Institutional Objects, Reductionism, and Theories of Persistence

Tobias Hansson Wahlberg

Lund University

DRAFT!

Abstract:

This paper addresses the issue whether institutional objects can be identified with

physical objects that have been ascribed certain status functions, as advocated e.g. by

John Searle in The Construction of Social Reality (1995). It argues that the prospect for

such identification hinges on how objects persist – i.e. whether they endure, perdure or

exdure through time.

1. Introduction

In day-to-day life, we routinely quantify over and refer to institutional objects such as nation-

states, governments, laws, corporations, currencies, contracts, etc. The ontological status of

institutional objects is far from straightforward, however. One reason for being puzzled about

their mode of being is this: If humanity were to disappear, or if we just stopped thinking that

there are nation-states, corporations and dollar bills, arguably there would no longer be (in

whatever sense they “are”) any entities of these kinds.1 Contrast this sort of mind-dependence

with the mode of existence of stars, animals and electrons (natural objects), or bricks, pieces

of paper and cups (artefacts). If we all went into a permanent coma, still there would be stars,

pieces of paper and cups.2 Artefacts may be mind-dependent in the sense that their coming-

into-being is in need of intentional thought, not only causally, but logically (Hilpinen, 2004;

Thomasson, 2007); but institutional objects seem to be mind-dependent, logically, in a

stronger way: some kind of ongoing representational activity appears needed in order for there

to be nation-states, corporations and dollar bills.3 Moreover, typical institutional objects seem

1 Compare the thought experiment offered by Searle (1995, p. 11), although his rough-and-ready test involves the idea that there had never been any human or sentient beings. 2 Mereological Nihilists (e.g. Wheeler, 1979) deny that there are any composite macroscopic objects; according to them there are only mereological simples. (Some more moderate nihilists allow that there are animate composites, e.g. van Inwagen, 1990a.) Global constructivists (Goodman, 1978, and Rorty, 1979, are candidates) and idealists (e.g. Berkeley, 1734/1998) think that all of external reality is constructed or mind-dependent. In this paper I will simply take for granted that nihilism, global constructivism and idealism are false doctrines.3 Searle can be read as suggesting that there is no distinction between institutional objects and artefacts in this regard (1995, pp. 11, 20-1). If he does hold this I think he is wrong; more about this issue below.

1

to require social recognition, not merely the recognition of a single individual – wherefore the

study of their mode of being is often referred to as “social ontology” (see e.g. Thomasson,

2003; Searle, 2010).

What, then, is the ontological status of institutional objects? Do they exist “out

there” in the external world, or are they merely some sort of shared mental constructs or

useful fictions? If they exist in the external world, how are they related to the mind-

independent aspects of reality? Can they be reduced to, i.e. identified with, natural objects or

artefacts that have been ascribed certain functions? Or are they sui generis entities that

supervene on natural/artefactual objects and human attitudes? (For simplicity, in what follows

I will often refer to natural objects and artefacts as mind-independent or brute, but the reader

should bear in mind that some artefacts may be mind-dependent in the way indicated above.)

Some writers apparently hold that a single answer will do for all sorts of institutional objects

(e.g. Searle, 1995). I find it prima facie plausible, however, that different answers are called

for depending on what specific kind of institutional object is in question (compare prime

ministers with complex international organizations like the E.U.).

In this paper, I investigate the tenability and scope of the externalist, reductive

view, more specifically as it is defended by John Searle in his (1995). I join Amie Thomasson

(2003) and others in arguing that although the theory is plausible in relation to institutional

objects such as dollar bills and presidents, the position is in trouble in relation to other kinds

of institutional objects. I show that the problematic class of institutional objects is larger than

is typically envisaged – e.g. by latter-day Searle (2006; 2010) who has made some

amendments to his original theory, due to the criticism of Thomasson and others. At least, the

class is larger as long as the commonsensical endurance theory of persistence (also known as

three-dimensionalism) is presumed, as it typically is (albeit tacitly) by current social

ontologists. My argument relies on Leibniz’s Law (LL) and I defend its applicability to

institutional objects against scepticism of such application that might arise due to certain

remarks made by Searle. I moreover discuss, and illustrate in some detail, how externalist

reductionism fares in the light of some currently popular revisionary theories of persistence:

the perdurance and stage theories (which are varieties of four-dimensionalism). I argue that

the prospect for externalist reductionism improves if either of these theories is adopted, but

that the improvement comes with certain costs.

Interestingly, it has since long been appreciated within the philosophy of

material constitution (i.e. constitution of natural objects and artefacts) that theories of

persistence are relevant to issues of reduction and supervenience (see e.g. Rea, 1997; Sider,

2

2001). Such considerations have as of yet to be applied to the metaphysics of institutional

objects, however. This paper aims at remedying this oversight and initiating such

philosophizing.

2. Institutional objects as reducible

In this section I focus on John Searle’s formulation of the externalist, reductive view, in

particular as it is defended in his seminal book The Construction of Social Reality (1995).

(The externalist, reductive position will from now on be referred to as “the reductive view” or

“reductionism”.) Searle’s 1995-theory is the most developed version of reductionism that I

know of and it has been highly influential. However, the principal difficulties identified are

applicable, I think, to any version of reductionism. (As indicated in Section 1, Searle has to

some extent restricted his reductionism in later writings (Searle 2006; 2010); I explain how

and why in Section 3.)

According to Searle, institutional concepts, such as “dollar bill”, “professor” and

“president”, are self-referential. By a self-referential concept he means a concept F such that

it is analytically true that for an object a – existing in the external world – to fall under it a has

to be believed to be, or regarded as or treated as, an F (Searle, 1995, pp. 32-4).4 Thus, on this

view, being represented as an F is logically prior to being an F (ibid., p. 13). Searle takes the

phenomenon of self-referentiality to explain our intuition that institutional objects are mind-

dependent: there would not be any Fs in the world if we did not think that there are any (cf. p.

33). He allows, however, that an object a can be of institutional kind F without it being

thought to be an F (i.e. without us having singular attitudes about a) – but in such cases,

Searle holds, we are accepting some general rule to the effect that all entities x, having certain

features G, H, I... (exemplified by a), count as instances of F, whereby they are Fs (pp. 32-3,

46).5 Hence, professors and dollar bills are indeed external objects, but their being professors

and dollar bills is not intrinsic to them; this is a status that is assigned to them (pp. 9-13).

Moreover, Searle maintains that institutional concepts express “status functions”

(pp. 40-1). Status functions are functions such that they could not be performed or had by an

object just in virtue of the object’s physical properties and abilities (if it has any) but go

beyond them; they presuppose the assignment of a new status to the object that is to perform

4 He denies that the idea involves vicious circularity (ibid., pp. 52-3). 5 For example, if a certain dollar bill has fallen from the printing presses into the cracks of the floor, and has never been thought of as a dollar bill, nevertheless it is a dollar bill and not just a piece of paper, because of a general rule applying to it (ibid., p. 32).

3

the function in question.6 Typically, status functions involve, either directly or indirectly, what

Searle calls deontic or conventional powers: rights, responsibilities, obligations, duties,

entitlements, etc. (pp. 100, 104-5).7 A bearer of a dollar bill, for example, has the right to use

it for payment in the U.S., because dollar bills are legal tender in the U.S. Moreover, Searle

seems to regard the observer-relative “status” and its associated status function as being so

intimately connected that in many cases he simply speaks of “status functions” rather than

“statuses” and their associated “status functions”. In what follows I will adopt this way of

speaking. Consequently, I will say that a predicate such as “being president of the U.S.”

expresses a status function rather than say it expresses a status with an associated status

function. (In many cases it seems that a status is in fact defined in terms of its status function

– for example, to be president of the U.S. is simply to have a certain status function, i.e. to

have certain deontic powers.)

