I’ve been doing a lot of thinking about these declarations since they started appearing, and I’ve come to the somewhat troubling conclusion that I don’t think I care if learning objects are dead or not. My primary interest always has been, and I suspect always will be, in increasing access to educational opportunity to people who have been denied that right for any of a variety of reasons. I loved the learning objects idea because the “write once, use anywhere” idea had a lot of economic appeal – once an object had been created for whatever reason, we could copy it (for free) and send it (for very close to free) almost anywhere around the world to be employed in the exercise of an individual’s right to education.

For a very long time now (in 1999, in 2000, and heck, NSF even gave me a CAREER award founded on this criticism in 2002) I’ve been saying that the idea of LEGO-like assembly of resources simply will not work from a learning perspective. The role of context is simply too great in learning, and the expectation that any educational resource could be reused without some contextual tweaking was either naive or stupid. I will here attribute learning objects’ inability to live up to the incredible hype and investment they received to the fact that the premise of the possibility of simple reuse was simply wrong.

An example.

The ultimate success story in the “write once, use anywhere” history of educational materials is the textbook. However, you will notice in this long and storied history that there has never been any confusion over whether or not a collection of algebra, algebra ii, geometry, trig, and calc textbooks could be “simple sequenced” and presented to a learner without additional contextualization and support. Or that the sections in one of these books could be simple sequenced (to become the textbook) for use by learners without significant contextualization and support. As I enjoy saying frequently, “libraries would never have evolved into universities” if all that education depended on were preexisting, high-quality resources.

So if learning objects are dead – and they may be – what is it that we should care about? As instructional technologists interested in further empowering people to exercise their right to education, what should be the focus of our design and research efforts? In a previous JIME article Stephen left the idea of learning objects behind and encouraged us to think simply about “resources,” and get away from the jargon of learning objects. There’s something to the idea of simplifying things that I like quite a bit. However, for my purposes (and I readily recognize they may not be your purposes) I have a need for something more than just resources. As I’ve thought about that need, I think it is best expressed as easily localizable resources.

In the first round of learning objects definition wars, I contributed “any digital resource that can be reused to mediate learning” as my best shot. In retrospect, the primary weakness of this definition was supposed to be the keyword it all hinged upon: “reuse.” Because the systems that authored, managed, and delivered learning objects were all software systems, the majority of the people doing the actual work on learning objects implementations were software engineers (or people parading as software engineers). “Reuse” was almost unanimously interpreted by this group as “technical interoperability” with no thought for the pedagogic, semiotic, or other contextual dimensions of the term. The whole learning objects field of work turned into a giant software engineering exercise aimed at answering the question “can your content send scores for true / false items to my management system?” Because the term reuse (as used by many more people than just me, I’m certainly not trying to hoard all the blame here) was only partially understood, learning never really got into learning object systems. If anything, they were “technically interoperable content systems.”

Now, for my money, the technical interoperability of content doesn’t need to go much further than “can be properly rendered by most web browsers.” (IMS or SCORM Content Packaging is nice since it gives us a way to move metadata around with content, but my last statement was about content.) When you really believe that reusing educational resources is a contextualization or localization exercise, and not a matter of intelligently slapping a “Next =>” button somewhere on the object, it turns out that you don’t need much more in terms of technical interoperability than what every good students knows at the end of an HTML course. Create your content in such a way that it will render properly in most browsers and don’t purposefully futz with your source code so that people have a hard time seeing what you’ve done (WebCT’s HTML, anyone?). Feel as you may about the GPL, WebCT and others might do well to remember its language here:

Accompany it with the complete corresponding machine-readable source code, which must be distributed…on a medium customarily used for software interchange…. The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable. GPL

What if all the effort and money spent hyping and building technically interoperable content systems had gone into better understanding the process of localizing educational materials, and developing whatever new tools were necessary to support that process? <sarcasm>Of course, there’s very little market for these processes and tools, because when you’re talking about supporting people who have been unable to exercise their right to education, you’re obviously talking about “poor people,” and how would you make any return on products developed for “poor people?” I mean, after all, how are they supposed to pay?</sarcasm>

So whether learning objects are dead or not, I couldn’t say. And to some extent, who cares? As long as people are willing to (1) openly share (2) educational materials that will (3) render properly in most web browsers, and they also (4) provide access to the unobfuscated source for the materials (especially for Flash files, Java applets, Photoshop images with many layers, and the like), I certainly don’t care. Argue about what to call them all you like – I’ll be busy trying to help someone somewhere figure out how to localize some of these things so that they can actually derive some value from them – maybe even improve their lives some. Won’t you help, too?

As I have said many times, I believe that the term “pedagogy-neutral” is not adequately descriptive, and instead the term “pedagogy-agnostic” should always be used. I purposely choose “agnostic” because of its religious implications: a pedagogy-agnostic standard “doesn’t know if there’s a pedagogy or not.” In other words, it is impossible to design a language which allows the expression of *any* pedagogy without simultaneously allowing the expression of *no* pedagogy.

