Theory of the firm: theortical perspectives from an innovation point of view.

ABSTRACT

This paper explores the nature of a firm from an innovation perspective. An organization may be thought of as a collection of networked resources. Developments in information technology enable increased interactions between these resources for information and knowledge generation. Because of these increased interactions, the value of each resource is likely enhanced. The possibility of generating new knowledge and information resources as a result of these interactions would affect the structure which organizations adopt to manage environmental uncertainty and to effect innovation. The basic question we explore is whether the need for innovation is the fundamental reason for firms coming into existence in this era of information economy.

1. INTRODUCTION

Many of the phenomena that we observe in economic and social organizational behavior might be well understood if we take into account that the actors involved have limited capability to understand fully the cause and effect of the phenomena. The behavior of the actors under this "limited rationality" would sensitize organizational behavior with respect to changes in actors' information processing capacity. This limitation in the information processing capability of human actors has been thought as the basis of the existence, structure, and functioning of economic organizations such as firms (e.g. Simon 1973). Literature in organization performance and behavior are closely linked to the view that an organization can be considered as an information processing system (Simon, 1973; Galbraith 1977; Tushman and Nadler, 1978).

This paper will focus on the effect of the advances in information technology on organizational resources and on the structure an organization adopts to facilitate a continuous process of innovation. We argue that organizational capabilities with respect to information processing and decision making are enhanced by information technology advances. The behavior of each member in an organization is likely to be influenced more by others because of the increased possibility of interactions for information and knowledge sharing. The increased capability of the members for interactions has implications on the evolution of structures which an organization adopts to manage environmental uncertainty and innovation.

Schumpeter's idea of innovation is essentially "carrying out new combinations" and in this respect, innovation is a broad concept (1934). The importance of firm's resources in carrying out these new combinations was pointed to by Penrose (1959). Her main point was that different combinations of existing resources and their deployments are ultimate issues of organization. The rapid advance of information technology has increased the possibilities of combining firm resources within and combining resources across firms and national boundaries. This is because IT advances have caused a dramatic increase and transportability of information (Sampler, 1998). Managing intangible resources like knowledge and information can have greater impetus for economic growth than tangible resources do, because intangible resources can be leveraged to add value to tangible input resources (Teece et al., 1997). Given the emerging and increasing concerns regarding advanced IT and its impacts on organizations, this theoretical study intends to explore the dynamics of the following research questions.

1. How do the advances in IT and communication technology affect the structure that an organization adopts to manage innovation?

2. What resources are critical for the organization for managing in an information economy? How do the characteristics of these resources affect the organization structure that best suits innovation?

3. What are the determinants of a firm level organization in contrast to a market level organization where the resources to be coordinated and organized are predominantly information and knowledge?

2. LITERATURE REVIEW

Organizational information processing is traditionally known under various characterizations as management information systems, intelligence systems, decision support systems, communication systems, and administrative control systems. We use the definition of organizational information processing according to Huber (1982) as "the carrying out of the ad hoc or prescribed activities through which organizational members assess, distribute, alter or use organizationally relevant information". The term "members" is intended to be defined broadly as an individual, a work group, a department or even a corporate division. The members may be called 'nodes' or 'units' in an information network or structure context.

Until the advent of modern information technology, an organization's structure was a relatively inflexible hierarchical channel for information flows (Norton and Lester, 1996). Multiple units organized in a hierarchical structure reduce complexity as well as simplify coordination of large systems with multiple heterogeneous tasks (Simon, 1981). Hierarchy primarily serves two functions; one it gives a pyramid of authority, and second, it decomposes an organization into departments or nodes that generally reflect the pattern of communication (Simon, 1983). The roles of individuals and groups are clearly delineated in traditional hierarchies. The formal and informal communication channels in an organization are subjected to the vagaries of the working relationships between the units in the hierarchy and this is instrumental in differing levels of communication effectiveness across organizations having identical hierarchical structure and environment. This problem of human hierarchy was pointed out by Simon (1981). "The real flesh and blood organization has many interpart relations other than the lines of formal authority and in many instances the formal hierarchy exists only on paper" (Simon 1981, p. 197).

The problem of the ideal form of organizational structure has been a vexing question for many reasons. For one, hierarchical forms of organizational structure have been the tradition. But, with the rise in information technology traditional hierarchical structures are increasingly replaced by matrix, hybrid, network and cellular structured organizations. Alternatives to hierarchical structure are proposed as a solution in an environment where firms are even going beyond the usual mantra of customization and now embarking on the process of continuous innovation. Without the developments in IT, coordination in these structures would have been next to impossible. "Information technology has begun to revolutionize organizational design by providing alternatives to hierarchy as the primary means of coordination" (Nadler 1992, p6).

