SPT v9n2 - Artefacts and Collective Intentionality

Number 2
Winter 2005
Volume 9

Artefacts and Collective Intentionality

Seumas Miller
Centre for Applied Philosophy and Public Ethics
(An Australian Research Council Funded Special Research Centre)
Charles Stuart University
Australian National University
University of Melbourne

Social reality is multifaceted and comprises at least social conventions and norms, social roles and relations, social institutions and social artefacts. John Searle’s ambitious project in his Construction of Social Reality is to show in general terms how social reality can be accommodated to physical reality ( Searle 1995 ). This raises a host of issues including how individual mentality and sociality are to be related, and how sociality and morality are to be related. Elsewhere I have argued against Searle’s anti-individualist conception of sui generis ‘we-intentions’ ( Miller 2001 , ch. 2). Specifically, I have argued in favour of what I term the collective end account of joint action according to which sui generis ‘we-intentions’ are neither desirable nor necessary. I have also suggested that Searle’s constructivist collective acceptance account of social forms emasculates central moral notions such as that of right and duty. On Searle’s account, human rights and correlative moral duties, for example, can exist only by virtue of collective acceptance of specific practices. Accordingly, in morally degenerate societies in which human rights are never respected, we must conclude that there simply are no human rights to be respected, for we have no external moral standpoint from which to make moral judgements in relation to collectively accepted social practices.

Searle’s account can also be assessed in terms of its adequacy at the level of more specific social forms, such as social artefacts and social institutions. Elsewhere I have argued against Searle’s account of social institutions and in favour of my own teleological account ( Miller 2001 , ch. 6). In this paper I want to consider Searle’s account of social artefacts. The social artefact category has a number of sub-categories, including technical objects. Technical objects are such things as screwdrivers, levers and the like, as opposed to coins, sceptres and such. (The latter might be termed “institutional artefacts” for reasons explained below.) I argue against Searle’s account of social artefacts. Moreover, I suggest that his general account is not well suited for the analysis of a number of sub-categories of social artefacts, including technical objects and institutional artefacts. I draw on my earlier work and also that of Peter Kroes in his paper “Screwdriver Philosophy: Searle’s Analysis of Technical Functions” ( Kroes 2003 ).

Artefacts and Functions

Artefacts such as chairs, tables, screwdrivers, levers, hammers, coins and so on are humanly designed and constructed physical objects that have certain functions. Their physical size, shape etc. have been designed so that they can be used for certain purposes. Accordingly, they have interacted causally with human beings and with other elements of the physical world at two stages, namely the stage of design and the stage of use. The screwdriver has been fashioned out of (say) wood and metal, and thus has been given a certain shape and size etc.. Moreover, a screwdriver is an implement that carpenters and others use to drive in and extract screws which are themselves artefacts that enable, for example, bits of wood to be held together tightly and in a stable manner. So artefacts include not only designed physical objects that causally mediate human agents and elements of the physical world that are not artefacts, many artefacts also mediate between human agents and other artefacts.

Naturally, there are objects, e.g., logs, which can be used as (say) chairs, but have not been designed as such. As Kroes points out, these involve cases of accidental as opposed to proper functions ( Kroes 2003 , 8). Accordingly, objects with proper functions are not only used to bring about purposes, they have been designed to serve those purposes. The properties of the physical object are designed in such a way that their causal powers are in the service of the human agents who use them.

This designed-in feature of artefacts also serves to distinguish artefacts from physical objects that are simply used to serve human purposes, but that do not have a function as such, e.g., trees that are cut down and used as fuel.

It is consistent with the above to hold that some token of an artefact type is an artefact notwithstanding the fact that the token has never in fact been used. On the other hand, if no tokens of a given design type have ever in fact been used because, contra their designer, there is in fact no use for them, then they are not really artefacts, or at least not paradigms of artefacts, but rather, say, mere ornaments.

The first general point to be made here is that in the paradigm case, at least, artefacts – in the sense of technical objects such as screwdrivers - have been causally acted on by human agents in the design phase, and in the use phase artefacts are both causally acted on and themselves causally efficacious in processes involving human agents bringing about certain outcomes. Specifically, in using an artefact an agent acts on an artefact, e.g., a screwdriver, in such a way as to cause the artefact in turn to cause an outcome, e.g., an extracted screw. Moreover, the designers and users of artefacts impose intentionality on artefacts in the sense, firstly, that the shape, size and other physical properties of an artefact have been intentionally brought about. Secondly, when artefacts are causally bringing about the outcome that they were designed to bring about this causal process is intended .

