Being ANTish in Aotearoa New Zealand: leaders assembling net-work

ABSTRACT This article takes up an ANTian sensibility to explore the enactment of a policy for educational collaboration in one region in the South Island of Aotearoa New Zealand (New Zealand). The case offers potential for considering the benefits of a sociology of associations (Latour 2005/2007): a Treaty-based bicultural nation, school atomisation consequential to a decades-long ‘system’ of self-managing schools, and geological actors in the form of damaging earthquakes in 2010 and 2011. The article considers the introduction of voluntary Kāhui Ako | Communities of Learning as a policy initiative intended to address achievement and equity concerns by providing support for collaboration. While the policy as articulated focuses on the aspirations and abilities of human actors in leadership roles, I take up ideas around actants, symmetry, alliance and translation to foreground other actors – both present and long absent – involved in myriad processes of policy enactment.


Introduction
From 1984 to the early 1990s a radical reform of mechanisms of government was undertaken by New Zealand's fourth Labour government and continued by successive National governments (Kelsey 1997).The reforms were suggested to be 'closest in the world to the application of "economic rationalism"' (Easton 1994, 78), an Australasian term related to global policy trends which emphasised economic competition as a public good and government intervention to correct any market failure as both inappropriate and inefficient (Stokes 2014).Within this context, wide-ranging education policy initiatives were implemented, including Tomorrow's Schools (Lange 1988).Under that policy, schools became fully self-managing, run by parent-elected Boards of Trustees.All departmental administration at the regional level was removed and the ten Education Boards abolished.Court and O'Neill (2011) have argued that while Treasury's arguments for market deregulation and elevation of parental choice in this process of decentralisation did not gain widespread support, the managerialist strategies used by the State Services Commissionargued to be a mechanism to promote greater public accountability of schoolswere able to be embedded during the 1990s.
While New Zealand was not the only developed country to respond to a perceived crisis of state schooling outcomes by introducing policies to decentralise education (Court and O'Neill 2011), it was unique in being: 'the only country that … built its national school system on schools operating on their own' (Wylie 2012, 4).Tomorrow's Schools generated atomisation and increasing inequality in New Zealand's education system; the skills and knowledgeboth academic and experientialto manage a school in a context that privileged tenets of economic rationalism are not evenly spread across New Zealand communities.Some Boards of Trustees found the trustee role difficult while othersgenerally those in affluent schools with a high percentage of professional parentsfound their role straightforward which, in turn, resulted in increased funding from high levels of enrolments.For example, the actions of 'aspirational families' contributed to increased funding in over-subscribed schools who, through selection policies and enrolment zones, could also mediate beyond-school factors by excluding the most disadvantaged students.By contrast, those schools that came to be labelled as 'failing' had falling rolls and were then required to take any student who wished to enrol (Thrupp 2010).
Echoing trends in other developed democracies, both new forms of accountability and ideological narratives of professionalism were active in the implementation of decentralisation (Gronn 2003).In the first generation of reform, these facilitated the State's ability to steer at a distance, in particular through moving from a concern with 'individual learning and equity imperatives towards a national standards and accountability-led education policy agenda' (Court and O'Neill 2011, 134).
Analysis of second-generation education reforms since the mid-1990s position the 'policy process as a whole as a discursive field or network of social practice' (Court and O'Neill 2011, 120) where individuals and organisations exhibit 'greater or lesser degrees of agency over the public policy process'.Yet, these 'individuals and organizations' and 'networks of social practice' continue to be conceptualised in terms of human actorsgovernment ministers, civil servants and policy advisors, consultants, principals, teachers, and parentswhile nonhuman actorselectoral cycles and political manifestos, historical events, intergenerational aspirations, school gates, funding instruments, the Transnational Leadership Package (Thomson, Gunter, and Blackmore 2021), digital technologies, rooms with doors and desks, buildings and so onfade away as actors.In this article, my aim is to foreground these faded actors to portray a more comprehensive reading of the enactment of collaboration as a policy narrative.

