Elucidating the Social -- Developing Social Process Tracing as an Integrative Framework
Saved in:
| Title: | Elucidating the Social -- Developing Social Process Tracing as an Integrative Framework |
|---|---|
| Language: | English |
| Authors: | Jonas Gejl Kaas (ORCID |
| Source: | Sociological Methods & Research. 2025 54(2):605-645. |
| Availability: | SAGE Publications. 2455 Teller Road, Thousand Oaks, CA 91320. Tel: 800-818-7243; Tel: 805-499-9774; Fax: 800-583-2665; e-mail: journals@sagepub.com; Web site: https://sagepub.com |
| Peer Reviewed: | Y |
| Page Count: | 41 |
| Publication Date: | 2025 |
| Document Type: | Journal Articles Reports - Evaluative |
| Descriptors: | Qualitative Research, Research Methodology, Social Science Research, Sociology, Social Environment, Influences, Epistemology |
| DOI: | 10.1177/00491241231220732 |
| ISSN: | 0049-1241 1552-8294 |
| Abstract: | The promise of process tracing methods is that they can help us better understand how things work in real-world cases. Despite the many advances in the past two decades, we contend that existing accounts result in either under-socialised accounts in which the moves made by actors are studied while neglecting the social dimension of action, or over-socialised accounts that are so focused on social context that they neglect the causal process of actors doing things. In this article, we attempt to bridge this divide by merging the conceptual language from the productive/generative account of mechanisms with realist ideas to develop what we term social process tracing (SPT) as a practical research method that captures both the moves performed by actors, the meanings they ascribe to them, and how the underlying social context shapes how social causal processes play out in real-world cases. After developing ontological and epistemological foundations, the article illustrates how SPT can be used in practice. |
| Abstractor: | As Provided |
| Entry Date: | 2025 |
| Accession Number: | EJ1473583 |
| Database: | ERIC |
|
Full text is not displayed to guests.
Login for full access.
|
|
| FullText | Links: – Type: pdflink Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwGSovdZTwXNfA4liTuwx0q7AAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDELKeKur3OJRWtzzvQIBEICBm0kcqvoDZnrAAi9_GjaTBnNF2HrEK6A73G6uwenNpJQ9fwCs6BRDhz5TOoNa--xB4zzuMqHL5gPCZR_oFWQ_iIS13CVWHpJD-q_Cauu8pKnXD9raFPsk8C6XY3V7Uxaq67VYrH2qomXZ5Y2T158sZZ55jY8w0f075xpQw6aLu0xnvS5s5crcgklFkJQKowSA4sYUcIwFc7RLbOHl Text: Availability: 1 Value: <anid>AN0184233958;som01may.25;2025Apr07.05:58;v2.2.500</anid> <title id="AN0184233958-1">Elucidating the Social – Developing Social Process Tracing as an Integrative Framework </title> <p>The promise of process tracing methods is that they can help us better understand how things work in real-world cases. Despite the many advances in the past two decades, we contend that existing accounts result in either under-socialised accounts in which the moves made by actors are studied while neglecting the social dimension of action, or over-socialised accounts that are so focused on social context that they neglect the causal process of actors doing things. In this article, we attempt to bridge this divide by merging the conceptual language from the productive/generative account of mechanisms with realist ideas to develop what we term social process tracing (SPT) as a practical research method that captures both the moves performed by actors, the meanings they ascribe to them, and how the underlying social context shapes how social causal processes play out in real-world cases. After developing ontological and epistemological foundations, the article illustrates how SPT can be used in practice.</p> <p>Keywords: process tracing; social process tracing; interpretivist methods; multimethod research; causal process; causal mechanisms</p> <hd id="AN0184233958-2">Introduction</hd> <p>How can low intelligence capacity of security forces produce mass violence against civilians ([<reflink idref="bib63" id="ref1">63</reflink>])? How can an epistemic community gain influence over policy ([<reflink idref="bib39" id="ref2">39</reflink>])? The promise of process tracing (PT) methods used in these and many other studies is that they can help us move beyond correlational claims through tracing the causal processes that link causes and outcomes together, enabling us to better understand how things work in real-world cases.[<reflink idref="bib7" id="ref3">7</reflink>] PT methods have come a long way in the past two decades, with significant developments in our knowledge of both what we are tracing at the theoretical level, as well as advances in our understanding of how we can trace causal processes empirically. Yet despite their many merits, existing PT methods run into theoretical and empirical challenges when we try to take the social dimension of causal processes seriously. We contend in this article that existing PT methods result in either under- or over-socialised accounts of causal processes.</p> <p>Most existing approaches to PT can be broadly categorised as positivist due to the emphasis on testing the correspondence between theoretical hypotheses and empirical observables in the form of events (e.g., [<reflink idref="bib4" id="ref4">4</reflink>]; [<reflink idref="bib6" id="ref5">6</reflink>]; [<reflink idref="bib11" id="ref6">11</reflink>]; [<reflink idref="bib24" id="ref7">24</reflink>]; [<reflink idref="bib40" id="ref8">40</reflink>]). In these approaches, there is little difference between studying the natural and social worlds. This means that social aspects of interaction between actors and how the underlying social context shapes how processes work are downplayed to varying degrees, resulting in what can be termed 'under-socialised' analysis. Yet when studying social processes, we are studying reasoning human beings who try to make sense of their interactions with other social actors ([<reflink idref="bib13" id="ref9">13</reflink>]: 20, 33; [<reflink idref="bib46" id="ref10">46</reflink>]: 159; [<reflink idref="bib50" id="ref11">50</reflink>]: 23). Observing the physical manifestations of action in the form of sequences of events merely tells us what happened, but not why.[<reflink idref="bib8" id="ref12">8</reflink>] Using a metaphor, only studying the observable moves of actors in social causal processes is like watching a group of people moving pieces on a game board without knowing what game they are playing and why they move pieces as they are doing. To understand and evidence <emph>why</emph> actions occurred and <emph>why</emph> they had the effects they had, we also need to probe the social context for action and how social actors understood action, tapping into the intentions behind the actions performed by actors in a particular social context – in effect studying the <emph>meaning</emph> behind actions ([<reflink idref="bib1" id="ref13">1</reflink>]: 195; [<reflink idref="bib17" id="ref14">17</reflink>]: 307; [<reflink idref="bib44" id="ref15">44</reflink>]; [<reflink idref="bib46" id="ref16">46</reflink>]: 160–1).</p> <p>There have been attempts to develop an interpretivist approach to PT, for instance by trying to understand patterns of action that are embedded in a social context as practices ([<reflink idref="bib46" id="ref17">46</reflink>]), or how capacities for action exist in a particular social context ([<reflink idref="bib41" id="ref18">41</reflink>], [<reflink idref="bib42" id="ref19">42</reflink>]). However, interpretivist PT approaches tend to result in more structural accounts that lose sight of the interactions and relations between social actors that are the causal drivers of processes. Because social structure then determines action, these approaches can be categorised as being 'over-socialised' analyses that downplay the process itself and what actors are doing therein (see also [<reflink idref="bib54" id="ref20">54</reflink>]). Returning to the moves/game metaphor, focusing on social context and how it shapes action is like reading the rulebook for a game without seeing the game itself being played.</p> <p>In this article, we attempt to bridge the divide between positivist and interpretivist PT approaches by developing Social PT (SPT) as a distinct methodology that should be thought of as a friendly amendment to both positivist and interpretivist accounts of PT. In theorising processes, we build on the positivist PT account of [<reflink idref="bib4" id="ref21">4</reflink>] and their focus on actors and their actions. We supplement this with conceptual language drawn from critical realism, where social structure is taken into account when theorising, while not losing sight of agency and processual dynamics ([<reflink idref="bib9" id="ref22">9</reflink>]; [<reflink idref="bib13" id="ref23">13</reflink>]; [<reflink idref="bib46" id="ref24">46</reflink>]; [<reflink idref="bib50" id="ref25">50</reflink>]). Critical realism on its own lacks a conceptual language for theorising causal processes themselves, focusing instead on underlying, unobservable mechanisms and context, with the result that critical realist empirical analyses of processes tend to postulate combinations of context and mechanisms that account for an outcome without actually evidencing the causal linkages in-between causes and outcomes. At the same time, the language about actors and actions in broadly positivist PT methods (see e.g., [<reflink idref="bib4" id="ref26">4</reflink>]) tends to black-box the <emph>causal linkages</emph> that bind the actions of actors together. Merely theorising that actor A (an expert) puts forward a report that diagnoses a policy problem, and actor B (a policy-maker) responds by putting forward a policy solution that tackles the diagnosed problem does not explain why the action of actor A leads to behavioural change in actor B. The causal linkage in this example is less explicit than it could be because we are not told <emph>why</emph> actor B would accept the diagnosis of actor A. To answer why one action led to response from another actor we would need to unpack the intentions, motivations and meanings underlying actions (O'Mahoney 2015; [<reflink idref="bib22" id="ref27">22</reflink>]). This is not just a theoretical problem, for if the causal linkages are not theorised explicitly, they cannot be evidenced empirically. To rectify this, we draw on ideas from critical realism to help us fill in this black-box both theoretically and empirically. In critical realism, actor motivations for action and how they perceive the actions of others are a key part of explaining why actions have the impact they have. Critical realists supplement theorisation about intentions and perceptions with other underlying mechanisms that help bolster the explanation for why a given action leads another actor to do something in response – thereby explaining a causal linkage.[<reflink idref="bib9" id="ref28">9</reflink>]</p> <p>At the epistemological level, critical realism offers an <emph>analytical dualist</emph> position that promises the benefits of combining observational evidence of interactions between actors and the importance of material structure with more interpretivist evidence that enables the analyst to tap into the 'social' dimensions of action and social structure (context). However, in practice critical realists tend to neglect the more observable, material dimensions of action and structure in their analyses (see e.g., [<reflink idref="bib15" id="ref29">15</reflink>]). As a consequence, we draw on positivist PT approaches for guidance on how to trace the observable manifestations of actors engaging in actions and the impact of the material dimension of structure. In order to enable both approaches to be used together, we suggest that the two different approaches are used in a supplemental fashion in which they help answer different questions as regards social causal processes: (<reflink idref="bib1" id="ref30">1</reflink>) positivist PT methods are used to ask what happened (<emph>what</emph> 'moves' were made), and what were the material roles and resources available to actors (the material dimension of the 'game'), and (<reflink idref="bib2" id="ref31">2</reflink>) interpretivist methods as used by critical realists are used to ask why actions were linked together (<emph>why</emph> 'moves' were made), and what the impact of social structure (<emph>how</emph> actors <emph>understood</emph> the 'game' being played).</p> <p>Note that we are not arguing for completely re-inventing the PT wheel. Indeed, there are many excellent published applications of PT that <emph>implicitly</emph> include aspects of the social dimension of action and context. SPT aims to make explicit at the theoretical level what is often taken for granted or black-boxed in many applications of PT as regards the impact of the 'social.' Moreover, when aspects of the social dimension are implicit in a theory, they are typically not evidenced empirically. SPT provides an analytical dualist position that can help scholars elucidate explanations for causal linkages and gather empirical evidence of social processes and understand how actors' understandings of the underlying 'game' being played impact how processes work.</p> <p>Before moving further, it is important to emphasise that this is a methodological article and not an exercise in social theory. Our intention is to develop SPT as a research method that is compatible with many different social science theoretical perspectives. As will become clear in the following, the theoretical language of SPT pushes us in the direction of theories in which both structure (social context) and process (action and agency) are important. But beyond this, many different mid-range social theories are compatible with SPT methods. For instance, if we are studying policy learning processes, a social process theory might build on theoretical ideas related to bounded rationality and risk adverse policy-makers in which actors learn through incremental trial-and-error. Or we might draw on social psychological theories related to epistemic authority to develop a social process theory in which policy-makers quickly recognise problem diagnoses provided by epistemic authorities (socially recognised experts). For both social process theories, we would want to unpack both what actors were doing, what intentions they had in performing these activities, and the underlying social context that enabled or constrained them.</p> <p>The article proceeds in three steps. We start with a brief overview of how the social dimension of processes has been treated in the existing PT literature, showing they result in either under- or over-socialised approaches. Afterwards, we develop a theoretical language to conceptualise the social aspect of processes and outline an analytical framework for tracing social causal processes that combines positivist PT with interpretivist methods. The article then illustrates how SPT can be used in practice by reconstructing what a published PT article would have looked like if SPT had been used. The article concludes with a discussion of how social processes can be 'generalised' and what is required to do so.