Elamur Research⌕
☰
home / papers / 2024 / Lateral Inversions, Word Form/Order, Unnamed Grammatical Entities and Ambiguities in the Constituency Parsing and Annotation of the Igala Syntax through the English Language

Lateral Inversions, Word Form/Order, Unnamed Grammatical Entities and Ambiguities in the Constituency Parsing and Annotation of the Igala Syntax through the English Language

Paper · 2024 · Mahmud Mohammed Momoh

Summary

The paper examines why Igala, a Nigerian language, is hard to parse, annotate and translate into English. It focuses on three phenomena. The first is "lateral inversion": possessive determiners come after the noun, so Igala literally says "child their died" for "their child died." The second is the lack of clearly defined prepositions. The third is apostrophe contractions that fuse two words into one. The author argues that Igala is mostly SVO, like English, yet still produces translation mismatches. Using a transfer-based approach (parse the source, transfer the structure, generate the target), the author hand-builds constituency trees and Penn Treebank-style bracketed annotations for a small set of Igala sentences and their English counterparts. The paper reports no quantitative evaluation. Its contribution is descriptive: it shows where Igala structure diverges from English annotation conventions and proposes a schema for a future machine translator. It also motivates a proper Igala corpus, since the author judges earlier parallel-corpus work insufficient.

Key points

  • Igala is described as mainly subject-verb-object. The possessive forms nwū (his/her/it), mī (my) and mā (their) are placed after the noun they modify, as in "Īyē mī wā" (mother my came), which means "my mother came."
  • Six tree graphs were produced: Figures I–III cover Igala expressions and IV–VI cover English. Bracketed annotations use underscore-attached POS tags such as NP1, VVD and NN1 in the Penn Treebank / Lancaster style.
  • The ≡ symbol marks direct translation equivalence. The ≅ symbol marks syntactic isomorphism, meaning the same meaning with a different word sequence.
  • Contractions like gw'ȯjī ("sat on") were split into gwū + ȯjī before annotation, because the fused form cannot be tagged cleanly. ȯjī literally means "head" and is used for "on/above/over."
  • The author concludes that Igala has no true prepositions in the strict sense, which argues against earlier work (e.g., Ilori 2015) that labelled certain words as prepositions. "kā" ("one") was treated as an article in the treebank.

Interesting ideas

  • I1. Undo orthographic contractions into separate words before tagging, so that every token gets a single well-defined grammatical category.
  • I2. Use two separate relations between source and translation: strict equivalence, and same-meaning-but-reordered isomorphism.
  • I3. Words that are grammatically nouns can do the job of adpositions; Igala's "head" used for "on" is one example. Such words are better analysed as relational nouns than forced into a preposition tag.
  • I4. Keep a literal word-by-word gloss next to the fluent rendering, so that mismatches in order and missing categories stay visible.
  • I5. Use transfer-based structure mapping (analyse the source tree, map it onto a target tree, then generate) as an inspectable alternative to opaque end-to-end translation.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

  • I1 (tokens): Mova's tokenizer could split clitic contractions in informal or dialectal English into separate tokens, so each one gets its own tag and lemma.
  • I2 (back): Translation and the text→graph→text compressor could label each output as either order-preserving or reordered-but-equivalent, which makes evaluation of reordering explainable.
  • I3 (world): The ontology could store a mapping from body-part nouns to spatial relations, so that phrases like "on top of" or "at the foot of" resolve to the same location relation.
  • I4 (learn): Gloss-vs-fluent pairs could serve as diagnostic test items that check whether the parser recovers the same tree when word order differs.
  • I5 (back): Translation through the meta-world could log each tree-transfer rule it applies, so every reordering decision is traceable.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 works

🧭 Topics 2Research subjects and languages the work is about.English · 2,379 worksSyntax & grammar formalisms · 221 works

Cite

Cite as Mahmud Mohammed Momoh. 2024. Lateral Inversions, Word Form/Order, Unnamed Grammatical Entities and Ambiguities in the Constituency Parsing and Annotation of the Igala Syntax through the English Language. Proceedings of the Fifth Workshop on Resources for African Indigenous Languages @ LREC-COLING 2024.

