ABSTRACT
By all accounts, the trend toward using information and communication technologies (ICTs) for the global expansion of organizations is growing swiftly. Many organizations have moved into global operations through captive offshoring (offshoring work to a branch office or wholly owned subsidiary in another country). In fact, a 2005 report by the McKinsey Global Institute has predicted that, by the decade’s end, US companies will employ more than 2.3 million offshore knowledge-intensive workers (Farrell & Rosenfeld, 2005). From the perspective of organizational communication research, increases in the captive offshoring of knowledgeintensive tasks that rely heavily on the use of ICTs are hardly surprising. Recent studies have consistently shown that technologies enable the global distribution of knowledge-intensive work by providing individuals access to crucial information (Sakthivel, 2005; Tractinsky & Jarvenpaa, 1995), improving knowledge transfer capabilities among corporate units in different countries (Bhagat, Kedia, Harveston, & Triandis, 2002; Werner, 2002), and permitting new organizational forms that more effectively handle spatial and temporal dispersion (Boudreau, Loch, Robey, & Straub, 1998; Monge & Fulk, 1999). A good number of authors who study organizations in which people in different parts of the world are working on highly interdependent tasks have concluded that global knowledge sharing is fraught with difficulties (Jarvenpaa & Leidner, 1999; Maznevski & Chudoba, 2000). The problem, as they often argue, is circular in nature. ICT infrastructure, such as telephone lines and the Internet, and software applications that run on them, like groupware, group decision support systems, and even complex computer simulation technologies, enable organizations to internationalize; they create the conditions that make it possible to offshore interdependent knowledge-intensive tasks. But those same technologies also pose major obstacles for global knowledge sharing. This circular problem is often implicitly understood in terms provided by the transmission metaphor of communication. ICTs create an easily accessible channel for information to flow through space and time. Channels are
always limited in their carrying capacity. Only a small extract from the entire body of knowledge necessary to perform a task can flow at a given time. Further, as the knowledge leaves its source (a sender) it becomes decontextualized and uprooted from the practical space that defines the contours of its meaning, and can best be termed “information.” Consequently, when such information arrives at its terminus (a receiver) it appears only in bits and pieces, and is abstracted from its original context and key referents. Because channels are limited in the types of content they can transmit, some amount of information must be left behind. Researchers of computer-mediated communication have long suggested that because the channels created by ICTs support text and voice predominantly, non-verbal cues are often omitted from transmission (Cornelius & Boos, 2003; Walther, 1994). The omission of non-verbal cues has been shown to impede mutual understanding by reducing personal self-disclosure and increasing conflict (Cramton, 2001; Hinds & Weisband, 2003). Further, channel limitations can preclude the transfer of entire spaces of knowledge. Knowledge gained from touch, smell, taste, or lived experience cannot be transferred through most ICTs available for use in today’s global organizations. A good deal of knowledge is acquired through these sensory stimuli, and when senders attempt to translate this knowledge into text and voice, receivers are not often able to acquire the contextual understanding necessary to use that knowledge in meaningful ways. Limitations of the transmission metaphor also direct our attention to problems that arise when knowledge is, at one time, encoded into a channel by someone with a particular constellation of socio-cultural understandings and later decoded by someone who does not share those same orientations. A persistent concern is that a sender from one culture might encode knowledge in a way that will not be decodable by someone from another culture (Gibson & Manuel, 2003; Zakaria, Amelinckw, & Wilemon, 2004). If such cultural misalignment occurs, knowledge is not transferred. A related concern is that culture shapes the way individuals cognitively process information (Markus & Kitayama, 1991; Nisbett, Peng, Choi, & Norenzayan, 2001). As a consequence, individuals who encode or decode information will sample it in different ways, assign different weights to what is sampled, and make distinct associations between different pieces of knowledge (Bhagat et al., 2002). The result is that the entirety of knowledge sent from one shore via ICTs may not be received on the other. Such a practically grounded and communicatively situated conceptualization helps us to explain why so many problems arise when knowledge is extracted from the context that defines it, forced into an ICT that mutates it, and decoded by someone who has neither the practical nor cultural experience to apprehend it. It is striking that, despite the subtlety with which these accounts are constructed, few researchers seem to consider that people who are attempt-
