In general we can say that social and collective properties and notions are man-made and that when people act on these notions a part of the social world becomes man-made. There are two important features of sociality and the collective creation of some central aspects of the social world (emphasized by such authors as Barnes (1983), Bloor (1996), Kusch (1996), and Searle (1995)). The first feature is that of the performative character of many social things (entities, properties). The second is the reflexive nature of many social concepts. My account adds to this list a third feature, the collective availability or "for-groupness" of collective social items.

I will argue in this paper that much of sociality is created through collective acceptance. I have elsewhere developed a "Collective Acceptance" account of sociality (see Tuomela and Balzer, 1997). The first section of the present paper will sketch this account. The second section discusses the philosophical content of the account in more detail. The third section applies the view to the problem of ontology of the social world.

According to the Collective Acceptance account of (collective) sociality, emphasized by many earlier authors and developed in detail in Tuomela and Balzer (1997), certain entities get their collective social status by being collectively created. For example, almost any kind of physical entity - for instance, squirrel fur - can "in principle" become money through the members of the collective in question accepting it as money. As soon as they cease to collectively accept it as money and to mutually believe that it is money, squirrel fur loses its status and function as money.

We must distinguish between a) the collective creation of an idea, b) collectively holding and maintaining it, and finally c) collectively realizing it or carrying it out. Collective acceptance relates to a) and b) in the first place. I have argued that collective acceptance entails holding a "we-attitude" of a suitable kind (cf. Tuomela, 1995, Chapter 1, Tuomela and Bonnevier-Tuomela, 1997 for the notion of a we-attitude). The we-attitudes that are needed for collective acceptance basically belong either to the intention-family or to the belief-family of attitudes. My account concentrates on intentional actions, since all collective acceptance can be or could have been intentional. (Even if in actual life collective acceptance may be non-intentional, it could have been intentional; and this suffices for the present account resembling political contract theories in this respect.) Since intentional actions arguably are based on intentions and beliefs, it suffices to deal with these notions. Intentions entail motivating wants or pro-attitudes in a broad sense, and in this respect the present account also presupposes underlying wants.

Consider collectively intentional collective acceptance. For instance, in a group there might be a (weak) we-goal to oppose the recent tax increase; viz. this is the group members' goal and they believe that the others share this goal and believe that this is a mutual belief among them. Another example involving collective acceptance as we-belief would be one where the group members believe that the earth is flat, believe that others believe so and also believe that this is mutually believed in the group. Collective acceptance in this kind of situation can be generally construed as acceptance either in the sense of conative commitment to a sentence or proposition s (intention to make s true or to uphold s as in the tax increase example) or doxastic commitment to s (the "acceptance" belief that s is true). Collective acceptance here is a) (weak) "we-acceptance", viz. each person comes to accept s, believes that the others accept s, and also believes that there is a mutual belief about the participants' acceptance of s. This we-acceptance can be either private acceptance (acceptance in the "I-mode") or acceptance in the "we-mode". In the latter case also a collective commitment (we-mode commitment) to s is entailed, and then we get the minimal sense of acceptance for the group (and the involved "we-mode" togetherness) which (intersubjectively) involves the group. Without the "we-mode" mutual belief there is not enough intersubjectivity and collective commitment for the application of the phrase 'for the group' and for saying that the participants are attempting to see to it collectively that the accepted content will become or - as the case may be - remain satisfied. Acceptance "for the group" can be viewed as entailed by acceptance in the we-mode (e.g. "We accept that s is true or correctly assertable for us in our group-related activities").

Stronger forms of collective acceptance "for the group" are norm-based, institutional acceptance and plan-based or agreement-based collective acceptance. An example of the first of these is the collective acceptance that driving when drunk is wrong and punishable, that anniversaries in a marrige ought to be celebrated, and perhaps also in some collective that squirrel fur counts as money. The last example is based on the social norm (the genesis of which I will not here try to give an account of) that everyone in the collective ought to treat squirrel fur as money. An example of plan-based or agreement-based collective acceptance is the group members' joint decision to elect a certain person as their leader. In general, acceptance for a group entails mutual belief in the acceptance, at least in "egalitarian" groups and in groups in which the normative structure in the group does not affect collective acceptance.

The following general thesis of sociality in a collective sense or of "collective-sociality" can now be proposed (Tuomela and Balzer (1997):

Collective Acceptance Thesis (CAT): The sentence s is collective-social in a primary sense in a group G if and only if the following is true for group G: a) the members of group G collectively accept s, and b) they collectively accept s if and only if s is correctly assertable.

In the analysans a) is the assumption of the categorical collective acceptance of s while clause b) is a partial characterization of the kind of collective acceptance that is needed here.

In symbols,

(CAT*) s is collective-social in a primary sense in G if and only

if

Forgroup(CA(G,s) & (CA(G,s) <-> s)).
 

