Authority caveat. Nobody on this site claims to know Navajo authoritatively. Every Navajo form is copied from a named published source with its example number; none is invented. Formal rules are hypotheses to be tested against Grant's texts.

SOURCED stated in a published source we opened HYPOTHESIS our formalization or inference GRANT'S CONJECTURE Grant's own claim OPEN needs data or reading

Background

2026-11-11 audit pass 1. The included working notes previously risked treating older authors’ phase/Agree/word-order models as the standard. They are historical hypothesis comparisons. Only the stipulated subject base in left Spec,vP and theta-role mapping are the starting assumptions here; V→v→T and independent subject raising are Grant’s conjecture to test. Source S-subdivision, light d, phases, A/B and telicity are not additional premises. See the corrected core; unresolved included derivations are recorded for pass 2.

2026-11-13 audit pass 7. Starting assumptions now: the subject base in left Spec,vP and theta-role mapping (VPISH frames where arguments start; it is not a word-order account), and, in tensed clauses, V→v→T plus subject raising Spec,vP→Spec,TP as the working default. The pass-1 note above, which called these movements Grant’s conjecture to test, is kept as written. See the 2026-11-13 note.

Note (2026-11-02): Grant has dropped the “roles, not words” thesis. Pages that test it are kept as a historical record. Current work: Light determiners.

The theory the working notes lean on, kept in one place: Merge and Agree, VPISH and the pre-VPISH sources, Dowty’s proto-roles, event structure, the Algonquian inverse, and a guide to the sources. Copied verbatim from the working notes.

Merge vs Agree · VPISH · Dowty · Event structure · Algonquian · Source guide · Thesis status

Merge and Agree inside the word

Copied verbatim from the working note The Navajo verb, start to finish: template, derivation, selection, Algonquian, phrasal verbs, §6. Are there Merge and Agree sites inside the word?

HYPOTHESIS From intra-word Merge/Agree (T40): Merge sites inside the word are few and at domain edges (bumped object, deictic subject, the ’ placeholder); inner slots are mostly Agree sites (MA-4). GRANT'S CONJECTURE "ordered slots function as successive Merge and Agree sites" holds partly. Grant's "classificatory stem is valued by prefixes" is in no opened source (T40h, OPEN). Algonquian's Root-v-Voice analysis (Oxford 2019) is a concrete published case of word-internal syntax, which makes the general idea respectable, not proven for Navajo.

VPISH, and which sources predate it

Copied verbatim from the working note Framework audit: which sources and which of our rules are pre-VPISH, transitional, or minimalist?, §1. Method and classes

ClassMeaning hereMarkers looked for on the pages opened
(a) pre-VPISH / flat S / transformationalClause built by subdividing S; or rules that move or delete NPs between deep and surface structure; subject not VP-internalFlat S with V, AUX and noun daughters; “deep structure”, “surface structure”, “transformation”, deletion/raising rules over S and NP phrase markers; INFL “governs” the subject with no VP for it
(b) transitionalGB-era, X-bar, early VPISH: heads and specifiers, but S-structure/LF, Case Filter, government or barriers; or functional projections without v/MergeIP, CP, V′, S′, COMP; Spec/Head; VP-internal subject mentioned; trace/chain, Case Filter, c-government, barriers, d-/s-structure
(c) minimalist / VPISH-awareMerge-built tree, external argument in Spec,v/vP, TP/CP above, no D-/S-structureMerge, v/vP with a Spec,vP subject, TP/CP trees, Agree or Spec–Head by feature, no mention of d-structure as a level
(n) not a syntactic-theory sourceEthnographic, semantic, typological or descriptive; no trees or rules over phrase markersAbsence of the markers above after reading the cited pages
(u) unverifiedKnown only through another work, or listed without openingNot placed. A classification “by date” is not made

How the markers were found. Text-layer keyword counts (INFL, VP-internal, Spec, X′, Merge, vP, TP/CP/IP, c-command, deep/surface structure) on the box copies, then page images of the pages where a hit or a tree occurs. Where a scan has no text layer (Hale 1999, W&J 2000, S&J 1989, Witherspoon), I looked at the page images only. A zero count on an OCR layer is weak evidence (Willie 1991's layer is noisy), and is stated as such. Page numbers are the printed ones unless the row says “PDF p.”.

Dowty’s proto-roles

Copied verbatim from the working note Theta roles as arguments: the proto-role schema, §2. Definitions HYPOTHESIS (built from sourced pieces)

Entailment inventory. SOURCED Dowty 1991:572, exx. (27)–(28) (page image checked).

Proto-Agent (27)Proto-Patient (28)
vol: volitional involvement in the event or state
sent: sentience and/or perception
caus: causing an event or change of state in another participant
mov: movement relative to the position of another participant
(indep: exists independently of the event named by the verb)
chg: undergoes change of state
inc: incremental theme
aff: causally affected by another participant
stat: stationary relative to movement of another participant
(dep: does not exist independently of the event, or not at all)

Dowty's Argument Selection Principle (1991:576, (31)): the argument with the most Proto-Agent entailments is lexicalized as subject, the one with the most Proto-Patient entailments as object. Corollary 1, (32), and Nondiscreteness, (34), are on the same page. Dowty adds that the ASP constrains which lexical predicates are possible; it is not a step in a syntactic derivation. That matters below: if the schema is going to do syntactic work, it needs more than the ASP.

Proto-role slot. A slot is a triple

ρ = ⟨ E_ρ , pos_ρ , φ_ρ ⟩
  E_ρ   ⊆ PA ∪ PP         entailment profile, fixed by the event (root + v)
  pos_ρ ∈ {Spec,V ; Comp,V ; Spec,v ; Obl(P)}     first-Merge site
  φ_ρ   = [π: _ , #: _ , rank: _ ]               unvalued at first Merge

Schema. For a root √ and voice/classifier head v, Σ(√, v) = {ρ₁ … ρₙ}, n ≤ 3. First-Merge sites follow Hale's (2003) structural argument positions:

  • sole argument in Spec,V for the labile/unaccusative type (Hale 2003 ex. (22));
  • Spec,v for the unergative type, e.g. cha ‘cry’ (Hale 2003 ex. (23));
  • Comp,V for the object of a transitive (Hale 2003 tree (8)).

Grant's distinctive restriction.

(G)  No lexical DP is ever first-Merged at pos_ρ for any ρ ∈ Σ.

Exponents. The values of φρ are spelled out by verb-internal pronominal exponents. In the sources these are the conjunct object and subject markers: sh- ‘1s’ and ni- ‘2s’ in Hale 2003 ex. (1); the third-person object forms yi-/bi-; the fourth person ji- (subject) and ho- (object) (Willie 1991 §4.1). They also include null exponents, e.g. the Ø 3rd-person subject in the imperfective (NLA Navajo Verb Modes).

Event structure

Copied verbatim from the working note Is event structure built inside the Navajo verb word, alongside the role slots?, §6. Rules

ES-1  Every event verb word has exactly one mode; Perfective/Imperfective/
      Progressive = viewpoint; others = neutral viewpoint.      SOURCED (SPF §3.1)
ES-2  Mode + subject = one portmanteau in M, above vP.          SOURCED (Hale ms.:10)
ES-3  Perfective ⇒ bounded; telic ⇒ completed; atelic perfective ⇒
      arbitrary endpoint.                                        SOURCED (SPF §4.1)
ES-4  Telicity is semantic (event structure), not a grammaticalized
      situation type in Navajo.                                  SOURCED (SPF §4.1; Smith 1996)
ES-5  ł alone = transitivity alternation, only for verbs with an internal
      theme (change of state); else y- + ł causative.            SOURCED (H&P 1996)
ES-6  The verb word assembles Σ (role slots) AND viewpoint + situation
      category before any noun is adjoined.                      HYPOTHESIS (from ES-1, ES-2,
                                                                 Willie 1991:51–52)
ES-7  Telicity can depend on material outside the word ('ałtso, incremental
      object).                                                    SOURCED fact (28a);
                                                                 incremental-theme reading HYPOTHESIS
ES-8  Aspect category restricts the role frame.                   GRANT'S CONJECTURE / OPEN

The Algonquian inverse

Copied verbatim from the working note The Algonquian inverse: same roles, a different subject slot (the "crossing flip", T52), §9. Bottom line

SOURCED Theta roles never change. In every Algonquian direct/inverse pair opened (A1–A8), the agent is the agent and the patient is the patient. The inverse changes which one controls the subject agreement, and Bruening stresses it is "not a passive".

SOURCED Two inverses. 3→1/2: obligatory, Agree only, no movement (shallow). 3′→3: optional voice, the patient raised over the agent, binding reversed (deep) (Oxford 2023).

HYPOTHESIS "Crossing flip" is the Morphic bot's own braid term. It fits only the deep inverse: a swap of structural order with the roles fixed.

HYPOTHESIS Merge vs Agree. Shallow = Agree; deep = Agree plus internal Merge. Theta roles are set at external Merge, which the inverse leaves alone.

HYPOTHESIS Grant's thesis. Algonquian ranks persons, not proto-roles. Blackfoot's sentience ranking (Bliss 2005) is the one place it touches a Dowty entailment.

HYPOTHESIS Navajo. yi-/bi- lines up with the deep inverse only (Willie 1991: 96–97).

OPEN Chemistry. None of these sources mentions chemistry, and this page claims no chemistry analog.

Source guide

The Sources page lists, run by run, what was opened and what is only cited. Pre-VPISH classes are in the framework audit.

Where the thesis stands

See About: the honest verdict and the essay.

Legacy link recovery
Legacy section links (scroll sideways); old articles remain available.
intra-word-merge-agreemerge-agree-originalsepp-2000-2001

Merge and Agree inside the Navajo verb word: which word-internal facts are Merge sites, which are Agree, what is a phase, and is anything [+MFS]-like? (T40)

Working note. Kept as written (URL, data and status labels unchanged). Filed under Merge and Agree inside the word; the consolidated argument is the essay merge-agree. Rule IDs: rules index.

Verbs on this page → dictionary: yishxash ‘bite’ · dabiiłtsą́ ‘see’ · yiyííbį́į́ʼ ‘pick (round objects, e.g. berries)’ · yiyííłtaʼ ‘count (plural objects)’ · yíghááh ‘join; with -tah ‘among’: go among’

Restated on the unified structure (2026-11-15). Prefix order is restated as Grant’s hypothesis that headedness is fixed locally by Merge site, so the sublexical domain is linearized opposite to the phrasal domain and the prefixes are successive word-internal Merge and Agree sites (GRANT’S HYPOTHESIS, tested on the sourced template in §3.2; fit/misfit OPEN). A possible sublexical V-to-T is a separate, untested GRANT’S HYPOTHESIS (§3.3). The head-direction mismatch argued below as rule sets A/B or a changeover is now stated as Grant’s Merge-site hypothesis and tested in §3.1–3.2. Position-class template statements below describe the order of prefixes; they are not the derivation (§3). Derivations below in which the verb or subject does not raise are labelled contrast; the working default raises V→v→T and the subject to Spec,TP. Status on the spine: see unified structure, §3. The data, judgments, translations and citations below are unchanged; the analysis below is kept as the record of what was argued then.

2026-11-11 audit pass 1. MA-, EF-, TP- and related mid-level rules import hypotheses about phasehood, probes and feature inheritance; none follows from VPISH plus theta roles. “No movement needed” in the old comparison is not evidence against Grant’s primary V→v→T derivation. Existing step lists are retained as historical hypotheses; target creation, locality and LF must be re-derived on a left-specifier vP/TP skeleton in pass 2.

2026-11-13 audit pass 7. An earlier dated note on this page treats V→v→T and subject raising as Grant’s conjecture or a comparator to test against PF/template rivals. Since 2026-11-13 they are the working default in tensed clauses: VPISH frames where the arguments start so that syntax and semantics line up; it is not a word-order account. PF/template accounts are kept only as labelled contrast, and the open question is what else has to move. The earlier note is kept as written. See the 2026-11-13 note and the pass 7 report.

Grant's note “Typology of Merge/Agree Phenomena” sets three levels apart: primitives (Merge, Agree, phases, locality), mid-level configurations (Rudin's [±MFS], Multiple Agree, feature inheritance, Edge Features on particular heads, phasehood choices, wh-clustering) and surface constructions. His working notes “Merge Sites and Agree Sites” (August 2026) propose that in a template language like Navajo “ordered morphological slots function as successive Merge sites and Agree sites inside the complex word” (§2, §7). This page takes that idea and sorts the word-internal findings already on this site by mid-level setting. It keeps the two operations apart throughout. Merge builds structure: an element is placed at a position, either fresh (External Merge) or moved (Internal Merge). Agree values features: a probe on a head finds a goal and copies its value, and nothing moves. Written 2026-10-19, 1:52pm PT.

Update 2026-10-19, 8:46pm (T40b). The originals behind this page's mid-level terms have now been opened: Rudin 1988, Hiraiwa 2001 and Chomsky 2008 in full, and Grewendorf 2001 as an abstract only. The results are on the borrowed terms checked against their originals. In short: the picture here (Agree inside, Merge at edges, [−MFS]-like) survives, with five corrections. (1) The leftover deictic (MA-2b) is OPEN between Agree and Merge, because Chomsky's inheritance allows both readings. (2) The ‘start to’ placeholder matches Chomsky's expletive EM (2008 p. 157) directly. (3) The object prefix is not a Hiraiwa Multiple Agree candidate, and SH-6 is favoured. (4) The PP being a phase is our extension, and an inactive-goal alternative is added. (5) Rudin's decisive test is the clitic position, and the disjunct order is not her kind of order. Rules MA-0′ to MA-6′ and derivations MA1′ to MA10′ are on that page. The rules and derivations below are kept as written at 1:52pm.

Update 2026-10-20, 4:46am (T41e). After Chomsky 2000/2001 (the older definitions behind “edge feature”), the ‘start to’ placeholder (MA-2c) is still a Merge site, but its head is PINCH, low in the word, not the top. Both texts allow expletive insertion only at a non-phase head. The bumped object (MA-2a) is Move to the same position. Grant's “whether a head has an EPP-feature” is confirmed as the 2000/2001 notion.

Update 2026-10-20, 12:46pm (T41f). The n. 12 datum cited in §6 (“an enclitic can split a preverb off”) is a split between two words: Hale's ka ‘sick’ stands as its own word (where the yes/no enclitic -ísh can go). So it no longer “points to divisibility” of the disjunct zone, and §6's lean is dropped (OPEN, no lean). The clitic test of T40a cannot be run with -ísh if the enclitic never enters the verb word (EN-1′, a prediction). T40a is closed into T43a/T43b.

Ground rules (Grant's). Every Navajo form below is already on this site, copied from a named source with its example number. The page links to where it was read. No new Navajo form is introduced and none is segmented beyond what its source prints. Each source and each earlier page of this site may be partly right. Only contradicting data count against an analysis. The mid-level terms ([±MFS], Multiple Agree, feature inheritance, Edge Feature, phase) are general-syntax notions. The originals (Rudin 1988, Hiraiwa 2001, Chomsky 2008, Grewendorf 2001) were not opened for this page. (They were opened at 8:46pm: see the definitions side by side.) The definitions used here are the ones in Grant's note, and no Navajo source opened on this site uses any of these five terms. Every mapping of a Navajo fact onto one of them is therefore ours and is at most HYPOTHESIS. Grant's first notes contained no astrophysics analogies (he sent them separately later; see §11). The note named for this task (“Merge_and_Agree_Sites_Exploratory_Notes”, text and PDF, 6 pp.) was searched in full, and so were the two raw chat exports in the same folder. None mentions astrophysics.

Notation (the site's own). Σ = role-schema step; A = word level (cycle D1 = QualP + VP-internal PPs = Willie's MVS, MVS-D1); B = clause level (D2). Spine (Hale 2003): Qual > M/S > v > V, with lexical V, P, N head-final and the functional heads head-initial (p. 32). Slots ρ = ⟨E, pos, φ, (δ)⟩ (SH-0). CYC = C1 non-heads before heads, C2 non-heads higher-first, C3 heads higher-first (spell-out rules §3). VPISH: the subject slot is first-merged in Spec,v. In Hale 2003 “There is no head-movement in the traditional sense” (p. 30) and arguments are “licensed in their thematic positions” (pp. 29–30).

Bottom line, in plain words

HYPOTHESIS Most of what happens inside the Navajo verb word is Agree, not movement. Three facts are the clearest. The ordinary subject prefix is pronounced on the Mode/Subject head. The subject's slot stays where it was built, and that head copies its person and number (Hale 2003 n. 11's portmanteau condition, read under VPISH: subject on the head). The postposition's own pronoun, including the inverse bi- in W&J (50b) bitah, is valued inside the postpositional phrase. The choice between yi-, bi- and zero is made inside the word (D1), before any noun arrives. None of these needs an element to move inside the word.

HYPOTHESIS The word-internal Merge sites are few, and they are all at the left edge of a domain. There are three candidates. (1) An object that is bumped to the far left when the object position is already taken (Hale, Munro & Platero's “bumping” into “disjunct prefix Position I”, n. 2). (2) The deictic part (fourth person ji-, indefinite ’-, areal ho-) of a subject's bundle, which attaches to the highest head of the word. That is Merge by adjunction, not a specifier landing. (3) The role-less placeholder ’ that ‘start to’ adds when the verb supplies nothing for its object requirement. That last one behaves like an expletive filling an Edge Feature or EPP requirement. Grant's “Merge sites inside the word” therefore survives, but in a narrow form: it holds for edges, and the template's inner slots are mostly Agree sites.

HYPOTHESIS Feature inheritance is a good model for the split subject. If the qualifier (the top of the word, standing where C stands in Chomsky's model as Grant's note reports it) hands person and number down to Mode/Subject (where T stands, Hale 2003 p. 10: M/S = Infl/T), then two things follow. Ordinary person and number are pronounced on Mode/Subject. The deictic feature stays on top, on the qualifier. That is exactly what Hale found twice (the deictic is “a prefix to E”, 1997; it “prefixes to the Qualifier element”, 2001). It derives the same strings as the site's rule SH-4 and adds nothing, so it is a restatement, not new evidence.

HYPOTHESIS Where Navajo looks like [−MFS], not [+MFS]. Rudin's [+MFS] means a position that is already filled can host more. The two cases in the sources where a word-internal position is already filled both show the opposite. Under ‘start to’ the object position is full, so the real object is pushed out to the far left. When a thematic prefix (so-) already holds the far left, the bumped element does not join it. It stays and fuses (HMP n. 4). Neither position takes two occupants. The disjunct prefixes do co-occur, but in a fixed order by height, not in Rudin's attraction order, and Hale reports that an enclitic can split a preverb off (2003 n. 12). So disjunct “clustering” fails at least one of Rudin's diagnostics as far as it can be checked. Whether the others hold is OPEN.

HYPOTHESIS Phases inside the word. Two domains behave like phases. The word itself (D1) is spelled out once and then counts as a single block (the site's spell-out-twice rule, CYC). The postpositional phrase is opaque to Mode/Subject's probe: a participant inside a PP never becomes the subject (IN-3, supported by every ‘start to’ form). One older analysis lets a theme escape a PP through its edge (Hale 1997: it raises from Spec,PP to Spec,MP). That is phase-edge escape in modern terms, though Hale does not call it that.

