due attention to the formation of a dual state.

The dual state: what it is and why it matters

11In the 1974 Preface to the belated German edition of The Dual State (1941) Ernst Fraenkel reflects on what originally stimulated the main idea of the work:

Anyone who did not shut his or her eyes to the reality of the Hitler dictatorship’s administrative and judicial practices, must have been affected by the frivolous cynicism with which the state and the [Nazi] party called into question, for entire spheres of life, the validity of the legal order while, at the same time, applying, with bureaucratic exactness, exactly the same legal provisions in situations that were said to be different. Based on the insights into the functioning of the Hitler regime that I gleaned from my legal practice, I believed to have found a key to understanding the National Socialist system of rule in the duality or concurrent existence of a “normative state” (Normenstaat) that generally respects its own laws, and a “prerogative state” (Maßnahmenstaat) that violates the very same laws. (Fraenkel 2017: xv)

12The kind of duality that Fraenkel described in the above passage had for him a more personal – or one might even say, existential – edge to it. On the one hand, as a Jewish person and a socialist, he was among those who were directly targeted by the Nazi prerogative state. On the other hand, as a practicing lawyer, he was still entitled to sit at the defense table of the courtroom and joined the everyday functioning of the normative state, which allowed him to observe the metamorphosis of the law from within. Thus, extrapolating from his point-of-view of tangible courtroom cases, Fraenkel offered a penetrating analysis of law and its relation to politics in the early years of the Nazi dictatorship.

13However, even though The Dual State was meant to analyze the immediate political context in which it was written, its lasting theoretical contribution tends to focus on the conceptualization of dual-system rule as an “ideal type,” one which sheds light on the “institutional logic” of permanent emergency (Meierhenrich 2018). This is also why perhaps Fraenkel himself subtitled the book as A Contribution to the Theory of Dictatorship, suggesting that the modern experience of dictatorship was tied to an apparent abuse of emergency powers to dismantle legal constraints on “the political” without necessarily abandoning the law.

14Drawing on Fraenkel’s work by abstracting his insights from their immediate historical context, I underscore three aspects of the dual state taken as an ideal type: (1) The self-referential primacy of the prerogative state, (2) the normative state functioning only when the prerogative state remains silent, and (3) the normative and the prerogative halves of the dual state not corresponding to two separate bodies.

15The self-referential primacy of the prerogative state could be defined as it alone deciding the scope of its own mandate. The prerogative state addresses itself exclusively to matters of national security, but what counts as a matter of national security is exclusively decided by the prerogative state (hence its essential relation to “securitization,” elaborated on further). Since its mandate is not subject to effective judicial control on the basis of foreseeable norms, the power of the prerogative state is limitless in principle. In effect, however, it acts within certain limits that bend at will, depending on the circumstances. In the dual state the limits of political power are not drawn by legal norms, but by the prerogative exercise of political power limiting the validity of legal norms. Or, in Fraenkel’s formulation, “the presumption of jurisdiction rests with the normative state,” whereas “the jurisdiction over jurisdiction rests with the prerogative state” (1941: 57).

16The second aspect is that the normative state speaks so long as the prerogative state remains silent. In other words, the normative state keeps functioning in accordance with extant laws and regulations in cases and spheres of life which are not currently subject to prerogative intervention. Even though the prerogative state can interfere with anything and anytime does not mean that it will interfere; what is currently irrelevant to the prerogative state falls under the jurisdiction of the normative state. The latter has a regulative function that is indispensable to the integration of modern complex societies. Hence, it is crucial for the regime’s societal legitimation. The normative state is productive in everyday life, thereby creating patterns of relative stability vital for the continuity of the dual state.

18Based on these ideal-typical features, one could define the dual state as a form of autocratic rule in which law continues to matter. Yet, its application and sphere of validity are rendered systematically unpredictable due to the extensive use of prerogative power. Each component of this definition helps us differentiate the dual state from other political constellations.

