Do Historians Study the Mechanisms of History? A Sketch

Abstract

In this exploratory sketch, I move across the boundaries of philosophy of historiography to social science and its philosophy. If we want to answer the central question of this chapter, we need to know what types of scientific problems historians are interested in, what history is, and what mechanisms are. I sketch the most prominent theories of social mechanisms in the context of wider ontological approaches. I investigate Mario Bunge’s “Emergentist Systemism,” “Critical Realism” in the tradition of Roy Bhaskar’s influential philosophy, and Daniel Little’s “Methodological Localism.” Since it turns out that mechanisms are taken to be rather different entities, the question is only answered trivially, but some problems are suggested that need to be separated if the debate shall not end up in “mechanism talk.” It is also suggested that philosophers of historiography can find in these debates what they are normally not interested in, that is, science-oriented philosophy of history.

1 The Question: Historians, Mechanisms, and Histories

If we want to approach an answer to the central question raised in the title of this chapter we need to deliver a lot that we are currently incapable of providing. Our question presupposes answers to some of the little but pertinent and notoriously unsolved problems of philosophy of so-called history.

First, we need to know what so-called historians do, that is, what types of problems they claim to solve, try to solve, or even solve in their research before they produce reports on former research, the products of scientific “history” (Topolski 1976). Second, it would be of interest to answer the strange or in these days even seemingly ridiculous question “What is history?” before it perhaps finally makes sense to have a look at whether at least some of the people that are called historians study the mechanisms of history. At this point, some account of what (social) mechanisms are would be of interest.

Unfortunately, the bigger part of philosophy of historiography can help us neither in solving the puzzle of what historians do nor of what it is that they perhaps study beyond the so-called historical sources, if we still dare to commit ourselves to realist assumptions at all. Fortunately, there is much and growing literature on social mechanisms in historico-social science discourses and the realist part of its philosophy that might help us in advancing towards an answer.¹

What I will try to do in the following is to present some of the independent theories regarding the ontology of (social) mechanisms proposed by philosophers of social science over the past 40 years and to sketch some implications these philosophies have with regard to answering the unpopular question “What is history?” or the rediscovery of that question. Since all these philosophies claim that hinting at some mechanism is central to (scientific) explanation of societal change or stasis, we get to the kernel of these issues as well, since one objection to realist models of mechanismic explanation in the sciences of history/historiography might be that there are no such “things.”

We start, in a rough chronological order, with “Emergentist Systemism” (ES), established by Mario Bunge,² and the role of mechanisms therein (Sect. 10.2) before we have a look at some aspects of “Critical Realism” (CR), initiated by Roy Bhaskar (Sect. 10.3). Then we approach the ontology called “Methodological Localism” (ML) by Daniel Little (Sect. 10.4). This selection is justified by the observation that these philosophers are prominently discussed in social scientific journals and their contributions are thus believed to be relevant to social scientific practice. Furthermore, they are all realists about mechanisms.

Since the common use of the term “mechanism” in philosophical and sociological literature in recent times suggests a degree of convergence in the discourse that might turn out to be misleading, I will situate the respective theories of mechanisms in the context of wider ontological systems, that is, some exegesis is unavoidable and necessary. At this point, however, I will deliberately ignore theories that derive more directly from within the social sciences.³ The resulting and admittedly wordy exegesis is necessary because due to conceptual ambiguities in the wide discourse around the notion of a “social mechanism,” it is quite unclear (i) what such a social mechanism might be, (ii) which problems are to be solved in these debates, (iii) what their solution is, and whether (iv) this solution is necessary. I am going to sketch issues around (i) and (ii). The biggest problem lurking in the background is, of course, that sociologists and historians are in no agreement about the objects of research for social science, about what is to be explained if anything, and how such explanations can and should be approached (see, e.g., Blaikie 1993). Because this seems to be so, my strategy in this sketch is basically to take nothing for granted, not even a single concept such as “history” that appears to be innocent. My faint hope is that thereby possible problems become clearer that are perhaps even worth solving and that scholars who share the experience of fearing to drown in these debates might find some rescue in the following lines.

The primary result will be meager. Given the following reconstruction, there is no obvious reason to believe that historians are not interested in mechanisms. The secondary result is exactly that it is by now unclear what a social mechanism is, and the differences in theories of social mechanisms are made explicit. The tertiary result is that if there is a problem about social mechanisms, this problem comes in a bundle with others fairly familiar from social theory. The question “What is a history?” might express one of those problems. This result is achieved by attending in detail to the differences in theories of social mechanisms.

Anyway, a word of caution seems to be in order at this point. I do not claim to be an expert on any of these philosophical systems. The apology I offer for discussing them nevertheless is that this literature has hardly ever been discussed together in a comparative fashion, although people rather frequently quote from each of these positions and others, as if it were unquestionable that those positions treat the same stuff behind the veil of the term “social mechanism” and, furthermore, as if it were unproblematic if they were not. My hunch is that this is questionable and problematic. This chapter might therefore be of interest to social theorists and researchers, including historians, who do not believe that this all just amounts to “mechanism talk.”⁴

2 Mechanisms and Emergentist Systemism (Mario Bunge)

What does history consist of? According to Bunge, the world is a world of concrete, that is, material things and consists of nothing else. Thus, the so-called social world is equally supposed to be a world of things. At its heart, the project of scientific historians, once called historiologists (1985, p. 193), and sociologists consists in the study of such “things” or “social matter” (1974, p. 445, 1981, p. 5).⁵

As it is well known or at least claimed quite often, many contemporary historians and perhaps all sociologists seem to be hardly interested in individual persons and their actions but rather in so-called social facts. But what is such a social fact? According to Bunge, a fact is the being in a state or a change in the state of any material thing, so that all facts are moreover singular and positive (1996, p. 17, 2006, p. 17). A nonsocial, for example, a mental fact, then, is the being in a state of a brain or a change in the state of a brain of a higher-order animal, given Bunge’s psycho-neuronal emergentist monism (1984). But what are social facts or the analogue to brains?

The basic pillars of the “systemic approach” (2006, p. 128) are the concepts of a system and of emergence (2001a, 2003a). According to Bunge, the world is a world of systems (1979a). Everything that is not a system is at least a part of one, or if it is not yet, or merely not for the moment, such a part, it will become a component of a larger whole.⁶ Whereas some physicists might study elementary particles that are things but not systems, the rest of the scientific community, including historiologists, studies complex concrete systems (1979b, 1992b, 1993b).

A concrete system is a bundle of real things held together by some bonds or forces, behaving as a unit in some respects and (except for the universe as a whole) embedded in some environment. (Bunge 1997, p. 415)

Human social systems are obviously composed of persons and the artifacts they have built (1996, p. 21). And they are constructed, maintained, altered, and primarily destructed by persons. Thus, to qualify as such, human social systems have to be composed of at least two persons engaging in some social relation or interaction, but they can be as big as is compatible with the laws of social systems, that is, laws that are assumed to exist by Bunge in contrast to many contemporary sociologists (2007). Such social systems can be short living as well as long lasting, either rather local or spread all over the globe. To give some examples, “families, afternoon coffee parties, football teams, school classes, working groups in organizations (Institutsarbeitsgruppen)” (Bunge and Mahner 2004, p. 86; any translation D.P.) or even the world economic system, if it qualifies as such, are social systems on Bunge’s account.

