Elamur Research⌕
☰
home / papers / 2021 / From Plenipotentiary to Puddingless: Users and Uses of New Words in Early English Letters

From Plenipotentiary to Puddingless: Users and Uses of New Words in Early English Letters

Paper · 2021 · Tanja Saily, Eetu Makela and Mika Hamalainen

Summary

The authors set out to find who coined or adopted new English words in personal letters, since earlier studies of lexical history leaned on printed texts written mostly by educated men. They used the Corpora of Early English Correspondence (CEEC), which carries social metadata for each letter. Heavy spelling variation meant most CEEC word forms could not be matched to the OED or MED, so they applied neural machine translation normalization, keeping only outputs that are real OED lemmas. Its errors were too frequent and too uneven across social groups for a fully automatic pipeline. They therefore treated machine output as suggestions, checked them by hand, and limited the work to two balanced samples (1640–1660 and 1760–1780). A word counted as a neologism if its first OED attestation came no more than 40 years before its use in the letter. Men used more new words, but in the 18th century women and lower ranks took part more. Letters between close friends were richest in neologisms. Both samples also show historical effects: the Civil War and the changing social role of letters.

Key points

  • Dictionary coverage. CEEC has about 150,000 distinct word forms, about 125,000 after removing foreign words, proper nouns and abbreviations. Only about 36,000 mapped directly to the OED/MED, and 88,000 (70%) did not. Words from lower-rank writers matched less often.
  • Normalization quality. NMT normalization with an OED-lemma filter recovered correct lemmas for 61% of the previously unmatched words. In a 17th-century check:
    • 75% of words received a lemma, but 30% of those lemmas were wrong;
    • 17% of the correctly lemmatized words had the wrong part of speech;
    • accuracy differed by gender and relationship type, not by rank.
  • Sampling. Letters were drawn stratified by writer gender, social rank and writer–recipient relationship. Candidates were words making their first CEEC appearance in a given letter. The 17th-century sample yielded 820 candidate types and the 18th-century sample 645, roughly 9% of their respective pools of 8,954 and 7,131.
  • Results. Manual verification in the FiCa tool, plus OED dates, gave 42 neologism types (53 tokens) for the 17th century and 21 types (one token each) for the 18th. Meanings were analysed through the hierarchy of the Historical Thesaurus of the OED.
  • Limits and future work. Some candidates were missing from the OED: 13 in the 18th-century sample and 4 in the 17th. Back-translation did not improve normalization, because the training variants come from dictionaries and differ in kind from the spelling variation in CEEC.

Interesting ideas

  • I1. Constraining a seq2seq normalizer so that only lemmas existing in a reference dictionary are accepted turns free generation into a much safer lookup step.
  • I2. Error rates were broken down by social group and register, revealing that the pipeline failed unevenly, which would silently skew any downstream comparison.
  • I3. A word only became a candidate when it first appeared in the corpus at that point, and it was then checked against a dictionary date window, combining corpus-internal novelty with an external reference.
  • I4. Mapping each word to nodes of a hierarchical thesaurus let the authors compare meanings at coarse or fine levels of the hierarchy.
  • I5. The failed back-translation showed that synthetic training data helps only if it imitates the real variation of the target texts, not a cleaner authoritative source.

For Mova

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

  • I1 (tokens): The en lemmatizer could normalize odd or archaic spellings with a candidate generator, then accept only lemmas found in WordNet/the lexicon, flagging the rest as unknown.
  • I2 (learn): Evaluation reports could slice parser and frame accuracy by genre (fables, school tests, word problems) so that a model that is good on average but weak on one text type gets caught.
  • I3 (world): New entity or term mentions could be detected as first occurrences in a text and then checked against the knowledge graph, separating truly unseen words from known ones that need linking.
  • I4 (world): Frame arguments could be grouped by WordNet/FrameNet ancestors at a chosen depth, giving explainable summaries such as "mostly emotion words" instead of word lists.
  • I5 (learn): Silver or synthetic data for parsers should be generated to match the noise and style of the texts Mova actually reads, with a held-out real sample checking that it helps.

