Thursday, May 05, 2005

The Ethics of Video Game-Based Simulation

Play the podcast (downloadable audio file in mp3 format)

While video game-based simulation is gaining ground in interactive, multi-user distributed online learning, underlying ethical and philosophical issues remain unaddressed. These are, in some cases, quite troubling, because learner lives can be at stake. In other cases, profound issues dealing with instructional strategies and design are often only tangentially touched upon. This article identifies the ethical issues involved in using simulation in education and training, explores some of the conceptual and philosophical bases, and proposes ways to address the problems and adjust instructional strategies.

Specifically, this article examines the role of video game-based simulation used to prepare members of the military for combat situations. As opposed to the simulations of the past which focused on driving fast cars and piloting aircraft, the new generation of video game-based simulations focus on player-player interaction, and derive much from first-person shooter games. In contrast with the earlier versions of games, in which the goal was to accumulate enough points to go to different worlds or "levels," the new games combine the "sim" world-building experience, with customizable avatars to reflect the player's constructed persona, weaponry, and team members. Incorporating artificial intelligence, the game allows computer-controlled soldiers the ability to respond to student commands from a matrix of possibilities, each selection weighted via algorithms that build in probabilities and randomness.

Perhaps the most well-known of these video game-based simulations are the Full Spectrum Warrior and Full Spectrum Command, which were developed by the University of Southern California's Institute for Creative Technologies (ICT), employing the skills of Oakland-based Legless Productions' project management, Santa Monica's Pandemic Productions for the game development, and Sony Imageworks for the art assets. Not only have these two games received accolades from the Army; they have also become full-fledged commercial successes. Full Spectrum Warrior runs on Microsoft Xbox, while Full Spectrum Command is a PC-based game, accommodating multi-player as well as single-player settings.

The result the ICT initiatives and other such video game-based simulation projects is something that military training centers and training directors enthusiastically endorse. They cite improved training efficiencies, cost effectiveness, and a fine-tuned alignment of training with the actual job to be performed, particularly in terms the development of cognitive skills needed in highly specialized activities utilizing highly technical equipment.

For example, Brigadier General Frank A. Panter, Jr., Commanding General of the Marine Corps Warfighting Lab, points out what are viewed as indicators of quality: number of realistic scenarios that a student will experience; replicating the stress of the real situation; engaging the affect by being perceived as "fun" and "challenging." Gen. Panter states that the "MCWL is currently experimenting with the Combat Decision Range (CDR). The CDR uses facilitated, multimedia-driven combat scenarios to build artificial experience. In a typical 45-minute training session, the Marine makes more tactical decisions than in a week in the field. CDR has been exposed to the operating forces through the infantry regimental training centers and other units. It is highly acclaimed by participants as a stressful, fun, challenging and very beneficial training tool" (Panter 2003).

He also points out that using video-game simulations as well as other types allows Marines to obtain a training experience of uniformly high quality, and to familiarize oneself with always-evolving high-tech equipment. Modeling likely scenarios and logistical challenges is also cited as an advantage of such training, as is the way that small unit leadership and fire team coordination are developed.

At this point, and important distinction should be made between video game-based simulation and the kinds of simulations employed by flight schools, air traffic control centers, and the military. One such simulation environment has gaming elements, but is large enough for individuals to move about in and develop the physical dexterity and cognitive skills they will need in the actual situation. Dubbed the MOUT, which is an acronym for Military Operations in Urban Terrain, the portable simulators fit in three shipping containers, with final dimensions of 8 ft wide by 9 ft high, and 20 ft long (Wilson, Nov 2003). They have been shipped to various field training locations, as well as to forward locations, such as in Afghanistan. The simulators have equipment that generates smells, sounds, and sights (lasers, etc.), and allow students to monitor themselves via cameras, recording equipment, motion detectors, and other sensors that allow individuals to "score" themselves.