The evolution of collaboration as a policy narrative
New Zealand's Education Review Office (ERO) was formed under the Education Act 1989 to provide and manage an accountability framework for New Zealand's self-managing schools.Soon after its implementation, ERO began to highlight disparities in school performance, including influential reviews of widespread failure in three of the poorest communities in New Zealand: Mangere Otara, East Coast, and Northland (Fancy 2007).By the late 1990s as evidence accumulated, various forms of collaborationboth organic collaborations initiated by schools, and State-initiated networkswere beginning to emerge in efforts to mediate the impact on educational equity of school atomisation.The first of the State initiatives, commencing in 1997, was the Strengthening Education in Mangere and Otara Project (Robinson and Timperley 2004).This was followed by School Improvement Clusters with approximately 10 per cent of schools involved by 2001.2008 saw the introduction of Network Learning Communities to support schools with the implementation of the revised New Zealand Curriculum (Ward and Henderson 2011).In 2010, NZ$7 million was allocated to the facilitation of Learning and Change Networks to support schools with the implementation of the contested policy of National Standards in reading, writing and mathematics (Annan and Carpenter 2015): in the context of economic rationalism these standards were deemed necessary to allow comparisonsincluding longitudinal comparisonsbetween students, schools, and nations (Lough 1990).
Since 2014 the major policy initiative around collaboration has been the implementation of Kāhui Ako | Communities of Learning, one of three components of Investing in Educational Success (IES) (Ministry of Education 2014).Kāhui Ako are voluntary collaborations of early childhood, school and post-compulsory education providers who secure funding for leadership roles that support collaborative work to address the priority challenge for the compulsory school system in New Zealand: achieving equity and excellence for all students.Kāhui Ako leadership roles are: a Community of Learning leader (usually a principal), Across-School Teacher roles, and Within-School Teacher roles.Each role is focused on 'achievement challenges'shared evidence-based goals with at least one associated target that are identified by the Kāhui Ako and subsequently endorsed by the Ministry of Education. 1 In 2021 over two thirds of New Zealand schoolssome 610,000 children and young peoplewere learning within 220 Kāhui Ako (New Appointments National Panel 2021).There has now been a substantive body of research into the limits and possibilities of these leadership roles and, more broadly, the effectiveness of the policy agenda (for example Charteris and Smardon 2018;Kamp 2019b

Materials and methods
In this paper, I draw on empirical research conducted with Kāhui Ako in one region during 2019-20.The region is Christchurch City, based around the second largest city in New Zealand, with a population of some 380,000.As a case, it offers an unusual opportunity to make visible the importance of nonhuman actors in policy enactment including, in this instance, a geological actor.While earthquakes are nearly an everyday occurrence in New Zealand, the magnitude 7.1 Darfield earthquake of 4 September 2010 was an extreme event.Thousands of aftershocks were also triggered, the most damaging of which was a magnitude 6.3 earthquake on 22 February 2011.Centred very close to the already damaged city, and striking in the middle of a working day when people were in the city, it took the lives of 185 people and caused extensive damage to the built environment.
In education, nearly every primary and secondary school in Christchurch was affected with a number of schools in the inner city and the eastern suburbs so damaged they were unable to reopen.For educators, the priority was to get students back at school; to achieve this, nine schoolsprimary, intermediate 2 and secondarybecame 'guest' schools sharing the site and facilities of a 'host' school (Ham et al. 2012).Teachers in every school were faced with providing elevated levels of emotional support for students and families in distress, a situation that continued for years (O'Toole 2018).Within the NZ$1billion rebuild of 153 Christchurch schools (Kenny 2020), modern learning environments and multi-school campuses that demand collaborative practice were progressively introduced (Fickel, Mackey, and Fletcher 2019), all of this occurring contemporaneously with the introduction of Kāhui Ako | Communities of Learning.All of these events tested leadershipwithin and beyond educational organisationsin novel ways.
The research underpinning this article was granted ethical approval from the University of Canterbury Educational Research Human Ethics Committee (Approval 2020/19/ ERHEC).With the endorsement of the Regional Office of the Ministry of Education, I invited a number of Community of Learning Leaders from the Christchurch City, Canterbury region (comprising 16 Kāhui Ako in total) to be interviewed for the purposes of informing a book Leading Educational Networks: Theory, Policy and Practice (Greany and Kamp 2022).Semi-structured interviews were conducted over a period of eight months.Interviews were also completed with Ministry of Education staff with specific responsibility as Lead Advisors for Kāhui Ako in the Canterbury region, with principals of schools involved in Kāhui Ako but not holding the Community of Learning lead role, and with Across-School Teachers.All interviews were transcribed verbatim and were returned to the interviewee for member checking before analysis commenced.Throughout 2021, I also completed observations at meetings of the Canterbury Cluster Leadership Network, a facilitated network supported by the Ministry of Education involving both Kāhui Ako networks and other school partnerships.Field notes were recorded during, and subsequent to, these meetings.Publicly available data on individual Kāhui Ako and their endorsed achievement challenges were accessed from the Ministry of Education website.External evaluation and research reports on the Kāhui Ako policy were also secured and used as secondary data (New Appointments National Panel 2022; 2021).Content analysis was undertaken with all data using an actor-network theory informed coding scheme.