Since status functions go beyond the sheer physics of objects they must,

according to Searle, be continuously assigned or accepted in order to be successfully had and

performed (p. 45). That is, it is not enough that we assign the status function to some object

once and then forget about the function, in order for the status function to be instantiated over

time.8

When assigning a certain status function to some object(s), we are applying or

accepting (if we are not inventing) some specific constitutive rule, or a system of such rules

6 It should be noted, however, that for Searle, all functions are “observer-relative” in that they are teleologically, not just causally, defined. Nevertheless, he accepts that many functions can be performed simply in virtue of objects’ physical properties, ones a goal or value has been specified (pp. 14-5). For example, a heart causes blood to flow through a body and relative to the goal of survival it “functions” to pump blood (this is a “nonagentive” function on Searle’s terminology); a corkscrew has certain physical characteristics and in virtue of these it can be used for unscrewing corks (an “agentive function”) (p. 20). Status functions form a special sub-class of functions whose members are, so to speak, further removed from the mind-independent aspects of reality: they may even be arbitrarily related to the physics of the things in question – fiat money being a telling example, according to Searle (pp. 40-2).7 For criticism of the step from functions to deontic powers, see Tuomela (1997); for Searle’s reply, see Searle (1997).8 In spite of his view of non-institutional agentive functions as not going beyond the sheer physics of things (apart from the teleological part) Searle seems to maintain that they too have to be continuously assigned in order to be had over time (Searle, 1995, pp. 11, 20-1). The exact reason for this is not made entirely clear, however. I would maintain, in any case, that the concepts expressed by predicates such as “is an urn” and the concepts expressed by predicates such as “functions as an urn” should be kept apart. The first kind of concept only seems to require a past or present imposition of some agentive function, while the latter may require, if Searle is correct, the present assignment of the agentive function in question. Thus, I think that when an archeologist discovers an old artefact it can be true to say that the artefact is (present tense) an urn, even though it has not been used or assigned the agentive function functions as an urn for centuries. It would be true to say this because the artefact was designed for the purpose of functioning as an urn – not because the archaeologist now assigns this agentive function to the object. Such a past assignment does not seem to be enough for the satisfaction of institutional kind concepts. A person who only used to be counted as a prime minister no longer is a prime minister (although it is true to say that the person is a former prime minister); an expired dollar bill is no longer a valid dollar bill (it is merely a former dollar bill). It appears, then, that artefactual and institutional kind concepts are different in this respect.

4

(pp. 27-8). Constitutive rules should be distinguished from mere regulative rules, Searle

maintains (ibid.). The latter rules regulate activities that can exist without the rules.

Constitutive rules, on the other hand, create the possibility of a new kind of activity, fact or

object. For example, according to Searle, the constitutive rules of chess (“such and such a

move counts as a legal pawn move”) create the possibility of chess playing, which is a

different activity from merely moving pieces of wood on a board (even if the latter

movements are indistinguishable from genuine chess moves, i.e. moves intentionally

governed by the constitutive rules of chess).9

Constitutive rules characteristically have the following form (p. 28; see also

Searle, 1969, pp. 33-7, for a general discussion):

X counts as Y in context C.

In the case of institutional objects (Searle, 1995, p. 44), the X term is a placeholder for an

expression referring to some object (or type of object), the Y term is a placeholder for an

expression expressing some status function, while the C term specifies the context in which

the “counts as” holds (if the status function is not conferred across contexts).10 For example,

on Searle’s view, the predicate “being a dollar bill” expresses a status function which is

assigned to certain pieces of paper issued by the Bureau of Engraving and Printing under the

authority of the U.S. Treasury; these bits of paper count as dollar bills, and thereby they are

dollar bills. Thus, the existence of dollar bills is dependent on the acceptance of some non-

empty set of constitutive rules (pp. 45-6). And since “an institution is a system of constitutive

rules” (Searle, 2010, p. 10) dollar bills are aptly called “institutional objects” (cf. Searle,

1995, p. 44).11

In the example above the status function being a dollar bill is assigned to an

artefact (a piece of paper), which is a mind-independent object (see Section 1). Searle holds,

9 See also Searle’s examples of people accidently going through the moves of American football (Searle, 1969, pp. 35-6) and unconscious robots behaving as if they had money (Searle, 2010, pp. 135-7). As an example of a regulative rule Searle mentions “drive on the right-hand side of the road” (Searle, 1995, p. 27); driving can exist without such rules. The distinction between regulative and constitutive rules is criticized by Ruben (1997); for Searle’s reply, see Searle (1997).10 If the term in the X-position denotes an event or a process we get an institutional event/process, e.g. a marriage or an election. In this paper I am exclusively concerned with the creation of institutional kinds of objects.11 Notice that Searle cannot, somewhat paradoxically, hold that institutions themselves (i.e. systems of constitutive rules) are institutional entities. At the very least, he cannot hold this of all institutions. Not every institution can depend for its existence on constitutive rules, on pain of an infinite regress of constitutive rules. It should be noted, though, that Searle is not very consistent in his use of “institution”. In his (2010, p. 91) he cites examples of institutions which clearly are not just systems of constitutive rules, but rather institutional objects: baseball teams, hospitals, universities, restaurants, governments, corporations, etc.

5

however, that the formula may be iterated (ibid., p. 80) so that a new status function is

assigned to an object which already is an institutional object, i.e. an object which has already

been assigned a status function. For example, the status function being president of the U.S.

may be assigned to (and only to) a citizen of the U.S. Searle insists, however, that such chains

of status function assignments must be grounded in, or start off with, mind-independent, brute

objects (pp. 55-6, 121).

In what sense, now, is this account reductive? It is reductive in that it identifies

institutional objects with brute ones. An institutional object is not an entity over and above a

brute object: an institutional object neither supervenes on, nor is it constituted12 by, a brute

object. Rather, an institutional object is numerically identical with a brute object. Of course,

not every brute object is an institutional one, on this scheme of things. Only those brute

objects that have been assigned a status function are institutional objects. More precisely, only

those brute objects that have been assigned a status function of a sort that defines and yields

an institutional kind of object – e.g. a professor, a dollar bill – are institutional objects.13 But it

is important to realize that the assignment of such a status function to a brute object does not,

on this account, create a new particular object that exist “on top” of the brute one, although

the status function as such is an institutional addition to the brute object. What happens is just

that a pre-existing object acquires a new status function, with the result that the pre-existing

object becomes a new kind of object (a dollar bill say, in addition to being a piece of paper).14

Thus, on Searle’s theory, an institutional object is identical neither with the conjunction or

sum of a brute object X and a status function Y, nor with the fact that X is a Y (X’s being a

Y). If any of these positions were endorsed, then the institutional object would be regarded as

containing the brute object as a proper part or constituent.15 But then it would not be the brute

X-object that counts as a Y and consequently is a Y. The institutional object would be

something over and above the X-object (although arguably overlapping it). But Searle

explicitly says that the constitutive formula is of the form X counts as Y.

12 In the sense of “constituted” discussed by e.g. Johnston (1992/1997) and Wiggins (2001), according to whom constitution is not identity.13 Consequently, brute objects that have only been ascribed a status function which is just an institutional property – i.e. not a kind or sortal property – do not qualify as institutional objects; they are mere brute objects that in addition have, what might be called, a “characterizing” institutional property. A candidate for being such a characterizing institutional property is being insolvent. I elaborate somewhat on this Aristotelian-Lockean distinction (see e.g. Strawson, 1959, pp. 168-9; Wiggins, 2001, pp. 8-9) below; the distinction is not explicitly made by Searle himself. 14 Of course, it may be that in some cases the status function is assigned simultaneously with the creation of the brute X-object. 15 For a discussion of the metaphysics of facts or states of affairs, see e.g. Armstrong (1997).

