At ESA, a friend of mine brought up how she disagreed with the way the Century model simulates soil decomposition.  As an empiricist, she disliked how Century was phenomenological and lacked mechanisms that are known to be important.  She was skeptical of how using such phenomenological models can lead to mechanistic insight.  And she thought Century was sometimes misunderstood and misused because modelers and empiricists don’t communicate well.

Even though her points were valid, I disagreed with her because I sometimes model.  And because it is more fun to argue than to agree.  So here it is, battles in science: modelers vs empiricists on how Century models soil decomposition.  The battle will go for 3 rounds, and I will judge who won each round because I’m an ecologist and that makes me impartial.  Let the battle begin!

Round 1: What Century is

Modelers argue that Century is a collection of mathematical abstractions; Century is a big, fancy hypothesis on how soil decomposition works.  And pieces of Century came from experiments.  Instead of criticizing how Century models decomposition wrong, empiricists should test the predictions of Century with data and suggest better equations to incorporate into the model, like how this study does.  Stop saying Century sucks because it doesn’t have microbes in it!

Empiricists argue that Century, with its three-pool structure, is too phenomenological to elucidate mechanisms if not obscure the important ones in the real world.  That study says Century fails to describe globally important soil carbon fluxes and pools because Century lacks mechanistic description of how microbes work. Century also lacks very important mechanisms that govern soil organic matter stabilization.  Century sucks because it hides how important microbes are!

Round 1 score: 10 – 9 modelers

Round 2: How Century is used

Empiricists argue that modelers use Century as a hypothesis without showing how to test it let alone actually evaluating it with real world data.  No empiricists can get away with publishing only a hypothesis and made up data, so why is that accepted for modelers?  Modelers are just waving their hands with Century and not doing real ecology!

Modelers counter that Century should be used to integrate piece-wise knowledge.  Century should be used to test which hypothesized processes and mechanisms are important in the functioning of soil, and to evaluate the consequences of how the hypothesized processes and mechanisms interact with each other.  These do not require modelers to get data.  Besides, usable data are hard to come by. The integration and generalization afforded by Century are part of ecology!

Round 2 score: 10 – 9 modelers

Round 3: How Century is explained

Empiricists argue that modelers don’t explain Century well enough.  Modelers need to explicitly say that Century is a heuristic tool.  Century models soil decomposition by dividing soil organic matter into three simple pools, ignoring that soils exist in gradients of materials, texture, chemistry, and microbes.  But modelers explain the model so poorly that some researchers think there are really three pools in soil that they can go measure somehow.  Just say Century is at best just an educated guess, if not just a guess!

Modelers counter that empiricists misunderstand Century because they fundamentally mistake knowledge for reality.  All knowledge is an abstraction of reality and not reality itself.  It’s not Century’s fault that empiricists misunderstand!

Round 3 score: 9 – 10 empiricists

So by my totally impartial judgment, by the score of 29 – 28, modelers win!  Leave a comment if you disagree or want to give another perspective.