Empirical insights
A number of key concepts from actor-network theory informed my analysis.Iintroduced the term 'assemblage' in the title of this article.My use of this term derives from Latour's (2005Latour's ( /2007) ) title for his introductory text on actor-network theory, Reassembling The Social.There, Latour makes the case for a 'sociology of associations' to eschew making assumptions concerning the composition of social action.Assemblage is both verb and noun; it refers to the gathering(s) of heterogenous actors who influence each other in ways that may, but will not always, result in the durability and extension of the assemblage-as-outcome.From a research perspective, a concern with assemblage and reassemblage demands explicitly tracing the association and interaction of diverse actors in their constantly shifting and often underestimated interactions.The value of such a perspective for those concerned with policy enactment and what makes it work is self-evident.Concerns with assemblage and its outcome(s) are guided by a number of principles.First, the principles of actant and of symmetry.Then, the principles of translation and alliance which, of necessity, draw us towards the distinction between network and 'network'.These principles will frame the presentation of the research insights: I begin by articulating my use of each concept, before presenting the insights.

Actants and symmetry
Already in this paper, I have introduced a number of nonhuman actorstechnologies of accountability, measurement and standards aligned with economic rationalism, parental aspirations, enrolment zones, earthquakes, digital devices, modern learning environments and so onthat were active in generating effects that mobilised the enactment of Kāhui Ako in Christchurch.'Actant' is a term from early actor-network theory that I find useful: it refers to the full range of potential agents in the network; an actant may, or may not, come to be seen to have a 'performing part' (Fenwick and Edwards 2010, 10) in the assemblage.Having a language for these 'potential' actors is useful in allowing space for those actors who we might assume to be the source of action, but who analysis indicates can be put aside as tracing proceeds.Actant is a semiotic definition: something that acts or to which activity is granted by others.It implies no special motivation of human individual actors, nor of humans in general.An actant can literally be anything provided it is granted to be the source of an action.(Latour 1996, 373) This implication of 'no special motivation of human individual actors' flags a commitment to 'symmetry' in actor-network theory.For Law (2009), actor-network theory 'borrows' from Kuhn (1970) and 'other sociologists of scientific knowledge'.For Kuhn, scientists worked through cases, exploring what 'formalisms' mean in practice.This is 'the basic methodological and philosophical principle' of actor-network theory: 'knowledge lies in exemplars and words are never enough' (Law 2009, 143-4).Thus, the fundamentally empirical nature of actor network studies: the minute tracing of processes of assemblage involving both humans and nonhumans and any other form of dichotomy: micro and macro, nature and culture, big and small (Landri 2021, 41).These approaches to knowledge generation align with a move towards posthumanism: for Law, actornetwork theory is an 'empirical version of poststructuralism' (Law 2009, 145).
Some examples from the research highlight this generalised symmetry with an array of nonhumans: digital platforms and devices, school buildings, clothing.For actor-network theorists, digital platforms and devices are not simply tools used by human actors.Indeed, they are often a source of human action in both productive and unproductive ways.In their study, Greany and Kamp (2022) demonstrated that while self-managing schools in New Zealand, as elsewhere, may in principle have autonomy in evaluating their work, in practice this is commonly directed by nonhuman actors that embody models of knowledge informed by the school improvement and effectiveness literature.As has been argued by Landri (2021, 52-3), digital 'platforms are not merely a neutral tool, as they orient the ways schools define themselves', reinforcing an 'epistemology of seeing', in the process exerting a disciplinary power (Foucault 1983).Yet, the case also illustrates how digital platforms have potential to act as a 'prime mover': an enabler of a more stabilised assemblage that may generate positive outcomes to all the actors in an event.Thus, innovation can occur through successful, repetitive associations of humans and nonhumans.
