In February 2007, late-night talk-show host Conan O'Brien challenged tennis savant Serena Williams to a match. O'Brien sprang the idea on her, stripped to tennis clothes, and humiliated the champion athlete to the delight of his audience. Williams, an Olympic gold medalist, winner of virtually every major tennis tournament in the world, could be good humored about it. She was not actually beaten by a less-than-athletic talk-show host in real tennis. The game played was a simulation using Nintendo's Wii system.
The action projected from a computer display as she and O'Brien manipulated controllers. It looked a little like real tennis; they both lunged and stroked their controllers not unlike one swings a tennis racket. The rules were the same, and the court was beautifully rendered. According to the CBS News Showbuzz Web site, even the projected players looked eerily like the real O'Brien and Williams.
Still, no one suggests that O'Brien would actually be able to play at Williams' level on a real tennis court. Wii does not require one to hit a real ball with a real racket; it never demands the physicality and timing of a true game played on a real court. Its seduction is that it allows one to imagine playing tennis at the level of Serena Williams but without the thousands of hours of practice it takes to master the craft. It's much easier, but still fun.
More than a Game
The training simulations now used in the U.S. Navy are much more than amusing games. These systems make real contributions toward overcoming significant challenges. For example, the Littoral Combat Ship training concept envisions the entire crew trained within a ship simulator. Some level of qualification is conceived as occurring before ever setting foot on board. P. M. Ludwick wrote on the Navy's official Web site in 2006 that the actual ship can be actively employed while the replacement crew trains ashore. The returns are the obvious fuel savings, reduced transit time into areas of operation, more effective individual deployment cycles, and less family separation.
Beyond that specific application, distributed learning systems (training with Web-based simulation) now permeate the training environment. As early as 1999, H. H. Bell wrote for the Association for Computing Machinery that the systems' ability to replace classroom instruction vastly reduces the costs associated with transferring Sailors and officers to and from schools distant from their duty stations and makes complex training accessible on a day-by-day basis. Flight training simulators have long played a vital role in gaining experience for pilots. A 2002 RAND Corporation monograph says that trainees can practice high-risk, difficult, and emergency procedures under the watchful eyes of skilled instructors and without physical risk to equipment or pilots.
But at some level, all these examples are still versions of O'Brien's Wii. At some point, like hitting a real tennis ball with a real racket, one must pilot a real aircraft, dock a real ship, or shoot a real missile. The obvious question is, "How much live training is required?"
However, beyond the obvious, there is another risk inherent to simulation. Too much simulation threatens a subtle but crucially important strength, one critical to maintaining perspective on simulation's value. Simulation, if overvalued, fails to create the culture of craftsmanship required in naval warfare. Beyond all the considerations of how much live training is enough, the issue of craftsmanship and expertise must be considered to avoid an undervaluation of real experience.
Before looking at simulation, the ideas and importance of expertise and craftsmanship bear some consideration. Craftsmanship has always been a fundamental and assumed part of maritime and naval practice. Normally associated with the idea of making crafts, or construction of useful objects, in the naval craft one sees it partly articulated with the words seamanship or airmanship. But the concept actually goes far beyond those two words or the idea of hand-crafting something.
The Need for Craftsmanship
In his 2008 book, The Craftsman, Richard Sennett studies the nature and value of craftsmanship as it relates to society. He says, "Craftsmanship names an enduring basic human impulse, the desire to do a job well for its own sake." Sennett sees craftsmanship as an innate element in everyone, "an enduring human impulse." He underscores the importance of craftsmanship, saying that it "cuts a far wider swath than skilled manual labor; it serves the computer programmer, the doctor, and the artist; parenting improves when it is practiced as a skilled craft, as does citizenship."
But Sennett does not portray craftsmanship as some sort of mystical state achieved through genius. He links it to the idea of expertise and makes a profoundly important observation in the consideration of simulation. He cites 10,000 hours as "a common touchstone for how long it takes to become an expert." These hours are filled with countless repetitions to hone gradually increasing skills until the hand and brain achieve mastery. More important, he places enormous value and responsibility on the expert, dividing expertise into two types, social and anti-social. The social expert both arises from and creates a productive community, or a community of craftsmanship. The anti-social expert acts against it.
Sennett's anti-social expertise requires little attention in the context of this discussion. Suffice it to say that this expertise acts against organizational excellence through the process of shaming others and what Sennett calls invidious comparison. Social expertise, however, is vital to organizational excellence. It seeks ever-improved processes, achieves high standards of performance, and encourages innovation. Unfortunately, social expertise is most vulnerable to harm from misuse of simulation in training.
