A processual framework for analysing liberal policy interventions in conflict contexts

The article proposes a heuristic framework based on processual sociology to analyse policy interventions aimed at change within conflict contexts. Such a framework is valuable because it creates an opportunity for a more open approach to empirical research that may allow us to research evolving processes and to see things we might miss otherwise. The article aims to complement goal-oriented and predominantly relational approaches and to contribute to debates that warn against the reification of actors and structures in research. It also points to a lack of attention to politics in the analysis of policy interventions. The argument derives from a discussion of transitional justice and peacebuilding and is empirically illustrated for the context of the Tunisian transitional justice process.


Introduction
This article proposes a framework based on processual sociology for analysing liberal policy interventions that aim to promote and steer change in conflict contexts. These policy interventions include, for example, efforts in peace-or statebuilding, transitional justice (TJ) or democracy promotion. I argue that such a processual framework allows us to broaden our analytical perspective and take a more open approach to empirical research (Abbott, 2016; see also Danielsson, 2020b;Menzel, 2020;Schroeder, 2018, on the necessity to do so) that helps to avoid 'the tendency to reify the actors and structures one sees' (Kennedy, 2016: 75) in analysis (see also Brown, 2020;Hirblinger and Simons, I propose a fresh perspective that expands our analytical horizon by moving away from the interveners 'theory of change' as the site of analytical departure (Sabaratnam, 2017: 17;Schroeder, 2018: 143-144; see also Paffenholz, 2021: 380), by moving away from a priori categorisation of actors and their relations (cf. Danielsson, 2020bDanielsson, : 1088 and towards a processual perspective. This approach allows us to analyse 'what is going on here?' (Schwedler, 2013: 28), broadening our analytical perspectives and avenues for empirical research without deciding beforehand which foci of analysis are the most relevant. The approach works out what characterises the processes we are studying and challenges an 'understanding that society [would be] developing more or less "automatically" towards a better social order' (Elias, 1978: 151). We can then see any diversions from plans of liberal policy interventions, not as mere unintended consequences or 'a result of deviant national or local politics' (Kappler, 2018: 136), but also as an essential characteristic of processual developments, and thereby helping to avoid analytical bifurcation (see, e.g. Go, 2017).
This article proceeds through three parts. First, I outline the argument about why a processual perspective is valuable and what it might offer for the analysis of policy interventions. Second, I propose a processual analytical framework for the analysis of internationalised processes of change, the centrepiece of this article. This framework combines insights from processual sociology drawn mainly from Norbert Elias' work and that of International Relations and Peace and Conflict Studies. It introduces two key concepts and three initial processual characteristics as starting points for empirical research. Third, I briefly illustrate this framework in an empirical context, before concluding with a summary of the article's argument.
The empirical illustration is based on original data collected concurrently to the evolving Tunisian TJ process. I conducted field research in Tunisia (almost 6 months in total between 2014 and 2018) and the United States (12 weeks in total in 2015 and 2019), and by phone and video (2020). The data include interviews with domestic and international TJ professionals, politicians and government officials, civil society representatives, the media, and truth commission members and staff (over 100 in total), as well as (participant) observation at, among others, conferences and truth commission events (see Supplementary appendix for a more detailed description of the research process). The 'process-concurrent' nature of the research process, and the dynamic, direct experience of the research topic, likewise warranted a processual approach that did not require an endpoint for analysis.

