The Individualism-Holism Debate in Outline PDF

Preview

Summary

The Individualism-Holism Debate in Outline Julie Zahle & Finn Collin Section of Philosophy, Department of Media, Cognition and Communication, University of Copenhagen [email protected] (Final Draft - forthcoming in Julie Zahle & Finn Collin (eds.) Rethinking the Individualism-Holism Debate. E...

Description

The Individualism-Holism Debate in Outline Julie Zahle & Finn Collin Section of Philosophy, Department of Media, Cognition and Communication, University of Copenhagen

[email protected] (Final Draft - forthcoming in Julie Zahle & Finn Collin (eds.) Rethinking the Individualism-Holism Debate. Essays in the Philosophy of Social Science, Synthese Library, Springer.

1 Introduction The individualism-holism debate is an old – but still vibrant – dispute within the philosophy of the social sciences and the social sciences themselves. Over the course of its history, there are three phases in which the discussion has been particularly lively (Udehn 2002:479). The first is around the turn of the nineteenth century with significant contributions by, among others, Emile Durkheim and Max Weber. The second phase is around the 1950s where the ardent defenses of methodological individualism by Friedrich Hayek, Karl Popper, and J.W.N. Watkins spurred on the debate. Finally, the third and last phase stretches from the 1980s and up till today with pioneering contributions by a number of theorists including Roy Bhaskar, Raymond Boudon, James S. Coleman, Jon Elster, Alan Garfinkel, Daniel Little, Harold Kincaid, and Philip Pettit.1 The individualism-holism debate has first and foremost revolved around two issues: -

What is the ontological status of social phenomena and, as part of this, their relationship to individuals? To what extent may, and should, social scientific explanations focus on individuals and social phenomena respectively?

The second question, in particular, has received a lot of attention. The methodological individualism-holism debate refers to the discussion of this issue, whereas the ontological individualism-holism issue denotes disputes relating to the first question. As part of these discussions, a number of other topics have been addressed too relating to meaning, confirmation, research heuristics, ethics, and the like. Accordingly, it is also possible to distinguish between individualism-holism debates on meaning, confirmation, research heuristics, and so on. The present collection of papers is in line with the tradition in terms of its focus. The contributions divide into two categories: One group focuses primarily on the ontological dispute, the other on the methodological dispute. The papers reflect the newest trends within these debates in the following ways: First, many papers point to basic concepts and distinctions that are widely applied within the debate. Rather than being uncritically taken for granted, it is suggested, these concepts and distinctions are in need of being carefully spelled out, questioned, or even replaced by more adequate conceptions. This approach is, among other things, adopted in relation to the notion of supervenience, the notion of agency and its domain of application, the distinction between individuals and social phenomena, and the distinction between explanations that focus on individuals and social phenomena respectively. Second, many of the papers strike out in a new direction by paying close attention to actual developments within social scientific research.

1

It should be stressed that Udehn simply lists the three periods. Moreover, his claim that the third phase of the debate is still ongoing is from 2002. We are responsible for the observation that the third phase extends to the present just as the addition of key figures within all three periods is our doing.

1

They explicitly state – and exemplify – a focus on issues raised by, or addressed within, the context of social scientific research, just as they make suggestions of relevance to social scientific practice. In these ways, then, the papers exemplify a rethinking of the debate that points to novel directions in which to take future philosophical discussions and future empirical work in the social sciences. In the following, we provide a rough systematic overview of the individualism-holism debate. As we go along, we also comment on the history of the debate. In section 2, we outline the ontological dispute and, in section 3, the methodological debate. Further, in section 4, we discuss the individualism-holism debate in relation to two kindred disputes.2 Against that background, we briefly present the contributions to this book.

2 The Ontological Individualism-Holism Debate The ontological individualism-holism debate concerns the ontological status of social phenomena (or facts) and, as part of this, their relationship to individuals (or facts about individuals). Ontological holists contend that social phenomena exist sui generis or over and above individuals, whereas ontological individualists deny this. Before looking into various interpretations of the idea that social phenomena exist sui generis, it is instructive briefly to characterize the notions of social phenomena and individuals. Within the individualism-holism debate more generally, it is common to distinguish between various kinds of social phenomena. Some of the most frequently mentioned ones are: a) social organizations, as exemplified by a nation, a firm, and a university; b) statistical properties like the literacy or suicide rate of a group of individuals; c) norms and rules as illustrated by the rule to drive to the right and the prohibition against sex with close family members; d) cultures such as the Mayan culture; and e) social structures as typically identified with one or several of the sorts of social phenomena already listed. The ontological debate has mainly concentrated on the issue of whether social organizations exist over and above individuals. In this connection, social organizations are simply referred to by way of terms such as “university,” “hospital,” and the like. It is less obvious how individuals should be characterized from the perspective of offering an ontological analysis of their relationship to social phenomena. Not surprisingly, therefore, opinions diverge on this question. Most notably, it is a matter of dispute exactly what kinds of interactions and interrelations between individuals it is permissible to invoke as part of ontological analyses of this sort. Also, and related to this, it is discussed whether it is admissible to mention individuals’ beliefs about social organizations like nations and firms. Or, to mention one last example, there are different views as to whether a specification of individuals should be taken broadly to include the objects they make use of and, more generally, the physical environment in which they find themselves. Throughout the history of the ontological debate, a variety of suggestions have been made as to what it takes for social phenomena to exist sui generis. Here are some of the many, and sometimes overlapping, ways in which this idea has been fleshed out: -

