Are Routine Manual Systems Genuine Information Systems?

Information systems can be computerised or manual, with many manual information systems such as paper ledgers or Kalamazoo accounting systems being noted in the literature (Benson and Standing, 2002; Boddy et al., 2005; Land, 1987; Stair and Reynolds, 2003). However, examples of manual systems given in the literature are generally of systems where written data is stored on paper and there is a relatively simple translation to a computerised model (Checkland and Howell, 1998; Land, 1987). Descriptions of systems in the literature and also real world experience suggests, however, that there is a category of systems whose members support the performance of manual work but, because they do not always produce paper to file, or appear to store data in a conventional way, are not seen as information systems and do not fit well with traditional definitions of an information system. In these systems, tasks are completed manually through practiced routines and there is limited reliance on information stored in, and retrieved from, either fixed computerised databases or paper based files. These manual systems may include some paper artifacts (cards or tokens, rather than files) or have physical components often using tools such as whiteboards (Schmidt and Simone, 1996), cards or coloured magnets (Lederman et al., 2003; Mackay, 1999; Wong, 2000) as information resources. While there may be some written information on these artifacts, they do not function as traditional data stores. Instead, the meaning of these artifacts is contained, for example, in their colour or placement. Included among the socio-technical systems exhibiting these traits are some hospital scheduling systems, air traffic control systems and systems in small manufacturing concerns.

These manual systems support work that is relatively complex but largely repetitive and routine and many of the human activities in these systems appear to be performed in a largely non-cognitive way. They comprise forms of activity where participants use their tacit knowledge of work practices and various cues perceived directly from the work environment to routinely perform manual tasks. In these systems actors use cues embedded in the physical components of the system to deduce information on how to act to complete goals. For example, a magnet of a particular colour on a whiteboard might promote a particular response. While such systems have been discussed previously, the information systems literature does not consider to any significant degree how these manual artifacts provide signals or cues for behaviour, how system participants respond to these cues, or the role that manual artifacts play in providing information for routine (non-cognitive) action. Particularly, previous research does not focus on the extent to which these systems can be viewed as genuine information systems rather than simply as discrete tools or aides memoire. In this paper, such systems will be termed routine, manual systems.

This paper examines the legitimacy of regarding these routine, manual systems as information systems. Information in these systems is not defined in terms of our traditional understanding of it as processed data so conventional definitions of IS do not appear to suit these systems. However, this paper claims they can still rightfully be called information systems. Because these systems do not store or process data in ways commonly documented in the IS literature we seek a way of explaining how these systems conform to a definition of information systems that is broader than that commonly found in the IS literature.

The method we use is to firstly examine traditional definitions of the term ‘information system’. From these definitions we extract four general concepts that we use to form the basis of a new definition: fact, transformation, signal and action. Then we present four example systems. The first is a conventional information system and is presented as an illustration of the type of system commonly regarded as an information system. This provides a basis for understanding why it may be difficult to also perceive a routine, manual system as an IS because, prima facie, they appear to be very different. However, we then describe three routine, manual systems and consider how the elements in our new definition of an information system appear in all three of these systems. Then we reflect, using the literature and our own observations, on the answer to the question: Are routine, manual systems genuine information systems?

We find that the three routine, manual systems form a class with common characteristics that can conform to a broader definition of information systems than is currently commonly used. In answering the question of whether or not they are information systems we stand to gain an important insight into their relationship to traditional information systems as well how they provide a source of information to their users in informing action.