1 Introduction

There’s an ongoing debate about what the right representation of phi features should be. One particularly controversial point in this debate is the status of third person, with there being a core divide between those who think that third person is the lack of more specified person features, a view that we can dub 3noP (mnemonic for “third person is not a person”), following Nevins (2007), and those for whom there is a distinct third person feature, a view we can dub 3yesP.

What kind of data would help us decide between 3noP and 3yesP? Under a 3noP theory, there is no way refer directly to third person to the exclusion of first and second, as the representation of third person is contained within the representation of first and second. 3noP would thus predict that there should be no morphosyntactic phenomenon that targets third person and only third person—let’s call this Invisibility:

    1. (1)
    1. Invisibility
    2. In no language can third person explicitly be targeted to the exclusion of first and second person in various kinds of morphosyntactic processes.

Thus, there should be no morphological process, like Impoverishment, that could target only third person. Nor could there be a syntactic process, like agreement, that targets only third person. In contrast, 3yesP does give us the formal tools to refer exclusively to third person, predicting that we should get counterexamples to Invisbility. The literature has provided us with morphological counterexamples to Invisibility—we get morphological processes that can see third person to the exclusion of first and second (Nevins 2007; Trommer 2008, a.o.). Here, I add to this body of work by providing an example of a syntactic counterexample to invisbility.

The empirical domain is the peripheral suffix in Algonquian. This suffix appears in the independent order, which is generally characteristic of matrix declaratives, and it surfaces linearly rightmost on the verb, to the right of various kinds of tense/mood/evidential marking. I’ll follow the majority of the generative literature on Algonquian in placing the peripheral suffix in C across the family (Halle & Marantz 1993; Branigan & MacKenzie 1999; Bruening 2001; Bliss 2013; a.o.).

In all Algonquian languages, the peripheral suffix displays an omnivorous agreement pattern (Nevins 2011), always agreeing with a third person, no matter if it’s a subject or an object (Xu 2020; 2021).1 Here, I’ll guide us through data from two Algonquian languages from different branches of the family, Blackfoot (Plains Algonquian; Frantz 2017; Goddard 2018) and Plains Cree (Central Algonquian; Wolfart 1973; Dahlstrom 1991; Okimāsis 2018), that show the most straightforward instantiation of this pattern: “the [peripheral] suffixes show agreement with the highest 3rd person argument (proximate, obviative, or plural)” (Bliss 2013: 234).2 After outlining the omnivorous third person agreement pattern, I’ll respond to some potential alternative accounts that attempt to analyze it away in order to save 3noP: agreement for animacy, agreement for obviation, and agreement for [D] features. I conclude that we really do genuinely have omnivorous third person agreement in Algonquian, and we must accordingly recognize the existence of third person features in the syntax.

2 Setting the stage

But before we dive into the weeds, let’s set up the theoretical backdrop a bit more to see why omnivorous third person agreement is particularly important for the 3noP vs 3yesP debate.

Let’s start with theories of person features. One influential proponent of 3noP is Harley & Ritter (2002) (though the insight is longstanding; see Forchheimer 1953 for discussion), who propose that third person lacks person features entirely. They propose the following representations of person features:

    1. (2)
    1. a.
    1. [π]: third person                                              Harley & Ritter (2002)
    1.  
    1. b.
    1. [part, addr]: second person
    1.  
    1. c.
    1. [part, auth]: first person exclusive
    1.  
    1. d.
    1. [part, auth, addr]: first person inclusive

Thus, they adopt a privative feature system, with the only person features being [part], [auth],3 and [addr]. Third person is underspecified for person features.

In contrast, Nevins (2007) argues for 3yesP, arguing for a theory with binary [±part] amd [±auth] features:4

    1. (3)
    1. a.
    1. [–part, –auth]: third person                                            Nevins (2007)
    1.  
    1. b.
    1. [+part, –auth]: second person
    1.  
    1. c.
    1. [+part, +auth]: first person

Under this representation, there is a dedicated third person feature: it’s labelled [–part]. We could also imagine a privative 3yesP theory, with privative [3], [part], [auth], and [addr] features—such a theory is proposed by Bondarenko (2020). Of course, these exemplars of 3noP and 3yesP do not exhaust the possibility space, but they will do to illustrate the two opposing camps.5

One of Nevins’s (2007) central arguments is the existence of 3-on-3 dissimilation effects. In the abstract, these involve the ungramamaticality of two linearly-adjacent third person clitics that can be fixed by removing the person specification of one of the offending clitics, thus resulting in the spellout of an underspecified form. Both the conditioning environment for these kinds of rules (adjacency to a third person clitic) as well as the target of the rule (a third person clitic) have to be able to make reference to some kind of feature found only in third person, and not first or second.

For a concrete example, let’s consider the case of Ubykh (Northwest Caucasian), discussed by Deal (2020).6 In Ubykh, the third person absolutive clitic ɐ́ = cannot linearly precede a third person singular clitic—instead, you either drop the absolutive clitic (which is independently possible—absolutive clitics are optional), or you realize it as the otherwise unattested form jɨ=:

    1. (4)
    1. Ubykh (Deal 2020: 18–19, 21)
    1.  
    1. a.
    1. *
    2.   3abs=
    1. Ø=
    2. 3sg.dat=
    1. n=
    2. 3sg.erg=
    1. twɨ
    2. give
    1. -n
    2. -prs
    1.   Intended: ‘[She] gives it to her.’
    1.  
    1. b.
    1.   
    2.  
    1. Ø=
    2. 3sg.dat=
    1. n=
    2. 3sg.erg=
    1. twɨ
    2. give
    1. -n
    2. -prs
    1.   ‘She gives (it) to her.’
    1.  
    1. c.
    1.   
    2.   3abs=
    1. Ø=
    2. 3sg.dat=
    1. n=
    2. 3sg.erg=
    1. twɨ
    2. give
    1. -n
    2. -prs
    1.   ‘She gives it to her.’

Deal (2020) argues that this really is a surface morphological restriction, not a syntactic one, as it arises only when third person ɐ́ = is linearly adjacent to a third person singular clitic, no matter the case or syntactic role of the offending clitics. Additionally, if there is another clitic intervening between the two third person ones, then there is no issue: ɐ́ = surfaces and is not replaced by jɨ=. Deal proposes that person features on a third person absolutive clitic are deleted when linearly preceding another third person clitic, and that jɨ= is the default exponent of an absolutive clitic, which only has a chance to surface when person features are deleted.

Data like this seem to push us away from 3noP in favor of 3yesP. However, there’s a possible response that the 3noP camp could make that’s combatible with this data, which Preminger (2019) nicely articulates. Preminger notes that all of the counterexamples to Invisibility in the literature involve morphological phenomena, and thus argues for a modular view: the representation of features across different modules—syntax, morphology, semantics—can differ. He proposes that 3noP holds specifically in the syntax, not the morphology: for the purposes of narrow syntactic processes, third person is the absence of more specified person features. Inspired by phonological redundancy rules, which insert predictable featural information into underspecified phonological representations (Archangeli 1984; 1988, a.o.), Preminger proposes that in the mapping from the output of the narrow syntactic derivation to the morphology we can “fill in” third person features into underspecified syntactic representations, thus rendering third person visible to the morphological component, and allowing us to capture effects like 3-on-3 dissimilation:

    1. (5)
    1. a.
    1. [part]SYN ⇒ [+part]MORPH
    1.  
    1. b.
    1. ¬∃[part]SYN ⇒ [–part]MORPH

All the syntax can see is [part], but the morphology can see both [+part] and [–part]. Let’s call this more nuanced version of 3noP syntactic 3noP.

Now the question becomes: what data would distinguish between syntactic 3noP and 3yesP? We would need to find cases of crucially-syntactic phenomena that exclusively target third person. One such phenomenon is agreement,7 and in particular the phenomenon of omnivorous agreement (Nevins 2011), which involves an agreement marker that seems to specifically seek out certain features (e.g. [part] or [pl]) no matter if they are found on the subject or object. A classic and by-now standard account of omnivorous agreement in a Probe-Goal model of agreement (Chomsky 2000; 2001, a.m.o.) is to make use of the idea that probing is relativized (Béjar 2003 extensively explores this intuition): probes are specified to look for certain types of features, and only interact with those features on a goal (e.g. wh, φ, etc.). Omnivorous agreement is thus analyzed as a probe relativized for a particularly “specific” feature: a probe relativized for [part], [upart], would only interact with part goals, skipping over any third persons, deriving the omnivority. Under this light, all agreement is at least slightly omnivorous: for instance, plain phi agreement skips over potential goals that don’t bear phi features.

One classic example of omnivorous person agreement is found in the Kichean Agent Focus construction, which appears in contexts where the transitive agent is Ā-extracted. In this construction, the verb will agree with first or second person arguments (i.e. it’s looking for the feature [(+)part]), no matter whether they’re the subject or object, as in the following Kaqchikel (Mayan) examples:

    1. (6)
    1. Kaqchikel (Preminger 2014: 18)
    1.  
    1. a.
    1. ja
    2. Foc
    1. rat
    2. 2sg
    1. x-at/*Ø-ax-an
    2. com-2sg/*3sg.abs-hear-af
    1. ri
    2. the
    1. achin.
    2. man
    1. >3
    2.  
    1. ‘It was you that heard the man.’
    1.  
    1. b.
    1. ja
    2. Foc
    1. ri
    2. the
    1. achin
    2. man
    1. x-at/*Ø-ax-an
    2. com-2sg/*3sg.abs-hear-af
    1. rat.
    2. 2sg
    1. 3>
    2.  
    1. ‘It was the man that heard you.’

In the Agent Focus construction, then, the verb agrees omnivorously with first or second person arguments, skipping over third persons if it has to.

If there is a third person feature, we should be able to specify a probe for it: we predict the existence of [u3] (or [u–part]) probes. Thus, 3yesP predicts the existence of omnivorous third person agreement: we should find instances of an agreement marker that specifically seeks out third persons, ignoring first and second if it has to. True omnivorous third person agreement has been assumed to not exist (e.g. Preminger 2019, a.m.o.), and this assumption has formed part of the argument for (syntactic) 3noP.8 Here I’ll argue that omnivorous third person agreement does exist. In doing so, I show that even the weakest form of 3noP—syntactic 3noP—is untenable. We must accept 3yesP, even in the syntax.9

3 The data

With the theoretical stage set, let’s now turn to the data from Blackfoot and Plains Cree peripheral agreement (analyzed as a probe in C), first examining configurations where there’s only one third person (intransitives, part>3, and 3>part), then configurations with two third persons (skipping part>part forms, which never feature the peripheral suffix). I’ll use first person exclusive forms as exemplars for part (other forms behave similarly). Throughout, I use “subject” to refer to the single argument of an intransitive and the external argument of a transitive, and “object” to refer to the internal argument of a transitive. I will bold the peripheral suffix when it appears, as well as the controller of agreement in the translation. I default to ‘she’ as the translation for third person singular. As a reference, Tables 1 and 2 list the possible exponents of the peripheral suffix in Blackfoot and Plains Cree.

Table 1

Blackfoot peripheral suffix.

in an.prox an.obv
sg -wa -wa -yini
pl -yi -yi -yi
Table 2

Plains Cree peripheral suffix.

in an.prox an.obv
sg -Ø -Ø -a
pl -a -ak -a

3.1 Intransitives

In intransitives, the peripheral suffix indexes the animacy, number, and obviation of third person subjects. If the subject is not third person, there is no peripheral suffix.

    1. (7)
    1. Blackfoot (Frantz 2017: 170)
    1.  
    1. a.
    1. nit-á’po’taki-hp-innaan
    2. 1-workai-P-1exc
    3. ‘weexc worked’
    1.  
    1. b.
    1. á’po’taki-yi
    2. workai-pl
    3. ‘they worked’
    1.  
    1. c.
    1. á’po’taki-yini
    2. workai-obv.sg
    3. ‘sheobv worked’
    1.  
    1. d.
    1. soká’pii-yi
    2. be.goodii-pl
    3. ‘theyin are good’
    1. (8)
    1. Plains Cree (Dahlstrom 1991: 20)
    1.  
    1. a.
    1. ni-pimipahtâ-nân
    2. 1-runai-1exc
    3. ‘weexc run’
    1.  
    1. b.
    1. pimipahtâ-w-ak
    2. runai-3-prox.pl
    3. ‘theyprox run’
    1.  
    1. c.
    1. pimipahtâ-yi-w-a
    2. runai-obv-3-obv
    3. ‘sheobv runs’
    1.  
    1. d.
    1. mihkwâ-w-a
    2. be.redii-3-in.pl
    3. ‘theyin are red’

While it may look like the 1exc agreement markers in the (a) examples are in complementary distribution with the peripheral suffix, they actually occupy a different slot in the verbal template—the central suffix, typically analyzed as occupying Infl/T (Halle & Marantz 1993; Coon & Bale 2014; Oxford 2014; a.o.). Below, we’ll see forms that feature both the central as well as the peripheral suffixes.

3.2 Mixed configurations: part>3 and 3>part

In configurations involving a third person subject and a part object, the peripheral suffix indexes the subject.10

    1. (9)
    1. Blackfoot (Frantz 2017: 61)
    1.  
    1. a.
    1. nits-ikákomimm-ok-innaan-a
    2. 1-loveta-inv-1exc-sg
    3. ‘she loves usexc’
    1.  
    1. b.
    1. nits-ikákomimm-ok-innaan-i
    2. 1-loveta-inv-1exc-pl
    3. ‘they love usexc’
    1. (10)
    1. Plains Cree (Okimāsis 2018: 187–188)
    1.  
    1. a.
    1. ni-pâhpih-iko-nân-Ø
    2. 1-laugh.atta-inv-1exc-prox.sg
    3. ‘sheprox laughs at us’
    1.  
    1. b.
    1. ni-pâhpih-iko-nân-ak
    2. 1-laugh.atta-inv-1exc-prox.pl
    3. ‘theyprox laugh at us’

In the (a) examples we have singular subjects and singular peripheral suffixes; in the (b) examples we have plural subjects and plural peripheral suffixes—thus, C varies according to the features of the third person subject. So far, all the data we’ve seen so far seems reducible to simple locality: C indexes the highest argument, and only happens to have overt exponents for third persons.

However, in configurations involving a part subject and a third person object, the peripheral suffix indexes the object. The probe skips over a part subject in preference for a third person.11 This is the core property we would expect from an omnivorous probe: ignoring more local potential goals in favor of more distant ones.

