Navajo “phrasal prescription” — theme, base, preverb, postposition
What’s claimed and how well supported
- Labels on this page
- SPECULATIVE ×6
- Strongest label used
- SPECULATIVE
- Main caveat (quoted)
- The chemistry half is ◆ SPECULATIVE pedagogy only.
Counted automatically from the labels printed on this page. Counts are a reading aid, not a grade of the page.
Terms on this page (4), plain-language definitions
- Valence (chem.)
- Bonding capacity / oxidation-related electron count of an atom in context.
- Dopant
- Intentional dilute impurity meant to tune electronic/optical/catalytic behavior without replacing the host as majority phase.
- Valency (ling.)
- Arity: how many actants a predicate takes; formal vs semantic valency can diverge.
- Host (matrix)
- Majority element/phase whose lattice dominates the material.
Definitions quoted from the Chem for linguists primer glossary.
Date: 2026-10-04 PT (rev.; first filed 2026-10-03 PT)
Status: Draft. Linguistics map first. The chemistry half is SPECULATIVE pedagogy only.
Audience: syntax/semantics (less chemistry).
Does not: announce a joint field, rewrite Bridge A, invent Navajo forms, glosses, or trees, invent concentrations or other chemical numbers, or treat Grant’s frame as a finding.
Prior project notes (not primary literature): Lexicalization; Derivational movement; Causative split; Zipf; secondary index. This note does not contradict sourced claims in those files. Where names for “base” and “theme” disagree across authors, both uses are recorded.
Abstract — sourced. Athabaskanists do not have one English label “phrasal verb” for a single Navajo unit. What they list is a verb base and a verb theme, but the names are swapped across authors (see §1). Hale treats a preverb and the verb stem as one lexical constituent inside one verb word, discontinuous on the surface in the way a Germanic particle verb is; Chee and Yazzie, and Chee 2025, use “phrasal verb” for a different object, a postposition plus a (often minimally inflected) verb word, which can also feed nominalization. That listed or lexical unit can carry valenceBonding capacity / oxidation-related electron count of an atom in context. (via the classifier) and can be used without an independent lexical noun phrase: subject person is an exponent on mode, and nominal arguments are optional. It is not a grid of empty slots that mode/tense then fills. Mode is obligatory inflection, fused with subject and often with stem shape. Navajo is not fully tensed. The picture “a prescription, with dopantIntentional dilute impurity meant to tune electronic/optical/catalytic behavior without replacing the host as majority phase. levels at places, realized by VPISH inside a tense” is not what these sources say. Young 2000’s own Theme→Base layering, as quoted below, matches Smith, Perkins & Fernald and does not match McDonough’s report of how the 1987 dictionary labels the listed classifier+stem piece.
Abstract — HIS, SPECULATIVE, not a finding. Grant’s frame, not a joint field: molecule ↔ phrase, atom ↔ word, quark ↔ morpheme (he also said maybe anyons; this note does not choose). On his view morphemes show up only by breakdown of a larger unit. His question: is a Navajo phrasal verb like a prescription for a molecule, with certain levels of dopant at certain places, realized through something like VPISH inside a tense? He says this is abstract. The sourced half above is the answer. A short speculative section below may liken a theta slot to a place a typed participant occupies. It does not assign levels or concentrations.
- 0. Constraints already in the corpus (one sentence each)
- 1. What the unit is called
- 2. Argument positions, without independent lexical items
- 3. Mode is not “the tense where VPISH realizes the prescription”
- 4. Morphemes “only by breakdown” — compatible description, not a particle claim
- 5. SPECULATIVE — places, not levels
- 6. Sources opened for this pass
- 7. Refused
0. Constraints already in the corpus (one sentence each)
Williams et al. 2015 “phrases” are random mixed-length contiguous spans inside punctuation-bounded clauses, not syntactic phrasal verbs (Zipf). Hale: Navajo has no verbal noun incorporation; Northern Athabaskan does; Navajo allows restricted noun-into-postposition incorporation (PA essay, as recorded in Lexicalization). On the drawn pronominal-argument analysis, arg1 merges in Spec,vP and does not move; the subject exponent is inside the verb word (Derivational movement). Bridge A (concentration windows ↔ valencyArity: how many actants a predicate takes; formal vs semantic valency can diverge./saturation) is a different strand; see Solubility windows; unrevised here.