OPEN Multiple Agree has candidates, but none is decided. The third-person object prefix depends on the subject (zero under a fourth-person subject in jishxash, yi- under a noun-linked subject in ʼashkii yishxash). The voice prefix compares two arguments' rank. Each could be one probe agreeing with two goals at once, or ordinary agreement plus a context-sensitive pronunciation rule. The strings are the same. No opened source has a form with one prefix that spells both the subject's and the object's person, which is the clean test.

GRANT'S CONJECTURE “Navajo has intra-lexical Merge sites, and possibly Agree too” is Grant's. The split proposed here (Agree for the inner slots, Merge for the edges) is our refinement, graded HYPOTHESIS. Nothing in the sources contradicts it, but no source states it.

On this page: 1 Sources · 2 The mid-level terms, kept apart · 3 Classification table · 4 Formal statements · 5 Derivations MA1–MA9 · 6 Is anything [+MFS]-like? · 7 Phases inside the word · 8 What each option gets right and misses · 9 Grant's notes: what is kept, what is changed · 10 Tasks T40a–j · 11 Questions from the astrophysics notes

1. Sources

SourceAccess this runUsed for
Grant, “Typology of Merge/Agree Phenomena: A Takeaway Note on Mid-Level Configurations of Core Syntactic Operations” (PDF, 2 pp.; attachment of 2026-10-19)Opened (full text)The three levels (§4); the definitions of [±MFS], Multiple Agree, feature inheritance, EPP / Edge Feature, phasehood and wh-clustering (§§2–3). Its source list (Rudin 1988; Grewendorf 2001; Bošković 1997–2002; Hiraiwa 2001; Chomsky 2008; Krapova & Cinque; Billings & Rudin) was not opened
Grant's Grok notes, “Merge Sites and Agree Sites: Exploratory Notes Toward a Separate Research Project” (August 2026 working draft, 6 pp.)Opened (text and PDF; same 3,268 words)Core claims (§2), Navajo “word-level specialization” (§3.3), “Agree sites” (§4), the template as Merge/Agree sites (§7), open questions (§10). No astrophysics analogies are present in this file or in the two raw chat exports beside it
Grant, “Handoff Document: Hierarchical Sites, Residual Memory, and Quark Logic” (October 2026; attachment of 2026-10-19, 2:32pm)Opened (full text)Questions only (§11). Its astrophysical figures are not cited; §8 (quark logic) and the kanji remark tabled
This site's pages (Navajo forms and rules, each read earlier on page images or text, with sources listed there)Re-readSubject on the head (Hale 1997, 2001, 2003; Willie 1991; Kibrik 2005); ‘Start to’: HMP ms. and the ‘start to’ test; inverse as Agree/EPP; clause spine; postposition units; A/B as spell-out; feature bundles
Not openedRudin 1988; Hiraiwa 2001; Chomsky 2008; Grewendorf 2001; Bošković; Rice 2000 (known here only secondhand); Young & MorganNot used as evidence

2. The mid-level terms, kept apart

Definitions as in Grant's note (§§2–3). The right-hand column says what would count as the configuration inside a Navajo word. That column is ours.

SettingOperationGrant's noteWhat would show it inside the word (ours)
[±MFS] (Rudin)Merge“in [+MFS] languages an already-moved wh … can serve as a further landing site, so later whs Merge to it and form a cluster. In [–MFS] languages no such intermediate host is available”A filled template position that takes a second occupant (or a cluster that moves and splits as a unit)
Edge Feature / EPP on a headMerge (triggered)“Whether a given head (T, v, C…) has an Edge Feature triggering Spec Merge”A position that must be filled, by a placeholder if nothing else is available
Multiple Agree (Hiraiwa)Agree“Whether one Probe can Agree with multiple Goals in a single step”One exponent valued by two slots
Feature inheritance (Chomsky 2008)Agree (which head probes)“Whether φ-features on C are inherited by T”; “Controls which head is the actual Agree site and what moves where”One bundle's features pronounced on two heads, higher and lower
PhasehoodLocality for both“Which heads count as phases (barriers to Agree and movement)”A domain whose contents a higher probe cannot see, or that is spelled out as one block
Wh-clustering (Grewendorf)Merge (feature-based [+MFS])“Whether a moved wh can itself host further whs”As [+MFS]

3. Classification table: each word-internal finding by mid-level setting

#Finding (already on the site)OperationMid-level settingGradeDerivation
W1The prefix template is ordered non-heads first, then heads, each higher-first (CYC); this rests on Hale's head-initial functional spine (p. 32)neither: linearization of a Merge-built structureNot a mid-level setting. It is how a phase is pronounced (spell-out), as Grant's notes themselves say (§3.1: “linearization is local”)HYPOTHESIS (CYC)all
W2Ordinary subject φ is pronounced on M/S (Hale 2003 n. 11; SH-2 PORT)AgreeProbe on M/S (the Agree site), goal ρA in Spec,v; no movement (VPISH)SOURCED condition; Agree reading HYPOTHESISMA1
W3Deictic subjects (4th ji-, indefinite ’-, areal ho-) are pronounced left of the qualifier (Hale 2001 (22a–c); Kibrik (8a))Agree for φ, then Merge (head-adjunction) of the leftover δFeature-inheritance model: Qual hands φ to M/S and keeps δHYPOTHESISMA2
W4The third-person object exponent depends on the subject (Willie (23b) jishxash vs p. 129 ʼashkii yishxash; SH-6)Agree, possibly with two goalsMultiple Agree candidate, or contextual allomorphyOPENMA3
W5The inverse on a postposition: W&J (50b) bitah yíghááhAgree inside PP (IA-3); Internal Merge only under IA-1*PP as an Agree domain; under IA-1*, escape needs a PP edgeHYPOTHESIS (IA-3 fits directly)MA4
W6‘Start to’ bumps the object to Position I (HMP n. 2; (10a) shináʼniidlóóh; IN-5′)Merge at the left edgeA filled position does not host: [−MFS]-like. Also, the PP is opaque to PORT: phase-likeSOURCED pattern; settings HYPOTHESISMA5
W7A thematic Position I prefix blocks the bump (HMP n. 4 sodiʼniiszin)blocked Merge; fusion[−MFS]-like at Position ISOURCED pattern; setting HYPOTHESISMA6
W8‘Start to’ needs an overt object; a verb with nobody low gets a role-less ’ (HMP (7b), (16a); IN-8, IN-9)Merge of a placeholderEdge Feature or EPP on INCH, satisfied by an expletive-like elementSOURCED requirement; EPP reading HYPOTHESISMA7
W9da- counts the agent although it precedes bi- (Willie (64a) dabiiłtsą́)AgreeOption e4 = da- shares the number M/S took (feature sharing); e2 = da- probes for a roleOPEN (trilemma on K5)MA8
W10A one-place verb whose participant starts low (areal (22c) hodildóóh; SH-8)Internal Merge (U-1) or Agree with the closest goal (U-2)The one place where Merge and Agree compete inside the word with the same stringOPEN (U-2 favoured on economy)MA9
W11The word is spelled out once and then ordered as one block (CYC, D1/D2; Hale n. 12: where the word stops is unsolved)localityPhase = D1HYPOTHESIS§7
W12The verb's arguments are feature bundles (FB-1 to FB-5)(the goals)Agree needs goals with features. Bundles are exactly that, with or without a nounGRANT'S CONJECTURE formalized as HYPOTHESIS§9

4. Formal statements (ours; all HYPOTHESIS)

Update 8:46pm: revised as MA-0′ to MA-6′ after reading the originals. MA-0's test would class subject agreement as Merge. MA-3 is not Chomsky's own version. MA-6's option is not Hiraiwa's Multiple Agree. MA-5 is our extension. See the revised rules and what changed.

MA-0 SEPARATE   Merge(α, β) builds structure. Agree(P, G) values P's unvalued features from G.
                A word-internal finding is classified as Merge only if an element is pronounced at
                a position other than where its role puts it, or a position is filled by an element
                with no role. Otherwise it is Agree plus CYC spell-out.
MA-1 AGREE-SITES  The probes inside D1 are M/S (phi of the slot it locally c-commands and governs:
                PORT, Hale 2003 n. 11), P (its own pronoun: AGR on P, Hale 1997), and the voice
                position (the yi-/bi-/0 choice, SH-6). No probe in D1 triggers movement of a slot,
                except under IA-1* (v's EPP in the inverse) and U-1.
MA-2 EDGE-SITES  The Merge sites inside D1 are (a) the left edge of the V-level non-heads
                (Position I: a bumped PP, IN-5'); (b) adjunction to the highest nuclear head (a
                leftover delta, SH-4); (c) the object position of INCH, filled by a role-less rho_D
                when no lower slot exists (IN-9). (a) and (c) are positions with a requirement.
                (b) has no requirement; it is where leftover material goes.
MA-3 INHERIT    (feature-inheritance restatement of SH-2 + SH-4.) The phase head of D1 (Qual)
                bears the subject probe [uphi, udelta]. It passes [uphi] to its complement head
                M/S and keeps [udelta]. M/S Agrees with rho_A (phi). Qual Agrees with rho_A's
                delta if it has one. Exponents: phi in M/S; delta adjoined to Qual (or DETR if
                present). Same strings as SH-2 + SH-4.
MA-4 NO-MFS     A template position holds one exponent. A second candidate either goes to a
                higher edge (IN-5') or fuses (' + ' -> ', HMP n. 4). No position is a host for a
                cluster. (HYPOTHESIS; two positions tested.)
MA-5 PHASE-PP   A PP inside D1 is a phase for M/S's probe (IN-3). Its edge (Spec,PP) is
                accessible (Hale 1997's theme raising from Spec,PP), its complement is not.
MA-6 MA-CAND    (OPEN.) Either the voice exponent of rho_X is valued by one probe that Agrees
                with rho_A (delta, phi) and rho_X (phi) in one step (Multiple Agree), or rho_X is
                valued alone and spelled by a context-sensitive rule that reads rho_A (SH-6).

5. Derivations MA1–MA9

MA1. Hale 2001 (24a) shidoołhosh ‘He/She will tickle me’: Agree only (as SD1)

  1. Σ Σ(√TICKLE) = {ρA, ρX}; φ(ρA) = 3, no δ; φ(ρX) = 1sg.
  2. A External Merge: V ghosh + ρX (Comp,V); v ł; ρA in Spec,v (VPISH); M/S; Qual d. Every slot is merged where its role puts it.
  3. A Agree (MA-1): M/S probes. ρA is the specifier of its complement and the closest goal, so M/S gets φ = 3 → w (“3rd:Perf”). No Internal Merge: ρA stays in Spec,v.
  4. A Spell-out D1 (CYC): sh ⌢ d ⌢ w ⌢ ł ⌢ ghosh = Hale's segment line ✓.
  5. Classification: one Agree site (M/S), no word-internal Merge site. The template order here is spell-out, not movement.

MA2. Hale 2001 (22b) shiʼdooʼniid ‘I was told (lit. someone told me)’: feature inheritance as a model of the split subject (as SD4)

  1. Σ ρA = ⟨φ = 3 (zero), δ = INDF⟩; ρX = 1sg.
  2. A External Merge: V niid + ρX; v d (Hale n. 6); ρA in Spec,v; M/S (w); Qual d with [uφ, uδ].
  3. A Inheritance (MA-3): Qual passes [uφ] to M/S. Agree M/S ⟶ ρA: φ = 3 (zero). Agree Qual ⟶ ρA: δ = INDF.
  4. A Merge by adjunction: the δ exponent ’ left-adjoins to Qual: [’ + d]. This is a word-internal Merge site (MA-2b), and the only one in this form.
  5. A Spell-out D1: sh ⌢ ’ ⌢ d ⌢ w ⌢ d ⌢ niid ✓ (Hale's segment line).
  6. Comparison. SH-2 + SH-4 give the same string with no inheritance step. MA-3 adds a reason why the leftover goes to the top head: that head held the probe first. It makes one further prediction: if a word has DETR above Qual (SD8, Kibrik (8a)), the phase head must be DETR, not Qual. Whether that is acceptable is T40c.

MA3. Willie (23b) jishxash ‘That person bit it’ vs p. 129 ʼashkii yishxash ‘The boy bit it’: a Multiple Agree candidate (as SD6)

  1. Σ Σ(√BITE) = {ρA, ρX}; φ(ρX) = 3. jishxash: ρA = ⟨3 (zero), δ = 4⟩. ʼashkii yishxash: ρA = ⟨3, no δ⟩.
  2. A External Merge, Agree M/S ⟶ ρA, as MA1. For jishxash, MA2's steps 3–4 (δ = 4 on Qual).
  3. A Option MA (Multiple Agree): one probe at the voice position Agrees in a single step with ρX (φ = 3) and with ρA (δ present or absent). Its exponent is y- if ρA is φ = 3 with no δ, and Ø if δ = 4.
  4. A Option CA (contextual allomorphy, SH-6): the probe Agrees with ρX only. At spell-out the exponent rule reads whether ρX is c-commanded in D1 by a δ-less third person.
  5. A Spell-out D1, either option: j ⌢ … ⌢ shxash; yi ⌢ … ⌢ shxash ✓ (not segmented further).
  6. B ʼashkii links to ρA in D2. It cannot change the D1 choice (“B is inert”, WF-5), so the deictic δ must already be in D1.
  7. What would separate MA from CA: a case where ρX is not third person but its exponent still varies with ρA's δ. (MA predicts the probe can still read ρA. CA ties the rule to the third-person exponent only.) No such form is in an opened source (T40d).

MA4. W&J (50b) dibé łį́į́ʼ bitah yíghááh: Agree inside the PP; Merge only under IA-1* (as L5, PU4)

  1. Σ ρA (the one who goes), ρX = P's object; both π = 3. Inverse: ρX [top], ρA [foc].
  2. A External Merge: P -tah + ρX (head-final PP). Agree inside PP: P's own agreement (Hale 1997: AGR on P) with ρX. Under IA-3 ρX already carries [top], so the exponent is b- → bitah.
  3. A External Merge: PP inside VP; v; ρA in Spec,v; M/S; Agree M/S ⟶ ρA (MA-5: ρX is inside the PP, not a goal); Qual.
  4. A IA-1* branch: v's EPP feature in the inverse should attract the topic slot (Internal Merge). ρX is in the PP's complement, which MA-5 makes inaccessible. So either (i) ρX first moves to Spec,PP, the PP edge (Hale 1997's position for a theme in a PP), or (ii) IA-1* fails for postpositional objects. Under (i), bi- is pronounced on P only if the lower copy is the one pronounced (T31e).
  5. A Spell-out D1: bitah ⌢ yíghááh.
  6. B łį́į́ʼ → [foc] = ρA; dibé → [top] = ρX. D2: dibé ⌢ łį́į́ʼ ⌢ [bitah yíghááh] ✓.
  7. Classification: Agree site inside the PP. The PP behaves as an Agree domain closed to M/S. A word-internal Merge step is needed only by IA-1*, and then it must go through the PP edge. That is one more reason the site's in-situ option IA-3 is cheaper here.

MA5. HMP (10a) shináʼniidlóóh ‘I start to get cold (about dawn)’: a left-edge Merge site, a filled position that does not host, and a PP opaque to Agree (as HD2)

  1. Σ Σ(√GET.COLD) = {ρ1}, first-merged in Spec,V; φ = 1sg.
  2. A External Merge: V dlóóh, ρ1; REP ná; INCH puts the thematic ’ (ρI) in the object position (HMP: “the Position IV indefinite object prefix … functioning here as a thematic element”).
  3. A MA-4 NO-MFS: the object position is full. ρ1 cannot join it. IN-5′: [PP ρ1 PINCH] is merged as the highest V-level non-head (MA-2a), above ná. In HMP's words the object is “bumped” “into the disjunct prefix Position I” (n. 2).
  4. A Agree M/S: Spec,v is empty. ρ1 is inside a PP (MA-5, IN-3), and ρI is thematic, so there is no goal. M/S takes the default (IN-7 (D); HMP: the subject position is “entirely empty … as it is regularly in persons other than first or second”).
  5. A Spell-out D1: shi ⌢ ná ⌢ ʼnii ⌢ (M Ø) ⌢ d ⌢ dlóóh ✓ = shináʼniidlóóh. HMP do not segment (10a); the alignment is the site's.
  6. Classification: one Merge site (left edge), a [−MFS]-like refusal (object position), a phase-like PP (opaque to M/S), and Agree failing to find a goal. It is the richest single form for this page.

MA6. HMP n. 4 sodiʼniiszin ‘I started to pray’: the left edge is full too (as HD6)

  1. Σ ‘pray’ = theme so=d-l-zin; ρA = 1sg in Spec,v; nothing below v, so INCH's requirement is met by ρD (IN-9 (c)).
  2. A MA-4: the thematic Position I prefix so holds the left edge. The candidate does not join it (no cluster). It stays in the object position and fuses with ρI: ’ + ’ → ’ (HMP: “hypothetically at least”).
  3. A Agree M/S ⟶ ρA (1sg). Spell-out: so ⌢ di ⌢ ’ ⌢ nii ⌢ … ⌢ zin ✓ (HMP give no further segmentation).
  4. Classification: the escape site is itself one-occupant. That is a second [−MFS]-like position. Fusion is not hosting: one exponent results.

MA7. HMP (16a) ʼiʼniichííl ‘it started to snow’ and (7b) ʼawééʼ ʼináʼniichah ‘the baby starts to cry’: an EPP-like requirement met by a placeholder (as HD3, HD7)

  1. Σ ‘snow’: no participant. ‘cry’: ρA in Spec,v only; nothing below v.
  2. A IN-8: INCH “must always be preceded by overt object agreement morphology” (HMP). Read as an Edge Feature or EPP on INCH: a position that must be filled.
  3. A IN-9: no slot below v, so a fresh role-less ρD = ⟨E = none, φ = 3, δ = INDF⟩ is Merged, spelled ’. It is expletive-like: it fills a position and has no role (HMP: “use of indefinite agreement is a reasonable alternative”).
  4. A Spell-out: ‘cry’ ʼi ⌢ ná ⌢ ʼnii ⌢ … chah ✓; ‘snow’ ʼi ⌢ ʼnii ⌢ … chííl ✓ (not segmented further than HMP).
  5. B ʼawééʼ links to ρA, not to ρD (ρD has no role to match).
  6. Classification: the clearest word-internal EPP effect in the sources. The contrast with (10d) biʼniichxʼǫʼ ‘(this meat) has started to spoil’, where the participant itself fills the position, is the one English shows between there-insertion and raising. That parallel is ours.

