The Nature of Science

Astronomy 10 - Vista College - Spring 2004

Dr. Korpela


As we saw in class, astronomy is the scientific study of the stars (and other objects in the heavens). Since astronomy is a science, we must discuss what science is and how science works.

Science is from Latin scientia meaning knowledge. Today it means a kind of knowledge that makes predictions that can be confirmed by rigorous observation and experiment.

Science isn't just a collection of facts, science is a way of knowing. Some of you might say "I can't do science", but you would never say "I can't do history". Historians look at facts of history, and try to learn about people's lives, find patterns and explanations behind events. Scientists look at observed facts of nature, to try to learn about the rules of the natural world. Naming the stars isn't enough, we want to understand them!

Figuring out the rules that nature follows can be difficult! As an analogy (Thomas Swihert contribution to Great Ideas for Teaching Astronomy), imagine a foreigner watching a game of American football. He or she knows nothing about the game, and wants to learn just by watching it! No rulebook is available, and no one is around to explain the game. For awhile, it'll just be confusion! Later, he or she will begin to see patterns that repeat, indications that there is organization, and evidence that there really are rules for the game. Similarly, astronomers want to learn about the game of reality in the Universe. There is no rulebook or teacher. They must observe Nature, find repeating patterns, indications of organization, and evidence of rules. It's a long, slow process.

So science is a process; the endeavour to understand the natural world through observation and reason. What characteristics should a good scientist have? My list includes: careful observation, curiousity, creativity, always questioning, willing to admit you're wrong, perceptive!

Babies make great scientists! (analogy from The Cosmic Perspective) They have all of these characteristics. They wonder about everything to begin with, and they're learning about the way the world works through observation! They spend a lot of time dropping objects, finding out that they fall. After LOTS of repetition, they decide that it is a rule that objects fall when dropped. But what happens when you give the baby a helium balloon, and she lets go of it? Surprise! It rises! The theory that all objects fall needs to be modified or discarded!

Scientific Method

The process I just described for the baby contains all the elements of science. Science is learning about nature through careful observation and trial-and-error experiments. Every scientific theory and conclusion must be supported by EVIDENCE obtained from experiments or observation (dropping every object that comes to hand!).

Scientists are simply trained to organize this thinking. They have formulated a more formalized description of the way science works, known as the scientific method. This is simply an organized approach to explaining observed facts with a model of nature. Any such model must be testable, and should be discarded or modified if it fails the tests.

The ancient Greeks developed this idea of using a model to explain and predict the behaviour of some aspect of nature. They combined this modelling with logic and mathematics to pursue knowledge in a way very similar to modern science.

In its idealized form, the scientific method contains several steps:

Question Based on Observations

Often the process of figuring out the rules of nature starts with a question, which is itself based on careful observation. "Why is the sky blue?" Curiousity is an important component here, since not everyone spends time looking at facts and wondering "Why?". Careful observation is also essential, since you don't want to base your scientific theory on observations that you later find out were incorrect!

What is an observed fact is not necessarily obvious, however. Implicit assumptions can creep in. It is observed that the Sun rises each morning, and we might say it is the same Sun each day, but the Egyptians believed that the Sun died each sunset, and was reborn each sunrise! Even if we believe it is the same Sun each day, for most of human history people thought it obvious that the Earth was stationary, and that the Sun was moving around it.


Once you have a set of observed facts, and a question, you need to propose a model to explain them. The proposed model often called a "hypothesis", which really means educated guess. Sometimes the model is a picture, and sometimes it is a "story" that we tell, but in both cases we use it to explain the observations. Sometimes scientists must be very creative to come up with a model that correctly explains the facts!

You might have more than one possible explanation! For example, there are two possible ways to explain the Sun's daily motion in the sky. 1) The Sun goes around Earth. 2) The Earth rotates, making it appear that the Sun is moving. Today, we know that (2) is correct. How?

In order to distinguish between competing models, it is necessary that a useful model should make predictions that can be tested through experiments or further observations. How will you ever gain confidence in your model if you can't test it? We'll talk more in a couple of lectures about the predictions of the two models above, and how we determined which was correct.

Experiment/Observation - Comparing to the Model

The predictions of the model are tested through experiments or further observations. Sometimes against observations that have already been made. The new results are compared with the predictions of the model. When a prediction of your model is verified, you gain confidence that the model truly represents nature. When a prediction fails, you must recognize that the model is flawed, and therefore must refine or discard it.

