Andy Rex ’77 explores one
of physics’ great puzzles

Andy Rex ’77 (above) of Fircrest, Wash., co-authored Maxwell’s Demon 2: Entropy, Classical and Quantum Information, Computing with Harvey S. Leff. An updated version of their 1990 book, the new volume—over half of which is completely new—includes progress in the field of physics since the last edition, as well as recent challenges to the second law of thermodynamics. Rex is a physics professor at the University of Puget Sound in Tacoma, Wash. (Photo by Ross Mulhausen)

It is known as Maxwell’s demon, and for the past 130 years it has both lured and perplexed the hundreds of physicists who have contemplated it. Andy Rex ’77 (right) first heard the siren call of this demon early in his career as a physics professor at the University of Puget Sound (UPS) and for the last 20 years has explored the many questions that it raises. In the following interview, conducting via e-mail with IWU Magazine editor Tim Obermiller, Rex explains his interest in one of physics' great puzzles.

The most obvious question, to start with, is: What the devil is Maxwell's demon?

Your question prompts me to say first of all that it has nothing to do with devil or demons in the religious sense. It’s purely a scientific phenomenon.

James Clerk Maxwell (1831-1879) was a Scottish physicist from Edinburgh who was educated at Cambridge and spent a few years on the faculty there before returning to Scotland. He was probably the leading physicist of the 19th century and worked and consulted with other luminaries of the day, especially William Thomson (also known as Lord Kelvin). Maxwell’s greatness stems from the fact that he made contributions in many areas of physics, but two outstanding ones in particular. First, he developed the theory that connected all electric and magnetic phenomena into a single entity. His electromagnetic theory is still the one all physics students study and is the basis for all we understand about electromagnetism.

His second great contribution was in heat and thermodynamics. There, he helped relate the invisible molecular motions to things we can measure, such as heat flow and temperature. One aspect of that theory is the relationship between the speeds of molecules in a gas to the gas’s observed temperature. Maxwell showed that at any temperature, the molecules in a gas would have a distribution of speeds, some faster and some slower, with the average speed related in a fairly direct way to the gas’s temperature. That’s when he got the following idea (published in 1871). Suppose a very small, agile creature (or perhaps some kind of automated valve) could control an opening between two chambers of gas, initially at the same temperature. This creature could open and close a door between the two chambers at just the right moments, to let faster than average molecules go in one direction, and slower than average molecules go in the other direction. After a while, the two chambers would be at different temperatures.

So what? Well, it turns out that this would violate what’s known as the second law of thermodynamics, because the temperature difference created this way could be used to run an engine. The creature/device that separated the molecules in the first place didn’t do any work (in the physical sense), so we would in essence get something for nothing, which is precisely what the second law says you can’t do. You could unplug your refrigerator (or any other electrical device, really) and replace it with one of these. No more energy problems!

But you can’t do this. The second law is basically right—or at least most people still think so. The reasons this demon (a name given by Maxwell’s friend Thomson a few years later) cannot work are complex and subtle and have occupied the thoughts of many physicists for over 130 years. And of course, that’s why we have this 500-page book—there are so many aspects to be considered, some of them related to other parts of physics, including quantum theory, which only came along much after Maxwell was gone. And of course there’s a certain allure to being able to overthrow something as sacred as the second law of thermodynamics. It’s the sort of challenge that can inspire a lot of physicists, and obviously still does. This allure (of a new theory) is probably greater than the inspiration of making some practical device (like a refrigerator) out of a demon. Since it operates on a molecular level, it would still be a big jump to be able to make a real machine like that using a demon, or a collection of them, because the second law still operates on a macroscopic level to create unavoidable losses of energy, for example the friction of moving parts. There’s another description, a little more concise than mine, which I particularly like, located at You might want to check it out now that you’ve read my basic description.

By the way, Maxwell did not intend his demon to be a challenge to the second law, even though it’s clear that the demon can provide that challenge. And herein is one of the more intriguing aspects of the second law. Suppose you have those two chambers of gas, as Maxwell described. Forget about the demon—just throw open the valve for a time, and then close it again. Since the molecular motion (and distribution of speeds) is perfectly random, it’s possible just by random chance that you will create a temperature difference this way. That sounds weird, but it’s because the second law represents only the tendency, or probability, that things will become more disordered in time. Left long enough, things are almost certain to evolve a certain way. For this reason, the second law is often associated with an “arrow of time.” Thermodynamic processes make sense only if they flow in one direction—for example, heat flows from warmer to colder objects, not the other way around. Heat just doesn’t flow from cold to warm, because it’s such a statistical improbability. But it could happen, just once, if you do enough experiments! The second law is unique in this way, as classical physical laws go, in that it represents only a tendency, not something definite. Of course when you get to quantum mechanics, things get more weird than that, but historically this is all well before quantum mechanics.

