Two letters in the 2 December 1974 issue of Physical Review Letters point readers towards each other. The Brookhaven paper notes that its authors learned of a similar result at SPEAR after completing their manuscript. The SPEAR paper acknowledges the Brookhaven finding while its own manuscript was being prepared.[2][3]
These were two routes to the same particle, now called J/ψ. Samuel Ting’s team fired protons at beryllium and identified electron–positron pairs among the debris. Burton Richter’s group collided electrons and positrons at Stanford. A very narrow production peak as they varied collision energy revealed an unexpectedly long-lived state.[4]
The Royal Swedish Academy of Sciences announced their shared physics Nobel on 18 October 1976, exactly fifty years before this article’s publication. The independence of their discoveries was central to that recognition. Their acknowledgements record a relationship that matters: parallel discovery can produce reciprocal credit.[4]
Yet those cross-citations confuse a familiar disruption index. In an April 2026 study, Munjung Kim, Sadamori Kojaku and Yong-Yeol Ahn calculated scores of −0.11 and −0.64 for the two papers; deleting their mutual citations counterfactually changed the scores to +0.085 and +0.80.[1]
How one reference changes the answer
The index originates in work on technological change by Russell Funk and Jason Owen-Smith. It asks whether later work cites a focal contribution while leaving its predecessors behind, or continues citing both. The first pattern scores positively; the second scores negatively. Subsequent works citing only the predecessors contribute zero to the numerator while still entering the denominator.[5]
Consider an intentionally small hypothetical network. Paper A cites its simultaneous counterpart B. Ten later papers cite both, with no other relevant papers. Each later paper contributes −1, so A scores −10/10=−1. Remove A’s citation to B and those same ten later papers count as citing A without its predecessors: +10/10=+1.[5]
The experiments and the ten later papers have not changed. Only the relation between the original pair has changed. This calculation illustrates how a measure designed to recognise replacement can mistake a fellow discovery for the intellectual past being replaced.[5]
Finding a shared future
The new method learns two vectors per paper from directed walks through the citation network. One represents its connections to earlier research; the other, its connections to later work. Distance between them estimates redirection. Papers with nearby future vectors have been used similarly by subsequent research, making them candidates for simultaneous discovery.[1]
Among 80 selected, highly cited papers checked manually, 34 had independently simultaneous counterparts and 30 were same-author companion publications. Those categories must remain separate. This is retrospective detection using later citations, not a forecast of breakthroughs.[1]
It can also miss equivalents recognised in disconnected communities, the authors note, citing the Cook–Levin theorem.[1]
The 1974 footnotes are still worth reading literally. Each team was giving readers news of another experiment that had reached the same startling neighbourhood. A reference can record a shared arrival as readily as an inherited idea.[2][3]
Sources & further reading
- Uncovering simultaneous breakthroughs with a robust measure of disruptiveness Science Advances · 1 Apr 2026 · Accessed 18 Oct 2026
- Experimental Observation of a Heavy Particle J Physical Review Letters · 2 Dec 1974 · Accessed 18 Oct 2026
- Discovery of a Narrow Resonance in e+e− Annihilation Physical Review Letters · 2 Dec 1974 · Accessed 18 Oct 2026
- Press release: The 1976 Nobel Prize in Physics NobelPrize.org · 18 Oct 1976 · Accessed 18 Oct 2026
- A Dynamic Network Measure of Technological Change Management Science · 22 Mar 2016 · Accessed 18 Oct 2026