WPbanner5.jpg (10176 bytes)
No. WP- 03-02
Power Issues in Knowledge Management

Hamid Ekbia
Rob Kling
School of Informatics
Indiana University
Bloomington, IN. 47405   
Center for Social Informatics  SLIS
Indiana University
Bloomington, IN. 47405

Note: This is a very extended version of “Power in Knowledge Management in Late Modern Times” which was accepted as a “Best Paper” for the Academy of Management Conference, to be held in Seattle, WA in August 2003.
Knowledge management was advanced in the early 1990’s as a new managerial reform suited to the rapidly changing and globally vast business environment. These reformers encouraged managers to treat as a critical source their employees’ knowledge, of which they themselves had minimally articulated and varying conceptions. The major common feature among these conceptions was their generally cognitive, epistemological, and often individualistic approach to the question of knowledge, which dispossesses them of other important issues, most notably “power.” Adopting a sociological approach in this paper, will reexamine issues of knowledge management, especially as they relate to power relationships inside and outside organizations. We apply a refined version of Foucault’s notion of a “regime of truth” to show the institutionally-specific processes, procedures, and mechanisms that are usually at work in the creation of statements about the social world that function as true. As examples, we distinguish three regimes of truth that, we argue, are at work in the functioning of publicly traded businesses in the U.S. — the financial reporting, analysts’ research, and business press regimes of truth. A brief look at knowledge-management literature will further manifest a fourth regime of scholarly research. The close examination of these multiple regimes will lead us to the overall conclusion that power relationships can systematically influence the statements about the social world that function as true. In the latter part of the paper, we will study the implications of this observation for the theory and practice of knowledge management.

The second half of the twentieth century is often characterized as times of great social and political uncertainty, rapid economical change, globalization, and hypercompetitive business environments. Under such circumstances, it is suggested, knowledge turns into the most important means of production (hence, the “knowledge society” and “knowledge economy”) and a strategic resource of the firm (hence, the “knowledge organization), which should be maintained, processed, integrated, disseminated, and shared (hence, “knowledge management”) by a new generation of competent workers who are trained and technically equipped for this very purpose (hence, “knowledge workers). The concept of knowledge, often in association with such other terms as “innovation,” “competency,” “flexibility,” and “competitive advantage,” has thus become a central concept in discussions of late modern times.

In the early and mid-1990s, a new managerial reform was advanced by a number of consultants and some academics -- knowledge management. Knowledge management has several roots -- including the conception of certain technical and professional occupations as knowledge workers, the growing discourses about information societies and even knowledge societies, and last, the claims of some computer scientists to develop “knowledge based information systems.” What is often missing from such discussions and discourses, however, is the close relation between “knowledge” and “power” as social processes that create and constrain each other in intricate ways. Except for cryptic allusions to the Hobbesian dictum that “Knowledge is power,” therefore, one hardly sees a coherent analysis of the above relation in discussions of knowledge economy, in general, and of knowledge management, in particular. And the Hobbesian dictum, as many have argued, is not only an outdated cliché for our times (e.g., Harraway 1991), it provides, as we shall argue, a lopsided view of the relation between power and knowledge, somehow sweeping its significance under the rug.

The purpose of this article is to examine how power relations, apply  to the theory and practice of knowledge management in modern organizations. The article continues in the next section with a brief review of the prevalent conceptions of knowledge management. We then apply a refined version of Michel Foucault’s notion of a “regime of truth” to introduce three regimes that are involved in the functioning of publicly traded companies.  Building on a study by Schultze and Leidner (2002) and a bried examination of business literature in the 1990’s, we further propose a regime of scholarly research. Our main conclusion is that power relationships play important roles in helping to understand some of the limits of what kinds of knowledge professionals with share with others,  these power relationships have not been integrated into theories of knowledge managemnt.. We will discuss the implications of this conclusion for the current theory and practice of knowledge management.     

What is knowledge that it should be managed?
Knowing, even in the case of moral judgment, consists only in using a method of manipulative skill for getting rid of the external conditions which interfere with, disturb, intrude upon the self and the knower. (Dewey 1989 [1939]: 585)
Knowledge management is one of the latest methods advanced by consultants, practitioners, and some academics to meet the demands of what is thought to be an increasingly dynamic and competitive business environment. The urge for global reach and the speed with which this should be achieved are often considered the main motives behind organizations’ interest in questions of what they know, who knows it, and what needs to be known (Alavi & Leidner 1999, Prusak 2001). The knowledge management reformers encouraged managers to treat their employees’ knowledge as a critical resource (Huseman & Goodman 1999; Stewart 1997; Sveiby 1997). But there seems to be a lack of a clear conception among managers and practitioners, who have rather different perspectives and expectations of knowledge management. In a study of more than a hundred managers and executives across the globe, for instance, Alavi and Leidner (1999) discovered that the conceptions about the scope of knowledge management vary from customer and client information, through competitor and product/market information, to human resource, finance, and strategy (p. 13).

In the academia, the underlying conceptions of knowledge, while varying across authors, seem to have been pragmatic, rather than developed with a refined philosophical analysis. Most authors invoke a triple hierarchy of data, information, and knowledge, which takes data as either uninterpreted symbols or some kind of “basic facts,” and information as data interpreted and vested with meaning. However, there is less of a consensus about knowledge.  Some authors, like Choo (1998) and Huysmans (2002), define knowledge in Platonic terms as "justified true belief." Some, like Alavi and Leidner (1999), draw upon Nonaka (1994) and Huber (1991) to characterize knowledge as "justified personal belief that increases a person's ability to take effective action." While none of these authors advances a conception of justification, Alavi and Leidner do link knowledge to an actor's framework for action. Some KM analysts treat knowledge as more integrated than information, while others favor a “pluralist epistemology,” that is more pragmatic (Stenmark 2002). But generally, the focus for most knowledge creation is "in people's heads," which sharply differentiates it from the “digitized, codified, and easily distributed” character of information (Prusak 2001: 1002). Based on this distinction, for instance, Prusak enumerates “judgment, design, leadership, better decisions, persuasiveness, wit, innovation, aesthetics, and humor” as valuable components of knowledge (ibid).

