In a 2007 article published in Armed Forces Journal entitled “A Failure in Generalship,” US Army Lieutenant Colonel Paul Yingling wrote a compelling indictment of currently serving general officers and the impact their failures have had on America’s national security. Yingling offered a solid understanding of the poor outcomes that the system, so defining of the Army officer corps, produces. What Yingling does not do is provide explanations about “why” the system functions in this manner. In our article, we suggest that the US Army’s institutionalized system of reward and promotion, which undergirds a long enduring culture of conformity, is specifically designed to produce the feckless generals that Yingling portrayed.[ii]
The US Army is a complex social system in which entrenched bureaucracies thwart even modest change. This professional institution also possesses an insular culture that has always resisted change, especially any reformation of the officer corps. Historically, strong evidence exists that persistent careerism has led to widespread, unacceptable behaviors among the Army’s officers. Examples include General William Westmoreland’s commission of the Army War College’s 1970 “Study on Military Professionalism.”[iii] Prompted by a note from Lieutenant General William Peers, the lead investigator into the My Lai atrocity, that “something had gone badly wrong within the Army’s officer corps,” the Army War College report confirmed Peers’ observations.[iv] However, it is telling that upon receiving the report, Westmoreland restricted its access and directed that the report be classified. Similarly, in the 1990s, the Army faced a serious exodus of company grade officers due largely to unbridled careerism among the field grade officer ranks.[v] More recently, in a 2011 article entitled “Why Our Best Officers Are Leaving,” author Tim Kane found from a survey of West Point graduates that eighty-two percent believed that the best officers in the Army were leaving. Keane noted that “the military personnel system—every aspect of it—is nearly blind to merit. Performance evaluations emphasize a zero-defect mentality, meaning that risk avoidance trickles down the chain of command.”[vi]
Despite several highly publicized transformations since Vietnam, no changes and reforms have ever addressed or resolved the fundamental problems of careerism or the Army’s culture of conformity. Consequently, the caliber of officers who advance in the Army’s promotion system is open to question. Indeed, the career advancement system, from lieutenant to general, conforms to a rigid mold that begins at selection and ends in a lucrative retirement. How many capable, highly intelligent officers have fallen by the wayside since Vietnam and the First Gulf War simply because they did not fit this mold?
“…the system has to permit more dissent without the sacrifice of careers as the price.”
— Brigadier General Douglas Kinnard, 1976[i]
Why the Culture of Conformity Persists
The Army’s officer promotion system is supposedly designed to objectively evaluate the performance of commissioned officers and to provide them with feedback on their potential. However, it is a closed, top-down system that has been in place for decades with only limited changes and improvements. Since the Vietnam War period, when careerism ran rampant throughout the Army officer corps, the Army has attempted to improve the Officer Evaluation Report (OER) largely to eliminate inflationary ratings. In 1969, for example, a study from the Office of the Deputy Chief of Staff for Personnel (ODCSPER), noted that “there exists a serious lack of confidence by officers in the value and usefulness of the report form.”[vii] The decade long (1972 to 1982) implementation of a new Officer Personnel Management System (OPMS) produced a system that failed to address careerism and an inflationary and subjective evaluation methodology.[viii] Indeed, in the down-sized post-Cold War Army of the 1990s, junior officer attrition was severe and officers were “unhappy, more selfish, and competitive, and less committed and cooperative. The Army’s leadership [was] slow to acknowledge and even slower to address these alarming trends.”[ix] The trends yet continued into the 21st century. A 2001 monograph from the School of Advanced Military Studies (SAMS) noted the feelings of disenfranchisement:
Two major contributing factors caused discontent and possible attrition among junior officers. The first, the lack of formal counseling from senior officers, has a decidedly negative impact on how junior officers view the US Army. . . The second is the perception that senior raters pool all Captains in their organization in order to build their rating profile. The Captains’ perception is that pooling produces standardized OERs ranking all Captains as center of mass performers during the times that they hold non-branch qualifying jobs. Pooling tends to disregard aptitude and talent according to the results from the focus group survey. Only Captains serving as company commanders are eligible for top ratings and this practice is viewed as grossly unfair.[x]
Today, four decades after Vietnam, the current evaluation system and OER appear to offer few improvements over previous methods. Similar to its predecessors, the current OER is a two-page document that reflects the assessments of the rated officer’s “rater” and “senior rater,” that is, the supervisor one and two levels up the chain of command. Officers receive evaluations annually unless there is a “change of rater” within the one-year time frame. Although the rater and senior rater provide both comparative and narrative feedback, the senior rater’s portion is the most important. Here, the senior rater categorizes the evaluated officer in one of four levels or “blocks”: above center of mass, center of mass, below center of mass/retain, or below center of mass/ do not retain. The senior rater is limited to placing no more than fifty percent of his or her rated officers in the above center of mass category. Most senior raters consistently approach the fifty percent threshold and very few reports fall into either of the below center of mass categories.
