Mediated Stateness as a Continuum: Exploring the Changing Governance Relations between the PLO and the Lebanese State

Since the 1960s, the Palestine Liberation Organization (PLO) has implemented an extensive governance project in Lebanon that is often regarded as contributing to the weakness of the Lebanese state. Challenging such zero-sum logic, this article explores the institutional interdependencies between the PLO and the Lebanese state and their different yet mutual interests in governance coordination. It conceptualises the relations between the PLO and the Lebanese state along a continuum of mediated stateness and thereby contributes to both the operationalisation of the notion of the mediated state and our understanding of the diverse empirical manifestations of the PLO’s governance in Lebanon.


Introduction
Governance, the acquisition of public authority through the provision of public goods related to security, welfare and representation, is not the prerogative of states. In many situations, the state is only one among several governance actors -and not always the most significant one. Social and political scholarship has consequently increasingly concerned itself with rendering visible and theorising forms of governance that go beyond the state. In recent years, academic thinking about state and non-state providers of public goods has moved beyond focusing on dichotomies, such as state and non-state, formal and informal, modern and traditional, coercive and consensual forms of governance, to showing overlaps, interdependencies and nuances.
The idea of 'mediated stateness' has played an important role in this development. Departing from the premise that state actors in (post-) conflict settings often strategically co-opt and engage non-state armed actors rather than fighting them, the mediated state conceptually paves the way for the notion of multi-layered governance that is central to this special section. This article develops such multilayeredness by applying the mediated state notion to the dynamic, and extensive, governance project that the Palestine Liberation Organization (PLO) has implemented in Lebanon since the late 1960s. It presents a dialogue between theory and evidence that enhances our understanding of the empirical manifestations of evolving PLO governance in Lebanon and further develops the analytical tool of the mediated state.
Lebanon is infamous for its 'weak state' and the (armed) presence of the PLO in the country is widely regarded as having contributed to this state weakness. In this regard, the PLO's governance activities in Lebanon are often cast as constituting a 'state-within-the-state' (Brynen 1989, p. 58, Beker and Van Oordt 1991, p. 16, Atzili 2010. Such zero-sum logic, however, does not do justice to the diverse empirical manifestations of the PLO's governance in the country. In the 1970s, the PLO reached the peak of its institutional activities and its governance often undermined the authority and sovereignty of the Lebanese state. After the Lebanese Civil War (1975-90) ended, however, the PLO's governance project took a more modest shape and became largely subservient to the recuperating Lebanese state.
In this article, I demonstrate how my interpretation of the mediated state concept can help account for these changes in PLO governance through exploring the relationship between the Lebanese state and the PLO. Mediated stateness thus contributes to understanding multilayered governance by putting a premium on explaining the interaction between state and non-state forms of governance, rather than merely emphasising competition and contestation. It by no means aims to replace a competition-centred approach with a cooperation-centred one, but seeks to complement it; my point in this article is not to claim that the PLO and the Lebanese state only coordinated, but to demonstrate that they also coordinated.
I present two case studies of mediated stateness corresponding to the two periods of PLO governance in Lebanon. Analysis of the first case is based on secondary sources, while the second case, concerning the PLO's current governance activities in Lebanon, draws from a specific in-depth case-study in South Lebanon (Stel 2015(Stel , 2016(Stel , 2017. 1 Through an extensive literature review, this article seeks to historically contextualise these contemporary ethnographic findings and re-conceptualise them as part of a governance continuum. Consequently, this article combines a literature-based macro-perspective with a more empirical micro-perspective to connect past and present forms of PLO governance in Lebanon.

Mediated Stateness
Since 1969 the PLO has had an extensive governance project in Lebanon, providing security, services and representation to the country's Palestinian refugees and, at times, even to Lebanese citizens. The continuation and form of such nonstate 2 governance can be understood from various conceptual perspectives. One could start out from the weakness of the state that has allowed for the rise of non-state governance, the premise of the 'failed state' paradigm. The idea of state failure, however, promotes an overly simplistic version of Weber's famous state-centric theory of governance. It puts an analytical premium on sovereignty, the idea that 'no other actor may gainsay the will of the sovereign state' (Van Overbeek 2014, p. 18). This makes it ultimately unable to 'theorise about arenas of competing multiple sets of rules, other than to term these as negative, as failures, weak states or even non-states' (Migdal and Schlichte 2005, p. 12, see also Hoehne 2009, Hoffmann andKirk 2013, p. 6).
Another perspective that helps to make sense of governance beyond the state is the idea of rebel rule (Weinstein 2007, Arjona et al. 2015, Kasfir 2015. The study of rebel rule signifies a 'revalorization of non-state forms of order and authority' (Meagher 2012(Meagher , p. 1073. It focuses on the autonomy and interests of governance actors that are not related to the formal state, such as rebels and insurgents. Yet just as the idea of the failed state has no eye for the non-state, the notion of rebel rule is mostly silent on how 'parallel' governance and 'parcelized' sovereignty (Wickham-Crowley 1987, pp. 475, 476, 494) relate to the state. 3 Governance actors, however, whether state officials or not, do not operate in a vacuum and the resources and repertoires that affect the ways in which they govern will be influenced by other actors in the governance arena they inhabit.
To understand non-state governance, therefore, I suggest turning to a third perspective, which puts a premium on interaction between governance actors. This allows for a more comprehensive and holistic understanding of how nonstate governance actors emerge and endure and why their governance takes the form it does. It is also, as such, much better geared towards capturing and explaining fluctuations in non-state governance projects. The concept of the mediated state, developed by Menkhaus (2006), is particularly expedient in theorising how states in hybrid political orders need not necessarily compete with other loci of authority, but often opt for a more pragmatic form of engagement that allows them to govern alongside and sometimes even through non-state actors. Menkhaus (2007, p. 78) defines a mediated state as a political order 'in which the government relies on partnership (or at least coexistence) with a diverse range of local intermediaries and rival sources of authority to provide core functions of public security, justice, and conflict management in much of the country.' The mediated state proposes that to maintain control over spatial or political 'hinterlands' state institutions 'partner with, co-opt, or sub-contract to whatever local nonstate authorities they can find' (Menkhaus 2007, p. 78).
The mediated state perspective goes beyond the mere simultaneity of state and non-state governance systems to explore their relatedness and helps to show how governance interaction is often not only competitive, but also functional and pragmatic (Menkhaus 2006, p. 6). As such, the mediated state concept is the most explicit concept at hand to study instances of 'complicity and overlap between state and non-state forms of political power' (Raeymaekers et al. 2008, p. 16). 4 Scholars working with the idea of hybrid political order explicitly explore the relations, interactions and institutional overlap between state and non-state forms of governance. Drawing on previous work on 'pluralism' and 'multiple sovereignty' (Wickham-Crowley 1987, pp. 473-475), the idea of hybrid political order emphasises the multiplicity of, and stresses the 'interstices' and 'symbiosis' between, the various 'power poles' represented by state and nonstate authorities (Bierschenk and Olivier de Sardan 1997, p. 441, Kingston 2004, p. 7, Raeymaekers et al. 2008p. 8, Wiuff Moe 2011, p. 145, Risse 2013.

A Genealogy of PLO Governance in Lebanon
The governance practices of the PLO in Lebanon constitute a potent example of mediated stateness. When it was forced out of Jordan in 1969, the PLO was an unwelcome rebel group that moved into Lebanon at a moment when the Lebanese state was struggling with institutional and political challenges. Until the 1982 Israeli invasion expulsed the PLO from Lebanon, it was more powerful than its host. In terms of welfare institutions and avenues of political representation, too, the PLO often outdid the Lebanese state during this period. By the end of the Lebanese Civil War in 1990, conversely, the PLO lost much of its governance clout, while the Lebanese state engaged in assertive post-war statebuilding. Yet, throughout both phases, the governance practices of PLO and the Lebanese state were intertwined and interdependent. The relation between state and non-state governance actors in this case evolved over time, shifting from a dominant PLO to a stronger state, and both phases constitute different forms of mediated stateness. In this paper, these two distinct periods in the governance relation between the PLO and the Lebanese state are analysed as cases within a continuum of mediated stateness.