Here the 'operator' CA represents the collective acceptance of s by G for G. Forgroup(G,s) means that s is correctly assertable (or, as a special case, true) for the group G in question (see Tuomela and Balzer, 1997, for discussion). (Forgroupness in general entails mutual belief concerning acceptance.) CA must be a performative achievement-expressing notion and 'acceptance' is general enough to cover both the creation and upholding of s and has achievement conceptually built into it. That a sentence is correctly assertable for G means, roughly, that the group members are collectively committed (in the "we-mode") to the sentence and hence treating it as correctly assertable (or true) in their various intellectual and practical activities in group contexts and when acting as group members.

Consider now briefly the notion of collective acceptance as characterized by (CAT*): Forgroup(CA(G,s) <-> s). First consider the implication from left to right:

Forgroup(CA(G,s) -> s) (Performativity)

This is true simply on the basis of the notion of collective acceptance, which is an achievement notion relative to the group's "intentional horizon." What is accepted by the group is correctly assertable or true for the group members.

Next consider the converse implication:

Forgroup(s -> CA(G,s)) (Reflexivity)

This gives a central and often emphasized "mark of the social": The truth of s for G makes reference to s itself within the sentence CA(G,s). This condition will be discussed in Section II.

We can say that a sentence is collective-social in a derived sense if it is not collective-social in the above primary sense but presupposes for its truth (for the group) that there are some relevant true (for the group) sentences which are collective-social in the primary sense. For instance, sentences using 'power' or 'wealth' are at least in some cases candidates for collective-social sentences in a derived sense. Latent or unilateral social influence are social features of the social world that would not - and correctly so - be cases of even derivatively collective-social features even when many agents are concerned. The same holds for "naturally" social emotions such as envy (unless they are specifically about socially constructed things). Many shared we-attitudes also are not socially constructed (for instance, shared we-fear can be regarded as a "natural" social phenomenon).

(CAT) leads also to an account of social institutions in the broad sense. Due to limitations of space, let me just briefly consider the matter and present the following suggestion (see Tuomela and Balzer, 1997):

(SI) A norm-entailing sentence s expresses a social institution in a primary sense in a collective G if and only if the members of G collectively accept s for G, with the understanding that collective acceptance for the group entails and is entailed by the correct assertability (or truth) of s.

Social institutions in a derivative sense can be characterized analogously. In them the sentence s is collective-social in the derivative (rather than in the primary) sense. Norm-based social power-relations could be cited as an example of social institutions in the derived sense.

My somewhat tentative thesis is that the family of intention concepts (including agreements and commitments) and acceptance beliefs (doxastic takings) are the basic attitudes needed to sustain (CAT), but a detailed defense is not possible here.

Searle's (1995) theory of the construction of social reality shares many features with my above account. His basic formula for collective acceptance in the context of social institutions is "We accept that S has power (S does A)" (Searle, 1995, p. 104, 111). This is understood to be implicitly entailed by my central acceptance sentence "We collectively accept s" (or CA(G,s)), but what is explicitly accepted in my account is the sentence s, e.g. s = squirrel fur is money, and not the underlying powers, rights and duties, concerning the possessors of squirrel furs and other members of the group.
 
 

What is the precise class of sentences s to which (CAT) is claimed to apply? Underlying my Collective Acceptance model is the general assumption that in each context of application one can distinguish between sentences whose obtaining or "truth" is entirely up to the members of the group (or up to their conceptual activities, especially to what they on metaphysical grounds can accept as true) and sentences whose truth is at least in part up to nature (as opposed to them), to the way the world is, and thus in part dependent on the causal processes occurring in the world "outside them." It is built into the nature of the equivalence in (CAT) that it is concerned with things which are up to the group members and are as "viewed" by them. Therefore, group members can collectively accept (for the group) the correct assertability or truth of some sentences, e.g. "Stars determine our fate," without making it to be the case that those sentences are true in the standard sense.

Generally speaking, social concepts and sentences are reflexive in the following sense (cf. Searle, 1995, Chapter 2. A collective-social sentence using a putatively social predicate (e.g. 'money,' 'leader,' or 'marriage') does not apply to real things (such as certain pieces of paper or squirrel furs in the case of 'money') unless collectively accepted and, so to speak, validated for that task. Let us consider money as an example. The predicate 'money' does not refer to itself but to coins, dollar notes, squirrel furs, and so on. Reference here means that 'money' correctly applies to those things. The loose talk about reflexivity in this context therefore should be understood as being about presupposition-stating sentences such as "Money is not money unless collectively accepted to be money." This is not a matter of what phrase to use but what the concept of money is. This concept is expressed by what the user of the predicate 'money' in English is entitled to say and, especially, extralinguistically do (and what he may be obligated to do). The concept of money thus connects with some deontic powers and obligations collectively bestowed upon those who use the predicate 'money' and who belong to the collective in question. The discussed presupposition (viz. that money is not money unless collectively accepted to be money) is central precisely because of the following assumed fact: It is up to the members of the collective - and nobody else - to bestow those extralinguistic deontic powers upon its members. This contrasts with sentences involving only physical predicates like 'tree' or 'heavy.' In their case it is not up to the members of the collective to do more than stipulate how to use certain linguistic phrases and, e.g. what word to use for trees.

