Knowing what we know
This chapter describes a theoretical framework for conceptualising the practices of ordinary knowledge, how we know things in our ordinary lives, and the boundaries of knowing. The framework is based on Russell Hardin’s theory of ordinary knowing and the notion of boundary objects developed by Susan Star and James Griesemer. Its premise is that the decisions we take make sense at the moment we make them, but at the same time what we can know perfectly is limited by many types of boundaries. The choices are not necessarily easy or effortless, but we feel they are sensible. The boundaries that limit our knowing can be technological, social and even physical. Those between different knowledge communities hinder the possibility of us knowing things but at the same time create a comfort zone, which makes things understandable for us. In the pursuit to be economic in our knowing, to focus on knowing relevant things, we can be unable and often unwilling to traverse many of the boundaries, which are volatile. They are plastic enough to adapt to the needs of specific communities but at the same time robust enough to traverse boundaries and make them recognisable and understandable on the different sides of the boundaries. The boundaries can be bridged using so-called boundary objects, which reside between different communities of knowledge. In addition to trying to span and cross the boundaries by ourselves, we can try to help others to cross their boundaries either together with them as peers, or as ‘experts’ with a different horizon of capabilities and limitations. The help can be given by teaching, guiding or coaching others to cross a particular boundary, or by helping them to learn where the boundaries are and how to cross them in general.
People know in many different ways. Knowing is based on science, institutions, religious beliefs and practices, and cultural and moral judgements among many other sources. Most of the time things are plainly known. It is very difficult to explain how and why some things are known or not known outside the scope of epistemologically rigorous knowledge systems based on a premise, albeit a theoretical one, of the existence of a super knower as Hardin (2009) suggests. Explanations of knowledge have a tendency to become either very simple, similar to Friedrich Hayek’s (1945) reasoning that prices are sufficient signals to coordinate human activity, or extremely complex, as in models with a large number of variables and explanations (e.g. Jakubik, 2011). They tend to be based on an assumed influence of individual factors (such as price) or on highly complicated models that are difficult to use in practice. Research shows that even experts of specific fields have significant difficulties in explaining why they happen to know certain things they know (Bouwman et al., 1987). The difficulties of providing rational explanation pertain even to such heavily regulated areas of work as policing or theoretically rigorous areas of knowledge like science (Latour and Woolgar, 1986). The observation that in ordinary everyday life contexts of work and leisure people do not know why they know or don’t know is seemingly trivial, yet it is almost always ordinary knowledge rather than specialist knowledge that affects how people decide to act and behave in their daily pursuits. We very seldom have real opportunities to rely on advanced specialist knowledge when we make decisions in our daily lives.
Although they lack epistemological diligence, the economic theories of knowledge have tried to address the fundamental question of knowing and decision-making in ordinary life. The question is: how do we behave when we know? In this context the notion of behaviour should not be understood as behaviour in a behaviourist sense, but rather in an everyday sense of doing things. The premise of knowing is that the ideal state of possessing complete knowledge is not a practical reality. Our knowledge is limited and mostly we do not even know how. Zygmunt Bauman (2000) has argued that in contemporary reality, which he calls ‘liquid’, uncertainty has extended to the aims of human activity. In the absence of traditional supreme societal authorities, knowledge is limited by the inability not only to achieve goals, but also to identify what is the significant nature of the goal.
In economics, the notion of information asymmetry (Rosser, 2003) refers to a situation in which one of the parties is more knowledgeable than the others. The typical context of analysing information asymmetry relates to transactions and financial decision-making. The principal models of information asymmetry explain the situations of adverse selection (Akerlof, 1970) and moral hazard. Adverse selection refers to decisions made on the basis of hidden or suboptimal knowledge of the real conditions of the other party. An example could be a decision to rely on what is a seemingly trustworthy social web service for safekeeping personal data while in reality the chosen web service provider would silently supply third parties with the data and perhaps would also be knowledgeable about the technical deficiencies of the web application used in the service. An example of a moral hazard could be an information sharing service originally started with good intentions to enable people to share their data, but that is used to secure criminal transactions. The two strategies to counter adverse selection are known as signalling and screening (ibid.). Signalling is the use of certain signals to communicate credible information. Wikipedia, the free encyclopaedia, signals credibility by opening up the edit history of articles and enforcing a complicated series of policies to ensure the quality of its contents. A typical strategy of signalling is to refer to formal qualifications and external evidence. Screening is the process of acquiring information. The less informed party tries to collect enough information in the course of time to balance the asymmetry of information (Stiglitz, 1975).