    1. (11)
    1. Blackfoot (Frantz 2017: 46–47, 56–57)
    1.  
    1. a.
    1. nits-ikákomimm-a-nnaan-a
    2. 1-loveta-3obj-1exc-sg
    3. ‘weexc love her’
    1.  
    1. b.
    1. nits-ikákomimm-a-nnaan-i
    2. 1-loveta-3obj-1exc-pl
    3. ‘weexc love them’
    1.  
    1. c.
    1. nits-ííkoon-ii-hp-innaan-a
    2. 1-take.downta-in.obj-p-1exc-sg
    3. ‘weexc took it down’
    1.  
    1. d.
    1. nits-ííkoon-ii-hp-innaan-i
    2. 1-take.downta-in.obj-P-1exc-pl
    3. ‘weexc took themin down’
    1. (12)
    1. Plains Cree (Okimāsis 2018: 145–146)
    1.  
    1. a.
    1. ni-mow-â-nân-Ø
    2. 1-eatta-3obj-1exc-prox.sg
    3. ‘weexc eat itan.prox’
    1.  
    1. b.
    1. ni-mow-â-nân-ak
    2. 1-eatta-3obj-1exc-prox.pl
    3. ‘weexc eat theman.prox’

In these examples, we can see that varying the number of the third person object causes the peripheral suffix to vary accordingly. Thus, the peripheral suffix always indexes third person features, no matter whether they come from the subject or the object—the characteristic property of omnivorous agreement.

That brings us to the end of the basic only-one-third-person forms. To summarize: whenever you have only one third person argument, the peripheral suffix agrees with it, skipping over first and second persons if necessary. This is precisely the behavior one would expect from an omnivorous third person probe.

3.3 Syntactic interlude I: The syntax of mixed configurations

But wait!—the staunch 3noP-er might say. How do we really know that C is skipping over the subject to agree with the object in part>3 configurations? What if the object actually has moved over the subject in those configurations (and only those) prior to C probing, thus rendering it the most local goal to C? Here’s a sketch of what one could propose:

    1. (13)
    1. The 3noP syntactic response
    1.  
    1. a.
    1.                                             Intransitive, agree with subject
    1.  
    1. b.
    1.                                          3>part, agree with subject
    1.  
    1. c.
    1.                part>3, agree with object

If that was right syntax of mixed configurations (configurations that mix part and third person arguments), then we wouldn’t need to say anything special about C: it would just agree with the most local goal, and happen to only have overt exponents for third person (similar to English 3sg -s).

I don’t think this is the right syntax of mixed configurations. Firstly, if we try to formally work out a syntax that derives the behavior in (13), it’s strikingly difficult to avoid replicating a need for syntactically-active third person features elsewhere. We would need to have a probe somewhere that would move the object over the subject in part>3 configurations. For sake of argument, let’s place it in Voice, the head that introduces the subject. We need Voice to probe down and move the object in part>3 configurations, but crucially to avoid doing so in part>3 configurations. It’s not immediately clear to me how we could derive this state of affairs without the probe being specified to move third person DPs only.12

But even if there were some non-stipulative way to derive (13) without having to make reference to third person features elsewhere in the syntax, cross-Algonquian data points against this picture. While I am not aware of clear evidence relevant to the question of what’s going on syntactically in part>3 configurations in Blackfoot or Plains Cree (or any of the other languages with the same agreement pattern), there is evidence from Odawa (a variety of Ojibwe, Central Algonquian), Listuguj Mi’gmaq (Eastern Algonquian) and Passamaquoddy-Wolastoqey (Eastern Algonquian) that these contexts lack object inversion (Rhodes 1994; Hamilton 2018; and Grishin 2022a; respectively). Strikingly, mixed configurations in those languages also show the same omnivorous third person pattern, with C always indexing the third person argument. While it is conceivable that Blackfoot and Plains Cree could diverge from these languages in the syntax of mixed configurations, there is as of yet no syntactic evidence from any Algonquian language for a syntax of mixed configurations like that in (13). Therefore, at least as a null hypothesis, it seems reasonable to assume parallel syntactic behavior across the family in this domain, especially given the fact that C agreement in mixed configurations is strikingly uniform across Algonquian (Xu 2020; 2021), and thus cannot be used as justification for a distinct syntax of mixed configurations in some Algonquian languages.

The data involves long distance agreement (LDA) constructions that show locality effects: the matrix attitude predicate can only agree with the highest argument in the embedded clause. In Odawa and Mi’gmaq, the relevant LDA construction involves complements in the conjunct order (a distinct agreement paradigm from the one this paper focuses on) of verbs like ‘know’, and in Passamaquoddy-Wolastoqey the relevant LDA construction involves complements in the subordinative mode (a subtype of the independent order that lacks C agreement) to verbs like ‘want’.13 I exemplify below:

    1. (14)
    1. LDA with part subject in part>3
    1.  
    1. a.
    1. Odawa (Rhodes 1994: 438)
    1.  
    1.  
    1. G-gikenm-in
    2. 2-knowta-2obj
    1. [cj
    2.  
    1. gii=
    2. pst=
    1. baashkzw-Ø-ad
    2. shootta-3obj-2sg>3cj
    1. ].
    2.  
    1. ‘I know that you shot him.’
    1.  
    1. b.
    1. Listuguj Mi’gmaq (Hamilton 2018: 116)
    1.  
    1.  
    1. Geji-u’l
    2. knowta-2obj
    1. [cj
    2.  
    1. gesal-Ø-t
    2. loveta-3obj-2sg>3
    1. ].
    2.  
    1. ‘I know that you love her.’14
    1.  
    1. c.
    1. Passamaquoddy-Wolastoqey (Grishin 2022a: 8)
    1.  
    1.  
    1. Roger
    2. Roger
    1. n-puwatom-a-ku-n
    2. 1-wantti-appl-inv-N
    1. [sub
    2.  
    1. nt-olintuwew-a-n
    2. 1-sing.tota-3obj-N
    1. Asawis
    2. John
    1. ].
    2.  
    1. ‘Roger wants me to sing to John.’
    1. (15)
    1. No LDA with 3 object in part>3
    1.  
    1. a.
    1.   Odawa (Rhodes 1994: 439)
    1.  
    1.  
    1. *N-gikenm-aa
    2.   1-knowta-3obj
    1. [cj
    2.  
    1. gii=
    2. pst=
    1. baashkzw-Ø-ad
    2. shootta-3obj-2sg>3cj
    1. ].
    2.  
    1.   Intended: ‘I know that you shot him.’
    1.  
    1. b.
    1.   Listuguj Mi’gmaq (Hamilton 2018: 116)
    1.  
    1.  
    1. *Geji-Ø-’g
    2.   knowta-3obj-1sg>3
    1. [cj
    2.  
    1. gesal-Ø-t
    2. loveta-3obj-2sg>3
    1. ].
    2.  
    1.   Intended: ‘I know that you love her.’15
    1.  
    1. c.
    1.   Passamaquoddy-Wolastoqey (Grishin 2022a: 9)
    1.  
    1.  
    1. *Roger
    2.   Roger
    1. ’-puwatom-uw-a-n
    2. 3-wantti-appl-3obj-N
    1. [sub
    2.  
    1. Husaw-ol
    2. John-obv.sg
    1. nt-olintuwew-a-n
    2. 1-sing.tota-3obj-N
    1. ].
    2.  
    1.   Intended: ‘Roger wants me to sing to John.’

In each example above, the embedded clause involves a part subject and third person object. The grammatical examples in (14) show the matrix verb agreeing with the part subject as if it were the object of the matrix clause. In contrast, the ungrammatrical examples in (15) involve the matrix verb agreeing with the third person object. Thus, we seem to find a locality effect here, with the matrix verb only able to show object agreement with the highest embedded argument—the part subject, not the third person object (see Rhodes 1994; Hamilton 2018; Grishin 2022a for detailed argumentation that the agreement pattern is really sensitive to syntactic height, and not e.g. hierarchical agreement or sensitivity to base-generated positions).

While it is true that Odawa, Mi’gmaq, and Passamaquoddy-Wolastoqey do not show the same peripheral agreement pattern as Blackfoot or Plains Cree across the board, they strikingly behave exactly the same as Blackfoot and Plains Cree (as well as the rest of the Algonquian family) when we restrict our attention to configurations with only one third person argument: C agrees with the third person, no matter if it’s the subject or object (see Xu 2020; 2021 for discussion of the cross-Algonquian variation in peripheral agreement):16

    1. (16)
    1. Odawa (Valentine 2001: 287)
    1.  
    1. a.
    1. n-waabm-aa-naan-ig
    2. 1-seeta-3obj-1pl-prox.pl
    3. ‘weexc see themprox’
    1.  
    1. b.
    1. n-waabm-ig-naan-ig
    2. 1-seeta-inv-1pl-prox.pl
    3. ‘theyprox see usexc’
    1. (17)
    1. Passamaquoddy-Wolastoqey (Sherwood 1983: 219)
    1.  
    1. a.
    1. n-wicuhkem-a-nnu-k
    2. 1-helpta-3obj-1pl-prox.pl
    3. ‘weexc help themprox’
    1.  
    1. b.
    1. n-wicuhkem-ku-nnu-k
    2. 1-helpta-inv-1pl-prox.pl
    3. ‘theyprox help usexc’

In the (a) examples, we see peripheral agreement indexing the third person object; in the (b) examples, we see peripheral agreement indexing the third person subject. This exactly parallels the behavior we see in Blackfoot and Plains Cree—I conclude that across Algonquian, peripheral agreement shows omnivorous third person behavior. I wait until Section 6.2 to discuss the variation we do see in Algonquian.

Thus, the 3noP-er’s syntax for mixed configurations seems unlikely, not only because it seems difficult to implement without just replicating a third person probe elsewhere in the syntax, but also given the comparative Algonquian picture. Though we don’t (to my knowledge) have relevant evidence from the literature on Blackfoot and Plains Cree, the data from Odawa, Listuguj Mi’gmaq, and Passamaquoddy-Wolastoqey strongly suggests that there is no object inversion in mixed configurations. Further research is required to develop relative c-command diagnostics for subject and object in mixed configurations in Blackfoot and Plains Cree.

3.4 3>3 direct and inverse

Let’s turn back to the agreement data in Blackfoot and Plains Cree. In contexts with two third persons in direct configurations (ones where the subject outranks the object; prox>obv, an>in), the peripheral suffix indexes the subject.

    1. (18)
    1. Blackfoot (Frantz 2017: 46–47, 58)
    1.  
    1. a.
    1. ikákomimm-ii-wa
    2. loveta-3obj-sg
    3. ‘sheprox loves her/themobv’
    1.  
    1. b.
    1. ikákomimm-ii-yi
    2. loveta-3obj-pl
    3. ‘theyprox love her/themobv’
    1.  
    1. c.
    1. ikóón-im-a
    2. take.downti-in.obj-sg
    3. ‘she took it/them down’
    1.  
    1. d.
    1. ikóón-im-i
    2. take.downti-in.obj-pl
    3. ‘they took it/them down’
    1. (19)
    1. Plains Cree (Okimāsis 2018: 132, 145)
    1.  
    1. a.
    1. mow-ê-w-Ø
    2. eatta-3obj-3-prox.sg
    3. ‘sheprox eats it/themobv’
    1.  
    1. b.
    1. mow-ê-w-ak
    2. eatta-3obj-3-prox.pl
    3. ‘theyprox eat it/themobv’
    1.  
    1. c.
    1. wâpaht-am(-w)-Ø
    2. seeti-in.obj-3-prox.sg
    3. ‘sheprox sees it/them’
    1.  
    1. d.
    1. wâpaht-am-w-ak
    2. seeti-in.obj-3-prox.pl
    3. ‘theyprox see it/them’

The (a) examples feature singular subjects, and the (b) examples feature plural subjects—peripheral agreement varies accordingly. Additionally, since the peripheral suffix is only able to index the number of the subject, the number of the object is ambiguous. Thus, we learn that the peripheral suffix prefers indexing the more local goal—a familiar locality effect that is appropriately captured by a third person probe.

Alternatively, with two third persons, we can get an inverse configuration (object outranks the subject; obv>prox, in>an). In this case, the peripheral suffix indexes the object.

    1. (20)
    1. Blackfoot (Frantz 2017: 62)
    1.  
    1. a.
    1. ots-ikákomimm-ok-a
    2. 3-loveta-inv-sg
    3. ‘she/theyobv love herprox’
    1.  
    1. b.
    1. ots-ikákomimm-ok-oaa-yi
    2. 3-loveta-inv-pl-pl
    3. ‘she/theyobv love themprox’
    1. (21)
    1. Plains Cree (Dahlstrom 1991: 23)
    1.  
    1. a.
    1. sêkih-iko-w-Ø
    2. frightenta-inv-3-prox.sg
    3. ‘sheobv/theyobv/it frighten herprox’
    1.  
    1. b.
    1. sêkih-ik-w-ak
    2. frightenta-inv-3-prox.sg
    3. ‘sheobv/theyobv/it frighten themprox’

Here, the (a) examples feature singular objects, the (b) examples feature plural objects, peripheral agreement varies in turn, and the number of the subject is ambiguous. These cases may seem to fly against the picture we’ve been building up so far: why are we skipping over the subject to agree with the object? The subject is third person, and thus should satisfy the probe’s needs.

Locality is restored once we realize that the 3>3 inverse crucially involves A movement of the object over the subject, following Bruening (2001; 2005; 2009), Bliss (2013), Hamilton (2015; 2018), Oxford (to appear), and Grishin (2022a), among others. The only assumption we need to make here is that this step of A movement happens before C probes, something that seems quite reasonable given C’s structural height and given the fact that CP is typically an Ā domain.

3.5 Syntactic interlude II: The syntax of the 3>3 inverse

To show that the inverse, but not the direct, involves A movement of the object over the subject, we can use the classic variable binding diagnostic: A movement should be able to feed variable binding. Put another way, while Ā movement of a quantifier can trigger Weak Crossover (WCO) effects, A movement of a quantifier does not (Postal 1971; Wasow 1972; Koopman & Sportiche 1983; a.m.o.). In both Blackfoot and Plains Cree, the 3>3 inverse reverses binding relations in exactly the way we’d expect from A movement.

Bliss (2013) demonstrates that in direct configurations in Blackfoot, universal quantifier subjects can bind into objects but not vice versa (22), whereas in the inverse, universal quantifier objects can bind into subjects but not vice versa (23). First let’s take a look at the direct configurations:

    1. (22)
    1. Blackfoot (Bliss 2013: 291)
    1.  
    1. a.
    1.   Amo-ksi
    2.   dem-pl
    1. aakííkoa-iksi
    2. girl-pl
    1. ohkaná-ísinao’sskip-ii-y=aawa
    2. all-kissta-3obj-pl=3pl.prn
    1. [
    2.  
    1. om-i
    2. dem-obv
    1. sááhkomaapi-i
    2. boy-obv
    1. ot-áákomimm-a-yi
    2. 3-loveta-3obj-obv
    1. proi
    2.  
    1. ].
    2.  
    1.   ‘Every girli kissed [the boy shei loved].’
    1.  
    1. b.
    1. *[
    2.  
    1. Ann-a
    2. dem-prox
    1. sááhkomaapi
    2. boy
    1. ot-áákomimm-ok-a
    2. 3-loveta-inv-prox
    1. proi
    2.  
    1. ]
    2.  
    1. ohkaná-ísinao’sskip-ii-y=aawa
    2. all-kissta-3obj-pl=3pl.prn
    1. amo-ksi
    2. dem-pl
    1. aakííkoa-iksi.
    2. girl-pl
    1.   ‘[The boy she*i,✔j loved] kissed every girli.’

