Laws of information systems

The Laws of information systems are a collection of observations, assertions and generalizations characterizing the behavior of people, hardware, software, and procedures enclosed in a certain scope (an information system). An information systems is an amalgam of scientific and humanistic disciplines including computer science, management science and social sciences. It is therefore characterized by laws as well as a number of principles. These span the range of transaction processing, effective systems, user interfaces and system development. These "Laws" should not be confused with Scientific or Physical Laws.
Laws
Transaction Volumes
The volume of transactions will increase with the stage of development of a society. The greater the development the greater the number of goods and services exchanged and the greater the number of transactions.
Symbol Systems
This is a corollary of the law of transaction volumes. The symbol systems by which messages are encoded grow more complex as a society evolves. Thus the 7 bit ASCII code gave way to the 8 bit version which is giving way to the 16 bit Unicode system. This trend is likely to continue when civilization expands beyond the confines of our planet and the volume of information exchanged becomes astronomical.
Technological Evolution
Technology seeks the most efficient form, unless otherwise constrained. Efficient form is defined qualitatively as one that is best adapted to its application or as one with the fewest problems. This is a variation on Darwin’s theory of evolution and is manifested in the case of memories, storage devices, databases etc.
Infinite Processing Needs
The information processing needs of an organization or society will always exceed its information processing capabilities. This can be seen as a converse of Moore’s law - doubling of processing power every two years. Whenever a new technology is introduced it is overwhelmed by new applications or increased usage. For instance, current technologies are not up to the task of processing images streaming from space probes in real time. This is also evidenced in the case of email, internet, search engines etc. that are being swamped by volume.
Good Systems
A good system produces benefits that are disproportionately high in comparison to the initial investment. Any complex system, including an information system is typically interconnected with other systems. So a good system has ripple effects which show up as unexpected benefits. The freeway system in the U.S. for example, led to the growth of the automotive, steel and motel industries. Another example is the Sabre system that has been designed for making reservations but has been used in crew scheduling and flight forecasting. As a corollary a bad system produces problems that are disproportionately high in comparison to its area of operation, a prime example here being the 64K memory limitation of DOS which for a long time stymied software developers.
Right Design
Every software that involves users has a “right” design. The“right design” refers to decomposition of functions into menus/controls. The fact that some types of software are intuitive while others are not leads to the belief that there is a “right design” for every software/IT application that is up to the designer to find.
Interconnected Systems
An interconnected system cannot be controlled unless each interconnect is individually controllable. A very simple example is provided here. In one email system, the “signature” text and message composition are wedded together so that the same signature is produced every time a message is composed. This is also known as “coupling.” It provides an option to select the default signature but doesn't allow this to be done dynamically. It should provide an option to select which signature is used, at the time of message composition.
Complex Interfaces
There cannot be a simple interface to a complex system (Here complexity is informally measured as number of menu options). This is a variation on the law of requisite variety which states that variety in a system should be at least as great as that found in its environment. Complex systems such as visual programming environments or CASE tools therefore cannot have simple interfaces.
Irregular Transactions
Information systems (transaction processing systems) which cannot process irregular transactions are doomed to fail. An irregular transaction is defined as one that deviates from the norm, in terms of items bought, conditions or constraints. Examples include registering for two courses that are scheduled to start at the same time or including a child safety seat in a car rental reservation.
Qualitative Decision Models
It is impossible to calculate outcomes with any certainty in a decision situation that involves qualitative variables. This is based on the theory of computability which states that a problem is computable if an algorithm exists, the algorithm is efficient/tractable and if there is a well defined solution state. Because of their very nature, qualitative problems lack well-defined state and hence the law.
Mental Models
The system’s model must not exceed the user’s mental model in complexity. The system’s model is the organization of features in the system whereas the user’s mental model is their conceptualization of the system. When the system’s model exceeds the user’s model, user will not be able to operate the system without extensive training and the result is often an implementation failure.
Accelerating Returns
The rate of progress of technology is accelerating to such an extent that it produces returns that are not linear, but exponential. Much like Moore’s law, every decade there is a doubling of progress and therefore technological advances will occur exponentially. Kurzweil expects that the next great evolutionary step of human kind will be the integration of human biological components with machines.
Brooks Law
Adding manpower to a late software project makes it late. Fred Brooks was the chief engineer overseeing the 360 project, which was one of the largest software projects ever undertaken. Based on his experience, Brooks came to the conclusion that putting additional programmers on a delayed project will not hasten its implementation because of the additional communications overhead. In fact, it had the tendency to reduce productivity.
Information Independence
Users should be able to access their information regardless of where it is physically located. This is a variation on the concept of distribution independence in databases that has been extended to include other types of information and other types of computing contexts.
Soft information principle
Information systems must incorporate soft information or they are doomed to fail. One way in which irregular transactions can be handled is to provide additional notes on the transaction. Those systems that do not accommodate such soft information may result in a transaction failure or may result in inconveniencing the user/consumer.
Reapportionment Principle
Tasks that can be performed by the system (in the context of software use) should be performed by it. Automatic filling of personal details from ss# or phone# in a customer registration form is one example. This is ultimately based on the simple economic principle of labor substitution to leverage productivity. It's a widely used principle nowadays.
Sharing User Information
All desktop systems must share information about the user. This is a corollary of the re-apportionment principle. To the extent that desktop systems require user information (such as email address, phone# etc.) it is advantageous for users to have the system obtain it from a common profile.
Information Responsibility Principle
Those who have information are obliged to share it with those who need it. This is a principle attributed to Peter Drucker in his 1988 HBR article, “The Coming of the Knowledge-Based Organization”. Since information is intangible, it is difficult for potential consumers of information to perceive its source and hence the principle. The principle implicitly assumes that reasons for excluding information specifically from a person do not exist.
Information Ownership
Owners of information must have access to it. This is a corollary of the information responsibility principle. When information changes, the owner has a stake in making the change in the system so it is reasonable to give them the access to do it. Many companies are web-enabling their systems, thereby illustrating this principle.

Main Menu

See Also

Main Menu