Lebanon is often regarded as a quintessential 'weak state' (Fregonese 2012, Kingston 2004, p. 5, Kingston and Zahar 2004, Migdal 2001. The country faces a legacy of colonial divide-and-rule, a brutal civil war and Israeli and Syrian occupation and its policy-making is almost perpetually gridlocked by its consociational political system (Long and Hanafi 2010, p. 676). This system institutionalised and reproduces the 'resilient strength' of Lebanon's sectarian communities that each have their regional strongholds, political parties, social institutions like schools, clinics and charity organisations as well as armed militias (Kingston andZahar 2004, p. 82, see also Faour 2007). Since its official installation in 1943, the Lebanese state has known various instances of collapse (Barak 2003, El-Khazen 2003, making the word 'Lebanonization' shorthand for the breakdown of state control (Migdal 2001, p. 136, Barak 2003. As a consequence, its sovereignty is often described as 'softening' (Ramadan 2008), 'virtual' (Picard 2012) or 'hybrid' (Bacik 2008, Fregonese 2012. Lebanon is also regularly depicted in terms of its 'states-within-the-state' (Kingston andZahar 2004, p. 81, Atzili 2010).
As such, the Lebanese state should not be reified as a single coherent institutional actor. While this article foregrounds the state's interactions with the Palestinian non-state rather than its internal logics, the amorphous nature of the Lebanese state has crucially shaped these engagements with the PLO. During the Civil War especially, the Lebanese state could hardly be considered more than the sum of the various Lebanese political-sectarian actors that vied for dominance both internally -considering the issue of who commanded the central state institutions -and externally -regarding the matter of who represented the country vis-à-vis other states, but also non-state institutions such as the PLO. Much of the clashes between the PLO and the Lebanese state were in fact confrontations between the revolutionary dynamics catalysed by the former and the interests of the conservative Lebanese elites -from all sects -that had captured the latter (Brynen 1990a, p. 162).
As Brynen (1990a, p. 163) puts it: 'The PLO could refrain from intervening in other inter-Lebanese political conflicts, but it could not prevent itself from becoming the focus of one' . Lebanon's Palestinian refugee community is thus intricately interlinked with characterisations of state fragility and hybridity (Czajka 2012, p. 238). Lebanon hosts some 400,000 Palestinians, constituting roughly 10 per cent of the country's population (before the arrival of the approximately 1.5 million refugees fleeing the Syrian war). 5 They are the remnants and descendants of the people who sought refuge in Lebanon when they were forcefully expelled from Mandatory Palestine during the 1948 Nakba that led to the creation of the state of Israel.
The PLO, established in 1964, has long functioned as the dominant 'institutional embodiment' of this Palestinian community (Sayigh 1997b, p. 20, Siklawi 2010. While a detailed analysis of the internal institutional dynamics of the PLO is beyond the scope of the current article, it is important to note from the outset that the PLO was (and is) an internally divided organisation that hosts various, often opposing, factions (Sayigh 1997b, p. 194). These different factions resulted in a largely instable and mostly centrifugal organisation which meant that the PLO routinely failed to operate as a cohesive or unitary actor. As we will see, for instance, the PLO's leading party, Fatah, perceived itself as the 'protector' of the Lebanese state and often pleaded a policy of restraint and non-intervention in its Lebanese operations, whereas many other factions did not agree with this approach and actively undermined it (Becker 1984, p. 100, Cobban 1984, p. 65, Hanf 1993. What is more, in many instances Lebanese-Palestinian interaction was shaped by individual initiatives stemming from competitive actions by rivalling Palestinian organisations rather than by any form of unanimous PLO policy. The PLO never amassed a 'universally accepted interpretation of Lebanon's political dynamics, ' which were interpreted as primarily sectarian by some factions and ultimately classed-based or geopolitical by others (Brynen 1990a, p. 162). Far more importantly, it was simply never able to control the activities of its individual organisations. Lack of internal 'cohesion, control and coordination' undermined effectiveness, created inconsistencies and eroded PLO credibility 'in the eyes of its opponents and allies alike' (Brynen 1990a, p. 171).
The PLO initially organised its resistance against Israel's occupation of Palestine from Jordan, but was ousted from the country in 'Black September' 1970. It then relocated to Lebanon which, from 1970 until 1982, served as 'the political and military centre of gravity of the Palestinian movement, ' providing a logistical base for military and civilian activities (Brynen 1989, p. 48, Hirst 2010. In 1982, the PLO was expelled from Lebanon by Israel. Officially, operation 'Peace for Galilee' had the objective to end Palestinian attacks on Israel from South Lebanon, but underlying that was the unstated yet evident aim to dismantle the PLO's governance project in Lebanon (Rubenberg 1983, p. 54, Brynen 1989. The expulsion meant that the vast majority of the PLO's armed forces as well as a large segment of its political leadership left Lebanon and relocated to Tunisia. 6 By the end of the Lebanese Civil War in 1990, the PLO had started to gradually rebuild some of its institutional structures and political presence in Lebanon, but it never recaptured the power position it held in the 1970s. My analysis of the governance interactions between the Lebanese state and the PLO focuses on the PLO's institutional heyday spanning the 1969-82 phase, also known as the organisation's 'Beirut era' (Cobban 1984, p. 124), and the post-1990 phase. 7

1969-82: Mediated Stateness Under Non-State Dominance
When the PLO landed in Lebanon in 1970, 8 the 'Palestinian Revolution' that the PLO represented was initially welcomed with enthusiasm by many Lebanese. Its pan-Arab, socialist and secular ideology combined with militant resistance against Israeli occupation appealed to Lebanese from different classes and sects (Cobban 1984, p. 47, Brynen 1989, p. 52, Beydoun 1992. Indeed, it was the widespread Lebanese popular support for the PLO and its cause that had 'effectively stymied Lebanese government efforts to suppress the fida'iyyin [guerrillas]' (Brynen 1990a, p. 161). A decade later, however, little was left of this solidarity. This was the result of both an effective Israeli divide-and-rule strategy -retaliation for Palestinian attacks launched from South Lebanon was systematically directed at Lebanese civilian targets so as to alienate the population from the PLO -as well as increasing misconduct by Palestinian revolutionaries (Becker 1984, Brynen 1989, see also Beydoun 1992, p. 36, Hirst 2010, p. 21, Siklawi 2010. Due to the weakness of the Lebanese state, the professionally institutionalised Palestinian public authority was perceived as fundamentally undermining its sovereignty (Sayigh 1997a, p. 668). The PLO, and eventually the broader Palestinian community in Lebanon, was considered a 'cuckoo' or a 'Trojan horse' that 'superseded' the state and held Lebanon hostage (Sayigh 1997a, p. 675, p. 49, p. 551, Hirst 2010. Notwithstanding this eventual disenchantment with the Palestinian Revolution, the PLO on the whole often worked 'in conjunction with rather than against the official Arab state system' (Cobban 1984, pp. 204, 205). Even at the time of the Lebanese state's disintegration, there was still an army, still a president, still an array of local officials and state-related power brokers that the PLO had to negotiate with. From 1975 onwards, the Lebanese state fragmented and disintegrated. It was headed by governments of various denominations over the years. Sometimes, the government consisted of anti-Palestinian forces (Siklawi 2010, p. 607). Other governments 'reinforced Lebanese relations with the Palestinian fida'iyin' (Siklawi 2010, p. 608). In general, the PLO's primary alliance was not so much with the state, but with one of the two alliances competing for state power throughout the Civil War: the Lebanese National Movement (Brynen 1989, p. 51, Hanf 1993, p. 187, Hirst 2010, p. 89, Siklawi 2010. At the same time, however, as outlined in this article, the PLO strategically maintained close ties with the official state, regardless of who was running it. At times, moreover, the ever-shifting allegiances of the Civil War period led to a convergence of the official state and the political parties and movements that the PLO allied with at particular junctions in time. In 1986, Lebanon's Christian leaders and the PLO opposed Syrian intervention together and President Gemayel renewed 70,000 passports for Palestinians in Lebanon. The Lebanese Forces also helped to 'transport armaments and fighters for the PLO into the camps in Beirut and the south' (Hanf 1993, p. 313).