In (22a) we have a proximate universal quantifier subject, ‘every girl’ (amoksi aakííkoaiks ‘the girls’ plus the quantificational prefix ohkana- on the verb) binding a variable contained inside a relative clause modifying the obviative object, omi sááhomaapii otáákomimmayi ‘the boy she loved’. This is acceptable under the bound reading, which is perhaps what one would expect given that subjects are first-merged c-commanding objects. In (22b), on the other hand, we are attempting to have an obviative universal quantifier object ‘every girl’17 bind into a proximate subject anna sááhomaapi otáákomimmoka ‘the boy that she loved’. The bound reading here is not available—only an unbound reading is.18 This indicates that in direct configurations, subject and object retain their first-merge c-command relationship.19

In the inverse, exactly the opposite state of affairs obtains:

    1. (23)
    1. Blackfoot (Bliss 2013: 291)
    1.  
    1. a.
    1.   Amo-ksi
    2.   dem-pl
    1. aakííkoa-iksi
    2. girl-pl
    1. ot-ohkaná-ísinao’sskip-ok-y=aawa
    2. 3-all-kissta-inv-pl=3pl.prn
    1. [
    2.  
    1. om-i
    2. dem-obv
    1. sááhkomaapi-i
    2. boy-obv
    1. ot-áákomimm-a-yi
    2. 3-loveta-3obj-obv
    1. proi
    2.  
    1. ].
    2.  
    1.   ‘[The boy shei loved] kissed every girli.’
    1.  
    1. b.
    1. *[
    2.  
    1. Ann-a
    2. dem-prox
    1. sááhkomaapi-wa
    2. boy-prox
    1. ot-áákomimm-ok-a
    2. 3-loveta-inv-prox
    1. proi
    2.  
    1. ]
    2.  
    1. ot-ohkaná-ísinao’sskip-ok-a
    2. 3-all-kissta-inv-sg
    1. ann-iksi
    2. dem-pl
    1. aakííkoa-iksi.
    2. girl-pl
    1.   ‘Every girli kissed [the boy she*i,✔j loved].’

In (23a) the proximate object is able to bind into the obviative subject, in stark contrast to what we saw above in the direct (22b). Conversely, as we can see in (23b) obviative subjects cannot bind into proximate objects, again contrasting with the direct (22a). This suggests that in the inverse, the object A moves above the subject, allowing the object to bind variables in the subject, and additionally suggests that the object cannot reconstruct for variable binding. Parallel conclusions are drawn from parallel data in other Algonquian languages by Bruening (2001; 2005; 2009) for Passamaquoddy-Wolastoqey, Hamilton (2015) for Mi’gmaq, and Oxford (to appear) for Oji-Cree.

Similar binding data can be adduced from the literature on wh questions and WCO in Plains Cree. Dahlstrom (1986; 1991) argues that Plains Cree does not display WCO effects with examples like the following:

    1. (24)
    1. a.
    1. Plains Cree (Dahlstrom 1986: 57)
    1.  
    1. a.
    1. Awînai
    2. who
    1. ê-sâkih-â-t
    2. ic-loveta-3obj-3cj
    1. [
    2.  
    1. proi
    2.  
    1. o-mâmâ-wa
    2. 3-mother-obv
    1. ]?
    2.  
    1. ‘Whoi loves hisi mother?’
    1.  
    1. b.
    1. Awînai
    2. who
    1. ê-sâkih-iko-t
    2. ic-loveta-inv-3cj
    1. [
    2.  
    1. proi
    2.  
    1. o-mâmâ-wa
    2. 3-mother-obv
    1. ]
    2.  
    1. ti?
    2.  
    1. ‘Whoi did hisi mother love ti?’

In (24a), we see that wh subjects can bind into the object, just like in English. However, in contrast to English, (24b) shows us that it’s possible to wh-extract the object and have it bind into the subject—this looks like it should result in a WCO violation, but it doesn’t.

But there’s an important confound here, as the above discussion (and my bolding) should have primed you to notice: in (24a) we have the direct marker -â (third person object agreement), but in (24b) we have the inverse marker -iko. In fact, all of the grammatical sentences in the literature on Plains Cree that are meant to show that Plains Cree lacks WCO feature the inverse, as Bruening (2001) notes. However, if the inverse involves the object A-moving over the subject, then that should allow the object to bind into the subject without any issues, just like in Blackfoot.

Interestingly, elsewhere in the literature, we can find examples of attempting to get wh objects to bind into subject in the direct, and these examples are ungrammatical, just like in Blackfoot: Blain (1997) provides the following data, which involve direct morphology and attempted “inverse” variable binding:

    1. (25)
    1. Plains Cree (Blain 1997: 216, 219)
    1.  
    1. a.
    1. *Awînai
    2.   who.prox
    1. [
    2.  
    1. proi
    2.  
    1. o-têm-a
    2. 3-dog-obv
    1. ]
    2.  
    1. kâ-nawaswât-â-yi-t
    2. rel-chaseta-3obj-obv.sbj-3cj
    1. ti?
    2.  
    1.   ‘Whoi is (it that) his*i,✔j dog (is) chasing ti?
    1.  
    1. b.
    1. *Awîni-hii
    2.   who-obv
    1. [
    2.  
    1. nâpêw
    2. man
    1. kâ-sâkih-â-t
    2. rel-loveta-3obj-3cj
    1. proi
    2.  
    1. ]
    2.  
    1. kâ-ocêm-â-t
    2. rel-kissta-3obj-3cj
    1. ti?
    2.  
    1.   ‘Whoi did [the man that loved her*i,✔j] kiss ti?

The result is ungrammatical. Thus, just like in Blackfoot, when you have direct morphology the WCO violation pops into view. This suggests that the inverse is a necessary ingredient in order to avoid a WCO violation, supporting the idea that the inverse involves object movement in Plains Cree.20

Example (25a) requires a bit of explanation: here, the matrix verb form indicates that we have a direct obv>obv configuration (from the combination of the direct marker -â (third person object agreement) and obviative subject agreement -yi). However, awîna is proximate, and given the verbal agreement, you might have expected obviative awînihi (note that awîna being proximate doesn’t automatically make the sentence unacceptable: it’s acceptable when awîna ‘who’ doesn’t corefer with the possessor of otêma ‘his dog’—i.e. when there’s no WCO violation). Blain says the following: “the default (proximate) form awîna ‘who’ is usually acceptable where there is no chance of ambiguity” (Blain 1997: 62). I think it’s conceivable that (25a) is a cleft—e.g. ‘who is it [that his dog is chasing]?’ (Blain actually argues that all Plains Cree wh questions are clefts)—and you get an “obviation reset” at the clause boundary.

Going back to the main point at hand: the variable binding and WCO data tell us that the inverse involves objects A moving over subjects. This results in the object now being the most local goal to the probe in C, resulting in the peripheral suffix indexing the object in the 3>3 inverse—a standard locality effect, just like in the 3>3 direct.

3.6 Summary of the data

Now that we’ve gone through all the relevant data, let’s take a step back and evaluate where we’ve gotten to. I’ve shown that the peripheral suffix in Blackfoot and Plains Cree displays an omnivorous third person pattern, agreeing with the closest third person, no matter if it’s the subject or object. A crucial point for this interpretation of the data is the idea that in the 3>3 inverse, the object A-moves over the subject. I provided independent justification for this based on the fact that WCO violations are obviated only in the inverse. Below, I sketch the various configurations. I use privative [3] and [part] features here, but that is not crucial—they could be replaced with [–part] and [+part], respectively, if one would like to adopt Nevins’s (2007) featural representations.

    1. (26)
    1. a.
    1.                                                                intransitive
    1.  
    1. b.
    1.                                          part>3
    1.  
    1. c.
    1.                                          3>part
    1.  
    1. d.
    1.                                          3>3 direct
    1.  
    1. e.
    1.                    3>3 inverse

The crucial data point here is thus part>3 configurations (26b), as those are the only configurations in which C isn’t just agreeing with the most local goal—there, C must skip over the part subject in preference for the third person object.21 So: omnivorous third person agreement does in fact exist, and thus we need a third person feature in the syntax, contra Preminger (2019). It seems we must favor 3yesP over 3noP.

4 Against alternative analyses

Or do we? Before we jump to conclusions, we need to do our due diligence and consider alternative analyses that can capture this data without abandoning (syntactic) 3noP. I’ll consider three such plausible attempts here: (i) agreement for animacy, (ii) agreement for obviation, and (iii) agreement for D features. To spoil the plot: agreement for animacy is straightforwardly falsified by the fact that first and second person must have syntactically-active animacy features, agreement for obviation fails no matter what your analysis of obviation is, and agreement for D features is falsified by a lack of asymmetry in the structure of part and third person pronouns in Blackfoot and Plains Cree (Wiltschko et al. 2015; Déchaine et al. 2015).22

4.1 Agreement for animacy?

The idea goes like this: first and second person are not contrastive for animacy, as they’re always animate, but we do get animacy contrasts for third person. So, we don’t need to specify first and second person for animacy, but we do need to do so for third person. Then, we could say that C probes specifically for animacy features, and this will result in the surface appearance of a third-person-only probe.

Unfortunately, that doesn’t work, as we can show that first and second person must be specified for animacy in the syntax. The relevant data comes from the distribution of finals—derivational morphology found across Algonquian that marks the transitivity of the verb and indexes the animacy of the intransitive subject and transitive object. To get a feel for the basic pattern, consider the Menominee verbs in (27). All of these verbs share the same root, panât- ‘be spoiled, spoil’, and they differ in terms of which finals they involve (there are many more finals than just these four, and finals rather idiosyncratically select for different roots, as well as often bear more concrete, lexical meaning, behaviors characteristic of derivational rather than inflectional morphology).

    1. (27)
    1. Menominee (Bloomfield 1962: 330)
    1.  
    1. a.
    1. panât-at(-w)
    2. be.spoiled-II-3
    3. ‘it is spoiled’
    1.  
    1. b.
    1. panât-ese-w
    2. be.spoiled-AI-3
    3. ‘she is spoiled’
    1.  
    1. c.
    1. panâc-eht-a-w
    2. spoil-TI-in.obj-3
    3. ‘she spoils it’
    1.  
    1. d.
    1. panâc-eh-æ̂-w
    2. spoil-TA-3obj-3
    3. ‘she spoils her’

For this particular root, Bloomfield provides forms with four distinct finals:

  • panâtat-, with the final -at, indicating an intransitive verb with an inanimate subject (inanimate intransitive, II);

  • panâtese-, with the final -ese, indicating an intransitive verb with an animate subject (animate intransitive, AI);

  • panâceht-, with the final -eht (which palatalizes the preceding /t/), indicating a transitive verb with an inanimate object (transitive inanimate, TI);

  • panâceh-, with the final -eh (which palatalizes the preceding /t/), indicating a transitive verb with an animate object (transitive animate, TA).

As should be evident, finals appear to be sensitive to the animacy features of what would essentially be the “absolutive” argument if Algonquian languages featured ergative case marking, to echo the way Rhodes (1976: 80) puts it.

At this point, we need to ask two questions: (i) is the selection of finals really sensitive to morphosyntactic animacy features, rather than notional animacy? and (ii) if only one of inanimate and animate is specified for animacy, which one is it?

The answer to question (i) is yes: the selection of finals is sensitive to grammatical animacy, not notional animacy. While many nouns across Algonquian are predictably animate or inanimate—for instance, nouns referring to notionally animate entities like people and animals are overwhelmingly animate23—there are several cases where animacy must simply be lexically-specified, most commonly for notionally inanimate nouns that are grammatically animate (see Dahlstrom 1995; Goddard 2002; and Quinn 2019 for nuanced discussion of animacy in Algonquian). Below I provide examples of unpredictably animate/inanimate nouns in Meskwaki (Dahlstrom 2015: 3–6, 3–7, 3–8):

    1. (28)
    1. Inanimates
    1.  
    1. meški
    2. anîpi
    3. ahtêhimini
    4. čîmâni
    5. trêkitêhi
    1. ‘blood’
    2. ‘elm’
    3. ‘strawberry’
    4. ‘canoe’
    5. ‘tractor’
    1. (29)
    1. Animates
    1.  
    1. atôwa
    2. meškwâwâhkwa
    3. wîtawîha
    4. atâpyâna
    5. âtamôpîna
    1. ‘blood clot’
    2. ‘red cedar’
    3. ‘raspberry’
    4. ‘wagon’
    5. ‘automobile’

As should be evident from comparison of the inanimates in (28) to the animates in (29), it’s not clear what, if anything, could semantically distinguish the two sets of nouns. This behavior is characteristic of animacy across the family.

And when we put these nouns into a sentence, the verb stem (verb root plus finals) tracks grammatical animacy:

    1. (30)
    1. Meskwaki (Dahlstrom 2015: 3–9)
    1.  
    1. a.
    1. Ne-mîči-p-ena
    2. 1-eat.TI-P-1pl
    1. ahtêhimin-ani.
    2. strawberry-in.pl
    1. ‘Weexc ate strawberries (an).’
    1.  
    1. b.
    1. Net-amw-â-p-ena
    2. 1-eat.TA-3obj-P-1pl
    1. wîtawîh-aki.
    2. raspberry-prox.pl
    1. ‘Weexc ate raspberries (in).’

In (30), we have two suppletive verbs, mîči- ‘eat in’and amw- ‘eat an’—we might imagine treating these as a portmanteau of an abstract EAT root plus a TI or TA final. Here, the choice of which verb stem to use is dependent on the grammatical animacy of the object, rather than its notional animacy.

The answer to question (ii) is that inanimate is more likely to be underspecified for animacy. Dahlstrom (2015) argues that “inanimate is the semantically unmarked, elsewhere category of the gender system” (Dahlstrom 2015: 3–11). She notes that, in contexts where you have a single (null) pronoun referring to a mixed set of inanimate and animate nouns, the result is inanimate: in the Meskwaki story wisakea osani okyeni osimeani okomeseani [Wisahkeha, his father, his mother, his younger brother, his grandmother], told by Alfred Kiyana (1913), we find the form mîčikwêna ‘whoever ate themin’, using the TI stem mîči- ‘eat’, to refer to a person who ate strawberries (inanimate) and fish (animate) (Dahlstrom 2015: 3–11).24 Goddard (2002) makes a similar point, noting that the inanimate indefinite pronoun kêkôhi ‘something’ can be used in contexts where it must be referring to an animate entity, providing the following example, coming from the story “Wapasaya’s Younger Brother” by Alfred Kiyana, where someone admits that they have not killed a (most likely human) enemy:

    1. (31)
    1. Meskwaki (Goddard 2002: 221)
    1.  
    1. Âkwi
    2. neg
    1. =mâh
    2. =you.see
    1. =nîna
    2. =1sg
    1. kêkôhi
    2. anything
    1. nehtô-yân-ini.
    2. killTI-1sg.cj-neg
    1. ‘I have not killed anything.’

