Contributor Metadata Approval Sheet/

Contributor Contract/Proof

Dear Contributor(s),

Thank you for publishing with us! Please find enclosed your contributor proof. In order for your contribution
to be published, it is essential to fill out the enclosed metadata form and then confirm the Contributor Con-
tract by placing check marks and data in the marked places. Any special written agreements with us stay valid
beyond that.

Finally, please correct the proof of your contribution.

Thank you for your cooperation!

De Gruyter

Appendices

- Title page of the Contributor Contract

- Contributor Metadata Form
- General Terms and Conditions for Contributor Contract
- Proof

Contributor Publication Agreement

Walter de Gruyter GmbH · Genthiner Str. 13 · 10785 Berlin · Deutschland
CONTRIBUTOR CONTRACT
Between

De Gruyter

and

Corresponding Contributor
Maria Mazzoli

Regarding
Chapter Title:
Michif mixed verbs: Typologically unusual word-internal mixing
Chapter DOI:
10.1515/9781501511257-005
Book Title:
New Perspectives on Mixed Languages

Creative-Commons- License:
---

a Contributor Contract has been concluded. Details regarding the Contributor Contract are specified in the
attached General Terms and Conditions for Contributor Contract.

Contributor Publication Agreement

Walter de Gruyter GmbH · Genthiner Str. 13 · 10785 Berlin · Deutschland
 3

Contributor Metadata Approval Sheet

Book title: New Perspectives on Mixed Languages
Chapter DOI: 10.1515/9781501511257-005
Chapter title: Michif mixed verbs: Typologically unusual word-internal mixing

Please check the metadata in the second column carefully and update/add information in the third column. Even if you
have already submitted this information to the publisher.

Contributor 1

Corresponding address according to available information To be changed

Salutation* (Mr/Ms/other)
Title*
Surname* Mazzoli
First Name* Maria
E-Mail*
Corresponding Author* (yes/no) yes

The given address is a(n) affiliation. private address.

(For private address omit Institution and Department!)

Institution University of Groningen

Department Faculty of Arts
Number* 26
Street* Oude Kijk in t Jatstraat
ZIP Code* 9712 EK
City* Groningen
State
Country* The Netherlands
*required

The data from this form will only be used to enable internal processes! This includes sending a notification via e-mail
as soon as your contribution is published on the De Gruyter website. You can then download a PDF for personal use
using the registration information provided therein. If free copies are part of the agreement with De Gruyter, the postal
address listed here will be used to send the copies.

Contributor Publication Agreement

Walter de Gruyter GmbH · Genthiner Str. 13 · 10785 Berlin · Deutschland
 4

Contributor Metadata Approval Sheet

Book title: New Perspectives on Mixed Languages
Chapter DOI: 10.1515/9781501511257-005
Chapter title: Michif mixed verbs: Typologically unusual word-internal mixing

Please check the metadata in the second column carefully and update/add information in the third column. Even if you
have already submitted this information to the publisher.

Contributor 2

Corresponding address according to available information To be changed

Salutation* (Mr/Ms/other)
Title*
Surname* Bakker
First Name* Peter
E-Mail*
Corresponding Author* (yes/no) no

The given address is a(n) affiliation. private address.

(For private address omit Institution and Department!)

Institution Aarhus University

Department
Number*
Street*
ZIP Code*
City*
State
Country*
*required

The data from this form will only be used to enable internal processes! This includes sending a notification via e-mail
as soon as your contribution is published on the De Gruyter website. You can then download a PDF for personal use
using the registration information provided therein. If free copies are part of the agreement with De Gruyter, the postal
address listed here will be used to send the copies.

Contributor Publication Agreement

Walter de Gruyter GmbH · Genthiner Str. 13 · 10785 Berlin · Deutschland
 5

Contributor Metadata Approval Sheet

Book title: New Perspectives on Mixed Languages
Chapter DOI: 10.1515/9781501511257-005
Chapter title: Michif mixed verbs: Typologically unusual word-internal mixing

Please check the metadata in the second column carefully and update/add information in the third column. Even if you
have already submitted this information to the publisher.

Contributor 3

Corresponding address according to available information To be changed

Salutation* (Mr/Ms/other)
Title* Mrs.
Surname* DeMontigny
First Name* Verna
E-Mail* kookume@[Link]
Corresponding Author* (yes/no) no

The given address is a(n) affiliation. private address.

(For private address omit Institution and Department!)

Institution
Department
Number* 143
Street* 8th Street
ZIP Code* R7A3Z7
City* Brandon
State Manitoba
Country* Canada
*required

The data from this form will only be used to enable internal processes! This includes sending a notification via e-mail
as soon as your contribution is published on the De Gruyter website. You can then download a PDF for personal use
using the registration information provided therein. If free copies are part of the agreement with De Gruyter, the postal
address listed here will be used to send the copies.

Contributor Publication Agreement

Walter de Gruyter GmbH · Genthiner Str. 13 · 10785 Berlin · Deutschland
 6

Checked and confirmed:

Contributor Metadata and Contributor Contract 16.02.2021

Date ____________________

(By inserting the check mark you confirm the conclusion of the Contributer Contract for your contribution. In case of
technical problems with the check mark, you can agree to the contract by typing “Confirmed” in the date field, e.g.:
“Confirmed, 17.4.2020”.)

Contributor Publication Agreement

Walter de Gruyter GmbH · Genthiner Str. 13 · 10785 Berlin · Deutschland
 7

GENERAL TERMS AND CONDITIONS d) To reproduce and make available your Work in electronic form, particularly as
eBook, database, and/or other forms of electronic media and within the scope of
FOR CONTRIBUTOR CONTRACT internet services or other online uses or in interactive multimedia production;
e) To make your Work available in public or closed user groups at the location and time
of their choice (for example, eBook use in libraries) as well as reproducing it on mo-
These General Terms and Conditions are an integral part of your Contributor Contract. The indi-
nitors or other scanners and to be printed by the user as many times as the user wishes,
vidual details of your agreement with us can be found on the sheet titled Contributor Contract.
in whole or in part, including as prepublication or in excerpt;
f) To reproduce and distribute your Work on any and all data carriers, for example
Our goal as a publisher is to be a competent partner to our contributors and to facilitate the quality, DVD, CD-Rom, flash drive.
reach, reputation, and validity of their work. In order to do so, both parties must commit to a set 3.2 Ancillary Rights: You also assign to us, for the purpose of optimally exercising the rights
of rules as follows: to your Work, the right:
a) To translate it into other languages and to use the translation;
1 Definitions b) To print it in whole or in part as a prepublication and/or as subsequent reprint, inclu-
ding in newspapers and periodicals (for example in reviews);
You: You as the Contributor of the Work. When there is more than one Contributor, the term “You”
c) To publish it in whole or in part in our other publications or publications of another
shall apply collectively and the provisions set out in these terms and conditions apply equally to
publisher, including in abridged form;
all Contributors.
d) To produce or license as a paperback, or as a popular or special or reprint edition;
e) To reproduce and distribute it in a collected works edition and in compilations, even
We: the Walter De Gruyter GmbH Publisher named in your Contributor Contract. after 20 years have elapsed since the first publication of the Work;
f) To reproduce and distribute it by all other means, including photocopying, photome-
Work: the content to be published as printed and electronic edition. chanical reprinting, or as Braille embossing;
g) Of rendition, including rights to recitation, performances, and broadcast in radio or
Contributor Contract: the actual agreement, which is issued to you as a cover page to these terms television media or internet;
and conditions. h) To transfer it, in full or in part, to sound recordings, image or image-sound recor-
dings as well as the right to their reproduction, distribution, and reproduction to the
public;
2 Your Rights and Obligations
i) To use it in collections for use in church, school, or instructional settings.
2.1 Grant of Rights: As a publisher, we take copyright very seriously. We must make sure that we 3.3 Extension of Rights: Unless otherwise specified in the provisions for contributors in the au-
have the right to use all parts of your Work. You warrant that you own the manuscript, and thors section of our website you grant all usage rights and ancillary rights to us as exclusive
that you are free to assign the rights without restriction and have acquired the rights to any rights without any restriction as to content or territory for the duration of the copyright. We
content as necessary. This applies to all parts, including any illustrations, indices, tables, tex- may exercise these rights but are under no obligation to do so to the extent described here.
tual excerpts, multimedia components, etc. Further, you warrant that you have not assigned 3.4 Unknown Types of Use: You grant us the exclusive and permanent rights without any res-
any rights that would conflict with the rights granted to us in this contract. You guarantee that triction as to content and territory for all forms of media of expression now known or that
your work, including all parts, does not infringe on any rights such as copyrights, performing will be developed in the future. The granting of rights extends to the exercising of rights
rights, trademarks, rights of privacy, or other third-party rights. You agree to reimburse us for through us or through the granting of these rights to a third party.
any costs or liabilities resulting from a breach of these warranties. 3.5 Transferral to Third Parties: You grant us the right to transfer all rights listed here to third
2.2 Publication Elsewhere: In order to support the marketing strategy for your Work, you agree parties and/or to license the Work to third parties. We require these rights in order to fulfil
not to publish the Work (or a substantially equivalent work) or make it otherwise available certain sales models such as online use through aggregators (platforms that curate content
to the public without first obtaining our written consent. for specific usage by customers and give us a share of the proceeds). We naturally require
2.3 Repository Policy: You are free to self-archive up to 10% of the Work (with the exception that all licensees provide appropriate attribution to you, the copyright holder..
of textbooks and databases) on your own website or in the repository of your university
or institution. In addition, you may place a link to the product page of your Work on our 4 Open Access
website [Link] in public repositories that allow a redirect. You can find a In the case that we have mutually agreed to publish your Work as Open Access, the following
complete list in the Rights and Licenses section of our website. conditions apply in addition:
2.4 VG Wort: As the laws regarding the German collecting society VG Wort are subject to
frequent change, we include here a translation of the recommended wording for contributor
4.1 You are free to publish your Work according to a Creative Commons license ([Link]
agreements provided by the German Publishers and Booksellers Association on exercising
[Link]), as of a date agreed upon with us. You choose the appropriate license
your rights through VG Wort. In the following paragraphs, you assign to us the right to
when discussing the contractual details with us (see Contributor Contract). You have the
exercise our rights with VG Wort. In order for you to profit from the copyright-holder share
choice between:
of the VG Wort distribution, you must sign an agreement directly with VG Wort as well as
a) CC-BY (Attribution)
register your Work upon publication. You may then decide if you would like to let us, as a
b) CC-BY-NC-ND (Attribution-NonCommercial-NoDerivatives).
publisher, participate in the distribution of those proceeds as well. We will provide more
4.2 We will provide you with the final version of your Work as a PDF file. This is the version
detailed information in a separate communication when your Work is published. We retain
that is published Open Access.
the right to make changes to the following paragraphs as a result of court rulings or changes
4.3 We will publish your Work as a freely available eBook on our website under the license you
to the law.
have chosen and which is shown in the Contributor Contract.
The Author grants to the Publisher, for the duration of the contract, all usage rights that may be 4.4 According to the license stated in 6.1, you are of course free to put the eBook version of
exercised at collecting societies such as VG Wort and VG Bild-Kunst according to their contrac- your Work on your own homepage, a university or institutional repository, or any website
tual stipulations. The granting of these rights serves the purpose of registration at the collection of your choice.
societies for mutual exercising of rights. The proceeds from the exercising of rights are distributed
by the collecting societies according to §27 Abs. 2 VGG and according to the shares defined in 5 Manuscript / Content
their distribution plan, and are distributed directly to copyright holders and publishers. The Au-
thor agrees that he or she must conclude an agreement directly with the collecting society in order 5.1 General:
to receive the (copyright-holder) share of the distribution. a) We and the editors are authorized to change your Work after prior discussion with
you, or to ask you to make changes in order to maintain the consistency of the series,
With regard to the statutory reimbursement claims stated in §63a Satz 1 UrhG, the following ap- the volume, or the database.
plies if the Author is registered to exercise rights through the VG Wort: the Author can, according b) In addition, to ensure quality of your Work before final acceptance, we are authori-
to §27a VGG, either upon registration or at a later time following publication, grant consent at the zed to ask a third party to review the Work. Should revisions be necessary, you will
collecting society for the Publisher to participate in the proceeds from the statutory reimbursement be asked to perform these. Should you be unable or unwilling to do so, we reserve
claims. The amount of the publisher share is based on the current distribution plan of the individual the right to deny acceptance or publication of your Work.
collecting society. If the Author is not registered to exercise rights, the Publisher can receive a c) Our customers and readers are interested in the most precise information possible
share of the distribution from collecting societies (publisher share) for the statutory reimbursement regarding your Work. Please inform us immediately as soon as you become aware
claims stated in §63a Sentence 1 UrhG only in the case that the Author has transferred these claims that you will not be able to deliver by the date agreed upon in the Contributor Con-
to the Publisher after the publication of the work. tract or that the manuscript deviates by more than 10% of the length agreed upon.
d) If your Work contains an index, you are required to create an index according to our
3 Grant of Rights technical specifications. If the publication date is delayed by more than 6 weeks as
a result of index preparation, we reserve the right to publish the Work without an
3.1 Rights of Use: You assign to us, for the purpose of ensuring the optimal distribution and index. In the electronic version of the Work, the full-text search replaces the need for
availability of your work the right: a separate index.
a) To reproduce, distribute, and make available your Work in printed form including as e) If you cannot adhere to the manuscript length or the submission deadlines agreed
print-on-demand; upon with you in your Contributor Contract (e. g. delayed delivery of the Work),
b) To produce machine-readable forms of your Work (including digitization) and to we reserve the right to set a new deadline or to plan a different form of publication.
store it electronically in all storage media, including in our own databases and those If the second deadline also cannot be kept, we reserve the right to deny publication
of third parties, such as Amazon or Google; or to reduce your royalty.
c) To make your Work available in part with “look inside” functionality, for example
on Amazon or Google;

