A Century After Hubble, JWST Rewrites the Story of Galaxy Shapes

One hundred years ago, Edwin Hubble arranged the galaxies he could see through the telescopes of his era into a simple, elegant diagram: a tuning fork with ellipticals on one side and spirals on the other. That classification scheme, the Hubble sequence, has framed nearly a century of thinking about what galaxies are and how they evolve. But a new meeting report published in Nature Astronomy by Sandro Tacchella of the University of Cambridge argues that the sequence is no longer merely a static filing cabinet for galaxy shapes. In the era of the James Webb Space Telescope (JWST) and the Atacama Large Millimeter/submillimeter Array (ALMA), morphology is being transformed into a quantitative probe of galaxy-formation physics, capable of testing ideas about gas accretion, star formation and stellar feedback at epochs when the universe was only a fraction of its present age.

The core of the argument is observational. Modern facilities are revealing that galaxies in the early universe were not the chaotic, shapeless protogalaxies that many theorists once expected. Instead, a substantial population of galaxies at high redshift appears dynamically mature, with rotating disks, concentrated star-forming regions and structural regularities that resemble, in broad strokes, the Hubble types we see nearby. The report emphasizes that this is a conceptual shift: if ordered structures were already in place when the cosmos was young, then the processes that shape galaxies, from the inflow of fresh gas to the turbulence driven by young stars, must operate quickly and efficiently, and the Hubble sequence itself becomes a fossil record of those processes rather than a late-time curiosity.

To understand why this matters, it helps to recall what the Hubble sequence actually encodes. Elliptical galaxies are triaxial, pressure-supported systems dominated by old stellar populations, while spirals are thin, rotationally supported disks with ongoing star formation organized into arms. Lenticulars occupy the transition zone. For decades, astronomers treated this ordering as an evolutionary sequence of sorts, though Hubble himself was careful to deny that the tuning fork implied a temporal path. Modern astrophysics has replaced that naive picture with a hierarchical one, in which galaxies grow through the accretion of gas and the merging of smaller systems, and morphology reflects the balance between those external drivers and the internal physics of the baryonic gas.

The technical language that has grown around this subject is now central to the field. Astronomers distinguish between the stellar mass surface density profiles of galaxies, their sizes measured at fixed stellar mass, their Sérsic indices, which quantify how centrally concentrated the light is, and their kinematics, the rotation velocities and velocity dispersions that reveal whether a galaxy is disk-like or dispersion-dominated. Each of these quantities can, in principle, be measured for galaxies whose light has traveled for more than ten billion years before reaching our telescopes. The report highlights how combining rest-frame optical imaging from JWST with cold-gas and dust observations from ALMA allows researchers to connect the stellar structure of early galaxies to the fuel reservoirs from which their stars form.

One of the most striking threads in the recent literature, and one that the report weaves through its discussion, concerns the abundance of disk galaxies at high redshift. Kinematic surveys of star-forming galaxies at redshifts between roughly one and three, the epoch when cosmic star formation peaked, showed that many of them are turbulent, thick disks with high gas fractions, supported partly by rotation and partly by random motions. Those studies, led in large part by near-infrared integral field spectroscopy on ground-based telescopes, established that disks existed early but were far more turbulent than their present-day counterparts. JWST has now pushed such structural and kinematic characterization to even higher redshifts, closer to the era of first galaxy assembly, and the emerging picture is one of surprisingly organized systems rather than a universal chaos.

Equally important is the population of compact, dense star-forming galaxies that dominated the high-redshift universe. Studies of their size evolution have shown that galaxies of a given stellar mass were typically much smaller in the past, and that the most massive systems assembled their dense cores first, a phenomenon often described as inside-out growth. Tacchella’s own earlier work contributed to this picture, demonstrating that star formation in massive galaxies appears to self-regulate: periods of intense star formation deplete or expel the gas supply, the galaxy temporarily quenches, and then renewed gas accretion restarts the cycle. Such cycles leave imprints on the structural properties of galaxies, linking morphology directly to the physics of feedback, the process by which supernova explosions, stellar winds and accreting black holes heat or eject the gas from which stars would otherwise form.

The report also situates these observations within the framework of modern simulations. Numerical models of galaxy formation, run on some of the largest supercomputers available, now resolve the interstellar medium of individual galaxies within a cosmological context, allowing theorists to predict how gas accretion from the cosmic web, turbulence driven by stellar feedback, and gravitational instabilities combine to set a galaxy’s size, thickness and spiral structure. Recent work cited in the report examines how such simulations reproduce the morphological diversity of galaxies across cosmic time, and where they still fall short. The comparison between simulated and observed morphologies is no longer a qualitative exercise; quantitative metrics, including non-parametric measures of concentration, asymmetry and clumpiness, as well as machine-learning classifiers trained on labeled images, allow statistical comparisons over thousands of galaxies.

That statistical approach is becoming essential as the data volumes grow. JWST surveys are imaging hundreds of thousands of galaxies in the rest-frame optical, the wavelength range where the older stellar populations and the overall structural skeleton of a galaxy are best traced. ALMA, meanwhile, maps the cold molecular gas and dust that fuel and obscure star formation, revealing the raw material from which the Hubble types are built. Combining the two gives a multiwavelength view in which morphology can be decomposed into its physical ingredients: a rotationally supported stellar disk, a turbulent gas layer, a bulge assembled through early collapse or mergers, and a halo of dark matter that sets the gravitational stage. The report argues that it is precisely this decomposition that turns morphology from a descriptive label into a diagnostic of the underlying baryon cycle.

The implications run in two directions. Looking backward, the presence of dynamically mature disks and massive, compact galaxies at high redshift places strong constraints on how efficiently baryons were assembled into ordered structures in the first billion years, constraining models of early gas accretion and the role of feedback in the youngest galaxies. Looking forward, the report suggests that the classical Hubble sequence should be understood as the late-time outcome of physical processes that were already operating when the universe was young: disks that formed early and survived, bulges that grew through mergers and secular evolution, and quiescent systems whose star formation was shut down as their gas supplies were exhausted or heated. In that reading, the tuning fork is not a museum piece but a summary of galaxy physics, one that JWST and ALMA are now reading out at every epoch of cosmic history.

A century after Hubble drew his famous diagram, the galaxies themselves have become the message. The structures we classify, spirals, ellipticals, lenticulars and their high-redshift progenitors, encode the accretion histories, feedback cycles and dynamical states of the gas and stars within them. As Tacchella’s report makes clear, the coming decade of observations, with deeper JWST imaging, wider ALMA mapping and ever more sophisticated simulations, will turn that encoding into a quantitative test of galaxy-formation theory, ensuring that the Hubble sequence remains not a relic of 1920s astronomy but a living framework, continuously revisited and refined across cosmic time.

Subject of Research: Galaxy morphology and the Hubble sequence as a probe of galaxy formation across cosmic time

Article Title: Beyond Hubble: revisiting the Hubble sequence across cosmic time

Article References: Tacchella, S. (2026). Beyond Hubble: revisiting the Hubble sequence across cosmic time. Nature Astronomy. https://doi.org/10.1038/s41550-026-02989-1

Image Credits: AI Generated

DOI: 10.1038/s41550-026-02989-1

Keywords: galaxy morphology, Hubble sequence, JWST, ALMA, high-redshift galaxies, galaxy disks, star formation, gas accretion, stellar feedback, galaxy evolution, early universe, cosmological simulations