You Have Died of Dysentery:

The Making of the Oregon Trail

Appendix 3: A Philosophy of Educational Software Design

This appendix consists of two parts. First we start with a transcript of a conference presentation that I gave in 1989. Then we conclude with a short essay that I wrote in 1992. Both parts of this appendix are helpful in explaining my philosophy of educational software design.

Computers and the Trend towards Non-Didactic Learning

A Presentation at the MECC'89 Conference

By R. Philip Bouchard / Bouchard Creations

November, 1989

Good morning! If you are looking for the presentation with the peculiar title, then you have come to the right room. My name is Philip Bouchard, and my talk is entitled “Computers and the Trend towards Non-Didactic Learning”. Before I get started, I’d like to briefly introduce myself. I’m a freelance educational software designer, and I’ve been designing and programming educational software since 1977. In 1981 I joined the MECC staff for 5 years, and I’ve been a freelance for the past 3 years.

In the past 5 years I have either designed, or been the principal designer, of 7 software products:

The Oregon Trail (1985) MECC
Word Munchers (1985) MECC
Number Munchers (1986) MECC
Odell Lake (1986) MECC
Checkerboard Trails (1988) Focus Media
Animal Trackers (1989) Sunburst
Weeds to Trees (1989) MECC

I have also assisted in the development of several teacher-training products.

My talk today consists of five parts:

  1. Introductory Thoughts
  2. Models of Non-Didactic Learning
  3. How Computers Fit In
  4. Software Demonstrations
  5. Summary & Questions

This outline also appears on your handout.

I.   Introductory Thoughts

I’ll start this presentation with three riddles, each of which I plan to relate to the issue of didactic vs. non-didactic learning.

My first riddle is a classic question which most of you have probably heard before:

Riddle #1:

If a tree falls in the forest, but no one is there to hear it, does it make a sound?

I’d like for each of you to decide whether the best answer is YES, or the best answer is NO. [Brief pause.] Okay, now let’s get a count of hands. How many of you decided on YES? [Count the hands.] And how many of you decided on NO? [Count again.]

So I see that some of you said YES, and some of you said NO. So how should we decide which answer is correct? Do we simply take a vote? Or do we find more rigorous way of making this decision?

In reality, the answer to this question depends upon the definition of a single word. Which word is that? [Pause.] Yes, you are quite right. The answer depends upon the definition of the word “sound”. So here are two different definitions of the word “sound”, both of which I obtained from the very same dictionary:

So now which answer to the riddle is correct? [Pause.] As you can see, the correct answer depends upon which definition for “sound” you are using. Therefore I can say that each of you chose a reasonable answer based on your understanding of the word “sound”.

Now imagine two teachers, each of whom put this riddle on a test for their students. Teacher A asked for the students to answer simply YES or NO. All YES responses were marked as wrong. Teacher B asked for the students to answer the question, and then to explain their answer. Any response that was adequately defended was counted as correct. Which teacher was following a didactic model of learning? [Pause.]

Yes, that is correct. Teacher A was following a didactic model, while Teacher B was following a non-didactic model.

My second riddle has a similar format to the first riddle:

Riddle #2:

If a teacher lectures to a room full of students, but no one listens, did any teaching occur?

Once again, more than one answer to this riddle is possible, based on how we define a single word. What word is that? [Pause.] Yes, the key word to define here is “teaching”. If someone answers YES to this riddle, then it suggests that they subscribe to a didactic model of teaching. But if someone answers NO to this riddle, then it suggests that they subscribe to a non-didactic model of teaching.

Myself, I prefer to define “teach” as “to facilitate learning in other people”. This definition for “teach” would, of course, imply that the answer to this riddle is NO.

Like my other two riddles, my last riddle has more than one possible answer:

Riddle #3:

What is the worst enemy of learning?

I would like each of you to select a word or phrase to describe what you feel is the biggest impediment to learning that our children currently face. [Pause.] Okay, let’s hear the answers that a few of you have selected. [Several people provide their answers.] Yes, those are all good answers to this riddle.