The traditional and controversial question concerning the existence of social wholes is therefore affirmed in ES. It seems to be equally important that there are two types of wholes according to Bunge. A system is not a mere aggregate or heap because it has emergent properties: most importantly the system structure that determines that the parts of the system are at least minimally integrated or cohesive, so that the system “behaves” as a whole in some respect.⁷ Due to this real or ontic integration or connection, for example, “a labor union is a social system and therefore just as concrete and real as its members” (1998a, p. 69). Mere aggregates or heaps as, for instance, a bunch of people accidentally attending a football (soccer) match “together” or a selection of people sharing some more or less relevant property (e.g., bourgeoisie) do not constitute systems, although they might qualify as wholes.⁸

Disentangling the short definition quoted above and thereby summarizing the former, every (social) system at any point in time is supposed to be minimally characterized by a definite composition, a definite structure, and a definite environment. This yields in scientific reconstruction what Bunge formerly called the minimal CESM-Model of a system (e.g., 1995, 14f.). Whereas the composition of a system is the set of its components that are as concrete and as real as the system as a whole, the structure (sometimes also called “organization” or “architecture”; 2004a, p. 188) of the system is the set of all the relations holding among the system components. The environment is said to be the set of relations the system components hold with things (systems) that are not part of the system. For example, in an army relations of command are part of the endostructure of the system, and relations of combat or supply belong to its exostructure (2003c, p. 277).

Skipping some interesting problems concerning systems ontology relevant to historico-social scientific practice,⁹ subsets of the overall structure of the system are what Bunge once called the “system structure (or bondage)” and the “spatial structure (or configuration)” of any system (1979a, p. 11). Whereas spatial and temporal relations are often of minor interest to historians and social scientists because they presumably make not much of a difference to the whole or its parts, social relations make, by definition, a difference to the relata because they are bonds, ties, couplings, or links.¹⁰

At this point we only have to mention that these relations are in part dynamic and causal. They are interactions if they are “root (or basic) social relations” (1979a, p. 222). But they might also consist merely in being a part of a whole, as in being a citizen or being married to someone, the latter being perhaps the prime example of a social relation that does not necessarily imply interaction. Other psychosocial bonds as affection and interest (1979b, p. 20), loyalty (2003a, p. 20), love (2004a, p. 189), or the rules inherent in or constitutive of social relationships (1979a, p. 196) hardly count as causal as such in Bunge’s system, in contrast to CR and ML.

Furthermore, different social systems might be related among themselves as wholes to constitute higher-level systems. If one system acts on another system, this is said to be one case of “social power.” Whole social systems might also be dynamically related to persons; this is the second type of social power in ES (2009b, p. 189). But barring miracles, every relation between social systems, that is every non-basic social relation, is executed by a person-to-person or basic social relation, that is, if it happens to be causal. A further reason to call interactions the basic social relations in ES is that social systems and the more permanent bonds that hold them together diachronically emerge from personal interaction (2001b, p. 134).¹¹

Thus, structures of systems in general and social structures specifically are not quite easy to grasp in social science or philosophy. According to ES, the latter are mediated ultimately by the brains of actors because social relations depend on them. In Bunge’s view, this is the truth of ontological individualism (1997, p. 453, 2008, p. 61). However, social relations are claimed to be emergent from interaction of people that are embedded in a social and natural environment, not on free floating ideas. Moreover according to ES, rather obvious but often ignored in traditional philosophy of historiography, social “systems have no brains, hence they have no intentions” (2004b, p. 376), although all their social properties and most of their changes are due to people acting and interacting.

Another important point for our purpose in addition to the reality of (social) wholes or “totalities” (1999, p. 8) as such is that they are claimed to possess properties of their own, that is, as wholes, some of which we did already encounter above, that is, societal or systemic properties, which are the reason to grant these totalities an ontological status of their own in the first place. In Bungean ontology, there are again two types of holistic properties, namely, emergent and resultant properties. What is an emergent property according to ES?

P is an emergent property of a thing b if and only if b is a complex thing (system) no component of which possesses P, or b is an individual that possesses P by virtue of being a component of a system (i.e., b would not possess P if it were independent or isolated). (Bunge 1996, p. 20, cf. 2003a, p. 14; Mahner 2001, p. 79)

Thus, there are two types of emergent properties. The former are called global or intrinsic emergent properties, the latter relational or structural emergent properties (Bunge and Mahner 2004, p. 79). If the property P is a property of a whole, but not emergent, it is called resultant. It is already possessed by the system components and merely aggregates to an unstructured whole.

Let us furnish a bunch of arbitrarily selected examples for these types of social properties. Global emergent properties are supposed to be social stratification, cohesion, mobility, stability, economic growth, form of government, political stability, mode of production, undergoing a social revolution, size of territory, or population. Examples of structural emergent properties are role, civil right, scarcity, price,¹² and every form of being coupled to other system components at all, so that every time a social system is built, at least two people are supposed to acquire structurally emergent properties (2003a, p. 78). In contrast to resultant properties (e.g., the total consumption of bubble gum in a society), emergent properties are ontologically irreducible to (though explainable by) properties of the system components, as is the total production of the economy of a society or being the goalkeeper in a football (soccer) team (1979b, p. 22; Mahner 2001, pp. 170–294).

Framed in a few fancy slogans reminiscent of the philosophical tradition, we perhaps get near the core of systemic ontology: There are not just capitalists but capitalist economies. There are not just protestants but people organized in churches. There are not only soldiers, but soldiers organized in nested military organizations with military subunits allegedly characterized by an emergent firing power manifesting in the field.

This is the static view of the world in Bungean ontology. Since we started out this paragraph with an utterly vague question about the constituents of history, we should not finish it without saying something about changes occurring in the world, since once upon a time the “historically minded” thought “history” in the potentially ontic and epistemic meaning to have something to do with change (Topolski 1976).

The static CES-Model of a system presupposes, of course, a version of ontological realism, the thesis that the (social) world does not care about it being studied – a thesis rather uncommon in contemporary sociohistorical academia (1993a). Given this theory, every concrete (social) system at every point in time possesses a definite number of properties. This is the state of the system at a given moment.

