Building with Jell-o: A Pragmatic Model of Learning

This little piece represents my personal philosophy and values when it comes to helping people learn. I originally wrote it for the class Foundations of Instructional Psychology and Technology at BYU.

I first conveyed my Jell-o theory to a few friends in  high school. First, I posited, you come to believe some fact about the world. Then you build on top of it, constructing additional ideas about how the worlds works. But beware, for these building blocks are not made of granite; they’re Jell-o. They’re shaky, amorphous elements that you’ll likely need to replace later on. Without realizing it, I had adopted a Jell-o flavored cognitivist framework for conceptualizing my own learning.

Truth and Lies

I came to Jell-o theory through wrestles with religious doctrine and high school physics. As a young child, I distinctly remember believing that God not only knew everything but caused everything, which conveniently made Him responsible for my mischievous behavior. Eventually this odd determinist model collapsed under the weight of stronger notions of free will and accountability that I found critical to Christian faith. Similarly, as an adolescent I came to believe in the solar-system model of the atom as taught in a classroom textbook. Imagine my chagrin when I learned a few years later that instead of orbiting their nuclei in neat concentric circles, electrons behave quite unlike any mass that we observe at a human scale, let alone a planet orbiting a star. These and other unexpected demolitions of my mental models inspired a sense of increased caution and flexibility with regard to what I thought I knew.

The Utility of Jell-o

Half-baked, gelatinous ideas present an interesting paradox: their inaccuracy precludes them from standing the test of time, but they are solid enough to build upon, at least in the short term. In his landmark work on user experience design, Norman (2013) illustrated this notion with the example of vehicle controls such as steering wheels and boat tillers, observing that one’s mental model of how the control works need not be very accurate in order to be useful. Indeed, most operators understand very little about the actual mechanism, but their knowledge of how it works is typically sufficient to operate it successfully, even if their model is mostly erroneous (Norman, 2013, p. 22).

In a complex and competitive world, the high cost of rational confidence may encourage us to stick with inaccurate but helpful models. For example, if a friend gets sick after eating a certain mushroom, I could spend time carefully isolating the variables to determine whether the mushroom was the culprit. But if hunger is driving me to a quick decision, a wobbly Jell-o model of mushroom causes illness is probably sturdy enough to support a safe dinner.

This pattern also helps explain why fallacious heuristics persist in human society. Stereotyping people by their appearance is irrational (and unfair), but it presents a useful shortcut for identifying potential enemies and allies in a matter of seconds. The trade-off for this incredible speed is severely diminished accuracy, which carries the risk of serious consequences. In the case of stereotyping, inaccurate judgments have led to such injustices as housing discrimination and unwarranted police brutality. These problems serve as sobering reminders that our mental models are often insufficient to help us realize our best ideals in difficult situations.

Not All Models Are Created Equal

Since our Jell-o models can have such grave ramifications, it makes sense that we invest so much energy in improving them. Hence the rationale for teaching: let’s help people replace a jiggly cube or two with something a little sturdier, so that they can better navigate their life as rational and virtuous citizens. But what does it mean for a model to be sturdy? The simplest measure of sturdiness is accuracy, meaning the fidelity with which a model reflects reality. This assertion assumes an objective reality, a trait that would confirm Jell-o theory as decidedly cognitivist (Ertmer & Newby, 2013).

But this conceptualization of accuracy is problematic, because we don’t carry reality around in our heads; we have Jell-o in our heads. My physics teacher had reason to believe that his model of probability clouds was more accurate than my solar system model, but like Thomson and Rutherford before him, his head was not filled with “correct ideas,” but with very mutable (albeit a more rigorous) Jell-o. Restrained by our lack of definitive knowledge, the best we can really offer each other is a normative understanding, an agreed-upon model that represents our best collective thinking for the moment. I find this approach refreshingly pragmatic, because it optimizes mental models by their usefulness in dealing with other people in the world. For example, normative language proficiency isn’t the power to speak correctly in some universal sense, but rather the ability to use language to successfully interface with most people in a given language environment.