The conclusion that Dahlstrom and Goddard draw from this kind of data is that inanimates must be unmarked. They can in principle refer to both animate and inanimate referents (as revealed by these examples), but inanimate markers are (usually) restricted to appearing with non-animates only because of some kind of morphological competition—for instance, the Subset Principle in Distributed Morphology (Halle 1997).

So, back to the original question: can we show that first and second person are specified for animacy? The fact: when you have a first or second person subject of an intransitive, or a first or second person object of a transitive, you use an animate final:

    1. (32)
    1. Menominee (Bloomfield 1962: 151, 155)
    1.  
    1. a.
    1. ne-pâhp-æhcenæ̂-m
    2. 1-fall-snag.AI-M
    3. ‘I fall’
    1.  
    1. b.
    1. ne-tæp-ahw-ek(-w)
    2. 1-pay-by.instrument.TA-inv-W
    3. ‘he pays me’

In both these examples we have an animate final: the AI final -æhcenæ̂ in (32a), which Bloomfield glosses as ‘snag, immerse’ (Bloomfield 1962: 315), and the TA final -ahw in (32b), which indicates the use of an instrument (Will Oxford, p.c.)—here, the instrument would presumably be money.

Since finals track grammatical animacy, and animate is the specified value out of the pair animate-inanimate, we must conclude that the finals here are sensitive to a specified [an] feature on first and second person DPs. Thus, first and second person are specified for animacy. If so, we cannot analyze the peripheral suffix as probing for animacy, as then it shouldn’t be able to skip over first and second person DPs, counter to fact.25

4.2 Agreement for obviation?

Another attempt: maybe C probes for obviation features, and first and second person are not specified for obviation? This kind of analysis is somewhat tricky to respond to, since there is no general consensus on what exactly obviation is, and there are several distinct proposals in the literature for how it’s represented morphosyntactically. Here, I’ll focus on three kinds of views, and show that none of them is able to formulate an unproblematic response to the omnivorous third person pattern. The three views are:26

  1. Proximate and obviative are in principle contrastive on all persons (Hammerly 2020).

  2. Proximate and obviative are only contrastive on third persons, and…

    1. …first and second person are always proximate (Bruening 2001; Lochbihler 2012; Oxford 2019).

    2. …first and second person are not specified for obviation (Bondarenko 2020).

For reference, let us call the first view Contrast Everywhere and the second 3 Only. There are two kinds of 3 Only we find in the literature: first and second person are always proximate (iia), and first and second person are not specified for obviation (iib). I’ll address these views in turn, considering how they would implement the obviation agreement idea, and arguing that the obviation agreement account of the apparent omnivorous third person pattern fails in each case. Contrast Everywhere and the first subtype of 3 Only (part is proximate) only make the argument for 3yesP stronger, and the second subtype of 3 Only reduces to positing a third person feature.

4.2.1 Contrast Everywhere

If we adopt Contrast Everywhere, we could say the following about the omnivorous third person pattern we find in Algonquian: in these languages, it’s just an accident that there is no obviation contrast on first and second person, in a similar way to how some languages lack gender/noun class distinctions in the first and second person. Then, we could say that the surface appearance of omnivorous third person agreement is actually the confluence of two independent factors: (i) first and second person just happen to not have obviation features in these languages; and (ii) C is probing for obviation features.

However, the core issue here is Blackfoot. Blackfoot displays the omnivorous third person agreement pattern, as we’ve seen, yet Blackfoot is the clearest example we have of a language that contrasts obviation on first and second persons.27 The relevant data are the independent pronouns, listed in Table 3 which show a contrast that uses the same suffixes found elsewhere as proximate (-wa) and obviative (-yi) markers.28

Table 3

Blackfoot independent pronouns (Wiltschko et al. 2015: 266).

prox? obv?
1 n-iistó-wa n-iistó-yi
2 k-iistó-wa k-iistó-yi
3 o-(ii)stó-wa o-(ii)stóa-yi
1exc n-iistó-nnaan-wa n-iistó-nnaan-yi
1inc k-iistó-nnoon-wa k-iistó-nnoon-yi
2pl k-iistó-waaw-wa k-iistó-waaw-yi
3pl o-(ii)stó-waawa-wa o-(ii)stó-waawa-yi

I hedge here because it’s not actually obvious that -wa and -yi really mark obviation on the independent pronouns, as their syntactic distribution doesn’t seem to obviously mirror the syntactic distribution of obviation on third persons. As Bliss (2013) puts it, “it is yet unclear what determines whether a local independent pronoun is marked as proximate or obviative” (Bliss 2013: 253); see also the discussion by Wiltschko et al. (2015).

Setting all that aside, let’s accept for the purposes of argument that Blackfoot distinguishes between proximate and obviative on first and second person. Thus, first and second person need to have obviation features (or at least, the overt pronouns need to be specified for obviation features). The problem: if you have a sentence with a first or second person overt pronoun and a third person noun phrase, the peripheral suffix will still always agree with the third person noun phrase in obviation and number:

    1. (33)
    1. Blackfoot (Wiltschko et al. 2015: 276)
    1.  
    1. a.
    1. N-iistó-wa
    2. 1-prn-prox?
    1. nit-ákomimm-a-yini
    2. 1-love-3obj-obv.sg
    1. an-i
    2. dem-obv
    1. nináá-yi.
    2. man-obv
    1. part >
    2.  
    1. ‘Iprox? love that manobv.’
    1.  
    1. b.
    1. Om-a
    2. dem-prox
    1. nináá-wa
    2. man-prox
    1. nit-ákomimm-ok-wa
    2. 1-love-inv-prox.sg
    1. n-iistó-yi.
    2. 1-prn-obv?
    1. > part
    2.  
    1. ‘That manprox loves meobv?.’

Thus, even though we could say the Plains Cree (and all the other Algonquian languages with the same peripheral suffix pattern) only “accidentally” shows omnivorous third person agreement, Contrast Everywhere forces us into saying that the Blackfoot peripheral suffix must be a true instance of omnivorous third person. Thus, if you accept Contrast Everywhere, you must also accept 3yesP.

4.2.2 3 Only

However, let’s say that you dispute that Blackfoot poses a real counterexample to the idea that the proximate-obviative distinction is limited to third persons only—thus, you accept 3 Only. If you follow Bruening (2001), Lochbihler (2012), and Oxford (2019) (among others) in saying that first and second person are always proximate, then the argument against obviation agreement is even stronger. This is because in part>3 contexts in which the third person object is proximate, both arguments would thus be proximate, and yet C still indexes the more distant third person object (11–12, repeated below):

    1. (34)
    1. Blackfoot (Frantz 2017: 46–47, 56–57)
    1.  
    1. a.
    1. nits-ikákomimm-a-nnaan-a
    2. 1-loveTA-3obj-1exc-sg
    3. ‘weexc love her’
    1.  
    1. b.
    1. nits-ikákomimm-a-nnaan-i
    2. 1-loveTA-3obj-1exc-pl
    3. ‘weexc love them’
    1.  
    1. c.
    1. nits-ííkoon-ii-hp-innaan-a
    2. 1-take.downTA-in.obj-P-1exc-sg
    3. ‘weexc took it down’
    1.  
    1. d.
    1. nits-ííkoon-ii-hp-innaan-i
    2. 1-take.downTA-in.obj-P-1exc-pl
    3. ‘weexc took themin down’
    1. (35)
    1. Plains Cree (Okimāsis 2018: 145–146)
    1.  
    1. a.
    1. ni-mow-â-nân-Ø
    2. 1-eatTA-3obj-1exc-prox.sg
    3. ‘weexc eat itan.prox’
    1.  
    1. b.
    1. ni-mow-â-nân-ak
    2. 1-eatTA-3obj-1exc-prox.pl
    3. ‘weexc eat theman.prox’

In these forms, if the first person subject were proximate, there’s no reason why C shouldn’t stop probing at the subject. Thus, in order to get C to ignore the first person subject and instead agree with the third person object, you’d have to say that C isn’t probing for proximate/obviative features, but rather third person. Under this version of 3 Only, the obviation agreement analysis simply fails to account for the data.

However, there’s another version of 3 Only out there: we could say that first and second person are just not specified for proximate or obviative, following Bondarenko (2020). She provides the following feature geometry for person:29

    1. (36)

And then in order to get C to probe for obviation features specifically, we would need to specify it as probing for [non-part]. But isn’t that just restating the third person agreement analysis? This is the crucial conceptual issue with the obviative agreement analysis under this version of 3 Only: how is it distinct from a third person probe? Bondarenko (2020) understands that [non-part] is just a third person feature, and follows Nevins (2007) in accepting 3yesP: for her, obviative and proximate are subnodes of third person.

To summarize the discussion about agreement for obviation: I’ve outlined three kinds of views of the featural representation of obviation. Contrast Everywhere states that person and obviation are orthogonal, just like person and gender, and that obviation could in principle be contrastive on all persons. If we accept this, then Blackfoot can only be analyzed as showing agreement for third person specifically, and not obviation, as Blackfoot independent first and second person pronouns show an (apparent) obviation contrast—yet the peripheral suffix still skips over them in preference for third persons. Alternatively, we could adopt 3 Only, which states that obviation is only contrastive in third persons. If we think that first and second person are always proximate, then first and second person should be visible to an obviation probe, and thus the fact that C can skip over part subjects to agree with third person objects could only be explained under a third person agreement analysis. Alternatively, if we believe that first and second person are universally not featurally specified for obviation, then it’s unclear in what sense an obviation probe would be distinct from a third person probe, as Bondarenko (2020) notes.

4.3 Agreement for D?

So, I’ve argued that we can’t reanalyze the agreement pattern discussed here as an animacy or obviation probe. There is one last option to consider, which is suggested by Oxford (2014; 2017b) and Xu (2021): agreement for the categorial feature [D].

Oxford (2014), in solving a problem related to his featural representations for various kinds of DPs in (Proto-)Algonquian, proposes that third person nominals are DPs, whereas first and second person nominals are φPs.30 Xu (2021) then capitalizes on this idea in accounting for the peripheral agreement pattern in Menominee (another language with the same peripheral agreement pattern as Blackfoot and Plains Cree).

Oxford provides the following independent argument for this conclusion: first and second person do not participate in obviation and absentative marking, unlike third person, and he proposes that we can understand these facts if those features must be located in D, and first and second person lack D. However, he doesn’t provide independent evidence that obviation and absentative features must be found in D, so this argument ends up somewhat circular.

While this proposal technically works, it doesn’t seem to hold up to the available evidence. Following diagnostics from Déchaine & Wiltschko (2002), Wiltschko et al. (2015) argue that all Blackfoot independent pronouns are φPs, and Déchaine et al. (2015) argue that all Plains Cree independent pronouns are DPs. Neither conclusion involves a structural asymmetry between part and third person pronouns, counter to what the D agreement analysis would want. Oxford (2017b) acknowledges this challenge, and suggests that “since emphatic pronouns are excluded from clausal argument positions, they do not participate directly in subject/object agreement” (Oxford 2017b: 18).

While it is true that overt pronouns in Algonquian languages very often appear in clause-external topic and/or focus positions—a fact that follows naturally from the fact that Algonquian languages are rampantly pro-drop—proposing that Algonquian overt pronouns cannot occupy clausal argument positions at all is too strong, as there are cases where overt pronouns do appear in clause-medial positions. For instance, in Plains Cree, overt pronouns can occur as a kind of “resumptive” element in a left-dislocation construction:

    1. (37)
    1. Plains Cree (Déchaine et al. 2015: 44, 51)
    1.  
    1. a.
    1. …mâka
    2. but
    1. John
    2. John
    1. wîya
    2. 3sg
    1. miywêyiht-am
    2. likeTI-in.obj
    1. êkw’-anima.
    2. top-that.in
    1. ‘…but John, he likes that.’
    1.  
    1. b.
    1. Nîsta
    2. 1sg.too
    1. nîya
    2. 1sg
    1. mîna
    2. also
    1. ni-wî-tôhtâ-n.
    2. 1-fut-goAI-part.sg
    1. ‘Me too, I’m also going to go too.’

In (37a), the third person pronoun wîya resumes the contrastively-focused John, and in (37b), the first person pronoun nîya resumes the additive pronoun nîsta ‘me too’, a contrastive topic. In these examples John and nîsta are clearly occupying left-peripheral Ā positions, but the pronouns wîya or nîya look more like they are clausal-internal, resuming the left-dislocated constituents.

We also see evidence of clause-medial positioning of overt pronouns in Meskwaki (which has the same peripheral agreement pattern as Blackfoot and Plains Cree): the emphatic series of independent pronouns can occur quite low, between preverb and verb (Dahlstrom 1988)—a use Dahlstrom calls “enclitic”:

    1. (38)
    1. Meskwaki (Dahlstrom 1988: 168)
    1.  
    1. Ne-kwayâši
    2. 1-already
    1. =kêh
    2. =but
    1. =nînâna
    2. =we.exc
    1. anehkâtî-pena.
    2. be.acquaintedAI-1pl
    1. ‘But we were already acquainted with each other.’

Here, the pronoun occurs below an aspectual modifier kwayâši ‘already’ and before the verb root anehkâti ‘be acquainted with each other’. Again we see that overt pronouns can appear in clause-medial positions, contra Oxford’s suggestion.31

It seems to me that the simplest conclusion is to accept the null hypothesis: overt pronouns can indeed occur in clausal argument positions. While they are usually interpreted as topics/foci and therefore usually appear in sentence-peripheral topic/focus positions (Dahlstrom 2017), they can appear overtly even in situ. I conclude that we do actually get agreement with overt pronouns in Algonquian, and therefore, following Wiltschko et al. (2015) and Déchaine et al. (2015) who find no syntactic difference between part and third person pronouns in Blackfoot and Plains Cree, the D agreement account cannot properly derive C’s preference for third person.

So: the alternative analyses of the omnivorous third person pattern have failed. We must accept 3yesP, even in the syntax.

5 Possible ways of working out the details

It is not the goal of this paper to advocate for a specific feature theory. My aim is more modest: to show that the omnivorous third person behavior of the peripheral suffix in Algonquian narrows down the set of existing (and future) theories of φ features to those compatible with 3yesP. I leave any further winnowing of theories to future work.