However, when Searle talks of artefacts and other social objects having intentionality imposed upon them he means that the fact that some physical object is an artefact, e.g., a screwdriver, is an observer-relative feature of the world. This seems to be either false or trivial. For the imposition of intentionality involved in artefacts is not necessarily that involved in standard cases of observer-relative phenomena, such as, say, visual perception; if someone ‘sees’ a red light when almost everyone else ‘sees’ a green light then – notwithstanding that the light is green – the person is having a visual experience of a red light. But just because the members of a hippie community under the influence of drugs sees what is in fact an ice cream cone as a screwdriver does not make it one. Nor for the same reason is it observer relative in the sense of relativism, i.e.: If A believes that p then p. The ice cream cone has not been designed to be and cannot be used as a screwdriver, notwithstanding the false beliefs of the members of the hippie community.

Given this standard sense of the term “observer-relative,” Searle would be taken to be claiming that beliefs and other cognitive states are a sufficient condition for being an artefact. As we have seen, this is clearly false. Kroes suggests that Searle, rather, intends to be claiming that such intentional states are merely a necessary condition for being an artefact. However, even this is doubtful, if the intentional states in question are the states of passive observers, as opposed to the states of causally active agents, i.e. if the states are beliefs, or other like cognitive states, as opposed to, say, intentions.

Suppose Robinson Crusoe designs and constructs an entirely new implement to break open coconuts, but falsely believes that the object is incapable of functioning as such; after years without being rescued Crusoe has become an incurable pessimist. Now suppose Man Friday finds the implement and tries to break open coconuts. In fact, Friday succeeds in opening the coconuts. However, under the influence of hallucinogenic drugs provided by Crusoe to ease the pain of social isolation, Friday believes that he never actually succeeds in opening any coconut. This implement is in fact an artefact, and has been designed and is used as such; but is not believed by anyone to be an artefact. The general point is that artefacts are not necessarily believed to be such; and so artefacts are not observer-relative in the sense that their being artefacts is dependent on someone believing them to be so.

On the other hand, artefacts necessarily involve human intentions . Artefacts are a primary locus of human agent causation. As such, the intentional dimension of artefacts is only observer-relative in the trivial sense that the intentions manifest in their design and use are, qua intentions, subjective states.

The second general point is that there appears to be a distinction between social and non-social artefacts. Doubtless, artefacts are in the paradigm case social objects. However, if Robinson Crusoe designs and uses a completely new device for breaking open coconuts, the Crusoe Coconut opener, and the device is never replicated or used or even known about by anyone else, then presumably the device is not a social object, though it is clearly an artefact. The third general point is that in order for artefacts – in the sense of technical objects – to be used successfully to achieve the outcomes for which they were designed, they have to have a number of properties, including properties relative to the outcome they are used to bring about, and relative to the agents who are to use them. Thus a screwdriver has to have a certain shape and size, if it is to be able to extract screws of a given shape and size. But it also has to have a certain size and shape, albeit within a range of possible sizes and shapes, if it is to be used by humans with hands of a certain size and shape, as opposed to, say, intelligent animals with huge tentacles but no hands.

Artefacts and Physical Objects

An important question arising here – and discussed by Kroes – concerns the necessity or sufficiency or otherwise of physical properties, as opposed to intentional properties, for something to be an artefact (technical object). Kroes convincingly argues that Searle’s analysis oscillates between defining artefacts in terms of their imposed intentionality and doing so in terms of those of their physical properties that enable them to be used as artefacts. Is an artefact a physical object different from other physical objects only in so far as it has intentionality externally imposed on it, e.g., a log used as a seat is an artefact, but one which is not so used is not a seat? Or is an artefact a physical object that has a set of physical properties that enable it to be used in various ways, even if it is in fact not so used? Such sets of physical properties may or may not have been designed-in.

If Searle is taken to be claiming that artefacts are simply physical objects whose properties enable them to be used in the service of human purposes then his definition is far too weak. Almost any physical object could be used for some human purpose. The very ground beneath us can be used as a seat; trees are chopped up, burned and used as fuel to heat us. Rocks could be used as parts of the walls of houses. However, none of these physical objects is an artefact.