1. What the unit is called
There is no Athabaskanist term that simply means “the Navajo phrasal verb.” Four labels are in the sources opened for this note. They are not synonyms.
| Label | Who uses it, for what | Grain |
|---|---|---|
| Verb base | McDonough, reporting how Young & Morgan’s dictionary actually builds words: classifier + stem. Listed; “not productive.” | Smallest listed verbal piece. Final syllable of the core verb. |
| Verb theme | Same McDonough paper: not a technical word-formation term in Young & Morgan. A looser name for the pieces that make the lexical meaning of a full word, “usually excluding agreement marking.” She also once folds a Model Paradigm cell into that loose unit. | Meaning package larger than, or other than, the bare base. Not a formal slot. |
| Verb theme / verb base (the other layering) | Smith, Perkins & Fernald 2007, citing Young & Morgan 1987: theme = classifier + stem, the basic lexical entry; base = that theme plus adverbial derivational prefixes; verb word = pronominal and mode prefixes plus the base. Footnote: sometimes the classifier is not thematic, and often other prefixes are. | Same pieces, names swapped relative to McDonough’s “base.” |
| Preverb + stem | Hale 2001, citing Rice 2000: one lexical constituent. Surface order splits them (preverb at the left edge, stem at the right), “much in the manner of” Dutch and German verb-particle combinations. Hale does not call this a phrasal verb. | Inside one verb word. |
| Phrasal verb | Chee & Yazzie (and Chee’s related child-speech description): a postposition plus a verb, often a short third-person form. Can add an object or an indirect object, or shift the meaning. About 17–19% of verb tokens in their child/caregiver sample occurred with a postposition. | Two pieces. Not the verb theme. |
Naming split, partly closed for Young 2000; still two dictionary-report uses. McDonough’s opened guide is explicit that Young & Morgan’s 1987 word-formation engine is Model Paradigm + verb base (classifier+stem), and that “verb theme” is not what that engine uses. Smith, Perkins & Fernald, citing the same 1987 book, use “theme” for classifier+stem and “base” for theme plus derivational prefixes. Young 2000 The Navajo Verb System: An Overview was still not opened as the book. A long quotation of Young 2000 pp. 57–59 (retrieved via Acta Lingweenie, 17 September 2011 post “Natlang Inspiration: Navajo WOD,” which reprints those pages) uses the Smith / Perkins / Fernald layering:
- Verb Theme = classifier + stem. Young’s printed example in that quotation: with Ø-Classifier the Verb Theme ØWOD is produced from root WOD (“flex, bend”).
- Verb Base = that Theme plus an adverbial derivational prefix. Same quotation: Theme ØWOD + adverbial 'ahá- (“apart”) yields Verb Base 'ahá-øwod (“bend apart, become disjointed”). With Ł-Classifier the Theme is transitivized (“cause to bend”), and the same adverbial yields a corresponding causative-transitive Base.
So Young 2000’s own Theme / Base chapters, as quoted, align with Smith, Perkins & Fernald, not with McDonough’s report of the 1987 dictionary’s “base” label for classifier+stem alone. Both McDonough’s report of 1987 practice and Young 2000’s overview labeling are real; they disagree on which English word names the classifier+stem piece. The Wikipedia “Navajo grammar” summary of Young 2000 says the same Theme→Base→inflection layering; it is a secondary digest and is not treated as a substitute for Young’s pages. Neither naming is Grant’s “phrasisized,” and neither is a stored English string of free words.
Chee 2025 adds a third “verb base” use — do not collapse it. Chee’s Journal of Child Language article (full text opened this pass) says the Navajo verb is traditionally divided into three domains — verb theme, conjunct prefixes, and disjunct prefixes — citing Young & Morgan 1987, Young, Morgan & Midgette 1992, and Young 2000. Her Table 1 labels the classifier+stem columns VERB THEME. Separately she writes that the minimal bisyllabic verb, “also referred to as the verb base (Young & Morgan, 1987; Young, 2000), is composed of the verb stem and the prefix string.” That “minimal word = base” use is not identical to Young 2000’s Theme+adverbial Base in the WOD quotation, and not identical to McDonough’s classifier+stem “base.” Recorded; not collapsed.
Preverb, not “phrasal verb,” for the verb-internal case. Hale 2001: the stem and the preverb are a unit in the lexicon (Rice’s point, which he accepts). The preverb is an adjunct, not a head. The object, on the pronominal-argument reading he assumes, is a complement of V. Heads in the extended projection (V, Voice, Mode/Subject, Qualifier) are spelled into a bipartite phonological skeleton. He argues that this spell-out, not head movement, puts the stem at the right edge and the preverb at the left. Rice’s alternative, as he reports it, does raise the verb to Voice and then to Mode/Subject. That disagreement is about how a lexical constituent is linearized. It is not a claim that the constituent is a chemical prescription.