Contributor Publication Agreement

Walter de Gruyter GmbH · Genthiner Str. 13 · 10785 Berlin · Deutschland
 8

5.2 Manuscript for Typesetting 9 Remainders

a) This option is for manuscripts that will be professionally typeset. You submit the
Because our warehouse capacities are limited, we regularly check if our printed stock can be re-
manuscript to be typeset as an electronic file, usually in the format docx, tex, rtf,
duced. Through digitisation and print-on-demand, we can keep the majority of our titles available
or indd. Formulas and tables should not be anchored in the manuscript as images.
and in print without keeping physical stock on hand. We reserve the right to pulp or sell any remai-
In addition, we require a PDF file or a definitive printout for comparison including
ning stock at a reduced price.
reproducible copies of illustrations / high-resolution image files.
b) After the manuscript has been prepared by our typesetter, we provide you with pro-
10 Legal Succession
ofs for corrections. In most cases, you will receive proofs in electronic format as a
PDF file or via a web-based online proofing system. We ask for your understanding 10.1 The legal situation must be defined in case of succession. Our obligations and duties accor-
that we generally provide one set of proofs for correction and a second set of proofs ding to the Contributor Contract will transfer for your heirs in full, with the exception of
for your approval for printing (imprimatur). paragraph 11.
c) The first set of proofs is provided for you to check the manuscript conversion to 10.2 In the event of death before the first edition of the Work is completed, we are entitled to
typesetting. Please check these proofs carefully for any mistakes (e. g., word breaks) withdraw from this agreement with no claims from your heirs. We may take over the partial
that may have occurred during the process. Kindly note that only minor content cor- manuscript and have it completed by a third party, taking your personal rights into account.
rections can be done at this late stage. If the index entries were not already anchored
in the manuscript, we will ask you to do this during the first correction stage. 11 Termination
d) The second set of proofs is simply to check that any corrections marked in the first
proof run have been carried out, and for you to provide your approval for printing 11.1 We are entitled to terminate the Contributor Contract in writing with immediate effect if
(imprimatur), if need be under the condition that final corrections be carried out our cooperation on the Work cannot be completed as planned, e. g. if you are unwilling or
before the work is printed. We will check these final corrections internally in order unable to complete the Work. If we have made substantial investments at the time of ter-
to ensure punctual publication of your Work. We retain the right to decide when the mination, all rights assigned remain with us. We reserve the right to reclaim any payments
manuscript is ready to print from a technical and typographical point of view. already made to you.
e) We ask for your cooperation in keeping the number of corrections at a reasonable 11.2 You are also entitled to terminate the agreement in writing with immediate effect, e. g. if we
and necessary level. Changes to the Work after the submission of the final manu- are liquidated. In this case, all rights revert back to you.
script are very costly. We cover the cost of corrections up to two per printed page, 11.3 In the case of termination, we retain the right to continue to distribute any copies of the
and reserve the right to charge you if changes made beyond that result in excessive Work that have already been produced. We may continue to make the electronic version of
costs and/or if these changes require additional print runs. This does not apply if your Work available on our platform in order to fulfil obligations to customers who have
corrections are required due to errors in the typesetting process or due to new acade- purchased those versions.
mic findings or legal decisions. Writing and Submitting via a Content Management 11.4 If your Work is published as part of a series or as a contribution to a volume or database,
System (CMS) we are entitled to terminate the Contributor Contract with immediate effect if the contract
f) Particularly in the case of encyclopedias, dictionaries and reference works, which with the editor expires or is terminated, or if the series or database has been terminated. In
are often published “ahead of print,” we may use a so-called Content Management this case, we will do our utmost to publish your Work as a stand-alone volume or to include
System (CMS), which is individually configured for each Work. it in another series.
g) We provide the CMS to you, and we request that you use it to write and submit your 11.5 You agree to forgo your extraordinary termination option according to German law for
Work. You will receive an invitation to the system and all relevant supporting guide- the non-publication of a contribution in a periodical, e. g. yearbook (§45 Para. 1 Ver-
lines and documentation. lagsgesetz).
h) Reviews by the series or volume editors or in-house editors as well as manuscript
revision and corrections are done directly in the CMS. Please follow the style and 12 Data Protection
content guidelines to avoid unnecessary delays. We are committed to the responsible handling and processing of the personal data we collect from
i) At the relevant point in the process, you will be contacted to provide your approval for you. Details can be found in our data protection policy for authors and editors on our website
publication of the content (not the form). Please make only those corrections which ([Link] A printed copy will be provi-
are absolutely necessary at this time. Changes to content are no longer possible, as ded to you upon request.
they can no longer be reviewed. If you do not respond to the call for approval by the
given deadline, approval will be assumed. 13 General Provisions
j) We support your work in the CMS with detailed instructions and guidelines, and are
personally available if there are problems. Any decisions regarding bug fixes or other 13.1 Unless otherwise agreed upon, all provisions of this agreement are valid and binding for
technical issues will be made solely by us. both your legal successors as well as our legal successors.
13.2 Revisions and amendments to the agreement must be made in writing. Oral agreements and
declarations have not been made.
6 Publication Subsidy / Open Access Fee
13.3 Should individual provisions of this agreement become invalid or unenforceable, this shall
6.1 A publication subsidy may be required for some works to be paid plus VAT if not mutually not affect the validity of the remaining provisions. You and we are bound to replace the in-
agreed in the Contributor Contract. valid provision with a valid one that corresponds to the purpose and meaning of the invalid
6.2 If we have come to a mutual agreement to publish your Work as Open Access, an Open one. The same shall be applicable to any gaps in this agreement.
Access fee is generally required. 13.4 The laws of the Federal Republic of Germany apply to this agreement, particularly the
6.3 The payment of the publication subsidy or the Open Access Fee is generally due upon German Civil Code, German Copyright Law, and German Publishing Law.
publication of the work. The publication subsidy or the Open Access fee is not subject to 13.5 Place of fulfilment and jurisdiction is the location of our headquarters in Berlin, Germany,
reimbursement. to the extent legally possible.
6.4 We will be happy to support you in applying for any outside funding by providing any
necessary calculations or paperwork, and will fulfil funder requirements.
Version last revised: 18.10.2019
7 Free Copies

If you provide us with your e-mail address you can download a PDF of your contribution from our
website. The number of free print copies is agreed upon in the contract concluded with your editor.
You may acquire further non-German copies of your own Work as well as all book titles from the
publisher, including eBooks, at a discount of 30% off the retail price. You are also entitled to a
discount of 20% off the retail price of individual subscriptions and single issues of journals. You
may not sell free copies or copies purchased at a discount.

8 Subsequent Editions, Revised Editions

8.1 A number of factors are relevant when considering a subsequent or revised edition, for
example, the economic feasibility of the work, as well as the topicality or the accuracy of
the scientific findings.
8.2 Should you wish to publish a new edition of your Work, please inform us and we will
review your suggestion in light of the criteria above.
8.3 Should we decide against publishing a new edition of your Work, we will contact you to
discuss alternative options.
8.4 Should we decide to publish a new edition of your Work, we will communicate to your our
decision and negotiate with you a publication plan. You agree to revise the Work according
to the current state of the academic discussion and submit it to us for publication.
8.5 Should you not be willing or able to perform the revisions yourself, you may suggest a third
party to us. We agree not to reject this suggestion without good reason. If you do not wish
to make a suggestion, we are authorized to assign the revision to a third party.
8.6 In the case of revision by one or more third party/parties, we are entitled but not obligated
to continue to use your name in the title of the Work.

Contributor Publication Agreement

Walter de Gruyter GmbH · Genthiner Str. 13 · 10785 Berlin · Deutschland
Maria Mazzoli, Peter Bakker and Verna DeMontigny
Michif mixed verbs: Typologically unusual
word-internal mixing
1 Introduction
The Michif language combines Plains Cree (autoglossonym Nēhiyawēwin) verbs
and Metis French nouns. It also shows the influence of other languages, mainly
Saulteaux and, more recently, English. Michif formed in the first decades of the
19th century in the area of the Red River basin (present-day Manitoba, Canada),
along with the emergence of the Metis, a bicultural nation resulting from the
encounter of French voyageurs and Indigenous women mainly of Ojibwe/Saul-
teaux descent (Bakker 1997). In all its varieties, Michif is severely endangered.
In this paper, the term ‘Michif’ refers only to southern mixed Michif. Today,
mixed Michif is spoken in traditional Metis communities in Manitoba, southern
Saskatchewan, and North Dakota, by around 100–150 speakers between Can-
ada and the USA (Mazzoli 2019).
In this paper, we illustrate the different Michif constructions that permit
word-internal mixing in the verbal domain, and which morphological slots in
the Michif verb template allow insertion of non-Algonquian material. We assess
whether stem-internal mixing has added in any way to the non-lexical Michif
verb morphology, and whether the stem-internal mixing attested in Michif dif-
fers in some way from similar phenomena in other mixed languages or compa-
rable borrowing situations. The paper is structured as follows. In section 2, we
provide a background of Michif as a mixed language. In section 3, we review
the mixed stem types. In section 4, we discuss our data with respect to the
Michif descriptive scholarship and the literature on mixed languages and lan-
guage contact. The conclusions follow in section 5.

