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
The Navajo verb, start to finish: template, derivation, selection, Algonquian, phrasal verbs, and a chemistry check
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). Light d, used below, is withdrawn (Grant, 2026-11-14); light n is an OPEN structural question.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. Position-class template statements below describe the order of prefixes; they are not the derivation (§3). The “roles not words” theory, where it appears below, is dropped. 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. The old overview blended observed prefix order with proposed syntactic heads. Use the corrected vP/TP core: subject first in left Spec,vP, then independently test V→v→T and subject raising. Qual/M/S, light d, NP-PRO, A/B and the unique word/clause boundary are not axioms. The current boundary tests remain OPEN; no telicity or movement follows just from a prefix.
2026-11-11 audit pass 2. Wording that used VPISH as the name of a spine, tree, skeleton or set of slots named a structure after the hypothesis. VPISH is a hypothesis name, never a node; the structure is a vP/TP spine with the subject first-merged in left Spec,vP. Only the wording changed.
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. See the pass 3 report.
2026-11-11 audit pass 4. In the prefix-by-prefix table for nishhozh the prefix glosses now carry the same hyphen as the prefixes (2SG.OBJ-, MODE-, 1SG.SUBJ-, CL-), so form and gloss line up one-to-one. Gloss only; no form changed. See the pass 4 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.
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.
A stand-alone guide (added 2026-10-27, ~11:29am PT) for readers who want the whole picture in one place. It pulls together results scattered over this site, adds newly opened Algonquian material (Oxford's Routledge chapter on Algonquian, §§4.1–4.2), and links to the detailed pages for every claim.
In plain words: A Navajo verb is one long word built from a stem at the right end and up to about nine slots of prefixes before it. Those prefixes carry who did what to whom (as small pronouns), the tense-like "mode", and the voice. Because the pronouns are inside the verb, many linguists say the verb's real arguments are those pronouns and the free-standing nouns are extras. That is the closest the evidence gets to Grant's idea that verbs take "theta roles, not words" as arguments, and it is close but not the same: the slots are pronouns that carry a role and person and number, and some nouns still behave like true arguments. Algonquian languages (Ojibwe, Cree, Delaware) build verbs the same "word-as-sentence" way, with a different ranking system. Navajo does have "phrasal verbs" (verb + postposition pairs with a special meaning). The chemistry comparison at the end is speculative and finds no real match.
Status badges: SOURCED (from a source opened on this site; citation given) · HYPOTHESIS (our analysis) · GRANT'S CONJECTURE · OPEN. No Navajo or Algonquian form on this page is new: every form is copied from a source already used on a linked page, with its example number.
Notation. Rule set Σ = role schema step; A = word-level (left-headed) rules; B = clause-level (right-headed) rules (see headedness). vP/TP spine (VPISH base positions): V → v (Voice, the "classifier") → M/S (mode/subject portmanteau) → Qual (Hale 2003; derivations). MVS = Willie's (1991) "minimal verb sentence", the verb plus a postposition unit.
Bottom line, in plain words
SOURCEDThe verb is a template: a stem with ordered prefix slots I–IX, disjunct (I–III) and conjunct (IV–IX). Hale 2003:14–15 ex. (10), after Young & Morgan 1987:37–38. Morphemes of wider scope sit further right, nearer the stem (Rice, via headedness §2).
HYPOTHESISGrant's thesis in its literal form ("verbs take theta roles, not lexical items") is not established. The strong version fails on this site's diagnostics; the mainstream weaker version (the verb's arguments are role-bearing pronominal prefixes; nouns link to them) is supported but contested; the slots carry person/number as well as a role, so "arguments are roles" is a simplification (thesis page §6).
SOURCEDAlgonquian is the nearest relative in type, not in family. Both put the participants in the verb word; Algonquian ranks persons (1,2 > 3 > obviative > inanimate) and marks direct vs. inverse with a "theme sign"; Navajo has only the deep (3-on-3) kind of inverse, yi-/bi- (Navajo-only claim: HYPOTHESIS). The specialist literature on Algonquian has largely moved away from a pure pronominal-argument analysis (Oxford, Routledge chapter, citing LeSourd 2006).
SOURCEDYes, Navajo has "phrasal verbs" in a real sense: verb + postposition pairs whose meaning shifts. Chee & Yazzie 2025 use the term and give five pairs; Willie 1991 shows verb + P forms a tight unit (the MVS). That these pairs are stored as lexical units is HYPOTHESIS.
SPECULATIVEChemistry: no correlate found for any verb-internal phenomenon; Bridge A stays locked. Precipitation is the "best-fit" family for the typed-participant idea only as a teaching analogy. No scholarly literature joins Merge/Agree with chemistry.
