Pith. sign in

REVIEW 4 major objections 6 minor 29 references

The Algebraic Structure of Morphosyntax

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read Word-internal and sentence-level structure are generated by the same algebraic operation, and Distributed Morphology's four operations reduce to two: fusion and fission.

desk verdict The algebraic scaffold is worth reading, but the advertised reduction of Distributed Morphology to fusion/fission is not proven and needs either serious strengthening or a more honest restatement. read the letter →

arxiv 2507.00244 v1 pith:S6FF3NZI submitted 2025-06-30 cs.CL math.QA

classification cs.CLmath.QA MSC 16T0518M70
keywords distributedmorphologymorphosyntaxMergeoperadHopfalgebranon-associativecommutativemagmafeaturebundlesoperadiccorrespondencesyntax-morphologyinterface
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that morphology is not a separate computational system from syntax: feature bundles are hierarchical binary trees generated by the same free non-associative commutative magma—a binary tree-building operation—that builds syntactic objects, and morphosyntactic trees arise by inserting those morphological trees at the leaves of syntactic trees. The insertion is governed by a compatibility relation between morphological feature bundles and syntactic lexical items or features, and is formalized as a colored correspondence between two algebras over the Merge operad (a formalism for composing operations with many inputs and one output). On this basis the paper reinterprets the four classic operations of Distributed Morphology: fusion and fission are the basic ones, moving the syntax-morphology boundary upward or downward, while impoverishment and obliteration are shown to be derived from fusion, fission, and the morphology coproduct. If correct, the account unifies word formation and sentence formation under one algebraic core and reduces the primitive operations of Distributed Morphology from four to two.

What carries the argument

The central object is the free non-associative commutative magma over the set of morphological features, whose elements are non-planar full binary rooted trees with features at the leaves and internal vertices labelled by unions of leaf features (Definition 2.1); this gives feature bundles an internal hierarchical structure. Around it sit the extended morphological objects of Definition 2.8 (binary trees with non-branching vertices and monotone bundle assignments), the Hopf algebra of morphological workspaces with coproduct, and the Merge operad, whose operations are abstract binary trees with unlabelled leaves. The load-bearing mechanism is the colored correspondence of algebras over an operad, which shows that inserting extended morphological trees at the leaves of syntactic trees, constrained by the syntax-morphology feature correspondence, is compatible with the operad actions on both syntactic and morphosyntactic trees. The structure-building operators then combine the coproduct with operadic insertion, and the Distributed Morphology operations are defined as transformations of these operators: fusion in (5.2), fission in (5.6), obliteration in (5.12), and impoverishment in Proposition 5.20. Proposition 5.21, showing that obliteration and impoverishment are obtainable from fission, fusion, and the coproduct, is what carries the paper's reduction claim.

What would settle it

A documented morphological rule that applies to two heads in one linear order but not the reverse (for example, fusion of subject and tense only when the subject precedes the tense head) would falsify the central claim, because the model's non-planar trees and set-valued bundle labels cannot distinguish the two orders.

Watch

Extended reading notes

Core claim

The paper claims that the morphology-syntax interface has a precise algebraic description: morphological objects are non-planar full binary trees whose internal vertices carry feature bundles obtained by unioning leaf features, and these form a free non-associative commutative magma over the feature set. Extended morphological objects, allowing non-branching vertices and arbitrary monotone bundle assignments, are the inputs inserted at syntactic leaves, subject to a syntax-morphology feature correspondence. The paper proves that the set of extended morphological trees is a colored correspondence between the algebra of syntactic objects and the algebra of morphosyntactic trees over the same Merge operad, and it constructs structure-building operations from the morphology coproduct and operadic insertion. It then shows that the Distributed Morphology operations of fusion and fission are the two primitives—fusion moves the syntax-morphology boundary up, fission moves it down—and that obliteration and impoverishment are combinations of fission, fusion, and the coproduct. The stated upshot is that morphology and syntax share the same core computational structure, with the boundary between them movable by a semigroup of post-syntactic transformations.

Load-bearing premise

The argument rests on the assumption that every morphological feature bundle relevant to Distributed Morphology can be faithfully represented as a (possibly extended) non-planar binary tree whose internal labels are unions of leaf features, so that no information needed for fusion, fission, impoverishment, or obliteration lives outside those tree labels.

