In the era of today’s large language models, questions about computational methods for aligning content across languages have a new urgency. One area of work focuses on syntactic alignment. In the DH world, the standard-setting project for syntactic annotation is the Universal Dependencies framework, with the goal of “consistent annotation of grammar (parts of speech, morphological features, and syntactic dependencies) across different human languages.”
advocates for some variation of Chomsky’s Universal Grammar are having a moment.
Universal?
syntax vs semantics: continuum…
Further complicated by choices for
Any computational analysis of natural-language syntax must choose a syntactic model suited to its purpose
But lanaguage models that have inferred syntactic models for a language also enable new kinds of study of a content in a single language.
My immediate focus comes from Complutensian Bible project, and desire to align features of the same text in Hebrew, Greek, Latin and Aramaic.
Further, the line between syntax and semantics is not always as crisply drawn as we might prefer. I
https://medium.com/in-medias-res/deponent-verbs-are-not-random-2ba94afade56
transitive vs. intransitive in IE languages. Morphologically passive verbs that are deponent verbs can take direct objects.
Citation
@online{smith2026,
author = {Smith, Neel},
title = {Models for Syntactic Analysis},
date = {2026-10-08},
url = {https://neelsmith.github.io/posts/2026-09-24-universal-syntax/},
langid = {en}
}