• The dual state is a form of autocratic rule in the sense that it defies the basic tenets or the “historically valid nuclear meaning” (Sartori 1962: 860) of constitutionalism―most notably, limited government, rule of law, and the effective protection of rights―at a systemic level. On the other hand, or to put it the other way around, a political regime is not a dual state insofar as the constitution in effect not only organizes the exercise of political power, but also restrains it regularly enough to meet the minimal normative standards of constitutionalism.

• Even though constitutionalism is substantively undermined in the dual state, formal law continues to matter, at least in a functional sense, as a way of stabilizing social action. If formal law does not apply in a political system, a dual state does not exist, but rather a condition of pure arbitrariness. That is to say, “the dual state concept is not an illuminating and appropriate description of political regimes when the prerogative state rules with such an amount of arbitrariness and terror that law is either completely annihilated or rendered virtually insignificant” (Suntrup 2020: 16).

• • 2 A conceptual and terminological clarification is in order. I use the term “authoritarian legalism” (...)

  While law continues to matter in the dual state, its application is systematically unpredictable due to the extensive use of prerogative power. There is a crucial difference between the predictable application of law even though its content is illiberal (which I call authoritarian legalism) and the unpredictable application or omission of legal norms, regardless of their content, depending on the changing expectations of the prerogative state (which I call dynamic uncertainty). The dual state is a regime of dynamic uncertainty rather than authoritarian legalism.2

19This ideal typical backdrop makes Fraenkel’s work contemporarily relevant, especially regarding the use and abuse of emergency powers, the forms of hybridity fostered by autocratization processes, and the function of law under autocratic rule in general. The recent turn to Fraenkel from a contemporary angle (e.g. Sakwa 2010; Meierhenrich 2018; Tekin 2018; Köker 2020; Suntrup 2020) is therefore not unpredictable.

The dual state in the making: deconstitutionalization and securitization

21The context of the public address is also instructive. Erdoğan made the above-mentioned speech approximately six months before the coup attempt in July 2016, meaning there was no legally declared state of emergency in effect. Nonetheless, it was a turbulent time, and a lot had happened since the AKP had lost its parliamentary majority in June 2015 general elections. Negotiations for a peaceful solution to Turkey’s longstanding Kurdish question was brought to an abrupt end by the Erdoğan administration proceeding the elections, and violence resurged between Turkish security forces and the outlawed Kurdistan Workers’ Party (PKK). As deadly armed clashes took place throughout the Kurdish provinces of Turkey, inter-party talks for a coalition government failed and the renewed elections in November 2015 resulted in the AKP’s regaining the parliamentary majority.

22During the armed conflict in Kurdish cities, provincial and district governors systematically imposed curfews on the grounds that it was a necessary measure to enforce security and to track down armed militants. In the meantime, however, human rights organizations documented severe corporal crimes by security forces, including “excessive use of force, killings, enforced disappearances, torture, destruction of housing and cultural heritage, incitement to hatred, prevention of access to emergency medical care, food, water and livelihoods” (OHCHR 2017). A few days before Erdoğan’s aforementioned meeting with the district governors, the most severe and intense human rights violations reportedly occurred during curfew, especially over the five weeks from December 11, 2015 to January 21, 2016 (HRFT 2016).

23When the officials in the meeting were told to set aside the law, they had already been doing so. Not only were the numerous violations against civilians illegal, but also all curfews enacted by governors in Kurdish provinces and districts were plainly unconstitutional. According to Turkish law, governors are authorized to impose a curfew only during a state of emergency, but there was no officially declared emergency at the time. What was in force was the prerogative state, acting outside of the law. Erdoğan’s double statement about the validity of law and the legitimacy of negating the law was in this sense a true proposition corresponding to brute reality.

