Source-linked AI summary
Fixed Suffix Dependency Ratio: Quantifying the Dual-Track Mechanism of Gender Assignment in Latvian Loanwords
Yelingyun Zhang, Atis Kapenieks
TL;DR
Research on English loanwords finds masculine bias, but pooled analyses can conflate fixed morphological rules with default assignment. This study uses FSDR to examine those mechanisms in Latvian nouns and finds greater suffix reliance among feminine loanwords, with masculine loanwords concentrated in free choice. The pattern is specific to loanwords and grows stronger in contemporary data.
Problem
Pooled loanword gender proportions cannot distinguish masculine bias arising from fixed morphological rules from bias arising through default assignment.
Source: §1 · p. 1Method
The study uses FSDR, the gender-specific conditional probability of carrying a fixed derivational suffix, to compare morphological reliance across genders and word types.
Source: §3.3 · p. 10, §6 · p. 20Results
Loanword FSDR asymmetry is +0.336, with feminine loanwords relying more on fixed suffixes and masculine loanwords more often in unconstrained areas; the asymmetry rose from +0.215 historically to +0.395 contemporarily.
Source: §1 · p. 2, §6 · p. 20Takeaways & Limitations
FSDR makes default gender quantitatively testable, and the Latvian findings identify masculine default as a contact-related pattern in the loanword subsystem.
Source: §1 · p. 2, §6 · p. 20The analysis uses written materials that reflect relatively formal, standardized Latvian, so it does not establish whether assignment patterns differ in spontaneous speech, social media, or code-switching.
Source: §5.3 · p. 19
Abstract
from arXiv · showhide
Existing research has repeatedly observed the tendency for English loanwords to cluster in the masculine gender across different recipient languages, yet the origin of this pattern remains difficult to determine, as fixed morphological rules and default assignments are frequently analysed together. This study proposes the Fixed Suffix Dependency Ratio (FSDR) to quantify the degree of reliance on fixed derivational suffixes across different genders, and to distinguish between morphological anchoring and free-choice in distribution. By examining 1,832 Latvian noun lemma types, the results reveal a significant FSDR asymmetry within the loanword system: feminine loanwords rely significantly more on fixed derivational suffixes, while masculine loanwords are more concentrated in the free-choice zone. This pattern exhibits loanword specificity and has become more pronounced in contemporary usage. FSDR therefore provides a quantitative framework for testing default gender and shows how masculine default can be activated and reinforced under language contact.
1 Introduction
Research on loanword gender often finds a masculine bias, but pooled proportions cannot distinguish fixed morphological assignment from default assignment. This study introduces FSDR to separate these mechanisms and examines their distribution in Latvian loanwords.
1 Introduction
- Pooling loanwords obscures whether masculine bias reflects target-language morphological rules, default assignment, or both.Overall masculine proportions alone cannot establish that masculine functions as the default; separating rule-constrained vocabulary is a necessary first step.Source: §1 · pp. 1–2
- FSDR measures gender-specific reliance on fixed derivational suffixes, distinguishing morphological anchoring from assignment in the unconstrained free-choice zone.The study uses this mechanism-based comparison to investigate whether loanword and native distributions differ and vary with contact conditions.Source: §1 · p. 2
- Latvian offers a useful comparison because its native lexicon is not masculine-skewed: masculine nouns are 44.6% overall and 42.4% in the free-choice zone.Its contact history has shifted toward English, whose nouns provide no source-language gender cues.Source: §1 · p. 2
- Feminine loanwords rely more on fixed derivational suffixes, while masculine loanwords concentrate in the free-choice zone, where their proportion is 65.8% versus 42.4% for native words.The loanword FSDR asymmetry is +0.336 and increased from +0.215 historically to +0.395 contemporarily; sensitivity tests and phonological controls support robustness.Source: §1 · p. 2
- The study presents FSDR as a framework for testing default gender and argues that contact can activate and reinforce masculine default in the loanword subsystem.This claim is grounded in the shift toward masculine forms within the free-choice zone despite the native baseline.Source: §1 · p. 2
2 Theoretical Background
The theoretical background distinguishes rule-governed gender assignment from default choices and motivates using suffixes to separate them. It reviews loanword evidence, Latvian’s gender system, and the shift toward English contact.