</p> <hd id="AN0184233958-3">Existing Approaches to Process Tracing: Either Under- or Over-Socialised Methods</hd> <p>In this section, we review the existing PT methodological literature,[<reflink idref="bib10" id="ref32">10</reflink>] showing that both positivist and interpretivist approaches lack a conceptual language that enables analysis that explicitly captures the social dimension of causal processes while at the same time not losing sight of the importance of action and agency. Existing approaches as a consequence produce either under- or over-socialised accounts.[<reflink idref="bib11" id="ref33">11</reflink>] We contend that the lack of a good conceptual language for process that elucidates the social dimension has epistemological implications, for if we cannot unpack <emph>social</emph> processes theoretically, how can we evidence them empirically? Further, without understanding the social aspects of social causal processes, we have difficulties in making explicit the causal linkages in the process themselves – and when they are not explicitly theorised, they tend to not be studied empirically, meaning that many approaches to PT end up producing accounts of events in sequence.</p> <hd id="AN0184233958-4">Positivist Process Tracing – Missing 'the Social'</hd> <p>Common to what can broadly be characterised as positivist approaches to PT is the ambition to study the process linking causes and outcomes together (e.g., [<reflink idref="bib4" id="ref34">4</reflink>]; [<reflink idref="bib6" id="ref35">6</reflink>]; [<reflink idref="bib24" id="ref36">24</reflink>]; [<reflink idref="bib62" id="ref37">62</reflink>]).[<reflink idref="bib12" id="ref38">12</reflink>] At the ontological level, existing positivist PT approaches can be differentiated into more 'minimalist' and 'disaggregated' treatments of processes at the theoretical level. Minimalist approaches conceptualise process in very abstract terms, often as intervening or mediating variables, whereas disaggregated approaches contend that we should unpack a causal process into its constituent parts composed of actors engaging in actions.</p> <p>We start with minimalist approaches. In the minimalist PT approach, as put forward by [<reflink idref="bib24" id="ref39">24</reflink>] but also others (e.g., [<reflink idref="bib6" id="ref40">6</reflink>]), causal processes at the theoretical level are treated as a form of intervening variables. Goertz for instance depicts theories very abstractly as X→M→Y. He describes [<reflink idref="bib28" id="ref41">28</reflink>] 'more elaborate' process theory that links economic inequality with a democratic transition as:</p> <p>In the example depicted in Figure 1, what actors are doing in the process is not unpacked due to the high level of theoretical abstraction. While the theory suggests that the outcome will occur whenever the three factors are present, the theory itself is so abstract that it tells us next to nothing about what the process would look like in any given case. We are not told <emph>who</emph> is involved in mass mobilisation, <emph>what</emph> these actors are actually doing in a causal sense, nor are we told about the interactions between elites and other societal actors engaging in mobilisation beyond that it leads at one point to acquiesce by elites. The 'social' dimension of the processes remains fully within a black-box.</p> <p>Graph: Figure 1. Goertz' depiction of Haggard and Kaufman's causal processes. Source: [<reflink idref="bib24" id="ref42">24</reflink>]: 50).</p> <p>At the empirical level, given that the abstract process theory tells us next to nothing about any given case beyond that organisational capacity is important, all that is required for 'confirmation' of the theory in any given case is that some form of grievance-based mobilisation took place, and that it is followed by elite acquiescence. In the book, the authors describe a set of empirical events for each 'case', and based on their event history, they claim that there is evidence of similar processes in transitions in such different contexts as Argentina, the Baltics, and Niger. However, in all of the cases it is evident from their descriptive event histories that there are very different actors engaging in very different forms of mobilisation in very different contexts in each case. In some cases like Argentina, unions were the key actors, with interactions between union leaders and members in organising strikes and protests. In other cases like in the Baltics, ethnonationalist actors played a dominant role in mobilisation, where the key social interactions were between the ethnonational leaders and followers and the foreign 'other'. In other cases like Niger, state worker unions played a key role in mobilising demands for change <emph>within</emph> government. No explanation is put forward for these differences in who is doing what, and there is little explanation for why events are linked together in a causal sense. Given the very high level of abstraction of process theories used in minimalist approaches to PT (see appendix 1 supplemental material also), the analytical result is a relatively shallow theory that leaves us in the dark about both the social interactions and relations between actors in social causal processes, and how the social context of a case shapes how the process plays out in any given case.</p> <p>The disaggregated approach to tracing causal processes put forward by [<reflink idref="bib4" id="ref43">4</reflink>] draws on conceptual language from the productive/generative account in the philosophy of science to unpack mechanisms into parts composed of entities (aka social actors) and activities (e.g., [<reflink idref="bib40" id="ref44">40</reflink>]).[<reflink idref="bib13" id="ref45">13</reflink>] However, the approach focuses only on observable activities and their hypothesised impact on other actors, evidenced using the traces left by activities in a case ([<reflink idref="bib4" id="ref46">4</reflink>]: 4–6, 55–6). At the theoretical level, the social dimension of action is not made explicit as it could be, especially in terms of how participants perceive the 'game' they are playing, why they perform particular actions, and how they perceive the actions of others. Yet without studying the <emph>meaning</emph> of actions and how they are embedded in a particular social context, we come short when trying to explain why actions were causally linked to responses from other actors that together compromise a social causal process (O'Mahoney 2015). A critical part of the causal explanation for the linkage – and thereby also the empirical evidence – that is missing here is how the 'moves' (activities) made are <emph>interpreted</emph> by actors and those impacted by them, and how these actors <emph>understand</emph> the underlying social 'game' they are situated in. Taken as a whole, reducing the actions of social actors to their mere physical empirical manifestations results in under-socialised accounts that have difficulties in evidencing the causal linkages in a process.</p> <p>For instance, [<reflink idref="bib63" id="ref47">63</reflink>]) develops a disaggregated process theory of actors engaging in activities for a process that is triggered by low intelligence capacities of security forces and leads to mass categorical violence against particular groups. The article is arguably one of the best examples of 'disaggregated' PT published to date. In the theorised process (depicted in Figure 2), the cause (low intelligence capacity and unrest) spurs the security forces to <emph>approach</emph> local civilian elites for assistance in gathering information. The local civilian elites then <emph>exploit</emph> this dependence to settle scores in relation to pre-existing local conflicts with a particular group by <emph>providing false information</emph> targeted against individuals from the group. The security forces <emph>use</emph> the (false) information provided to <emph>detain</emph> and <emph>interrogate</emph> individuals from the targeted group, resulting in an escalating cycle of torture and violence in which the many false confessionals derived from torture lead to even more detained and tortured individuals. Further, an increase in the number of detainees strains the capacities of the security forces, leading them to take extreme steps like extra-judicial killings to clear out prisons, resulting in the outcome of mass categorical violence.</p> <p>Graph: Figure 2. Winward's collaboration with elites process theory. Source: Adapted from Winward (2021: 558–562). Explanation: Actors are depicted as boxes with underlined text, actions as italic text.</p> <p>While making actors and actions explicit for key interactions is a good starting point, theorising a social <emph>causal</emph> process is more than depicting a sequence of actions in a chronological narrative, in which temporal succession can be conflated with the actions being linked in a causal sense ([<reflink idref="bib50" id="ref48">50</reflink>]: 141). Returning to the move/game metaphor, actors and actions are merely the moves made in the game. Instead, theorising a social <emph>causal</emph> process requires <emph>explanations</emph> of <emph>why</emph> an action is <emph>linked</emph> in a productive causal relationship with another social actor doing something in response. Therefore, a good <emph>social</emph> process theory should explain <emph>why</emph> an action triggers another action in response by making explicit the causal linkage between actions and responses. In the Winward example, the process theory as described in the article does not fully explain the linkages. For instance, why do local elites provide false information targeted against particular groups? Why would security forces uncritically accept the information without seeking to verify it with other sources? The causal aspect of the processual explanation could be strengthened by theorising <emph>explicitly</emph> the <emph>causal linkages</emph> in the process that can explain why actors act and respond in particular ways. To do this requires that we also attempt to understand the game being played and how this shapes the moves that actors make and how others respond. Making causal linkages more explicit also improves our ability to make causal inferences because analytical attention is directed towards evidencing the linkages, at least indirectly (O'Mahoney 2015).</p> <hd id="AN0184233958-5">Interpretivist Process Tracing – Missing the 'Process'</hd> <p>There have been several attempts to develop more interpretivist versions of PT. Indeed, many of their claims about the importance of the 'social' resonate with the arguments in this article, especially about the need to draw on evidence about how actors understand a particular social context (e.g., [<reflink idref="bib41" id="ref49">41</reflink>]: 35; [<reflink idref="bib46" id="ref50">46</reflink>]). However, existing interpretivist PT approaches tend to become over-socialised accounts due to a lack of conceptual language for action and process, and in particular, they tend to miss the <emph>interactions</emph> between actors that drive causal processes forward.</p> <p>[<reflink idref="bib46" id="ref51">46</reflink>] develops what he terms 'practice-tracing', in which practices are defined as habitual patterns of actions performed by actors. Echoing the point about the meaning of social action, Pouliot argues compellingly for adding an extra layer of 'meaningfulness' by studying the stock of unspoken assumptions and tacit know-how of practitioners to understand their intentions and beliefs as regards sets of practices that they perform, using interpretivist discourse or ethnographic observation methods. At the same time, by focusing on understanding the meaning of practices, actions become separated analytical from the causal processes themselves that link causes and outcomes. Not surprisingly, the approach lacks conceptual language for causal process and the actions and interactions therein, which means they cannot be evidenced empirically</p> <p>[<reflink idref="bib41" id="ref52">41</reflink>], [<reflink idref="bib42" id="ref53">42</reflink>]) puts forward an interpretivist variant of PT in which causal processes are nested in what he terms 'constitutive' explanations. The constitutive element includes the structural conditions, latent dispositions and causal capacities that come from a given social context, whereas the causal element is the moving parts. Norman defines causality as the relations between events, focusing on '...how specific actions and events can alter the context in which they appear' ([<reflink idref="bib41" id="ref54">41</reflink>]: 87). However, he then suggests that interpretive PT uses what he terms 'contrastive' explanation that build on counterfactual comparisons. These can include before/after comparisons, or 'normal' events compared with an 'abnormal' event that changes the trajectory of developments. He suggests a three-step analytical procedure that:</p> <p></p> <ulist> <item> maps change in the broader institutional setting, for example by asking actors of 'normal' actions, and then what constituted the 'abnormal' and thus conflict between actors at a later time ([<reflink idref="bib42" id="ref55">42</reflink>]: 19);</item> <p></p> <item> explores how change reconstituted actor self-understandings</item> <p></p> <item> and how the changed self-understandings gave rise to particular actions</item> </ulist> <p>In the descriptions of the procedure and application of it, there is a focus on timing and sequences of events, but as with many other PT approaches, he does not unpack conceptually what links events together in a causal sense. For instance, he discusses a 'rupture' mechanism that explains how '...shifts in the self-understandings of agents prompt them to interpret their environments in new ways' ([<reflink idref="bib41" id="ref56">41</reflink>]: 87). The mechanism (i.e., causal process) is depicted in Figure 3, below.</p> <p>He writes that, 'The <emph>rupture mechanism</emph> explains the shift through which habitual or scripted action becomes reflected upon and a matter of contestation. Structure from this point of view moves from connoting shared understandings and taken-for-granted ways of acting to representing itself to specific actors in terms of opportunities and constraints' ([<reflink idref="bib41" id="ref57">41</reflink>]: 88). While he does discuss agency of actors (p. 86), the actual theorisation and empirical analysis focuses on changes in context, resulting in an over-socialised explanation in which what social actors are actually doing is neither theorised nor evidenced empirically. This prevents us from understanding why a clash leads to the emergence of particular institutional preferences, or why these preferences would lead to institutional conflict.</p> <p>Graph: Figure 3. Norman's 'rupture' mechanism. Source: [<reflink idref="bib41" id="ref58">41</reflink>]: 88).</p> <p>In order to develop PT accounts of social causal processes that avoid becoming either under- or over-socialised, and that can explain and evidence the causal linkages in a process, we need to build on the strengths of existing PT approaches, supplemented with insights from critical realism to develop a conceptual language that enables us to make sense of activities and context, and empirical methods that enable us to both understand what is going on and why. It is to this task that we now turn.