A common feature among these accounts is their generally cognitive, epistemological, and often individualistic character. One major exception is for those analysts who discuss cultural knowledge -- knowing the norms of a group or an organization, so as to be able to accurately interpret the behavior of other people as well as to behave appropriately. But, as in the case of knowledge, there seems to be a lack of clear understanding as to what precisely constitutes organizational culture and how this could be learned.  In Alavi and Leidner (1999) study, for example, most of the managers believed that cultural and managerial aspects accounted for the bulk of the issue. However, as the authors argue, “the responses were nebulous in terms of specific cultural implications, perhaps indicating a root concern for the absence of concrete ideas on how to address it” (p. 12).

These ambiguities and the absence of concrete ideas drive many critics to interpret knowledge management as “nonsensical” (Wilson 2002), an oxymoron (cf. Skyrme 2002), or a cacophony (Nunn 2001), and other skeptics to take it as a “re-badging” of earlier information and data management methods (cf. Prusak 2001: 1002), or simply as the latest fad among what consultants have wrought in order to replace declining revenues from previous methods like business process reengineering (Harmon 2001). While   there might be some value in these skeptical assessments, we also believe that the practices of sharing and withholding professional  knowledge in organizations is of  significant managerial importance.

We want here to take issue with the abstract and nebulous notions of knowledge and culture that many of KM accounts have to offer --abstracted, that is, from the social processes, procedures, and techniques that give rise to knowledge and are, in turn, bred by it. As Schultze and Leidner (2002) conclude from their comprehensive survey of KM literature, except for a very trivial percentage, none of the KM formulations explicitly engage issues of power, and the ways that power relationships between organizations or between actors can influence both the character of beliefs that actors treat as knowledge, and the conditions under which they will share knowledge. This abstract notion of knowledge as something that is always good, useful, and efficacious has come under scrutiny by contemporary thinkers such as Foucault, who see knowledge in a different t light, especially in its relation to power.     

Foucault on Power, Truth, and Knowledge

In the last analysis, we must produce truth as we must produce wealth, indeed we must produce truth in order to produce wealth in the first place. (Foucault 1980: 93–94)

Like other political philosophers, Foucault is interested in understanding the relation between power and the discourses of truth. Unlike those philosophers, however, he begins his survey with a concrete down-to-earth question: “What type of power is susceptible of producing discourses of truth that in a society such as ours are endowed with such potent effects?” (1980: 93). And rather than seeking the answer in the intentions of people --“why certain people want to dominate, what they seek, what is their overall strategy” --he suggests, we should study power in its external visage, where “it installs itself and produces its real effects” (ibid: 97). But where exactly are these places of installation and effect? Not, again, in individual hands or specific localities:

Power must be analyzed as something which circulates, or rather as something which only functions in the form of a chain. It is never localized here or there, never in anybody’s hands, never appropriated as a commodity or piece of wealth. Power is employed and exercised through a net-like organization… In other words, individuals are the vehicles of power, not its points of application (ibid: 98) 

This relational formulation has two immediate consequences. First, it affords a morally neutral conception of power, not as something that only weigh on us as a repressive force, but that “traverses and produces things, induces pleasure, forms knowledge, [and] produces discourse” (ibid: 119). Secondly, and in a sharp reversal of dominant views, it permits a parallel understanding of knowledge, not as something that is always good and efficacious, but also “subject to constant economic and political incitement … produced and transmitted under the control, dominant if not exclusive, of a few great political and economic apparatuses (university, army, writing, media)” (ibid: 131-132). It is with these parallel and conjoined conceptions of power and knowledge that Foucault postulates for each society a “regime of truth” as:

the types of discourse which it accepts and makes function as true; the mechanisms and instances which enable one to distinguish true and false statements, the means by which each is sanctioned; the techniques and procedures accorded value in the acquisition of truth; the status of those who are charged with saying what counts as true. (ibid: 131)  

This notion of a regime of truth, we believe, provides a useful conceptual tool for understanding the mechanisms, techniques, and procedures at work in the production and distribution (management, if you will) of knowledge in organizational settings.  If, as Foucault suggested, “we must produce truth in order to produce wealth,” there should inevitably exist, in parallel with business firms, sites of production of truth, which are the institutional milieu for regimes of truth. To analyze and understand these regimes and their institutional embeddings, though, a sociological, rather than epistemological, approach is needed. This is the approach that we aim to pursue in this article. We will treat as "knowledge" statements (and thus beliefs) that people will act upon as if they were true and that are produced by a regime of social truth. Thus, by this definition, many individuals who purchased stock in companies like WorldCom and Enron in 1999 "knew" that the stock would gain in value for at least a few years. Technically, we might say that these stock investors just had mere speculations or beliefs, and not knowledge. But we shall illustrate how there were organized systems of social knowing that gave their beliefs a higher degree of trustworthiness.

For this purpose, we are going to define a key concept-- an "institutionally specific regime of social truth" -- as a structured constellation of organizations, procedures, laws, and techniques that produce statements about the social world that function as true. For example, which team wins a game of college basketball is not determined by some consensus of observers watching the game. The National Collegiate Athletic Association develops rules through sport-specific rules committees, and changes some rules each year. These rules are interpreted by referees during the course of game play .  These rules cover diverse topics, including the designation of eligible players, courts, equipment, scoring, the use of court side camcorders and when they may be consulted, fouls, and penalties. Decisions of the referee, during the game, are final. In a significant way, the truth of the statement "Ohio State lost to Purdue, 53 to 51" is not determined only by the players and their game play, but also by the current set of rules and referees judgements. The statement "Ohio State lost to Purdue, 53 to 51" functions as true in a  world of action -- for example by influencing which of these teams may compete in a major end-of-season tournament.

Similarly, the statement that "O.J. Simpson is not guilty of murder" was produced by a legal regime of truth that included extensive Los Angeles County Superior Court rules rules for investigation, the development of charges, procedures for selecting a jury, rules of evidence and witness examination, and rules for jurors' behavior during the trial. The statement that "O.J. Simpson is not guilty of murder" functions as true in a world of human action; O.J. Simpson was not sent to jail for allegedly murdering his wife.