The dynamics of the system consist then, according to Bunge, in the changes of state of the system in the epoch of its existence. In ES a change of state of a thing is an event. A process is a sequence of changes of state or events. And last but not least, the history of the thing or system is supposed to be nothing but the total sequence of its states during its journey through the “course of history,” the totality of its changes in properties, including the acquisition (or processual emergence) of relatively or absolutely new properties (2003a, p. 13):

Whereas a process (or partial history) of a thing is any (ordered) sequence of states of (or, alternatively, events in) a thing, the (total) history of a thing is the ordered set of all its successive states (or events). (Bunge and Mahner 1997, p. 19, cf. 2004, p. 58; Bunge 2003c, p. 128)

At least some interesting implications of ES for any philosophy of history are noteworthy, although they are inconspicuous. Although there are quite many, we will list only three of them. First, every history is the history of a thing and every event occurs in some thing. In contrast to what can frequently be found in philosophy of historiography, social theory, and historiology, there are no processes in themselves, properties of events, histories of events, or evolving histories, and society is not a process.¹³ Furthermore, history is not “a tale told by an idiot” (Sewell 2005, p. 102), but strictly speaking nonexistent for being a summative concept representing all histories of concrete things (1996, p. 26), which are on their part as real as anything could possibly be. For not being things, histories of things do not have properties and therefore do not change or do anything at all (Bunge and Mahner 2004, p. 68). Second, according to this explicit ontology, most of the talk about history implicit in philosophy of historiography is either nonsensical or false. The latter holds for “history was in the past” (Tucker 2012, p. 277) or “All history is the history of thought” (Collingwood 1994, p. 215), the former for common slogans as “the end of history,” “the return of history,” or “the course of history.”¹⁴ Third, since societies are considered to be real in ES and to have genuine properties that might change or even be newly acquired, there is a real history of every society or of any social system of whatever scale that can in principle become an object of study. This might seem trivial but it is not given that this is impossible in ontological individualism or “historical individualism” (2000b) because on that account there is no fall of any empire since there are no empires in the first place.¹⁵

Accordingly, we can finally say what social facts are in Bunge’s philosophy. They are states of social systems or their changes. Social change is simply defined as change in societal properties (1979a, p. 235). And although social states and their changes occur outside any individual’s mind, they depend on social interaction and so finally on the activation of the furniture of the minding brains, to put it in Bungean terms, of concrete persons (2008, 60f.) or the “individual-in-society” (1979b, p. 17). This individual is on its part characterized by structural emergent properties and reacts furthermore to social facts as she perceives them, although these are not reducible to her interpretations of reality (1997, p. 453).

In order to provide us with further material to contrast ES with the other ontologies involving mechanisms, two further implications should be noted. Since societies do, according to Bunge, not hover above individuals that compose them and, moreover, social relations are constituted by interacting persons, whereas the set of all relations among the system components represents its structure, all of these are no candidates for being causes. Especially, “[t]here is no action of the whole on its parts (…)” (1979a, p. 39). More concisely, because in ES events (changes) are the relata in causal relations, neither properties nor states or conditions are considered as causes. The expression “a causes b” is therefore a short version of “a change in the state of thing a produces a change in the state of thing b” (Bunge and Mahner 2004, p. 95). Since Bunge conceptualizes the productive or generative aspect of causation in terms of different forms of energy transfer (1996, p. 31, 1999, p. 27), it seems to follow that only concrete persons as they are working with, talking to, or shooting each other cause something in the so-called social world, alongside changes in technological and natural systems, of course. The former quote thus continues: “rather, there are actions of some components on others” (1979a, p. 39). In a nutshell, social properties are real, given ES, but do not cause social or psychic events.¹⁶

After all these seemingly just preliminary remarks, what is a social mechanism according to ES? Is it a more or less stable complex thing (e.g., a family, the German economy, a state, a university, a mafia family) or a process (e.g., child rearing, production, growth, innovation, social control, academic tenure, dealing drugs) or a property (e.g., cohesion, productivity, structure)?¹⁷ Though we have to shorten the story considerably, in this ontology the concept of a mechanism as a structured object is unnecessary due to the concept of a concrete complex system, its components, and its structure.¹⁸ Consequently, “mechanisms are not things but processes in systems” (2009, p. 19). Moreover, “[e]very mechanism is a process, but the converse is false” (1999, p. 24), since not every systemic change is a mechanismic process, for example, economic growth, the spread of brilliant ideas among social systems or dancing. Therefore, mechanisms are not identical to the history of a system, the boundary concept of change (of a thing) in Bunge’s system. Mechanisms are ultimately supposed to be a subset of the totality of processes occurring in a concrete complex system, namely those that involve the properties that are essential to the kind of system we are concerned with, their specific processes or functions that keep the system alive or “running”, “make it work” the way it “normally” does or “what it is” (e.g., 2003b, p. 146).¹⁹ Accordingly, further basic assumptions of this ontology are that every (social) system is endowed with at least one specific function, mechanismic or essential process (1995, p. 43; 2010, p. 376), that there are mechanisms of change and mechanisms that prevent or control change or keep the system in a state. That is to say, there are mechanisms for “either the emergence of a property or another process” (2003a, 20f.; 2010, p. 379). A social mechanism is, of course, just a mechanism in a social system, that is “a process involving at least two agents engaged in forming, maintaining, transforming or dismantling a social system” (1999, p. 57).²⁰

Where are mechanisms when it comes to analyze the so-called social world? Are they to be found on the side of agency or structure? In ES they are neither to be found in the heads of people, that is why they cannot be reenacted or verstanden, nor in a mysterious whole above people or social structure, but “in or among” social systems and a part of the systemic processes unfolding therein (1999, p. 57). But social interaction is not only said to be the “source of system,” that is, of their diachronic emergence under the condition that one essential process gets running, but also the “fuel of mechanism” (2001b, p. 134). More interesting and explanatory accounts of social systems or their histories thus also have to include mechanisms in CESM-models (1998a; 2002).²¹

Let us finally deliver some suggestive examples that might help us in answering the main question. The specific function or part of the mechanism of the postal system is to distribute the mail (1995, p. 43). The finance authorities collect taxes or fail therein, and in schools, pupils are taught and teachers managed (Bunge and Mahner 2004, p. 75). And as any enthusiast for movies believes to know, blackmail, drug dealing, and intimidation of judges are specialties of mafia families and, according to Bunge, mechanisms or part of the total mechanism that keeps these systems running (2008, p. 53).

3 Mechanisms and Critical Realism (Roy Bhaskar)

What does history consist of in CR, that is, the philosophy that gained wide influence in social theoretical discourses and philosophy of social science over the past decades?²² At first glance, there seem to be some similarities to what we found in Bungean ES. The world is said to consist of “things” or “mechanisms.”²³ The overall conception of this ontology is also materialist but not reductionist, because psychological and social levels or the properties of its members are supposed to be genuinely real, that is, “emergent” (1986, p. 104, 1989, p. 91) from the preceding levels.²⁴ As in the case of Bunge, emergence is an ontological, not an epistemological, concept (1978, p. 113). The resulting ontology is therefore called “synchronic emergent powers materialism” (SEPM).²⁵ Moreover, Bhaskar at times equally distinguishes between mere aggregates (heaps) and what he calls “totalities,” which are said to be “characterized by an emergent principle of structure” (1994, p. 80). Although this suggests that structures are properties of totalities, on the same occasion totalities are said to be structures.²⁶

In SEPM “people and society” are accordingly supposed to be “radically different kinds of thing” (1979, p. 42), or put in slightly different terms, “while the properties and powers of individuals and societies are necessary for one another, they are irreducible to one another” (1989, p. 63). However, whether ES and CR are compatible is yet an open question, given the first tendentious quotes that seemingly equate things with, in Bungean parlance, “their” mechanisms, while the second equate totalities with structures, whereas the concept of a power is furthermore absent from ES.²⁷

The things that are supposed to constitute the world according to Bhaskarian CR are “causal agents.” Causal agents are those entities endowed with causal powers.