Acknowledgements: Maria Mazzoli conducted part of this research with the support of the
University of Bremen and the European Union FP7 COFUND under grant agreement n° 600411.
She thanks Nicole Rosen, Dale McCreery, Dennis Davey, David “Doc” Brian, Lawrie Barkwell,
Harvey Pelletier, and Heather Souter. The authors also thank Marivic Lesho for proofreading
AU: Please
the text and the anonymous reviewers and Eeva Sippola for their helpful comments. provide the
missing affili-
Maria Mazzoli, University of Groningen, [Link]@[Link] ation details
Peter Bakker, Aarhus University, linpb@[Link] for the au-
thor “Verna
Verna DeMontigny
DeMontigny”.

[Link]
122 Maria Mazzoli, Peter Bakker and Verna DeMontigny

The data come from the Michif dictionary by Laverdure and Allard (1983)
and, when the source is not specified, from Mazzoli’s fieldwork corpus (collected
in collaboration especially with Verna DeMontigny). The spelling conventions
adopted here follow those used in Rosen and Souter (2009, 2015). All occurrences
from Laverdure and Allard (1983) are reported in their original form and repro-
duced in the new spelling. In the examples, Michif phrases/morphemes deriving
from French are in italics, while Cree-derived phrases are in plain font. For the
sake of readability, we will use the terms ‘Cree’ and ‘French’ to refer to Michif
morphemes/words or phrases that are in fact derived from Plains Cree or derived
from Metis French.

2 Background: Michif mixed structure

Michif resulted from the introduction of Metis French noun phrases into the
structures of Plains Cree. Structurally, it is classified as a mixed Verb-Noun lan-
guage (Bakker 2017; Meakins and Stewart to appear) that has a mixed lexicon
and, arguably, also a mixed grammar. The great majority of its verbs are etymo-
logically derived from Plains Cree, as well as the personal pronouns and the
demonstratives. Virtually all Michif nouns, adjectives, and articles are etymo-
logically derived from French. Bakker (1997: 137–141) calculated that around
90% of the total number of nouns in Michif are of French origin, while the per-
centage of the Plans Cree verb roots ranges from 86% to 99%. Calculated using
a (noun-dominated) 100-word Swadesh list, 52% of the entries have French roots
and 29% are from Cree, while 19% of the entries have both Cree and French
equivalents (Bakker 2019). However, in order to evaluate the mixed nature of
Michif, it is crucial to assess also how mixed the grammar of Michif is. The three
crucial questions to unpack and illustrate the mixed nature of Michif grammar
are as follows:
1. Is the syntax of the noun phrase mixed?
2. Is the phonology of Michif mixed?
3. Is the morpho-syntax of the verb phrase mixed, or restructured with respect
to Plains Cree due to contact with French?

Issue (1) has been addressed in the recent literature. Gillon and Rosen (2018)
discuss it from a generative point of view and claim that Michif noun phrases
have a mixed syntax, exhibiting mostly Algonquian-derived features and a few
French-derived features (but cf. Thomason 2019). Sammons (2019) identifies
Michif noun phrase structure as typologically unusual with respect to attested
 Michif mixed verbs: Typologically unusual word-internal mixing 123

strategies of borrowing as well as to general systems of noun classification

in the world’s languages. Michif ‘French sex-based’ and ‘Cree animacy-based’
genders are surprisingly equivalent to their sources, with little simplification or
development to arbitrary gender assignment. Even though the functional load
of sex-based gender in Michif appears to be weakened (with respect to French
and with respect to the load of the animacy-based classification), its productiv-
ity is confirmed by the existence of novel English forms with both genders
(Sammons 2019: 231). Michif noun phrase syntax does not belong entirely either
to French or to Cree but rather is mixed fairly evenly and shows an original and
typologically rare set-up.
Issue (2) has received some attention in the recent literature as well. This work
has brought a deeper understanding of the phonology of Michif, given the evident
stratification of Michif lexicon. Papen (2017) recently reaffirmed the claim for a
split phonology for Michif, while Prichard and Shwayder (2014) go against that
claim. Concerning vowels, if Michif speakers operationalize two separate vowel
systems at a time (Cree and French), this would bring the total number of vowels
for Michif to 25. After examining the possible conflicting sites, however, Rosen,
Stewart and Sammons (2020) identify only two French vowels that differ signifi-
cantly from their Cree counterparts (/ɛ, ɔ/), while the rest are assimilated to the
Cree system. According to this analysis, Michif’s vowel system largely reflects that
of Cree. However, their study does not include the nasal vowels.
The morpho-syntax of the Michif verb phrase (issue 3) is another relevant
topic for a global assessment of the mixed grammar of Michif. The Michif verbal
inflectional system is almost equivalent to that of Plains Cree, with very few
modifications (Rhodes 1977; Bakker 1997: 261–263; Gillon and Rosen: 179–185).
The general verb structure and derivational schemes are described by Mazzoli
(accepted), who illustrates a few restructuring processes with respect to Plains
Cree. However, an account of mixed verb formations in Michif is currently miss-
ing in the literature, and the analysis provided in this paper sheds light on the
constraints that originated mixed verbs. Cases of mixed Cree-French verb stems
are relatively rare in Michif (Bakker 1997: 243–244; Antonov 2019). The paucity
of French verbs in Michif has been attributed to the polysynthetic morphologi-
cal type of the Michif verb, which is derived from Plains Cree (Bakker 1997: 246,
2006; Gillon and Rosen 2018: 15). “Polysynthesis” is a complex notion in typo-
logical linguistics (Fortescue, Mithun and Evans 2017). In the case of Michif,
polysynthetic verb formation has to do with: (1) a high ratio of morphemes per
word, (2) polypersonal agreement in the transitive verb, (3) a tendency to have
non-root (bound) lexical affixes within the stem, (4) verb root serialization, and
(5) denominal formations, exclusively from French nouns (discussed in this
paper in section 3.2).
 124 Maria Mazzoli, Peter Bakker and Verna DeMontigny

In Michif, as in Algonquian languages, there are four main verb stem classes
based on the intersection of transitivity and animacy [Link] class requires
different inflectional markers: inanimate intransitive (II), animate intransitive
(AI), transitive inanimate (TI), and transitive animate (TA). To these, one can add
semitransitive verb stems (Dryer 2007), which are semantically (ambiguously)
transitive verbs that are morphologically intransitive but can take objects (abbre-
viated here as AI+O).
Mazzoli (accepted) provides a template in 12 slots for the Michif verb, which
AU: Please can be summarized as follows (cf. Goddard 1990; Wolfart 1996; Bakker 2006):
clarify
whether this
is Wolfart (1) (prefixes) + (PREVERB) + [INITIAL+FINAL] + (DERIVATIVE) + [suffixes]
1996a or
1996b. [STEM]

In this paper, our analysis of mixed verbs adopts the schema in 1, which dis-
tinguishes regarding role, position, and function between ‘stem-internal finals’
(in primary stems) and ‘derivatives which modify stems’ (in secondary stems).
Morpho-phonologically, stem finals and derivatives behave differently: cases of
hiatus are resolved differently within the stem (coalescence) and the stem’s bor-
ders (epenthesis, vowel juxtaposition, cf. section 4.1). In 1, components that are
necessary to form a grammatical verb form are in square brackets, while com-
ponents that are optional are in parentheses. Underlined elements are slots
that can be occupied by non-Algonquian (French and English) nominals. They
are: the primary stem (before derivatives), the initial slot, and the preverb slot.
In section 3, we provide a survey of mixed verbs in Michif.

3 French in the Michif verb

The Michif verb is essentially of Cree origin, but some French verbs are also at-
tested. For instance, one finds French copular verbs (e.g. sete ‘it was’, se ‘it is’,
iyave ‘there was/were’) and a few French stems. This section focuses on how
French lexical material is inserted into a Cree verbal frame, resulting in mixed
stems which are inflected and derived like all the other Michif verbs of Cree origin.
The insertion of French material within Cree inflected verbs is generally
one among several possibilities available to Michif speakers, as in:

(2) li peer ishiikat-ee-w John pii Mari

DEF.M.S [Link] [Link]-DIR-IND.3>4 John and Mari
‘the priest blessed John and Mary (sprinkling holy water)’
 Michif mixed verbs: Typologically unusual word-internal mixing 125

(3) li peer binidiktiooñ kii-miy-ee-w John pii

DEF.M.S [Link] blessing [Link]-DIR-IND.3>4 John and
Mari miina
Mari also
‘the priest gave a blessing to John and Mary (signing the cross with the
hand)’

(4) li peer (li) binii kii-miy-ee-w

DEF.M.S [Link] (the.M.S) blessing.M [Link]-DIR-IND.3>4
lii zaañfaañ
[Link] [Link]
‘the priest gave a blessing to the children (signing the cross with the hand)’

(5) li peer li-binii-w lii zaañfaañ

DEF.M.S [Link] NMLZ-bless-IND.3S [Link] [Link]
‘the priest gave a blessing to the children (signing the cross with the hand)’

In (2), a Cree TA verb is used to denote the act of sprinkling holy water with an
aspergillum; in (3), an analytic form made of the Cree TA verb for ‘to give’ and
the French noun for ‘blessing’ is created; in (4), a similar analytical form is cre-
ated using the Cree verb for ‘ to give’ and the French nominalized infinitive li
binii (lit. ‘the to bless’ < Fr. bénir ‘bless’); and finally, in (5), a mixed verb is
used, consisting of the infinitive French form of the verb ‘bless’ (binii) preceded
by the singular masculine French article li. When mixed and non-mixed forms
are available, speakers’ familiarity with the competing forms varies greatly de-
pending on their location and specific language use within their networks.
Example (4) li binii (lit. ‘the to bless’) and (6) li brodii (lit. ‘the to embroider’)
‘the embroidery’ illustrate a Michif-specific nominalization strategy that forms
nouns out of French infinitive verbs (cf. section 3.1):

(6) nahayhtawwun li brodee chi-oushtawhk

nayeehtaa-wan li brodii chi-osht-aa-hk
articulated-II.3S DEF.M.S embroidery.M [Link]
‘embroidery is intricate’ (lit. ‘it is complicated to make embroidery’)
(Laverdure and Allard 1983: 147)

Mixed verbs with the insertion of non-Algonquian material have been created
in most cases to fill lexical gaps with respect to introduced technologies, cul-
ture, and customs, as in (5) and (7).
126 Maria Mazzoli, Peter Bakker and Verna DeMontigny

(7) bakwaat-ee-n (ee)-li-surfii-t la net tultañ

[Link]-IND.1S (CNJ)-[Link]-CNJ.3S DEF.F.S net.F always
‘I don’t like that s/he is browsing the net all the time’ (< Eng. surf ‘browse
the internet’)

All the mixed stems are intransitive (Antonov 2019). The large majority are animate
intransitive (AI), and just two types among the mixed constructions discussed in
this paper are intransitive inanimate (II) (viz. those with Cree final -payi and deri-
vational suffix -iw-an). In all cases, the mixed stems conform to the inflectional
and derivational paradigm common to the other Michif verbs (but cf. Bakker 1997:
243). The use of conjunct prefixes (8), personal pronominal prefixes (9), preverbs
(10), derivational morphemes (11), obviation and possessed subject markers (12),
person, number, possessed-subject, and animacy-based gender agreements (11),
(12), (13) is regularly attested, and indicated in bold here:

(8) miyeuyiht-am sooñ zhardaeñ ee-zhalii-ihkee-t

[Link].3S 3POSS.M.S [Link] [Link]-CNJ.3S
‘s/he likes to make her/his garden nice’ (< Fr. joli ‘nice’)

(9) nd-aschinii-n avek mooñ mari

[Link]-IND.1S with 1POSS.M.S [Link]
‘I argue with my husband’ (< Fr. obstiner ‘to quarrel’)

(10) ati-la-vyeey-iwi-w
[Link]-IND.3S
‘she is becoming an old woman’ (< Fr. la vieille ‘the old lady’)

(11) bakwaat-ee-n binii-kaashoo-chik li mooñd

[Link]-IND.1S [Link]-CNJ.3PL DEF.S.M people.M
‘I don't like that the people have been blessed’ (< Fr. bénir ‘to bless’)

(12) pooree-iwuniyiw sa zhawmb

puurii-iw-an-iyi-w sa zhaañb
[Link]-II-POSS.S-3S 3POSS.S.F [Link]
‘gangrene set in her leg’ (< Fr. pourri ‘rotten’)
(Laverdure and Allard 1983: 108)

(13) bakwaat-ee-n John pii Mari beegii-chik

[Link]-IND.1S John and Mari [Link]-CNJ.3PL
‘I don't like that John and Mary are stuttering’ (< Fr. bégayer ‘to stutter’)
 Michif mixed verbs: Typologically unusual word-internal mixing 127

In general, either nouns or nominalised verbs are inserted into the Cree frame,
but on a limited number of occasions, (etymologically) prepositional phrases
are attested, introduced by the French article li. The French expression en retard
‘delayed’ (in bold in 14) is undoubtedly monomorphemic for Michif speakers.