24I take the turbulent context of the post-election period in 2015 as a critical juncture in the formation of the dual state in Turkey. There is disagreement among scholars of Turkish politics regarding how to date the country’s authoritarian turn, with the search for turning points being criticized as a “fallacy” (Erensü, Alemdaroğlu 2018). In a sense, such disagreement is hardly surprising given the incremental nature of the autocratization process and the multiplicity of factors that have fostered it. To avoid misunderstanding, in pointing out the post-election period of 2015 as a critical juncture, I do not mean to feature it as the turning point in Turkey’s authoritarian trajectory in general. I agree that even the period of allegedly liberal reforms in the early years of AKP was imbued with repressive practices, especially at social and political margins, that the small-scale or capillary emergencies have been all too pervasive in the everyday life of discriminated groups, with police violence remaining a common occurence. In addition, the instrumentalization of the judiciary for shifting political purposes has a well-recorded history over the entire AKP era. Large-scale political trials put on far before 2015 (especially those commonly known as “Ergenekon,” “Balyoz,” and “KCK” trials, which were handled by special courts under the political patronage of Erdoğan’s administration) attest to the fact that AKP has extensively used and abused law to criminalize or decriminalize its shifting foes and allies (Kaynar 2021). In short, some of the elements that make their way into the formation of the dual state, albeit in a new constellation, were already in place before 2015. What is it, then, that makes the post-election period in 2015 a critical juncture? In the post-election period, I argue processes of “deconstitutionalization” and “securitization” interlocked with one another in such a way as to foster a de facto regime best qualified as a dual state.

25“Deconstitutionalization” – in Turkish anayasasızlaştırma, a term coined after the Gezi protests (Kaboğlu 2013) – refers to a process in which the Constitution is stripped of its binding force as a result of deliberate and systematic breaches by constitutional organs refusing to comply with certain constitutional provisions as they see fit without facing legal consequences (Gözler 2016). This does not imply the abolishment of the Constitution as such, but the undermining of constitutionalism as a normative principle premised on the separation of powers and the protection of fundamental rights. In other words, the Constitution neither ceases to exist nor falls into complete desuetude in a process of deconstitutionalization, rather it loses its capacity to effectively restrain political power due to intense practices of “constitutional bad faith” (Pozen 2016).

26It is hard to pinpoint when exactly deconstitutionalization begins because it is not an “event” such as a coup or a revolution but rather a “process.” The constitutional referendum in 2010, the inter-party negotiations for a new constitution from 2011 to 2013, the violent repression of Gezi protests in 2013, and Erdoğan’s presidency in August 2014, all contributed to Turkey’s deconstitutionalization in their own ways. The constitutional amendment in 2010 was a prime instance of “abusive constitutionalism” (Landau 2013) that facilitated deconstitutionalization by dismantling certain checks and balances and redesigning the judiciary for partisan purposes (Gözler 2017). The seemingly democratic inter-party negotiations for a new constitution from 2011 to 2013 also facilitated deconstitutionalization. Once the negotiations had failed, as they were meant to fail from the outset on account of unrealistic consensus requirements (Tombuş 2019), the existing constitution was politically delegitimized without being replaced by another. The Erdoğan administration’s repressive response to Gezi protests in 2013 was a further step in the process of undermining “constitutional supremacy” (Kaboğlu 2013). In the face of extensive and extralegal police violence enabled by a “sovereign decision,” constitutional guarantees regarding the protection of rights had turned into parchment barriers. Having won the presidential election in 2014, Erdoğan persistently ignored the constitutional norms of the parliamentary system, urging that with a president directly elected by the people, Turkey’s system of government had now changed in practice; the constitution must conform to this fact, not vice versa (Hürriyet Daily News 2015). The process of deconstitutionalization was already underway when AKP lost its parliamentary majority in the June 2015 general elections.