However, it is worth taking the time at this point to consider various possible feature theories compatible with the data presented here, how they can account for an omnivorous third person probe, as well as the typological predictions they make. I’ll discuss three pre-existing 3yesP-compatible ways of working things out: (i) Bondarenko’s (2020) privative feature geometry (39);32 (ii) Nevins’s (2007) binary feature system (40); and (iii) Harbour’s (2016) binary feature system (41):

    1. (39)
    1.               (Bondarenko 2020: 10)
    1. (40)
    1. a. [+auth, +part]: first person exclusive              (Nevins 2007: 305)
    2. b. [+auth, +part, addr]: first person inclusive
    3. c. [–auth, +part, addr]: second person
    4. d. [–auth, –part]: third person
    1. (41)
    1. a. [–part, +auth]: first person exclusive              (Harbour 2016: 96)
    2. b. [+part, +auth]: first person inclusive
    3. c. [+part, –auth]: second person
    4. d. [–part, –auth]: third person

While Bondarenko (2020) doesn’t explicitly present how her feature geometry maps to different person specifications, it is clear from her discussion that this is what she intends:

    1. (42)
    1. a. [π, part, auth]: first person exclusive
    2. b. [π, part, auth, addr]: first person inclusive
    3. c. [π, part, addr]: second person
    4. d. [π, 3(, prox)]: proximate third person
    5. e. [π, 3, obv]: obviative third person

Note that a crucial assumption here is that [part] and [3] cannot appear in the same feature bundle: this would correctly derive restriction of the feature [3] to third persons only, and also derives a restriction of the obviation contrast to third persons only—putting [obv] on a [part] feature bundle would entail that that feature bundle also contains [3], which is impossible.33

First, some broad discussion of these different feature systems. Bondarenko’s system is essentially a standard privative feature geometry (à la Harley & Ritter 2002 or Béjar & Rezac 2003 or a host of others) with a 3 node tacked on. As in a standard feature geometry, the existence of a lower node N on a feature bundle entails the existence of all of the nodes that dominate N: thus, if you contain [obv], then you also contain [3] and [π].34 As mentioned above, Bondarenko’s system needs some way of preventing [part] and [3] from appearing in the same feature bundle—as far as I can tell, this simply has to be stipulated. While this could be derived semantically, by endowing [part] with the semantics ‘includes a speech act participant’ and [3] with ‘does not include a speech act participant’, that would predict that we can’t use third person forms to include speaker and addressee, counter to fact—for instance, generalizations involving third person expressions like all humans usually are meant to include speaker and addressee.35

Nevins’s system is a classic binary feature system with [±part] and [±auth] features that has a privative [addr] added on to account for the exclusive-inclusive contrast; the fact that [addr] is privative while [±part] and [±auth] are binary is exploited by Nevins to account for certain asymmetries between first and second person. Nevins adopts a standard semantics for [±part] and [±auth]: [+part] entails that the referent contains a speech-act participant, and [+auth] entails that the referent contains the author (and the negative values are not given a semantics).

Finally, Harbour’s system is a modification of the classic part-auth binary feature system to allow those two features alone to generate the “standard quadripartition” (the 1exc-1inc-2–3 contrast). For him, the semantics of the features involves operations on the person lattice (a representation of all possible referents) that are crucially ordered: for instance, putting things very roughly, [+part] adds all speech act participants to the representation of person, and [–part] removes all speech act participants (I refer the reader to Harbour 2016 for the formal details). Thus, a representation like [–part, +auth] (for first person exclusive) involves first removing all speech act participants from the representation of person, then adding back in only the speaker—resulting in a representation with only the speaker, and not the addressee, deriving first person exclusive. Harbour’s system is thus unusual in that first person exclusive is specified [–part].

How could we implement an omnivorous third person probe in each system? In the Bondarenko system, the answer is simple: we just define a [u3] probe: it’ll look for and copy features from the closest accessible goal with the feature 3. In the Nevins and Harbour systems, which are binary, we need to be able to define probes that are relativized to negative features, [u–F]. Once we allow that, we can specify a probe [u–part] in the Nevins system, which will pick out only third persons. Note that we can’t use that probe in the Harbour system, as the representation of first person exclusive also contains the feature [–part]. Instead, we would have to make use of a composite probe (Coon & Bale 2014; van Urk 2015), [u–part,–auth], which looks for goals contiaining both features: this would result in a probe that skips over potential goals that only contain [–part] (e.g. first person exclusive) or only contain [–auth] (e.g. second person), instead agreeing with the closest goal with both those features—in other words, the closest third person.

If we adopt a binary feature system and allow ourselves the ability to relativize goals to negative features, we might wonder about the typological predictions we make. For instance, are there probes relativized to [–auth], which would result in omnivorous 2/3 agreement, skipping over first persons? Or if we adopt Harbour’s system, where first person exclusive is specified [–part], can we have a probe relativized to [–part] only, which would result in omnivorous 1exc/3 agreement, skipping over first inclusive and second person? To the best of my knowledge, such systems do not exist, though of course future research and insights from the descriptive and/or typological literature may yet fill out these potential cells of the typology. Assuming however that such probes should be ruled out on principled grounds, this would constitute an argument against the binary feature systems in preference for a Bondarenko-type privative system, as her system wouldn’t make those kinds of overpredictions.36

What about obviation and animacy? Following Harley & Ritter (2002), we can simply say that noun class features (like gender and animacy) are a distinct node from person features in the feature geometry, something like the following:

    1. (43)

We would just need to ensure that first and second person are always specified for the feature an, as discussed in Section 4.1. It is worth noting that my analysis is incompatible with the analysis of inanimates given by Lochbihler (2012) and Oxford (2014), who propose that inanimates are wholly unspecified for person features (in contrast to third person animates, which have a Pers feature). Since for me inanimates need to be specified for third person so that they can be targeted by the third person probe on C, we cannot say that inanimates wholly lack person features. The analysis of animacy features as orthogonal to person features (Harley & Ritter 2002) correctly allows inanimates to be visible to C under my proposal.

As for obviation, Bondarenko (2020) proposes that obviation features are a subnode of third person features, but Nevins (2007) and Harbour (2016) do not address obviation. There are a wide range of often rather strikingly divergent proposals for how to analyze obviation (Bruening 2001; Richards 2010; Lochbihler 2012; Oxford 2017a; 2019; Hammerly 2020; a.o.)—my proposal that the probe on C in Blackfoot and Plains Cree is a third person probe is compatible with a wide range of theories of the featural representation of obviation, as long as C can copy obviation features onto itself. Importantly, however, since C is just specified for third person, in order for it to copy animacy and obviation features, we need to say that agreement is featurally coarse: the probe copies all the features of the goal, not just the ones specified by the probe (Preminger 2014; Hammerly 2020; 2021).37

To recap: I’ve shown how there are a wide range of theories of person that are compatible with my proposal for peripheral agreement in Blackfoot and Plains Cree as an omnivorous third person probe. We can specify the probe in C as [u3] under Bondarenko’s (2020) feature system, [u–part] under Nevins’s (2007) feature system, and a composite probe [u–part,–auth] under Harbour’s (2016) feature system. In all cases, feature copying must be coarse, copying over animacy and obviation features as well as person features.

6 Consequences and extensions

To end the discussion, I would like to briefly take up two additional relevant threads that I think are worth bringing up, even if only cursorily and somewhat incompletely: the notion of the defaultness of third person, and capturing the range of variation in peripheral agreement across Algonquian.

6.1 Defaultness

It has often been noted (rightly, I think) that third person occupies some kind of default status in language:

    1. (44)
    1. Defaultness
    2. In contexts that don’t seem to involve any person features at all, like expletives and the realization of failed agreement (default agreement), we invariably get third person forms in language after language.

3noP, the idea that third person is underspecified, offers a natural account of Defaultness: in contexts that lack person features, we can only insert underspecified exponents, and underspecified exponents are (definitionally) third person.

While underspecification is a compelling and intuitive way to analyze Defaultness, it’s not the only formal tool we can use to analyze default effects—a conclusion that has been richly explored in the phonological literature on markedness (e.g. Calabrese 1995; de Lacy 2006; a.m.o.)—see also the discussion by Nevins (2007) in the domain of person features. In that sense, Defaultness is not as compelling an argument for 3noP as it might seem. One very basic option to capture Defaultness in a privative 3yesP system is to say that in contexts that truly lack person features, like expletives and default/failed agreement, you insert negative values of binary features as a default. This kind of rule would have to simply be stipulated in the system, and perhaps one might worry about its explanatory adequacy, but it would work to capture the fact that expletives and default agreement are invariably third person. Note that Preminger (2019) also stipulates this kind of rule in order to convert underspecified, privative syntactic representations of person to binary representations in the morphology. Another alternative, proposed by Ackema & Neeleman (2013; 2018; 2019), is to appeal to semantics, noting that the semantics of first and second person would be incompatible in default contexts which lack referents.38 I don’t take a strong stance here on what the right analysis of Defaultness should be—I just want to note that underspecification doesn’t have a monopoly on Defaultness. The real prediction of 3noP is Invisibility, and I hope to have shown that that Invisibility is wrong even in the syntax, as omnivorous third person probes do exist.

6.2 The cross-Algonquian picture

So far, I have only mentioned the cross-Algonquian picture in Section 3.3, when discussing the syntax of mixed configurations across Algonquian. There, I noted that other Algonquian languages share the same agreement pattern as Blackfoot and Plains Cree in configurations with only one third person argument: C agrees omnivorously with the third person. However, striking differences emerge when considering configurations with multiple third person arguments. In this section, I’d like to provide a sketch of how we could capture the variation across the family in a unified way under the assumption that C is an omnivorous third person probe.

The standard result when there are multiple matching accessible goals is a locality effect: all else being equal, probes agree with the closest matching goal. This is exactly what we saw in Blackfoot and Plains Cree, with C agreeing with the highest third person in 3>3 configurations (the subject in 3>3 direct, the object in 3>3 inverse). But, strikingly, other Algonquian languages show the exact opposite pattern, seeming to prefer agreeing with the lower third person. Let’s consider the following examples from Southwestern Ojibwe (Nichols 1980) and Passamaquoddy-Wolastoqey (Sherwood 1983) to see this in action:

    1. (45)
    1. Southwestern Ojibwe (Nichols 1980: 289, 292)
    1.  
    1. a.
    1. o-waapam-aa-waa-n           dir
    2. 3-seeTA-3obj-pl-obv
    1. ‘theyprox see herobv’
    1.  
    1. b.
    1. o-waapam-iko-waa-n         inv
    2. 3-seeTA-inv-pl-obv
    1. ‘sheobv sees themprox’
    1. (46)
    1. Passamaquoddy-Wolastoqey (Sherwood 1983: 217, 219)
    1.  
    1. a.
    1. Ø-wicuhkem-a-wa-l           dir
    2. 3-helpTA-3obj-pl-obv.sg
    1. ‘theyprox help herobv’
    1.  
    1. b.
    1. Ø-wicuhkem-ku-wa-l         inv
    2. 3-helpTA-inv-pl-obv.sg
    1. ‘sheobv helps themprox’

In the (a) examples, we have direct configurations, and C agrees with the object; in the (b) examples, we have inverse configurations, and C agrees with the subject. This is exactly the opposite of what we saw with Blackfoot and Plains Cree: here, we seem to violate standard assumptions about locality by indexing the lower goal.

Thus, we can make the following split between Highest and Lowest languages: languages where C indexes the highest third person, like Blackfoot and Plains Cree, and languages where C indexes the lowest third person, like Ojibwe and Passamaquoddy-Wolastoqey.39 So far, we’ve seen how an omnivorous third person probe [u3] can capture the behavior of C in Highest languages—what about Lowest languages? How should we analyze lowest-preference?

Grishin (2022b) argues that various existing analyses of lowest-preference—a low probe with Cyclic Expansion (Béjar 2003; Rezac 2003; Béjar & Rezac 2009), syntactic inversion (Myler 2017; Colley 2018), case-sensitivity with ergative-absolutive case (Bobaljik 2008; Preminger 2014), and the Activity Condition (Oxford 2014; Hammerly 2020; 2021; Xu 2021)—all fail to capture all of the facts of Algonquian Lowest languages. Instead, he proposes to analyze Lowest languages as having an insatiable third person probe in an interaction-satisfaction theory of Agree (Deal 2015; 2021), agreeing with all accessible third persons in the clause, resulting an a multiply-valued probe. Drawing a parallel with multiple case-assignment, where there are languages that only expone the outermost, last-assigned case (like Niuean in hyperraising; Seiter 1980; Massam 1985; Béjar & Massam 1999; see also Pesetsky 2013), he proposes that a similar parametrization can apply in the case of probes as well, with there being a general microparameter Expone Outermost40 that regulates the exponence of a multiply-valued head, whether multiply-valued for case or phi features:41

    1. (47)
    1. Expone Outermost
    2. Given a multiply-valued head H⟨[A], [B], …, [C]⟩, expone only the outermost feature bundle [C].

On this view, the Highest languages have a C probe whose satisfaction condition is identical to its interaction condition, [int:3, sat:3], resulting in copying the features from only the highest third person, but the Lowest languages all involve insatiable probes, [int:3, sat:—], agreeing with all accessible third persons, along with a rule of Expone Outermost. The result: C in Highest languages will show agreement with the highest third person, and C in Lowest languages will show (overt) agreement with the lowest third person—but they’ve actually interacted with all third persons on the way there. Thus, we can derive the contrast between Highest and Lowest languages by simply modulating properties of the probe on C. In all cases, however, the probe is restricted to third persons only: we must accept 3yesP.

7 Conclusion

A classic argument for 3noP found in the literature is Invisibility: the idea that no morphosyntactic process can explicitly target third person to the exclusion of first and second. There have been arguments that morphological Invisibility is wrong (Nevins 2007; Trommer 2008)—there are morphological processes (like Impoverishment) that need to have access to a distinct third person featural representation. This has led some to propose a revised view of 3noP—syntactic 3noP—which states that third person is underspecified specifically in the syntax (Preminger 2019). Here, I’ve presented a counterexample to syntactic Invisibility—the omnivorous third person agreement pattern found in Algonquian, specifically focusing on the pattern we see in languages like Blackfoot and Plains Cree. I then showed that various alternative analyses that could save syntactic 3noP—agreement for animacy, obviation, or [D] features—failed in various ways. Animacy agreement is straightforwardly falsified by the behavior of finals, obviation agreement fails under various analyses of the representation of obviation—either it fails to predict the data (Contrast Everywhere and one kind of 3 Only), or it reduces to third person agreement (the other kind of 3 Only)—and D agreement is falsified by the syntax of Blackfoot and Plains Cree pronouns (Wiltschko et al. 2015; Déchaine et al. 2015). I conclude that we really do have omnivorous third person agreement in Blackfoot and Plains Cree (and the rest of Algonquian). Syntactic 3noP cannot be upheld: we must accept 3yesP.