To say that x has the power to do φ is to say that it will do φ in the appropriate circumstances in virtue of its nature (e.g. structure or constitution); that is to say it will do it in virtue of its being the kind of thing that it is. (Bhaskar 1978, p. 237)

These natures are on other occasions also called “essences” or straightforwardly “structure.” Thus, to ascribe a power (to some thing) amounts to distinguishing accidental from essential properties of the thing. Only if the “intrinsic structure or essential nature of a thing” changes that the powers and tendencies of the thing change. This would not be the case if only some conditions for its manifestation or relations to other things changed (1978, p. 97).

A causal agent is then nothing more, but also nothing less, than “anything which is capable of bringing about a change in something (including itself)” (Bhaskar 1978, p. 109). These things or agents are the bearers of at least two types of causal power, namely, (i) powers and (ii) liabilities. The former are held to be capacities to produce changes actively, whereas the latter are conceived to be capacities to suffer or passive powers (1978, p. 87). To use a common example, a fire is supposed to have the power to burn people that are liable to be burned.²⁸

Causal powers, in turn, are the foundation of tendencies. There were supposed to be two types of tendencies in the 1970s (1978, p. 230); later on Bhaskar distinguished seven types (1994, p. 83). Tendencies, if exercised, are said to ground the normic behavior of things (1978, p. 106) and are among others the referents of normic law statements. As it seems, Bhaskar shares Bunge’s belief that there are “causal laws, generalities, at work in social life” (1979, p. 27). But what are tendencies more exactly? We find that tendencies “are roughly powers which may be exercised unfulfilled” (1978, p. 98).

Tendencies may be possessed unexercised, exercised unrealized, and realized unperceived (or) undetected by men; they may also be transformed. (Bhaskar 1978, p. 18)

If tendencies are possessed unexercised, they seem to be mere powers. The thing is said to possess the tendency even though it is not yet tending to do anything. Tendencies are “dynamized” powers or powers “set in motion,” although exercised powers (tendencies) need not manifest themselves in open systems (1978, p. 50), given that other tendencies might counteract. If a power is triggered, it (or the thing bearing it) is claimed to tend towards its manifestation. It acts “transfactually” and would actualize if it is not counteracted by other actualized powers. In other words and terminology that is hardly used by critical realists, powers seem to be dispositions and tendencies are dispositions that are triggered or released and manifest themselves only ceteris paribus.²⁹

In summary, we can say that whereas the basis or foundation of tendencies seem to be powers or that tendencies are thought to be a mode of being of powers, these on their part seem to have a basis from which they are supposed to emerge synchronically. The bases of powers are the “natures” of things (1978, p. 178) or their “real essences” that are supposed to be their “intrinsic structures” (1978, p. 174).

If our reading is not structurally beside the point, we can preliminarily picture the basic outline of this ontology in the following way:

that is, roughly, structures ground powers which are the foundation of tendencies.³⁰

Even at this point, before we even got near the social ontology of CR, we can sketch some hypotheses of CR’s philosophy of history. First, history is not just a sequence of some such events, but something in “the course of” which something real persists or even radically changes or is transformed as far as to eventually produce qualitative novelty.³¹ Secondly, given that Bhaskar believes that something persisting is the basis for a “genuine concept of change, and hence history” (1979, p. 47), the question arises what this something is that endures, changes, or is transformed. In his later work, causal powers are said to be “processes-entified-in-products” (1993, p. 52) and Bhaskar even goes as far as to write about the “presence of the past,” its causal efficacy and of the “presence of the future” (1993, 140ff.). Accordingly, one answer to the question which stuff is transformed during the “course of history” and even determines in some sense “the future” that we can extract from CR is that this stuff is causal powers. A “historical event,” then, is not any event whatsoever as in ES, but an event that significantly changes or transforms historical things and their powers (1979, p. 24).

Where do mechanisms enter the ontological picture? A “generative mechanism is nothing other than a way of acting of a thing. It endures and under appropriate circumstances is exercised as long as the properties that account for it persist” (1978, 51f.), that is, as long as the natures, essences, or intrinsic structures do not change significantly. “Mechanisms are enduring; they are nothing but the powers of things. Things, unlike events (which are changes in them), persist” (1978, p. 221).³²

On this reading we can substitute the former summary of basic CR ontology by the following schema so that mechanisms take the place of powers:

that is, roughly, structures are the foundations of mechanisms which ground tendencies.³³

Whereas in Bungean systemism mechanisms are actual or manifest and furthermore processes, in Bhaskarian realism they are, or seem to be, dispositional properties or powers. Yet on another reading, they might be something in between. In Plato Etc. we find the statement that structures possess causal powers, “which, when triggered or released, act, as generative mechanisms” (1994, p. 23). Given this reading, mechanisms are powers triggered or dynamized, the role taken above by tendencies. Accordingly, the concepts of power and tendency are said to “come together in the concept of generative mechanism, which may be either or both” in standard CR literature (Hartwig 2007, p. 57).

When stimulated, released or enabled, then, powers and generative mechanisms are tendencies (…). Where a thing just is its powers and tendencies (mechanisms), these are the same as structure. (…) Otherwise mechanisms and structures are distinct, i.e., mechanisms (powers and tendencies) are of (instantiated in) structured things. (Hartwig 2007, p. 57)

On the second reading, then, we get something like this schema:

Structures (or natures or essences) → powers → mechanisms

that is, roughly, that structures ground powers, which, when triggered, transfactually act as mechanisms or are mechanisms if triggered.³⁴ In the case of people, who, in the later work (1993, p. 165), are said to be an example of things that just are powers, we get:

that is, some structures are mechanisms (or ensembles of powers). With a little slip of the pen, we might summarize the forgoing in the following schema:

Mechanism (thing, structured thing, structure) → mechanism (power) → mechanism (tendency).³⁵

Given the foregoing, what are mechanisms in CR? Are they complex things, properties, or processes? As far as I can see, this happens to be rather unclear. They sometimes appear to be complex “things,” sometimes events, whereas the primary referents are non-manifest properties capable of “doing” something or “bringing about” changes in things. Given this multiplicity of meanings of the mechanism concept in CR, it is not surprising that one gets the impression that much of the literature interprets these CR mechanisms implicitly as complex objects (mechanisms as systems)³⁶ characterized by recurrent processes (mechanisms as processes)³⁷ due to the properties of the component things (mechanisms as powers) and their relation, organization (mechanisms as structures or structural powers), or interaction.

But this reading seems to be slightly beside the point, because given the former interpretation and the strict distinction between the domains of “the real” and “the actual,” the former including mechanisms (powers), the latter encompassing events as changes in things, hardly anything is happening in our world yet, given that even triggered powers (mechanisms or tendencies), though pinched towards moving, might, by definition, not end up in actual changes in the world, even when they are said to be acting in some sense (transfactually), since they might remain unrealized though exercised, if they happen to be exercised at all. Bungean history is, as it seems, still dormant in the story told until now.