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{momoh2024lateral,
  title = {{Lateral Inversions, Word Form/Order, Unnamed Grammatical Entities and Ambiguities in the Constituency Parsing and Annotation of the Igala Syntax through the English Language}},
  author = {Mahmud Mohammed Momoh},
  booktitle = {Proceedings of the Fifth Workshop on Resources for African Indigenous Languages @ LREC-COLING 2024},
  year = {2024},
  doi = {10.63317/292bjjf6tmna},
  url = {https://doi.org/10.63317/292bjjf6tmna}
}

References

28 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Abraham Sunday Unubi. (2017). “The use of conjunctions in English and Igala: A contrastive analysis”. International Journal of Advanced Multidisciplinary Research, 4(8), Pages 34-64. find
  2. Atadoga, F. T. (2011). ‘Igala Morphology’. In Omachonu G. S (ed.) Igala language Studies. Saarbrucken: LAP Lambert Academic Publishing, Pages 76-102. find
  3. Attabor Theophilus Ocheja. (2019). “Syntactic Interference: A Study of Ígálá and English Noun Phrases in Malachai 1:6 and Mathew 2:1”. Journal of Literature, Languages and Linguistics, 60, Pages 40-46. ↗
  4. Ayegba Sani Felix, Osuagwu O.E, and Njoku Dominic Okechukwu. (2014). “Machine Translation of Noun Phrases from English to Igala using the Rule-Based Approach”. West African Journal of Industrial and Academic Research, 2(1): 18-28. find
  5. Ayegba Sani Felix, Abu Onoja, and Musa Ugbedeojo. (2017). “English – Igala Parallel Corpora for Natural Language Processing Applications”. International Journal of Computer Applications, 171(9): 1-6. ↗
  6. Bonami Olivier, Godard Danièlle, and Marandin Jean-Marie. (1999). “Constituency and word order in French subject inversion”. In Gosse Bouma, Erhard Hinrichs, Geert-Jan M. Kriujff, and Richard Oehrle (eds.) Constraints and Resources in Natural Language Syntax and Semantics. CSLI Publications, Stanford, USA, Pages 1-20. find
  7. Borsley D. Robert, Tallerman Maggie, and Willis David. (2007). The Syntax of Welsh. Cambridge University Press, Cambridge, UK, Pages 1-102. ↗
  8. Brunato Dominique, and Dell-Orletta Felice. (2017). “On the order of words in Italian: a study on genre vs complexity”. In Proceedings of the Fourth International Conference on Dependency Linguistics, Pisa, Italy, September 18-20, Pages 25-31. find
  9. Comer J. William. (2021). "Russian’s Most Frequent Words and Implications for Vocabulary Instruction," Russian Language Journal, 71(1), Emmanuel Edoja Achor, and Christiana Akor. (2015). “Exploring the Use of Igala Language in Teaching Statistics to Samples of Selected Primary Six Learners”. In Proceedings of ISTE International Conference on "Towards Effective Teaching and Meaningful Learning in Mathematics, Science and Technology", University of South Africa, Pages 55-66. find
  10. Fransen Theodorus. (2020). “Automatic Morphological Parsing of Old Irish Verbs Using Finite-State Transducers”. Leeds Working Papers, 1: 15-28. find
  11. Frazier Lyn. (1980). “Parsing and Constraints on Word Order”. University of Massachusetts Occasional Papers, 6(5): 177-198. find
  12. Gudivada N. Venkat, and Arbabifard Kamyar. (2018). “Computational Analysis and Understanding of Natural Languages : Principles, Methods and Applications”. In Gudivada N. Venkat, and Rao, C.R. Handbook of Statistic 38. https://www.sciencedirect.com/topics/computer -science/translationmodel#:~:text=The%20language%20model%2 0can%20be,using%20the%20notion% 20of%2 0alignments. ↗