Editorial extensions

If this is right

  • Word formation and sentence formation share a single generative core: the same free non-associative commutative magma builds both morphological and syntactic trees, so no separate morphological Merge operation is needed.
  • The morphology-syntax interface is an insertion operation: extended morphological trees are placed at syntactic leaves whenever the root feature bundle matches the leaf's lexical item or syntactic feature, making morphosyntactic trees an algebra over the same Merge operad as syntax.
  • The four Distributed Morphology operations reduce to two: fusion and fission are the basic boundary-moving primitives, and obliteration and impoverishment are derived from them together with the morphological coproduct.
  • The syntax-morphology boundary is movable: fusion and fission generate a semigroup acting on morphosyntax-building operations, so a single pre-Externalization computational structure underlies typological differences in where word formation meets syntax.
  • Because the model is pre-Externalization, vocabulary insertion, planarization, and language-specific filtering act later on the same morphosyntactic trees, keeping the core algebraic structure language-independent.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If fusion and fission are truly the only primitives, every Distributed Morphology derivation should be decomposable into a sequence of boundary-moving steps; one could test this by compiling attested fusion, fission, and impoverishment rules into words of the post-syntactic semigroup and checking whether equivalent rules have a unique normal form.
  • The model's non-planarity entails a testable prediction: pre-Externalization morphology should be insensitive to linear order, so an attested rule that applies only to one order of two heads would force order data into the magma or correspondence.
  • The semigroup action suggests a quantitative typology: the minimal number of fusion or fission steps needed to realize a language's surface word structure from a common syntactic spine could serve as a measure of morphological synthesis, placing polysynthetic and analytic languages on a single spectrum.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper develops an algebraic model of the morphology-syntax interface in the tradition of Distributed Morphology, building on a previously proposed Hopf-algebra/operadic model of syntactic Merge. Morphological feature bundles are represented as non-planar binary trees generated by a free non-associative commutative magma, with internal vertex labels obtained by union of leaf features. A second class of 'extended' morphological objects allows non-branching vertices and more general bundle assignments, and is used to define morphosyntactic trees by inserting morphological trees at syntactic leaves. The paper then reformulates the DM operations of fusion, fission, obliteration, and impoverishment as transformations of morphosyntactic trees or, equivalently, as transformations of the assembly operations that build them. The central advertised claim is that fusion and fission are the basic operations, with obliteration and impoverishment derived from them together with the coproduct, yielding a movable syntax-morphology boundary and a semigroup description of DM operations.

Significance. If the construction is correct, the paper would provide a unified mathematical language for morphology and syntax, connecting Merge-based syntax, operads, Hopf algebras, and DM operations. The definitions are explicit, the Swahili, Arabic, and Basque examples illustrate the intended applications, and the paper is careful to separate the pre-Externalization computational core from language-specific filtering. A notable strength is that the framework makes falsifiable structural predictions: for instance, all DM operations should be expressible through fusion/fission plus the coproduct, and the semigroup generated by the post-syntactic operations should describe the movable boundary. These claims are, however, conditional on the tree representation of feature bundles and on the proof of the reduction theorem, both of which need further support.

