top of page

Physics: Theory or Practice?

Writer's picture: Mishkat BhattacharyaMishkat Bhattacharya

This post is about a choice people in physics often have to make: whether to pick theory or experiment for a career. Some comments based on my experience of 25 years (but none of it in any administrative positions):


i) Natural Selection: If you have a gift/temperament for one but not the other, then the choice might be automatically made for you. I have heard it phrased as follows: "Theory is for you if you prefer spending an afternoon chasing down a minus sign in your calculations; experiment is for you if you would rather spend that time detecting a pressure leak in your apparatus."


If you are equally gifted/interested in both, some areas of physics may be able to accommodate you. Certainly a number of people in Atomic, Molecular and Optical physics do both theory and experiment, and for me that was one of the initial attractions (before I found out I was better suited to theory).


In Condensed Matter physics this might not be so easy. I know of physics departments where the CM theorists sit on the top floor while the experimentalists are all in the basement (they haven't been banished; the experiments are just more stable there).


In High Energy physics and Astronomy, it seems quite difficult to practice both theory and experiment at the same time. The respective training and techniques are rather specialized and combining them in one career seems quite impracticable.


ii) Collaborative/social aspects: Theorists have, and often exercise, the option of working by themselves. I have written some single-author papers. However, theorists also love to talk shop and nowadays many of them exploit fully the possibilities of establishing nonlocal collaborations, since computers and the internet make the exchange of information between virtually any two locations on the planet quite easy. I have found this useful not only in maintaining ongoing collaborations but using students and ex-postdocs as resources for onboarding/training new recruits to my group.


I have found fewer experimentalists who work by themselves; in academia, the professor usually has a group of students; in the research institutes there are postdocs and other technical people; similar situations apply in industrial research. In the universities the experimental groups are typically larger than theoretical groups. Experimentalists are typically the engines of the department: they occupy more real estate, they bring in more money, they train more students and postdocs. Of course, exceptions are always there.


Experimental collaborations occupy a spectrum. If big science is involved (particle accelerators, telescopes, nuclear reactors, plasma confinement) then there are typically one or more locations (e.g. FermiLab, CERN, BICEP) at which collaborating groups converge to work.


For groups doing table-top experiments, the collaborations often emerge out of the division of labor: one group does the fabrication, perhaps, another the experiment, the third the assay. In these cases, students and postdocs who have graduated out are not typically available to hands-on train new group recruits, but this can be handled by timing hiring appropriately.


iii) Flexibility in research directions: In theory it is often relatively easy to switch research topics. Sometimes, of course, this requires a major investment in learning new theoretical techniques; still, the investment is not as expensive - in terms of time and money - as it would be if a similar change were to be made by an experimentalist. In my own career I have added on relatively new directions four or five times over a quarter of a century .


On the other hand, acquiring new equipment and going through the learning curve for implementing a new platform is a nontrivial task for an experimentalist and few of them do this on a regular basis.


iii) Reception in the community: As a practicing theorist I have found that my work is received much more credibility when it is supported by experimental data. If I say at the beginning of my talk that our theory is borne out by such-and-such experiment, then the degree of skepticism from the audience is typically very low, if not absent.


On the other hand there are all kinds of objections about any standalone theory that I might present: how do I know this approximation is correct? How do I know that series converges? How do I know my starting assumptions are reasonable? Am I sure what I am proposing can be implemented experimentally?


I am not complaining too much; all this makes for a lively discussion and I enjoy the attention our work gets as a result of it. Still, the difference experimental evidence makes is interesting and noticeable.


In contrast, a well done experiment has the stamp of finality - see Kapitsa's well known quote on the topic - while a theory is 'just a theory'. So I think on average experimentalists enjoy a better reception from the community (of course there are exceptions), while the theorists enjoy an occasional high when their predictions are confirmed (or they present a tight theoretical achievement such as exactly solving a model for the first time).


Conclusion


Both theory and experiment have aspects that are enjoyable and aspects that are challenging. I personally enjoy doing theory very much; there are few things that get me going like a new idea. Since I have a PhD in experiment, I also find it natural to talk to experimentalists. I dare say they like talking to me as well -:).



Recent Posts

See All

Comments


Responsible comments are welcome at mb6154@gmail.com. All material is under copyright ©.

© 2023 by Stories from Science. Powered and secured by Wix

bottom of page