Why do we need a processual approach?
Although 'change is the norm in all life and all fields', its process does not get much attention in the study of international/world politics (Crawford, 2018: 233; my emphasis).
As Jackson and Nexon (1999) state, '(. . .) theories of processes and relations are better suited to address certain questions, most notably those involving change in global politics' (p. 291). Although they focus their analysis and concept development on states, Jackson and Nexon also call for them to be applied to particular phenomena. This article focuses on one set of phenomena: policy interventions in global politics that are part of, and aim at, inducing/catalysing change in a particular direction. Such interventions constitute internationalised processes of change 1 that attempt to alter norms and institutions. 2 They assume the existence of a certain malleability in those countries regarded as transitional states (McAuliffe, 2017).
In the analysis of global politics, particularly as it relates to change, Jackson and Nexon distinguish between two modes of analysis: 'substantialism' and 'processual relationalism'. 3 They consider the latter a more accurate term than simply 'relationalism', since it analyses 'configurations of ties', which are 'not static "things" but ongoing processes' (Jackson and Nexon, 1999: 292). While relational analyses have become more popular in global politics and peace and conflict studies (e.g. Boege and Rinck, 2019;Hirblinger and Landau, 2020;Jones, 2016), the processual aspect of the dyad has gained less attention. 4 Research that is situated within a processual relational approach, and aims to push the study of interventions towards more open empirical options, takes more seriously the 'epistemic commitment' (Danielsson, 2020b: 1086; see also Brown, 2020) that acknowledges (policy) interventions as co-constituted processes, often places the relational at centre-stage (e.g. Danielsson, 2020aDanielsson, , 2020b. Recently, this lack of a distinct focus on processes has likewise been identified for peace and intervention studies and organisation studies by Moe and Geis (2020) -who share my finding that a processual perspective often highlights useful insights that might otherwise be missed. 5 Moe and Geis, however, offer a heuristic that draws on one processual feature, namely friction (which I also make use of for the heuristic proposed here), and one static feature -hybridity -which has been criticised for reifying binaries and for its focus on outcomes (Danielsson, 2020a: 118).
In this context, I propose an alternative heuristic to complement those approaches focusing on relations. I propose to shift the focus towards processes. Although relations are indeed central features of processual analysis, they constitute just one feature and serve as analytical tools. Looking at processes in their own right may allow us to see things that we would otherwise overlook (Abbott, 2016: xii). This alternative heuristic approach joins recent proposals that encourage more open empirical research (Brown, 2020;Danielsson, 2020b;Menzel, 2020). I draw on Abbott's processual sociology to develop the argument, but to develop the framework, I mainly draw on Eliasian processual sociology (in combination with insights from other fields) because it fits well with processes that entail initiatives to induce, catalyse or shape change -the focus of this article. Such a processual approach may offer valuable contributions in three areas: a fresh analytical perspective that may allow us to see things we might miss otherwise, the epistemic commitment to avoid reifying actors and structures in analysis, and a way to research evolving processes concurrently.

