Civilization III: A Tool for Remediating Understandings in World History

Civilization III: A Tool for Remediating Understandings in World History

The Civilization series was invented and designed by Sid Meier, who is frequently credited as a founding father of computer strategy gaming. Sid Meier's Civilization (the game’s full name) was designed with Bruce Shelley and first released in 1991 by MicroProse Software. Civilization builds on the appeal of historical empire building in his earlier Railroad Tycoon, but takes on a much broader time scale, attempting to simulate the growth of a civilization over its entire 6000-year history. Civilization and the sequel, Civilization II (1996) are both widely considered as two of the most compelling and significant games ever produced. The third installment in the Civilization series, Civilization III, was released in October of 2001 and has also been met with critical praise and strong sales figures.

Developed by Firaxis and published by Infogrames, Civilization III is a turn-based strategy game where the player commands a civilization over the course of its 6000 year history, managing its natural resources, finances and trade, scientific research, cultural orientation, political policies and military. Significantly, the Civilization series gives the player control over a band of people for its entire civilized history, placing her in the role of a supreme leader for which there is no historical analog. As such, Civilization, on the one hand, enables the player to ask and pursue tantalizing historical trajectories across broad time-scales, while, on the other, risks producing and fostering misconceptions about politics or history. The original Civilization manual introduces the game in the following terms:

Civilization casts you in the role of the ruler of an entire civilization through many generations, from the founding of the world's first cities 6,000 years in the past to the imminent colonization of space. It combines the forces that shaped history and the evolution of technology in a competitive environment … If you prove an able ruler, your civilization grows larger and even more interesting to manage. Inevitable contact with neighbors opens new doors of opportunity: treaties, embassies, sabotage, trade and war” (Shelley, 1991, p. 7, quoted in Friedman, 1999, p. X).
Although some educators have argued that Civilization might be an interesting tool for studying social studies, the game was originally designed as an entertainment product and has yet to be studied seriously as an educational resource. Friedman’s (1999) study of Civilization is the most serious academic consideration of the game to date. In it, he argues that much of the allure of Civilization comes from the player mastering the rules that define the game. Insomuch as the rules embodied in the simulation reflect accurate models of social phenomena, it would seem that even casual players can develop understandings of the game world that have analogs to phenomena in the real world. Unfortunately, however, there have been no studies examining this, let alone the suitability of such a game for use in educational settings.

Game Play

The game play behind the Civilization series is notoriously complex; Civilization III ships with a 231-page manual. A single game can take over twenty hours, and an expert player can easily log thousands of hours of game play. The game play is robust enough to support multiple strategies, and players approach the game quite differently according to their own idiosyncratic values and tastes. For example, some players prefer to dominate the world through a powerful military, others prefer to build strong economies, technological superiority, or a strong intelligence, while others simply build dominant cultures in order to colonize others. There are numerous different strategies a player might adopt to win the game, and most players enjoy experimenting with different approaches. Civilization III is designed for such divergence, including at least six different win conditions: (a) conquering all rivals, (b) constructing a spaceship and successfully colonizing Alpha Centauri, (c) dominating the world through controlling a vast majority of the planet’s land, population, and resources, (d) controlling the United Nations and being voted the victor, (e) culturally dominating the world through cultural hegemony, and finally (f) achieving the highest score if no other win conditions are met. The player loses the game if her civilization is wiped out through military force or another civilization attains a win condition.

Most of the game play involves planning the geographic growth and evolution of the civilization; securing, developing and exploiting natural resources; building and maintaining military forces; waging war; conducting trade and political negotiations; budgeting; directing scientific efforts; choosing an appropriate type of government; maintaining civic happiness; and, perhaps most importantly, managing each cities’ production. For each turn, the player must decide whether to focus production on building military forces, creating workers or settlers that can help develop the civilization, implementing civic improvements (e.g., libraries, factories, banks, or cathedrals), or even developing one of the Great Wonders of the World, (e.g., The Pyramids or Darwin’s Voyage). Each structure so produced has a resulting effect on the “performance” of the civilization, modifying the civilization’s food, economic, scientific, industrial, or cultural production. In addition, players also manage armies that can be used to explore uncharted territories, to defend against attacking civilizations, or to attack other civilizations. Finally, players can use settlers, workers, and engineers to irrigate fields, build roads, clear forests, build railroads, or construct fortresses. Thus, in Civilization, by controlling the food production, natural resource gathering and production, taxation, cultural production, economic policies, military policies, domestic priorities, and foreign policies of an entire civilization from the year 4000 BC to the present, player learn about the fundamental yet complex conditions under which civilizations endure.

Starting the Game

The player begins the game in an unknown corner of the globe. For the purposes of this study, all players use maps based on the Miller projection of Earth, although players can choose to use maps of fictitious planets in regular game play. The player must first choose where to build a city (See Figure 3.1), forcing the skilled player to consider several variables contributing to the strategic geographical location of a city, including relative access to food sources, potential for exploiting natural resources, access to n
avigable waterways, and strategic land barriers. The player also begins the game with a basic worker unit who can irrigate fields, build roads, or explore new territories.