The most recent and disconcerting change to the system is that only four of the ten officer grades – major through brigadier general – receive this definitive feedback. The other six grades only receive narratives. Consequently, the value of the report is subject to the writing ability of the rater and senior rater as well as the reader’s interpretation. At the same time, this evaluation system is the primary determinant of promotion, assignment, and retention in the Army officer corps.
The military retirement system is an all or nothing system in that no vesting occurs after a reasonable period of service as in most civilian retirement plans. Therefore, if an officer chooses to leave or is forced to leave prior to completing twenty years of service, he or she receives no retirement benefits, immediate or deferred. Assuming that an officer retires after twenty years of service, lives to age seventy-two, and receives a pension of $3000 per month, at stake are $1,080,000 in addition to medical and other benefits as a military retiree. For most officers who come from middle or lower middle class socioeconomic backgrounds, financial security is an economic reality that is impossible to ignore. Consciously or unconsciously, commitments to a twenty-year career and the benefits that await affect behaviors and thought processes. The “up or out” promotion system creates tremendous competition and anxiety, especially among mid-grade officers. The institution’s culture of conformity shapes officers into molds that resemble the traits preferred by the raters and senior raters. Consequently, espoused and practiced values often diverge. Espoused values such as candor, courage, and integrity as practiced are stressed under these immense pressures. These three subsystems – the initial selection process, the officer evaluation and promotion system, and the ‘all or nothing’ retirement system – give structure to Yingling’s observation that “it is unreasonable to expect that an officer who spends twenty-five years conforming to institutional expectations will emerge as an innovator in his late forties.”[xi]
The barriers to the creation and development of the kind of generalship that Yingling advocates are very real. Any expression of substantive disagreement with superiors who seek affirmation rather than information from a subordinate could be career ending. Instances of disagreement or providing disconfirming data often result in the subordinate being labeled a negative thinker who is disloyal to the boss or is a non-team player ‘who just doesn’t get it.’ The path of least resistance is to go along, affirmed by the commander and the institution, and assure a million dollar payday. In essence, officers pursuing a military career don a “million-dollar muzzle” in order to pass through the promotion gates that will take them to twenty years of service. Certainly, there is no conscious or malicious intent behind the thought processes that create this behavior. On the contrary, grounded in a real desire to serve the nation, it simply becomes a rational exercise of enlightened self-interest and institutional affirmation. Sadly, few organizations, least of all an authoritarian one such as the military, hold their internal provocateurs dear.
Conformity is a powerful force. In the case of the Army officer corps, it presents an obstacle to significant change. Yet change is a permanent condition that occurs at speeds unanticipated in years past. In the 21st century, the rate of change is so frequent that the effective application of military force is often determined by how quickly the Army can adapt to shifting contexts, fluid situations, and most certainly complex problems that have no easy technical solutions. Indeed, “adaptation” has become a popular buzz word in the Army’s current vernacular. But do Army senior leaders truly understand what adaptive challenges mean?
Finding solutions to adaptive challenges requires decision-making processes that are antithetical to the traditional authoritarian role of “the Commander.” Ronald Heifetz has stated that adaptive work is required when the problem is not easily defined and the solutions are not readily discernable.[xii] In such circumstances, successful adaptive work requires authority figures, in our case senior Army officers, to ask or solicit the right questions rather than offering preconceived solutions. Adaptive leaders resist pressures to orient people too quickly and they allow norms to be challenged. In our view, authoritative practices, such as the “commander’s intent,” may be quite appropriate for technical problems but could serve to derail successful responses to adaptive challenges. Unfortunately, the Army’s culture of conformity impedes units from becoming learning organizations (critical to problem definition and solution implementation), and creates a climate where the “devil’s advocate,” regardless of rank, is ostracized. The unforeseen requirements of the Army’s global mission into the 21st century will demand officers who challenge the prevailing conventional wisdom at all levels. “Yes men” who mimic the behaviors of their bosses and predecessors should not be welcomed.