Lack of information is not, however, the only obstacle that hinders ‘knowledgeability’ (the possession of knowledge). Information science researchers have acknowledged the significance of uncertainty as an essential quality of human information seeking and use (T.D. Anderson, 2006). Anxiety and uncertainty have been found to be major sources of failure in information seeking (Case, 2002: 106), but – paradoxically – also necessary premises of success and innovation (ibid.; Kuhlthau, 1993). Another elementary observation of the obstacles and anxieties of information seeking is that despite the tendency to prefer certain types of knowledge – for example, scientific or religious – different things are known in different ways. The tendency may be partly explained as a form of external uncertainty or indeterminacy, as Ahmad M. Kamal and Jacquelyn Burkell (2011) suggest, but the preferences do not always relate to biases or the lack of skills as they propose, especially when there are multiple correct answers to individual questions and problems. There are multiple modes of knowing (Hardin, 2009). A scientist expects a high degree of rigour of the presentation and argumentation of scientific knowledge. At the same time, the scientist may take for granted hearsay about his neighbours or rely on a review of a household product written by an unknown critic on a website. You might say that the latter mode of knowing is biased and based on a lack of skills, but this type of a value judgement fails to appreciate the variety of ways in how things can be known in the two cases. There are certain rules that make a scientific argument plausible, but the usefulness of a household product is not guided by norms following a similar logic.
The third pragmatic observation is that it is possible to make relatively good decisions based on simplifications of knowledge. As already noted, Hayek argued in his essay ‘The use of knowledge in society’ (1945) that prices are sufficient signals to coordinate human activity. He suggested that changing prices provide people with information that helps them to make decisions about buying, investing or selling assets; for instance, they decide to buy a book or to travel to a foreign country. In many cases, decisions are based on very basic and in some senses remarkably suboptimal categories of knowing (Hardin, 2009). A classic example is voting behaviour. Joseph Schumpeter (1950) argued that in democratic countries voters rely on arguments and analyses that would be readily recognised as ‘infantile’ within the person’s sphere of interest and expertise. The problem with similar weak signals on the social web is that they are not readily recognisable and relatable to the entities of physical reality. There is no separate digital reality that would be a complete state of otherness, but as the discussions of intellectual property rights and the fair price of digital commodities illustrate, economic theories of knowing based on the exchange of physical commodities do not necessarily translate well into the domain of digital information.
Russell Hardin (2009) has proposed a theory of the economics of ordinary knowledge that discusses the typical inadequacy of knowledge in decision-making. Even if individuals would like to be fully rational, it is difficult for them because they lack comprehensive knowledge. The opposite concept, discussed in his earlier book (ibid.), is the problem of predicting the outcomes of action: as Hardin argues, ignorance may in fact be a sign of ‘rational’ (in an everyday sense) behaviour. He argues against Schumpeter’s proposition by positing that it is rational not to put too much effort into choosing the right electoral candidate, because one vote is not very significant. Hardin’s argument is based on an observation that is well known in information science research. When seeking knowledge people tend to use the most convenient method to find it (Connaway et al., 2011). The tendency has been conceptualised with slightly varying emphases within different theoretical frameworks – including Zipf’s Principle of Least Effort (Bruce, 2005; Saracevic, 2007) and the influential studies on the information search process by Carol Kuhlthau (1991, 2004) – and from the uses and gratifications perspective (Case, 2007: 154–7). The findings of Byström and Kalervo Järvelin (1995) on the influence of task complexity on information seeking and use, and more recently by Wilkinson et al. (2012) on how people adapt their reading behaviour according to time pressures, provide further empirical evidence on the economics of information and knowing behaviour. But in contrast to the prevalent assumption, convenience is not necessarily only a question of making a compromise between the quality of information and the ease of accessing it, as researchers have tended to argue for several decades (e.g., Taylor, 1968). The compromise is much more in the eyes of the beholder than of the observed individual.
As there are some apparent problems with Hardin’s argument, it is useful to question whether his theory is actually a theory per se (Leppälä, 2011). For an individual to vote in an election, although that person is just one among 50 million others, can hardly be described as irrational behaviour, in contrast to what Hardin seems to suggest. The choice to vote might be conceived to be based on a different form of rationality than the economic rationality proposed by Hardin. As an individual act, voting does not make much sense. However, as a contribution to a certain crowd wisdom, and especially as a form of communicative action (Habermas, 1984a), voting may be seen as deeply rational behaviour. The rationale is not to believe that I am right as an individual voter (and all others are potentially wrong), but that all the votes together and voting as a practice result in a status quo that is largely beneficial for me as an individual and for others. Besides ignoring Habermasian ideas of the public sphere and communicative action, Hardin may also be criticised for overlooking a large corpus of literature on philosophy, psychology and information science that might not only contradict the proposed theory but also support his arguments.