The question is: Does the call for alternative forms of organization really address the basic problem of organization? The basic problem of an organization is information processing for decision making. What emanates from a node could range from merely conveying a piece of information to another node or making strategic product-market choices. Increasingly organizations are called upon to operate in highly dynamic and unpredictable markets and environments. Goals and activities facing each individual node of the organization are far less clear. Decisions are responded to by non-linear feed back which create ambiguity and uncertainty. In this scenario, any fixed organizational structure whether hierarchical or other forms of structure would not work. Continuous adaptation to the changing environment requires continuous reorganization of the organization decision structure. This continuous reorganization is carried out by a combination of exploration of new possibilities which might include experimentation, search and innovation and exploitation of old certainties which might include refinement, selection, efficiency (March, 1991). What organization structures are suited for such a complex and continuously innovating system performance is the subject we turn to next

2.1 Emergent hierarchy in Organization Structure

As Simon (1962) suggested, complex systems tend to organize hierarchically, since it is the adaptive form for finite intelligence to assume in the face of complexity. In this type of hierarchy, stable intermediate forms or subsystems are organized in such a way that interactions between the intermediate forms are weak whereas interaction within any particular stable form is strong. These stable forms provide new building blocks for further construction of the hierarchy. Stable forms in an organization context can be nodes or even clusters of nodes responsible for any particular activity, like a department or a product development team. For example, competencies (as possessed by a product development team) for instance may act as catalysts for new product development efforts by an organization. These stable forms can be engineered and pre-designed in an organization context if the environmental "uncertainty" is expected to follow in a certain predictable manner. For systems operating under greater flux, it might be the case that managers find it impossible to predict, design, and control all the activities. Investment in research and development to develop competencies carry a lot of risk and uncertainty when the environment is turbulent. Under these conditions of greater complexity, hierarchical structures should emerge as dictated by the adaptability requirements. These are hierarchical because if activities are completely random, complex tasks will take an infinite amount of time to complete and therefore a hierarchical pattern has to emerge rather than be imposed (Morgan, 1996, p416).

It should not be construed that these emergent hierarchies within an organization are beyond the purview and control of managers. Strategic managers may be handicapped by their inability to plan for organization wide adaptation programs because non-linear responses are difficult to predict and control. The role of managers would be to establish and modify the direction and the boundaries within which self-organized solutions can evolve (Meyer et al. 1998). Managers will set constraints upon local actions, observe outcomes, and tune the system by altering the constraints, all the while in effect changing the way the system which they are managing self organizes with increasing or decreasing degree of complexity.

The above analysis may indicate that as organizations get complex or as they operate in complex environments, managers will be devising strategies with more and more unpredictable outcomes. Or, to state the other side of the same coin, it is akin to a situation where managers loose more and more strategic leverage as the complexity increases. Strategic hits will diminish and strategic misses proliferate. Well-intentioned trial and error tinkering with the boundary conditions and adding or changing agents of the organization might find the strategist unwittingly surveying the heap of a catastrophic avalanche that resulted from his action. Radical innovations (useful) might be well due to evolutionary mistake rather than from conscious decisions. Social and economic systems like organizations do not have the luxury to recover from a fatal blow unlike biological systems where the reproductive capacity of unaffected organisms ensures that the species recover from the catastrophe. In summary, organizational theory development is yet to capture the outcomes that are likely to occur as managers tinker with the organizational structure and strategy (Anderson, 1999). In essence, the ability of the top management to organize the whole organization hierarchically will diminish. Emergent hierarchy at lower system level would be the norm.

2.2 Intelligent Subsystems of Organizations

The difficulty of the strategist to manage a complex organization may not be much of a practical concern if the subsystems that comprise the organization can look for evolutionary combinations within the organization or from outside. It is risky for CEO (organizational strategist) to look for evolutionary combinations. But, for the lower subsystems of the organizational hierarchy, the risk will be lower if they align with other subsystems for the pursuit of goals that may even be different from the organizational purposes. This is because lower subsystems can spot complimentary linkages with other potential partners much more efficiently than they would if they were controlled by the strategist at the top of the organization hierarchy. This is similar to the concept of decentralized information processing. The more information processing power and intelligence these lower subsystems are endowed with, the more strategic flexibility the parent organization will have. As the independence of the strategies of each subsystem increases, it would progressively be the case that the parent organization will be devoid of any major influence on the subsystems. Different business units of the organization will pursue different local goals with progressively lesser connection and complimentarity with other business units. However, this sort of evolution is only possible if these subsystems are given the administrative freedom to optimize at their own level.