Our characterization of an "institutionally specific regime of social truth" refines Foucault's society-wide regime of truth. For Foucault, there was one regime of truth in contemporary Western societies -- science. For us, there are numerous and varied regimes of social truth that render statements, such as "IBM's annual report may be trusted" as functionally true. To provide a concrete sense of what this means, we will first examine three major organized "regimes of social truth" that were pertinent to making justified judgments about the value of stock in the 1990s:
  1. A regime of financial reporting, that includes a firm’s annual reports and the reports of official auditors that a client firm's reports are trustworthy;
  2. A regime of research into a firm’s financial prospects by major investment banks and stock brokers’ research result in recommendations to buy sell or hold specific stock;
  3. A regime of inquiry into business practices that are reported in the mainstream business magazines.
To make the discussion even more concrete, we use the example of Enron, and the kinds of statements of “truth” that were made about it within each of these regimes, to show the expected influence of economic power in the production of such statements.

Regimes of Regimes Social Truth in American Business During 1990's

The organizations of the post-capitalist society must constantly upset, disorganize, and destabilize the community. (Drucker 1995: 80)

The main push behind the “knowledge economy,” as we pointed out earlier, is claimed to be global reach in a timely fashion that meets demands of a rapidly changing environment. Under such circumstances, it is suggested, knowledge becomes the primary resource for individuals and for the economy, giving traditional factors of production (land, labor, and capital) a secondary status (Drucker 1995: 76). To become productive, however, knowledge should be integrated into a common task by an organization, whose putative function is “to put knowledge to work --on tools, products, and processes; on the design of work; on knowledge itself” (ibid: 77). Furthermore, since knowledge changes fast, and since the social environment tends to maintain stability and slow change, “every organization has to build the management of change into its very structure,” playing the role of a destabilizer of the environment (ibid). In short, the primary role of the modern organization, according to Drucker, is to innovate and to do this, in the spirit of Schumpeterian “creative destruction,” in a purposeful, disruptive, and persistent manner (ibid).

This recipe of change and disruption was hardly followed in the 1990’s in as a faithful a manner by any other firm as it was by Enron, whose publicly open strategy was “to open the markets”(ref), “to define new commodities (derivatives)” (O’Reilly 2000), and “to shake the whole [business] culture” (Hamel 1997). Founded in 1984, as the merger of Houston Natural Gas and InterNorth, which carried out the exploration, production, and transfer of oil and gas, Enron Corporation morphed into one of the largest US corporations — $100 billion in revenue, $60 billion market value, and twenty thousand employees around the globe. According to dominant narratives, this was made possible by the adoption of a new market strategy that leaned toward “knowledge and innovation”, rather than the traditional ownership of physical assets. The central vision in Enron’s enterprise was to fully use the financial and derivatives markets: to buy a commodity that somebody wanted to sell, and then sell it for a profit to someone who wanted to buy it. It began with oil and natural gas, and expanded to a variety of derivatives, from electric power generation and pipeline capacity to broadband communication and the freight capacity of modular containers. By and large, Enron succeeded in materializing this vision in many areas, trading more than 1000 commodities in the futures market of later 1990’s. Dominant narratives tended to attribute, by and large, Enron’s success to its innovative strategy (Hurt 1996, Chan & Mauborgne 1999, Sweat 2000, Sweetman 2001). We want to argue, however, that Enron owed its “success” to the simultaneous functioning of a number of regimes of truth, three of which we are going to discuss here.

The Regime of Financial Reporting
Financial accounting is the branch of accounting concerned with providing useful information about publicly-traded firms to external users, such as investors and creditors, who do not have access to the internal financial information of a business, and who should, therefore, depend on such financial statements to decide where to invest or loan money.  The current structure of financial accounting in the U.S. was largely defined by the Securities Act of 1933 and the Securities Exchange Act of 1934, which gave the authority to prescribe the methods and rules of preparation of financial statements to the Securities and Exchange Commission (SEC). For historical reasons that are relevant but beyond our scope here, however, SEC has for more than sixty years almost completely abdicated the responsibility of setting the rules to Financial Accounting Standards Board (FASB), a body comprised of seven representatives from the private sector. SEC has maintained the role of enforcing those rules for SEC itself. The members of FASB, which is comprised of investors, business people, and accountants, are selected by an oversight body called the Financial Accounting Foundation (FAF). FASB formulates the generally accepted accounting principles (GAAP) that governs what is reported in financial statements and how it should be measured --for example, whether a manufacturer/trader should include expected sales/contracts for the following year(s) in their current report, and if they do, how should these be measured (current market value or future predicted price). Although the rules set forth in GAAP do not have the legal status of “laws,” they have to be followed by publicly traded firms when they report to SEC. Nonpublic firms, in search of capital investment, also have often to  prepare their statements according to GAAP. In short, GAAP establishes what some authors call the “constitutive rules” of communication and reporting in the securities market (Abolafia 1998).

To ensure the accuracy of the statements, a third body -- namely, accounting firms comprised of certified public accountants (CPAs) -- is in charge of auditing financial statements issued by businesses. CPAs are also organized in the American Institute of CPAs (AICPA). Although there are thousands of accounting firms in the U.S., a large portion of the auditing of big corporations was until recently performed by the “Big 5.” The main responsibility of auditors is to make sure that financial statements are issued in conformity with GAAP. The combination of these multiple bodies (SEC, FASB, AICPA, etc.), rules (GAAP), and procedures (reporting, auditing, election of members) is what we call the regime of financial reporting. As the above brief description demonstrates, the overall function of this regime of social truth is to create justified financial statements about the economic performance of business firms, and to make these statements public so that, in their specific institutional milieu, these statements function as true. Within this regime, power is  implicated in devising GAAP, in the SEC’s enforcement of them, and, most conspicuously, in auditing statements.  What complicates this process of knowledge and truth creation, however, is the flexibility built into GAAP, which allows a choice among different methods that may often produce incompatible results and truths about the performance of firms (Mumford & Comiskey 2002). Nonetheless, until the major financial scandals of 2001-2002, many investors relied upon the veracity of audited annual reports, and repetition of key claims from them in newspaper and magazine stories, investment newsletters, etc. They "functioned as  true" to the extent that many investors used these accounts as a basis for spending their money on the stocks of specific firms.