6

For the same reason I think it would be a mistake to interpret Searle as saying

that institutional objects exist partly in our minds (this is why I refer to his view as externalist,

see Introduction). Searle does sometimes write things such as “the attitude we take toward the

phenomenon is partly constitutive of the phenomenon” (Searle, 1995, p. 33; my emphasis);

but I think this is simply a misleading and unfortunate way of expressing the idea that

institutional phenomena are mind-dependent. The thesis is that there would not be any

institutional (kinds of) objects unless we thought there were any – but this claim does not

entail the position that proper parts of institutional objects exist in our minds. If the relevant

X-objects (e.g. certain pieces of paper) exist outside our heads, i.e. do not overlap with our

brains/minds, then the relevant institutional objects (dollar bills) exist wholly outside our

minds. This is because it is only the X-objects (pieces of paper) that are counted as

institutional objects (dollar bills), not the X-objects + our representations of institutional

objects. Thus, if I have a dollar bill in my pocket, I do not have a part of it in my head; the

dollar bill itself is wholly outside my mind. True, institutional concepts are relational, even if

monadic, in the sense that they involve the idea that their satisfiers stand in certain relations to

people having certain attitudes towards them; but that does not entail that their satisfiers

somehow extend beyond themselves into the minds of the people holding the attitudes. The

monadic predicate “is a father” is clearly relational, but that does not mean that the members

of its extension somehow extend beyond themselves to their children.

Thus, I read Searle as holding that a dollar bill simply is a piece of paper –

namely, a piece of paper (an X-object) which counts as a dollar bill (has status function Y). If

we name the dollar bill “a”, and the piece of paper which is counted as a dollar bill “b”, we

have two co-referential names, referring to a single thing. Again, a is not an entity over and

above b (e.g. b + status function Y). It is b (= a) which is counted as a dollar bill, and which

consequently falls under the predicate “is a dollar bill”. Likewise, the president of the U.S. is

identical with a certain U.S. citizen who counts as the president of the U.S.; and the U.S.

citizen is identical with a specific human being who counts as a U.S. citizen. In the end, the

president of the U.S. simply is a human being (Barack Obama) who counts as the president of

the U.S. Thus we have: the 44th president of the U.S = Barack Obama. Since the chain of

identities (if there is a chain) must terminate in mind-independent objects on this theory,

institutional objects are in the end reduced to (in the sense of identified with) mind-

independent ones (cf. Smith, 2003b, p. 18).

I realize that Searle does not call himself a reductionist about institutional

objects. But as far as I can see, this refusal stems from the fact that he denies that the

7

intentional acts of imposing status functions to objects are reducible. First, he holds that such

acts typically are in the “we-mode” rather than in the “I-mode”, and that we-mode

intentionality cannot be reduced to I-mode intentionality (pp. 23-26). Second, intentional

states cannot, according to Searle, be reduced to brain-states, although they are realized in the

brain (Searle, 1983). But an irreducibility stance towards intentionality can be combined with

the view that institutional objects are identical with mind-independent ones (namely those that

have been ascribed status functions). It is because Searle is committed to the latter thesis that I

call his theory reductive.16

Moreover, I am fully aware that Searle, in his response to one of Barry Smith’s

critical articles (Smith, 2003a), expresses dissatisfaction with generic talk about institutional

objects (a category which encompasses entities such as presidents, dollar bills, lawyers, etc.)

in contrast to generic talk about institutional and social facts (encompassing “entities” such as

the fact that Barack Obama is the president of the U.S.) (Searle, 2003). But this dissatisfaction

seems to stem from the fact that talk of institutional objects can give the impression that we

are discussing a class of objects which are distinct from mind-independent ones – which they

for Searle empathetically are not.

He writes:

In my hand I hold an object. This one and the same object is both a piece of paper

and a dollar bill. As a piece of paper it is a non-social object; as a dollar bill it is a

social [institutional17] object. So which is it? The answer, of course, is that it is

both. But to say that is to say that we do not have a separate class of objects that we

can identify with the notion of social object. Rather, what we have to say is that

something is social object only under certain descriptions [e.g. when described as a

dollar bill] and not others [when described as a piece of paper] 18 […] Thus there is

only one object that is both a piece of paper and a dollar bill, but the fact that it is a

piece of paper is not the same fact as that it is a dollar bill, even though they are

16 Again, at this stage of the paper I am referring to Searle’s pre-2006 view – that is, before he accepted “free-standing Y terms”; see the next section for Searle’s revised view.17 In his response to Smith, Searle is using the wider term “social object”, presumably because that is the expression used by Smith in his critical article.18 Searle here appears to be suggesting that an object is a social/institutional object only under certain descriptions. This is unnecessarily defensive (unless he just means that an object is institutional relative to an acceptance of a constitutive rule). I do not think he needs to relativize this status to descriptions. Given that the object is accepted as having a certain status function it is an institutional object, period. There is no contradiction involved in saying that the object is both mind-independent and institutional. For the same reason, I do not think Searle should write “The open sentence ‘X is a social object’ is not extensional with respect to substitutability” (Searle, 2003, p. 303). For more on this issue, see Section 4.

8

both facts about one and the same object. […] Again, when I am alone in my room,

that room contains at least the following “social objects”: a citizen of the United

States, an employee of the State of California, a licensed driver, and a taxpayer. So

how many objects are in the room? There is exactly one: me. (Searle, 2003, pp.

302-303, my emphasis)

It seems to me, then, that in his 2003-article Searle disavows the notion of institutional object

in favour of institutional fact because he wants to forestall what he takes to be a misconceived

picture of institutional reality: what is true, according to Searle, is that institutional facts are

distinct from brute facts; not that institutional objects are distinct from brute objects. It should

be noticed that in The Construction of Social Reality (1995), and in papers after his 2003-

paper, Searle allows himself to talk quite freely of social and institutional objects. Here are

some examples:

I said that the form of the constitutive rule was “X counts as Y in C”; but as I am

using this locution, that only determines a set of institutional facts and institutional

objects where the Y term names something more than the sheer physical features of

the object named by the X term (1995, p. 43-44). Such material objects as are

involved in institutional reality, e.g., bits of paper, are objects like any others, but

the imposition of status-functions on these objects create a level of description

where it is an institutional object, e.g., a twenty dollar bill. The object is no

different; rather, a new status with an accompanying function has been assigned to

an old object […] (1995, p. 57) We are talking about the mode of existence of

social objects such as the United States of America, the San Francisco Niners

football team, the University of California and the Squaw Valley Property Owners

Association (2006, p. 12)

Indeed, as I have indicated above, I think Searle is committed to the existence of institutional

objects – even if he regards institutional facts as being in some sense more fundamental than

institutional objects, and he sometimes speaks sceptically (but incoherently, I think) of

institutional objects as being “just placeholders for patterns of activities” (Searle, 1995, p. 57).

By (collectively) ascribing a certain status function F to a mind-independent object a an

institutional fact is created, that a is F, or that a is an F. But notice that if (and this is the

normal case) the term “F” expresses a status function that defines an institutional kind of

9

object (roughly, if the term is a count-noun and not just an adjective), then “F” prescribes

what sort or kind of object a is. And if a is prescribed as being an institutional kind of object

(a president, say), then a is an institutional kind of object (i.e., it is not merely a mind-

independent object having a “characterizing” institutional property, cf. Strawson, 1959, p.

168). If so, a is both an institutional and a mind-independent object: it falls under two

different sorts of kind terms, “president” (an institutional kind term) and “human being” (a

brute kind term).19

3. Difficulties with the reductive view

Reductionism is plausible, I think, in relation to the kind of examples discussed by Searle in

his pre-2006 writings on institutional reality. We do think of institutional objects such as

dollar bills, presidents and lawyers as concrete entities that can be seen, touched and heard,

and as having the same physical properties as their underlying “realizers”. We even take them

to have the same temporal properties as their realizers. If some person x becomes a barrister,

we do not regard this event as involving the creation of a new particular object. The age of the

person and the age of the barrister are taken to be the same. Likewise, if the person would

cease to be a barrister in the future, we would not regard this as the death or ceasing to be of

an institutional object, the barrister. Prima facie, the current referent of “the barrister”, i.e. the

person, would still exist, although no longer as a barrister – i.e. she would no longer satisfy

the present-tensed “is a barrister”, but she would still be in existence. Because the barrister

and the person prima facie have the same properties, LL (the principle that says that if entities

a and b are numerically identical, then whatever is true of a is true of b and vice versa) does

not seem to be a threat to their identification – i.e. reductionism appears plausible for such

cases.

