Model-Dependent Realism in Biology

Biology has hit a wall. Biological systems are simply too complex to model and understand on a cellular level. Is there a way out? There seems to be.


Compared to the other sciences, progress in biology was measly during the medieval period, simply because the technology of that time simply could not satisfy all the experimental requirements of this science.

With the advent of modern technology, however, biology today holds the position of the most rapidly-advancing science. It promises to change the way we live and look at life. We’ve seen results already.

But such rapid advancement today confronts us with a greater problem: biological systems at the microscopic level are simply too complex to model, let alone understand. 

How do we get over this?

Model Dependent Realism

Although a key scientific philosophy much older than its name, the term “model-dependent realism” first found mention in The Grand Design, a 2010 book by Stephen Hawking and Leonard Mlodinow.

Model-dependent realism is the idea that we perceive the world around us through our senses, and then try to derive from our perceived observations an accurate mathematical model that succeeds in explaining what is happening, and predicting how similar future occurrences shall unfold.

Contrary to claims made by critics, model-dependent realism doesn’t discard the idea of objective reality. It accepts that there is an objective reality.

However, our senses are not infallible, and neither are our methods of observation and experimentation.  

This philosophy suggests that so long as a model we humans devise is successful at explaining and predicting events in nature, it ought to be accepted as an accurate description of nature. 

For example, how do we know that quarks exist? None of us have ever directly observed quarks. According to this philosophy, we accept this mathematical model because it is successful at explaining happenings in nature.

Have any of us ever seen an atom? Have we seen protons, neutrons, and other subatomic particles? No. We accept mathematical models that involve them simply because they are successful in explaining and predicting events in nature.

A consequence of this philosophy is that we humans must come to terms with a very uncomfortable truth: we will never really know, or see what the world really is. The best we can do is devise and stick to models of reality we create to explain and predict the world around us.

A Point in Conflict

The next obvious question would be, what happens if a model is later shown to be inaccurate?

For the sake of argument, let’s say a model becomes “inaccurate” when it fails to explain or predict certain occurrences in nature.

Do we then completely discard the theory? What does model-dependent realism have to say?

Let’s take Newton’s Theory of Universal Gravitation, and Einstein’s Theory of General Relativity to illustrate this point. They are definitely different models. Yet, they describe the same thing, gravity. The question which then arises is, which is correct?

Let’s take a closer look. Newtonian Gravity considers gravity to be a force, and relativistic gravity consider gravity to be an intrinsic character of spacetime, that is, the curvature of spacetime.

The curvature of space-time. Image Credit: Wikimedia Commons

Newton’s theory effectively explains and predicts gravitational interactions for everyday purposes. However, in certain situations, results predicted by Newton’s theory are (slightly) inconsistent with observations. It is in these cases that Einstein’s model is able to predict results that match observations.

Which is wrong, if at all? Considered separately, both Newton’s and Einstein’s model are successful at explaining gravitational interactions. It’s just that Newton’s model falls short in certain situations.

Enter model-dependent realism. Einstein’s model is correct. And so is Newton’s model, in most situations. Where Newton’s model fails, we use Einstein’s model. Where results predicted by both agree with observations, both are correct. In such a situation, we simply use whichever model we find convenient.

That is the very essence of model-dependent realism.

Where We Get Stuck

The very breakneck progress in biology I’m talking of has opened our eyes to a new reality. A reality that challenges the very foundations of the scientific principles we use to do biology.

Biology is rapidly advancing today. We have reached knowledge levels in biology no one could have dreamed of a century ago. We’re studying genetics. The brain.

Oh yes, the brain. That’s where we get stuck.

Biologists traditionally had the approach of trying to explain signalling in the brain by studying the physiology of neurons, the biochemistry involved; basically, using classical biology.

However, if you attempt to understand the brain using classical biology at the level of cells, you suddenly find yourself lost in a maze of gibberish you can’t make head or tail of! Why? Simply because it is too complex. More complex than what the best of our present mathematics can model. 

You just can’t proceed. You’ve hit a roadblock, and seemingly an insurmountable one.

Right now, it’s the brain we’re stuck with. Tomorrow, it might be something else. Later, some other item might pop up on that list.

