Most stories about galaxy evolution treat supermassive black holes as cosmic regulators that blast away gas and shut down star formation across entire galaxies. A new preprint challenges that tidy picture. Using the Northern Extended Millimeter Array (NOEMA) as part of the Surveying the Whirlpool at Arcseconds with NOEMA (SWAN) project, researchers mapped dense molecular gas in M51—the iconic Whirlpool galaxy—at 180-parsec resolution. They combined these observations with optical integral-field spectroscopy from the VENGA survey to separate gas ionized by the galaxy’s modest Seyfert nucleus from gas ionized by young stars.

The team, led by Mallory Thorp, found that AGN-dominated regions show elevated emission in HCN, HNC, and HCO+ (1–0) lines beyond what dense-gas abundance alone would predict. This points to excitation by X-rays generated when the weak jet slams into surrounding material—a “two-stage” feedback process. Yet the influence stays tightly confined to the nuclear zone. Even the highest emission-line-ratio values, which also track shock tracers like HNCO/CO, suggest any molecular outflow is limited in both radius and impact. If N₂H⁺ is taken as the cleanest cold-dense-gas tracer, the broader interstellar medium of the spiral arms appears largely untouched.

This preprint (submitted April 2026, not yet peer-reviewed) offers something mainstream coverage routinely misses: quantitative spatial limits on feedback in a typical low-luminosity AGN. Popular articles often cite powerful quasar-driven winds seen by JWST or ALMA and imply every black hole does the same heavy lifting. In reality, feedback efficiency depends strongly on accretion rate and jet power. M51’s nucleus is comparatively weak, and the 180-parsec beam still averages over many giant molecular clouds, a limitation the authors acknowledge.

Placing the result alongside two other real studies sharpens the picture. García-Burillo et al. (2019, arXiv:1905.05777) used ALMA to map a more luminous Seyfert, NGC 1068, and found kiloparsec-scale molecular outflows with clear evidence of jet-driven disruption—far more extended than in M51. Meanwhile, theoretical work from the EAGLE simulation suite (Schaye et al. 2015, updated in Crain et al. 2020) relies on tunable “ AGN heating” knobs to match the observed galaxy stellar-mass function; the SWAN constraints imply those knobs may be set too aggressively for galaxies hosting only modest nuclear activity. Veilleux et al. (2020) reviewed multiphase outflows and noted that only about 10–20 % of the energy from low-luminosity AGN couples effectively to the cool gas reservoir—consistent with the localized excitation seen here.

Taken together, these sources paint a more nuanced portrait: AGN feedback is real but scale- and luminosity-dependent. For many spiral galaxies like the Whirlpool, stellar feedback and internal dynamics may still be the dominant governors of star formation. The SWAN results therefore supply concrete observational anchors that galaxy-evolution models have lacked, showing that the question of how black holes and galaxies co-evolve is far from settled.