If reductionism holds for all institutional objects, then there are no institutional

objects which go in and out of existence due to human attitudes and decisions. What happens,

rather, is that certain mind-independent objects begin and cease to satisfy certain relational

predicates (the institutional ones) due to human attitudes and decisions (or the absence

thereof). For example, if humanity were to disappear or stopped thinking that there are dollar

bills there would no longer be any dollar bills (i.e. there would no longer be any objects of

that kind), but the particular pieces of paper that once were the dollar bills (i.e. the entities that

19 Entity a will no doubt fall under other kind terms as well: “person”, “female” (or “male”), “animal”, “mammal”, etc. Which of these terms express brute kinds, and which do not, is of course up for debate (cf. Searle’s discussion of “human being” in his 2010, Ch. 8). In the text I have assumed for the purpose of brevity that “human being” expresses a brute kind.

10

used to satisfy “is a dollar bill”) would still be in existence. If reductionism about institutional

objects is true in general, then institutional count-nouns are, in David Wiggins’s terminology

(Wiggins, 2001, p. 30), merely phased-sortals: an object can begin, and cease, to be an

institutional object; an institutional object need not always be an institutional object.

However, reductionism about institutional objects does not seem plausible

across the board. In many cases there does not seem to be a suitable mind-independent X-

object with which the alleged institutional object can be identified. Amie Thomasson has

objected that many institutional objects seem to be too abstract to sustain such identification.

Entities as the U.S. Constitution, General Motors, or Calvinist doctrine obviously have

close relations to individual material objects, but in no case is there some particular

material object (with additional social properties) with which we can identify the

entity in question […] In the case of the U.S. Constitution, for example, even if the

original document in the Archives were destroyed, we certainly not thereby declare

that the United States was a nation without a constitution; a fire may destroy an

original document of historical interest, but not the U.S. Constitution itself. […] In

fact, it seems that the social and institutional world is becoming increasingly abstract

[…] This makes it increasingly important that a comprehensive social ontology

include abstract social objects as well as concrete ones. (Thomasson, 2003, p. 273)20

In later writings (2006; 2010) Searle has admitted that his original theory is implausible in

relation to prima facie abstract institutional objects. He now says that “we have to allow for

the attachment of status functions to abstract entities” (2010, p. 95). Using the terminology of

Barry Smith (2003b), Searle maintains that there can be constitutive rules with “freestanding

Y terms” (2006, p. 22; 2010, pp. 97-100). That is, he thinks that we can create institutional

objects of kind Y without there being a mind-independent, concrete X-object which is

counted as being of institutional kind Y. His examples are corporations and money

represented in computers and by credit cards (2006, p. 22; 2010, p. 101). (There is no money

in the computer, neither in the credit card, and the money represented need not exist

externally in the form of concrete currency.) It is not altogether clear, however, what the

20 Perhaps Thomasson is arguing here that because these institutional objects cannot be identified with material ones they are abstract. But I think the argument can be made to go in the reverse order: these institutional objects cannot be identified with concrete physical objects because they are abstract. In any case, the latter line of thought is the way I construe the objection here.

11

ontological status of these abstract institutional objects is for Searle.21 Regarding corporations,

for example, he writes: “there is an entity Y, the corporation, that has the function, even

though the entity is, as they say, a ‘fictitious’ entity” (2010, p. 100). And regarding money he

writes: “It is, for example, a mistake to treat money and other such instruments as if they were

natural phenomena like the phenomena studied in physics, chemistry, and biology. The recent

economic crisis makes it clear that they are products of massive fantasy.” (ibid., p. 201; my

emphasis) Does Searle think of abstract institutional objects as having external existence or

not? Interesting as this question is I will not pursue it further here.

I shall now argue that it is not just intuitively abstract institutional objects, such as

laws, electronically represented money, and corporations (and I would add states and

supranational entities here), that seem to refuse to be amendable to Searle’s original “X counts

as Y” formula. Some more mundane, concrete-looking institutional entities – i.e. institutional

objects that prima facie consist entirely of concrete entities, such as persons and/or buildings –

are also obstinate, such as governments, football teams, and universities.22

A first problem is that the underlying, mind-independent level in these latter cases

(governments etc.) prima facie consists of pluralities (even though the institutional objects are

conceptualized as units), while the X-term in Searle’s formula apparently is an expression

referring to a single thing or unit (he writes “X counts”, not “the Xs count...”). However, since

a single thing (such as an institutional object) cannot be identical with many things (pace

Baxter, 1988; see e.g. van Inwagen, 1994; Byeong-Uk, 1999),23 I think Searle is in fact well

off putting the formula the way it is stated – at least if the aim is reductive identification. But

what, then, are the mind-independent concrete units to which the singular X term refers in

these cases? The best suggestion, I think, is that the X term in these cases refers to a

mereological sum of people and/or buildings. Mereological sums are supposed to be concrete

wholes or units that exist independently of human attitudes (for the technical concept of a

mereological sum, see Simons, 1987). Moreover, on the standard conception of mereological

sums, any two objects, no matter how disparate and unrelated, form a mereological sum, i.e.

mereological summation is unrestricted. Given the latter, there will always be a mereological 21 Notice, however, that “money” seems to be a mass term, not a count-noun. You cannot ask: How many moneys do you have? You have to ask: How much money do you have? Nor can you say: This is a money.22 Cf. Ruben (1997); however, he does not explicitly distinguish between abstract and concrete institutional objects. Rather, he alleges that there are many kinds of institutional objects that resist Searlean reductionism. Some of his examples are concrete- and some abstract-looking. He does not go into details in the article in question. 23 Could Searle say “the Xs count as a unit U”, and then “U counts as Y”? He might try, but I think it would be an incredible move to suggest that because we count a plurality of objects as a unit they are a unit. I think I can understand how we can assign status functions to objects via rules of the form X counts as Y; but the idea of creating units of pluralities through such formulas is of a different magnitude.

12

sum ready to serve as referent for the X term in an act of status function assignment of the

kinds under consideration.

Other candidates for being the referent of the X term in the problematic cases at issue

are sets or (if we are concerned with people only) groups (see e.g. Searle, 1995, pp. 97, 104). I

think both of these candidates can be discarded at an early stage of inquiry, however. Sets are

unsuited because they are abstract (see e.g. Quine, 1960, p. 233), or worse fictitious (Hrbacek

and Jech, 1999, p. 1). Here we are trying to find real, mind-independent and concrete units

with which the institutional objects in question can be identified. Invoking groups would be

troublesome for the following reasons. First, groups are typically not conceived of as

metaphysical units by philosophers and social scientists, at least not by those working in the

methodological individualism tradition.24 Second, realists about groups sometimes distinguish

between groups in the sense of “mere collections” (i.e. groups that cannot survive any change

of membership) and groups that are such that they can survive changes of membership (see

e.g. List and Pettit, 2011, p. 31). If we postulate groups in the first sense then it seems that we

are simply back with the aforementioned proposals, because this notion of group appears to

involve either the idea of a set or the concept of a mereological sum, although restricted to

pluralities of persons. And then there would not be much gained by using the term “group”

(and methodological individualists will think we are speaking of mere pluralities). The second

group-concept is usually characterised in terms of people having some kind of group-attitude

towards the members of the group (see e.g. Ruben, 1985, pp. 19-21; Gilbert, 1989, Ch. IV) –

i.e. the second group-concept appears to be self-referential in Searle’s sense.25 But if that is

the case then groups of the second kind may very well be some kind of institutional objects, in

particular if they involve a deontology of commitments, obligations and rights, which they

seem to do, either directly or indirectly (cf. Gilbert, 1990). At the very least, they do not seem

to be mind-independent only (cf. Ruben, 1985, p. 22). But if they are institutional objects,

there must be some mind-independent concrete unit that is counted as a group, if we are to

hold on to Searle’s “X counts as Y”. And the best candidate for this mind-independent role, I

suggest, is, again, a mereological sum of persons. So we are simply back with mereological

sums. Consequently, in what follows I will proceed on the assumption that the mind-

24 See List and Pettit (2011) for a historical survey, but also for an opposing, realist view.25 List and Pettit think it is logically possible for there to be groups in the second sense without anyone having such group-attitudes. However, they write that “we are not aware of any examples of new group agents coming into existence among human beings in this way” (List and Pettis, 2011, p. 33). Moreover, one can question how realistic List and Pettit’s view of “group agents” really is. They use Daniel Dennett’s “intentional stance” (Dennett, 1987) as a vehicle for defining and discovering group agents (List and Pettit, 2011, pp. 19-24, 32), but it is debatable what features of the world are discovered when adopting the stance (see e.g. Churchland, 1988; Searle, 1988).