As Brynen (1990b, p. 209) writes: Multiple centres of authority co-existed and even competed, while organization and discipline suffered. Reflecting both this and the crisis-laden atmosphere of the times, Lebanese decision-making was generally ad hoc and fragmented in nature. Those supposedly 'unified' decisions that were taken were usually little more than paper exercises.
Yet to simply conclude that the state collapsed is too convenient an analysis. Barak (2003, p. 319) importantly highlights the 'remarkable continuity manifested by [the state's] formal institutions during the conflict. ' Kingston and Zahar (2004, p. 90) also note that Lebanon's wartime 'cantonization' was in many ways 'illusive' and the state remained a key referential power. This scattered but continuously functioning Lebanese state was importantly connected with the governance practices of the PLO. These connections have both formal manifestations, as evidenced in the Cairo and Melkart Agreements, and practical expressions, as becomes clear in the everyday governance practices regarding security, welfare and representation of the PLO throughout this period. The Cairo Agreement, signed in 1969 by the leader of the PLO and the commander of the Lebanese army, formalised the parameters of the Palestinian armed activity in Lebanon by supporting the PLO's resistance against Israel; acknowledging the right to residency, employment, and movement of all Palestinians in Lebanon; and sanctioning the formation of local governance committees as well as the presence of weapons within the camps (Sayigh 1997a, p. 192, Meier 2010. As such, the Cairo Agreement, especially in hindsight, was branded overly 'indulgent' and a sign of PLO dominance over the Lebanese state (Czajka 2012, p. 241, see also Brynen 1989, p. 50). Clearly, there was a discrepancy between the de jure and de facto meanings of the Cairo Agreement. While the official agreement reconciled Lebanese sovereignty and Palestinian autonomy, such 'squaring of the circle' in reality proved impossible and the text was rarely honoured by either side (Hanf 1993, p. 166, see also Becker 1984, p. 101). What matters here, however, is that the Agreement represented not just about a doomed attempt to resolve mutually exclusive interests, but also contained unprecedented logics of coordination and alignment. The Cairo Agreement is not simply the formalisation of PLO supremacy in Lebanon. In fact, the text of the agreement explicitly stated that Lebanese authorities, 'both civil and military, shall continue to exercise all their prerogatives and responsibilities in all areas of Lebanon in all circumstances' (Czajka 2012, pp. 241, 242, see also Cobban 1984, p. 47). The Agreement's opening declaration reads: In consonance with the bonds of brotherhood and common destiny, relations between Lebanon and the Palestinian revolution must always be conducted on the bases of confidence, frankness, and positive cooperation for the benefit of Lebanon and the Palestinian revolution and within the framework of Lebanon's sovereignty and security. (Brynen 1990a, appendix) While largely utopian, the Cairo Agreement was nevertheless a key demonstration of engagement and interaction between the Lebanese state and the PLO; the first formal negotiation and establishment of their governance relations (Sayigh 1997a, p. 20, Suleiman 2006. For at least 15 years, it 'was to provide the formal basis for Palestinian-Lebanese relations' (Sayigh 1997b, p. 192). After its signing, a 'higher political committee for Palestinians, ' headed by a Fatah official, was set up to 'regulate daily relations with the government' (Sayigh 1997b, p. 192, see also Brynen 1990b, p. 208). The Palestinian local governance committees were likewise established with the aim to care for the interests of Palestinians residing in these camps 'in cooperation with the local Lebanese authorities within the framework of Lebanese sovereignty' (Brynen 1990a, appendix). In the words of a PLO leader at the time, the Cairo Agreement was 'the first official recognition of the Palestinian revolution […] a document whereby we could deal directly, officially with the Lebanese authorities' (Khalaf 1987cited in Brynen 1990b. Nor was the agreement a one-time occasion. In 1973, it was reinforced and updated through the 'Melkart Protocol' (Brynen 1989, p. 51, Siklawi 2010. And in 1976, at an Arab Summit meeting, the Lebanese president again 'formally agreed to reiterate the validity of the agreement reached in Cairo eight years previously' (Cobban 1984, p. 81).
The parameters of interaction between the Lebanese state and the PLO that were formulated in the Cairo Agreement indicate that in these interactions, the PLO had the upper hand during this period. The 1970s saw the PLO's institutional heyday in Lebanon (Brynen 1989, p. 58, Beker and Van Oordt 1991, p. 16, Ramadan 2008, p. 666, Atzili 2010. Analysts agree that the Palestinian 'political, military and organizational achievements' in Lebanon and its bureaucracy and taxation systems were remarkably extensive (Brynen 1989, p. 51, see also Rubenberg 1983, p. 78, Sayigh 1997b) -even if the PLO's governance regime in reality mostly did not reflect the idealised welfare state the PLO claimed it was (Becker 1984, pp. 141, 142). Hirst (2010, p. 87) notes that: With its quasi-governmental bureaucracies, welfare and medical organizations, social, cultural and educational institutions, research centres and the economic planning or industrial development boards of what was fast becoming the wealthiest resistance movement in the world, it engaged in functions that ranged far beyond the requirements of 'armed struggle. ' While it presented a stark contrast with the weak Lebanese state, this dominance did not result in unilateral governance on the side of the PLO, as is sometimes suggested (Khalil 2013, p. 2). 9 While PLO actions indeed at times corroded Lebanese state sovereignty, its often supposed 'virtual autonomy' is waylaid not only by the text of the Cairo Agreement, but also by the PLO's actual governance on the ground (Czajka 2012, p. 242). In many ways the PLO sought to align its governance project -its provision of security, welfare and representation -with that of the Lebanese state (and vice versa). The PLO's strategy in this regard, particularly evident in the early 1970s, depended on two major pillars: restraint and communication (Brynen 1990a, p. 162). It sought to regularise the behaviour of the guerrillas operating under its umbrella through its military police and invested in a continuous dialogue with its Lebanese political adversaries.
In terms of the provision of security, Brynen (1989, p. 52) estimates that the PLO's military forces encompassed over 10,000 'well-armed fighters in semi-regular formations' (the Palestine Liberation Army) that were backed up by thousands more militia. Remarkably, some of these cadres were Lebanese nationals (Hirst 2010, p. 87). In addition, the PLO's leading party, Fatah, commanded 'an increasing number of marginal but compliant Lebanese militias' (Brynen 1990b, p. 212, Hanf 1993. The PLO and the Lebanese army fought each other on many occasions and it is not the aim of this article to paints an overly benign picture of the relations between the PLO and the Lebanese state and its institutions (Siklawi 2010, pp. 604, 605). However, there were also instances of intense military cooperation between them and these hold significance for understanding PLO-Lebanese interactions as well. The Cairo Agreement, through which 'the Lebanese Government had transferred security responsibilities in the camps to the Palestinian Armed Struggle Command' (Siklawi 2010, p. 601), also stipulated coordination of all PLO military action against Israel with the Lebanese High Command (Beker andVan Oordt 1991, p. 12, Sayigh 1997b, p. 192). 10 In line with this, the Lebanese army assisted the PLO and, 'faced with the strength of popular sentiment' , offered training to Palestinian refugees (Sayigh 1997b, p. 188).