By citing Michael Howard that “in structuring and preparing an army for war…. the important thing is not to be too far wrong, so that you can put it right quickly,” Yingling stressed the criticality of visionary generals. However, “putting it right quickly” requires senior officers who possess the capability and capacity for adaptive thinking. Unfortunately, the Army officer education system, which plays a critical role in an officer’s advancement into the senior ranks, does not promote reflection and adaptive thinking. For example, in her 2002 book on the Army War College, Judith Stiehm found that although the Army War College curriculum explored “change,” it was not provocative. Guest speakers, on average two per week throughout the year, were all conservative and “one doubts that students were ‘provoked’ by any of the speakers.” [xiii] Stiehm observed the heavy influence of the Army Chief of Staff and noted that the Chief selects the Army War College Commandant, whose “views are likely to be experienced as directives.” Among the students in her cohort, all but one shared the same Meyers-Briggs personality type indicator (judging and reasoning over feeling and seeing possibilities).[xiv] Overall, Stiehm described the Army’s preparatory school for colonels and generals as a “culminating experience” rather than a launching experience. In our view, reflection and adaptive thinking are missing; not surprising when one considers that all of the students and faculty have successfully worn the million-dollar muzzle.
The Army officer corps is actually a conglomeration of subsystems organized as occupational specialties (e.g. infantry, cavalry, etc.) which exhibit fraternal loyalties that overtly and covertly enforce conformity. Compounded by historically rich unit heritages, “mission first” pressures have become institutionalized and are taken for granted among all officers as a given. To cite a popular colloquium: Failure is not an option. For officers looking to win the approval of their senior commanders, means versus ends can become a very real dilemma. In such an environment, dissension is viewed as a virus of sorts. Even in retirement, former general officers are expected to wear a muzzle. For example, the 2006 so-called “revolt of the generals” – the public call by six retirees for Secretary of Defense Donald Rumsfeld’s resignation – sparked a very vocal reaction from serving general officers who believed that retired generals should never voice such dissent.[xv] This mentality of rewarding conformity and reinforcing the status quo also goes a long way toward explaining the Army Senior Mentor Program. The program is “careerism on steroids,” where retired three- and four-star officers earn $440 an hour and up to $179,000 a year as Pentagon “consultants” while also receiving full pensions and employment as defense contractors.
The million-dollar muzzle is explained by nothing more sophisticated than informed self-interest in response to a social structure. This self-interest is further explained by several more broadly applied and academically informed social theories. Three theories are particularly relevant: the normalization of deviance, outlining how organizations develop cultures to unconsciously justify errant or amoral behavior; the immunity to change, which explains the real reason why change is so difficult and temporary; and the “undiscussability” of the “undiscussable.”
A culture of conformity can have disastrous consequences and outcomes. At the very least, conforming behaviors create a type of “groupthink” that may send decision-making processes down the wrong paths and block adaptive thinking. When this happens we see the “normalization of deviance.”[xvi] Diane Vaughan, a sociologist specializing in organizational failures, coined this term in her exploration of the 1986 Challenger launch disaster. Vaughan found that a culture of conformity was the fundamental root cause of the accident. Such a comparison is fair because NASA and the US Army share many similarities. Both are large technological, bureaucratic government organizations. They each have elite and distinct cultures whose members view themselves as belonging to a profession and to professional institutions. Both have enjoyed high levels of public trust. With the Army’s strong emphasis on force protection, it has shared prominence with NASA as an exemplar of risk assessment and operational safety.
