Papex

Open problem · Physics · Open

Generations of matter

Why are there exactly three generations of quarks and leptons?

Summary

Quarks and leptons come in three families that are identical in their interactions but differ in mass; why there are exactly three generations is not explained by the Standard Model.

Why it is open

Each generation contains two quarks, a charged lepton and a neutrino; a fourth generation would add one more such set [8]. Only the first generation seems to play a crucial role in the everyday world, and Masahiro Ibe, Alexander Kusenko and Tsutomu Yanagida call the question of why the Standard Model contains three "one of the most important outstanding questions in modern particle physics" [11]. They name two kinds of answer: a fundamental theory "which requires three generations as a consistency condition", or a selection argument of the kind they propose [11].

The problem has a second part: why the generations have such different masses. In the Standard Model the masses come from Yukawa couplings (the strengths with which each fermion couples to the Higgs field), which the model leaves "completely undetermined" [14]. Zhi-zhong Xing notes that the nine charged-fermion masses span nearly six orders of magnitude, a puzzle he calls the flavour hierarchy problem [14]. One striking regularity among the charged-lepton masses has its own page, the Koide formula.

Experiments bear on how many generations there are. They count the light neutrinos through the decays of the Z boson and through the early universe [6, 13, 15], and they limit additional heavy generations through the Higgs boson [8, 9]. No body sets a bar for an explanation. The sources point to what one would need: a theory that requires three generations rather than merely allowing them [5, 10, 11], and evidence for the extra symmetries or dimensions that such theories invoke, which so far "remain hypothetical" [11].

What is known

Why more than two (1973). Makoto Kobayashi and Toshihide Maskawa concluded in 1973 that "no realistic models of CP-violation exist in the quartet scheme", that is with four quarks, without new fields; CP violation is a difference in the behaviour of matter and antimatter [1]. With three families of quarks, Xing explains, the Yukawa matrices cannot all be made real, and weak CP violation "arises naturally" from three-family fermions interacting with both gauge and scalar fields [14].

Counting light neutrinos at the Z (2006–2020). The four LEP experiments recorded 17 million Z decays, and the SLD experiment 600 thousand [6]. Assuming that the only invisible Z decays are to neutrinos coupling as the Standard Model expects, they found Nν = 2.9840 ± 0.0082 light neutrino species, about two standard deviations below three, a result that "fixes the number of fermion families with light neutrinos to the observed three" [6]. In 2020 Patrick Janot and Stanisław Jadach recomputed the Bhabha scattering cross section used to measure LEP's luminosity, which moved the result to Nν = 2.9963 ± 0.0074; "The 20-years-old 2σ tension with the Standard Model is gone," they wrote [13].

The early universe (2018). The cosmic microwave background constrains Neff, the effective number of relativistic species other than photons; any additional light particles produced well before recombination would add to it [15]. Planck's final analysis, combined with baryon acoustic oscillations, gave Neff = 2.99 ± 0.17, consistent with the standard value of 3.046, which is slightly above 3 because the three neutrinos had not completely decoupled at electron–positron annihilation [15].

A fourth generation (2012). Abdelhak Djouadi and Alexander Lenz noted that, given the LEP count, a fourth neutrino that barely mixes with the others must be heavier than about half the Z mass, and that ATLAS and CMS had excluded fourth-generation down-type and up-type quarks below about 600 and 560 GeV [8]. Using the excesses that ATLAS and CMS had reported near 125 GeV in early 2012, they showed that, with the exact next-to-leading-order electroweak corrections, the Higgs signals "unambiguously exclude the perturbative fermionic fourth generation case", if the excesses seen at the LHC and the Tevatron are indeed due to a Higgs boson [8]. Otto Eberhardt and colleagues compared global fits with three and four generations and found the four-generation model excluded at 5.3 standard deviations, or 4.8 without Tevatron data on Higgs decays to bottom quarks [9]. Both results concern a sequential fourth generation with perturbative couplings and one Higgs doublet [8, 9].