(14) keemaekouhow ay-li awn rtawr-iwit awsha meena

kii-maakoh-oo-w ee-li-aañrtaar-iwi-t
[Link]-IND.3S [Link]-CNJ.3S
aasha miina
already again
‘s/he was given a reprimand for being late again’ (< Fr. en retard ‘late’)
(Laverdure and Allard 1983: 263)

There are four types of constructions that host French material:

a) French verbs nominalized in their infinitive form, and inflected as AI+O,
e.g. li-brodii-w ‘s/he embroiders’ (see section 3.1).
b) Denominal verb formation, where French nouns or nominalized adjectives
are combined with the Cree derivatives, e.g. la-maegr-iwi-w ‘s/he is skinny’
(see section 3.2).
c) French nominals or nominalized adjectives combined with Cree AI finals,
e.g. li-paapaa-i-payi-w (see section 3.3).
d) French nominals in the preverbal slot (see section 3.4).

The great majority of the occurrences in our data pertain to either group (a) or (b)
(section 3.1 and section 3.2). Occurrences of types (c) and (d) are less frequent.

3.1 The frame li-X-ii: French nominalized infinitives

as AI+O Michif stems
French verbs are integrated in their infinitive forms as full Michif primary stems
that can be further subjected to secondary derivation (cf. [11] and [18]–[21]).
They are conjugated as animate intransitive verbs, irrespective of their original
transitivity (15), (16), (17). The French stems invariably appear as infinitives,
where the final long vowel -ii corresponds to the French infinitive marker -er,
as noted by Bakker (1997: 242). In the examples in this section, we gloss the
relics of the infinitive as ‘INF’. In section 4.1, an alternative glossing will be pro-
posed (‘AI’ animate intransitive). In addition, infinitive French forms are intro-
duced by what is etymologically the masculine singular article li-, which marks
128 Maria Mazzoli, Peter Bakker and Verna DeMontigny

nominalization and therefore is glossed as NMLZ. The nominalizer is occasion-

ally dropped, e.g. in (17), (26).1

(15) bakwaat-ee-n ee-li-brod-ii-yaan

[Link]-IND.1S [Link].1S
‘I hate embroidering’
brodii AI< Fr. broder ‘to embroider’, ambitransitive

(16) kii-li-sup-ii-naan
[Link]
‘we (INCL) had supper’
supii AI < Fr. souper ‘to eat supper’, ambitransitive

(17) sapraañ kiihtwam chi-arañzh-ii-kaasho-t,

[Link] again [Link]-CNJ.3S
chi-li-arañzh-ii-yahk
[Link]
‘this has to be rearranged, we (INCL) have to rearrange it’
arañzhii AI+O < Fr. arranger ‘to arrange’, transitive

In (20), we observe an exception to the usual pattern, also identified in Antonov

(2019: 33). The stem dooñt- ‘to tame’ (<Fr. dompter) is introduced by the nomi-
nalizer and ends in -ii, as in the other cases. However, the nominalized infini-
tive li-dooñt-ii is conjugated as a transitive animate verb in (18), and it takes
direction marking (19), as well as passive (21) and medio-passive derivations
(20). This is because the nominalized infinitive stem is further derived into a
transitive verb through the derivational transitivizing suffix -h, which attaches
regularly to animate intransitive stems in Cree and Michif (cf. Mazzoli accepted,
here in section 4.4 and section 4.5).2

1 Li is also dropped in other contexts, especially before nominalized adjectives in denominal

constructions, e.g. la farin mweeñzii-iwi-w (< Fr. moisi) ‘the flour is moldy’; li log purii-iwi-w
‘the log is rotten’; tooñ kapoo purii-iw-an ‘your coat is rotten’.
2 A further example of a French infinitive conjugated as a TA is given in Antonov (2019: 33),
taken from the GDI online Michif dictionary: biniiwahwaaw (original spelling: biniiwahwow),
‘to bless someone’, where the TA final -ahw- ‘do by tool, external medium’ is used. However,
this seems to be an isolated occurrence.
 Michif mixed verbs: Typologically unusual word-internal mixing 129

(18) lidoonteehik tee zhvoo

li-dooñt-ii-h-ik tii zhvoo
NMLZ-tame-INF-TR-IMP.2S>3PL [Link] horses
‘domesticate your horses’
(Laverdure and Allard 1983: 78)

(19) li-dooñt-ii-h-ee-w lii zhvoo

NMLZ-tame-INF-TR-DIR-IND.3>4 [Link] horses
‘s/he tames horses’

(20) lii zhvoo li-dooñt-ii-h-aa-w-ak

[Link] horses [Link]-IND.3-PL
‘the horses were tamed’ (lit. ‘they tame the horses’)

(21) li-dooñt-ii-h-ikaasho-w li zhval

[Link]-IND.3S DEF.M.S horse
‘the horse has been tamed’

Example (26) illustrates the use of the English verb stem beeg ‘to beg’, and
highlights the emergence of a specific pattern:

(22) sartaeñ ga-waapam-aa-naan (ee)ka-chi-li-beeg-ii-t

certain [Link]-DIR-IND.1PL>3 [Link].3S
pur larzhaañ
for money
‘we (EXCL) will make sure that he will not beg for money’
beeg-ii AI < beeg < Eng. beg, transitive

The strategy for forming Michif AI stems from nominalized French infinitives
becomes a frame to accommodate loan verbs from English. The abstract frame
is given in (23), and below we give further examples with the English verbs can,
surf, collect and rob (24)–(27) (the French frame is in italics, English loans in
bold, and Cree in plain font).

(23) prefixes – li – English verb – ii – AI endings

130 Maria Mazzoli, Peter Bakker and Verna DeMontigny

(24) bakwawtaen shi-li can-iyawn daw li tawn payzawn

bakwaat-ee-n chi-li-kan-ii-yaan daañ li
[Link]-IND.1S [Link].1S in DEF.M.S
taañ peezaañ
weather heavy
‘I don't like to can in humid weather’
(Laverdure and Allard 1983: 137)
kan-ii AI < kan < Eng. can ‘preserve food’, transitive

(25) li-surf-ii-w li net

[Link].3S DEF.M.S net.M
‘s/he is surfing the net’
(Gillon and Rosen 2018: 129)
surf-ii AI < surf < Eng. surf, ambitransitive

(26) saprawn chikalekteeyen wiya ouhchi

sapraañ chi-kalekt-ii-yen wiiya ohchi
needed [Link].2S him/her from
‘you have to collect (wages) from her’
(Laverdure and Allard 1983: 64)
kalekt-ii AI < kalekt < Eng. collect, transitive

(27) noo keekishkayistam la bawnk aykee-li robeehk

noo kii-kishkeeyiht-am la baañk
NEG [Link].3S DEF.F.S [Link]
ee-kii-li-rob-ii-hk
[Link]
‘s/he was unaware of the robbery at the bank’
(Laverdure and Allard 1983: 143, Papen 2005: 78)
rob-ii AI < rob < Eng. rob, transitive

Examining the loan verbs from English, it is evident that the frame in (23) has
grammaticalized as the pattern for verbal borrowing. In section 4.1, we discuss
in more detail the role of li- and -ii in this emerging frame.
 Michif mixed verbs: Typologically unusual word-internal mixing 131

3.2 The bulk of Michif mixed verb stems:

Denominal verb formation
Michif denominal verb constructions combine French nouns and the Cree deriv-
atives -iwi ‘to be’, -hkee ‘to make’, and -hkaashoo ‘to pretend’. Such constructions
are sometimes analyzed in the literature as cases of (light verb) noun incorpo-
ration (Gerdts and Marlett 2008, Barrie and Mathieu 2016). We will discuss how
Michif denominals relate to regular noun incorporation (henceforth NI), and to
English-like denominal verb formation realized through conversion or category-
changing derivatives (e.g. in English, operation > operationalize). Then, in section
4.1, we will discuss how Michif denominal verb formation fits into the general
structure of the Michif verb that we presented in (1).

3.2.1 Mixed verbs with French nominals and Cree -iwi ‘to be’

Animate intransitive forms with the Cree suffix -iwi ‘[Link]’ cover the copular
meanings of attribution (‘s/he is X-y’), ascription (‘s/he is an X’), and identifica-
tion (‘s/he is X’). In Michif, -iwi combines with French (or English) noun
phrases or nominalized adjectives only (29–32). Occurrences such as (28), with
a Cree nominal, are almost unattested in Michif (cf. Laverdure and Allard 1983:
227, koohkooshiwi- ‘be a pig’, from Cree kôhkôs).

(28) mahîhkan-iwi-w
[Link]-IND.3S
‘s/he is a wolf’
(Plains Cree, Wolfart 1996: 428) AU: Please
clarify
whether this
(29) li pawpaw-iwiw pour see zawnfawn is Wolfart
1996a or
li-paapaa-iwi-w pur sii zaañfaañ 1996b.
[Link]-IND.3S for [Link] children
‘he is the father of her children’
(Laverdure and Allard 1983: 92)

(30) kaa-li-beebii-iwi-yaan
[Link]-CNJ.1S
‘when I was a baby’
132 Maria Mazzoli, Peter Bakker and Verna DeMontigny

In denominal verb formation, the nouns appear with articles featuring sex-based
gender (31–32) and the definite-indefinite distinction (33–34):

(31) no li kawnsayr-iwiw li toomer

no li-kaanseer-iwi-w li tumer
NEG [Link]-IND.3S the.M.S [Link]
‘the tumor is benign/not cancerous’ (< Fr. le cancer ‘the tumor’)
(Laverdure and Allard 1983: 40)

(32) la shayayr awn fayr ka-la rouy-iwiw

la shayeer aañ feer ka-la-ruy-iwi-w
the.F.S pail.F of iron [Link]-IND.3S
‘the metal pail will rust’ (< Fr. la rouille ‘the rust’)
(Laverdure and Allard 1983: 278)

However, the use of a definite or an indefinite article and aeñ ~ en (masculine

and feminine, respectively) is not indicative of a certain semantics, out of con-
text. For instance, (33) and (34) are considered equivalent by DeMontigny, both
meaning ‘Lorna is an old lady’.

