Fixed Suffix Dependency Ratio: Quantifying the Dual-Track Mechanism of Gender Assignment in Latvian Loanwords
Yelingyun Zhang, Atis Kapenieks
cs.CL
Quick verdict
Proposes FSDR metric: Latvian feminine loanwords lean on fixed suffixes (0.598 vs 0.262), masculine default is contact-induced; solid, diachronic part weaker.
Summary
The paper introduces the Fixed Suffix Dependency Ratio (FSDR), defined as P(fixed-suffix | gender), to separate loanword gender assignment driven by fixed derivational suffixes from assignment made in a free-choice zone where defaults may operate. Using 1,832 Latvian noun lemmas (530 loanwords, 1,302 native), annotated via a fine-tuned mBERT loanword classifier and the Stanza pipeline with human audits, the authors find feminine loanwords far more suffix-dependent than masculine ones (FSDR 0.598 vs 0.262; asymmetry +0.336, bootstrap CI [0.254, 0.414]), while native words are near-symmetric (-0.042). In the free-choice zone loanwords are 65.8% masculine versus 42.4% for native words (chi-square = 43.53), supporting a contact-induced masculine default in a language whose native lexicon does not skew masculine. Asymmetry grew from +0.215 (up to 1985) to +0.395 (after 1985), consistent with intensification after English replaced gendered source languages. Robustness checks cover three suffix-definition variants, deduplication, and stem-final phonology. Limitations include a diachronic design confounded with genre (books vs Wikipedia/textbooks), borderline diachronic statistics (z = 1.655, p = 0.098), no multiple-comparison correction, and no animacy or semantic-domain controls.
Review complete
Assessment at a glance
4
risk flags identified
● high 0
● medium 2
● low 2
4 strengths · 4 weaknesses · 5 author questions
AI text signal = Binoculars zero-shot detector score over 20 sampled passages (Fast-DetectGPT co-reported) — absence does not prove human authorship, and a flag does not imply misconduct.
Levels reflect verification complexity — not a quality score. Evolve never rates your paper.
Key claims under review
- FSDR = P(Fixed|Gender) separates suffix-coerced gender assignment from default-driven assignment at the data level, making default gender an empirically testable object rather than a residual label.
- Latvian loanword gender shows a strong dual-track structure: feminine loanwords depend far more on fixed suffixes (FSDR 0.598 vs 0.262; asymmetry +0.336, CI [0.254, 0.414]), a pattern absent in native words (−0.042).
- The masculine default is contact-specific: in the free-choice zone loanwords are 65.8% masculine versus 42.4% of native words, and this cannot be explained by the native baseline or stem-final phonology (OR=2.18 for loanword status).
- The asymmetry intensified from +0.215 (≤1985) to +0.395 (>1985), consistent with masculine default strengthening after English replaced gendered contact languages and created a 'gender vacuum'.
Strengths
A genuinely operational solution to the default-gender attribution problem
FSDR = P(Fixed|Gender) directly operationalizes Kilarski's (1997) 'dustbin category' critique (Sections 2.1, 5.1), separating suffix-coerced from free-choice assignment at the data level instead of arguing it conceptually.
Clean internal baseline that prior cross-linguistic work lacked
Latvian's native lexicon is 55.4% feminine overall and only 42.4% masculine in the free-choice zone (Table 1), so the loanword shift to 65.8% masculine in the same zone cannot be read off a native masculine skew — the observational condition Section 2.3 shows German/Russian/Polish/Spanish lack.
Above-average robustness work for a corpus study
Internally consistent, audited annotation
Weaknesses
Diachronic design confounded with genre
The headline diachronic claim rests on the weakest statistics
No animacy or semantic-domain control in the free-choice zone
Label noise plausibly correlated with the fixed/free boundary
Full Review Report
$1.99 · one-timeUnlock the complete preliminary analysis Evolve produced for this paper.
Report preview
Scope & Claims
The paper targets a real methodological gap in loanword gender research: aggregate masculine rates conflate suffix-coerced assignment with default assignment — precisely Kilarski's (1997) 'dustbin category' critique. Its instrument is FSDR = P(Fixed|Gender), the probability that a noun of gender G carries one of 51 fixed derivational suffixes, plus an Asymmetry index (FSDR_Fem − FSDR_Masc), applied to 1,832 Latvian noun lemmas (530 loanwords, 1,302 native; Section 3.1). Four claims are advanced: (1) feminine loanwords are far more suffix-dependent than masculine ones (asymmetry +0.336, bootstrap CI [0.254, 0.414]); (2) the asymmetry is loanword-specific (native −0.042); (3) the masculine default manifests in the free-choice zone (65.8% vs 42.4% native; χ²=43.53); (4) asymmetry intensified from +0.215 (≤1985) to +0.395 (>1985), attributed to the shift from gendered contact languages to English — the 'gender vacuum' framing of Section 2.5.
The remaining sections cover Methodology Assessment, Statistical Evidence, Results & Reproducibility, Limitations & Risks, and Verdict & Recommendations.
- All 4 strengths & 4 weaknesses, analyzed in detail
- 4 risk flags with severity, section & reviewer questions
- Methods & statistical evidence assessment
- 5 questions a peer reviewer would ask
- Full six-dimension evaluation report & recommendations
One-time purchase · this paper only
One-time purchase, this paper only. AI-generated preliminary analysis — not peer review.
Frequently asked questions
What is the Fixed Suffix Dependency Ratio (FSDR)?
FSDR is the probability that a noun of a given grammatical gender carries a fixed derivational suffix (e.g., Latvian -īb- always yields feminine, -um- always masculine). Comparing FSDR across genders shows which gender's members are assigned by fixed morphological rules versus left to open, free-choice assignment where defaults can operate.
Do English loanwords in Latvian really become masculine?
Partially. Overall loanword gender looks fairly balanced, but that hides a split: loanwords with fixed suffixes are 68.6% feminine, while loanwords without them are 65.8% masculine — versus only 42.4% masculine among comparable native Latvian words. So the masculine tilt concentrates exactly where no suffix rule forces an outcome.
Why is Latvian a good language for testing default gender?
Unlike German, Russian, Polish or Spanish, whose native noun systems already skew masculine, Latvian's native lexicon is majority feminine (55.4%) and only 42.4% masculine in the free-choice zone. A masculine tilt appearing only among loanwords in that same zone therefore points to a contact-induced default rather than a pre-existing native bias.
Is the paper's data and code available?
The annotated Latvian borrowing dataset from Wikipedia (Zhang et al., 2025, Zenodo DOI 10.5281/zenodo.17380900) is publicly available, and the 51-suffix inventory is printed in the paper. However, the paper does not state a release of the full 1,832-lemma dataset (including the textbook and historical-text portions) or of the analysis code.
About this review
This is an AI-generated preliminary review by Evolve.Pub, produced from the openly available text of the paper. It is not peer review and does not represent the authors' views. The canonical record of the paper is at the publisher link above.
- Figures 1 and 2 are referenced but their plotted content cannot be inspected in the text extraction; all headline numbers cited in this review are taken from the extracted Tables 1-4 and body text.
Go deeper on this paper
Wrote a paper? Stress-test it before reviewers do.
Upload your manuscript for the same structured scrutiny — reasoning, statistics, evidence, figures, and reproducibility.