The Oxford Latin Syntax

Volume II of the Oxford Latin Syntax deals with the syntax and pragmatics of complex sentences in Latin and other topics that transcend the simple clause (which is the content of Volume I). The volume starts with a chapter on subordination in general, followed by chapters on subordinate clauses that function as argument or as satellite in their sentence. Separate chapters are devoted to subordinate clauses governed by nouns and adjectives and to relative clauses. In addition there are chapters on coordination, comparison, secondary predicates, information structure of clauses and sentences including the use of emphatic particles, word order, and various discourse phenomena such as sentence connection. As in Volume I, the description of the Latin material is based upon texts from roughly 200 BC to AD 450. The Latin texts that are discussed are provided with an English translation. Supplements contain further examples to illustrate the main text. The grammatical framework used is mainly that of Functional Grammar but the description is accessible for readers without a modern linguistic background.

Grammar Engineering for the Ekegusii Language in Grammatical Framework

The knowledge-driven economy uses technology, thereby increasing the demand for language tools and resources to acquire and distribute the knowledge. Such tools and resources are scarce for the under resourced, spoken Bantu languages. This paper develops a computational grammar for the Ekegusii language in the Grammatical Framework (GF) to bridge the gap. The grammar development uses a bottom-up and modular-driven approach. A machine translation experiment was set up to evaluate the grammar resulting in BLEU and PER of 55.95% and 19.49%, respectively. This work contributes by providing computational grammar for an under-resourced language, thus providing a platform for analysis and synthesis, plus a machine translation within the GF ecosystem.

Abstract Syntax as Interlingua: Scaling Up the Grammatical Framework from Controlled Languages to Robust Pipelines

syntax is an interlingual representation used in compilers. Grammatical Framework (GF) applies the abstract syntax idea to natural languages. The development of GF started in 1998, first as a tool for controlled language implementations, where it has gained an established position in both academic and commercial projects. GF provides grammar resources for over 40 languages, enabling accurate generation and translation, as well as grammar engineering tools and components for mobile and Web applications. On the research side, the focus in the last ten years has been on scaling up GF to wide-coverage language processing. The concept of abstract syntax offers a unified view on many other approaches: Universal Dependencies, WordNets, FrameNets, Construction Grammars, and Abstract Meaning Representations. This makes it possible for GF to utilize data from the other approaches and to build robust pipelines. In return, GF can contribute to data-driven approaches by methods to transfer resources from one language to others, to augment data by rule-based generation, to check the consistency of hand-annotated corpora, and to pipe analyses into high-precision semantic back ends. This article gives an overview of the use of abstract syntax as interlingua through both established and emerging NLP applications involving GF.

Digitale Grammatik der litauischen Sprache

Der Artikel beschreibt die ersten Schritte beim Schaffen der litauischen digitalen Grammatik. Dazu benutzt man die Software “Grammatical Framework“, die für das Schaffen der digitalen Grammatiken bestimmt ist. 2018 sind die digitalen Grammatiken für 34 Sprachen geschaffen. Sie alle werden in Resource Grammar Library aufbewahrt. Dort sind auch die Grammatik der lettischen, sowie der estnischen Sprachen, aber die litauische Grammatik steht in dieser Liste noch nicht. Der wichtigste Zweck der digitalen Grammatiken ist die maschinelle Übersetzung. Es wäre sehr nützlich beim Übersetzen unter mehreren Sprachen eine universale Grammatik zu haben. Im Artikel werden die Versuche beschrieben, eine solche Grammatik zu schaffen, aber es gelang bis jetzt noch nicht wegen der großen Unterschiede, die die einzelnen Sprachen aufweisen. In der maschinellen Übersetzung werden die statistischen sowie die neural-Network Methoden jetzt mehr geläufig. Im Artikel beschreibt man ihre Vorzüge und Nachteile. Die Beispiele der Übersetzung mit Hilfe der digitalen Grammatik der litauischen Sprache sind im Artikel aufgeführt und mit den Übersetzungen von Google verglichen.

John Palsgrave’s description of French word-formation

The comprehensive nature of John Palsgrave’s endeavour to analyse and describe the French language in Lesclarcissement de la langue francoyse (1530) also encompasses the formation of words. Whereas Chapter 6 focused on his pioneering achievement as a grammarian and lexicographer, this chapter describes his most impressive work as a sixteenth-century lexicologist analysing the word-structures of a vernacular. The coining of words is embedded in a word class-based grammatical framework. For each word class, e.g. nouns, adjectives, verbs, adverbs, etc., he discusses the formative processes (derivation and composition), specifies the formal patterns (with their changes to the base), paraphrases the meaning of each formation, and then provides a good number of examples. Exceptions are also pointed out.