Moreover, the definition of artefacts in terms of useable physical objects is inadequate for making out a distinction between artefacts in the sense of technical objects, and artefacts of other kinds, e.g., coins and the crowns of kings.

Nor will recourse to Searle’s notion of the imposition of collective intentionality deal with this problem. i For if screwdrivers and chairs are physical objects upon which functions have been collectively imposed, so are coins and the crowns of kings.

As Kroes points out, the definition of artefacts requires recourse to both sets of properties, i.e., physical properties and intentions; however, the difficulty lies in spelling out the precise relationship between them.

Without attempting to provide an adequate full-blown account of artefacts, I think the following is a serviceable definition for my purposes here. In the paradigm case, artefacts are physical objects whose physical property set has been designed by members of a social group to be used by agents in that group to achieve certain outcomes, and they are so used. Such paradigm cases include what might be regarded as non-institutional artefacts such as screwdrivers (technical objects), as well as institutional artefacts such as coins.

The distinction between institutional and non-institutional artefacts is problematic, especially given that both institutional and non-institutional artefacts have both a physical and an intentional dimension. However, my suggestion is that the distinction can be made out in terms of the different purposes or functions of these two categories of artefact. Roughly speaking, non-institutional artefacts, such as screwdrivers, have as their purpose or function to bring about changes in the physical world, e.g., to bring it about that screws are driven in or extracted. By contrast, institutional artefacts qua institutional artefacts have as their purpose or function to bring about changes in the institutional world, e.g., to bring it about that person A now owns the product he exchanged his coins for.

Status-Functions ii

According to Searle, institutional facts, including facts about institutional artefacts such as coins, are (ultimately) facts about physical objects, or states of affairs or events, upon which, what he terms, status-functions have been collectively imposed ( Searle 1995 , 41). I say “ultimately” because although status-functions can be collectively imposed on pre-existing institutional facts, any such iterated structure of status-functions must terminate at some point in physical objects or events (or more precisely, must terminate in what he terms, brute facts) ( Searle 1995 , 27, 34).

Something has a status- function if it has, or those who use it have, deontic powers. Thus a police officer has a status-function, and therefore a set of deontic powers, including rights to stop, search and arrest people under certain conditions. A five-dollar bill is a piece of paper (a physical object) the bearers of which have various deontic powers, including the right to exchange the bill for goods to the value of five dollars. These statusfunctions, and therefore deontic powers, have been collectively imposed in the sense that the relevant members of a community accept or agree to or otherwise treat the objects or persons that possess these status-powers as if they do in fact possess them. But in accepting or so treating, for example, the police officer as if he has the right to arrest people, the police officer comes to have that right. By Searle’s lights, if no one ever paid any attention to police officers they would cease to have any deontic powers and therefore any status-function; indeed they would cease to be police officers. Similarly, if no one were prepared to exchange five-dollar bills for goods then these bits of paper would cease to have any status-function, and the bearers of them would cease to have any deontic powers.

Searle’s account makes use of four basic notions: (1) imposition of functions; (2) the deontic powers of institutional persons and objects; (3) the distinction between constitutive and regulative rules; (4) collective intentionality. iii Let us look more closely at these notions, beginning with functions and deontic powers.

Functions and Deontic Powers

Searle’s notion of function concerns what it is that an individual (individually) or a group (collectively) imposes on a physical phenomenon. For example, if members of a community began to sit on a log then the log would in effect have become a bench. So a function – that of being used to sit on – would have been collectively imposed on a physical object. However, as we saw above, Kroes has pointed out that logs and the like are not artefacts in the paradigm sense of technical objects. For example, artefacts are designed rather than simply having a function individually or collectively imposed upon them by way of being used.

On Searle’s conception all functions ultimately depend on the existence of physical objects on which functions are imposed. However, some functions – such as the function of being a chair – depend on the specific physical properties of the object on which they are imposed. Thus a log can become a bench only if it has a certain size and shape; giant redwood trunks are not serviceable as seats. By contrast, a key feature of institutional facts is that they involve functions that are not able to be imposed simply by virtue of the specific physical properties of the phenomena on which they are imposed. Rather possession of the function exists by virtue of the collective character of the imposition ( Searle 1995 , 39). So being a chair is not an institutional fact; rather, its functionality exists by virtue of its specific physical properties. On the other hand, being a medium of exchange is an institutional fact; its functionality supposedly exists by virtue of collective imposition rather than specific physical properties.

