Looking at both types of systems, it is observed that the basic elements of both the traditional system and the routine manual system have the characteristics of facts. In both types of systems elements stand for something within the community of the system. In the traditional system there is a credit limit in a table of customers, in the manual routine system there is the manner in which cards are placed in an allocation box. Both elements stand for something as well as establishing a rule determining how the element is understood. While the elements in the traditional system are easily called data, in the non-traditional system using the term data, which has previously been more closely associated with numbers, words, symbols or events, creates expectations of something different in the situations encountered. Facts however, can be understood as observable objects or events (Checkland and Howell, 1998) or even ‘a situation that exists’ (Oxford English Dictionary). In this, the term fact can include the data found in traditional systems or the situations found in routine, manual systems.

Traditional information systems are said to contain data, which is then processed. In processing, data is altered by first selecting it and then in some way organising, manipulating or aggregating it. In a system like Causeway, data such as an invoice amount is selected in a deliberative manner from an invoice. The data is then inserted into the accounts receivable master file, which is updated. Thus, traditional numeric data is processed.

In routine manual systems, a course of action that is both similar and different to processing in traditional systems occurs. A situation in a system, such as the existence of a coloured magnet on a whiteboard, is perceived directly by a user and the user reacts by, for example, instigating a development such as moving the magnet to a different place on the whiteboard. At the end of this activity the board reflects a newly developed situation, not dissimilar to the updating of a master file, and further response can occur. While what happens to the whiteboard may not commonly be seen as processing, what is common to the two systems is the idea of transformation discussed earlier from the systems literature (Land, 1973), and also referred to in the semiotics literature (e.g. Liu, 2000). Transformation occurs when systems participants are faced with cues from their environment, which may be data or situations, and the participants then define and redefine what to do next, either processing data or developing a situation, altering the system each time to transform it to a state closer to the participants’ goal or objective. When a fact from either type of system is presented for manipulation, a transformation can occur. Thus, transformation is common to both types of systems.

As stated previously a conventional understanding of output includes the kind of elements, such as customer information, cheque totals, and credit listings, seen in the Causeway system. It does not traditionally include elements such as the way a ticket is placed on a table, the busyness of a room, or the emptiness of a slot on a whiteboard. However, earlier in this paper output was also described as something that signals or projects processed facts. Based on this description, semiotics provides a possible bridge across both types of systems. It is argued that elements within systems transfer messages (signs or signals) and that this communication results in action based on the transferred sign. A sign is defined as ‘anything that conveys information because it stands for something else within a community’ (Stamper, 2001). The role of a sign ‘is to establish a habit or general rule determining … the way the sign is to be “understood” on the occasions of its use’ (MacEachren, 1995).

What unifies the elements of both types of system is the appropriateness of the application to each of the term sign, or its further extension from the semiotics literature, signal. According to MacEachren (1995) ‘when a sign token mechanically or conventionally triggers some reaction on the part of the receiver, it is said to function as a signal’. Signals are seen as containing ‘pragmatic information’ (Stamper, 2001) that has the potential to change action and, in this sense, ‘signal’ is a more appropriate term for what is evident in a dynamic system than is ‘sign’.

Stamper (2001) suggests that actors learn to employ particular repertoires of behaviour and that particular signals come to stand for these repertoires. In the ICU Ward Management Case, for example, a coloured magnet of a particular kind might come to trigger a particular repertoire. In systems where concepts are perspectival, behaviours are triggered for individuals who process the information in a way that is significant to them. Similarly, the representation of situations that may be transitory provides a signal that assists in the navigation of a changing environment. A magnet indicating or signalling that a patient may be discharged creates a cue to initiate or prepare for action. Additionally, situations that no traditional interpretation of what constitutes an information system would consider to be output, such as the phenomenon of ATC controllers gathered round a table, can be seen as signals that can be transformed into information for action. These new situations that occur following the transformation of previous situations, signal action in the same way that a credit limit in a database provides a signal that can be acted upon.

In routine, manual systems whole situations with multiple aspects may be significant as signals. For example, the way a flight strip is placed is interpreted within the context of a flight strip holder, which is in turn interpreted within the context of a room full of flight controllers. In traditional systems, conversely, output is abstracted and de-contextualised and the importance of the broader context outside of a particular table or database is marginal. However, while the form and breadth of representation in both types of system are different, they are similar in providing a signal for action. Thus, the term signal is a term that encompasses both the output of traditional systems and the new situations, following previous situation development, found in routine, manual systems.

Finally, the response to, or feedback resulting from, the signals in both types of systems provides the fourth element of an information system: an impetus for action. In all systems the signals produced provide the cues for action necessary to keep the system functioning as a goal attaining entity. In Causeway, for example, processed data brings the database to a new state providing the feedback for the next activity. In the ICU ward, for example, a situation indicating possible availability of a patient for transfer is developed into a new situation, a definite availability evoking a response from a staff member. In the two types of systems what happens after the signal is apparent is different. In traditional systems, output generally provokes a cognitive response, based on the information the new state provides. That is, the original data is manipulated in some way and the user thinks about the result of processing (the output) before acting. In the routine, manual systems, new situations are often perceived much more directly (Lederman et al., 2003) in the sense that a movement of a coloured magnet on a whiteboard is detected and responded to in a way that is largely reactive and non-cognitive, and quite different from the deliberative reading of a new value in a table. However, whether the fourth stage is a traditional cognitively based or a reactive response, both eventualities correspond to action, the fourth common feature of the two types of system. Thus action is a unifying feature found in both types of systems following the initial three elements.

These ideas are expressed in Table 2 below, which provides the values missing in Table 1. Table 2 expresses what is common in both types of systems and unifies these common elements under a larger, universal heading. Where there is data input, such as a cheque amount in a traditional system, in a routine, manual system the input is a situation that occurs, such as a magnet placed on a board. Both of these elements provide facts to enter into the next stage in the system. In a traditional system the data is processed by, for example, adding or multiplying whereas in a routine manual system it is developed in some way that augments or diminishes its meaning such as moving the magnet across the board. In all cases a transformation occurs where the state of the system is now changed. In traditional systems this changed state is represented as output; in routine, manual systems it is represented as a new situation. However, in both cases what is evident is a signal for action by users. Finally, in the traditional system an item of data output such as a cheque total produces an action such as a confirmation of the total. Similarly in the routine, manual system the new situation cues a response such as a rearrangement of other magnets or a review of bed allocations. Both feedback and response are unified in the impetus for action that they provide.

Reviewing Table 2 our claim is that all of the systems discussed above exhibit the qualities of information systems. While the difference between the system elements are analytical in routine manual systems rather than clearly separated in space and time in the way that making a change in a database and producing a new report might be in a traditional system, a unified view of both types of systems can still be presented. It may be harder to identify an element such as an output in a routine, manual system than in a traditional system. However our claim is that such elements, although having a different form, have universal qualities that can be found and aligned across both types of systems. These universal elements include facts and transformation, which encompass the traditional concepts of information systems, data and processing, as well as the situations and situation development found in routine, manual systems. When transformation occurs a change in the system takes place creating data output or the evolution of a new situation. This elicits action and moves the system one step closer to goal fulfilment. Fundamentally, what makes a system an information system is the existence of four things: facts that can be transformed by users into signals which promote action. Thus, we argue that through these shared features all of these systems can rightfully take the same label: information system.