My own nomination as the answer to this riddle is… [brief pause] BOREDOM. Students simply don’t learn when they are not interested. You can lead a horse to water… You know the rest.

And yet it is such an irony that boredom should be a major impediment. After all, children are natural learning machines. Children are BORN to learn. Kids are naturally curious, and kids are naturally inclined to experiment. This is especially obvious with preschoolers. They pepper us with questions: “Mommy, what is that?” “Daddy, why are you doing that?” And they constantly conduct little experiments, such as:

And when the kids reach their teenage years – that is when their experiments REALLY give their parents gray hairs.

So how can we enlist these natural inclinations to help us teach? My personal answer is this: My goal is to help students to learn to love to learn. If I can just do that, they WILL learn. And they will continue to learn. They will, in fact, become life-long learners. And that is the most valuable lesson that I could possibly give them.

II.   Models of Non-Didactic Learning

So now it's time to define Non-Didactic Learning. But before I can do that, I first need to define Didactic Learning.

In Didactic Learning, the teacher is expected to be the font of all knowledge:

That's a really heavy role. In this model, knowledge flows from the teacher to the learner. The role of the teacher is to broadcast the knowledge, like radio waves, and the role of the learner is to soak it up. Unfortunately, both of these roles can be quite difficult. Furthermore, this model implies that learning is an extremely passive process.

By contrast, Non-Didactic Learning implies a very active learner. It results in a more learner-centered model. It suggests that what is important is NOT what is being broadcast across the room, but what is going on in the head of the learner.

The teacher in a non-didactic environment has 3 important tasks:

  1. Motivate the students.
  2. Provide adequate resources.
  3. Provide adequate guidance.

This last point is just as critical as the other two. In a non-didactic setting, children have a great deal of opportunity to explore and experiment. But it is not really FREE exploration, not completely, or else many students will flounder aimlessly. They need that gentle guidance to give them a direction.

There is more than one possible model of non-didactic learning. In fact, I have identified six, and I'm sure that there are more. There is a great deal of overlap between these models. A single, good non-didactic activity may fit neatly into several of these categories. The six models I’d like to discuss are:

  1. Socratic dialog
  2. Problem solving / Higher-order thinking skills
  3. Inquiry-based learning
  4. Cooperative learning
  5. Apprenticeship
  6. Constructivism / Consciously modeling reality
Socratic Dialog

As you know, the Socratic Dialog has been around for a very long time. Like the didactic model, this model puts the teacher in a high-profile role. But unlike the didactic model, where knowledge flows from the teacher to the learner, the Socratic dialog assumes that the seeds of the knowledge already reside in the learner. It is the role of the teacher to lead the learner through a dialog that enables the learner to pull that knowledge out, to organize the fragments into a coherent whole, and to verbalize it.

Problem Solving / HOTS

These days there is a lot of interest in the terms “problem solving” and “higher-order thinking skills”, or HOTS for short. This model is somewhat broader and vaguer than the other models listed here. But again it implies active learners, because the students are not exercising higher-order thinking skills if they are simply absorbing facts from a lecture.

Inquiry-Based Learning

What does the word INQUIRY make you think of? [Pause.] For me, the word makes me think of questions. In inquiry-based learning, students formulate questions, then find ways to answer them. The students themselves need to come up with the questions, so that they have ownership of them. There are two main ways to answer these questions. The students can design and conduct an experiment, or they can seek out the answer from a resource such as a book, a database, or an expert.

Cooperative Learning

Cooperative learning implies that two or more learners are jointly engaged in the quest for a particular piece of knowledge. This obviously requires active learners, and certainly can't happen in the middle of a lecture.


Admittedly, “apprenticeship” might sound to some people like a didactic model. And when done poorly, it can be. But a good apprenticeship activity puts the learners into an active role as they mimic the teacher. The teacher, in turn, models a process for the students to follow, rather than abstractly discussing the process. The emphasis is not on a formal, idealized process, but on a real, practical approach to solving problems. The students learn how to make decisions, not just follow a cookbook approach. Notice that this is very much like how we build computerized expert systems – by observing a human expert and learning how to make good decisions.