Goal-orientation, teleological appeal
This article posits that the 'process gap' in the analysis of internationalised processes of change stems from a strong analytical focus on goals, outcomes and effects that mirror approaches in practice, as well as a certain neglect of politics in the analysis. For internationalised processes of change, practice, as well as research, often follow a teleological, goal-oriented logic (Jones, 2021;Kappler, 2018;Wesley, 2008) -or, as Moe (2010 puts it, focus on what ought to be. This logic may be related to a tight link between scholarship and practice (e.g. through collaboration) or to the fact that programming patterns from practice are often mirrored in research because they orient (academic) assessment. 'Oughtness' is a criterion for judging something that carries with it a certain performativity (Abbott, 2016: 30-31). Thus, mirroring oughtness in research logics bears the risk of mirroring (not analysing) performativity in turn. In any case, Abbott (2016: 161) notes that there is 'little reflection on why we have chosen the particular forms and timings of 'outcome' that we have in fact chosen', which makes outcome a 'value-laden' characteristic rarely reflected upon in research.
In the following, I will outline the goal-oriented logic in an example from the overlapping fields of TJ and peacebuilding. There are other fields of policy interventions, such as democracy promotion, which are similarly overlapping and for which a similar logic could be identified, but that goes beyond the scope of this article.
TJ is essentially about processes of change. However, the focus in TJ is often on the goals, outcomes and effects of TJ efforts (Jones, 2021). The concept is closely linked to the 'transition paradigm' to liberal market democracy (Arthur, 2009) and exhibits goalorientation, both in practice and in scholarship. TJ efforts, which are introduced after conflict or violent rule, usually have (implicit or explicit) teleological ends, such as justice, peace and democracy, against which they measure 'success' (Jones, 2021: 4). Thus, one prominent strand of research either explores -normatively -what TJ should deal with or how it assesses -empirically -the extent to which certain goals have been reached. There is also a wealth of scholarship that discusses, from both quantitative and qualitative perspectives, the outcomes and effects of TJ measures for peace, human rights records, democracy and political institutions (for an overview see Salehi, 2022a: 48). This work includes critical assessments of success and failure that question the parameters according to which this is measured (Hirsch et al., 2012). Although there has been a shift in the field with regard to what TJ is assumed to be aiming to achieve, McAuliffe (2017: 11) identifies a continuing presumption 'that transformative outcomes can be produced, legitimised and stabilised as matters of intention and design, and that transitional justice can catalyse desired economic dynamics and outcomes'.
Scholars have paid even scanter attention to how TJ interacts with the 'transition' (i.e. the political processes it ought to complement and render more just (McAuliffe, 2017;McGrattan, 2009:)). McGrattan (2009) assumes TJ to be 'implicitly prescriptive and seeks to instil several practical, theoretical and policy interventions into areas identified as 'transitional'' (p. 165). One critique is that TJ research often 'write(s) politics out of the equation' (McGrattan, 2009: 166; see also Kochanski, 2020). With the focus on TJ as a project that unfolds according to a specific plan -or at least tries to -the political changes that this project should be facilitating may be neglected. A processual approach helps keep politics in our analytical equation as we observe 'projects' of change unfold in transitional contexts.
In the broader field of peacebuilding, we can observe similar dynamics. Since the 1990s, international peacebuilding has been an essential part of global governance, aimed at 'transform[ing] war-torn states into those that could sustain peace through rule of law, market-based economies, and liberal democracy' (Campbell, 2018: 1-2). 6 This clear goal of peacebuilding practice also serves as an anchor and guide for a prominent strand of peacebuilding research. Peacebuilding efforts rest upon the assumption that 'an engineered process of simultaneous statebuilding and democratization can bring modern political order to post-conflict states' (Barma, 2017: 1). Common programming patterns long assumed that 'if one develops a project description with the right analysis, the right strategy, the right project aims, and the right measurement indicators and anticipate the right risks, that project will achieve its desired outcome' (Campbell, 2018: 8;see also de Coning, 2020). And Kappler (2018) shows that, in peacebuilding, intervention is framed as linear and aimed at progress, while deviation from this linearity and 'backwardness' is assumed to appear from outside the intervention. These assumptions point to a teleological 'theory of change' that has been underlying international peacebuilding efforts. Meanwhile, the understanding has prevailed among academics and policy-makers that linear paradigms should neither reflect nor shape peacebuilding (Paffenholz, 2021; see also de Coning, 2020) and that there may be changes in the approaches to practice. However, according to Paffenholz (2021), these changes have been rather 'cosmetic' (p. 367).
These general trends in the practice are also mirrored at times in research that focuses on 'mandates, mechanisms and immediate outcomes' (Barma, 2017: 3). As with TJ, there has been plenty of research engaging broadly -and more or less critically -with the goals, outcomes and effects of liberal peacebuilding. And similarly to the dynamics in TJ outlined above, scholars have found that there is the danger of missing out on politics in the analysis when the dynamic context of peace missions remain unaddressed (e.g. Barma, 2017;Campbell, 2018: 15;Moe and Geis, 2020;Paffenholz, 2021). By critiquing how narratives of linearity and progress are orientalising and used to justify the intervention, Kappler (2018) also argues for moving beyond such evaluatory questions.
Autesserre's (2014) ethnography of peacebuilding represents one important work taking on the question of 'how' and not only 'what', albeit also mainly focusing on questions of efficiency and effectiveness. Various contributions that deal with practices and relations between agents claim that they are about processes, or that they spell out the need for 'an analytical sensitivity to process' (Danielsson, 2020b(Danielsson, : 1086. Authors working at the intersection of research and policy have emphasised the complex, non-linear and dynamic nature of peacebuilding and made suggestions of how to better account for them. De Coning(2020), for example, has recently made a conceptual case for 'adaptive peacebuilding' by rejecting the linear theory of change on which liberal peacebuilding has long been built and drawing instead on complexity theory 7 to develop a theory of change, geared towards policy-makers, that achieves more effective outcomes. In a, in her own words, more utopian vision, Paffenholz (2021) argues for discarding notions of 'success' and 'failure' and in favour of a paradigm of 'perpetual peacebuilding' (p. 377). 8 Thus, the focus is rarely on the processes themselves, either in their own right or as a starting point for analysis; attention is still focused primarily on the actors and their practices, or on the relations between them. Or the focus is on how to improve peacebuilding, so that it fits better with complex realities -a more prescriptive analytical perspective than I would suggest, because it nonetheless transports 'oughtness' to some degree. 9 Thus, to use Danielsson's (2020a: 119) wording, these 'are still useful but need a further push' towards processual thinking.