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

 Institutions

🎓University of HelsinkiHelsinki · Finland33 works · 4,705 citations

📚 Shelves 3 Arts and Humanities · 9,296 works Language and Linguistics · 9,206 works Literature and Literary Theory · 340 works

🧭 Topics 1 English · 1,373 works

Cite

Cite as Tanja Saily, Eetu Makela and Mika Hamalainen. 2021. From Plenipotentiary to Puddingless: Users and Uses of New Words in Early English Letters. arXiv:2103.09926.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.31885/9789515150257.15 · arXiv 2103.09926

@misc{saily2021plenipotentiary,
  title = {{From Plenipotentiary to Puddingless: Users and Uses of New Words in Early English Letters}},
  author = {Tanja Saily, Eetu Makela and Mika Hamalainen},
  year = {2021},
  doi = {10.31885/9789515150257.15},
  eprint = {2103.09926},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.31885/9789515150257.15}
}

References

13 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. 1. Brewer, C.: Treasure-house of the language: The living OED. Yale University Press, New Haven (2007) Users and Uses of New Words in Early English Letters 2. Carretta, V.: Sancho, (Charles) Ignatius (1729?–1780), author. In: Oxford Dictionary of National Biography. Oxford University Press, Oxford, online edn. (2021). https://doi.org/10.1093/ref:odnb/24609 3. Chakravarthi, B.R., Rani, P., Arcan, M., McCrae, J.P.: A survey of orthographic information in machine translation. arXiv preprint arXiv:2008.01391 (2020) 4. Clark, G. (ed.): Correspondence of the Foundling Hospital inspectors in Berkshire 1757–68, Berkshire Record Society, vol. 1. ↗
  2. Berkshire Record Society, Reading (1994) 5. Devlin, J., Chang, M.W., Lee, K., Toutanova, K.: BERT: Pre-training of deep bidirectional transformers for language understanding. In: Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers). pp. 4171–4186. Association for Computational Linguistics, Minneapolis, Minnesota (2019). https://doi.org/10.18653/v1/N19-1423 6. Dixon, R.M.W.: The rise and fall of languages. ↗ our card
  3. Cambridge University Press, Cambridge (1997) 7. Domingo, M., Casacuberta, F.: Modernizing historical documents: A user study. Pattern Recognition Letters 133, 151–157 (2020) 8. Dubossarsky, H., Hengchen, S., Tahmasebi, N., Schlechtweg, D.: Time-out: Temporal referencing for robust modeling of lexical semantic change. In: Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. pp. 457–470. Association for Computational Linguistics, Florence, Italy (2019). https://doi.org/10.18653/v1/P19-1044 9. Duong, Q., Hämäläinen, M., Hengchen, S.: An unsupervised method for OCR post-correction and spelling normalisation for Finnish. arXiv preprint arXiv:2011.03502 (2020) 10. ECCO: Eighteenth Century Collections Online. Gale (nd), https://www.gale.com/primary-sources/eighteenth-century-collections-online 11. EEBO: Early English Books Online. ProQuest (nd), https://proquest.libguides.com/eebo 12. Grieve, J., Nini, A., Guo, D.: Analyzing lexical emergence in Modern American English online. English Language and Linguistics 21(1), 99–127 (2017). https://doi.org/10.1017/S1360674316000113 13. Hämäläinen, M., Rueter, J.: Finding Sami cognates with a character-based NMT approach. In: Proceedings of the 3rd Workshop on the Use of Computational Methods in the Study of Endangered Languages Volume 1 (Papers). pp. 39–45. Association for Computational Linguistics, Honolulu (2019), https://www.aclweb.org/anthology/W19-6006 14. Hämäläinen, M., Säily, T., Rueter, J., Tiedemann, J., Mäkelä, E.: Normalizing early English letters to Present-day English spelling. In: Proceedings of the 2nd Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature. pp. 87–96. Association for Computational Linguistics, Santa Fe, New Mexico (2018), http://aclweb.org/anthology/W18-4510 15. Hämäläinen, M., Säily, T., Rueter, J., Tiedemann, J., Mäkelä, E.: Revisiting NMT for normalization of early English letters. In: Proceedings of the 3rd Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature. pp. 71–75. Association for Computational Linguistics, Minneapolis, USA (2019). https://doi.org/10.18653/v1/W19-2509 16. Hämäläinen, M., Wiechetek, L.: Morphological disambiguation of South Sámi with FSTs and neural networks. In: Proceedings of the 1st Joint Workshop on Spoken Language Technologies for Under-resourced languages (SLTU) and Collaboration and Computing for Under-Resourced Languages (CCURL). pp. 36–40. ↗ our card
