The ‘language’ of informatics: The nature of information systems

International Journal of Information Management 29 (2009) 92–103

Contents lists available at ScienceDirect

International Journal of Information Management journal homepage: www.elsevier.com/locate/ijinfomgt

The ‘language’ of informatics: The nature of information systems Paul Beynon-Davies Cardiff Business School, Cardiff University, Colum Road, Cardiff CF10 3EU, United Kingdom

a r t i c l e

i n f o

Keywords: Informatics Information System Universals Second World War

a b s t r a c t This is the second paper in a series examining the fundamental nature of informatics. The aim of the current paper is to provide a more detailed account of the concept of an information system based upon an earlier paper entitled Informatics and the Inca. The paper also builds upon the content of the first paper in this series entitled Neolithic Informatics: The Nature of Information. We ground the discussion in a significant case from the Second World War: that of the Warning Network. The Warning Network was a system that contributed to victory of the Royal Air Force in the Battle of Britain. Through examination of this case we establish the idea of an information system as a semi-formal ‘language’ necessary for the coordination and control of activity in various forms of human organization. © 2008 Elsevier Ltd. All rights reserved.

One may draw the conclusion that the decisive factor in this war is not so much the weight of the material used, as a High Command who knows how to use it best. (Report on the Battle of Britain from the German Air Historical branch in 1940, cited in Puri (2006)) 1. Introduction We have used the term informatics previously as a convenient umbrella term to stand for the overlapping disciplinary areas of Information Systems, Information Management and Information Technology. This is in opposition to the tendency to re-brand Computer Science as Informatics and hence constrain the area of interest solely to that of ICT and particularly its incarnation in modern digital computing and data communications. In a previous paper (Beynon-Davies, 2007) we argued that the locus of the discipline of informatics is better placed with the concept of an information system. However, only a brief conceptualisation of the concept of an information system was provided in this early paper. The current paper is the second of a series of papers, all of which have the aim of developing a clearer and more sophisticated systematics for the area following Kling and Allen (1996) we refer to as organizational informatics. Organizational informatics is interested in the place of information, information systems and information technology in various forms of human organization (Beynon-Davies, 2002). Systematics is that branch of enquiry devoted to taxonomy: the process of describing, defining, identifying, classifying, and naming of things. Gregor (2006) refers to

E-mail address: [email protected]. 0268-4012/$ – see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijinfomgt.2008.11.002

taxonomic or analytic theory as one of her five types of theory in Information Systems. Taxonomic theory ‘goes beyond basic description in analyzing or summarizing salient attributes of phenomena and relationships among phenomena’. As such, taxonomic theory can act as a foundation for further work seeking to provide causal explanations of phenomena, testable propositions and predictive statements. There have been a limited number of attempts at systematics within the discipline of Information Systems and Informatics more generally (Alter, 2005). However, most such attempts take as their starting point the development of terminology from the base-line of modern information and communication technology (ICT). In previous papers (Beynon-Davies, 2007; Beynon-Davies, 2009a) we have taken a different direction, which we continue with in the current paper. Our overall aim is to identify a number of universal characteristics of the core elements of Informatics: information, information systems and information technology. We work from the premise that these key elements are a natural consequence of the need for humans to communicate and coordinate activity. Hence, we would expect that information representation, information systems and even information technology exist across time, space and human cultures. To use the metaphor of generative grammar (Chomsky, 1998), the surface forms of information, information systems and information technology are likely to be generated by some deep structures or common universal characteristics, heavily associated with the cognitive and social makeup of Homo sapiens. Our method of exploration is therefore to utilise cases from the historical, anthropological and palaeontological evidence to seek to determine the essence of information, information systems and information technology. In terms of mining the historical record, our aim has much in common with that of Mason, Mckenney, and Copeland (1997) who

P. Beynon-Davies / International Journal of Information Management 29 (2009) 92–103

argue for the importance of historical studies within the overall methodology of a discipline such as Information Systems. However our aims in interpreting historical material are much broader than those proposed by Mason et al. From our position, historical cases are useful for a number of reasons. First, they act as evidence of the universality of information and information systems across time, space and culture. Second, they demonstrate the ingenuity of humans in using many different forms of artefact for recordkeeping, data processing and communication (Crowley & Heyer, 2003). Hence, we argue for the universal nature of information technology (or ICT) in human society. Third, such cases provide an intellectual distance from considerations of modern ICT in definitions of information systems and information. They allow us to begin to determine something of the essence or universal nature of the component elements of informatics. The first paper in the series entitled Neolithic Informatics: The Nature of Information, made the case for considering information in terms of the idea of signs and sign-systems. For this purpose we adapted and tried to make more accessible, theoretical elements from semiotics in general and organizational semiotics (Stamper, 2001) in particular. The current paper builds upon this earlier discussion and introduces and adapts theoretical elements from what has become known as the language-action approach or perspective (Winograd & Flores, 1986). As DeMoor (2002) indicates the language action approach and organizational semiotics have clear synergies. However, they are rarely combined in attempts to define the nature of information systems. To help ground the discussion we move the historical period considered forward substantially from that covered in BeynonDavies (2009a) and Beynon-Davies (2007) to that of the Second World War. This means that we consider the concept of information system within a familiar and comparatively recent Western context. Grace Hooper, the creator of the business programming language COBOL, once said that ‘Life was simple before World War II. After that, we had systems.’ The Second World War was noteworthy not only for the scale of this conflict; it was also an incubator for a whole range of innovations in ‘systems’ that have affected human societies across the World up to the present day. An examination of a case from this period is therefore useful in making sense of organizational informatics as a systemic subject. We would argue that the key interest informatics has with the relationship between data, information, decision-making and action is a reflection of the necessary inter-relationship between three forms of system which are evident across forms of human organization in numerous societies: activity systems, information systems and ICT systems. We consider the relationship between these three types of system in the historical setting of the Warning Network of the Royal Air Force (RAF). This allows us to demonstrate the ways in which these systems successfully interacted and played a role in Allied victory during the Second World War. During the summers of the late 1930s the RAF’s Fighter Command created a socio-technical system that has been claimed to have contributed to victory in a decisive battle against the German Luftwaffe – the Battle of Britain – in 1940. Since control of the skies was an essential pre-condition of a successful sea-borne invasion, this victory contributed in part to the decision of German High Command to abandon the planned invasion of Britain and turn its attention eastwards to the Soviet Union. This, in turn, created space for the invasion of continental Europe by the Allies in 1944. The reason for discussing this historical case is because the remoteness of time enables us to consider the issue of what constitutes an information system in a fresh light. In particular, a consideration of this example helps us to understand some of the important differences between information systems and ICT systems, as well as their role within activity systems.