major comments (4)
  1. [§5.3–5.4, Proposition 5.21, Definition 6.1] The central reduction claim is not established. Proposition 5.13 defines full-bundle obliteration directly through the coproduct and the unit 1 of the magma, without any use of fusion or fission. The proof of Proposition 5.21, which is supposed to show that obliteration and impoverishment are derived from fission, fusion, and the coproduct, assumes without proof that for the required split Bv = B ⊔ B′ there exist A and a partition (B1, B2) with (Bi ∪ A, αi) ∈ ΓSM, i.e. that the set P_{A,αℓ,α}(Bv) of Definition 5.6 is nonempty. It also assumes that the trees Sℓ,Bi∪A produced by the algorithm of Definition 5.6 satisfy the monotonicity and leaf-containment conditions of Definition 2.8. Since ΓSM is an arbitrary relation, these conditions can fail. Moreover, Definition 6.1 still lists the transformation (6.3) as a generator of the DM semigroup SDM alongside fusion (6.1) and fission (6.2), which is in direct tension with the claim that fusion and fission are the sole basic operations. The reduction claim should either be proved under explicit hypotheses on ΓSM or stated as a conjecture with the required hypotheses made precise.
  2. [§5.1–5.2, Propositions 5.3 and 5.9] The 'unique map' characterizations of fusion and fission are not proven and are not well-posed as stated. The commutative diagrams in Propositions 5.3 and 5.9 mix arrows of different types: the horizontal maps are morphosyntax-assembly operations from V(WM) to V(WMS), while the vertical maps are transformations within V(WM); no categories or uniqueness criteria are specified. In the absence of a universal property or an explicit construction of a map making the diagram commute, the word 'unique' is unsupported. At minimum, the authors should state the source and target categories of the diagrams and prove either existence and uniqueness or replace 'unique' by 'canonical under the given construction.'
  3. [§2.1, §2.4, §5.1, Example 5.10] The load-bearing representation assumption is not independently justified. Definition 2.1 models feature bundles as non-planar binary trees with internal labels obtained by union, and Definition 2.8 allows additional bundle assignments, but the paper does not provide an empirical argument that DM operations depend only on this tree structure. This is not a purely formal point: fusion in Definition 5.2 is described as applying to two feature bundles in 'two different adjacent syntactic leaves,' yet the model is non-planar and has no notion of adjacency or linear order; similarly, the obliteration rule of Example 5.10 is stated in terms of two clitics 'within an auxiliary M-word,' which requires linear or morphological word-internal information. The paper defers planarization and adjacency to Externalization, but the DM operations it formalizes are claimed to act before Externalization. The authors should clarify how adjacency and word-internal linear relations are represented, or justify why the operations in question are well-defined on non-planar trees.
  4. [§3.5, Theorem 3.16] Theorem 3.16, which asserts that the extended morphological objects form a colored correspondence between the algebras SO and MS over the Merge operad, is verified by a brief compatibility sentence rather than a proof. The statement is central to the paper's architecture, since the correspondence is what replaces a morphism of algebras over an operad. The proof should explicitly verify the diagrams of Definition 3.14 (and the colored version of Definition 3.15) for all arities n and all sequences k1,...,kn, and should specify how the domains in (3.18) are exactly matched by the syntax-morphology correspondence ΓSM. As written, the argument is a plausibility check rather than a theorem.
minor comments (6)
  1. [§2.3, Definition 2.5] The definition contains a likely typo: 'We also write V(FMO0) for the vector space spanned by the forests in the set V(FMO0), where components are in the set MO' should presumably be 'FMO0' in the second occurrence. The word 'staisfying' should also be corrected to 'satisfying.'
  2. [§5.1, Example 5.1] The Swahili examples appear to be missing from the manuscript, leaving only the surrounding prose; please include the actual paradigm sentences or remove the empty display.
  3. [§5.2, Example 5.4] The Arabic agreement table is garbled in the displayed text: the row for 2F in the plural and the row for 3M appear interleaved, and the bold marking of person affixes is not visible. The table should be reformatted so that the paradigm is legible.
  4. [§5.2, Definition 5.6, Eq. (5.7)] Equation (5.7) contains a stray comma in the list 'S1, . . . , Sℓ,B1∪A, Sℓ,B2∪A, , . . . , Sn'; this should be cleaned up.
  5. [§5.4, Eq. (5.17) and (5.18)] The notation 'FvΦA,(B,B′)' is introduced without a precise definition of which fusion term is applied at the root vertex v; please define this operation before using it in the display.
  6. [References] In the introduction, the citation '[21, 26, 27, 6, 21, 22, 29]' lists [21] twice; remove the duplicate and check the reference list for other repeated entries.

Circularity Check

2 steps flagged · score 5.0 of 10

Impoverishment and obliteration are defined via fission and the coproduct, so Proposition 5.21's derivation restates the definitions; the operadic correspondence itself is self-contained.

  1. self definitional [Section 5.4, Proposition 5.20; used in Proposition 5.21]
    "Proposition 5.20. The impoverishment operations I_{B⊂Bv} : V(WMS) → V(WMS) and I_{Bv/B} : V(WMS) → V(WMS) have two cases: ... In the first case, for B ⊂ Bv at the root v of an extended morphological tree S, the operation I_{B⊂Bv} acts as (5.15) I_{B⊂Bv}(γSO,MO(T,S,S1,...,Sn)) = γSO,MO(T,SB′,S1,...,Sn), where Bv = B ⊔ B′ and SB′ is the subtree of the fission of S according to Bv = B ⊔ B′."