Despite their seemingly contradictory nature, Hardin’s reasoning resonates closely with the Habermasian notion of communicative rationality (Habermas, 1984a) and especially how it rebuts the instrumental form of rationalism. Hardin is less interested in communication per se than in the personal economical behaviour of individuals he describes in terms that are close to the instrumental ideas of the classic form of rational choice theory. Hardin seems to suggest that, in general, people tend to be rather instrumental in their rationality, although he emphasises that the aims of their instrumental behaviour can be very diverse and even irrational from the perspective of an outsider. The reason why people are rational in a communicative sense for Habermas (and irrational for Hardin) or in the instrumental sense of the classical rational choice theory depends on their conception of the type of knowledge that is relevant in a given situation. The rationale of human action may be to reach instrumental goals or to communicate or, perhaps even more so, to economise between multiple forms of rationality.
The basic theoretical assumption of the present volume is to see human beings as profoundly sensible in their constantly changing individual and shared contexts and situations of rationality and economics. The balance between different forms of rationality depends on multiple factors. An economically rational basis for a decision is a subjective and cultural concept as much as it can be a measurable one. As Lawrence Grossberg (2010) has aptly remarked, the measures and ‘economics’ of things are not necessarily related to the classical assumptions and theories in the field of economics. Contextual and situational reasons can make an economical (financial) decision deeply uneconomical (for instance, socially). The criticised emphasis of the measurable forms of instrumental rationalism and a relative disinterest in the importance of communicative action, communality and other contextual factors is clearly visible in the information science literature. Information seeking and use has been seen as a highly goal-oriented activity and even the accounts that focus on the rationality of (instrumentally) irrational information activity tend to describe it as a convenient survival strategy instead of being a perfectly sensible and (in an everyday sense) economic pattern of behaviour. Only relatively recently – in the context of the broader emphasis on everyday-life information seeking, information and emotions, and serendipity – has the role of intuition and (in an instrumental sense) seemingly irrational non-directed information-seeking behaviour received more attention among researchers as a legitimate pattern of activity (e.g., Björneborn, 2008; McKenzie, 2003; Nahl and Bilal, 2007). However, even then, the comparison of instrumental and serendipitous modes of information seeking implicitly takes the first as a yardstick and risks ‘colonising’ the latter as an alternative quasi-instrumental ‘strategy’.
The notion of ‘satisficing’, the idea of the preference for ‘good enough’ knowledge coined by Herbert A. Simon (1955), has often been cited in information science literature as a rationale for imperfect information seeking. The research on satisficing behaviour in information seeking and use may be credited for an analytical rigour and elaborateness that is lacking in Hardin’s theory. For instance, Prabha et al. (2007) and Zach (2005) have identified a detailed set of quantitative and qualitative criteria for how people decide what is enough when they seek information. Following Hardin, the problem with the approach of elaborating detailed criteria and the entire notion of satisficing is that the ideas of ‘good enough’ or ‘true enough’ make perfect philosophical sense (and are instrumentally rational), but are almost too evident in practical contexts. It has been demonstrated that people gather and process information as long, and only as long, as it is necessary to make a decision (de Lange et al., 2011), but there is nothing inherent in the situation that would let us make a distinction between ‘good’ and ‘good enough’. Almost all ordinary knowing is merely satisfactory, but ‘good enough’ knowledge is only rarely conceptualised as explicitly ‘good enough’ (Hardin, 2009). Whether information is ‘good’ or deficient in some way becomes an explicit issue only rarely, when we know the haphazard nature of a situation, such as in early warning decisions studied by Chun Wei Choo (2009). If asked, people may admit the possibility that their knowledge is not perfect, but in practice, as Hardin argues, there is little real difference between whether knowledge is ‘good’ or ‘good enough’, because we usually have very limited resources to decide this for sure. In practice, knowledge is mostly ‘good’ or ‘bad’, or even both at the same time. In information science research, a typical indication of a satisfactory state of knowledge (‘good enough’) is the moment when a person stops seeking more information. Chandra Prabha et al. (2007) list available time as a factor that limits information seeking, but it is easy to conceive a long list of other factors that limit explicit and implicit information- seeking activities in everyday life contexts that do not appear in formal information-seeking contexts. The simplest ones are perhaps that information is not found or something else had to be done instead.