Notes

  1. There is a clear diachronic explanation for this: the peripheral suffix was originally a nominal suffix that marked a nominal’s animacy, number, and obviation status—thus, it only ever appeared on third person noun phrases. The Proto-Algonquian independent order can be internally reconstructed as arising from nominalized verbs (Goddard 2007), thus accounting for the appearance of nominal morphology on verbs and the restriction of the peripheral suffix to third person forms only. In fact, all of the agreement structurally outside of the reflexes of these historic nominalizers (the “formative” suffixes) is historically derived from nominal morphology, and in many Algonquian languages is still synchronically identical to nominal morphology. [^]
  2. C agreeing with the highest third person is widespread across Central Algonquian, found also in Menominee (Bloomfield 1962), Meskwaki (Goddard 1994; 1995; Dahlstrom 2015), Miami-Illinois (Costa 2003), and across the Cree-Innu-Naskapi dialect continuum (e.g. in Innu; Clarke 1982; Drapeau 2014). In other Algonquian languages, C appears to index the lowest third person, with a few variations (see Xu 2020; 2021 for discussion of the different patterns). I briefly discuss the Lowest pattern in Section 6.2. [^]
  3. Harley & Ritter (2002) use “Speaker”, but I adopt the modality-neutral label auth, for “author”. [^]
  4. To capture inclusive first person, Nevins adds a privative addr feature: [+part, +auth, addr]. [^]
  5. A common elaboration to Harley & Ritter’s (2002) 3noP theory is one in which third person is not completely underspecified, but the featural representation of third person is a subset of first and second person. Béjar (2003) proposes a view like this, with the feature [π] (for “person”) found in the representation of all person values, both third person as well as first and second. Third person is thus not completely underspecified, but it is underspecified relative to first and second person—it lacks certain featural distinctions that first and second person have. Another version of 3yesP involves a theory in which there isn’t a dedicated third person feature, but the featural representation of third person isn’t a subset of first and second person. Harbour (2016) argues for such a theory, with third person represented as [–part, –auth], sharing [–part] with exclusive first person, and sharing [–auth] with second person. As third person is not underspecified in any sense, this is still a 3yesP theory. [^]
  6. Nevins (2007) actually focuses on the “spurious se” rule in Spanish (Romance), where a third person dative clitic le(s) becomes the “reflexive” clitic se when it precedes another third person clitic. I do not reproduce this classic example here, as Alcaraz (2018) demonstrates that there are numerous syntactic distinctions between spurious se and dative le(s) that pose a significant challenge to any morphological dissimilation account (see also Fábregas & Cabré 2021, who take up Alcaraz’s challenge and put forth a syntactic account of spurious se). Outside of Spanish and Ubykh, similar restrictions are found across other varieties of Romance (Bonet 1995; Pescarini 2010; a.o.), in Arabic (Semitic, Walkow 2012), Kambera (Malayo-Polynesian, Klamer 1997), and Caquinte (Arawakan, Drummond & O’Hagan 2020), though in several of these cases the repairs don’t involve dissimilation. [^]
  7. Bobaljik (2008) argues that agreement is postsyntactic; see Arregi & Nevins (2012) and Preminger (2014) for responses affirming that (at least one component of) agreement occurs in the narrow syntax. [^]
  8. Preminger (2019) notes that the distribution of the Menominee “formative” suffix (to use Goddard’s (2007) term for it) seems to show an omnivorous third person agreement pattern, citing Trommer (2008). Roughly speaking, there’s a slot in the verb that can appear as -w, -m, or -n in various verb forms, and in simple cases -w seems to appear whenever there’s any third person argument in the clause. However, there are a number of complications. Firstly, Trommer (2008) doesn’t provide a syntactic analysis of this phenomenon, but rather a morphological, OT-based analysis, and thus strictly speaking doesn’t make an argument against syntactic 3noP. Secondly, as Trommer (2008) discusses in great depth, -w does not appear in all contexts that contain third person arguments—for instance, it’s absent in (most) negative verb forms, part>in verb forms, and verb forms that contain only inanimate arguments—and thus it’s not actually very straightforward to say that -w spells out omnivorous third person agreement (and this is part of the reason why Trommer’s OT analysis with competing contraints works so well). [^]
  9. This argument is prefigured by Bondarenko (2020), who proposes that the Passamaquoddy (Eastern Algonquian) theme sign (agreement on Voice) is specified for third person in the independent order—she does this to derive the distribution of the inverse marker in the independent. However, there are several plausible alternative analyses of inverse marking that don’t involve reference to third person features (Bruening 2001; 2019; Hammerly 2020; a.o.), so this isn’t conclusive. [^]
  10. I do not provide inanimate subject forms for Blackfoot because inanimates cannot be the subject of a transitive in Blackfoot (Frantz 2017). I do not provide in>part object forms for Plains Cree because they lack peripheral suffixes—a similar state of affairs holds in part>in forms. I assume this is due to an idiosyncratic Impoverishment rule, following similar proposals by Halle & Marantz (1993) and Xu (2021) for Potawatomi and Menominee, respectively. [^]
  11. Note that I do not provide Plains Cree forms with a part acting on an inanimate object: these forms lack a peripheral suffix, just like the 3>part forms, as mentioned above. I appeal to Impoverishment here as well. [^]
  12. Placing it above both subject and object, e.g in Infl, doesn’t seem like it would help either, as that just adds on another syntactic issue to solve, in addition to the problem of third person vs. part objects: why does this probe skip over the subject in preference for moving the object? [^]
  13. Passamaquoddy-Wolastoqey also has another LDA construction into conjunct complements that’s found under verbs like ‘know’, in parallel with Odawa and Mi’gmaq, but in this LDA construction the matrix verb can agree with any argument of the embedded clause (Bruening 2001; LeSourd 2019), like in Blackfoot (Bliss 2009), Innu (Branigan & MacKenzie 2002), Kitigan Zibi Algonquin (Lochbihler & Mathieu 2016), and Meskwaki (Dahlstrom 2015). [^]
  14. My glossing of the Mi’gmaq example here significantly diverges from Hamilton’s analysis. Hamilton treats the -t suffix as indexing third person; here I adopt the more diachronically conservative analysis that treats it as a 2sg>3 portmanteau form, cognate to the Odawa -ad suffix and going back to Proto-Algonquian 2sg>3cj *-at. While this is homophonous with the third person suffix -t, they do diverge under negation, where the third person suffix has an allomorph -g but the (at least historic) 2sg>3 portmanteau remains -t. [^]
  15. Again, my glossing of the Mi’gmaq example here significantly diverges from Hamilton’s analysis. Hamilton treats the -’g suffix as indexing third person (the apostrophe here indicates lengthening of the preceding vowel); here I adopt the more diachronically conservative analysis that treats it as a 1sg>3 portmanteau form (from Proto-Algonquian 1sg>3cj *-ak). This is almost but not quite homophonous with the -g allomorph of the third person suffix (which doesn’t lengthen preceding vowels). [^]
  16. I leave out Listuguj Mi’gmaq here as I do not have access to verbal paradigms for the modern language; for early 20th century Listuguj Mi’gmaq, see Pacifique (1939), translated and edited by Hewson & Francis (1990), where peripheral agreement behaves just as shown here in mixed configurations. [^]
  17. Though the Blackfoot plural marker -iksi is ambiguous between proximate and obviative, we know that amoksi aakííkoaiks ‘the girls, every girl’ is obviative here because the subject is proximate and we have the direct marker -ii (third person object agreement) on the matrix verb. [^]
  18. The failure of binding here isn’t a proximate-obviative issue—the intended bound variable in the relative clause is obviative, as indicated by the proximate object anna sááhkomaapi ‘the boy.prox’ and the presence of inverse marker -ok on the embedded verb, so we don’t have an issue of an obviative trying to be coreferent/bind a proximate. [^]
  19. One might wonder whether word order is relevant here. Bliss (2013) argues that that is not the case. [^]
  20. Blain actually pursues another analysis of the Plains Cree WCO data—her analysis involves analogizing Plains Cree WCO to the phenomenon of weakest crossover (Lasnik & Stowell 1991; Demirdache 1997). However, I’m not sure that that parallel is warranted, given the quantificational nature of wh items. [^]
  21. Bruening (2001) and Hamilton (2015) propose that in 3>part configurations, which also involve in the inverse marker, you also get objects moving over subjects. If so, the 3>part forms would also involve C skipping over a more local part goal to index third person, which would only strengthen my argument. However, Hamilton (2018), Oxford (to appear), and Grishin (2022a) argue that 3>part configurations do not involve object movement, unlike the 3>3 inverse. While I side with Hamilton (2018), Oxford (to appear), and Grishin (2022a) here, this point ultimately doesn’t affect the core of my argument. [^]
  22. An anonymous reviewer suggests that we could make use of the Activity Condition (Chomsky 2000; 2001; Oxford 2014; Hammerly 2020; 2021; Xu 2021), following a suggestion by Hammerly (2021), to analyze the behavior of C in languages like Blackfoot and Plains Cree. The idea is that we make it so that Infl (but not Voice) has the power to deactivate arguments it agrees with, and Infl prefers agreeing with part DPs, so that by the time C gets around to probing, it only has third persons left to agree with. This approach is simply a nonstarter for for Blackfoot, as we’ve already seen examples where Infl and C both clearly agree with the same argument—this occurs in the 3>3 inverse, where both heads agree with the object (20, repeated below):
      1. (i)
      1. Blackfoot (Frantz 2017: 62)
      1.  
      1. a.
      1. ots-
      2. 3-
      3. Infl
      1. ikákomimm
      2. loveta
      3.  
      1. -ok
      2. -inv
      3. -Voice
      1. -Ø
      2. -sg
      3. -Infl
      1. -a
      2. -sg
      3. -C
      1. ‘sheobv love herprox’
      1.  
      1. b.
      1. ots-
      2. 3-
      3. Infl
      1. ikákomimm
      2. loveta
      3.  
      1. -ok
      2. -inv
      3. -Voice
      1. -oaa
      2. -pl
      3. -Infl
      1. -yi
      2. -pl
      3. -C
      1. ‘she/theyobv love themprox’
    Additionally, in Meskwaki, Kickapoo, and Cree, in obv>obv configurations (both direct and inverse), both Infl and C index the obviative feature of the higher third person (here illustrated with direct forms in Meskwaki, as it preserves the number distinction in obviatives that makes the pattern clearer):
      1. (ii)
      1. Meskwaki (Goddard 1994: 190)
      1.  
      1. a.
      1. wâpam
      2. seeta
      3.  
      1. -ê
      2. -3obj
      3. Voice
      1. -ni
      2. -obv
      3. Infl
      1. -w
      2. -3
      3. Infl
      1. -ani
      2. -obv.sg
      3. C
      1. ‘sheobv sees her/themobv’
      1.  
      1. b.
      1. wâpam
      2. seeta
      3.  
      1. -ê
      2. -3obj
      3. Voice
      1. -ni
      2. -obv
      3. Infl
      1. -w
      2. -3
      3. Infl
      1. -ahi
      2. -obv.pl
      3. C
      1. ‘theyobv see her/themobv’
    From elsewhere in the paradigm, we can see that in Meskwaki (Goddard 1994), Kickapoo (Voorhis 1974), and Plains Cree and South East Cree (Kang 2017), this -ni morpheme (-yi in Plains and South East Cree) only appears with obviative subjects. We can conclude that -ni and -ani/-ahi are both indexing the subject, and thus Infl can’t have “deactivated” the subject, contrary to the predictions of the Activity Condition account. [^]
  23. Dahlstrom (1995) notes two counterexamples to this in Meskwaki: the inanimate collective noun mîčipêhi ‘game’ (as in wild animals hunted for food), and the inanimate noun čînawêtîweni ‘kin’ (which is an extension of the word’s original sense of ‘kinship system’). [^]
  24. We can’t use agreement with (overt) coordinations to test for markedness because Meskwaki features closest conjunct agreement (Dahlstrom 2015: 3–9). [^]
  25. An anonymous reviewer suggests that we can save the animacy agreement analysis if we allow finals to also be sensitive to notional animacy, as well as grammatical animacy. It’s not clear to me how exactly we could work this out formally: we can’t say that animate finals impose a semantic restriction on the relevant argument, as we’ve seen that animate finals can and must appear with grammatically-animate notional inanimates, like Meskwaki wîtawîha ‘raspberry’. However, playing devil’s advocate, one could imagine a system where finals first try to match the grammatically-specified animacy of the absolutive argument (which will only succeed with third persons, by hypothesis), but failing that, match its notional animacy (which will only succeed with part arguments, by hypothesis). It’s unclear to me how we could formalize that (in particular to derive the crucial result that finals would need to be (sometimes) sensitive to notional animacy, but C would need to never be sensitive to notional animacy, in order to prevent notionally-animate part arguments from being interveners), and moreover it’s unclear to me how this proposal could be falsified. [^]
  26. There is at least one other kind of relevant featural analysis in the literature: obviation is contrastive only on third person, first person is always proximate, and second person is always obviative (Bliss & Jesney 2005). I set this view aside here. [^]
  27. The only other potential example I know of is the language isolate Ktunaxa (also known as Kutenai), but there the picture is even muddier than in Blackfoot; see Hammerly (2020) for discussion. [^]
  28. There’s a clear diachronic explanation for this fact: the pronominal stem iistó is cognate to (o)isto ‘body’, and the independent pronouns originated as possessed forms of the word for ‘body’. Thus, historically, these pronouns were just third person noun phrases, and would participate in the proximate-obviative contrast just like other third person noun phrases in any other Algonquian language. [^]
  29. Again, I’ve relabeled spkr (speaker) to auth. [^]
  30. The issue for him is getting the representation of the indefinite agent in the Algonquian “unspecified agent” construction to be distinct from proximate third, obviative third, and inanimate third person—they pattern differently in agreement, and indefinite agents seem to (at least descriptively) rank below first and second person but above third person. He ends up proposing that indefinite agents are φPs, like first and second person pronouns, but lack (privative) [part], like third persons. This idea is similar to Legate’s (2014) proposal for the syntax of the Icelandic “grammatical object passive” construction, also known as the “new passive” or “new impersonal”, which involves an expletive subject, a passive nonagreeing verb form, and an accusative-marked internal argument—she also proposes that the expression of the agent in this kind of impersonal construction is a φP. [^]
  31. Newell & Piggott (2014: 339, fn.9) propose that similar data in Ojibwe can be accounted for as some kind of word infixation/tmesis à la English expletive infixation. However, this seems perhaps unlikely for Meskwaki, as multiple arguments can appear between preverb and verb:
      1. (i)
      1. Meskwaki (Dahlstrom 1987: 68)
      1.  
      1. Êh-nawači=
      2. ic.aor-stop-to=
      1. meko
      2. emph
      1. o-nesêmâw-ani
      2. 3-tobacco-obv.sg
      1. wâpeški-nenosôškaši-mîšâm-eki
      2. white-buffalo.hoof-sacred.pack-loc
      1. =pakin-â-č-i.
      2. =throwta-3obj-3cj-aor
      1. ‘He stopped to throw his tobacco on the White Buffalo’s Hoof sacred pack.’
    See Dahlstrom (1987) for more arguments against an incorporation account. [^]
  32. I’ve relabelled her non-part to 3 for simplicity. Additionally, while she presents a prox feature in her geometry, she doesn’t commit to its existence; thus, I’ve put it in parentheses here. [^]
  33. This conclusion is desirable under the assumption that the -wa/-yi contrast on Blackfoot overt part pronouns, which historically derives from a proximate-obviative contrast on possessed third persons, has been reanalyzed synchronically. [^]
  34. Though note that for Bondarenko (2020), as for Coon & Keine (2021), this entailment condition is relaxed on probes. [^]
  35. Another way of deriving the incompatibility of [part] and [3] is to mix binary and privative feature systems, like Nevins: we can say that [part] and [3] are actually binary [±part], while all the other features are privative. The use of the binary features bakes the stipulation into the representation, as negative values are typically defined formally as the absence of the positive value: [–F] :=¬∃[+F] (cf. Nevins 2007: 288). [^]
  36. Adding a third person feature also make predictions about PCC effects (as Nevins 2007 discusses). However, given the enormous range of extant analyses of the PCC (e.g. Anagnostopoulou 2003; 2005; Béjar & Rezac 2003; Nevins 2007; Coon & Keine 2021; Deal 2021, a.m.o., which only represent a tiny slice of the literature), giving full justice to the consequences of these feature representations for the PCC would take us too far afield and I leave this topic for further research. However, it is worth noting that adding just a privative 3 feature (à la Bondarenko 2020) to the feature gluttony account of the PCC proposed by Coon & Keine (2021) predicts the existence of a reverse Weak PCC (*part>3), which exists in Washo (Arregi & Hanink 2022a). [^]
  37. Note that this is incompatible with Béjar & Rezac (2009) and Coon & Keine’s (2021) version of featural coarseness, which involves copying over all the features on the goal that entail the probe’s specification. If animacy and obviation features don’t dominate the third person node, then those features wouldn’t be copied over to C under this version of featural coarseness. [^]
  38. This idea would require that person features on probes be interpreted, contrary to popular intuition (see Rezac 2016 for discussion). In order to flesh this idea out, we would also need to say something about how to compose the person features on a probe with the referent they are copied from, in particular to account for cases where the same agreement slot can agree with multiple different syntactic positions, i.e agreement displacement, to use Béjar’s (2003) term. A potential way forward could involve appealing to the literature that argues for syntactically representing indices (Rezac 2004; Elbourne 2005; 2008; Schwarz 2009; Hanink 2021; a.o.), especially the idea that indices can be copied in agreement (Rezac 2004; Clem 2021; Arregi & Hanink 2022b), saying that the featural coarseness of agreement involves copying over indices as well. If both indices and φ features are interpreted on probes, then we would predict that any conflict between them should be ruled out—thus, for instance, a second person feature would be incompatible as part of the same feature bundle as an index that points to an entity other than the addressee. I leave it to further research to work out the details of a proposal of this shape, as well as to consider its theoretical and typological ramifications. [^]
  39. This is an oversimplification. The Lowest group of languages is not homogeneous, and divides into at least three subgroups: (i) languages where C indexes the lowest third person (Mi’gmaq, Passamaquoddy-Wolastoqey); (ii) languages where C indexes the lowest third person but themes of ditransitives are totally blocked from agreement (Ojibwe-Potawatomi, Shawnee, Cheyenne); and (iii) languages where C (more-or-less) agrees with the lowest definite/specific third person object (Delaware, Mahican, Wampanoag, Western Abenaki, Proto-Algonquian). There is also additionally variation in whether the objects of AI+O (“pseudotransitive”) verbs—verbs that have an intransitive final, and yet still take an object argument—are able to be agreed with. For instance, in Odawa and Eastern Ojibwe, C is able to agree with those objects, but in Oji-Cree, C can’t (Xu 2020). I leave out fuller discussion of this rich range of variation across the family for reasons of space; I refer the interested reader to Xu (2020; 2021), who proposes that different Algonquian languages vary in the domain of accessibility for C, with languages like Ojibwe having smaller accessibility domains for C than languages like Passamaquoddy-Wolastoqey, and that definiteness-sensitive languages like Delaware have a rule of object shift that brings only definite objects “into view” of C. Under this way of thinking of things, the generalization that emerges for Lowest languages is that C always agrees with the lowest third person within the domain of accessibility. See Grishin (2022b) for a fuller implementation of this idea. [^]
  40. The name Expone Outermost is due to Amy Rose Deal. [^]
  41. A similar insight is briefly suggested by Bruening & Rackowski (2001) as well, though with a different implementation involving multiple agreement “overwriting” previous copied values. [^]