  4. European Language Resources Association, Marseille, France (2020), https://www.aclweb.org/anthology/2020.sltu-1.5 17. Hengchen, S., Tahmasebi, N., Schlechtweg, D., Dubossarsky, H.: Challenges for computational lexical semantic change. arXiv preprint arXiv:2101.07668 (2021) T. Säily et al. 18. Keune, K.: Explaining register and sociolinguistic variation in the lexicon: Corpus studies on Dutch. Ph.D. thesis, Radboud University Nijmegen (2012), https://hdl.handle.net/2066/101694 19. Kyle, C.R.: L’Estrange, Sir Hamon (1583–1654), politician. In: Oxford Dictionary of National Biography. Oxford University Press, Oxford, online edn. (2005). https://doi.org/10.1093/ref:odnb/67340 20. Lijffijt, J., Säily, T., Nevalainen, T.: CEECing the baseline: Lexical stability and significant change in a historical corpus. In: Tyrkkö, J., Kilpiö, M., Nevalainen, T., Rissanen, M. (eds.) Outposts of historical corpus linguistics: From the Helsinki Corpus to a proliferation of resources. Studies in Variation, Contacts and Change in English, vol. 10. ↗
  5. VARIENG, Helsinki (2012), https://varieng.helsinki.fi/series/volumes/10/lijffijt_saily_nevalainen/ 21. Mäkelä, E., Lagus, K., Lahti, L., Säily, T., Tolonen, M., Hämäläinen, M., Kaislaniemi, S., Nevalainen, T.: Wrangling with non-standard data. In: Proceedings of the Digital Humanities in the Nordic Countries 5th Conference, Riga, Latvia, October 21–23, 2020. CEUR Workshop Proceedings, vol. 2612, pp. 81–96. ↗
  6. CEUR-WS.org, Aachen (2020), http://ceur-ws.org/Vol-2612/paper6.pdf 22. McCrae, J.P.: Identification of adjective-noun neologisms using pretrained language models. In: Proceedings of the Joint Workshop on Multiword Expressions and WordNet (MWE-WN 2019). pp. 135–141. Association for Computational Linguistics, Florence, Italy (2019). https://doi.org/10.18653/v1/W19-5116 23. MED: Middle English Dictionary. University of Michigan (nd), https://quod.lib.umich.edu/m/med/ 24. Nevalainen, T.: Early Modern English lexis and semantics. In: Lass, R. (ed.) The Cambridge history of the English language, III: 1476–1776, pp. 332–458. ↗
  7. Cambridge University Press, Cambridge (1999). https://doi.org/10.1017/CHOL9780521264761.006 25. Nevalainen, T., Raumolin-Brunberg, H., Keränen, J., Nevala, M., Nurmi, A., Palander-Collin, M., Kaislaniemi, S., Laitinen, M., Säily, T., Sairio, A.: CEEC, Corpora of Early English Correspondence. Department of Modern Languages, University of Helsinki (1998–2006), https://varieng.helsinki.fi/CoRD/corpora/CEEC/ 26. Nevalainen, T., Säily, T., Vartiainen, T., Liimatta, A., Lijffijt, J.: History of English as punctuated equilibria? A meta-analysis of the rate of linguistic change in Middle English. Journal of Historical Sociolinguistics 6(2), 20190008 (2020). https://doi.org/10.1515/jhsl-2019-0008 27. Oxford English Dictionary: OED Online. Oxford University Press (nd), http://www.oed.com/ 28. Palmer, C.C.: Measuring productivity diachronically: Nominal suffixes in English letters, 1400–1600. English Language and Linguistics 19(1), 107–129 (2015). https://doi.org/10.1017/S1360674314000264 29. Partanen, N., Hämäläinen, M., Alnajjar, K.: Dialect text normalization to normative standard Finnish. In: Proceedings of the 5th Workshop on Noisy User-generated Text (W-NUT 2019). pp. 141–146. Association for Computational Linguistics, Hong Kong, China (2019). https://doi.org/10.18653/v1/D19-5519 30. Peters, M., Neumann, M., Iyyer, M., Gardner, M., Clark, C., Lee, K., Zettlemoyer, L.: Deep contextualized word representations. In: Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers). pp. 2227–2237. Association for Computational Linguistics, New Orleans, Louisiana (2018). https://doi.org/10.18653/v1/N18-1202 31. Raumolin-Brunberg, H.: Social factors and pronominal change in the seventeenth century: The Civil-War effect? In: Fisiak, J., Krygier, M. (eds.) Advances in English historical linguistics (1996), Trends in Linguistics: Studies and Monographs, vol. 112, pp. 361–388. ↗ our card