27While deconstitutionalization is often a necessary condition of the dual state formation, it is not necessarily a sufficient one. For the rise of a prerogative state exercising emergency powers outside the law, the political sphere must also fall under influence of securitization. On a rudimentary level, “securitization” refers to discursive and institutional practices by means of which certain issues are redefined as a matter of existential threat to national security and thereby taken out of the sphere of public debate and ordinary democratic politics (Buzan et al 1998; Balzacq 2011). When successful, it helps to justify the use of emergency powers in relation to the “securitized” issues and/or groups, and enables the “securitizing actor” to act in ways that would be otherwise impermissible both de jure and de facto. In the aftermath of the June 2015 elections, and with the end of peace talks, securitization in Turkish politics fully engaged (Özen et al. 2021), encompassing the entire public sphere, while at the same time firmly interlocking with the deconstitutionalization process that had already been underway.

28The curfews were one instance, among others, where emergency powers were put into use without legal grounds, and the concurrent existence of two modalities of state action – one in accordance with the law, the other outside the law – came to be routinely visible. Another case in point is the internationally known “Peace Petition” released by Academics for Peace on January 11, 2016. The petition, endorsed by 2210 academics, publicly denounced the gross human rights violations committed by Turkish security forces during the curfews and urged the decision makers to resume peace negotiations. In a troubling response, President Erdoğan lashed out at Academics for Peace and accused them of supporting terrorism. Higher education authorities, most notably the Higher Education Council (YÖK) and the Inter-university Board (ÜAK), publicly stated that the petition was “irreconcilable with academic freedom,” that “supporting terrorism cannot be tolerated in any way,” and that “academic freedom cannot be abused to pose a threat to the existence, security and survival of the country” (HRFT 2019). University rectors retaliated against the petitioners by launching unlawful administrative investigations, issuing suspensions, and firing and/or blacklisting academics (Altıparmak, Akdeniz 2017, Tekin 2019). Some of the petitioners, especially those in small towns, were personally threatened by fascist groups and were forced to leave their respective towns for safety reasons. In one striking example, a threatened scholar who reached out to the police was denied protection (Bianet 2016a). This case was a textbook example of successful securitization, demonstrating the extent to which public authorities held legitimate dissent as detrimental to “national survival” and immediately identified it with “terrorism” (Baser et al 2017, Tekin 2019).

29In the post-election period in 2015, deconstitutionalization and securitization entrenched one another, thereby generating a devastating cycle. The more the Constitution lost its capacity to restrain political power, the easier it became to crimininalize opposition and to use counterterrorism measures against dissenting voices in civil and political society. In other words, deconstitutionalization facilitated the securitization of democratic politics. In turn, the effective or successful process of securitization provided decision makers and public authorities with a convenient pretext for “setting the law aside,” as President Erdoğan urged the district governors to do. Simply put, securitization legitimized deconstitutionalization. The outcome of this self-entrenching cycle was the formation of a dual state, in which constitutional norms were not altogether suspended but rather applied or omitted ad hoc, depending on political decisions and alleged necessities of national security. Turkey’s de facto dual state rapidly grew into a full-fledged institutional reality with the aforementioned abuse of emergency powers after the failed coup attempt on July 15, 2016.

State of emergency unleashed

30Turkey underwent a nationwide state of emergency from July 2016 to July 2018. The state of emergency was unbridled in a twofold sense. First, the executive branch went far beyond the limits of emergency powers provided in the Constitution, and usurped both legislative and judicial authority through emergency executive decrees. Second, sharply deviating from its former jurisprudence, the Constitutional Court refused to review whether or not the emergency decrees were in conformity with the Constitution, thus granting the executive branch with a carte blanche to do anything as long as the executive considers it as part of emergency measures.

31During the state of emergency, executive decrees were the main instrument of the prerogative state. The first decree of its kind, Emergency Decree No. 667, declared on July 23, 2016, established a set of unprecedented procedures for the unconstitutional and extralegal practices to be enacted, thereby creating a “manual” that served as the “real constitution” of the prerogative state. For instance, it provided for the immediate dismissal of public employees who were “deemed” to be affiliated with “terror organizations or those structures, formations or groups which the National Security Council has decided as acting against the national security of the State” (Emergency Decree No. 667, Articles 3 and 4) without due process. However, Article 15 of the Constitution explicitly states that “no one shall be held guilty until so proven by a court ruling” and that this norm applies in all circumstances, including “war, mobilization, and state of emergency.” Thus, the Emergency Decree created an extralegal sphere in which the presumption of innocence, a non-derogable right protected by the Constitution, was no longer valid.

