Is the scientific method in physics substantially different?

[yolo420wildebeast]

Hi,

I have two degrees. A BA in physics and a master's in computer science. I've taken more than 30 hours in upper level biology courses and a decent number of hours in mid level chemistry. One thing that I can attest to is that all of the disciplines have a different feel to their respective learners. I wanted to start a discussion on the uniqueness of physics and especially modern physics because the only thing that a researcher would understand that an engineer like myself would not understand is the appropriate application of the scientific method. The physics of research, in other words.

So how do the physics of research differ in the disciplines? Though they are almost certainly the same on some level, I can form an argument supporting the idea that there are differences that must be acknowledged, respected, and even internalized so that we can have success.

If I may, I would like to ask that we pay particular attention to that last word, "internalized". If physics demands something different in return for success, what is it and how do we internalize that difference? What are the patterns (of thought and behavior and inquiry) we follow to find success?

Hypothetical answers welcome. Multi-disciplinary answerers more than welcome.

Peace be with you,
WildBeast

 

[Dr. Courtney]

Most physics emphasizes more quantitative comparisons than other sciences. A model with a 2% error in its prediction may be abandoned if the experimental uncertainty is less than that and a more accurate model is proposed. 2% error does not usually cause much concern in other disciplines, especially biology, but also some areas of chemistry and other physical sciences (geology, climatology, etc.)

At the same time, I have found that thinking like a biologist has been very useful for me in physics. I was able to spot category and classification errors other physicists had missed while so strongly focused on reducing the errors and uncertainties in the quantitative predictions and measurements.

Then after a couple decades in physics, I was able to return to biology and use the quantitative thinking of a physicist to improve the state of the science on several problems where progress was stagnant due to an overabundance of qualitative/category thinkers and not enough quantitative analysis.

My bottom line is that the scientific method should be invariant across disciplines: predicting the outcome of repeatable experiments is the gold standard and ultimate arbiter between competing models. This tends to be more quantitative in physics due to the nature of the models and the training of the contributors. But I would not make the mistake of confusing a trend or a tendency with a fundamental difference in the epistemology of the method itself. Science cannot be reduced to what scientists do.

 

[TomServo]

At the same time, I have found that thinking like a biologist has been very useful for me in physics. I was able to spot category and classification errors other physicists had missed while so strongly focused on reducing the errors and uncertainties in the quantitative predictions and measurements.

Then after a couple decades in physics, I was able to return to biology and use the quantitative thinking of a physicist to improve the state of the science on several problems where progress was stagnant due to an overabundance of qualitative/category thinkers and not enough quantitative analysis.
.

I'm not quite following, do you think you can provide examples of these two paragraphs?

 

[Dr. Courtney]

I'm not quite following, do you think you can provide examples of these two paragraphs?

At the same time, I have found that thinking like a biologist has been very useful for me in physics. I was able to spot category and classification errors other physicists had missed while so strongly focused on reducing the errors and uncertainties in the quantitative predictions and measurements.

Sure. When I arrived at MIT, the group I was joining (with Dan Kleppner) had just published a paper showing experimentally that diamagnetic lithium had Poisson-type energy level statistics. However, since the odd-parity lithium has a small quantum defect (0.05) and is regarded as a hydrogenic atom, they assumed that the corresponding classical dynamics of the system was regular (not chaotic). They had a model that showed great agreement with their experiment, and the Poisson-type statistics corresponded well with what was expected of quantum systems with regular (not chaotic) classical analogues. (Diamagnetic hydrogen does not become chaotic until much higher energies/magnetic fields than their experiment). See: http://journals.aps.org/prl/abstract/10.1103/PhysRevLett.62.893

It took me a while, but I eventually realized that the earlier paper had made a category error: the system under study in that paper (diamagnetic lithium at those energies and field strengths) actually demonstrated strong classical chaos. Consequently, the earlier paper purporting to support the correspondence of Poisson energy level statistics with regular motion had actually disproven it by demonstrating Poisson energy level statistics with chaotic motion:

We present a theoretical study of the connections between quantum and classical descriptions of lithium in a magnetic field. We find that the localized nature of the ionic core causes a breakdown in the generic connections between energy-level statistics and classical motion: classical chaos is observed in a regime where the energy-level distribution is Poissonian. The breakdown arises because the source of chaos is localized on a node of the odd-parity wave function. In addition, we employ the principles of periodic-orbit theory to identify classical trajectories with spectral periodicities. We identify a series of core-scattered recurrences in the Fourier transform of the spectrum and quantitatively describe them by a simple model. Finally, we introduce recurrence counting and demonstrate how it can help relate the classical dynamics to the quantum spectrum. © 1996 The American Physical Society.

See: http://journals.aps.org/pra/abstract/10.1103/PhysRevA.53.178

There were a few other discovers I was a part of in those days that were more about finding a recurrence peak in a spectrum that was not predicted at all to be there (according to the early closed orbit theory) than about quantifying the amplitude or frequency of peaks in the spectrum to many significant figures. (We also did confirm frequencies and amplitudes to higher accuracy and resolution than had been done before for some peaks.) But the "Hey, there was not predicted to be a peak there" is more of a category discovery - a category error in the early theory. Of course, the theorists we were working with were very good, and the theory was quickly patched up in each case to account for and predict the frequency and magnitudes of the new peaks.

Then after a couple decades in physics, I was able to return to biology and use the quantitative thinking of a physicist to improve the state of the science on several problems where progress was stagnant due to an overabundance of qualitative/category thinkers and not enough quantitative analysis.

Rather than ask the category question, "Can an explosive blast cause a traumatic brain injury?" my wife and I asked the quantitative question, "How large an explosive blast is required to cause a traumatic brain injury?"

See: https://arxiv.org/ftp/arxiv/papers/1102/1102.1508.pdf
AND: http://www2.technologyreview.com/printer_friendly_blog.aspx?id=26368

Eventually, the broader research community started to think about the question the same way.

See: https://phys.org/news/2017-01-blast-greater-accuracy-brain-injury.html

Another application of our quantitative thinking in biology was discovering widespread errors in the weight-length parameters in fish at FishBase.org.

See: https://arxiv.org/ftp/arxiv/papers/1104/1104.5216.pdf

After considering weight-length relationships in fish for several years, eventually we came up with a more accurate approach:
http://www.macrothink.org/journal/index.php/ast/article/download/5666/4506

Of course, a good quantitative approach is also powerful for addressing category questions like, "Do Rainbow Trout and Their Hybrids Outcompete Cutthroat Trout in a Lentic Ecosystem?"

See: https://arxiv.org/ftp/arxiv/papers/1305/1305.1882.pdf