What makes this regime of social truth even more complicated, in a way threatening its functionality, is that auditors are paid by the same corporations that they audit, and in more than one way --e.g., by providing consulting services. This is the origin of what has come to be known as an inherent conflict of interest in the auditing regime.   As is now public knowledge, it is precisely what happened during the 1990’s between Enron and the auditing firm Arthur Andersen, which, according to reports and documents of numerous instances, had a decisive role in maintaining the glorious image of a innovative and efficient business for Enron. In a fake contract between Enron and Merrill Lynch in December 1998, for instance, Enron managed to raise its stock evaluation by 40%, and Enron executives earned about $100 million in stock sales, and Merrill Lynch was reimbursed by $8 million for the cancellation of the contract. However, Arthur Andersen did not ask, as it should have, for rewriting the books after the nullification of the contract.

Dominant narratives of these events attribute these pitfalls, in a sweeping manner, to the loopholes in GAAP, or to the complexity of the current accounting rules that makes their understanding difficult. Robert Herdman, the Chief Accountant of SEC, for instance, attributes them to the “rule-based,” as opposed to “principle-based,” character of FASB’s standards that “provide extremely detailed rules that attempt to contemplate virtually every application of the standard, …[and] make it more difficult for preparers and auditors to step back and evaluate whether the overall impact is consistent with the objectives of the standard” (Herdman 2002: 5). He thus recommends, for sake of greater transparency in financial reporting, a principled-based system of standards that reflects “the economic substance, not the form, of the transaction” (ibid). While it is not clear how such a shift would generate more transparency, it certainly fails to address the power relations that shaped the dynamics of links between Enron and Arthur Andersen, and, hence, the flow of knowledge within the regime of financial reporting as it affected the public.  At the heart of these relations were lucrative consulting contracts that are the major sources of income for auditing firms, and can only be afforded by a large corporation such as Enron, not by a small business. While the accounting rules are otherwise the same for both companies, it is only the lure of such contracts that makes them bend in one way rather than the other, giving rise to contrasting truths about them.  

The Regime of Investment Analyst Reports
The other main regime of social truth about the future financial prospects of American business is investment  analysts’ research into the performance of publicly traded firms that is reported to the public through established channels such as bankers' bulletins to their investor clients,    major analysts talks to investors and stockbrokers, investment analysts comments in the mass media, etc. Analyst reports play a very critical role, as it were, in shaping the public image of companies and, hence, the decisions of investors and creditors, determining in a serious way the financial viability of firms. Although there are no officially established rules and guidelines similar to GAAP for analyst research and reporting, there are “regulative” or “cognitive” norms of behavior (Aboafia 1998) that are meant to be observed by all analysts. One such norm, for instance, is “research integrity” in the sense of analysts’ commitment to produce unbiased reports based on objective criteria about the firms they study. What is supposed to guarantee such integrity is the analyst’s independence from the firm in terms of source of income. In other words, the analyst should be paid by an entity separate from the firm for the research they conduct about the firm. As a matter of fact, except for a small minority of “independent” firms, Wall Street banks and brokerage firms are the main employers and providers of income for investment analysts. Many of the major brokerage firms' clients generally rely upon the firms research reports for investment decisions: they 'function as true."

There is, however, a potential for conflict of interest through the dual activities of banks and brokerage firms that hire them --namely, their investment and/or securities trading, on the one hand, and their market-analysis activities, on the other. To prevent this conflict from prevailing and affecting analyst opinion, a “firewall” is supposed to be set up between the participant in the two activities --namely, traders and analysts. This collection of institutions (creditors, banks, securities firm, analysts), procedures (research methods, incomes, firewalls), and mechanisms (reports, bulletins, etc.) is what we call the regime of analyst reporting. As with financial reports of the previous regime, the major role of this regime is to produce statements that, in their institutional setting, are taken as statements of knowledge and truth by all the participants. 

Both of the above measures (objectivity and independence) are, however, subject to financial and political pressure in a way that undermines the sought-for integrity of analyst research. The independence of analysts from firms is violated indirectly by banks’ dual activities. Lending money is the major source of income of banks, and the expectation of return on such investment is what ties them closely to the firms in which they invest. It would only be “natural” for banks, then, to try to increase the chances of receiving their money (the principal) and the income accrued thereupon (the interest). This, in turn, implies that they should try their best to keep the wheel of the firms turning, a goal that would be obviously met if the firms make profit. Banks, however, do not have control over the internal operations of firms. Where they do have control is to maintain a positive image of the firms in the eyes of big and small investors. And this is where analysts and their reporting regime of social truth come in, as was revealed in cases such as Enron and WorldCom. It is being documented that bank executives did, in fact, use their power to penetrate the firewalls, basically dictating their analysts’ opinion on Enron in numerous occasions where signs of trouble in terms of Enron’ performance were in sight (Barboza 2002). Thus, a famous analyst is reported to have promoted WorldCom stocks in return for investment-banking deals (Creswell 2002). And when analysts refused to do so, not only were they subjected to pressure from their own employers, they were directly coerced by executives of firms on which they conducted research (Goodwyn 2002; Spitzer 2002).    

In addition to direct coercion, which undermined analyst independence, firms use their power and influence to affect the “objectivity” of analysts’ reports.  For instance, a conference call by Enron executives to counter a “rumor” about the performance of Enron Capital and Trade Resources brought together 170 analysts under one roof in less-than-a-day notice. Explaining their “system of computerized trading controls,” Enron executives reassured the analysts that “the company was actually making money,” reversing the course of events that led to a subsequent rise in Enron’s stock value (Hurt 1996). The favorable reception by analysts of the “knowledge” that these executives “shared” with them could only be understood in terms of Enron’s clout and influence in Wall Street institutions. Although truth was, on the surface, instantly created and communicated, its trace has to be found in the power relations that are part and parcel of the regime that generates the truth.

