Pith. Sign in

REVIEW

Rooted staggered fermions: good, bad or ugly?

PaperPDF from arXiv

data files

ReviewThe objections and verdict, as JSON Paper breakdownThe authors' claims, conditions and measurements, as JSON Signed receiptProof this review has not been altered
X Bluesky LinkedIn Reddit Hacker News
Stephen R. Sharpe

High Energy Physics - Lattice · Submitted 15 Oct 2006 · 86 references

Pith's verdict: A fair, well-organized review that lands on 'ugly' rather than 'bad', with the load-bearing assumption (finite-part polynomiality in §8.1) flagged by the author himself.

T0 · AI review

How this review was made: An AI referee (selfhosted/deepseek-v4.1-flash) read the paper on November 11, 2026 and answered a fixed list of questions about it. T0 is the starting level every paper gets. The levels describe how deeply Pith has checked the paper, never the authors.

  1. T0 AI review An AI read the paper and answered a fixed list of questions.
  2. T1 Audit Automatic checks that the references and files the paper relies on exist and match.
  3. T2 Compute The paper's code or calculations re-run with the same results.
  4. T3 Formal statement The claims written out in a formal math language a computer can read.
  5. T4 Proof Those claims proved, and the proof checked by computer.

How Pith checks papers →

· 2 major objections · 6 minor · Ask about this paper · The compiled record · Read the full review →
Challenge this review Comment (0) Read on arXiv ↗
Cite ▾
arXiv IDhep-lat/0610094 (version 1)
Pith IDVT66UBGD, Pith's permanent record of this paper and its review
CitationStephen R. Sharpe. "Rooted staggered fermions: good, bad or ugly?." arXiv:hep-lat/0610094 (2006). https://pith.science/paper/VT66UBGD
Cite this reviewBibTeX for Pith's review of this paper
Claim paper

Why Pith says this

Highlight any text to comment on it.

Referee report

How Pith checks papers → · Full audit of this paper →

The paper

The paper claims

Rooted staggered fermions are non-local and unphysical at nonzero lattice spacing, but the non-locality plausibly vanishes in the continuum limit, so the method is 'ugly' rather than 'bad'.

The reading

Lattice QCD simulates the strong force by discretizing space-time. Staggered fermions are cheap to simulate but come in groups of four identical copies ('tastes'), while real QCD has one up, one down and one strange quark. The MILC collaboration handles this by taking the fourth root of the fermion determinant, one root per quark flavor. That trick has produced a large, widely used set of lattice configurations and many quoted results for quark masses, decay constants and form factors. The worry is whether a fourth root of a determinant still corresponds to a sensible local quantum field theory. Bernard, Golterman and Shamir showed it cannot: no local single-taste theory reproduces the rooted determinant at nonzero lattice spacing. So the rooted theory is non-local away from the continuum limit. This talk reviews the year's work arguing that the non-locality is confined to the cutoff scale and disappears as the lattice spacing goes to zero. Three lines of argument are assembled: perturbation theory (rooting multiplies fermion loops by a non-integer factor, which is renormalizable and matches the desired continuum scheme), Shamir's renormalization-group blocking analysis (the taste-breaking part of the blocked Dirac operator is bounded and shrinks with the lattice spacing), and Bernard's chiral perturbation theory construction (the correct low-energy theory is 'rooted staggered chiral perturbation theory'). The author also rebuts Creutz's objections, argues there is no separate 'valence rooting' problem, and flags a genuinely new problem at nonzero chemical potential, where eigenvalue quartets can straddle the origin and make the phase of the rooted determinant ambiguous.
Extended reading notes

Core claim

Rooted staggered fermions are 'ugly' rather than 'bad': the rooted theory is non-local and unphysical at a not equal to zero, but the non-locality plausibly vanishes as a goes to zero, leaving QCD with one taste per flavor, at the price of fitting to rooted staggered chiral perturbation theory. The author states this explicitly in section 12: 'rooted staggered fermions are unphysical for a not equal to 0, yet the results go over to those of a physical, single-taste theory in the continuum limit.'