The trade-off between such factors as the cost of searching and the quality of information (Fu and Gray, 2006) is seldom a real compromise. The findings of Allen (2011) from a study of police officers’ information activities highlight this issue by showing how intuition, or ephemeral observations and very subtle fragments of information, tend to be a very effective basis for making informed decisions. However, according to the instrumentally rational formal procedures of police work established to avoid arbitrariness and racial discrimination, this type of decision- making is irrational and the information used as a basis for decision- making is uncontroversially ‘bad’ and imperfect, and can be characterised almost as a belief rather than information.
In contrast to Hardin’s proposal (2009: 26), it may be reasonable to make a distinction between knowledge and information, and belief, as his critics (e.g., Hindriks, 2010; Leppälä, 2011) have prudently pointed out. People certainly act differently if they believe they have enough petrol to drive home from if they know that they have enough, because they just left a service station. This holds true even if both the belief and the knowledge might be equally uncertain in practice. What if the petrol gauge or the petrol pump is out of order? The reason why Hardin refuses to see the distinction is that it is often very difficult to make and there are many things people think they know even though they actually only ‘believe’ they do. Many believers conceive their religious beliefs to be knowledge based on complex evidence. Ultra-relativists, on the other hand, may be inclined to perceive (or believe) all knowledge as a belief. Essentially, the problem is that it is equally difficult to know if something is known as to know why something is known or not. Not only the knowledge but also its justifications tend to be ‘good’ instead of being readily categorisable as perfect, ‘good’ or ‘good enough’. The key difference between knowledge and a belief, however, is that they are two different expressions used by people to describe how they know. In contrast to what Hardin (2009: 26) suggests, calling something a belief expresses not merely the general inadequacy of human knowledge, but also the degree of the perceived certainty of a specific piece of knowledge.
The practical difficulties of distinguishing between ‘good’ and ‘good enough’ are one aspect of the problematic nature of the instrumental rationality of Simon’s notion of satisficing. The second and a more fundamental problem stems from the difficulty of knowing if we know and how to decide when we know enough. Yazdan Mansourian et al. (2008) studied unsuccessful information searching of academics and identified three types of failures: unexpected failures (the searcher was surprised that information was not found), unexplained failures (the searcher could not explicate why searching failed) and predicted failures (the searcher expected that information would not be found). Suboptimal information-seeking performance may also be perceived to be acceptable (Fu and Gray, 2006), because a satisfactory state of knowledge is very much dependent on the experienced context of knowing. As Hardin notes, people tend to rely on a very basic understanding of a physical or technical phenomenon. In contrast, much more detail is expected of medical knowledge. It is sufficient for my friend to get an email when I press the ‘Send’ button even if I would have a hard time explaining what actually happened. In contrast, if people become seriously ill, they tend to be interested in very elaborate information about their disease. Blaise Cronin (1991) made an important point when he made public that he was appalled at the poor quality of service provided by libraries in the beginning of the 1990s. He drew attention to the fact that people get answers to their questions or find the books they are looking for only 50–80 per cent of the time. He contrasted this success rate with that of travel agents, who almost always manage to book tickets correctly, and how infrequently air travellers’ luggage goes astray (perhaps disproving popular myth). Although his observations are correct, the problem with making this comparison is that while in air travel the notion of ‘good’ is fairly well defined, in much ordinary and even specialist information seeking ‘good’ is measured in completely different terms.
However, despite the critique, the main points of Hardin’s theory, the economic ‘rationality’ of seeming ignorance and coexistence of fundamentally very different modes of knowing, are significant if they are properly put into context. In contrast to the largely negative ideas of information seeking and use as a task or problem-oriented activity presented in the information science literature, Hardin’s argument is fundamentally positive. Marian Dörk et al. (2011) criticise the tendency to see information seeking as an endeavour rather than a positive phenomenon. Before this Jarkko Kari and Jenna Hartel (2007) began a discussion about the possible dimensions of ‘positive approaches to information science’. Their central argument of the notion of the ‘higher things’ of Kari and Hartel and the concept of ‘information flaneur’ of Dork et al. is to look beyond the traditional criticism of how people act with information and to perceive information interactions as a possibly positive phenomenon. Information seeking can make sense in many different ways and it might be too hasty to judge a person irrational or ineffective without properly understanding the context of their information-related activities.
