To know for a fact

Many familiar social facts - like money and marriage and monarchy - are facts only by human agreement. They exist only because we believe them to exist. That is the thesis, at once startling yet obvious, that philosopher John Searle explores in The Construction of Social Reality.

Take, for example, the money in your pocket. It is not just your subjective opinion that those particular pieces of paper are money. It really is cash, with an out-in-the-world objective value, which you can use to pay for goods and services. But it is also only money because we all accept that it is money. There is no intrinsic physical property which makes it money: money only exists because human society collectively assigns this particular function to certain pieces of metal and paper and calls them coins and bank-notes. Searle sets himself the task of analysing the logical structure of socially determined things like money, which he terms "institutional facts".

His analysis depends upon three key notions. The first is "collective intentionality", by which he means simply the ability of conscious agents to take up a common attitude to something. Whenever this happens, say when a group of people watches the sun go down or a mother and baby play with a rattle, a "social fact" is created.

The second notion is "assignment of function". This occurs when a certain object is deemed to have a particular purpose. Consider a small plank of wood with four sticks of wood firmly fixed at right-angles to it. If the object is assigned the function of being sat on, it is a stool. If it is used for placing coffee cups, it is a coffee table.

In Searle's terminology, an undefined wooden object is a "brute fact" which exists independently of anyone's attitude towards it; but a table or chair is a "social fact" since it is created - qua table or chair - by collective intentionality, that is by its social acceptance as one or other piece of furniture. In such cases the identity of the object depends on its use, on the function it has been assigned. Searle therefore gives the name "functional fact" to all social facts of this type.

Most functional facts, although they acquire their functions by collective intentionality (eg we accept that this is a table), depend on their physical properties in order to function. So you could assign the function "table" to a flat piece of wood or iron or stone, but not to running water. There is however a special subclass of functional facts that are characterised by the fact that they do not depend on their physical properties in order to function. These are Searle's "institutional facts", such as money, which function by virtue of a new status imposed on them according to constitutive rules in the form "X counts as Y in C".

"Constitutive rules" are the third and most distinctive of Searle's three key notions. An institutional fact does not depend on any intrinsic physical quality of the phenomenon on which the function is assigned.

The effect of this assignment of function is that the phenomenon concerned acquires a new status, and this is what enables it to function effectively. This status can always be expressed in the following form: in a particular context (C), a particular phenomenon (X) will be treated - or counted - as having a certain function (Y). This formula, "X counts as Y in C", is the form taken by all "constitutive rules". Searle adds: "The 'counts as' locution is crucial in this formula because since the function in question cannot be performed solely in virtue of the physical features of the X element, it requires our agreement or acceptance that it be performed."

Thus, Clinton (X) counts as President (Y) in the United States (C) only because the people accept him as such. Notes issued by the Bank of England (X) count as legal tender (Y) in the United Kingdom (C) only so long as people have confidence in the currency. The necessary public recognition of the constitutive rule may be a spoken or written form of words, or a symbol, but it must be there.

There is a pointer here to a practical application of Searle's analysis. Some of the confusion in our social affairs can be traced back to an ambiguity - deliberate or otherwise - in the status of certain facts. Searle himself gives the example of Korea in the 1950s: The United States authorities at the time were anxious to avoid calling it "the Korean War", because it would have been unconstitutional to have embarked on an official war without going through certain procedures (which had not been observed). So it was called by the vaguer term "Korean conflict" and designated a "United Nations police action". That kept it constitutional. The fighting and killing were just the same, but the status of "conflict" was a way of evading the rights and duties attaching to the institutional fact "war" in both American and international law.

A contemporary example might be the status in Britain today of couples cohabiting without being legally married. There is a growing feeling that the rights and responsibilities in such cases should be the same as for married couples.

This can be seen in Searle's terms as a developing collective acceptance of cohabitation as an institutional fact. The difficulties arise from the lack of a clear constitutive rule to define the context (C) in which the phenomenon of two people living together (X) "counts as" a couple (Y) with the same rights and duties as a married couple. The complicating factor in this particular example is that one of the reasons people live together rather than get married is precisely to avoid the "institutional" aspect of marriage.

Searle's analysis suggests that there are intrinsic reasons why it is not possible simultaneously to enjoy the benefits and avoid the obligations of marriage simply by changing the name and omitting the ceremony.

In the second half of his book, Searle tackles a number of issues related to his main thesis, in particular the nature of reality.

Although he is arguing that far more facts than we generally realise are dependent on human acceptance, he is adamant in defending realism, ie "the view that there is a way that things are that is logically independent of all human representations."

This is a very carefully worded definition of a notoriously slippery word. Realism, he says, is not a theory of truth or of knowledge. All our knowledge could be mistaken and all our alleged truths false and still realism would survive. In a typically provocative assertion he declares that even if it should turn out that Mt Everest never existed, "realism remains untouched".

Searle is keen to affirm realism because he is very sensitive to the charge that his theory of institutional facts "seems to produce either a vicious infinite regress or a vicious circle." If, as the antirealists maintain, all facts - not just institutional ones - are observer-dependent, then Searle sees no way out of this circularity and regress.

However, it is not clear to me that when realism is defined in such a rarefied way, shorn of all associations with knowledge and truth, there is enough substance left to fulfil the task Searle demands of it. The weakness in his normally water-tight argument emerges in his use of the term "brute fact". He has a schematic hierarchy of facts, in which his first major division distinguishes "nonmental brute physical facts" from "mental facts". Yet when explaining why his theory demands the existence of brute facts, he says that these might manifest themselves as thoughts in people's heads.

Now on any definition, thoughts are surely mental facts? Again, as an example of a brute fact he gives "There is snow on Mt Everest". Yet, as we have just seen, he also says that realism would survive even if Mt Everest did not exist! So realism, as he defines it, is no guarantee of the kind of brute fact which he describes in setting out his theory.

Searle's discussion of realism and brute facts reminds me of the way many theologians talk about God. All the obvious objections to the existence of God are side-stepped by defining God in some highly abstract way such as "that which is prior to all that is"; but as soon as they use the term in relation to everyday life God is given much more concrete attributes. In precisely the same way, Searle avoids all the usual objections to realism by insisting that it "is not a theory of truth, it is not a theory of knowledge, it is not a theory of language," but actually uses it as though it were.

It is one of Searle's greatest gifts that he sets out his arguments with such clarity that the possible weaknesses are easily spotted. His challenge to the rest of us is to probe those weak spots with as much honesty and rigour as he has shown in developing his persuasive thesis. Only thus will our understanding of our social world be advanced.

Anthony Freeman is managing editor, Journal of Consciousness Studies, and a priest in the Church of England.