There is no doubt that training with large-scale, multi-sensory, role-playing simulation equipment is effective. However, truly distributed training is not available for everyone if there is a reliance on simulation equipment. Thus, PC-based simulations based on video games are viewed as ideal. In many cases, the video game-based simulation is used in conjunction with larger, full-scale simulation units. For example, Full Spectrum Command was first used in conjunction with the Army's MOUT at Fort Benning, Georgia and classroom instruction (Wilson, Nov 2003). The students were asked to develop mission plans as a homework assignment, then to come to class prepared to discuss them and to input them into the game. Later, they emulated some of the conditions and contingencies in the MOUT's onsite large simulator.

However, when one moves into the area of PC-based video game-inspired multi- or single-player simulation, the issue becomes a bit muddy. Ethical issues abound, a few of which are listed below.

Training Efficacy: Does gaming-based simulation really prepare people in the way they need to be prepared?
Perhaps the first and most obvious ethical issue is that of training efficacy. Obviously, it is unethical to claim that a video game is a simulation when it is not. A distinction between "simulation" and "scenario" must be observed, and the simulation must replicate lived experience in more than simply the visual. It must mimic the kinaesthetic by having similar controls. Further, decision-making must incorporate the same factors and conditions as the world that is simulation. For example, Full Spectrum Warrior contains a playbook with such features as "Crossing a Street Under Fire." This can be intriguing to a person who has been watching television footage of action in Baghdad, for example. It can seem like a godsend for a soldier who is about to be sent into a situation where he / she will have to cross a street under fire. Now, one has to hope that the instructions and the scenarios in the simulation are realistic. Does it really happen this way? It had better do so, or lives will be lost.

The simulation claims to prepare the players for real-life situations by presenting the most likely scenarios, equipment, and avatars possible. To reiterate a point, to not do so is fraudulent and harmful to individuals who may experience physical consequences if they do not act appropriately.

* Equipment in the game must be as faithful as possible to the appearance, functionality, and uses (predictable and unpredictable) as what will likely be encountered in real life. The controls need to be as realistic as possible as well.

* The scenarios need to be as realistic as possible. They must be designed and tested by people who have actually been in the places involved.

* The avatars - the characters in the game - must be as realistic as possible, with appearance, actions, response times, and scripts as accurate as possible. The algorithms that control the "dialogue buckets" must have some randomness at the beginning, but ideally the system should "learn" the player through artificial intelligence.

Illusion of Randomness
It is vitally important that the players maintain the illusion of randomness so that the scenarios, actions, and other items expand the number of envisionable possibilities. The better simulations in the PC-game-based world are those that involve multi-player actions, and which are customizable.

Randomness, or at least a high degree of unpredictability, allows individual players to develop creative problem-solving skills and to make decisions quickly.

The problem with most games is that they feature "charms" and magical amulet-type pieces which, unfortunately, do not exist in the real world. The player who is used to being able to gain lives, power, or magical abilities by grabbing rings, flashing jewels, amulets, or other items, may experience a kind of cognitive disconnect when confronted with the real-life situation.

The other problem is that of the seductiveness of entertainment. The person who will spend hours playing an engaging video game, or who enjoys the game-inspired simulation may or may not make the connection to the real world. The pleasure and "reward" stimuli in the simulation / game environment are not likely to be the same in the real world.

Further, the real-world "rewards" - the social interaction and human bonding - employ different techniques in the real world. To take an extreme example, let's take a look at first-person shooter games. According to James Paul Gee, in the realm of the video game, "shooting" is "having a conversation" and a kind of "social interaction grammar" comes to the foreground. Needless to say, a first-person shooter game-based simulation and a live-fire practice activity employ different social interaction grammars. Shooting a person is not a good way to bond. It doesn't even send a very good message.

The ethical problems of mixing grammars are severe, yet no one seems to be addressing this. One cannot "translate" first-person shooter game to live fire and expect positive results. And yet, that is precisely what the intellectual model assumes when it proposes such a tactic for transfer of knowledge. Invincibility is not merely entertaining, it invites hubris and/or mission-endangering grandiosity.