It should be noted that according to Searle functions can be individually imposed on physical objects by individuals. This would be the case if, for example, Robinson Crusoe alone used the log as a bench. Further, if a log had the function of a bench collectively imposed on it, then its being a bench would be a social fact ( Searle 1995 , 88). However, the fact that the log was a bench would not be an institutional fact; for its being a bench does not depend on the collective character of the imposition of this function.

Here we should reiterate that in the case of artefacts at least, the physical properties themselves have been individually or collectively imposed viz. in the design phase. Moreover, the above-mentioned way of drawing the distinction between institutional and non-institutional artefacts is problematic since both chairs and coins are physical objects and not just any physical object could be a chair or a coin. Moreover, both chairs and coins are designed and used by collectives; so both involve collective imposition. However, in the case of non-institutional artefacts typically the physical properties need to be relatively specific. But why is this so?

Presumably this is because the outcome that they are intended to bring about, e.g. extracting a screw, is a purely physical outcome. Such artefacts need to have specific physical properties because of the nature of their function. By contrast, institutional artefacts, such as coins, typically do not need to have specific physical properties. Here it is a matter of degree. A set of large buildings could not serve as coinage. But why is there this latitude in the case of institutional artefacts? Presumably because the function of institutional artefacts is to bring about institutional changes, e.g., ownership. But why should this entail greater latitude? Perhaps because there is greater latitude in the design of institutions. Arguably, the purposes that institutions serve are more open-ended and vague than the purposes served by artefacts. But from this is does not obviously follow that there is great latitude in the design of institutional artefacts – as opposed to the design of the institutions themselves.

Following Searle, perhaps institutional artefacts invariably have a symbolic dimension, since institutional change necessarily involves communication, and institutional artefacts necessarily have a communicative function as part of their overall function. Certainly, there is great latitude in the choice of symbols. However, institutional artefacts have other non-symbolic functions in addition to their symbolic function. So the question arises as to whether or not these non-symbolic functions confer latitude in the design of institutional artefacts. Consider coins. Their principal function is to be used as a medium of exchange. But this purpose could be served by a wide variety of physical objects, so designers have considerable latitude in the design of mediums of exchange. On the other hand, coins need to be reasonably lightweight and durable, if they are to serve their purpose. So the latitude in their design is not all that great. Moreover, the designers of some non-institutional artefacts, such as tables and chairs, also have considerable latitude.

A further point here is that many non-institutional artefacts have a symbolic dimension, e.g., expensive dining tables. So it is by no means clear that the symbolic dimension of institutional artefacts is what explains the apparently greater levels of latitude that exist in the case of institutional, but not noninstitutional artefacts.

Ultimately, Searle intends to rely on his conception of status functions with deontic powers to distinguish between institutional and non-institutional facts, including artefacts. Accordingly, says Searle, what makes a physical object, such as a piece of metal or paper, a dollar coin or dollar note, i.e., an institutional artefact, is not its physical properties but rather collective imposition of that status function and associated deontic powers.

According to Searle, in the case of institutional facts the imposition of a function consists not only in the assignment of a function, but also of a status with deontic properties ( Searle 1995 , 100-1). These deontic properties – rights and duties and the like – are regarded by Searle as powers. If a police officer has a right to arrest you, then he has a power to do so; if you have a right to goods worth five dollars in exchange for your five-dollar bill, then you have a power to receive those goods. And to reiterate, these status functions or powers are possessed by virtue of the agreement or acceptance of the community; the powers are collectively imposed and maintained.

The tight connection that Searle makes between function imposition and deontic properties is problematic. Searle asserts that the notion of a function necessarily brings with it values and deontology. Thus if the purpose of the heart is to pump blood, then it ought to pump blood ( Searle 1995 , 15-6, 19). No doubt, purposes and ends can be assessed in the sense that they can be said to have been realised or to have failed to be realised. Moreover, if one has an end, then, other things being equal, one ought to enact the means to that end. So functions, goals and the like bring with them normativity in the minimal sense of the normativity of instrumental rationality. But the claim that functions and purposes bring with them normativity in any stronger sense has not been shown. Specifically, Searle has not shown that his deontic properties, such as rights and duties, can be derived from the notion of a function. Nor has he shown that deontic properties can be derived from his notion of function taken in conjunction with collective intentionality. Indeed, it would seem that his own examples of tool using social animals is sufficient to demonstrate that deontology is not generated by functionality, whether collectively imposed or not.