2.1 Gender assignment mechanisms: morphological coercion and cognitive defaults
- Gender assignment is layered: semantic and formal rules apply before default assignment, but pooled loanword proportions make rule-based and cognitive-default outcomes difficult to distinguish.Separating morphologically determined words from those reflecting speakers’ choices is needed to attribute observed bias.Source: §2.1 · pp. 3–4
2.2 The role of derivational suffixes in the gender system
- Derivational suffixes can determine gender before open assignment, making fixed-suffix words a diagnostic category for examining remaining bias.The text distinguishes suffix-determined forms from words whose gender must be resolved through agreement or default rules; if bias persists among unconstrained loanwords but not native terms, it is more consistent with an independent default tendency.Source: §2.2 · pp. 4–5
2.3 Loanword gender in contact linguistics
- English loanwords lack source-language gender cues, and masculine assignment is reported across languages; native masculine biases in many recipient languages complicate attribution to contact.Suffix analogy and default can operate as parallel allocation pathways, as prior work also suggests.Source: §2.3 · pp. 5–6
2.4 The Latvian morphological and gender system
- Latvian has two genders and productive derivational suffixes, while its native noun lexicon is 44.6% masculine; this offers a baseline for examining loanword bias.Nouns with derivational suffixes are classified as fixed-suffix, and those without them as free-choice for observing default tendencies.Source: §2.4 · pp. 6–7
2.5 The diachronic context: from isomorphic mapping to gender vacuum
- Latvian contact shifted from gendered German and Russian toward English, whose lack of grammatical gender creates a gender vacuum for loanwords.The text links absent gender cues with greater reliance on the recipient language’s assignment mechanisms and motivates comparing native and loanword patterns over time.Source: §2.5 · pp. 7–8
3. Data and methods
The methods operationalize the distinction between fixed morphological assignment and free choice using a Latvian noun dataset, suffix classification, and FSDR. Statistical tests assess gender patterns, asymmetry, and changes across word types and periods.
3.1 Dataset and annotation
- The dataset contains 1,832 Latvian noun lemma types, including 530 loanwords and 1,302 native words, analyzed by type rather than token frequency.Items came from Wikipedia, educational materials, and historical texts; the sample is divided into Historical and Contemporary periods.Source: §3.1 · p. 8
- Loanword and gender labels were produced with NLP tools and reviewed in a stratified sample; validation found 0.93 Macro-F1 for loanword classification and 100% accuracy for extracted gender features.The source reports a separate held-out loanword-classification F1 of 0.899.Source: §3.1 · p. 9
3.2 Suffix classification and matching procedure
- Fixed-suffix nouns are assigned through morphological or semantic rules, while free-choice nouns lack those suffixes and can reflect phonological, semantic, or cognitive defaults.Suffix matching strips inflectional endings, uses longest-match-first, and applies a minimum root length for single-character suffixes; a reviewed sample and three suffix-definition variants assess classification reliability and sensitivity.Source: §3.2 · pp. 9–10
3.3 The Fixed Suffix Dependency Ratio (FSDR) and statistical framework
- FSDR is the conditional probability that a noun of a given gender carries a fixed derivational suffix; the Latvian asymmetry subtracts masculine FSDR from feminine FSDR.The measure operationalizes differing reliance on morphological rules versus default tendencies across genders.Source: §3.3 · p. 10
- Bootstrap confidence intervals assess asymmetry, while chi-square and binomial tests examine zone differences and free-choice loanword proportions; regression and phonological controls test interactions and alternatives.The analyses compare loanwords with native words and examine whether asymmetry changes by period.Source: §3.3 · pp. 10–11
4. Results
The section applies FSDR to Latvian noun data, comparing gender distributions across fixed-suffix and free-choice zones, then tests diachronic change and robustness.