  13. Hornby S. Albert. (2015). Oxford Advanced Learner’s Dictionary of English, (Ninth Edition). Oxford University Press, Oxford, UK, Page 1210. find
  14. Ikani, F. E. (2011). ‘Sense and Meaning Relations in Igala’. Omachonu G. S (ed.) Igala Language Studies. Saarbrucken: LAP Lambert Academic Publishing, Pages152-176. find
  15. Ileana Paul, and Postdam Eric. (2024). “Malagasy Framing Demonstratives and the Syntax of Doubling”. Glossa: A Journal of General Linguistics, 9: 1-37. ↗
  16. Ilori, Folorunso. (2015). “Prepositions in Igala”. Ihafa: A Journal of African Studies, 7(1): 139-160. find
  17. Joshua Attah, Ayegba Sani Felix, and Orah Richard Ojochegbe. (2020). “Example Based Machine Translation of English to Igala Language”. International Journal of Trend in Research and Development, 7(1): Pages 32-35. find
  18. Jurafsky Daniel, and Martin H. James. (2009). Speech and Language Processing. https://en.wikipedia.org/wiki/Transfer-based_machine_translation#cite_note-slp-1 ↗
  19. Kalin Laura. (2014). “The syntax of OVS word order in Hixkaryana”. Natural Language and Linguistic Theory, 32: 1089-1104. ↗
  20. Mahmud Mohammed Momoh. (2023). “Vowels and the Igala Language Resources”. In Proceedings of the Fourth workshop on Resources for African Indigenous Languages , pages 106-114. ↗
  21. Marcus Mitchell, Beatrice Santorini, and Marcinkiewicz Mary Ann. (1993). “Building a Large Annotated Corpus of English: The Penn Treebank”. Association for Computational Linguistics, 19(2), pages 113-330. ↗
  22. Marcus Mitchell, Kim Grace, and Marcinkiewicz Mary Ann., et al. (1994). “The Penn Treebank : Annotating Predicate Argument Structure”. https://aclanthology.org/H94-1020.pdf ↗
  23. McCone K. (1997). The Early Irish verb. Revised edition with index verborum. An Sagart, Maynooth, Ireland, Page 17. find
  24. Million Meshesha, and Yitayew Solomon. (2018). “English-Afaan Oromo Statistical Machine Translation”. International Journal of Computational Linguistic, 9(1): 26-31. find
  25. Minhui Choi, and Emily Schmidt. (2015). “Postpositions and Word Order Variation in Korean”. Linguistic Portfolios, 4(10): 108-115. find
  26. Mun Seongmin, and Desagulier Guillaume. (2022). “How do Transformer-Architect ure Models Address Polysemy of Korean Adverbial Postpositions?”. In Proceedings of Deep Learning Inside Out (DEELIO 2022): The Third Workshop on Knowledge Extraction and Integration for Deep Learning Architectures, Dublin, Ireland, Pages 11-21. find
  27. Namboodiripad Savithry, Kim Dayoung, and Kim Gyeongnam. (2017). “English-dominant Korean-speakers show reduced flexibility in constituent order”. Chicago Linguistic Society, 53: Philpot, W.T.A. (1935). “Notes on the Igala Language”. Bulletin of the School of Oriental Studies, 7(4): Pages 897-912. find
  28. Santorini Beatrice. (1990). Part of Speech Tagging Guidelines for the Penn Treebank Project. https://www.cis.upenn.edu/~bies/manuals/tagg uide.pdf Wang, L., Tu, Z., and Zhang, X., et al. (2017). “A Novel and Robust Approach for Pro-drop Language Translation”. Machine Translation, 31: 65-87. ↗

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/momoh-2024-lateral-inversions-word-form-order/