A processual framework for analysis
Often, analyses focusing on processes consider something from 'the end' backwards. Trying to find a posteriori explanations for one particular pathway, they may obscure dynamics along the way. Processes of political change -such as those that peacebuilding and TJ interventions should induce, catalyse or shape -are dynamic, not static. Researching them from the end does not seem to be the most suitable approach (see also Paffenholz, 2021: 378). In contrast, the processual approach suggested here draws on ideas that assume that one should regard the developing society as the subject of inquiry while considering that the possibilities of inquiry may change during the process of inquiring itself (Elias, (1983: 383). This approach is particularly well-suited for use in combination with an interpretive research design, that is, one constituted by a logic of inquiry that has the researcher constantly going 'back and forth in an iterative-recursive fashion between what is puzzling and possible explanations for it' (Shwartz-Shea and Yanow, 2012: 27). The processual approach encourages the researcher to conduct process-concurrent research, and it is well-suited to researching both processes that are still developing and those that offer the benefit of hindsight.
Following Soss (2018: 21), I adopt a nominal view on 'casing a study'. 10 I take the analytical framework developed here as a starting point while recognising the possibility that my subject of inquiry may prove to be a case of something else also. As part of the interpretive research process, the guiding assumptions I introduced may be revisited in conjunction with empirical findings, and further theoretical insights may be generated for a particular phenomenon. This means that we can identify other -substitutive or complimentary -characteristics of processes for a particular field and, if one wants to further systematise the study (because the analysis warrants further 'scholarly machinery' (Abbott, 2016: 1)), 11 we could combine them with a stage model.
In the following, I introduce two key concepts and three processual characteristics that can serve as analytical tools and starting points for empirical analysis. The first key concept of (con-)figuration is relational. The basic understanding used in this article follows Elias ( [1986), who talks of figurations, but there are several related conceptualisations that go in similar directions; these are often termed configurations. The second key concept, project is about goal-orientation and plans. I develop the notion by drawing first on Jackson and Nexon (1999) and Kennedy (2016). The three processual characteristics that I develop, and that may serve as starting points (or guiding assumptions) for analysing internationalised processes of change, are informed by theory and by empirical insights through the interpretive research process. First, in contrast to common programming patterns (cf. Campbell, 2018: 8), planned processes of change -or projects -usually interplay with unplanned political and social dynamics. In particular, a processual approach does not consider divergence from a plan as only 'unintended consequences'. Instead, it considers them to be an essential part of processual developments. Second, internationalised processes of change are not linear and are sometimes marked by simultaneous trends and counter-trends. Third, the conflicts and frictions driving and defining the process are essential characteristics of internationalised processes of change. Finally, I illustrate the framework, using insights from empirical research on TJ in post-revolutionary Tunisia.

(Con-)figurations
The term configuration is often used for a set of concepts mobilised to describe relations or ties, the patterns among them, and essentially 'an aggregation of processes' (Jackson and Nexon, 1999: 304). Jackson and Nexon subsume under this notion not only Bourdieu's field, and discursive formations as described by, for example, Foucault, and Laclau and Mouffe, but also Elias' notion of figuration, which I mainly draw on and will introduce here in a bit more detail. Essentially (con-)figurations, as I understand them in this article, are about dynamic actor and power relations.
(Con-)figurations have decisive influence on the dynamics of social and political processes and changing power relations. They emerge '[t]hrough the force of essential interdependence' (Elias, [1986: 101; my translation) of actors. They have structural features and function as representatives of a particular order (Elias, [1986: 100-101). Figurations are also shaped by the transmission of knowledge: actors enter specific symbolic worlds of existing figurations, which may alter those figurations. But actors also acquire specific societal symbols of knowledge or language when they enter the figuration. (Con-)figurations, and thus social processes, are shaped by interdependent and changing actors and power relations. (Con-)figurations can form on different levels and in different sizes and may cut across various actor groups. They can be formed by individuals, but also by other units/actors (e.g. states) and are not confined by often differentiated categories of 'global' and 'local'. Alliances may develop across these groups, so might conflict and friction that drive and define the process. Thus, organisations and institutions can also form, or be part of forming, figurations, either as a collective actor or by way of their representatives. Figurations also transcend state boundaries (Elias, (1981: 48-49). For example, they can form through rotating experts in the international sphere or in digital space and thereby shape internationalised processes of change.
The notion of (con-)figurations, and the features linked to how they are formed and shaped, offer an analytical tool and several entry points that may prove useful in the analysis of internationalised processes of change. Dynamics and processes displayed in empirical analysis may fluctuate between different levels, depending on the analytical aspect in question: sometimes individual actors are important for building an understanding of the particular processual dynamics in question. At other times, greater structures of connectedness are crucial, including the partial autonomy they can develop from individual actors (Elias, [1986: 100-101).
Internationalised processes of change are embedded in, and evolve from, pre-existing figurations. These may be states, communities in conflict, or groups of professionals, to name a few. Actors involved in these processes also form specific figurations themselves. To give a concrete example, an initial international TJ figuration emerged in the late 1980s (Arthur, 2009), shaped by knowledge transmission. From there, the knowledge was transmitted to other individuals and figurations and potentially developed or transformed. A TJ figuration is made up of a diverse range of actors. Indeed, it is important not to rule out prematurely who may be participating, because to do so would be to narrow the field of vision of our analysis and, potentially, to reify ontological and epistemological binaries (Danielsson, 2020b).