Figure 3.1: The beginning of the game.
Once the player has established a city, she begins to manage her civilization’s natural resources, which are abstracted into four primary game economies: food, natural resource production (shields), trade, and culture. Each one is an abstraction of several "real-world" production processes that vary across geography and time: Food production includes fishing, subsistence farming, hunting, herding, and the domestication of animals for food. Production (shields) includes both the production and refinement of goods. Trade refers to the generation of all goods or services that can be traded for profit which the player may then allocate toward gold (hard currency), knowledge (scientific discovery), or luxuries (See Figure 3.2). Finally, and perhaps most controversially, Civilization III (unlike earlier Civilization games) attempts to model "cultural growth and production by assigning value to cultural institutions, such as libraries, schools, churches, or works of art. Much of the game play involves balancing cities’ production so that each of these functions (population growth, production capacity, financial solvency, scientific discovery, citizen happiness, and cultural growth) are met, enabling the civilization to be competitive with other civilizations. In Civilization III, more than its predecessors, no one civilization can dominate the game along all axes simultaneously. Rather, civilizations depend on one another for strategic and luxury resources, for trading scientific discoveries, and occasionally even for financial support. Negotiations and trade are critical components of game play (See Figure 3.3)

Figure 3.3.Negotiation Screen

Once the player has built her first city, discovered one or two technological advancements, developed one or two military units, and possibly built a city improvement such as a temple, she might build a second. As the game continues, other complex variables come into play: Players must control political corruption, interact with other civilizations (played by the computer), secure strategic resources (such as aluminum, uranium, or oil), and even monitor global warming. These many factors create conditions under which no two games are exactly alike.

Modifying Civilization III for Classroom Use

Working from a world map developed by the game community at Apolyton.net, I created a unique game scenario for this unit. My goal was to design a scenario that would be an idea model for examining world history. As such it had to help students gain insights about social phenomena while also being usable within a learning context (Colella, Klopfer, & Resnick, 2001). My goal was not necessarily to construct a model accurate enough to predict historical events but rather one that would elucidate important social studies concepts. The main changes I made to the game design as shipped, therefore, included (a) substituting historically non-plausible civilizations for plausible ones (e.g. taking out the Americans and adding the Incans), (b) adding groups that the students would want to play (the Bantu¹⁰), (c) adjusting geographic bonus resources to more accurately reflect the technical and cultural advantages of particular civilizations and regions (e.g. giving the fertile crescent added food bonuses), (d) disadvantaging civilizations that were less developed as of 4000 BC (e.g. removing technologies from the Aborigines), (e) reducing randomness in combat, and (f) making the game more playable for novices (enhancing technological discovery rates, and giving additional starting bonuses). I used Diamond’s Guns, Germs, and Steel for much of this intellectual work – for example, determining which peoples were poised to potentially invent agriculture at 4000 BC. In order to balance the game for novices, I read through posts about developing game modifications at Civilization III at apolyton.net and spent approximately 40 hours, play testing various modifications, including trial runs on the school’s Pentium II 300 machines (64 MB RAM). Finally, working with the MEDIA School technology coordinator Norm Eng, I installed Civilization III and its accompanying patches on the school machines, loaded my modifications, and debugged the installation, which took approximately ten hours because of the school’s network controls.

Civilization III as an Historical Simulation Tool

Although Civilization was designed as an entertainment game, the historical, geographic, and political simulation elements of Civilization III make it an intriguing educational resource. Success in the game demands that players "master" geography through such strategies as focusing food production in agriculturally advantageous areas, using physical boundaries as natural borders, and securing natural resources. Players confront political dilemmas such as whether to pursue isolationist politics, enter complex alliances for protection, or gain precious resources through military force (which may also mean waging war to protect an ally). Finally, players can view how their history grows and evolves over time, exploiting the visualization tools included in the game (i.e., maps, charts, and graphs) to view how their civilization grows culturally, geographically, scientifically, and politically. In this section, I explore the pedagogical possibilities of Civilization III as a tool for studying world history. How the game is actually used by students in classroom settings, given the paucity of research in this area, is difficult to predict (and, indeed, the point of this study); therefore, the following discussion outlines the potentials of using Civilization III for learning rather than offering some definitive argument for its usefulness.

There is no “one” way to win Civilization, which makes game play an open-ended task that leads to divergent outcomes. One player may attempt to win the game through military domination, building a strong economy and strong military; another may be very peaceful and attempt to win primarily through political ends. From an educational standpoint, this open-endedness is interesting because it may dissuade players from competing against one another. After all, different players with different tastes will set different constraints for themselves and therefore will (purposefully or not) use the game to explore different ideas (e.g., trying to win the game without waging war). That people play Civilization III differently creates opportunities for educators to use contrasting gaming experiences as the basis for critical reflections and debriefing, enabling players to analyze not only what strategies work in the game and why but also for what ends.

The critical reader may question the notion of using a simulation to study world history, when very few historians engage in this kind of enterprise.¹¹ In the physical sciences, models and simulations have long been used as a part of inquiry. The idea of observing some phenomenon, building a mathematical model to represent important aspects of it, and then testing it against reality is centuries old in the physical sciences; physicists and astronomers have been building mathematic representations of systems at least as long ago as Copernicus and Newton.¹² More recently, Wolfram (2002) has argued that the digital affordances of computers are revolutionizing science so that scientific inquiry can be thought of as a process of observing phenomena, defining rules, and building and testing models rather than one of defining and testing hypotheses through controlled experiments. Charles Sanders Peirce (1878/1986) refers to this recursive process of observation, model generation, and evaluation of the predictive capacity of the model as abductive inquiry. Such abductive inquiry, uncommon in history, is more commonly employed in political science, economics, and increasingly, anthropology (Edmonds & Hales, n.d.; Wolfram, 2002).