The Regime of Business-Press Reporting
The next, and probably most visible, regime of truth that we are going to examine is that of business journalism.  Tensions, conflicts of interest, and demands of political and economic pressure are often too explicit that elaborating on them might at times seem to be a statement of the obvious. Despite its visibility, however, the processes of knowledge creation in this regime are the least defined and established among the ones examined thus far. Although journalists and reporters have their own norms of conduct --e.g., the integrity of reports, the validity and authenticity of information sources, cross-checking of sources, non-plagiarism, and so forth  there is a great deal of flexibility involved in how these norms are observed. This blurriness of norms and rules, though, does not exclude journalism from our definition of a regime of truth, for, by its very nature, journalism is in the enterprise of generating statements that, in the specific social environment that consumes those statements, function as true.     

These observations probably apply to business journalism more than any other category, because the major source of income for such journals is the advertisements by large business corporations. This is especially true in times of economic growth like the 1990’s, when industry records in terms of the number of ad pages (mainly by dot.com companies) were broken in both general-interest and business magazines. A major tension in here, therefore, as a business journalist has put it, is the “tension between the need to please advertisers and the need to please readers” (Longman 2002: 19). Building on his own experience with the “bubble” of the 1990’s, when journalists “were lured more by inflated salaries than any deep curiosity about how business works,” Longman shares an episode from one of the major publications in the U.S. where the owner basically prohibited the use of the word “bubble” in any writing that dealt with economy of the time.  He also points out instances of “very aggressive” behavior by Enron executives in the face of emerging stories about the company’s financial performance in 1999 (ibid: 20). Other journalists have provided similar views and accounts, albeit after the fact (Madrick 2002). The story of the attempts by Enron executives to block the publication of the report by Bethany McLean about the company’s accounting practices is well documented (Sherman 2002).

This might explain why Enron was touted as one of the most innovative companies five years in a row by Fortune magazine (1997, 2000), and its leaders were praised as “revolutionary,” “genius,” and “visionary” by the same journal and others such as Forbes, Wall Street Journal, Financial Times, and Business Week. This pattern of behavior in the generation of truth by the business press could only be understood as the outcome of knowledge/power relations in a web that encompasses journalists, editors, and reporters, but also the owners of publishing houses, the firms that provide income for the journals, and managers and executives who have stake in the image provided by journals of their firms.

Knowledge Management Revisited
We observe that while the three regimes discussed above are presumably established to create statements about the social world that function as true, each of these regimes systematically distorted their statements because of the expected influence of economic power: lucrative consulting by the big auditing firms, the prospects of new lucrative investment deals for firms whose stocks a banks' analysts treated favorably, and the expectation of important advertising revenues by creating a stream of  business stories that were -- on balance -- extremely favorable to large businesses and their managers. In short, power relationships can systematically influence the statements about the social world that function as true.

Of course, there is little mystery about the value of social actors to manage knowledge in the larger social world -- to create "spin" for new products, programs and services. In fact, in dramatalurgical sociological theories --such as Goffman's theories of social interaction --manipulative self-presentation is in some ways routine (19xx). In practice, however, this becomes trouble to the extent that it is grossly misleading and deceptive --i.e., while it might be acceptable for a newly wealthy person to show off to his friends by buying a top-of-the-line BMW, a mediocre realtor who borrows that BMW to impress potential clients with his success is misleading. By the same token, Enron’s, WorldCom’s, Tyco’s, and other similar behavior was, indeed, troublesome because it was grossly misleading, and the above regimes of social truth contributed, in a mutually interdependent manner, to this misleading behavior to the extent that they were affected in an adverse fashion by power relations circulating among the links that connect their respective entities. The question is, what are the implication of all of this for knowledge management.

We suggested a characterization of knowledge at the outset, according to which knowledge consists of statements and beliefs that people act upon as if they were true. Knowledge management, from this viewpoint, would entail managing the ways that a firms' employees will share social knowledge, or knowledge whose revelation can have important social consequences for those who believe it.  We have seen throughout the above discussion that power relations are closely involved in the management of knowledge in the above sense: they are constitutive elements of the processes and mechanisms that determine which knowledge is shared with whom and in what ways. What kinds of knowledge can be shared is determined by power relations. Consequently, the moral value of knowledge should also be determined according to the kinds of power relations that are involved in its knowledge; and this, as we emphasized at the outset, does not necessarily have to be negative.

Our use of Enron as the prime example so far might give the impression that power relations only have negative repercussions, leading to the conclusion that theoretical accounts based upon such relations are cynically pessimistic or only applicable to cases of failure. To avoid such an unwarranted conclusion, for the  remainder of this article we would like to argue against it, from both a theoretical and practical perspective. Thus, in the next section, building on a recent work by Schultze and Leidner, we will discuss the theoretical ramifications of incorporating power in accounts of knowledge management. We will show that scholarly work, such as the research literature about knowledge management, in fact, constitutes another regime of social truth, with certain contributions to the functioning of business enterprises.  Then, in section 7, we will look back at a few specific cases reported in the literature in order to show how this perspective can provide us with a better grip on the practical issues involved in the management of knowledge, whether we are talking about successful or unsuccessful cases of its application.

The Scholarly Regime of Social Truth
In a recent study of information systems (IS) research literature, Schultze and Leidner (2002) have used a taxonomy suggested by Deetz (1996) to analyze the theories and discourses in IS in terms of their motivating assumptions and methodologies, and have come up with four categories, as follows:
1.    Normative discourse focuses on the discovery of technology solutions to knowledge problems;
2.    Interpretive discourse is generally concerned with organizational practices, and examines knowledge and technology to the extent that they affect such practices;
3.    Critical discourse concerns itself with the power relations and inequalities that are inherent in organizational and societal structures;
4.    Dialogic discourse focuses on the disciplinary practices that operate to create order, knowledge, and power effects.

In their statistical analysis of the literature, Schultze and Leidner notice that of the 94 articles (later filtered down to 78) found and classified in their study, only one was written from a critical-discourse perspective, and two from the dialogic viewpoint; the rest were dominantly normative and partly interpretive, implying that “the negative implications of knowledge, namely its disciplining and dominating effects are left largely unexamined” (p. 230). This has led the authors to note “tendencies to adopt an optimistic view of the role of knowledge in organizations and of the role of information systems in enabling knowledge management”, and to forewarn of the dangers facing research in this area to “become unduly myopic and closed to new ideas” (pp. 230–231). They have hence encouraged IS researchers “to wrestle with the difficult issues of power and conflict that knowledge management might incite” (ibid).