Similarly to Hardin, in this book I do not attempt to make a contribution to the philosophical debate on the premises of knowing, knowledge or information, but rather to describe and conceptualise what is happening with information and knowing in contemporary society from a perspective that springs from observations and analysis of how people interact with information they find in the literature. Hardin’s premise is that all people act rationally according to their own assessment of what they are doing in their particular context of knowledge. People are not consciously foolish without a reason, even though much of the literature in information science and other disciplines is based on a premise that some parts of our behaviour are more or less irrational (and, to put it bluntly, plainly stupid) because the possible rationale of our irrationality is not considered a contextual matter. Librarians, teachers and parents are worried about the young people who are behaving irrationally by relying on dubious information sources. The young people should know better, they think, but in fact, it is librarians, teachers and parents, all of us really, who should know better.
What is it then that we should know better? First and foremost it would be useful to give greater emphasis to the fact that our everyday rationality is not always instrumental or even communicative. A course of action is considered rational if an individual perceives it to be sensible at the particular moment when he or she decides to take it. Following Hardin, even if people felt they were making an outright idiotic decision at the moment when they made it, my argument is that there is always a reason, a rationale, for acting foolishly. Making this assumption does not entail ‘userism’, as a form of uncritical and unreflective toadyism, or an extreme relativism (criticised by Suominen, 2007). It may be taken rather as a call for a need to be curious and serious about the practices of knowing and reasons for how and why we know and act as we do. The focus of the viewpoint is that if we attempt to understand how people behave in the present digital information culture, navigate the abundance of information and cope with the transience of resources, we have to see the rationale of the observed behaviours in their own contexts outside the traditional frame of institutionalised information seeking and rigid forms of instrumental or communicative rationality. Asking what people think about digital information services or why libraries are relevant or not does not provide answers to that question, nor does any approach that uses goal-oriented information seeking as an intrinsic yardstick of the analysis of other ways of knowing things. There must be another approach.
The essence of the predominant discourses of our time about the information society and the social web is the idea of liberation. We are no more confined by geographical or hierarchical distances. In contrast to earlier objective forms of knowledge, anyone can edit Wikipedia, microblogging has helped oppressed people to fight for their freedom and ICTs allow us to work and pursue our interests whenever and wherever we like. Lawrence Lessig (2004) writes about free culture. Volunteer programmers create impressive free and open software applications such as the GNU/Linux operating system, LibreOffice or Mozilla Firefox. Volunteers of the OpenStreetMap project created a complete up-to-date map of Haiti in a matter of days after the earthquake in 2010 (Lin, 2011). The ideals of freedom and liberation can be traced throughout the rhetoric of the information society and the social web. Esther Dyson, one of the leading figures of information ideology, wrote that the ‘net offers us a chance to take charge of our own lives and redefine our role as citizens of local communities and of a global society’ (1998: 14).
Despite the prevailing rhetoric of libertarianism, the information society is one of boundaries and the social web is a web of confines that divides as much as the virtual links connect us to each other. Despite heavy investment in information and knowledge management in organisations, the development of an inclusive information society and the lowering of digital divides on a societal level there is still a thicket of boundaries that limit our knowing (Hassan, 2008: 59–61; Swart and Harvey, 2011). People continue to prefer face-to-face communication, but in many ways the ICTs have lessened the barrier of geographical distance and made the differences of time the most significant problem of synchronous global communication. Europeans work late in the evenings to be able to have meetings with their American colleagues after they have arrived at their workplaces in New York or San Francisco in the morning. All European residents of the virtual world Second Life know that the de facto standard time of the world, Pacific Day Time, means that events scheduled at noon start hopelessly late in the evening. Similarly, Americans need to get up earlier to catch their European meetings and to work late to participate in the Asian ones. Besides obvious barriers such as time and place, there are also other more subtle barriers caused by cultural and social differences. We are not always as free to act as it may seem.
In the past information science research has attempted to explain various types of information boundaries, often conceptualised as information barriers. Different studies have identified a broad range of boundaries from personal to interpersonal, environmental and information-related. One of the early studies of success and failure in information seeking was conducted by Renata Tagliacozzo and Manfred Kochen (1970) in the context of known-item searches. They classified failures as collection failures (a book does not exist) and users’ failures (user failed to find a book). Later on, Yazdan Mansourian and Nigel Ford (2007) found that searchers themselves attribute their failures to the problems caused by internal barriers (e.g., ability or effort) and external barriers (e.g., luck or information not being available). In their study, the failures were caused by a number of individual obstacles such as inability to select appropriate search key words, inability to narrow down search terms, failures of search tools, unavailability of information, lack of skill and experience, perceived impossibility of finding out everything, lack of effort, lack of time and effort to filter the results, lack of patience, user errors, system errors and failure to recognise a failure or success. In another study, Reijo Savolainen and Jarkko Kari (2006) studied barriers in terms of gaps and gap bridging in web searching from a sense-making theory point of view. They found 11 gaps with three major explanations: problematic content of information, insufficient search competence and problems caused by the search environment. A major cause of what is thought to be failure in information seeking is often assumed to be related to the problems of defining and understanding the problem that triggers the information- seeking activity (Marchionini, 1997, 55). Anxiety has been found to be a source of failure in information seeking (Case, 2002: 106), but, paradoxically, a degree of stimulation has been observed to be necessary to cross the barriers of knowing (Case, 2002: 106; Kuhlthau, 1993). Marzena Swigon (2011) has recently reviewed a large corpus of literature on the barriers of knowing. She classifies the obstacles as personal characteristics, interpersonal factors, environmental limits and limits connected with information resources. Instead of barriers, she prefers the term ‘information limits’ to describe the ‘obstacles hindering, delaying or preventing access to information’ (Swigon, 2011: 366).