Lessons From History

Science can broadly be divided into 2 categories:

  • Physical Sciences: That is, all branches of physics and chemistry
  • Life Sciences: All branches and fields of biology

This is elementary middle-school stuff. Looking at this, a question may arise in one’s mind, why is it that physical sciences have been subdivided into two broader fields, and not life sciences?

That’s precisely what I am talking about. Physics deals with interactions between the forces and particles of the universe. Whereas, chemistry deals with interactions between substances, which are actually composed on particles.

See where I’m heading? No? Let’s take an example. A double displacement reaction in chemistry.

AgNO3 (aq.) + NaCl (aq.) = AgCl (ppt.) + NaNO3 (aq.)

Silver (I) nitrate in aqueous solution reacts with sodium chloride in aqueous solution, to produce silver (I) chloride precipitate and sodium nitrate in aqueous solution. 

Chemistry says that when the reactants are dissolved in water, they dissociate into their respective ions.

AgNO3 = Ag+ + NO3

NaCl = Na+ + Cl

The anions are then exchanged. That is, the chloride ion combines with the silver ion to form silver chloride, and the nitrate ion combines with the sodium ion to form sodium chloride. That’s what the concept of a double displacement reaction says.

Ag+ + Cl = AgCl

Na+ NO3–  = NaNO3

Now, suppose there was no concept of a double displacement reaction in Chemistry. You’d be lost! Looking at the reactants, you’d probably think of concepts like the atomic model, electronegativity, dipole moments, and so on. Or you could go a step further, and use laws of quantum physics (just sayin’). Using them, probably after an extremely long and perverse calculation, you’d arrive at the same expected result (or perhaps the calculations will get so complex that you’d get stuck there).

You’d see it’s extremely inconvenient to do that.

So, what do we do? Instead of using all these concepts individually to weigh in every situation, we devise and use a generalized concept that is as good as predicting such scenarios: the concept of a double displacement reaction.

So chemistry, it turns out, is actually just another (simplified) model of the same interactions that physics describes! 

What Biology Needs is Model-Dependent Realism

Let’s get back to our talk of the brain. What I said was: if you attempt to understand the brain using classical biology at the level of cells, you suddenly find yourself lost in a maze of gibberish too complex to model and make sense of.

So, what’s the solution?

Go a few levels up from the microscopic level to a slightly more macroscopic level. Think of and model the brain as an electronic circuit. Does it help? Well, yes it does!

Since signalling interactions in the brain are too complex to model on a cellular level, jumping a few levels up and instead modelling the brain as an electronic circuit simplies things just enough for us to model and make sense of.

Nice one. Biologists are far from knowledgeable in electronics. The obvious solution then would be to bring in people qualified in electronics. That’s precisely what we’re doing.

Instead of asking people who’ve studied classical biology to study signalling in the brain, we’re bringing in people qualified in electronics to collaborate with the biologists in doing these studies. Because, as I said, classical biology alone is insufficient to study neural signalling.

It’s the same system. Both models are accurately predicting possible outcomes. We accept whichever one is convenient. In this case, it’s the electronic circuit model.

Closing Note

It’s not just the brain.

As biology progresses, we come across many occurrences that classical biology, acting at the cellular level, is insufficient to explain. We need more generalized models. Otherwise, we get lost in a maze of complexity.

Studying life is, for the first time, turning out to be a truly inter-disciplinary field. Classical Biology alone is insufficient. We need to amalgamate along with it, fields of chemistry, physics, and electronics, and perhaps, even more.

So it’s not just the biologists studying life. It’s also the engineers, physicists, electricians, chemists, and so on. They are an integral part of our biological research.

The human body, and life in general, is mind-bogglingly complex and amazing. The brain studied with physics was insufficient. Enter chemistry. Still insufficient. Enter classical biology, amalgamated with physics and chemistry. Still insufficient. Enter electronics. We’re getting somewhere, but also nowhere anytime soon.

Using physics, chemistry, classical biology, and all of these combined, we still get lost. Electronics is something, but not the entire thing. That’s pretty much all fields of science exhausted. 

Where are we heading? Would studying life require us to devise newer fields of science we don’t have today? We really don’t know.