The PLO also assisted the Lebanese army. Fatah formed a special military police unit to 'assist the Lebanese army in supervising the entry of guerrillas and supplies from Syria' (Sayigh 1997b, p. 193), but also to restore local order, even breaking up intra-Lebanese neighbourhood conflicts (Hanf 1993, p. 329). At the request of Prime Minister Karami, the PLO helped monitor cease-fires in Tripoli and Beirut in 1975 (Sayigh 1997b, p. 369). Several official 'PLO-Lebanese liaison committees' conducted joint patrols in areas that had witnessed fighting (Sayigh 1997b, p. 362). In areas controlled by the PLO, the Lebanese police continued to function under PLO command. They were 'allotted limited duties and called the "social police"' (Becker 1984, p. 145).
The PLO's military might was closely tied to its second governance function: representation (Sayigh 1997b, pp. 27, 28). Internally, since its establishment in the early 1960s, the PLO became the vanguard of a renewed Palestinian self-awareness and offered a rallying point for pan-Arab solidarity, Palestinian national(ist) identity and the pursuit of political dignity and emancipation (Siklawi 2010, p. 600). Politically, the National Council, the Central Council, and the Executive Committee provided a legislative and executive infrastructure, for instance issuing identity cards (Rubenberg 1983, p. 55, Shiblak 1997. The Palestine National Fund managed the financial component of the PLO's governance operations (Rubenberg 1983, p. 59) and the Revolutionary Council maintained order among the organisation's different factions (Beker and Van Oordt 1991, p. 15).
Socially, the PLO created a range of trade unions and social institutions (Brynen 1990b, p. 216, Sayigh 1997b. Through its Department of Mass Organizations, the PLO encompassed 10 national unions (for workers, women, teachers, students, writers and journalists, lawyers, engineers, youth, artists, and doctors and pharmacists) that were represented in the National Council (Rubenberg 1983, p. 71). These unions were primarily concerned with the internal organisation of Lebanon's Palestinian society, but at times also engaged with institutions of the Lebanese state. The General Union of Palestinian workers, for instance, 'forged links with Lebanese trade unions' (Brynen 1990b, p. 216) and was successful in bargaining with Lebanese counterparts to arrange 'compensation for Palestinian dockyard workers unemployed as a consequence of the Lebanese civil war' (Rubenberg 1983, p. 72, Brynen 1990b. With marginal success, the Palestinian teachers' union lobbied the Lebanese Ministry of Education to 'end the myriad of restrictions imposed on Palestinians' in terms of education (Brynen 1990b, p. 215). 11 The PLO also operated an Information Bureau that had its own daily Arabic newspaper, a bi-weekly international journal that was published in French and English, and a news agency (Rubenberg 1983, p. 78, Brynen 1990b. While these outlets were primarily geared toward the PLO's Palestinian constituency, they were also circulated among Lebanese officials and civilians. Externally, the PLO was 'obsessively insistent' in 'obtaining from both Arab and non-Arab governments recognition of the PLO as the sole legitimate representative of the Palestinian people' (Sayigh 1997b, p. 21) -which was partly why the Cairo Agreement was such a defining moment (Brynen 1989, p. 50). Its Political Department, directed by 'the equivalent of a foreign minister, ' headed the PLO's diplomatic mission and represented the organisation at the United Nations and other international meetings (Rubenberg 1983, p. 78). In its quest for international recognition, the PLO considered the Lebanese state a prerequisite rather than an obstacle. When PLO leader Yaser Arafat addressed the United Nations General Assembly in 1974, he was 'introduced to the world' by the Lebanese President (Cobban 1984, p. 63). Fatah, the leading PLO party, routinely issued passports and visas in the name of the Lebanese state and used its official stamps (Becker 1984, p. 190). Becker (1984, p. 193)

described how
From the late 1960s onwards, there was an 'agreement' between the PLO and the Lebanese government, imposed by the PLO, by which foreigners could arrive at Beirut airport and be allowed into Lebanon without having to go through the usual immigration formalities. The daily list was handed to the passport and customs authorities.
According to Brynen (1989, p. 52), 'West Beirut had become the headquarters of the Palestinian leadership and the nerve centre of a vast international network of diplomatic and information offices and personnel' -a network that was also geared towards coordination with the Lebanese state. Cobban (1984, p. 77) describes how the PLO bureaucracy kept as many lines of communication open as possible to Lebanese state agencies as well as the politburos of Lebanese political parties. One of the chief Fatah leaders of the time captured his organisation's emphasis on interaction and diplomacy vis-à-vis the Lebanese state rather dramatically when he stated that 'our hope lies with dialogue and through dialogue, the only language that can pervade the whole of Lebanon' (Cobban 1984, p. 68).
It is in this light, also, that the PLO's ambition -or, more accurately, the ambition of the then dominant faction within the PLO -to be the 'guarantor' of the Lebanese state (rather than only its nemesis) should be seen (Sayigh 1997b, p. 323, see also Miller 1983, p. 37). The PLO in many ways needed a functioning, albeit ideally compliant, Lebanese state to maintain its sanctuary on Lebanese territory (Hanf 1993, p. 226). From the start, the PLO internally stressed the importance of interaction with local state representatives such as mukhtars 12 and mayors (Becker 1984, p. 181). When the Lebanese state grew weaker and weaker, ironically of course partly because of many of the PLO's factions' own conduct, the PLO continued to insist on 'dealing only with the Lebanese authorities' (Sayigh 1997b, p. 362). One senior PLO official tellingly explained that 'the Palestinian revolution adopts a defensive posture and does not try, under any circumstances, to move to the offensive, because we are committed to the sovereignty of Lebanon and to its security, stability, and national unity' (Sayigh 1997b, p. 365). 13 Even when Israel's 1982 invasion of Lebanon forced the PLO to withdraw its cadre from Lebanon, coordination with the Lebanese state remained a key aspect of the PLO's modus operandi (Hanf 1993, pp. 263, 264). In the negotiations over the terms of their evacuation, the PLO insisted on 'holding all negotiations for any evacuation from Beirut with the Lebanese president rather than with Sharon [the commander of the Israeli army]' (Cobban 1984, p. 122).
The PLO's representative institutions were closely related to an elaborate social service system that constituted the third pillar of the PLO's governance project in Lebanon. This 'civilian institutional infrastructure' of the PLO included hospitals, clinics, factories, cultural bureaus, art galleries, orphanages, schools and a research centre (Rubenberg 1983, p. 54, see also Brynen 1990b, pp. 216-219, Sayigh 1997b. The Palestine Red Crescent Society operated 11 hospitals and 60 clinics all over Lebanon that provided health care for a nominal fee (or even for free) (Brynen 1990b, p. 218, Rubenberg 1983). The Institution for Social Affairs and Welfare for the Families of Martyrs and Prisoners offered a range of social services to the families of people working for the PLO (Rubenberg 1983, p. 75). This institution also supervised the Popular Committees that provided public services such as electricity, water and sanitation inside the refugee camps (Rubenberg 1983, p. 76, Brynen 1989. The PLO's Palestine Martyrs Works Society offered vocational training (Rubenberg 1983, p. 66, Brynen 1990b). In addition, the PLO's Education Department also ran kindergartens, offered scholarships for universities, and organised adult literacy campaigns (Rubenberg 1983, p. 77). Finally, through its Department of Information and Culture the PLO facilitated activities in the realm of theatre, cinema, dance, photography and popular and fine arts (Rubenberg 1983, pp. 69, 70).
This civilian infrastructure of the PLO operated parallel to that of the Lebanese state. Yet, there are ample indications that these state and non-state service systems were aligned. The fact that 'the PLO generally left responsibility for civilian and economic affairs to the Lebanese government agencies in the areas it held or shared' suggests that there must have been routine coordination between local state officials and PLO representatives about these matters (Sayigh 1997b, p. 449) -especially considering that much of the welfare services that the PLO provided were also accessible for Lebanese civilians (Rubenberg 1983, p. 62, Brynen 1989, p. 52, Weighhill 1997. When the Lebanese state further disintegrated, the PLO and the Joint Forces increasingly took over state services for Lebanese citizens (Brynen 1989, p. 52). During Israel's 1982 blockade of Beirut, in a joint operation 'PLO and Lebanese government agencies countered by ensuring that local hospitals and communications centres received regular supplies of fuel; their teams also drilled new artesian wells, and toured residential neighbourhoods with mobile generators to pump water for household use' (Sayigh 1997b, p. 535).