</p> <hd id="AN0184233958-6">The Foundations of SPT Methods</hd> <p>In this section, we develop a conceptual vocabulary for SPT that enables the theorisation of social causal processes in ways that focus analytical attention on what actors are doing and the causal linkages between their actions, while at the same time taking the social dimension of context seriously. At the theoretical level, the conceptual language we develop in SPT builds on the disaggregated approach from positivist PT, in which causal processes are composed of a series of entities (actors) and activities (actions) that provide the productive relationship between a cause to an outcome. To theorise causal linkages and the importance of social context, we supplement the actors/actions language with critical realist language about the importance of social context, and the meaning and intentionality behind actions.</p> <p>We then discuss the epistemological implications of making the social explicit in theorising causal processes in SPT. The epistemological starting point is the critical realist argument that studying the social world is fundamentally different from research of the natural world because social actors produce and reproduce the social world within which their actions and relations are embedded ([<reflink idref="bib9" id="ref59">9</reflink>]; [<reflink idref="bib50" id="ref60">50</reflink>]; [<reflink idref="bib46" id="ref61">46</reflink>]; [<reflink idref="bib13" id="ref62">13</reflink>]). We also follow the critical realist rejection of a strong constructivist position (e.g., [<reflink idref="bib8" id="ref63">8</reflink>]), in which there is no material reality that is not constructed by actors. Instead, we suggest that SPT is most in alignment with a moderate social constructivist position that ascribes causal significance to social meaning while also accepting a 'reality' out there independent of our knowledge of it, thereby acknowledging that material structures also shape action and interactions (e.g., [<reflink idref="bib31" id="ref64">31</reflink>]: 6). We develop an analytical dualist methodology for SPT that enables tracing of interactions between actors in a way that evidences the causal linkages in a causal process, thereby enabling causal inferences. Further, the methodology can arguably better capture the importance of both the material and ideational dimension of social context of action because it is theorised explicitly. To develop a practical methodology for SPT, we provide a set of questions that act as the analytical framework for tracing social processes.</p> <p>Note that a social process theory is always an abstraction from the almost infinite complexity of any particular case. The essence of the causal process is distilled in more abstract form from the empirical record in terms of who (key actors), what and how (key activities and causal principles linking them together) during key episodes. For all but the simplest of causal processes, every case is strictly speaking 'unique'. However, this does not mean that similar social processes might not be at work in other cases in relatively similar contexts. What is required when theorising a process is that for each key episode in the process, the actors, actions and linkages ('moves') and context (the 'game' being played) are abstracted enough from case specifics that we can think in terms of <emph>functional equivalences</emph> between what actors might do in similar (but not identical) case contexts. Instead of describing the particular actor by name, we would use a more abstract categorisation such as 'expert' that could play a functionally equivalent (though not identical) role in another case in a similar context (i.e., 'game'). Actions would also be abstracted from the particular by using more abstract terms such as 'offering incentives', which might involve money in one context, and public recognition in another. For both, a similar motivation might underly the actions. In the example of [<reflink idref="bib39" id="ref65">39</reflink>] article on expertise and policy influence, one part of the process is that an epistemic community (actor) promotes their favoured policy (action). In the Polish and Czech cases that she investigates, the identity of key actors in the epistemic community are somewhat different, and the actions they take in promoting are also slightly different (in Poland educational trainings were held, whereas in the Czech Republic conferences and talks were held). But based on the detailed case knowledge of the author, she asserts that the actors and actions are functionally equivalent in causal process terms in the two cases.</p> <p>A good social process theory is simple enough that it only details key episodes, but at the same time includes enough processual attributes (who, what, when and how) that it is possible to establish whether the process worked in similar or different ways in other cases. If an initial case study confirmed the existence of the social process, another relatively similar case can be selected. The process theory will then have to be 'particularised' in relation to the empirical context of the new case, with the scholar deploying deep case knowledge of both cases to assess whether there was functional equivalence at the process level in terms of who, what and how in key episodes linking the cause and outcome together. We return to this question in the conclusion of this section.</p> <hd id="AN0184233958-7">Theorising Social Causal Processes</hd> <p>We define social causal processes as a sequential series of episodes composed of actors (the who) engaging in actions that bind a cause (or set of causes) (the what and why they are linked) and an outcome together in a productive (i.e., generative) relationship. Note also that social processes do not have to be necessarily linear, and there can be feedback loops and more complex forms of interaction. We also suggest focusing on the <emph>key episodes</emph> in a social process. These are defined as being particularly crucial from a causal perspective, and where we have theoretical reasons to expect that the process might have played out differently if different actions had been taken, or the context was different ([<reflink idref="bib58" id="ref66">58</reflink>]: 88–92). Using a metaphor, when making a movie, writers and directors spend considerable time thinking about where to spend precious movie time and special effects, prioritising scenes that are viewed crucial and exciting for the storyline. Similarly, when theorising a social causal process, there are particular episodes of interaction that will be more interesting in causal terms because they are particularly surprising from a theoretical/empirical perspective. Returning to the Winward example (Figure 2 above), key episodes include security forces approaching potential collaborators, and the civilian elites providing false information. Without these interactions, the process would arguably have broken down.</p> <p>Actions are what actors do in a process; they are what binds them to other actors in a social process. To be part of a causal process, the action of one actor triggers an action in response from other actors. This means that action is understood in <emph>relational</emph> terms. Actions can be speech acts (speech or statements in conversations), or physical actions such as striking or putting forward a proposal in a negotiation. Social <emph>actors</emph> can be individuals or collective. The only condition for theorising collective actors is that the later have properties and orientations that enable them to <emph>do things</emph> that can impact other actors in a process, and that their activities as a collective cannot be reduced to the sum of the individuals (i.e., there is emergence). Individual social actors could be policy entrepreneurs or individual decision-makers, whereas collective actors might be epistemic communities or states. We now turn to developing a conceptual language for making causal linkages more explicit in social causal processes than they are in existing PT approaches.</p> <hd id="AN0184233958-8">Causal Linkages – Context, Intentions and Perceptions, and Causal Principles</hd> <p>Causal linkages can be thought of as explanations of an 'if-then' type linkage between input (action(s) of one actor) and output (response(s) from another actor or actors). They can be thought of as 'because' clauses, in abstract terms Actor A did Action X 'because...' (causal linkage). The conceptual language we use for theorising causal linkages is inspired by critical realist understandings of context and mechanisms that for them are the foundation for understanding social behaviour. Critical realists emphasise the importance of context, and in particular how it shapes the motivations behind action and how actors respond. Critical realists talk about the micro-foundations underlying actions and linkages as 'mechanisms', defined as, '... that which can cause something in the world to happen and in this respect the mechanisms can be of many different kinds' ([<reflink idref="bib13" id="ref67">13</reflink>]: 47). Given that it can be confusing to talk about mechanisms as micro-foundations when many process tracers use the term mechanism for the aggregate causal process linking causes and outcomes, we utilise in this article the terms <emph>intentions/perceptions</emph> to refer to what critical realists understand as the motivational mechanisms linking actions together. Further, we follow [<reflink idref="bib10" id="ref68">10</reflink>] in using the term <emph>causal principles</emph> for all other explanations for how actors behave in particular contexts that can provide the causal linkages in a process (i.e., what critical realists term all other types of mechanisms). Returning to the Winward example, civilian elites providing false information describes who and what took place, but not why it is causally linked with the previous part. Here we could unpack the motivations behind civilian elites providing this information to understand why actions were linked in a causal sense.</p> <p>Before we proceed, it could be questioned whether it is necessary to theorise causal linkages explicitly. Here we contend that being as explicit as possible about the hypothesised causal linkage has both theoretical and empirical benefits. First, trying to make linkages explicit makes the causal explanation more open to logical scrutiny. Second, trying to make linkages explicit helps the analyst identify different (and often unexpected) theoretical literatures that can be drawn in to provide ideas about what causal dynamics can explain the linkage between actions and responses. Third, making the linkage as clear and specific as possible enables us to evidence it, at least indirectly. If the linkage is not theorised, it is difficult to claim that we have any evidence of causal linkages. In contrast, if we find evidence in a case that suggests that local elites had strong motivations to provide false information, this enables us to make the inference that there was a causal linkage between security forces approaching local elites and the provision of false information. Finally, making linkages between actions explicit helps when assessing potential functional equivalence between what actors are doing in different cases.</p> <p>Returning to the move/game metaphor used in the introduction, social context can be understood as the distribution of material and ideational resources between actors that shapes what social 'game' is being played and the relative power and interests of different actors therein ([<reflink idref="bib1" id="ref69">1</reflink>]: 151–2; [<reflink idref="bib13" id="ref70">13</reflink>]: 47). Social context does <emph>not</emph> determine action; instead individuals or collective actors have some form of <emph>agency</emph>, for it is possible for people to creatively resist, repudiate, suspend or try to circumvent the underlying social context and the roles they possess (e.g., [<reflink idref="bib1" id="ref71">1</reflink>]: 195).[<reflink idref="bib14" id="ref72">14</reflink>] Instead, what game is being played shapes what actors want, what moves are possible and their chances of succeeding. For instance, if policy-makers recognise that they are facing a crisis, they are acting in a social context with very different rules and resources than in non-crisis, routine policy-making situations.</p> <p>The underlying social context can be conceptualised in SPT as composed of <emph>both a material and ideational dimension.</emph> The <emph>material dimension</emph> is the physical manifestations of context (often but not always observable), including physical or human resources as they are embedded in rules, roles and institutions in a particular social setting. However, rules are not only things out there, but have to be interpreted by actors to have impact. The <emph>ideational dimension</emph> is the social side of rules, roles and institutions that shape human behaviour through what they <emph>mean</emph> to actors (e.g., [<reflink idref="bib1" id="ref73">1</reflink>]; [<reflink idref="bib54" id="ref74">54</reflink>]). The underlying social context cannot be reduced to the sum of actions of individuals, as the '...social constitutes its own stratum and hence society has other properties and powers than the individuals constituting it' ([<reflink idref="bib1" id="ref75">1</reflink>]; [<reflink idref="bib54" id="ref76">54</reflink>]: 53). In other words, context has 'emergent' properties. Combined, the material and ideational dimension can be thought of as the social context for action, producing motivations for action based on the relative positions of actors, and also shape the ability of actors to achieve these interests by acting as constraints and enabling resources ([<reflink idref="bib1" id="ref77">1</reflink>]: 200–17; [<reflink idref="bib46" id="ref78">46</reflink>]: 120). There is a strong <emph>relational</emph> aspect of social context, in that it is not possession of resources or roles per se, but instead how they relate to what is possessed by other actors ([<reflink idref="bib17" id="ref79">17</reflink>]: 291–2). Porpora summarises this relational aspect clearly by writing, '...people act from a social position related to other social positions...' ([<reflink idref="bib46" id="ref80">46</reflink>]: 116–8).</p> <p>Motivations for action and how actors perceive the actions of others are shaped by the social context. To understand why actions occurred and why they had the effects they had, we therefore need to probe how social actors understood actions, tapping into the intentions behind the actions performed by actors in a particular social context – in effect studying the <emph>meaning</emph> behind actions ([<reflink idref="bib1" id="ref81">1</reflink>]: 195; [<reflink idref="bib17" id="ref82">17</reflink>]: 307; O'Mahoney 2015; [<reflink idref="bib46" id="ref83">46</reflink>]: 160–1). Actors ascribe <emph>meaning</emph> to action both in terms of those performing it and those on the receiving end. The <emph>sender</emph> has <emph>intentions</emph> when performing an action, but just as important, how the <emph>receiver perceives</emph> the action also determines its impact. This has important epistemic implications, as only by shedding light on <emph>both</emph> intentions and perceptions can we make inferences about causal linkages in a social process.