4.1 Gender distribution and FSDR asymmetry
- Feminine loanwords had higher fixed-suffix dependency than masculine loanwords (FSDR 0.598 vs 0.262; asymmetry +0.336), unlike native words (−0.042), while free-choice loanwords were 65.8% masculine.Native free-choice words were 42.4% masculine.Source: §4.1 · pp. 12–13, Figure 1 · p. 12, Table 1 · p. 12
4.2 Diachronic change in FSDR asymmetry
- Loanword FSDR asymmetry rose from +0.215 historically to +0.395 contemporarily, while native-word asymmetry remained close to zero across both periods.The period increase was approximately 84%; the reported bootstrap 95% CI for the difference was [0.008, 0.357].Source: §4.2 · pp. 13–15, Figure 2 · p. 14, Table 2 · p. 14
4.3 Robustness and control analyses
- The FSDR asymmetry stayed positive under all three suffix definitions, and free-choice masculine proportions remained above 60%.Removing duplicate lexemes across periods left asymmetry unchanged at +0.336.Source: §4.3 · p. 15, Table 3 · p. 15
- Phonological controls found loanword consonant–gender mappings unlike native patterns, despite stem-final consonants still playing a role in gender assignment.Loanword status remained an independent predictor of masculine assignment after consonant control (OR = 2.18, p = 4.92 × 10⁻⁶); masculine entropy was lower in all three samples.Source: Table 4 · p. 15, §4.3 · pp. 16–17
5. Discussion
The discussion presents FSDR as a way to test default gender by separating fixed-suffix assignment from free choice and comparing loanwords with native words. In Latvian, loanwords show a masculine bias in the free-choice zone unlike native words, and the discussion links stronger asymmetry to contact with a genderless source language.
5.1 Quantifying the dual-track mechanism and empirically testing default gender
- FSDR separates fixed-derivation outcomes from free-choice assignments, addressing concerns that default gender can become a residual label.The discussion also compares loanwords with native words as an internal baseline.Source: §5.1 · pp. 17–18
- 65.8% of masculine loanwords in the free-choice zone contrasts with 42.4% of native words, alongside a +0.336 loanword FSDR asymmetry.The discussion treats this divergence from the native baseline as evidence that the masculine bias is specific to the loanword subsystem.Source: §5.1 · p. 18
- When English, which lacks grammatical gender, becomes the main contact source, the discussion says the gender vacuum magnifies the contrast between fixed rules and free choice.The time-series results show increased FSDR asymmetry.Source: §5.1 · p. 18
5.2 Cross-linguistic perspective and the methodological contribution of FSDR
- FSDR comparisons across languages require separating fixed rules, comparing loanwords with native words in the same zone, and accounting for each language’s native baseline.The framework is described as transferable to languages with fixed gender-assignment rules, different suffixes, native lexical bases, and contact histories.Source: §5.2 · p. 19
5.3 Limitations and future research
- The analysis uses formal written Latvian and lemma types, leaving spontaneous speech, code-switching, token frequency, and several free-choice factors for further study.Potential competing factors include lexical domain, syllabic structure, translation equivalents, and terminological standardisation.Source: §5.3 · pp. 19–20
6. Conclusion
The conclusion presents FSDR as a way to distinguish morphological anchoring from default assignment in Latvian loanwords. It finds a loanword-specific masculine concentration in free choice that has intensified in contemporary usage, while proposing broader comparative applications for the framework.
6. Conclusion
- FSDR asymmetry for Latvian loanwords is +0.336, with a confidence interval that excludes zero; feminine loanwords rely more on derived forms while masculine loanwords more often fall outside fixed derivational constraints.The conclusion says loanwords and native words differ in the same free-choice domain, and reports that suffix sensitivity and phonological controls support the finding.Source: §6 · p. 20
- FSDR asymmetry has intensified in contemporary loanwords, with no corresponding change in the native lexicon.The conclusion suggests contact pressure primarily affects loanword assignment as words enter Latvian, rather than the noun gender system as a whole.Source: §6 · p. 20
- FSDR makes default gender quantitatively testable and can be applied to compare loanwords with native baselines across other gendered contact languages.The conclusion names Lithuanian and Polish as possible targets for further application.Source: §6 · p. 20