Even if Lebanese coercive state power was regularly deployed against the PLO during the above-described period, from 1969 to 1982 the PLO's institutional and military might as well as the potency of its pan-Arab revolution overall stood in stark contrast with the weakness and ideological bankruptcy of the Lebanese state. An exploration of the formal and practical dimensions of PLO governance during this era, however, reveals that this dominance cannot be explained through a 'simple dyad of refugee power versus state power' (Brynen 1990b, p. 209). Rather than pure PLO strength, it was 'PLO-Lebanese relations [that] became the major determinants of Palestinian activity in Lebanon for more than a decade' (Brynen 1990b, p. 209). The involvement of the PLO in the governance realms of security, welfare and representation was a specific form of mediated stateness because even if the PLO had the upper hand over the Lebanese state and even if its engagement with the Lebanese state was by no means always benevolent, it did not seek to defeat or destroy it, but rather aimed to manage and use it. In fact, PLO governance at times propped up the faltering Lebanese state. This also meant that the Lebanese state continued to shape Palestinian governance as well (for instance by demanding information on the PLO's military operations and by manipulating PLO diplomacy).
This should serve as a reminder that the PLO was a 'state-in-exile' rather than a 'state-within-the-state' (Sayigh 1997b, p. 448). The PLO's intended state, after all, did not lie in Lebanon. Its governance programme in Lebanon was a means towards the end of establishing a 'real' state in Palestine, not an end in itself. For Hirst (2010, p. 107), the Palestinian governance project in Lebanon was a platform to 'win the world's recognition and then to hold its own, come what may, until the world accorded them their permanent one: their Palestinian state in Palestine itself' (see also Brynen 1989, p. 59). This external focus explains much of the rationale behind the PLO's governance. Although certainly not constituting unitary dominance, the PLO's influence over the Lebanese state while co-operating with it in delivering governance as such provides a case of mediated stateness near one end of a continuum based on the relative power of the two parties.

and After: Mediated Stateness Under State Dominance
When the PLO leadership was expelled from Lebanon, the institutional infrastructure of the PLO in Lebanon began to crumble and its welfare structure slumped. 'The Israeli invasion of Lebanon and the evacuation of the PLO from Beirut in summer 1982 effectively ended the Palestinian armed struggle and the process of state building' there (Sayigh 1997b, p. 28, see also Beker and Van Oordt 1991, p. 1). While remaining weak, with the end of the Lebanese Civil War, the Lebanese state gradually started to recapture its sovereignty and revive its governance functions. In contrast to the situation of PLO dominance described above, this meant that reinstating PLO governance in Lebanon was now subject to state control. In the governance relations between the Lebanese state and the PLO, in other words, the power positions by and large were reversed. When, in 1991, after being surrounded by the Lebanese army in his southern stronghold, Arafat tried 'to make a grand exit by "donating" all arms to the Lebanese army' he soon understood that he was living in the past (Hanf 1993, p. 621).
After Israel had defeated the PLO militarily, 'the full legacy -and cost -of earlier PLO policy in Lebanon made itself felt' and long-time foes and disillusioned former allies made sure to prevent a resurgence of a strong PLO in Lebanon (Brynen 1989, p. 53, Czajka 2012. 14 This development was bolstered by the abrogation of the Cairo Agreement in 1987, which the Lebanese Chamber of Deputies declared 'null and void as if it had never existed' (Brynen 1990a, appendix; see also Siklawi 2010, p. 610). Tellingly, when, in 1991, the PLO demanded a new agreement with the Lebanese Government on 'Palestinian status' in Lebanon, 'none of the representative political groupings showed any interest' (Hanf 1993, pp. 619, 620). The ambition to keep the PLO down was also enshrined in the Ta'if Accord that formally ended the Lebanese Civil War in 1990 (Mallat 1990). The Ta'if Accord, drafted not merely as a peace agreement but as an instrument in the 'struggle to regain sovereignty' (Saseen 1990, p. 57) that was directed against the PLO, explicitly refuses any form of 'permanent settlement' or 'implantation' of non-nationals -a reference commentators unanimously consider to refer to the country's Palestinian community (Meier 2010, p. 148).
However, even after the Lebanese Civil War, the Palestinian presence in Lebanon has remained intricately linked with the country's struggle for sovereignty (Hanafi 2011, p. 35, Picard 2012. The Palestinians constituted a convenient 'Other' against which the heterogeneous Lebanese in search for unity could identify (Hanf 1993, p. 621, Haddad 2004, Sfeir 2010. This was vividly illustrated once more by the 2007 Nahr el-Bared crisis during which the Lebanese army destroyed large parts of the Nahr el-Bared camp in North Lebanon to eliminate militants hiding there. This operation was steeped in a discourse that stipulated that 'Lebanese sovereignty over these security islands had to be established' (Czajka 2012, p. 250). When the army entered the camp, to give a poignant example, it was seen to erect Lebanese flags over some of the rare buildings still standing in the shelled refugee camp (Czajka 2012, p. 250).
Indeed, for the Lebanese state 'the function of the Palestinians today is to be held up as the party whose behaviour can explain what happened to the Lebanese' during the Civil War (Beydoun 1992, p. 52). In practice, this meant that Lebanese policy has not merely treated the post-war Palestinian refugee community in its country with political stringency, but has implemented a comprehensive procedure of socio-economic, legal and spatial strangulation (Sayigh 1995, Allan 2014. Palestinians in Lebanon have been withheld civil rights and are severely discriminated against on the labour market. Since 2001, moreover, they are prohibited from registering real estate (Haddad 2004, p. 479, Knudsen 2007).
Yet, despite defeat, discrimination and strangulation, Lebanon's Palestinians are still there. And they still govern and are governed. The PLO's governance project in Lebanon has become far more modest since 1982, but it has not disappeared. In particular, it has changed in three important ways. First, rather than operating its institutional structures for representation, security and welfare on a national level, in territories that extend beyond the official refugee camps, the PLO now limits its governance provision to these camps. Second, whereas during its heyday the PLO provided services and gave political voice to Lebanese communities as well, it now solely focuses on its Palestinian constituency. Third, these developments have shifted the nucleus of governance from the national to the local level -from the political and military leaders directing the revolution and its institutional base from within their headquarters in Beirut's 'Fatahland' (Beydoun 1992, p. 36, Hirst 2010 to the Popular Committees operating as the 'municipalities' of the camps (Kortam 2011, p. 203, see also El Ali 2011, p. 28).
Popular Committees were installed through the Cairo Agreement as the PLO's instrument to organise governance inside the camps. They fall under the Lebanese office of the PLO's Department of Refugee Affairs that oversees a Central Follow-Up Committee on the national level, five regional Popular Committee offices and a Popular Committee in each camp. As Knudsen and Hanafi (2011, p. 9) pointedly surmise, 'despite their appealing name, "Popular Committees" neither represent popular vote nor popular sentiments but are vested power bases of non-local political patrons. ' Notwithstanding their lack of capacity, resources and popular support, however, Popular Committees can still be found in every Palestinian camp. In 70 per cent of the camps, Popular Committees are the 'major co-ordinating bodies within the communities' (Jacobsen and Khalidi 2003, p. 185).
Today's Popular Committees cannot stand in the shadow of the extensive governance project that the PLO ran before. They also do not govern the camps unilaterally considering that the United Nations Relief and Works Agency for Palestine Refugees in the Near East (UNRWA) provides the bulk of education and health care and various NGOs and Palestinian religious institutions and political parties take on a large part of governance in the realm of representation and welfare. Yet, the Popular Committees, due to their affiliation with the PLO, are broadly recognised as representing camp communities -in fact, the significance of the Popular Committees has only increased in comparison to the 1970s as it is now one of the few, rather than one of the many, PLO institutions in Lebanon.