To be entertaining and to trigger the fantasy-projection aspect of video games in order to have the player want to identify with the protagonist, the video game must feature main characters of superhuman flexibility, strength, and pain-free injury. The characters must be very "Lara Croft" in order to "sell" - but what are the ethical implications of identification with an avatar that dies painlessly, accompanied by a somber musical signature. There may be an adrenaline-charged moment or two, but they are generally pleasurable. The gaming-based PC simulation does not incorporate thirst, hunger, nausea, pain, dizziness, shock, or other conditions under which a fighter must work in the real world.

Further, when a Lara Croft action figure is employed, the player identifies with the fantasy-qualities and loses sight of the real mission at hand. That is probably unavoidable, given that only with great difficulty is the simulation addressing problem-solving issues that are connected to something. In other words, the "situated learning" that occurs is only situated within the identity of the player; there are no actual tasks and embodied action (the memory of lived experience) does not factor in at all. Needless to say, the player becomes engaged in a fantasy identity construction activity, and pays little or no attention to anything else.

Translate from 3D to 4D in the real world

Navigating real highways and roads using Mapquest is probably on of the best "situated meaning" simulation activities on the Internet. More simulation games should employ the techniques used there. Note that the virtual world represented by Mapquest is at all times connected to a tangible reality that the player must interact with. It is clear at all times that the graphical depiction of reality corresponds to a tactile reality; and, that errors in decision-making have an immediate negative consequence (wrong roads, getting lost).

Abstract notions are rooted in metaphors for embodied experience (Lakoff 1987) and comprehension is grounded in perceptual simulations that prepare agents for situated action (Barsalou 1999).

While Kress and van Leeuwen and others suggest that naturalistic color, lighting, etc. "count as real" (Kress and van Leeuwen 2003), Mapquest's simplified graphical representation and the user's easy identification of that as having a corresponding reality in the real world seems to argue the opposite point.

In the case of warrior-preparing video-game based simulations such as "Full Spectrum Command," the very naturalistic color lighting, and perspective encourage identification and a belief (a suspension of disbelief) in the verisimilitude of the representation.

The Dangers of Dehumanization

Playing a well-designed video game is a pleasurable social experience. To be honest to the experience, a video-game simulation of war should be unpleasant, uncomfortable, and horrifying, as is war itself. However, to fill players with horror, fear, nausea and pain would probably be viewed as demotivating, and it would be difficult to have popular acceptance of the technique as an effective training tool.

Further, many have argued that video game-based simulations encourage dehumanization, by encouraging the players to view other players as objects rather than sentient beings. Oddly enough, this may actually be a benefit, particularly to those in the infantry who are on the front lines and will actually be called upon to see the person they kill (as opposed to artillery or aerial-delivered bombs). Killing is often requisite upon the ability to consider the person a "target" and to disassociate oneself from the action.

Undetected Obsolescence

The training efficacy of a simulation is a function of its relevance to real-world situations. In the world of asymmetric war, particularly in urban settings, the techniques and approaches constantly change. In fact, this is the key to tactical advantage. Thus, a game must be modifiable and adaptable in order to accommodate new equipment, tactics, and players.

Although the designers may shrink at incorporating ethically unsavory tactics and enemy actions, it is necessary to do so, particularly when there are attempts at disguise, entrapment, booby traps, and improvised explosive devises. If the designer does not accommodate the ever-evolving world, the simulation is worse that useless; it is dangerous. SMEs are Not God, but they do create worlds, and if they get it wrong, you can die.

The quality of a scenario is only as good as its verisimilitude, and its ability to represent the cast, the conditions, and the equipment the players are likely to encounter in the real world. Authenticity and relevancy are key factors. The subject matter expert team needs to "ground truth" often. In designing the simulations, the project managers must ask questions: Who decides what the scenarios are? What are the decision-making points? How many options are presented? Does the system employ adaptive learning? When and how are modifications allowed?

These questions are vital to preserving the integrity of the experience and avoiding the temptation to focus on the entertainment value, rather than focusing on the most important decision-making, leadership, and physical skills to be developed.