Projects
The second key concept introduced here is inspired by, but not entirely synonymous with, Jackson's and Nexon's understanding of projects. They define projects as units with agentic features. However, and critically for the framework I develop here, in their understanding a project 'unfolds, or at least tries to unfold, according to some generic plan' (Jackson and Nexon, 1999: 292). 'Projects [as something that should be obtained or achieved] determine what people will count as a gain or loss' (Kennedy, 2016: 67). Thus, for the framework developed in this article, it is useful to label planned processes of change based on policy interventions, as they have been outlined above, as projects. This notion of plan allows us to capture the sense of goal-orientation, without making it the main (or sole) logic on which we base our research.
Labelling these processes as projects is often done regardless, not in an analytical manner, as a particular configuration or unit that represents bundles of processes (Jackson and Nexon, 1999: 308), but rather in a narrower sense, as a breakdown into certain bundles of measures based on financing schemes, implementation agencies, project management logics and so on. However, the latter understanding may be helpful for informing the former, and linking these two perspectives on projects provides critical insights. Krause (2014), for example, identifies the logic of 'good projects' that dominates humanitarian work in global governance. She analyses how international(ised) non-governmental organisations translate their values and interests into practical work. She argues that relief work 'is a form of production' (Krause, 2014: 4) with the 'project' as its main output (mainly produced for donors). In this context, with the focus on management tools, the specific meaning relates to the narrower understanding of projects. Yet, the teleological notion of 'good projects' gives insights on projects in the broader sense. Another example is Schlichte's (2012: 5) 'Cubicle Land' concept of bureaucracy in internationalised rule, in which he defines 'projects' as 'short-term policies stipulating attainable goals ("project purposes")'. He concretely refers to management logics and policy cycles, thus also relating to the narrower understanding of project. However, Schlichte also draws implications for 'projects' in the broader sense by linking his understanding to broader political processes: 'projects' undermine utopias and overarching 'expectational concepts', for example, peace, democracy and justice, that shape internationalised rule (Schlichte, 2012: 14).
To provide a concrete example drawing on my research on the Tunisian TJ process, it is useful to differentiate analytically the 'project' of the planned, institutionalised TJ efforts from the broader notion of TJ as a process of political and societal change. Such a differentiation, even when lines may in fact be blurry, allows for a more nuanced analysis. The understanding of project employed here moves beyond the mere definition of work packages but relates to the configuration or bundle of processes that make up the TJ project -as an initiative unfolding according to a specific plan. In Tunisia, the project (the configurations and bundle of processes) includes professionals, politicians and civil society actors dealing specifically with TJ, dialogue and law-making processes, as well as the establishing institutions, the work of the truth commission, attempts at contestation and so forth. This is different to the understanding of TJ as something that 'happens' in societies and (ideally) leads to peace, democracy and societal reconciliation, without paying attention to its planned component.

Interplay between the planned and unplanned
The proposed framework first suggests that internationalised processes of change are characterised by an interplay between planned/institutionalised and unplanned/spontaneous dynamics (cf. Elias, 1977: 138-139). These interplays most likely lead to a processual outcome that is different to the planned one (cf. Elias, 1977: 148). Change is in part 'driven by institutionalized processes' (Crawford, 2018: 233). Change can therefore be induced through planned socio-technological measures, such as policy-making (or 'engineering' (Barma, 2017: 1)). This characteristic therefore includes the interplay between internationalised 'projects' and processes of political transition. As mentioned above, underlying programming patterns often inform research, because they influence what kind of questions are asked and gear research in a particular direction. However, the framework posits that these planned measures do not by themselves determine whether and what kind of change will actually occur. That emerges only in conjunction with unplanned processes, through spontaneous social and political dynamics.
Looking at the example of internationalised peacebuilding efforts, Barma (2017) identifies the 'interaction between the international interventions and the domestic elites with whom they work' (p. 7). Boege and Rinck (2019) similarly show, through the examples of Bougainville and Sierra Leone with a focus on relationality, how liberal peacebuilding played out in an unplanned manner. Instead of examining resistance against, or adaptation of a (presumably) received international script of what peace missions should look like, the proposed perspective takes the interplay between different actors and their changing power relations and preferences (whether they are domestic or international stakeholders and decision-makers) as an essential part of processual development. Processes can develop immanent dynamics (Elias, 1977: 145) and figurations can achieve partial -but never full -autonomy from individuals and other figurations and, therefore, from individual and collective political preferences (Elias, [1986: 100-101). Consequently, when there are diversions from plans (and the related 'theory of change' 12 ), these 'diversions' are not only unintended consequences (which they may be, from a certain perspective) but are actually essential to the process. Rumours and conspiracy theories play a role alongside what seems 'rational' (Elias, [1986; Jackson and Nexon, 1999: 302).