Constructivism / Consciously Modeling Reality

In a nutshell, constructivism is the idea that students construct their own understanding and knowledge of the world through personal experience, and by reflecting on those experiences. This philosophy of education is gaining traction primarily in the sciences. Rather than discussing all the variants of constructivism, I will focus on my own particular version of it, which I call “consciously modeling reality”. There is a lot that I could say about this model – I could easily talk for a full hour on just this one topic. But instead I’ll give you the 2-minute version.

“Consciously modeling reality” means that we explicitly tell our students that we are teaching them models of reality, rather than absolute reality. This approach brings out into the open all of the hidden assumptions behind the way that we structure information. To follow this philosophy, we must openly admit that the standard models we use when teaching are, to some degree, artificial and arbitrary – but quite useful anyway. By contrast, in didactic learning, we attempt to define an absolute reality. We provide the learner with pre-created models that we claim to be the only truth, and therefore we discourage any questioning of these models.

“Consciously modeling reality” can be summarized with three basic ideas:

Modeling Reality vs. Defining Reality

  1. Acquiring knowledge is dependent upon creating mental models to organize information in a meaningful way.
  2. These models differ from the reality upon which they are based. Hence, they are to some degree arbitrary.
  3. Our most important form of mental model is the taxonomy.

So what does this all mean? Idea #1 suggests that we are awash with information, but we cannot learn any of it until we create appropriate mental models upon which to hang the various bits of information to which we are exposed.

Ideas #2 says that models are simplifications of reality. Several different models – all reasonably valid – could be drawn from the same reality. But just because our models are arbitrary doesn’t mean that they are not important. On the contrary, they are essential, because no learning occurs without them.

Idea #3 reflects the pervasiveness of taxonomies – also called classification systems – in all stages of our educational system. Throughout the curriculum, we constantly rely on taxonomies to convey facts and concepts. Consider the following common taxonomies:

All of these taxonomies are somewhat arbitrary. When consciously modeling reality, we tell our students up front that these are models – very useful models – but that other models are possible. We discuss the reasoning behind the details of the model, and we give the students a chance to critique the model. Some students might even enjoy the challenge of trying to create a solid alternative model, or researching alternative models that already exist.

III.   How Computers Fit In

Now that we have gone over several different models for non-didactic learning, let's discuss how computers fit in.

First of all, I firmly believe that computer-based electronic media have AT LEAST as much potential for assisting learning as does the medium of print. We are just beginning to unlock the potential of this medium. And just as with the medium of print, there are many different ways to use this medium. Some ways work well with didactic models of learning, while other ways are better suited to non-didactic models. Even the technology terms we use can have implications as to our model of learning.

For example, consider these seven words, which can be combined to form nine different acronyms for the use of computers in education:

The nine possible acronyms are: CAl, CAE, CAL, CBI, CBE, CBL, CMI, CME, and CML. Let's consider just two of these acronyms: CAL and CMI.

CAL (computer-assisted learning) sounds like a non-didactic approach, with the learner at the center and the computer a mere assistant. There is also an implication that the teacher still has an important role.

But CMI (computer-managed instruction) sounds like a didactic model, with the computer at the center, controlling the learner and "delivering instruction" to him or her. Furthermore, it sounds like the teacher has been mostly cut out of the loop – replaced by a computer! Sadly, the more heavily managed the instruction, the more it tends to emphasize lower-level thinking skills.

So then, what uses of the computer promote non-didactic learning? I see three principal categories:

  1. Student tools
  2. Simulations
  3. Thought-provoking games

Student tools have received a great deal of attention in recent years. And the number of kinds of tools has greatly multiplied. In fact, we now seem to have a cornucopia of excellent student tools:

  1. Word processor
  2. Database program
  3. Probeware (also known as “laboratory interfacing” or MBL)
  4. Paint & draw programs
  5. Desktop publishing
  6. Multimedia presentations
  7. Telecommunications
  8. Other (spreadsheets, statistics packages, spelling checkers and other writing aids, etc.)