Abbreviations

1 = first person, 2 = second person, 3 = third person, abs = absolutive, addr = addressee, af = agent focus, ai = animate intransitive, an = animate, aor = aorist, appl = applicative, auth = author, cj = conjunct, com = completive, dat = dative, dem = demonstrative, dir = direct, emph = emphatic, erg = ergative, exc = exclusive, foc = focus, fut = future, ic = initial change, ii = inanimate intransitive, in = inanimate, inv = inverse, loc = locative, n = N formative, neg = negative, obj = object, obv = obviative, p = P formative, part = participant, pl = plural, prn = pronoun, prox = proximate, prs = present, pst = past, rel = relative, sbj = subject, sg = singular, ta = transitive animate, ti = transtitive inanimate, top = topic.

Acknowledgements

I thank Tanya Bondarenko, Amy Rose Deal, David Pesetsky, Norvin Richards, Will Oxford, participants at the MIT MorPhun reading group, and three anonymous reviewers for their valuable comments and insights. Blame me for any mistakes, misconceptions, or misrepresentations.

Competing interests

The author has no competing interests to declare.

References

Ackema, Peter & Neeleman, Ad. 2013. Person features and syncretism. Natural Language & Linguistic Theory 31(4). 901–950. DOI:  http://doi.org/10.1007/s11049-013-9202-z

Ackema, Peter & Neeleman, Ad. 2018. Features of Person: From the Inventory of Persons to Their Morphological Realization (Linguistic Inquiry Monographs 78). Cambridge, MA: MIT Press. DOI:  http://doi.org/10.7551/mitpress/11145.001.0001

Ackema, Peter & Neeleman, Ad. 2019. Default person versus default number in agreement. In Franco, Ludovico & Marchis Moreno, Mihaela & Reeve, Matthew (eds.), Agreement, case and locality in the nominal and verbal domains, 21–54. Berlin: Language Science Press. DOI:  http://doi.org/10.5281/zenodo.3458062

Alcaraz, Alejo. 2018. The spurious vs. dative problem. In Berns, Janine & Jacobs, Haike & Nouveau, Dominique (eds.), Selected papers from ‘Going Romance’ 20, Nijmegen (Romance Languages and Linguistic Theory 13), 5–20. Amsterdam: John Benjamins. DOI:  http://doi.org/10.1075/rllt.13.02alc

Anagnostopoulou, Elena. 2003. The Syntax of Ditransitives: Evidence from clitics. Berlin: Mouton de Gruyter.

Anagnostopoulou, Elena. 2005. Strong and weak person restrictions: A feature checking analysis. In Heggie, Lorie & Ordóñez, Francisco (eds.), Clitic and Affix Combinations: Theoretical perspectives (Linguistik Aktuell 74), 199–235. Amsterdam: John Benjamins. DOI:  http://doi.org/10.1075/la.74.08ana

Archangeli, Diana. 1984. Underspecification in Yawelmani Phonology and Morphology: Massachusetts Institute of Technology dissertation.

Archangeli, Diana. 1988. Aspects of Underspecification Theory. Phonology 5(2). 183–207. DOI:  http://doi.org/10.1017/S0952675700002268

Arregi, Karlos & Hanink, Emily A. 2022a. Reverse weak PCC in Washo. In Bakay, Özge & Pratley, Breanna & Neu, Eva & Deal, Peyton (eds.), Proceedings of NELS 52, 43–56. GLSA.

Arregi, Karlos & Hanink, Emily A. 2022b. Switch reference as index agreement. Natural Language & Linguistic Theory 40. 651–702. DOI:  http://doi.org/10.1007/s11049-021-09521-2

Arregi, Karlos & Nevins, Andrew. 2012. Morphotactics: Basque Auxiliaries and the Structure of Spellout (Studies in Natural Language and Linguistic Theory 86). Dordrecht: Springer. DOI:  http://doi.org/10.1007/978-94-007-3889-8

Béjar, Susana. 2003. Phi-Syntax: A Theory of Agreement: University of Toronto dissertation.

Béjar, Susana & Massam, Diane. 1999. Multiple Case Checking. Syntax 2. 65–79. DOI:  http://doi.org/10.1111/1467-9612.00015

Béjar, Susana & Rezac, Milan. 2003. Person Licensing and the Derivation of PCC Effects. In Pérez-Leroux, Ana Teresa & Roberge, Yves (eds.), Romance Linguistics: Theory and Acquisition. Selected papers from the 32nd Linguistic Symposium on Romance Languages (LSRL), Toronto, April 2002 (Current Issues in Linguistic Theory 244), 49–62. Amsterdam: John Benjamins. DOI:  http://doi.org/10.1075/cilt.244.07bej

Béjar, Susana & Rezac, Milan. 2009. Cyclic Agree. Linguistic Inquiry 40(1). 35–73. DOI:  http://doi.org/10.1162/ling.2009.40.1.35

Blain, Eleanor. 1997. Wh-constructions in Nêhiyawêwin (Plains Cree): University of British Columbia dissertation.

Bliss, Heather. 2009. Structuring Information in Blackfoot: Against an A’-Agreement Analysis of Cross-Clausal Agreement. In Jones, Susie (ed.), Actes du congrès annuel de l’Association canadienne de linguistique 2008. Canadian Linguistics Society.

Bliss, Heather. 2013. The Blackfoot configurationality conspiracy: Parallels and differences in clausal and nominal structures. Vancouver, BC: University of British Columbia dissertation.

Bliss, Heather & Jesney, Karen. 2005. Resolving hierarchy conflict: local obviation in Blackfoot. Calgary Working Papers in Linguistic 26(Spring). 92–116. DOI:  http://doi.org/10.11575/PRISM/28978

Bloomfield, Leonard. 1962. The Menomini Language. New Haven: Yale University Press.

Bobaljik, Jonathan D. 2008. Where’s Phi? Agreement as a Post-Syntactic Operation. In Harbour, Daniel & Adger, David & Béjar, Susana (eds.), Phi Theory: Phi-features across interfaces and modules (Oxford Studies in Theoretical Linguistics 16), 295–328. Oxford: Oxford University Press.

Bondarenko, Tanya. 2020. Feature gluttony for Algonquian: agreement in Passamaquoddy. Generals paper, MIT.

Bonet, Eulàlia. 1995. Feature structure of Romance clitics. Natural Language & Linguistic Theory 13. 607–647. DOI:  http://doi.org/10.1007/BF00992853

Branigan, Phil & MacKenzie, Marguerite. 1999. Binding relations and the nature of pro in Innu-aimun. In Tamanji, Pius & Hirotani, Masako & Hall, Nancy (eds.), Proceedings of NELS 29, vol. 1, 475–485. UMass Amherst: GLSA.

Branigan, Phil & MacKenzie, Marguerite. 2002. Altruism, A’-movement, and object agreement in Innu-aimûn. Linguistic Inquiry 33. 385–407. DOI:  http://doi.org/10.1162/002438902760168545

Bruening, Benjamin. 2001. Syntax at the Edge: Cross-Clausal Phenomena and the Syntax of Passamaquoddy: MIT dissertation.

Bruening, Benjamin. 2005. The Algonquian Inverse is Syntactic: Binding in Passamaquoddy. Manuscript, University of Delware.

Bruening, Benjamin. 2009. Algonquian Languages Have A-movement and A-Agreement. Linguistic Inquiry 40. 427–445. DOI:  http://doi.org/10.1162/ling.2009.40.3.427

Bruening, Benjamin & Rackowski, Andrea. 2001. Configurationality and Object Shift in Algonquian. In Gessner, Suzanne & Oh, Sunyoung & Shiobara, Kayono (eds.), Proceedings of the Workshop on the Structure and Constituency of Languages of the Americas 5, vol. 5, 71–84. Vancouver, BC: UBC Working Papers in Linguistics.

Calabrese, Andrea. 1995. A constraint-based theory of phonological markedness and simplification procedures. Linguistic Inquiry 26. 373–463.

Chomsky, Noam. 2000. Minimalist inquiries: The framework. In Martin, Roger & Michaels, David & Uriagereka, Juan & Keyser, Samuel Jay (eds.), Step by Step – Essays in Minimalist Syntax in Honor of Howard Lasnik, 89–156. Cambridge, MA: MIT Press.

Chomsky, Noam. 2001. Derivation by Phase. In Kenstowicz, Michael (ed.), Ken Hale: A Life in Language, 1–52. Cambridge, MA: MIT.

Clarke, Sandra. 1982. North-West River (Sheshātshīt) Montagnais: A Grammatical Sketch (National Museum of Man Mercury Series 80). Ottawa: National Museums of Canada. DOI:  http://doi.org/10.2307/j.ctv170qn

Clem, Emily. 2021. Cyclic Expansion in Agree: Maximal Projections As Probes. Linguistic Inquiry DOI:  http://doi.org/10.1162/ling_a_00432

Colley, Justin. 2018. Object movement derives object preference. In Hucklebridge, Sherry & Nelson, Max (eds.), Proceedings of NELS 48, vol. 1, 149–162. Amherst: GLSA.

Coon, Jessica & Bale, Alan. 2014. The interaction of person and number in Mi’gmaq. Nordlyd: Special issue on Features 40(1). 85–101. DOI:  http://doi.org/10.7557/12.3235

Coon, Jessica & Keine, Stefan. 2021. Feature Gluttony. Linguistic Inquiry 52(4). 655–710. DOI:  http://doi.org/10.1162/ling_a_00386

Costa, David J. 2003. The Miami-Illinois language. Lincoln, Nebraska: University of Nebraska Press.

Dahlstrom, Amy. 1986. Weak crossover and obviation. In Nikiforidou, Vassiliki (ed.), Proceedings of the 12th Annual Meeting of the Berkeley Linguistics Society, 51–60. Berkeley: Berkeley Linguistics Society. DOI:  http://doi.org/10.3765/bls.v12i0.1880

Dahlstrom, Amy. 1987. Discontinuous constituents in Fox. In Kroeber, Paul & Moore, Robert (eds.), Native American Languages and Grammatical Typology, 53–73. Bloomington: Indiana University Linguistics Club.

Dahlstrom, Amy. 1988. Independent pronouns in Fox. In Shipley, William (ed.), In Honor of Mary Haas: From the Haas Festival Conference on Native American Linguistics, 165–194. Berlin and New York: Mouton de Gruyter. DOI:  http://doi.org/10.1515/9783110852387.165

Dahlstrom, Amy. 1991. Plains Cree Morphosyntax. New York: Garland Publishing.

Dahlstrom, Amy. 1995. Motivation vs. predictability in Algonquian gender. In Pentland, David H. (ed.), Papers of the 26th Algonquian Conference, 52–66. Winnipeg: Univerisity of Manitoba.

