🔭 astrophysics

Pegasi Ascendant: Ranking Constellation Genitives on their Aesthetic Merit

Using an active sampling survey of 74 professional astronomers analyzed through a Bayesian framework, this paper ranks the aesthetic merit of the 88 official constellation genitives, finding that "Pegasi" is the most pleasing while "Gruis" is the least, with notable discrepancies between self-reported preferences and actual rankings.

Original authors: Pranav Nagarajan

Published 2026-04-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Pranav Nagarajan

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are at a party where everyone is arguing about which flavor of ice cream is the absolute best. Some say "Vanilla" is classic, others scream "Rocky Road," and a few are weirdly obsessed with "Sardine Sorbet."

Now, replace "ice cream" with constellation names (like Orion or Pegasus), but specifically the fancy Latin versions astronomers use when talking about stars (like Orionis or Pegasi).

This paper is a very serious, scientifically rigorous, yet completely silly study by a Caltech astronomer named Pranav Nagarajan. He wanted to settle the ultimate debate: Which of the 88 official constellation names sounds the most beautiful when you say it out loud?

Here is the breakdown of his "scientific" journey, explained simply:

1. The Problem: A Sky Full of Names

There are 88 constellations in the sky. Astronomers use their Latin "genitive" forms (the possessive version, like "of the Horse" becoming Pegasi) to name stars and planets.

The Question: Which of these 88 names just sounds the coolest?
The Stakes: The author claims that if a name doesn't sound good, it fails to "move the soul," which is a problem for science outreach.

2. The Method: The "Taste Test"

Instead of asking people to just pick their favorite (which is easy to lie about or change your mind on), the author built a digital game.

The Game: He asked 74 professional astronomers (students, postdocs, and professors) to play a "This or That" game.
The Rules: The computer would show two names at a time (e.g., "Is Pegasi better or Centauri?"). The user had to pick the one that sounded nicer.
The Magic: The computer was smart. It didn't just ask random questions. It used a fancy algorithm (called "Active Sampling") to figure out exactly which pair of names would teach it the most about the user's taste. It was like a DJ mixing songs to find the perfect playlist for a crowd.

3. The Results: The Hall of Fame and the Hall of Shame

After collecting all the answers, the computer crunched the numbers and produced a definitive "Tier List."

🏆 The Top 5 (The "Cool Kids"):

Pegasi: The winner! The author thinks this is because winged horses are cool, and there's a famous exoplanet named after it.
Centauri: A strong second, likely because everyone knows Alpha Centauri (our nearest star neighbor).
Andromedae: Sounds melodic and is home to our nearest big galaxy.
Delphini: Because dolphins are universally loved.
Carinae: Sounds like "Carina," a name meaning "beloved," plus it holds a very bright star.

🥀 The Bottom 5 (The "Uncool Kids"):

Gruis: The absolute loser. It sounds like the word "gruesome," even though it means "Crane" (a bird). The author jokes it lost because it sounds like a bad horror movie.
Muscae: The Fly. It's the only constellation named after an insect, and people just don't think flies are pretty.
Horologii: The Clock. It sounds too much like a boring instrument.
Equulei: The Foal. It has too many "u" sounds and doesn't roll off the tongue.
Normae: The Carpenter's Square. The author jokes that people associated it with "normies" (regular people) rather than cool space tools.

4. The Twist: Expectations vs. Reality

Here is the funniest part of the study:

Before the game started, the author asked everyone, "What is your favorite constellation?"
The Result: Most people shouted "Orionis" (Orion). It's the most famous constellation in the Northern Hemisphere.
After the game: When forced to compare it against others, Orionis only came in 7th place.
The Lesson: People think they like Orion because it's famous, but when you actually listen to the sound of the word compared to others, it's not the prettiest. It's like saying you love a famous pop star, but when you hear a new, unknown band, you realize the new band actually has better music.

5. The "Generational Gap"

The author also noticed that different groups of astronomers liked different things:

Students and Postdocs (younger astronomers) mostly agreed with each other.
Faculty (older professors) had very different tastes, but there were so few of them that the results were a bit shaky.
The Theory: Maybe growing up in the field changes your taste, or maybe the hiring process for professors just selects for people who like boring names like Normae.

6. The Conclusion

The author concludes that Pegasi is the most beautiful name in the sky. He admits this is a "April Fools" style paper (even though it's written in 2026), but he insists that the science behind the ranking is real.

The Takeaway:
If you want to sound like a cool astronomer, stop saying "Orionis" and start saying "Pegasi." And whatever you do, don't say "Gruis" unless you want to be the last person at the party.

(Note: The paper was written by a real person using real scientific methods, but the topic is intentionally humorous to celebrate the quirks of language in science.)

Based on the provided draft paper "Pegasi Ascendant: Ranking Constellation Genitives on their Aesthetic Merit," here is a detailed technical summary.