93

2. The Warning Network In this section we discuss the case of the Early Warning Network of RAF Fighter Command. This case has been described and analysed before in a number of different ways in extant and recent literature. For instance, it has been portrayed as an exemplar of good operations management (Kirby & Capey, 1997), good strategy-making (Grattan, 2005) and good information systems design (Holwell & Checkland, 1998b). Our interpretation of the case bears most similarity with that of Holwell and Checkland (1998a, 1998b) who portray the Warning Network as a successful example of a precomputerised information system. However, we differ from other sources in two fundamental ways. First, we have attempted to include more information on the actual operation of the Warning Network as well as details of its context than available in other sources. Second, we interpret the case not solely as an information system, but as demonstrating the important interaction of people, activities and technologies in systems of various forms, working within an environmental context. This method of presentation and analysis allows us to factor out the essential features of an information system from technological systems on the one hand and social systems on the other. It also enables us to provide a more balanced assessment of the ‘value’ of such systems, particularly the key information system, to the military victory of the time (Robinson & Wilson, 2002). Air Marshall Huw Dowding established RAF Fighter Command in 1935. At that time he was extremely aware of the limitations of the British Air ministry in meeting the minimum target levels set for fighter aircraft production. Dowding therefore looked for other ways of providing an advantage to his fighter aircraft in an air battle. He therefore set up an organizational structure within Fighter Command which would enable it to sense enemy aircraft quickly and respond and intercept such aircraft within minutes. During the early 1930s accepted military strategy for air defence was to fly so-called ‘standing patrols’ on flight paths likely to intercept bombing raids by an enemy. This constituted an extremely expensive military strategy in that fighter aircraft had to be kept permanently in the skies. Not surprisingly, this strategy was eventually replaced with the use of interceptor flights that could take-off quickly and attack incoming bomber raids. However, the key question remained, how was an air force to determine the precise position of incoming enemy aircraft in sufficient time to enable effective interception? The key solution to this problem involved the utilisation of radio technology to detect aircraft—a technology that became known as radar. This technology developed out of an observation in 1934 that an aircraft flying through radio beams across Kiel harbour reflected them to produce an image of a battleship. This insight stimulated research and development both in Germany and Britain to attempt to exploit this technology for military purposes. By 1935, both the Germans and British had access to this technology and indeed German radar was technically superior to its British equivalent at the time. The crucial difference was that the British were better able to utilise the technology within systems of air defence. The British were able to gain what would become known in modern management jargon as competitive advantage from this technology (Porter & Millar, 1985). In June of 1940 Air Marshall Huw Dowding’s Fighter Command faced a number of major challenges. The RAF had lost 500 operational fighter aircraft and 300 pilots in the air battle over Flanders and France. Having 620 operational fighters remaining, Fighter Command was therefore 50% below the 1200 target established in 1939 as that needed to win an air battle with the Luftwaffe over the UK. Dowding therefore had to find ways of utilising his limited resources to maximum effect.

94

P. Beynon-Davies / International Journal of Information Management 29 (2009) 92–103

The first step in the process was the establishment of an effective command and control structure. A headquarters for Fighter Command was established at Bentley Priory in Stanmore just north of London. This had overall strategic control of operations. The organization of Fighter Command was divided into four geographical groups covering major parts of the country, each group being controlled by a Group station. Number 10 group covered South West England and South Wales with its HQ at Box in Wiltshire, number 11 group covered London and South East England and was based at Uxbridge in West London, number 12 group covered the Midlands and was centred on Watnall in Nottinghamshire, while number 13 group covered the remainder of England to the North and was based at Kenton Bar in Newcastle. Each group was in turn divided into a number of sectors with a sector HQ at each of the airfields. For instance, group 11 was divided into seven sectors. Group HQs had tactical control within their area and Sector HQs had control of pilots when airborne. The second step in this process of achieving advantage was RAF Fighter Command constructing an effective system of ‘information technology’ that could sense the whereabouts of aircraft. This involved the establishment of 50 radar stations and 1000 observation posts strategically placed around the coastline. There were actually two chains of radar station, one for detecting highflying aircraft and one for detecting low-flying aircraft. These radar stations could detect aircraft flying at an altitude of as much as 30,000 feet and 150 miles distant. This was supplemented with a chain of posts manned by civilian volunteers observing incoming aircraft, known as the Observer Corps. Radar stations, observer posts and fighter airfields were all connected to Fighter Command headquarters by dedicated Post Office tele-printer (tele-writer) and telephone lines. Not surprisingly, in traditional accounts of this conflict, radar is portrayed as the ‘killer application’ which drove the strategy of air defence by the RAF (Grattan, 2005). This interpretation tends to under-emphasise the place of other important information technologies of the day. For instance, the data generated by radar such as height readings were often approximate. There was also a time lag of some 4 min between interpreting radar data and scrambling a fighter squadron and this compared to the six or so minutes it took for enemy aircraft to cross the English Channel. Because of this radar data was also supplemented and corrected by other channels. A chain of wireless listening stations known as Station Y and their control centre at Station X provided a significant channel of data for this purpose. These agencies, for instance, intercepted radio messages from Luftwaffe crews and this communication traffic frequently revealed the destination, size and timing of Luftwaffe raids. Another key advantage of this channel was that data was also typically passed on to RAF Fighter Command within 1 minute of being heard (Puri, 2006). The third crucial step was the creation of an effective system in which the ‘information technology’ could be utilised. During the summers between 1936 and 1939, a series of teams formed from physicists, engineers and RAF personnel engaged in a series of practical exercises with the aim of solving the fundamental problem of turning raw data received from radar, observer posts and other channels into information for pilots to fly to the precise point at which to intercept enemy raids. The teams were unable to be told of the development in RADAR at this time for security reasons. Hence, they were given the brief of designing an activity system and associated information system that could utilise data on the bearing, distance and altitude of enemy planes coming to them at regular intervals from a mysterious source. Their overall objective was to decide on the best way of turning such data into information in order that courses of interception could be established for fighter aircraft.

