Are Galaxies “Going Bananas” at Early Cosmic Times?

Avi Loeb
4 min readJan 8, 2024


Examples of prolate (cigar-like) shapes of galaxies in Webb telescope images. (Credit: Pandya et al. 2023)

In a recent paper titled, “Galaxies Going Bananas: Inferring the 3D Geometry of High-Redshift Galaxies with JWST-CEERS”, Viraj Pandya and collaborators argued that the shapes of dwarf galaxies in Webb telescope data show an elongated banana-like fraction that increases from about 25% in the nearby Universe at redshifts 0.5–1 to about 50–80% in the early Universe within the redshift range of 3–8, about a billion years after the Big Bang. The data was taken from the Cosmic Evolution Early Release Science survey, abbreviated as CEERS.

When saw this analysis, it occurred to me that the reported evolution might be the result of a selection bias, because galaxies must exceed a surface brightness threshold to be detectable against the background noise in the sky. As I subsequently showed in a new paper, this favors the detection of edge-on galactic disks for galaxies at low-luminosities and high-redshifts which are near the detection threshold.

In an edge-on configuration, the emitted light from a galaxy would be concentrated into a thin region on the sky, making it easier for the galaxy to exceed the background noise and be detected. The projected surface brightness of a circular galactic disk which is tilted relative to the line-of-sight would be enhanced by the ratio of the long-to-short axes in the projected ellipsoidal image. The term `surface brightness’ represents the observed radiation energy per unit time per unit angular area on the sky.

Cosmology resembles archaeology. The deeper we look into the Universe, the more ancient is the layer that we uncover. The reason is simple. The farther a galaxy is from us, the more time it takes its light to reach us and so the older is the image we see, representing the way the galaxy looked when it was young. Cosmic expansion makes galaxies at great distances appear fainter. Their apparent surface brightness dims inversely with the fourth power of the expansion factor since the light was emitted by them until it was received.

The two primary factors that dictate the intrinsic surface brightness of a galaxy are its mass in stars and its physical size. The sizes of galaxies are observed to increase with increasing stellar mass. At a given stellar mass, they are observed to decline with increasing redshifts because the Universe was denser at earlier times.

Altogether, these scaling relations combine to make the surface brightness of galaxies lower for galaxies at higher redshifts and lower stellar masses. As a result, the emission from early dwarf galaxies is detected only out to a surface-brightness contour that exceeds the detectability threshold relative to the background noise.

For dwarf galaxies near the threshold of detectability, one expects an enhanced occurrence rate of elongated shapes at high-redshifts and low-luminosities. In my paper, I showed that if low-mass galaxies show this bias at redshifts above 5, then galaxies which are ten times more massive would show the same bias only at redshifts above 9. This result could help explain the reported prevalence of prolate shapes of the faintest galaxies detected in the Webb images, whereas no such bias is found among the brightest galaxies.

In conclusion, high-redshift galaxies might not be “going bananas”. The reported statistical prevalence of banana-like shapes in the images of dwarf galaxies at high redshifts might simply reflect a surface-brightness bias and not an intrinsic property of these faint galaxies. When Viraj Pandya saw my paper, he clarified that they studied the completeness of their sample in Appendix B of their paper and concluded that the bias cannot explain their statistical results.

Of course, there might also be astrophysical explanations at play, such as an enhanced fraction of merging galaxies at early cosmic times which would lead to elongated tidal tails, or stretching of some background images by gravitational lensing along the line-of-sight. However, the surface-brightness bias of faint early galaxies must be corrected before any astrophysical conclusions are drawn on the basis of raw images.

When you think correctly about the Universe it makes sense, even though “The Universe is under no obligation to make sense to you” according to Neil deGrasse Tyson. As a young kid I was frustrated by not getting meaningful answers to my questions from the “adults in the room”. And so, I became a practicing scientist so that I would be able to answer the questions myself.


Credit: Chris Michel (October 2023)

Avi Loeb is the head of the Galileo Project, founding director of Harvard University’s — Black Hole Initiative, director of the Institute for Theory and Computation at the Harvard-Smithsonian Center for Astrophysics, and the former chair of the astronomy department at Harvard University (2011–2020). He chairs the advisory board for the Breakthrough Starshot project, and is a former member of the President’s Council of Advisors on Science and Technology and a former chair of the Board on Physics and Astronomy of the National Academies. He is the bestselling author of “Extraterrestrial: The First Sign of Intelligent Life Beyond Earth” and a co-author of the textbook “Life in the Cosmos”, both published in 2021. His new book, titled “Interstellar”, was published in August 2023.



Avi Loeb

Avi Loeb is the Baird Professor of Science and Institute director at Harvard University and the bestselling author of “Extraterrestrial” and "Interstellar".