One of the Christchurch Kāhui Ako demonstrates the mobilising agency of nonhumans in collaborative endeavours by humans.The Kāhui Ako includes seven schools two secondary, two intermediate and three primaryand thirteen early childhood education providers.Over 2,500 of the students in the Kāhui Ako are based in one campus alone.This campus exists because of the presence in the network of the aforementioned geological actor; in the rebuild process after the Christchurch earthquake the two formerly separate secondary schools (one a girls' school, the other a boys' school) now share a new campus: This spectacular, colourful campus enables each school to retain its culture, identity and teaching spaces while facilities such as administration, auditorium, gym and cafe are shared by both schools.There are physical bridges that link the two teaching 'wings', [the boys' school] on the left and [the girls' school] on the right as you enter the campus.Each school is visually distinct, yet closely intertwined given the way the architecture acts.There are multiple Kāhui Ako-funded 'within' and 'across' school teacher roles on the campus and thus the chance for 'casual conversations and then the opportunities that arise from those happen much more often'.(Greany and Kamp 2022, 150) This emerging architecture was spoken of by teachers as a source of action that encouraged them to act in unfamiliar ways, implementing an array of mechanisms to 'be visible', building unusual stores of social capital (Bourdieu 1986).This in turn provided a platform for the joint practice development (Hargreaves 2011;2010;Hargreaves and Fullan 2012) that has been argued to lie at the heart of educational collaboration.The word 'colleague' has taken a new meaning, referring to all stakeholders in the Kāhui Ako, including actors from the local university and local industry.Clothing can now be seen as an actor with one respondenta teacher at secondary school levelexplaining that, in beginning his Kāhui Ako Across School role, he took on relief teaching for other Kāhui Ako members including those at early-childhood and primary level: 'I got out of this suit because you can't be a [kindergarten] teacher if you're in a suit'.Having invested his time working in early childhood and primary schools the question arose of how to sustain the collaboration without dependence on his presence.Here, digital platforms act: an app was introduced to the assemblage.The app acted to connect humans, smartphones, fitness regimes, school timetables and commitments to learning: They didn't need to ask their principal's permission, they didn't need to book a time with me; we didn't need to align our calendars or whatever.They could click on Spotify, they could click on Apple, they could go on their phone.They're at the gym, on their way to school, and they could listen to a conversation and be part of that collaborative conversation.(Greany and Kamp 2022, 151) Translation and alliance Translation, and the processes of mediation by which change occurs, is a central principle in tracing the enactment of policy.Actor-network theory is focused on enabling stakeholders to see the diverse range of actants and exploring their work as either 'intermediaries' or 'mediators'.An entity that is an intermediary 'transports meaning or force without transformation': whatever is its input will also be its output.Intermediaries are rare; very few objects exert no influence whatsoever by their presence in an assemblage (Latour 2005(Latour /2007, 38-40), 38-40).Intermediaries often project an appearance of solidity, 'a black box appearance [that] gives the impression that things are naturally and inevitably just as they are, the smooth exterior obscuring the internal machinations' (Kamp 2019a).There are, however, prolific, always complex mediators that 'transform, translate, distort, and modify the meaning of the elements they are supposed to carry' (Latour 2005(Latour /2007, 39), 39).These 'unfaithful intermediaries' (Michael 2017, 180) not only disrupt or destroy associations, but may also 'proliferate and complicate them'.This reference to proliferation introduces the principle of alliance.A sociology of associations does not take one object or actor to be inherently stronger or more powerful than another; 'actants gain in strength only through their alliances.As long as no one reads Mendel's papers, his breakthroughs in genetics remain weak' (Harman 2009, 15).Thus, 'the system' is not necessarily any more powerful than other actors, particularly those actors who orient themselves towards heterogeneous alliance and collaboration.
