There has in recent times been an increased interest in the ontology of trust and money, and how to model those [1, 2, 11]. Still, we believe that there are a few puzzling existence and identity aspects of money that seem hard to reconcile. Money, it is said, is debt [8]. But how then can it be that one can burn a bank note and lose its worth forever? Do we burn the debt by setting fire to the note? Do we really lose money in the fire? Or can it be recreated somehow? The purpose of this discussion paper is to highlight some properties of money and ask ourselves what consequences those properties have for conceptual modelling. Amaral et al [2] identify three commonly used criteria that anything called money must fulfil. Money is: - “a medium of exchange: a means of payment with a value that everyone trusts; 'money is used as an intermediary in trade to avoid the inconveniences of a barter system, i.e. the need for a coincidence of wants between the two parties involved in the transaction', [5]. Borrowing an example from [13], the statement 'I bought this shirt for 20 euros' reports the use of money as a medium of exchange. - a unit of account: 'money acts as a standard numerical unit for the measurement of value and costs of goods, services, assets and liabilities', [5]. An example, mentioned in [13], that reports the use of money as a measure of value is the statement 'My car is worth 10,000 euros'. - a store of value: 'money can be saved and retrieved in the future', [5]. The statement 'I have 1,000 euros in my bank account', [13] reports the use of money as a store of value.”
In this paper, we discuss money in the context of a three level model having a physical level, a social level, and an informational level, see figure 1. The physical level contains, as the name suggests, all the objects that make up the physical world. This includes, for example, humans, magnetic areas on disks, or pieces of paper. The social level contains objects that are recognised, in a wide sense, as being parts of a society, for example, persons or organisations. The informational level is tucked in between the physical and the social levels, and it contains information about objects in both the other levels. The point of those levels is the assumption that the notion of money is somehow related to all of them. We should also note that there are no watertight barriers between the levels. For example, an information system exists arguably on all the three levels.
There are correspondences between the levels. What is a human on the
physical level corresponds to a person on the social level. And when discussing money,
corresponding to a person on the social level is an account on the informational
level. However, the correspondence relation between the levels is a somewhat
complicated issue. What we first of all want to express is that everything on the
social level or the informational level is directly or indirectly grounded in
something on the physical level, [
Correspondences between the physical level and the informational level is
indirect. The reason for this is that we view accounts as information items
signifying persons and not humans. The point of an account is to represent the
economic information about a person and make use of this information. So, the
correspondence relation between the those levels is signification. The economic
information is a sign of something on the social level. In the
Resource-EventAgent (REA) ontology terms, an account can be viewed as an economic agent,
[
There is however a complication in that items on the physical level may carry economic information that can be used by persons on the social level. For example, a euro bill is a physical item carrying information that expresses an economic value. From a modelling point of view this could mean that such an item should be positioned on the informational level or, more likely, the informational content is split from the physical item and placed in the informational level. We will not do that but conclude that the correspondence relation between the physical level and the social level can (besides constitutes) also be a signification relation just as it is between the informational level and the social level. A corollary is that items on both the informational level and on the physical level can be used by persons on the social level as economic information. 3
An obligation, in this model, is an item on the social level. It may come about
as a moral duty to fulfil a promise. When I promise someone to do something
for her I become obligated to fulfil that promise; to deliver on my promise. In
our view an obligation can be created, kept, and destroyed, but it cannot be
transferred. It cannot be separated from the person who has it. As such it may
well be modelled, in the UFO sense, as a mode of a person, [
Even if a social debt cannot be separated from a person, information about it can. One way of codifying and expressing social debt is as physical money. One may say that physical money (shells or bills, for example) is a very concrete part of a decentralised information system whose purpose is to track debt. A more efficient and more centralized system for this is the banking system. In a bank a person is represented in a book keeping system as an account and a person’s social debt is represented in the content of the account. Although banking is by no means the only way to represent info about someone’s economic status, we use those banking (or accounting) terms when discussing the informational level.
In figure 2 we sketch some possibilities regarding money. As mentioned, social
debt can be created from a promise made from one person to another. The debt is
a social item and it is inseparable from the person that holds it. Other plausible
ways to view social debt are that it might be seen as a relationship between
two persons or perhaps even between two persons and a society. As modelled
in figure 2, social debt should be understood as being a part of person B. As
such it cannot be considered to be money as it fails Amaral’s first criterion
(transferability) [
Information about social debt, though, can be seen as money. (see [
One may rightfully wonder about how a single bill on the physical level can keep track of a person’s social debt. What is the connection? First of all, the bill exists because there is debt in the world and physical bills can represent that debt, [8, p. 49]. However, the bill in B’s pocket didn’t necessarily come about as a representation of B’s particular debt but it came about as a representation of someone’s debt. B has it in his pocket because by convention it may be accepted as a form of payment in exchange for a good or a service. One of the main ideas behind the bill is just that the control of it can be transferred to someone else as a payment. That is, B relinquishes control of the bill in order to compensate for a debt he has towards A. In this way, it is reasonable to say that the bill selected for a payment is an expression of B’s debt towards A.
Note the ontological problems that arise when we can have info about social debt is coupled more or less hard to a carrier. The € and the N’s are info about person B’s social debt towards A and therefore both are forms of money. A difference between the forms is that the € (which can be thought of as a money bill) is a physical item. As such if it is destroyed then the info about the social debt is also gone. A question is whether one can conclude that social debt disappears when the evidence of it disappears. This can be contrasted with the N’s on the informational level. If account B’s information should be erased for some reason (e.g. in a disk crash), then this information can be restored from info in account A or from a backup tape.
Another ontological question is that according to the above we have that a social debt can be modelled as a UFO mode. We have hypothesised that information about the mode is money. On the information level we have said that the information can be copied and that there may exist multiple copies of it. On the physical level we say that physical items (e.g. bills) under someone’s control can be viewed as assets while the same items used in a payment is representing a liability (i.e., a debt). When modelling money this is challenging and shows that although money is information about a social debt, the way by which the info is handled affects how money can be modelled. 4
In everyday life, money is typically viewed as being anonymous, i.e., not having identity. As the saying goes, money does not smell. So, the question arises how this lack of identity can be modelled on various levels. On the physical level, there are objects, such as shells, notes and coins. For these, there exists some identity principle, as for any physical objects. Thus, these objects do have identity, which seems to contradict the perception of money as anonymous. However, the identity principle may be more or less easy to apply, depending on the type of physical object. For example, for most notes, there is a serial number printed on each one of them, which can be used in an identity principle, such as “two notes are the same if they have the same serial number”. However, for other monetary objects, such as coins, there is typically no serial number. In practice, it then becomes difficult to identify such monetary objects. This situation illustrates that identity and mechanisms for identification have a practical side, meaning that people sometimes introduce tailor-made properties to be used for supporting identification. How can such properties and mechanisms be represented in an ontology for money?
Another issue is whether it is possible to devise a form of money, say some
kind of electronic money, where identification is ruled out (without resorting
to some anonymizing procedure). Considering the case of bitcoin, each token
representing a bitcoin, or subdivision thereof, is actually identifiable through a
number. But would it be possible to design a scheme in which electronic money
could be transferred without keeping its identity? Possibly, such a scheme could
make use of the fact that electronic money is not specifically dependent on some
unique physical object. Instead, the information representing money can migrate
between physical objects, and even be encoded in multiple physical objects, as
is the case in, for example, a distributed ledger. In other words, the information
is a generical dependent continuant, in the terms of BFO [