(1) In a constitutional state, the emergency regime is not an arbitrary, lawless regime. It is a rule of law regime, limited with the constitutional provisions and the law in general. It stands under the control of the legislative as well as the judicial power. (2) Its aim is to protect the ordinary law order with means of extraordinary measures, but always within the limits of law. (3) Emergency decrees of the executive power should not exceed the limits of the aim of the proclamation of the state of emergency. They are allowed only to regulate operative matters in terms of state of emergency. (4) The general laws cannot be amended with emergency decrees since they lose their validity after the end of the state of emergency regime. The executive may only act with measures, but not with general regulations. (5) Emergency decrees only prevail to the regions that state of emergency has been declared in, for the specified state of emergency period. The objective is to resolve the cause of the state of emergency. (6) Even though the Constitution excludes the emergency decrees from judicial control, the Constitutional Court is authorised to determine the characteristics of them, meaning if they are actually in the scope of the state of emergency. Otherwise an extralegal area could be established. (7) During the judicial control, not only the Articles 119 to 122 provisions, but the Constitutional provisions as a whole have to be taken into consideration.

34In short, according to the former opinion of the Court, the executive could not issue emergency decrees for reasons unrelated to the declared purposes of emergency, nor could they be applied outside the emergency location, nor in such a way as to be valid after the emergency was over; otherwise, the Court emphatically urged, emergency decrees could be used to create an extralegal sphere and the constitutional regime of the state of emergency would degenerate into an arbitrary reign of executive power.

35When the Court changed its opinion in October 2016, and ruled that Article 148 must apply unconditionally, this proves paramount to the abovementioned degeneration. The new jurisprudence of the Court confirmed the primacy of the prerogative state: regardless of how the Constitution limits the emergency regime as to time, location and purpose, an emergency decree became what the executive branch decreed it to be. Unsurprisingly, the executive branch made the most out of this open check. For instance, Emergency Decree No. 676, the first regulative decree released two weeks after the Court’s ruling amended the Higher Education Law, abolished rectoral elections at public universities, and authorized President Erdoğan to appoint university rectors as he saw fit. While there is no relationship between the official purpose of the emergency and the appointment procedures of university rectors, the opinion of the Constitutional Court nonetheless affirms that any proposal could be a matter of emergency and hence exempt from judicial review so long as the executive deems it an “emergency decree.”

Institutional cohabitation and dynamic uncertainty

36The primacy of prerogative power over normative regulations is a key indicator of dual state formation. Having removed judicial review on emergency decrees, the Constitutional Court’s fateful decision endorsed this self-referential primacy, thereby paving the way for a two-track institutional reality in which public institutions function in two modes at once: one in accordance with extant legislation, the other in accordance with extralegal measures. I describe this as “institutional cohabitation” of normative and prerogative functions, i.e., the concurrent existence of two modalities of state action within the same institutional setting.

37The political history of modern Turkey is not unfamiliar with extraordinary institutions performing prerogative functions. For example, the notorious State Security Courts (Devlet Güvenlik Mahkemeleri), first established in 1973 but abrogated by the Constitutional Court in 1976, were restored after the military coup of 1980 and incessantly operated until 2004. They functioned alongside the regular judicial system, had their own procedures, and addressed political and criminal offenses deemed to be detrimental to the security of the state. These courts were meant to perform prerogative functions, and they did so in a special capacity. They were separate institutions endowed with special powers. In that model of what might be called “institutional differentiation,” extraordinary adjudication was demarcated from the ordinary judicial system, and the latter remained relatively intact, thereby ensuring a predictable stability.