Michel Callon's (1986) non-linear 'moments' of problematisation, interessement, enrolment, and mobilisation has provided the 'basic vocabulary of the [translation] approach' (Landri 2021, 8).The introduction of Kāhui Ako, offers an apt illustration of using the principles of translation and alliance to explore the enactment of network policies.'Problematization' is the point at which there is articulation of a problem or issue.As detailed in the introduction of this paper, from the get-go critics had profiled the unintended consequences of New Zealand's radical approach to self-managing schools.In response, various small-scale network initiatives had been attempted by successive governments; organic collaborations had been initiated by communities of schools.The fifth National Government, elected in 2008, was a prime mover in initiating problematisation through establishing the Investing in Educational Success Working Group and consequent policy.With a significant allocation of the Vote Education budget allocated to the policy, the problem of atomisation and its consequences was articulated in such a way to become a key player in the education landscape.An obligatory passage point was established: a 'single locus that could shape and mobilise the local network' and 'have control over all transactions' (Bijker and Law 1992, 31).However, the policy as enacted can clearly be understood as an effect of a particular assemblage.Political interests were part of the web of relations that formed the policy as actor.In his overview, Thrupp (2018) details how the initial announcement of the policy by the then Prime Minister -John Key, leader of the right-leaning National Partywas made to business leaders rather than educators.The assemblage of private sector stakeholders, high levels of prescription in the policy, particularly related to the agenda around National Standards and the managerialist imperative within them, along with the diverse actions of educators who on principled grounds either opposed the policy or 'entered the tent' to mediate it (Thrupp 2018) was argued to be essential to such success as can be attributed to Kāhui Ako in the long term.
'Interessement' is the moment whereby an entity acts to stabilize the identity of other entities in the problematization.For Kāhui Ako, the introduction of compulsory achievement challenges, endorsed by the Ministry of Education, achieves this stabilization.In the absence of the achievement challenge that is submitted by the Kāhui Ako and endorsed by dedicated staff in the Regional Office of the Ministry of Education, there is no formalization of the network and no access to dedicated funding.Conversely, the presence of an endorsed achievement challenge in the assemblage draws other actors in technologies of governance which confine the possibilities for focus 3 and structure, and preclude engagement and investment in other collaborative endeavours.However, this stabilization of a particular form of identity is never complete.Mediation can take the form of interference and, while the Ministry can translate the goals of an existing collaboration that wishes to be recognized as a Kāhui Ako, the collaboration can equally translate the goals of the Ministry.Thus, for all actors there is tangible labour required to maintain a beneficial assemblage: 'continuous effort is required to hold it together, to bolster the breakages and counter the subterfuges'.(Fenwick and Edwards 2010, 11) The third moment of 'enrolment' captures processes of negotiation that further stabilises the web of relations.Roles and relationships are formally assigned.Negotiation strengthens the network but can equally weaken it: actors 'may connect in ways that lock them into a particular collective, or they may pretend to connect, partially connect, or act disconnected and excluded even when they are connected' (Fenwick and Edwards 2010, 120).As Graham Harman (2009, 116) notes: Actors are defined by their relations, but precisely for this reason they are cut off in their own relational microcosms, which endure for only an instant before the actor is replaced by a similar actor.The work of mediation must be done at every moment to restore or maintain the links between actors.
In the Christchurch region, this form of labour was apparent, as was the sometimes benefit of being cut off from one's relational microcosm.One secondary school, already a member of a Kāhui Ako which reflected its historical connections nonetheless took up an associate membership with a second Kāhui Ako.For the purposes of a key actorfundingthere was no benefit to this second enrolment, a point I elaborate on in the following paragraph.Yet, this enrolment opened the way for a shared future involving both Kāhui Ako which could not have been imagined previously, particularly given the vast difference in the socio-economic standing of these two educational communities.With the Kāhui Ako policy and the current principal at the secondary school as fully-fledged actors, the web of relations has evolved from two networks of primary schoolsone loose network, and one tight network with a long historyto a network of networks that connects early childhood education, and primary and secondary schools that ranged from the then decile two to decile ten 4 , in the process undermining the strength of deciles, formerly a prime mover in categorising schools and narratives about the potential of their students.