Dahlstrom, Amy. 2015. Meskwaki (Fox) Syntax. Manuscript, University of Chicago.

Dahlstrom, Amy. 2017. Seeking consensus on the fundamentals of Algonquian word order. In Macaulay, Monica (ed.), Papers of the 45th Algonquian Conference, 59–72. East Lansing: Michigan State University Press.

de Lacy, Paul V. 2006. Markedness: Reduction and preservation in phonology. Cambridge: Cambridge University Press. DOI:  http://doi.org/10.1017/CBO9780511486388

Deal, Amy Rose. 2015. Interaction and satisfaction in φ-agreement. In Bui, Thuy & Özyıldız, Deniz (eds.), Proceedings of NELS 45, 179–192. Amherst, MA: GLSA.

Deal, Amy Rose. 2020. 3-on-3 restrictions and PCC typology: a reply to Pancheva and Zubizarreta (2018). Manuscript, UC Berkeley.

Deal, Amy Rose. 2021. Interaction, Satisfaction, and the PCC. Linguistic Inquiry. DOI:  http://doi.org/10.1162/ling_a_00455

Déchaine, Rose-Marie & Cardinal, Toni M. & Johnson, David & Kidd, Anne-Marie. 2015. Plains Cree Personal Pronouns. In Valentine, J. Randolph & Macaulay, Monica (eds.), Papers of the 43rd Algonquian Conference, 28–57. Albany: SUNY Press.

Déchaine, Rose-Marie & Wiltschko, Martina. 2002. Decomposing pronouns. Linguistic Inquiry 22. 409–442. DOI:  http://doi.org/10.1162/002438902760168554

Demirdache, Hamida. 1997. Dislocation, Resumption and Weakest Crossover. In Anagnostopoulou, Elena & Van Riemsdijk, Henk & Zwarts, Frans (eds.), Materials on Left Dislocation (Linguistics Today 14), 193–231. Amsterdam: John Benjamins. DOI:  http://doi.org/10.1075/la.14.12dem

Drapeau, Lynn. 2014. Grammaire de la langue innue [Grammar of the Innu language]. Montreal: Presses de l’Université du Québec.

Drummond, Emily & O’Hagan, Zachary. 2020. Morphological and syntactic person restrictions in Caquinte. In Asatryan, Mariam & Song, Yixiao & Whitmal, Ayana (eds.), Proceedings of NELS 50, vol. 1, 197–206. Amherst: GLSA.

Elbourne, Paul. 2005. Situations and individuals. Cambridge, MA: MIT Press.

Elbourne, Paul. 2008. Demonstratives as individual concepts. Linguistics and Philosophy 31(4). 409–466. DOI:  http://doi.org/10.1007/s10988-008-9043-0

Fábregas, Antonio & Cabré, Teresa. 2021. Towards a syntactic account of ungrammatical clitic sequences and their repairs. In Hinzelin, Marc-Olivier & Pomino, Natascha & Remberger, Eva-Maria (eds.), Formal Approaches to Romance Morphosyntax (Linguistische Arbeiten 576), 91–116. Berlin and Boston: De Gruyter. DOI:  http://doi.org/10.1515/9783110719154-004

Forchheimer, Paul. 1953. The category of person in language. Berlin: Walter de Gruyter. DOI:  http://doi.org/10.1515/9783111562704

Frantz, Donald G. 2017. Blackfoot Grammar. Toronto: University of Toronto Press 3rd edn.

Goddard, Ives. 1994. Leonard Bloomfield’s Fox Lexicon: Critical Edition. Winnipeg: Algonquian and Iroquoian Linguistics.

Goddard, Ives. 1995. Notes of Fox (Mesquakie) Inflection: Minor Modes and Incompletely Described Morphemes. In Pentland, David H. (ed.), Papers of the 26th Algonquian Conference, 124–150. Winnipeg: University of Manitoba.

Goddard, Ives. 2002. Grammatical Gender in Algonquian. In Wolfart, H. C. (ed.), Papers of the 33rd Algonquian Conference, 195–231. Winnipeg: Univerisity of Manitoba.

Goddard, Ives. 2007. Reconstruction and history of the independent indicative. In Wolfart, H. C. (ed.), Papers of the 38th Algonquian Conference, 207–271. Winnipeg: University of Manitoba.

Goddard, Ives. 2018. Blackfoot and Core Algonquian Inflectional Morphology: Archaisms and Innovations. In Macaulay, Monica & Noodin, Margaret (eds.), Papers of the 47th Algonquian Conference, 83–106. East Lansing: Michigan State University Press. DOI:  http://doi.org/10.14321/j.ctt1x76dj5.8

Grishin, Peter. 2022a. ECM in Passamaquoddy: Implications for the inverse. Manuscript, MIT. https://ling.auf.net/lingbuzz/006575.

Grishin, Peter. 2022b. How to agree with the lowest DP in Algonquian. Manuscript, MIT.

Halle, Morris. 1997. Distributed Morphology: Impoverishment and fission. In Bruening, Benjamin & Kang, Yoonjung & McGinnis, Martha (eds.), Papers at the interface (MIT Working Papers in Linguistics 30), 425–449. Cambridge, MA: MIT Working Papers in Linguistics.

Halle, Morris & Marantz, Alec. 1993. Distributed Morphology and the pieces of inflection. In Hale, Kenneth & Keyser, Samuel Jay (eds.), The view from Building 20, 111–176. Cambridge, MA: MIT Press.

Hamilton, Michael David. 2015. Phrase structure in Mi’gmaq: A configurational account of a “non-configurational” language. Lingua 167. 19–40. DOI:  http://doi.org/10.1016/j.lingua.2015.07.003

Hamilton, Michael David. 2018. On ordering and reordering arguments in Algonquian. In Macaulay, Monica & Noodin, Margaret (eds.), Papers of the 47th Algonquian Conference, 107–128. East Lansing: Michigan State University Press. DOI:  http://doi.org/10.14321/j.ctt1x76dj5.9

Hammerly, Christopher. 2020. Person-Based Prominence in Ojibwe: University of Massachusetts Amherst dissertation.

Hammerly, Christopher. 2021. A verb raising analysis of the Ojibwe VOS/VSO alternation: Lessons for feature copying and movement. Manuscript, University of Minnesota, Twin Cities.

Hanink, Emily A. 2021. DP structure and internally headed relatives in Washo. Natural Language and Linguistic Theory 39. 505–554. DOI:  http://doi.org/10.1007/s11049-020-09482-y

Harbour, Daniel. 2016. Impossible Persons. Cambridge, MA: MIT Press. DOI:  http://doi.org/10.7551/mitpress/9780262034739.001.0001

Harley, Heidi & Ritter, Elizabeth. 2002. Person and number in pronouns: a featuregeometric analysis. Language 78. 482–526. DOI:  http://doi.org/10.1353/lan.2002.0158

Hewson, John & Francis, Bernard. 1990. The Micmac Grammar of Father Pacifique. Winnipeg: Algonquian and Iroquoian Linguistics.

Kang, Eun Bi. 2017. Morphological variation in Cree inflection. Toronto Working Papers in Linguistics 39.

Klamer, Marian. 1997. Spelling out clitics in Kambera. Linguistics 35. 895–927. DOI:  http://doi.org/10.1515/ling.1997.35.5.895

Koopman, Hilda & Sportiche, Dominique. 1983. Variables and the Bijection Principle. The Linguistic Review 2(2). 139–160. DOI:  http://doi.org/10.1515/tlir.1982.2.2.139

Lasnik, Howard & Stowell, Timothy. 1991. Weakest Crossover. Linguistic Inquiry 22(4). 687–720.

Legate, Julie Anne. 2014. Voice and v: Lessons from Acehnese. Cambridge, MA: Cambridge University Press. DOI:  http://doi.org/10.7551/mitpress/9780262028141.001.0001

LeSourd, Philip S. 2019. Raising and long-distance agreement in Passamaquoddy: A unified analysis. Journal of Linguistics 55(2). 357–405. DOI:  http://doi.org/10.1017/S0022226718000415

Lochbihler, Bethany. 2012. Aspects of Argument Licensing: McGill University dissertation.

Lochbihler, Bethany & Mathieu, Éric. 2016. Clause Typing and Feature Inheritance of Discourse Features. Syntax 19(4). 354–391. DOI:  http://doi.org/10.1111/synt.12126

Massam, Diane. 1985. Case theory and the projection principles: Massachusetts Institute of Technology Ph.D. dissertation.

Myler, Neil. 2017. Cliticization feeds agreement: a view from Quechua. Natural Language and Linguistic Theory 35. 751–800. DOI:  http://doi.org/10.1007/s11049-016-9351-y

Nevins, Andrew. 2007. The representation of third person and its consquences for personcase effects. Natural Language & Linguistic Theory 25. 273–313. DOI:  http://doi.org/10.1007/s11049-006-9017-2

Nevins, Andrew. 2011. Multiple agree with clitics: person complementarity vs. omnivorous number. Natural Language & Linguistic Theory 29(4). 939–971. DOI:  http://doi.org/10.1007/s11049-011-9150-4

Newell, Heather & Piggott, Glyne. 2014. Interactions at the syntax-phonology interface: Evidence from Ojibwe. Lingua 150. 332–362. DOI:  http://doi.org/10.1016/j.lingua.2014.07.020

Nichols, John David. 1980. Ojibwe Morphology: Harvard University dissertation.

Okimāsis, Jean L. 2018. nēhiyawēwin: paskwāwi-pīkiskwēwin [Cree: Language of the Plains]. Regina: University of Regina Press.

Oxford, Will. 2014. Microparameters of Agreement: A Diachronic Perspective on Algonquian Verb Inflection: University of Toronto dissertation.

Oxford, Will. 2017a. Proximate DP, obviative KP: Balancing the morphosyntax and pragmatics of obviation. In Monti, Andrew Alexander (ed.), Actes du congrès annuel de l’Association canadienne de linguistique 2017. Canadian Linguistics Society.

Oxford, Will. 2017b. Size matters: Deriving person hierarchies from nominal structure. Manuscript. https://ling.auf.net/lingbuzz/004534.

Oxford, Will. 2019. Inverse marking and Multiple Agree in Algonquin: Complementarity and variability. Natural Language & Linguistic Theory 37. 955–996. DOI:  http://doi.org/10.1007/s11049-018-9428-x

Oxford, Will. To appear. A tale of two inverses. Syntax.

Pacifique, Père. 1939. Leçons grammaticales théoriques et pratiques de la langue micmaque [Theoretical and practical grammatical lessons on the Mi’gmaq language]. Sainte-Anne de Ristigouche, P. Q: Bureau de messager Micmac.

Pescarini, Diego. 2010. Elsewhere in Romance: Evidence from clitic clusters. Linguistic Inquiry 41(3). 427–444. DOI:  http://doi.org/10.1162/LING_a_00003

Pesetsky, David. 2013. Russian Case Morphology and the Syntactic Categories (Linguistic Inquiry Monographs 66). Cambridge, MA: MIT Press. DOI:  http://doi.org/10.7551/mitpress/9780262019729.001.0001

Postal, Paul M. 1971. Cross-Over Phenomena (Transatlantic series in linguistics). New York: Holt, Rinehart and Winston.

Preminger, Omer. 2014. Agreement and its failures (Linguistic Inquiry Monographs 68). Cambridge, MA: MIT Press. DOI:  http://doi.org/10.7551/mitpress/9780262027403.001.0001

Preminger, Omer. 2019. What are φ-features supposed to do, and where? Talk at Thirty million theories of features, University of Tromsø.

Quinn, Conor. 2019. Productivity vs predictability: Evidence for the syntax and semantics of Animate gender in four Northeastern-area Algonquian languages. In Mathieu, Éric & Dali, Myriam & Zareikar, Gita (eds.), Gender and Noun Classification, 249–265. Oxford: Oxford University Press. DOI:  http://doi.org/10.1093/oso/9780198828105.003.0012

Rezac, Milan. 2003. The fine structure of Cyclic Agree. Syntax 6. 156–182. DOI:  http://doi.org/10.1111/1467-9612.00059

Rezac, Milan. 2004. Elements of Cyclic Syntax: Agree and Merge: University of Toronto dissertation.

Rezac, Milan. 2016. On the (un)interpretability of phi-agreement. Manuscript. https://ling.auf.net/lingbuzz/003165.

Rhodes, Richard A. 1976. The Morphosyntax of the Central Ojibwa Verb: University of Michigan dissertation.

Rhodes, Richard A. 1994. Agency, inversion, and thematic alignment in Ojibwe. In Gahl, Susanne & Dolbey, Andy & Johnson, Christopher (eds.), Proceedings of the 20th Annual Meeting of the Berkeley Linguistics Society, 431–446. Berkeley: Berkeley Linguistics Society. DOI:  http://doi.org/10.3765/bls.v20i1.1436

Richards, Norvin. 2010. Uttering Trees. Cambridge, MA: MIT Press. DOI:  http://doi.org/10.7551/mitpress/9780262013765.001.0001

Schwarz, Florian. 2009. Two types of definites in natural language: University of Massachusetts Amherst dissertation.

Seiter, William. 1980. Studies in Niuean syntax. New York: Garland Publishing.

Sherwood, David Fairchild. 1983. Maliseet-Passamaquoddy Verb Morphology: Yale University dissertation.

Trommer, Jochen. 2008. Third-Person Marking in Menominee. In Harbour, Daniel & Adger, David & Béjar, Susana (eds.), Phi Theory: Phi-Features Across Modules and Interfaces (Oxford Studies in Theoretical Linguistics 16), 221–250. Oxford: Oxford University Press.

Valentine, J. Randolph. 2001. Nishnaabemwin Reference Grammar. Toronto: University of Toronto Press.

van Urk, Coppe. 2015. A uniform syntax for phrasal movement: A case study of Dinka Bor: MIT PhD Thesis.

Voorhis, Paul H. 1974. Introduction to the Kickapoo Language. Bloomington: Indiana University.

Walkow, Martin. 2012. Goals, Big and Small: University of Massachusetts Amherst dissertation.

Wasow, Thomas. 1972. Anaphoric relations in English: Massachusetts Institute of Technology dissertation.

Wiltschko, Martina & Marshall, Valerie & Matheson, Andy & Vincent, Audra. 2015. Independent Pronouns in Blackfoot. In Macaulay, Monica & Valentine, J. Randolph (eds.), Papers of the 43rd Algonquian Conference, 266–286. Albany: SUNY Press.

Wolfart, H. C. 1973. Plains Cree: A grammatical study. Transactions of the American Philosophical Society 63(5). 1–90. DOI:  http://doi.org/10.2307/1006246

Xu, Yadong. 2020. The cline of the peripheral agreement and implication about object types. 52nd Algonquian Conference, University of Wisconsin-Madison.

Xu, Yadong. 2021. Deriving degrees of accessibility in Algonquian peripheral agreement. WSCLA 25, Sogang University.