---
title: Where do the nouns sit? Condition C and the limits of Hale's argument
date: '2026-10-27'
updated: '2026-10-27'
authors: Grant Golland (research by Grok Bot)
abstract: 'Hale 1999 reports Navajo sentences in which a null subject cannot be the
  same person as a noun inside another phrase, and treats this as Condition C. If
  the reasoning holds, nouns sit within the subject''s reach, against the claim that
  they are linked from outside. This essay audits that argument. It is valid, but
  it depends on ten assumptions, none of them established for Navajo, and Speas reports
  a sentence of the same shape with the opposite judgment. The essay also records
  a correction: on 2026-10-15 the site said the feature-bundle hypothesis was ''wrong'',
  and that was too strong.'
status: 'HYPOTHESIS (D6: compatible, undecided; Hale''s argument valid but conditional
  on A1–A10)'
---

# Where do the nouns sit? Condition C and the limits of Hale's argument

*Grant Golland (research by Grok Bot)*, 2026-10-27. Status: HYPOTHESIS (D6: compatible, undecided; Hale's argument valid but conditional on A1–A10).

**Abstract.** Hale 1999 reports Navajo sentences in which a null subject cannot be the same person as a noun inside another phrase, and treats this as Condition C. If the reasoning holds, nouns sit within the subject's reach, against the claim that they are linked from outside. This essay audits that argument. It is valid, but it depends on ten assumptions, none of them established for Navajo, and Speas reports a sentence of the same shape with the opposite judgment. The essay also records a correction: on 2026-10-15 the site said the feature-bundle hypothesis was 'wrong', and that was too strong.


*Restated on the unified structure (2026-11-15): nouns are arguments merged in their theta positions, the subject raised to Spec,TP by the working default; where nouns sit for Condition C is to be re-run on that spine, and the adjunction analyses below are labelled contrast. See <https://navajo-decoded.grant-golland.workers.dev/unified-structure/#us-clause>. The essay’s data, judgments and citations are unchanged; it is kept as the record of what was argued then.*

## 1. The question

The feature-bundle hypothesis (FB, Essay 1) includes a clause called FB-5. It says a null bundle can never trigger Condition C on a noun, because the noun is linked to its slot rather than merged under it.

Hale 1999 reports two cases that seem to show the opposite **[SOURCED]**:

- In (10), the null subject of 'give' cannot be 'my father', who is named inside 'my father's very horse'.
- In (11a, b), the null subject cannot be 'your younger brother', who is named inside a clause under 'about it'.

In both cases the participants are people, so rank plays no part (pp. 172–173).

Do these cases show that nouns sit within the subject's reach?

## 2. Related data

Platero's coreference judgments are of the same kind. Coreference fails where a noun would follow and be c-commanded by its antecedent **[SOURCED]**:



> **(1)** *[ashkii bizhéʼé] ashkii yiyiiłtsą́ *  
> boy his:father boy see   
>   
> — Platero 1978:86, (3.18), (3.19); 87, (3.20) (page images checked; (3.20) at 400 dpi); copied from the working note “What comes first decides: Platero 1978 on who "his" can mean, and a no-movement route to W&J (70b) (T59a)” <https://navajo-decoded.grant-golland.workers.dev/possessor-command/>, ex. (8)





> **(2)** *[ashkii bííghahgi] ashkii tłʼiish yiyiiłtsą́*  
> boy beside boy snake see  
>   
> — Platero 1978:86, (3.18), (3.19); 87, (3.20) (page images checked; (3.20) at 400 dpi); copied from the working note “What comes first decides: Platero 1978 on who "his" can mean, and a no-movement route to W&J (70b) (T59a)” <https://navajo-decoded.grant-golland.workers.dev/possessor-command/>, ex. (9)





> **(3)** *Ashkii [ashkii bííghahgi] tłʼiish yiyiiłtsą́. *  
> boy boy 3-beside snake 3OBJ(yi).3SBJ.PFV.see   
> coreference "is impossible"  
> — Platero 1978:86, (3.18), (3.19); 87, (3.20) (page images checked; (3.20) at 400 dpi); copied from the working note “What comes first decides: Platero 1978 on who "his" can mean, and a no-movement route to W&J (70b) (T59a)” <https://navajo-decoded.grant-golland.workers.dev/possessor-command/>, ex. (10)



## 3. The argument

Hale's argument is valid. The site's derivations D1–D5 confirm both halves **[HYPOTHESIS]**:

- Nouns within the subject's c-command give *i ≠ j*.
- Nouns adjoined above the subject, as in Hale's own tree (9), give free coreference.

So the argument rules out one placement: nouns hanging above the subject.

It needs ten assumptions, A1–A10, and none is independently established for Navajo **[HYPOTHESIS]**:

- **A3** ("an adjunct is outside the subject's reach") meets counter-evidence in English.
- **A5** (stable judgments) meets Speas's (4.119), a sentence of the same shape with the opposite judgment.
- **A10** is needed for (11b) and fails on the site's own analysis of psych verbs.
- **A6** is a stipulation.

The better question is timing, not adjunct status: is the noun already in the tree when the hidden subject reaches its final position? Eleven combinations of attachment site and subject position were tested **[HYPOTHESIS]**:

- six predict Hale's judgment for (11a);
- five predict Speas's;
- none predicts both.

**Two surprises** **[HYPOTHESIS]**:

1. In Hale 2003's structure, the subject pronoun is merged in Spec,v and does not move. A noun attached to vP after that point is therefore outside the subject's reach.
2. Hale 2003 places overt nouns "adjoined to the full verb word (or verbal clause)" (p. 27). With his Spec,v subject, that predicts *free* coreference in his own 1999 sentences. This tension is derived by the site; Hale does not claim it.

## 4. Reconciliation and correction

At 12:46pm on 2026-10-15 the site wrote that FB-5 "is wrong" and that the D6 verdict was "withdrawn". At 1:07pm that wording was corrected as too strong. D6 is **compatible, undecided** **[HYPOTHESIS]**.

FB-1 to FB-4 are untouched. If the A-set holds, FB needs three conditional additions **[HYPOTHESIS]**:

- **LINK-IN**: a linked noun sits inside the projection of its head;
- **FB-5′**: Condition C applies over unlinked coreference;
- **FB-6′**.

A note on older notation: Jelinek 1984's Warlpiri tree is pre-VPISH flat-S notation (V, AUX, CPP with no VP). It gives no verdict on A3, and an earlier statement that it did was withdrawn **[HYPOTHESIS]**.

## 5. Status

A3 survives only in a restated form, A3′. Sentence (10) is the cleanest case and the next one to test with speakers. Where Navajo nouns attach (CP, TP/M, vP, PP or a left-peripheral specifier) remains open **[OPEN]**.


## References

1. Hale, Ken 1999. Conflicting truths (pp. 171–173). [opened]
2. Speas 1986. Dissertation, Ch. 2 and Ch. 4, (4.15a), (4.119)–(4.121). [opened]
3. Hale, Ken 2003 (manuscript), pp. 27–30. [opened]
4. Jelinek, Eloise 1984. Warlpiri, tree (28), pp. 50–51, 73. [opened]
5. Baker. Agreement/dislocation paper, n. 16. [opened]
6. Baker 1996 (trees). [not opened]
7. Platero, Paul 1978. Dissertation, pp. 80–87. [opened]

Canonical HTML: https://navajo-decoded.grant-golland.workers.dev/essays/condition-c/