  8. Mouton de Gruyter, Berlin (1998). https://doi.org/10.1515/9783110804072.361 Users and Uses of New Words in Early English Letters 32. Renouf, A.: Tracing lexical productivity and creativity in the British media: ‘The chavs and the chav-nots’. In: Munat, J. (ed.) Lexical creativity, texts and contexts, Studies in Functional and Structural Linguistics, vol. 58, pp. 61–89. ↗
  9. John Benjamins, Amsterdam (2007). https://doi.org/10.1075/sfsl.58.12ren 33. Säily, T.: Sociolinguistic variation in English derivational productivity: Studies and methods in diachronic corpus linguistics, Mémoires de la Société Néophilologique de Helsinki, vol. XCIV. ↗
  10. Société Néophilologique, Helsinki (2014) 34. Säily, T.: Change or variation? Productivity of the suffixes -ness and -ity. In: Nevalainen, T., Palander-Collin, M., Säily, T. (eds.) Patterns of change in 18th-century English: A sociolinguistic approach, Advances in Historical Sociolinguistics, vol. 8, pp. 197–218. find
  11. John Benjamins, Amsterdam (2018). https://doi.org/10.1075/ahs.8.12sai 35. Säily, T., Mäkelä, E., Hämäläinen, M.: Explorations into the social contexts of neologism use in early English correspondence. Pragmatics & Cognition 25(1), 30–49 (2018). https://doi.org/10.1075/pc.18001.sai 36. Säily, T., Mäkelä, E., Hämäläinen, M.: Neologism dataset for “From plenipotentiary to puddingless: Users and uses of new words in early English letters” (2021). https://doi.org/10.5281/zenodo.4578940 37. Säily, T., Mäkelä, E., Kaislaniemi, S.: Cha before tea: Finding earlier mentions in a corpus of early English letters (part 1). In: Oxford English Dictionary blog. ↗ our card
  12. Oxford University Press, Oxford (2019), https://public.oed.com/blog/cha-before-tea-finding-earlier-mentions-in-a-corpus-of-early-english-letters-part-1/ 38. Säily, T., Suomela, J.: types2: Exploring word-frequency differences in corpora. In: Hiltunen, T., McVeigh, J., Säily, T. (eds.) Big and rich data in English corpus linguistics: Methods and explorations, Studies in Variation, Contacts and Change in English, vol. 19. ↗
  13. VARIENG, Helsinki (2017), https://varieng.helsinki.fi/series/volumes/19/saily_suomela/ 39. Somervell, T.: Public and private, real and fictional: The rise of women’s letter-writing in the eighteenth century. Bluestocking: Online Journal for Women’s History (2011), https://blue-stocking.org.uk/2011/03/01/public-and-private-real-and-fictional-the-rise-of-womens-letter-writing-in-the-eighteen 40. Wein, S.: Classification and analysis of neologisms produced by learners of Spanish: Effects of proficiency and task. In: Proceedings of the The Fourth Widening Natural Language Processing Workshop. pp. 88–91. Association for Computational Linguistics, Seattle, USA (2020). https://doi.org/10.18653/v1/2020.winlp-1.22 41. Wolfson, N.: The bulge: A theory of speech behavior and social distance. Penn Working Papers in Educational Linguistics 2(1), 55–83 (1990), https://repository.upenn.edu/wpel/vol2/iss1/3/ 42. Zalmout, N., Thadani, K., Pappu, A.: Unsupervised neologism normalization using embedding space mapping. In: Proceedings of the 5th Workshop on Noisy User-generated Text (W-NUT 2019). pp. 425–430. Association for Computational Linguistics, Hong Kong, China (2019). https://doi.org/10.18653/v1/D19-5555 43. Zarnoufi, R., Jaafar, H., Bachri, W., Abik, M.: MANorm: A normalization dictionary for Moroccan Arabic dialect written in Latin script. In: Proceedings of the Fifth Arabic Natural Language Processing Workshop. pp. 155–166. Association for Computational Linguistics, Barcelona, Spain (2020), https://www.aclweb.org/anthology/2020.wanlp-1.14 ↗

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/saily-2021-plenipotentiary-puddingless-users-uses-new/