The evolution of organizations with increased likelihood of divorce by the lower subsystems may not have a biological analogue. But this sort of development has to be anticipated because of the following reasons: 1) organizational subsystems are knowledge structures and 2) information technology advances increase the likelihood of strong alliance of potential complimentary partners (subsystems or knowledge structures) for creations of a new idea or innovation. These ideas are explained further below.

The function of the firm to integrate inputs and diverse specialized areas of knowledge to perform a productive task is acknowledged in literature (Grant, 1996a). Firm competencies are a result of combination of input and knowledge based resources that exist in a hierarchy of integration (Grant, 1996b). Firms are also viewed as 'clusters' of input and knowledge based resources (Galunic and Rodan, 1998). Thus, it is apparent that we can characterize the hierarchy that we discussed in the previous section is fundamentally related to how knowledge and information is integrated in a hierarchical fashion within the firm. But still the fundamental question is why the subsystems of the organization are willing to pursue their own "illicit" relationships seeking out complimentary combinations? The answer might lie in the theory of transaction cost analysis itself. The fact that these subsystems were initially organized under the firm in the first place because the organization would have incurred a higher transaction cost if these subsystems were outside the firm. To make this argument an economically viable one, transaction cost analysis should have also considered the opportunity cost of the subsystem when it allowed itself to be organized under the firm. Once, this opportunity cost adjustment is allowed explicitly in the cost calculus, we can have a scenario where, if an incentive is allowed for a subsystem to divorce itself from the firm, then firm level organization would give way to a market level organization of this subsystem. The incentive in this case would be an innovation possibility for the subsystem to exploit by combining with other knowledge systems outside the firm.

In other words, if a knowledge subsystem or a resource could be better combined with other complimentary knowledge subsystem or resources, then that possibility has to be accounted for when we consider the emergence of organization structures. The original organization may disappear in the process of this strategic realignment of subsystems under its control. It is not very infrequent that we see in the market place that some organization is a target of take over by predators, precisely because some of the competencies of the target could be better utilized by the acquirer. It is also observed that in many knowledge intensive firms, the departure of any particular partner usually takes away along with him/her in entirety particular type of competency possessed by the firm. All these events should be viewed as cascades of changes that affect organizations that characterize the existence as a dynamic equilibrium. The changes caused to the organization in this way are through a type of subsystem level dynamics.

2.3 Resources for Innovation: Information and Knowledge

The use of information resources by firms has assumed a central role in our economy. Much of the wealth that is created now is due to technical progress and innovations.. The characteristics of information--knowledge specificity and time specificity (Sampler, 1998)--distinguish them uniquely from other resources. The specificity of knowledge is similar to the concept of tacit knowledge. Tacitness generally describes the extent to which knowledge is not codifiable (Galunic and Rodan, 1998). Tacit knowledge is personal, context specific, and therefore hard to formalize and communicate whereas explicit or codifable knowledge is transmittable in formal and systematic language (Nonaka and Takeuchi, 1995). Furthermore, intangibles like specific knowledge is expensive to transfer across because it cannot be easily aggregated meaningfully (Hayek, 1945). Although it may be difficult for the firm to extract significant value from any single stock of information/knowledge created in isolation, an ongoing information system for processing of a cumulative and dynamic knowledge could be a source of competitive advantage. This can create a strong "informational barrier to entry" (Glazer, 1991) against other organizations that might want to access the same knowledge base. It so happens that when the organization structures itself hierarchically, the subsystems of knowledge structures that emerge will have tacit components. Once the tacitness becomes highly prized in the market place we might observe the sort of subsystem level dynamics as discussed above. Advances in information technology increases the likelihood that the tacit subsystems or resources are able to contribute to innovations within the firm, because increased possibility of interactions and combinations of resources through the information system within the organization should generate more knowledge resources, but equally important would be the possibilities that might open up for this subsystem to leverage its competencies by strategic realignment with subsystems outside the firm. This will alter the constitution of an organization, since the structure of an organization's information system is closely linked to organizational change (Brynjolfsson and Mendelson, 1993). Imposition of hierarchy from top management in information and knowledge processing would not work, as coordination by such a hierarchy will be inefficient for managing tacit resources (Grant, 1996b). A full investigation of the various possibilities would answer the research questions posed in the beginning.