</p> <p>Intentions and perceptions are of course not the only explanations for why actions trigger responses. Here we follow the pragmatic position of most critical realists who define causal principles (aka mechanisms) as any social science theory that can explain why an actor would respond in a particular way. For instance, we might draw on cognitive psychological theories relating to epistemic authority (e.g., [<reflink idref="bib35" id="ref84">35</reflink>]) as a causal principle explaining the linkage between an action (<emph>problem diagnosis put forward</emph>) of an expert (actor A) and the action in response (<emph>uncritical acceptance of diagnosis</emph>) by a policy-maker (actor B).</p> <p>As an example of the analytical value-added of making causal linkages explicit, [<reflink idref="bib12" id="ref85">12</reflink>] investigated a group of social actors in a <emph>barrio</emph> (neighbourhood) in Venezuela who engaged in widespread looting of stores and warehouses. While the term <emph>looting</emph> describes the action itself, it leaves us in the dark as regards <emph>why</emph> social actors responded with looting, i.e., why it was taking place. What actions by others triggered the looting, and why were these actions causally linked to the looting? In their work, [<reflink idref="bib12" id="ref86">12</reflink>] then explored the reasons the social actors themselves gave for the looting actions, probing motivations and perceptions of actors. The social actors explained the actions as acts of protest because the government had created a situation where people did not have enough to eat. Here, '...the meaning that people ascribed to the mass stealing helps to explain why they did what they did. Participants and witnesses did not interpret these actions as simply "looting" or "stealing"; nor did they regard others who engaged in these activities as looters or thieves' ([<reflink idref="bib22" id="ref87">22</reflink>]: 81). Here perceptions of injustice and desperation are an important part of the causal linkage in the process.</p> <p>This example also raises an important point. While the meaning of action had shared, intersubjective elements amongst the residents of the <emph>barrio</emph>, the owners of warehouses and law authorities would not perceive the looting actions to mean the same thing. Indeed, in almost all social interactions, there will be contrasting and conflicting perspectives of what social actions mean depending on whether one is looking at the sender or receiver of action. A critique here would be that if the meaning of action that forms part of the explanation of causal linkage depends on perspective, no 'objective' measurement of intentions, motivations and meanings is possible, making it futile to study empirically (e.g., [<reflink idref="bib36" id="ref88">36</reflink>]: 40). Yet for scholars interested in explaining why things happen in social processes, these tensions and contradictions in the intersubjective meanings ascribed to actions are not matters of concern. Instead, they are an analytical goldmine. Exploring why different actors explain actions in the way they do enables us to shed more light on complicated and multifaceted social processes and the underlying power relations as they play out in the real world ([<reflink idref="bib22" id="ref89">22</reflink>]: 82; [<reflink idref="bib13" id="ref90">13</reflink>]). At the same time, it is also obvious that other forms of evidence can <emph>also</emph> be used in parallel to shed light on the causal linkage between actions of the authorities and the responses of residents. In the case of the looting actions, we might theorise that the grievance-motivation would have different observable implications than a greed-driven linkage. We could then investigate whether looters went only after very valuable items, and whether the actions were accompanied by other manifestations of protest (e.g., slogans being chanted) to see whether there was evidence pointing in the direction of grievances or greed.</p> <p>Bringing it all together, a social process theory makes actions and actors explicit, as well as the causal linkages between them that stem from the social context and how they shape the motivations/perceptions and how other causal principles operate. In abstract terms, this is depicted in Figure 4, below. While a figure should keep the information at a minimum for clarity reasons, the accompanying text can provide fuller descriptions of actions, and the underlying theorisation of the causal linkages between actions and responses for key episodes in enough detail that they can be empirically assessed.</p> <p>Graph: Figure 4. Conceptual language for theorising a social causal process.</p> <p>As an example of what one key episode of a social causal process theory could look like can be seen in [<reflink idref="bib20" id="ref91">20</reflink>] article on the processes that led to people participating in the Rwandan genocide. In the article, Fujii develops a relatively abstract process theory based on detailed empirical fieldwork. In the article, she does not detail the social process in explicit terms. However, it is possible to reconstruct key episodes from the article in social process theory terms because she provides enough details about actors, actions and the linkages (motivations etc). One key episode of the process involves group leaders who deploy pressure in the form of threats and other tactics of forcible recruitment to push persons with whom they have kinship or other strong social ties to participate who otherwise did not want to take part in violence. In Figure 5, we illustrate what this episode might have looked like if theorised using the conceptual language of SPT. To understand the causal linkage in the episode, Fujii uses interpretive interviewing techniques to build a more abstract explanation of joiner motivations and perceptions of threats, and the social context within which the actions occur. Further, she draws on [<reflink idref="bib27" id="ref92">27</reflink>]) theory of how a social context of 'embeddedness' can lead joiners to respond with participation in violence. Translated into SPT terms, the causal linkage is a combination of motivations, perceptions and causal principles in a relatively abstracted social context (interpersonal kinship networks). The added value of making Fujii's social process theory more explicit is both theoretical (e.g., enabling greater scrutiny of the logic behind the claimed linkage), and empirical (e.g., enabling more focused attention to be directed towards assessing whether the postulated causal linkage actually worked as theorised).</p> <p>Graph: Figure 5. A key episode in Fujii's 'power of local ties' process theory in SPT terms.</p> <hd id="AN0184233958-9">The Empirical – an Analytical Dualist Epistemology for SPT</hd> <p>How can we trace social causal processes empirically? The conceptual language of context and action developed above has epistemological implications, mandating the adoption of methodologies that enable us to trace causal process and the patterns of action and what they mean (the 'moves' and intentions/perceptions) within specific social contexts (the 'game'). This implies adopting an analytical dualist epistemological approach that draws on both positivist PT methods, supplemented with interpretivist methods as used by critical realist to elucidate causal linkages. The crucial value-added here is that by trying to elucidate evidence of social causal linkages, stronger causal inferences can be made using SPT. In contrast, positivist PT methods tend to only trace actors and actions, resulting in an analysis that assumes causation based on sequences of events. Additionally, as we discussed in the previous section, a social process theory is always an abstraction from case-specific details. Assessing empirically whether it functioned as theorised in any given case requires considerable case knowledge to be able to assess whether actors performed the causal 'functions' detailed in the theory.</p> <p>In the following, we develop how more positivist and interpretive methods can be used in a <emph>supplemental</emph> fashion because they are used to ask <emph>different questions</emph>. The overall approach is illustrated in Figure 6, below.</p> <p>Graph: Figure 6. Methods for evidencing social causal process theories.</p> <p>Positivist PT methods can be used to observe <emph>what</emph> actors are doing, and <emph>what</emph> roles and resources actors have in relation to each other in more material terms (social context) ([<reflink idref="bib4" id="ref93">4</reflink>]; [<reflink idref="bib6" id="ref94">6</reflink>]), generating what can be termed '<emph>experience-distant</emph>' evidence ([<reflink idref="bib23" id="ref95">23</reflink>]). The term experience-distant denotes empirical material that is in the form of 'objective' observables (e.g., a proposal was put forward) where the interpretation of what the evidence means is based on researcher assessment that is external to the lifeworld of actors. Collecting experience-distant evidence involves observing actions through the empirical traces they leave in the empirical record. Here we are asking 'what' type questions about actions. The traces of actions can take many different empirical forms, ranging from the content of documents, accounts from participants, or even statistics produced by public authorities.</p> <p>As regards evidencing causal linkages (because...), we argue for a combination of positivist PT 'tracing', and interpretivist methods to understand both intentions/perceptions and causal principles, and how they are produced and shaped by social context. At first glance, it might seem as if collecting evidence of sender-receiver intentions and perceptions for the causal linkages in a social process can be done using positivist PT methods – for instance by conducting a series of 'semi-structured' interviews, or by deploying a survey.</p> <p>However, probing intentions and perceptions is not so simple or straightforward. Instead, it requires a hermeneutic interpretative methodology that goes beyond merely gathering information, but involves an epistemological commitment to '...care to glean the meanings that the people under study attribute to their social and political reality' ([<reflink idref="bib51" id="ref96">51</reflink>]: 5) that transcends the distinction between data generation and analysis. In other words, it is not just about getting information, but it is also <emph>accessing meaning</emph> through combining an awareness of details (e.g., <emph>which</emph> words did the actor use when he criticised the bill? What was the <emph>tone</emph> of his voice?) with an orientation towards the <emph>meanings</emph> of these details in the particular social context (e.g., was she being dismissive or seeking to compromise?). This can be termed 'experience-near' evidence that is used to attempt to tap into the lifeworld of participants to reveal more about the intersubjective understandings of actions, motivations and social context ([<reflink idref="bib23" id="ref97">23</reflink>]; [<reflink idref="bib13" id="ref98">13</reflink>]: 30, 33).</p> <p>Experience-near evidence goes beyond merely asking people why they did something. It can be used to understand <emph>why</emph> activities took place, and in particular in probing the causal linkage between the actions of one actor and their impact on another actor. Without empirically exploring the meaning behind action, we would miss the social dimension of action and why things happened as they did. At the same time, often people do not know why they did something, and there is a large body of evidence in psychological research that suggests we seldom are able to 'objectively' attribute why we did something (see the chapters in [<reflink idref="bib28" id="ref99">28</reflink>] for a good overview; also [<reflink idref="bib57" id="ref100">57</reflink>]; O'Mahoney, 2015). But this is not the point. Using interpretive methods like ethnography, interpretive interviewing[<reflink idref="bib15" id="ref101">15</reflink>] or textual analysis, the goal is to probe actor <emph>understandings</emph> of their intentions and how they perceived the actions of others to shed light on their reflections about their goals, desired outcomes, and other possible courses of action – all of which can help us understand what an activity meant both for the actor who performed it and the target(s) who are impacted by it and triggered to further actions and interactions. It is of course important to also be aware that any given activity can be motivated by multiple goals, or a cluster of activities might be towards one goal. Further, humans often describe their own actions in situational terms ('I had no choice, the context made me do it'), whereas they ascribe more intentionality to the actions of others ([<reflink idref="bib33" id="ref102">33</reflink>]).</p> <p>Taking meanings of action more seriously raises a critical difference between positivist PT and SPT (and with other more interpretive methods). In SPT, the research is a human being trying to understand the meaning that other human beings ascribe to action in a specific social context, meaning that a critical awareness of the scholar as the main instrument of research is also required in SPT ([<reflink idref="bib52" id="ref103">52</reflink>], 85; [<reflink idref="bib22" id="ref104">22</reflink>]). Being exposed to actors in the researched setting, scholars often stumble on puzzling actions that need to be explained, unexpectedly encounter actors that need to be included, or that material and social resources the scholar considered valuable are perceived as useless by the actors. Consequently, the most frequent outcome is revision of the theorisation and research design ([<reflink idref="bib22" id="ref105">22</reflink>]).</p> <p>However, when learning about how things worked in the social world, the researcher must also be aware of how researcher <emph>positionality</emph> influences the degree of access and interpretations of what have been observed and interactions in the setting. By positionality we mean the demographic (e.g., gender, race, socioeconomic status), locational (physical location in the research setting) as well as theoretical 'position' (experience distant concepts) of the researcher in the research setting ([<reflink idref="bib52" id="ref106">52</reflink>]). Being aware of how these forms of positionality influence access and interpretations of the encounters with actors in settings is important, and may be used actively and strategically to enhance embeddedness and further ground one's knowledge-claims in context, thereby also enhancing the trustworthiness of interpretations of the researcher ([<reflink idref="bib22" id="ref107">22</reflink>]; [<reflink idref="bib47" id="ref108">47</reflink>]: 224). Note that there is not a quick-fix for the issue of positionality, for example by merely treating it as part of the inherent empirical uncertainty of real-world research that downgrades our ability to update our confidence in our inferences. Instead, it requires reflexivity throughout the course of a research process.