These ‘experiments’ established a quick and effective means of bringing two groups of aircraft travelling at different speeds together at the same point in the sky using something that became known as the ‘Tizzy angle’. They also established the need for a filter room with its own ‘map table’ to ensure the accuracy of information passed to the operations room (see below). The eventual system of air defence that was created was given a series of different names such as warning and control system, early warning network or the control and reporting system. We shall refer to it as the early warning network or Warning Network for short (Holwell & Checkland, 1998b). Much literature portrays the Warning Network as allowing an initially under-strength RAF to successfully compete with a numerically greater force of enemy aircraft. While not denying the contribution of the Warning Network to RAF operations one should be careful not to over-play its importance. For instance, Puri (2006) makes the point that the calculation of relative strength should discount bombers, dive-bombers and twin-engine fighters, all of which lacked the capability of speedy and rapid manoeuvre required of the dog-fighting characteristic of the Battle of Britain. If these aircraft are excluded, the RAF’s 700 or so fighter force faced an opposition of 800 Luftwaffe single-engine fighters at this time. Nevertheless, an effective system of sensing, interpretation and response was undoubtedly important to Dowding’s strategy. Data from the two chains of radar stations were telephoned to Fighter Command HQ. This data went first to a filter room, manned 24 hour a day, where members of the Women’s Auxiliary Air Force (WAAF) turned such data into useful information. Because of the possibility of human error in the use of RADAR detection equipment the quality of the data was first assessed and if necessary corrections made to it. Radar plots and sightings were then turned into intelligence on the likely strength, position, height, speed and direction of both enemy aircraft and friendly aircraft. This filtered information was then passed on next door to the Fighter Command Operations Room (see Fig. 1). Within the operations room filtered information was coded onto a visual display by other members of the WAAF, who we shall refer to as plotters. The display itself consisted of using tokens and a large plotting table to model the disposition of aircraft, both enemy and friendly, in the sky. The plotting table consisted of a large irregularly shaped table on which was painted a large-scale map of the parts of the UK controlled by the particular operations room. A small wooden block or token was used to represent a group of friendly and enemy aircraft, such as a RAF squadron. Such tokens were moved around on the plotting table by plotters using croupier sticks. The placement of tokens on the plotting table clearly indicated the last known position of groups of aircraft in the sky. The movement of tokens across the plotting table indicated the direction of the aircraft. These tokens were also used to indicate the height and strength of aircraft units. RAF units were represented by triangular or wedge-shaped wooden tokens. Numerals slipped into tracks on the sides of such tokens revealed both the units’ altitude in thousands of feet and the number of ‘angels’, or friendly fighters, in the formation. In contrast, disc-shaped tokens were used to represented German aircraft units. Each of these particular tokens showed a code number and the attack size of the unit, such as ‘30+’ or ‘40+’. It was also possible to code the time a particular aircraft group had been in some area of air-space using such tokens. A specially built clock known as the plotting clock acted as a way of time-stamping the information on the plotting table within the operations room. Three colours – yellow, red and blue – were used on the clock-face to code 5 min intervals. On first identification a token was given the colour associated with the current 5 min interval, e.g., red. This provided instant recognition of the recency of

P. Beynon-Davies / International Journal of Information Management 29 (2009) 92–103

95

Fig. 1. Elements of operations rooms within the Warning Network.

data and by inference how long the aircraft group had been within airspace. In addition to the plotting table and tokens, a display called the ‘tote’ was used to record the state of enemy raids and the state of readiness of particular RAF squadrons. For instance, particular states of readiness for a squadron included available 30 minutes, available five minutes, take-off or cock-pit readiness—2 minutes or in the air. The actual state for a particular squadron was indicated by illuminating the particular area on the tote. The colour used for the illumination matched the time interval on the plotting clock in which the communication was received. A glowing white bar signified a communicated state more than ten minutes old. This rendered it unreliable. One or more operations controllers were also present in the operations room, usually seated in an observation gallery above the plotting table and opposite the tote and plotting clock. They made decisions on the basis of the ‘real-time’ updates being made both to the plotting table and the tote. Such decisions initiated the telephoning of instructions through to appropriate controllers at group and sector level, as well as other bodies such as Anti-Aircraft command, the Observer Corps, the BBC and civil defence organizations such as those sounding air-raid warnings. Hence, instructions could consist of commands for air-raid warnings to be sounded in threatened areas, fighter squadrons to be scrambled, and commands for air-sea rescue to pick up pilots ditched in the English Channel. The operations room at group level worked in the same way except that the maps used represented the group air-space. Group HQs also received data (aircraft sightings) from Observer Posts. This was first filtered at group level and then passed on to command and sector HQs.

The sector operations rooms were set up as two units. The first unit duplicated the picture at command and group level by copying the positioning of counters and updates to the HQ tote and plotting table at sector level. This allowed operations controllers at sector level to continually sense their sector’s place in the larger operational picture. The second unit plotted at sector level on the map the exact location of their own planes from their radio transmissions. From here aircraft were assigned to a particular raid and their interception courses were continually plotted using compass, ruler, pencil and paper. The sector operations controller scrambled selected aircraft on command from the group HQ. Once in the air, command passed over to the flight leader until combat was over. A diagrammatic representation of the Warning Network is illustrated in Fig. 2. This diagram identifies key information flows (labelled arrows), information-handling activities (labelled boxes) and information stores (represented by open boxes). Work in the filter and operations rooms is frequently portrayed within movies as sedate activity. In practice, activities were hectic. Numerous incoming messages had to be sorted, prioritised, coded and disseminated at a breathtaking pace. Late information could send a valuable squadron of hurricanes or spitfires looking for hostile targets on bearings and at altitudes which had been vacated by the enemy. The Warning Network described above contributed to successful action on the part of the RAF during the period from July to the end of October 1940. On Sunday 15th September 1940 the system was severely tested (Holwell & Checkland, 1998a). A hundred German bombers crossed the Kent coast at 11:30 that day. Seventeen squadrons from three groups of the RAF went to intercept them. At 14:00 the same day a second wave came in and was met by 31

96

P. Beynon-Davies / International Journal of Information Management 29 (2009) 92–103

Fig. 2. The flow of information through the Warning Network.

squadrons (over 300 planes in all). At the end of the day RAF losses were 27 aircraft with 13 pilots killed. The Luftwaffe lost 57 aircraft. The article by Holwell and Checkland has the unfortunate title, An Information System Won the War. Robinson and Wilson (2002) argue that two linked claims are presented in this paper. First, that the Warning Network enabled Allied victory in the Battle of Britain. Second, that this victory made possible the winning of World War II. Rightly they express concerns over this interpretation and argue in opposition that the linkages between the existence of the Warning Network, eventual victory of the Battle of Britain and successful invasion of continental Europe are subject to much historical debate. For instance, success in the Battle of Britain was due to interaction of numerous factors, many of which were environmental or contextual in the sense that they occurred outside of the sociotechnical system described in this section. For instance, significance has been attributed to the poor strategic decision-making by the Luftwaffe during this period as compared to an effective strategy produced and executed by the RAF (Grattan, 2005). During early