1. The verb template
SOURCED Hale 2003:14–15 ex. (10), after Young & Morgan 1987:37–38 (Y&M itself not opened):
Position
Slot
Zone
What it does (sourced gloss)
I
Preverb (incl. postposition + object)
disjunct
Lexical/adverbial; forms one lexical unit with the stem though at the opposite end (Hale 2001:678, like Dutch/German particles)
Root + aspect host; stem shape varies by mode/aspect
SOURCED Kibrik 2005 segments a long form: nihideestsił 'I will move on the buttocks to a point' = ni-hi-di-ghi-s-l-tsił TERM-SERIATIVE-INCEPTIVE-PROGR-1SG-classifier-move.sitting (headedness §2.2, K2). Surface form and underlying slots differ a lot: prefixes fuse and delete.
SOURCEDStem and mode. Three modes express aspectual viewpoint: Perfective, Imperfective, Progressive; the Progressive is only for motion bases (Smith, Perkins & Fernald 2007 §3.1, quoted at distance-measure §3.3). The stem is root + an aspect host; -ł marks future and progressive (Hale 2003 n. 10). The full stem-set tables (which stem goes with which mode, verb by verb) are in Young & Morgan 1987, OPEN: not opened.
Merge vs Agree, briefly.SOURCED (Chomsky 1995, 2000 as summarized on intra-word Merge/Agree): Merge takes two pieces and builds a bigger one (structure). Agree builds nothing: a head with an unvalued feature (a "probe") finds a matching element (a "goal") and copies its value (e.g. person, number).
Σ The stem/theme fixes a two-role schema: tickler (proto-agent), tickled (proto-patient). HYPOTHESIS (the Σ step is this site's model, theta schema).
AMerge V (ghozh) with v/Voice (ł): the classifier is the voice head. SOURCED analysis: Hale 2003 bracketing (7).
AMerge the object pronoun (2SG) into the verb phrase as the inner argument; it is spelled ni- in slot IV. HYPOTHESIS: object = a Merge site at the edge (MA rules).
AMerge M/S (mode + subject) over v+V; Hale: M/S selects v+V. SOURCED.
Agree: the subject slot is valued 1SG (spelled sh-); the mode slot is valued imperfective-type (here Ø). These are valuations, not new structure. HYPOTHESIS (MA-4 "no Merge from slot").
A Qualifier (VI) would Merge above M/S if present (Hale: Qual selects M/S); none here.
Spell-out (CYC): prefixes surface right-to-left from the stem as ni-Ø-sh-ł-ghozh → nishhozh (ł + gh fuse). Phonology of the fusion: from Hale's form; we do not derive it.
B At clause level, an overt noun (none here) would link to a pronoun slot; Navajo clause order is right-headed SOV.
2.2 yiztał vs biztał: the yi-/bi- alternation
Willie & Jelinek 2000:273 (61a), (62b)
łį́į́ʼ
tsé
yiztał
horse
rock
3.OBJ(yi)-kicked
‘The horse kicked the rock.’
ʼashkii
łį́į́ʼ
biztał
boy
horse
3.OBJ(bi)-kicked
‘The boy was kicked by the horse.’
Σ Same schema both times: kicker, kicked. SOURCED the roles do not change (Algonquian inverse; Willie 1991).
Σ Choice: direct (yi-) or inverse (bi-); the choice is made at Σ on this site's analysis (inverse §5). HYPOTHESIS.
A Merge V+Voice, then object, then M/S, as in 2.1.
Hale's condition: yi- appears only when a real 3rd-person doer sits structurally above the object; otherwise bi- (Hale's general b-). SOURCED. With bi-, the patient is the higher ("topical") argument and the first noun is the kicked one.
B Nouns link by order to the pronoun slots: first noun ↔ higher slot. Two sources give the same noun order with opposite prefixes, so the "above" is between hidden pronoun slots, not spoken nouns. HYPOTHESIS on SOURCED data.
SOURCED Willie 1991:122 (20a–c): yishxash / bishxash show the same alternation (derivations D4). Whether binding reverses with bi- as in Algonquian "deep" inverse: OPEN.