</p> <p>The causal principles underlying causal linkages can be drawn from any relevant social science causal theory that provides a plausible explanation. There are many cognitive and social psychological causal principles such as confirmation bias whose operation we cannot study directly in the field, but whose workings have been studied extensively in laboratory experiments. While these findings cannot act as evidence that it worked in our particular case, these findings make it more plausible that it might be operative in the case in hand, as long as the social context is one in which it could have worked. Further, even if we cannot evidence a causal principle directly, they can be theorised and studied 'as if' it has a real existence that is independent of our knowledge of it ([<reflink idref="bib13" id="ref109">13</reflink>]). Here making the causal principle explicit enables the analyst to cast the net widely for any observable implications of a particular linkage, even if they are relatively indirect evidence. For instance, an action motivated by greed might take a different form that a grievance motivated action.</p> <p>Studying how social context (i.e., the 'game') shapes motivations/perceptions and how causal principles operate requires a combination of positivist PT and more interpretivist methods. There are elements of context such as institutional roles and relations between actors that can be relatively easily studied through using positivist PT methods. For instance, an actor might hold a powerful elected office, or a state as an actor might have a large number of allies in a formal treaty that is an important resource that can be leveraged in its relations with foes.</p> <p>However, the underlying social context and how it shapes behaviour cannot be studied solely based on 'objective' observables. Instead, it is also important to assess how actors interpret the 'game' they are playing, and the roles and resources that emanate from these intersubjective understandings using more interpretive methods ([<reflink idref="bib17" id="ref110">17</reflink>]: 304). For instance, actors acting within an organisation would have different resources depending on whether there is a culture of formal adherence to the rules (e.g., privileging actors higher in the formal hierarchy) versus a more informal culture of 'getting the job done' that might privilege actors with particular skills like expertise or problem-solving. For collective actors, we might engage in multiple interviews throughout an organisation to understand its culture, and potentially also multiple conflicting understanding of the 'game' being played.</p> <p>Given the relational nature of context, in which roles and resources are only important in relation to those possessed by other actors, the understanding of the game amongst participants is intersubjective. This does not mean that participants are necessarily consciously aware of what game they are playing, and there can be shared tacit or implicit knowledge that can only be uncovered through hermeneutic interpretative methodologies that seek to uncover the, '...taken-for-granted, unspoken, yet shared and known "rules" for acting and interacting' ([<reflink idref="bib63" id="ref111">63</reflink>]: 12; [<reflink idref="bib46" id="ref112">46</reflink>]). Additionally, there can be tensions about what game is actually being played in any given situation. For example, in the European Parliament, individual elected representatives can differ in whether they view themselves as part of a more national-focused 'game' in which they are representing their national party in Europe, or they are taking part in a EU-level 'game' where loyalty should lie with the transnational European party they are part of in the European Parliament. Further, what 'game' is being played might change over the course of a process as either material and/or ideational roles and resources change. For example, a leader who oversteps their legitimate authority might lose the trust of the followers in subsequent interactions, thereby undermining the relational resource that enabled them to lead in the first place.</p> <p>Figure 7 summarises sets of key questions to be asked to evidence different aspects of a social causal process theory. Now we turn to a step-by-step guide for conducting SPT in practice.</p> <p>Graph: Figure 7. Sets of questions to be asked when evidencing social causal process theories.</p> <hd id="AN0184233958-10">A Step-by-Step Guide for Engaging in Social Process Tracing</hd> <p>This section presents a step-by-step guide for how SPT and the analytical framework developed in the previous section can be used in practical research. We illustrate how to move from initial case selection and theorisation of social processes to how different types of evidence can be used in a supplemental fashion. We conclude with a discussion of the challenges of generalisation and SPT. To illustrate how SPT can be used in practice, we use a running example in which we reconstruct the research design of a SPT study co-written by one of the authors ([<reflink idref="bib5" id="ref113">5</reflink>]). The case involved flawed policy learning in the Spanish banking crisis in 2012.</p> <p>Before we proceed, it is important to note that while the following presents SPT in a step-by-step manner, in practice SPT involves an iterated, <emph>abductive</emph> design that moves back-and-forth between a process theory and empirics ([<reflink idref="bib52" id="ref114">52</reflink>], 26–34; [<reflink idref="bib59" id="ref115">59</reflink>]). When developing a new social process theory, an initial process theory sketch is developed based on initial empirical probing, logical brainstorming, and searches in the existing literature. The sketch acts as a form of analytical baseline that is juxtaposed against the empirics in a particular case and what we otherwise know from relevant theoretical literatures, and then continually refined as it is developed. Elucidating how an actor interpreted a specific action in a particular social context requires the SPT-researcher to re-visit existing interpretations of the actor's intentions and perceptions as the researcher's own case-specific knowledge expands ([<reflink idref="bib59" id="ref116">59</reflink>]: 21–34; [<reflink idref="bib52" id="ref117">52</reflink>]: 27–34). For example, after discussing with participants what would be a considered common or appropriate behaviour in a particular type of meeting (i.e., what 'game' is being played), a SPT-researcher might discern that a speech act that she initially thought represented a statement of subservience from a group member to the meeting leader was merely a ritual statement that meant no more in the particular context than that the speaker was being polite. Admittedly, updating previous interpretations based on newly acquired knowledge is not a hallmark of SPT, but an expectation of good research practices found in social science research in general. However, as SPT stresses the importance of the relevant actors' respective understandings of the game being played and how they understand the moves being made <emph>within a specific social context</emph>, the social process tracer will become more knowledgeable about how the social process worked in the particular case – but never comprehend completely – as iterations between process theory and empirics continue. Following [<reflink idref="bib52" id="ref118">52</reflink>]: 30), SPT knowledge-claims represent 'momentary stopping points' and not conclusive and final findings. When the process 'stops' and conclusions are presented depends on the research goals and the state of knowledge in the field. In some situations, it can be good enough to stop when a relatively shallow understanding of how a process worked in a particular case is achieved, backed by relatively weak evidence. In other situations the social process theory found in one case might be assessed in another case to see whether similar (but not identical!) processes are at play.</p> <hd id="AN0184233958-11">Initial Case Selection and Initial Process Theorisation</hd> <p>PT methods focus on studying positive cases, understood as cases where causes/outcomes and contextual conditions for the operation of the process are all present ([<reflink idref="bib24" id="ref119">24</reflink>]; [<reflink idref="bib4" id="ref120">4</reflink>]). However, given SPT's openness to learning about how actors themselves understand what action and context mean, what theoretical process a given case is actually a 'case of' can shift during the course of research as we learn more about how the process worked. This means that case selection in SPT is closer to what Soss terms 'casing', in which the researcher continually problematises the relationship between abstract theoretical categories and the observations and interpretations of them ([<reflink idref="bib57" id="ref121">57</reflink>]: 23). Being open to changing what the case is a 'case of' means that the temporal or spatial bounds of a case can also change. Further, different theoretical literatures might become more/less relevant in the course of research ([<reflink idref="bib57" id="ref122">57</reflink>]: 23–24).</p> <p>Once a positive case has been selected, an initial round of empirical work should be conducted to map what happened, in terms of who were the key actors and episodes, and what were these actors doing. Based on this, an initial process theorisation sketch should be developed that can act as the jumping off point for the first round of empirical analysis. The first step in theorisation should be to engage in a wide-reaching survey of the literature to see whether there are existing theories that can serve as the starting point in theorising what might link a cause (or set of causes) with an outcome in a particular context. Often it will be found that there are not off-the-shelf process-level theories that can be used. Or if they exist, they might have been theorised in very abstract, minimalist terms; e.g., <emph>mobilisation processes</emph> amongst social actors are theorised to bind grievances (cause) and an outcome like conflict, without telling us more about who is theorised to be doing what and why. This said, even a shallow one-liner theorycan help focus attention on the most important processual dynamics and linkages.</p> <p>After surveying the literature, an initial brainstorming exercise can be conducted to theorise a plausible causal process linking the cause(s) and outcome together, using the conceptual template presented in Figure 7 (previous section). Process theorisation should be thought of as an iterated dialogue between theories and empirics from the case. Often what links parts of a process together will often be unclear initially, creating a new set of questions that should be explored by searching across other literatures for inspiration for potential answers. Further, it might not be apparent what episodes are 'key' at this moment, but identifying these episodes is a crucial aspect of theoretical abstraction from the gritty details of a given case.</p> <p>Note that at this stage of SPT, it is usually easier to theorise the 'material' dimension of the process, as we still have not started exploring how actors themselves understood the game being played and the meaning of the moves therein. This said, some attention should still be given to speculation about the ideational dimensions of structure and action. The result of initial theorisation should be a process theory sketch that contains preliminary ideas about the key parts of the process in terms of actors, activities and the causal l binding them together, and that include some ideas about the relational, social context in terms of situational logics.</p> <p>What can this look like in practice? Using a published example of PT as an example, the following reconstructs what the research process would have looked like if SPT had been deployed ([<reflink idref="bib5" id="ref123">5</reflink>]). In the example, the case of the EU's management of the Spanish banking crisis in 2012 was initially selected because we identified an empirical puzzle while we were cataloguing empirically what happened in the case. We kept finding key EU actors referring to the lessons learned from the Irish collapse in October 2010 when they were trying to understand the causes of the Spanish banking crisis in spring 2012, which was surprising given that the Spanish situation was quite different from what caused collapse in Ireland.</p> <p>Interested in explaining this puzzle, and based on the empirical cataloguing of what happened, we narrowed the processual focus to the role played by policy learning when diagnosing problems, and how analogies were used. However, as existing research tended to black-box actual policy learning processes (e.g., [<reflink idref="bib14" id="ref124">14</reflink>]), we had to develop an initial process theory sketch from scratch that could help us account for why smart policy-makers, backed by teams of policy experts, allowed themselves to be cognitively constrained by lessons learned from past cases that might not fit the case in hand. The initial process theory we developed in the research is described in Figure 8. However, key episodes of the process remained in a black-box at this point of the research (denoted by grey text and question marks). Additionally, note that the initial process sketch is more in the form of <emph>what</emph> happened, and the causal linkages between episodes were not very clear before we undertook the second, more detailed round of empirical fieldwork.</p> <p>Graph: Figure 8. The initial process theory sketch between policy failure and premature consensus.</p> <hd id="AN0184233958-12">Empirical Probing and Theoretical Revision</hd> <p>The initial process theory sketch is then used as the basis for questions that are asked of the empirical record. The researcher can start empirical research by asking whether there is any evidence suggesting that the theorised process sketch actually took place, whether and how the process theory needs to be revised, including potentially a new 'casing' of the phenomena itself. For the abductive-minded social process tracer, it is not a problem to revise theories or research questions. Instead, revisions show that the research process is progressing through the mediation between experience-near and -distant knowledge, which also enables an update of the prior understanding of what it is a 'case of'.</p> <p>At the most basic level, the researcher should thus always have two questions in mind when engaging in SPT: what happened and how were the actions in key episodes linked together? The social dimension should be considered for key episodes of the process (keeping in mind that the 'game' may change during a process), and not just be discussed as a condition at the beginning and/or end of the analysis. Based on the set of SPT questions (see Figure 7, above), we now elaborate on how SPT can be used to produce and combine experience-distant and –near empirics to evidence social process theories.