As opposed to textbooks, which contain one state-sanctioned narrative, Civilization III is an open-ended sandbox, a possibility space wherein players can explore and actualize any one of several different narrative outcomes. To be sure, Civilization III is a rule-bounded space that represents physical, social and cultural systems in very specific ways. For example, Flood plains produce 150% more food than grasslands; grasslands produce 200% more food than hills. Democracies have less corruption and produce more efficient workers than monarchies. Listing the number of rules and relationships encoded into Civilization III is beyond the scope of this dissertation, let alone this chapter.

Emergent Properties of the Game System

Using Civilization III as a simulation tool changes the method of studying history from one of memorizing facts and mastering sanctioned narratives to one of defining the terms and rules of the system (either similar to or contradictory with those thought to describe events in real life) and then exploring its emergent properties. One of the more intriguing aspects of Civilization III as a learning tool is that no two games are alike. Early decisions in the game might have important ramifications hundreds of game years later. The specific trajectory of events comprising any one game may have no analog in real life history; however, the rules binding the game – the core variables included in the model, the interrelationships among political, economic, and geography systems – allow players to gain insights into each of these areas and therefore the unfolding of history itself. On the other hand, sometimes players’ games do simulate actual historical events (and as reported in Case VI, such as the barbarian ransacking of China (as well as Europe) in the first millennium AD, the building of the silk road in China, French colonization of Canada, or the relatively late discovery of Aborginal Australia. Regardless, the interplay of underlying causal variables (rules) within the game reveal important aspects of how history “gets made.” In the next section, I explore Civilization III as an historical possibility space, highlighting several properties of the game that merit deeper consideration. I follow with an exploration of the implications of using such a tool in social studies education.

Problems with Civilization III as a Learning Tool in History (and a Few Solutions)

Because Civilization III is intended to be an entertainment product, it is designed to support good game play rather than good historical simulation. For example, if a player builds the Pyramids, she receives a free granary in every city. There simply is no real historical analogy. Native populations (called “Barbarians” in the game) are dealt with questionably as well, presented overwhelmingly as warring tribes with no real “culture” to speak of. At the same time, however, other cultures that one might consider native (e.g. Aztecs, Iriquois, or Zulus) are treated as “real” cultures. Notably, Civilization III also attempts to operationalize cultural influence as a game mechanic, giving players’ points for creating institutions that add to a culture’s influence. As Barkin (2002) points out, this notion of culture is theoretically problematic as it combines French definitions of culture (as in “to become cultured”) with traditionally German definitions of culture (as collective ways of being).

Characterizing and operationalizing each civilization is also potentially problematic. Distinguishing features of civilizations are distilled down into relatively simple game mechanics. For example, Americans receive industrial and expansion bonuses, including special fighter jets – and each civilization begins with roughly the same resources at roughly the same time (6000 BC). Obviously, “America” was utterly meaningless as a concept until 5600 years later, in at least the late 16th century. Less obviously, it was also well into the 17-18th centuries before a notion of a “France” or a “Germany” developed. In response to these conditions, I modified the starting civilizations to only include the Incas, Iroquois, Aztecs, Egyptians, Bantu, Persians, Carthaginians, Oceanic peoples, the Chinese, Russians, Indians, Japanese, Babylonians, Celts, Saxons, and Goths. I intentionally included civilizations historically thought of as Barbarians (such as the Celts and the Goths) in order to help counter misconceptions about such tribes as well as to raise the question, “What does it mean to be ‘civilized’ and what does it mean to be ‘barbaric’?”

Jared Diamond’s Guns, Germs, and Steel (1999) persuasively argues for the importance of food packages¹³, particularly agricultural, in giving rise to specializations in the division of labor, the creation, collection, and trade of goods, population densities, and, ultimately, the growth of civilization. Diamond analyzes how agriculture emerged through the careful cultivation of crops that are suitable for domestication, the nutritional value of various crop packages, the native animals available for domestication, differences in local climate’s capacity for supporting agriculture, and the comparative advantage or disadvantage of other hunting options. To aggregate these many factors into one variable, the “food production of land” as Civilization III does, glosses over many important local conditions and technological discoveries that gave rise to agriculture. However, by beginning the game at the onset of agriculture and assuming that all civilizations began at 4000 BC, the game avoids the problem rather than addressing it. In Civilization III, one simply assumes that each civilization created a food package that meets its nutritional needs and therefore only accounts for advantages or disadvantages from civilization to civilization by assigning those in especially fertile areas extra food production bonuses. While nothing is inherently inaccurate in such a depiction of agriculture, the black-boxing of so many variables into the simple notion of “agriculture” or “geographical production bonuses” may be misleading.

Moreover, while most texts agree that agriculture emerged at between 10,000 BC and 13,000 BC, Civilization III begins at 4000 BC and presumes adequate agricultural production within each civilization by this time. To compensate for the “headstart” of civilizations such as the Egyptians or the Phoenicians, who formed into civilizations relatively early, I gave those civilizations bonus workers and settlers. In addition, I awarded river valleys where such agricultural technologies originated extra crop bonuses. I also adapted the map so that horses and cattle can only be found in historically appropriate areas cotton is also represented as a natural resource. Many critics have noted that sheep, an historically important livestock, are not included in the game. Unfortunately, I have yet to make any modifications in the game to rectify this inaccuracy at this time.