It can thus be said that the alleged reflexivity of collective and social concepts strictly speaking is not directly concerned with the entities that the concepts (predicates) apply to. Rather, we may say that a collective-social concept is conceptually reflexive or "self-conceptual" in the sense that it presupposes itself; and this can be explicated in terms of truth (for the group) as follows. When for a social predicate q a sentence q(a) is true (for the group) this presupposes the collective acceptance of a as q in the group. Thus, if q(x) expresses that x is an item of money and a stands for a squirrel fur, then the statement q(a) can be true (for the group) only if squirrel furs are, in fact, collectively accepted as money or "made" money in the group. A similar point can be made about, meanings of words, leaders, marriages, property rights, and so on.
 
 

It follows from the Collective Acceptance account that the parts of the social world it applies to are collective-socially constructed and therefore man-made. This, however, needs qualifications. Let us consider the matter.

It can be, and has been argued, that reality is criterially connected to causality in the sense that an entity cannot be real unless capable of occurring in singular causal "inquirer-independent" contexts (viz. in claims of the form C(f,f'), C standing for causation and f, f' being inquirer-independent facts related to the entity in question). Here inquirer-independence is independence of an inquirer's mind (or the "ideally rational" scientific community's "mind", viz. attitudes, views, etc., or best-explaining theory, or what have you). Roughly speaking, the inquirer-independence of causation here can be understood in the sense of causation in a world similar to ours but in which there are no inquirers with minds. (Cf. Tuomela, 1985, Chapters 4-7, for a discussion of this from the point of view of scientific realism.) Note that our present criterion for an inquirer-independent world "out there" allows that the world contain creatures possessing minds (e.g. intentions and beliefs). In addition to inquirer-independent reality "out there" we have a group's (any group's, large or small) point of view. From a group's point of view the social institutions and other collectively constructed and upheld things in that group are collectively "mind-dependent" in the sense of being constitutively dependent on the group's acceptance and thus its attitudes. These things - dependent on the group's "mind" - can nevertheless be said to be "socially real" in the group (viz. intersubjectively real and belonging to the group's posited "public space"), and they are also real in the sense of being independent of an external inquirer's point of view. (There are of course also things which are dependent only on a particular person's mind - e.g. a person remembering an appointment.)

From a group's point of view there can thus be things which depend for their existence (creation, recreation, and maintenance) on intentional group activities, depending thus on the underlying intentions and beliefs of the group members (cf. the "duality models" of Giddens, 1984, and Bhaskar, 1989). Note that the group members generally need to have right thoughts about e.g. money and the school, etc., when they act, but they need not of course think that by so acting they contribute to the maintenance and renewal of the involved institutions.

As said, the Collective Acceptance account presupposes the following dichotomy: There is a viable distinction to be made between sentences whose truth (or, more generally, correct assertability) is entirely up to "us" (viz. up to the group members or indeed any human beings or beings capable of operating as the view requires) collectively considered (or, rather, up to our conceptual activities) and sentences whose truth is at least in part up to the way the inquirer-independent world causally is. This assumption presupposes that sense can be made of the causal processes occurring in the world "out there." Thus, according to this view, group members can collectively accept (for the group) the truth of some sentences, e.g. "Stars cause our fate," without making it the case that those sentences are true in the standard sense. Using other well-known philosophical terminology, we can try to say that here (CAT) applies only to the entities and features belonging to the conceptual order rather than to the causal order. While on the right track, this is not quite right, for (CAT) has ontic import. It connects with the inquirer-independent causal order by giving the participating group members rights (permissions) and duties (obligations). Accordingly, the fact that the group members, qua group members, have rights and duties entails their being (conditionally) disposed to act in certain specific ways. This is an entailment of ontic content.