The present work was, indeed, also motivated by similar observations about a noticeable failure among IS researchers (and organizational and management theorists in general) in taking account of power relations in the management of knowledge in organizations. It could, therefore, be construed as a response to the above call by Schultze and Leidner. In fact, we aim to push their observation further and suggest that, above and beyond the risk of myopia, the scholarly business literature has its own regime of truth that has contributed at times, albeit unintentionally, to the creation of misleading social truths.

The scholarly regime of social truth differs in some important ways from the regime of business press reporting regarding the selection of authors, the editorial processes for selecting articles,  and the funding models. While there are numerous variations in detailed practices, the following discussion outlines some of the important features of these two different regimes of social truth. A business magazine, such as Fortune, pays a staff of full-time journalists who write articles based on complex informal negotiations with some of the editors. The journalists' articles may be modified based upon a review by the magazine's editors. In contrast, the articles that are published by peer-reviewed scholarly journals are rarely written by the journals' employees. The authors are usually employees of universities and research institutes. When authors submit articles to a peer-reviewed journal, they may be screened by an editorial staff for suitability for the journal. After an initial screening, they are sent to several scholars with appropriate expertise for review, and who are supposed to serve as independent judges of the quality of an article. While they serve as advisors to the journal's editor, their written reports can also advise authors about ways to improve their articles for publication.

The first peer reviewed journals, first published in 1665, were the  Journal des Scravans and the  Philosophical Transactions of the Royal Society. (Weller, 2001). Until World War II, peer review was most common in  journals of the natural sciences, although today it is commonplace for scholarly journals in all fields. Peer-reviewing practices can vary in important details, such as the number of referees, whether authors can recommend referees, and whether referees know the authors' identities. Further, a journals' acceptance rate may influence the quality of its articles. Some management journals publish less than 20% of the articles submitted (often with some revision), while others may publish a majority of the submitted manuscripts.

Peer review cannot insure the "truth" of a published article. Peer reviewers cannot insure that survey data was properly coded and that statistics were properly calculated. In the experimental natural sciences, peer reviewers cannot insure that meters were properly calibrated and that data was accurately transcribed.  In reviewing  managerial case studies, peer reviewers cannot revisit organizations to insure that authors' claims fairly and accurately reflect the organizational practices that are reported. Peer reviewers can examine an article for its likely importance, its coherence, the plausibility of its methods, the extent to which major conclusions are well grounded either in the data presented or in the conceptual argument that is advanced, etc.

Editorial peer-review has been subject to many criticisms about peer-reviewers’ biases in favoring authors at more elite institutions and in strongly viewing articles through the lenses of their favorite theories. The characteristics of the peer-review system -- and possible improvements -- have been most systematically studied in the field of medicine. Since 1989, the Journal of the American Medical Association and  the British Medical Journal have cosponsored a series of international Congresses on Biomedical Peer Review (Rennie, 2002). Despite its flaws, most scholars seem to consider some form of editorial peer-review the most reliable regime for producing trustworthy scholarship. A brief look at the scholarly literature of the 1990’s would confirm the over-optimism that lurks behind this view.      

Chan and Mauborgne (1997) characterizes Enron and a number of other firms (CNN, Wal-Mart, Compaq, IKEA, Charles Schwab & Co., SAP, etc.) as “value innovators,” as opposed to “technology innovators,” in that they all “focus on expanding existing markets or creating new ones --not beating the competition” (ibid).  Enron, for instance, according to these authors, “has struck upon repeated value innovations, lowering the cost of gas and electricity to customers by as much as 40 percent to 50 percent… while dramatically reducing its own cost structure by, for example, creating the first national spot market for gas in which commodity swaps, future contracts, and other complex derivatives effectively stripped the risk and volatility out of gas prices” (ibid: 46). The crucial question here is to ask about the sources of information (or should we say “knowledge”?) of these authors about, for instance, the reduction in the cost of gas or the cost structure within Enron.

Given what we now know about both of the above claims, and given that the authors do not specify any other sources of information, we are justified to assume that the source was indeed Enron itself. The statement by the authors that, “Enron exemplifies the transition from the production to knowledge economy,” as a statement of social truth could, then, be understood only in this light. Had Enron been a minor company with less of an image, the authors would have certainly, in line with sound scholarly work, depended on alternative sources of information before drawing conclusions about the claims. 

Other writers went even further, suggesting that innovation could be routinized in standard procedures emerging from the lower echelons of a firm.  Sweetman, for instance, counts Enron among the “nimble companies” that “in a world of ceaseless change, … ensure that their strategies will evolve as rapidly as the opportunities arising around them” (2001: 10). The mechanisms for carrying this out, according to the author, are “flexibly organized decision-making mechanisms called Strategy Innovation Routines (SIRs)…[that] emerge as a company struggles with crisis” (ibid). The way the author uses Enron’s example to describe the “emergence” SIRs is worth quoting at length:

When deregulation caused gas prices to fluctuate wildly, Enron s CEO encouraged a small team to borrow from Wall Street’s risk-management techniques and offer fixed-price and inflation-indexed gas contracts to producers and buyers of gas. To avoid potentially disastrous losses, Enron created a comprehensive risk-management system. By ensuring that all possible risks had been addressed in fixed-price gas deals, the company devised a highly effective method for vetting new ideas to relevant departments in the organization. Once the system was proven, Enron leaders encouraged entrepreneurial people within the company to take other new ideas through the risk-management process to prove their viability. The result: successful innovation for long-term contracts, for trading electricity, for setting up power grids in other countries and for reorganizing the entire power-supply system for utilities. (Sweetman 2001; emphasis added)

This passage calls for close attention in many respects, not the least of which is the language it employs in describing the process of the “emergence” of SIR.  Take the assertion that a “small team” presumably borrowed risk-management techniques from Wall Street, created contracts, and, after ensuring that all possible risks are taken into account (how?), disseminated this innovation to other departments. This is obviously anything but a spontaneously emerging pattern described by the author as “rarely designed deliberately by senior managers” (ibid). The language of “encouragement” used in characterizing the relationships between managers and employees seems to be intended to obscure this discrepancy, but recent accounts of the dominant climate in Enron shows otherwise. Cruver, a former Enron employee, for instance, reports that Enron’s social milieu encouraged behavior similar to a casino gambler who risks his life savings hoping to get incredibly lucky. Hard work and talent were not enough to keep one employed. The rule, according to Cruver, was always “no bad news.” The latter requirement was nonnegotiable and took priority over everything else.