To return to the football analogy, perhaps you had a working model for the rules of the game, and suddenly you see a new play, or penalty, that doesn't fit into your model. Perhaps you can simply slightly change one part of your "rulebook", and resolve the issue. On the other hand, perhaps this one piece of information indicates to you that you've been going down the wrong path completely. You must be willing to admit you were wrong, throw out your "rules" and start over.

Through experiment and observation, you try to gather evidence for your model. You need to "make a case" for your hypothesis. You wouldn't want to convict a murderer based on an "educated guess" without supporting evidence! You don't want to accept a scientific theory or model without evidence, either. In addition, you can't be selective in considering evidence; you have to test your theory against ALL available evidence. You must always question your model against all the observed facts. You can't say "I didn't see that football play that makes my `rulebook' invalid" just because you want to believe your model.

Note that in Astronomy, we can rarely do experiments. We can't, for example, say "I think I know why the Sun shines, and I can prove it if I just make another Sun almost like our own, but with one thing changed!" We're restricted to what we can see out there, what nature has provided for us. This is different from most other sciences, where you frequently have more control over setting up your experiment.

Scientific Theory

You may believe your model, but you have to convince others too! Maybe they can come up with a different (equally valid, or possibly better) explanation. You must share your observations and your conclusions with other scientists!

This means your observations have to be carefully done, repeatable by others, and your explanations have to be logical and precise. Perhaps they will find a new model that more easily explains your observations. Perhaps they will find something wrong with your observations (some effect that you didn't consider), that allows your results to agree with the model. In any case, scientists publish articles in peer-reviewed journals. This means that their work is only published after other scientists have looked it over and agreed that it was done correctly!

A model achieves the status of a "scientific theory" after a broad range of its predictions have been repeatedly verified Theories are NEVER proven true beyond all doubt. They are always subject to testing and challenges.

Sample Application to Astronomy

To study the stars, astronomers gather the required data (brightness, size, mass, etc.). They then build a model of a star, by calculating the properties a star must have according to the rulebook on stars. Finally, they compare the calculated properties with the observed properties of real stars. If the comparison is poor, then they see how they can modify the rules to improve the comparison. When it is good, they gain confidence that their ideas of Laws of Physics pertaining to stars are in good agreement with those that Nature uses to make real ones.

Notes About Science

The scientific method is an idealized description of how science proceeds. It isn't always this clean.

Sometimes your model doesn't make a prediction, but merely explains existing observations more completely than previous models. Or just more simply. In general, scientists believe that if two models both explain the observed facts, then the simplest is the most likely to be correct (unless proven otherwise).

In addition, experiments and observations are subject to errors. Perhaps you made a mistake making the measurement. Perhaps there is an effect you didn't consider that changes the result of the experiment. Keep this in mind when you accept "facts".

Although we'd like to think of science as objective, personal biases and beliefs can creep in. In the past, astronomers truly believed that there was civilization on Mars, and made detailed drawings of their observations of Martian "canals". Biases can also affect your interpretation. You must always question "How do I know this?". The scientific community at any given time contains implicit assumptions and biases that affect how it perceives the results of experiments.

Nevertheless, over many years, the actions of many scientist contribute to our knowledge of the world in a way that does follow the scientific method.

Science is a way of knowing. it isn't the only way of knowing. "Nonscience" is knowledge sought through processes not related to the scientific method. These include (but are not limited to) intuition, traditions, scriptures, story, and faith. Science cannot tell us anything about the validity of knowledge acquired through these processes, since they don't include the element of experimental testing. Scientific knowledge is knowledge acquired through the scientific method.

"Pseudoscience" attempts to search for knowledge in ways that at first seem scientific. But the results are not rigorously tested and verified. For example, at the beginning of every year psychics make predictions about the year to come. An analysis of their success rate tells us that they almost never come true! Many years ago, "Pyramid Power" was a hot topic. The claim was that the pyramidal shape could "focus cosmic forces" on whatever object was placed inside, allowing one to preserve fruit, sharpen razor blades, and so on. While it may be true that fruit placed under a pyramid lasts longer than fruit which is left exposed, simple experiments show that any shape can protect the fruit in the same way!

Concluding thoughts

The scientific view of the world

"What distinguishes science from other forms of learning is its emphasis on predicting and testing (experimenting) as a way of sorting out ideas which sound good, but are false, from ones which present a more accurate model of how the world actually works". (D. Duncan, astronomer)

"We especially need imagination in science" (Maria Mitchell (1818-1889); Astronomer and first woman elected to the American Academy of Arts and Science)

Science can be creative! "How can I test this? What explains that?"

As you read about ANY science, look for the evidence in the form of measurements or observations. EVERY theory or conclusion should have supporting evidence.