September of 1940 the Luftwaffe decided to switch from bombing military targets such as airfields to attacks on London. This, it is argued, provided Fighter Command with a much-needed breathing space in which to restore its strength. On the other hand, Dowding made the strategic decision to allow group 11 to bear the brunt of much of the fighting during the Battle of Britain. Other groups bolstered the defence when needed but also acted as a safe haven for squadrons in need of recuperation. The result of this strategy of rotation was a defence that never committed more than was necessary for the job at hand. However, this is not to deny the contribution of the Warning Network to eventual success both in the Battle of Britain and as a platform for further Allied success in the war in Europe. As Robinson and Wilson cogently put it, the Warning Network ‘played a necessary but not a sufficient role in the Battle of Britain’ (Robinson & Wilson, 2002). In a sense this interpretation is reminiscent of continuing arguments over the value of information technology and information systems to organizations (Strassman, 1990). Both academics and practitioners in the informatics area have a

P. Beynon-Davies / International Journal of Information Management 29 (2009) 92–103

natural tendency to over-inflate the contribution of technology and information to organizational success. In contrast academics and practitioners within the business and management area frequently under-estimate the role of technology and information to successful management. More generally, good information systems are a necessary but not a sufficient condition for organizational success. This is because the nature of information systems involves the complex inter-twining of information, decision and action in continual interaction within some environment.

3. Information systems In the previous paper in the series we argued that information can be conceived of in terms of signs and sign-systems. There is a close synergy between this perspective and what has been referred to as the language-action approach (Weigand, 2006; Winograd & Flores, 1986). In this section we utilise aspects of the language-action approach and combine this with our conception of sign-systems from the previous paper. This merging and adaptation of two bodies of literature helps us define some of the essence of what an information system constitutes. However, it should be noted that we diverge from and expand upon both perspectives in a number of important ways which we discuss below. Stamper (2001) defines a sign as being ‘anything that ‘conveys’ information because it stands for something else within a community of users’. Signs mediate between the physical and the social world. The discipline of semiotics is devoted to the study of signs and consists of four sub-areas – pragmatics, semantics, syntactics and empirics – that represent various facets of the concept of a sign which span between the social and the technical (Beynon-Davies, 2009b). Pragmatics is concerned with the purpose of communication. Semantics is concerned with the meaning of a message conveyed in a communicative act. Syntactics is concerned with the structure of the signs used to represent a message. Empirics is concerned with the physical form used to carry or store a message. Within this paper we establish some of the universal features of an information system which builds upon our previous established conception of information. A communicative act involves a person in formulating an intention and in communicating that intention to another person in the form of a sign (Beynon-Davies, 2009a,b). The syntax and semantics of a sign fundamentally involves the way in which it is represented as a symbol or set of symbols and the ‘stands for’ relation between the symbol, its referent and concept. This ‘meaning triangle’ (Ogden & Richards, 1923) should not be interpreted as establishing that signs have an inherent meaning. A sign can mean whatever a particular community of users chooses it to mean. The emphasis here is on community of users. Humpty Dumpty in Lewis Carroll’s Through the Looking Glass says that a word ‘means just what I choose it to mean—neither more nor less’. In reality, of course, individuals cannot define their meanings and intentions in isolation from other individuals. The very act of communication implies a mutual understanding and agreement to a common ‘protocol’ consisting of associations between signs, meanings and intentions within the members of some social group. However, the association between sign and meaning is not an immutable one. The same sign may mean different things in different social contexts. As interpreters of signs humans are extremely proficient at assigning the correct interpretation for a sign in a particular context. For instance the same gesture (forming an ‘o’ with the first finger and thumb of the hand and holding it up with palm facing outwards) can be used to communicate different intentions, depending on the cultural context. In the USA it means a-ok. In France it means zero. In Japan it means money. In Tunisia it means I’ll kill you! Likewise the same intention can be communicated using

97

a number of different signs. For instance, the intention ‘I love you’ can be communicated with a verbal message, a blown kiss, or a written message containing a graphic representing a heart—‘♥’. As the previous examples demonstrate, signs are used in various ways in human life. We would argue that in terms of organizational informatics we are particularly interested in the way in which signs are used to make decisions and generate action. This is clearly the realm of the social world. However, informatics is also interested in the relationship between signs and the physical world—particularly in the ways in which signs are embodied in ‘technology’. The current paper concentrates on the linkage between signs and the social world. The next paper in the series considers the linkage between signs and the technological world. Morris (1964) created a sophisticated model of signs founded in the notion of action. More recently John Searle has built on Morris’ work and has created a detailed theory of so-called speech acts (Searle, 1970). For Morris, semiosis, the process of using signs, consists of five elements. The first element is the sign itself which sets up in the interpreter the disposition to act in a certain kind of way (which he referred to as the interpretant) to a certain kind of object (the signification) under certain conditions (the context). To help understand the relationship between these elements it is useful to consider a case of non-human communication, the study of which is now referred to as animal or zoosemiotics. In the 1950s and 1960s Karl Von Frisch carried out a series of studies which revealed evidence of communication amongst European honey bees. When a honeybee scout discovers a useful source of nectar it flies back to its hive and then performs a ‘dance’, observed by other bees. The details of the dance vary depending on the distance to the source of nectar and on the particular species and variety of bee. In the most publicised case the bee performs a tail-wagging dance in the form of a squashed figure eight with a straight middle section. The bee signifies to other bees the source of nectar by varying or modulating a number of the elements of the dance. The time the dancing bee takes to complete the figure eight indicates the distance to the nectar source. The longer the time; the longer the flight to the source. The level of excitement of the bee in the dance indicates the quantity of nectar. The greater the level of excitement then the more nectar there is at the source and hence the more bees are needed to collect it. The orientation of the ‘straight’ part of the figure eight represents the direction of the source with respect to the position of the sun. For instance, if the ‘straight’ section is oriented at 80 degrees to the left of straight up then the bees are instructed by this to fly toward a point 80 degrees towards the left of the sun (see Fig. 3). Morris (1964) portrays this example in terms of the process of semiosis. ‘Karl von Frisch has shown that a bee which finds nectar is able on returning to the hive to ‘dance’ in such a way as to direct the other bees to the food source. In this case the dance is the sign; the other bees affected by the dance are interpreters; the disposition to react in a certain kind of way by these bees, because of the dance, is the interpretant; the kind of object toward which the bees are prepared to act in this way is the signification of the sign; and the position of the hive is part of the context’. So the bee’s communicative act (it’s dance) signifies to other bees its intention. They take action on the basis of the message received. The main area of debate surrounds whether examples such as this from the animal world indicate the presence of information in the process of semiosis. Most biologists would propose that bees will act instinctively in response to a fellow bee’s dance. In other words, ‘programming’ in the bee’s innate biological makeup will predispose it to react. This implies that the bee does not interpret the data received and decide to act in a certain way–no information is interpreted and no decision-making interposes between data and action in such a case.