Load-bearing premise

Shamir's second renormalization-group assumption, in section 8.3: that the taste-breaking part Delta_n of the blocked Dirac operator is bounded above and vanishes as the fine lattice spacing goes to zero, so that the expansion of det[D_n]^(1/4) in eq. (8.7) is legitimate. The author calls this 'the least well developed part' of the argument and says it is important to check that no surprises arise in perturbation theory for the non-standard RG setup. If Delta_n does not vanish, the second factor in eq. (8.7) survives the continuum limit and the non-locality does not disappear.

Original abstract

I give a status report on the validity of the so-called ``fourth-root trick'', i.e. the procedure of representing the determinant for a single fermion by the fourth root of the staggered fermion determinant. This has been used by the MILC collaboration to create a large ensemble of lattices using which many quantities of physical interest have been and are being calculated. It is also used extensively in studies of QCD thermodynamics. The main question is whether the theory so defined has the correct continuum limit. There has been significant recent progress towards answering this question. After recalling the issue, and putting it into a broader context of results from statistical mechanics, I critically review the new work. I also address the related issue of the impact of treating valence and sea quarks differently in rooted simulations, discuss whether rooted simulations at finite temperature and density are subject to additional concerns, and briefly update results for quark masses using the MILC configurations. An answer to the question in the title is proposed in the summary.

The compiled record

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Compiled by classifier-1.13-20260917 via OpenRouter, question bank abstract-p1-20260923, 2026-09-23: 1 request, 6,147 input tokens, $0.0003 reported.

What the abstract itself states, as probabilities: claims first 0.03 · states limitation 0.18 · human participants 0.02 · controlled comparison 0.04 · randomized assignment 0.01 · states quantitative result 0.06 · compares with prior results 0.39 · mentions code or data availability 0.02.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; .

I give a status report on the validity of the so-called ``fourth-root trick'', i.e. the procedure of representing the determinant for a single fermion by the fourth root of the staggered fermion determinant. This has been used by the MILC collaboration to create a large ensemble of lattices using which many quantities of physical interest have been and are being calculated. It is also used extensively in studies of QCD thermodynamics. The main question is whether the theory so defined has the correct continuum limit. There has been significant recent progress towards answering this question. After recalling the issue, and putting it into a broader context of results from statistical mechanics, I critically review the new work. I also address the related issue of the impact of treating valence and sea quarks differently in rooted simulations, discuss whether rooted simulations at finite temperature and density are subject to additional concerns, and briefly update results for quark masses using the MILC configurations. An answer to the question in the title is proposed in the summary.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#sentencerolethe authors' own workstates the main conclusionstrength
s0 I give a status report on the validity of the so-called ``fourth-root trick'', i.e. the procedure of representing the determinant for a single fermion by the fourth root of the staggered fermion determinant. background 0.58 0.23 0.03 unclear 0.84
s1 This has been used by the MILC collaboration to create a large ensemble of lattices using which many quantities of physical interest have been and are being calculated. background 0.98 0.07 0.02 asserted 1.00
s2 It is also used extensively in studies of QCD thermodynamics. background 1.00 0.07 0.02 asserted 1.00
s3 The main question is whether the theory so defined has the correct continuum limit. background 0.55 0.06 0.03 asserted 0.65
s4 There has been significant recent progress towards answering this question. background 0.90 0.08 0.03 asserted 0.99
s5 After recalling the issue, and putting it into a broader context of results from statistical mechanics, I critically review the new work. contribution 0.77 0.61 0.03 asserted 0.56
s6 I also address the related issue of the impact of treating valence and sea quarks differently in rooted simulations, discuss whether rooted simulations at finite temperature and density are subject to additional concerns, and briefly update results for quark masses using the MILC configurations. contribution 1.00 0.76 0.04 asserted 0.52
s7 An answer to the question in the title is proposed in the summary. result 0.54 0.65 0.57 unclear 0.90

What kind of paper the classifier reports this is: review 0.92 · evidence from literature 0.58.