You've mentioned that this topic has held the fascination of people even beyond the realm of physics, including musicians and writers. Why do you think this subject holds such a wide and sustaining appeal?

This is hard to say, since I have just the physicist’s perspective. I guess it’s a catchy name, for one thing. Maybe it’s the idea of such a great man as Maxwell, being able to put his finger on something like this but never quite be able to explain it, and all the ongoing attempts to explain it in the intervening years. People like a mystery, and this demon has been rather elusive. Anything that can baffle so many great minds has to get some attention.

Some of this also has to do with the idea of entropy, which I somehow avoided in my answer above but should have mentioned. One way of stating the second law of thermodynamics is that entropy always tends to increase. (Entropy is a rather technical thing, but it’s sometimes taken as the measure of disorder or randomness in a system.) People tend to be fascinated by entropy and the idea of things becoming more disordered in time, and so they are intrigued by any effort to reverse this law of entropy increase.

You write that this second edition of your book looks at Maxwell's demon from the standpoint of new theories and information being developed through quantum mechanics and computing. What kinds of new revelations have these approaches yielded to this 130-year-old debate?

On the computing side, there’s a lot of overlap. If you think back to Maxwell’s original idea, the demon had to make some kind of measurement to determine where a particular molecule is and where it’s going, in order to operate the trap door at the right moment. How is that information gathered? It has to be in some physical process, which is where some of the subtlety lies, because in a careful accounting at this stage, you might find that’s where you have to pay for the energy you get out at the end. Then, once you have the information, you have to process it. That’s where the computing side comes in. One of the most important ideas to come out of all this thinking about Maxwell’s demon came from a brilliant physicist named Rolf Landauer (1927-1999), who often stressed: “Information is physical.” That is, the gathering, processing, and discarding of used information are all physical processes, where you have to keep track of physical quantities like energy and entropy. Landauer’s most significant contribution was to show that in certain processes, the demon can operate as imagined at the gathering and processing stages, but it’s in the discarding of information—in effect resetting the processor for the next operation—that the demon fails to beat the second law (and is forced to generate some entropy in order to make up for any entropy reduction elsewhere).

Quantum mechanics comes into play because the whole notion of “information” is different there. Classical states are definite—your binary “bit” of information is a 0 or a 1. But in quantum mechanics there is uncertainty, and some states that are partly one thing and partly the other. This adds a whole new layer of complexity to the computations and bookkeeping of things like energy and entropy.

How realistic is the hope that problems raised by Maxwell's demon will ever be proven experimentally or solved in some other way? What would the implication be for science if it were finally resolved?

There are actually a few people who have proposed new experiments that might resolve many of the issues. Obviously, they haven’t been done yet, or the issue of Maxwell’s demon would be decided by now. Some of these experiments are rather formidable technically, and when you are looking for variations on a molecular scale, it’s always tough to do with macroscopic machinery. What I always tell people is that I will stick to the second law until someone gives me a computer that I can plug into my bath water instead of the electrical outlet. Of course, the second law could be overturned. This happens all the time in science, and it’s one of the more exciting moments when it happens. An example familiar to many people is how classical dynamics was overturned by Einstein’s theory of relativity. As I say, it would be exciting, and this is how whole new fields of study are opened.

Where does your own personal interest in this subject stem from? (An Illinois Wesleyan reference would be nice but not required!)

I have to confess it doesn’t come from Illinois Wesleyan, but it did come about in a way that anyone from Illinois Wesleyan might appreciate. It was very early in my teaching career at Puget Sound, maybe 1983 or 1984. I was teaching a class in modern physics and discussing the Maxwell molecular distribution in thermodynamics. I wanted to mention Maxwell’s demon to my class, because I had heard a little about it and thought they might find it interesting. About an hour before the class I decided to look up why Maxwell’s demon can’t work as imagined, so I could tell my class. Well, an hour wasn’t enough, and here it is almost 20 years later and it’s still an open question!

I did some more research and wrote an article or two. Then in 1986 (I think) I was giving a paper on Maxwell’s demon at a conference, and that’s where I met Harvey Leff, who was also talking about Maxwell’s demon in the same session. We teamed up first to write a resource letter, and then because we were frustrated that the literature on Maxwell’s demon was so diffuse, we decided to write the book (first published in 1990) to make many of the ideas related to the demon available in a single reference.