1. Problem Statement

Despite the ubiquity of constellation genitive forms (e.g., Orionis, Centauri) in astronomical nomenclature (Bayer/Flamsteed designations, exoplanet naming, meteor showers), there is no consensus regarding their aesthetic merit. The paper addresses the subjective question: Which of the 88 official IAU constellation genitives sounds the most aesthetically pleasing? The author posits that auditory impact is crucial for science outreach and seeks to establish a definitive "tier list" based on professional opinion.

2. Methodology

The study employed a quantitative, data-driven approach to rank the 88 genitives using a pairwise comparison survey administered to professional astronomers.

Survey Design & Platform:
- A web application was developed using Python (Flask), SQLite, and JavaScript, hosted on PythonAnywhere.
- Participants: 74 professional astronomers (47 graduate students, 17 postdoctoral researchers, 10 faculty members), primarily from North America and Europe, with a focus on stellar astronomers.
- Structure: The survey consisted of a series of head-to-head comparisons. Before the main task, participants optionally reported their pre-existing favorite.
Active Sampling Algorithm (ASAP):
- Initialization: The initial set of comparisons was generated as a uniformly random spanning tree over the 88 nodes (genitives) to ensure every item appeared at least once.
- Dynamic Selection: For subsequent comparisons, the Active Sampling for Pairwise Comparisons (ASAP) algorithm (Mikhailiuk et al. 2020) was utilized.
- Mechanism: The algorithm maintains a global $88 \times 88$ pairwise comparison matrix $M$ . It selects batches of 87 pairs that form a minimum spanning tree of the inverse of information gain, thereby maximizing the expected information gain for the next ranking update.
Statistical Framework:
- Bayesian Inference: A fully Bayesian framework was used to estimate scores. Each genitive's score was modeled as a normal random variable.
- Output: The system computed a mean score and $1\sigma$ uncertainty for each genitive after every response.
- Validation: A leave-one-out analysis was performed to test the robustness of the final ranking against individual outliers.

3. Key Results

The study produced a definitive ranking of all 88 constellation genitives based on the aggregated survey data.

Overall Ranking:
- Top Rank: Pegasi emerged as the most aesthetically pleasing genitive with a score of $1.032 \pm 0.124$ .
- Top 5: Pegasi, Centauri, Andromedae, Delphini, and Carinae.
- Bottom Rank: Gruis received the lowest score ( $-1.002 \pm 0.127$ ).
- Bottom 5: Gruis, Muscae, Horologii, Equulei, and Normae.
Discrepancy Between Self-Report and Comparative Ranking:
- When asked to self-report a favorite before the survey, Orionis was the most popular choice (10 respondents).
- However, in the comparative ranking, Orionis placed 7th ( $0.765 \pm 0.125$ ). This suggests that open-ended preference questions yield different results than forced-choice pairwise comparisons, which better probe relative preferences.
Demographic Analysis (Career Stage):
- Generational Differences: There is tentative evidence of differing aesthetic tastes across career stages.
  - Students/Postdocs: Showed higher agreement with each other. Pegasi performed strongly in these groups.
  - Faculty: Showed lower agreement with students/postdocs. Notably, Indi was rated negatively by students, neutrally by postdocs, and positively by faculty.
- Data Limitations: The faculty subgroup had a low edge density ( $\approx 4\%$ ) compared to students ( $\approx 23\%$ ), resulting in high uncertainties that preclude robust statistical inference for this specific subgroup.
Robustness:
- The final top 10 list remained stable in a leave-one-out analysis, changing by only one position in the worst-case scenario.
- The final pair comparison matrix had an edge density of $\approx 29\%$ , indicating the data was not sparse.

4. Discussion & Interpretation

The author provides qualitative speculation on why certain genitives ranked high or low, citing phonetic appeal, cultural significance, and astrophysical relevance:

High Rankings: Attributed to "cool" imagery (Pegasi/winged horses), proximity to Earth (Centauri/Alpha Centauri), and melodic phonetics (Andromedae).
Low Rankings: Attributed to awkward phonetics (Equulei with double 'u'), association with mundane objects (Normae/carpenter's square), or unappealing subjects (Muscae/fly, Gruis/crane sounding "gruesome").
Machine Learning Failure: An attempt to use VADER sentiment analysis failed, as all genitives were scored as neutral, highlighting the need for human expert evaluation in this domain.

5. Significance & Conclusion

Scientific Contribution: While the topic is inherently subjective and humorous (noted as an "April Fools" paper in the acknowledgments), the work demonstrates a rigorous application of active learning and Bayesian statistics to a non-traditional scientific problem.
Methodological Demonstration: It serves as a case study for using the ASAP algorithm to efficiently rank large sets of items with minimal user effort.
Cultural Insight: The paper highlights a disconnect between immediate recognition (Orionis) and comparative aesthetic preference (Pegasi), suggesting that familiarity does not always equate to aesthetic superiority in scientific nomenclature.
Future Work: The author calls for a larger, more diverse sample (specifically more faculty members) to confirm generational taste differences and improve the statistical power of subgroup analyses.

Note on Context: The paper is dated April 2, 2026, and explicitly acknowledges in the "Acknowledgments" section that it was written as an April Fools' joke, though it utilizes genuine scientific methodology and real data collection.