In terms of representation, the PLO bylaw reiterates that the Popular Committees are the 'official representative of the camps vis-à-vis foreign, national and all other organizations' . The committees keep a basic administration of the population of the camps that helps to make them the 'key interlocutor for everyone' (Martin 2011, p. 185) and play a crucial gatekeeper role between these communities and other actors, such as the Lebanese state, UNRWA and NGOs (Stel 2016). Residents and aid organisations stress that anyone who enters the camps 'has to pass through them [the Popular Committee]. ' 15 The Popular Committees' role in representation is further illustrated by the monthly magazine it publishes and its role in hosting festive occasions, rallies and demonstrations in the camps. The importance of coordination with the Lebanese state is also evident in the vision of the Central Committee for the Popular Committees in Lebanon. It commands its regional offices to 'work towards the activation and improvement of the relations with the neighbourhood, especially with the municipalities in the cities and the surrounding Lebanese villages' 16 This suggestion is clearly taken to heart by Popular Committees that have established regional 'follow-up committees, ' consisting of committee representatives and political leaders. These 'serve as an intermediary between the Palestinian factions controlling the camp and the Lebanese state' (Czajka 2012, p. 248, see also Chabaan 2014, Stel 2017. 17 Concerning security, the Popular Committees play a rudimentary role in the administration of justice and closely coordinate with Security Committees that are made up of armed fighters belonging to the different factions of the PLO (El Ali 2005, p. 86). The Security Committees maintain order in the camps and their representatives coordinate with local Lebanese political leaders to prevent conflict between Palestinian and Lebanese communities (Long and Hanafi 2010, p. 685). They are expected to 'keep the peace' and prevent 'intra-factional fighting' (Long and Hanafi 2010, p. 685). Local conflicts between Palestinian and Lebanese residents are often solved through the mediation of Lebanese local authorities (such as mukhtars) and politicians alerted by the Popular Committees (Stel 2017, p. 54). In a report aptly titled 'How the PLO and Lebanese authorities became unlikely friends, ' Lambert (2016) describes how security coordination between the PLO and the Lebanese state grows ever more intense. The PLO's Security Committees have received training and support from Lebanese security forces in their attempt to maintain order in Lebanon's most notorious refugee camp (Ain al Hilweh in Sidon). In this context, there have also been high-level meetings between a 'Higher Palestinian Security Committee' and a former Lebanese Prime Minister to support the unremitting efforts of the Palestinian factions and forces and their coordination together and with the Lebanese government, in order to establish an atmosphere of security and tranquility in the camp and prevent its use against its inhabitants and residents, and against Lebanon's security and stability (Lebanese Palestinian Dialogue Committee (LPDC) Newsroom 2016).
With regard to welfare, the Popular Committees operate the camps' communal water wells, coordinate with the public and private providers of electricity and supervise the collection and management of waste (Hilal 2010, p. 32, Chabaan 2014, p. 34, Yassin et al. 2016. Here, 'under the table' coordination with local Lebanese authorities is widespread (Stel 2017, p. 191). Kortam (2011, pp. 202, 203) for instance describes how the Committee in Shatila camp in Beirut regularly met with the Beirut municipality, the Lebanese Water Company and Électricité du Liban. In her study on the informal 'adjacent areas' to Lebanon's refugee camps, Hilal (2010, p. 54) similarly concluded that interactions with municipalities to smooth the provision of basic urban services exists even though they are characterised by the absence of 'formal mechanisms for coordination or intervention. ' My own work extensively establishes that coordination between Popular Committees and local Lebanese authorities such as municipalities, mukhtars and public utility companies is commonplace in local endeavours to maintain infrastructure, access electricity and organise waste management (although it mostly remains under the radar due to its informal, indirect, irregular and asymmetrical nature) (Stel 2015(Stel , 2016. As one Popular Committee member rhetorically asked me: 'How would you solve issues if you don't sit together?' 18 Popular Committees thus coordinate (in different ways and to different degrees) with representatives of the Lebanese state (El Ali 2011). This is evident in the on-the-ground dynamics summarised above, but also in the more formal dimensions of governance. Official diplomatic relations between the PLO and the Lebanese state have steadily intensified after the war. 19 The Nahr El-Bared crisis that emerged in 2007 has been a watershed moment in this development. While the Lebanese army caused what many observers have considered excessive human suffering and displacement, the PLO firmly backed the Lebanese army. After the initial destruction, there was, moreover, extensive coordinated action of the PLO and the Lebanese state regarding relocation, humanitarian assistance, and relief payments to Palestinian victims were jointly issued by the PLO and the Lebanese state. This 'unconditional' support for the Lebanese Government during this crisis allowed the recently reinstalled 20 PLO representative to Lebanon 'to enhance ties with the Lebanese ruling alliance' (Knudsen 2011, pp. 105, 106). As Ramadan (2009, p. 161) explains, in today's current geopolitical setting, 'if the Lebanese government is able to construct a new model of demilitarised camp governance, Fateh/PLO will want to be the Palestinian partner responsible for those camps. ' Mansour and Yassin (2010, p. 28) corroborate this and conclude that throughout the crisis the PLO 'served to extend the government's control. ' In 2011, the Palestinian Embassy in Lebanon was finally re-opened as well, initiating additional avenues for diplomatic engagement (Czajka 2012, p. 244). Since then, both the Lebanese state and the PLO have regularly reiterated the value of cooperation and coordination. The PLO stated the need to create 'new relations with our Lebanese brothers' and Lebanese state discourses 'seem to tend toward a renewed, and more urgent, interest in the establishment of a more conciliatory relationship between Palestinian refugees and the Lebanese state' (Czajka 2012, pp. 250, 251).
This new era in the history of the PLO's presence in Lebanon (Suleiman 2006, Hilal 2010) has its institutional embodiment in the Lebanese Palestinian Dialogue Committee (LPDC). The LPDC was created in October 2005 as an inter-ministerial committee to address the situation of Palestinian refugees in Lebanon (Hilal 2010, p. 35). According to its website, it seeks to 'strengthen the interaction between the Lebanese government and the various Palestinian community representatives […] promoting channels of communication, enhancing relations and connecting Palestinian camps and communities with their neighbouring municipalities. ' 21 The LPDC has met with widespread criticism (Knudsen 2011, p. 102). As a contested diplomatic vehicle, it ultimately 'failed to make any significant change in the lives of Palestinians in Lebanon' (Suleiman 2010, pp. 15, 16), also because it has no representation on the local level and does not engage with refugee populations or their direct representatives. Its politicisation -the LPDC gained much of its relevance in the wake of the Nahr el-Bared crisis that was described above -and exclusion of important oppositional forces gained it the nickname of 'Lebanese Palestinian Monologue Committee. ' 22 As the first ever officially institutionalised dialogue structure between Lebanese and Palestinian leaders, however, the LPDC nevertheless remains a symbolic and institutional 'turning point in the historically troubled relationship' between the PLO and the Lebanese state (Czajka 2012, p. 251).
What is significant about the governance activities by the Palestinian Popular Committees working in the camps and the simultaneous revival of diplomatic ties between the PLO and the Lebanese state is that, despite the apprehensions of many Lebanese, they do not challenge the Lebanese state. Post-war PLO governance in Lebanon indeed constitutes a form of mediated stateness that in fact at times bolsters the Lebanese state. 23 What has changed, then, is not the occurrence of interaction between the Lebanese state and the PLO, but rather the relative power positions in this interaction and hence the consequences for Lebanese political authority. Coordination in the realms of security, representation and welfare between local-level Lebanese and Palestinian authorities continues, but now occurs mostly on the terms of the Lebanese state. Following the logic of mediated stateness, the Lebanese state needs the PLO and its Popular Committees as an interlocutor through which to meet the core postwar objective of 'reestablishment of state control over the refugee community' (Sayigh 1995, p. 42). Considering that Lebanon's post-war governments have consistently faced a severe lack of resources and capacities to actually govern, the PLO and its Popular Committees, are a crucial element of the Lebanese state's attempts to control or contain nonstate spaces.