It is tempting to perceive boundaries as a primarily negative phenomenon, but in practice they play a very fundamental role in how we make sense of the world. Even if Swigon’s (2011) viewpoint is to see information limits as (negative) obstacles, calling barriers ‘limits’ can be useful. ‘Limit’ is a less negative term than the more common concept ‘information barrier’. Limits can also demarcate a positive place of isolation. Lawrence Grossberg refers to the notion of place using the concept of territory, which he borrows from Gilles Deleuze and Felix Guattari. He develops it further as a ‘context of lived reality’ that describes ‘a complex set of affective articulations and registers that constitute different ways of living in already socially determined locations, different possibilities of the forms and configurations of investment, emplacement and orientation, change and security’ (2010: 34). A place is an ‘expressive organization of socio-spatio-temporal investments’ that transforms ‘extensive space-time through intensive relations into a livable space-time’ (ibid.). Ultimately the boundaries constitute our everyday life as a territory of knowing that maps the possibilities of how we can know (2010: 242). According to Grossberg’s conceptualisation, the limits become a resource rather than a hindrance. The territory of knowing is a context for knowing things rather a prison of ideas.
The notion of everyday life as a territory of knowing is related to other spatial conceptualisations in information science. It forms a ‘round’ in which people live, according to Elfreda Chatman (1999). In contrast to the notions of information use environment (Taylor, 1991) and for instance various concepts of information ecology (Capurro, 1990; Davenport, 1997), the Grossberg-influenced territory of ordinary knowing emphasises the process of knowing. In addition, a bounded territory of knowing is primarily a personal space and only secondarily a social environment. It differs from the concepts of information horizon (Sonnenwald, 1999) and information source horizon (Savolainen and Kari, 2004) by its emphasis of the significance of explicable boundaries. Even if the spatial continuum of potential knowing extends beyond the boundaries, I argue that a closer look at the boundaries that shape the current horizon is useful precisely because the limits of the visible horizon are not always clear to us. We know they are there; we place them there; but we do not necessarily recognise them if we are being asked to do so, in much the same way as people tend to have difficulty in elaborating the taboos of everyday life, according to the classical observations of Mary Douglas (1966). The territory may appear as a horizon even if the boundaries limit what we see. It is a part of the system of categorisations that makes the world comprehensible for us. Eviatar Zerubavel (1991) described boundaries as ‘mental fences’, which help us to make sense of and conceptualise our life world. Geoffrey Bowker and Susan Star (2000) provide a well-known elaborate account of the depth of the consequences of erecting these classificatory boundaries. An intentional or unintentional categorisation, an act of delimiting a thing outside the boundaries, can change our life world, make it invisible, make it incomprehensible and change radically how we conceptualise and treat it.
Despite their seeming enormity and complexity, no boundary is inherently impermeable. In the digital environment, many of them are possibly more elastic than the physical boundaries they have replaced. We can lower boundaries, transform them and cooperate with others across them whether the barriers are physical, social or individual. Making the change is an effort, however, and something that helps us to bridge the difference. Star and Griesemer (1989) made an interesting observation when they studied the cooperation of amateurs and professionals in the early years of Berkeley’s Museum of Vertebrate Zoology at the beginning of the twentieth century. They could identify several different types of abstract and physical objects that helped the two groups to cooperate effectively. Star and Griesemer called these entities boundary objects. They described boundary objects as translation devices and argued that the shaping and maintenance of boundary objects is central to developing and maintaining coherence across communities. Boundary objects are a precondition for communication, cooperative work, and having and reaching mutual goals. Star and Griesemer identified four different types of boundary objects: repositories of things (‘ordered piles of objects’), ideal types (e.g., a diagram, an atlas), coincident boundaries (objects with the same boundaries but different contents (e.g., the ‘idea’ of the state of California and its meaning for different individuals) and standardized forms (e.g., forms and other devices for standardising working methods) (ibid.). Paraphrasing the soft systems theory of Peter Checkland (2000), a theory that uses the notion of system to describe the systemic dynamics of technological and human systems and their hybrids, the boundary objects may be seen to play a significant role in helping to cross systemic boundaries and tying related systems together to form larger systems.