3. A MODEL OF KNOWLEDGE EXCHANGE BETWEEN TWO NODES

To understand the evolution of organizational structures that can manage knowledge and information resources, we examine closely the mechanics of the interaction between two tacit resources. The increasing importance of knowledge resources in the organization of resources within the firm would motivate us to explore further the characteristics of these resources. The description of tacit knowledge according to Nonaka and Takeuchi (1995) has been termed as loosely formulated and hence overlooking the finer variants of tacit knowledge (Boisot, 1998, p 56). It is difficult to know all the nuances of the context of the originating knowledge and even if it is known, the cost of codifying can be tremendous to incorporate all the nuances and hence a full identification of the tacitness may be very costly. Thus a particular knowledge can be made 'explicit' for a node within a firm, but the same kind of 'explicitation' may still be inadequate to a node outside the firm which is not conversant with the context of the source knowledge or the history of the firm. The economics of the codification is thus dependent upon the context of the originating knowledge and the context of use of this codified knowledge. How this context dependent knowledge interacts is the process we want to understand.

For simplicity of exposition, if we consider the firm as the interaction of two nodes that exchange information and knowledge, the following representation shown in figure 1 is suggested.

[FIGURE 1 OMITTED]

In the above, node A and node B have different contexts for holding knowledge and exchange of information. For example, both the nodes occupy different organizational contexts of responsibility. All the knowledge required for exchange can be broadly divided into two groups, one which is context dependent and the other which is not. We posit that this context free knowledge is easily codified (if needed) and available in usable form for all potential users. Thus the components EA2 and EB2 of nodes A and B can be made 'explicit' in its fullest sense. For an example, knowledge based on general physical principles like the 'resistivity' or 'conductance' of materials can be easily communicated without the fear of contextual subjectivity. Components EA1 and EB1 are also explicit in a sense and readily communicable, but they have significant context related information. An example could be the exchange of work related knowledge for a simple shop floor operation between two supervisors. The cost of communication will be reduced considerably, if the context of A and B are identical in this case of exchange. Components TA2 and TB2 are the 'tacit' components which are again context dependent. Cost of communication will not reduce considerably even when the contexts are made identical, because the major part of the cost is the effort for making the mental models of A and B into a codified form. These mental models are quite unique even in similar contexts and hence codifying cost is highly specific with respect to the holder of the knowledge. Sharing of knowledge between two highly skilled researchers having competency in different areas in a research lab is an instance of this (e.g. physicist vs. metallurgist in a weapons lab).

Components TA1 and TB1 are the type similar to the personal knowledge of Polanyi (1966). They remain elusive and uncodifiable. They mediate the formation of all knowledge and hence form a very interesting component. In our discussions here, a node is 'tacit' if it possesses at least one tacit component above and 'explicit' if it has only the explicit components.

The exchange of knowledge and the consequent new knowledge creation is a dynamic interplay between the components of each node's knowledge. The exchange process will involve decisions whether to convert the tacit component into explicit form before exchange because explicitness increases the likelihood of detection of new resource combinatory opportunities (Galunic and Rodan, 1998). Knowledge that is difficult to codify cannot be easily explicated and hence would be difficult to be combined with other knowledge in the firm (Teece, 1981). If codification preserves the original context specific information, then the new codified knowledge can be moved within and across the firms and applied to different contexts and making use of the full know how. There is no need for the existing knowledge to be codified preserving hundred percent of the original information. Codification can stop when the marginal cost of codifying exceeds the marginal benefit as a result of codification.

An alternative to codification and transfer is to move the holder of the tacit knowledge. Transferring people possessing tacit knowledge to socialize with others to inspire new knowledge creation is suggested by (Nonaka and Takeuchi, 1995). Some authors go further to stress that the key to "tapping into and unlocking" individual tacit knowledge requires a "highly interactive social process between people in a co-located face-to-face work environment" (Dan Holtshouse, 1998 p. 279). Co-location is needed because much of the tacit knowledge that is shared and exchanged is accomplished through close interaction. Discovery of tacit resources is difficult because this involves interaction and experiencing with people having tacit knowledge and on most occasions this involves interaction of people across different competencies and functions. All different ways of transfer have their own cost and benefit dimensions.

