Defining religion has led me to a number of sub-discussions. One of these is the discussion about social kinds, and this is how I see this issue presently:
By ‘social kinds’ I wish to refer to kinds that involve human purposes or activities. Accordingly, social kinds are not just social kinds: there are several sub-categories, and it must be considered whether different definitional strategies are needed regarding these sub-categories. As far as I can judge, there is no fixed terminology and no customary way of demarcating or distinguishing between the social kinds. For this reason, and because of the complexity of the matter, it is impossible to give a complete overview over the issue here. Instead, I will focus on three types of social kinds (‘institutional kinds’, ‘artifactual kinds’, and ‘cultural kinds’) and indicate what distinguish them ontologically and conceptually and why the distinctions matter to defining social kinds.
A considerable amount of the attention on social ontology has been devoted to matters relating to collective intentionality. This approach has been particularly influential in discussions about institutional kinds (such as money, presidents, contracts etc.) which plausibly rely on collective intentions. Money, for instance, ceases to exist if people do not recognize it as such, and the reason for this is that institutional and artifactual kinds are causally dependent on being created (and sustained) intentionally.
Defining artifactual kinds (such as knife, chair, umbrella etc.) have often been considered philosophically problematic (most notably in Wittgenstein). Part of the reason is that although artifacts are physical objects, they cannot be treated in the same way as natural kinds. In contrast, they admit of multiple realization in terms of both materials, shapes and composition (think of the difference between a bean bag chair and wooden chair). An obvious alternative way of defining them is the through their function, i.e. what they are made for or (can be) used for. The view can be substantiated by the fact that we tend to see purpose as the standard of evaluation (e.g. the knife is meant for cutting – the good knife cuts well).
Some, however, have objected to functional definitions that, for instance, a broken clock is still a clock, that a table does not turn into a chair even if it is used for sitting on, and that nothing distinguishes the function of a cup from that of a bowl. Accordingly, an alternative approach replaces function as the determining characteristic for artifactual kinds with the successful intention of the creator to make something that belongs to the kind in question.
The intentional and the functional explanations may at first sight seem to underline the differences between artifactual kinds and cultural kinds (such as religion, mother tongue, gender, and family). Because, in contrast to artifactual kinds, family and mother tongue are hardly definable in terms of a function or intention. For instance, mother tongue is not a result of an intentional creation, and it appears that even dysfunctional families count as families. Interestingly, however, psychological experiments seem to suggest that the acquisition and use of artificial kind concepts is less straight forward than suggested by the functional or intentional accounts. For instance, even if objects are presented as fulfilling the function of e.g. an umbrella or boat, they are not accepted as such if they differ significantly in appearance. Likewise, Bloom claims that people – at least non-philosophers – tend to agree that an object which looks like and can be used for a chair is in fact a chair even though it was the result of a lightning striking a rock and thereby creating the object. Consequently, the conceptual structure of artifactual kind terms may be more complicated than is normally assumed. This indicates that conceptualizing artifactual kinds may have more in common with conceptualizing other types of social kinds such as cultural kinds (e.g. religion or family) than would otherwise have been expected.
Defining the nature of cultural kinds (and the meaning of the concepts that represent them) is particularly complicated because they evolve through time and vary between cultures, and I think this is the reason that concepts of cultural kinds are often vague concepts with loose boundaries. Even so, it appears definitions of phenomena within one culture are usually precise enough to enable comparison between cultures and decisions about whether a cultural kind concept (e.g. family, conflict, language, music, etc.) applies to phenomena in another culture. In a way, cultural kinds are manmade because they are causally dependent on the existence of humans and human purposes, but, as stated above, in general not with a specific function in mind. This means that the analysis from collective intentionality does not directly apply to them, and this makes it more difficult to provide a satisfactory definition.
In addition to this, the problem of realism concerning social ontology needs to be considered. As within other domains, views on social kinds range from eliminativism to robust realism, and a core issue here is what realism concerning social ontology entails. When it comes to artifactual and institutional kinds, it is not unlikely that their existence is dependent on conceptualization – at least, many have argued that, for instance, money (and maybe also chairs) cannot exist without being conceptualized. In contrast to the case of artifactual and instutional kinds, I find it plausible that the existence of cultural kinds such as religion, gender, or art is not dependent on conceptualization, even though they are dependent on the existence of humans and human purposes and intentions. An important consequence of this is that realism concerning cultural kinds is less demanding than realism concerning e.g. institutional kinds. Often mind-independence is seen as a condition for realism. As this, however, automatically excludes realism concerning e.g. mental states, it has been suggested that instead of seeing realism in terms of mind-independence, realism should be seen as the position that something is real “… if the fact that it exists is not in itself a matter of opinion or conjecture” (cf. Oderberg 2007). On this view, realism concerning religion would be at least prima facie plausible in spite of its mind-dependence, while realism concerning money could prove harder to uphold because of its likely dependence on conceptualization.
So even if realism concerning artifactual and institutional kinds proves hard to defend, the same does not necessarily apply to cultural kinds. Amie Thomasson has argued that artifactual and institutional kinds fail to fulfill the dual demand of standard realism, namely independence of the mental and having natural boundaries (which is a condition for the possibilities of ignorance and error usually associated with realism). Nevertheless, she also shows that Searle in his analysis of social ontology overlooked that there may be “…entities (and kinds of entities) depending on mental states of various kinds, without their depending on any beliefs about them” (Thomasson 2003). As examples she mentions superstition and racism, and urges that even though these are not mind-independent, they can be unknown within a society and “… require substantive social and scientific discovery for their existence as well as nature” (Thomasson 2003). Other examples include jokes or insults: it is possible for someone to tell a joke or insult someone without having the concepts. In my opinion, religion counts as yet an example.