Enrolment and the formalisation of roles and relationships clearly highlights the impact of funding as an actor.As my own earlier research has evidenced (Kamp 2013), there are two immediate effects of the presence of funding as an actor.First, at the level of the assemblagethe Kāhui Akothe presence of funding as an actor can, as Fenwick and Edwards also found, result in enrolment of actors whose motivations are focused on access to funding, rather than the call to action, whatever that call may be.Second, Bourdieu (1986) argues that in any network, a form of delegation is institutionalised by which the network is able to focus its volume of social capital into the hands of an agent whom is charged with authority to represent the group.In Kāhui Ako, teachers who are appointed to formal roles as Across School Teachers or Within School Teachers are granted staff release time that can readily slip towards a perception that work of the Kāhui Ako, its joint practice development is 'their' responsibility, rather than the shared investment of all.Thus, in Christchurch, there were groups of schools that would not engage with this major policy initiative, refusing this stabilisation of their identity, despite the potential for Kāhui Ako to become an obligatory passage point for collaboration in education across New Zealand.For the schools in these collaborations, the absence of government and its technologies of governance was seen to be positive, eliminating actors that brought very particular relations into play whilst also opening the possibility for skilled leaders to be involved in formalised roles 5 : The biggest challenge I think in the design of Kāhui Ako is the appointments or the criteria around the across-school teachers, and even in part the within school, although the withinschool were easier to manage and mould.But the across-school teachers was a challenge because [they] … did not have leadership experience to a degree where they understood even successful leadership.They might have had an experience where they led an activity or a leadership task, but they didn't quite understand how to lead others in a way that values expertise, builds collaborative, shared understandings.(Principal) The fourth non-linear moment is 'mobilization'.Mobilisation occurs when issues of representation are resolved.Mobilisation renders formerly immobile entities, mobile.In the Christchurch cases, both human and nonhuman actors were apparent here.Lead principals, and principals of early childhood centres, schools and post-compulsory providers were potentially spokespersons for the collective, able through their positional leadership to mobilise formerly immobile entities including staff, learning spaces, digital platforms, teaching resources, and so on.Ministerial guidelines, school bulletins and newsletters, innovative learning environments (or traditional classrooms), student data and research papers were also active in mobilisation, with different levels of visibility with different actants.Thus, the extent to which a network retains a sense of 'realness' depends on a sustained effort on multiple levels and in multiple directions to attract allies: 'the more attachments it has, the more [the network] exists and the more mediators there are the better' (Latour 2005(Latour /2007, 217, original emphasis), 217, original emphasis).

Network
What is important in the word network is the word work.(Latour 2004, 83, original emphasis) In 1996, Bruno Latour spoke to three misunderstandings about actor-network theory that drew from the common usage of the word 'network'.Given the case in this article is one concerned with collaborationwith networking -Latour's advice merits revisiting.He suggested the first misunderstanding was to 'give [the word 'network] a common technical meaning in the sense of a sewer, or train, or subway, or telephone "network"'.The second was that actor-network theory had 'very little to do with the study of social networks'.Such studies focus on social relations of individual human actors, a stance from which actor-network theory fundamentally departs as it not only 'extends the word actor … to non-human, non-individual entities' but also stresses that 'attachments are first, actors are second (Latour 2005(Latour /2007, 217), 217).The third misunderstanding was that there is nothing but net-works, there is nothing in between them, or, to use a metaphor from the history of physics, there is no aether in which networks should be immersed.(Latour 1996, 369-70) These misunderstandings are particularly pertinent in considering what actor-network theory can contribute to policy enactment in the case I have presented in this article.As has been suggested in other research on policy assemblages (for example, Gorur 2011), policy is at the other end of the continuum from actor-network theory's sensitivities; policy is aligned with these 'misunderstandings'.Policy documents and guidelines do focus on the common, technical, structural meaning of the word 'network', and its human/social members, rather than transient webs of relations and associated net-work to maintain alliances that involves both human and nonhuman actors.One might argue policy is conceptualised as 'the aether' in which and from which practice, be it collaborative practice or some other form of practice altogether, is immersed.
Yet in the case presented in this paper, policy development and enactment are not far removed from local practice, given the actors introduced early in this paper.First, te Tiriti of Waitangi | the Treaty of Waitangi.Education in New Zealand has, not before time, become increasingly accountable to the principles derived from te Tiriti: partnership, participation and protection.All educators and policy actors are trained, and work within, the requirement to engage with and embrace a Māori worldview and that world viewte ao Maoriis fundamentally collective (Macfarlane and Macfarlane 2019).This was evident in my research where leaders often made reference to strengthening links with local iwi and hapū [tribes and subtribes] in seeking to meet their achievement challenges.Second, New Zealand's accountability framework in education remains less 'high stakes' than those in other jurisdictions and its government has, for decades, worked by means of guidelines rather than prescription (Wylie 2012), effectively enabling flexibility in policy enactment.Tomorrow's Schools with its approach of 'steering from a distance' continues to act despite a major review, not least because some teachers and principals have no experience of any system other than that of three decades of selfmanaging schools.Geological and geographical actors remain.The profound upheavals during and subsequent to the earthquakes of 2010 and 2011 made human actorswherever they were in the education systemact in innovative ways and mobilised the enactment of the collaborative policy agenda.
