--- title: "Transfer Learning for Creole TTS — Strickland 2026" description: "Creole languages typically take their vocabulary from a lexifier and much of their phonology and syntax from substrate languages." authors: ["Emmett Strickland", "Marc Evrard", "Valentina Fedchenko"] year: 2026 venue: "Proceedings of the SIGUL 2026 Joint Workshop with ELE, EURALI, and DCLRL: Towards Inclusivity and Equality: Language Resources and Technologies for Under-Resourced and Endangered Languages" source: "https://aclanthology.org/2026.sigul-1.11/" doi: "10.63317/5d5qjmokuvmc" licence: "CC BY-NC 4.0 (ELRA, LREC proceedings)" licence_url: "https://creativecommons.org/licenses/by-nc/4.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/strickland-2026-transfer-learning-creole-tts-pilot/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/strickland-2026-transfer-learning-creole-tts-pilot/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/10/) / Transfer Learning for Creole TTS: A Pilot Study on Whether Substrate Phonologies or Lexifier Vocabularies Matter More # Transfer Learning for Creole TTS: A Pilot Study on Whether Substrate Phonologies or Lexifier Vocabularies Matter More **Authors:** [Emmett Strickland](https://md.research.elamur.ai/free/en/authors/emmett-strickland/), Marc Evrard, Valentina Fedchenko · **Year:** 2026 · **Venue:** Proceedings of the SIGUL 2026 Joint Workshop with ELE, EURALI, and DCLRL: Towards Inclusivity and Equality: Language Resources and Technologies for Under-Resourced and Endangered Languages · **Licence:** [CC BY-NC 4.0 (ELRA, LREC proceedings)](https://creativecommons.org/licenses/by-nc/4.0/) **Links:** [Publisher's page](https://aclanthology.org/2026.sigul-1.11/) · [PDF](https://aclanthology.org/2026.sigul-1.11.pdf) · [DOI 10.63317/5d5qjmokuvmc](https://doi.org/10.63317/5d5qjmokuvmc) · [OpenAlex](https://openalex.org/W7166510728) **Citations:** 0 citations (openalex-snapshot · 2026-10-14) ## Summary Creole languages typically take their vocabulary from a lexifier and much of their phonology and syntax from substrate languages. This raises a practical question: which parent language gives a better starting point for transfer learning in text-to-speech? This pilot study tests that question under matched low-resource conditions for two creoles: Nigerian Pidgin (Naijá, pcm), which is English-lexified, and Guadeloupean Creole (gcf), which is French-lexified. The authors recorded about 30 minutes of studio read speech from one female native speaker per language. They then fine-tuned VITS base models, built with CoquiTTS on grapheme input and trained on English, French and Yoruba, which produced six systems. Evaluation combined MCD, PESQ and informal listening. French-based models came out ahead for both creoles, but the difference was statistically significant only for Guadeloupean Creole. Yoruba-based models were generally the weakest, mainly because they handled consonant clusters poorly. The authors suggest that historical relatedness may predict transfer less well than phonotactic compatibility and orthographic depth. This is encouraging for languages that lack a close, well-resourced relative. ## Key points - **New datasets:** two single-speaker creole TTS datasets (~30 min each, 22,050 Hz, RØDE Wireless GO II microphone). - Recordings were segmented into sentences or breath groups. - About 10% of each dataset was held out as a gold-standard test set. - Speaker consent covers internal research use only. - **Base models:** trained on LJ Speech (English), SIWIS (French, "neutral" emphasis subset only) and ÌròyìnSpeech (Yoruba). - Base training used 3000 samples, 1000 epochs and batch size 32. - Fine-tuning used 300 creole samples for 600 epochs. - **Naijá results:** French gave the best mean MCD (English 17.36, Yoruba 17.33) and the highest mean PESQ (French 1.08, English 1.06, Yoruba 