This post is something I have been kicking around in my head for maybe a year.
What is the purpose of science? Think about that for a minute. Why do people do science? What drives them to do it? What do they get out of it? Does it answer any of the same questions as religion?
Is the purpose of science to discover truth?
I think getting clear on the goals of science is the most important place to start when discussing science and religion. If we don’t get a pretty accurate picture of what science is really trying to do, we wont be able to understand how we should use the data we get from science.
Science, first and foremost is a mathematical tool that we use to superimpose impose order on the world, to take things which are not perfect circles and treat them as perfect circles, to take things which are not uniform and treat them as uniform, to ignore entire categories of force, and to get to a balance between simplified assumptions that make the math easier to do, and accuracy of results that are remotely useful.
Take a look at this graph from http://fivethirtyeight.blogs.nytimes.com/ that shows how some real world data stacks up to the math equation version of the same data.
Not a great curve fit, but can still be useful. And that here is the goal. not perfect prediction, but utility.
So in order to achieve this goal of creating useful math equations that come close enough to the real world, science tends to have two driving principles that play out.
First, Science will evaluate everything based on utility. And things that are useful, even if not accurate and known not to be accurate, will still be taught. This can be seen in any number of places. What is the gravitational constant for acceleration? What is the number you used for pi last time you did an equation?
Everything boils down to utility. Is newtonian physics accurate for any other velocities than the ones we see in human experience? Does 3.1415 work well enough for what you are solving?
First, odds are you were not solving a real world problem but more likely a story problem. Something like, “find the density of this orange if it weighs 1.1lbs and has a radius of 2.2in, (express your answer in metric units). And will it float?”
Second, The “real” answer of course varies from orange to orange, would require doing some displacement measures using fluids and precise weights, may vary based on temperature and altitude, and the ultimate answer could most accurately be determined by actually floating the orange in water.
But getting the real answer is a lot of work and only marginally more useful than the over simplified, sea level, room temperature, perfect sphere, uniform density answer.
So the first thing to keep in mind is that science is not looking for truth, it is looking for something that is useful and easier to do that an actual measurement.
The second thing that science is looking for is reductionism. This can be seen in the previous utility answer. But it is a specific criteria for evaluating new theories as well.
If you come up with a new theory of gravity, or quantum mechanics, you will need to do a few things to get your theory (aka math equation) accepted. First you will need to have data that shows your new theory is as accurate as the current models that are out there. Ie. It has the same baseline utility as the current equations. And then as a second hurdle it will need to do something new. Usually there needs to be some new and unpredicted by the current models event, or feature that people can test for and find. Or if it doesn’t predict anything new, it must be simpler.
If your equation, no matter how great, is the same complexity or greater as existing models (equations) and doesn’t give any new utility it doesn’t have a chance of being adopted.
This preference is shown in the principle known as Occum’s Razor:
“simpler explanations are, other things being equal, generally better than more complex ones”
The side effects of this are, that in many circumstances science looks to mask complexity and over simplify the world, treating organic and natural things that are irregular and occasionally chaotic as if they are uniform for the purpose of creating a utility equation.
So does that mean science is useless? No, of course not. But you need to understand how it is created and what its limits are and how to work around them.
The most experienced people at doing this are engineers. They are the people that are trained to bridge the gap from the real world to the math equation world and make the jump back into reality. And I am not just saying this because I was an engineer, but because I— like you— depend on engineers to design the things that I use everyday, like my car, microwave oven, bridges, elevators etc. And I am glad that there are people who navigate this gap the real world and the math version of the world successfully on a daily basis.
So how do engineers deal with the utilitarian simplification of science?
In engineering you start with a real problem, not a word problem. How do I build a bridge from here to there? And that is a specific place. With real specific dirt, and real specific temperature ranges, and rain fall and all of that. The furthest thing from a pristine Pythagorean geometric world you can imagine.
So the first thing most engineers and engineering students will do is draw a simplified version of the problem and decide which things they will ignore. Basically prepare to start translating the world in which they need to find a solution into something that starts to look like an equation. In the case of doing an house you might not need a geological survey of the bedrock, like you would with say a skyscraper. When building an average bicycle you might not need the same level of finite element modeling that you would for a racing version or that you would for a crash test model of a car.
In our bridge example they are going to start with surveys, and get really clear on the grid points for a math model of the place where the bridge is going to go.
But regardless of what you are going to start designing as an engineer you first need to start by making baseline assumptions. Speed ranges you will cover, temperature variations etc. Model a spring and a damper, or just a spring? How much math and computer cycles do you want to spend to get that level of an answer. Then once that is done, you know have a basic math model of the environment and you have an idea of what math models from science you will or won’t be able to use.
So now that you have that you will need to go to the books, internet resources etc. And find a good set of equations for your use. If it is a large project you will have done this before and probably have a team and good process to create a design.
You will take a stab at the design. You sit with the architects, designers, clients etc. And come up with a first draft of the solution that you want to exist in the real world. To solve you problem in your context.
So you and your team will take your solution, convert it to a math model, and combine it with the previous math model for your proposed site. Then you will add in the math models that represent the use and loads you expect to see on the new solution and design.
You then will run the equations and if it works out. It almost never does in phase one. So you make adjustments to you design, and the models of your design and you rerun the equations. And eventually you get to where the equations predict success.
So if the models predict sucess what’s next?
Now knowing that there are simplified utility equations, an engineer will do two things before they actually build this in the real world.
They will add a fudge factor sometimes called a safety margin to the design. For example it the analysis says it will require 4 bolts, they may add a 5th just to be sure. Or add a millimeter here or maybe an inch there.
Almost never will a design that you see make it into the real world be designed to be just as big or as sturdy as math equation world would predict they need to be. Engineers know better that to trust equations and the materials that they are given to be mathematically perfect. That is not the real world.
The other thing that they will do is test the computer results out. Build a scale model and put it in a wind tunnel. Create a prototype test mule for a car to run through a proving ground.
Think about it, would you want to buy a car that had only had crash test done on a computer simulation? No.
Now I know that they are getting better, and more and more accurate all the time. But they are what all science is, and attempt to have utility. The purpose of having an accurate simulation is save car makers money. If they can get a pretty close model of what the actual crash test will do, by running a computer simulation for 8 hours instead of building a car to crash each time, their costs are thousands of dollars per test and not 10’s or 100’s of thousands per test.
But at the end of the day, the real proof no matter how well the simulations look is in crashing an actual car.
So in summary, when you look at science and the discussions in public of what science is doing, keep in mind what the goals of the process are. It is not to measure things as they are. It is to create an artificial mathematical curve fit of the real world that while not perfect, has some utility value for people like engineers. So that when someone has an idea that they want to create and a problem they need to solve, they can do it more quickly that if they were going to use trial and error.
Science does not have a place for, nor does it care about the place where we think and live. It is concerned with the math version of this world. And forget about places that are not in this world. For example the place where the idea for this article lived for the past year.
Think about that.
And let me know if you agree or disagree and why.