the Ministry [moved] from being a, 'here's the manual and this is the compliance' through to 'we are a service agency for education.How can we make your idea work'?That ethos is still my complete experience of the Christchurch Ministry.I love the Christchurch Ministry.(Principal) This overview gives insight into the array of actors as they move through and around the four moments.With sufficient allies, it may be that the policy becomes routine and, thereby, 'real'a fully functioning part of 'the social' for as long as the web of relations 'holds'.As Landri (2021, 83) explains, 'In [ actor-network theory], the word network does not qualify a closed thing but a dynamic assemblage of entities, a set of successful translations'.When this dynamic assemblage gains some sense of stabilisation and begins to act 'as one', it can be thought of as a 'black box' (Latour 1987, 131).Policy documents can, in striving to appear authoritative, be composed as if to imply that practicein this case, educational collaborationsis a black box before the event.This black box is simple, unproblematic, input equalling output, not a big deal, smooth in its function, amenable to control.However, this effort at black boxing serves neither policy nor practice well.It gives the impression of some form of closure or completion (Landri 2021), in the process diminishing expectations of necessary dissent (Thrupp 2018), minimising appreciation of the criticality of learning, and the constantly evolving effort that is involved in holding together potentially paradoxical webs of relations in a given policy area.
In previous work (Kamp 2013), I have written about the implications of all of this for government, and governance.That book -Rethinking Learning Networksinvolved me in a case study of the implementation of a policy of learning and employment networks in the state of Victoria, Australia.Intended as a three-year initiative, Local Learning and Employment Networks have stabilised and have become 'real' in the context of the Victorian education system.My analysis there is pertinent to the discussion in this paper in its argument that the insights from poststructuralism pose a learning challenge for governments in their policy enactment efforts: Thus this is a learning challenge for governments … conceiving of an expanded language for governance, one developed in the process of situated learning within the network-rhizome of which it is a part … .Clearly, the parallel, unarticulated challenge that is fundamental to releasing the potential of networks lies equally in capacity-building for government itself.In this new paradigm for the governance of networks there is a concern for outcomes other than the 'distinctive [government] approach to the definition of success' (Seddon 2005: 41), outcomes that are not solely pre-specified but also flow from informal accountability among network actors (Romzek 2012).… For me, the significant word here is 'vibrantly' it is a word that connects with the kind of energy and movement that successful networks focused on sustainable change evoke, it connects with the desire which is 'never separable from complex assemblages that necessarily tie into molecular levels' (Deleuze and Guattari 1987: 215), that is, the desire that is fundamental to rhizomatic thinking.(Kamp 2013, 89-90) If nonhuman actors such as procedures and guidelines in a context such as New Zealand can be perceived as 'stifling' policy enactment (Kamp 2019b), then we need to trace the ways in which actors such as procedures, guidelines, precedents and accountability frameworks of various forms make teachers, principals, Boards of Trustees, parents and students act, particularly when they come into dynamic assemblages with all of the other actors in a given event.

Discussion
In what remains of this paper, I want to discuss how actor-network theory's forms of mediation enable us to 'see' the enactment of the Kāhui Ako policy in new ways, and to consider the implications of this for leaders, leadership, and leadership development.Successful policy enactment is more likely when stakeholders acquire a shared goal.This is mediation in its guise as 'interference'.In this, each actor interferes with the original goal of the other, in the process moving policy-connected entities from co-operation to achieve individual goals, to collaboration and a shared goal that is the focus of all.This leads to mediation in its second guise as compositionthe composite goal becomes the common achievement of each of the actors.Thus, in Kāhui Ako, principals commented on seeing 'one of their kids'.However, the student that prompted this comment was not from the school the principal led.Rather, the achievement of all students wherever they may be enrolled in the Kāhui Ako was the concern of every leader; the school uniform was no longer a fully-fledged actor in regard to signalling the spectrum of educational opportunity.In its third guise, mediation is black boxing: the more something succeeds, the less it can be understood.By focusing only on inputs and outputs, rather than the complexity and messiness inherent in policy enactment, we lessen our own ability to understand what, and who, makes us act as we do.Finally, in its fourth and most important guise, mediation is delegation.This refers to how action is delegated to nonhuman objects; in the context of education delegation is profoundly evident such as in the use of standardised tests, assembled in contexts far removed in time and place from the lives of teachers, students and parents, potentially making them act in ways that would not be of their choosing and against which they may react (Nespor 2002).Delegation also illuminates how a 'prime mover' can be absent, yet profoundly present: long ago actions can remain active in our day-to-day lives (Latour 1993).New Zealand offers a profound example in this regard: the social consequences of injustices related to the failure to honour te Tiriti o Waitangi continue to reverberate in educational settings; in the case study discussed in this paper, the movement of the Earth's plates also continue to reverberate, in multiple assemblages.