Non-linearity, simultaneity or 'change does not equal progress'
As a second characteristic of internationalised processes of change, the framework suggests that they are non-linear, reversible and in constant disaccord with counter-processes. Social processes, as Elias (1977Elias ( : 147, 1978 understands them, may indeed be directional, but they are neither automatic nor linear nor non-reversible. Directional does not mean unidirectional. The directions of these processes can even be antipodal (Elias, [1986: 104), shaped by trends and counter-trends. These antipodal directions can be interlinked, and can both be part of the same development, as trends and counter-trends occur simultaneously. Thus, 'change does not equal progress' (Crawford, 2018: 237) and should not, therefore, be analysed solely based on a teleological 'theory of change' (see also Kappler, 2018). Rather, to see what changes and how requires us to pay attention to non-linearities, trends and counter-trends that shape processes of change. Antipodal developments, impulses and multidirectional forces, and opposing trends may be interlinked, and processes of change may be reversible. The non-linear, reversible developments often result from, or go along with, power shifts. These shifts are not necessarily absolute, but can be temporal, relative adjustments subject to change.
While processes of change can move in the same direction at different times, there can also be simultaneous developments in different directions. For instance, in TJ there can be a trend in favour of seeking justice and accountability simultaneous with a countertrend against these efforts. These trends and counter-trends also imply the non-linearity of processes of social and political change and indicate that developments are reversible. Looking concretely at TJ, this means that the norm of dealing with the past, even if it seems to have become a 'societal standard of behaviour', is reversible and that new antagonisms can occur. In the peacebuilding field, Eppert and Sienknecht (2017: 117) found simultaneous processes of securitisation and de-securitisation of the United Nations Mission in Iraq, and they considered these potentially 'mutually obstructive processes'. Eppert and Sienknecht show that this was possible because of a structural decoupling between mandating and implementing organisation. Another example would be de-pacification -and even peace-breaking practices -through which 'the peace process is impeded, shaped, deviated, and transformed in new unexpected directions' (Visoka, 2016: 52).

Conflicts and friction
The third characteristic of the framework suggests that conflict and friction are key drivers for, and defining components of, internationalised processes of change that can be both disruptive and productive (cf. Elias, 1977Elias, , 1978Elias, , 2003Elias, [1987Elias, [1986). They influence power structures and the changes that power structures undergo -and may in turn be triggered by these changes and dynamics. This relationship holds true in particular for TJ and peacebuilding, as these phenomena emerge per se from, and are embedded in, conflict. Conflict plays a particular role in social change and development (Elias, 1978), as social order in a processual view cannot take shape in its absence (Abbott, 2016: 204).
Conflict can usefully be linked with the notion of friction as it has been harnessed by the social sciences in general (notably Tsing, 2005) and for Peace and Conflict Studies in particular (e.g. Björkdahl and Höglund, 2013;Kappler, 2013;Moe and Geis, 2020). Friction should also be understood in a processual sense (Björkdahl and Höglund, 2013), as an aid for analysing 'global connections' across differences (Tsing, 2005) and observing 'how global ideas pertaining to liberal peace are charged and changed by their encounters with post-conflict realities' (Björkdahl and Höglund, 2013: 292). Conflict and friction can be both productive and disruptive: they can slow down change, when power struggles suppress the impulses that drive the change (Elias, 1978: 22), or they can keep things in motion (Tsing, 2005: 5-6). Frictional encounters can also trigger feedback loops, leading to new processes and structures. Processes of change can in turn trigger conflict, as 'affinity and understanding for new ways of speaking and thinking never develop without conflict with older and more familiar ones' (Elias, 1978: 20-21).
These points are crucial for TJ and peacebuilding politics, because a transition towards a new political order may bring power shifts and the reduction of power differentials; after all, challenging existing power structures is one of the main aspirations leading to change and transitions in the first place. Then, as Elias (1977: 139) states, conflict and friction are likely to increase when power differentials are reduced to the benefit of previously weaker actors. This article does not assume that alliances and lines of conflict conform to fixed categories, forming only along a local/domestic-international divide (cf. Danielsson, 2020b). Instead, these dynamics are cross-cutting (see also Kappler, 2013), and some domestic actors may have more overlapping interests with certain international actors than with domestic political rivals.
To sum up, the processual heuristic introduces two key concepts ((con-)figuration and project) and three processual characteristics (interplay between the planned and unplanned; non-linearity, trends and counter-trends; conflict and friction). The two key concepts embrace goal-orientation and relationality, respectively, but an initial focus on processes allows us to see other things (or the interplay of different dynamics) by identifying processual characteristics. I will now provide a selective empirical illustration, drawing on TJ in Tunisia. Although I differentiate between the three characteristics for this purpose and highlight certain instances for their illustration, they are of course entangled.

