Black-hole mass gap: secondary components, uncertainty and competing models

The black-hole mass-gap result reported in April 2026 depends on which member of a merging pair is being counted. The evidence in the identified study concerns the smaller, secondary component. It is not a universal empty interval in the masses of every observed black hole.

That distinction was largely lost in the earlier article, which converted a particular population analysis into broad claims about what the latest catalogs conclusively establish. A corrected explanation needs both the result and its interpretation, including a later proposed alternative for producing a similar pattern.

The smaller member carries the reported gap

Hui Tong and colleagues' paper, published in Nature and available in an updated May preprint version, analyzed the fourth LIGO–Virgo–KAGRA gravitational-wave transient catalog, GWTC-4. It reports a lower gap boundary of about 44 solar masses, with a 90-percent credible range extending four solar masses lower and five higher.

The authors say the gap appears in secondary masses, m2, while it is absent from the primary-mass distribution, m1. Here primary means the more massive member of the pair, so m2 is less than or equal to m1. Those labels describe present masses, not necessarily which object formed first.

This is why a very massive primary does not automatically contradict a feature reported in the secondary population. Combining both components into one undifferentiated histogram can hide the particular relationship an analysis is trying to describe.

The interpretation involves stellar birth and later growth

The theoretical motivation is pair instability: some very massive stellar cores can undergo processes that prevent them leaving black-hole remnants in an expected mass interval. The relevant expectation concerns formation through that route, rather than a prohibition on a black hole ever reaching those masses by any process.

Tong and colleagues interpret the mass and spin patterns as evidence for a subpopulation of hierarchical mergers, in which a primary black hole is itself a product of an earlier merger. That is their proposed population interpretation, not a direct observation of each component's entire history.

An observed gravitational-wave event constrains properties of a merger. Connecting those properties to a prior stellar life or an earlier merger adds an inference about formation. The inference can be supported by multiple observations without becoming a photographed record of every earlier stage.

A later model offers another way to make a cutoff

In a preprint first submitted in April and revised in September, Aleksandra Olejak and Jakob Stegmann examine whether binary evolution can mimic features attributed to the pair-instability gap. Their population-synthesis and simplified calculations explore how mass transfer can limit the less massive black hole while allowing a heavier companion.

They argue that a similar secondary-mass cutoff can arise without requiring a low pair-instability boundary. That is an alternative model result, not proof that pair instability does not occur. The preprint explicitly leaves pair instability as a possible interpretation.

This corrects the earlier article's claim that apparent gap events necessarily favor one special history over ordinary binary evolution. A feature's existence and the uniqueness of a proposed cause are separate questions. Reproducing a pattern in another model matters even when it does not determine which model best describes every real system.

A credible interval is not a universal physical constant

The reported boundary comes from a named analysis of a named catalog, with uncertainty. It is not equivalent to measuring an immutable threshold at exactly the same mass in every stellar environment. A broad illustrative theoretical range and an inferred population boundary also answer different questions.

The earlier article supplied an essentially settled 50-to-120-solar-mass desert and swept together different catalog claims. The corrected account preserves the specific GWTC-4 result instead of presenting it as an exhaustive summary of every subsequent analysis or observing run.

Likewise, more detections do not mechanically strengthen one interpretation. New data can sharpen a feature, alter it or reveal that a model was too restrictive. The result depends on how measurements, detectability and the population model are combined, not just on the number of events.

The April reporting was a starting point, not a final verdict

The original Ars Technica account emphasized how focusing on smaller components could expose a boundary obscured by previous mergers. That is a concrete methodological idea worth retaining. It does not justify the earlier article's unsupported numerical claims about all first-generation objects or a guaranteed explanation for every outlier.

The useful conclusion is therefore precise: one published analysis reports a secondary-mass gap and interprets it through hierarchical growth and pair-instability physics; a later preprint explores a competing binary-evolution mechanism. Keeping the component labels, catalog, uncertainty and model assumptions explicit makes the research intelligible without declaring the physical history of every heavy black hole settled.