User:Ric Phillips/sandbox

For the article on Digital Identity...

The origins of digital identity
The practice of recording information about people to identify them predates the invention of computing technology. The problems those earlier practices answered are no different to the problems solved by digital identity. Digital identity is simply a means to enable computers to perform a common task that is not specific to information technology. To build computers that could interact effectively with external agents computer scientists used concepts and techniques that existed long before the invention of computers.

Identification as a behaviour
Our ability to interact with our environment depends on our ability to identify things in that environment. And we have different ways of identifying things. For example we can identify something by its kind. We may identify an object as a couch, a chair, or just something that can be sat on. We can also uniquely identify an object apart from its class. For example identifying that a chair is the exact same one one we bought because it is in our living room and has a particular scuff mark on the seat.

Identification then may be selection, classification, or a combination of the two.

We also identify people by selection or classification. For example when we identify a friend by recognising their unique features we are selecting them from the class of all other people. When we identify a soldier by recognising their uniform we are assigning them to the class of soldiers. And if we know a soldier personally, we select them from, and assign them to the class of soldiers at the same time.

Identification as a behaviour of computers is no different. A computer operates in an environment that includes its users and, when it is networked, other computers. To be useful a computer must be able to at times select an external agent from a class, assign an external agent to a class or do both. For example a computer running a banking application will need to decide whether a user is a customer or staff member - assignation to a class - and display the correct interface accordingly. The same computer will need to know which particular customer a user is - selection from a class - so that it can provide information to that user about his or her money.

The problem of identification
In everyday life we identify things with little effort or conscious reflection. We can identify the people and objects we interact with because we have learned and can remember who and what they are. The ease with which we carry out identification may give the impression it is a trivial, or simple function. However, the class of neurological disorders called agnosia demonstrates that identification is a complex and vitally important capability.

In social environments we get to know people over time through encounter, introduction and referral. We then maintain and apply that knowledge using inborn cognitive abilities. In small societies human memory and social relationships suffice to allow identification. --- an important aspect of identity is the role of identification and classification - for all practical purposes classification may substitute for recognition.

However, even in small scale societies, the problem of identification is often factored into do not degrade our ability to identify others. It is when societies grow to sizes where identification cannot rely on personal introductions and memory that techniques begin to appear that use recorded information as a means for identification.

The development of identifying information
The use of recorded information, and its relationship to the growth of civilisation, is what defines the difference between history and prehistory. Identification of persons in social relationships is one of the earliest uses of recorded information. in many of those cases records were not designed and kept solely for the purposes of identification. Rather, identification was either are prerequisite or byproduct of the record. However, those innovations from which modern identity records developed may be grouped into two classes: credentials and registries.