MA8. Willie (64a) dabiiłtsą́ ‘They/he (was)were seen by them/*him’: da- as shared agreement (as K5)

  1. Σ ρA plural (W&J gloss PL-3ACC-3NOM-saw); ρX 3; inverse.
  2. A External Merge; Agree M/S ⟶ ρA (φ including plural number).
  3. A e4 as feature sharing: da- has an unvalued number feature and Agrees with M/S, not with a slot. It receives the agent's number second-hand, so its position above b- does not matter. This is a chain of two Agree steps, not Multiple Agree. e2: da- probes for ρA's role directly.
  4. A Spell-out D1: da ⌢ b ⌢ heads → dabiiłtsą́ ✓.
  5. Classification: Agree. Under e4 this is the inheritance pattern again (a value passed from head to head), applied to number. The site's own derivation does not decide between e2, e3 and e4 (T30 tasks).

MA9. Hale 2001 (22c) hodildóóh ‘It (e.g., cloud) starts to drift’: where Merge and Agree compete (as SD5)

  1. Σ ρ1 = ⟨φ = 3 (zero), δ = AREAL⟩. Whether ρ1 starts in Spec,v or Spec,V is unknown (SH-8).
  2. A U-1 (Internal Merge): if ρ1 starts in Spec,V, it raises to Spec,v inside D1, where M/S locally governs it, and then Agree.
  3. A U-2 (Agree only): M/S Agrees with the closest slot in its domain, wherever it is (IN-4 PORT-CLOSEST).
  4. A Either way, δ = AREAL is left over and adjoins to Qual (MA-2b): [hw + d]. Spell-out: hw ⌢ d ⌢ Ø ⌢ l ⌢ dóóh ✓.
  5. Classification: same string, different operation. The ‘start to’ facts favour U-2 on economy: there the low participant's subjecthood is bled with no movement. Only a movement diagnostic would decide (T37b, T40e).

6. Is anything [+MFS]-like?

Update 8:46pm: checked against Rudin 1988. [+MFS] needs adjunction to a filled SpecCP (pp. 480–481). Her decisive [−MFS] test is a clitic right after the first element (pp. 461–466), which Hale 2003 n. 12 reports for ka. Her fixed order is subject before non-subject (p. 475), so the disjunct order below counts for neither type. See Rudin's diagnostics applied.

CandidateWhat the sources showRudin's diagnostics (as in Grant's note)Verdict
Object position (IV) already filled by INCH's ’The object does not join it. It is bumped to Position I (HMP n. 2; (2b) bi-ná-ʼnii-sh-…, (10a))No second occupantHYPOTHESIS [−MFS]-like
Position I already filled by thematic so-The bumped element stays and fuses (HMP n. 4; sodiʼniiszin)No second occupant; fusion yields one exponentHYPOTHESIS [−MFS]-like
Hale 2003 n. 13: an oblique takes “the highest prefixal position when not in competition with a preverb”A preverb at the edge blocks the oblique's jump (cited on spell-out rules (f2))Competition, not clusteringHYPOTHESIS [−MFS]-like (third case, same pattern)
Several disjunct non-heads in one word (bumped PP + ná-; preverb + iterative + plural, Hale's trees (11), (31b))They co-occur at one level, ordered by height (C2), each adjoined to the projection of the highest nuclear head (Hale pp. 16–17)Order: fixed by height and class, not by attraction order (no Superiority analog in any opened form). Indivisibility: Hale 2003 n. 12 reports that the preverb ka “can be separated from the rest of the verb by the polar interrogative enclitic” (forms printed in a broken font, not reproduced). Clitic placement after the whole group: unknown. Multiple extraction: not applicableOPEN, leaning “not a cluster”. Multiple adjunction to one projection has the shape of multiple specifiers, but the one diagnostic that can be checked (n. 12) points to divisibility

HYPOTHESIS Summary. Wherever the sources show a word-internal position that is already filled, the second candidate goes elsewhere or fuses. That is the [−MFS] behaviour Grant's note describes (“no such intermediate host is available”), transposed to template positions. The disjunct zone allows several elements, but as separate adjuncts in a fixed order, not as a cluster built on a first-moved element. The test that would show a genuine cluster: does the polar enclitic go after the first disjunct or after the whole disjunct group? It needs forms with two or more disjuncts and the enclitic (T40a).

7. Phases inside the word

DomainBehaves like a phase becauseAgainst, or unknownGrade
D1 = QualP + VP-internal PPs (the word / MVS)Spelled out once, then ordered as one non-head in D2 (CYC); nouns link after it and cannot change its choices (WF-5; MA3 step 6); W&J: the MVS is “a noninterruptible constituent … the domain of the voice alternation” (p. 270)The boundary is stipulated; Hale: “when to stop building onto the skeleton” is unsolved (n. 12); preverbs can be separate wordsHYPOTHESIS
PP inside D1Opaque to M/S's probe (IN-3; every ‘start to’ form); has its own Agree (P's pronoun)Whether lexical PPs are in D1 at all is T31f; the edge (Spec,PP) is accessible only in Hale 1997's analysisHYPOTHESIS
vP—Nothing on the site needs vP to be a phase inside the word. M/S Agrees into vP (with Spec,v and, under U-2, below it)OPEN: no evidence either way

HYPOTHESIS Grant's notes ask (§10, Q4): “Are phases themselves the natural domains that delimit the possible heights of Merge and Agree sites?” On the Navajo word, a partial yes. Every Merge site found (MA-2a, b, c) is at the edge of D1's V-level material or on D1's highest head. Every Agree site (M/S, P, voice) is inside D1 and probes down. That is the distribution a phase-edge account predicts. It is a fit, not a test.

8. What each option gets right and misses

OptionGets rightMisses or costs
Grant's notes as stated (“ordered morphological slots function as successive Merge sites and Agree sites”, §7)The template is the place where the decisions are made; the inverse “is fixed at a specific height inside the conjunct domain” (§4) fits IA-3 and SH-6“Successive” Merge sites slot by slot would mean each prefix is a landing site. The sources show most prefixes are pronounced where their slot was built (no head movement, Hale p. 30). The notes also say “the classificatory stem … carrying … Proto-Patient properties … must be valued” by the classifier and prefixes (§4). That is not in any opened Navajo source (T40h)
This page (Agree inside, Merge at the edges)Separates the two operations; places every word-internal finding; keeps all stringsAdds no new prediction except MA-3's DETR case (T40c) and the enclitic test (T40a); the Multiple Agree question is left open
A movement-heavy alternative (every prefix raises to its template position, as antisymmetry requires: spell-out rules §5)Gives every prefix a Merge site, as Grant's “successive Merge sites” suggestsThe object must raise above the subject, which Hale considered and rejected (2003:11–12), and da- then counts the wrong slot (K5)

9. Grant's notes: what is kept, what is changed

  • GRANT'S CONJECTURE → HYPOTHESIS “Specialized Merge sites and Agree sites are local decisions fixed at particular heights or morphological domains” (§11). Kept: §3 places each word-internal finding at a height. Changed: inside the Navajo word the sites are mostly Agree. The Merge sites are edges.
  • GRANT'S CONJECTURE “The yi-/bi- alternation is the morphological reflex of an Agree relation that reads the relative ranking of two arguments” (§4). Kept as a Multiple Agree candidate (MA3, W4). It is equally compatible with the in-situ topic analysis (IA-3), where one slot carries [top].
  • GRANT'S CONJECTURE Feature bundles (FB-1 to FB-5). Agree needs goals with φ and δ features, present whether or not a noun is linked. Every derivation above values features on bundles in D1 before any noun arrives. So the bundle picture is what makes the Agree-heavy classification work.
  • HYPOTHESIS Grant's notes distinguish the classificatory stem series from the position-IX classifier (§7). Nothing on this page depends on that distinction; it is recorded, not tested.
  • Grant mixes Merge and Agree in places (e.g. the template slots as “Merge sites and Agree sites” without saying which is which). Rule MA-0 is the separation used here.

11. Questions from the astrophysics notes

What this section is. Grant later sent his astrophysics handoff, “Hierarchical Sites, Residual Memory, and Quark Logic” (2026-10-19 attachment, opened). It is used here only as a source of questions. Its astrophysical figures (for example the share of second-generation black holes and the age of a galactic merger) are not sourced by this site and are not cited as findings. Its quark-logic section (§8) and its kanji remark stay tabled. Three of its claims are turned into questions about the Navajo verb word: early/low sites “tend to be restrictive: they fix core properties” while late/high sites “tend to be selective: they organize or adjudicate already-fixed material without fully rewriting it” (§4); “combination events leave residual properties that later events can read” (§2); and “Phases function analogously to causal horizons”, since “The Phase Impenetrability Condition freezes the complement while leaving the edge accessible” (§7). The handout itself names Navajo: “classificatory stem lexicalizes Proto-Patient properties early; conjunct slots handle inverse, agreement, and aspect” (§3). That is Grant's claim, tested below. A chemistry counterpart of each question is on the Chemical·Theta Roles scout page.

11.1 Early-restrictive vs late-selective sites

Layer (inside out)What the sources show it doingRestrict, select, or add?Grade
Stem / theme (the classificatory stem; the role schema Σ)Fixes properties of the patient slot that limit which noun can link: Willie's (22) yiyííbį́į́ʼ needs “round objects” and (24) yiyííłtaʼ “plural objects” (pp. 61–62; thesis D9). The roles in the action are fixed here. Under ‘start to’, “the role in the action is unchanged by INCH” (HD2)RestrictsSOURCED for linking restrictions (Willie); “Proto-Patient properties lexicalized early” as a general claim is GRANT'S CONJECTURE, supported here only by these two glosses
Event layer (v, the ł-causer, the inceptive qualifier, INCH, DETR)Adds participants or slots: the ł-causer adds a causer (causative split); ‘start to’ adds a thematic ’ and, when nothing low exists, a role-less placeholder (MA7); it also turns an unaccusative's sole participant into an object (HMP (11); MA5)Adds and re-maps (not merely selects)HYPOTHESIS; the facts are SOURCED (HMP; Hale & Platero 1996 via Gessner)
Agreement and voice (M/S, P, the yi-/bi-/Ø choice, da-)Picks which already-built slot is pronounced on the head (PORT) and which is marked as the topic (IA-3); da- counts the agent in both voices (Willie pp. 82–85), so voice does not change rolesSelectsHYPOTHESIS (SOURCED facts)
Clause (D2: nouns, topic/focus)Nouns link to existing slots by feature match and position; they cannot change the word's choices (WF-5; MA3 step 6)SelectsHYPOTHESIS (SOURCED linking rules: Willie p. 53; H&P 1974 (28))

HYPOTHESIS Result. The handout's asymmetry holds at the two ends and fails in the middle. The innermost layer restricts (Willie's sortal glosses) and the outer layers select (voice, agreement, noun linking). But the event layer does not just select among fixed material: it adds a causer, adds a placeholder slot, and re-maps an unaccusative participant to object. So the Navajo word is “restrict, then add, then select”, not “restrict, then select”.

HYPOTHESIS Against Grant's STRATA conjecture (theta roles, then event roles, then lexical selection): the order matches the three outcomes exactly. The theta-role layer restricts. The event-role layer is the one that adds, which is what an event-structure layer is expected to do (it introduces causers and phases of the event). Lexical selection (nouns, D2) is the last and purely selective layer. On this reading STRATA is the better fit and the handout's two-way split is a simplification of it. Nothing opened contradicts either; the one thing that would is a later (D1-top or D2) operation that changes a role in the action. None is known on the site.

11.2 Residual memory read by later operations

ResidualWhere it is leftWhat later reads itGrade
The deictic δ of a fourth-person / indefinite / areal subjectAt the top edge of the word, adjoined to Qual or DETR (SH-4; MA2)Inside D1, the object exponent reads it (Ø, not yi-, in jishxash: MA3). Across clauses, the fourth person is what introduces a new agent in a =go chain, and only from the first clause (Willie Ch. 5 §5.1, pp. 132–140, on 4th across clauses). So a later clause reads itSOURCED facts; “left at the edge, read later” is HYPOTHESIS
Voice (yi- vs bi-)In the word's exponents (the finished D1)Noun linking at D2: “the NP immediately preceding” a yi- verb is its object and before a bi- verb its subject (Hale & Platero 1974 (28), p. 17); W&J: the verb complex “determines the interpretation of the adjoined nominals” (p. 270)SOURCED
The inverse on a postposition (bitah)On the PP, inside D1D2 links dibé to the bi- object ((50b); W&J: coindexed but “not a constituent of the PP”)SOURCED (W&J (50b))
Whether ρX moved inside D1 (IA-1* vs IA-3), or the low participant raised (U-1 vs U-2)Interior of D1Nothing later reads it: both options give the same strings, and no D2 rule is sensitive to the differenceHYPOTHESIS

HYPOTHESIS Fit with the PIC. What later operations read is always something pronounced on the word's exponents or at its edge (the deictic on the top head, the voice prefix, P's pronoun). What they never read is how the interior was derived. That is the PIC pattern (“freezes the complement while leaving the edge accessible”, in the handout's words) applied to D1. One caveat: the voice prefix is not at the top edge of the word; it is an object exponent inside it. So the generalization that fits is “later operations read D1's output, not D1's derivation”, which is weaker than an edge-only PIC.

11.3 Phases as horizons

HYPOTHESIS D1 acts as a horizon in the handout's sense for derivational history (11.2, last row), not for content: its exponents stay visible to D2. Hale's caution applies: where the word stops is unsolved (2003 n. 12), so the position of the horizon is stipulated. The PP inside D1 is a horizon for M/S's probe (IN-3) but not for noun linking at D2, which reaches P's object (50b). OPEN whether any Navajo operation is blocked by D1 in the way the PIC blocks Agree into a complement: T40g.

Changes made elsewhere

Links added from clause spine §11 and open questions (row T40a–j), and from the home page. Sources for this page: sources, added 2026-10-19 (1:52pm).

10. Tasks T40a–j

TaskWhat it needs
T40aRudin's indivisibility / clitic diagnostic: forms with two or more disjunct prefixes and the polar interrogative enclitic (Hale 2003 n. 12, Platero p.c.; legible page or Young & Morgan)
T40bAnswered 2026-10-19, 8:46pm on the originals page. Open Rudin 1988, Hiraiwa 2001, Chomsky 2008 and Grewendorf 2001 and check the definitions used here against them
T40cConditionally answered 8:46pm (selection condition; → T41a). MA-3 with DETR: is the phase head DETR when di- is present (Kibrik (8a) ʼá-zh-dí-l-zhééh)? Does anything else treat DETR as the top of the word?
T40dNarrowed 8:46pm: Hiraiwa's Multiple Agree is excluded (MA3′). Multiple Agree vs contextual allomorphy: a form where the object is not third person and its exponent still varies with the subject's δ; or a single prefix spelling both subject and object person
T40eU-1 vs U-2 (= T37b): a movement diagnostic for a low participant inside the word
T40fPP edge: is there any Navajo form where a postpositional object behaves as if at the PP edge (Hale 1997's Spec,PP theme) under the inverse?
T40gPartly answered 8:46pm: in Chomsky 2008, v*P is a phase and unaccusative vP is not (MA9′). Is vP a phase inside the word? Needs a probe on M/S that is blocked by vP, or one that is not
T40hGrant's notes §4: valuation of the classificatory stem's “Proto-Patient” features by prefixes. Is there a Navajo source for agreement between a classificatory stem and a prefix?
T40iA third test of MA-4: any template position that holds two exponents at once (a counterexample would make that position [+MFS]-like)
T40jPerson Case Constraint, listed in Grant's note as a related family member: do Navajo combinations of a local-person object with a deictic subject show gaps? Not checked

The borrowed terms checked against their originals: what Rudin 1988, Hiraiwa 2001 and Chomsky 2008 actually define, and what that changes inside the Navajo verb word (T40b)

Working note. Kept as written (URL, data and status labels unchanged). Filed under Merge and Agree inside the word; the consolidated argument is the essay merge-agree. Rule IDs: rules index.

Verbs on this page → dictionary: yishxash ‘bite’ · dabiiłtsą́ ‘see’ · yíghááh ‘join; with -tah ‘among’: go among’

Restated on the unified structure (2026-11-15). The framework is restated in §1: VPISH (arguments merged in the verb’s scope, subject in Spec,vP) with theta roles, and V→v→T with the subject raised to Spec,TP as the working default in tensed clauses; each pattern is what else moves, with an LF whose scope matches c-command. Light d, used below, is withdrawn (Grant, 2026-11-14); light n is an OPEN structural question. The head-direction mismatch argued below as rule sets A/B or a changeover is now stated as Grant’s Merge-site hypothesis and tested in §3.1–3.2. Derivations below in which the verb or subject does not raise are labelled contrast; the working default raises V→v→T and the subject to Spec,TP. Status on the spine: see unified structure, §1. The data, judgments, translations and citations below are unchanged; the analysis below is kept as the record of what was argued then.

2026-11-11 audit pass 1. The source definitions of Merge/Agree/phase are brief contrasts, not axioms imposed on Navajo. The MA/EF rules, d identity, phase heads and placeholder attachment are hypotheses. Copying a definition does not establish a Navajo head or movement path. The primary vP→TP subject derivation and V→v→T head chain, including all LF/locality checks, remain to be compared explicitly with PF assembly in pass 2.

2026-11-11 audit pass 3. Status badges now use only the label set SOURCED, HYPOTHESIS, SPECULATIVE, OPEN, NONE and GRANT'S CONJECTURE; verdict or task words that had been styled as badges (for example Opened, PRE-VPISH, DONE, SUPPORTS, PARTLY ANSWERED) are now plain text or a label with a qualifier. Local box file paths were removed from this page; files inside the repository are given by repo-relative path, and the source citation next to each removed path stays as the reference. See the pass 3 report.

2026-11-13 audit pass 7. An earlier dated note on this page treats V→v→T and subject raising as Grant’s conjecture or a comparator to test against PF/template rivals. Since 2026-11-13 they are the working default in tensed clauses: VPISH frames where the arguments start so that syntax and semantics line up; it is not a word-order account. PF/template accounts are kept only as labelled contrast, and the open question is what else has to move. The earlier note is kept as written. See the 2026-11-13 note and the pass 7 report.

The page Merge and Agree inside the Navajo verb word sorted the site's word-internal findings using five mid-level terms from Grant's typology note: Rudin's [±MFS], Hiraiwa's Multiple Agree, Chomsky's feature inheritance, the Edge Feature, and phases. Its originals had not been opened. This page opens them, compares each definition with the wording used, tightens the formal rules MA-0 to MA-6, and re-runs the nine derivations, plus one more. The two operations stay apart throughout. Merge builds structure: it puts an element at a position, either a fresh element (External Merge, EM) or a copy of one already in the structure (Internal Merge, IM, which is movement). Agree values features such as person and number: a probe on a head finds the closest matching goal and copies its value. Nothing moves under Agree. Written 2026-10-19, 8:46pm PT.