-

-

2

The causal overriding criterion: Social phenomena such as social organizations exist over and above individuals insofar as they have causal powers that are independent of, and override, the causal powers of individuals. The translation criterion: Social phenomena such as social organizations exist over and above individuals insofar as terms, like “nation” and “school,” that refer to these phenomena cannot be translated into statements about individuals. The composition criterion: Social phenomena such as social organizations exist over and above individuals insofar as they are not merely composed of ensembles of individuals.

Sections 2- 4 draw and expand on Zahle (2007) and Zahle (2013).

2

-

-

The determination criterion: Social phenomena such as social organizations exist over and above individuals insofar as individuals do not non-causally determine what kinds of organizations, properties, and the like, are being instantiated. The agency criterion: Social phenomena such as social organizations exist over and above individuals insofar as they qualify as group agents that have attitudes supervenient upon the attitudes of individuals.

In his writings from around the turn of the nineteenth century, Durkheim is commonly taken to appeal to the causal overriding criterion when maintaining that social phenomena exist sui generis (see Durkheim 1897). The opposite view is particularly associated with Popper’s and Watkins’ work from around the 1950s. Both Popper and Watkins saw the claim that social phenomena have causal powers that override the powers of individuals as going together with the historicist view that societies go through stages of development according to deterministic laws of historical development, that is, laws which individuals cannot change in any way (see Popper 1964[1957] and Watkins1973[1957]). Popper argued that there cannot be such laws and as such he rejected the idea of the sui generis existence of social phenomena relative to the causal overriding criterion. Turning to the translation criterion, it was appealed to in a much discussed paper by Mandelbaum from 1955 (Mandelbaum 1973[1955]). Mandelbaum maintained that it is impossible to translate terms which refer to social phenomena into terms which refer to the thoughts and actions of individuals only. This shows, he contended, that social phenomena are not identical with specific individuals having certain thoughts and performing given actions. Hence, he concluded that social phenomena should be regarded as existing sui generis. Nowadays there are few, if any, theorists who defend the theses of ontological holism or individualism by appeal to the causal overriding and the translation criteria. Instead, the majority of theorists tend to be concerned with the composition and determination criteria. More specifically, most participants in the debate regard themselves as ontological individualists by reference to one or both of these criteria: They subscribe to the view that social phenomena are composed of and/or non-causally determined by individuals. But what exactly does it mean to say that social phenomena are non-causally determined by individuals? This question has received a lot of attention in the recent debate. One influential suggestion is to spell out this relation as one of supervenience. Applied to social phenomena, this is the contention that there can be no difference at the level of social phenomena without there being a difference at the level of individuals. Roughly speaking, this means that if individuals with certain properties, standing in certain relations, etc. instantiate a university, then they cannot cease to do so unless some change takes place as to their properties, interrelations, and so on. This way of spelling out the notion of supervenience is only one among many. Moreover, various other notions, such as the notions of emergence, realization, and grounding, have been brought into play in the attempt to cash out the notion of non-causal determination. Their exact specification, too, is a matter of discussion. The wide acceptance of the view that social phenomena are non-causally determined by individuals means that the most recent criteria of sui generis existence are made in full compatibility with this point. To illustrate this observation, consider the agency criterion as currently defended by Christian List and Pettit (see, e.g. List and Pettit 2011). The thought behind the criterion is that in order for social organizations and the like to qualify as group agents, it must be possible to ascribe to them certain attitudes supervenient upon the attitudes of individuals. It is very difficult, though, to pin down the exact individual attitudes that these attitudes of group agents supervene upon. For this reason, List and Pettit contend, social phenomena that are group agents should be regarded as existing over and above individuals. In discussions of the ontological status of social phenomena, the term “explanation” is sometimes used to refer to ontological analyses of how individuals must be related, what properties they

3

must have, and so on, in order to constitute social phenomena of various sorts. Explanations of this sort map out non-causal or synchronic relationships between social phenomena and individuals. They should be distinguished from explanations which map out the causal or diachronic relations between events or states involving social phenomena and individuals. It is explanations in the causal or diachronic sense that is the focus of the methodological individualism-holism debate.