(33) Lorna la-vyeey-iwi-w

Lorna [Link]-IND.3S
(< Fr. la vieille ‘the old woman’)

(34) Lorna en-vyeey-iwi-w

Lorna [Link]-IND.3S
(< Fr. une vieille ‘an old woman’)

In addition, occurrences like (35) and (36) feature an indefinite concept but use
li- to introduce the noun, so that the meaning of the determiner of the inte-
grated noun phrase is weakened:

(35) noo gawmyeustaen ka-li besoon-oowiyawn

no gaa-miyeuyiht-ee-n kaa-li-besooñ-iwi-yaan
NEG [Link]-IND.1S [Link]-CNJ.1S
‘I wouldn’t like to be a twin’ (< Acadian Fr. le besson ‘the twin’)
(Laverdure and Allard 1983: 341)
 Michif mixed verbs: Typologically unusual word-internal mixing 133

(36) li bek seukree ayow [. . .] apray aykee-li meshawn-iwit

li bek seukree ayaa-w apree
DEF.M.S beak.M sugared [Link]-3S after
ee-kii-li-meshaañ-iwi-t
[Link]-CNJ.3S
‘he has a sweet voice for his wife after being mean’ (< Fr. adj. méchant
‘mean’)
(Laverdure and Allard 1983: 175)

3.2.2 Mixed verbs with French nominals and Cree -iw - an ‘to be’

Inanimate intransitive forms with the II suffix -iw-an ‘[Link]-II.0’ cover the copu-
lar meaning attributive (‘it is X-y’), as in (37) and (38):

(37) la broo-iwun li laek

la-broo-iw-an li laek
[Link]-II.0 DEF.M.S [Link]
‘there is froth on the lake’ (< Fr. ?)
(Laverdure and Allard 1983: 105)

(38) la brem-iwun
la-brem-iw-an
[Link]-II.0
‘it is foggy’ (< Fr. la brume ‘the fog’)
(Laverdure and Allard 1983: 99)

As in the case of -iwi, the article introducing the non-Algonquian material is

always singular, and it agrees in sex-based gender with the noun it introduces.

3.2.3 Mixed verbs with French nominals and Cree -(i)hkee ‘to make’

In Michif, the suffix -(i)hkee~aa meaning ‘to make’ combines with French nouns
(39)–(40) or nominalized adjectives (41) to make AI stems:
134 Maria Mazzoli, Peter Bakker and Verna DeMontigny

(39) meena ka-li boudaen-ihkayw

miina ka-li-budaeñ-hkee-w
again [Link]-IND.3S
‘s/he’ll sulk again’ (< Fr. le boudain ‘the blood sausage’; idiomatic ex-
pression)
(Laverdure and Allard 1983: 321)

(40) ga-li-shmaeñ-hkaa-n
[Link]-IND.1S
‘I’ll build a road’
(adapted from Rosen 2007: 205)

(41) miyeustum a lawntoor sa maizoon ay-li zhalee-ihkayt

miyeuht-am alaañtuur sa meezooñ
[Link].3S around 3POSS.F.S [Link]
ee-li-zhalii-hkee-t
[Link]-CNJ.3S
‘s/he likes to beautify her yard’
(Laverdure and Allard 1983: 38)

The introducing article agrees in sex-based gender with the noun, and an ep-
enthetic -i- surfaces after consonant-final nouns like bet in (42):

(42) la-bet-ihkee-w
[Link]-IND.3S
‘she is making silly things’ (< Fr. la bête ‘the silly, stupid woman’)

The suffix -(i)hkee is not attested with Cree nouns (43):

(43) o-maskisin-ihkee-w
[Link]-3S
‘he/she makes shoes’ (< Plains Cree maskisin ‘shoe’)

3.2.4 Mixed verbs with French nominals and Cree -(i)hkaashoo ‘to pretend’

The Cree element -(i)hkaashoo means ‘to pretend, to make oneself’ and is com-
mon in Michif with Cree verb stems:
 Michif mixed verbs: Typologically unusual word-internal mixing 135

(44) paahpi-hkaashoo-w
[Link]-IND.3S
‘s/he pretends to laugh’

Most commonly, however, -(i)hkaashoo combines with French or English nomi-

nalized adjectives or nouns preceded by a definite article (45), and occasionally
with French infinitives (46):

(45) li-dzhaab-ihkaashoo-w
[Link]-IND.3S
‘s/he is acting like the/a devil’ (< Fr. le diable ‘the devil’)

(46) aschin-ii-hkaashoo-w
[Link]-IND.3S
‘s/he is pretending to argue’ (< Fr. obstiner ‘to quarrel’)

In the above examples of denominal formations, we have seen some French ad-
jectives used in mixed stems (36), (41)–(42). Adjectives are not incorporated in
mixed constructions, but they are used in nominalized constructions in both
genders (47)–(48). The French integrated into the Cree verb in Michif has the
noun phrase as its base.

(47) gakway-la maegriwin

gaakway-la-maegr-iwi-n
[Link]-IND.1S
‘I’m trying to slim down’ (female subject)’ (< Fr. maigre ‘slim’)
(Laverdure and Allard 1983: 257)

(48) mitouni la prop-iwiw

mitoni la-prop-iwi-w
very [Link]-IND.3S
‘she is very proper’ (< Fr. propre ‘clean’)
(Laverdure and Allard 1983: 243)

As far as gender is concerned, we have seen that nouns and nominalized adjec-
tives agree in sex-based gender with their article. The article and the noun’s mas-
culine/feminine gender may also appear to agree with the subject’s gender (47)–
(48). However, when the subject is coded for non-natural sex-based gender, as in
136 Maria Mazzoli, Peter Bakker and Verna DeMontigny

(49), la rosh ‘the stone’, and (50), ta shmiiz ‘your shirt’, the agreement between
the French article and the subject is blocked:

(49) li-purii-iwi-w la rosh

[Link]-IND.3S DEF.F.S [Link]
‘the rock is rotten’
* la-purii-iwi-w la rosh

(50) li-purii-iw-an ta shmiiz

[Link]-II.0 2POSS.S.F [Link]
‘your shirt is rotten’
* la-purii-iw-an ta shmiiz

3.2.5 Denominal verb formation and noun incorporation

We will now discuss how Michif formations with -iwi, -hkee~aa, and -hkaashoo
compare to denominal verbs with category-changing affixes and noun incorpo-
ration (NI). Denominal verb formation takes different shapes in the world’s lan-
guages. In English (Clark and Clark 1979), there is a productive process of zero
derivation that converts nouns into verbs, in a way that the meaning of the
event is related to the noun in a variety of ways, and sometimes unpredictably.
English, as well as other languages, also uses category-changing affixes like de-,
-ize, -ify, and -ate. Speakers interpret affixes in denominal verbs in terms of se-
mantic categories that permit redefinition of the internal semantics of the noun
as a verb (Gottfurcht 2008). In Michif, denominal verbs appear in construction-
specific environments, and they are never simple conversions. There are specific
verbalizing morphemes that form a single word with the noun. The suffixed lexi-
cally intransitive verbs in the form of affixes are attached to nouns that are their
predicative or direct arguments. A predictable meaning is associated with the
construction in most cases.
Denominal verb formation in Michif can also be compared to Noun Incorpo-
ration (NI). Mattissen (2017) and Jacques (2012: 1208) propose a strict definition of
NI as the combination of a noun and a verb, on the condition that (1) the resulting
construction constitutes a morphological and phonological word, which occurs in
finite forms; (2) both the nominal and the verbal roots in question exist also as in-
dependent stems; and (3) the combination represents a productive pattern that is
optional, with a non-idiomatic compositional meaning. However, incorporating
languages show varying degrees of adherence to the definition (Fortescue, Mithun
and Evans 2017). In the strict sense, NI does not occur in Michif, while it does
 Michif mixed verbs: Typologically unusual word-internal mixing 137

occur in Plains Cree (Mellow 1990), albeit with limited productivity. In fact, in
Michif, it is not possible to incorporate any French noun into a verbal Cree struc-
ture that includes an otherwise autonomous Cree verb stem. However, denominal
verbs in many Algonquian languages have been treated as cases of NI (Mathieu
2013; Barrie and Mathieu 2016; Johns 2017 for Inuktikut), where the verbalizing suf-
fixes function as light verbs that incorporate noun phrases. Although the Michif
verbalizing suffixes in (29)–(50) do not occur as autonomous verb stems, the de-
nominal constructions do meet the other requirements for NI. Michif denominal
constructions constitute morphological words, since the noun is integrated within
the Cree verb template. The combination is a productive pattern, as demonstrated
by the use of French and English noun phrases, and its semantics is compositional
(with the exception of a few idiomatic expressions, cf. [39]).
In summary, denominal verbs in Michif appear to share some features with
NI, with the exception that Michif verbalizing suffixes are not independent verb
stems. Verbalizing suffixes like -iwi, -iw-an, -(i)kee, and -(i)hkaashoo function
like light verbs that incorporate French noun phrases. Further research should
explore the referentiality and (stranded) modifiability of the noun phrase in
Michif denominal constructions, which are common attributes of incorporated
nouns (Smit 2005).

3.3 French nominals combined with Cree finals

The Cree elements discussed in this section are identified as finals because they
are found in Michif mainly in combination with Cree initials, differently from
the denominal suffixes discussed in section 3.2.

3.3.1 French nominals with concrete finals of perception, AI (-inaakoshi)

and II (-inaakw-an)

The morphemes -inaakoshi ‘appears, looks like’ and -inaakwa ‘appears, looks
like’ are intransitive finals of perception related to sight. They probably consist
of a final related to vision and the eye, -n-, and a derivational morpheme of
‘middle attributive predication’, -aakoshi (AI) or -aakw-an (II). In Michif they
combine with French elements:

(51) toolee-naakoshi-w
[Link]-IND.3S
‘s/he is/looks ugly’ (< Fr. tout laid, lit. ‘very ugly’)
138 Maria Mazzoli, Peter Bakker and Verna DeMontigny

(52) toolee-naakw-an
[Link]-II.0
‘it is/looks ugly’

(53) dilet-inaakw-an
[Link]-II.0
‘it looks milky’ (< Fr. du lait ‘some milk’)
(Laverdure and Allard 1983: 177)

These finals also commonly combine with Cree initials (e.g. ishinaakoshiw AI
‘s/he looks thus’; shoohkinaakoshiw AI ‘s/he looks strong’; miyonaakwan VII ‘it
is beautiful, it looks good’, teepinaakwan VII ‘it is still seen (in the distance)’, ‘it
is still in sight’).

3.3.2 French nominals combine with Cree concrete final -payi ‘[Link]’

The morpheme -payi is a concrete final in Plains Cree and Michif that combines
with a great variety of initials to form intransitive stems. As a stem-internal
component, -payi has fuzzy semantics related to ‘movement’, ‘change of state’,
and ‘unexpected/sudden event’ (54):

(54) li paeñ ohpi-payi-w

DEF.M.S [Link] [Link]-IND.3S
‘the bread leavens’

In Michif, unlike in Cree, -payi combines with a group of French nouns and
nominalized adjectives introduced by li. There is an epenthetic vowel i between
the consonant-final French material and the final -payi:

(55) li simawn ka-li jeur-ipayin

li simaañ ka-li-dzheur-i-payi-n
DEF.M.S [Link] [Link]-II.0
‘the concrete will harden’
(Laverdure and Allard 1983: 125)

(56) li zhardaeñ ka-li-zhalii-payi-n

DEF.M.S [Link] [Link]-II.0
‘the garden will get nice’
 Michif mixed verbs: Typologically unusual word-internal mixing 139

(57) liveer-i-payi-n
[Link]-II.0
‘it will get winter’

(58) mooñ freer kaa-li-paapaa-i-payi-w

1POSS.M.S [Link] [Link]-IND.3S
‘my brother is going to be a dad’

In the Michif mixed constructions, the semantics of the final -payi are fully inter-
pretable, and -payi functions as a kind of light verb, incorporating noun phrases
(similarly to the ones in section 3.2).