A more plausible source of some deontic properties is rules in the sense of explicitly formulated directives issued by authorities to perform certain kind of actions under certain circumstances. However, Searle claims that ‘rules’ in the very general sense of general policies can be abused, and hence have a normative status ( Searle 1995 , 48). As I have argued elsewhere, I do not accept the proposition that general policies or habits necessarily generate standards of conduct – as opposed to mere regularities in behaviour – and therefore deontic properties. Certainly Searle offers no demonstration of his claim.

At any rate, I suggest that it is a non sequitur to move from individual or collective imposition of functional properties to deontic properties. For it is perfectly possible for a person or an object to have a set of functions – and be treated as having those functions – without any associated (noninstrumental) deontology. For example, a set of individuals might use a certain sort of relatively rare shell as a medium of exchange, and do so notwithstanding the fact that no-one had any desire to possess these shells independent of the fact that they could be used as a medium of exchange. In this scenario all that is required is that each exchange shells for goods, and goods for shells, intend to continue to do so, and believe that all the others do and intend likewise. Of course it would add greatly to the stability of this arrangement if these shells were somehow authorised as an official medium of exchange, and if a (rule constituted) system of rights and duties in relation to the exchange of these shells was introduced and enforced. However, such a deontological structure is not a necessary feature of the system of exchange. iv Here we have artefacts and imposition of collective intentionality, yet there are no deontic powers, and hence (arguably) no institutional artefacts.

Moreover, even where a deontological framework of the kind described by Searle has been adopted, the relevant deontic powers do not subsume, take the place of, provide the basis for, or go hand in hand with, the substantive functions of the objects or persons in question. Here by substantive functions I mean the functions that are central and necessary to the object or person being the kind of thing that it is. Consider an incompetent surgeon who is incapable of performing a successful operation on anybody, and who largely avoids doing so, or when he absolutely has to, always ensures that he is part of a team comprised of other competent surgeons and nurses who actually do the work. By virtue of being a fully accredited surgeon this person has a set of deontic powers, including the right to perform surgery, and others have deontic powers in relation to him, including the right that he performs operations competently and with due diligence. Moreover, these deontic powers are maintained in part by (say) the Royal College of Surgeons, his colleagues and the community. However, the surgeon simply does not possess the substantive functions of a surgeon. The deontology is there but the substantive functionality is not. Accordingly, it is simply false to claim, as Searle must do, that he is a surgeon ( Searle 1995 , 49-50). v If someone cannot perform, and knows nothing about, surgery, he cannot be a surgeon, irrespective of whether he has the highest professional qualification there is, is treated as if he were a surgeon, and indeed is widely believed to be the finest surgeon in the land.

Now consider a currency that has been devalued by virtue of, say, a collapse in the economy, so that notwithstanding government declarations, very few are prepared to exchange goods for the official currency; the currency has official status, but has largely gone out of use. So the currency has deontic powers underpinned by the government, but no longer has substantive functionality. Accordingly, deontic powers are at best a necessary condition for being an institutional artefact.

I conclude that in relation to institutional facts, including artefacts, Searle’s notion of deontic power – whether collectively imposed or not – does not necessarily subsume, take the place of, provide the basis for, or go hand in hand with, substantive functionality. In the case of artefacts, as well as institutional roles, substantive functionality and deontic powers can come apart; there can be substantive functionality without deontic powers, and deontic powers without substantive functionality. However, institutional artefacts and institutional role cannot simply be defined in terms of deontic powers. For if there is no substantive functionality then there is no artefact, institutional or otherwise, and if there is no substantive functionality there is no institutional role. Thus deontic powers are at best a necessary condition for being an institutional artefact.

Let us turn now to Searle’s notion of collective intentionality, and the related notions of social acceptance and social agreement.

