Feature-Geometric Harmonic Serialism and the RUKI Sound Law
Bibliographic record
Abstract
The RUKI Sound Law, wherein the reflex of Indo-European *s comes down as [ʂ] in Sanskrit after [r u k i], has remained an open question in linguistics for over a century. The problem with this change is that [r u k i] do not form a natural class: there is no way to define them as a set according to articulatory features which both include all of them, and exclude every other segment. This project aims to provide a solid answer to the RUKI problem using current analytical techniques. Specifically, it proposes a constraint-based analysis using a modified version of Harmonic Serialism (McCarthy 2016) which has integrated the Clements & Hume (1995) version of feature geometry. It expands on the interpretation that feature changes constitute a single operation in HS GEN. We must explicitly attach a feature theory to the model so that we know which features are available in the derivation. This dissertation proposes Feature-Geometric Harmonic Serialism (FGHS) to address this issue. This dissertation also addresses concerns about constraint-based diachronic phonology, examining and challenging many of the criticisms levied against it. While FGHS is not itself capable of solving all the concerns, it has the potential to develop into a model which does solve the problems. The formal analysis of the RUKI Sound Law argues that it is a telescoping of at least two changes: the first was a general place-assimilation of *s to the place if the preceding segment, with the assimilated allophones later being phonemicized into the singular [ʂ]. This analysis is backed by a statistical analysis of a large sample of Sanskrit tokens of [s ʂ] which show that the RUKI set should not be defined as a natural class, but is instead as an “accidental class” shaped by distributional patterns rather than featural characteristics. *s only appeared after a subset of segments, and The RUKI Set were the only ones which would be able to assimilate /s/ to anything other than its organic dental place. Through this analysis, this dissertation shows that derivational-serial constraint-based models are entirely viable for historical research.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.009 |
| Scholarly communication | 0.002 | 0.005 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.007 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".