One of the more interesting intellectual problems with Civilization III is that, because the game includes an historically accurate model of how civilization advances occurred, every game is tied to the same set of technological advancements. Restated, technological advancements are linear and predetermined. Although no two games are alike, particular discoveries – such as the wheel, monarchy, or horseback riding – are present in every game, always in the predetermined sequence, and always, therefore, shaping each civilization’s simulated evolution over time. On the one hand, some (for example, Diamond) argue that the affordances of some inventions (particularly the wheel) are so apparent and so reproduced across multiple civilizations in time that they ought to be treated as at least highly probably if not inevitable. Other improvements, such as communism or atomic theory, may seem less so. Others have argued (e.g. Barkin, 2002) that there may have been other hypothetical paths not taken: For example, imagine that a Native American civilization remained the dominant culture occupying North America. Under such circumstances, there could easily be different technological forms developed and distributed globally. Such questions are not problematic; rather, they are central questions within the domain and suggest the very reason that Civilization III is an intriguing historical resource. Playing Civilization III tends to recruit deep, intellectually valuable ways of thinking about world history and to produce the kind of social structures that nurture and sustain such discourse.

Finally, there are several minor simulation biases that can be raised about the game. For example, some game fans believe that the historical importance of tin is overlooked, or that militiamen are not given adequate fire-power. Other long-standing debates hinge on whether or not it is theoretical possibility for primitive battle units to defeat technologically advanced ones. For example, it is inevitable that at some point in a game an ancient war unit, such as a group of axe-wielding warriors, defeats a modern military unit, such as a battleship. Critics chastise the game for such moments, considering them inherent limitations to the historical simulation; Meier and others, on the other hand, argue that such improbable events can and do, in fact, occur (as in Afghanistan) through superior execution of strategy, equipment malfunction, or sometimes, just luck. I argue that such simulation errors present fruitful learning opportunities for students. A primary goal of using a game like Civilization III is to engage students in critique and deconstruction of the game. Moments of failure or surprise can become “teachable moments” that encourage students to critically reflect on how history is written, recorded, and modeled. Encouraging students to critically evaluate the simulation’s design and assumptions in this way is, for example, one of the primary goals of the instructional unit used in this study.

Building a World History Curriculum for Civilization III ***

A driving question behind this dissertation is what role might a game such as Civilization III play in a world history classroom and what instructional strategies are useful in supporting learning through game play with a game such as Civilization III. Although no real research exists on how games might be appropriated within school contexts, findings from other similar approaches (e.g. VR modeling and simulation, Barnett, Barab, & Hay, 2001) suggests that how the software is used, specifically how it supports and affords students’ inquiry and is simultaneously appropriated by local cultures needs to be studied (Squire et al., 2003). Consistent with this theoretical approach, I examine game play (and education) as cultural historical practice, and I am reluctant to suggest a “instructional design theory” in the abstract before examining how it is appropriated in several contexts. Thus, a central commitment to this research is that it is far more useful to build such an emerging pedagogy based on empirical studies of the software in actual use than to design instructional programs and theories a priori.

Having said that, that viewing learning as a socio-cultural enterprise in the tradition of Vygotsky and through the lens of activity theory (and informed by existing educational research) suggests certain instructional approaches over others. Although I had no set instructional design in mind, and in fact consistent with collaborative models of curriculum development strove to design materials, activities, and unit plans in collaboration with teachers. How these attempts at collaboration unfolded is detailed in each case; in the following section, I describe my initial best guess at what might make an effective game-based learning unit (See Table 3.2). In the conclusions, I will compare the findings from the case studies against this framework.

Days	Game Play	Activities
1-3	Set up the game, do tutorials	Read and discuss background information
4-8	Play on realistic maps	Generate questions
9-12	Purposeful game play	Research, create maps, timelines
13-16	Recursive play	Examine outcome of simulations, compare to history, create presentations
17-18	Finish games	Discuss presentations, aggregate findings across games

Table 3.2 Suggested Curricular Outline for Civilization III
Days 1-3: Setting up the game as a problem space, learning the game tutorial. In the first week of activities, students’ game play will focus on learning to play the game. They will start by doing the tutorial (with simpler maps and scenarios), then, by the end of this period, move on to custom-made maps (which are more complex). This game play would be augmented through readings on ancient civilizations which provides additional background information, seeds students with questions for further research, and foregrounds issues they will wrestle with (See Appendix F). These readings could be used as homework, introductory exercises, back-up lessons in case computers fail, or as activities during downtime. Given the engaging nature of Civilization III, it is not unreasonable to expect a few students to even seek out information on their own in encyclopedias or through the web in order to augment their game play (i.e. learn where civilizations grow the quickest). The focus of these first few days will really be learning to play the game, and the teacher will be largely consumed with structuring activities (i.e. the tutorial) to help students learn the game and then by answering questions.

Days 4-8. Game Play and Question Generation. By day four, students should be confident with the basic controls and will be ready to use Civilization III as a tool for inquiry into world history. Students should be familiar enough with game concepts to begin using a full, realistic map and game scenarios (which are more complex than the tutorial). I expect that Civilization III will still be opaque to some, and the game’s complexity overwhelming, so these first few days will also include plenty of time for students to explore the game interface and concepts, perhaps using the Civilopedia.