To Sennett, "the well-crafted institution will favor the sociable expert." This well-crafted organization is one that will develop "whole human beings, . . . it will encourage mentoring, and it will demand standards in language that any person in the organization can understand." What he means is that the standards will be transparent or public, but at the same time intelligible to non-experts.
This notion of standards is a vital principle to craftsmanship as a whole and critical to Sennett's idea of a well-crafted organization. Craftsmanship, he writes, "focuses on objective standards, on the thing itself." It depends on the expert to be able not only to define the standards, but articulate them in a manner everyone can embrace. The expert, then, sets the standards of the craft, and the craft's efficacy is dependent on those standards. Without the expert, no well-crafted institution can exist.
The Elements of Expertise
This is no small responsibility for an expert nor is it a simple achievement to become one. To understand this complexity, Sennett first describes elements common to all expertise. Expertise is characterized by intrinsically valuable skills in analysis and observation. He would identify these as applicable across a broad variety of fields; being an expert is an expertise in itself. In addition, the expert must have profound knowledge of some particular field and become conversant "with knowledge that allows (him or her) to see beyond the elements of a technique to its overall purpose and coherence. . . . It is the knowledge in which the making and fixing are parts of a continuum."
Beyond even that, sociable expertise maintains the characteristics inherent in a well-crafted organization: objective, transparent standards, focus on the whole human being, good practices, and mentoring. The point is that craftsmanship in an institution is nurtured by its experts, and their expertise reflects an institution that values and demonstrates objective, transparent standards. Moreover, in the case of sociable expertise, standards rise out of the expert's ability to see beyond the limits of technique to the technique's actual purpose.
Simulation as a tool, then, appears to have great potential to support the culture of craftsmanship described here. The expert could use the simulation tools, the computer, and the training environment to shape technique, all the while setting objective, transparent standards that reflect the actual purpose. The expert could do this because he or she is inculcated with this standard and has both the knowledge and experience to judge performance.
Simulation Savings
According to a December 2004 National Defense article by S. I. Erwin, if simulation can allow ships to skip going to sea for just one three-week period, 4,000 barrels of fuel might be preserved. A 2006 Department of the Navy operations and maintenance estimate said that if just 1 percent of combined training-related expenses from the Fiscal Year 2007 budget proposal could have been saved through simulation, $27 million could have been preserved for other uses.
The nature of modern naval warfare is beginning to resemble something like an Internet game. If so, then simulation produces fidelity nearly matching the actual conditions of naval warfare. And in a sense, simulation is and always has been pervasive in training for warfare; nearly the only training not involving some level of simulation is that training found on an actual battlefield, according the previously mentioned RAND report. Computer simulation, it can be argued, simply adds another level. Finally, simulation might be a superb tool to achieve exactly the repetition required to gain expertise. Looking at these benefits, the only important determination would seem to be deciding the minimum live training required.
RAND analyzed that very issue. While the report applies a detailed analysis and develops both metrics and methodologies, the frustration inherent to the study focuses on two elements, proficiency and objectives. The monograph struggles with the concept of proficiency and how to determine both the nature and shape of the training required to achieve proficiency. The study states:
[I]t is hard to measure training proficiency. The training matrices used by the Navy designate minimum proficiency qualification standards based on the professional, albeit subjective, judgment of the trainers. If the standards are achieved, so is training readiness. Thus, the goal of unit training is not to increase proficiency but to meet a minimum standard.
In the end, looking at tradeoffs to attain this vague proficiency, the study fell short of making concrete recommendations. It concluded: "Different answers could emerge based on what the desired objective is. The goal must be determined. Is it to reduce cost in achieving maximum proficiency, or is it to reduce the cost of achieving a given proficiency threshold?"
In other words, the study pointed out that the transparent, objective standards inherent to sociable expertise could not be found to measure proficiency. Instead, only a minimum standard could be identified. Thus, it could be inferred that the sociable expertise crucial to a Sennett's well-crafted institution was absent or negated. The study did not consider the elements inherent in establishing the expertise Sennett advocates.
Computer-Aided Design
Beyond the inferred issue of expertise, Sennett directly tackles other issues regarding simulation. He examines computer-aided design (CAD) as it applies to the craft of construction and architectural design and notes that CAD brought with it immense advantages of speed and the ability to resolve complexity and presentation. But he also notes two unanticipated limitations.
