About

These pages display stories recorded, transcribed and translated by Alan Corbiere and the Anishinaabemowin Revival Program at M’Chigeeng First Nation. We hope that they will be a useful resource for learners, speakers and scholars of Nishnaabemwin.

Each Nishnaabemwin sentence is accompanied by an English translation, and beneath it, a detailed ‘interlinear’ analysis. It is our hope that by making these analyses available, learners of Nishnaabemwin can have an easier time understanding what is going on in a sentence. This might just mean getting a sense for which word means what, or you could figure out how to properly modify words with suffixes and prefixes (and combine the words into a sentence!).

Work on this collection is ongoing, and we are fixing bugs and adding features all the time … so be sure to check back in! And if you notice something incorrect, missing or just plain strange, please reach out so we can fix it. The analysis was mostly done by machine (more on that in a bit), and the machine can fail to recognize a word or it can give the wrong analysis. Many of these missing cases have been fixed by hand, though we could not figure out some of them to our satisfaction. Mistaken analyses are rarer, as far as we know, but it is hard to detect them.

Interlinear Analysis

The analysis takes each word in a sentence and displays further information about that word on a line directly underneath it … so the analysis is ‘inter’ (between) ‘linear’ (lines). The very first line simply repeats the original sentence, albeit without punctuation and in all lowercase letters. The second line is the first analysis line. Here you will find part of speech information (adverb, noun, verb, animacy, etc) and grammatical information (subject/object aka who is the do-er and the do-ee of the verb, possessor, etc). Hover your mouse over an analysis to get an explanation of the grammatical codes! The third line gives a very terse translation of the word into English (it is only one word long!). These translations were produced by an AI system (Microsoft Copilot) using the definitions in the dictionary portion of this site. They have not been extensively checked, so they may seem a bit odd at times (if so, please let us know!). Also, some words in the stories are not in the dictionary yet, so no terse definition is available even if there is a grammatical analysis. If a word was homophonous with another word in the dictionary, both terse definitions are shown, separated by slashes. If the word is a compound, the terse definitions of each compound are shown, separated by a hyphen.

Computational Analyzer

The analyses were produced by a digital analyzer of Nishnaabemwin that can stick prefixes and suffixes onto root words, and make adjustments to make sure the resulting word is spelled properly. For the most part, we used words from the dictionary, but we added/subtracted/modified entries as the need arose. Simplifying slightly, when the analyzer is given a word, it checks if the word is in the set of prefix-root-suffix combinations. If the answer is ‘yes’, it gives a summary of what it found (which is translated into the grammatical summary you see), and if the answer is ‘no’, it says it couldn’t find anything.

Notice that the analyzer tries to make sure that a word is spelled “properly”. This is a tricky question in any language, and especially in Nishnaabemwin. Nishnaabemwin can be written many ways, but probably the most popular systems are the “double vowel” system devised by Chuck Fiero and used in Richard Rhodes’ Eastern Ojibwa-Chippewa-Ottawa Dictionary (the Fiero/Rhodes system), and the very similar, but crucially different, system developed by Mary Ann Corbiere (the Corbiere system) and used in the Nishnaabemwin Online Dictionary. At its most basic level, the analyzer works with the Fiero/Rhodes system, and there is a modification that allows it to use the Corbiere system. Ultimately though, spelling in Nishnaabemwin is really hard. Further modifications to compensate for non-standard spellings have been made, but each change away from the core standard makes the analyzer more bloated and less reliable.

The computational analyzer used to produce these interlinearizations is based on an initial version developed at the University of Alberta in 2015-2017, which was then ‘forked’ (further developed independently) by Dustin Bowers in 2022-2024 to produce the current model (https://github.com/bowersd/otw). If you want to analyze your own texts with the analyzer, there is an online version of it here.

Text Analysis and Programming:

Dustin Bowers

Web Application Programming:

Delasie Torkornoo, Holden Madagame

Story Editor:

Alan Corbiere

Transcriptions & Translations:

Alan Corbiere, Evelyn Roy, Albertine Migwans, and Alvin Ted Corbiere

Funding

National Endowment for the Humanities, USA, FN-291243-23 (2023-2024)

Canada Research Chair Grant to Alan Corbiere, York University (2022)

M’Chigeeng First Nation, Canada

Funding to develop the web application:

Social Sciences and Humanities Research Council of Canada (grant # 435-2014-1199)

Social Sciences and Humanities Research Council of Canada (Partnership grant # 895-2019-1012)

Social Sciences and Humanities Research Council of Canada (Insight grant)

National Research Council of Canada (Canadian Indigenous languages technology project, 2024)

Funding supporting the initial development of the analyzer:

Social Sciences and Humanities Research Council of Canada (Partnership Development grant # 890-2013-0047)

Supporting Institutions:

Carleton University, Canada

University of Wisconsin–Madison, USA

Reed College, Portland, Oregon, USA

Temple Beth Israel, Eugene, Oregon, USA

Copyright notice:

Stories posted under license from M’Chigeeng First Nation

Acknowledgments

Many thanks to Rand Valentine, Eli Franz, Zoe Kyriacopolous, Natasha Cruz, and Amy Debassige, for their help in improving the analyses and for acting as our quality control team.