I expect that students will have many questions about the game and history. I hope to encourage and capitalize on this curiosity by encouraging students to ask questions which can become the seeds for their research projects. I will lead structured question-asking activities which should draw out students’ questions, facilitate knowledge sharing, and scaffold their inquiry process. Specifically, I will distribute note pads, encouraging students to write down questions about the game and about history. Using butcher paper, we will compare and organize and students’ questions. My hope is that this question-generation process will encourage a culture of inquiry and pique their curiosity in specific historic events or concepts. As a class, we will discuss these questions, suggesting how they might be studied and which ones captivate students’ interest. If authentic question-asking is at the heart of any meaningful inquiry-based learning project (Barnett, et al. 2000), then a crucial function of this component of the unit is to encourage students to ask questions that are generative, that lead to deep questions central to the study of world history that are also meaningful to them. Engendering such a culture of inquiry and question-asking is notoriously difficult; however, my hope is that Civilization III might serve as a vehicle for supporting question-asking.

Days 9-12. Purposeful Game Play and Research. By Days 9-12, student should be comfortable with the game controls and concepts, begun to create and identify goals in the game (and strategies for achieving those goals) and beginning to make connections their game play and existing knowledge of history. For example, a student playing as the Iroquois might note that he or she is more isolated than those students playing in the Middle East. Some students may have reached as far as 1000 AD, and having started over the game a few times (which is common practice), perhaps is developing a broad sense of historical timescales, particularly in regards to Great Wonders or technological improvements. Between Days 9-12, I imagine students creating maps and timelines of their games and comparing them to world history charts. For example, a player playing as the Iroquois might be surprised that his civilization has grown quite large, and decides to create a map of his civilization, and then compare it to historical maps and figures. Or, a student interested in technology might make a timeline of technological discoveries, comparing the years that technologies were discovered in his game with the actual dates in history. I envision students using the Internet for much of this research, but particularly Hyper History Online (http://www.hyperhistory.org), a website with over 2000 files (including maps, charts, and timelines) on world history. These tasks should flow organically from students’ questions about the games (i.e. a question about discovering technologies more quickly might lead to questions about which civilizations discovered which technologies). I envision some class time being devoted to working on these projects, although they seem best fit for homework assignments.

Days 13-16. Recursive Game Play, Simulated History, and Creating Presentations. By the fourth and fifth weeks, students should have a mastery over the basic game concepts and already begun using the game as a springboard for studying history through generating questions and comparing their game play to history. Now, I imagine students beginning to play recursively, that is failing in the game, generating hypotheses about the game system (and possibly history) and then restarting their games under new conditions to test these theories. I envision my role as that of helping students articulate how their theories about the game that they are exploring (e.g. it is easier to play in the Middle East) may reflect underlying theories of world history (i.e. playing the game in order to achieve technological victory may be easier in locations geographically conducive to trade). A challenge during this phase will be to help students play the game on two levels: to at one time become involved in their game play and at the same time thinking about history more broadly.

One way that I hope to structure students’ activities so that they engage in historical thinking is by requiring them to make a 7 minute presentation at the end of the course addressing one of the questions that they identified earlier in the unit. Sample projects might be the role of the Nile in Egyptian history, the historical importance of the silk road in China, or disease and population growth in the Americas. My hope is that students would not just look up information in books or charts, but that they would use their games as simulations to explore these ideas, recursively asking questions to the game system, examining historical data, and then eventually forming historical arguments. This kind of historical thinking, historical thinking in terms of arguments bolstered by simulations and multi-media owes some intellectual debt to Wineburg (2000) as well as “hypothetical history” as an emerging mode of inquiry (c.f. Cowley, 1999), but is also quite unique and hitherto unexplored in relation to games and as a pedagogical model.

Days 17-18: Discussion within Communities of Inquiry. As students create their presentations, I hope to create a community of inquiry whereby students ask questions, share information and present work to their peers. Time and space allowing, I would like to gather students who played as the game civilization (i.e. the Iroquois or Egypt) and have them create an artifact (posterboard or presentation) about their civilization. My goal would be to have students reconcile their experience of their civilization with other students’, creating an artifact that spans each of their games. Students will analyze one another’s games, exchange theories of history, and negotiate meanings. Consistent with a pragmatic approach, the goal of these activities will be to have students articulate understandings, negotiate meanings, and expose their ideas to others’ within a community of inquiry.

Summary. This section suggests one possible pedagogical model for using Civilization III in world history classrooms. Central to my approach (and my philosophy of education in general) is that the key to producing meaningful learning is to engender a community of inquiry whereby students ask personally meaningful questions, guided by teachers and other students. As described in Chapter II, I ascribe to a socio-cultural view of learning, taking learning as social practice and learning can very much be thought of as a process of enculturation. The pedagogical model suggested here is rooted in this socio-cultural approach to learning and pulls from several other complimentary pedagogical approaches, including modeling-based constructionism (e.g. Barnett, Barab, & Hay, 2001), knowledge building communities (Scardamalia & Bereiter, 1994), goal-based scenarios (1994), and project-based learning (Barron et al, 1998). Central to the project is the notion that bringing commercial games into classrooms may create unanticipated consequences and result in the need for very particular instructional approaches, although a good deal can be learned by building on existing understandings from the learning sciences (Squire, 2002). The next section explains my approach to collecting data in order to investigate these issues.
Data Collection

Within a design experiment framework, I use a combination of naturalistic, qualitative methods to address the research questions listed above. Drawing from Stake’s case study methodology (1995), I develop narrative descriptions of each case in order to give the reader a sense of how events at each site unfolded, particularly what factors went into any design changes made. Consistent with Cobb et al. (2003), I present the case in terms of turns in classroom practice with each case chunked by shifting classroom dynamics.