</p> <p> <emph>Experience-distant empirics</emph> deal with evidencing actions (i.e., what 'moves') and the material dimension of context (roles, resources and institutions). To evidence action, the analysis should start by operationalising expectations about the potential observables that might be expected to be found in the case if an action was performed. Here the analyst is speculating about the data-generating process associated with the actions in each episode of interaction in the process, asking what empirical traces might be left in the case that could confirm or disconfirm that an action took place (for more on how to do this, see [<reflink idref="bib4" id="ref125">4</reflink>]: 155–222). Further, experience-distant empirics are collected regarding the relations between actors by looking for empirics concerning the relative roles and resources of actors that might enable/constrain their actions. For example, the 'assessing' action was expected to leave a paper trail in the form of studies and other analytical knowledge products.</p> <p>At the same time, <emph>experience-near empirics</emph> aim to shed light on what actions mean (i.e., intentions of sender and perceptions of receiver) in a particular context and the understandings of the game being played amongst actors. Several interpretive methods may be relevant to produce experience-near empirics (see [<reflink idref="bib53" id="ref126">53</reflink>]). The social process tracer could for instance conduct ethnographic fieldwork, which entails the researcher immersing herself in the empirical setting over extended periods and observing what is going on while also talking to the actors (for good introductions, see [<reflink idref="bib7" id="ref127">7</reflink>]; [<reflink idref="bib17" id="ref128">17</reflink>]). Central to the ethnographic approach is the 'ethnographic sensibility' that is combining an awareness of details with an orientation towards the meanings of these details in the particular context ([<reflink idref="bib43" id="ref129">43</reflink>]). Having this sensitivity helps social process tracers understand what actions means to actors in the given setting, that is, learning about the game being played. More specifically, doing ethnographic fieldwork allows the social process tracer to observe <emph>what</emph> is going on <emph>while</emph> it is going on ([<reflink idref="bib32" id="ref130">32</reflink>]),[<reflink idref="bib16" id="ref131">16</reflink>] thereby also generating experience-distant empirics (what are the actors doing, who are they talking to about what, etc.?), but more importantly in the context of SPT, ethnography enables the researcher to directly observe how the actors understand, negotiate and establish 'the game'. Moreover, combining ethnographic observations with different forms of ethnographic/interpretive interviewing enables the social process tracers to examine the intentions and perceptions of the different actors – thereby getting closer to answering the why questions concerned with the social meaning that actors ascribe to specific actions ([<reflink idref="bib7" id="ref132">7</reflink>]; [<reflink idref="bib22" id="ref133">22</reflink>]; [<reflink idref="bib37" id="ref134">37</reflink>]). From such interviews, the scholar might learn that seemingly important actions were not intentional or unfolded quite differently from what some actors thought they would.[<reflink idref="bib17" id="ref135">17</reflink>]</p> <p>When evaluating the overall strength of evidence, it is important to note that SPT uses different methods to evidence different things about a causal process. Experience-distant empirics are used to evidence the moves that were made and the material dimension of context. In contrast, experience-near empirics and researcher interpretation of them are used to shed light on the social dimension of action, and the underlying social game being played in a particular context. Instead of evidential standards related to the empirical correspondence between theory and empirics, users of experience-near evidence use the term 'trustworthy' inferences to refer to degree of thoroughness involved in interpreting and re-interpreting findings, as well as the degree of transparency about how this was achieved ([<reflink idref="bib52" id="ref136">52</reflink>]). These different evidential standards mean that the overall inferences about how the process worked cannot be neatly summarised, especially in the form that some Bayesian process tracers suggest (e.g., [<reflink idref="bib19" id="ref137">19</reflink>]). Further, the different types of evidence <emph>supplement</emph> each other in complex ways, meaning that we cannot estimate the probative value of any individual piece of evidence. Instead, the SPT analyst should be prepared to engage in an often-painstaking <emph>evidential bricolage</emph> using the many different types and pieces of evidence to paint an overall picture of how the process worked in a case. At the same time, what is important in practical terms is for the analyst to flag where there is strong and weak evidence, respectively, and be able to indicate the evidential basis for the different inferences about moves and the game.</p> <p>Returning to the SPT example, the study first used experience-distant evidence to catalogue what actions took place. As regards the first key episode of the process (experts assessing), we reviewed the academic literature both before the Irish collapse (policy failure) and afterwards, attempting to identify who published the first studies of the causes of collapse and what impact the studies had in the literature. We found that the first academic study was by two IMF economists, and we were then able to document that their diagnosis was quickly absorbed in the academic and think tank literature as a more abstract 'lesson' of the dangers of too much interdependence between banks and sovereigns that was stripped from the particular context of the Irish case (i.e., overgeneralised). To understand <emph>why</emph> the lesson was over-generalised, we looked into the cognitive psychological literature on learning and found evidence that when people 'learn' causal lessons from the past, it is in an abstracted form that is stripped of the context in which the lesson worked ([<reflink idref="bib25" id="ref138">25</reflink>]: 1344–5; [<reflink idref="bib49" id="ref139">49</reflink>]). Based on this causal principle, we then expected empirics in the form of a lack of discussion of context when policy-makers (and experts) were discussing the applicability of lessons from other cases. We were able to evidence that there was a lack of discussion through participant interviews.</p> <p>What then puzzled us was why the over-generalised lesson diffused so quickly from experts in the EU bureaucracy and think tanks to policy-makers. Through interviews with participants, we built a preliminary explanation based on the idea that they were members of a shared epistemic community that policy-makers trusted. However, the questions we asked still dealt primarily with who did what (i.e., moves of the game).</p> <p>We used interpretive interviewing methods to understand more about the game being played, and how actions of different actors were intended and perceived. We asked expert insiders about their everyday practices – e.g., who they typically met with and what types of interactions they engaged in. We listened carefully to what actors were telling us, noticing the types of words used and inquiring about them. Through these interviews, the involved actors helped us understand the epistemic community within which the actions were embedded, and what the rules of the social game being played within this community, and how the community interacted with policy-makers. From the interviews we learned that while there were formal institutional barriers between actors, in practice there was a small informal network of economic experts from the EU institutions (European Commission and the European Central Bank), the International Monetary Fund, and the Bruegel think tank that engaged in close dialogue on the banking/sovereign issue and within which lessons moved freely. Afterwards, we used the new knowledge about the informal epistemic community of economic experts to qualify our prior understanding of why particular actions (e.g., the publication by a single think tank expert) could resonate so strongly within this community. This enabled us to shed some light on how this 'game' between experts and the ensuing game between experts and policy-makers worked.</p> <p>This combination of cataloguing what happened and why, coupled with probing of the game being played, resulted in a final process theory after successive iterations of empirical analysis and theoretical revision in which we were able to unpack the causal linkages between the key episodes, depicted in Figure 9. Of course, it can be argued that the final process theory is unnecessarily complex, and far from the goal of 'parsimonious' theories. However, by making causal linkages explicit in the theorised process, we were able to evidence them empirically – for some at least indirectly – thereby enabling causal inferences to be made. This ability to actually evidence causal linkages is a major advance in relation to existing PT methods, which tend to conflate a sequence of actions with causation without evidencing the actual linkages between actions.</p> <p>Graph: Figure 9. Revised process theory of policy failure and premature consensus. Source: Inspired by [<reflink idref="bib5" id="ref140">5</reflink>]).</p> <hd id="AN0184233958-13">And so What – Moving Beyond the Single Case</hd> <p>The result of the analysis of a particular case with SPT is an evidence-backed, abstracted social process theory that explains who was doing what and why for key episodes. Given that it is an abstraction, the process can in principle also be present in other cases. Yet this might seem to contradict critical realist ideas about the importance of social context for how things work. However, by applying notions related to establishing functional equivalences between what actors are doing in relatively similar social contexts, it is possible to go beyond the particular context of the single case. Naturally, before other cases have been studied, we cannot assume that similar processes actually were operative in other cases. To be able to make evidence-backed claims that similar processes were operative requires a <emph>process-level comparison</emph> of another case that enables the assessment of functional equivalences for key episodes. Here the analytical dualist approach of SPT is a crucial value-added in exploring potential functional equivalences between two cases.</p> <p>Engaging in a processual comparison requires first that another case is identified that has a similar social context, and the cause(s) and outcome were (at least in principle) present. For instance, in the example of policy failure and learning discussed above, we might select another case of policy failure/crisis and possible premature consensus in a policy area in which there was an epistemic community that was trusted by policy-makers, and policy-makers were facing a new crisis.</p> <p>The purpose of the second case study is not to 'test' whether the process theory is valid per se, as the first case study provided evidence that it was valid in that case. Instead, the second case study can be either be used to explore the external validity of the processual theory derived from the first case through <emph>assessing similarities</emph>, or it can be a juxtapositional comparison in which lessons from the first case are contrasted with another case to help understand what happened in the second case through <emph>assessing differences</emph> ([<reflink idref="bib50" id="ref141">50</reflink>]).</p> <p>Assessing whether a theorised causal process worked in a similar fashion in the second case requires that the social process theory from the initial case is 'particularised' in relation to the new case. This involves translating the theoretical abstractions for each key episode of actors, activities and linkages in key episodes into case-specific empirics for the new case. For example, instead of 'policy-makers', we would have to detail what these looked like in the new case (e.g., it could be decision-makers in the European Council). The critical move when particularising is deploying extensive case knowledge of both cases to assess whether the actors and the activities they performed in their interactions during key episodes played causally similar roles, that is, assessing functional equivalences. Establishing functional equivalence has a qualitative/interpretive character, building on detailed empirical case knowledge of what action means in the context of the two cases that are then compared.</p> <p>While the process will never be identical, the processual comparison enables an evidence-backed claim that similar processes were operative. Using SPT methods, we would be able to justify that we knew enough about actors, actions, linkages in context in the two cases to be able to assess similarity/differences. If we find differences, are there contextual differences that produced the differences? For instance, if we used the social process theory from the policy learning article discussed above to assess another case, we might have found that while policy-makers in the original case accepted and used a given lesson without much reflection given the severity of the crisis, in another crisis case we might find that more reflection was undertaken because there was more time before a decision had to be taken. If we found this, we would flag the importance of time as a contextual factor in the social process theory in the first case.</p> <p>Finding similar processes operative in two cases would not enable a generalisation to all other potential cases. Developing an externally valid processual generalisation for a population of cases requires further assessment of other cases, using procedures such as a 'snowballing outwards' strategy of further empirical probing of other cases ([<reflink idref="bib4" id="ref142">4</reflink>]: 138–144).</p> <p>Assessing differences involves using the process theory from the first case as a comparative baseline against which the case-in-hand can be compared ([<reflink idref="bib50" id="ref143">50</reflink>]). For instance, particular episodes of action and interaction and how causal principles (e.g., confirmation bias) worked in one social context (e.g., EU decision-makers trying to comprehend Russian actions in Ukraine in 2022) can be compared with how things worked in another case that might be a different context (e.g., US decision-makers trying to comprehend Soviet actions during the Cuban missile crisis). The analytical goal would not be to refine a more general process theory based on findings from new cases, but instead to gain more explanatory traction on the second case through comparison of dissimilarities.