Mentor No More: Forgetting how mentorship happens on the ground (trust, etc.)
Although video game-based simulations claim to develop team-building and leadership skills, this is not always the case. The temptation is always to "cowboy" the game and to indulge identity fantasies rather than to develop the skills that will lead to missions accomplished, lives saved. Mentoring is vital. Individuals need to be able to see a seasoned leader and to be able to emulate him or her. Further, trust is developed in the team, and through shared danger, rescue, and extended bonding. This is difficult, if not impossible, to achieve in the video-game simulation world.

Perhaps the most persistent ethical issue remains in the area of training efficacy. Other issues are briefly alluded to, and include access, integrity of the experience, dehumanization, and instructional design. This article assumes video game-based simulation for military applications for individuals who are already either active duty or reserve forces. It does not touch on the ethics of using an entertaining experience or satisfying social interaction to recruit individuals to enlist for an experience that is utterly devoid of the glamour, non-stop excitement, pain-free, and freedom from loneliness, anxiety, and discomfort of the game. Further, in a game, no one really dies.

first published in

Aldrich, C. (2004). Simulations and the future of learning. San Francisco: Pfeiffer / John W. Wiley.
Barsalou, L. W. (1999). Language comprehension: Archival memory or preparation for situation action. Discourse Processes 28: 61-80.
Bruer, J. T. (1993). Schools for thought: A science of learning in the classroom. Cambridge, Mass.: MIT Press.
Churchland, P. M. (1989). A neurocomputational perspective: The nature of mind and the structure of science. Cambridge, Mass.: MIT Press.
Erikson, E. (1968). Identity, youth and crisis. New York: Norton.
Freire, P. (1995). The pedagogy of the oppressed. New York: Continuum.
Harris, P. (2003). SimsSimsSims - Simulation: The game is on. Training and Development October 2003: 46-51.
Kress, G., and van Leeuewn, T. (1996). Reading images: The grammar of visual design. London: Routledge.
McDonald, Andy. (2004). Moving to the forefront. Military Training Technology. 9 (3): July 19, 2004. Retrieved August 16, 2004.
McPherson, M. and Baptista Nunes, M. (2004). The failure of a virtual social space (VSS) designed to create a learning community: lessons learned. British Journal of Educational Technology. 35 (3): 305-321.
Nicol, D. J. and Boyle, J. T. (2003). Peer instruction versus class-wide discussion in large classes: A comparison of two interaction methods in the wired classroom. Studies in Higher Education 28 (4): 457 - 473.
Panter, F.A. (2003). Interview. Military Training Technology. 8 (1): Jan 1, 2003. Retrieved August 16, 2004.
Pelligrino, J. W., Chudowsky, N., & Glaser, R. (2001). Knowing what students know: The science and design of educational assessment. Washington, D. C.: National Academy Press.
Pinker, S. (1999). How the mind works. New York: Norton.
Poole, S. (2000). Trigger happy: Videogames and the entertainment revolution. New York: Arcade.
Rip, M. R., and Hasik, J. M. (2002). The precision revolution: GPS and the future of aerial warfare. Annapolis, MD: Naval Institute Press.
Shinn, S. (2003). The leader within: Be, know, do. BizEd. Nov-Dec 2003: 30-35.
Wolf, M. and Perron, B. (Eds.). (2003). The video game theory reader. Bristol, PA: Taylor & Francis.
Taylor, C. (1994). The politics of recognition. In C. Taylor, K. A. Appiah, S. C. Rockefeller, M. Waltzer, and S. Wolf (1994), Multiculturalism: Examining the politics of recognition. Ed. By A. Gutman. Princeton, N.J.: Princeton University Press, pp. 25-73.
Wilson, J. R. (2003) Fixed urban training. Military Training Technology. 8: (4), November 19, 2003: Retrieved August 16, 2004.
Yan, Z.; Hao, H.; Hobbs, L. J.; and Wen, N. (2003). The psychology of e-learning: A field of study. Educational Computing Research. 29 (3): 285-296.