While making a distinction between what is up to us collectively to achieve and what is up to nature is basically right, I think, considering the matter epistemologically, that its precise content is not an a priori knowable matter. Here is a point which supports the a posteriori status of the distinction. Assume that a group g collectively accepts, among other things, causal statements (both singular and general, e.g. that a bridge collapsed because of a heavy truck crossing it or that smoking causes cancer). Here we have embedded causality in a broader frame of collective acceptance by putting a non-social claim, viz. a claim about a non-social causal connection, in the same "acceptance-box" as, say, the statement about stars determining our fate - a merely social claim with an incorrect causal content. In more general terms, the embedding here goes as follows. We start with a comprehensive system of causal relations including relations of a purely social nature as well as other relations. In this system we can delineate the social causal relations as those which satisfy the Collective Acceptance model. Now a good epistemic procedure should be able to distinguish between those collectively accepted causal connections which are not inquirer-independent in the meant sense from those which are. I submit that the scientific method (perhaps already our current view of it) is able to do the job. Thus, for instance, by means of theorizing and testing the constructed theories we are led to a warranted rejection of the claim that stars determine our fate, or that "similia similibus curantur" (the central claim of homeopathy), or that smoking is not a causal factor of lung cancer (the central claim up to now of tobacco industry). This a posteriori method does (or tries to do) "from within" what the older metaphysical views do from an external point of view and in an a priori sense.

Is the social world real (viz. is full-blown realism about the social worl warranted)? As seen, the matter crucially depends on what is meant by the social world and by something being real. Basically, at least realist can regard the social world as real in the inquirer-independence sense, viz. the social world exists independently of an inquirer's mind or of the inquirers' community's viewpoint. However, a part of the social world is dependent on collective acceptance, as emphasized, and is in this sense "socially real". As a consequence, in order to be intelligible (in the sense of being correctly explainable) at least this part of the social world must be conceptualized as its inhabitants conceptualize it (squirrel fur may be money for medieval Finns but not for others).

According to our collective acceptance account, social institutions, qua some kind of collections of position-involving normative structures, are causally effective ultimately only via the group members' minds and actions. This is a kind of ontic individualism or "interrelationism" (see Tuomela, 1995, Chapter 9). I cannot here argue for this view, which, let me emphasize, does allow for social entities which are supervenient on individuals' minds and actions (etc.). However, in general social concepts, especially institutional ones, can be regarded as irreducible primitives. Let me note that social institutions can have and often do have causal impact via the participants' (in collective acceptance) thoughts and thus subjectively (in the group members' beliefs) qua social institutions, or at least their central nervous systems (in non-intentional cases). The "internalized" rights and duties related to institutional entities like money or institutional positions (e.g. teacher) can, accordingly, in this embedding involve causal connections independent also of the group members' minds.

In "non-normative" cases (cf. leader, esteem, status) based on collective acceptance in the sense of mutual acceptance belief (viz. the acceptance of something as true for the group) the analogous holds, for collective acceptance always is group-relative, viz. it relates the constructed and recreated things to the group (thus to the mental life of the group members; cf. Tom is our leader only in so far as accepted by us as our leader).

Physical social artefacts such as church buildings, cars, chairs, and books exist as causally effective entitities. They can enter causal connections not only qua having suitable physical features but also, and in the present context in an important sense, qua being artefacts expressing normative or non-normative collective practices (see Tuomela, 1998, for qua-causation).

Finally, there are social properties and relations which can be regarded as real in a more naturalistic sense and which correctly fall outside the scope of the Collectie Acceptance model. For instance, Tom's being jealous of Jane is an example of such a non-constructed social fact. Another example is provided by some shared collective attitudes (or we-attitudes) in the "I-mode" (but not those in the "we-mode") in the sense of Tuomela (1995). A concrete example would be our we-fear that a lion will attack us.
 
 

References
 

Barnes, B., 1983, 'Social Life as Bootstrapped Induction', Sociology 17, 524-45.

Bhaskar, R., 1989, Reclaiming Reality: A Critical Introduction to Contemporary Philosophy, Verso, London

Bloor, D., 1996, 'Idealism and the Social Character of Meaning', manuscript

Giddens, A., 1984, The Constitution of Society, Polity Press, Cambridge

Kusch, M.,1996, 'The Sociophilosophy of Folk Psychology', Studies in History and Philosophy of Science 28, 1-25

Searle, J., 1995, The Construction of Social Reality. London, Allen Lane, The Penguin Press

Tuomela, R., 1985, Science, Action, and Reality. D. Reidel Publishing Company, Dordrecht, Boston, and Lancaster

Tuomela, R.,1995, The Importance of Us: A Philosophical Study of Basic Social Notions, Stanford Series in Philosophy, Stanford University Press, Stanford, Calif.

Tuomela, R. 1998, 'A Defense of Mental Causation', forthcoming in Philosophical Studies (April 1998).

Tuomela, R. and Balzer, W., 1997, 'Collective Acceptance and Collective Social Notions', manuscript

Tuomela, R. and Bonnevier-Tuomela, 1997, 'From Social Imitation to Teamwork', in Holmström-Hintikka, G, and Tuomela, R. (eds.), Contemporary Action Theory, vol. II: Social Action), Kluwer Academic Publishers, Dordrecht, pp. 1-47