The mediating role of the context imparts 'relativeness' to tacit and explicit knowledge. Explicitness is achieved by codifying the original information by incorporating the full information and information context of the source knowledge. If we assume that it is almost impossible to know all the circumstances that surround a particular event of knowledge creation, then the complete codification into explicitness is impossible to achieve. A second assumption is to consider the second component of 'tacitness' (TA2 and TB2) as being 'explicit' from the point of view of the holder of the information. The holder of this component has a relatively clear idea of the nature and the attributes of this tacit knowledge. This means that if an 'explicit' knowledge comes from outside, this 'tacit' source can use that explicit information like an explicit versus explicit knowledge interaction process. However, this explicit vs. explicit interaction is not similar to the simple interaction of two bodies of explicit knowledge envisaged by Nonaka and Takeuchi (1995), but a much richer concept where the 'explicit' vs. 'explicit' interaction is from the perspective of the 'tacit' information holder. Also the 'explicitness' of a particular piece of knowledge as viewed by the source and the user may be different, as there may be different interpretations possible as mediated by the context. However, the motive behind all 'explicitation' efforts is to make the difference between these two interpretations as negligible as possible, but with usual cost-benefit analysis in mind.

3.1 Determinants of The Exchange Process Between Nodes

Nonaka and Takeuchi (1995) contend that the key to organizational knowledge creation is the conversion of 'tacit' knowledge into 'explicit' knowledge. They also envisage other secondary conversions between 'tacit' to 'tacit', 'explicit' to 'explicit' and 'explicit' to 'tacit'. According to these authors, the key to innovation is the process of generating knowledge within the firm and thereby the key determinant for innovation is the ability of the firm to transform the tacit knowledge within the firm into explicit knowledge that can be shared and combined by all users. The approach of these authors is to describe the process of organizational knowledge creation (like conversion of tacit to explicit, etc), but here we go one step further to outline the reasons for one process to be preferred over another.

We start with the mechanics of the transfer of codified knowledge between a tacit source node and a recipient node which for simplicity we assume has only explicit knowledge. We assume that the tacit knowledge of the source node can be made explicit by identifying the maximum possible 'n' dimensions or attributes of the tacit knowledge. The percentage of these 'n' dimensions which are made explicit can be called the degree of codification. The tacit source can choose the degree of codification for a particular recipient node or a set of nodes depending on the marginal cost benefit analysis. The recipient nodes will first value the codified knowledge from the source by seeing the value creation possibilities by interacting and combining with its own resources. Each recipient node might require only a particular combination of source attributes for combining with its own set of explicit attributes. Since the source node cannot know the optimal degree of codification before the exchange process (ex ante), we assume that the source node will incrementally make its attribute explicit till the marginal benefit of 'explicitation' equals the marginal cost of codification. Thus after codification, certain attributes or combination of attributes might remain useless for the recipient nodes. The optimal degree of codification will be determined by the extent to which the recipient nodes value each particular attribute or combination of attributes of the source node. The cost of exchange consists largely of the cost of codification and to some degree the cost for transfer of knowledge.

An alternative exchange process would be the tacit node serving as the information processor for other explicit nodes. In this mechanism, the tacit node will not codify its own knowledge, but it processes its knowledge creation process with both tacit knowledge and the explicit knowledge of the recipient nodes. Even if the information processing cost is equal to the cost of codification in the previous case described above, the chances of identifying superior combinations are more in the second because the tacit node works under the full information set of all the possible combinations of attributes of the tacit node and the recipient nodes. Thus if the tacit node has adequate information processing power, then transfer of knowledge from explicit source to tacit source should be made for processing. For an analogy, product development teams of a firm frequently incorporate the 'explicit' knowledge about customers. Imagine the unfeasibility of asking the shop floor supervisor to 'put together a product', although he or she might be the closer to customer and hence knows the needs more than the development team, but most often does not know how to meet the needs efficiently even if given all the requested information. This leads to the following proposition.

P1. A tacit knowledge node will be a more efficient processor for creating innovation in a knowledge exchange with an explicit node than an explicit knowledge node which will process codified tacit knowledge transferred from the tacit knowledge node.

This advantage of the efficiency of the tacit node has many implications. We discussed above the advantage of designating the tacit node as the prime processor and the disadvantage of making all nodes into explicit. For some tacit nodes to function as processor, the tacit knowledge of other interacting processors has to be made explicit. Proposition 1 implies a shift in focus from 'explication' all sources of knowledge because the cost of this would far outweigh the benefits. Compared to this, the same benefits or more can be achieved under a situation where the tacit node can be allowed to receive and process explicit knowledge from outside which is already made available for different reasons. The tacit node will have economies of scale as the more explicit information it has processed, the more efficiently it can process future relevant information in a sense increasing the 'discriminatory' power to enable the node to identify the useful ones from the useless. Also, by virtue of having processed more information, the tacit node has internalized a large amount of knowledge and hence 'tacitized' body of knowledge. This will increase the primacy of this tacit node as the center for efficient processing of outside information and creation of new innovation possibilities or value by combining with its own existing knowledge and largely tacit knowledge from the perspective of outside knowledge sources. This efficiency of the tacit nodes renders themselves as the nucleus of innovation or value creation process. This may be summarized by proposition 2.

