cs.LGAug 27, 2026

Terrain signatures in Welsh settlement names

Authors: Oktay Karakuş, Can Eyupoglu

Organizations: Department of Computer Science and Informatics, Cardiff University, Cardiff CF24 4AG, UK · Department of Computer Engineering, Turkish Air Force Academy, National Defence University, Istanbul 34149, Türkiye

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Landscapes are named, but whether names retain measurable environmental information beyond broad geographic structure is rarely tested. We analysed 3,757 Welsh settlements using a frozen, source-audited 24-element lexical framework, preregistered outcome-specific models and geographically structured validation. The central comparison contrasted 101 settlements carrying high-terrain elements (bryn or mynydd) with 139 carrying low-terrain elements (cwm or pant). High-terrain names occupied locations 24.4 m higher relative to their 2-km surroundings (95% CI, 10.8--38.1 m; Holm-adjusted pp = 0.00137). The association remained positive across prespecified 1-, 2- and 5-km neighbourhood definitions and was reproduced using an independently produced elevation source (24.1 m; 95% CI, 10.6--37.6 m). Adding terrain-name polarity to a non-lexical spatial and settlement baseline reduced geographically held-out mean squared error by 4.63%, 6.22% and 7.30% under 10-, 25- and 50-km spatial blocking, respectively, although improvement varied among held-out regions. River-related names provided weaker, directionally consistent evidence, while the preregistered woodland model was non-estimable. Residual spatial structure, unresolved name language and the absence of independent external replication limit interpretation. Selected Welsh settlement-name categories therefore retain measurable information about present-day terrain within Wales, without establishing individual etymology, causal naming, historical environmental memory or transferability to other naming systems.

Figures & tables

Panel A | Effect-estimate robustness
AnalysisNNAdjusted effect (m)95% CI (m)Raw pp
OS Terrain 50, 1 km240+19.2310.57 to 27.901.37×10−51.37\times 10^{-5}
Primary OS Terrain 50, 2 km240+24.4310.77 to 38.080.000456
OS Terrain 50, 5 km240+25.897.56 to 44.220.00563
Copernicus GLO-90, 2 km240+24.0710.59 to 37.560.000466
One record per normalised name197+21.254.54 to 37.970.0127
Table 2: Table 2 | Robustness and geographically held-out prediction of the terrain-name association.
HypothesisLexical contrastOutcomeNN (groups)Adjusted effect (95% CI)Raw ppHolm ppResult
H1afon / nant vs other eligible settlementslog⁡(1+river distance)\log(1+\text{river distance})3,757 (49/3,708)−0.433-0.433 ( −0.778-0.778 to −0.087-0.087 )0.01420.0284Direction agrees with H1
H2bryn / mynydd vs cwm / pant2-km local terrain position240 (101/139)+24.43 m (10.77 to 38.08)0.0004560.00137Direction agrees with H2
H3coed / llwyn vs other eligible settlements1-km woody-cover fraction2,366 (43/2,323)Not estimable—1.000 †Model did not converge
Table 1: Table 1 | Preregistered tests of settlement-name and environmental associations.
Figure 4: Figure 4 | Study design linking Welsh settlement names to environmental evidence. Lexical exposures and environmental outcomes were constructed independently before confirmatory testing. The frozen lexical system defined preregistered water (H1), terrain (H2) and woodland (H3) hypotheses, which were evaluated using outcome-specific adjusted models with normalised-name-group cluster-robust uncertainty and Holm correction across H1–H3. The principal H2 terrain contrast was subsequently challenged through neighbourhood and independent-source measurement robustness, repeated-name and restricted-randomisation analyses, and geographically held-out prediction. The latter compared the non-lexical baseline M0 with M2, which adds H2 terrain-name polarity, using separate 10-, 25- and 50-km five-fold blocking designs with tuning confined to outer-training observations. Copernicus terrain-source robustness and geographically held-out prediction were post-registration analyses. Spatial validation was conducted within the Welsh analytical population and does not constitute external validation.

Explore similar work

CardsList

  1. TEMPLAR Wales: A georeferenced environmental and toponymic dataset of Welsh settlements

    Aug 27, 2026Oktay Karakuş, Can Eyupoglu

    Department of Computer Science and Informatics, Cardiff University, Cardiff CF24 4AG, UK · Department of Computer Engineering, Turkish Air Force Academy, National Defence University, Istanbul 34149, Türkiye

  2. Georeferencing Non-Gazetteered Place Names using Biological Specimen Records

    Aug 7, 2026Aneesha Fernando, Surangika Ranathunga, Kristin Stock +2Geospatial Reasoning

    School of Mathematical and Computational Sciences, Massey University, Auckland, New Zealand · Joint Centre for Disaster Research, Massey University, Wellington, New Zealand · School of Computer Science and Informatics, Cardiff University, United Kingdom

  3. Geolocating News about Extreme Climate Events: A Comparative Analysis of Off-the-Shelf Tools for Toponym Identification in German

    May 5, 2026Brielen Madureira, Mariana Madruga de Brito, Andreas NieklerNamed Entity Recognition

    LeipzigLab - Climate Discourse, Leipzig University, Leipzig, Germany · Helmholtz Centre for Environmental Research - UFZ, Leipzig, Germany · Leipzig University, Leipzig, Germany +1