    Impoverishment is defined as 'the subtree of the fission of S'. Proposition 5.21 then 'proves' that impoverishment is obtainable from fission, fusion, and the coproduct. The proof is a sequence of maps applied to this fission-based definition, so the advertised reduction is true by construction rather than by an independent derivation from the external DM notion of impoverishment.

  2. self definitional [Section 5.3, Proposition 5.13; Section 5.4, Proposition 5.21]
    "Proposition 5.21. The operations of Obliteration and Impoverishment are obtainable from combinations of fission, fusion, and the coproduct Δρ. Proof. The case of Obliteration of a full feature bundle is already described in the proof of Proposition 5.13 in terms of the primitive part of the coproduct, so we focus on Impoverishment."

    Obliteration was introduced in Proposition 5.13 by replacing a morphological tree S with the unit 1, using the primitive part of the coproduct. Proposition 5.21's proof of the reduction for full obliteration simply refers back to that coproduct-based definition rather than deriving obliteration from fusion or fission. Thus the claimed derivation uses exactly the coproduct structure that was already used to define the operation, making the reduction a restatement.

full rationale

The paper is largely a self-contained mathematical construction: it defines a magma of morphological trees with leaf-union labeling, extended morphological workspaces via a comodule, an operadic correspondence between syntactic and morphosyntactic algebras, and fusion/fission transformations checked by commutative diagrams. These components are not fitted to data, and their internal consistency is argued from the definitions rather than from a self-citation chain. The circularity is concentrated in the advertised central claim that obliteration and impoverishment are 'mathematically provable' as derived from fusion/fission plus the coproduct. Impoverishment is defined in Proposition 5.20 as a subtree of a fission, and obliteration is defined in Proposition 5.13 using the coproduct; Proposition 5.21 then proves they are obtainable from exactly those ingredients. That is a definitional restatement, not an independent reduction of the external DM operations. The tension is visible in Definition 6.1, where the DM semigroup is generated not only by fusion (6.1) and fission (6.2) but also by the obliteration/impoverishment generator (6.3), so the paper's own formalism does not implement a semigroup generated by fusion and fission alone. There are also unproven existence assumptions in the fission-based derivation, such as nonempty partition sets P_{A,αℓ,α}(Bv) and quotients satisfying Definition 2.8; these are correctness gaps rather than circularity. The operadic/interface construction and the formalizations of fusion and fission have independent mathematical content, so the overall circularity is partial and definitional, not global.

Assumptions & free parameters 3 free parameters · 7 assumptions · 3 invented entities

The model has no numerical parameters fitted to data. Its free choices are the feature inventory, the matching relation ΓSM, and the tree representation of bundles. The axioms are mostly inherited from the authors' earlier syntactic model or chosen ad hoc to make the DM interpretation work. The invented entities are formal devices without independent falsifiable support.