P2. Efficiency of the tacit nodes in creating value or innovations will result in a situation where increasingly more innovations will be centered on the efficient tacit notes. This specialization in processing a particular type of information is largely due to the cost differential of processing own tacit information versus processing of the same tacit knowledge by an outside node.

Transfer between two explicit nodes will be the simplest case in terms of processing cost to identify the valuable combinations. The other extreme case, the transfer between two tacit nodes, will be the most costly. The actual mechanism of exchange would consist of some degree of 'excplicitation' of knowledge of each of the nodes to begin with and with this initial knowledge of each other, these nodes can proceed in an iterative way to further identify the areas of their own knowledge relevant to value creation and proceed forward to communicate this with the other node/nodes. We can imagine the exchange and creation of knowledge between two tacit sources as similar to the conversation between two strangers from two countries getting to know each other and about their respective countries. The start of the conversation would be based on common knowledge and then the conversation may proceed iteratively to further explore the tacit details of each other with cues taken from the shared or explicit common knowledge. If the two conversationalists have come together for a specific purpose based on the promise shown by the initial shared knowledge, then the value creating collaboration of these two individuals will further 'explicitize' the tacit knowledge to some degree in the process of identification of the value creation possibilities. The optimal result will be achieved by 'explicitation' the knowledge of the person having the least tacit knowledge fully (in a situation where these two have come together for a specific purpose, and based on arguments of P1), and allowing the other to process this explicit knowledge as discussed above. However the most likely result is a case where both participants have some degree of 'tacitness' left over. This problem leads us to the following proposition.

P3. Although knowledge is created as a result of the exchange of information between tacit knowledge nodes, efficient identification of the innovation or value creation possibilities require 'explicitation' of the least tacit node so that the resultant explicit knowledge can be processed by the tacit node serving as an specialist node.

As a tacit node process more and more explicit information from outside, the efficiency of processing should ideally go up because of the 'discriminatory power'. But due to the inherent information processing limitations of a node, as the processing power may not be able to be increased in proportion to the rate at which new information has to be processed, there is a decrease in the ability of the node to integrate new explicit information from the environment. This is because in a real life situation, the tacit node might be organized actually as a research division in a firm or as a collection of experts in a consultant firm where knowledge processing is the primary function. New information flowing into a research department may be increasingly assimilated better by a few entities within the department rather than by the whole department. Following this analogy, any new explicit information flowing to a node may increasingly be assimilated better by some cluster (a specific knowledge group with the node) within the node than by other clusters within the parent node. This process will in effect result in sub-nodes developing with a node and therefore a relative increase in tacitness of the parent node. The creation of these tacit sub-nodes may be a result of the communication difficulties within pockets of specific knowledge once the parent node increases in knowledge size and complexity, even if we assume the processing capabilities are roughly similar among these sub-tacit nodes. Once more and more distinct sub-nodes takes shape within the parent node, the difficulties of exchange of information between 'tacit' nodes arise and will lead to underutilization and redundancy of the organization resources. The degeneration into sub-tacit nodes can be alleviated to some extent by having tighter coordination and communication routines. However, the cost of coordination and communication can go up and thereby limiting the effective size of the organization. Hence the following proposition is due.

P4. The size of the firm is limited by the extent to which the firm can process the information from outside and create value or innovations by combining its tacit knowledge with outside information. The advantages of having more innovation or value creation possibilities by processing more information is counterbalanced by the increasing costs for coordination and communication cost for managing the 'tacit' knowledge nodes within the firm.