Given that social mechanisms were in ES said to be neither social things nor persons but systemic processes, where are mechanisms to be found in CR? Since they are (primarily) powers, mechanisms are where powers are to be found. And according to CR’s social ontology, there are social structures or societies and persons, which are both claimed to be ensembles of powers. These types of powers finally meet each other in “processes” or “rhythms.”³⁸ In the social sphere

process [is] where structure meets events; that is, in the study of the mode of becoming, bestaying and begoing of a structure or thing, i.e. of its genesis in, distantiation over and transformation across space-time. Process is not an ontological category apart from structure and event; it just is a structure (or thing), considered under the aspect of its story (sic!) – or formation, reformation and transformation – in time. (Bhaskar 1986, p. 215)

In ways similar to Little’s ML framework, in the social sphere “social structures” or “social forms” (1983, p. 85) fuse in actions with agential causal powers (mechanisms) and natural causal mechanisms (powers) to lead to changes in the maintenance or transformation of “social structures” (mechanisms). Societal change is thus a change in or a transformation of societal powers. A society, a causally inert system of systems in ES, is said to be “a complex and causally efficacious whole – a totality (…) which is being continually transformed in practice” (1989, 87f.). Although:

Society (…) is an articulated ensemble of tendencies and powers which, unlike natural ones, exist only as long as they (or at least some of them) are being exercised; are exercised in the last instance via the intentional activity of human beings; and are not necessarily space-time invariant. (Bhaskar 1989, p. 79, cf. 1978, p. 196)³⁹

It is noteworthy that the powers in the social case are dispositions that are always actual or at least acting transfactually as tendencies or exercised powers, but are never purely dispositional, given the former quote.⁴⁰

In the famous “Transformational Model of Social Activity” (TMSA) or the later “Social Cube,” social life is conceptualized as work on a preexisting social world, that is, as work on “social structure.” This move yields the hypothesis that “in every process of productive activity a material as well as an efficient cause is necessary” (1979, p. 43, 1986, p. 119). Accordingly, in classical CR “social structures” are supposed to sponsor the world with “social material causes” (1993, p. 155) that “govern, enable and constrain” (1986, p. 130) individual action. Accordingly, these powers are, first of all, believed to be predating these actions, whereas persons change, reproduce, and transform “social structures” that enable and constrain their actions. The truth of social individualism according to Bhaskarian CR is that “people are the only moving forces in history” (1989, p. 81).

Since nothing makes things clearer than examples, let us have a look at some arbitrarily selected instances of social structures that carry powers or simply are powers. Examples for “social entities” are “institutions, traditions, networks of relations and the like” (1989, p. 175); the former two are also on occasion called “emergent social things” (1993, p. 54). Structures are, furthermore, said to be “religious rites established by the practices of the long dead” (1994, p. 95); “the economy, the state, the family” (1989, p. 4); “Nazism, bureaucracy and (…) capitalism”; and “buildings we have, the stock-market, the whole financial economic system”; they are claimed to be “everything that is there before any given voluntaristic act” (2001, 28f.), as are “languages,” “systems of belief, cultural and ethical norms” (1978, p. 196). Anything that constrains or enables individual actions is structural and powerful: even “stories are social structures” (2001, p. 36) as well as “the age structure of a population, or the occupational structure of a population, or the academic status of a population or perhaps the class structure of a population” (ibid. p. 37). Furthermore, “social structures and their generative mechanisms” are said to be exemplified by “ways of cooking, making micro-chips or production generally” (1993, p. 155).

The heterogeneity of these examples of social structures would not be problematic if these structures were not believed to be mechanisms or powers, which are what distinguishes causal agents, namely, “anything that is capable of bringing about a change in something” (1978, p. 109), from non-agents. If our story above is correct, in order to get powers in CR, we need “the key concept of a causal agent” (1978, p. 77). The question is whether we find those agents with essences and emergent powers in these allegedly social examples, for example, traditions, stories, norms, and social relations.

The problem seems rather obvious. First, if there are no such complex objects to be found whereof those powers are properties, CR faces the problem that nothing seems to justify the assumption of “social causal powers” any more.⁴¹ If we do not need agents as structured objects or systems in Bungean terms to get powers, then we seem to face in social ontology a different concept of powers than in the materialist ontology for the natural sciences.⁴² Second, it seems to be controversial to frame “social material causality” in terms of the original causal powers account, since this seems to be in slight conflict with the productive or generative account of causation.⁴³ Anyway, it might be worth believing that a fistful of dynamite has the power to blow one’s brains out and actually does so if it is triggered and, finally, acts undisturbed. But, as any realist knows, this seems to be something different from the case in which a fistful of dollars results in my buying a cuckoo clock if it is handed over the counter. One of the powers of a fistful of dollars might be to be liable to be burned, but that one of its powers is to buy a clock might be controversial, though everybody seems to understand what is going on in this social episode in common sense terms.⁴⁴ Third, if we stick to the belief that powers have to confront triggers or releasers or what have you, we might ask whether people are the conditions for the manifestation of a social causal power (mechanism), for example, of a norm. Or we can ask whether social causal powers (mechanisms) are merely the conditions for the manifestation of individual powers. Since a condition for the manifestation of a disposition of a thing is another powerful thing that is disposed to trigger the disposition of the former if they happen to join one another in an event, we might say that personal and social powers have somehow been made for one another. But however we twist and turn, in most examples of social causal powers, we lack the second and furthermore social “agent” and a property that might be considered as a candidate for a disposition. What is, say, the disposition of, or dispositional about, “the age structure of a population”? Of course, if it would turn out that there are no plausible candidates for social powers, there would be no social mechanisms, given CR ontology. But, of course, this is an open question.

4 Mechanisms and Methodological Localism (Daniel Little)

Daniel Little is another outstanding philosopher who has worked on the problem of sociohistorical explanation, mechanisms, and social ontology fairly independently of the authors discussed so far since the 1980s.⁴⁵ Methodological Localism revolves around three ontological hypotheses: (i) the social causation thesis, (ii) the microfoundations thesis, and (iii) the agency-structure thesis. The basic methodological thesis is that sociohistorical scientific explanations do actually invoke “mechanisms” and also have to do so in order to be appropriate. This is the epistemic counterpart to the ontological hypothesis that “social causation” works somehow but only and solely through individual agency or action. A microfoundational explanatory account then basically provides an answer to the question how macro-powers get to manifest their dispositional natures in micro-action. The main assumptions are the following:

Social structures and institutions have causal properties and effects that play an important role within historical change (the social causation thesis). They exercise their causal powers through their influence on individual actions, beliefs, values and choices (the microfoundations thesis). Structures are themselves influenced by individuals, so social causation and agency represent an ongoing iterative process (the agency-structure thesis). (Little 2010, p. 97; cf. 2007, p. 358; 2009, p. 169)

Methodological Localism is in somewhat more detail the hybrid out of the theses that “[s]ocial entities supervene upon individuals” (2007, p. 367) or “upon individuals and institutions” (2010, p. 56). Those social entities are “the sum of the constellation of socially situated individuals and institutions” (2007, p. 354) so “that all social facts are carried by socially constructed individuals in action” (2009, p. 159), which leads to “the idea that the causal nexus of the social world is the behaviours of socially situated and socially constructed individuals” (2011, p. 293).