39During the two-year state of emergency, 6,081 academics were dismissed and banned from academic employment and civil service for life (6,055 from public universities and 26 from private universities) through emergency decrees. Dismissals were based on arbitrary accusations of “affiliation with a terrorist organization,” which were not verified before a court of law. The number of academics dismissed from public universities amounted to 4.57 percent of all academic staff employed at Turkey’s public universities at the time. This figure is proportionally lower than but comparable to the academic purge in Nazi Germany, where the loss was estimated 14 to 16 percent (Hartshorne 1937: 87-95). What is more striking than the cumulative number of dismissed academics, however, is the way in which this extralegal act of the prerogative state was institutionally regulated and the extent to which universities themselves were involved in it.

40Following the failed coup attempt on July 15, 2016, the Council of Higher Education (Yükseköğretim Kurulu - YÖK), the supreme administrative body in charge of Turkey’s higher education institutions, convened all university rectors for an anomalistic meeting in Ankara. Proceeding the meeting, the Council asked all faculty deans at Turkish universities, both public and private, to resign from their administrative positions, and a total of 1,577 deans complied without exception (Bianet 2016b). The official statements of the Council stated that it assumed responsibility for “coordinating and orienting universities” in accordance with emergency measures in general, and for ensuring the “cooperation” of university administrators with “local governers, police forces and other judicial bodies” in particular (YÖK 2018: 16).

41Despite its supervising and coordinating role, however, it was not the Council itself that prepared the particular purge lists. Rather, as several statements by higher education authorities, including university rectors, clearly confirmed (BBC Türkçe 2017a, 2017b), the university administrations decided who were to be stigmatized and dismissed without due process as “having affiliations with a terrorist organization”. One might ask, in this case, how many public universities actually took part in the academic purge as an institutional agent of the prerogative state. The short answer is all but two. When the state of emergency began, there were 108 active public universities in Turkey. Among them, only two institutions abstained from the academic purge and dismissed no one (Boğaziçi University and Social Sciences University of Ankara). The remaining 106 universities were actively involved in the process. The frequency of dismissals by each university reveals a wide range extending from one dismissal (Galatasaray University) to 271 dismissals (Süleyman Demirel University) in a single institution. The purge was more prolific in some institutions as opposed to others, but almost all public universities were engaged in performing prerogative functions. It was not an institutionally differentiated agent of the prerogative state that solely carried out the academic purge, but universities themselves.

42A further manifestation of the dual state is represented in the systematic erosion of legal predictability. In a comprehensive study of Fraenkel’s work, Jens Meierhenrich argues that the very notion of the dual state is an “analytical shorthand for the institutional logic of a particular kind of rule” (2018: 182). At the center of this institutional logic is the dynamic division of labor between the prerogative and the normative states, based on the primacy of the former. The scope of the normative state is constantly drawn and redrawn, shrinks and extends, according to the prerogative state’s shifting priorities and ad hoc needs. The institutional logic of the dual state undermines legal predictability without abandoning law as such. Thus, it is important to distinguish between the predictable application of law even though its content is illiberal (authoritarian legalism) and the unpredictable application or omission of legal norms regardless of their content (dynamic uncertainty). As stated before, the concept of “authoritarian legalism” refers to the substantial content of the law being incompatible with or detrimental to liberal democratic rights. However, the principle of legality is nonetheless adhered to, i.e., the law is sufficiently general, public, prospective, coherent, clear, stable, and practicable (Fuller 1964). In contrast, the unpredictable application or omission of legal norms, regardless of their content, is another issue. It creates a condition I deem “dynamic uncertainty” which undermines the very principle of legality itself. The institutional logic of the dual state fosters dynamic uncertainty rather than authoritarian legalism. This applies to the current condition in Turkey.