Update 2026-10-20, 4:46am (T41e). Chomsky 2000 (“Minimalist Inquiries”) and 2001 (“Derivation by Phase”) have now been opened: the older definitions behind “edge feature”. Four changes to this page. (1) MA-2′(c) is revised. Both texts allow an expletive only in the extra position of a T-like, non-phase head; a phase head may fill its extra position only by raising (MI (5a), p. 102; p. 109). So the ‘start to’ placeholder cannot satisfy the Qual-layer phase head's own feature, as written below. It belongs to PINCH, the postposition-like piece at the top of the verb phrase (new EP-2; this was IN-8's original location). The bump is Move to that same position. (2) MA1′'s reason (the phase wall) is MI's version of the PIC; under DbP's version, M/S can see the object and the closer subject wins (EP-6). The strings are the same. (3) MA-5b gets its original mechanism (defective intervention and default [3person], MI pp. 123–128), but it needs an activating feature on Navajo slots, which MI/DbP tie to structural Case (EP-5). (4) MA-3a holds only if the deictic δ is not part of the person/number set, since MI p. 124 values that set “as a unit” (EP-7). Grant's EPP/Edge Feature wording is confirmed as the 2000/2001 notion (MI p. 109; DbP p. 41). The rules and derivations below are kept as written.

Update 2026-10-20, 12:46pm (T41f). Hale's n. 12 forms are now read (where the yes/no enclitic -ísh can go): ka naash[á] ‘I am sick’ and ka-[á]sh nanin[á] ‘are you sick?’. Hale gives ka as a preverb that can “appear as separate words”, so the enclitic split is between two words, not inside one. The bottom-line “[−MFS]-like” verdict below and the §6 clitic row therefore go back to OPEN. The enclitic does meet Rudin's condition (it attaches to the first constituent: NLA yes/no prose, (4), (6), (10a), (12)), but by EN-1′ it can never reach inside the verb word, so it cannot test the prefix zone (EN-6). The text below is kept as written.

Ground rules (Grant's). No new Navajo form is introduced. Every form below is already on this site, from a named source with its example number, and the page links to where it was read. None is segmented beyond what its source prints. The three originals are general-syntax works. None of them mentions Navajo. Every mapping of a Navajo fact onto their terms is therefore ours, graded HYPOTHESIS at most. What the originals say is graded SOURCED, with the page number. Only contradicting data count against an analysis; a source's silence does not. No earlier analysis (Hale, Willie, HMP, Kibrik, or this site's own rules) is assumed to be fully right.

Notation (the site's own, as on the Merge/Agree page). Σ = role-schema step (which slots the verb has and their roles in the action); A = word level (cycle D1, rule set A: the prefixal, left-headed functional spine); B = clause level (cycle D2, rule set B: right-headed clause syntax and noun linking). Spine (Hale 2003): Qual > M/S > v > V. Slots ρ = ⟨E, pos, φ, (δ)⟩ (SH-0), where φ is person/number and δ the deictic value (fourth person, indefinite, areal). CYC = the site's spell-out ordering (§3). VPISH: the subject slot is first-merged in Spec,v. New abbreviations from the originals: EF = edge feature (Chomsky 2008), v* = the v of a verb with full argument structure (transitive or experiencer), PIC = phase-impenetrability condition, DIC = defective intervention constraint (Hiraiwa 2001).

Bottom line, in plain words

HYPOTHESIS The Merge/Agree page's main picture survives the originals, but it is now smaller and sharper. Most of what happens inside the word is still Agree (copying person and number), and the places where something is really built or moved are still at edges. One of its three “Merge sites” has to be downgraded to OPEN. That is the fourth-person, indefinite or areal piece of a subject, which sits at the top of the word. Chomsky's own version of feature inheritance allows two readings of it, and they give the same strings. On one reading it is agreement morphology on the top head (Agree). On the other, the top head's edge feature moves it there (Merge).

HYPOTHESIS The ‘start to’ placeholder is the best match found anywhere on the site between a Navajo fact and a mid-level term. Chomsky (2008, p. 157) proposes that when an edge feature cannot be satisfied by moving anything, it “must be satisfied by EM, necessarily of an expletive since no argument role can be assigned”. In the same passage he adds that when no goal is accessible, the head shows “default morphology”. That is exactly what Hale, Munro & Platero describe. The role-less ’ appears with ‘cry’ and ‘snow’ because nothing else can fill the slot. In ‘I start to get cold’ the Mode/Subject head shows a default third person because its only candidate is inside the postposition-like piece.

HYPOTHESIS The “Multiple Agree” candidate is not Hiraiwa's Multiple Agree. Hiraiwa's operation is one probe valuing one kind of feature on several goals at once. The probe's own morphology then shows only the closer goal (p. 79 n. 12: “a closer goal feature determines agreement”; no language shows “multiple agreement on the side of probe/head”). The Navajo object prefix (yi- vs zero) reads two different things from two different slots: the object's person and the subject's deictic value. So either it is a context-sensitive spelling (the site's SH-6, now favoured), or it is a counterexample to Hiraiwa's generalization. It cannot be his operation as he defines it.

HYPOTHESIS [−MFS]-like still holds, but for a better reason. In Rudin's framework no ordinary position can take a second occupant, so “a full position refuses a second one” proves nothing by itself. [+MFS] is about extra elements adjoining to a filled specifier (pp. 480–481). Rudin's decisive diagnostic for the [−MFS] type is a clitic placed right after the first fronted element (pp. 461, 463, 466). Hale reports exactly that for the one disjunct case on the site: the polar enclitic can split the preverb ka off from the rest of the verb (2003 n. 12). The fixed order of disjunct prefixes does not count either way. Rudin's fixed order is subject before non-subject (p. 475), not an order by prefix class.

HYPOTHESIS Phases: Chomsky's own list gives one result for free. His phases are CP and v*P, and an unaccusative's v is not a phase (p. 143). Mode/Subject's probe therefore sees only the subject at the edge of a transitive vP, never the object. In a one-participant verb whose participant starts low, it reaches the participant directly, with no movement. That is the site's “closest slot” option U-2, and Chomsky lists exactly this case (“passive/unaccusative in situ”, p. 155). The movement option U-1, as the site wrote it, has no trigger in his system. A postposition phrase is not on Chomsky's list, so the site's “PP is a phase” is our own extension. The same texts offer a second explanation: the postposition's own agreement uses up its object, which then cannot agree with anything higher.

GRANT'S CONJECTURE “Navajo has word-internal Merge sites, or possibly Agree sites” is Grant's. After this check, the word-internal Agree sites are firmer than the Merge sites. Two Merge sites remain certain: the bumped object and the placeholder. The third is OPEN. Nothing here contradicts the conjecture.

On this page: 1 Sources · 2 Definitions side by side · 3 Rules MA-0′ to MA-6′ · 4 What changed, rule by rule · 5 Derivations re-run (MA1′–MA10′) · 6 Rudin's diagnostics applied · 7 Gets right / misses · 8 Grant's frame · 9 Tasks T41a–h

1. Sources

SourceAccess this runUsed for
Rudin, Catherine. 1988. On multiple questions and multiple WH fronting. Natural Language and Linguistic Theory 6: 445–501Opened: full PDF (57 pp., scanned journal pages with a text layer) from a University of Chicago course-readings URL, read as extracted text; page numbers from the printed headers (PDF p. n = journal p. 444 + n)[±MFS] definition (70), structures (71), diagnostics table (69), clitic predictions (34), order facts, substitution vs adjunction in SpecCP (pp. 480–482)
Hiraiwa, Ken. 2001. Multiple Agree and the Defective Intervention Constraint in Japanese. MITWPL 40 (HUMIT 2000): 67–80Opened: full PDF (14 pp., born-digital) from web.mit.edu (a 24.956 course folder), extracted textAgree and Match (1), (2); DIC (3), (9); Multiple Agree (7), (8); Icelandic default agreement (31)–(33); n. 12 on probe-side morphology
Chomsky, Noam. 2008. On phases. In R. Freidin, C. P. Otero & M. L. Zubizarreta (eds.), Foundational Issues in Linguistic Theory, 133–166. MIT PressOpened: the chapter PDF (MIT Press eBook portion, hosted on a Caltech personal page), extracted text with printed page numbers. Only the chapter's own pages 133–166 and notes are used. The PDF also contains the start of the next chapter, which is not usedEF (p. 139); EM vs IM (p. 140); phases CP, v*P and the claim that only phase heads drive IM (p. 143); inheritance (pp. 143–144, 148–149); A′-position as EF-attracted (p. 150); EF is indiscriminate (p. 151); goal spelled out in situ or at the probe (p. 155); v vs v* (p. 156); EF inheritance, default morphology and expletive EM (p. 157)
Grewendorf, Günther. 2001. Multiple wh-fronting. Linguistic Inquiry 32: 87–122SOURCED (abstract only) Publisher abstract (MIT Press / DOI page) read through search results. The full text is behind a paywall and was not openedOnly the abstract's statement of the cluster mechanism
Grant, “Typology of Merge/Agree Phenomena” (2 pp.)Re-opened (text, [local working copy, not in the repository])The definitions checked in §2
This site's pagesRe-readMerge/Agree page (rules, MA1–MA9); subject on the head (SH-2–SH-6, SD7, SD8, Willie (31f)); HMP ms. (IN-0, IN-5′, IN-7 (D), IN-8, IN-9); inverse as Agree/EPP (IA-1*, IA-3); A/B as spell-out (Hale 2003 n. 12)
Not opened—Chomsky 2000, 2001, 2004 (cited in the originals); Bošković; Richards 1997; Miyagawa 2005 (known here only through Chomsky's summary on p. 156); Young & Morgan. None is used as evidence

2. Definitions side by side

Left: the wording in Grant's note (and on the Merge/Agree page). Middle: what the original says, with page. Right: the consequence.

TermGrant's note / the Merge/Agree pageThe original SOURCEDVerdict and consequence
[±MFS]“Multiply Filled Specifier”; “in [+MFS] languages an already-moved wh … can serve as a further landing site, so later whs Merge to it and form a cluster. In [–MFS] languages no such intermediate host is available.” Page: “a position that is already filled can host more”“MULTIPLY-FILLED SPECCP” (p. 448). (70): “SpecCP can contain multiple Wh-words at S-structure in Bulgarian and Romanian ([+MFS] languages), but only one in Serbo-Croatian, Polish, and Czech ([-MFS] languages)” (p. 478). (71b): [−MFS] still fronts every wh. The others sit “in or adjoined to IP” (pp. 447, 471). Mechanism (p. 480): “A. A single Wh-phrase is substituted into the SpecCP position … B. An additional Wh-phrase or phrases may be adjoined to SpecCP. C. The adjunction in (B) is to the right.” A flat multi-slot SpecCP is rejected because it “requires substitution into a node that already contains one or more Wh-phrases” (p. 481)Grant's note is right about the two types. Two refinements. (i) In Rudin, [−MFS] does not mean “no host”. The extra elements go to a different position (IP). (ii) In her framework a substitution position never takes two occupants. Only adjunction to a filled specifier can give [+MFS]. So the Navajo test must be about adjunction (rule MA-4′)
[+MFS] diagnostics“Superiority, no intervention, clitics after whole group, multiple extraction”; [−MFS]: “Freer order, intervention possible, clitics after first wh”Table (69), p. 478: multiple wh-extraction from a clause; wh-island violations; “clitics follow first Wh-word” ([−MFS] +); “parentheticals, adverbs, particles after first WH” ([−MFS] +); “free NOM/ACC Wh-word order” ([−MFS] +). Clitics: “In the [+MFS] languages, clitics are placed after the whole series” (p. 461). The prediction holds “in the case where clitic placement is by encliticization to the first constituent of the clause” ((34), p. 463), and the clitic evidence is “inconclusive for the [+MFS] languages” because their clitics are pre-verbal (p. 466). Order: “restricted at least to subject > nonsubject” in [+MFS] (p. 475), derived from the leftmost wh being the “head of SpecCP” (p. 482)Grant's list matches (69), except that it leaves out wh-islands. Two cautions apply. The clitic test needs an enclitic that attaches to the first constituent. And the fixed order Rudin means is an order by grammatical function. Both cautions matter for Navajo (§6)
“Superiority is Attract Closest”Grant's note: “Superiority is Attract Closest on the initial Agree step with C”Rudin derives the order from Comp-indexing by the head of SpecCP plus a split ECP and an LF binding condition ((65)–(68), pp. 477, 482–483), not from Attract ClosestA later restatement, not Rudin's. It is not used below
Wh-clustering (Grewendorf)“Whether a moved wh can itself host further whs”; “feature-based implementation of the [+MFS] cluster effect”Abstract only: multiple fronting “consists of moving a single wh-cluster to [Spec, CP]”; “wh-elements can act as landing sites for wh-movement due to morphological properties of wh-words”The Merge/Agree wording in Grant's note (“an already-moved wh … can serve as a further landing site”) is closer to Grewendorf's mechanism than to Rudin's. In Rudin, later whs adjoin to SpecCP and no cluster moves. Both are kept, under separate names
Multiple Agree“Whether one Probe can Agree with multiple Goals in a single step.” Page MA-6: one probe Agrees with ρA (δ, φ) and ρX (φ)(7), pp. 69–70: “MULTIPLE AGREE … with a single probe is a single simultaneous syntactic operation; AGREE applies to all the matched goals at the same derivational point.” The probe is “[+multiple]” and keeps matching “until the probe locates all the matching goals within an ‘accessible’ domain” (p. 70). Match: “Matching is feature identity”; “Locality reduces to ‘closest c-command’” ((2), p. 68). n. 12 (p. 79): “There exists no language that shows … multiple agreement on the side of probe/head”; “It is a universal fact that a closer goal feature determines agreement.” Icelandic: with a closer dative goal the agreement on T is “default 3rd person singular” (p. 79; (31)–(33), p. 78)Grant's one-line definition is right. The page's MA-6 candidate is not Hiraiwa's operation. It mixes two features (δ and φ), and its result shows on the probe side for both goals, which n. 12 says no language does. MA-6′ renames it
Feature inheritance“Whether φ-features on C are inherited by T.” Page MA-3: Qual “passes [uphi] to … M/S and keeps [udelta]”“Agree and Tense features are inherited from C, the phase head” (pp. 143–144). It goes to “the LI selected by C (namely T)” (p. 144). “C has two probes: the edge feature EF … and an Agree-feature (φ-features)” (p. 148); v* passes its Agree feature to V (p. 148). Obligatory or optional: “unsettled” (p. 149). When φ shows at T without tense, it is “just a morphological effect of agreement” (p. 144). EF may be inherited too (p. 157, tentative)Chomsky passes the Agree feature as a whole. What the phase head keeps is EF, which drives Merge, not an Agree probe. MA-3 as written (keep [uδ]) is an extension, not his model. MA-3′ states both versions. Inheritance goes only to the head the phase head selects, which gives a new prediction (T41a)
Edge Feature / EPP“Whether a given head (T, v, C…) has an Edge Feature triggering Spec Merge”Every lexical item has EF: “an LI has a feature that permits it to be merged. Call this the edge feature” (p. 139). “IM should be driven only by phase heads” (p. 143). Raising to Spec-T is “derivative, via inheritance” (p. 144). EF is “indiscriminate: it can seek any goal in its domain” (p. 151). A position reached by EF is an A′-position (p. 150). EPP restated: “If nothing is raised, then the inherited edge feature of T must be satisfied by EM, necessarily of an expletive since no argument role can be assigned”, and with “no accessible NOM, then T will have default morphology” (p. 157)In Chomsky 2008 what varies is not whether a head has EF, since every item has one. What varies is whether IM to it is possible (phase heads, or inherited). The wording in Grant's note matches the older EPP-feature idea (Chomsky 2000/2001, not opened). The p. 157 passage is a direct model for the ‘start to’ placeholder and default (MA7′, MA5′)
Phasehood“Which heads count as phases (barriers to Agree and movement)”. Page: D1 and the PP are phase-likePhases are “CP and v*P”; v* is “the functional head associated with full argument structure, transitive and experiencer constructions”, one of several choices for v (p. 143). TP is not a phase; DP might be, “put aside” (p. 143). “Probe into an earlier phase will almost always be blocked by intervention effects” and “It may be … that PIC holds only for the mappings to the interface” (p. 143). “v typically permits both long-distance agreement and raising, v* does not” (p. 156). “In a probe-goal relation, the goal can be spelled out only in situ (under long-distance Agree) or at the probe (under internal Merge)”, with “passive/unaccusative in situ or with movement” as the standard illustration (p. 155)Grant's wording is right. PP is not on Chomsky's list, so MA-5 is our extension, and MA-5′ adds a non-phase alternative. The v/v* split settles part of T40g (MA1′, MA9′)

3. Rules MA-0′ to MA-6′ (ours; all HYPOTHESIS unless marked)

MA-0' SEPARATE   Merge(X, Y) = {X, Y}. EM if Y is not part of X; IM (movement, a copy) if
                 Y is part of X (Chomsky 2008 p. 140). Agree(P, G): P's unvalued features
                 are valued by the closest matching G that P c-commands (Match = feature
                 identity, locality = closest c-command: Hiraiwa (1), (2)).
                 CLASSIFYING AN EXPONENT e AT HEAD H:
                  (i)  Agree   if e spells only features that H probes for (agreement
                               morphology: "a morphological effect of agreement", p. 144);
                  (ii) IM      if e spells features of a goal that H does not probe for,
                               or the goal's own category (the goal "spelled out ... at the
                               probe (under internal Merge)", p. 155);
                  (iii) EM     if e spells an element with no source in the structure
                               (a placeholder / expletive, p. 157).
                 Replaces MA-0's test "pronounced away from where its role puts it",
                 which would wrongly class PORT (subject phi on M/S) as Merge.
MA-1' AGREE-SITES Probes in D1: M/S (derivatively, by inheritance from Qual: MA-3'), P (its
                 own pronoun), and, only if a voice head exists, a voice probe (else the
                 yi-/bi-/0 choice is a spelling rule on rho_X: MA-6'). M/S's probe stops at
                 the edge of a phasal v*P, so in a transitive it reaches only Spec,v*.
                 In an unaccusative v is not phasal (p. 143), so M/S reaches the low slot
                 directly (= U-2, IN-4 PORT-CLOSEST = Hiraiwa's (2c) closest c-command).
                 IM in D1 is driven only by a phase head's EF (p. 143): Qual's EF, or v*'s
                 EF (IA-1*, which then lands in an A'-position, outer Spec,v*: p. 150).
MA-2' EDGE-SITES (a) The bumped PP (IN-5'): Merge, either EM of the PP high (IN-5' as
                 written) or IM by the EF of the Qual-layer INCH (IN-0: n, ii are Qual-layer
                 heads). Merge either way.
                 (b) The leftover delta (SH-4): Agree (MA-3a) or IM (MA-3b). OPEN.
                 (c) The placeholder rho_D (IN-9): EM of an expletive that satisfies the EF
                 of the Qual-layer INCH when nothing below v can (p. 157). Since Qual is
                 D1's phase head (MA-3'), this EF is the phase head's own: no inheritance.
MA-3' INHERIT    Qual = the phase head of D1. It passes its Agree feature [uphi] to the
                 head it SELECTS, M/S (p. 144: "the LI selected by C (namely T)").
                 MA-3a PARTIAL-AGREE: Qual keeps a [udelta] Agree probe; the delta exponent
                       is agreement morphology on Qual (Agree). Not in Chomsky (he passes
                       the Agree feature whole): an extension.
                 MA-3b EF-RAISE: Qual keeps EF (p. 148); EF raises the delta-bearing part of
                       rho_A from Spec,v* (the edge of v*P, accessible) to Qual (IM). The
                       landing site is an A'-position (p. 150). Needs delta to be a separate
                       syntactic object (SH-4 already pronounces it "as its own element").
                 Both: the phi-passing head must be the head that selects M/S. If DETR and a
                 separate Qual are both present (DETR > Qual > M/S), the phase head
                 that passes phi is Qual, not DETR (T41a).
MA-4' NO-ADJOIN  No word-internal position takes a second element ADJOINED to its occupant
                 (Rudin's [+MFS] needs adjunction to a filled specifier, pp. 480-481). A
                 second candidate goes to a different position (as Rudin's [-MFS] extras
                 go to IP, (71b)), here a HIGHER one (Position I, IN-5'), or fuses
                 (' + ' -> ', HMP n. 4). The one zone with several adjoined elements (the
                 disjuncts) shows Rudin's [-MFS] clitic pattern (Hale 2003 n. 12; §6).
MA-5' PP-OPAQUE  M/S's probe never values from a slot inside a PP (IN-3). Two sources:
                 MA-5a PHASE: PP is a phase (our extension; not on Chomsky's list, p. 143).
                 MA-5b INACTIVE: P has already Agreed with its object (P's own pronoun),
                       so the object is inactive (the inactivity condition as
                       restated on p. 150; Hiraiwa (3)); a closer inactive
                       goal also blocks lower goals (Hiraiwa (9)) and the probe shows
                       default morphology (Hiraiwa pp. 78-79; Chomsky p. 157).
                 They differ only if an active goal lies BELOW the PP (5a: reachable;
                 5b: blocked, default). No such form on the site (T41c).
MA-6' TWO-GOAL   (Narrowed.) Hiraiwa's Multiple Agree is excluded for the object
                 exponent: it needs one matched feature (Match = identity) and shows only
                 the closer goal on the probe (n. 12). The exponent reads rho_X's phi AND
                 rho_A's delta. So it is either (a) SH-6, a spelling rule on rho_X that
                 reads rho_A in D1 (favoured: under SH-4 rho_X is pronounced on itself, so
                 no probe is needed), or (b) a non-Hiraiwa two-goal mechanism (none opened),
                 which would also be a counterexample to n. 12.