Obviously, at least at first glance, this is rather uncontroversial. And equally salient, this ontological model is almost a reinvention of Bhaskar’s TMSA. Accordingly, “influence” should here strictly be read in terms of causation.⁴⁶ We should consequently emphasize again the hypothesis that there is social stuff and that this stuff is furthermore supposed to be a bearer of “causal powers” and a producer or generator of individual action and derivative social facts.⁴⁷ Given the problems inherent in CR ontology, much depends on the ML theory of social structures and their powers, the mechanisms of CR.

In accordance with Bhaskar, who claimed that “social mechanisms and structures generating social phenomena” (1986, p. 122) are only relatively enduring, Little claims in accordance with most historians that “all social structures are historically rooted; so there is no fixed ‘essential’ nature of a state or economy” (2010, p. 75). That is, “historical individuals” (2010, p. 42) or “historical entities” (ibid. p. 47) always “‘morph’ over time” (ibid. p. 62). The central problem that guides the search for a theory of social mechanisms and mechanismic explanation is therefore the same as in Bhaskar’s TMSA framework, “Agents constitute structures; and agents are in turn constituted by structures,” which is then interpreted as some form of “ongoing mutual influence” or causation “within and across generations” (2007, p. 356).⁴⁸ The three main assumptions quoted above say in a nutshell that “[m]acro entities exercise causal properties through the individuals who constitute them at a given time” (2007, p. 366). Given that Little shares with Bhaskar and Bunge the aim of finding a middle way between ontological and methodological individualism and holism (2007, p. 346), we stick to the tradition established above by asking what does the so-called social world and perhaps history consist of and what are mechanisms? More to the point, what are social facts, macro-entities, or social structures in ML? What is it for such a “social thing” (2010, p. 72) to possess a causal property? And how do they cause individual actions and derivative social effects?

First of all, according to ML, there are “social things,” for example, “relations, institutions, practices, organizations” (2010, p. 72), “historical individuals” or the “concrete social formation” (2010, p. 47), and furthermore “things as revolutions or capitalist economies” (2010, p. 42). At times, Little seems to suggest that “things” are that type of entity that bears “causal powers”.

Second, in ML society, that is, a system of systems in ES and an ensemble of powers and internally related social positions in CR is thought to “consist of specific social, economic, and political institutions, mentalities and systems of beliefs and values, and higher-level structures that are composed of these institutions, practices, and mentalities.” Agents are moreover claimed to “constitute” or “populate” these “social factors” and to act “within the context of these structures.” Thus they “affect the future states of the system while being prompted or constrained by existing structures and mentalities” (2007, p. 353). This, then, amounts to the hypothesis that individuals are always “socially situated” in the sense that their “domain of choice” is restricted by existing “social institutions,” that is, these decisions are caused by the given circumstances and their powers; in contrast to ES and in accordance with CR. Agents are furthermore said to possess “social properties,” that is, they “exist in social relations and social institutions.”⁴⁹ They are, moreover, “socially constructed” in the sense that their furniture of mind is acquired through interaction (2007, 353f., 2009, p. 174). Thus, agents are sometimes claimed to be in “social states” exemplified by “beliefs, intentions, reasoning, dispositions and histories” (2010, p. 59; 2007, p. 352).⁵⁰

In summary, “the social” consists, according to ML, of individual agents that constitute, compose, or “embody” institutions. “Social institutions and organizations come together to constitute complexes of institutions” (2007, 354f.). Such complexes of institutions are called structures, which are the constituents of social formations, which are said to be the “comprehensive social entity at the macro-level” or “large systems” (2010, pp. 56–58).

What is a social structure, given its centrality as a causal agent and patient in “the agency-structure” thesis? In his early work, Little (1989, p. 24) believed a social structure to be “a set of constraints and incentives imposed on individual conduct and embodied in patterns of individual behavior.” Examples of social or historical structures in his recent work are human organizations as, for example, a “rail system” (2011, p. 286) and “the fiscal system of the ancient regime” (2010, p. 3). A “fascist movement” and a “market” (2010, p. 88) are also called structures as are “the revolution of 1848” (2010, p. 81) and “large complexes of rules and practices” (2010, p. 75). Although it might seem that structures in ML sometimes are complex concrete things (formal organizations or systems in Bungean systemism), they sometimes resemble events or processes or even sets of rules or patterns of rule-governed individual actions. The last paradigm comes near the theory advanced in the earlier writing:

a social structure is a system of geographically dispersed rules and practices that influence the actions and outcomes of large numbers of social actors. (Little 2010, p. 73)

Because anything else would probably amount to a harsh form of holistic idealism, it is claimed that each social entity is “constituted by the socially constructed individuals who make it up, through their beliefs, values, interests, actions, prohibitions, and powers” (2010, p. 56). As it seems, on the one hand, social entities are not necessarily concrete systems as in Bunge’s systemism, in which systems are straightforwardly composed of men and women of flesh and blood. They are not composed of or constituted by actors’ beliefs or values.

On the other hand, Little claims that agents “populate” “social factors,” “institutions,” and “higher-level structures”; what seems to make some sense if these are concrete objects. But those structures then are said to consist of “institutions, practices, and mentalities” (2007, p. 353), which can hardly be said to be populated or to be constituted by people. In any case, it remains unclear what this could mean. In order to clarify this and since formations are also said to be constituted by structures and these by institutions, it is worth to get a clearer picture of what institutions are according to ML:

An institution (…) is an embodied set of rules, incentives, and opportunities that have the potential of influencing agents’ choices and behavior. An institution is a complex of socially embodied powers, limitations, and opportunities within which individuals pursue their lives and goals. A property system, a legal system, and a professional baseball league all represent examples of institutions. Institutions have effects that are in varying degrees independent from the individual and “larger” than the individual. (Little 2007, p. 352)

Institutions, then, seem to be almost the same as what Little once called social structures, which was in part to be expected given that above structures were said to be “complexes” of such smaller institutions.

Although individuals in the end “embody” the whole social world, his position is intended not to be the same as ontological individualism, since (i) the individual is by itself social or socially constructed (2010, p. 58) and (ii) “social arrangements and circumstances affect individual action,” that is, through the social situation (2007, p. 360). Because of this allegedly causal influence, social stuff is believed to exist in the first place.⁵¹ Even more, large social stuff such as villages are supposed to be “lodged with a larger political, economic, and natural environment” that “influence and constrain” what is going on in the village. Thus, Little does not subscribe to what he calls “ultra-localism,” the theory that the social world is exhausted by face-to-face-interaction (2007, p. 349), which, we should add, would considerably limit our possible thoughts about real histories.⁵²

But how do such social facts or powers cause individual action and social facts? Little subscribes to the view he terms “causal realism”:

a thesis about the reality of causal mechanisms or causal powers. (2010, p. 101; 2011, p. 275)

Given the former section, the question arises how causal powers and mechanisms are related, given that in CR they are roughly the same and such passages in Little’s work suggest a similar reading. Little is not very explicit about his theory of causal powers, though it is remarkably different from Bhaskar’s due to an absence:

What is it to attribute a causal power to an entity? It is to assert that the entity has a dispositional capacity to bring about specific types of outcomes in a range of causal fields. To have a causal power is to have a capacity to produce a certain kind of outcome in the presence of appropriate antecedent conditions. (Little 1998, p. 205)

What are the “entities” endowed with such causal powers, given that these, in early CR, were supposed to be complex objects or concrete particulars: “Only things and materials and people have ‘powers’” (Bhaskar 1978, p. 78).