Tuesday, May 03, 2005

Network-centric Warfare and Implications for Distributed Education

Play the podcast (downloadable mp3 file)

The concept of “network centric warfare” has totally pervaded military training and policy publications, and, although there are operational and philosophical critiques of the concept, the fact remains that it continues to influence the way that the military thinks of itself and its activities.

It represents a profound epistemological change – not only in terms of how strategy is conceptualized, but the entire notion of what constitutes meaning and knowledge.

In a certain way, the notion of network-centric warfare privileges knowledge and knowledge management, suggesting that victory has to do with successful management and dissemination of data.

We’ve seen spectacular successes. Desert Storm is often cited as an excellent example of coordinated information and joint force efforts resulting in overwhelming force, and a quick, harsh, decisive victory.

The rise of remote sensing (which evolved quantum leaps from where it was in the late 1970s with LANDSAT satellite imagery, and the ability to generate false color composites and map vegetation, landmasses, and human activity trends in new ways) coupled with highly mobile, handheld GPS units not only transformed land navigation, it also led to breathtaking detail (if not precision) of surface features and heat signatures.

Tracking people, natural phenomena, static and mobile features (buildings, equipment, etc.) while coordinating logistics, materials and “fire” could lead the impression that the war of the future (or of today) would be something from a science fiction film. Certainly the footage that the average civilian like myself would see reinforced that notion. “Surgical strikes” of insurgent hideouts in Kosovo, with scary fire coming from gunships, and remotely directed attacks on distant targets, ranging from Sarajevo to wherever, made warfare seem like a techies’ game.

According to John Gilligan, the U.S. Air Force’s Chief Information Officer, the accuracy and precision of air strikes have become even more astonishing. In a Air Force press release, Gilligan was cited as saying the following: In Desert Shield/Desert Storm, 40 percent of the munitions that we delivered were precision munitions. Today it's over 90 percent. Our accuracy rate is phenomenal. And when you talk to those who are doing the planning of weapons delivery, we're able to gain insight in terms of what's the physical composition of the target. What are the wind speeds? What's the angle, et cetera? They can predict with extraordinary accuracy what is going to be the collateral damage. Now that's not just what's the impact point, it's what's the collateral damage. All of that is information-based. All of that information is available through the network” (US Fed News, 28 March 2005).

And then comes Fallujah. And then Najaf, Sadr City, and all the other places where the other ubiquitous term, “asymmetric warfare,” comes into play.

Needless to say, in this case, network centric warfare, according to the people who are typically on the ground – Marines, Infantry, Army National Guard, Navy Seabees, even Air Force personnel charged with driving convoys – has come under fire.

In the April 2005 issue of the Proceedings of the U.S. Naval Institute, Captain Tim Feist of the U. S. Marine Corps takes issue with the inflated expectations and hints of invincibility that the “network centric warfare” model suggests. In “Transformation Has Limits,” he points out the failures of networks and information, and argues that the raw material and data are often faulty, as are the networks themselves, and the assumptions used to build the models. There is also a suggestion that an over-reliance on it can lead to strategic disasters. This not only has to do with what we are commonly led to think of as “insurgents” – snipers, suicide bombers, child soldiers, and the trickery of booby-traps and human shields. It also has to do with activities in the "informal economy."

What is often not mentioned are the virtual economic blockades that arise due to illicit profiteering – mercenaries, arms-trafficking, black-marketing, etc. How many ambushed convoys and supply lines are actually types of piracy? How many kidnappings of truck drivers were not expressions of holy war, but of a turf battle over which contractor would receive the lion’s share of contracts? How many of the contractor killings were attempts to drive out the lower-paid workers who came in from Nepal, Kenya, Pakistan, and Bangladesh? The informal economy is not just about a business opportunity. It has strategic implications.

Corruption – whether originating due to a local “kleptocrat” leader, or de facto “pirates” who take advantage of anarchy and chaos to peddle guns, water, and false documents, along with trafficking humans to provide “safe” passage to somewhere else – can defeat a network centric war plan, as well as “boots on the ground.” Asymmetric warfare is not just about small, unexpected guns against the big ones, it’s also about duplicity, double-dealing, and greed.