1.05). - Friedman tests were not significant (p = 0.22 for MCD, p = 0.51 for PESQ). - **Guadeloupean Creole results:** a Friedman test gave p = 0.0006, a significant difference between models. - The gcf models outperformed their pcm counterparts overall. - The authors attribute this partly to gcf's shallow orthography and to the gcf speaker's experience reading creole texts. - **Future work:** the authors plan to repeat the experiment with phonetized input to separate the effects of phonology and orthography. They also want to study how tone marking in spelling interacts with prosody and to evaluate the base models themselves. *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: [Linguistics & languages](https://md.research.elamur.ai/free/en/shelves/linguistics/) ## Cite The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{strickland2026transfer, title = {{Transfer Learning for Creole TTS: A Pilot Study on Whether Substrate Phonologies or Lexifier Vocabularies Matter More}}, author = {Strickland, Emmett and Evrard, Marc and Fedchenko, Valentina}, booktitle = {Proceedings of the SIGUL 2026 Joint Workshop with ELE, EURALI, and DCLRL: Towards Inclusivity and Equality: Language Resources and Technologies for Under-Resourced and Endangered Languages}, year = {2026}, doi = {10.63317/5d5qjmokuvmc}, url = {https://doi.org/10.63317/5d5qjmokuvmc} } ``` - [← previous Transcription Accuracy in the Icelandic Gigaword Corpus: Evaluating Automatic and Manual Annotation](https://md.research.elamur.ai/free/en/papers/mechler-2026-transcription-accuracy-icelandic-gigaword-corpus/) - [next → Transformer Attention as a Unified Model of Encoding and Retrieval in Human Sentence Processing](https://md.research.elamur.ai/free/en/papers/parker-2026-transformer-attention-as-unified-model/) - [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/10/) ## References 6 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. Enoch O Aboh. 2016. Creole distinctiveness: A dead end. Journal of Pidgin and Creole Languages, 31(2):400–418. [find](https://scholar.google.com/scholar?q=Enoch+O+Aboh%2E+2016%2E+Creole+distinctiveness%3A+A+dead+end%2E+Journal+of+Pidgin+and+Creole+Languages%2C+31%282%29%3A400%E2%80%93418%2E) 2. Peter Bakker, Aymeric Daval-Markussen, Mikael Parkvall, and Ingo Plag. 2011. Creoles are typologically distinct from non-creoles. Journal of Pidgin and Creole languages, 26(1):5–42. [find](https://scholar.google.com/scholar?q=Peter+Bakker%2C+Aymeric+Daval%2DMarkussen%2C+Mikael+Parkvall%2C+and+Ingo+Plag%2E+2011%2E+Creoles+are+typologically+distinct+from+non%2Dcreoles%2E+Journal+of+Pidgin+and+Creole+languages%2C+26%281%29%3A5%E2%80%9342%2E) 3. Ayo. Bamgbose. 2000. A grammar of Yoruba, volume 5. Cambridge University Press. [find](https://scholar.google.com/scholar?q=Ayo%2E+Bamgbose%2E+2000%2E+A+grammar+of+Yoruba%2C+volume+5%2E+Cambridge+University+Press%2E) 4. Derek Bickerton. 1998. Creole languages, the language bioprogram hypothesis, and language acquisition. In Handbook of child language acquisition, pages 195–220. Brill. [find](https://scholar.google.com/scholar?q=Derek+Bickerton%2E+1998%2E+Creole+languages%2C+the+language+bioprogram+hypothesis%2C+and+language+acquisition%2E+In+Handbook+of+child+language+acquisition%2C+pages+195%E2%80%93220%2E+Brill%2E) 5. Michel DeGraff. 2005. Linguists’ most dangerous myth: The fallacy of creole exceptionalism. Language in society, 34(4):533–591. [find](https://scholar.google.com/scholar?q=Michel+DeGraff%2E+2005%2E+Linguists%E2%80%99+most+dangerous+myth%3A+The+fallacy+of+creole+exceptionalism%2E+Language+in+society%2C+34%284%29%3A533%E2%80%93591%2E) 6. Gölge Eren, Coqui TTS Team, et al. 2021. Coqui tts. Zenodo. 5. \#\# Conclusion [find](https://scholar.google.com/scholar?q=G%C3%B6lge+Eren%2C+Coqui+TTS+Team%2C+et+al%2E+2021%2E+Coqui+tts%2E+Zenodo%2E+5%2E+%5C%23%5C%23+Conclusion)