</p> <hd id="AN0184233958-14">Conclusion</hd> <p>This article developed SPT as a case study approach that elucidates the social dimension of causal processes. As discussed in section 'Existing Approaches to Process Tracing: Either Under- or Over-Socialised Methods', existing approaches to PT are either <emph>under-socialised</emph> because they neglect the social dimension of action (i.e., focusing only on the 'moves' of actors), or are <emph>over-socialised</emph> because they lose sight of the processes themselves in their focus on social context (i.e., the underlying 'game' being played). Further, existing approaches have difficulties in making causal inferences through tracing how a process works because causal linkages tend to be black-boxed, resulting in them not be evidenced empirically.</p> <p>In the article, we developed a theoretical language for theorising social processes, building on both critical realism and positivist PT approaches. We recognise that explicitly theorising the causal linkages and underlying social context for key episodes is very demanding. In practice, we therefore suggest that scholars might choose to focus only on a small number of key episodes. Further, instead of including all the information in a figure, it can be included in the textual description of the social process theory. At the same time, if linkages are not made explicit theoretically, how can we evidence them as we would not know what empirical material was claimed to be evidence of?</p> <p>We developed the epistemological side of SPT for how social processes can be evidenced using a combination of experience-distant and experience-near evidence. Taken together, SPT has the ambition of both studying the moves being made by actors and the underlying game being played amongst actors in a specific social context. We do not claim that SPT should be used in all research situations. Existing PT approaches have comparative advantages in some research situations (e.g., interpretive PT can be useful in capturing shifts in discourses). SPT should be used when scholars are interested in capturing more fully the social dimension without losing the focus on actions and interactions, and/or when they are trying to evidence more fully the causal linkages (at least indirectly) in social causal processes.</p> <p>While we believe that the SPT analytical framework developed in this article contributes to both the PT and broader case study literature, there are also a range of unanswered questions and avenues for further development. For instance, how should the abductive aspect of SPT case studies be presented when discussing how the research was conducted? While book-length work can walk the reader through the different stages in ways that are analogous to how ethnographic work is presented (i.e., thick description that leaves enough of the scaffolding in place to see how it was built), this can be very difficult in an article due to word limits – although it can be presented in an Online Methodological Appendix in the supplemental material. Further, and even more challenging, given the ambition to use different types of evidence in supplementary ways, SPT studies can risk leaving both more positivist and interpretivist scholars dissatisfied. More positivist PT reviewers might object to the lack of clear identification of the probative value of each piece of evidence, or be troubled by the probing of actor intentions, while an interpretivist reviewer might be uncomfortable with the ambition of theorising a more abstract process theory that lays bare the causal linkages instead of sticking with a contextualised thick description. When do we stop collecting evidence in any given case? Additionally, while we shed some light on the issue, we still lack good answers for how to use the lessons from SPT case studies can travel to other cases. When are social causal processes 'similar enough' in two cases?</p> <p>At the end of the day, there are no easy answers to these and many other questions. But this should not stop process tracers who are interested in elucidating the 'social' from engaging in SPT and developing it further. We will only move forwards through learning by doing.</p> <hd id="AN0184233958-15">Supplemental Material</hd> <p>Graph: Supplemental material, sj-docx-1-smr-10.1177_00491241231220732 for Elucidating the Social – Developing Social Process Tracing as an Integrative Framework by Jonas Gejl Kaas, Mathilde Cecchini and Derek Beach in Sociological Methods &amp; Research</p> <ref id="AN0184233958-16"> <title> References </title> <blist> <bibl id="bib1" idref="ref13" type="bt">1</bibl> <bibtext> Archer Margaret.1995. Culture and Agency. The Place of Culture in Social Theory. Cambridge: Cambridge University Press.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref31" type="bt">2</bibl> <bibtext> Beach Derek.2021. " Evidential Pluralism and Evidence of Mechanisms in the Social Sciences." Synthese. 199:8899–919.</bibtext> </blist> <blist> <bibl id="bib3" type="bt">3</bibl> <bibtext> Beach Derek.2022 "Process-Tracing Methods in Social Science." Pp. 1–21 in Oxford Research Encyclopedias of Politics: Qualitative Political Methodology. Oxford: Oxford University Press.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref4" type="bt">4</bibl> <bibtext> Beach Derek, Pedersen Rasmus Brun. 2019. Process-Tracing – Foundations and Guidelines. 2nd Edition. Ann Arbor: University of Michigan Press.</bibtext> </blist> <blist> <bibl id="bib5" idref="ref113" type="bt">5</bibl> <bibtext> Beach Derek, Smeets Sandrino. 2022. " Once Bitten, Twice Shy. The Overgeneralization Trap and Epistemic Learning After Policy Failure." Politics and Policy. 50(6):1067–284.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref5" type="bt">6</bibl> <bibtext> Bennett Andrew, Checkel Jeffrey. 2014. Process Tracing: From Metaphor to Analytical Tool. Cambridge: Cambridge University Press.</bibtext> </blist> <blist> <bibl id="bib7" idref="ref3" type="bt">7</bibl> <bibtext> Bernard H. R.2011. Research Methods in Anthropology : Qualitative and Quantitative Approaches. Lanham, MD: AltaMira Press.</bibtext> </blist> <blist> <bibl id="bib8" idref="ref12" type="bt">8</bibl> <bibtext> Bevir Mark, Blakely Jason. 2018. Interpretive Social Science: An Anti-Naturalist Approach. Oxford: Oxford University Press.</bibtext> </blist> <blist> <bibl id="bib9" idref="ref22" type="bt">9</bibl> <bibtext> Bhaskar Roy.1978. A Realist Theory of Science. Brighton: Harvester.</bibtext> </blist> <blist> <bibtext> Cartwright Nancy.2021. " Rigour Versus the Need for Evidential Diversity." Synthese. 199:13095–119.</bibtext> </blist> <blist> <bibtext> Collier David.2011. " Understanding Process Tracing." PS: Political Science &amp; Politics. 44(4):823–30. doi: https://doi.org/10.1017/S1049096511001429</bibtext> </blist> <blist> <bibtext> Coronil Fernando, Skurski Julie. 1991. " Dismembering and Remembering the Nation." Comparative Studies in Society and History. 33(2):288–337.</bibtext> </blist> <blist> <bibtext> Danermark Berth, Ekström Mats, Karlsson Jan Ch. 2019. Explaining Society. Critical Realism in the Social Sciences. 2nd ed. London: Routledge.</bibtext> </blist> <blist> <bibtext> Dunlop C. A., Radaelli C. M. 2017. " Learning in the Bath-Tub: The Micro and Macro Dimensions of the Causal Relationship Between Learning and Policy Change." Policy and Society. 36(2):304–19.</bibtext> </blist> <blist> <bibtext> Edwards Paul K., O'Mahoney Joe, Vincent Steve, Eds. 2014. Studying Organizations Using Crit</bibtext> </blist> <blist> <bibtext> Eldar-Vass David. 2012. The Reality of Social Construction. Cambridge: Cambridge University Press.</bibtext> </blist> <blist> <bibtext> Emerson R. M., Fretz R. I., Shaw L. L. 2011. Writing Ethnographic Fieldnotes. 2. edition ed.Chicago: The University of Chicago Press.</bibtext> </blist> <blist> <bibtext> Emirbayer M.1997. " Manifesto for a Relational Sociology." American Journal of Sociology. 103(2):281–317.</bibtext> </blist> <blist> <bibtext> Fairclough Norman.2010. Critical Discourse Analysis: The Critical Study of Language. 2nd ed. London: Routledge.</bibtext> </blist> <blist> <bibtext> Fairfield Tasha, Charman Andrew E. 2017. " Explicit Bayesian Analysis for Process Tracing. Guidelines, Opportunities, and Caveats." Political Analysis. 25(3):363–80.</bibtext> </blist> <blist> <bibtext> Fujii L. A.2008. " The Power of Local Ties: Popular Participation in the Rwandan Genocide." Security Studies. 17(3):568–97.</bibtext> </blist> <blist> <bibtext> Fujii L. A.2015. " Five Stories of Accidental Ethnography: Turning Unplanned Moments in the Field into Data." Qualitative Research. 15(4):525–39.</bibtext> </blist> <blist> <bibtext> Fujii L. A.2018. Interviewing in Social Science Research: A Relational Approach. London: Routledge.</bibtext> </blist> <blist> <bibtext> Geertz C.1974. " "From the Native's Point of View": On the Nature of Anthropological Understanding." Bulletin of the American Academy of Arts and Sciences. 28(1):26–45.</bibtext> </blist> <blist> <bibtext> Gentner D., Loewenstein J., Thompson L., Forbus K. D. 2009. " Reviving Inert Knowledge: Analogical Abstraction Supports Relational Retrieval of Past Events." Cognitive Science. 33:1343-1382.</bibtext> </blist> <blist> <bibtext> Goertz Gery. 2017. Multimethod Research, Causal Mechanisms, and Case Studies: An Integrated Approach. Princeton, NJ: Princeton University Press.</bibtext> </blist> <blist> <bibtext> Granovetter M. 1985. " Economic Action and Social Structure: The Problem of Embeddedness." American Journal of Sociology. 91(3):481-510.</bibtext> </blist> <blist> <bibtext> Haggard Stephan, Kaufman Robert R. 2016. Dictators and Democrats: Masses, Elites, and Regime Change. Princeton, NJ: Princeton University Press.</bibtext> </blist> <blist> <bibtext> Heckhausen J., Heckhausen H. 2018. Motivation and Action. 3rd ed. Cham, Switzerland: Springer.</bibtext> </blist> <blist> <bibtext> Hedström Peter, Ylikoski Petri. 2010. " Causal Mechanisms in the Social Sciences." Annual Review of Sociology. 36:49–67.</bibtext> </blist> <blist> <bibtext> ical Realism – A Practical Guide. Oxford: Oxford University Press.</bibtext> </blist> <blist> <bibtext> Jacobs Alan M.2014 "Process Tracing the Effects of Ideas." Pp. 41–73 in Process Tracing. From Metaphor to Analytic Tool, edited by Bennett A., Checkel J. T. Cambridge: Cambridge University Press.</bibtext> </blist> <blist> <bibtext> Jerolmack C., Khan S. 2014. " Talk is Cheap: Ethnography and the Attitudinal Fallacy." Sociological Methods &amp; Research. 43(2):178–209. doi: https://doi.org/10.1177/0049124114523396</bibtext> </blist> <blist> <bibtext> Jervis Robert.1976. Perceptions and Misperceptions in International Politics. Princeton, NJ: Princeton University Press.</bibtext> </blist> <blist> <bibtext> Kaidesoja Tuukka. 2021 "Three Concepts of Causal Mechanism in the Social Sciences." in Norms, social structures and policy changes: Essays in honor of Hannu Ruonavaara edited by Erola J., Naumanen P. Uppsala: Uppsala universitet.</bibtext> </blist> <blist> <bibtext> Krebs R. R., Jackson P. T. 2007. " Twisting Tongues and Twisting Arms: The Power of Political Rhetoric." European Journal of International Relations. 13(1):35-66.</bibtext> </blist> <blist> <bibtext> Kruglanski A. W., Raviv D. Bar-Tal, Raviv A., Sharvit K., Ellis S., Bar R., Pierro A., Mannetti L. 2005. " Says Who?: Epistemic Authority Effects in Social Judgment. " Advances in Experimental Social Psychology. 37:345–92.</bibtext> </blist> <blist> <bibtext> Lamont Michèle, Swidler Ann. 2014. " Methodological Pluralism and the Possibilities and Limits of Interviewing." Qualitative Sociology. 37(2):153–71. doi: https://doi.org/10.1007/s11133-014-9274-z</bibtext> </blist> <blist> <bibtext> Löblová O.2018. " When Epistemic Communities Fail: Exploring the Mechanism of Policy Influence." Policy Studies Journal. 46(1):160–89.</bibtext> </blist> <blist> <bibtext> Machamer Peter, Darden Lindley, Craver Carl F. 2000. " Thinking About Mechanisms." Philosophy of Science. 67(1):1–25.</bibtext> </blist> <blist> <bibtext> Mahoney James.2015. " Process Tracing and Historical Explanation." Security Studies. 24(2):200–18.</bibtext> </blist> <blist> <bibtext> Norman L.2016. The Mechanisms of Institutional Conflict in the European Union. Oxon, UK: Routledge.</bibtext> </blist> <blist> <bibtext> Norman L.2021. " Rethinking Causal Explanation in Interpretive International Studies." European Journal of International Relations. 27(3):936–59.</bibtext> </blist> <blist> <bibtext> O'mahoney J. 2015. " Why did they do that?: the methodology of reasons for action." International Theory. 7(2):231-262.</bibtext> </blist> <blist> <bibtext> Pader E.2014 "Seeing With an Ethnographic Sensibility: Explorations Beneath the Surface of Public Policy." Pp. 194–208 in Interpretation and Methode, edited by Schwartz-Shea P., Yanow D. Armonk: M.E. Sharp.</bibtext> </blist> <blist> <bibtext> Porpora Douglas V.2015. Reconstructing Sociology: The Critical Realist Approach. Cambridge: Cambridge University Press.</bibtext> </blist> <blist> <bibtext> Pouliot Vincent.2014 "Practice Tracing." Pp. 237–59 in Process Tracing: From Metaphor to Analytical Tool, edited by Bennett Andrew, Checkel Jeffrey. Cambridge: Cambridge University Press.</bibtext> </blist> <blist> <bibtext> Reyes V.2020. " Ethnographic Toolkit: Strategic Positionality and Researchers' Visible and Invisible Tools in Field Research." Ethnography. 21(2):220–40.</bibtext> </blist> <blist> <bibtext> Sagi E., Gentner D., Lovett A. 2012. " What Difference Reveals About Similarity." Cognitive Science. 36(6):1019-1050.</bibtext> </blist> <blist> <bibtext> Sayer Andrew.2000. Realism and Social Science. London: Sage Publishing.</bibtext> </blist> <blist> <bibtext> Schaffer Fred.2018. " Two Ways to Compare." Qualitative and Multi-Method Research. 16(1):15–20. doi: https://doi.org/10.5281/zenodo.2562165</bibtext> </blist> <blist> <bibtext> Schatz Edward.2009. Political Ethnography: What Immersion Contributes to the Study of Power. Chicago: University of Chicago Press.</bibtext> </blist> <blist> <bibtext> Schwartz-Shea Peregrine, Yanow Dvora. 2012. Interpretive Research Design. New York: Routledge.</bibtext> </blist> <blist> <bibtext> Schwartz-Shea Peregrine, Yanow Dvora. 2014. Interpretation and Methode. Armonk: M.E. Sharp.</bibtext> </blist> <blist> <bibtext> Sewell William H.1992. " A Theory of Structure: Duality, Agency, and Transformation." American Journal of Sociology. 98(1):1–29.</bibtext> </blist> <blist> <bibtext> Sewell W. H. 2005. Logics of History: Social Theory and Social Transformation. Chicago: University of Chicago Press.</bibtext> </blist> <blist> <bibtext> Small M. L., Cook J. M. 2023. " Using Interviews to Understand Why: Challenges and Strategies in the Study of Motivated Action." Sociological Methods &amp; Research. 52(4):1591–1631. doi: https://doi.org/10.1177/0049124121995552</bibtext> </blist> <blist> <bibtext> Soss Joe.2018. " On Casing a Study Versus Studying a Case." Qualitative and Multi-Method Research. 16(1):1. doi.org/10.5281/zenodo.2562167</bibtext> </blist> <blist> <bibtext> Steel Daniel.2008. Across the Boundaries: Extrapolation in Biology and Social Science. Oxford: Oxford University Press.</bibtext> </blist> <blist> <bibtext> Tavory I., Timmermans S. 2014. Abductive Analysis: Theorizing Qualitative Research. Chicago: University of Chicago Press.