3.3.3 French nominalized adjectives combine with Cree -(i)shtikwaan-ee

‘[Link]’

The Michif element -(i)shtikwaan-ee/aa ‘to have a head like’ combines the Cree
medial -(i)shtikwaan and the AI final -ee/aa. It is used in Michif in a limited
number of occurrences where the stems are entirely from Cree (59)–(60):

(59) lee wawpishtikwawn

lii waap-ishtikwaan
[Link] light-head
‘the blond people’
(Laverdure and Allard 1983: 45)

(60) machi-shtikwaan-ee-w
bad-head-AI-IND.3S
‘s/he is stubborn’

In a few occurrences, it combines with French elements:

(61) aen kanawr kaw-li vayr-ishtikwawnayt

aeñ kanaar kaa-li-veer-i-shtikwaan-ee-t
INDEF.M.S [Link] CNJ-the.M.S-green-EPEN-head-AI-CNJ.3S
‘a duck that has a green head, a mallard’
(Laverdure and Allard 1983: 170, Bakker 1997: 245)
140 Maria Mazzoli, Peter Bakker and Verna DeMontigny

Echoing the example in (61), DeMontigny proposed (62), commenting that this
is a word she guesses “one could use”, in confirmation of the productive nature
of the word formation process:

(62) ruzh-i-shtikwaan-ee-w
red-EPEN-head-AI-IND.3S
‘s/he has a red head’

3.3.4 French nominals combine with the Cree abstract final -ishi

The morpheme -ishiis a Cree abstract final used to form AI verb stems. This final
is widespread in Michif with Cree initials:

(63) noohtay mashkawishiw

noohtee-mashkaw-ishi-w
[Link]-IND.3S
‘s/he wants to be strong’
(Laverdure and Allard 1983: 319, Bakker 1997: 240)

(64) gishkishin kaw-apisheeshishiyawn

gishk-ishi-n kaa-apishiish-ishi-yaan
[Link]-AI-IND.1S [Link]-CNJ.1S
‘I remember my childhood’
(Laverdure and Allard 1983: 61)

In Laverdure and Allard (1983), a few instances are attested of AI verbs com-
posed by the stem megr and the AI final -ishi:

(65) maegrashin
megr-ishi-n
[Link]-IND.3S
‘s/he is skinny’ (< Fr. maigre ‘skinny’)
Laverdure and Allard (1983: 23)

However, DeMontigny finds this final unacceptable with French nouns or nomi-
nalized adjectives. DeMontigny prefers meg-iwi-w, li meg, or kaawaakatishi-w to
(65), and she considers (66)–(71) equally unacceptable:
 Michif mixed verbs: Typologically unusual word-internal mixing 141

(66) * labet-ishi-w
(intended) ‘she is stupid’

(67) * liboss-ishi-w
(intended) ‘he is bossy’

(68) * libet-ishi-w
(intended) ‘he is stupid’

(69) * libet-ishi-n
(intended) ‘I/you am/are stupid’

(70) * lametres-ishi-w
(intended) ‘she is a teacher’

(71) ?? lipovr-ishi-w
(intended) ‘s/he is poor’

Therefore, in Michif, -ishi can be combined with Cree initials, but not with French
noun phrases.

3.4 French used in the preverbal position

The adverbial specification of manner in aeñ kwashooñ ‘a pig’ or kwashooñ ‘pig’

is taken from the expression kom aeñ kwashooñ ‘like a pig’, shortened, and used
in some Michif occurrences to modify a verb. These constructions are very rare.
In (72), kwashooñ follows the personal pronoun prefixes for speech act partici-
pants in the independent order. It is also suffixed with -i either to avoid the con-
sonant cluster or to form a preverb, as in the analysis by Mellow (1990), who
referred to the equivalent Plains Cree structure as ‘lexical noun incorporation’:3

(72) ki-kwashooñ-i-miichisho-n
[Link]-IND.2S
‘you eat like a pig, greedily’

3 kohkoos-i-miitiso-w ([Link]-IND.3S) ‘s/he eats like a pig’ (Mellow 1990: 248).

142 Maria Mazzoli, Peter Bakker and Verna DeMontigny

In a similar fashion, meelii- (< Fr. mêler ‘to mix’) is used in (73) following the
future marker ka-, and ipee- (< Fr. épaix ‘thick’) in (74):4

(73) ka-meelii-oshihtaa-n li morkii?

[Link]-IND.2S DEF.M.S [Link]
‘will you mix the plaster?’
(Laverdure and Allard 1983: 232)

(74) ipay-oushtaw la soup aen pchi braen

ipee-oshihtaa la sup aeñ pchi braeñ
[Link] DEF.F.S [Link] INDEF.M.S little bit
‘thicken the soup a little’
(Laverdure and Allard 1983: 331)

3.5 Summary

In Michif, a number of French nouns and nominalized adjectives are used in a

Cree verbal frame. French infinitives are used as primary verb stems (section 3.1),
and nominals can be derived into verbs through verbalizing suffixes that func-
tion like incorporating light verbs (section 3.2). On rarer occasions, French
nominals are used as initials in combination with Cree finals (section 3.3) or as
modifiers in a preverbal position (section 3.4) (cf. the general structure of the
Michif verb presented in [1]).

4 Discussion
Michif allows several forms of word-internal mixing within the verb. We will
now discuss the data presented in section 3 in the light of the descriptive litera-
ture on Michif and other Algonquian languages, and in relation to the literature
on language contact.

4 DeMontigny points out that this form is unknown to her. In her family, they would use kish-
pakin-a la sup.
 Michif mixed verbs: Typologically unusual word-internal mixing 143

4.1 Michif mixed verbs and their internal structure

4.1.1 Michif mixed forms and the general structure of the Michif verb

The French infinitive marker -er is the source of the long vowel -ii in occurrences
like li-binii (DEF-bless, ‘to bless’) and li-aranzhii (DEF-arrange, ‘to arrange’),
which are AI+O verb stems in Michif (section 3.1). The stem is a nominalized in-
finitive, which is a Michif-specific strategy not possible in French. The analysis of
English loans such as those in (24)–(27) raised the issue of the status of the long
vowel -ii before the inflectional AI endings of the mixed verbs. Given that an ex-
haustive list of Plains Cree/Michif stem-internal finals is not currently available,
it is unclear whether -ii is a productive Cree AI final.5 Well-known Plains Cree
abstract AI finals do include alternant -ê/â (e.g. piikishkw-ê-w, biikiskw-â-n, ‘s/he
speaks, I speak’), -â (e.g. nip-â-w, ‘s/he sleeps’), -i (e.g. wâp-i-w, ‘s/he has sight’),
and -ô (e.g. kit-ô-w, ‘s/he calls’). Cowan (1990: 846) mentions an AI final in -ii for
Plains Cree. The online Michif dictionary (Rosen 2019) lists fourteen AI verbs
with stems ending in -ii, although it is not clear whether the final long vowel
should be considered a category-defining final that assigns features of transitivity
and animacy to the stems. Bloomfield (1984)’s Cree dictionary lists 87 verbs end-
ing in -îw, all of them intransitive, and only eight of them animate. In Michif
mixed stems, -ii- could be a non-Cree-derived final, born out of the reanalysis of
the French infinitive -er, and functioning productively to form AI stems with
non-Algonquian-derived elements, nominalized through the nominalizer li- (cf.
also Gillon and Rosen 2018: 131–132).
French nominalized infinitives in these mixed formations are always trans-
posed and inflected as animate intransitives (AI) in Michif. We confirmed this
through eliciting conjunct inflected forms for each of the examples, which re-
sulted in all the forms consistently inflected as AI. The reason why speakers inter-
pret and inflect nominalized non-Algonquian infinitives as animate intransitive
stems could be related to either:
1) The ease of interpreting the final long vowel of the French infinitives -ii
into an AI final, due to its similarity to other Cree AI abstract finals consist-
ing of vowels (i.e. -ee, -ee/aa, -aa, -oo, -i).
2) The versatility of the AI verb inflection paradigm, given that in Cree and
Michif a group of morphologically AI verbs can be syntactically transitive
and take patients (Oxford 2017: 29).

5 Jeffrey Muehlbauer made an excellent survey of them: [Link]

A_morpheme_index_of_Plains_Cree
144 Maria Mazzoli, Peter Bakker and Verna DeMontigny

3) The rich available paradigm for deriving transitives from intransitive stems
(including the causative/valency increaser -h and other morphemes listed
in Figure 3 in Mazzoli accepted; cf. also Bakker 2006).

In the case of the French denominal verb formation presented in section 3.2 (e.g.,
li-paapaa-iwi-w, [29]), the analysis of hiatus resolution between the French noun
phrase and verbalizing suffixes like -iwi and -iw-an (‘to be’) indicate their relative
positions with respect to the verb template in (1). In Michif and Cree, vowel hia-
tus is tolerated at the stem boundaries with the preverbs (to the left) and the de-
rivatives (to the right). When the two components are internal to the stem, hiatus
must be resolved through assimilation. This is consistent with other Algonquian
languages, e.g. Oji-Cree (Slavin 2012: 242) and Ojibwe (Mathieu 2013). In the
Michif examples in (75) and (76), the hiatus is not resolved, and we observe
vowel juxtaposition between the vowel-final preverb miyo- and the vowel-initial
verb stems ayaa- ‘to have’ and aapachi- ‘to use’. In both occurrences, miyo-
‘good’ is used as stem-external preverb that modifies primary verb stems:

(75) Bachees miyou-ayow

Bachiis miyo-ay-aa-w
Bachiis good-be-IND.3S
‘John is in good health’
(Laverdure and Allard 1983: 354)

(76) miyou-awpachihikawshoow
miyo-aapachih-ikaashoo-w
[Link]-IND.3S
‘they take good advantage of him/her’ (lit. ‘s/he is taken good advantage of’)
(Laverdure and Allard 1983: 18)

In (77) and (78), the hiatus is resolved through assimilation, and therefore the
stem-internal initial miyo- assimilates with the finals -ishi and -ee(yi)ht to form
the stems miyoshi- ‘to be good’ and miyeuht- ‘to like’:

(77) miyoshiw
miyo-ishi-w
[Link]-IND.3S
‘s/he is good, s/he is nice’
(Rosen 2019)
 Michif mixed verbs: Typologically unusual word-internal mixing 145

(78) miyeuht-aakwan kaa-kiishoshi-hk

miyo-ee(yi)ht-aakw-an kaa-kiishoshi-hk
[Link]-II.0 [Link]
‘Ah! It feels good to be warm’
(Laverdure and Allard 1983: 19)

In denominal formations, if the combination between the French noun phrase

and Cree verbalizing suffix resolves hiatus through assimilation, it indicates
that the two components merge to form a new stem (i.e. they occupy the posi-
tions of initial and final). If French noun phrases and Cree verbalizing suffixes
do not resolve the hiatus, then their boundary is not stem-internal.
The data on Michif denominal verbs show consistently that adjacent noun
phrases and verbalizing suffixes tolerate hiatus and just show vowel juxtaposi-
tion (see [29]–[30], [36]–[37]). Therefore, the boundary in question patterns like
a stem-external boundary rather than a stem-internal one: Michif denominal
verbalizing suffixes, or light verbs, pattern like derivatives (and not finals) that
apply to noun phrases used as primary stems.
In the cases described in section 3.3 of French nominals combining with
Cree finals, the French material is used as an initial component in internally-
mixed stems. Nominals used in the preverbal position appear to be inserted
into a stem-external position to the left of the stem.
In sum, only selected slots in the general Michif template allow for inser-
tion of non-Algonquian material, and these are either (1) the primary stem, (2)
the initial slot, or (3) the preverbal position.