The first he describes as a disconnect between simulation and reality. He found that projects did not take into account the actual sensation of the elements. As a result, some created unusable outdoor areas, as CAD provided a "poor substitute for tactile experience." Sennett places these shortcomings squarely on the architects' individual failure to make actual site visits and surmises they were lured from this by the convenience of simulation. His more subtle and interesting point is that these errors are at least partly because of the nature of the simulation; CAD tends to hide problems rather than expose them. It is now accepted that no matter how good the simulation in architectural design, one must have tactile knowledge of the site to understand how the building fits into the environment.
The second limitation with CAD he calls overdetermination. CAD is so capable it designs to a level of efficiency unattainable to a human alone. On the surface this may seem desirable, but it takes away from the process "the tactile, the relational and the incomplete." Sennett's point is that it removes those learning elements from the craftsman. The result can be outcomes that are too narrowly designed to allow adaptation. In other words, the CAD makes the design inflexible and the outcome inexpert and unusable. It replaces the physical and intellectual element of expertise with simulation of those elements. It makes Conan believe he really can play tennis as well as Serena.
The O'Brien Effect
Overdetermination and the disconnect with reality are two very real dangers in warfare training with simulators. At one level, unless properly applied, they can have the obvious effect of training the student to excel at simulation without being able to excel at the actual task. This might be termed the Conan O'Brien effect, except that O'Brien had no illusions he could beat Williams in a real tennis match.
But imagine what it would be like if no one knew anything about real tennis. Imagine if the simulation was so good, the casual observer would think that activity actually represented a demonstration of the skills one needs to beat Serena Williams at her game.
The sophisticated simulators in use in the Navy today could create that very illusion. Here's why the safeguard against this outcome, sociable expertise, is also threatened deeply by the overuse of simulation. Without investing the tens of thousands of hours and the thousands of repetitions it takes to create true expertise and investing those hours in live activity, social expertise will only know how to simulate. There may be expertise, but it will not meet Sennett's criteria of seeing "beyond the elements of a technique to its overall purpose and coherence." In short, unless someone is around who really knows how to play tennis, all the mentoring focuses on the use of the Wii.
At another level, the assessment of combat effectiveness and efficacy might be harmed through the assessment of simulation. Not only will the practitioner gain a useless expertise, but the assessment of the very nature of battle may simply depart from reality. Enormous benefits are to be gained from simulation in training for naval warfare. Costs can be cut. Repetitions can be attained to raise proficiency. High-risk evolutions can be practiced. But if used without a firm commitment to live training, simulation risks creating expertise in simulation and craftsmen whose skills are incomplete. If sociable expertise and a culture of craftsmanship can be maintained, however, objective standards will reflect real values and proficiency. The result will be culture where the ideals of craftsmanship, seamanship, and airmanship are maintained by social expertise in the real skills needed to meet demands of naval warfare. Serena Williams knows this for certain. Conan O'Brien also knows he can't match her in a real contest. Our Navy needs to be at least that realistic.
H. H. Bell, (1999). The effectiveness of distributed mission training. Association for Computing Machinery. Communications of the ACM, 42(9), 73-78. Retrieved May 1, 2008, from ABI/INFORM Global database. (Document ID:44310384).
Department of the Navy (DON) (2006), Operations and maintenance, Navy, O-1a exhibit, Fiscal Year (FY) 2007 budget estimates submission, justification of estimates. Washington DC, Government Printing Office.
S. I. Erwin (December 2004), "Sailors move from classrooms to shipboard simulators," National Defense, 89 (613), pp. 57-58. Retrieved May 16, 2008, from ProQuest Military Module database (Document ID: 769783411).
Judy Faber, (2007), "Conan beats Serena at tennis?" Showbuzz CBS News Web site, Retrieved 28 May 2008.
J. W. Hedge, W. C. Borman, M. J. Bourne, "Designing a system for career development and advancement in the U.S. Navy," Human Resource Management Review, Volume 16, Issue 3, September 2006, pp. 340-355.
P. M. Ludwick, (2006) "New LCS training process allows crew members to man ship anytime, anywhere," U.S. Navy Official Web site Washington, DC: Department of the Navy, Retrieved 31 May 2008.
J. Schank, H. Thie, C. M. Graf, J. Beel, J. M. Sollinger, (2002) Finding the right balance, simulator and live training for Navy units. Santa Monica, CA: RAND Corporation. Retrieved 28 April, 2008.
Richard Sennett, The Craftsman, (New Haven & London: Yale University Press, 2008).