13

independent, concrete unit to which the status function is assigned in the problematic cases at

issue is a mereological sum. (Let me add that I think that the proposal of grounding, say, a

government in a preexisting group of people (in the second sense) gets things back to front; in

a well-functioning democracy at least, it is because the relevant persons have become

members of the same government that the persons involved may form a group in the second

sense. Also, please notice that groups in the second sense will in any case not do for

universities, consisting as they do not only of people but buildings etc.)

But now a second problem crops up: a government – or a football team, a university –

and a merological sum of persons/buildings typically do not have the same temporal

properties.26

Consider the case of a government. We take governments to be able to persist through

the replacement of at least some of their members. At time t a certain government g may be

constituted of a certain mereological sum of people s, and at some later time t´ g may be

constituted of a distinct sum of people s´ (some minister may have died and been replaced by

another one, s and s´ not being numerical identical). In relation to such a case I suppose Searle

would want to say that at time t, s counts as the government of the country (with the effect

that that s = g), and at time t´, s´ counts as the very same government (with the effect that s´ =

g).27 But due to the transitivity of identity, we should then have s = s´, in contradiction with

our premise that s ≠ s´. Moreover, according to LL, if entities a and b are numerically

identical, then whatever is true of a is true of b and vice versa. But here it is true of g that it

persists from t to t´, although this is not true of s (not if some part/person of s has ceased to

be). And it is true of g that it was the government at t, although this is not true of s´. So, given

LL, the government cannot be identified with s, nor can it be identified with s´. And as the

reader may easily verify for herself, similar logical problems affect the identification of

football teams (like England’s national football team) and universities with certain

mereological sums of people or/and buildings.

It seems, then, that mereological sums are, in the end, not suitable reductive bases for

the concrete-looking kinds of institutional entities mentioned. And since we have, for good 26 The problem is analogous to Wiggins’s classic case of a tree and the sum of cells constituting it at a certain time (Wiggins, 1968/1997). This kind of problem may be vaguely alluded to in the quote by Thomasson above. If so, my contribution here is that I make the problem fully explicit. Also, see Ruben (1985, p. 16) for a similar objection in relation to identifying institutional objects with sets. I think sets are doomed to begin with (if we want real and concrete reduction bases) since they are either abstract or fictitious. 27 On the standard conception of identity, identity statements with rigid designators are atemporal in the sense that it is, at best, redundant to qualify them with time clauses: if the identity holds between the referents “at some time”, it holds simpliciter. I here follow the standard conception of identity. For a deviant view, allowing for temporary identities, see Gallois (1997). For criticism of the temporary identities view, see Sider (2001, pp. 165-176).

14

reasons, discarded the other alternatives (sets and groups) it appears that reductionism can

only be true of a fairly restricted class of institutional objects, including entities such as

lawyers, dollar bills and presidents. Institutional objects such as corporations, laws, states,

governments, universities and football teams appear to be, at best, supervenient entities.

4. Leibniz’s Law and institutional objects

At this stage I need to address a worry that might arise regarding my reliance on LL. In the

preceding section, I took for granted that LL governs not only mind-independent objects, but

institutional ones as well. At first blush this might seem innocent and unproblematic because

how could any particular objects a and b be identical if something is true of a (or b) which is

not true of b (or a)? LL is a principle that looks impossible to deny and virtually all analytic

philosophers accept it.28 However, Searle has repeatedly argued that statements about

institutional reality are somewhat peculiar in that they “are not always extensional. They fail

such tests as the substitutability of coreferring expressions (Leibniz’s Law)” (Searle, 2010, p.

119; see also Searle, 1995, pp. 28-9). Suppose that Searle is correct about this – does that

overthrow the argument of the preceding section?

I do not think it does. To begin with, Searle and I mean different things by

“Leibniz’s Law”. I take LL to be a material principle governing identity: it is about objects in

the world and whether or not “they” are identical. For Searle, LL is a meta-linguistic principle

which can be used to test whether or not a sentence is extensional.

Here is how LL is stated by Searle: “if two expressions [names, predicates,

sentences] refer to the same object [individual, property, truth value] they can be substituted

for each other in a sentence without changing the truth value of the sentence” (Searle, 1995, p.

18). If a sentence fails this test, does not adhere to the principle, then the sentence is not

extensional but intensional. It is a well known fact that many sentences are intensional – for

example modal sentences and sentences reporting beliefs. (Note that some philosophers take

the phenomenon of intensionality to show that the expressions, when occurring in intensional

contexts, are not genuinely referring expressions after all – see e.g. Quine, 1960, §30.) Now,

Searle maintains that sentences predicating functions to objects are intensional too (1995, pp.

18-9). More specifically, he argues that “X counts as Y in C” is intensional by putting forth

examples where instances of the schema change truth value when the Y term, standing for a

28 Is not LL refuted by the humdrum observation that objects change their intrinsic properties over time? In my (2007) I show that it is not, even if objects are taken to endure through time.

15

certain status function, is substituted for an allegedly coreferential Y term (ibid., pp. 28-9;

2010, p. 119).

Grant Searle that “X counts as Y” fails to satisfy the meta-linguistic version of

LL in the way suggested.29 Does such failure show that the material version of LL does not

hold for institutional objects? No. There is still every reason to think that if the objects

referred to by the singular expressions in place of the schematic “a” and “b” in the material

version of LL are one and the same object (institutional or not), then whatever is true of the

“first” object is true of the “second” object, and vice versa. For example, if a = b then it will

be true of “both” of them that they count as Y if anyone of them does, and it will be true of

both of them that they fail to count as Y´ if anyone of them fails (even if Y = Y´).

The application of the material version of LL to institutional object would only

be threatened if it could be convincingly shown that the metalinguistic version of LL in some

cases fails where coreferring singular expressions (referring to an institutional object) are

substituted within the context of institutional or non-institutional predicates. By disquoting,

we would in effect have examples of the form: is true of a that it is F but it is not true of b that

it is F and yet a = b (and here it would be crucial that we take “F” as expressing the same

property when annexed to “a” as when it is annexed to “b”, if we are to have a

counterexample to the material version of LL). But Searle has not shown that there are such

cases. True, in his (2003) reply to Smith (2003a), more specifically in the context of the

discussion of the identity of the dollar bill and the piece of paper in his hand (see above,

Section 2), Searle declares: “The open sentence ‘X is a social object’ is not extensional with

respect to substitutability” (Searle, 2003, p. 303). (The idea seems to be that it true to say

“The dollar bill is a social object”, false to say “The piece of paper is a social object”, and yet

true to say “The dollar bill is identical with the piece of paper”.) But the reason why he says

this is apparently that he thinks that, although the dollar bill is the piece of paper, there is a

contradiction in saying, in one breath, using only one singular expression: “The piece of paper

is a mind-independent object and a social (or an institutional) object”. And the reason why he

thinks this is a contradiction is, it appears, that he thinks that falling under a predicate

standing for a mind-independent kind entails that the object is a non-social object. (Remember

that he wrote: “As a piece of paper it is a non-social object”. He explicitly says: “one and the

same thing can be a social object relative to one description, and a non-social object relative

29 In fact I find his chosen examples highly problematic, but there is no room to go into details here. The phenomenon, if genuine, is probably due to the fact that the part “counts as Y” reports some kind of intentional attitude of the agents ascribing the status function to the X-object (see Searle, 1995, p. 29); compare belief reports.