Thinking clearly and objectivelywould only place you into harm's way. The immediate goal of higher stock prices and bonuses were all that mattered. Let somebody else worry about tomorrow (Cruver 2002).

In brief, the flow of “knowledge” in Enron seems to have closely followed the channels in which power circulates within and outside the organization. Legal questions about the depth and scope of “knowledge” of Enron's prior management in inflating the value of the company’s assets for the purpose of avoiding the negative impact of a write-down also make sense only if we view knowledge on all dimensions, especially as it relates to power (Eichenwald 2002).  And it is exactly on this point that scholarly accounts of companies like Enron in the 1990’s, by and large, failed. To the extent that scholars writing about these issues relied on these companies as the main sources of information, they were grossly misled. To the extent that they downplayed the conflicts, secrecies, and power games involved in the strategies of these companies, scholarly writings were hugely misleading. Power relations were implicated in the scholarly regime of truth in more than one way.  

Making Sense of Knowledge Management in Practice
How, then, does this view affect the real practice of knowledge management in businesses and organizations? A great deal, we believe; and to demonstrate this we would like to draw on some of the case studies reported in the literature, taking a fresh look at them from the perspective advanced above. 

Cases of Failure
Barth (2000) is a study of a number of knowledge management initiatives that, according to the author, have gone wrong. One such initiative was taken in the research and development group of the food manufacturing Pillsbury Co. of Minneapolis, where 500 scientists, technologists, technicians and support staff served fourteen business teams. A consistency problem with waffles triggered a proposal by a scientist for the creation of a knowledge-sharing forum among different departments. Subsequently, the IT department created a virtual space for this purpose, and the initiators launched a campaign by ending e-mail invitations and asking questions, to which they did not get any response for six months --a clear sign of failure for the initiative. The person in charge of the KM project at Pillsbury explains the failure in terms of lack of incentive for people to invest time and energy to solve “other people’s problems” (Barth 2000: 1). Pointing out, furthermore, that, “some vice presidents frowned upon the idea of ‘their’ people working to help other teams,” he concludes that the “culture” did not reward and recognize knowledge-sharing activity.        

In another case of a large constituent lobbying organization, the management sought to attract more membership by promoting knowledge sharing among employees. So they hired a chief knowledge officer, brought in two consulting firms, and installed a sophisticated intranet system, to find out in the process that collaboration among employees is not enhanced. A KM consultant called up to fix the situation explains the reason in this manner; “There was a sense of never cooperating with another department for fear that if they did, another department might get more money next year while their budget is slashed”(ibid: 3). And the author reiterates the previous conclusion that efforts to improve the situation “were doomed unless the company’s culture changed”(ibid).

Similar conclusions are drawn by Barth(2000) from the study of other cases, invoking always the organizational “culture” as the barrier to the successful implementation of knowledge sharing. Other analysts also use a similar language when explaining cases of KM success and failure. In discussing the issues facing KM projects around the globe, Alavi and Leidner, for example, suggest that a successful implementation of KM requires “profound cultural renovations because, traditionally, organizations have rewarded their professionals and employees based on their individual performance and know-how” (1999: 21).  The language of "culture," it seems, is one that enables analysts to allude to issues of power without discussing specific power relationships. What has happened in both of the cases mentioned above, in our view, is simply that department heads refused to share their knowledge and know-how with other department because they did not want to instigate a reduction in their budget and power with their own hands. Failing to take particular note of this simple fact in all its particularity would result in the persistence of problems, though in more disguised forms. A look at a successful case of knowledge management reinforces this observation. 

A Successful Case
Pan and Scarbrough (1999) report on their study of Buckman Laboratories, which in their judgment represents a successful implementation of KM. Critical of the existing literature that, according to these authors, tends to be “over prescriptive… failing to address important empirical trends,” they define knowledge as “multilayered and multi-faceted, comprising cognition, actions and resources” (p. 360). Based on this view and adopting a socio-technical approach, Pan and Scarbrough then propose an model of knowledge management the details of which is not crucial for our purposes here. What is crucial, however, is that this model enables them to analyze their case in a very concrete fashion that is most clearly reflected in the idea, expressed by lab’s director, that “employees who shared their knowledge would be the most influential and would be sought by others within the company” (p. 369). A top manager explains the mechanism for this in the following manner:

At the heart of knowledge-sharing activities in Buckman is a climate of continuity and trust. This is the most difficult aspect of knowledge-sharing to achieve. If you can’t do it, you can’t succeed. We grew up learning to hoard knowledge to achieve power. (ibid)  
In summing up their study, Pan and Scarbrough take note of the importance of considering “the political dimension of leadership, including the application of rewards and sanctions to overcome resistance” (p. 370).  They draw our attention to what they consider to be “the most important leadership role played by the lab’s leader in the knowledge management arena” --namely, “his ability to ‘manage the managers’ and thus to enroll them as enthusiastic practitioners of knowledge management” (ibid). This view and the language used in its description (influence, power, resistance, enrolling, etc.), we believe, gets to core issues in knowledge management, explaining not only the success of the case under study, but providing also a greater handle on these issues, in contrast with the more diffuse concept of “organizational culture.” 

Complaining about the “go-go culture” of the 1990’s has turned into a blithe intellectual habit of our times  a phenomenon that is understandable in light of persistent economic downturn and its effects on typical middleclass individuals. What is beneficial about such complaints is that they remind us of the rampant extravagance that infected many in the U.S. in unprecedented ways. What is misguided about them, however, is their moral neutrality and social inefficacy: they put the blame on the diffuse notion of “culture” --which is, by the way, a far cry from the rather precise conception of the same word by anthropologists --without articulating the heavily contrasting roles that different actors in different power relationships had in the creation of that culture and, more importantly, in the exploitation of its material fruits.