98

P. Beynon-Davies / International Journal of Information Management 29 (2009) 92–103

Fig. 3. The signification of bees.

However, within the human sphere signs are used in a much more complex way because of the more sophisticated nature of human languages as sign systems and the importance of human interpretation to information representation and transmission using such languages. In its broadest sense the term language refers to an agreed system of signs used to convey messages between a set of communicative agents. Language is important because as Stephen Pinker states, ‘a common language connects the members of a community into an information-sharing network with formidable collective powers’ (Pinker, 2001). It is possible to type languages in a number of different ways, such as whether they are natural/artificial, formal/informal and what medium of communication they use. Certain languages such as English are described as natural languages. They have grown ‘naturally’ amongst a speech community and hence as a matter of course are continuously changing. Certain other languages are invented or created—Esperanto being a particular example of an attempt to create an expressive language which would be easy to learn and hence might act as a universal human language. It is conventional to distinguish between spoken and written forms of natural languages. Frequently the conventions of spoken and written language are considerably different; as is the case in spoken and written Welsh. However, we need also to add to this categorisation the non-verbal or body languages, consisting of nonverbal cues and gestures which convey meaning between human actors (Morris, 1979). Most natural languages have a complex, continuously changing syntax and semantics. As such they are informal languages.

Other languages are designed languages in that they have a highly specified and limited formal syntax and semantics. Hence, they are referred to as formal languages. The most notable examples of such languages are programming languages used in the construction of ICT systems such as Java or Perl. Information systems are systems for using signs in the sense that they act as a communication medium between different people, sometimes spatially and temporally distant. It is therefore possible to consider the nature of an information system from the point of view of language and action (Auramaki & Hirschheim, 1992). However, as we shall see, information systems are best described as semi-formal languages: some of their features are designed; some of their features emerge in continuous human interaction. At the semiotic level of pragmatics, (Beynon-Davies, 2009a) information systems can be seen to be semi-formal languages that are used to create, control and maintain social action. On the one, hand an information system is tied to action. On the other hand an information system is tied to a system of artefacts. The ‘language’ of some information system includes formal messages that create, control and maintain social interactions in an organizational context. Such messages not only serve to make statements about the world. They are also used to give orders, make promises or commitments, classify things, etc. The ‘language’ of some information system, just like all languages, therefore consists of messages with a correct syntactic structure, a meaningful semantic context and a significant pragmatic use. Take the example of a system which records information relating to orders for particular products by particular customers. This system can be conceived as a number of discrete acts of communi-

P. Beynon-Davies / International Journal of Information Management 29 (2009) 92–103

cation between customers and the organization. In the act of placing an order with the company the customer commits herself to receiving goods ordered and paying for products selected. The company in turn commits itself to delivering goods to the customer in an agreed time-period and in a satisfactory state for an agreed price. These various commitments are likely to be stored in records maintained both by the customer and the supplying organization. The values in the records act as a persistent memory of the commitments negotiated. Within the realm of the language-action approach, the structure of some information system is considered in terms of ‘speech acts’ and conversations (Searle, 1970). The main idea is that engaging in some communication, such as uttering a sentence is the performance of an act. In the language-action tradition these acts of communication are referred to as speech acts, even though such acts are not restricted to the use of spoken language. The term speech act would also be taken to cover written texts and the use of other signs such as gestures, flags, etc. The language-action tradition also refers to speakers and hearers as agents of a communication even though a communication may, for instance, be written. Therefore, to avoid confusion, we prefer to use the more encompassing terms: sender, receiver and communicative act. This terminology also accords with that introduced in our first paper on the nature of information (Beynon-Davies, 2009a). Within the language-action tradition, communicative acts are seen as the basic unit of human communication and are categorised into numerous different types. The most important type of communicative act for the purpose of defining the nature of information systems is the illocutionary act. An illocutionary act is an intentional act of communication. It is performed whenever an actor creates and sends some message in an appropriate context with certain intentions. For instance, a human actor might make spoken statements such as ‘I promise to write a letter’, ‘I refuse to pay a bill’. Searle (1975) identifies a number of different forms of illocutionary acts: assertives, directives, commissives, expressives and declaratives. Assertives are communicative acts that explain how things are in the world, such as reports and assertions. Such acts commit the speaker to the truth of the expressed proposition. Directives are communicative acts that represent the senders’ attempt to get a receiver to perform an action, such as requests, questions, commands and advice. As such, they cause the receiver to take a particular action. Commissives are communicative acts that commit a speaker to some future action such as promises, oaths and threats. They are communicative acts that represent a speaker’s intention to perform an action. Expressives are communicative acts that represent the speakers’ psychological state, feelings or emotions such as apologies and criticisms. They express a speaker’s attitudes and emotions towards some proposition. Finally, declaratives are communicative acts that aim to change the world through the communication itself, such as baptism, pronouncing someone husband and wife and sentencing a prisoner. Hence, they are usually communicated against a normative background and are frequently institutionalised. The structure of any illocutionary sign act according to Searle consists of three parts: content, context and illocutionary force: • Content refers to the propositional content of the message. Hence, the content of the commissive act – ‘I promise to write a letter’ – is ‘to write a letter’. • Context is defined in terms of the sender, receiver, time, place and the ‘world’ or accepted range of understandings within which the illocutionary act takes place. Hence the context defines a message from a sender to a receiver at time T and in place P. World is used to collect together a range of other features of context which are important for understanding the meaning of a mes-

99

sage, such as channel (speech, writing, etc.), code (language style) and message-form (chat, debate, sermon, etc.). • Illocutionary force is used to represent the kind of commitment a speaker makes when he says something and the direction of fit between the world and the propositional content of the communicative act. For instance, a business order has a sign-to-world direction of fit as it is intended to change the world (precipitate some action), and the sender commits himself to the future action of paying an invoice in return for receiving some goods or services. Illocutionary acts of communication normally occur in ordered sequences. These sequences are referred to as ‘conversations’, and have much in common with the idea of dialogues or a stream of discourse (Clark, 1996). Again, to avoid association solely with speech, these sequences are best referred to as communicative patterns. The major feature of such communicative patterns is their ‘game-like’ character. In other words, a particular communicative act creates the possibility of usually a limited range of communicative acts as response. For instance, in verbal discourse between two human actors a question is normally responded to with an answer, an assertion with a statement of assent or disagreement, a statement of thanks with an acknowledgement and an apology with an acceptance. In this view, information systems are seen as systems which create, maintain and fulfil communicative acts. Consider, for example, the case of the Warning Network. The information system supporting the activity of Fighter Command could be interpreted as a number of inter-linked patterns of communication between actors within the different social groups involved in joint endeavour: radar operatives, observation corps operatives, filter room staff, operations room plotters, operations room controllers, aircraft pilots, and so on. Hence, elements of the communicative pattern from such an information system, includes communicative acts such as • A radar operator makes a statement of the expected strength, altitude, position and direction of an aircraft group to filter room staff (assertive). • A member of the observer corp. reports the expected strength, altitude, position and direction of an aircraft group to filter room staff (assertive). • A station Y operative reports the content of signals intelligence to filter room staff (assertive). • A filter room operative collates incoming data and confirms the strength, altitude, position and direction of an aircraft group to members of operations staff and the operator of the tote (declarative). • A plotter within the operations room confirms the last-known details of an aircraft group by plotting it on the plotting table (declarative). • The operator of the tote confirms the current state of an aircraft group on the tote (declarative). • An operations controller requests a group controller to change the state of readiness of particular sectors (directive). • A group controller requests a sector controller to change the state of readiness of a particular squadron (directive). • A sector controller instructs a squadron leader to change the state of readiness of his squadron (directive). • A sector controller confirms that a squadron has scrambled to a group controller (commissive). • A group controller confirms that a squadron has scrambled to a HQ controller (commissive). • A squadron leader indicates to group control that enemy aircraft have been sighted (assertive). Each of these communicative acts can only occur at particular points in patterns of communication between radar operators,

