What is the difference between theoretical science and experimental science

I was talking to a chemist today and he says he is 100% theoretical, not experimental. However his work involves computer models and computer simulation, which are based on an understanding of the actual physical properties of atoms. To me this seems like a speeded up version of physically mixing chemicals, so why is this considered theoretical and not experimental?

What is theoretical science then, is it science where you use computers and mathematics to prove and investiage theories instead of observation and experiments? Wouldn’t using computers be a form of experimental science since the code is written to match real world situations?

The thing is, a simulation on a computer, is still a model, it is theoretical. It may be based on the laws of physics, and have constraints from observations, but it is not “real”. An experimentalist looks at the real world, and uses the models, i.e. the theory, to help interpret what is going on. Experimental results are used to verfy/disprove the models.


So are the proofs used in theoretical science computer modeling and mathematics while the proofs in experimental science are physical experiments? What other methods of proving are there in both theoretical and experimental science?

He said he was working in collaboration with experimental chemists, so I assume he makes the models on his computer and the experimentalist interprets them in real world situations.

No. Theoretical science can prove that something is not contrary to the laws of nature as we understand them at this point. This does not imply that just because the simulation says it can happen, that it will happen. The theory also needs to include results of observations. There are no solid “proofs” per say in theoretical science, apart from the model/theory agreeing with the real data, i.e. experiments.

What do you mean? How do you prove your pet theory/model (if you’re a theorist)? Simple – look at what the experiments say. If they do everything your model says the experiments should, chances are you’re onto a winner, if not, sorry, try again. For experimental science, you get a set of observations, and see if the results of the observations match your favorite theory.

However, there is no absolute “proof”. A couple of hundred years ago, everyone thought that all mechanics/dynamics could be considered using Newtonian methods alone; now we know there are regimes in which this is untrue. Another thing your theory has to do is not only predict something new, but match all the other observational results out there, i.e. your theory cannot be inconsistent with what we already know.

Not quite. Its probably an iterative process. He produces a physical model. The experimentalists produce results, which he tries to reconcile with his models, making alterations as neccessary. The experimentalists probably produce more results, and everything starts again.

Continuing from Angua’s explanation, one can further say that some scientific work is more observational than theoretical or experimental. Although all experimentation implies observation (or else what’s the point?), there are some circumstances in which one can merely observe and not control the variables. An example would be astronomy in its most common form (I’m not trying to step on Angua the Astronomer’s toes here!).

As a case study, consider the identification of helium in our own sun. 19th century astronomers such as Joseph Norman Lockyer and Pierre Janssen noted (via observation) that the sun’s emission spectrum contained atomic lines that could not be accounted for by any known terrestrial element. They postulated that the sun contained a new element, which they called “helium” after the Greek word for “sun”. Decades later, laboratory experiments were carried out on the gas that we now know as helium, and the atomic emission lines were found to match those in the previously-observed solar spectrum. Bingo! The combination of observation and experiment identified the gas in the sun as the same one as found (in much smaller quantities) on earth. Later theoretical advances (such as quantum theory) enabled scientists to identify exactly which transitions in the helium atom were responsible for the various observed spectral lines.

Think of observation, theory, and experiment as the tripod that holds up scientific research. Without any one of them, the endeavor would fail. As Angua said, without experimental results, the theoretician has no way of verifying his/her computer models. There’s a lot of overlap between disciplines, however, and although one can find people who claim to be “100% theoreticians” it’s unusual to find “100% observers” or “100% experimentalists”. Even “lab rats” (of the human variety, that is) run simulations from time to time.

For a current astronomical mission in which the experimental, observational, and theoretical aspects are easy to identify, take a look at NASA’s “Deep Impact”. This is a probe (launched January 12, 2005) that will make rendezvous with comet Tempel 1 on July 4 of this year (it’s a US mission, and the date is no coincidence…). It’s not just going to observe the comet, however; the probe will launch an “impactor” that will smash into the comet. As it approaches the comet, the impactor’s instruments will relay images back to the “flyby spacecraft”. After impact, the resulting ejected matter will be observed by the flyby spacecraft, and the results sent back to Earth, where the data will be analyzed. This will lead theoreticians to postulate new models of what is occuring in the comet (and what caused its formation), and experimentalists will devise new laboratory experiments to test these hypotheses. The experimental results will refine the hypotheses, and the dance will continue until a next-generation probe is sent to a comet at some time in the future.

[Personal note: a colleague of mine is part of the Deep Impact scientific team, and will be in Hawai’i on July 4 to observe the mission’s climax. I will be involved in the laboratory experiments that will be conducted in the months following the “Impact”. It’s fun stuff!]

Old joke: In theory, there’s no difference between theory and practice. In practice, there is.