What was examined, and what was not

sources read
  • arXiv listing abstract v1
not examined
  • full text
  • figures and tables
  • references and citing sentences
  • relations to other papers
levels run
  • P abstract
not run
  • S sections
  • T body sentences
  • C citations
  • R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

Figures
Figure 1. Update on pion mass results using rooted improved staggered fermions. The scale is set with r1. Z-factors account for logarithmic scaling of quark masses. See text for further explanation. mu/md = 0.42(0)(1)(0)(4) [0.43(0)(1)(0)(8)]. (6.3) The new results (still preliminary) include the super-fine and medium-coarse lattices [40], and should be compared to the old results [shown in parentheses] which used only the coarse and fine lattices [2, 3, 42]. The four errors are from (i) statistics, (ii) fitting and other systematics of the simulations, (iii) perturbation theory, and (iv) the…
Figure 2. Results for correlation-length exponent for d = 2 Ising model with long-range interactions. e.g. Ref. [47]. There have also been numerical simulations, e.g. Refs. [48, 49], which, as can be imagined, are challenging for non-local interactions. I do not have time or competence to discuss all this work, and note only one simple point. In the RG analysis, after blocking, the spins are replaced by scalar fields, which have a φ 4 interaction, and a bare propagator whose inverse at long distances has the form: G0(k) −1 ∼ r + jσk σ + j2k 2 +... , (7.2) with r, jσ and j2 calculable…
Figure 3. Results for the strong coupling constant, in the potential scheme, versus scale [52]. the valence correlation functions that are calculated in practice give physical results in the continuum limit? I discuss all of these issues (and more) in the following. 8.2 Numerical results on scaling I begin by briefly showing what is known about the non-perturbative β−function with rooted staggered quarks. This has been studied in Ref. [52] by calculating Wilson loops numerically and comparing to two-loop (rooted) perturbation theory. This allows an extraction of the coupling constant at a…
Figure 4. RG framework and notation. Details are discussed in text. 8.3 Renormalization group analysis In this section I describe recent work of Shamir, who studies rooting using an RG frame￾work [16]. The main conclusion is that, given certain assumptions, the non-locality can be bounded and shown to vanish when a → 0, leaving a theory which is in the correct universality class. (To make this more precise I will need to set up the RG framework.) This is a strong, positive claim for rooted staggered fermions, and, not surprisingly, a key question is the plausibility of the assump￾tions. In…
Figure 5. Condensates in the Schwinger model plotted versus quark mass in dimensionless units [21]. Results for Nf = 1 and 2 are shown, using both (rooted) staggered and overlap fermions. Each result shows a central value and a 1-σ error band. The asterisk shows the known continuum result for the Nf = 1 massless theory. The “(−1) ν curve is explained in the text. term by hand, and set θ = π. This is equivalent to adding a weight term (−1) ν to the determinant. While this might be numerically challenging for QCD (due to the sign problem) it is tractable for the Schwinger model, as shown by the…

Figures reproduced from arXiv:hep-lat/0610094 by the authors.

References
How the literature uses it

58 arXiv papers cite this in 71 sentences

Each citing sentence is taken from the citing paper's LaTeX source and typed by Pith's classifier. A label appears at probability 0.8 or more (0.95 for support), a contest at 0.7 or more; the quoted words are the citing authors'. How each label was checked · JSON

Papers citing it · 44 sentences without a confident label

Uses its method or tool 1 paper
Cites it as support 4 papers
  • Indeed, it does: in the theory with $N_f$ flavors and $n_r$ replicas, the total number of quarks on any closed loop is equal to $4N_fn_r$, which, for $n_r=1/4$ (which corresponds to taking the fourth root of each of the $N_f$ staggered determinants) is precisely equal to $N_f$ [ref] .

    QCD with rooted staggered fermions · Continuum limit: a renormalization-group framework. · 0.98

  • Then contributions from connected and disconnected correlators cancel the divergent behavior; with the correct combinatoric factors [ref] , the cancellation holds even with the rooted determinant of Eq. ([ref]).

    Staggered fermions, zero modes, and flavor-singlet mesons · Introduction · 0.98

  • The issue of the validity of rooted staggered quarks—i.e., whether the known lattice artifacts (nonunitarity and nonlocality at physical scales when $a≠0$ [ref] ) persist as $a→0$—is separate, and there is a large body of analytic and numerical evidence [ref] that these artifacts vanish or decouple in the continuum limit for any strictly positive quark mass(es).

    Reply to: `Comment on: " 't Hooft vertices, partial quenching, and rooted staggered QCD" ' · Introduction · 0.98

Cites it as background 22 papers

Discussion

Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.