“Phrasal verb” is used, for the other construction. Chee & Yazzie reserve that English phrase for postposition + verb in ordinary speech: the postposition can add an argument to an intransitive or an indirect object to a transitive, and some pairings shift or extend the meaning. They also say the template is an analytic chart. Fusion blurs prefix boundaries, and the spoken word is not the sum of the separated pieces. That supports a careful version of “morphemes show up by taking a larger unit apart.” It does not make the postposition+verb pair into a listed molecular formula, and it does not identify that pair with the verb theme or with Young’s Verb Base.
Chee 2025 (full text) sharpens the same label, still not a prescription. Melvatha R. Chee, “Navajo Verbs in Child Speech,” Journal of Child Language 52(5):1010–1035 (2025), DOI 10.1017/S0305000924000229 (Cambridge Core PDF opened). She argues that children prefer minimally inflected, often bisyllabic, third-person verb words not only because they are simpler, but because those forms are productive as main verbs and are used to create phrasal verbs and nouns. Her “phrasal verb” is again postposition + verb, with meaning shift or extension. Examples she prints (her glosses; not re-analyzed here): nizhóní ‘s/he/it is beautiful’ beside shił nízhóní ‘I like it’; bééhózin ‘it is known’ beside shił bééhózin ‘I know it’; hózhǫ́ beside baa hózhǫ́; hólǫ́ ‘it exists’ beside shee hólǫ́ ‘I have’. She also derives nouns by adding a nominalizer to such a pair: adą́ ‘eating’ + bikáá’ ‘on top of it’ → phrasal verb bikáá’ ádą́ ‘eating on top of it’ → noun bikáá’ ádą́ní ‘table’ (with inserted -n- for CV structure); similarly dah naat’a’ → dah naat’a’í ‘flag’, na’oolkił → na’oolkiłí ‘clock’. So the postposition+verb unit is real in discourse and in derivation. It is still two pieces, not Young’s Theme→Base layering inside one verb word, and it is not a dopant-level prescription realized by VPISH in a tense.
Hale’s earlier point, already on file, stays one sentence: this postpositional material is not verbal noun incorporation. Navajo lacks that; incorporation of a bare noun is into the postposition, and it is restricted.
2. Argument positions, without independent lexical items
Partly yes. Not as an empty prescription.
Smith, Perkins & Fernald: a well-formed Navajo clause may be only a verb word. Nominal arguments are optional. Pronominal prefixes (subject and object) and a mode prefix are added to the base; they are not the base.
McDonough, on Young & Morgan’s practice: subject markers are not independent prefixes in the word-formation system. Mode (and she writes that there is no tense in Navajo) is inflected for person and number. Subject marking is an exponent on mode. The core verb is that mode–person cell plus the verb base. She also says valence is not a property of the mode paradigms. It appears when the classifier+stem is attached. Young & Morgan, as she reports them (their p. 117), split the classifier’s work into valency, where it marks argument structure, and theme, where it does not. The boil pair she reprints is the familiar one: yibéézh ‘it’s boiling’ beside yiłbéézh ‘he’s boiling it.’ She does not treat ł as productive in general. The causative scout already separates that short alternation from the longer unergative causative and from listed thematic ł. This note does not redraw it.
So the listed base can specify transitivity without a stack of lexical noun phrases. The person who saturates the subject role is, in this descriptive system, the person-number cell of the mode paradigm, inside the verb word. Hale 2001, assuming Jelinek’s hypothesis, goes further: the object prefix is the object, and third-person subject lives in the Mode/Subject portmanteau; a full noun phrase is an adjunct, like the deictic “subjects,” which he says are not the real subject. The movement note already records the live disagreement (Speas / Perkins / Platero: the prefixes are agreement). Either way, the saturator that is always there is morphology on the verb, not a required independent lexical item.
What that is not. The theme/base is not described as a list of vacant theta positions waiting to be filled. McDonough’s loose “theme” usually excludes agreement, except where a particular prefix string has been lexicalized into one dictionary meaning (her bi’dishnééh entry: the indirect-object and indefinite-object pieces are agreement and still part of that entry’s meaning). Those are listed formatives, not open sites. Hale keeps three relations apart: argument structure (verb and object), adjunction (preverb, iterative, plural), and the nuclear heads (Voice, Mode/Subject, Qualifier). A preverb is not an argument slot. A postposition, in Chee & Yazzie, can add an argument, and it does so as a separate word, not by mode conjugation.