String compactifications (1985–2019). In 1985 Philip Candelas, Gary Horowitz, Andrew Strominger and Edward Witten studied ways in which ten-dimensional superstring theories could leave four dimensions with unbroken supersymmetry; "this condition permits only a few possibilities", and one gives a model "with four standard generations of fermions" [2]. In 2006 Volker Braun, Yang-Hui He, Burt Ovrut and Tony Pantev built a compactification of the E8 × E8 heterotic string whose observable sector has "three families of quarks and leptons, each family with a right-handed neutrino", and no exotic matter [7]. In 2019 Andrei Constantin, He and Andre Lukas estimated by extrapolation about 10²³ such string standard models on one large class of Calabi–Yau spaces (the six-dimensional shapes used for the compact dimensions), and about 10⁷²³ on another [12].

Main approaches

Anomaly cancellation in a larger gauge group. In "3-3-1" models the weak SU(2) × U(1) symmetry is enlarged to SU(3) × U(1) [3, 4], and the third quark family is treated differently from the first two [3]. Strongest evidence: Paul Frampton showed that anomaly cancellation, the condition that quantum effects do not spoil the gauge symmetry, then "requires that the number of families be equal to the number of quark colors" [3]. Main difficulty: the model needs new particles, among them exotic quarks with electric charges 5/3 and −4/3 and doubly charged gauge bosons [4], and Ibe and colleagues note that the extra symmetries such explanations rely on "remain hypothetical" [11].

Extra dimensions and strings. If Standard Model fields live in extra dimensions, the consistency of the higher-dimensional theory can restrict the number of generations [5]; see the page on extra dimensions. Strongest evidence: Bogdan Dobrescu and Erich Poppitz proved that with two universal extra dimensions, global anomaly cancellation requires more than one generation, and that if all generations have the same gauge charges and chiralities "there must be three generations" [5]. String compactifications with exactly three families exist [7]. Main difficulty: the extra dimensions remain hypothetical [11]; the first string compactifications gave four generations [2], and the scans that count string standard models impose three chiral families as a constraint rather than deriving it [12].

Discrete symmetries. Flavour models built on discrete groups such as A4 or S4 place the three families in a three-dimensional representation [14]; a stronger claim is that a discrete symmetry fixes the number. Strongest evidence: Jason Evans, Ibe, John Kehayias and Yanagida showed that if the supersymmetric Standard Model embeds minimally in a grand unified theory, an anomaly-free discrete R-symmetry larger than R-parity needs more than two generations, and that in classes of such theories "three generations is uniquely required" [10]. Main difficulty: the result rests on supersymmetry, grand unification and the discrete symmetry [10], which Ibe and colleagues count among the hypothetical ingredients [11].

Anthropic selection. If right-handed neutrinos produce both the matter–antimatter asymmetry and the dark matter, then among vacua with different numbers of generations only those with three or more meet the conditions for observers to exist; the argument makes predictions for X-ray observations and neutrinoless double-beta decay [11]. Strongest evidence: leptogenesis, the origin of the matter–antimatter asymmetry in right-handed neutrino decays, needs at least two right-handed neutrinos; a third can be the dark matter; and anomaly freedom of the gauged B − L symmetry ties the number of right-handed neutrinos to the number of generations [11]. Main difficulty: the argument favours "three or more families", not exactly three, and it assumes both an ensemble of vacua and that the right-handed neutrinos supply the asymmetry and the dark matter [11].

Open sub-questions

  1. Room in the invisible width. With Janot and Jadach's corrected luminosity, how large a Z branching fraction to invisible particles other than three Standard Model neutrinos do the LEP data still allow [6, 13]?
  2. A sharper Neff. What precision on Neff would separate the standard value of 3.046 from exactly 3, compared with Planck's ±0.17 [15]?
  3. Fourth-generation loopholes. Which fourth-generation scenarios outside the perturbative, one-Higgs-doublet case excluded in 2012 remain compatible with Higgs data [8, 9]?
  4. Three among string vacua. In the line-bundle scans behind the 10²³ estimate, how is the number of chiral families distributed when three is not imposed [12]?
  5. Sterile-neutrino dark matter. Which mass and mixing range for the third right-handed neutrino as dark matter, required by the anthropic argument, survives current X-ray searches [11]?
  6. Symmetry and masses together. Can a flavour symmetry that places the three families in a triplet also reproduce the observed charged-lepton mass ratios [14]? See Koide formula.

Work on this in Papex

If you have a testable answer to one of the sub-questions above, you can draft it as a private hypothesis in Papex, or prepare the draft from Claude or ChatGPT with the Papex connector. You review the exact version, sign it and decide whether to publish it; a public record that names this problem is listed on this page. A Papex record shows who signed which version and when. It is not peer review and does not mean the hypothesis is correct.