Update 2026-10-20, 4:46am. MA-2′(c) is superseded by EP-2, which puts the placeholder at PINCH, a non-phase head; MA-2′(a)'s bump is Move to the same position. MA-1′ now has two versions (PIC1/PIC2, EP-6). MA-5b = EP-5, with a cost. MA-3a requires δ ∉ φ (EP-7). MA-3′'s selection condition, MA-4′ and MA-6′ are unchanged; MA-6′ also gains MI p. 124 as a source.

4. What changed, rule by rule

RuleBefore (1:52pm)AfterWhy
MA-0Merge = pronounced away from its role position, or role-lessMA-0′: three-way test on exponents (Agree / IM / EM)The old test would class PORT as Merge (Chomsky p. 144, p. 155)
MA-1Probes: M/S, P, the voice positionMA-1′: M/S derivative; the v*P edge limits it; unaccusative v not phasal; IM only by phase-head EFChomsky pp. 143, 148, 150, 156
MA-2Three Merge sites (a, b, c)MA-2′: (a), (c) Merge; (b) OPENMA-3′ has two Chomsky-compatible readings with the same strings
MA-3Qual keeps [uδ]MA-3a (extension) vs MA-3b (Chomsky's EF), plus a selection conditionChomsky passes the Agree feature whole and keeps EF (pp. 144, 148)
MA-4A position holds one exponent (two positions tested)MA-4′: no position takes an adjoined second element; the disjunct zone shows the [−MFS] clitic patternRudin pp. 461–466, 480–481
MA-5PP is a phaseMA-5a (phase, our extension) vs MA-5b (inactive goal)PP is not in Chomsky's list (p. 143); activity (p. 150; Hiraiwa (3), (9))
MA-6Multiple Agree vs contextual allomorphy, OPENMA-6′: Hiraiwa's Multiple Agree excluded; SH-6 favoured; OPEN only against a non-Hiraiwa mechanismHiraiwa (2), (7), n. 12

5. Derivations re-run (MA1′–MA10′)

Each derivation is the Merge/Agree page's derivation with the revised rules. Steps are labelled by rule set: Σ role schema, A word (D1), B clause (D2). Each step names its operation: EM, IM, Agree or Spell-out.

MA1′. Hale 2001 (24a) shidoołhosh ‘He/She will tickle me’ (as MA1, SD1)

  1. Σ Σ(√TICKLE) = {ρA, ρX}: ρA has Proto-Agent entailments in the action (causation, sentience), ρX Proto-Patient ones; φ(ρA) = 3, no δ; φ(ρX) = 1sg.
  2. A EM: V ghosh + ρX (Comp,V); v* ł (transitive, so a phase head, p. 143); ρA in Spec,v* (VPISH; EM to v*'s specifier satisfies v*'s EF, p. 144, n. 32).
  3. A EM: M/S; Qual d. Inheritance (MA-3′): Qual passes [uφ] to M/S, the head it selects.
  4. A Agree (MA-1′): M/S probes. The v*P complement (VP) is behind the phase edge, so ρX is not a candidate. ρA at the edge is the closest goal, so φ = 3 → w. No IM.
  5. A Spell-out D1: sh ⌢ d ⌢ w ⌢ ł ⌢ ghosh ✓ (Hale's segment line).
  6. Change from MA1: the reason M/S never takes the object's φ is now Chomsky's v*P phase, not a stipulation. Same string.

MA2′. Hale 2001 (22b) shiʼdooʼniid ‘I was told (lit. someone told me)’: two readings of the split subject (as MA2, SD4)

  1. Σ ρA = ⟨φ = 3 (zero), δ = INDF⟩; ρX = 1sg.
  2. A EM: V niid + ρX; v d; ρA in Spec,v; M/S (w); Qual d.
  3. A Inheritance: Qual → M/S [uφ]. Agree M/S ⟶ ρA: φ = 3 (zero).
  4. A Reading MA-3a: Qual's kept [uδ] probe Agrees with ρA (closest goal): δ = INDF, spelled ’ on Qual as agreement morphology. Reading MA-3b: Qual's EF IMs the δ-part of ρA from Spec,v (an edge, so accessible) to Qual, an A′-position.
  5. A Spell-out D1 (both readings): sh ⌢ ’ ⌢ d ⌢ w ⌢ d ⌢ niid ✓.
  6. Classification: Agree (3a) or IM (3b); OPEN. MA2 called this “a word-internal Merge site (MA-2b)”. That is now one of two readings.

MA3′. Willie (23b) jishxash ‘That person bit it’ vs p. 129 ʼashkii yishxash ‘The boy bit it’: Hiraiwa's test (as MA3, SD6)

  1. Σ {ρA, ρX}; φ(ρX) = 3. jishxash: ρA = ⟨3 (zero), δ = 4⟩; ʼashkii yishxash: ρA = ⟨3, no δ⟩.
  2. A EM, inheritance, Agree M/S ⟶ ρA as MA1′; for jishxash, MA2′ step 4 (δ = 4 at Qual, either reading).
  3. A Hiraiwa's Multiple Agree tried: a [+multiple] probe would match goals bearing one feature type ((2a) “feature identity”). The exponent depends on ρX's φ (only a third-person object alternates) and on ρA's δ (zero under δ = 4). Those are two different features. And on Hiraiwa's own account the probe's morphology should show only the closer goal (n. 12), which here is ρA in Spec,v. ✗ Not his operation.
  4. A MA-6′ (a), SH-6: ρX is pronounced on itself (a non-head, SH-4), and its third-person exponent is y- if a δ-less third person c-commands it in D1, zero if that slot has δ = 4. No probe is involved.
  5. A Spell-out D1: j ⌢ … ⌢ shxash; yi ⌢ … ⌢ shxash ✓ (not segmented further).
  6. B ʼashkii links to ρA in D2 and cannot change the D1 choice (WF-5).
  7. Result: MA-6 narrows. The yi-/zero choice is a spelling fact unless some mechanism other than Hiraiwa's is found. T40d's test (an object that is not third person whose exponent still varies with the subject's δ) now separates only SH-6 from that unnamed mechanism.

MA4′. W&J (50b) dibé łį́į́ʼ bitah yíghááh: P's own Agree, and IA-1* as v*'s edge feature (as MA4)

  1. Σ ρA (the one who goes), ρX = P's object; both 3; inverse: ρX [top], ρA [foc].
  2. A EM P -tah + ρX. Agree inside PP: P values its own pronoun from ρX. Under IA-3 ρX carries [top], so the exponent is b- → bitah. ρX is now inactive (MA-5b).
  3. A EM PP inside VP; v; ρA in Spec,v; M/S; Qual. Agree M/S ⟶ ρA (ρX is unavailable on either MA-5a or MA-5b).
  4. A IA-1* branch, restated: EF of a phase head is “indiscriminate” and lands its goal in an A′-position (pp. 150–151). If the topic slot moves, the mover is v*'s EF (if ‘go’ has a v*; Chomsky's v* is for transitive and experiencer verbs, p. 143, so a one-participant motion verb may have no phasal v at all) or Qual's EF. Chomsky adds that a raised phrase “is identified as a topic by the final position it reaches” (p. 151). That fits IA-1*'s topic slot. Inactivity (step 2) does not stop EF, because EF is not Agree. So IA-1* no longer needs the PP-edge step (i) of MA4.
  5. A Spell-out D1: bitah ⌢ yíghááh.
  6. B łį́į́ʼ → [foc] = ρA; dibé → [top] = ρX; D2: dibé ⌢ łį́į́ʼ ⌢ [bitah yíghááh] ✓.
  7. Result: IA-3 (no movement) is still cheaper. IA-1* is now a coherent EF movement in Chomsky's terms, but the strings are the same, so it is not supported over IA-3.

MA5′. HMP (10a) shináʼniidlóóh ‘I start to get cold (about dawn)’: default agreement from an inactive or hidden goal (as MA5, HD2)

  1. Σ {ρ1}, a change-of-state (Proto-Patient) participant, first-merged in Spec,V; φ = 1sg.
  2. A EM: V dlóóh, ρ1; REP ná; INCH's thematic ’ (ρI) in the object position; Qual-layer heads n, ii (IN-0).
  3. A MA-4′: ρ1 does not adjoin to the filled object position. Merge (MA-2′ a): [PP ρ1 PINCH] as the highest V-level non-head, above ná. P's own Agree values the PP's pronoun from ρ1 (1sg → shi).
  4. A EM v d (unaccusative, so not phasal), M/S; inheritance Qual → M/S. Agree: M/S finds no active goal. Spec,v is empty, ρ1 is inside the PP (MA-5a hidden / MA-5b inactive), and ρI is a thematic object. → default (IN-7 (D); HMP: the position is “entirely empty … as it is regularly in persons other than first or second”; Chomsky p. 157: “default morphology”; Hiraiwa's Icelandic default 3sg, pp. 78–79).
  5. A Spell-out D1: shi ⌢ ná ⌢ ʼnii ⌢ (M Ø) ⌢ d ⌢ dlóóh ✓ (alignment the site's; HMP do not segment (10a)).
  6. Change: the default now has the same shape as a sourced general mechanism (agreement failing on an inactive or hidden goal). Same string.

MA6′. HMP n. 4 sodiʼniiszin ‘I started to pray’: fusion is not adjunction (as MA6, HD6)

  1. Σ theme so=d-l-zin; ρA = 1sg in Spec,v; nothing below v → ρD (IN-9 (c)).
  2. A MA-4′: Position I is held by so. The candidate neither adjoins to it nor moves there. It stays and fuses with ρI: ’ + ’ → ’ (HMP: “hypothetically at least”).
  3. A Agree M/S ⟶ ρA (1sg). Spell-out: so ⌢ di ⌢ ’ ⌢ nii ⌢ … ⌢ zin ✓.
  4. Grewendorf's cluster (abstract): an element “act[s] as [a] landing site … due to morphological properties”. Here two identical ’ elements merge into one exponent instead of one hosting the other. That is the opposite of a cluster. HYPOTHESIS.

MA7′. HMP (16a) ʼiʼniichííl ‘it started to snow’ and (7b) ʼawééʼ ʼináʼniichah ‘the baby starts to cry’: Chomsky's expletive EM (as MA7, HD3, HD7)

  1. Σ ‘snow’: no participant. ‘cry’: ρA (sentient, atelic) in Spec,v only.
  2. A EM: V; v; (‘cry’) ρA in Spec,v; M/S; Qual-layer INCH (n, ii) with its requirement IN-8 (“must always be preceded by overt object agreement morphology”, HMP), read as INCH's EF.
  3. A Nothing below v can be raised to satisfy it (no slot there; ρA in Spec,v never satisfies INCH, HMP). → EM of an expletive: a fresh role-less ρD ⟨E none, φ 3, δ INDF⟩. Chomsky p. 157: EF “must be satisfied by EM, necessarily of an expletive since no argument role can be assigned”.
  4. A Agree M/S ⟶ ρA (‘cry’: 3rd, Ø). ‘snow’: no goal → default.
  5. A Spell-out: ‘cry’ ʼi ⌢ ná ⌢ ʼnii ⌢ … chah ✓; ‘snow’ ʼi ⌢ ʼnii ⌢ … chííl ✓ (not segmented further than HMP).
  6. B ʼawééʼ links to ρA, not to ρD (ρD has no role in the action to match).
  7. Result: SOURCED mechanism (Chomsky) + SOURCED pattern (HMP); the match between them is HYPOTHESIS. One mismatch: Chomsky's case is an inherited EF on T, a non-phase head. Here INCH is in the Qual layer, the phase head's own layer, so no inheritance is needed (MA-2′ c).

MA8′. Willie (64a) dabiiłtsą́ ‘They/he (was)were seen by them/*him’: da- is not covered by either original (as MA8, K5)

  1. Σ ρA plural (W&J gloss PL-3ACC-3NOM-saw); ρX 3; inverse.
  2. A EM; inheritance; Agree M/S ⟶ ρA (φ incl. plural).
  3. A Option e4 checked: da- copies number from M/S. In Chomsky, features pass only from a phase head to the head it selects (p. 144), and Agree targets a goal with features of its own. A disjunct copying an already-valued feature from M/S is neither. Option e2 (da- probes ρA directly): ρA in Spec,v* is at the phase edge and is the closest φ-goal, so Hiraiwa's locality (2c) allows it.
  4. A Spell-out D1: da ⌢ b ⌢ heads → dabiiłtsą́ ✓.
  5. Result: e2 needs nothing beyond standard Agree. e4 needs a head-to-head mechanism that neither original provides. That is a theory-internal lean to e2, not data (T41g).

MA9′. Hale 2001 (22c) hodildóóh ‘It (e.g., cloud) starts to drift’: U-1 loses its trigger (as MA9, SD5)

  1. Σ ρ1 = ⟨φ = 3 (zero), δ = AREAL⟩; start position unknown (SH-8).
  2. A U-1 (IM of ρ1 from Spec,V to Spec,v): in Chomsky only phase heads drive IM (p. 143), and the v of a one-participant verb whose participant starts low is not v*, so nothing triggers this step. Chomsky's own raising for unaccusatives goes to Spec-T by inheritance (p. 149, (11)), which here would be Spec,M/S, not Spec,v. So U-1 as written has no trigger.
  3. A U-2 (Agree only): M/S's probe reaches ρ1 directly because v is not phasal (“passive/unaccusative in situ”, p. 155; “v typically permits both long-distance agreement and raising”, p. 156). φ = 3 (zero).
  4. A δ = AREAL: MA-3a (Agree on Qual) or MA-3b (IM by Qual's EF). Either way it is pronounced on Qual: [hw + d]. Spell-out: hw ⌢ d ⌢ Ø ⌢ l ⌢ dóóh ✓.
  5. Result: U-2 is now favoured on two independent grounds, economy (as before) and Chomsky's architecture. It is still not favoured on data: no form separates U-1 from U-2 (T40e stays OPEN).

MA10′ (new). Willie (31f) hojishxash ‘That person bit that person’ (p. 129): two deictics, and only the subject's goes up (as SD7)

  1. Σ ρA = ⟨3 (zero), δ = 4⟩ (the biter); ρX = ⟨3 (zero), δ = 4⟩ (the one bitten).
  2. A EM: V + ρX (Comp,V); v*; ρA in Spec,v*; M/S; Qual; inheritance; Agree M/S ⟶ ρA (φ = 3, zero).
  3. A MA-3a: Qual's [uδ] Agrees with the closest δ-goal, which is ρA at the v*P edge. ρX is farther away and, in any case, inside the phase. MA-3b: Qual's EF is indiscriminate (p. 151) but can only see the v*P edge (PIC), so it reaches ρA's δ and not ρX's.
  4. A ρX's φ is taken by no head, so SH-5 keeps its whole bundle in one object exponent ho.
  5. A Spell-out D1: ho (object, a non-head, C1 first) ⌢ ji (subject δ at the top head) ⌢ … shxash ✓ (Willie gives no segmentation; the alignment is the site's, as in SD7).
  6. Result: both readings derive the order. The form shows why SH-5's “only a subject has a separate deictic position” follows from Chomsky's phase edge. It does not separate MA-3a from MA-3b.

The selection condition and Kibrik (8a) ʼá-zh-dí-l-zhééh (as SD8)

SOURCED Chomsky: inheritance is “by the LI selected by C (namely T)” (p. 144). HYPOTHESIS So the head that hands φ to M/S must be the head right above M/S. In Kibrik (8a) DETR dí is the only head above M/S (no separate qualifier in the segmentation), so DETR is both the top head and the selector: zh ⌢ dí ✓. SH-4 (“the highest nuclear head”) and MA-3′ (“the head that selects M/S”) agree on every form on the site. They come apart only in a word that has a DETR and a separate qualifier above M/S, together with a deictic subject. SH-4 predicts δ ⌢ DETR ⌢ Qual … and MA-3′ predicts DETR ⌢ δ ⌢ Qual … (schematic only; no such form has been opened: T41a). This answers T40c's “is the phase head DETR?” conditionally. Under Chomsky's selection condition it is DETR only when nothing intervenes between DETR and M/S.