4.1.2 French denominal verb formation: What noun phrase structure

is incorporated?

The Michif mixed denominal constructions described in section 3.2 are special
because the incorporated noun requires the presence of the article, which varies
both in sex-based gender and definiteness:

(79) maachi-li-vyeu-iwi-w
[Link]-IND.3S
‘he is getting old’
* maachi-vyeu-iwi-w
146 Maria Mazzoli, Peter Bakker and Verna DeMontigny

(80) ni-maachi-la-vyeey-iwi-n
[Link]-IND.1S
‘I am getting old’ (female speaker)

(81) en-vyeey-hkaasho-w
[Link]-IND.3S
‘[somebody] is pretending to be an old lady’

The article li is used to nominalize and incorporate verbs or adjectives, some-

times in quite complex forms:

(82) la galet li vayr di gree-iwiw

la galet li-veer-di-grii-iwi-w
DEF.F.S [Link] [Link]-IND.3S
‘the bread is moldy’
(Laverdure and Allard 1983: 181)

(83) la parsonn li pchee megr feb inawkoushiw

la parson li-pchii-megr-feb-inaakoshi-w
DEF.F.S [Link] [Link]-IND.3S
‘the person looks puny’
(Laverdure and Allard 1983: 245)

In other Algonquian languages, the noun phrases incorporated in similar for-

mations require some structure. Mathieu (2013) argues that Ojibwe incorporated
nominals are complex noun phrases that potentially include singular number
marking, diminutive and pejorative morphology, and qualifying adjectives as
well as possessives. In Michif, there are more restrictions on the specific nomi-
nal structure that can be incorporated into a verb. Definite articles do not de-
note definiteness of the incorporated noun phrase, as shown in (33)–(36).
Animacy features of the noun may be retained, but there is no superficial
agreement with any other elements of the noun phrase or the entire verb
phrase. In fact, demonstratives are never incorporated into these denominal
constructions in Michif (nor in Ojibwe, Mathieu 2013). Accordingly, (84b) and
(85b) are ungrammatical:

(84) a. (ni-)lur-ihkaashoo-n
(1-)[Link]-IND.1S
‘I am pretending to be a bear’
 Michif mixed verbs: Typologically unusual word-internal mixing 147

b. * lur-ana-ihkaashoo-n
[Link]-IND.1S
(intended) ‘I am pretending to be like that bear’

(85) a. ka-la-meezooñ-ihkee-w
[Link]-IND.3S
‘s/he is making a house’
b. * ka-anima-la-meezooñ-ihkee-w
[Link]-IND.3S
(intended) ‘s/he is making that house’

Possessives are not attested in our data in incorporated nominals, and numer-
als are not attested either.
Within the denominal mixed verbs, the plural article lii is unattested and
ungrammatical, even with nouns that are otherwise attested and used almost
exclusively in the plural:

(86) * lii-bulet-ihkee-w
[Link]-IND.3S
(intended) ‘s/he’s making meatballs’
(Gillon and Rosen 2018: 134)

Apparently, Michif incorporated noun phrase structure is much less extensive

than the structure incorporated in Ojibwe, since it does not include possessives
and numerals (Mathieu 2013). Table 1 below sums up the noun phrase features AU: The
cross referen-
incorporated in denominal constructions in Michif. ces to fig-
ures, tables
and equa-
Table 1: Features of the incorporated noun phrase in Michif denominal constructions. tions are
highlighted
for the au-
Feature Attested Specifics thor/editor to
check and
articles YES mostly obligatory
confirm its
definiteness (of the YES BUT it appears neutralized correct place-
ment. These
article)
highlights
will be re-
gender (of the noun and YES
moved in the
the article) next stage.
Please make
animacy (of the noun) N/A
changes if
plurality NO ruled out as ungrammatical necessary.
148 Maria Mazzoli, Peter Bakker and Verna DeMontigny

Table 1 (continued )

Feature Attested Specifics

prepositions NO only forms like aañritaar, unanalysed by speakers and

preceded by the nominalizer li-

demonstratives NO ruled out as ungrammatical

possessives NO not found in the data

numerals NO not found in the data

It is likely that denominals represented the first context for the formation of
mixed French/Cree verb stems in Michif, and the great variety of these types of
constructions demonstrates that denominal verb formation is a productive sub-
set of the morphological Cree lexicon in Michif.

4.2 Typologically unusual patterns of borrowing

and word-internal mixing in Michif

The presence of French elements in the Michif verb is marginal yet productively
embedded in an Algonquian frame. The Cree polysynthetic verb frame is diffi-
cult to penetrate, probably due to the pivotal role of the stem-internal finals,
which mark obligatory categorization with respect to transitivity and animacy,
for all the stems to be inflected. Every non-Algonquian or mixed verb stem
would have to receive this type of categorization to be able to be inflected. In
what follows, we will compare the use of French nouns in Michif to phenomena
of borrowing, code-switching, and other processes of word-internal mixing in
other known mixed languages.
The patterns of use of French nominals in the Michif verb seem different
from other known instances of borrowing, code-switching, or creole and mixed
language formation. French nouns have been borrowed in Montagnais along
with a variety of French determiners, like la, les, l’ and even with des and tes as
well as zero marking, although both the articles and the possessives were in-
cluded as part of the phonologically integrated string of the borrowed French
nouns in that language (Drapeau 1980). The same happened to some Michif
nouns with the indefinite, definite, and plural determiners (e.g. l/n/arzhañ ‘the
money’, l/n/zotomobil ‘the car’, and l/nuuvraazh ‘the work’). However, Michif
also has morphologically identifiable definite, indefinite, and plural determin-
ers, and a set of possessives, which carry the overt marking of sex-based gender
 Michif mixed verbs: Typologically unusual word-internal mixing 149

and number in the noun phrase. Moreover, in typical borrowing situations, loans
are often assigned a default or a natural value of a nominal classificatory cate-
gory (Kilarski 1997). For example, in Montagnais, in a more recent round of
phonologically non-integrated borrowings by Montagnais-French bilinguals,
the plural article les is used as a default to flag foreign borrowings, even when
this would be incorrect in French (e.g. les kafe ‘the coffee’ Drapeau 1980).
This is different from Michif, in which the sex-based gender of the French
nouns in Michif is maintained productively and pretty much equivalent to its
source, with only 14.4% of Michif gender values differing from their French
equivalents (Sammons 2019: 232–234).
Michif differs from other known cases of intense borrowing from Romance
languages, such as in Chamorro and Tetun Dili, which do not borrow gender
from the European language. In general, it is uncommon for languages to borrow
nouns with the markers of the original gender, and certainly not across the board
as in Michif. The fact that Michif maintained the French gender system and that
denominals are formed only on complex French noun phrases sets this language
apart typologically from other contact languages (Good 2012, Corbett 1991).
Michif also shows peculiar behavior in denominal verb formation (section 3.2).
The borrowing of nouns to form denominal verbs is known to be a common
cross-linguistic process (Gardner-Chloros and Edwards 2007).Gerdts and Marlett
(2008: 414) describe many languages from North America that employ denominal
verb constructions as a productive source of borrowing, mentioning examples
from Yaqui, Halkomelem, White Mountain Apache, Seri, and Nuuchahnulth. Val-
entine (2001: 419) notes productive denominal constructions from English bare
nouns in Ojibwe with the light verbs -ke ‘make’ and -i ‘have’. However, the major-
ity of Ojibwe denominals are still formed from Algonquian nouns, while Michif
forms denominal verbs only from French noun phrases. Also, borrowed nouns in
denominal formations in Algonquian languages mostly feature the incorporation
of bare roots (Gerdts and Marlett 2008: 415) and thus behave differently from
Michif, in which the entire category of denominal verbs has been restructured
and adapted to the incorporation of complex French noun phrases (Table 1 in
section 4.1).
Moreover, Michif features a peculiar type of mixed stem combining French
nominals and Cree finals. Due to their fuzzy semantics and uncertain productiv-
ity, Cree finals in Michif are not commonly used in mixed forms. However, the
data illustrated in section 3.3 testify that some stem-internal Cree finals are at-
tested in combination with French. That is the extent reached by word-internal
patterns of mixing in Michif. Among the Algonquian languages, the only other
known example is Mi’kmaq, which borrows verbs using the Algonquian final -
ewi-t (Bakker and van der Voort 2017: 420).
150 Maria Mazzoli, Peter Bakker and Verna DeMontigny

The existence of constraints on borrowing in highly synthetic languages

has been exaggerated in the literature (Bakker and van der Voort 2017). Our
study adds to this by showing that borrowing is not blocked in polysynthetic
languages. However, speakers do prefer borrowing nouns, while borrowing
verbs into polysynthetic languages appears to be particularly uncommon. A
cross-linguistically widespread strategy to inflect loan verbs is that of using
an inflected verb of support (Matras 2007: 47; Wohlgemuth 2009; Haspelmath
and Tadmor 2009, who survey borrowings in a sample of 41 languages, very
few of which are polysynthetic). For example, in Montagnais, borrowed verbs
are never integrated morphologically. The French verbs in Montagnais are ut-
tered in their infinitive form and associated with the native forms of the verb
tuutam ‘to do’ to help integrate the loan (Drapeau 1980, 1995). Michif adopted
a different strategy, integrating the French nominalized infinitives into the
Cree verbal structure. This triggered the reanalysis of the infinitives as bipar-
tite stems with an originally created categorizing AI final -ii to accommodate
English loans.
Theories of borrowing and language mixing have attributed the variability
in the outcomes of language contact to the typological compatibility of the
languages involved (Myers-Scotton 1993; Field 2002) and the existence of
borrowing hierarchies, claiming, for instance, that nouns are the easiest to
transfer, while inflectional morphology is the hardest (Muysken 2000). For
word-internal mixing, the syntactic boundedness of the borrowed morpheme
has proved to be a relevant factor, and Gardani (2008) observed that contextual
inflection (agreement) is more difficult to borrow than inherent inflection. Recent
literature has singled out some cases of morphological inflectional borrowings
among creole and mixed languages. For the mixed language Gurindij Kriol, Mea-
kins (2011) provides evidence of the borrowing of contextual inflection, the most
uncommon case of grammatical borrowing. Clements and Luís (2015) illustrate
a case of exceptional morphological borrowing in Korlai Indo-Portuguese
(a Portuguese-based creole language in contact with Marathi), which bor-
rowed a Marathi non-finite verb form to create a new inflectional class, specif-
ically for integrating loan verbs. The Korlai case is very similar to the Michif
one, with the new AI final -ii developed from the French infinitive marker -er.
Michif patterns with cases of intense language contact and other known
mixed languages, which further contribute to classifying Michif as a specific
case of mixed language formation.
 Michif mixed verbs: Typologically unusual word-internal mixing 151

5 Conclusions
This paper adds a puzzle piece to the discussion of the mixed nature of Michif
from the perspective of the morpho-syntax of the verb phrase, which is often
considered to be entirely Cree-derived. Our analysis showed that the Cree verb
structure persists when elements from non-Algonquian languages are inte-
grated into Michif mixed verb formations. French influence on the Cree verb in
Michif is mostly confined to lexical components within the verb, either as entire
stems, as preverbs, and, in a few instances, as initials. However, at least two
new functional morphemes appear to have emerged in Michif. The first is a new
element -ii, which we argue to be an AI final, derived from the reanalysis of the
French infinitive marker -er, and possibly influenced by Cree animate intransi-
tive verbs ending in -îw. The second is the nominalizer li- that is prefixed to
nominalized verbs or adjectives in noun phrases incorporated into a Cree verb
frame.
Finally, we argued that Michif strategies of accommodating French nomi-
nals are typologically rare, also with respect to other contact languages, which
places Michif among other cases of intense language contact and known mixed
languages. Infact, unlike other Algonquian languages in contact with French,
Michif adapted a variety of French nominalized verb stems to function as ani-
mate intransitive (AI) verbs. Also, while denominal verbs are commonly formed
out of the bare roots of borrowed French nouns in many Algonquian languages,
Michif replaced the entirety of denominal forms with new mixed constructions
formed from complex French noun phrases and a set of Cree derivatives.