Vaughan found that NASA decision-makers proceeded with their work as if nothing was wrong even when they were repeatedly faced with evidence that something was indeed wrong. She attributes the leadership failures to the institution’s values and culture. NASA officials witnessed a myriad of small infractions that, over time, added up to something big–and never recognized that the small infractions were serious or that they themselves were not making sound decisions.[xvii]
Similarly, Army senior leaders, as fully vested members of an insular organizational culture, appear to promote an ever-entrenched state of groupthink. Despite formal decision-making processes that outwardly appear inclusive, members of the command routinely and often unknowingly acquiesce to the views of their senior commanders. Senior commanders then play out their expected roles as all-knowing, strong decision-making leaders by seeking affirmation rather than information or honest subordinate input. Consequently, poor decisions and other consequential outcomes, such as infractions of core values, are far too easily rationalized and explained away. Self justifications abound as the followers–the lower and middle-level leaders–“spin” the incidents to put the institution in the best light or to protect “the old man” and the unit or command. Even when official investigations of crimes find officers guilty or culpable, seldom are those involved in the leadership chains of command held accountable to the same degree and proportions as those below them. Recent examples include the war crimes committed at Abu Ghraib. As the incident came to light, the Army was quick to point out that the perpetrators were enlisted reservists, and no soldier higher than staff sergeant faced convictions, despite culpability reaching into the general officer ranks, both active duty and reserve.[xviii] Similarly in Afghanistan, the death of eight soldiers in the Battle of Wanat two years ago were attributed in Army official reports as platoon-level mistakes rather than the alleged neglect and poor planning and support at battalion, brigade, and division levels of command. Finally, the heavily reported cover-up of the death of Pat Tillman reached well into the general officer ranks. In the Tillman investigation, the two-star general implicated later obtained four-star level command in Afghanistan.
In their book entitled “Immunity to Change,” Robert Kegan and Lisa Laskow Lahey argue that our individual beliefs combined with collective mindsets in organizations create a powerful immunity to change.[xix] First, they suggest three levels of leader development: the socialized mind, the self-authoring mind, and the self-transforming mind. Yingling’s idealized general would exhibit the self-transforming mind that “can stand back from its own filter and look at it, not just through it,” and a mind that “both values and is wary about any one stance, analysis, or agenda.”[xx] Unfortunately, both of these capabilities are discouraged, if not destroyed, by the system that produces the Army’s generals. Second, Kegan and Lahey argue that the barrier to change is often the untested, usually unspoken “collective big assumption.”[xxi] For the million-dollar muzzle, that assumption is that officers will not complete a twenty-year career if they do not adapt to the authoritarian, top-down, compliant behavior of the Army culture.
A third well-established set of social theory that helps explain the “million-dollar muzzle,” comes from the work of Chris Argyris. Argyris enriches this discussion with his construct of “dialogues” that is bounded by “undiscussables”: the catalog of issues, topics, view points, and questions that cannot be raised in organizations without adverse consequences being visited upon those who dare to raise them. Argyris writes:
In order to achieve organizational excellence, learning, competence, and justice are a much more realistic foundation than are morale, satisfaction, and loyalty. The first foundation, learning, pinpoints how errors are detected and corrected, especially errors that are complex and potentially embarrassing and threatening. Competence means solving problems in such a way that they remain solved problems and increase the organization’s capacity for future problem solving. Justice is based on a set of values and rules–in this case about organizational health–that apply equally to all employees, no matter what their organizational position.[xxii]
We would argue that the million-dollar muzzle serves to thwart learning, competence, and justice as described by Argyris and hurts morale and trust. Officers who don the muzzle simply become slaves to loyalty.
As Stephen R. Covey, the internationally respected author of “The 7 Habits of Highly Effective People,” noted, “systems will overwhelm rhetoric or good intentions ‘at the end of the day.’”[xxiii] The systems that we have described here collectively drive accessions, evaluations, assignments, schooling, and promotions. Together, they form an insular culture protected by a handful of institutional stewards who are deeply entrenched and unlikely to change without fundamental reforms of these systems and structures. Yingling made several recommendations to address “A Failure in Generalship” but predominantly saw hope for improvement in the actions of Congress. We strongly disagree. True reformation of systems and structures will require the removal of persistent organizational barriers to change and a radical new approach to senior leader development.
In our view, Congress experiences similar forms of bureaucratic inertia that obstruct effective change. In fact, it has a vested interest in the status quo. Consequently, the Army can expect little change initiation from that direction. However, a truly adaptive and transformational leader appointed as Secretary of the Army could initiate and direct a fresh review of Army officer professionalism. As with many professional institutions, cultures are difficult to change. Since Vietnam, the Army’s conservative culture has been impervious to effective change despite well intentioned efforts to push the Army through several transformations. It is telling that the most significant findings found in the Army War College Study on Professionalism in 1970, specifically poor leadership, risk aversion, and unbridled careerism, remain true today. After four decades, we hold little hope that Congress or the officer corps can sufficiently reform itself. To expect that it can is like asking turkeys to arrange a traditional Thanksgiving Day dinner.
