Decision structures are clusters of people supported by technology who carry out roles involving the collection, processing, storing, disseminating and ultimately using of information. Such information flows by means of observing, reading, talking and listening. The information is used to suggest alternative courses of action and to choose from among the alternatives presented. In addition to carrying out these activities, decision structures are also bound together by values, beliefs and commitments, all of which work together to form the context in which the information is handled and decisions are made. Almost everything of importance that happens in our lives -birth, schooling, marriage, working, commerce, leisure, and death - is mediated through one or more of our society's myriad of decision structures. Some decision structures are commonplace, such as the one allowing a customer to purchase lunch by means of a credit card. Others are momentous, such as the one providing an ailing patient a blood transfusion in the midst of a heart by-pass operation. An essential feature of every decision structure, however, is that in order to make the decisions expected of it, and to take the action, it must receive and process personal information, sometimes a great deal of it.

There are an enormous number of possible decision structures in a society. A little mathematical logic can be used to establish a quantitative limit. Suppose a person - i.e. a first party - lives in a society comprised of N other people. In effect this person faces a lineup of N other people. With respect to a given piece of information about himself, he can mark a "1," meaning yes, if he wants one of the N individuals to have his or her personal information (this makes these individuals members of the second party); or, he can mark a "0," meaning no, if he does not want that individual to have the information (thereby making them members of the third party). Using this mechanism there are 2' possibilities - in most societies, a very large number. This reflects all of the combinations of 1's and 0's that can result.

One of these combinations is a string of 0's indicating that this is very private, contemplative information -tacit, hermit like or Freudian - about the person that he or she does not want to share with anyone. This is the kind of information, for example, protected by Roe v. Wade. It is exclusively first party information: Westin's Solitude. Releasing it to others would result in shame, embarrassment, social sanction or some other undesirable outcome. All other combinations represent possible decision structures to which the first party's personal information is shared with and entrusted to other party.

An especially important such structure is the one represented by a string of all 1's. This structure represents society as a whole. In the language developed above this is the fourth party. The information about an individual it receives is completely open and public. It is information the first party intends to share, or must share, with any one and every one of the N other members of the society.

These two possibilities - all 1's and all 0's - are extremes. Nevertheless, they represent familiar and understandable situations. In the variegated culture in which we live, however, the greatest challenge to privacy lies not at these extremes but in the middle, that is, in the lattice containing the remaining 2^N minus 2 possible decision structures. In these structures some personal information is shared with some people and not with others. Consequently, a dividing line must be drawn between the second party and the third party. Ideally die dividing line becomes a sort of "firewall" which keeps third party intruders out and personal information entrusted to the second party in.

It is intratractable and ultimately absurd to systematically examine all of the 2^N combinations. Nevertheless it is instructive to explore some of the various possibilities. Consider one-to-one relationships. In this hypothetical society a person may engage in up to N of them. Among these N other individuals there may be, for example, one or more very special second parties - a spouse, a lover, or a confidant - with whom a person wishes to have a friendly, loving and trustful relationship. Their intimate relationship is also a form of decision structure, albeit often a rather informal one. Usually it is a mutual relationship in which the both parties wish to share certain personal information with each other on a systemic basis, but with no one else. It would be an intrusion or invasion of their privacy if any one or more of the other N - 1 people, say by means of a hidden microphone, somehow acquired the information they shared between themselves.

Privacy is often discussed in these terms. Alan Westin, for example, describes intimacy in the following way: "the individual is acting as part of a small unit that claims and is allowed to exercise corporate seclusion so that it may achieve a close, relaxed, and frank relationship between two or more individuals." (p. 31) This point-of-view also underlies philosopher Charles Fried's observation that privacy is a "necessary context for [forming] relationships which we would hardly be human if we had to do without - the relationships of love, friendship, and trust." (Schoeman, 1984, p. 206) Drawing on this understanding Fried goes on to define privacy as "that aspect of social order by which persons control access to information about themselves." (p. 206)

Conditions of intimacy do not always involve the symmetric exchange of personal information. Some very important relationships are asymmetrical. The priest/penitent, therapist/patient, reporter/source, doctor/patient, and lawyer/client dyads are cases in point. Personal information is shared in these relationships mostly one way between two people. Relationships involving asymmetric information exchange have two significant characteristics. First, the recipient of personal information gains a power advantage over the giver - information is power. Second, the giver - penitent, patient, or client, however could not receive the services desired if he or she did not reveal a great deal of sensitive information about themselves to the service provider. For these reasons, many societies accord decision structures that are based on the asymmetric exchange of personal information conditions of privileges. This means that personal information received by this type of decision structure should not be disclosed to others except in rare and exceptional circumstances.

Dyads, however, are not the only form of social groupings. Institutions, organizations, associations, groups of all kinds - combinations extracted from the 2^N - N other possibilities - form the remaining fabric of decision structures which may handle personal information. This vast array of multiple person grouping are also of two types: helping and controlling.