3 The Methodological Individualism-Holism Debate Why is the European economy in recession? Why has the birth rate in Tanzania recently gone up? And why do revolutions tend to be followed by famines? The methodological individualism-holism debate is about the proper focus of social scientific explanations that are advanced in response to questions like these. More specifically, it concerns the extent to which social scientific explanations may, and should, focus on individuals and social phenomena respectively. The discussion takes various forms. Among these, there are in particular two which stand out: The dispensability debate, as it may be called, and the microfoundations debate. 3.1 The Dispensability Debate The debate about dispensability revolves around the question of whether individualist (or individual level) or holist (or social level) explanations may, and should, be dispensed with within the social sciences. There are three basic positions on this issue: Methodological individualism: Individualist explanations should be advanced. Holist explanations may, and should, be dispensed with. Strong methodological holism: Holists explanations should be offered. Individualist explanations may, and should, be dispensed with. Weak methodological holism: Not only individualist but also holist explanations should be put forward. Neither individualist nor holist explanations may, and should be, dispensed with. In order further to characterize the methodological debate, it is instructive to go over four dimensions along which particular methodological individualist and holist positions may vary. To begin with, particular positions may differ with respect to their notion of explanation. All theses of methodological individualism and holism express a view as to what is the proper focus of explanation. By implication, they involve some view of what an explanation is. In line with the general trend in philosophy of science, there was a point at which most theorists endorsed the Covering Law model of explanation. According to this model, a scientific explanation takes the form of a deductive or inductive argument which shows why an event in need of explanation was to be expected. Today, this view is rarely, if ever, espoused. Instead, a number of different notions of explanation are appealed to in the debate. For instance, some participants in the debate embrace the erotetic model of explanation which states that an explanation is an answer to a why-question. Others subscribe to the causal-information view of explanation which has it that an explanation provides information about the causal process leading to the event in need of explanation. Whatever their notion of explanation, participants in the methodological debate make a distinction between individualist and holist explanations. It is widely concurred that individualist explanations revolve around individuals, their beliefs, actions, etc. whereas holist explanations are centered round social phenomena such as social organizations, structures, and the like. Beyond this point, however, agreement ceases: There is no general consensus as to how, more precisely, to circumscribe individualist and holist explanations. Here are some of the issues that have been discussed: What kinds of relations and 4

interactions among individuals should individualist explanations be allowed to refer to? Does an explanation really qualify as holist if it describes the rules and norms within a group of individuals? Are individualist explanations only permitted to refer to particular individuals or may they describe types of individuals too? The debate on the proper distinction between individualist and holist explanations has primarily concentrated on the category of individualist explanations. One reason for this is probably that as the debate evolved, methodological individualists have opted for ever broader conceptions of individualist explanations. As such, they have continuously brought up the question of the proper circumscription of these explanations. Both individualist and holist explanations may be divided into different types. And not surprisingly, there are various conceptions as to what constitutes an adequate or satisfactory type of individualist or holist explanation. For example, one type of individualist explanation that has been highly popular among methodological individualists is explanations informed by rational choice models. These state how rational individuals acted in light of their beliefs and desires. Another type of individualist explanation consists in accounts of individuals’ actions by appeal to their dispositions to act in certain ways in certain circumstances. Turning to holist explanations, the advancement of accounts that point to the statistical properties of groups has been very prevalent among methodological holists as have explanations specifying how social organizations bring about various effects. Needless to say, there are multiple more or less finegrained ways in which to differentiate between types of individualist and holist explanations. It should be stressed that the rejection of the adequacy of, say, explanations informed by rational choice theory or explanations by appeal to statistical properties does not amount to a refutation of methodological individualism or holism as such. All it amounts to is the repudiation of the adequacy of a particular type of individualist or holist explanation. Conceptions of adequate individualist and/or holist explanations typically go together with a preference for the use of one or several forms of explanations. Three commonly mentioned forms of explanations are functional explanations, intentional explanations, and straightforward causal explanations, as they may be called. In the past, methodological holists have often favored the use of functional explanations. Accordingly, they have explained the continued existence of various social phenomena by reference to their function, or effect, in some society. By contrast, methodological individualists frequently make use of intentional explanations: They explain actions by appeal to individuals’ reasons or motivations for their actions. Today, functional and intentional explanations are often regarded as a species of causal explanations while being contrasted with straightforward causal explanations. When applied by methodological holists, straightforward causal explanations state, say, that a rise in unemployment caused a rise in crime. Again, different and more fine-grained classifications of forms of explanation have also been suggested. As these considerations bring out, there is a variety of ways in which particular individualist and holist positions may differ. It is possible to distinguish between particular individualist and holist positions as regards their notion of explanation, their distinction between individualist and holist explanations, their conception of what constitutes adequate types of individualist and holist explanation, and their preference when it comes to different forms of explanation. Turning now to a description of the dispute itself, there are few proponents of strong methodological holism. As a result, the debate mainly plays itself out between methodological individualists and weak methodological holists. As both parties agree that individualist explanations should be advanced, methodological individualists put all their efforts into showing that holist explanations may, and should, be dispensed with, whereas weak methodological holists concentrate on showing that holist explanations may not, and should not, be dispensed with. A number of arguments have been offered in support of, and against, these stances.

5<...