However, a progressive Secretary of the Army, one with little obligation or allegiance to politicians and the institutional Army, could initiate a deep, introspective review that would lead to meaningful reforms. First, reform the officer evaluation system to include a viable 360-degree review of evaluated officers. The current evaluation system gives total control over the evaluation of the officer to the rater and senior rater, resulting in only a one-dimensional perspective of the rated officer. The 360-degree evaluation system integrates the officer’s peers and, most importantly, his or her subordinates. How can an officer’s leadership abilities be evaluated without the input of the followers? The 360-degree evaluation distributes power, broadens perspective, and democratizes an autocratic process. Additionally, the OER system should offer definitive (e.g. blocked rankings) of all 10 officer grades rather than the four discussed previously.
Second, restructure the retirement system to vest officers at the ten- or twelve-year point. This change in the retirement system would be a large step in loosening the million-dollar muzzle. Midgrade officers could speak up with integrity and courage without placing everything on the line. The costs associated with this change could be mitigated by restricting this benefit only to officers who were in the top block (as a result of the 360-degree evaluation process), those high performers most likely to have dissenting views and great potential.
Third, overhaul the curriculum and the educational goals of the Army War College. Instill a culture of reflection that results from a system of academic rigor. The Army War College is currently structured as an institution that perpetuates “the good” at the expense of “the great.”[xxiv] We believe that the Army War College experience should be a final test rather than a ticket punch. Today, the Army War College is essentially a culminating experience for senior officers who have successfully worn the million-dollar muzzle. At issue here is an apparent confusion between training and learning. To date, the social and behavioral sciences have produced numerous studies to show that true transformational and adaptive leaders head up and promote learning organizations. Training results in the successful performance of technical tasks. Learning has occurred when new adaptive behaviors lead to sustainable change. To this latter point, explore the military history of other nations, past and present, to seek out exemplars of successful general staff schools that adopted similar measures. The German General Staff system of the inter-war period is one example that comes to mind.
A successful revamp of the Army War College educational system would begin at the top. Specify that the commandant assignment is a terminal position in order to free him or her from the pressures of perpetuating expected institutional viewpoints and have the commandant report directly to the Secretary of the Army. Additionally, the Army War College should have a deputy hired from academia with a PhD in the humanities to strengthen the curriculum, to act as a provocateur, and to oversee the academic evaluation system. The deputy commandant would also strengthen the faculty, inject more rigor into the curriculum, and revamp the extensive guest speaker program to ensure participation by a wide range of professionals who are critics of the existing Army culture and perspectives. Students should be forced out of their comfort zones by speakers such as Noam Chomsky, Andrew Bacevich, Chris Argyris, Ronald Heifetz, and David Walker, for example. Historically, the speakers have been predominately status quo advocates who generally patronize the institution and the students.
Finally, the most meaningful role that Congress could play would be to emplace a program of mandatory national service. One option of the program would include service in the Army. Despite claims to the contrary, the All Volunteer Army has not succeeded in fielding a quality Army that is representative of the society it serves. In fact, the majority of our soldiers today come from the third and fourth socioeconomic quintiles of our citizens, while the first socioeconomic quintile is virtually absent–some may say AWOL.[xxv] The result is that the best and brightest of our nation never experience military service and are therefore never in the pool of candidates to aspire to or achieve general officer rank. Even if they leave after the initial enlistment they would bring a heightened level of curiosity, morality, introspection, and creative thinking while serving out their contracts. This new voice from the bottom, combined with a bold, progressive Secretary of the Army at the top, would essentially squeeze the existing culture from both ends to affect true reform and transformation. Engaging this first socioeconomic quintile also produces a more informed citizenry, democratizes the decisions regarding military issues, and creates a pool of civilian leaders capable of making better decisions around military and national defense issues (in the 112th Congress only 20.9% of the members are veterans).[xxvi] We have devolved from a nation where” all gave some, some gave all” into one where “some gave all, most give nothing.” At the very least, a serious debate on mandatory national service would question the fundamental assumptions that drove the 1973 decision to launch the All Volunteer Army and would bring to light the realities of a 2011 world. These realities were reflected in Yingling’s 2007 concern for “the long war.” He warned then that “the quantity and quality of manpower required may call into question the viability of the all volunteer military.” We could not agree more.