Empirical illustration
Until the ouster of president Zine el-Abidine Ben Ali in January 2011, Tunisia had been under authoritarian rule since its independence from French colonialism in 1956.
The rule of both Ben Ali and his predecessor, the country's first president Habib Bourguiba, was based on violence and repression. They curtailed civil and political liberties and systematically marginalised the regional strongholds of their political challengers, especially in the country's south and centre. After the fall of the Ben Ali regime, Tunisia quickly began to develop a 'new political architecture', 13 and new political figurations formed. The country elected a National Constituent Assembly (NCA) in 2011, which also functioned as a legislature, and adopted a new constitution in January 2014. It then held free and fair parliamentary and presidential elections at the end of 2014. For the 2011 elections, the decrees regulating the elections functioned as means for vetting: the 'old regime' was excluded from running for office. The moderate Islamist Ennahda emerged as the most popular party and formed a coalition (the Troika) with two smaller secularist parties. In the 2014 elections then, which were not vetted anymore, Nidaa Tounes, a party that absorbed many members of the 'old regime', won the most seats in parliament, as well as the presidency. Transitional struggles mostly concerned the future shape of the Tunisian state, along with questions of inclusion and exclusion and access to power. These struggles materialised in frictional encounters and open conflict within political institutions, for example, within the NCA, as well as protests and violent clashes (most often between police and protesters) 'on the street'.

Interplay between the planned and unplanned
After the fall of the regime in 2011, one could observe a general trend towards justice and accountability, as the country very quickly started dealing with its authoritarian past. TJ was initiated through domestic ad hoc measures, such as military trials for human rights violations carried out during the uprising, civilian trials for economic crimes of previous years, investigation commissions and compensation measures. These measures were followed by a planned TJ project, the policy intervention aimed at fostering peace, democracy and the rule of law, which was designed and established with substantial involvement of international TJ professionals -in particular from the International Center for Transitional Justice (ICTJ) (the most important, globally operating TJ non-governmental organisation (NGO)), United Nations Development Programme (UNDP) and Office of the United Nations High Commissioner for Human Rights (OHCHR). 14 The planned project (at least) partially emerged out of the ad hoc measures. A Tunisian law professor who served on the ad hoc investigation commission tasked with investigating corruption and embezzlement, coordinated the exploration of a potential engagement of the ICTJ in the country, which then prompted the formation of Tunisian civil society associations to engage with TJ. 15 Planning and institutionalisation was advanced by a figuration formed of political and civil society actors, and international TJ professionals. They set up a nation-wide consultation process, during which people also could acquire knowledge necessary to enter the TJ figuration. This culminated in the development of an ambitious, far-reaching TJ law that was passed by the NCA at the end of 2013. The law provided for dealing with almost six decades of violent rule, a broad range of violations, including economic crimes and socio-economic marginalisation, the establishment of a Truth and Dignity Commission, Specialised Chambers in the Tunisian court system and a reparations fund. Such a project corresponded to the dominant approach in professionalised TJ at that time, that is, a 'holistic approach', but only became possible because of transitional political dynamics, the interests of those political elites that were in power at that time, and the (temporary) exclusion of the 'old regime' from political processes (Salehi, 2022b: 103ff). The planned project therefore was shaped by the interplay between dominant approaches in international expertise with (frictional) unplanned political and social dynamics. The latter, therefore, are not merely diversions from plans, but essential part of the process.