When using a standard which is “truly pedagogically neutral,” a person must be able to express an instance of any particular pedagogy. If this is the case, then a person must of necessity also be able to express no pedagogy whatsoever. If the language of expression places no constraints on the expressor (let’s pretend that its possible for such a standard of expression to be agreed upon), so that they can express any pedagogy they desire, the language is incapable of keeping them from expressing anything else – because there are no constraints on expression! Expressions of great pedagogy and no pedagogy may both be well-formed and validate.

One might say, “judging whether or not the pedagogy is effective,” (as if this phrase is meaningful without specific reference to specific individuals and specific content) “is not our goal. Our goal is to enable the machine interpretable expression of the pedagogy itself.” This leads us to abstract coneptual languages like IMS-LD. Using this language we can model any one or set of individuals acting in specific roles, in specific environments, with specific resources, as they perform certain activities.

Note that we can model not only teaching in this way. As the IMS-LD spec points out, we could model all of drama using these contructs (actors play parts amongst scenery advancing a storyline). But in addition to all the teaching ever done and all the drama ever performed, one quickly comes to see Wertch’s analytic construct of “person acting with mediational means in a sociocultural context” or Engestrom’s notion of activity theory – people using socioculturally grounded tools to work toward goals. In other words, we see that high-level, abstract conceptual languages like IMS-LD are, in fact, capable of modeling the whole of human activity. What of it does not involve people, resources, environments, and activities?

And so IMS-LD is not only pedagogically-neutral, it is pedagogically agnostic – capable of modeling in machine interpretable format the wide range of human activities, any specific one of which may be neatly within or far beyond the tiny realm of teaching.

Is this a good thing or bad? Realizing that the primary focus of the standard is modeling human activity does raise certainly questions. For example, why would we model human activity? In order to better understand the roles humans play, so that we can neatly replace them with “intelligent” machine subsititutes? Is it because people are so excited to use the automated support option when they call Dell after their Windows machine blows up? In asking over 2,000 people now if, when they need support, they choose the autoamted system or a real-live human support engineer, guess how many hands I have seen go up for the autoamted system? About 8. Ooh! What a demand! Let’s replace other humans who provide support too! Like learning support!

Why would we remove the communications from telecommunications? “It is still communications, surely” some will argue. Excellent. As long as your machine teacher sending me bits counts as communication, then surely my machine student sending back the “right” bits when prompted by the machine teacher will still count? We could learn something from the communications from machine to machine. If the teachers were to broadcast, error-check, and continue rebroadcasting and error-checking until they could confirm that the student accurately received the message, then perhaps something akin to communications would be going on. But we are so obsessed with the technology… determining the XML binding, making sure our XML validates, making sure it looks beautiful after being XSLT’ed, etc.

Why would we turn the greatest enabler of social interaction into a simple data download service? It’s like the great sequence of scenes in Real Genius. At a top school (I always thought it was supposed to be CMU, for some reason) a lecture room is full of students and a professor. But before long, the professor has been replaced by a reel-to-reel, and students simply scrawl down notes from the recorder. By the end, the reel-to-reel plays in the front to a room empty of all but miniature tape recorders. That is the direction all this automation madness is going, you know.

Now, I’m not against automation where it makes sense. I love an exciting computer game as much as the next person – I form hypotheses, test them out in the game environment, receive immediate feedback, tune my hypothesis, test them again, receive feedback again, and so on. But this and other automated systems
are not complete learning environments. They’re a temporary activity I engage in before going back to my primarily social learning interactions with real human beings.

This doesn’t mean I think distance education is a bust. On the contrary, I believe massively multilearner environments hold great promise (for some people learning some content), modeled on things like World of Warcraft. Why? They’re “the best of all possible worlds” – they place opportunities for automated, immediate feedback in a highly social context. But don’t get me wrong. There is no “best approach” for every individual and all content. But I think these MMLEs hold great promise for many people and a sizable family of content.

Likewise I don’t believe that IMS-LD is a bust. I do think, however, it could be put to more fruitful purposes. For example, when people carry out “educational research” of the horse race vareity (you know, my pedagogy is better than yours), IMS-LD could play a vital role. A key to good empircal work is replicability, which we almost never get in educational research reporting. You get garbage like “our constructivist approach worked better than traditional classroom instruction” without enough detail to ever dream of replicating either approach, let alone the study. If all reporters of educational research used IMS-LD to describe their methods, it would go a long way toward bringing needed rigor to our field.

Someone will, of course, rightly point out that, because it is a human activity, individuals engagements in MMLEs can be modeled in IMS-LD, too. Well, good. I hope someone does it.