The astute observer can see a very clear parallel between this situation and that of distributed education. While the disruptive technology has already happened – the reliance on networks and network information to make decisions and coordinate action (of all types) – the fact remains that the technology can be rendered useless by ground activities when and if the goal is to control activities in urban centers on the ground.

Network centric warfare can also be rendered useless by incompetence or hubris in the data gathering process. The old “garbage in – garbage out” tenet still holds.

What does this have to do with military training and education? In a network-centric approach, the following elements hold true:

First, Training / educational content is distributed over networks: “Coordinated autonomy” is possible through a central network. Entire learning network systems can be in place, with collaboration, contribution, and decentralized access and contribution to a network-housed structure, where the content is housed.

Second, Simulation and video-game based training will take place far from the actual place of battle: Learning will take place far from where the instructor is, and the participants may be separated in time and in space. Yet, they will be united through a massively multiplayer system that allows interaction and coordination. Knowledge gained will be useful, or useless – depending on its relevancy and “freshness.” In such cases, being “stale” but not knowing it will result in a very perniciously deleterious product. In a word, it can kill.

Third, Data used to develop training and “situated” learning will come from far away: The inputs used to create learning modules, learning objects, and the integrated systems will be contributed to the network from widely divergent sources. Coordination is key. Security must constantly be examined. Security will probably always be one step behind the hackers. Even now, researchers are exploring how to probe enemy communication webs in wartime. The same applies to e-learning.

Fourth, Training created by people who have second-hand or third-hand knowledge can be flawed: Extreme situatedness is a must. Data to build learning programs that is collected on site and uplinked through handhelds in situations as directly related to the real thing as possible is critical. This used to be called “ground truthing.” Today is falls into the category of collaborative, multi-nodal course / training development. Training must be reviewed and updated continuously, as circumstances and conditions (as well as tactics) continuously change.

Fifth, Corruption can destroy the integrity of a learning system. Corruption takes many forms in terms of training and education. Academic dishonesty, theft of intellectual property, and “short-cutting” to simply pander to perceived business opportunities rather than providing a high-quality product are all issues to be addressed. Duplicity, disinformation, political backstabbing, and black-marketeering may not play the same role in an education system as they do in a war-torn country. Nevertheless, one has to say that the sterile view from above (in the control rooms of the networks) has little or no connection with the ugly streets of war where wild dogs fight over human carcasses. Does the analogy hold in higher education or in corporate training? You, the reader, can decide.

To backtrack a bit and to provide grounding for readers who may not be aware of the nuances of network centric warfare, the following is offered.

At a recent briefing and update, David Alberts, director of research and strategic planning in the Office of the Assistant Secretary of Defense for Networks and Information Integration, provided a broad, theoretical overview of network centric warfare. The architect of what was often referred to as information technology warfare, Alberts has influenced policymakers in the Pentagon for a number of years. Here are his words, from the press release:

Network centric warfare has four simple tenets. It all starts with the concept of a robustly networked force. That robustly networked force leads to increased information sharing.

Second, increased information sharing enhances not only the quality of information but encourages collaboration and increases what we call shared awareness.

Third, increased collaboration and shared awareness enables self-synchronization.

Fourth, all of that together dramatically improves mission effectiveness and ability.

These four tenets serve to define a value chain that links the full spectrum of material and non-material investments to operational effectiveness and agility.

Although these tenets are very general and have broad applicability, there were many people who thought that it was only about the conduct of high intensity warfare. So to better convey the notion that these principles are general, we started using the term network centric operations sometimes instead of and sometimes with network centric warfare.
(US Fed News, 28 March 2005)

The controversy is only beginning. Certainly it will continue to gain momentum as casualties mount, and call into the question the efficacy of a plan that assumes that remotely directed fire and remotely gathered information can truly result in total control of the ground environment.

The argument in education is only beginning, too. Can we assume that remotely delivered educational experiences that take place completely in a digital environment can truly result in total control of the learning outcome?

We’ll see.

Blog Archive