Stake’s case study technique is particularly useful in studies exploring contexts such as the ones investigated in here, where the very circumstances in which the research is conducted are unstable: Which students are in class on any given day is unpredictable and the extent to which those who do attend are willing to participate in the research (and in what ways) must be tacitly negotiated throughout each and every class period. Entering such an environment with a detailed and inflexible agenda and pedagogical design would be unwise. Not only would you alienate (if not lose entirely) most of your voluntary participants but you would also overlook emergent needs, patterns, activities, behaviors, and issues within the very context and activity structure you had hoped to understand. Stake’s case study methodology is most suitable in such contexts because it is responsive to emergent findings, emphasizing the importance of not over-prescribing data collection and research procedures but rather allowing data collection to follow emerging themes (cf. Stake, 2003). He writes,

“Initial research questions may be modified, or even replaced in mid-study by the case researcher. The aim is to thoroughly understand [the case]. If early questions are not working, if new issues become apparent, the design is changed. Malcolm Parlett and David Hamilton (1976) called it progressive focusing.” (Stake, 1995, p. 9)

Through progressive focusing – changing a design in ways responsive to emergent findings – researchers are particularly apt to catch the unintended (and not just intended) consequences of an educational intervention. Stake writes, “the real business of the case study is particularization, not generalization” (1995, p.8). For him, the purpose of case research is interpretation of the local, followed by development (when necessary) and application of research methods and techniques that are in response to the needs of the case.

In the work presented here, I take a fairly moderate stance on researcher intervention and interpretation. On the one hand, as an educator and not just a researcher in each setting, I felt obligated to modify and, at times, even remove wholesale various elements or aspects of the instructional activities when and where it seemed in the students’ best interest. In order for my instructional design to be responsive to local needs in this way, I had to act as interpreter within the field. On the other hand, whereas Stake’s methodology relies heavily on the continuous interpretation of the researcher in situ, I attempted to minimize my role as interpreter while in the field in order to protect the integrity of the original data. Generally, my more moderate solution was to structure on-site field notes around straight observations and then, after triangulating and supplementing our observations among the teacher(s), the hired researcher, and myself, to then use these observations to generate cases (cf. ethnomethodological approach, Silverman, 1993). Each primary case document – including unmitigated observations, interviews, and student artifacts – then became the basis on which any inferences were made. Because case documents were compiled after each day on site, I was able to use the observations as the basis for making pedagogical decisions. Inferences beyond that – ones which bore on the research questions themselves – were made during the data analysis phase subsequent to data collection (see the Data Analysis section below). A description of the specific observation, interview, and student artifact collection methods follows.

Consistent with Stake’s (1995) case study approach to data collection and reporting, I use observations, interviews, and document analysis. Stake (like many qualitative researchers) argues that researchers should use multiple data sources to generate more trustworthy and credible data. This process, frequently called triangulation (Denzin, 1989), thereby increases the degree of validity of the assertions generated by the research, as the researcher builds arguments based on less “contestable data” (Stake, 1995). A description of these specific method and the inter-relationships among them follows.

Observations

Because I played an active role as a participant-observer in each environment, I ran risks of both becoming vested in the project and missing important interactions occurring away from me. In order to ameliorate this potential source of bias, I employed a trained researcher to attend each class, take field notes, and discuss emerging findings between and after sessions. My original intention was to have the second researcher videotape the classes; however, such data collection became unfeasible as students resisted the intrusion (as in the Media case) or participant consent could not be secured (as in the YWCA case).

Data collection focused on two objectives: 1) Capturing students’ game play activities, including what year they were in, what civilization they were playing as, and, crucially, what game practices they engaged in, and 2) capturing social interactions, including the content of the talk among participants, the character of the social interactions surrounding the game play, how knowledge was shared among them, what meanings became taken-as-shared with in groups of players, instances when players use Civilization III to discuss understandings of other social phenomena (and instances when they didn’t), and occasions when cooperative or competitive play emerge.

Both researchers adopted an active stance as educators in the classroom, choosing to probe students’ understandings with questions and observations. We made regular use of informal interviews to query and clarify observations and learn more about players’ own understandings of their activities and context. Such regular participation as facilitators/instructors in student’s activities “risks” engaging players in reflective activities known to promote learning and therefore seriously compromising the generalizability of any findings. However, given the exploratory nature of this research and the double role we ended up needing to play, I chose to err on the side of the students’ benefit and simply be diligent and systematic in recording our participation in student activities so that, in later generating inferences and conclusions, I might be able to at least partially disentangle the effects of our presence from the shape and trajectory of the naturally occurring activities overall. I then purposely wrote the cases in a manner that might allow readers to accomplish such disentanglements when and where I was unable, making explicit note of when and how we intervened in activities and shaped students interactions with each other and/or the game.

Interviews

Students were interviewed throughout the study in order to probe researcher observations and hypotheses and to gain participants’ views in order to triangulate their interpretations against our own. Both researchers asked frequent formal and informal questions to students, often eliciting explanations of students’ actions – both what they were doing and why. In addition to frequent informal interviews, nine participants from the Media case(s) and four students from the YWCA case were given semi-structured interviews following the unit that included performance tasks. We queried interviewees about their attitudes toward social studies and probed their conceptual understanding of social studies phenomena, using both discussion and performance tasks dealing with historical timescales, maps and texts (see Appendix D for the interview protocol). We also gave students dynamic assessments (Brown, 1992), including novel historical, geographical, and political scenarios in order to understand their thinking on related topics that they had not yet encountered. All interviews were audiotaped and transcribed.