The concept of boundary objects has been applied to different contexts, primarily in information systems and computer supported cooperative work research (Lee, 2007), but also in other research communities (e.g., Kuhn, 2002). This concept is not the only boundary-related theory or spatial metaphor used in information science research (e.g., Savolainen, 2009) or in other scholarly fields from education (Wertsch, 1983) to literature (Hayles, 1999) and anthropology (Zerubavel, 1991), but during the last two decades it seems that this particular concept has proven its usefulness in explaining communication and cooperation between communities in very different types of contexts. Studies have showed that different artefacts including visual representations (Henderson, 1991), cancer (Fujimura, 1992), technical standards, geographic information systems (Harvey and Chrisman, 1998), maps (Lin, 2011), activities (Macpherson et al., 2006) and documents (Østerlund, 2008) may function as boundary objects. The boundary objects are typically described to form ‘liminalities’ (thresholds) between communities, but some researchers (e.g. Giorgi and Redclift, 2000; Oppermann, 2011) have demonstrated the applicability of the concept also in the analysis of bounding discourses.
Researchers have made several distinctions between various types of boundary objects and boundaries. In addition to the four types of boundary objects discussed by Star and Griesemer (1989), Michael Briers and Wai Fong Chua (2001) introduced a visionary boundary object. Visionary boundary objects are conceptual and cannot be argued against. Institutionalised codes or ‘best practices’ illustrate such objects. Karin Garrety and Richard Badham (2000) distinguish between primary (technology) and secondary (physical and abstract) boundary objects. The latter enable communication between communities. Paul R. Carlile (2002) notes that communities need different types of boundary objects in altering situations and makes a distinction between syntactic (repositories), semantic (standardised form and methods) and pragmatic (models, maps) boundary objects.
Even if the term ‘boundary object’ may suggest a relative stability, the dynamism of boundary objects has been discussed in several studies. Drawing on Peter L. Berger and Thomas Luckmann, Uri Gal et al. (2004) emphasise the dynamic nature of boundary objects and their relation to social infrastructures. Peter Mambrey and Mike Robinson (1997) discuss the dynamics of boundary objects from the perspective of the changes that occur in the status of documents when they cross boundaries from one community to another. According to their observations, the changes that occurred in boundary crossings did not occur within the limits of the boundaries and vice versa. Despite their versatility in boundary crossings, Eswaran Subrahmanian et al. (2003) have showed that organisational changes disrupt existing common ground and consequently destroy existing boundary objects. New boundary crossings require new effort to establish new boundary objects. The observation seems to suggest that boundary objects are resilient to internal change, but less tolerant to changes in their bordering communities. Star perceives the emergence and assimilation of boundary objects as a cyclical process of infrastructural standardisation and the emergence of residual categories (Star, 2010).
Kathryn Henderson (1999) observed that related translatory objects – objects that help members in a community understand each other – also reside within communities. She called these objects ‘conscription devices’. Boland and Tenkasi (1995) developed the notion of internal and interfacial translatory devices further by introducing the concepts of ‘perspective taking’ and ‘perspective making’. Helena Karsten et al. (2001) discuss the combined conceptual apparatus and suggest that boundary objects and perspective taking are related to translation between communities while conscription devices and perspective making function within communities. Acts of perspective taking and perspective making emphasise the dynamic nature of negotiating translatory objects and the social organisation of knowledge in and between communities.
The notions of perspective taking and perspective making highlight a dimension of translation that Etienne Wenger (1998) emphasised. He stressed that besides objects, boundaries and connections across boundaries are matters of practice. As Charlotte Lee (2007) remarks, the idea has parallels in earlier boundary objects literature. In their seminal article, Star and Griesemer (1989) refer to an additional category of boundary crossing, ‘methods standardisation’, which is very similar to the notion of shared practices. According to Wenger (1998), boundaries are crossed and negotiated in an explicit and implicit process of brokering (mediation). The perceived significance of boundary making and negotiation has been discussed especially in terms of boundary practices (Vashist et al.: 2011), a notion introduced by Wenger (1998: 114–15). Instead of assuming an empathetically practice-oriented point of view, I chose to look at boundaries as ‘quasi-thingish’ objects and to focus on the implications and consequences of these appropriated objectifications and how they relate to knowing, instead of merely looking into the variety of ongoing practices once more.