Collective Intentionality, Collective Acceptance and Constitutive Rules

Searle’s apparatus of status-functions is created and sustained by what he terms collective intentionality. As we have just seen, his view that the substantive functions of objects and persons is subsumed by, or necessarily goes hand in hand with, deontic powers is incorrect. Moreover, it is now clear that collective intentionality cannot generate the substantive functions of many institutional artefacts and persons. What of the relationship between collective intentionality and deontic powers? Let us get a little clearer on the notion of collective intentionality.

Collective intentionality is for Searle a primitive notion expressed by locutions such as “we-intend.” A we-intention is not reducible to an individual intention, nor to an individual intention in conjunction with other individual attitudes such as individual beliefs ( Searle 1995 , 24-6). It will be evident from my discussion in the opening section of this paper that I do not accept Searle’s conception of primitive we-intentions. I take the view that an analysis is possible using only individualistic notions, including especially the above-mentioned notion of a collective end. In any case Searle’s notion of primitive we-intentionality is under-theorised. He has not provided an account of it, analytical or otherwise.

But that aside, what of the role of collective intentionality – however it is to be understood – in relation to the creation and maintenance of deontic powers? Consider the deontic powers associated with the utterance of sentences, such as the right of hearers that speakers will aim at the truth. Searle must hold that such a right can only exist if a group adopts some sort of we-attitude, e.g., we-accept or we-agree or we-belief or we-intention, to these deontic powers. And presumably the same goes for other moral rights that may attach to artefacts, such as intellectual property rights in relation to inventions or cultural rights in relation to social artefacts created by a particular social group such as the Australian aborigines. In the latter case if the larger Australian community does not accept or we-agree to the contemporary minority aboriginal population having custody and other rights to ancient indigenous artefacts, does it follow that there are no such moral rights?

Arguably all humans have a right to life, vi hearers have a moral right that speakers aim at the truth, and individual and group ‘inventors’ of artefacts have intellectual and cultural property rights to those artefacts. Moreover these moral rights exist quite independently of whether the larger community in which those individuals or groups live we-intends that this be the case, or accepts that it is, or whatever. So the source of at least some central deontic properties, namely, some moral properties, is evidently not collective intentionality.

Naturally, whether or not deontic properties, including rights, are respected might ultimately turn on the attitude of the community. So general community ‘acceptance’ of deontic properties, including rights – in a weak sense of acceptance – is necessary for the maintenance of the deontological framework in that community. So much is trivially true.vii

I conclude that Searle has not established that all or most deontic properties associated with institutional artefacts and persons are (so to speak) ontologically dependent on collective intentionality ( Searle 1995 , 104). viii

I now want to focus attention on the connection between collective acceptance and substantive functionality. I will begin with the special case of institutional authorities.

What is it to collectively accept or collectively treat someone as if he or she is a leader? Presumably it is for people to obey her directives because they believe that they ought to do so. But now what is the actual or realised substantive functionality of a leader? Surely it is in large part getting people to obey one’s directives because they believe that they ought to do so. So in the case of leaders, and other institutional authorities, I suggest that Searle’s notion of collective acceptance is redundant; it collapses into actual or realised substantive functionality. To possess the functional properties of a leader is (in large part) to have one’s directives obeyed by one’s followers. But to have one’s directives obeyed by one’s followers is to be collectively accepted.

What of the relation of collective acceptance and substantive functionality in the case of institutional artefacts, e.g., coins or notes. Once again I suggest that Searle’s notion of collective acceptance collapses into the actual or realised substantive functionality of the institution. ix If people exchange dollar notes for the purpose or end of receiving goods (and exchange goods for the purpose of receiving dollars) then they have ‘accepted’ dollars in the only important sense of that term; so there is no need for an additional notion of collective intentionality or collective acceptance or treating dollars as if they were dollars.

And there is this further point. Persons using social artefacts, including as a medium of exchange, typically act in conformity with conventions, and indeed, social norms. So the substantive functionality of these institutions is realised in accordance with conventions and social norms. Accordingly, there is collective acceptance in the sense of conformity to conventions and social norms. When we use a medium of exchange we typically do so in part by conforming to the relevant conventions and social norms. So much is obvious.