On Little’s account “events, conditions, structures”; “institutions, ideologies, technological revolutions, communications, and transportation systems”; “properties, conditions, and events” (1998, 198f.); and “various social forces” (2011, p. 275) as “social classes” (2010, p. 83) are supposed to be, as well as incentives and opportunities (1998, p. 206), and “rumors” (1991, p. 19) are possessors of causal powers. In short, every necessary condition is in accordance with early CR in the end a cause (1998, p. 200) and, by definition, a bearer of dispositional capacities.

This account and many examples given lack something which resembles a concrete social thing or system that bears the powers and produces something, so that there must be another trick than from some emergence basis in a complex thing to endow “social factors” with causal powers that somehow produce actions of individuals. How do such “things” as rules, opportunities, incentives, or mentalities come to possess powers to cause changes in concrete things according to ML?

The first theory embodied in ML might be named the theory of instantaneous power acquisition:

Institutions and other aspects of social organization acquire their causal powers through their effects on the actions and intentions of the individuals involved in them – and only from those effects. (Little 1991, p. 19)

A rather strange but perhaps suggestive analogy might help here. For example, when the sun, although perhaps no social thing, causes modern people to go “sunbathing” by actually shining, because people decide to do so, it acquires the dispositional capacity to cause people to go sunbathing in the right and a variety of circumstances and instantaneously exercises it. This seems to be incoherent if causal powers are supposed to be intrinsic, dispositional, and therefore possessed even if not actualized and possessed before the thing causes anything or is triggered by some event. If this is not to be presupposed, we face a somewhat different concept of a power.

The second theory about the way the ascription of social causal powers can be justified might be termed the explanatory account of social dispositions. On this reading, “the causal capacities of social entities are to be explained in terms of the structuring of preferences, world views, information, incentives, and opportunities for agents” (2009, p. 170; 2007, p. 361; 2010, p. 106). If we can somehow explain microfoundationally how the shining sun provides modern people with an opportunity for sunbathing, that is, if we can sketch the “pathway” of how it manages to do that through people’s heads, we are justified to claim that it simply possesses the causal capacity to bring about sunbathing people. But, at least originally, the “powers” of “things” were to be explained by their composition, structural organization, or their intrinsic natures, not by cultural accidents. Again, here we do not seem to face the concept of a causal power that grounded the realist tradition in philosophy of social science, but something else.

In a third reading we find what might be termed the relational or plausibility account of social causal powers or the theory of their rational calculation dependence. In this theory, that is similar to the former and perhaps explicates what “structuration” means, it is maintained that the causal powers of social stuff “derive from the incentives, powers and knowledge that these institutions provide for participants” (2010, p. 106), that is, they “derive from the structured circumstances of the individuals who make up those entities, and from nothing else” (2007, p. 358). The result of the plausibility account is that social things “possess causal powers in a derivative sense: they possess characteristics that affect individuals’ behavior in simple, widespread ways” (2007, p. 362; 2010, p. 106). According to this account, for example, the protestant ethic as a supposedly supervenient social entity, or perhaps an embodied mentality, causes a whole bunch of actions in or through individuals who happen to believe in protestant doctrines and are therefore incited by these. Furthermore, this has “historical changes” as unintended effects, if these people actually happen to act on those doctrines and thus instantiate the powers of Protestantism.

The problem with this account is that it is somehow plausible but quite fluffy, if we stick to the idea that also in the historico-social sciences or their ontology something explicit and perhaps exciting is associated with social causal powers or properties of some “thing.”⁵³ But such social powers reside somewhere between the circumstances of acting people; the incentives, constraints, or opportunities; and their relation to the agents themselves, that is, their perception or interpretation of “social things.” Virtually lost is what was thought to be in need of justification, given the realist tradition in philosophy of social science, namely, social things, their powers, and thereby productive social causation.

Put in different words, we might get the impression that the social powers metaphysics in philosophy of social science is grounded in theories about how “the social” constrains and enables actions. Though this might be perfectly alright in explanatory contexts, it seems to be questionable if this is enough to populate the world with social powers and causes.⁵⁴ Given that Little states that social causal capacities “are entirely defined by the current states of psychology, norm, and action of the individuals who currently exist” (Little 2007, p. 347; 2009, p. 166; 2010, p. 61), critics of realist social ontologies might simply want to expel the powerful ghost without a machine altogether or assign him a powerless status. Reification of “the social” might not be “the attribution of causal powers to entities without an understanding of the mechanisms through which those powers are expressed” (2007, p. 350), but the attribution of causal powers to entities that are not concrete systems or what was formerly called a powerful particular.⁵⁵

After all these preliminaries, what are mechanisms and what is their function in ML? Are mechanisms complex things, processes, properties, or do they embody the whole spectrum? The question suggests itself, given that in the quotation above it seemed as if powers were roughly mechanisms, whereas in the ending of the last paragraph, mechanisms are said to be something where powers are “expressed,” which suggests that mechanisms are events or processes, that is, they are supposed to be exactly what they are explicitly not supposed to be in CR.

The role for mechanisms in ML is first of all to “mediate” social causation (1998, p. 203), that is, to be the medium of the “expression” of macro-powers and to be the connection between macro-cause and macro-effect (2011, 278f.). Although it is slightly misleading, we can picture this the following way:

Macro → Mechanism → Macro.

Accordingly, the first theory of mechanisms in ML was straightforwardly that of a causal chain.⁵⁶ This line of thought seems to persist in Little’s recent writing: “A causal mechanism is a series of events or processes that lead from the explanans to the explanandum” (2007, p. 357). In the case of social causal mechanisms, it is “a set of social conditions, constraints, or circumstances combined to bring about a given outcome” (2009, p. 168).

Given the ontological microfoundations thesis and causal realism, “individual actors embody this causal process” (2007, p. 347; 2010, p. 61), that is, the process that is central to the agency-structure thesis.

Social structure → Individual actions → Social structure.