Abbreviations
0 inanimate subject
1 first person
2 second person
3 third person
4 third person obviative (fourth)
AC actor
AI animate intransitive
AN animate
ART article
CNJ conjunct
DEF definite
DIR direct (TA)
EPEN epenthetic vowel
152 Maria Mazzoli, Peter Bakker and Verna DeMontigny

EXCL exclusive
F feminine
FUT future
II inanimate intransitive
IMP imperative
INAN inanimate
INCH inchoative
INCL inclusive
IND independent
INDEF indefinite
INF infinitive
M masculine
[Link] middle-reflexive
MED medio-passive
NEG negative
NMLZ nominalizer
O object
PASS passive
POSS possessive
PL plural
PRED predicative attributive
PST past
S singular
SBJ subjunctive
TA transitive animate
THE theme sign
TI transitive inanimate
TR transitive

References
Antonov, Anton. 2019. Loan verb integration in Michif. Journal of Language Contact 12(1). 27–51.
Bakker, Peter. 1997. A language of our own: The genesis of Michif, the mixed Cree-French
language of the Canadian Métis. New York: Oxford University Press.
Bakker, Peter. 2006. Algonquian verb structure: Plains Cree. In Grazyna Rowicka & Eithne
Carlin (eds.), What’s in a Verb?, 3–27. Utrecht: LOT.
Bakker, Peter. 2017. Typology of mixed languages. In Alexandra Y. Aikhenvald &
R. M. W. Dixon (eds.) The Cambridge handbook of linguistic typology, 217–253.
Cambridge: Cambridge University Press.
Bakker, Peter. 2019. Review of Gillon, Carrie & Nicole Rosen, with Verna DeMontigny. 2018.
Nominal contact in Michif. Oxford University Press. International Journal of American
Linguistics 85(1). 163–165.
Bakker, Peter & Hein van der Voort. 2017. Polysynthesis and language contact. In Michael
Fortescue, Marianne Mithun & Nicholas Evans (eds.), The Oxford handbook of
polysynthesis, 408–427. Oxford: Oxford University Press.
 Michif mixed verbs: Typologically unusual word-internal mixing 153

Barrie, Michael & Eric Mathieu. 2016. Noun incorporation and phrasal movement. Natural
Language and Linguistic Theory, 34(1). 1–51.
Bloomfield, Leonard. 1984. Cree-English lexicon. 2 Volumes. New Haven: Human Relations
Area Files.
Clark, Eve V. & H. H. Clark. 1979. When nouns surface as verbs. Language 55(4). 767–811.
Clements, Clancy J. & Ana R. Luís. 2015. Contact intensity and the borrowing of bound
morphology in Korlai Indo-Portuguese. In Francesco Gardani, Peter Arkadiev & Nino
Amiridze (eds.), Borrowed Morphology. 219–240. Amsterdam/Berlin: John Benjamins.
Corbett, Greville G. 1991. Gender. Cambridge: Cambridge University Press.
Cowan, William. 1990. Bloomfield, structuralism, and the Algonquian languages. In Hans-Josef
Niederehe & E.F.K. Koerner (eds.), History and historiography of linguistics: Proceedings
of the fourth International conference on the history of the language sciences, Trier,
24–28 August 1987. Volume 2: 18th–20th Century, [Studies in the History of the Language
Sciences 51(2)]. 833–848. Amsterdam: John Benjamins.
Drapeau, Lynn. 1980. Les emprunts au français en montagnais. In Inuktitut et langues
Amerindiennes au Quebec. Special issue of Cahiers de linguistique, 10, 29–49. Quebec:
Presses de l’Université du Québec.
Drapeau, Lynn. 1995. Codeswitching in caretaker speech and bilingual competence in a Native
village of Northern Quebec. International Journal of the Sociology of Language 113. 157–164.
Dryer, Matthew. 2007. Clause types. In Timothy Shopen (ed.), Language typology and
syntactic description. Volume. I: Clause structure, 224–275. Cambridge: Cambridge
University Press.
Field, Fredric 2002. Linguistic borrowing in bilingual contexts. Amsterdam: John Benjamins.
Fortescue, Michael, Marianne Mithun & Nicholas Evans (eds.). 2017. The Oxford handbook of
polysynthesis. Oxford: Oxford University Press.
Gardani, Francesco. 2008. Borrowing of inflectional morphemes in language contact. Frankfurt
am Main: Peter Lang.
Gardner-Chloros, Penelope & Malcom Edwards. 2007. Compound verbs in codeswitching:
Bilinguals making do? International Journal of Bilingualism 11. 73–91.
Gerdts, Donna & Stephen Marlett. 2008. Introduction: The form and function of denominal
verb constructions. International Journal of American Linguistics 74. 489–510.
Gillon, Carrie & Nicole Rosen, with Verna DeMontigny. 2018. Nominal contact in Michif.
Oxford: Oxford University Press.
Goddard, Ives. 1990. Primary and secondary stem derivation in Algonquian. International
Journal of American Linguistics 56. 449–483.
Good, Jeff. 2012. Typologizing grammatical complexities: or Why creoles may be
paradigmatically simple but syntagmatically average. Journal of Pidgin and Creole
Languages 27(1). 1–47.
Gottfurcht, Carolyn A. 2008. Denominal verb formation in English. Evanston: Northwestern
University dissertation.
Haspelmath, Martin & Uri Tadmor (eds.). 2009. Loanwords in the world’s languages:
A comparative handbook. Berlin: De Gruyter Mouton.
Jacques, Guillaume. 2012. From denominal derivation to incorporation. Lingua 122. 1207–1231.
Johns, Alana. 2017. Noun incorporation. In Martin Everaert & Henk C. van Riemsdijk (eds.), The
Wiley Blackwell Companion to Syntax, 2nd ed. Hoboken, NJ: John Wiley and Sons, Inc.
Kilarski, Marcin. 1997. The case of the unmarked gender in loanwords. Studia Anglica
Posnaniensia, 32. 89–99.
154 Maria Mazzoli, Peter Bakker and Verna DeMontigny

Laverdure, Patline & Ida Rose Allard. 1983. The Michif dictionary. John C. Crawford (ed.).
Winnipeg: Pemmican Publications.
Mathieu, Eric. 2013. Denominal verbs in Ojibwe. International Journal of American Linguistics
79(1). 97–132.
Mattissen, Johanna. 2017. Sub-types of polysynthesis. In Nicholas Evans, Marianne Mithun &
Michael Fortescue (eds.), The Oxford handbook of polysynthesis, 70–98. Oxford: Oxford
University Press.
Matras, Yaron, 2007. The borrowability of structural categories. In Yaron Matras & Jeanette
Sakel (eds.), Grammatical borrowing in cross-linguistic perspective, 31–73. Berlin:
Mouton de Gruyter.
Mazzoli, Maria. 2019. Michif loss and resistance in four Metis communities (Kahkiyaaw
mashchineenaan, “All of us are disappearing as in a plague”). Zeitschrift für Kanada-
Studien, 69. 96–117.
Mazzoli, Maria. Accepted. Secondary derivation in Michif: Beyond the traditional Algonquian
template (presented with the glossed text of “La Sandrieuz” in Appendix). In
Danae M. Perez & Eeva Sippola (eds.), Postcolonial varieties in the Americas. Berlin:
Mouton de Gruyter.
Meakins, Felicity 2011. Borrowing contextual inflection: Evidence from northern Australia.
Morphology 21(1). 57–87.
Meakins, Felicity & Jesse Stewart. To appear. Mixed languages. In Salikoko Mufwene & Anna
Maria Escobar (eds.), Cambridge handbook of language contact. Cambridge: Cambridge
University Press.
Mellow, Dean. 1990. Asymmetries between compounding and noun incorporation in Plains
Cree. In William Cowan (ed.), Papers of Twenty-first Algonquian Conference, 247–257.
Ottawa: Carleton University.
Muysken, Pieter 2000. Bilingual speech: A typology of code-mixing. Cambridge: Cambridge
University Press.
Myers-Scotton, Carol. 1993. Duelling languages: Grammatical structure in codeswitching.
Oxford: Clarendon Press.
Oxford, Will. 2017. Algonquian grammar myths. Toronto Working Papers in Linguistics
39. 1–37.
Papen, Robert A. 2005. On developing a writing system for Michif. Linguistica Atlantica
26. 75–97.
Papen, Robert A. 2017. On confirming the split phonology hypothesis for Michif. Presented at
the 49th Algonquian Conference, Université du Québec à Montréal.
Prichard, Hilary & Kobey Shwayder. 2014. Against a split phonology of Michif. University of
Pennsylvania working papers in linguistics 20(1). 271–280.
Rhodes, Richard A. S. 1977. French Cree: A case of borrowing. In William Cowan (ed), Actes
du huitième congrès des algonquinistes, 6–25. Ottawa: Carleton University.
Rosen, Nicole. 2007. Domains in Michif phonology. Toronto, ON: University of Toronto
dissertation.
Rosen, Nicole & Heather Souter. 2009. Piikishkweetak aañ Michif! Winnipeg, Manitoba: Louis
Riel Institute.
Rosen, Nicole & Heather Souter. 2015. Piikishkweetak aa’n Michif! [2nd ed.]. Winnipeg,
Manitoba: Louis Riel Institute.
Rosen, Nicole. 2019. Michif Online Dictionary. [Link] (accessed
26 April 2020).
 Michif mixed verbs: Typologically unusual word-internal mixing 155

Rosen, Nicole, Jesse Stewart & Olivia Sammons. 2020. How ‘mixed’ is mixed language
phonology? An acoustic analysis of the Michif vowel system. Journal of the Acoustical
Society of America 147(4). 2989–2999.
Sammons, Olivia N. 2019. Nominal classification in Michif. Edmonton: University of Alberta
dissertation.
Slavin, Tanya. 2012. The syntax and semantics of stem composition in Oji-Cree. Toronto:
University of Toronto dissertation.
Smit, Niels. 2005. Noun incorporation in Functional Discourse Grammar. In Casper de Groot &
Kees Hengeveld (eds.), Morphosyntactic expression in Functional Grammar, 87–134.
Berlin: Mouton de Gruyter.
Thomason, Sarah Grey. 2019. Review of Gillon, Carrie & Nicole Rosen, with Verna DeMontigny.
2018. Nominal contact in Michif. Oxford University Press. Language 95. 806–811.
Valentine, Randolph. 2001. Nishnaabemwin reference grammar. Toronto: University of Toronto
Press.
Wohlgemuth, Jan. 2009. A typology of verbal borrowing. Berlin: Mouton de Gruyter.
Wolfart, H. Christoph. 1996a. Sketch of Cree, an Algonquian language. In Ives Goddard (ed.),
Handbook of North American languages 17, 390–439. Washington, DC: Smithsonian
Institution.
Wolfart, H. Christoph. 1996b. Sketch of Cree, an Algonquian language. In Ives Goddard (ed.),
Handbook of North American Indians 17. Languages, 390–439. Washington, DC:
Smithsonian Institution.