100

P. Beynon-Davies / International Journal of Information Management 29 (2009) 92–103

filter room staff, operations room plotters and various levels of operations controller. Each communicative act also calls into play a limited number of other communicative acts in response. This means that a graph should be able to be produced which documents the conventional pattern of communication within the Warning Network. The language-action perspective has proven influential in certain sectors of informatics and has particularly stimulated endeavour in Computer-Supported Cooperative Work (Winograd & Flores, 1986), business process modelling (Dietz, 2006) and more recently in communication support systems (Te’eni, 2006). It has also been much discussed as a valid approach within the theory of Information Systems (Lyytinen, 1985). However, as Winograd (2006) acknowledges, the perspective has not really had any significant impact upon approaches to the design of what he refers to as information systems, but we shall refer to as ICT systems. There are a number of reasons for this. Part of the reason is the emphasis within this sub-discipline upon the development of formal notations for representing communicative patterns that are complex and frequently in opposition to conventional techniques adopted in the analysis and design of information systems and ICT systems. Hence, the data or information flow diagram illustrated in Fig. 2 would not be seen as adequate from the language-action perspective, because it is seen to reify communication as information flow and as a consequence obscures the role of human actors in this process. However, from our point of view, the main difficulty with the approach is the lack of a clear conception of social context on the one hand (Suchman, 1994) and technology on the other. Many attempts to apply the language-action approach to informatics explicitly leave out considerations of the linkage between an information system and its activity system. They also tend to lack an explicit conception of the ‘record’ and/or ‘transaction’ within communicative patterns. These limitations suggest the need to adapt subtly the conception of an information system implicitly adopted in the language-action approach for our purposes. The word information can be traced to the Latin verb Informare, meaning ‘to shape’, ‘to form an idea of’ or ‘to describe’ (Hobart

& Schiffman, 1998). Dietz (2006) plays with the root of this verb (forma) and defines three sets of actions that we adapt for our purposes to correspond both with the three types of systems referred to above and to the layers of the semiotic ladder. Information systems deal with informa – the content and meaning of signs. Informa therefore crosses the semantics and pragmatics levels of signs on the semiotic ladder. An information system is an example of a communication system and hence can be seen to consist of human actors engaging in communicative or informative acts (Fig. 4). Forma is a term used to refer to the ‘substance’ that carries a sign. Forma therefore crosses the technical and empirics levels on the semiotic ladder. Within an information system much communication is stored, manipulated and transmitted via information technology – these are all formative or data acts. Humans utilise formative acts to extend the reach of their communicative acts across time and space, as well as between multiple actors. The purpose of informative acts is to ensure coordination of activities. Performa is used to refer to the use of communication within social action. Acts of production or performative acts correspond to the ‘work’ of people in collective interaction. The linkage between an information system and an ICT system occurs through the concept of a transaction. Associated with production or performative acts within activity systems is a corresponding flow of communicative or informative acts. These communicative acts generate data structures that record some coherent unit of activity, typically an event within some activity system or between activity systems. Such transactions typically write to the records of some ICT system. The purposes of human agents engaging in some activity system are encoded as intentions within the communicative acts of some information system. These intentions are also decoded in a twopart process which involves reading data from the ICT system as transactions and interpreting information from the corresponding information system in support of decision-making and action. In essence, the idea of the distinction between data (formative) acts, communicative (informative) acts and production (performative) acts is a convenient division to highlight the different roles

Fig. 4. Formative acts, informative acts and performative acts.

P. Beynon-Davies / International Journal of Information Management 29 (2009) 92–103

associated with the more encompassing concept of a sign act. A sign act has intent (pragmatics), expresses meaning (semantics), has structure (syntactics) and has a physical representation (empirics). As such, a given sign act bridges between the social world (activity system) and the technical world (ICT system) through the mediating realm of human communication (information system).

4. Discussion Besides its critical historical importance, the case discussed in this paper of the Warning Network is extremely useful in offering key insight into the characteristics of information systems in general and consequently the nature of informatics. Broadly, the case is useful in emphasising the importance of three types of systems to human organization: activity systems, information systems and ICT systems. The case is also useful in demonstrating the inherent interaction between these three types of systems in any conception of the structuration of human organization (Walsham & Han, 1991). Structuration theory speaks of the ‘duality of structure’. On the one hand, the structure of social institutions is created by human action and is only evident in human action. Through human interaction, social structures are reproduced but may also change. On the other hand, human action is constrained by the way in which humans utilise institutional structure as a resource in interpreting their own and other people’s actions. This cyclical process Giddens calls the process of structuration (Giddens, 1986). Information systems as semi-formal languages can be seen as having many of the features of a social institution. A ‘communication protocol’ common to a group of persons is a necessary pre-condition for communicative acts. Such a ‘language’ can hence be described and studied somewhat independently of given communicative acts. In this sense the information system can be seen to have an existence independent of the humans using it. However, the information system is only really evident in actual communicative acts. People use a sign-system as a resource for communication and this sign-system is produced and re-produced through communicative acts. Over time, communicative acts may change the structure of the sign-system itself leading to a new basis for the institutional resource. Informatics as a discipline is particularly interested in questions relating to the value of information systems and ICT. Such questions can only be answered in terms of the activity systems in which these systems are embedded. We have defined an activity system as a set of activities (performative acts) performed by a group of persons in fulfilment of some defined purpose. Activity systems are designed systems (Checkland, 1987) and are typically modelled in terms of sets of processes with embedded control. Activity systems are also social systems. They consist of people engaging in coordinated and collaborative action. In the case of the Warning Network the activity system consisted of the command and control of the fighter aircraft of the RAF. Information and communication technology is any collection of artefacts used to support aspects of an information system. The information system in this case was reliant on aspects of information technology. However, the ICT described in the example was subtly different from modern information and communication technology. In the case of the Warning Network this information technology comprised radar, the telephony network and devices such as the plotting table, tokens, plotting clock and tote. Many systems in organizations are examples of socio-technical systems. A socio-technical system is a system of technology used within a system of activity. Information systems are primary examples of socio-technical systems. Information systems consist of ICT used within some human activity system. They therefore span between ICT and human activity. Part of the human activity will