6. Rudin's diagnostics applied to the word

Rudin's diagnostic ((69), p. 478; clitics pp. 461–466)Navajo word-internal analogueWhat the sources showReading
Can a filled specifier take an adjoined second element? (pp. 480–481)The object position under INCH; Position I with so-No. The object goes elsewhere (HMP n. 2; (10a)) or fuses (n. 4)HYPOTHESIS [−MFS]-like (MA-4′)
Where do the extra elements go? ([−MFS]: “in or adjoined to IP”, (71b))The bumped PPHigher, to Position I (HMP n. 2)HYPOTHESIS same logic (a different position), opposite direction (up, not down)
Clitic right after the first element ([−MFS] +, p. 461; needs “encliticization to the first constituent”, p. 463)Polar interrogative enclitic in a word with a disjunct preverbHale 2003 n. 12: the preverb ka “can be separated from the rest of the verb by the polar interrogative enclitic” (forms in a broken font, not reproduced)HYPOTHESIS [−MFS]-like on Rudin's decisive test. Two caveats. Whether ka was the first of several disjuncts cannot be seen. Whether the enclitic attaches to the first constituent in Rudin's sense is not shown (T41f)
Fixed order ([+MFS]: “restricted at least to subject > nonsubject”, p. 475)Fixed order of disjunct prefixes by class and height (C2)Fixed, but by morpheme class (preverb, iterative, plural), not by grammatical functionHYPOTHESIS not Rudin's kind of order: counts for neither type (changed from the 1:52pm table, which counted it as a property)
Parentheticals/adverbs after the first element ([−MFS] +)Material between disjunctsOnly the enclitic of n. 12 is knownOPEN
Multiple extraction; wh-island violations ([+MFS] +)—No word-internal analogue: prefixes are not extractedNot applicable

HYPOTHESIS Summary. On the diagnostics that can be applied, the Navajo word patterns with Rudin's [−MFS] type. The one point that once looked [+MFS]-like, the fixed order, turns out not to be the kind of order Rudin's diagnostic is about. The 1:52pm verdict on the disjunct zone (“OPEN, leaning not a cluster”) becomes “HYPOTHESIS: [−MFS]-like on the one applicable decisive test”, with T41f as the check.

Update 2026-10-20, 12:46pm (T41f). Caveat 2 of the clitic row is answered yes: -ísh is a first-constituent enclitic. Caveat 1 is moot, because ka is outside the verb word. The row's verdict is now OPEN (not testable with this enclitic if EN-1′ holds), and the summary's “[−MFS]-like on the one applicable decisive test” is withdrawn. See where the yes/no enclitic -ísh can go, §7.

7. What each side gets right and misses

SideGets rightMisses or costs
Grant's typology noteAll six definitions are accurate at the level of a one-line summary. The [±MFS] diagnostics match Rudin's table (69). It places Multiple Agree, inheritance, EF and phases at the right levelIt leaves out wh-islands. It states “clitics after the whole group” without Rudin's own caveat (inconclusive, p. 466). It describes [+MFS] in Grewendorf's terms (“a moved wh hosts further whs”), not Rudin's (adjunction to SpecCP). “Whether a head has an EF” is pre-2008 wording. The “Attract Closest” account of Superiority is not Rudin's
The Merge/Agree page (1:52pm)Agree inside, Merge at edges. The [−MFS]-like verdict. The EPP reading of the placeholder (now matched by Chomsky p. 157). Its decision to leave Multiple Agree OPEN, now narrowedMA-0's test misclassifies agreement. MA-3 is not Chomsky's version. The δ site was over-confidently called Merge. The voice exponent was called a Multiple Agree candidate, which Hiraiwa's definition excludes. The PP was called a phase without saying that this is our extension. The disjunct order was counted as a Rudin property
Chomsky 2008 applied to Navajov vs v* gives U-2 and SH-5 without stipulation. Phase-head EF gives the placeholder and the bump. Inheritance gives the split subjectIt says nothing about a phase inside a word, or about a P-phase. It does not decide MA-3a vs MA-3b. Its EF moves phrases, not single features, so MA-3b needs δ to be a separate object
Hiraiwa 2001 applied to NavajoClosest c-command states PORT-CLOSEST exactly. Default agreement under an inactive goal fits the ‘start to’ defaultMultiple Agree has no Navajo instance on the site

8. Grant's frame

GRANT'S CONJECTURE Word-internal Merge or Agree sites. After the check, Agree sites inside the word are firmer, since each now matches a definition in the originals (M/S by inheritance, P's own Agree, default under an inactive goal). Two Merge sites are certain: the bumped object (Merge in either reading) and the placeholder (EM of an expletive). The deictic is OPEN between the two operations. Merge builds structure; Agree values features. The deictic case is exactly where the strings cannot tell which one applied.

GRANT'S CONJECTURE Feature bundles and the weak thesis. Every probe on this page Agrees with a bundle ρ in D1, before any noun arrives (B). Chomsky's own formulation needs goals with features, not nouns. The expletive case (MA7′) is a slot with features and no role, and the noun ʼawééʼ cannot link to it. That supports the weak thesis (roles live on slots, nouns link later) and says nothing about the strong one.

HYPOTHESIS Dowty and theta vs event structure. Nothing in this check changes anyone's role in the action. Proto-Agent ρA starts in Spec,v(*), and a change-of-state Proto-Patient ρ1 starts low. INCH adds an onset to the event and moves no role. The event-structure layer and the theta roles stay compatible descriptions of the same participants.

GRANT'S CONJECTURE STRATA (theta roles first, then event-structure roles, then noun selection). Chomsky (p. 140) pairs EM with “generalized argument structure (θ-roles …)” and IM with “discourse-related properties”. On this page the role layer is all EM (Σ slots at their positions), and so is the event layer (INCH, its placeholder, the bumped PP). The only possible IM (MA-3b, the deictic) is discourse-related, and noun linking comes last in D2. The STRATA order fits that pairing. It does not separate STRATA from the adjunct view or the inside-the-clause view, because none of them predicts a different string here.

HYPOTHESIS The changeover from rule set A to rule set B. If Qual is the phase head of D1 (C-like, with M/S = T, Hale 2003 p. 10), then the word is a CP-like phase. In Chomsky's terms (p. 143), Transfer at the word's phase boundary is the point where rule set A's output is handed on. Rule set B then sees only the word's edge and exponents (“output, not derivation”, Merge/Agree page §11.2). That is a fit, not a test: where the word stops is still unsolved (Hale 2003 n. 12).

9. Tasks T41a–h

TaskWhat it needs
T41aSH-4 vs MA-3′ selection: a printed form with a DETR and a separate qualifier above M/S and a deictic (4th/indefinite/areal) subject. SH-4 predicts δ ⌢ DETR ⌢ Qual; MA-3′ predicts DETR ⌢ δ ⌢ Qual (schematic). Kibrik 2005, Young & Morgan
T41bMA-3a vs MA-3b (Agree vs EF-movement of the deictic): any A′-property of the top-edge deictic (e.g. scope or discourse effects that do not come from its person value), or a deictic of something other than the Spec,v slot appearing at the top
T41cMA-5a vs MA-5b (PP as phase vs inactive object): a word with an active goal below a PP that M/S could otherwise reach
T41dGrewendorf 2001 full text (abstract only so far): his diagnostics for cluster formation, to test MA6′'s fusion reading
T41eChomsky 2000, 2001 (the EPP-feature definitions behind the wording in Grant's note, and Match/Agree as Hiraiwa cites them)
T41fRudin's clitic test, refined from T40a: a legible form with two or more disjunct prefixes and the polar enclitic, plus a statement of whether that enclitic attaches to the first constituent (Hale 2003 n. 12, Platero p.c.; Young & Morgan)
T41gda-: is there any form where da- counts a plural that M/S does not spell (which would favour e2 over e4)? (joins T30, T10f)
T41hIs ‘go’ (W&J (50b)) built with a phasal v*? Chomsky ties v* to transitive and experiencer verbs (p. 143). For IA-1* this decides whose EF would move the topic slot

Update 2026-10-20, 4:46am. T41e answered and T41h partly answered on the 2000/2001 page (‘go’ is not v* by either definition). T41c is sharpened (EP-5 vs MA-5a; → T42c). New tasks T42a–h are there.

T40 statuses after this page: T40b answered (this page). T40c conditionally answered (§5, selection condition) → T41a. T40d narrowed (MA3′). T40e OPEN, with a second theory-internal ground for U-2. T40g partly answered: in Chomsky's model v*P is a phase and unaccusative vP is not, and every site fact fits. T40a refined as T41f. T40f, T40h, T40i, T40j are unchanged.

Changes made elsewhere

Update notes on Merge and Agree inside the word (lede, rules, §6, tasks). There is a new row on open questions (T41a–h), a sources section (added 2026-10-19, 8:46pm), and a home-page entry. No verb-dictionary entries changed and no new Navajo forms were added.

The older definitions behind “edge feature”: what Chomsky 2000 and 2001 say about the EPP-feature, Match/Agree, activity and phases, and what that changes inside the Navajo verb word (T41e)

Working note. Kept as written (URL, data and status labels unchanged). Filed under Merge and Agree inside the word; the consolidated argument is the essay merge-agree. Rule IDs: rules index.

Verbs on this page → dictionary: yishxash ‘bite’ · yíghááh ‘join; with -tah ‘among’: go among’

Restated on the unified structure (2026-11-15). The framework is restated in §1: VPISH (arguments merged in the verb’s scope, subject in Spec,vP) with theta roles, and V→v→T with the subject raised to Spec,TP as the working default in tensed clauses; each pattern is what else moves, with an LF whose scope matches c-command. The head-direction mismatch argued below as rule sets A/B or a changeover is now stated as Grant’s Merge-site hypothesis and tested in §3.1–3.2. Position-class template statements below describe the order of prefixes; they are not the derivation (§3). Derivations below in which the verb or subject does not raise are labelled contrast; the working default raises V→v→T and the subject to Spec,TP. Status on the spine: see unified structure, §1. The data, judgments, translations and citations below are unchanged; the analysis below is kept as the record of what was argued then.

2026-11-11 audit pass 1. MI/DbP conditions are source-theory claims, not premises of the current VPISH-plus-theta regime. The old expletive/phase and TP rules are conditional hypotheses; a word’s exponent order does not identify T or a phase boundary. Compare the primary V→v→T and separate subject raising with template/PF rivals. Full target-before-Move and LF re-derivations of the retained lists remain OPEN.

2026-11-11 audit pass 3. Local box file paths were removed from this page; files inside the repository are given by repo-relative path, and the source citation next to each removed path stays as the reference. See the pass 3 report.

2026-11-13 audit pass 7. An earlier dated note on this page treats V→v→T and subject raising as Grant’s conjecture or a comparator to test against PF/template rivals. Since 2026-11-13 they are the working default in tensed clauses: VPISH frames where the arguments start so that syntax and semantics line up; it is not a word-order account. PF/template accounts are kept only as labelled contrast, and the open question is what else has to move. The earlier note is kept as written. See the 2026-11-13 note and the pass 7 report.

Last night's page, the borrowed terms checked against their originals, opened Rudin 1988, Hiraiwa 2001 and Chomsky 2008. Two older works were still unopened: Chomsky 2000 (“Minimalist Inquiries”, MI) and Chomsky 2001 (“Derivation by Phase”, DbP). The wording in Grant's typology note comes from them (“whether a given head has an Edge Feature triggering Spec Merge”), and so do Hiraiwa's definitions of Match and Agree. This page opens both, sets their definitions beside the ones the site has been using, tightens the rules, and re-runs the derivations that change. The two operations stay apart throughout. Merge builds structure: it puts an element at a position, either a fresh element (“pure Merge” in MI's term, External Merge in 2008's) or a copy of one already in the structure (Move; Internal Merge). Agree values features such as person and number: a probe on a head finds the closest matching goal and copies its value, and nothing moves. MI's Move is the two together plus “pied-piping” (p. 101). Written 2026-10-20, 4:46am PT.

Ground rules (Grant's). No new Navajo form is introduced. Every form below is already on this site, from a named source with its example number, and the page links to where it was read. None is segmented beyond what its source prints. MI and DbP are general-syntax works, and neither mentions Navajo. Every mapping of a Navajo fact onto their terms is therefore ours, graded HYPOTHESIS at most. What the originals say is graded SOURCED, with the printed page number. Only contradicting data count against an analysis; a source's silence does not. No earlier analysis (Hale, Willie, HMP, Hiraiwa, Chomsky, or this site's own rules) is assumed to be fully right.

Notation (the site's own, as on the originals page). Σ = role-schema step (which slots the verb has and their roles in the action); A = word level (cycle D1, rule set A: the prefixal, left-headed functional spine); B = clause level (cycle D2, rule set B: right-headed clause syntax and noun linking). Spine (Hale 2003): Qual > M/S > v > V. Slots ρ = ⟨E, pos, φ, (δ)⟩ (SH-0): φ is person/number, δ the deictic value (fourth person, indefinite, areal). INCH = the ‘start to’ prefix ʼnii-; PINCH = its postposition-like requirement (IN-8); ρI = INCH's own ’ in the object position; ρC = the slot that satisfies PINCH; ρD = the role-less placeholder ’. New abbreviations from the originals: EPP-feature = a head's requirement that its specifier be filled (MI p. 102); Expl = expletive; MoM = “Merge over Move”: simpler operations preempt Move (MI pp. 101–104); PIC1 = MI's phase-impenetrability condition (21); PIC2 = DbP's revised version (11); activity = a goal can take part in Agree only while it still has an unvalued feature of its own (DbP (3a)).

Bottom line, in plain words

SOURCED Grant's wording is the 2000/2001 idea, faithfully stated. In MI and DbP, having an “EPP-feature” is something a head may or may not have. T's “might be universal”; for the phase heads v and C it “varies parametrically among languages and if available is optional” (MI p. 109). DbP p. 41: “Something distinguishes among heads H that require, allow, or disallow that H head an occurrence of K.” That is exactly the question in Grant's note. The 8:46pm page's verdict, “pre-2008 wording”, is now confirmed with page numbers. Only the name “Edge Feature” is 2008's. In 2008 every lexical item has one.

HYPOTHESIS The main change concerns the ‘start to’ placeholder. It must belong to a low, non-phase head, not to the top of the word. Both MI and DbP allow an expletive to be inserted only into the extra position of a head like T. A phase head (C, or a transitive v) may fill its extra position only by moving something up from inside its own phase: “If H is v/C, XP is not introduced by pure Merge” (MI (5a), p. 102); “pure Merge from outside Ph is barred” (p. 109). Last night's rule MA-2′(c) put the placeholder in the qualifier layer, which the site treats as the word's phase head, and said “no inheritance is needed”. Under both definitions that is exactly the excluded case. Hale, Munro & Platero print placeholder forms (‘the baby starts to cry’, ‘it started to snow’), so the data stand and the rule has to move. It moves back to where the site's own IN-8 first put it: to PINCH, the postposition-like piece at the top of the verb phrase. Two more facts fit a T-like head there. INCH's requirement is obligatory (HMP: it “must always be preceded by overt object agreement morphology”), as T's EPP may be, while a phase head's is optional. And the position it opens carries no role, which MI (6) requires for an expletive.

HYPOTHESIS Two of IN-9's stipulations now follow. First, why does the placeholder appear only when nothing below v can fill the slot? MI's “Merge preempts Move” would insert the placeholder whenever it is available. But in a verb with a low participant, that participant would then have nowhere to be pronounced, because INCH's own ’ already holds the object position. So only the bumping derivation survives (EP-3). Second, why is the subject of ‘cry’ never used to fill the slot? PINCH sits below v, and movement only goes up, so a participant that starts in Spec,v is out of its reach.

HYPOTHESIS The default third person in ‘I start to get cold’ now matches MI's own Icelandic case, with one cost. In MI, if the closest goal for T has already been agreed with and is “inactive”, it still blocks T from looking further, and T shows a default (“[3person] (the default choice)”, p. 128; “if such an associate is lacking, T is default”, p. 126). In shináʼniidlóóh, PINCH has already agreed with ‘I’, so Mode/Subject finds an inactive goal and shows the default. The cost: in MI and DbP, what keeps a goal active is structural Case (DbP p. 6: a noun “is active only when it has structural Case”). The site's slots have no Case. Using this mechanism means adding an abstract activating feature to every slot. The rival, “the postposition phrase is a phase”, is still our own extension, so the two remain OPEN (T41c).

HYPOTHESIS Phases: one reason changes, no string changes. In MI the object inside a transitive verb phrase is sealed off as soon as the phrase is complete. In DbP a T-like head can still reach it until the next C-like head arrives: “T in the domain of Z can agree with an element within its complement” (p. 14). On DbP's version, Mode/Subject never shows the object's person because the subject is closer, not because of a wall. Every form on the site comes out the same either way (T42a).

HYPOTHESIS Hiraiwa's citations check out, with two refinements. His Match (2) is MI (40), p. 122, word for word. But Chomsky himself narrows “feature identity” to “identity of the choice of feature, not of value” (MI p. 124), and DbP to “Non-distinctness: same feature, independently of value” (p. 5). MI p. 124 also says a head's person-and-number set is valued “as a unit”, and that “The φ-features of T do not agree with different NPs”. That strengthens last night's MA-6′ (the object prefix is not one probe agreeing with two goals). It also adds a commitment to MA-3a: the deictic δ must be a feature separate from person, as the site's slot notation already writes it.

GRANT'S CONJECTURE Grant's frame. MI p. 127 separates the two systems in so many words. Roles come from “pure Merge of XP to substantive LIs”, and agreement (“checking theory”) is a different system with different properties. On this page every position that ‘start to’ adds carries no role, and every participant keeps the role it got where it first merged. That fits STRATA's order (roles first, the event layer next, nouns last), and theta roles and event-structure roles stay compatible. It does not separate STRATA from its rivals.

On this page: 1 Sources · 2 Definitions side by side · 3 Hiraiwa's citations checked · 4 Rules EP-0 to EP-8 · 5 What changes, rule by rule · 6 Derivations EP1–EP8 · 7 Gets right / misses · 8 Grant's frame · 9 Tasks T42a–h and T41 statuses

1. Sources

SourceAccess this runUsed for
Chomsky, Noam. 2000. Minimalist inquiries: the framework. In R. Martin, D. Michaels & J. Uriagereka (eds.), Step by Step: Essays on Minimalist Syntax in Honor of Howard Lasnik, 89–155. Cambridge, MA: MIT Press (= MI)Opened: 71-page scan of an e-book printout of pp. 89–155 (document-delivery copy in a University of Arizona course folder). The scan has no text layer, so it was OCR'd on the box (tesseract 5.5). Page numbers come from the printed “Page N” headers. Printed p. 94 is missing from the scan. The quotations for pp. 102, 103 and 109 were checked against the page images; the rest are from the OCR, which is clean on these pages. Notes are not usedMerge, Agree, Move and the preference order (pp. 101–102); EPP-features (p. 102); (4)–(6), properties of CFCs (pp. 102–103); Merge over Move and subarrays (pp. 104, 106); phases (pp. 106–108); (17) inert phase head; PIC (21) (p. 108); (24) phase-head EPP, optional (p. 109); probe/goal, (40) Match, (41) equidistance (p. 122); activity, (42) defective intervention (p. 123); “one fell swoop” φ-deletion and value-free identity (p. 124); Expl = [person] only (p. 125); (45) EPP satisfaction and default T (p. 126); θ-theory vs checking theory (p. 127); Icelandic default [3person] (p. 128); (50) (pp. 129–130)
Chomsky, Noam. 2001. Derivation by phase. In M. Kenstowicz (ed.), Ken Hale: A Life in Language, 1–52. Cambridge, MA: MIT Press (= DbP)Opened: 28-page scan, two book pages per sheet, pp. 1–52 including notes (Georgetown faculty page). No text layer, so OCR'd on the box. Page numbers from the printed running heads; the quotation on p. 14 was checked against the page image(2) three kinds of uninterpretable feature (p. 4); Match = identity, then non-distinctness (pp. 4–5); (3a) activity, (3b) φ-completeness, Case as the activating feature (pp. 5–6); Tdef, Expl = [person] (p. 7); (5)–(6), EPP and φ-completeness (pp. 8–9); EPP-feature defined, Move preempted (p. 10); strong vs weak phases (p. 12); PIC (7)–(11) (pp. 13–14); T can probe into the v*P complement (p. 14); (14) Maximize matching effects (p. 15); (43)–(46) equidistance and the phonological edge (pp. 27–28); (60)–(61) optional EPP on v* (p. 35); EPP as the property of heading an occurrence (pp. 40–41); n. 8 v* (p. 43)
Hiraiwa, Ken. 2001. Multiple Agree and the Defective Intervention Constraint in Japanese. MITWPL 40: 67–80Re-opened (the extraction from 2026-10-19, 8:46pm)His citations of Chomsky 2000/2001, (1)–(5), checked in §3
Chomsky, Noam. 2008. On phasesAs recorded on the originals page (not re-opened)pp. 139, 143, 144, 148, 155, 157, for comparison
Grant, “Typology of Merge/Agree Phenomena”As recorded on the originals page ([local working copy, not in the repository])The EPP/Edge Feature line compared in §2
This site's pagesRe-readHMP ms. (IN-5′, IN-7 (D), IN-8, IN-9; HD1–HD7); originals page (MA-0′–MA-6′; MA1′, MA4′, MA5′, MA7′, MA10′); Merge/Agree page
Not opened—Chomsky 2004 (“Beyond explanatory adequacy”), where the EPP-feature becomes the edge feature; Grewendorf 2001 (still paywalled, T41d); Young & Morgan. None is used as evidence

2. Definitions side by side

Left: what the site and Grant's note have been using. Middle: what MI and DbP say, with page. Right: what follows.