From an ANTian perspective, to be a 'realistic whole' is not the starting point but, rather, a provisional achievement of a composite assemblage (Latour 2005(Latour /2007, 208), 208).Such an orientation sits in tension with the forms of accountability that policy enactment often demands, particularly in its early stages.Yet there is 'no straightforward determinism through which technologies and materialities make the social durable' (Landri 2021, 87).For those who take up the principles of actor-network theory, there is a need to privilege the process and practice over the dictates of structure and procedure.And this, by necessity, lays down a challenge to those humans charged with policy development and enactment.
The privileging of practice also has implications for leaders, and leadership preparation.If policy enactment involves a focus on dynamic assemblage, and if both sustainability and power rest in an understanding of how to acquire and maintain allies, then relationship work and collaboration takes centre stage.Professional learning and development around leadership must be part of the assemblage, regardless of where leadership lies.Such a perspective is not in tension with current leadership trends towards understanding and fostering leadership as diffused and shared.However, what will be new is a clear appreciation, and the skills and knowledge required, to follow and engage with all the actorshuman and nonhuman, micro and macro, present and pastin a shared commitment to a productive level of policy stabilisation.Notes 1.All achievement challenges are publicly available on the interactive Education Counts website: https://www.educationcounts.govt.nz/know-your-col 2. Intermediate schools are a bridge between primary school and secondary school.Primary education in Aotearoa starts at Year 1 and continues through to Year 8. Years 7 and 8 can be completed either at a primary school or at a separate 'intermediate school': 'These schools allow students to compete, connect and engage with other students from different schools at their level.We are able to provide competitive opportunities on a scale that full primaries are simply unable to do' https://www.studywithnewzealand.govt.nz/en/study-options/education-system 3. The requirement for achievement challenges to focus on National Standards was subsequently removed, but a focus on the National Curriculum remains.
4. School deciles ranked schools from decile one to decile ten, based on socio-economic indicators.Some 10 per cent of schools are classified as either decile one or decile ten.Deciles were intended to target funding for state and state-integrated schools to mediate barriers to learning faced by students from lower socio-economic communities: the lower the school's decile, the higher the funding.Decile 1-7 schools could receive extra funding if they did not ask for donations; 95 per cent adopted this policy.From January 2023, decile funding is to be replaced with Equity Funding calculated through use of an Equity Index. 5.The procedures precluded teachers holding positions of responsibility, (for example, a Deputy Principal) from applying for these roles.

Disclosure statement
No potential conflict of interest was reported by the authors.

Notes on contributor
Dr ; Ministry of Education 2017; Ministry of Education 2018; New Appointments National Panel 2022; 2021).
Annelies Kamp has 25 years of experience in senior leadership roles in Crown agencies, industry training and the not-for-profit sector.Her academic career has been undertaken in Australia, Ireland and New Zealand.At the University of Canterbury she was Head of School and Deputy PVC.She is now Adjunct Associate Professor in Leadership.She has published six books: Wellbeing: Global Policies and Perspectives (2023), Leading Educational Networks: Theory, Policy and Practice (2022 with Toby Greany), Education Studies in Aotearoa: Key Disciplines and Emerging Directions (2019), Re-Assembling the Pregnant and Parenting Teenager (2017 -available fully open source), A Critical Youth Studies (2015 with Peter Kelly), and Rethinking Learning Networks -Collaborative Possibilities for a Deleuzian Century (2013).