101

involve use of ICT systems. The information provided by the information system will also drive decision-making leading to further action within the organization. In other words, information systems constitute communication systems designed to support human activity with the aid of technology. It is important to recognise that information systems have existed in many forms of human organization prior to the invention of modern ICT, and hence information systems do not need modern ICT to exist. The value of historical cases lies in the way in which we can distance ourselves from modern ICT and consider the essence of what information systems constitute. However, in the modern, global and consequently complex organizational world most information systems of the present day rely on hardware, software, data and communication technology to a greater or lesser degree. Some of the key lessons from the Warning Network case lie at the heart of the nature of an information system as a socio-technical system. An information system is fundamentally concerned with communication in support of human activity using artefacts to represent, store, manipulate and transmit data. The essence of an information system therefore lies not purely in the technology or in the activity: it lies in the way in which technology is used in support of purposeful action. For instance, radar by itself was not key to effective control of the skies over Britain during this period. Instead, it was the way in which the information technology was used within an encompassing information system to support purposeful action that proved key. Good information systems are therefore critical to effective human action. The information system within the Warning Network was set up by the RAF to enable them to better coordinate their actions in relation to beating off the mass raids of the German Luftwaffe. Hence, the key utility or value of this information system was established in relation to the effectiveness of action reliant upon it. Information systems are systems of communication. They involve people in producing, collecting, storing and disseminating information. The information system of the Warning Network involved collecting data from channels such as radar, organizing this data for military-decision-making and the dissemination of both decisions and data to airfields. Consider the place of the operations room at RAF Fighter Command. Here various elements such as the makeup and positioning of counters on the plotting table and the signals on the tote acted as a sign-system (Beynon-Davies, 2009a,b) for relevant human actors. The state of such signs at any one point in time was used as a ‘real-time’ record or model of the disposition of aircraft within the air defence battle. On the basis of the key actors’ interpretation of this record, tactical decisions were made in relation to the deployment of fighter aircraft. However, information systems are subtly different from general systems of human communication: an information system is a specialised sub-type of a communication system. The key difference between an information system and a communication system can be illustrated with an example. In the children’s game ‘telephone’ a chain of children is required to pass on a verbal message. The ‘sender’ of the message whispers it in the ear of the first child. The first child then whispers it in the ear of the second child and so on down the chain until it reaches the ear of the ‘receiver’ child, the last person in the chain. Usually, the message is highly distorted by the time it gets to the end of the chain. Hence a message like ‘I love you’ can end up as ‘Joe hates Gill’. Therefore, this game normally demonstrates some of the limitations of human verbal communication in groups. Communication theorists would explain the distortion in the message through the concept of ‘noise’. However, when substituting written for verbal communication the communication is less subject to distortion. Hence, getting each child to write the message transmitted separately each time and pass it on to the next

102

P. Beynon-Davies / International Journal of Information Management 29 (2009) 92–103

child reduces problems for instance of poor hearing and misinterpretation. Passing of course the same written message along the line should minimise distortion. This highlights the usefulness of records in group communication. Hence, we need to make a clear distinction between a communication system and an information system. An information system is a communication system in which messages are encoded, transmitted and decoded in the form of transactions and stored in the form of records. Transactions are highly formalised records of communicative acts flowing along defined communication channels. Records are formalised items of persistent communication. They have a life over and above the communication process within which they took part. For example, the use of tokens on the plotting table of the operations room at RAF Fighter Command is an important example of the representation of data in ‘records’ of communication. A given token was created by one of the plotters within the operations room as a record of a communication received from one of the WAAF operatives in the filter room. The makeup and use of a particular token upon the plotting table was used to signify a number of things to all present in the operations room: • The shape of the token signified the type of aircraft. A triangular or wedge-shaped wooden block was used to represent a group of allied aircraft. A cylindrical or disc-shaped wooden block was used to stand for a group of enemy aircraft. • The altitude and number of aircraft in a particular group was signified by annotations to the token. Two set of numbers were slotted into each triangular block: one on each face. One number referred to the altitude last reported for the group. The other number indicated the reported strength of the group. • The recency of the record was indicated by the colour assigned to a token. The approximate time data was received about a particular aircraft group was coded in terms of the three colours calibrated against the plotting clock. Each token had an appropriate coloured ‘flag’ attached to it. • The positioning of the token against the map of the plotting table signified the last known position of the aircraft group within the defined air-space. This example highlights another key feature of an information system. An information system is characterised not only by one-to-one or dyadic communication. One particular advantage of the use of records within communication is that the presence of such persistent artefacts facilitates one-to-many and many-tomany communication within a group of actors over a period of time. It turns individual memory into social memory. In terms of such many-to-many communication in particular, an information system can be said to ‘model’ critical aspects of its encompassing activity system. For instance, in a way the total collection of tokens on the plotting table, their makeup, and their positioning signified to both plotter and more importantly to operations controllers the current disposition of a particular air battle. 5. Conclusion The general theme linking this series of papers is that core elements of informatics are universal in nature. On this basis we propose that a clearer and more sophisticated systematics for the core concepts of information, information systems and ICT is possible. We further propose that clearer definitions of core terminology are not just of theoretical interest. Clearer terminology has practical consequences. For instance, within discussion of the Warning Network the issue of the value of information, information systems and ICT arose at a

number of points. This demonstrates that concern with the value of information-handling activity and the technology on which it relies is nothing new. However, with the increasing penetration of ICT into organizational experience the issue of value has become ever more prominent. In recent times, the heads of both private and public sector organizations have continually and repeatedly raised questions as to the value of ICT to their organizations. Many billions of dollars has been invested globally in ICT but companies still find it difficult to put an accurate monetary value on the return to the organization from this investment. In many ways this apparent paradox of ICT (Brynjolfson, 1993; Brynjolfson & Hitt, 1998) is due to a poorly developed systematics for the area. In other words, it is typically due to the way in which ICT is conceptualised by business managers, by technologists and by many academics. Such groups have a tendency to overtly focus on the technology itself and ignore crucial questions relating to its appropriate application. To play devil’s advocate, it is possible to argue that ICT alone (for example, as a set of computer systems situated on communication networks) is worthless to most organizations (unless of course you are a business which produces or sells ICT!). Instead, the value of ICT emerges in its relationship with organizational activity through the mediating force of systems of information. This means that to ensure that ICT has value, organizations must consider how this technology is fit for purpose. In other words, they must ensure that ICT infrastructure in organizations matches the ongoing information needs of activities within and between organizations. In the first paper we argued for considering information as a universal feature of human organization and introduced the idea of an information system as a communication system with necessary record-keeping. In this paper we have considered the primary nature of an information system as a socio-technical system. On the one hand, this serves to identify the purpose of an information system in support of coherent decision-making and action. On the other hand, it serves to locate the place of information technology as a system of artefacts which serve to signify models of the world. It is to this latter issue that we turn in the next paper of the series.