TermSite / Grant's note (with 2008 as recorded)MI 2000 and DbP 2001 SOURCEDVerdict and consequence
EPP-feature (later “edge feature”)Grant: “Whether a given head (T, v, C…) has an Edge Feature triggering Spec Merge.” 2008: every lexical item has EF (p. 139)MI p. 102: “Each CFC also allows an extra Spec beyond its s-selection: for C, a raised wh-phrase; for T, the surface subject; for v, the phrase raised by object shift (OS). For T, the property of allowing an extra Spec is the Extended Projection Principle (EPP). By analogy, we can call the corresponding properties of C and v EPP-features”. “EPP-features are uninterpretable (nonsemantic, hence the name)”. MI p. 109: “The EPP-feature of T might be universal. For the phase heads v/C, it varies parametrically among languages and if available is optional.” (24): “The head H of phase Ph may be assigned an EPP-feature.” DbP p. 10: “an uninterpretable selectional feature (by definition, an EPP-feature), which requires Merge in [Spec, LB(K)]”. DbP p. 41: “Something distinguishes among heads H that require, allow, or disallow that H head an occurrence of K.” DbP (61a), p. 35: “v* is assigned an EPP-feature only if that has an effect on outcome.”Grant's line is an accurate statement of the 2000/2001 notion, and the question “does this head have one?” only makes sense there. In 2008 the question becomes “can this head drive movement?” (phase heads, or by inheritance). For Navajo the useful 2000 distinction is obligatory (T-like) vs optional (phase heads)
Who may fill the extra position with an expletiveMA-2′(c): the placeholder ρD = “EM of an expletive that satisfies the EF of the Qual-layer INCH … Since Qual is D1's phase head …, this EF is the phase head's own: no inheritance.” 2008 p. 157: expletive EM satisfies “the inherited edge feature of T”MI (5a), p. 102: “If H is v/C, XP is not introduced by pure Merge.” p. 103: “Pure Merge is Merge that is not part of Move”; (6): “Pure Merge in θ-position is required of (and restricted to) arguments”, so pure Merge in [Spec, T] is restricted “to Expl”. p. 109: since the phase head is inert beyond its phase, “the EPP must be satisfied by raising within Ph: pure Merge from outside Ph is barred.” (45b), p. 126: “The EPP can be satisfied by i. Merge of expletive ii. Merge of associate iii. Merge of α closer to T than the associate”. DbP p. 7: “the EPP-feature of Tdef is satisfied by Merge of expletive” or “by raising”MA-2′(c) as written is excluded by both texts, and 2008's own case is T (non-phase, inherited) too. Since the placeholder forms are printed (HMP (7b), (16a)), the requirement must sit on a non-phase head. New EP-1, EP-2 (§4)
Merge over MoveNot used on the siteMI p. 101: Move is “more complex than its subcomponents Merge and Agree”; p. 102: “Merge or Agree (or their combination) preempts Move, which is a ‘last resort’”. p. 104: “Preference of Merge over Move selects the former”, and the choice “depends on whether or not an expletive is available in the initial lexical array”. p. 106: “If the subarray in active memory does not contain Expl, then Move can take place …; if it does, Merge of Expl preempts Move.” DbP p. 10: Move is “preempted where possible by the simpler operations Merge and Agree”Applied to INCH, this would insert the placeholder whenever it is available. Combined with “the object position is full” (HMP n. 2), it derives when the placeholder appears (EP-3)
MatchHiraiwa (2a): “Matching is feature identity” (cited to MI p. 122). MA-0′, MA-6′ use itMI (40), p. 122: “a. Matching is feature identity. b. D(P) is the sister of P. c. Locality reduces to ‘closest c-command.’” p. 124: “We therefore understand ‘feature identity’ in (40a) to be identity of the choice of feature, not of value.” DbP p. 5: “Accordingly, Match is not strictly speaking Identity, but Non-distinctness: same feature, independently of value.”Hiraiwa quotes (40) exactly. Chomsky's own refinement (same feature type, any value) changes no site verdict. The object prefix still reads two feature types (φ of the object, δ of the subject), so MA-6′ stands
Agree and the φ-set as a unitMA-3a: Qual passes [uφ] to M/S and keeps [uδ]. MA-6′: no probe-side agreement with two goals (Hiraiwa n. 12)MI p. 124: “We take deletion to be a ‘one fell swoop’ operation, dealing with the φ-set as a unit. Its features cannot selectively delete: either all delete, or none. The φ-features of T do not agree with different NPs”. MI p. 124 also: “verbs agree with nouns, not conversely”MA-6′ gains an earlier source than Hiraiwa n. 12. MA-3a survives only if δ is not part of the φ-set. The site's slot notation already separates them (ρ = ⟨E, pos, φ, δ⟩), but this is now a commitment, not a convenience (T42b)
Activity and defective interventionMA-5b: the object inside a PP is inactive after P's own Agree; Hiraiwa (3) DICMI p. 123: “uninterpretable features render the goal active … The operations Agree and Move require a goal that is both local and active.” (42): “α > β > γ … β and γ match the probe α, but β is inactive so that the effects of matching are blocked.” p. 127: after Case is deleted the phrase is still “‘visible’ to a probe, allowing defective intervention effects”. DbP (3a), p. 6: “Goal as well as probe must be active for Agree to apply.” “φ-features of N are interpretable; hence, N is active only when it has structural Case”; once valued it is “frozen in place”. DbP p. 27: “inactive nominals induce intervention effects”MA-5b now has its original mechanism. But in both texts the activating feature of a nominal goal is structural Case. Navajo slots carry none on this site, so MA-5b needs an added abstract feature (ours). A cost, not data (EP-5)
φ-completenessNot usedDbP (3b), p. 6: the probe “must have a complete set of φ-features (it must be φ-complete) to delete uninterpretable features of the paired matching element”. MI p. 124: “only a probe with a full complement of φ-features is capable of deleting the feature that activates the matched goal”For EP-5, PINCH's agreement must be φ-complete (person and number). Untested (T42f)
Default agreementIN-7 (D) (HMP: the position is empty “as it is regularly in persons other than first or second”); 2008 p. 157 “default morphology”; Hiraiwa pp. 78–79MI p. 126: with Merge of an α closer to T than the associate, if “such an associate is lacking, T is default.” p. 128: Icelandic remote nominative “allows number agreement but not first/second person agreement. That would follow if the [person] feature of T reduces to [3person] (the default choice) when it attracts quirky Case or Expl”The original of Hiraiwa's Icelandic default. The default is third person, which is what IN-7 (D) needs
ExpletiveρD = ⟨E none, φ 3, δ INDF⟩ (IN-9)MI p. 125: Expl's feature set “must be less than a full φ-set, hence optimally just [person]”. DbP p. 7: “pure expletives of the there-type should have no other formal features”. MI p. 128: “it-type Expl with a full complement of φ-features” also existsρD has δ = INDF beyond person, so it is not a pure there-type expletive. Whether it is it-type, or not an expletive at all, is OPEN (T42f)
PhasesMA-1′: CP and v*P (2008 p. 143); unaccusative v not phasalMI: “A phase is CP or vP, but not TP or a verbal phrase headed by H lacking φ-features and therefore not entering into Case/agreement checking: neither finite TP nor unaccusative/passive verbal phrase is a phase” (pp. 106–107). p. 108: “verbal phases being limited to transitive v with φ-features and EA”. DbP p. 12: “phases are CP and v*P”; “call the former strong phases and the latter weak. The strong phases are potential targets for movement: C and v* may have an EPP-feature”. n. 8, p. 43: “Call v with full argument structure v*: transitive v or experiencer.”The v/v* split the site took from 2008 is already in MI. Neither text says how to class an unergative. By MI's own criterion (φ-features and EA), a v with no object to agree with is not a phase (EP-8, T41h)
PICMA1′: “The v*P complement (VP) is behind the phase edge, so ρX is not a candidate” for M/SMI (21), p. 108: “In phase α with head H, the domain of H is not accessible to operations outside α, only H and its edge are accessible to such operations.” DbP (11), p. 14: “The domain of H is not accessible to operations at ZP; only H and its edge are accessible to such operations”, ZP the next strong phase. p. 14: “The probe T can access an element of the domain YP of HP: the PIC imposes no restriction on this. But with Σ = ZP (so that Z = C), the probe Z cannot access the domain YP.” “T in the domain of Z can agree with an element within its complement, for example, with the in-situ quirky nominative object of its v*P complement.”MA1′'s reason holds under PIC1 only. Under PIC2, M/S (T-like) can see ρX, and the closer ρA is what wins. Qual (C-like) cannot see it under either version, so MA10′ is unchanged (EP-6)
Equidistance / phonological edgeHiraiwa (4), (5)MI (41), p. 122: “Terms of the same minimal domain are ‘equidistant’ to probes.” DbP (43)–(44), p. 27: “(44) Terms of the edge of HP are equidistant from probe P”; (46), p. 28: “The phonological edge of HP is accessible to probe P”, with intervention evaluated at the next strong phaseNo site rule uses equidistance. Recorded for §3 only
Roles vs agreementSTRATA (Grant); 2008 p. 140 pairs EM with θ-structureMI p. 127: “θ-structure and similar semantic roles are based on pure Merge of XP to substantive LIs or their projections. Checking theory, in contrast, involves uninterpretable features of functional categories”; “The two theories differ in nature and implementation”. MI (6), p. 103 (above)Two separate systems already in 2000. Used in §8

3. Hiraiwa's citations checked

Hiraiwa 2001Cited asIn the originalVerdict
(1) AGREE: “uninterpretable features of α and β are checked/deleted”“cf. Chomsky 2000”MI p. 122: “The erasure of uninterpretable features of probe and goal is the operation we called Agree.” DbP p. 5 replaces erasure-only with valuation: unvalued features get values, then deleteA fair paraphrase of MI (marked “cf.”). DbP's valuation version is the one the site's rules assume (a probe copies a value)
(2) Match a–c“Chomsky 2000:122”MI (40), p. 122, identicalExact. Chomsky narrows (a) on p. 124 and in DbP p. 5 (§2)
(3) DIC: “β is inactive due to a prior Agree with some other probe”“cf. Chomsky 2000:123”MI (42), p. 123: “β is inactive so that the effects of matching are blocked”. Inactivity = the goal's Case already deleted (pp. 123, 127)“Due to a prior Agree” is Hiraiwa's gloss, and it is accurate for MI: Case is deleted under Agree
(4) “[Only (K.H.)] the phonological edge of HP is accessible to probe P”“Chomsky 2001”DbP (46), p. 28: “The phonological edge of HP is accessible to probe P.”Exact, and Hiraiwa marks his own addition “[Only (K.H.)]”
(5) Equidistance: “Terms of the edge of HP are equidistant from probe P”“Chomsky 2000:122 & footnote 77”MI (41), p. 122 is worded “Terms of the same minimal domain are ‘equidistant’ to probes”. Hiraiwa's wording is DbP's restatement (44), p. 27Content right; the wording he quotes is DbP's, not MI's
Icelandic default, (30)“Chomsky 2000:130”MI pp. 128–130 (default [3person], p. 128; (51), p. 130)Consistent

4. Rules EP-0 to EP-8 (ours; all HYPOTHESIS unless marked)

EP-0 THREE-FEATURES  A head may carry (i) a phi-probe (drives Agree), (ii) an EPP-feature
                 (requires its Spec to be filled: drives Merge), or both; a goal takes part
                 in Agree or Move only while ACTIVE (it still has an unvalued feature of its
                 own).  (SOURCED: DbP (2a-c), p. 4; MI p. 121-122; DbP (3a), p. 6.)
                 Refines MA-0': an exponent at H is Agree-morphology if it spells what H's
                 phi-probe copied; it is a Merge site if it fills H's EPP position.
EP-1 WHO-FILLS   (SOURCED: MI (5a), (6), (24), pp. 102-103, 109; (45b), p. 126; DbP p. 7.)
                 (a) Phase head (C, v*): its EPP-feature is optional and can be satisfied
                     only by raising from inside its own phase. No pure Merge, so no
                     expletive.
                 (b) Non-phase head (T-like): its EPP-feature may be obligatory and can be
                     satisfied by raising OR by pure Merge of an expletive.
                 (c) Pure Merge into a theta-position is for arguments only, so an
                     expletive goes only into a non-theta position.
EP-2 INCH-EPP    (revises MA-2'(c); restores IN-8's location)  INCH's requirement IN-8 is
                 an obligatory EPP-feature on P_INCH, a NON-PHASE head merged at the top of
                 the V-level (below v, above the disjuncts such as na).  The position it
                 opens (written on the site as the PP's object rho_C) is a non-theta
                 position.  It is filled by
                   IM (Move) of the closest slot below P_INCH (= IN-5' BUMP), or
                   pure Merge of rho_D (= IN-9's placeholder; EP-1 (b), (c)).
                 NOT the Qual-layer phase head: that would make the placeholder pure Merge
                 into a phase head's EPP position, which EP-1 (a) excludes, while HMP
                 print placeholder forms ((7b), (16a)).
                 Restated structure (ours): [P_INCH-P rho_C [P_INCH' P_INCH [VP ... na ... V]]];
                 C2 still spells rho_C first, as IN-5' requires.
EP-3 MOM-FILL    (derives IN-9's two stipulations)
                 (a) If rho_D is in the active subarray, Merge of rho_D preempts Move
                     (MI p. 106).  If a slot rho_L is first-merged below P_INCH, rho_L then
                     has no position: the object position is held by rho_I (IN-1; HMP
                     n. 2), and nothing adjoins to an occupant (MA-4').  -> crash.
                     Unless rho_L can fuse with rho_I (' + ' -> ', HMP n. 4, only for an
                     indefinite ').  So rho_D converges only when nothing is below P_INCH.
                 (b) Move goes only to a c-commanding position.  A slot in Spec,v is merged
                     ABOVE P_INCH, so it can never satisfy P_INCH ("never the Spec,v
                     participant").  Areal ho- (a subject delta at the top, SH-4) likewise.
                 (c) If two slots are below P_INCH, the CLOSEST one is moved (MI (5b-iii):
                     "if there is no closer candidate"; (40c)).  Schematic prediction:
                     Spec,V before Comp,V (T42e).
EP-4 OBLIGATORY  P_INCH's EPP is obligatory (HMP: "must always be preceded"), like T's
                 ("might be universal", MI p. 109), not like a phase head's ("optional").
                 With nothing below and no rho_D available: crash.  Agrees with EP-2.
EP-5 DEFAULT-DI  (gives IN-7 (D) a sourced mechanism; one of two)  P_INCH Agrees with the
                 slot it attracts (its "object agreement morphology", HMP) and, being
                 phi-complete, deactivates it (DbP (3b); MI p. 124).  M/S then probes; in
                 an unaccusative (v not phasal) its closest matching goal is that inactive
                 slot -> defective intervention (MI (42)) -> M/S default [3person]
                 (MI pp. 126, 128).
                 COST: activity needs an activating feature on rho (MI/DbP: structural
                 Case).  The site's slots have none, so we must add an abstract [uK] to
                 rho (ours).  Rival: MA-5a, PP (or P_INCH-P) as a phase (our extension).
                 OPEN between them (T41c, T42c).
EP-6 PIC-CHOICE  M/S never takes the object's phi.  Two reasons, same strings:
                 PIC1 (MI (21)): VP is sealed once v*P is complete.
                 PIC2 (DbP (11)): M/S (T-like, inside the C-like D1) can see into VP until
                 Qual is merged; rho_A at the v*P edge is the closer goal (MI (40c)).
                 Qual (C-like = Z) cannot see VP under either.  They differ only for a
                 transitive whose Spec,v* slot is not a matching goal (T42a).
EP-7 PHI-UNIT    (SOURCED: MI p. 124.)  A phi-set is valued and deleted as a unit; one probe
                 does not agree with two goals.  MA-6' stands.  MA-3a (Qual keeps [udelta],
                 passes [uphi]) is consistent only if delta is NOT a phi-feature; MA-3b has
                 no such condition (T42b).
EP-8 PHASAL-v    A verbal phrase is a (strong) phase only if its v has phi-features (an
                 object to agree with) and an external argument (MI pp. 106-108), i.e.
                 v* = "transitive v or experiencer" (DbP n. 8).  Unergatives: neither text
                 says; by MI's criterion, no.
RuleRests onStatus
EP-0DbP (2), (3a); MI pp. 121–123SOURCED definitions; the restatement of MA-0′ is HYPOTHESIS
EP-1MI (5a), (6), (24), (45b); DbP p. 7SOURCED (as a statement of MI/DbP)
EP-2EP-1 + HMP (7b), (16a) printed placeholders + IN-8HYPOTHESIS. It follows from SOURCED definitions plus SOURCED forms, but the mapping of INCH onto “head with an EPP-feature” is ours
EP-3MI pp. 104, 106; HMP n. 2, n. 4; MA-4′; IN-5′HYPOTHESIS. It turns IN-9's “otherwise” and “never Spec,v” clauses into theorems. Every printed HMP form fits
EP-4HMP §2; MI p. 109HYPOTHESIS, a second, independent fit for EP-2
EP-5MI (42), pp. 124, 126, 128; DbP (3); HD2HYPOTHESIS; needs an added feature on ρ; OPEN vs MA-5a
EP-6MI (21); DbP (11), p. 14OPEN between PIC1 and PIC2; no string separates them
EP-7MI p. 124SOURCED principle; the consequence for MA-3a is HYPOTHESIS
EP-8MI pp. 106–108; DbP n. 8SOURCED criterion; its application to Navajo ‘go’ is HYPOTHESIS

5. What changes, rule by rule

RuleBefore (2026-10-19)AfterWhy
MA-2′(c)Placeholder = EM of an expletive satisfying the EF of the Qual-layer INCH, the phase head's ownMA-2″(c) = EP-2: pure Merge of an expletive into the EPP position of PINCH, a non-phase head at the top of the V-level. Still a Merge siteMI (5a), p. 109; DbP p. 7; 2008 p. 157 (T, inherited) all put expletive Merge on a non-phase head. HMP's printed placeholders force the move
MA-2′(a)Bump = EM of the PP high, or IM by the Qual-layer EFIM (Move) to PINCH's EPP position (EP-2). The participant is first-merged in its role position (MI (6)), so the bump has to be MoveMI (6): roles come only from first (pure) Merge; MI p. 109
IN-8INCH has a P-like requirement; Pnull carries itUnchanged, now supported over MA-2′(c)'s Qual locationEP-1, EP-4
IN-9Stipulates the placeholder “otherwise” and “a slot in Spec,v is never ρC”Both clauses derived (EP-3 (a), (b)). One new schematic prediction (EP-3 (c))MoM; c-command
IN-7 (D), MA-5′Default when no eligible slot; MA-5a (phase) vs MA-5b (inactive)MA-5b = EP-5, with MI's own Icelandic mechanism, at the cost of an abstract activating feature. Still OPEN against MA-5aMI (42), pp. 126, 128; DbP (3)
MA-1′M/S reaches only the v*P edge (phase)Two versions, PIC1 and PIC2 (EP-6). Same stringsDbP (11), p. 14
MA-3′MA-3a vs MA-3b OPENStill OPEN. MA-3a now requires δ ∉ φ (EP-7)MI p. 124
MA-6′Hiraiwa's Multiple Agree excluded (n. 12)Excluded with an earlier source too (MI p. 124)EP-7
MA-0′Three-way exponent test (Agree / IM / EM)Same test, now stated through feature types (EP-0)DbP (2)

HYPOTHESIS MA-3′'s selection condition is untouched. With the placeholder moved down to PINCH, the Qual layer keeps its role as D1's phase head. Its Agree feature still goes to the head it selects, and M/S still agrees with the subject in a ‘start to’ transitive (HMP (2b) -sh-, 1sg). Had the placeholder stayed in the Qual layer, the qualifier heads of INCH (n, ii, directly above M/S) would have had to be both the phase head and a T-like head at once. MI does not allow that combination, which is the clash this page removes.

