A multilingual dictionary database of Algonquian derivational morphemes

Algonquian Word-Structure Basics

(Written for the Nisinoon Project; last updated July 2, 2021)

Monica Macaulay


This document provides an overview of word structure in Algonquian languages.

Caveat: The view in this document is influenced by my familiarity with some of the Central Algonquian languages (especially Menominee); it may not be quite right for some of the other languages, especially the Plains languages, which are really different. But this should work for most of the languages. I will make updates as I learn more! (Input appreciated!)


Lexical Categories (Parts of Speech)

The classic way to divide up Algonquian lexical categories at the very highest level is according to whether they inflect or not (see next section for distinction between Derivation vs. Inflection):

(1) Do inflect (take endings)

(2) Do not inflect

Sometimes a few other things are included under the “do inflect” heading, e.g., pronouns, negators, etc. Particle is usually a vast category, and some authors have subdivided it further, adding, for example, prepositions (see, e.g., Oxford (2007) for discussion of the category particle in Innu-Aimun.

Notice that there aren’t any adjectives or adverbs (although some authors do use these labels).

Derivation vs. Inflection

The distinction is a little fuzzy around the edges, but basically:

In Algonquian languages you pretty much always have to have inflection on verbs (except in some cases where we say there’s a zero morpheme).

This project is concerned with the derivational morphemes of Algonquian.


The basic derivational morphemes in Algonquian languages are called components, and Nisinoon is creating a cross-Algonquian dictionary of these components. There are three types, which correspond to where in the stem they appear. The following is from my in-progress Menominee grammar.

Initial Medial Final
Figure 1. Menominee Stem Components

There are (at least!) two ways of looking at the make-up of stems: one view is that a stem can be formed of an initial by itself; an initial and a final; or an initial, a medial, and a final (see, e.g., Goddard 1990). The other view is that initials and finals are obligatory, but sometimes the final is a zero morpheme just carrying lexical category (Cowell and Moss 2008, for example, take this view for verb stems; see also Wolfart 1973).

Some authors allow for double medials; other authors do not. Authors who want to rule them out will say one of the (apparent) medials is really part of another component. Either way, though, it’s not common.

Finals can just give a word its lexical category (in which case they’re called ‘abstract’), or they can have some lexical meaning in addition to providing a lexical category (in which case they’re called ‘concrete’). Some authors divide finals into a concrete part and an abstract part (see below, on formatives).

Examples (components in slashes in the third line; this line does not include inflectional morphology):

  1. Menominee


    1. paemet‑



    2. ‑aehkw‑



    3. ‑ape



    S/he sits sideways.

    Bloomfield 1975:208

  2. SW Ojibwe


    1. ozhaash‑



    2. ‑sag‑



    3. ‑aa



    It is a slippery floor.

    Nichols 2015

  3. Delaware


    1. kwən‑



    2. ‑askw‑



    3. ‑at



    It is long grass.

    O'Meara 1990:250

  4. Blackfoot


    1. siso‑



    2. ‑ap‑



    3. ‑ittaki



    shred (the hide) into strips

    Frantz & Genee 2015

(In these examples, the components are given in their underlying forms, which differ in some cases from the forms as they show up in the actual word.)

Components occur in all lexical categories, not just verbs. But the particles tend to be fairly simple in structure, and the nouns are generally simpler than the verbs. It’s in the verbs that you see the really complex combinations.


Our standard for hyphens is:

However, other authors don’t necessarily follow this. We enter the data their way but convert it to the standard above in the project orthography.


Nouns in Algonquian languages can be animate or inanimate.

Verb Types

Verbs in Algonquian languages are categorized in two ways:

The two intersect like this (again, from my in-progress Menominee grammar):

Table 1. Verbs and Animacy
Subject Object
Animate Intransitive AI animate
Inanimate Intransitive II inanimate
Transitive Animate TA animate
Transitive Inanimate TI inanimate

That is, intransitive verbs are categorized by the animacy of their subjects (and of course they don’t have objects), and transitive verbs are categorized by the animacy of their objects. The abbreviations given for each type are used extensively.

This matters because there are different verb paradigms (sets of endings) for each type. And for our purposes, verb finals will create not just verb stems, but verb stems of a particular type (e.g. a TA verb).

Some of the types have subcategories, which are usually numbered — the AI and TI verbs have these, so you’ll see e.g. AI2 or TI1. In some of the languages you’ll see, e.g., TI1a and TI1b.

Paired Verb Stems

Usually, verbs come in pairs — AI and II, TA and TI. For example, in Menominee:

    1. pāpaehcen he, she, it (an.) falls (AI)

    2. pāpaehnaen it (inan.) falls (II)

    Bloomfield 1975:200

    1. na͞ewa͞ew s/he sees him, her, it (an.) (TA)

    2. na͞emwah s/he sees it (inan.) (TI)

    Bloomfield 1975:154

You can see that they differ, but only a bit. Usually the difference comes in the final, which makes sense, because that’s the component that gives a stem its category.