Whereas researchers coming from objectivist epistemologies might treat interviews as socially neutral, I view these interviews as social discourse – co-constructed dialogues between researcher and participant. This social perspective acknowledges that both parties have profound impact on how an interview unfolds as both shape responses and adapt to one another based on gestures, intonations, and language in a struggle to develop and maintain intersubjectivity. Thus, I did not hold back from nodding or commenting on students’ responses, although I did try to avoid leading them toward specific answers. Aware of the possibility that students may try to please (or resist) me, I framed the interviews as opportunities for students to help me understand what they learned (or didn’t) through playing the game. The interviews were not constructed as tests or assessment situations, and students did not seem to approach them as such.

Student Artifacts

Each day, I collected students’ log sheets (See Appendix C) and inscriptions created in support of their game play. On most days in the unit, participants wrote a short description of what happened that day in a journal, noting in particular any major game decisions they had made. We tried a variety of activities in each case, ranging from student timed-writings to having students submit questions on post-it notes in order to support learning. The relative success of these activities are detailed in the cases; regardless of their pedagogical success, however, they did generate documents that were particularly useful for triangulating researcher/teacher observations and students’ actual goals and activities.

Data Analysis

For each case, observations, transcribed interviews, and student artifacts were combined into a primary case document. In this way, I use multiple data sources for each case as a way to generate more trustworthy and credible data. Information gleaned from one source was checked against and used to supplement other sources – a process frequently called triangulation (Denzin, 1989) – in order to increase the degree of validity of the assertions generated. First, I translated each primary case document, roughly 300 pages on average, into a coherent narrative flow of events by combining chronologically relevant data from multiple sources together and removing redundancy. Working through several iterations, I then reduced the text by removing extraneous or theoretically uninteresting occurrences¹⁴ – what Stake (1995) calls winnowing and sifting the data. Incongruent text which did not match my study objects were noted and included in the case for the reader to examine.

I kept these reduced primary case documents in narrative form, consistent with an historical-cultural approach, in order to report, in a contextualized fashion, what practices emerged and how learning occurred within the environment. As such, the reader can vicariously experience (Stake, 1995) the decision-making processes involved in the cases, perhaps using this experience as the basis for drawing generalizations toward other contexts. Narrative-based research methods allow the researcher to uncover and share the many design decisions that went into the study, hopefully illuminating the practical and theoretical issues in using gaming technologies to support learning. Following Stake’s (1995) work, each case narrative is based on incontestable observations, observations that most any ordinary observer might agree with if present, as closely as possible. I then shared each case narrative with both the second researcher and the local teacher from each case as a way to verify the accuracy. Typically, the second researcher had substantive feedback while the case’s teacher had little beyond a few minor factual corrections.

Using the constant comparative method (Glaser & Strauss, 1967), I analyzed each case based on emergent themes and patterns that bear on the original research questions, scanning the data for counter-factual evidence and competing contradictions and informally verifying the inferred patterns with the second researcher as they were established. Specifically, the analysis focused on: (1) the everyday gaming and social practices that emerged in each classroom, (2) moments of engagement or lack thereof, (3) students’ displayed understandings of historical phenomena and how the game might (or might not) be remediating those understandings, (4) the affordances and constraints of Civilization III as a tool for learning within each educational context, and (5) how the game aligned or misaligned with schooling practices already in place at each site. I apply two distinct theoretical lenses to each case narrative in order to make tease out subtle yet significant patterns in the data that might bear on these five issues: Cobb et al.’s (1999) notion of “taken-as-shared meanings” and activity theory.

Taken-as-Shared Meanings

Consistent with Cobb and colleagues (1999), I searched for taken as shared meanings that arise in each case narrative. Taken-as-shared meanings are knowledge arising from activity that participants accept as understood and use as tools for solving problems. The notion of taken-as-shared meanings embodies a socio-culturalist epistemology in which knowledge is viewed as socially constructed and arising through use. If knowledge is something results through interaction in the world as mediated by social structures, then research learning needs to focus on sites where communities generate and accept shared meanings as sites for learning, taking particular note of spaces in which knowledge is taken as understood and is used as the basis for solving future problems. For example, in their studies of mathematics classrooms, Cobb and colleagues (1999) identify places where students use mathematical principles for solving more complex mathematical problems. Cobb looks for places where meanings are articulated and reified, and then discussed in problem solving. In this study, I examine places where understandings emerge through participation in game playing communities and examine how these understandings are used to inform game play as well as other learning tasks. Consistent with Cobb, I look for places where knowledge is taken as shared, situations I attempt to probe further in interviews.

Activity Theory Analysis

Consistent with earlier work characterizing activity systems (e.g. Barab et al., 2002), I analyzed the cases through an activity theory framework, using this framework to characterize the tensions in each activity system, and suggest ways that the system might evolve to support more robust learning exercises. Activity theory analysis is an ideological approach to data analysis, one that takes an existing social theory and then analyzes data through that lens. Such ideological approaches (feminist or Marxist theories are other examples of ideological approaches) offer the researcher powerful tools for gaining insight into a phenomena, saving the researcher work in having to retheorize situations in each setting. By basing an analysis on a particular theory, researchers can build on existing theory which can then be used to better understand human behavior. Ideological approaches also risk steering the researcher in an ideologically-driven direction, biasing the researcher toward seeing particular relationships or where there are none. In this study, I consciously chose to write my case studies before doing the activity theory analysis so as to minimize ideological bias.