With the developments in information technology, the economies of firm size and structure are changed appreciably. As consumers and markets possess more information processing capacity, chances are that 'tacit' knowledge sources are made into 'explicit' sources. For example, the 'tacit' knowledge of the experience of a customer with a brand new auto model is increasingly made explicit by the Internet where there are forums where the consumer can share his/her experience and have discussions with other users of the same product. These interactions of 'tacit' sources as enabled by the web can create 'explicit' sources of knowledge for new buyers of the same model. Also, the omnipresence of the web has made it more likely that a potential user will come across this useful 'explicit' information. Information is changed not in quality but in quantity as well. Because of the processing power of the web, specialists and experts can increase their stock of 'tacit' knowledge appreciably by incorporating the processed or 'expliticized' information from other tacit sources. As we discussed above, these specialists have thus built up a resource barrier with respect to the 'tacitness" of the knowledge they posses and their ability to discriminate useful information from the useless coming from the environment. Because these 'tacit' nodes can process information efficiently compared to other novice processors, they can offer their services at a discount (by offering more quality information per dollar). Also, because of more interconnection possibilities with other potential users of information, the 'specialist' node can enjoy economies of scale and market its resource at a competitive price. This competitive advantage of processing by tacit knowledge resources is basically due to the fact that knowledge based resources exhibit increasing returns to scale compared to conventional resources (Arthur, 1990). Conventional resources and even explicit knowledge show a decreasing return to scale. Increased return to certain knowledge resources is due to the fact that once knowledge is created, it can be utilized for all other subsequent innovations without effectively increasing further investment in creating such resources. For example, once an auto manufacturer masters gasoline engine technology, it can be applied to all future models without any additional investment in technology. The more the number of autos produced with this technology, the more is the rate of return. Also, it is also easy to diversify into related products based on knowledge gained from some product. Thus there is a strong economic need to create and process knowledge resources especially tacit resources. All these aspects will give rise to the rise of these 'specialist' or expert nodes in the form of specialist firms or expert centers within a firm offering custom services in an increasing scale in this information economy. The following proposition summarizes this.

P5. Developments of information technology will enable the 'tacit" information holders to process outside information better and we would see an increase in the scale of 'expert" or 'specialists' firms offering custom services.

We posed the question before about the fundamental reason why the knowledge subsystems of the organization are willing to align with other complimentary resources within the firm or outside the firm. Although we mentioned that the reason could be the savings in transaction costs, the recognition of the transaction cost and its origins are seen even clearer in the context of knowledge and information exchange between tacit and explicit nodes. The benefits to be realized from the exchange would be the innovation possibilities open to tacit and explicit nodes of knowledge. At first, nodes are organized within a firm because the savings are favorable compared to the case when they are outside the firm. As soon as another node tacit or explicit emerges within or outside the firm, which is complimentary to the first, it gives rise to an innovation possibility. The benefits and the cost involved for this innovation should now alter the prior working relationship of these nodes, if there is significant improvement in the returns with a new strategic realignment between nodes. Some of this strategic realignment of competencies is observable in the Silicon Valley type of business model. Brynjolfsson and Mendelson (1993) describes the silicon valley business model as where value is generated by ever-changing coalitions, where each member of a coalition specializes in its area of core competence and leverages it through the use of tactical and strategic partnerships. The competencies involved might reside in different firms to begin with and they are brought together for a specific purpose of an innovation. It is also known that many such coalitions have turned out to be independent firms once they are successful. Even without going into the collaborative effort of firm level coalitions, it is a well known fact that many start ups in the high tech sector are the result of joining together of specific individuals possessing complimentary sources of expertise. This phenomenon of reorganization at the possibility of creating an innovation is in many ways the reason for many firms to come to exist or to continue to exist. Instead of looking at only the cost of conducting transactions, as the traditional transaction cost theory does in explaining the existence of firms, once we account for the benefit of conducting transactions through realignment and combinations of complimentary knowledge resources, we can explain the innovation creation and creation of knowledge within firms in a better sense than on pure cost considerations. The possibility of innovation is much more fundamental for the continued existence of the firms as well as the establishment of new firms. The fundamental asset for innovation is knowledge and the character of this knowledge whether tacit or explicit explains the mechanism of organizing the resources that underlie the structure of most firms. In view of these arguments, the following proposition can be stated.

P6. Innovation possibilities arise from the combination of knowledge resources. Firms come to exist if they have more capability than the market to process these resources. The advantage of tacit resources to process information in a cost-effective manner would be the building block of firms.

4. CONCLUSION

From the recognition of the prime resources of innovation, knowledge and information, this paper discusses the mechanism by which these resources affect an organization for its innovations. The central theme was the close interaction possibility of the resources with the emergence of advances in information technology. The characteristics of the resources, for example "tacitness" is central to the information processing structures that emerge that affect the dynamics of the collaborative structures for innovation. Cost effective information and knowledge generation processes by "tacit" subsystems serve as nucleus for firms. Fundamental building block for firms is their ability to initiate and process tacit information and generate innovations.