16

to another description” (Searle, 2003, p. 303).30) But I argued in Section 2 that there is no

contradiction per se in the notion that an object falls under a predicate standing for a mind-

independent kind and a predicate standing for a mind-dependent kind. If the piece of paper is

a dollar bill (i.e. falls under the institutional predicate “is a dollar bill”) then the piece of paper

is a mind-independent object and an institutional object, period. We do not have to say that it

is only under certain descriptions (e.g. when explicitly referred to qua dollar bill) that the

referent is an institutional object. And that is all very well, because it leaves the material

version of LL unscathed.

But could it not be the case – returning to Searle’s formula “X counts as Y in C”

– that there are instances of the formula where the X term cannot be substituted for a

coreferential term salva veritae? By disquoting, we would then have examples of the form: X

counts as Y but X´ does not, and yet X = X´. Searle has not offered such examples, though.31

And in fact, I do not think that he could offer such examples and yet hold that the singular

terms are referring directly to an external object. For it seems to me that the substitution

condition could fail only if the X position in the formula were in the scope of “counts as”

(which it does not appear to be). In such a case the whole formula would just be a de dicto

report or representation of the content of collective intentionality; i.e. the formula should

rather be put as “Persons P1, ..., Pn count X as Y in C”. The X term would then not be in a

“purely referential position”, as Quine would put it. If “X counts as Y” is a shorthand for

“Persons P1, ..., Pn count X as Y in C” then the truth value of the formula would hinge on what

specific X term is used because the term would serve the function of partly representing the

content of peoples’ attitudes, rather than referring directly to an external object.32 (Cf. Searle’s

own account of the intensionality of many reports of mental states in his 1983, pp. 22-6 and

Ch. 7; see also his account of the “aspectual shape” of intentional states in his 1992, p. 155.)

30 It should be noticed, however, that, strictly speaking, there is only a contradiction here if “x is a non-social object” is taken to mean “it is not the case that x is a social object”. If it just means that “x is a mind-independent object” then there is no contradiction because, as I argue in the text, it can be consistently said of an object that it is mind-independent and social (or institutional). For a general and detailed discussion of the importance of distinguishing different senses of expressions of the form “x is non-F”, see my (2007). 31 Moreover, I do not see how such a phenomenon, per se, would refute my argument of Section 2. As such, all the phenomenon would show is that there are counterexamples to the material version of LL involving institutional truths. It would still have to be shown just why the material version of LL cannot be applied the way it is applied in the specific argument. Just how could it be the case that g = s and g = s´ when s ≠ s´ and g persists from t to t´ but s does not and g was the government at t but s´ was not? Merely pointing at some counterexamples to the principle in general will not do.32 Here is an example of what can happen if we understand the formula in the de dicto dense: The sheriff and his deputies count Jesse James as a criminal but they do not count Mr. Howard as criminal, although Jesse James = Mr. Howard. In fact, in chapter 4 of Searle’s (1995), when the “X counts as Y” formula is analyzed in terms of deontic or conventional power, something similar to the above de dicto formula is put forth: “We accept (S has power (S does A)” (p. 104). “S” does not appear to be in a purely referential position in this formula.

17

Thus, if the “X counts as Y” formula is understood in a de dicto sense then it should not be

understood as, first, specifying an entity X in the external world, and then reporting how this

entity X (de re) is thought of or treated as or represented as by agents external to it. The

consequence is that we cannot disquote with the result that we are talking (directly) about an

external object X (= X´); rather we are merely reporting the content of intentional attitudes.

(The X term may not even have a referent outside the minds of the people thinking about the

“X-object”.) And then we will not have counterexamples to the material version of LL, which

is about external objects, not representations of external (or even non-existent) objects.

I conclude that so far no good reason has been offered for thinking that the

material version of LL does not govern institutional objects.

4. Revisionary theories of persistence to the rescue?

There is another caveat, however. The reasoning in the preceding sections has been conducted

with a tacit view of objects as persisting by enduring. Objects endure through time if they are

wholly present at distinct times as numerically the same three-dimensional entity (Johnston,

1987; Simons, 1987, Ch. 5; Wiggins, 2001). This view of persistence is often characterized

as being in line with common sense. I think it is safe to say that it is generally presupposed by

writers on social ontology.33 However, philosophers of time and persistence have developed

alternative, revisionary theories of persistence. The most influential ones are the perdurance

and stage theories. I will argue that if either of these theories is adopted, the anti-reductive

conclusion may be resisted – at least for the second class of problematic objects discussed in

Section 2 (i.e. the concrete-looking ones).

Consider the perdurance theory (see e.g. Quine, 1960, p. 171; Smart, 1963, p.

133; Lewis, 1976/1983; 1986, pp. 202-4). Briefly put, on this theory, objects are four-

dimensional, i.e. they have not only spatial parts but temporal parts as well. Consequently,

they extend both in space and time. Objects, being regarded as mereological sums of their

parts, are held to persist through time by having temporal parts at various times, i.e. by

perduring. Change over time is analysed as consisting of distinct proper temporal parts of

four-dimensional objects having distinct properties. The semantics implicit in this theory is

that ordinary sortal predicates, such as “person”, and ordinary names, such as “Al”, name the

perduring four-dimensional sums/wholes (often called “space-time worms”, if the sums are

33 An exception appears to be Tuomela (2007, p. 146), who seems to be adopting, at least hypothetically, something like the perdurance theory of persistence in relation to groups. He does not, however, discuss the kinds of problem I address with perduring institutional objects.

18

spatiotemporally continuous), not the proper temporal parts of the wholes (this is called the

maximality principle).34

Return now to the problem of the persisting government g which changes its

ministers over time (see Section 2). If perdurantism is adopted, we are in a position to hold

that there is, after all, a mind-independent, concrete X-object with the same temporal

properties as g. However, the X-object cannot be held to be a mereological sum of persons.

Given perdurantism, the X-object will be a mereological sum of proper temporal parts or

segments of persons. Here is why: Our government g began to exist at time t, but its ministers

predate t (although perhaps not as ministers, but definitely as persons). The pre-t segments of

the persons must consequently be excluded from the sum in question; otherwise g would have

pre-t temporal parts and would predate time t. Likewise, most of the ministers probably

outlive the cessation of g. (The government ceases to be after some unfavourable election,

say). The post-g segments of the four-dimensional persons must consequently be excluded

from the sum; otherwise g would perdure into its non-existent future. Finally, one minister,

we stipulated, ceased to be part of the government at time t´ and another person took over

his/her role in the government from that time onwards. The first person’s post-t´ segment (if

there is one, perhaps as a corpse) must be excluded from the sum, and the latter person’s pre-t

´ segment must likewise be excluded. We now have a perduring mereological sum whose

spatial parts are perduring segments of four-dimensional persons. This mereological sum is,

given perdurantism, the natural candidate for being Searle’s X-object to which “g” refers.35

The reader may easily verify for herself that similar accounts can be given for

the other types of problematic, concrete-looking institutional entities mentioned. However, a

drawback (if one dislikes counterintuitive results) with this type of reductive account is that,

intuitively, we think of a government (or a football team, a university) as consisting of

persons, not segments of persons (or/and buildings). This shortcoming is avoided if stage

theory is adopted.