My respondents indicated that in Palestinian camps 'the reality on the ground, the current situation, is a consequence of mutual interests. ' 24 This aligns with other accounts of 'the tacit complicity between institutional stakeholders on the Palestinian political scene and the Lebanese government in maintaining the status quo' (Allan 2014, p. 203). Indeed, a local PLO official went as far as claiming that 'we help the Lebanese government to control. ' 25 In practice, most state institutions work on this premise as well. Municipalities, mukhtars and utility companies routinely work with Popular Committees. Even the former president of the LPDC found it self-evident that on 'construction, infrastructure, electricity, water, sewage […] we call them [the relevant Popular Committee] directly. ' 26 In the post-Civil War era, then, the governance of the PLO in Lebanon is still determined by a logic of mediated stateness. Its local provision of security, welfare and representation is closely aligned with and shaped by Lebanese local governance dynamics. Formally, although the Cairo Agreement was abrogated, governance interaction was reinstated in the mission of the LPDC. Replacement of the Cairo Agreement, which many saw as enforced by the PLO, with the LPDC, a voluntary arrangement initiated by the Lebanese Government, illustrates the important shift in power in the relation between the two governance actors. After 1982, the PLO in Lebanon was decimated. It exchanged its revolutionary ideology for a more pragmatic survival mode. Conversely, the Lebanese state, despite its ongoing 'fragility' , rebuilt its army and institutions and adopted a statebuilding creed. The PLO, in this new context, does not distance itself from its governance functions in Lebanon's refugee camps, but routinely assures Lebanese that Palestinians 'are guests in Lebanon and would not interfere in domestic affairs' (Czajka 2012, p. 249). The first post-war PLO representative to Lebanon has regularly reiterated that the PLO 'supports Lebanon's sovereignty and endorses any decisions Lebanese authorities made' (Czajka 2012, p. 249).
With such alternated power positions, the parties' respective interests changed, but their mutual stake in governance interaction remained. As outlined above, the PLO needs the Lebanese state to enable and facilitate its governance activities on the ground and the Lebanese state requires the PLO as a partner to control the country's infamous refugee camps. Since it is now mostly state institutions that dictate the terms of interaction, however, the Palestinian governance project is no longer a threat to Lebanese sovereignty, but in subtle ways actually serves to underpin it. The authority of the Lebanese state and that of the Popular Committees are intertwined rather than in contention. In comparison with the pre-1982 phase, their interactions after 1990 provide a case of mediated stateness near the other end of a continuum based on the relative power of the two parties.

Conceptual Implications
The spatial, demographic and hierarchical dimensions of the PLO's provision of security, welfare and representation in Lebanon have changed significantly over time. The PLO was present throughout the country and even hegemonic in vast regions in the 1970s, but limited its governance to the refugee camps after 1982. It served Lebanese and Palestinians alike in its heyday and focused solely on Palestinian refugees after its institutional demise. The PLO organised predominantly on a national level from 1969 to 1982 and re-organised on a local/ camp level during its re-emergence after 1982. It is partly the PLO's changing relation with the Lebanese state that has determined these shifts in governance.
During its historical trajectory, PLO governance in Lebanon first dominated the Lebanese state and then was dominated by it. Yet the protracted idea, as once voiced by former Lebanese President Fuad Shihab, that 'Lebanon will either repress the Palestinians or be repressed by them' (Beker and Van Oordt 1991, p. 11, see also Haddad 2004, Czajka 2012, is far too dichotomous to be helpful in understanding how the PLO's sophisticated governance programme emerged and evolved.
In the 1969-82 period, the PLO trumped the Lebanese state, already fragmented by partisan strife, both in terms of resources, with its superior military might and institutional clout. It built on its ideological appeal as a revolutionary liberation movement. Conversely, in the post-1990 period, especially since 2005, the Lebanese state drew on a more legitimate repertoire revolving around post-war sovereignty and statebuilding. It's still limited resources nevertheless outweighed those of the defeated PLO that had to prioritise refugee survival over liberation and revolution. Despite, or perhaps because of, these relative power shifts, however, the governance of each actor significantly shaped that of the other.
State and non-state forms of governance here are thus importantly interdependent. In the earlier period, the PLO may have been the dominant actor, but it still needed the Lebanese state as a sanctuary and, at times, as a diplomatic extension of its institutional structure. As such, and despite simultaneous conflicts and competition, the PLO at times propped up or 'guaranteed' the Lebanese state. Unlike Menkhaus' original mediated state idea, then, the mediated governance of the PLO shows that mediation goes both ways: states manipulate non-state governance actors, but non-state actors also use states to govern.
In the later period, the Lebanese state often overrode the PLO in its governance. But rather than actually doing away with the PLO, the Lebanese state at times partly governs through the PLO. Practically, it uses the PLO to control its 'extra-state' spaces and populations. Symbolically, it casts the Palestinian community in Lebanon as a convenient 'Other' in the post-war nationbuilding process. This 'dialectical relationship' between Lebanon's state and non-state governance actors thus 'pushes them toward coexistence rather than dissolution' (Kingston andZahar 2004, p. 81, see also Hanf 1993). This means that while the interests of state and non-state actors over a historical governance continuum may often be fundamentally different, they can be compatible enough to produce different forms of mediated stateness in which governance actors not only govern against each other but also with and through each other. It is thus too simplistic to assume 'a zero-sum situation in which any prerogative or control secured by [the non-state] is interpreted as an encroachment on the power of [the state]' (Weighhill 1997, p. 304; see also Beker and Van Oordt 1991, p. 1).
The PLO has sometimes qualified as a rebel ruler while it did not function as one at other moments. The Lebanese state might have displayed features of state failure at some moments and not at others. Mediated stateness, however, has been a constant. In this article, therefore, I explored the changing governance relations between the PLO and the Lebanese state along a continuum of mediated stateness in order to better understand the changing manifestations of PLO governance in Lebanon. This in turn enables a further operationalisation of the intuitively relevant but conceptually rather unrefined notion of the mediated state (Stel 2015). Menkhaus did not structurally work out the typology of the mediated state he put forward in a conference paper in 2006 or in his empirical articles on Kenya (2008) and Somalia (2007). Moreover, the mediated state is built exclusively on African cases. Apart from offering empirical material to substantiate the relevance of the mediated state idea, then, this article allows for a more elaborate conceptual operationalisation.
Building on the empirical insights presented above, I suggest redefining mediated stateness as a form of governance interaction whereby state and non-state governance actors are (partly) dependent on each other and therefore (partly) shape each other's governance projects. The manifestation of such mediated stateness, I propose based on the case-study presented here, is likely through an (explicit) acknowledgement of each other's existence and relevance in the form of formal agreements, that can also outline a de facto division of governance roles, and/or a (more implicit) alignment of daily governance practices in the realms of security, welfare and representation that is apparent in informal arrangements. Studying the PLO's governance project in Lebanon suggests that the form and substance of mediated stateness can then be gauged by looking at the spatial, demographic and hierarchical components of governance; i.e. to what extent governance practices are territorially bounded, limited to a specific constituency and consigned to a specific level (local, national, transnational) of governance. Borrowing from the broader rebel governance literature (Hagmann and Péclard 2010), changes in these components of mediated governance, as shown above, can be explored through mapping the fluctuations in the resources, mostly armed capacity and material assets, and repertoires, or mobilising ideologies, of each governance actor.
This more explicit operationalisation makes three contributions to Menkhaus' original mediated state idea. First, whereas Menkhaus appears to assume that state and non-state governance actors share an abstract interest in maintaining order, my operationalisation acknowledges that the interests of state and nonstate governance actors will often be fundamentally different but may nevertheless sometimes be compatible. Second, my elaboration moves beyond the rather static view put forward by Menkhaus and can help capture variation and developments in governance interaction modes. Third, while mediation is often associated with equality or symmetry, the case study here reminds us that mutual relations do not have to be benign to be productive. Shifts in the power relations between state and non-state governance actors help explain the specific forms that mediated stateness takes. Not only do states at times govern through non-states, but non-state governance actors sometimes govern through states as well.