A final paradox in the “failure of generalship” may be manifested in the question why, in recent memory, no general officer has resigned in protest or has been dismissed for incompetence. Can it be that the system is so effective and insightful that the senior leaders of our Army are not subject to the same moral, intellectual leadership and performance dynamics as their peers in other industries and sectors of our society? Or is it more likely that we have created a system where the million-dollar muzzle molds a culture of conformity to produce the generals that Yingling so aptly described?
[i] Douglas Kinnard, “The Vietnam War in Retrospect: The Army Generals’ Views,” in Journal of Political and Military Sociology, 1976, Vol. 4, (Spring): 17-28, 28.
[ii] Paul Yingling, “A Failure in Generalship,” in Armed Forces Journal, May 2007.
[iii] Department of the Army, Study on Military Professionalism, Carlisle, PA: US Army War College, 30 June 1970.
[iv] Brian M. Linn, The Echo of Battle: The Army’s Way of War, Cambridge, MA: Harvard University Press, 2007, 43.
[v] David McCormick, The Downsized Warrior, New York, NY: New York University Press, 1998, 136.
[vi] Tim Kane, “Why Our Best Officers Are Leaving,” The Atlantic, January/February 2011, 81. Interestingly, a zero defects mentality is noted in every decade since Vietnam.
[vii] As reported in The Officer Personnel Management System–OPMS. Department of the Army, 1972, E-2.
[viii] David McCormick, The Downsized Warrior, New York, NY: New York University Press, 1998, 104. Revisions in 1972 and 1975 failed to address the inflation problem.
[ix] Ibid., 118.
[x] Marvin W. Williams, The Relationship of the Officer Evaluation Report to Captain Attrition, School of Advanced Military Studies, United States Army Command and General Staff College, Fort Leavenworth, Kansas, 2001, 2.
[xii] Ronald Heifetz, Leadership Without Easy Answers, Cambridge, MA: Belknap Press of Harvard University Press, 1994.
[xiii] Judith Hicks Stiehm, The U.S. Army War College: Military Education in a Democracy, Philadelphia, PA: Temple Univ. Press, 2002. 141-144
[xiv] Ibid., 148. These findings are consistent with a similar Air Force study of the same period that found a majority of leaders who prefer logical thinking and structure (TJ). See Dianna Lea Williams, “Frequencies of Myers Briggs Type Indicator (MBTI) Among Military Leaders” in The Journal of Leadership Studies, 1998, vol. 5, number 3. While one could argue that military service attracts such personalities, the glaring absence of the other 14 types suggest that STJs are most likely to conform to the prevailing culture.
[xv] See Perry Bacon, Jr., “The Revolt of the Generals,” Time Magazine, April 16, 2006.
[xvi] Diane Vaughan, The Challenger Launch Decision: Risky Technology, Culture, and Deviance at NASA. Chicago, IL: University of Chicago Press, 1996, 266-272.
[xvii] Ibid, see especially Chapter 9: “Conformity and Tragedy.”
[xviii] We distinguish here between criminal convictions and non-judicial punishment. It is true that BG Karpenski was demoted to one rank lower (Colonel) but was permitted to retire with full benefits.
[xix] Robert Kegan and Lisa Laskow Lahey, Immunity to Change: How to Overcome It and Unlock Potential in Yourself and Your Organization, Boston, MA: Harvard Business Press, 2009.
[xx] Ibid., 19-20.
[xxi] Ibid., 58.
[xxii] Chris Argyris, Overcoming Organizational Defenses: Facilitating Organizational Learning. Boston, MA: Allyn and Bacon, 1990, xi.
[xxiii] Stephen R. Covey, The 7 Habits of Highly Effective People:
[xxiv] See Jim Collins, Good to Great: Why Some Companies Make the Leap and Others Don’t. New York, NY: Harper Collins, 2001. “Good is the enemy of the great,” 1.
[xxv] Kathy Roth-Douquet and Frank Schaeffer, AWOL: The Unexcused Absence of America’s Upper Classes From Military Service – and How It Hurts Our Country, New York, NY: Harper Collins, 2007.