Personal and social powers are also said to be part of the metaphysical “substrate” of the mechanism that brings about social processes. They provide those processes with fuel that is burnt only and solely in actions. The story can be abridged insofar as in the social world “causal mechanisms are constituted by the purposive actions of agents within constraints” (2011, p. 273). But finally, we get Little’s recent account of causal mechanisms:

A causal mechanism is (i) a particular configuration of conditions and processes that (ii) always or normally leads from one set of conditions to an outcome (iii) through the properties and powers of the events and entities in the domain of concern. (Little 2010, p. 102; 2011, p. 277)

Accordingly, a mechanisms is in ML roughly what is called a process or an event in CR, the place where in CR individual and social mechanisms (powers) get actualized, which are here mainly to be found under clause (iii). At the same time they resemble slightly what in ES is called a concrete system, because what is there called a component of a system (and perhaps also what figures as the environment of a system and as its structure in ES) is in ML a part of a mechanism. But a clear notion of a social thing seems to be absent from ML as it is from CR, although examples for mechanism such as “the feudal manor, the collective farm, the Wall Street law firm” (2011, p. 284) are straightforwardly systems in Bunge’s rather clear sense, not ES mechanisms. Therefore, the little process schemas above are slightly misleading because in the end social structures and social institutions are as well part of a social causal mechanism in ML as are persons that constitute structures, institutions, and formations, which together bring about the “behavior” of the social mechanism (ii). Yet, similar to Bunge, ML social mechanisms happen neither to be running in persons nor in “structures” but are somehow the resultant of both, that is, of actions within “constraints.” The difference is that ML mechanisms might also be said to be constituted by or composed of persons and social structures, which is, of course, impossible in ES.

And, finally, what is history? Little clothes his philosophy of history into the metaphor of a pathway: “history is an accumulation of pathways and roadways that embed human action over time” (2010, p. 9). But isn’t that, roughly, a social structure or a social mechanism?

5 Do They or Do They Not?

At the workshop that anteceded the publication of this volume, the question arose how central mechanisms are to the historian or history. If we want to refrain from answering this question by the traditional stories of philosophy of historiography that tell us that historians are by their essence, that is, by definition, interested in individual actions, singular or unique “big events,” the aesthetics of narratives, or what have you, we would need to know what people that are sometimes called historians do, what mechanisms are, and what history is.

Since I do not claim to know what so-called typical historians do, we have to speculate about this aspect of our question. If we follow Bunge’s materialism, we even have to discard the whole question. Why? Simply because it does not make any sense to ask about the mechanisms of history because the latter is conceptual according to ES. No single overarching history exists, although it is perfectly plausible to investigate the history of any thing or system or to hypothesize about the mechanism for an aspect of a history of such a thing.⁵⁷ We also saw that the leading philosophers of social mechanisms disagree remarkably on what mechanisms are and on further basic ontological assumptions, so that it is not by any means clear what a social mechanism is supposed to be.

Social mechanisms turn out to be processes occurring essentially in a kind of social system (ES), social dispositional properties or powers (CR), or configurations of processes and events, conditions, powerful people, and powerful social entities that regularly bring about social processes (ML) (Problem I: What is a social mechanism?). We furthermore witnessed notorious differences about the place of causation and determination in ontologies of the so-called social world, which looks rather different through these philosophies. For instance, “social entities” (Problem II: What is the furniture of the “social world”?) turn out to be causal agents or factors for some, not for others (Problem III: What is causation?). Thereby we also should have seen that the merits of philosophical theories of social mechanisms, causation, and social explanation are in the end only assessable if they are analyzed in their respective philosophical or social theoretical environments. ⁵⁸ In any case and in any of these philosophies, central notions such as “constraint” and “enablement” or the thought that the “environment does not act on a person, but rather through a person” (Bunge 2009a [1959], p. 181), what Marx famously called Zwang der Verhältnisse, remain worth disputing, although these are classic questions of social theory and philosophy (Problem IV: What is determining or causing change or stasis in “the social world”?).

It is hardly worth mentioning that the “structure vs. agency” issue is the most central problem in social theory and social research, and it figures prominently in mechanism discourses. More basically, this amounts also to the question what does change in “the social world,” if anything (Problem V: What is a history?). If there are no “social entities” or “social structures,” there is no social change, if change presupposes something that changes. Of course, different configurations of responses to these problems lead also to different opinions about the possible explananda of social or historiographical (geschichtswissenschaftlich) research or about the stuff merely to be described and, finally, to different norms of how sociological or historiographical explanations have to be framed, for example, whether those explanations are causal explanations, and if the answer is yes, in which way. (Problems VI: What is to be understood and how is it accordingly to be explained? See also Blaikie 1993). In a nutshell, given the former sketch of a comparison of some of the main positions on social mechanisms, it seems to be so that the problem of social mechanisms comes with a whole bunch of others, and the sketch suggests that they should perhaps be more clearly separated or related in future debates. Otherwise, it might be that those scholars who believe we are dealing here with fruitless “mechanism talk” are right.

But let us answer the main question: If we presuppose the ontology of mechanisms in ES and therefore the whole framework, it is arguably beyond doubt that many historians study mechanisms that make past or present systems tick and made them what they were, became, or still are.

If we presuppose the ontology of mechanisms in CR, it is arguably beyond doubt that many historians study mechanisms (social structures) or invoke individual or social powers (mechanisms) in explanations of individual or social events. Though we equally found out that “[i]t is not by any means obvious what the concept of mechanism refers to in philosophical and critical realist frameworks (…)” (Kurki 2008, p. 177). This diagnosis is mirrored in Bhaskar’s statement (1978, p. 49) that a “generative mechanism” is “a ‘real something’ over and above and independent of patterns of events.” Moreover, ascribing powers to “emergent” social entities seemed to be problematic.⁵⁹ Though it was not the central point in our discussion, CR clearly implicates some philosophy of history.

If we presuppose the ontology of mechanisms in ML, it is arguably beyond doubt that many historians study mechanisms,⁶⁰ although the same problems that were diagnosed in connection with CR concern the ML framework. Perhaps because of the affiliations to realist traditions, the concept of mechanisms as powerful properties and mechanisms as processes is often not clearly drawn. In summary, there is no reason whatsoever to believe that historians do not study mechanisms. This is the meager but positive result of this sketch.

But these answers are obviously utterly unsatisfying because the following question is which of these ontological frameworks or philosophies of history, if any, is adequate and why, given that they disagree on almost any central category, although we did only visit the main realist positions in social philosophy.⁶¹ This is the negative result of this sketch. While their merits can only be further evaluated in comparison with accounts of mechanisms coming from other parts of social science and philosophy as well as in confrontation with historiological practice, to borrow Bunge’s term, the merits of this exercise might be found in the simple observation that we did something resembling an ontology of history without engaging in speculation about the “course of history” and whether it moves in circles or squares, that is, we did not set out “historiosophical schemes” (Sztompka 1991, p. 182). Thus, our discussion suggests, among other “things,” the need for an ontology of history, if the main question shall be answered on occasion. For short, there seem to be not many reasons to bury any philosophy of history from the start.

Finally, let us dwell a moment upon the philosophical problems that result from the claim, explicit in at least two of the sketched ontologies and furthermore as often supposed to be trivially true as it is straightforwardly denied, that (all) things have histories, that is, that they have properties that change. Those problems basically are which are those (social) things that have histories? What are their (social) properties that change? Why do they change or remain constant? Those questions do not seem to be far away from some of the questions some scientific historians set out to answer: What exists or existed? How does it change or how did it change? Why did it change or remain what it was or even still is?⁶² And if I am not totally wrong, those are the more basic questions that are at issue in the debate over social mechanisms, although they are often hidden.

Obviously, the formulation above presupposes that things have histories. But isn’t it a central implicit assumption of much philosophy of historiography that events have histories? Perhaps this might be one puzzle for philosophy of history, although it might also turn out to be ill posed.