3. Mode is not “the tense where VPISH realizes the prescription”
The picture in the question is not what the sources say.
What they say instead:
-
The lexical unit is already a unit, then it is inflected. McDonough’s formula, which she says is Young & Morgan’s: a Model Paradigm cell (mode, person, number) plus a verb base yields the verb word. The base is not the output of that combination. Stem shape also alternates with mode and aspect; for event bases the stem is not fixed across modes (Smith, Perkins & Fernald). That is conjugation of a listed piece, including how the stem is pronounced in that mode. It is not a tense node assembling a molecule from a recipe.
-
Mode is not tense. McDonough: there is no tense in Navajo; mode is what is inflected for person and number. Smith, Perkins & Fernald: every verb word has one mode (imperfective, perfective, progressive, future, customary/usitative, iterative, optative). Those modes are not one semantic category. Future is the only mode whose main job is temporal location, and even it is not a full tense: it is unavailable for state verbs, it occupies the single mode slot so it is not stacked on a viewpoint, and its absence means “no information about the future,” not past. Past and future particles sit outside the verb word and are optional. With no future mode, no particle, and no time adverb, temporal location is inferred from boundedness (bounded default past, unbounded default present). They conclude Navajo is not a fully tensed language.
-
Where syntax papers put Mode, the operation is not VPISH. Smith, Perkins & Fernald assume mode sits on a high projection, “the Navajo version of InflP,” and that this projection is a landing site for head movement of the verb. They say Hale 2000 is similar but not identical. That is movement of the verb head, if one adopts it. Hale 2001, opened here, rejects traditional head movement for the preverb–stem discontinuity and uses phonological spell-out. The project’s drawn analysis (Derivational movement) is the pronominal-argument one: arg1 is merged in Spec,vP and is not moved to a subject position outside the verb word. The flagged “raise arg1 to Spec of Mode” overlay was already marked as not drawn, because the opened PA essay does not use it. “Realized by VPISH inside a tense” stacks that refused step on top of a tense Navajo does not have, and on top of a recipe the lexicon notes do not state.
Rice 2000 itself was not opened. Hale’s report of her movement account is used only as his foil. No tree is added here.
4. Morphemes “only by breakdown” — compatible description, not a particle claim
McDonough: the template’s morphemes are culled from fully inflected words, and building words by concatenating the template overgenerates. Young & Morgan do not use the template as the word-formation device; paradigms and listed bases do. Chee & Yazzie: the template shows an abstract arrangement; the spoken verb is not equivalent to those pieces laid end to end, because prefixes delete, merge, or change. Children produce the stem early; the stem is the salient right edge.
That is a fair linguistic reading of Grant’s remark that morphemes show up by taking a larger unit apart. It is a statement about analysis and about fusion, not a finding that a morpheme is a quark, an anyon, or an atom. His quark-versus-anyon alternative is not chosen.
SPECULATIVE Cross-domain section. Open after the materials.
5. SPECULATIVE — places, not levels
⚠ SPECULATIVE pedagogy. Not a finding. Not a joint field. No concentrations. Bridge A untouched. Does not force a match the Athabaskan sources do not make.
The doping–theta notes already on file (secondary index; secondary index) treat theta roles as which participant types a predicate requires, with Navajo as the exhibit of those types on the verb rather than as a stack of lexical noun phrases. The stress-test’s best-fit family was substitutional electronic doping, and only as a cast: a hostMajority element/phase whose lattice dominates the material. lattice as the domain, a dopant as the introduced participant, carriers as something the useful reading entails, vacancies or compensating defects as optional spoilers. Solid-solution strengthening and precipitation were collapsed for that analogy, because those foreigners are not dopants in the honest vocabulary. Process clocks were not equated with mode or aspect.
If a picture is wanted after that map, a theta slot is like a kind of place a typed participant occupies — domain versus introduced participant versus an optional defect — and not like a level of anything at that place. The sources opened here never say the verb base, the theme, the preverb–stem pair, or the postposition+verb pair is a prescription of dopant levels. Mode/person conjugation is how a listed base is spoken as a particular verb word. It is not a diffusion step, and it is not VPISH. Continuous amount versus a discrete slot was already the collapse pressure in the typed-participant note. This note adds no number to it.