ing to transfer knowledge and share it with others may not even agree on what knowledge is. Imagine the following (and heretofore entirely unrealistic) scenario. Person A wants to transmit a dog (metaphor for knowledge) to person B. Person A finds a teleporter that breaks the dog down into subatomic particles, which are sent across space and time and reassembled on the other side. He puts the dog in the transporter and presses the “send” button. In our current conceptualization of knowledge transfer, we say that a problem of transmission occurs if person B receives the dog but it has only three legs, it has no tail, or it has two heads. Indeed, our theories can explain, in intricate and varied ways, why the dog that went into the transporter is not as complete as the dog that came out on the other end. But what if person B receives a monkey? She calls person A and says, “I thought you were going to send a dog, but I got a monkey.” Person A responds, “I did send a dog, check it again.” She checks. The monkey is the same weight, the same size, and the same color as the dog person A says he sent – but it’s a monkey. In this scenario, the problem is not with the transfer process; no content was atrophied in the transmission. The problem is that persons A and B do not agree on what constitutes a dog. If they continue to have different ideas of what a dog is, no amount of refinement or improvement in the transfer process that eliminates channel noise (e.g., more advanced technology, more description of the dog’s characteristics, etc.) will help. Where one person sees a dog, the other person will see a monkey. The knowledge-sharing problem that this metaphor raises is one of semantic incompatibility. It comes as no shock to most that people from different cultures often have markedly distinct understandings of concepts like family, fidelity, faith, and friendship. So why wouldn’t people from different cultures also have divergent understandings about what counts as knowledge? If people have unique understandings of what knowledge is, an important attending concern is that they will have different perceptions of where knowledge lies. Just as dogs lie on the ground and monkeys (the arboreal ones, at least) lie on branches, thing X, which is considered to be knowledge by person A, is likely to lie in a different place than thing Y, which is understood to be knowledge by person B. The issue of where knowledge lies is of great importance for theories of knowledge sharing in global organizations, particularly those engaged in offshoring. In most offshoring arrangements (even those that are captive), there is little direct interpersonal exchange of knowledge. Instead, people on one shore simply point each other to locations where they are likely to find the knowledge they need to do their tasks (Dibbern, Goles, Hirschheim, & Jayatilaka, 2004). Research shows that people variously point to images or models (Leonardi & Bailey, 2008), knowledge management systems (Carmel & Agarwal, 2002), or written documentation and contracts (Gopal, Sivaramakrishnan, Krishnan, & Mukhopadhyay, 2003) as loci for knowledge. But if people from different cultures have distinct conceptualizations of
what counts as knowledge, and, consequently, unique beliefs about where knowledge lies, one person may look in a location designated by another and find nothing there that she considers knowledge. Or worse, when pointed toward a particular location by a colleague on a different shore, she might grow frustrated, thinking “of course there is no knowledge here,” and form disparaging opinions about that colleague’s competence, which, consequently, can cause her to question the soundness of the organization’s offshoring plan altogether. The goal of this chapter is to explore how cultural differences in perceptions of where knowledge lies can have adverse effects on global knowledge sharing in organizations. Rather than take for granted that people on different shores each conceptualize knowledge similarly and look for it in the same places, I begin with the assumption that cultural differences may compel workers in the same company to think of knowledge in distinct ways and to look for it in diverse places. This approach focuses attention on how people within the same organization come to think about knowledge in varied ways. Drawing on empirical examples from a large automotive engineering firm, I explore how different conceptualizations held by engineers in the US, Mexico, and India about what knowledge is and where it lies stifled sharing, impeded learning, and led to general animosity among members of the organization who needed to work collaboratively to design and test vehicles. I conclude this chapter by discussing the implications of these findings for theories of technology, knowledge sharing, and organizational communication.