free parameters (3)
  • MO0: finite set of morphological features = unspecified; assumed finite
    The model takes an arbitrary finite inventory of features such as [±PL] and [uPL]; its content is never fixed, so any specific morphosyntactic claim depends on this choice.
  • ΓSM: syntax-morphology feature correspondence = assumed surjective subset of P(MO0) × SO0; later relaxed
    This matching relation between feature bundles and lexical items determines which insertions are allowed; it is chosen by hand and not derived.
  • Binary and bivalent feature representation = modeling choice
    The paper chooses non-planar full binary trees and bivalent features (Remark 2.3) to simplify the algebra; this choice shapes all subsequent operations.
assumptions (7)
  • domain assumption The syntactic model of [17]: free non-associative commutative magma SO, Hopf algebra coproduct Δ, Merge as Hopf algebra Markov chain, and syntactic objects as algebra over the Merge operad.
    The entire morphology construction is built on this prior framework, cited and summarized in §1.1 and §3.2.
  • domain assumption V(F≤2_MO0) is a graded connected Hopf algebra and V(FMO0) is a bicomodule over it with ρR = Δρ (Remark 2.6 and Lemma 2.9).
    The comodule structure that generates non-branching morphological trees is inherited from [17] and not reproved here.
  • ad hoc to paper Feature bundles are tree structures with internal labels obtained by union (2.2).
    This is the paper's core modeling assumption; it converts bundles from sets to hierarchical trees.
  • ad hoc to paper Extended morphological objects (T,B) satisfy Bw ⊆ Bv and contain all leaf features (Definition 2.8).
    This extension allows non-branching vertices and extra features needed for DM operations; the assignment rules are chosen by hand.
  • domain assumption The syntax-morphology correspondence ΓSM is surjective onto SO0, except when obliteration is considered (Remark 5.12).
    Surjectivity ensures every syntactic object can carry morphology; dropping it is used to model obliteration.
  • domain assumption Head functions and labeling algorithms for syntactic objects from [17] determine the root label in fusion (αv = α_{h_{T'}(v)} in Definition 5.2).
    Fusion depends on this prior notion of head labeling to relabel the fused syntactic leaf.
  • domain assumption Distributed Morphology operations as described by Halle and Marantz are the target empirical operations.
    The paper chooses DM over Nanosyntax as the linguistic framework and assumes its operation inventory.
invented entities (3)
  • Extended morphological objects ]MO = T≤2_MO0,B
    purpose: Provide morphological inputs to syntactic leaves that can contain non-branching vertices and bundles not fully determined by leaves, enabling DM operations.
    No falsifiable handle outside the formalism; existence is justified by the need to model fission and impoverishment.
  • Correspondence of graded algebras over an operad
    purpose: Replaces morphisms of algebras over operads to relate syntactic and morphosyntactic trees while changing the number of leaves.
    A new mathematical definition introduced in Definition 3.14; no independent empirical or mathematical evidence provided beyond its role in the model.
  • Distributed Morphology semigroup SDM
    purpose: Organize DM operations as transformations on the set of morphosyntax-building operations, with fusion and fission as generators.
    An algebraic packaging of the paper's own constructions; no external evidence.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Algebraic Structure of Morphosyntax." pith.science (2026). https://pith.science/paper/S6FF3NZI

@misc{pith2026250700244,
  author       = {Pith},
  title        = {Pith review of: The Algebraic Structure of Morphosyntax},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S6FF3NZI}},
  note         = {Machine review of arXiv:2507.00244}
}
read the original abstract

Within the context of the mathematical formulation of Merge and the Strong Minimalist Thesis, we present a mathematical model of the morphology-syntax interface. In this setting, morphology has compositional properties responsible for word formation, organized into a magma of morphological trees. However, unlike syntax, we do not have movement within morphology. A coproduct decomposition exists, but it requires extending the set of morphological trees beyond those which are generated solely by the magma, to a larger set of possible morphological inputs to syntactic trees. These participate in the formation of morphosyntactic trees as an algebra over an operad, and a correspondence between algebras over an operad. The process of structure formation for morphosyntactic trees can then be described in terms of this operadic correspondence that pairs syntactic and morphological data and the morphology coproduct. We reinterpret in this setting certain operations of Distributed Morphology as transformation that allow for flexibility in moving the boundary between syntax and morphology within the morphosyntactic objects.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

29 extracted references · 28 canonical work pages

  1. [17]

    Marcolli, M., Chomsky, N., and Berwick, R. C. (2025a). Mathematical Structure of Syntactic Merge: An Algebraic Model for Generative Linguistics . MIT Press

  2. [1]

    and Nevins, A

    Arregi, K. and Nevins, A. (2012). Morphotactics: Basque auxiliaries and the structure of spellout . Springer Dordrecht

  3. [2]

    Baunaz, L., Haegeman, L., De Clercq, K., and Lander, E. (2018a). Exploring nanosyntax. Oxford University Press

  4. [3]

    Baunaz, L., Lander, E., De Clercq, K., and Haegeman, L. (2018b). Nanosyntax: the basics. Oxford Studies in Comparative Syntax , pages 3–56. Publisher: Oxford University Press

  5. [4]

    Bobaljik, J. D. (1997). If the head fits...: on the morphological determination of Germanic syntax. Publisher: Walter de Gruyter, Berlin/New York Berlin, New York

  6. [5]