Non-linearity, (simultaneous) trends and counter-trends
However, as the TJ project developed, it was simultaneously undermined by a lack of willingness to dismantle repressive structures at a deeper level. The aforementioned conflicts in parliament and 'on the street' brought the constitution-making process to a halt. There was, therefore, a simultaneous trend towards elite deal-making, which prioritised short-term conflict-resolution over longer term change and TJ. One of my interview partners described this simultaneity of trends and counter-trends with regard to the 2013 National Dialogue -the conflict-resolution forum that broke the deadlock in the constitution-making process and for which four Tunisian civil society organisations jointly received the 2015 Nobel Peace Prize. On the one hand, 'the Quartet [of organisations that initiated the National Dialogue] tried to ignore transitional justice', 16 and the logic of elite deal-making that the Dialogue was based on closed off avenues to further structural change because it led to a re-inclusion of the 'old regime' in politics. Although the support for justice and accountability of those political actors that had pushed for the development of the planned TJ project was not abandoned outright, it did become less consistent through the rapprochement with the 'old regime'. 17 On the other hand, the compromise deal that they had made helped pave the way for further institutionalisation of a (conflictive) TJ project by fixing it in the constitution.
In a changing political environment, political (con-)figurations, preferences and positions of power shifted. After the 2014 elections, as one of my interview partners put it, 'the governments included members of the old regime. Transitional justice is against their own interests. So transitional justice lost its flame'. 18 The institutionalised TJ project faced substantial challenges from the political sphere: What actually happened is that the system that ruled Tunisia since 2014 tried to block the transition process and transitional justice. The government issued a clear order: All departments of the state were not to deal with the Truth and Dignity Commission. 19 This pressure obliged the (international and domestic) TJ professionals, who supported the planned project, to take a stance against the government. In general, one of my interlocutors reflected that 'just providing technical assistance' is not possible 'when you are pushing for a human rights agenda '. 20 Nevertheless, in this interplay between the planned, internationalised TJ project and (unplanned) social and political dynamics running counter to it, the process developed (to some degree) an immanent dynamic -a limited degree of independence from shifting domestic power structures and political preferences. TJ was continuously performed even though there were efforts to obstruct the process. To give one concrete example, public hearings were held despite determined political opposition from the government and the presidency. 21 These public hearings were described to me by an interview partner as 'the best gain' 22 achieved by the truth commission.

Conflict and friction
In the previous sections, we saw that conflicts and frictions between different political actors and 'on the street' characterised the process, as well as between the TJ institutions and the political sphere. TJ itself generated new frictions and conflicts that were both productive and disruptive, through its immanent dynamic and continuous performance as well. In addition, conflict and friction -but also alliances -do not need to be linear, 'horizontal' or 'vertical'. They may cut across different actor groups and be subject to change. For example, though TJ professionals initially worked with both political actors and civil society to initiate and design the planned TJ project, frictions emerged between some parts of civil society and TJ professionals when civil society actors believed that these professionals were not advocating forcefully enough for civil society's inclusion in the project. 23 Then again, as mentioned above, with declining political support for the TJ project, the cooperation between political actors and TJ professionals became more conflictive. (For a more detailed empirical analysis of the characteristics see Salehi, 2022a).

Conclusion
The article proposed a processual heuristic for analysing liberal policy interventions in conflict contexts. Drawing on examples from TJ and peacebuilding, it showed a certain goal-orientation and a teleological appeal in the practice that is reflected in some strands of research. Thus, a focus on substance dominates, which may obscure dynamics that would better be explained by focusing on processes. From a goal-oriented perspective, there is often little left to say, except that interventions did not work out as planned, although a lot of things happened. A processual approach allows us to research 'what has been going on here?' by analysing the processual characteristics of these internationalised processes of change in conversation with the dynamic context. It therefore complements existing research on liberal policy interventions by providing a novel entry point, allowing for a broader research approach, and paying closer attention to the politics that shape processes of political transition.
The framework proposed here developed, and briefly illustrated, three initial processual characteristics for empirical research: the interplay between the planned and unplanned; non-linearity, simultaneous trends and counter-trends; and conflict and friction. Open empirical research could identify other, or additional, processual characteristics that could further contribute to the conceptual development of internationalised processes of change.
Sienknecht. I presented earlier versions of this paper at the DVPW International Relations section meeting 2020, the ISA Annual Convention 2021, the EISA Pan-European Conference 2021, as well at the WZB's Global Governance colloquium and at the Marburg team colloquium. I thank all participants for productive discussions, in particular the discussants Tobias Berger and Vincent Durac. I thank Robert Benson for careful language editing.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the German Ministry of Education and Research [grant number 01UC1301] and a postdoc fellowship of the A.SK Foundation.

Supplemental material
Supplemental material for this article is available online.

Notes
'people' she is talking about. Although it becomes clear in her case studies that she does not have a uniform, apolitical vision of society, I would like to emphasise again that people may want to build different kinds of societies. 9. For a critical account on approaches based on complexity theory and systems thinking that aim to approve peacebuilding see Bächtold (2021), who argues that they undermine accountability to those targeted by interventions. 10. This differs from the dominant 'realist' view on case studies, which 'positions the researcher as an outside observer who identifies and selects from cases made available by the real world' (Soss, 2018: 21). This is also the perspective that, according to Bennett and Checkel (2014), resonates better with 'classical' process tracing. Thus, processual analysis with a nominal view on 'casing a study' needs to rely on a different approach than process tracing.