Even more concerning in terms of the independence of the judiciary, there have been instances where there were serious concerns about judges being pressured to reverse their decisions, and where judges who issued decisions acquitting and/or ending the arbitrary detention of civil society actors and HRDs [human rights defenders] have been suspended from those cases and replaced with others. (OBS & İHD 2021: 23-24)

44Unjustified administrative restrictions on civic freedoms is yet another compelling case in point. In a recent study on “protest bans” imposed by local governors, for instance, Arslanalp and Erkmen (2020) explore how the frequent use of this specific administrative measure creates a dynamic form of emergency-like suspension of rights across national territory, which they coin “mobile emergency rule.” The protest bans work, they argue, by producing ambiguity:

Despite all the talk of rule of law, these governmental practices violate its basic tenets like clarity, prospectiveness, generality, and stability. (…) Ambiguity manifests itself in the shifting terrain of what, how, where, when, and by whom the government allows protesting. What is permitted one month, might be banned in the next month. The kind of activities disallowed may expand or contract in a matter of few days. In some cases the wording of the ban is ambiguous, allowing the authorities to interpret it as it suits their intentions. Moreover, ambiguity also pervades the enforcement of the ban by the police forces. (Arslanalp, Erkmen 2020: 956)

45The reports of human rights organizations confirm this argument with solid evidence, particularly with respect to the wide range of ad hoc restrictions on civic freedoms, including freedoms of expression, assembly, and association (HRFT 2021a, 2021b, 2021c; OBS & İHD 2020, 2021). The message arising from these reports is conclusive: emergency rule has become permanent in Turkey even though the state of emergency was officially lifted in July 2018.

Conclusion

46The idea of a dual state offers an avenue to analyze the seemingly paradoxical category of “permanent emergency” in concrete institutional terms. One lingering question is whether or not permanent emergency along the model of a dual state is something new in modern Turkey, a justified question, given the extensive use of emergency powers in Turkey’s political and legal history.

47In an incisive discussion of Turkey’s “tradition of authoritarian constitutionalism,” Işıksel draws attention to the fact that, in addition to earlier episodes of martial law or states of siege, “the state of emergency declared in 1987 ostensibly to restore order in two Kurdish provinces was renewed a total of 46 times in up to 14 provinces, and lasted until the end of 2002” (2013: 718). She argues that the 1982 Constitution was imbued with a sense of emergency. Nonetheless, making these remarks before the drastic wave of deconstitutionalization in 2015, Işıksel also notes the extent to which the Constitution plays a binding role in the conduct of Turkey’s politics: “the contemporary Turkish political system, in spite of its pronounced authoritarian features, seems to operate within the bounds of its own constitutional order,” thereby displaying a considerable degree of “constitutional discipline” (2013: 704).

48A twofold assessment of this sort, on one hand critically drawing attention to the deep-rooted authoritarian elements in Turkish constitutionalism, and on the other, soberly acknowledging its certain assets, is not unique in the literature. Bülent Tanör, a prominent constitutional scholar and a staunch critic of the 1982 Constitution, for example, made the following observation in early 1990’s in the context of his seminal study on Turkey’s “problem of human rights” (1994: 149).

Despite all deficiencies and arbitrary leanings, there exists [in Turkey] a notion of “constitutional government” along with its institutions and rules. The separation of powers and the parliamentary regime are factual realities. The 1982 Constitution did not abolish them. As for rule of law, judicial independence and judicial review (i.e., administrative courts and the Constitutional Court), even though there have been regressions, the extant assurences and resources are still of considerable degree.

49In the present context, however, neither an established notion of “constitutional government” nor the “separation of powers” can be said to exist in reality. In the same vein, the type of “constitutional discipline” emphasized by Işıksel has become antiquated. The process of deconstitutionalization, intertwined with nationwide securitization since 2015, culminating in the formation of a dual state in the context of an unbridled state of emergency, drastically changed Turkey’s constitutional reality. Echoing the foregoing distinction between “authoritarian legalism” and “dynamic uncertainty,” the Turkish government’s current condition would not necessarily qualify as constitutionalism with a strong authoritarian grasp, but rather as the demise of constitutionalism as such. Compared to the historical record of Turkish constitutionalism, therefore, Turkey has been experiencing something new.