Have you or Professor Leff had any memorable reactions or feedback to either editions of your book?

The response has been fantastic. We got several positive reviews of our first edition, in good journals such as Science and American Journal of Physics. We’ve also gotten invitations to speak at other conferences. The most memorable was just last summer, at the conference entitled Quantum Limits to the Second Law (at the University of San Diego). There we met quite a few luminaries whose work we have referenced in the past, or with whom we had communicated by mail. Several people we really respect told us how much out first book had meant to their research, and that really meant a lot to us. By the way, I hope you say something about what a great colleague Harvey has been. He’s an outstanding physicist and great person to work with in every respect. I simply could not have done this without him.

Are you working on any other projects that might be of interest to our readers?

I have just started writing a new textbook for the course typically called “College Physics.” The course is a staple of most colleges (including IWU), a full-year introductory sequence in physics intended for all those who want or need physics but who don’t have the calculus required to take “University Physics,” the introductory course for physicists and engineers. The way it works out is that College Physics ends up being the course for most students who intend careers as health professionals of one kind or another—physicians, dentists, veterinarians, physical therapists—plus a few liberal arts students who are just interested in a real physics course.

What I’m trying to do with my book is first to write it in a more compelling and engaging manner than the current books, but second to take better advantage of interactive tools that can be developed as an online companion to the course. This will include course management tools for the professor, automated homework grading, plus an interactive tutorial system. Those three features all exist separately in various commercial services and software packages, but what the publisher and I are trying to do is bring them all together in a single package that will be more convenient for both professors and students.

I am under the impression that Puget Sound was somewhat similar to Illinois Wesleyan in its size and mission. Is that a coincidence, or are you attracted to that kind of smaller, liberal-arts environment as a teacher?

I was absolutely sold on the liberal-arts college model, based on my experiences as a student at Wesleyan. So when I was finishing my graduate degree and looking for a job, I focused on liberal arts colleges. I had nothing against the University of Virginia (where I did my graduate work), or research universities in general, but that type of career wasn’t what I wanted. Nor did I look much at industrial or government research labs, because I enjoyed the teaching I did in graduate school and wanted to give it a try. Now it’s 21 years later, and I’m still here.

Do you think that physics has become regarded as "too hard" a subject in most college students minds? How to generate enthusiasm for a subject like this?

This question may occur to you (and others) because physics enrollments have been down across the country. I don’t think it’s regarded as too hard. Most students at places like IWU and UPS like a challenge and are up to it. My colleagues and I have talked about the enrollment situation—which by the way isn’t a serious problem here, although the numbers are down a little from a decade ago. We aren’t sure why this is happening. It could just be a normal cycle. Another possibility is that students are getting a little more focused on professional tracks, and unless that profession is engineering, this takes you away from physics. One thing we do (to generate enthusiasm) is to give students a range of options. Not everyone will be interested in the two standard introductory courses (one calculus-based, the other algebra-based). So we offer alternatives like astronomy and physics of music and light/color to interest liberal arts students. They flock to these courses. I teach a course in early 20th century physics and culture for our science-in-context program (“Copenhagen to Manhattan”). This course is always team taught with a professor from the humanities, so we look at the interplay between modern physics and the rest of the world, starting around 1900 and going through the development of the atomic bomb in 1945.

Lastly, can you tell me a little more about your life outside the classroom--family, hobbies, etc.?

Following up on your first question above, I really like it here in the Pacific Northwest. When I came, I had no idea what to expect. My job interview at UPS was the first time I had set foot in the state of Washington. But now I would have to think hard before I’d live anywhere else. Sure, it’s rainy and dark much of the year. But in the summer months, when a professor has some flexibility, it’s beautifully warm and dry, just when you folks (and almost everyone else) are sweltering. So I like to get out and play golf a couple of times a week, usually with colleagues from other departments at UPS. I also enjoy running year-round. I’m married to Sharon Rex, a banker, who is a native of this area. She likes to sing and is a member of the Tacoma Symphony Chorus. We love to travel—some of our favorite places are San Francisco, Vancouver, and Orlando. And of course there’s Seattle, a short drive away, where we enjoy a lot of cultural events. We share Seattle Mariner season tickets with a group of friends and so spend many of our summer evenings and afternoons there. We have a 16-year-old daughter, Jesse, who is a junior in high school and (I am proud to say) one of the top high school flute players in the region.