6. Derivations EP1–EP8

Steps are labelled by rule set: Σ role schema, A word (D1), B clause (D2). Each step names its operation: pure Merge (EM), Move (IM), Agree or Spell-out. All forms are already on the site with their sources. No new segmentation is introduced.

EP1. HMP (7b) ʼawééʼ ʼináʼniichah ‘the baby starts to cry’: an expletive at a non-phase head (as HD3, MA7′)

  1. Σ Σ(√CRY) = {ρA} (sentient, no endpoint → Spec,v by LINK-1); the complement is the conflated noun cha, not a slot (H&P/HMP). φ(ρA) = 3.
  2. A pure Merge: [V N cha V]; REP ná; INCH's ρI in the object position (IN-1); PINCH with its obligatory EPP-feature (EP-2, EP-4).
  3. A EPP of PINCH: no slot below PINCH (EP-3 (a)), and ρA is not merged yet and will sit above it (EP-3 (b)), so Move is impossible → pure Merge of ρD ⟨E none, φ 3, δ INDF⟩ into the non-θ position (EP-1 (b), (c)). This is MI's “Merge of expletive” (45b-i), at a non-phase head as MI (5a) requires.
  4. A pure Merge: v; ρA in Spec,v (VPISH); M/S; Qual [n ii]. Agree: M/S's closest goal is ρA (above PINCH-P) → 3rd, Ø.
  5. A Spell-out D1 (as HD3): ’ ⌢ ná ⌢ ’ ⌢ nii ⌢ … ⌢ chah → ʼináʼniichah ✓ (unsegmented by HMP; alignment the site's).
  6. B ʼawééʼ links to ρA. ρD has no role in the action, so no noun can link to it.
  7. Change from MA7′: same string; the expletive's head is PINCH, not the Qual phase head.

EP2. HMP (16a) ʼiʼniichííl ‘it started to snow’: the same, with no participant at all (as HD7)

  1. Σ Σ(√SNOW) = ∅.
  2. A pure Merge: V; ρI; PINCH. Nothing to Move → pure Merge ρD (EP-3 (a), EP-4). v (no specifier; not a phase, EP-8); M/S; Qual.
  3. A Agree: M/S's only candidate is ρD. If ρD is inactive after PINCH's Agree, M/S shows the default (EP-5); if it is hidden by a phase (MA-5a), likewise. Third person either way.
  4. A Spell-out: ’ ⌢ ’ ⌢ nii ⌢ chííl ✓ (as HD7).

EP3. HMP (2b) bi-ná-ʼnii-sh-dl[į́į́]h ‘start to drink it’: Move, and why the placeholder cannot be used (as HD1)

  1. Σ {ρA (drinker; Proto-Agent) = 1sg, Spec,v; ρX (what is drunk; Proto-Patient) = 3, Comp,V}. Both roles come from first Merge (MI (6)).
  2. A pure Merge: V dlį́į́h + ρX; REP ná; ρI takes the object position; PINCH.
  3. A Derivation (i), ρD not in the subarray: PINCH's EPP → Move ρX (the closest slot below, EP-3 (c)) to its position; Agree PINCH ⟶ ρX (its object agreement; ρX then inactive, EP-5).
  4. A pure Merge: v* d; ρA in Spec,v*; M/S; Qual. Agree: M/S ⟶ ρA, 1sg → sh. Under PIC2, ρX would be visible, but ρA is closer (and ρX is inactive anyway).
  5. A Spell-out (as HD1): bi ⌢ ná ⌢ ’ ⌢ nii ⌢ sh ⌢ d ⌢ dlį́į́h ✓ (HMP's segmentation; bi overt because oblique).
  6. A Derivation (ii), ρD in the subarray: pure Merge ρD preempts Move (MI p. 106). ρX stays in Comp,V. Its exponent position is the object position, which ρI holds, and it cannot adjoin there (MA-4′). Fusion is not available, since ρX is not an indefinite ’. ρX is left with no position → crash. So *’ ⌢ ná ⌢ … (placeholder plus a definite object) is predicted not to exist. This is schematic. The printed form is (i).

EP4. HMP (10a) shináʼniidlóóh ‘I start to get cold (about dawn)’: Move, then a default from an inactive goal (as HD2, MA5′)

  1. Σ {ρ1} (change of state; Proto-Patient) first-merged in Spec,V; φ = 1sg.
  2. A pure Merge: V dlóóh, ρ1; REP ná; ρI; PINCH. Move ρ1 to PINCH's position (EP-2); Agree PINCH ⟶ ρ1 (1sg). ρ1 is now inactive (EP-5, if ρ carries [uK]).
  3. A pure Merge: v d (unaccusative: not a phase, MI pp. 106–107, EP-8); M/S; Qual.
  4. A Agree: M/S probes. Its closest matching goal is ρ1, which is inactive. By MI (42) it blocks the search, and nothing else is in range anyway. → default, MI's “[3person] (the default choice)” (p. 128) = IN-7 (D). MA-5a gives the same result by making the phrase a phase.
  5. A Spell-out (as HD2): shi ⌢ ná ⌢ ’ ⌢ nii ⌢ (M Ø) ⌢ d ⌢ dlóóh ✓ (alignment the site's; HMP do not segment (10a)).
  6. A Failed alternative with ρD in the subarray: ρD preempts Move; ρ1 stays in Spec,V. ρD (closer, inactive) blocks M/S → default, so ρ1's 1sg is spelled nowhere: the object position is held, and M/S is not available → crash (schematic). The prediction is that no ‘start to’ unaccusative pairs a placeholder ’ with the participant spelled somewhere else (T42d). HMP's 5 + 8 unaccusatives all shift (u10), which fits.

EP5. HMP n. 4 sodiʼniiszin ‘I started to pray’: the fusion escape (as HD6)

  1. Σ theme so=d-l-zin; ρA = 1sg in Spec,v; nothing below v (HMP's “cf. (7b)”).
  2. A PINCH's EPP: nothing to Move → pure Merge ρD (EP-3 (a)). IN-5′'s blocking clause: Position I is held by so, so ρD is pronounced in the object position and fuses with ρI (’ + ’ → ’).
  3. A Agree M/S ⟶ ρA (1sg). Spell-out (as HD6): so ⌢ di ⌢ ’ ⌢ nii ⌢ … ⌢ zin ✓.
  4. EP-2 adds nothing to why so blocks the left edge (T39c stays open). The relation between PINCH's EPP position and where its filler is pronounced is a spell-out matter (C2), not a syntactic one. HYPOTHESIS.

EP6. Hale 2001 (24a) shidoołhosh ‘He/She will tickle me’: PIC1 vs PIC2 (as MA1′)

  1. Σ {ρA = 3, Spec,v*; ρX = 1sg, Comp,V}.
  2. A pure Merge: V ghosh + ρX; v* ł; ρA; M/S; Qual d.
  3. A Agree, PIC1: VP is sealed once v*P is complete, so M/S sees only ρA → 3. PIC2: M/S can see ρX (DbP p. 14), but ρA c-commands it and matches, so ρA is closest (MI (40c)) → 3.
  4. A Spell-out: sh ⌢ d ⌢ w ⌢ ł ⌢ ghosh ✓ (Hale's segment line).
  5. Result: both give the string. They come apart only if Spec,v* held a slot that does not match M/S's probe. No such form is on the site (T42a).

EP7. Willie (31f) hojishxash ‘That person bit that person’: the top head and the φ-unit (as MA10′)

  1. Σ ρA = ⟨3 (zero), δ = 4⟩; ρX = ⟨3 (zero), δ = 4⟩.
  2. A pure Merge through Qual. Agree M/S ⟶ ρA (φ = 3, zero). Under PIC2, M/S could see ρX, but ρA is closer.
  3. A Qual (C-like) cannot see VP under PIC1 or PIC2 (DbP p. 14: “with Σ = ZP (so that Z = C), the probe Z cannot access the domain YP”). So δ at the top can only be ρA's (MA-3a or MA-3b), as before.
  4. A MA-3a as written splits ρA's features: φ to M/S, δ to Qual. By MI p. 124 a φ-set is valued as a unit. So MA-3a is coherent only if δ (the fourth person) is not a φ-feature. The site's notation already writes it apart. If the fourth person were a person value, M/S would have to spell it, and here M/S is the plain third-person zero (Willie gives no segmentation; alignment the site's, as SD7).
  5. A Spell-out: ho ⌢ ji ⌢ … shxash ✓.
  6. Result: MA10′ unchanged. MA-3a vs MA-3b now turns partly on whether the fourth person is a person value (T42b).

EP8. W&J (50b) dibé łį́į́ʼ bitah yíghááh ‘The horse went among the sheep’: is ‘go’ a phase? (T41h; as MA4′)

  1. Σ ρA (the one who goes) in Spec,v; ρX = the postposition's object; both 3; inverse: ρX [top].
  2. A pure Merge P -tah + ρX; Agree inside PP → bi (IA-3).
  3. A v: is it v*? MI p. 108 limits verbal phases to “transitive v with φ-features and EA”, and DbP n. 8 to “transitive v or experiencer”. ‘Go’ has an external argument but no object for v to agree with (ρX is P's). So by both criteria v is not a strong phase and has no EPP-feature of a phase head (EP-8).
  4. A Consequence for IA-1*: if the topic slot moves at all, the mover must be Qual (the C-like phase head). Under MI (24), Qual's EPP can be met only by raising from inside its phase, and ρX is inside. That is allowed if the PP is not itself a phase (MA-5b) and blocked if it is (MA-5a). IA-3 (no movement) remains cheaper.
  5. A Spell-out D1: bitah ⌢ yíghááh; B łį́į́ʼ → ρA, dibé → ρX ✓ (as MA4′).
  6. Result: T41h is partly answered. By the definitions, ‘go’ is not v*. Whether Navajo's v for ‘go’ has any φ-features of its own is a data question.

7. What each side gets right and misses

SideGets rightMisses or costs
Grant's typology noteThe EPP/Edge Feature line is a faithful one-line statement of MI/DbP: heads differ in whether they “require, allow, or disallow” a specifier (DbP p. 41)It uses 2008's name for a 2000/2001 notion. It does not mention the phase-head/T split (MI (5a), p. 109), which turned out to be the part that matters most for Navajo
Hiraiwa 2001(2) is MI (40) verbatim; (3) is a fair gloss of MI (42); (4) is DbP (46) with his addition marked(5) quotes DbP's wording under an MI citation. (2a) “identity” is the version Chomsky himself narrows on p. 124
The originals page (2026-10-19, 8:46pm)The v/v* split, inheritance, and the placeholder as an expletive. MA-6′'s exclusion, now also supported by MI p. 124MA-2′(c) placed the expletive at a phase head, which MI/DbP exclude. Its own “mismatch noted” line (Chomsky's T is not a phase head) was the symptom. MA1′'s reason was PIC1 only
HMP / IN-8 (2026-10-19, 12:46pm)Locating the requirement on a postposition-like piece (“behaves like an ‘incorporated’ postposition”), which is the location MI/DbP allowIN-9 stated the placeholder's distribution as a list; EP-3 derives it
MI/DbP applied to NavajoEPP-type distinctions (obligatory vs optional; who may insert an expletive) organize the ‘start to’ facts. Defective intervention gives the default. MI's θ vs checking split matches the site's slot-and-role designActivity depends on structural Case, which Navajo slots do not show. Neither text classes unergatives. Neither says anything about structure inside a word, so every application here is an extension

8. Grant's frame

GRANT'S CONJECTURE Word-internal Merge or Agree sites. After this check, one of the two certain word-internal Merge sites, the placeholder, has a sharper address. It is at PINCH, low in the word, not at the top. The bump is Move to the same position. Both are Merge (they build structure), and they sit at the edge of the verb phrase, not at the word's outer edge. The Agree sites are unchanged (M/S, P's own pronoun, and now PINCH's object agreement). Merge builds structure; Agree values features such as person and number. Nothing on this page contradicts the conjecture.

GRANT'S CONJECTURE Feature bundles and the weak thesis. MI's and DbP's goals are feature sets (“the φ-set of N”), not words. Every Agree on this page is with a slot bundle ρ in D1, before any noun arrives. The placeholder is a bundle with features and no role, and nothing in MI or DbP requires it to be a noun. That supports the weak thesis (roles live on slots, nouns link later) and says nothing about the strong one. One new point: if EP-5 is right, Navajo slots need an activating feature that plays the part Case plays for MI's nouns. If Grant's bundles are the arguments, that feature belongs in the bundle (ours, HYPOTHESIS).

HYPOTHESIS Dowty, and theta vs event structure. No participant's role in the action changes. Roles come from first Merge (MI (6): pure Merge in a θ-position is for arguments only), and INCH's position is non-θ. The ‘start to’ layer adds an onset to the event and a position, but no role. MI p. 127 already treats θ-structure and agreement as separate systems, so the theta roles and the event-structure layer stay compatible descriptions of the same participants.

GRANT'S CONJECTURE STRATA (theta roles first, event-structure roles as another layer, noun selection later). On this page the order is forced. Roles by first Merge come first; PINCH can only Move a slot that is already there (EP-3); and nouns link in D2. MI's MoM adds a fit: the event layer's own element (ρD) is preferred over moving a participant whenever the result converges. As before, STRATA is not separated from the adjunct and inside-the-clause views, because none of them predicts a different string here.

HYPOTHESIS The changeover from rule set A to rule set B. DbP pp. 13–14: a strong phase's domain is spelled out at the phase, and its head and edge “are spelled out if they remain in situ. Otherwise, their status is determined … at the next strong phase”; (10): a phase “is interpreted/evaluated at” the next one. If D1 is a C-like phase, the word's interior is handed to rule set A's spell-out at D1, and what stays at the word's edge is evaluated again at D2 by rule set B. That matches “residuals read at D2” on the Merge/Agree page §11.2. This is a fit, not a test. Where the word ends is still open (Hale 2003 n. 12).

9. Tasks T42a–h and T41 statuses

TaskWhat it needs
T42aPIC1 vs PIC2 (EP-6): a transitive whose Spec,v* slot is not a matching goal for M/S. Does M/S then ever spell the object's person? (Young & Morgan paradigms)
T42bIs the fourth person a person (φ) value or a separate deictic feature? Any form where M/S itself spells a fourth-person subject differently from third would make δ part of φ and rule out MA-3a (EP-7)
T42cEP-5 vs MA-5a: is there any evidence that Navajo slots carry an activating feature (any case-like distinction), or a form where a slot already agreed with by a P is still found by M/S? (joins T41c)
T42dEP-3 (a) / EP4 prediction: no ‘start to’ unaccusative with a placeholder ’ and the participant spelled elsewhere; no transitive with a placeholder and a definite object (schematic). Check HMP 2000 (printed) and Young & Morgan 1980G:320–321
T42eEP-3 (c): a ‘start to’ form of a verb with two slots below v (e.g. a ditransitive or a verb with Spec,V and Comp,V). Which one is bumped? Prediction: the closer (higher) one
T42fIs PINCH's agreement φ-complete (does the bumped prefix show number, e.g. a plural participant)? EP-5 needs it to be. Is ρD a there-type expletive ([person] only) or it-type? Does a placeholder ever show anything but the indefinite ’?
T42gIs INCH's requirement obligatory in every mode (EP-4)? HMP's (2) shows five modes, all with the prefix; check the remaining modes and aspects in Young & Morgan
T42hChomsky 2004 “Beyond explanatory adequacy”: the step from the EPP-feature to the edge feature, and whether the phase-head/T split on expletives survives it (2008 p. 157 suggests it does)

T41 statuses after this page: T41e answered (this page). T41h partly answered (EP8: not v* by the definitions; the data question remains). T41c sharpened: MA-5b now has its sourced mechanism (EP-5) at the cost of an activating feature (→ T42c). T41a unchanged; MA-3′'s selection condition is untouched by EP-2. T41b unchanged, but MA-3a now depends on T42b. T41d, T41f and T41g are unchanged.

What else changed on the site

Update notes on the borrowed terms checked against their originals (lede, MA-2′(c), MA-1′, MA-5′, §9), ‘Start to’ in HMP's own words (rules IN-8, IN-9) and Merge and Agree inside the word (lede). There is a new row on open questions (T42a–h), a sources section (added 2026-10-20, 4:46am), and a home-page entry. No verb-dictionary entries changed and no new Navajo forms were added.