3. What a Navajo verb selects, compared with other languages
Notion
Meaning
Navajo
English-type languages
c-selection (category)
Which syntactic category a head takes as complement
Inside the word: Qual selects M/S, M/S selects v+V (Hale 2003) SOURCED. At clause level the verb's "complements" are pronominal slots, not NP/PP categories HYPOTHESIS (PAH)
Verb picks NP, PP, CP complements (put needs NP + PP) — standard textbook background, OPEN(not opened here)
s-selection (meaning)
Semantic type/sort of argument
Strong and stem-coded: classificatory stems require round objects (Willie (22) yiyííbį́į́ʼ) or plural objects ((24) yiyííłtaʼ) (Willie 1991:61–62) SOURCED
Mostly via separate verbs/collocations (drink needs a liquid)
Theta roles
Participant roles (agent, theme…)
Carried by the pronoun slots; nouns are linked to them (PAH: Jelinek 1984; Willie 1991) SOURCED as analysis, contested
Assigned to NPs in argument positions
SOURCEDJelinek and Baker. Jelinek's Pronominal Argument Hypothesis: in languages like Navajo the affixes are the arguments and full nouns are adjuncts. Baker 1996 (polysynthesis parameter, Mohawk) makes a similar claim, but Baker does not treat Navajo as fully polysynthetic, so he cannot be cited as direct support for Navajo (thesis page; Speas on Navajo argues for more configurational structure).
GRANT'S CONJECTURE "Navajo verbs take theta roles, not lexical items, as arguments." Verdicts (thesis §6): T-STRONG (nouns are never arguments; roles alone decide linking) fails; T-MID/feature-bundle version survives with key tests pending; T-WEAK (= PAH) is mainstream but contested; T-ROLE plausible, not shown. The literal wording is not supported: the slots are pronouns that bear roles and also carry person and number.
4. How Algonquian fits
SOURCED From Algonquian inverse and Oxford's Routledge chapter (newly used here, §§3–4):
Template. Ojibwe: prefix · stem · theme sign · negation · central · mode · peripheral, e.g. ni-waapam-ikw-ssii-naa-token-ak 'they might not see us'. Unlike Navajo, the main inflection is suffixal.
Stem = initial + (medial) + final. The initial carries the core meaning, an optional medial is noun-like, the final sets category and animacy; Oxford notes the parallel to Root + v in Distributed Morphology. Compare Navajo stem + classifier (V + v).
Four verb classes set by the final (Table 3, Proto-Algonquian *pank-): AI 'fall (animate)', II 'fall (inanimate)', TI, TA. The text layer prints TI *pank-ihšim- 'to drop an animate' and TA *pank-ihθet- 'to drop an inanimate'; these glosses look swapped relative to the class labels. OPEN: flagged, not corrected.
Selection by animacy. The stem selects the primary object's animacy (TA vs TI) but not a secondary object's (Delaware (9) nə-ne·w-a·-w-ak 'I see them', nə-mi·l-a·-w-ak 'I give something to them'); finals agree in animacy with the logical object regardless of the hierarchy (Bruening 2005).
Secondary finals derive causative, benefactive, reciprocal, reflexive, detransitive stems (Table 6), comparable in function (not form) to Navajo classifier/causative changes.
Preverbs (Ojibwe (8), Nichols 1980): ni- kii= nanta= kikkentaan 'I sought to know it', ranging from concrete ('seeking') to grammatical ('past'); compare Navajo slot I.
Hierarchy and inverse. 1,2 > 3 > obviative > inanimate. 3→1/2 inverse is shallow (Agree only); 3′→3 is deep (the patient raises; binding reverses). Willie 1991:96–97: the Navajo inverse "functions like obviative marking". Navajo has only the deep type HYPOTHESIS.
Pronominal arguments? Affixes are sometimes treated as the arguments (Mithun 1999), but LeSourd 2006 argues against it and many specialists reject a non-configurational analysis because of structural asymmetries (Oxford). This weakens any "Algonquian supports PAH" argument for Navajo.
Syntax. Oxford 2019: v = final, Voice = theme sign; head movement gives Root-v-Voice.
5. Phrasal verbs and lexicalization
SOURCED Chee & Yazzie 2025:22, 34–35 call verb + postposition pairs "phrasal verbs": naashné 'I play' → bee naashné 'I make him cry'; ánáshwoʼ 'I usually walk away out of sight' → biká ánáshwoʼ 'I usually help him'; honiigaii 'The area became hotter' → bitah honiigaii 'He became sick, ill' (see postposition units). Willie 1991: verb + P is the MVS, not interruptible, the domain of the verb's subcategorization (p. 50 (44)).
HYPOTHESIS These pairs are stored as units, like English look after. Difference from English: the Navajo P carries its own object pronoun (bi-, yi-) and sits before the verb. Also distinct: Navajo's slot-I preverbs are lexically bound to the stem inside the word (Hale 2001:678). Whether Young & Morgan list the pairs as lexemes: OPEN. Related: interrupted MVS, PP by verb, instrument voice.
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.
7. SPECULATIVE: chemistry-overlap table
SPECULATIVE Bridge A is locked. No joint Merge/Agree × chemistry literature was found (merge-agree-chemistry-probe-v0). No row claims a joint field.