The genius of Michel Foucault was in his persistently sincere, though not necessarily precise, articulation of the relation between money, knowledge, and power. He sought, as we mentioned earlier, to explain this relationship at the locus of its concrete effect, not in blurry terms. His notion of a regime of truth was, in fact, a useful shorthand for those relationships and the processes and mechanisms that make them effective. Our purpose in this article was to expand and elaborate this shorthand to show how power relationships systematically influence the statements about the social world in an institutionally specific manner --hence the multiple regimes of truth suggested here.

We showed, through the processes, procedures and techniques involved in these regimes in the 1990’s, how each one systematically distorted their statements because of the expected influence of economic power lucrative consulting by the big auditing firms, the prospects of new lucrative investment deals for firms whose stocks a banks’ analysts treated favorably, and the expectation of important advertising revenues by creating a stream of business stories that were -- on balance -- extremely favorable to large businesses and their managers. The cases of Enron provided us with ample evidence to drive this point home. 

We further showed, using the results of Schultze and Leidner’s (2002) study and our own brief survey, the scholarly research has also contributed to the distortion of statements about lavish business practices in its own (innocent) way. This observation suggested this method of producing statements about the social world as yet another regime of truth where power relationships are squarely implicated. Moreover, based on the above conclusions, we took a fresh look at few case studies reported in the knowledge management literature, showing that a discussion of power relationships gives a greater handle on these issues, in contrast with the more diffuse concept of “organizational culture.” Lastly, as Walsham (2001) has argued, we believe that the role of technology, especially new information and communication technologies, in organizational processes should not be seen in isolation from power relationships.

Abolafia, M. Y. (1998). Markets as cultures: an ethnographic approach. The Laws of the Markets. M. Callon. Oxford, Blackwell Publishers/The Sociological Review: 69-85.
Alavi, M. a. D. E. L. (1999). "Knowledge Management Systems: Issues, Challenges, and Benefits." Communications of the Associations for Information Systems 1.
Barboza, D. (2002). Ex-Executives Say Sham Deal Helped Enron. New York Times. NewYork: NY.
Barth, S. (2000). "KM Horror Stories." Knowledge Management http://www.destinationkm.com/default.asp?ArticleID=923
Creswell, J. (2002). Banks on the Hot Seat. Fortune. 146: 79+.
Deetz, S. (1996). "Describing Differences in Approaches to Organization Science: Rethinking Burrell and Morgan and their Legacy." Organization Science 7(2): 191-207.
Drucker, P. F. (1995). Managing in a Time of Great Change. New York, Truman Talley Books/Plume.
Foucault, M. (1980). Power/Knowledge: Selected Interviews & Other Writings 1972-1977. New York, Pantheon Books.
Goodwyn, W. (2002). Profile. All Things Considered. Washington D.C., National Public Radio.
Hamel, G. (1997). Turning Your Business Upside Down. Fortune. 135: 87+.
Hart III, H. (1996). Power Players. Fortune. 134: 94+.
Herdman, R. K. (2002). Testimony Concerning The Roles of the SEC and the FASB in Establishing GAAP. Before the House Subcommittee on Capital Markets, Insurance, and Government Sponsored Enterprises, Committee on Financial Services, Securities and Exchange Commission.
Huseman, R. C. a. J. P. G. (1999). Leading with Knowledge: The nature of competition in the 21st century. London, Sage Publishers.
Longman, P. J. (2002). Bad Press. Washington Monthly. 34: 18-21.
Mulford, C. W. a. E. E. C. (2002). The Financial Numbers Game: Detecting Creative Accounting Practices. New York: NY, John Wiley & Sons, Inc.
Nunn, A. (2001). "Interpreting the 'Knowledge Economy' Cacophony: the extension of commodification to information production, dissemination and storage." Information for Social Change(14).
O’Reilly, B. (2000). The Power Merchant. Fortune. 141: 148+.
Pan, S. L. a. H. S. (1999). "Knowledge Management in Practice: An Explanatory Case Study." Technology Analysis & Strategic Management 11(3): 359-374.
Rennie, D. (2002). Editorial. Fourth International Congress on Peer Review in Biomedical Publication, American Medical Association.
Schultze, U. a. D. E. L. (2002). "Studying Knowledge Mannegement in Information Systems Research: Discourses and Theoretical Assumptions." MIS Quarterly 26(3): 213-242.
Sherman, S. (2002). Enron: Uncovering the Uncovered Story. Columbia Journalism Review.
Skyrme, D. J. (1997). Knowledge Management: Making sense of an oxymoron, David Skyrme Associates.
Stenmark, D. (2002). Information vs. Knowledge: The Role of Intranets in Knowledge Management. 35th Hawaii International Conference on System Sciences, Hawaii, IEEE.
Stewart, T. A. (1997). Intellectual Capital: The New Wealth of Organizations. New York, Doubleday.
Sveiby, K. E. (1997). The New Organizational Wealth: Managing and Measuring Intangible Assets. London, Berrett Koehler.
Sweat, J. (2000). "Information: The Most Valuable Asset." Information Week(803): 213-224.
Walsham, G. (2001). Making a World of Difference: IT ina Global Context. New York, John Wiley & Sons.
Weller, A. C. (2001). Editorial Peer Review: Its Strengths and Weaknesses, Information Today, Inc.
Wilson, T. D. (2002). "The nonsense of "Knowledge Management"." Information Research 8(1).
(Foucault 1980; Drucker 1995; Deetz 1996; Skyrme 1997; Stewart 1997; Sveiby 1997; Abolafia 1998; Alavi 1999; Huseman 1999; Pan 1999; Barth 2000; Sweat 2000; Nunn 2001; Walsham 2001; Weller 2001; Mulford 2002; Rennie 2002; Schultze 2002; Stenmark 2002; Wilson 2002)

(Hart III 1996; Hamel 1997; O’Reilly 2000; Barboza 2002; Creswell 2002; Goodwyn 2002; Herdman 2002; Longman 2002; Sherman 2002)