This new approach to the mediated state, however, requires careful differentiation between non-state governance actors that rebel against the state that hosts them and those, like the PLO, that use their host state as a 'sanctuary' to wage a liberation struggle vis-à-vis a third state (Brynen 1990a). As demonstrated in this article, the default connection between non-state governance and violent conflict is too simplistic. Different situations before, during and after war are equally important in understanding the manifestations and implications of governance beyond the state. Governance is not always a stake in violent conflict; it is often an aspect of post-war ordering, with all the contention, but also deal-making, that entails (Hoffmann and Kirk 2013, p. 20, see also Hagmann and Péclard 2010, pp. 555, 556). Where non-state actors have established themselves as public authorities that are not primarily interested in seceding from, toppling, or taking over the central state apparatus so much as governing a specific area or constituency, governance can cease to be a zero-sum endeavour. Interaction between state and non-state governance actors then becomes more probable and more important to investigate.
Both conceptually and empirically, then, much can be learned about the power, institutions and intentions of respective state and non-state governance actors by investigating their mutual relations. The changing relations of the PLO and the Lebanese state illustrate how the significance of governance beyond the state lays not so much in its occurrence as in the strategic functions it performs for both types of governance actors. If we are to understand why state and nonstate governance actors collide, collude or compromise we should investigate not only how they clash and compete but also where and in what ways they institutionally converge or coordinate. Notes 1. During twelve months of fieldwork, I conducted 232 interviews and five focus groups with Lebanese and Palestinian politicians, policymakers and local authorities as well as residents, community leaders, non-governmental organisations (NGOs) and women and youth committees (Stel 2017). I also generated elaborate observational data and collected documentary sources. Fieldwork was conducted in two informal refugee camps, or 'gatherings, ' in South Lebanon. Gatherings differ from official refugee camps in that they are illicitly built on Lebanese lands and as such have less political and institutional autonomy. Interaction between Palestinian and Lebanese governance actors is therefore more likely to occur in the context of these gatherings. The dynamics of governance interaction and mediated stateness described for the gatherings, however, could often be corroborated for official camps as well (Stel 2017, pp. 209, 210). 2. Labelling governance actors 'state' and 'non-state' begs the question of where the boundaries between these categories can be drawn. Exploring the interaction, overlap and even symbiosis between 'state' and 'non-state' governance actors requires a constructivist conceptualisation of the state that follows Abrams (1988) vision on the state as simultaneously a system and an idea; both a material phenomenon and a social imaginary (Hagmann and Péclard 2010, p. 543). In a narrow sense, the state as a system then consists of various state agencies that can be considered governance actors (Klem 2012, p. 30, Stel andNdayiragije 2014). In a more comprehensive sense, the state as an idea (on how to legitimately govern) can be referred to and appropriated by any governance actor -whether it is part of the state system or not (Boege et al. 2009, p. 92, Hagmann and Péclard 2010, pp. 540, 541, Stel 2016. Distinguishing between the state system and the state idea helps to avoid reproducing the misguided state/society or state/non-state dichotomy that may be analytically convenient but is empirically untenable. 3. Several scholars working within the rebel rule paradigm, such as Klem (2012), Mampilly (2011) and Terpstra and Frerks (forthcoming) point out the significance of rebels' co-optation of pre-war state institutions and devote attention to the institutional overlap that can occur between state and rebel governance modalities. Overall, however, within the rebel rule perspective there is relatively little discussion about the relations between state and non-state governance actors -and where this is discussed, the emphasis tends to be on competition and separation. 4. This section is based on Stel (2015). The idea of mediated stateness is indebted to Migdal's (1988, p. 144) seminal thesis on the dialectic between 'weak states' and 'strong societies' where 'strongmen' deliver the social stability and mobilisation that 'statesmen' need. It also resonates closely with Scott's interpretation of indirect rule (Mamdani 1996). Scott (2009, p. 209) notes that in the face of significant non-citizen communities, such as refugees, states 'require a stable, reliable, hierarchical, "graspable" social structure through which to negotiate or rule. ' In attempting to control or contain 'extrastate spaces, ' states need an interlocutor 'through whom instructions can be conveyed, ' who 'can be held responsible for political order' and who, if need be, 'can be held hostage' (Scott 2009, p. 31, 209). 5. Numbers are highly contested (Czajka 2012, pp. 239, 240). More than 100,000 Palestinian refugees fled to Lebanon in 1948 (Brynen 1989, p. 205). In 1950, there were 127,600 Palestinians registered in Lebanon (Suleiman 2006, p. 6). Hirst (2010, p. 87) estimates the Palestinian community in Lebanon counted some 340,000 people in the 1970s. In 2005, 404,170 Palestinians were registered with UNRWA (but it is widely assumed that not all of these registered people actually reside in Lebanon) (Suleiman 2006, p. 6). 6. Its departure left the Palestinian civilian population unprotected from the Lebanese Phalangist militias that, under the auspices and with the active support of the Israeli army, in 1982 massacred over three thousand Palestinian civilians in the Sabra and Shatila refugee camps in Beirut (Czajka 2012, pp. 242, 243).
7. The 1982-90 period that separates these two phases is characterised by a lack of governance interaction between the PLO and the Lebanese state as the PLO was all but absent from the country. As such, it is considered beyond the scope of the current argument. 8. The PLO already had had an office in Beirut since 1964, but Lebanon only became its headquarters after the organisation's expulsion from Jordan in 1970 (Brynen 1990b, p. 207). 9. Lebanon's wartime 'cantonization' meant that the PLO was not the only non-state governance actor (Kingston and Zahar 2004, p. 81). Hirst (2010, p. 207) describes how there were at least 10 'mini-states, or "cantons", ' that were 'all usurping taxcollecting and other functions of the state, all supposedly defending their own "subjects" against the others, and all demanding tribute and protection money in return for that service' (see also Cobban 1984, p. 99). 10. The Agreement orders 'The establishment of posts of the Palestinian Armed Struggle [PASC] inside the camps for the purpose of cooperation with the local committees to ensure good relations with the Lebanese authorities. These posts shall undertake the task of regulating and determining the presence of arms in the camps within the framework of Lebanese security and the interests of the Palestinian revolution. ' It also refers to 'Establishing a joint command control of the Armed Struggle and the Lebanese Army;' to 'Appointing Armed Struggle representatives at Lebanese Army headquarters to participate in the resolution of all emergency matters;' and stipulated that 'The Lebanese Army shall facilitate the operation of medical, evacuation, and supply centers for commando activity' (Brynen 1990a, appendix). 11. And the General Union of Palestinian Students included students that were affiliated with various Lebanese parties so as to compete in elections for student representation at Lebanese universities (Brynen 1990b, p. 217). 12. Mukhtars are state representatives that perform social and administrative services on the neighbourhood or village level. 13. While these words may have often belied the actual conduct of some PLO factions, they signal the significance of exploring the institutional relations between the Lebanese state and the Palestinian non-state. 14. The PLO also faced more intra-Palestinian challenges and lost its hegemonic position within the Palestinian polity (Brynen 1990b, p. 220 gives the PLO an edge over its intra-Palestinian competitors as well (Knudsen 2011, p. 107). 18. Author's interview with Popular Committee member, Tyre, 13 June 2013. 19. And especially since 2005 saw the end of the de facto Syrian occupation of Lebanon, which determined most of Lebanon's politics in its first post-war era. Syria had also tightly controlled the PLO in Lebanon until that time (Hanf 1993, p. 295). 20. The PLO representative office in Lebanon, which had been closed since 1982, was only re-opened in 2006 (Knudsen 2011, p. 105