</bibtext> </blist> <blist> <bibtext> Trachtenberg Marc.2006. The Craft of International History. Princeton, NJ: Princeton University Press.</bibtext> </blist> <blist> <bibtext> Waldner David.2015. " Process Tracing and Qualitative Causal Inference." Security Studies. 24(2):239–50.</bibtext> </blist> <blist> <bibtext> Winward M.2021. " Intelligence Capacity and Mass Violence: Evidence From Indonesia." Comparative Political Studies. 54(3–4):553–84.</bibtext> </blist> <blist> <bibtext> Yanow Dvora, Schwartz-Shea Peri. 2014. Interpretation and Method. New York: Routledge.</bibtext> </blist> </ref> <ref id="AN0184233958-17"> <title> Footnotes </title> <blist> <bibtext> In this article, we develop a methodological framework for social process tracing, using theoretical and epistemological arguments. We do not generate or analyse any datasets in this article.</bibtext> </blist> <blist> <bibtext> The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.</bibtext> </blist> <blist> <bibtext> The authors disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work was supported by the Samfund og Erhverv, Det Frie Forskningsråd (grant number 0133-00115B).</bibtext> </blist> <blist> <bibtext> Jonas Gejl Kaas https://orcid.org/0000-0002-0522-8640 Mathilde Cecchini https://orcid.org/0000-0003-3946-9116 Derek Beach https://orcid.org/0000-0003-1869-0145</bibtext> </blist> <blist> <bibtext> Data sharing not applicable to this article as no datasets were generated or analysed during the current study.</bibtext> </blist> <blist> <bibtext> Supplemental material and appendix for this article are available online.</bibtext> </blist> <blist> <bibtext> Note that we do use the term 'causal process' throughout this article to denote what links causes and outcomes together in a causal relationship. Given that the term causal mechanism is used when relevant, but given that it is used in quite different ways in the literature, we prefer the umbrella term 'causal process'. In particular, while critical realists use the term to refer to the causal linkages in a process, whereas many process tracers use it to refer to a whole process, using the term creates more confusion than clarity. Further, we use the terms context and structure as synonyms in this article.</bibtext> </blist> <blist> <bibtext> [4] are a bit more nuanced on this point. However, given that in their framework neither the motivations and intentions behind action nor the social context for action are made explicit, theories based on their framework tend to focus on what happened but are less clear about why it happened.</bibtext> </blist> <blist> <bibtext> For critical realists, mechanisms are <emph>not</emph> directly observable, with the result that they tend to be assumed in empirical analysis. Here we are inspired by PT thinking, in which explicit theorisation of linkages gives us clues about what types of observables might be left, even if they are very indirect evidence of the linkage. For instance, in the example above, the causal linkage might be the epistemic authority of actor A as theorized in cognitive psychology. Indirect observables of this behavioural mechanism could include whether actor B accepted the expert judgement quickly with little debate or deliberation.</bibtext> </blist> <blist> <bibtext> Online appendix 1 in the supplemental material provides a more comprehensive overview of positivist and interpretivist approaches. See also Beach ([2], [3]).</bibtext> </blist> <blist> <bibtext> In the following, we will not review the broader debates about the nature of causal mechanisms found in the philosophy of science and amongst social science methodologists, focusing instead on how the different understandings of mechanisms have been implemented in actual PT methodologies. For good reviews of the debates within social science, see [29] and [35]).</bibtext> </blist> <blist> <bibtext> We do not assess process tracing approaches that focus exclusively on the empirical side of tracing.[19] for instance develop a Bayesian-inspired approach to evaluating evidence, but never tell us what we are tracing at the theoretical level.[30] focuses on the types of data that can be used to measure the impact of ideas, but does not discuss conceptually the nature of the ideational processes being traced.</bibtext> </blist> <blist> <bibtext> Given that we are studying reasoning human beings in social processes, we replace the terms entities and activities with actors and actions.</bibtext> </blist> <blist> <bibtext> The importance of <emph>particular</emph> social contexts implies that causal relations are fundamentally <emph>local</emph> (i.e., case specific) in SPT ([50]: 114–8). At the same time, recognising the particularity of how a social causal process played out in the particular case context does <emph>not</emph> imply that aspects of the process cannot resonate with how things worked in other cases. We return to this point in the final section.</bibtext> </blist> <blist> <bibtext> By interpretive interviewing, we mean approaches to interviewing that are oriented towards uncovering the meanings and meaning making (e.g., perceptions and intentions) of research participants ([22]) instead of interview methods that are aimed at gathering information.</bibtext> </blist> <blist> <bibtext> Another advantage of real-time research is that it avoids post hoc reconstructions in which the process seems much more orderly and determined because the outcome is known. If possible, studying a process in real-time helps the research understand the 'fog of war' that often is present in most political processes.</bibtext> </blist> <blist> <bibtext> Sometimes it is not possible to observe actors in context, nor is it feasible to interview them. In that case, social process tracers can turn to analysis of the textual testimonies left by the process. Various methods of interpretive textual analysis – ranging from more traditional historical methods to critical discourse analysis – can be used to generate and analyse data about human meaning-making ([56]: Chapter 1; [61]; [18]).</bibtext> </blist> </ref> <aug> <p>By Jonas Gejl Kaas; Mathilde Cecchini and Derek Beach</p> <p>Reported by Author; Author; Author</p> <p></p> <p>Jonas Gejl Kaas is Assistant Professor at the Department of Strategy and War Studies at the Royal Danish Defence College. In addition to his keen interest in social science methods and methodologies, Kaas' research focuses on Russian and EU defence, security and foreign policy. Currently, Kaas is investigating the opportunities and challenges associated with Denmark's recent entry into the EU's security and defence institutions.</p> <p>Mathilde Cecchini, PhD in Political Science. Cecchini is an Associate Professor at the Department of Political Science and Public Management, University of Southern Denmark. Her research focuses on discrimination, equality, and equity in the street-level bureaucracy. She employs ethnographic and qualitative research methods in her work. She has published in journals such as the Journal of Public Administration Research and Theory, Social Science and Medicine and Health Risk &amp; Society.</p> <p>Derek Beach is a professor of Political Science at Aarhus University, Denmark, where he does research on research methodology and European integration. He has authored articles, chapters, and books on research methodology, policy evaluation, international negotiations, referendums, and European integration, and co-authored the book Process-tracing Methods: Foundations and Guidelines (2019, 2nd edition, University of Michigan Press). He has taught case study methods at numerous workshops and PhD level courses throughout the world, and conducted evaluations at the national and international level. He was an academic fellow at the World Bank's Independent Evaluation Group in spring 2022. He is an academic coordinator of the Methods Excellence Network (MethodsNet).</p> </aug> <nolink nlid="nl1" bibid="bib63" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib39" firstref="ref2"></nolink> <nolink nlid="nl3" bibid="bib11" firstref="ref6"></nolink> <nolink nlid="nl4" bibid="bib24" firstref="ref7"></nolink> <nolink nlid="nl5" bibid="bib40" firstref="ref8"></nolink> <nolink nlid="nl6" bibid="bib13" firstref="ref9"></nolink> <nolink nlid="nl7" bibid="bib46" firstref="ref10"></nolink> <nolink nlid="nl8" bibid="bib50" firstref="ref11"></nolink> <nolink nlid="nl9" bibid="bib17" firstref="ref14"></nolink> <nolink nlid="nl10" bibid="bib44" firstref="ref15"></nolink> <nolink nlid="nl11" bibid="bib41" firstref="ref18"></nolink> <nolink nlid="nl12" bibid="bib42" firstref="ref19"></nolink> <nolink nlid="nl13" bibid="bib54" firstref="ref20"></nolink> <nolink nlid="nl14" bibid="bib22" firstref="ref27"></nolink> <nolink nlid="nl15" bibid="bib15" firstref="ref29"></nolink> <nolink nlid="nl16" bibid="bib10" firstref="ref32"></nolink> <nolink nlid="nl17" bibid="bib62" firstref="ref37"></nolink> <nolink nlid="nl18" bibid="bib12" firstref="ref38"></nolink> <nolink nlid="nl19" bibid="bib28" firstref="ref41"></nolink> <nolink nlid="nl20" bibid="bib31" firstref="ref64"></nolink> <nolink nlid="nl21" bibid="bib58" firstref="ref66"></nolink> <nolink nlid="nl22" bibid="bib14" firstref="ref72"></nolink> <nolink nlid="nl23" bibid="bib35" firstref="ref84"></nolink> <nolink nlid="nl24" bibid="bib36" firstref="ref88"></nolink> <nolink nlid="nl25" bibid="bib20" firstref="ref91"></nolink> <nolink nlid="nl26" bibid="bib27" firstref="ref92"></nolink> <nolink nlid="nl27" bibid="bib23" firstref="ref95"></nolink> <nolink nlid="nl28" bibid="bib51" firstref="ref96"></nolink> <nolink nlid="nl29" bibid="bib57" firstref="ref100"></nolink> <nolink nlid="nl30" bibid="bib33" firstref="ref102"></nolink> <nolink nlid="nl31" bibid="bib52" firstref="ref103"></nolink> <nolink nlid="nl32" bibid="bib47" firstref="ref108"></nolink> <nolink nlid="nl33" bibid="bib59" firstref="ref115"></nolink> <nolink nlid="nl34" bibid="bib53" firstref="ref126"></nolink> <nolink nlid="nl35" bibid="bib43" firstref="ref129"></nolink> <nolink nlid="nl36" bibid="bib32" firstref="ref130"></nolink> <nolink nlid="nl37" bibid="bib16" firstref="ref131"></nolink> <nolink nlid="nl38" bibid="bib37" firstref="ref134"></nolink> <nolink nlid="nl39" bibid="bib19" firstref="ref137"></nolink> <nolink nlid="nl40" bibid="bib25" firstref="ref138"></nolink> <nolink nlid="nl41" bibid="bib49" firstref="ref139"></nolink> |
|---|---|
| Header | DbId: eric DbLabel: ERIC An: EJ1473583 AccessLevel: 3 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 0 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Elucidating the Social -- Developing Social Process Tracing as an Integrative Framework – Name: Language Label: Language Group: Lang Data: English – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Jonas+Gejl+Kaas%22">Jonas Gejl Kaas</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-0522-8640">0000-0002-0522-8640</externalLink>)<br /><searchLink fieldCode="AR" term="%22Mathilde+Cecchini%22">Mathilde Cecchini</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0003-3946-9116">0000-0003-3946-9116</externalLink>)<br /><searchLink fieldCode="AR" term="%22Derek+Beach%22">Derek Beach</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0003-1869-0145">0000-0003-1869-0145</externalLink>) – Name: TitleSource Label: Source Group: Src Data: <searchLink fieldCode="SO" term="%22Sociological+Methods+%26+Research%22"><i>Sociological Methods & Research</i></searchLink>. 2025 54(2):605-645. – Name: Avail Label: Availability Group: Avail Data: SAGE Publications. 2455 Teller Road, Thousand Oaks, CA 91320. Tel: 800-818-7243; Tel: 805-499-9774; Fax: 800-583-2665; e-mail: journals@sagepub.com; Web site: https://sagepub.com – Name: PeerReviewed Label: Peer Reviewed Group: SrcInfo Data: Y – Name: Pages Label: Page Count Group: Src Data: 41 – Name: DatePubCY Label: Publication Date Group: Date Data: 2025 – Name: TypeDocument Label: Document Type Group: TypDoc Data: Journal Articles<br />Reports - Evaluative – Name: Subject Label: Descriptors Group: Su Data: <searchLink fieldCode="DE" term="%22Qualitative+Research%22">Qualitative Research</searchLink><br /><searchLink fieldCode="DE" term="%22Research+Methodology%22">Research Methodology</searchLink><br /><searchLink fieldCode="DE" term="%22Social+Science+Research%22">Social Science Research</searchLink><br /><searchLink fieldCode="DE" term="%22Sociology%22">Sociology</searchLink><br /><searchLink fieldCode="DE" term="%22Social+Environment%22">Social Environment</searchLink><br /><searchLink fieldCode="DE" term="%22Influences%22">Influences</searchLink><br /><searchLink fieldCode="DE" term="%22Epistemology%22">Epistemology</searchLink> – Name: DOI Label: DOI Group: ID Data: 10.1177/00491241231220732 – Name: ISSN Label: ISSN Group: ISSN Data: 0049-1241<br />1552-8294 – Name: Abstract Label: Abstract Group: Ab Data: The promise of process tracing methods is that they can help us better understand how things work in real-world cases. Despite the many advances in the past two decades, we contend that existing accounts result in either under-socialised accounts in which the moves made by actors are studied while neglecting the social dimension of action, or over-socialised accounts that are so focused on social context that they neglect the causal process of actors doing things. In this article, we attempt to bridge this divide by merging the conceptual language from the productive/generative account of mechanisms with realist ideas to develop what we term social process tracing (SPT) as a practical research method that captures both the moves performed by actors, the meanings they ascribe to them, and how the underlying social context shapes how social causal processes play out in real-world cases. After developing ontological and epistemological foundations, the article illustrates how SPT can be used in practice. – Name: AbstractInfo Label: Abstractor Group: Ab Data: As Provided – Name: DateEntry Label: Entry Date Group: Date Data: 2025 – Name: AN Label: Accession Number Group: ID Data: EJ1473583 |
| PLink | https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1473583 |
| RecordInfo | BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.1177/00491241231220732 Languages: – Text: English PhysicalDescription: Pagination: PageCount: 41 StartPage: 605 Subjects: – SubjectFull: Qualitative Research Type: general – SubjectFull: Research Methodology Type: general – SubjectFull: Social Science Research Type: general – SubjectFull: Sociology Type: general – SubjectFull: Social Environment Type: general – SubjectFull: Influences Type: general – SubjectFull: Epistemology Type: general Titles: – TitleFull: Elucidating the Social -- Developing Social Process Tracing as an Integrative Framework Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Jonas Gejl Kaas – PersonEntity: Name: NameFull: Mathilde Cecchini – PersonEntity: Name: NameFull: Derek Beach IsPartOfRelationships: – BibEntity: Dates: – D: 01 M: 05 Type: published Y: 2025 Identifiers: – Type: issn-print Value: 0049-1241 – Type: issn-electronic Value: 1552-8294 Numbering: – Type: volume Value: 54 – Type: issue Value: 2 Titles: – TitleFull: Sociological Methods & Research Type: main |
| ResultId | 1 |