The main implication of the paper is that innovations, especially based on knowledge creation, would be "emergent" based on the identifications of such possibilities by the knowledge subsystems involved in the process. Top-down management would be handicapped in creating innovations. Transaction cost theory explains the need for firms level of organization because the need to eliminate opportunism due to asset specificity. But, as we discussed, tacit subsystems would better leverage their capability although they are highly asset specific, if these systems are free to combine and compliment. The elimination of opportunism by firm level organization can therefore choke innovations for the larger market and customers. These implications are worth further scrutiny, especially in the light of IT advances.

REFERENCES

Anderson, Philip, "Complexity theory and Organization Science", Organization Science, Vol. 10 (3), 1999, 216-232.

Arthur, Brian W, "Positive Feedbacks in the Economy", Scientific American, 1990, 92-99.

Boisot, Max M, Knowledge Assets, Oxford University Press, 1998.

Brynjolfsson, Erik and Haim, Mendelson, "Information systems and the Organization of Modern Enterprise", Journal of Organizational Computing, 1993.

Galbraith, Jay R, Organization Design, Reading, MA: Addison Wesley, 1977.

Galunic, D.C. and Rodan, S, "Resource recombinations in the firm: Knowledge structures and the potential for Schumpeterian Innovation", Strategic Management Journal, Vol. 19, 1998, 1193-1201.

Gareth Morgan, Images of Organization, Sage Publications, Thousand Oaks, 1996.

Glazer, Rashi, "Marketing in an Information- Intensive Environment: Strategic Implications of Knowledge Asset," Journal of Marketing, 55, 1991, 1-19.

Granovetter, Mark, " Economic Action and Social Structure: The Problem of Embeddedness," American Journal of Sociology, Vol. 91(3), 1985, 481-510.

Grant, Robert M, "Prospering in Dynamically competitive environments: Organizational capability as knowledge integration," Organization Science, Vol. 7(4), 1996a, 375-386.

Grant, Robert M., "Toward a knowledge-Based theory of the firm", Strategic Management Journal, Vol. 17 (Winter Special Issue), 1996b, 109-122.

Hayek, F. A., "The use of Knowledge in Society", American Economic Review, Vol. 35(4), 1945, 519-530.

Holtshouse, Dan, "Knowledge research issues", California Management Review, Vol. 40(3), Spring, 1998, 277-280.

Huber, George, "Organizational information Systems: Determinants of their performance and Behavior", Management Science, Vol. 28(2), 1982, 88-110.

Meyer, Alan, Peter Frost and Karl Weick, "Overture--The Organization Science jazz Festival: Improvisation as a metaphor for organizing," Organization Science, Vol. 9 (5), 1998, 569-576.

Nadler, David A. "Introduction: Organization architecture: A metaphor for change," in Organization Architecture. Design for changing organizations, Nadler, D.A., Gerstein, M.S., Shaw, R.B., eds., The Jossey Management series. 1992.

Nonaka, Ikujiro and Takeuchi, Hirotaka, The knowledge-creating company: how Japanese companies create the dynamics of Innovation Oxford University Press, 1995.

Norton, Melanie and Lester, J, "Digital accessibility: Information value in changing hierarchies", American Society for Information Science. Bulletin of the American Society for Information Science, Washington, Vol. 22(6), 1996, 21-25.

Penrose, Edith T. The Theory of the Growth of the Firm, Wiley, New York, 1959.

Polyani, M. The Tacit Dimension, Doubleday & Co., New York. 1966.

Sampler, Jeffrey L, "Redefining industry structure for the information age," Strategic Management Journal, Vol. 19, 1998, 343-355.

Simon, Herbert A, "Applying Information technology to organization design", Public Administration Review, 1973, 268-278.

Simon, Herbert A, The Sciences of the Artificial, Cambridge, MIT Press MA, 1981.

Simon, Herbert A "The impact of electronic communication in organization", in Models of bounded rationality, Vol. 3, MIT Press 1997.

Teece, David J, G. Pisano, A. Shuen, "Dynamic capabilities and Strategic Management," Strategic Management Journal, Vol. 18(7), 1997, 509-553.

Teece, David J, "The market for know-how and efficient international transfer of technology," Annals of the American Academy of Political and Social Science, Vol. 45 (8), 1981, 81-96.

Tushman, M. and Nadler, D.A, "An Information Processing approach to Organization Design, "Academy of Management Review, Vol. 3(3), 1978, 613-624.

Author Profile

Dr. Joseph Richards earned his Ph.D. at the Syracuse University in 2001. Currently he is a professor of Marketing at California State University, Sacramento.

Dr. Seung Bach earned his Ph.D. at the University of Tennessee, Knoxville in 2002. Currently he is a professor of Management at California State University, Sacramento.