Where a normative model exists, it behooves us to help others toward it. The cognitivist roots of Jell-o theory suggest that this may be accomplished by introducing cognitive dissonance—by stressing and stretching a learner’s model with new ideas that require accommodation, or remodeling their Jell-o (Seifert & Sutton, 2016). The Socratic method offers a compelling technique for helping a learner stress-test a mental model and reflect aloud on the implications. Indeed, requiring a learner to articulate his understanding provides a convenient way to monitor his Jell-o building activity (Chinn & Sherin, 2014).

Direct experience and experimentation are also powerful catalysts for mental remodelling. For example, learners might reevaluate their notion of density as they observe objects with equal weights behaving differently when immersed in water (Chinn & Sherin, 2014). Techniques like quantitative domain mapping and microgenetic methodologies can help designers and instructors understand a learner’s likely path through a series of increasingly normative models, even suggesting occasional detours through less-accurate models that point the learner in the right (i.e. normative) direction (Bunderson, McBride, & Wiley, 2009; Chinn & Sherin, 2014).

The Long Game: Training Jell-o Architects

Still, not all domains are centered on a predominant normative model. While few would dispute the global consensus around the distributive property of multiplication, views on capital punishment vary widely enough that no clear norm prevails. In fact, most of life’s important questions lack an obvious normative answer. Such questions reinforce the importance of the Jell-o theory itself: the meta-model that explains our knowledge-building activities. When learners are endowed with awareness and understanding of their own learning, they are free not only to come to their own conclusions (build with Jell-o), but to examine how they came to them and deliberately refine them (architect their Jell-o).

This awareness allows Jell-o architects to tackle unanswered questions, sometimes for their themselves and often to the persuasion of others. It them to test competing models against each other in search of a compelling framework for understanding the world. It allows them to identify and adjust for their own biases. Experienced architects can even temporarily adopt replicas of other people’s models to try them out, or as Aristotle might say, entertain their ideas without accepting them.

Such architects are capable of clear communication and intense empathy, appreciating why others might feel and think the way they do. Their empathy in the face of misunderstanding is compassion without condescension, tempered by the humility of remembering how frequently their own understanding has required renovation. Such humility scales with their learning, as their expanding knowledge inevitably widens their view of what they don’t yet understand.

Equipped not with all the answers, but with the tools to seek them (and the know-how to get by until they do), Jell-o architects naturally make the best teachers. They are role models for excellent learning, exhibiting traits and tactics that learners can emulate even before they understand why they are valuable. As these learners mature, architects can recruit them as apprentices to assist in the construction of new frameworks and theories, building new knowledge and a new generation of knowledge architects in the process. The ultimate objective of Jell-o theory is just that—to help people become stewards of their own learning who can learn from anyone and any situation, and help others do the same.


Bunderson, C.V., McBride, R.H., & Wiley, D.A. (2009). Domain theory for instruction: Mapping attainments to enable learner-centered education. In C.M. Reigeluth and A.A. Carr-Chellman (Eds.), Instructional-design theories and models: Building a common knowledge base (pp. 327-347). Hillsdale, NJ: Lawrence Erlbaum Associates.

Chinn, C.A., & Sherin, B.L. (2014). Microgenetic methods. In R. K. Sawyer (Ed.), The Cambridge Handbook of the Learning Sciences (pp. 171-190). Cambridge, GB: Cambridge University Press.  

Ertmer, P.A., & Newby, T.J. (2013). Behaviorism, cognitivism, constructivism: Comparing critical features from an instructional design perspective. Performance Improvement Quarterly, 26(2), 43-71. doi: 10.1002/piq.21143

Norman, D.A. (2013). The design of everyday things. New York, NY: Basic Books

Seifert, K., & Sutton, R. (2016). Behaviorism and constructivism. In Foundations of learning and instructional design technology (Foundational theories). Retrieved from