In 2018 I published a peer-reviewed article titled, “Azimuthal transverse single-spin asymmetries of inclusive jets and charged pions within jets from polarized-proton collisions at √s = 500 GeV,” in the journal, Physical Review D. Physical Review D is the journal of the American Physical Society (APS) covering particles, fields, gravitation, and cosmology. In the field of physics, APS journals are among the highest profile; and, for a long article, such as this, Physical Review carries arguably the highest impact. The article details several first-of-their kind measurements from data collected by STAR. I was the lead analyst of the measurements and the lead author of the paper, drafting the manuscript, myself. According to collaboration bylaws and the publication policy, all collaborators are listed as co-authors in alphabetical order on all STAR papers, including this one. As with most STAR papers, publication of this article was a years-long effort with a grueling peer-review process. For the sake of those interested, in the following sections, I briefly describe the physics content of the article. Feel free to skip this, it won’t hurt my feelings! In the final section, I provide reflection on my experience leading the analysis and the effort that led to its publication.
The Physics
To our best understanding, the building blocks that make up most of what we see around us, namely protons and neutrons, are themselves made up of particles we call “quarks” held together by particles we call “gluons.” While our understanding of these “strong” interactions has improved dramatically over the last 50 years, many wonderful mysteries remain. We understand that every proton exhibits what we call an “intrinsic angular momentum,” in other words, each proton behaves like a spinning top. What that means for the quarks and gluons inside the proton remains less clear. Quarks and gluons also exhibit a “spin” property, but how their spins are arranged inside the proton is still an open question. For example, let us say a proton’s spin points “up” relative to the direction it is moving. It might be natural to ask, “how many of the quark spins now point ‘up’ compared to those that point ‘down’?” The name we have given this particular property is “transversity,” as it describes the “transverse” spin alignment of the quarks. One of the reasons physicists care about questions such as these is that their answers shed light on the laws of nature on their most fundamental level. In the strange world of quantum mechanics, the rules appear very different to what we experience on the every-day, “macroscopic” level, e.g. particle properties can only take on certain values, only certain kinds of arrangements are allowed, etc. Understanding these arrangements gives us new insight into these rules. They also reveal to us pictures of nature at a level beyond what can be seen by human eyes. The proton has a diameter of around 10-15 meters or 40 quadrillionths of an inch. The ability to paint pictures of our world at such a level is truly a wonder!
In the 1990s theoretical physicists began proposing ideas for how we might answer some of the questions posed above, e.g. what is the proton transversity? A physicists named John Collins developed an idea, colloquially known as the “Collins mechanism,” for how we might access it experimentally. If a quark is initially polarized “up” or “down,” it is fairly easy to calculate what its polarization should be after a collision. For years, physicists have observed that when quarks scatter off of each other in highly energetic collisions of protons, those quarks convert into a collimated spray or “jet” of particles that can be measured by our detectors. Collins proposed that the arrangement of particles inside these jets could depend upon the direction of the spin of the quark that created them. For example, if the scattered quark has spin “up” relative to its direction of motion, it may spray more particles to the “left” than to the “right.” Such differences in the collision products, what we call “asymmetries,” can be measured experimentally. If an experiment controls for the spins of the protons and observes asymmetries in the particle production, the Collins mechanism provides a way to connect these asymmetries to transversity.
The first experiments to put this idea into practice were electron-proton scattering experiments in the mid-to-late 2000s. Collins’s idea seemed to work, and by the start of the 2010s the first theoretical models for transversity were published. This was a tremendous stride, but it became obvious rather quickly that the information provided by experiments was dreadfully imprecise. For one, the experiments used electron beams to strike a stationary proton target, which severely limits the available energy. These interactions were also heavily biased toward a single kind of quark, giving us little information on the rest. Furthermore, the detectors of these experiments were not designed to catch the full “jet” of particles produced in the collision, rather only a single particle. While this still enabled asymmetry calculations, it limited what could be gleaned from the collisions. For these reasons, the need for more and diverse experiments became quickly obvious.
STAR turned out to be an ideal detector to join the fray. STAR utilizes colliding beams of protons rather than stationary targets, which dramatically increases the available energy. The collisions at STAR eliminate nearly all the bias toward a single kind of quark. The detector was designed to have very wide and uniform detector coverage so that full jets of particles can be measured. Leveraging these strengths, our analysis was able to show that the same asymmetries found at the electron-proton experiments were also present in the STAR proton-proton collisions, a first-of-its-kind observation. Not only does it provide a path for improved understanding of transversity, it also provides experimental evidence that Collins’s mechanism is “universal,” i.e. it appears to exist independent of the type of interaction.
Since publication, our paper has been cited by nine other studies in the field. STAR has also collected more precise data, and this topic continues to play a major role in upcoming RHIC runs. My colleagues and I are in the process of writing follow-up papers to publish more data, and results continue to generate excitement within the field.