All of these put power in the hands of the student. All of these result in an active learner. All can be easily incorporated into various models of non-didactic learning. But you have already heard plenty about student tools at this conference, so let’s take a look at the other two categories:

Simulations and Thought-Provoking Games

Simulations and thought-provoking games are two categories with a great deal of overlap. A good simulation is often a thought-provoking game, and a good thought­ provoking game is often a simulation. But not always!

A simulation can be completely open, permitting free exploration, or it can be directed, thereby taking on some game-like characteristics. If these game-like characteristics are fully developed, then the product becomes a simulation game.

Games that are NOT simulations run the gamut from simple drills at the level of flash cards to highly thought-provoking activities at the level of chess. While games at either end can have value, it is the thought-provoking games that best support non-didactic learning.

The best of these simulations and thought-provoking games have certain characteristics in common:

1.    The activities are highly motivating.

For example, The Oregon Trail is certainly motivating to kids.

2.    The activities define meaningful criteria for success.

That is, meaningful from the standpoint of the learner. In The Oregon Trail, students readily identify with the goal of surviving all the way to Oregon. And if they do survive, they readily identify with the game-like score system.

3.    The activities permit degrees of success.

In The Oregon Trail, students can feel some success each time they get farther on the trail. Once they reach Oregon, they voluntarily start over, trying to gain more points, often at harder difficulty levels.

4.    The activities require the acquisition of skills and knowledge to achieve higher degrees of success.

This is critical. This is where the educational objectives come in. The students are highly motivated to repeat the activity, to get better and better at it. But to achieve these higher levels, the students ought to have to learn something.

5.    The software provides resources to assist in learning these skills and knowledge.

This is important too, but the software does not have to be the sole resource for the student. If the game motivates the students, then they will seek out other resources on their own, even if only each other. This tendency to turn to other resources should be encouraged.

When the learner is interacting with a good simulation or game, there should be lots of interaction with other resources:

Furthermore, the computer activity should be just part of a complete lesson that includes introductory activities and follow-up activities. Most important of all is the follow-up discussion; this step is just as important as the computer activity itself.

A good follow-up discussion...

1.    is based on a common experience

The computer activity provides a common basis for discussion.

2.    brings out many possible strategies

In a good simulation or game, there are many possible strategies. No single strategy is "correct," although some may be better than others. Each strategy has its own strengths and weaknesses.

3.    permits learners to verbalize their thinking

This is critical for effective learning, and greatly aids retention.

4.    can alternate with iterations at higher difficulty levels

For example, there could be one discussion after students have experienced the “easy” level, another discussion after the student have done the medium level, and a final discussion after the students have tackled the hardest level.

5.    can be a Socratic dialog

Because the students have already been exposed to most of the key ideas while engaged in the computer activity, a Socratic dialog can be a particularly effective method to assist the students in connecting the dots, consolidating and reinforcing the incipient knowledge in their heads.

Note that a good simulation or thought-provoking game can accommodate almost all models of non-didactic learning.

IV.   Software Demonstrations

[not transcribed]

V.   Summary

Below are some, but not all, of the points covered or implied by my talk today:

  1. Our job as teachers is to facilitate learning.
  2. Children are naturally inclined to learn, but they need to be motivated.
  3. Good teachers are good motivators.
  4. Non-didactic learning can improve motivation and retention.
  5. There are several complementary models for non-didactic learning.
  6. Non-didactic learning strongly encourages the construction of essential mental models.
  7. Student tools, simulations, and thought-provoking games are uses of the computer that facilitate non-didactic learning.
  8. A good follow-up discussion is an essential part of an effective HOTS activity.

My final thought is another classic bit of language that many of you have heard before. Help me fill in the blanks:

I hear and I __________.
I see and I __________.
I do and I __________.

[Pause.] Yes, that is it: “I hear and I forget. I see and I remember. I do and I understand.”

That concludes my presentation!

