top of page

This isn't rocket science!

When people talk about things that are complex, they usually talk about rocket science, which is more generally referred to as “aerospace engineering” (, or “the primary field of engineering concerned with the development of aircraft and spacecraft”.

Now, that’s a complex field, and obviously has a large number of very smart people in it, but it’s really only the tip of the iceberg of our various space programs. Even setting aside the contributions from other scientific fields, rocket scientists need to understand disciplines such as aerodynamics, propulsion, and avionics, all of which are large enough to have sub-disciplines and enough scope to make it impossible for any individual to be expert in all areas.

But this is all the easy stuff, really. We’ve understood the basic principles of classical mechanics ( and gravitation ( since at least the time of Isaac Newton (, and had a practical understanding of ballistics for much longer.

Rocket scientists (engineers, really) base a lot of their work on the same basic principles that Newton used, and most of the work is around building the tools to use those principles to get us to wherever we want to go.

Since the time of Newton, our understanding has increased dramatically, but it’s important to note that relativity and quantum mechanics, did not really “prove Newton wrong” in any way. They mainly refined the equations (or limited their scope) to take into account situations that Newton would have had no way to address.

For most practical purposes, Newton’s equations are accurate until you reach a significant fraction of the speed of light, or look at particles at the atomic scale. (That said, rocket scientists need to make adjustments for relativistic effects, mainly due to the scale involved in space travel)

How can anyone say this stuff is easy, though? We have a lot of very smart people and vast numbers of complex calculations and extremely precise measurements, and it’s really, really hard.

Exactly. The calculations are often quite hard, and there are certainly a lot of them, but it’s relatively easy because these are things we can measure to high degrees of precision and define precisely.

When we design something like a rocket, we understand what we need it to do to (eg, generate X amount of force for Y seconds with a payload of Z, or whatever), and can measure all of the components very precisely. The number of calculations, and the number of tools and processes certainly make this hard, but not REALLY hard.

So, if rocket science isn’t so hard, what is?

Humans. Understanding ourselves is vastly more difficult than understanding physics, precisely because we don’t even understand what to measure, let alone how. We’re really still at the “blind men and the elephant” ( stage of understanding how humans think and behave.

I have previously commented on metacognition (, and trying to understand human behaviour has been a part of the human pursuit of knowledge – probably for almost as long as there HAS been human pursuit of knowledge.

But how far can you get if you can’t measure it?

Hard sciences are an “easy” (relatively speaking, of course) place to start because concrete objects are relatively easy to measure. Concepts like mass, volume, and speed are easily demonstrable, and the scientific understanding we have gained is based on vast numbers of experiments of greater and greater precision. In recent years, experiments which claim to contradict basic principles of physics have turned out to be either incorrect or deliberately misleading. “Free energy” or “perpetual motion” machines ( are invariably scams.

But humans? What is a thought? What is an emotion? How do they work? How can they be measured? CAN they be measured directly at all, or do we need to figure out some clever indirect method of measuring?

Some things, we can measure.

In the early 19th century, people would measure the contours of the skull, and attempted to associate this with the mental traits of a person. Unfortunately, phrenology ( was quickly discredited and survives only as an historical pseudoscience.

Nowadays, we can measure many things about the physical brain, using techniques ranging from x-rays, thorough Magnetic Resonance Imaging (MRI), and numerous others. We have also learned a lot about the brain, how it works, and how it interacts with the body.

But when it comes to measuring thoughts, or truly understanding the systems for how thoughts are formed, and how emotions work, or how people will behave under certain circumstances, we are barely scratching the surface. Advertising, social media, propaganda, influence operations, social engineering, and numerous other fields focus on trying to get people to do something, but these are mainly practical disciplines, using techniques that have been shown to work (to varying degrees), without necessarily understanding why they work.

When we drop a weight from a given height, we can state with virtual certainty that it will fall, and can determine (if all of a small set of variables are understood and/or controlled) exactly where it will land, to within a tiny variance.

When we consider radioactive decay (, we cannot predict when a specific atom will decay, but we CAN predict, with an extremely high degree of confidence, the half-life of a given substance. This is due to the vast number of particles involved, and the fact that the average result from large numbers of random trials tend to approach the expected value over time. (

But what about when a given person reads a blog post on how complicated humans are?

How will the person react? Amusement? Interest? Annoyance? Some combination of several different reactions? Can we even list all of the likely reactions?

What are the contributing factors to the reaction, and what is their significance? Is the person tired? What time of day are they reading? Are they cold? Hungry? Is English their first language? Are they young? Old?

And, even assuming we can identify all of the variables, can we measure them? How? Currently, most behavioural and psychological research measurements are based on things like brain activity (eg, MRI), physiological reactions (such as pupil dilation), or subjective information from things like surveys or interviews. How precisely can we measure the speed of a train by asking the passengers?

In the end, though we have learned a vast amount about human behaviour, we are just barely scratching the surface, and any conclusions we make are tentative and subject to vast amounts of additional validation. We can make predictions, and some of them turn out reasonably well, but the variability is still very high, and cannot be compared to the precision of predictions made in hard sciences like physics.

Statistically, things are a bit better, but not much. If we are dealing with large numbers of people (hundreds, thousands, millions), we can estimate probabilities with (sometimes) reasonably high confidence, but that confidence will have very large “error bars”, and cannot even begin to approach the precision or confidence possible with physical processes. “99%” confidence is a distant dream for psychological research, let alone the many additional “nines” of precision that we have for many processes in the hard sciences.

In our current environment of polarization, this is actually quite encouraging. On all sides, we hear that X is going to happen, or Y is going to happen, and that it is inevitable, but that confidence is an illusion (or, sometimes, simply a lie). The danger is that treating predictions like that as inevitable makes a person less likely to look for other ways to solve our problems – they are out there, but not if we don’t look for them.



bottom of page