Simplifying somewhat, one may say that stage theory retains the metaphysics of

the perdurance theory but revises its semantics (for details, see Sider, 1996, 2001; Hawley,

2001). The four-dimensional, perduring sums and space-time worms of perdurance theory are

accepted, but sortal predicates are taken to be satisfied by three-dimensional, instantaneous

34 Thus, according to the maximality principle, proper temporal parts of objects of ordinary kind F are not themselves objects of kind F. The denial of the maximiality principle would lead to overpopulation of Fs in the world. For discussion, see Lewis (1976/1983) and Hawley (2001).35 During the life-time of g the spatial parts of the sum meet on a regular basis, in virtue of their proper temporal parts, and make decisions; the decisions are made within the confines of a perduring building which houses the meetings of the successive governments of the state in question.

19

temporal parts of the sums, not by the four-dimensional sums themselves. Likewise, an

ordinary name (when indexed to, or evaluated at, a time) is taken to refer to an instantaneous

temporal part of a four-dimensional sum, not to the four-dimensional sum of which the

instantaneous temporal part is a proper part. An ordinary object (i.e. a satisfier of some

ordinary sortal-predicate, a referent of an ordinary name) is consequently identified with a

single instantaneous temporal part – or “stage”, as they are called by stage theorist – of a four-

dimensional sum. Moreover, according to stage theory, an ordinary object will typically stand

in many types of temporal counterpart relations to other stages/objects within the four-

dimensional sum of which it is a part. For example, a certain stage may be both a lump of clay

and a cup. Within the four-dimensional sum in question there may be other stages which are

lumps of clay and/or cups. If these other stages have the right characteristics and stand in the

appropriate relations36 to the first stage, then these other stages will be the first stage’s

temporal statue-counterparts and lump-of-clay-counterparts. Our first stage/object can then be

said to “persist” (although in various senses, and over different time-intervals) in virtue of

having different kinds of temporal counterparts at other times. The technical term used for this

kind of persistence is “exdurance” (see Haslanger, 2003).

Let me illustrate the theory in some detail in order to make it clearer: Consider a

process which we would normally describe in the following way: a lump of clay is formed

into a cup; subsequently the lump of clay is squashed, with the result that the cup is destroyed,

although the lump of clay survives the event (even if flattened). Stage theorists will want to

analyze the scenario as follows (cf. Sider, 2001, pp. 200-1): At some time midway through the

process there is a stage that is both a lump of clay and a cup. This stage stands in the lump-of-

clay-counterpart relation to an earlier stage which is a lump of clay but not a cup – the

midway stage does not, therefore, stand in the cup-counterpart relation to the earlier stage.

Moreover, the midway stage also stands in the lump-of-clay-counterpart relation to a later

stage that is a lump of clay but not a cup; the midway stage thus fails to stand in the cup-

counterpart relation to this later stage. If we pick out our midway stage as a lump of clay, then

we can say that it (qua lump of clay) persists through the whole process (in virtue of its earlier

and later lump-of-clay counterparts). If we pick out the same stage as a cup, then we will deny

that it (qua cup) persists through the whole process (because of the lack of cup-counterparts

located at the appropriate earlier and later times). There is no contradiction involved in this

because on stage theory the exact meaning of “persist” (or “exdure”) is context-sensitive: 36 Stage theorists typically want to remain neutral about the exact nature of the temporal counterpart relation (see e.g. Sider, 2001, p. 194), but I think it is natural to suppose that it involves spatio-temporal continuity, causation, and similarity relations (see my 2011).

20

when attached to or associated with “cup” it means “has temporal cup-counterparts”, when

attached to or associated with “lump of clay” it means “has temporal lump-of-clay-

counterparts”, and when attached to or associated with “stage” it means “is identical with a

non-contemporary stage”. Alternatively, we could introduce distinct persistence-terms that

express the various persistence modes (cup-persistence, lump-of-clay-persistence etc.) in a

context-free way.37 The latter would allow us to say, using a single subject expression, that the

midway stage (explicitly picked out as a stage) cup-persists over a certain interval, lump-of-

clay-persists over a longer interval, and does not stage-persist at all.

Now, stage theory applied to our government example would result in something like

the following. At time t there is a three-dimensional mereological sum of people, s. This sum

is qua stage instantaneous – alternatively put, it does not stage-persist. s counts as the

government at t. At time t´, there is a three-dimensional mereological sum of people, s´,

numerically distinct from s. s´ too is, qua stage, instantaneous. At t´, s´ counts as the

government. At t, “g” refers to s, at t´ “g” refers to s´. s´ is one of s’s temporal government-

counterparts, but it is not a temporal mereological-sum-of-people-counterpart of s; likewise

for s in relation to s´. At t it is true to say “g = s” even though it is true to say “g will persist to

t´, but s will not”. The two utterances are compatible because although “g” and “s” are co-

referential at t the latter utterance simply means that g has a government-counterpart at t´ and

that s does not have a mereological-sum-counterpart at t´. (Alternatively, the latter utterance

can be taken to mean that g will government-persist to t´ and that s will not mereological-

sum-of-people-persist to t´.) At t´ it is true to say “g = s´” even though it is true to say “g was

the government at t, but s´ wasn’t”. The two utterances are compatible because although “g”

and “s´” are co-referential at t´ the latter utterance simply means that g has a government-

counterpart at t and that s does not have mereological-sum-counterpart at t which is the

government at that time.

Thus, assuming stage theory, a prima facie concrete institutional object such as a

government turns out to be not only reducible, but reducible to a mereological sum of people

– which, ceteris paribus, might be considered an advantage over the perdurance theory. The

price for such a reduction, however, is a plethora of persistence modes and the notion that

names for persisting objects change their reference over time. The price for both kinds of

reductionism (i.e. the perdurance and the stage theory variants) is, obviously, a rejection of

the commonsense conception that objects endure through time.38 I leave it to the reader to 37 For detailed discussion of this issue, see my (2008).38 Some would also argue that by endorsing either the perdurance or the stage theory, one has to accept the B-theory of time, while endurantism goes with presentism, which is the commonsense view of time. However, in

21

decide whether these costs are worth paying; here I wanted to make clear what the alternative

conceptions are, since they have not been addressed in the literature on social ontology so far.

5. Concluding remarks

In this paper I have tried to show that theories of persistence are relevant to the question of the

ontological status of institutional objects. More precisely, I have argued that Searlean-style

reductionism can hold only for a sub-class of institutional objects, assuming endurantism.

This sub-class is smaller than is typically envisaged by critical writers such as Thomasson

(2003) and Smith (2003a; 2003b). In order to escape this stronger anti-reductive conclusion

an endurantist-reductionist will either have to adopt some deviant conception of identity, such

as Geach’s relative identity thesis (Geach, 1980/1997) or Gallois’s temporary identity theory

(Gallois, 1998), or some deviant conception of the logic of sortal concepts, such as Burke’s

dominance theory (Burke, 1994/1997). Merely denying that the material version of Leibniz’s

Law is applicable to institutional objects is not a credible move, I have argued. If social

ontologists are prepared to give up on endurantism (while retaining absolute identity and a

standard conception of sortal concepts), the prospects for enlarging the class of reducible

institutional objects are improved. However, in the case of perdurantism, the reduction base

will typically not consist of the sought-for types of objects, but rather temporal segments of

such objects.39 Stage theory promises to deliver the sought-for reduction, but only at the price

of a plethora of persistence modes and the notion that names for persisting objects

continuously change their reference. Again, let me stress that the purpose of this paper has

been neither to promote nor to discredit some specific view of persistence, but to initiate a

much-needed process of charting hitherto unmapped territories within social ontology.

8. References

Armstrong, D. M., 1997, A World of States of Affairs, Cambridge: Cambridge University

Press.

Baxter, L. M., 1988, “Many-One Identity”, Philosophical Papers, Vol. 17, No. 3, pp. 193-

my (2009b) I argue that all three theories of persistence require the B-theory, and that presentism should be given up on independent grounds.39 A further worry not addressed in this paper is that perduring entities may not have the right modal properties to sustain identification with ordinary macroscopic objects (van Inwagen, 1990b; Heller, 1990, Ch. 2; Hansson Wahlberg, 2009a). The problem might very well generalize to institutional objects. However, this issue cannot be sufficiently addressed here, but deserves a paper of its own.