A Philosophy of Educational Product Design

By R. Philip Bouchard / Bouchard Creations

Written in 1992

I am driven by a vision. I envision a world where people of all ages love to learn, and where the learning is fun. I believe that learning should always be a great, exciting adventure, not a bitter medicine to swallow. And I believe that this world should be filled with materials and teachers that combine learning with fun. So what do I do for a living? I create highly engaging computer games that just happen to be educational. You might say that I have made a career in “interactive educational entertainment”.

I see no conflict at all between the twin goals of 1) creating a game that is decidedly educational, and 2) ensuring that this game is tremendous fun to play. My goal is to create educational products so enticing that children and adults are irresistibly drawn in. They taste a degree of immediate success, and yet see that it is possible to do better. They are driven to try again and again, with each iteration improving their skills and knowledge. And the whole while they are having a wonderful time. As a result they gradually and subconsciously develop a more positive attitude towards learning; they in fact learn to love learning.

My products are educational in the sense that they provide opportunities to explore scientific concepts, episodes in history, current societal issues, or other thoughtful material. I never try to disguise the fact that my games have educational value, but the main reason that people play the games is to have fun. Nevertheless, most people are pleased to be learning something “educational” while playing an entertaining game. Furthermore, after playing the game for awhile, many of them develop a true interest in the subject matter of the game.

I firmly believe that the educational elements of a game should be completely integrated into the game, not pasted on in an artificial and obtrusive way. What many people dislike about certain educational games is not the content, but this sense of artificiality and obtrusiveness. Even some of the best educational software on the market suffers from this problem. For example, in many products the educational content consists of a series of pop-up questions that interrupt the action. Each question (typically) has only one right answer, and you must answer correctly to continue the game. Furthermore, the questions themselves are only faintly related, if at all, to the game itself.

By integrating the content completely into the game, the design gains several distinct advantages. First of all, the game becomes more fun. Second, the content material becomes much more meaningful. Third, instead of drilling strictly on educational “facts”, players must make intelligent decisions on how to apply these facts in various settings. In other words, they exercise higher-level thinking skills.

But how is this done? How can the designer of an educational game incorporate the educational content into the structure of the game itself? For me, the starting point for a product design is not a topic, but an activity concept. It is not an issue of how to present information to be learned, but how to create an environment in which the player interacts with available information to make interesting decisions. In other words, the game itself must be the learning activity. Hence, my products usually incorporate a rich and elaborate environment for exploration and experimentation.

A key aspect of my designs is that the players should feel that THEY are in control. Most people are delighted with the opportunity to be in control. They often respond to this sense of empowerment by becoming intensely self-motivated – devising experiments, testing various strategies, and trying alternative approaches. It is through such higher level applications, rather than the parroting of facts, that true learning occurs.

A related point is that every problem should have more than one solution, and every solution should consist of more than one decision. For example, in The Oregon Trail, your degree of success depends upon the net effect of many decisions. While some solutions may be more efficient than others, no single decision is the sole cause of success or failure. In fact, most decisions in themselves are neither right nor wrong. Instead, the decisions must be integrated into a coherent strategy, because each decision affects the environment for subsequent decisions. Just as there is more than one way to prepare a salad or build a treehouse, there is more than one way to succeed at The Oregon Trail.

One other key aspect of my designs is that success is always relative. No matter how good a player gets, it is always possible to do better. In other words, there is no such thing as perfection, nor any arbitrary line between success and failure. Everyone can taste a degree of success, even on the first try. After that, the goal becomes to beat one’s own personal best, striving to do better or go farther with each attempt. And with each iteration, the player improves his or her skills and knowledge.

In summary, my key points are:

  1. Learning should be fun.
  2. Good games can be both fun and educational.
  3. All the “educational” elements in an educational game should be completely integrated into the fabric of the game.
  4. Good educational games involve higher level thinking skills, not just the presentation or memorization of “facts”.
  5. In my products, I like to provide a rich environment for exploration and experimentation.
  6. A good game gives the player the feeling of being in control.
  7. Every problem should have more than one solution.
  8. Games should be designed so that, whenever possible, success is relative.

*** End of Appendix 3 ***