6. Sources opened for this pass
| Source | URL | Used for |
|---|---|---|
| McDonough, “How to Use Young and Morgan’s 1987 The Navajo Language” (Rochester CLS working paper) | https://www.sas.rochester.edu/cls/assets/pdf/working/fall-19/3mcdonoughsecondpaper.pdf | Base = classifier+stem; theme not a Y&M word-formation device; no tense; subject as exponent on mode; valence on the base, not on the mode paradigm; template is not the engine |
| Hale 2001, “Navajo Verb Stem Position and the Bipartite Structure of the Navajo Conjunct Sector,” Linguistic Inquiry 32(4):678–693 | https://ling-phil.mit.edu/papers/hale/papers/hale021.pdf | Preverb+stem as a lexical constituent, particle-verb comparison, spell-out rather than head movement; object as complement under PA. Navajo examples in the text layer lost diacritics; none are reprinted |
| Smith, Perkins & Fernald 2007, “Time in Navajo,” IJAL 73(1):40–71. DOI 10.1086/518334 | https://works.swarthmore.edu/cgi/viewcontent.cgi?article=1012&context=fac-linguistics | Theme/base/verb-word layering (names as they use them); optional nominals; one mode per verb word; not fully tensed; future vs particles; their Infl and verb-head landing-site assumption. OCR-damaged Navajo lines are not reprinted |
| Chee & Yazzie, “Comparing Child and Adult Language Use…,” LD&C Special Publication 30. http://hdl.handle.net/10125/74821 | https://scholarspace.manoa.hawaii.edu/server/api/core/bitstreams/02415785-e98a-489f-a502-f32dc56d3350/content | “Phrasal verb” = postposition + verb; template analytic; fusion. Their examples are their data, not re-glossed here |
| Chee 2025, “Navajo Verbs in Child Speech,” Journal of Child Language 52(5):1010–1035. DOI 10.1017/S0305000924000229 | Cambridge Core PDF (opened 2026-10-04 PT) | Same “phrasal verb” = postposition + minimally inflected verb; meaning shift; feed into nominalization (bikáá’ ádą́ní ‘table’); three domains theme/conjunct/disjunct; her separate “minimal bisyllabic = verb base” wording recorded, not collapsed |
| Young 2000 pp. 57–59 (Theme / Base), as reprinted in Acta Lingweenie (17 Sep 2011), “Natlang Inspiration: Navajo WOD” | https://acta-lingweenie.blogspot.com/2011/09/ (quotation of Young 2000) | Verb Theme = classifier+stem (ØWOD); Verb Base = Theme + adverbial ('ahá-øwod). Aligns with Smith/Perkins/Fernald naming. Young 2000 book PDF itself not opened |
| Wikipedia, “Navajo grammar” (secondary digest; cites Young 2000) | https://en.wikipedia.org/wiki/Navajo_grammar | Same Theme→Base→inflection summary; secondary only; not a substitute for Young’s pages |
Not opened as the book/PDF, so not quoted beyond the limits above: Young & Morgan 1987 as the book; Young 2000 as the UNM volume itself (only the pp. 57–59 quotation via Acta Lingweenie); Young, Morgan & Midgette 1992; Rice 2000; Hale 2000. Hale’s PA essay and the movement note remain the sources for “no verbal noun incorporation” and for “arg1 does not move,” as already filed.
7. Refused
- Inventing Navajo forms, glosses, trees, or a repair of Hale’s or Smith et al.’s damaged diacritics.
- Collapsing McDonough’s “verb base” with Smith, Perkins & Fernald’s “verb base,” or with Young 2000’s Verb Base, or with Chee 2025’s “minimal bisyllabic = verb base.”
- Treating Chee & Yazzie’s phrasal verb as the verb theme, or Hale’s preverb–stem constituent as their phrasal verb.
- Treating Williams et al. 2015 phrases as phrasal verbs.
- Any concentration, dopant level, or other chemical number.
- Choosing between quark and anyon.
- Identifying mode with tense, or mode with a process clock, or VPISH with the way a theme is realized.
- A joint field. A Bridge A revision. Inventing a Young 2000 quotation beyond the Acta Lingweenie reprint of pp. 57–59.
v0 — 2026-10-03 PT; rev. 2026-10-04 PT. Young 2000 Theme→Base (as quoted) matches Smith/Perkins/Fernald; Chee 2025 full text confirms postposition+verb “phrasal verb” and nominalization path. Prescription-in-a-tense picture still not the sources’ picture. No joint field.
Corpus note: navajo-phrasal-prescription-v0.md. REAL naming and unit map. HIS prescription frame is SPECULATIVE. Not Bridge A. Not merged with lexicalization, movement, the causative scout, Zipf, or the VPISH molecule picture.