A final point about Searle’s notion of collective intentionality and its relation to rules: In his later formulations it turns out that the we-intentional state in question is conventional in character; he speaks of conventional powers ( Searle 1995 , 104). Searle nowhere offers an analysis of conventions. However, he does distinguish them from rules. For Searle, conventions, by contrast with rules, are arbitrary ( Searle 1995 , 28). Moreover, as we saw above, he takes the view that rules are necessarily normative phenomena.

There are a number of unanswered questions here. What exactly is the nature of rules and conventions, and precisely how are these phenomena to be differentiated?

While Searle does not explain the nature of conventions or the nature of rules in general, he does offer an account of the distinction between constitutive rules and regulative rules. Moreover, constitutive rules have a key role in his account of institutions. Unlike regulative rules, constitutive rules “create the possibility of certain activities” ( Searle 1995 , 28). Constitutive rules have the form ‘X counts as Y in C’ where Y is the function imposed on X in conditions C ( Searle 1995 , 28). So constitutive rules are, according to Searle, very important in the construction of social institutions and institutional artefacts.

Now it is by no means clear that some conventions could not have the form “X counts as Y in C’. On Searle’s view of conventions as arbitrary, a convention-governed imposition of Y on X (if it were possible) would necessarily be a matter of arbitrary choice. But this seems to be so in many instances. Consider a convention among military strategists to use certain shaped pebbles to stand for troops in their strategising concerning troop movements.

More important, it is by no means clear that the notion of a rule – as opposed to a certain realised rule content viz. that persons count something as something else – is actually doing any work here. I argued above that function imposition does not necessarily bring with it values and deontology. Now I suggest that function ‘imposition,’ whether collective or not, does not require rules, or even conventions. Consider the abovementioned example of military strategists collectively imposing functions on physical objects. If they used toy soldiers that looked like real soldiers then the arbitrary character of their decision would diminish. So by Searle’s lights there would not be a convention. But equally there might be no rule. It might be a one-off episode, never to be repeated. The point is that it is the notion of counting one thing as another thing that is doing the work; whether they count x as y in accordance with a convention or rule or by virtue of some other collective device is not important.

This latter point is really a special case of the more general point made earlier, namely, that collective acceptance collapses into actual or realised substantive functionality. For what is critical is that the relevant individuals perform the action – viz., count the pebble as a soldier, use the paper as a medium of exchange, and so on. What is critical is not constitutive rules, but rather substantive functionality. x


Searle, John. 1995. Construction of Social Reality . London: Penguin.

Miller, Seumas. 2001. Social Action: A Teleological Account . New York: Cambridge University Press.

Kroes, Peter. 2003. Screwdriver Philosophy; Searle’s analysis of technical functions. Techne 6, no.3 (Spring): 21-34.

Searle, John. 1997. Reply to Raimo Tuomela, in Responses to Critics of The Construction of Social Reality. Philosophy and Phenomenological Research 57, no.2: 449-458.

Tuomela, Raimo. 1997. Searle on Social Institutions. Philosophy and Phenomenological Research 57, no. 2: 435-442

i For an account of collective intention and related notions see Miller, Social Action Chapters 1 and 2.

ii An earlier version of much of the material on Searle on institutions – as opposed to artefacts - in the following sections appeared in Miller op.cit. Chapter 6.

iii Searle ascribes a certain priority to the institution of language - language is somehow constitutive of institutional reality. Searle Construction of Social Reality chapter 3. Searle account of institutions also involves, what he calls, background abilities. ibid. chapter 6.

iv Searle in a later paper (“Reply to Raimo Tuomela” Philosophy and Phenomenological Research 57, no 2 1997) claims that the point of the deontic power is to enable the performance of the function. But the function of money viz. exchange can be performed by shells (without deontic status), and the function of surgeons by non-accredited persons with surgery skills. That is, it is false that deontic powers are necessary for the performance of these functions.

v Searle oscillates between claiming that there is a collective imposition of substantive functionality and claiming (in effect) that there is a collective imposition of additional, ancillary deontic powers. I do not dispute the latter.

vi As Immanuel Kant pointed out, the right to life underpins all other rights, since it is a necessary condition for their enjoyment.

viii In Searle’s terms, “the Y content is imposed on the X element by collective acceptance.”

ix Raimo Tuomela makes a similar point in his “Searle on Social Institutions” Philosophy and Phenomenological Research 57 no 2, p2.

x Thanks to anonymous referees from Techné for criticisms of an earlier draft of this paper.