Key references

  1. Kobayashi, M. & Maskawa, T. (1973). CP-violation in the renormalizable theory of weak interaction. Progress of Theoretical Physics 49(2), 652–657. 10.1143/PTP.49.652 · Checked 16 Oct 2026
  2. Candelas, P., Horowitz, G. T., Strominger, A. & Witten, E. (1985). Vacuum configurations for superstrings. Nuclear Physics B 258, 46–74. 10.1016/0550-3213(85)90602-9 · Checked 16 Oct 2026
  3. Frampton, P. H. (1992). Chiral dilepton model and the flavor question. Physical Review Letters 69(20), 2889–2891. 10.1103/PhysRevLett.69.2889 · Checked 16 Oct 2026
  4. Pisano, F. & Pleitez, V. (1992). SU(3) ⊗ U(1) model for electroweak interactions. Physical Review D 46(1), 410–417. 10.1103/PhysRevD.46.410 · Checked 16 Oct 2026
  5. Dobrescu, B. A. & Poppitz, E. (2001). Number of fermion generations derived from anomaly cancellation. Physical Review Letters 87(3), 031801. 10.1103/PhysRevLett.87.031801 · Checked 16 Oct 2026
  6. ALEPH, DELPHI, L3, OPAL and SLD Collaborations, LEP Electroweak Working Group, and SLD Electroweak and Heavy Flavour Groups (2006). Precision electroweak measurements on the Z resonance. Physics Reports 427(5–6), 257–454. 10.1016/j.physrep.2005.12.006 · Checked 16 Oct 2026
  7. Braun, V., He, Y.-H., Ovrut, B. A. & Pantev, T. (2006). The exact MSSM spectrum from string theory. Journal of High Energy Physics 2006(05), 043. 10.1088/1126-6708/2006/05/043 · Checked 16 Oct 2026
  8. Djouadi, A. & Lenz, A. (2012). Sealing the fate of a fourth generation of fermions. Physics Letters B 715(4–5), 310–314. 10.1016/j.physletb.2012.07.060 · Checked 16 Oct 2026
  9. Eberhardt, O., Herbert, G., Lacker, H., Lenz, A., Menzel, A., Nierste, U. & Wiebusch, M. (2012). Impact of a Higgs boson at a mass of 126 GeV on the standard model with three and four fermion generations. Physical Review Letters 109(24), 241802. 10.1103/PhysRevLett.109.241802 · Checked 16 Oct 2026
  10. Evans, J. L., Ibe, M., Kehayias, J. & Yanagida, T. T. (2012). Nonanomalous discrete R symmetry decrees three generations. Physical Review Letters 109(18), 181801. 10.1103/PhysRevLett.109.181801 · Checked 16 Oct 2026
  11. Ibe, M., Kusenko, A. & Yanagida, T. T. (2016). Why three generations? Physics Letters B 758, 365–369. 10.1016/j.physletb.2016.05.025 · Checked 16 Oct 2026
  12. Constantin, A., He, Y.-H. & Lukas, A. (2019). Counting string theory standard models. Physics Letters B 792, 258–262. 10.1016/j.physletb.2019.03.048 · Checked 16 Oct 2026
  13. Janot, P. & Jadach, S. (2020). Improved Bhabha cross section at LEP and the number of light neutrino species. Physics Letters B 803, 135319. 10.1016/j.physletb.2020.135319 · Checked 16 Oct 2026
  14. Xing, Z.-z. (2020). Flavor structures of charged fermions and massive neutrinos. Physics Reports 854, 1–147. 10.1016/j.physrep.2020.02.001 · Checked 16 Oct 2026
  15. Planck Collaboration: Aghanim, N., Akrami, Y., Ashdown, M., Aumont, J., Baccigalupi, C. et al. (2020). Planck 2018 results. VI. Cosmological parameters. Astronomy & Astrophysics 641, A6. 10.1051/0004-6361/201833910 · Checked 16 Oct 2026

Page reviewed 16 Oct 2026

Source

Open, as listed by Wikipedia · Unsolved problems in physics. Checked 08 Oct 2026.

The status is the source’s. Papex does not decide whether a problem is solved.

Records on this problem

No public record addresses this problem yet.

More in Physics