Reflection
Beyond the specific scientific insights, this experience has helped me develop significantly as a scientist and as a teacher. In this section, I will reflect on a few of the lessons I learned and how they have shaped me.
Thinking Creatively
When STAR collected the data published in our paper, no one was thinking about the Collins effect. The data were solely intended to provide a background study for a completely different set of measurements. I began analyzing these data as “practice” for a Collins measurement in a future set of data. The lessons I have learned from this are many. First, I am reminded of the importance of doing my job well, at all times. If STAR had not taken these runs seriously and as a result compromised the data, we would have missed out on a tremendous discovery. Second, I see in this a lesson on the importance of thinking creatively. Had we limited ourselves only to the “official” goals of the run, we would have missed out on a discovery that then became one of the primary “official” goals of at least three future RHIC runs. When I teach a given class, I certainly begin with a goal in mind, a map of how to reach the goal, etc. However, each new class of students is unique with different strengths, needs, interests, etc. This experience reminds me not to be so rigid in my approach that I miss out on opportunities for innovation, e.g. in how I teach, curriculum development, how best to prepare students for future success.
Eureka! vs. Huh?!
This next lesson deserves a bit of a story…
When I began analyzing these data, we expected to see nothing “exciting,” i.e. no noticeable “signal” for the Collins effect. My preliminary analyses showed exactly that–all Collins asymmetries were consistent with “zero.” I presented the preliminary results at multiple international conferences, generating no surprise amongst the external physics community. I decided to publish the results, anyway, as “zero” is still meaningful even if we rarely throw parties over it. Before declaring the analysis “final,” I decided to revisit my data selection criteria. I had preliminarily based them on old analyses that used different detectors at STAR. I optimized the criteria with simulated data, specific to this set of detectors. In the process, I discovered that without sacrificing accuracy, I could include a significant amount of previously excluded data by relaxing one particular criterion. The simulation had no Collins signal built in, so I was able to tune the criterion purely by optimizing precision and accuracy, i.e. “blind” to the Collins effect signal. I reset the criteria in the software, set the computers to process the data, and went home for the night.
When I returned to my office, the next morning, I pulled the data together and ran my plot generator. What I saw, very nearly knocked me off my chair. I was staring at the very first Collins effect signal ever seen in proton-proton collisions. Evidently, the preliminary selection criterion we had all thought reasonable was actually excluding the most “interesting” data! Of course, I didn’t believe the new result, yet. I spent the next week or so trying to destroy the signal. Despite my best efforts, I could not make it go away. In fact, I actually made it larger! After an hour-long phone call with my advisor, we had to admit the simplest explanation was that it was real.
Isaac Asimov is credited with variations of stating that discovery is rarely heralded with “Eureka!” but far more often “That’s funny…” In this case, the saying is certainly apropos. Discovery was not made because we carefully designed an experiment for it and carried it out as planned. On the contrary, we were explicitly not expecting a discovery; we were simply lucky to look where one happened to be waiting. Among the lessons I learn from this is that discovery is never guaranteed. We are not nature’s creator, consequently nature is not at our whims. An experiment may seem perfectly designed yet not reveal what we wished to find. On the other hand, we often find illuminating answers to questions we did not ask. Regardless, what is most important is to ask the question and seek an answer. Furthermore, I am reminded of the importance of honestly evaluating my preconceived notions. It was time-consuming to reevaluate my selection criteria. Since I already had the answer I expected, the temptation to save time and skip right to publication was real. A scientist needs to be able to recognize assumptions they make and be willing to question even the most well-motivated assumptions. I believe this is an important life skill, even for non-scientists. Increasingly our culture is seemingly unable–or unwilling–to distinguish between “truths” and unquestioned assumptions. Sadly, academia often finds itself on the leading edge of suppressing questions rather than encouraging debate. This experience reminds me that a willingness to question myself may not only reveal where I am misguided but will simultaneously strengthen my confidence in what is true.
Mystery
During the internal (STAR) peer-review process for the paper, one of my collaborators had the following suggestion:
“…the underlying mechanism driving these effects remains something of a mystery…”
Rephrase to
“The origin of these asymmetries is not yet fully understood”
This is, after all, a science paper.
I respectfully declined the suggestion, and the paper was ultimately published with my original phrasing. However, the suggestion that science papers should not talk of “mystery” has stuck with me as characteristic of a peculiar mindset within a portion of the physics community. Many, perhaps, most of us–myself included–are irritated when we are stumped by something; and we generally are not satisfied until we arrive at understanding. However, some of my colleagues seem uncomfortable with the idea that mystery plays an important, even irreducible, role in science. Each discovery reveals questions we never knew existed. If we are honest, this is great news! For one thing, the well of experiments will never run dry. At times, we have difficulty admitting that we are not fonts of infinite wisdom and that nature owes no obligation to reveal its secrets. In this respect, it is perhaps easier (in principle, at least) for the Christian to admit and celebrate our station in the wondrous universe we believe God created. I am grateful to work in a community that supports and encourages me to view myself and my role through the lens of faith.