References Alter, S. (2005). Architecture of Sysperanto: A model-based ontology of the IS field. Communications of the Association for Information Systems, 15, 1–40. Auramaki, E., & Hirschheim, R. (1992). Modelling offices through discourse analysis: The SAMPO approach. The Computer Journal, 35(4). Beynon-Davies, P. (2002). Information systems: An introduction to informatics in organisations. Houndmills, Basingstoke: Palgrave. Beynon-Davies, P. (2007). Informatics and the Inca. International Journal of Information Management, 27(5), 306–318. Beynon-Davies, P. (2009a). Neolithic Informatics: The nature of information. International Journal of Information Management, 29(1). Beynon-Davies, P. (2009b). Business information systems. Houndmills, Basingstoke, UK, Palgrave. Brynjolfson, E. (1993). The productivity paradox of information technology. Communications of the ACM, 36(12), 67–77. Brynjolfson, E., & Hitt, L. (1998). Beyond the productivity paradox. Communications of the ACM, 41(8), 49–55. Checkland, P. (1987). Systems thinking, systems practice. Chichester: John Wiley. Chomsky, N. (1998). On language. New York: The New Press. Clark, H. (1996). Using language. New York: Cambridge University Press. Crowley, D., & Heyer, P. (Eds.). (2003). Communication in history: Technology, culture, society. Boston: Pearson Education. DeMoor, A. (2002). Language/Action meets organisational semiotics: Situating conversations with norms. Information Systems Frontiers, 4(3), 257–272. Dietz, J. L. G. (2006). The deep structure of business processes. Communications of the ACM, 49(5), 59–64. Giddens, A. (1986). The Constitution of Society: Outline of the theory of structuration. London: Polity Press. Grattan, R. F. (2005). Strategy in the battle of Britain and strategic management theory. Management Decision, 43(10), 1432–1441. Gregor, S. (2006). The nature of theory in information systems. Management Information Systems Quarterly, 30(3), 611–642.

P. Beynon-Davies / International Journal of Information Management 29 (2009) 92–103 Hobart, M. E., & Schiffman, Z. S. (1998). Information ages: Literacy, numeracy and the computer revolution. London: John Hopkins University Press. Holwell, S., & Checkland, P. (1998a). Information, systems and information systems. Chichester, UK: John Wiley. Holwell, S., & Checkland, P. (1998b). An information system won the war. IEE Proceedings Software, 145(4), 95–99. Kirby, M., & Capey, R. (1997). The air defence of Great Britain, 1920–1940: An operational research perspective. Journal of the Operational Research Society, 48(6), 555–568. Kling, R., & Allen, J. P. (1996). Can computer science solve organisational problems? The case for organisational informatics. In R. Kling (Ed.), Computerisation and controversy: Value conflicts and social choices. San Diego: Academic Press. Lyytinen, K. J. (1985). Implications of theories of language for information systems. MIS Quarterly, March(9), 61–74. Mason, R. O., Mckenney, J. L., & Copeland, D. G. (1997). An historical method for MIS research: Steps and assumptions. MIS Quarterly, 21(3), 307–319. Morris, C. (1964). Signification and significance. Cambridge, Mass: MIT Press. Morris, D. (1979). Manwatching: A field guide to human behaviour. London: Harry N. Abrahams. Ogden, C. K., & Richards, I. A. (1923). The meaning of meaning. London: Routledge and Kegan Paul. Pinker, S. (2001). The language gene. Harmondsworth, Middx: Penguin. Porter, M. E., & Millar, V. E. (1985). How information gives you competitive advantage. Harvard Business Review, 63(4), 149–160. Puri, R. (2006). The role of intelligence in deciding the battle of Britain. Intelligence and National Security, 21(3), 416–439. Robinson, B., & Wilson, F. (2002). Soft systems methodology and dialectics in an information environment: A case study of the battle of Britain. Systems Research and Behavioural Science, 20(3), 255–268. Searle, B. J. R. (1970). Speech acts: An essay in the philosophy of language. Cambridge: Cambridge University Press.

103

Searle, J. R. (1975). A taxonomy of illocutionary acts. In K. Gunderson (Ed.), Language, mind and knowledge (Vol. 7) Minneapolis. Stamper, R. K. (2001). Organisational semiotics: Informatics without the computer? In L. Kecheng, R. J. Clarke, P. Bogh Anderson, & R. K. Stamper (Eds.), Information, organisation and technology: Studies in organisational semiotics. Dordecht, Netherlands: Kluwer. Strassman, P. (1990). Business value of computers. New Canaan, Conneticutt: The Information Economics Press. Suchman, L. (1994). Do categories have politics?: The language/action perspective reconsidered. Computer Supported Cooperative Work, 3(2), 177–190. Te’eni, D. (2006). The language-action perspective as a basis for communication support systems. Communications of the ACM, 49(5), 65–70. Walsham, G., & Han, C.-K. (1991). Structuration theory and information systems research. Journal of Applied Systems Analysis, 17. Weigand, H. (2006). Two decades of the language-action perspective. Communications of the ACM, 49(5), 45–46. Winograd, T. (2006). Designing a new foundation for design. Communications of the ACM, 49(5), 71–73. Winograd, T., & Flores, F. (1986). Understanding computers and cognition: A new foundation for design. Norwood, NJ: Ablex Publishing. Paul Beynon-Davies is currently Professor of Business Informatics at Cardiff Business School, Cardiff University. Before taking up an academic post, Professor Paul Beynon-Davies worked for several years in the informatics industry in the UK. He still regularly acts as a consultant to the public and private sector particularly in the area of information and communications technology (ICT) and its impact on organizational performance. He has published widely in the field having 10 books and over 60 academic papers to his name. He was engaged in a number of projects related to the impact of ICT on the economic, social and political spheres, both in the UK and Europe.