    Caha, P. (2009). The nanosyntax of case. Publisher: Universitetet i Tromsø

  7. [6]

    and Keine, S

    Coon, J. and Keine, S. (2021). Feature gluttony. Linguistic inquiry , 52(4):655–710. Publisher: MIT press journals-info@ mit. edu

  8. [7]

    and Hall, D

    Cowper, E. and Hall, D. C. (2002). The syntactic manifestation of nominal feature geom- etry. In Proceedings of the 2002 annual conference of the Canadian Linguistic Association , pages 55–66. Canadian Linguistic Association Toronto

Show all 29 references
  1. [8]

    Cowper, E. A. (2005). The geometry of interpretable features: Infl in English and Spanish. Language, 81(1):10–46. Publisher: Linguistic Society of America

  2. [9]

    and Noyer, R

    Embick, D. and Noyer, R. (2007). Distributed morphology and the syntax—morphology interface

  3. [10]

    and Marantz, A

    Halle, M. and Marantz, A. (1993). Distributed morphology and the pieces of inflection. In Hale, K. and Keyser, S., editors, The view from building 20 , pages 111–176. The MIT Press, Cambridge, MA

  4. [11]

    and Marantz, A

    Halle, M. and Marantz, A. (1994). Some key features of distributed morphology. In Carnie, A., Harley, H., and Bures, T., editors, MITWPL 21 , pages 275–288

  5. [12]

    Harbour, D. (2008). Discontinuous agreement and the syntax-morphology interface. Phi theory: Phi-features across modules and interfaces , pages 185–220. Publisher: Oxford University Press Oxford

  6. [13]

    Harley, H. E. and Ritter, E. (2002). Person and number in pronouns: A feature- geometric analysis. Language, 78:482 – 526

  7. [14]

    and Nahmad, H

    Haywood, J. and Nahmad, H. (1965). Arabic grammar of the written language . Lund Humphries

  8. [15]

    Hewett, M. (2023). Allomorphy in Semitic discontinuous agreement: Evidence for a modular approach to postsyntax. Natural Language & Linguistic Theory , 41(3):1091–1145. Publisher: Springer

  9. [16]

    and M¨ uller, G

    Kleine, S. and M¨ uller, G. (preprint). Impoverishment

  10. [18]

    Marcolli, M., Huijbregts, R., and Larson, R. K. (2025b). Hypermagmas and colored operads: heads, phases, and theta roles

  11. [19]

    violations

    Marcolli, M., Larson, R., and Huijbregts, R. (2024). Extension Condition “violations” and Merge optimality constraints

  12. [20]

    and Larson, R

    Marcolli, M. and Larson, R. K. (2025). Theta theory: operads and coloring. arXiv:2503.06091

  13. [21]

    Martinovi´ c, M. (2022). Feature geometry and head splitting in the Wolof clausal pe- riphery. Linguistic Inquiry , 54(1):79–116. Publisher: MIT press journals-info@ mit. edu. MORPHOLOGY-SYNTAX INTERF ACE 45

  14. [22]

    McGinnis, M. (2007). Phi-feature competition in morphology and syntax. Publisher: Oxford University Press

  15. [23]

    Nevins, A. (2016). Lectures on postsyntactic morphology. Ms. University College Lon- don

  16. [24]

    Noyer, R. (1997). Features, Positions and Affixes in Autonomous Morphological Struc- ture. Garland Publishing

  17. [25]

    Starke, M. (2009). Nanosyntax: A short primer to a new approach to language. Nordlyd, 36(1):1–6

  18. [26]

    Svenonius, P. (2006). Case alternations and the Icelandic passive and middle. Passives and impersonals in European languages . Publisher: Citeseer

  19. [27]

    Tosco, M. (2007). Feature-geometry and diachrony: the development of the subject clitics in Cushitic and Romance. Diachronica, 24(1):119–153. Publisher: John Benjamins

  20. [28]

    Watson, J. C. (1993). A syntax of Sanani Arabic , volume 13. Otto Harrassowitz Verlag

  21. [29]

    Xu, Y. (2018). A double-person feature analysis of Algonquian first person plurals. In Proceedings of WCCFL, volume 35. Department of Linguistics, Yale University, New Haven, CT 06511, USA Email address: isabella.senturia@yale.edu Department of Mathematics and Department of Co...

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.