Consistent with Engeström (1999), I developed depictions of each activity system. Using these depictions, I identified primary and secondary contradictions in the activity system. These activity system analyses serve to illuminate issues in the design of game-based learning environments and help educational designers understand the problem space of designing game-based learning environments. In reporting my analysis, I foreground the reporting of contradictions, as the tools, resources, and social structures are considered in the case studies.

Using the constant comparative method (Glaser & Strauss, 1967), I developed characterizations of each activity system, scanning the data for counter-factual evidence or competing contradictions. I shared these contradictions with the second researcher who agreed that these were primary contradictions driving activity in each situation. The only substantive difference was in the YWCA case, where both researchers had difficulty generating contradictions and felt that the analytic frame of activity theory may have created contradictions where there were none.

Developing Conclusions

After developing an interpreted narrative for each case, I examined learners’ experiences across cases to develop assertions about how practices emerged across settings, the role of social interaction in mediating activity, how engagement developed in each setting, and what role gaming played in learning. Because there were substantial differences across cases (as opposed to cases matched on all but one or two variables), I treat them as collective case studies (Stake, 1995) – multiple case studies combined in order to illuminate a broad range of research themes. In effect, collective case studies allow the researcher to see past the particularities of a single case.

I use the constant comparative method (Glaser and Strauss, 1967) to develop conclusions – qualified, data-driven statements that address research questions and have theoretical and / or practical importance (Mabry, 1999). As in developing the case studies, I searched for disconfirming evidence or counter-evidence to the claims provided here. As I developed conclusions, I iteratively scanned both the primary case documents and the interpretive narratives of each case. Consistent with the exploratory nature of this research, I attempted to delineate how Civilization III worked as an effective instructional resource as well as where it was problematic. My conclusions attempt to account for the variety of students’ experiences both within each case across cases. Places where the instructional unit failed to capture students’ interest or produce understandings are noted alongside locations of productive learning. Drawing from Cobb’s notion of iterative design, the goal of this study is not to argue for Civilization III as a resource in world history classrooms; rather, the purpose is to understand the properties of these instructional designs and illuminate underlying theoretical and practical issues in developing game-based learning environments.

Petite and grand generalizations. The conclusions of this study should be regarded as petite generalizations – generalizations about the activity in these particular, situated cases that may (or may not) have direct import for the reader’s own contexts. As Lincoln and Guba (1985) suggest, qualitative case studies cannot generate findings or assertions with universal applicability; instead, their power lies in situating the reader within narratives that allow the reader to make inferences about her own situation. Case studies are simply not well suited to producing generalizable results as the uniqueness of the case always persists. Still, as Stake (1995) argues, case studies can, in some instances, suggest problems with, or productive modifications to, broader theories, often suggesting productive avenues to explore in future research or caveats to those in the field.

Although I do suggest places where these data suggests holes or modifications in existing theory, the conclusions contained herein are petite generalizations, not grand ones. Consistent with Stake (1995), I believe that methods other than the case study are necessary for actually causing modification to grand theories rather than merely suggesting what their direction might be. Within the design experiment paradigm, I have argued that differing methods need to be employed at different times in order to understand design space, and that we are most likely to see advancements in theory through sustained research efforts across multiple contexts (Jenkins, Squire, & Tan, in press). Stake (1995) makes a similar argument, referring to the use of multiple research traditions or the collaboration of multiple researchers on one data set as triangulation of methodologies. This study is but a first step in understanding how digital game-based learning environments might function, particularly in world history, in order to provide a foundation for further research and theory in this emerging field.

Theory triangulation. By using several different theoretical approaches to a data set, one can get deeper, richer understandings of both practical and theoretical issues at play in a case. Stake (1995) calls this theory triangulation. In this study, I draw from two distinct research traditions: (a) Stake’s responsive case study approach, and (b) activity theory. Each offers unique perspectives and ways of looking at the data. Because ideological research approaches such as activity theory can be theoretical blinders, masking important trends in data and organizing the research toward certain conclusions, I developed my case studies before performing my activity theory analysis.¹⁵ After developing each case and then examining each context from an activity theory point of view, I generate conclusions, petite generalizations that draw from both the case studies and the activity theory analysis. Combining these two approaches not only yields a deeper, richer understanding of the data but also suggest differences in each theoretical approach, giving future researchers perhaps a more nuanced appreciation for how each theoretical system works and what each reveals and obfuscates. I close the dissertation by suggesting implications of this study. In particular, I try to tie these experiences to the emerging discourse on gaming and learning.

Directory: kdsquire
kdsquire -> Toward a Media Literacy for Games
kdsquire -> Running Head: Higher Education of Gaming The Higher Education of Gaming
kdsquire -> Theory By Design
kdsquire -> Critical Design Ethnography Running Head: Critical Design Ethnography
kdsquire -> Curriculum & Instruction 614: Instructional Computing II
kdsquire -> Spring 2004 Course Syllabus
kdsquire -> Augmented reality simulations on handheld computers
kdsquire -> Environmental Detectives – The Development of an Augmented Reality Platform for Environmental Simulations
kdsquire -> Video Games in Education