22

216.

Berkely, B., 1734/1998, A Treatise Concerning the Principles of Human Knowledge, (ed.)

Dancy J., Oxford: Oxford University Press.

Burke, M. B., 1994/1997, “Preserving the Principle of One Object to a Place: A Novel

Account of the Relations Among Objects, Sorts, Sortals, and Persistence Conditions”, in

Rea M. C. (ed.) Material Constitution: A Reader, 1997, Oxford: Rowman & Littlefield

Publishers.

Byeong-Uk, Y., 1999, “Is Mereology Ontologically Innocent?”, Philosophical Studies, Vol.

93, No., pp. 141-160.

Dennett, D. C., 1987, The Intentional Stance, Cambridge, MA: The MIT Press.

Churchland, P., 1988, “The Ontological Status of Intentional States: Nailing Folk Psychology

to its Perch”, Behavioural and Brain Sciences, Vol. 11, No. 3, pp. 507-508.

Gallois, A., 1998, Occasions of Identity – The Metaphysics of Persistence, Change, and

Sameness, Oxford: Oxford University Press.

Geach, P. T., 1980/1997, “Reference and Generality (Selections)”, in Rea M. C. (ed.)

Material Constitution: A Reader, 1997, Oxford: Rowman & Littlefield Publishers.

Gilbert, M., 1989, On Social Facts, Princeton: Princeton University Press.

Gilbert, M., 1990, “Walking Together: A Paradigmatic Social Phenomenon”, Midwest Studies

in Philosophy, Vol. 15, No. 1, pp. 1-14

Goodman, N., 1978, Ways of Worldmaking, Indianapolis: Hackett Publishing Company.

Hansson, T., 2007, “The Problem(s) of Change Revisited”, dialectica, Vol. 61, No. 2, pp.

265-274.

Hansson Wahlberg, T., 2008, “Can I be an Instantaneous Stage and yet Persist Through

Time?”, Metaphysica, Vol. 9, No. 2, pp. 235-239.

Hansson Wahlberg, T., 2009a, “4-D Objects and Disposition Ascriptions”, Philosophical

Papers, Vol. 38, No. 1, pp. 35-72.

Hansson Wahlberg, T., 2009b, Objects in Time – Studies of persistence in B-time, Lund:

Media-Tryck.

Hansson Wahlberg, T., 2011, “Can Persistence be a Matter of Convention?”, Axiomathes,

Vol. 21, No. 4, pp. 507-529.

Haslanger, S., 2003, “Persistence through Time”, in Loux M. J. and Zimmerman D. W. (eds.)

The Oxford Handbook of Metaphysics, 2003, Oxford: Oxford University Press.

Hawley, K., 2001, How Objects Persist, Oxford: Oxford University Press.

Heller, M., 1990, The Ontology of Physical Objects – Four-dimensional Hunks of Matter,

23

Cambridge: Cambridge University Press.

Hilpinen, R., 2004, “Artifact”, Stanford Encyclopedia of Philosophy,

http://plato.stanford.edu/entries/artifact, 2011-06-08.

Hrbacek, K. and Jech T., 1999, Introduction to Set Theory – Third edition, Revised and

Expanded, London: Taylor & Francis.

Johnston, M., 1987, “Is there a Problem about Persistence?”, Aristotelian Society

Supplementary Volume, 61, pp. 107-135.

Johnston, M., 1992/1997, “Constitution Is Not Identity”, in Rea M. C. (ed.)

Material Constitution: A Reader, 1997, Oxford: Rowman & Littlefield Publishers.

Lewis, D., 1976/1983, “Survival and Identity” in Philosophical Papers, Volume I, 1983,

Oxford: Oxford University Press.

Lewis, D., 1986, On the Plurality of Worlds, Oxford: Blackwell Publishing.

List, C. and Pettit, P., 2011, Group Agency, Oxford: Oxford University Press.

Quine, W. V. O, 1960, Word and Object, Cambridge, MA: The MIT Press.

Rea, M., (ed.), 1997, Material Constitution: A Reader, Oxford: Rowman & Littlefield

Publishers.

Rorty, R., 1979, Philosophy and the Mirror of Nature, Princeton: Princeton University Press.

Ruben, D-H., 1985, The Metaphysics of the Social World, London: Routledge & Kegan Paul.

Ruben, D-H., 1997, “John Searle’s The Construction of Social Reality”, Philosophy

and Phenomenological Research, Vol. 57, No. 2, pp. 443-447.

Searle, J. R., 1969, Speech Acts – An Essay in the Philosophy of Language, Cambridge:

Cambridge University press.

Searle, J. R., 1983, Intentionality, Cambridge: Cambridge University Press.

Searle, J. R., 1988, “The Realistic Stance”, Behavioural and Brain Sciences, Vol. 11, No. 3,

pp. 527-529.

Searle, J. R., 1992, The Rediscovery of the Mind, Cambridge, MA: The MIT Press.

Searle, J. R., 1995, The Construction of Social Reality, London: Penguin Books.

Searle, J. R., 1997, “Responses to Critics of The Construction of Social Reality”, Philosophy

and Phenomenological Research, Vol. 57, No. 2, pp. 449-458.

Searle, J. R., 2003, “Reply to Barry Smith”, American Journal of Economics and

Sociology, Vol. 61, No. 1, pp. 299-309.

Searle, J. R., 2006, “Social Ontology: Some Basic Principles”, Anthropological Theory, Vol.

6, No. 1, pp. 12-29.

Searle, J. R., 2010, Making the Social World: The Structure of Human Civilization, Oxford:

24

Oxford University Press.

Sider, T., 1996, “All the World’s a Stage”, Australasian Journal of Philosophy, Vol. 74, No.

3, pp. 433-451.

Sider, T., 2001, Four-dimensionalism: An Ontology of Persistence and Time, Oxford:

Clarendon Press.

Simons, P., 1987, Parts: A Study in Ontology, Oxford: Oxford University Press.

Smart, J. J. C., 1963, Philosophy and Scientific Realism, London: Routledge & Kegan Paul.

Smith, B., 2003a, “The Ontology of Social Reality”, American Journal of Economics and

Sociology, Vol. 61, No. 1, pp. 285-299.

Smith, B., 2003b, “John Searle: From Speech Acts to Social Reality”, in Smith B. (ed.) John

Searle, Cambridge: Cambridge University Press.

Strawson, P. F., 1959, Individuals: An Essay in Descriptive Metaphysics, London: Routledge.

Thomasson, A., 2003, “Foundations for a Social Ontology”, Proto-Sociology: An

International Journal of Interdisciplinary Research, Vol. 18-19, pp. 269-290.

Thomasson, A., 2007, “Artifacts and Human Concepts”, in Margolis E. & Laurence S. (eds.)

Creations of the Mind, 2007, Oxford: Oxford University Pess.

Tuomela R., 1997, “Searle on Social Institutions”, Philosophy and Phenomenological

Research, Vol. 57, No. 2, pp. 435-441.

Tuomela R., 2007, The Philosophy of Sociality: The Shared Point of View, Oxford: Oxford

University Press.

van Inwagen, P., 1990a, Material Beings, Ithaca: Cornell University Press.

van Inwagen, P., 1990b, “Four-dimensional Objects”, Noûs, Vol. 24, No. 2, pp. 245-255.

van Inwagen, P., 1994, “Composition as Identity”, Philosophical Perspectives, Vol. 8, pp.

207-220

Wheeler, S. C., 1979, “On that which is not”, Synthese, Vol. 41, No. 2, pp. 155-173.

Wiggins, D., 1968/1997, “On Being in the Same Place at the Same Time”, in Rea M. C. (ed.)

Material Constitution: A Reader, 1997, Oxford: Rowman & Littlefield Publishers.

Wiggins, D., 2001, Sameness and Substance Renewed, Cambridge: Cambridge University

Press.

25