In some cases you find sets of four, where all four verb types are based on a common initial, e.g. Menominee maehkīhotaew it is painted red (II), maehkīhosow it (animate) is painted red (AI), maehkīhonaew s/he paints him, her, it (animate) red (TA), maehkīhotaw s/he paints it red (TI).

Primary vs. Secondary Derivation

What was described in the Components section is what is known as primary derivation, where an initial, an optional medial, and a final form a stem. But you can take a stem created this way and add another final to it to create a larger stem — this is called secondary derivation. (Authors use different terms for what I’m calling the stem here; some call it an initial or a derived initial.) These are illustrated below (again, from my Menominee grammar (Macaulay, in prep):

Figure 2. Primary Derivation
Initial Medial Final
Figure 3. Secondary Derivation
Initial Medial Final

Some finals can only be used in primary derivation, some can only be used in secondary derivation, but some can be used in both.

Following are some examples from Menominee:

  1. kōhkawaew (AI) it (animate; for example, a wagon or canoe) tips over

    Stem: kōhkawae‑ [initial kōhkā‑ tip over + ‑āwa͞e AI final]

    Secondary final: ‑makat II verb

    New word: kōhkawaemakat (II) it tips over

    Bloomfield 1975:103

  2. cēkataham (TI) he or she sweeps it clear, cleans it with a broom

    Stem: cēkatah‑ [initial cēk‑ near, next to (?) + ‑atah by stick]

    Secondary final: ‑ka͞e AI verb; indefinite action

    New word: cēkatahekaew (AI1) he or she sweeps

    Bloomfield 1975:42

  3. wāqnenam (TI) he or she throws light on it, lights it up

    Stem: wāqnen‑ [initial wāqN‑ light + ‑aen by hand]

    Secondary final: ‑kan N; instrument, product, place, etc.

    New word: wāqnenekan (N) lamp, candle

    Bloomfield 1975:267

  4. sūniyan (N) money

    Stem: sūniyan‑ [initial sōni‑ silver + ‑ān N final (?)]

    Secondary final: ‑ikamekw N; house, building

    New word: sūniyanikamek (N) bank

    Bloomfield 1975:244 & Sarah Skubitz 4/3/00

These examples represent the standard kind of secondary derivation, where a final is added to a stem. Very rarely, you’ll find someone who allows secondary derivation to add a medial as well as a final (e.g. Drapeau 1980:317).


Traditionally, Algonquianists have recognized another level or layer of derivation, saying that components can themselves be made up of more than one piece. There’s no standard name for this sub-component element; we’ve adopted the word formative for it.

In this project, we believe that what looks like sub-component elements are really historical relics, and should not be treated as synchronically present in a word’s analysis. But it’s important for the database to be true to the sources we get the data from, so if they include formatives in their analysis, we are entering them.

The types of formatives can include:

This list (examples appear below) comes from Bloomfield’s (1962) grammar of Menominee; other authors may use other terms or include other types. Some of the formative types are claimed to be meaningful; others are just phonological material that attaches to a component (the accretions and extensions).

The example below comes from a paper I coauthored on this topic (Macaulay and Salmons 2017) and shows where each type of formative attaches to the component it goes with.

Menominee Stem Structure

Examples of formatives (all from Menominee):

For authors who divide finals into two parts, a concrete part and an abstract part (mentioned above), the prefinal corresponds to the concrete part.

The problem we have with the notion of formative has to do with the definition of morpheme. A morpheme is supposed to be the smallest unit of sound and meaning or function. But if components are morphemes, then what are the formatives? Morphemes shouldn’t be made up of smaller parts, especially meaningful parts (like with the premedials). But if the formatives are the morphemes, then what status does the component have? If it’s not a morpheme, what is it?

Deverbal Formations

Deverbal in the sense it is used here means “formed from a word” (not formed from a verb).

The traditional Algonquianists also see derivation happening to create morphemes. They very correctly recognize the relationship between full words (or stems or roots) and morphemes, but take it the extra step to say that the morpheme is derived from the word. The problem is, if this is a synchronic operation, where in the grammar could it occur?

For example, Bloomfield believed that a number of finals were derived from stems or words (“deverbal finals”). He cites the word mahka͞esen shoe, moccasin and notes that in a word like maeqtekuahkesen wooden shoe we find a final of the form ‑ahkaesen shoe. For him, the fact that they’re similar means the latter is derived from the former — apparently synchronically, although he’s (frustratingly) never specific.

Although we believe that this kind of relationship is purely historical, when an author says that something is deverbal, we make a note of it, and note the source, if they give it.


Acknowledgment of Support

This material is based upon work supported by the University of Wisconsin-Madison Graduate School and the National Science Foundation DLI-DEL program under grant number BCS1953103. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.