Despite the acknowledged theoretical and practical benefits of the notion of boundary objects, it has a number of ambiguities and complexities. Boundary objects are not only objects of translation. Carlile (2002) emphasises that boundary objects can be used jointly to transform knowledge by proposing alternative views. Gal, Yoo and Boland (2004) show how boundary objects function as resources to form and express social identities. Their transformative and often purposeful nature is also highlighted by the conceptualisation of bounding entities as discourses (e.g., Oppermann, 2011) instead of communities. The transformative and political dimension of boundary objects is also emphasised by the conceptualisation of boundary objects as authored entities. They are attributable to individuals, collectives or an essentially non-identifiable emergent form of authorship (Huvila, 2012). The authorship of a boundary object is different from the authorship of the particular entity that serves as a boundary object. Besides being authored, boundary objects can be seen also as political entities that express antagonistic tendencies between bordering communities (Huvila, 2011b). The common observation of these studies is that despite their seemingly consensual nature, boundary objects are transformative devices and more specifically instruments for advancing agendas and accommodating competing opinions. These agendas can be our own or be set by others who are helping us to cross the boundaries as coaches and teachers by ourselves or as experts and specialists by breaking the fence on our behalf.
The usefulness of boundary objects in the context of this book is based on the usefulness of boundary objects as conceptual devices for exposing and understanding various types of boundaries and boundary crossings between communities and many other types of entities. The notion helps one question the existence and implications of obvious and invisible parameters that affect what people choose to do and not to do. My point of departure, which is slightly different from Star and Griesemer’s original notion, is that boundary objects can be helpful in exposing and understanding the liberating and deeply anti-boundary assumptions made in the context of the information society and social web debates. A simplistic view of boundary objects would be to perceive ICTs and more specifically the technologies of the social web as principal boundary spanners and enablers of the libertarian boundlessness endorsed by the proponents of the information society and social web. The approach assumed in the present volume is rather different, however. ICTs are undoubtedly major catalysts and have the capability to cross major impediments, but at the same time they catalyse the emergence and strengthening of a multitude of new boundaries, which are no less meaningful than those in the past.
A central argument of this book is that the ways we try to be economic with our information seeking and use are contextual to our personal territories of knowing. The self-perceived rationality of an individual is influenced and delimited by the existence of visible and invisible boundaries. Information seeking can be fundamentally conceptualised as a boundary-crossing activity. For different reasons, we try to bend and cross the borders of our understanding and capabilities in order to gain new insights and perform things we have been unable to do before. Helping others to find and use information is another type of boundary activity, which incorporates external and internal transformations of the territory that forms the context of our ordinary knowing. We can try to help others to cross their boundaries either together with them as peers, or as ‘experts’ with a different horizon of capabilities and limitations. The help can be given by teaching, guiding or coaching others to cross a particular boundary or by assisting them to learn where the boundaries are and how to cross them in general. As Boland and Tenkasi (1995) have suggested, the crossings require us to have individual and communal perspectives.
Keeping the metaphor of economic boundary crossings in mind, our next step is to question some of the major assumptions related to our contemporary information environment and its limits. Acknowledging others’ observations of the major boundaries to seeking information and knowing things, the information environment and its infrastructures (Chapter 4), the social context of knowing (Chapter 5), ourselves (Chapter 6) and the information itself (Chapter 7), this book investigates the borders of the economy of our knowing. What are the things that we can find? What are the major barriers that constrain our ways of becoming informed and understanding? How are we are supposed to overcome those obstacles with the help of various types of boundary objects? It is necessary to assume a broad understanding of what a boundary object can be and how it can behave, essentially pushing the limits of earlier definitions in the direction of dynamic boundary negotiating artefacts. A boundary crossing is related not only to a mere boundary object, but also to the nature of the negotiation of boundaries, whether it is internal perspective making or external perspective taking (Karsten et al., 2001). The boundaries and the bounded soft system are always boundaries and systems ‘at hand’. They are related to the current situation. The understanding is similarly broad as the perception of the economics of ordinary knowing. As Charlotte Lee (2007) suggests, boundary objects should not be treated as black boxes. All boundary objects are complex and inseparable from social negotiation processes within and between communities and, therefore, empathetically purposeful constructs. The transformative capability of boundary objects means they not only passively bridge but also actively negotiate perceptual and practical differences between communities at the level of the interaction of discourses.