[Kra98a]: More Structural Analogies between Pronouns and Tenses

[Kra98a] [Par73] [Kle94] [Abu88] [Abu97] [Ogi89]

[Kra98a]'s SALT VIII paper extends [Par73]'s tense–pronoun analogy in three directions: an aspect-based decomposition of English simple past, SOT deletion via zero tenses, and zero forms with locality constraints. The substrate machinery (deletion mechanism + Kratzer-named lexical entries used by Fragments) is at Semantics/Tense/SOT/Decomposition.lean; this study file collects the paper-anchored cross-references and the empirical chain theorems connecting Fragments → Theory → Data → Empirical judgments.

Architectural note

The kratzerEnglishPast / kratzerGermanPreterit / kratzerZeroTense lexical entries live at the Theories layer (Tense/SOT/Decomposition.lean) because Fragments/{English,German,Italian}/Tense.lean consume them via the Fragments → Theories import direction. The "Fragments import Theories, never Studies" layering discipline forces the substrate placement; this Studies file collects the paper-anchored cross-paper claims and bridge theorems that don't need to be Fragment-visible.

Section 7 decomposition (English simple past = perfect + present)

The cornerstone empirical contrast ([kratzer-1998] Section 7, ex (40), p. 16) is the English/German out-of-the-blue diagnostic: English simple past is acceptable as a deictic past tense ((40a) "Who built this Church? Borromini built this church."); the German simple past (Präteritum) is deviant in the same context ((40b)); the German present perfect (Perfekt) fills the deictic slot ((40c)). Kratzer concludes: "Since the simple past in English can be used in out of the blue utterances describing past events, it must be a way of spelling out perfect aspect and present tense together" (p. 18).

The empirical data live as Examples.ex40a/ex40b/ex40c in the generated block below, with Examples.all exposing the full Kratzer98 example list. The chain theorems below verify that the Fragment entries' canBeDeictic predictions agree with each example's empirical judgment.

Fragment ↔ Example agreement: deictic-vs-anaphoric tense

The chain Fragment → Example replaces the previous Fragment → frame → datum chain (which was a six-conjunct rfl tower over hand-stipulated Reichenbach integers). The empirical anchor is now the LinguisticExample.judgment field of the corresponding Kratzer98 numbered example, which is verifiable from the paper itself.

Predictions tested:

• kratzerSimplePast.canBeDeictic = true ↔ Examples.ex40a.judgment = .acceptable (English simple past, out of the blue).
• kratzerPreterit.canBeDeictic = false ↔ Examples.ex40b.judgment = .ungrammatical (German Präteritum, out of the blue).
• kratzerPerfekt.canBeDeictic = true ↔ Examples.ex40c.judgment = .acceptable (German Perfekt, out of the blue).

theorem Kratzer1998.english_simple_past_chain : English.Tense.kratzerSimplePast.tensePronoun.constraint = Tense.present ∧ English.Tense.kratzerSimplePast.hasPerfect = true ∧ Examples.ex40a.judgment = Features.Judgment.acceptable

English simple past = perfect + present. Per Kratzer §7 (p. 18), the English simple past decomposes as PRESENT-tense head over PERFECT aspect. The Fragment-level encoding (constraint = .present + hasPerfect = true) is verified directly in Fragments/English/Tense.lean; this theorem isolates the bridge claim that needs both Fragment-side and Example-side facts: the Fragment encoding predicts deictic usability, and Kratzer's out-of-the-blue example (40a) ("Who built this Church?…") is .acceptable.

theorem Kratzer1998.german_preterit_chain : German.Tense.kratzerPreterit.tensePronoun.constraint = Tense.past ∧ German.Tense.kratzerPreterit.hasPerfect = false ∧ Examples.ex40b.judgment = Features.Judgment.ungrammatical

German Preterit = genuine past pronoun. Per Kratzer §7 (ex (40b), p. 16): the German Präteritum requires a contextually salient past time, behaving like an anaphoric pronoun. The Fragment encodes this as kratzerPreterit.tensePronoun.constraint = .past

• hasPerfect = false; the empirical anchor is Examples.ex40b (deviant out of the blue, star per Kratzer).

theorem Kratzer1998.german_perfekt_chain : German.Tense.kratzerPerfekt.tensePronoun.constraint = Tense.present ∧ German.Tense.kratzerPerfekt.hasPerfect = true ∧ Examples.ex40c.judgment = Features.Judgment.acceptable ∧ German.Tense.kratzerPerfekt.tensePronoun.constraint = English.Tense.kratzerSimplePast.tensePronoun.constraint ∧ German.Tense.kratzerPerfekt.hasPerfect = English.Tense.kratzerSimplePast.hasPerfect

German Perfekt = perfect + present (same decomposition as English simple past). Per Kratzer §7 (ex (40c), p. 16): the Perfekt fills the deictic-past slot that the Preterit cannot. The chain asserts BOTH the empirical agreement on (40c) AND the cross-Fragment parallelism (Perfekt's tense head + perfect-aspect coincide with kratzerSimplePast's), which is the substantive content of "same decomposition."

theorem Kratzer1998.zero_tense_chain : (Tense.SOT.Decomposition.kratzerZeroTense 1).constraint = Tense.present ∧ Tense.Overtness.fromBinding (Tense.SOT.Decomposition.kratzerZeroTense 1).mode true = Tense.Overtness.zero

Zero tense surface properties. Per Kratzer §4 (p. 10–11): English has two indexical tenses (present, past) and a zero tense. The substrate lemmas zero_tense_is_present and zero_tense_overtness in Tense/SOT/Decomposition.lean carry the underlying claims; this theorem just binds them locally for cross-reference.

Agreement with Ogihara on the simultaneous cell

theorem Kratzer1998.deletion_agrees_with_zero_tense_binding {Time : Type u_1} (m : _root_.Time.ReichenbachFrame Time) : Tense.SOT.Decomposition.applyDeletion m = Tense.simultaneousFrame m m.eventTime ∧ (Tense.SOT.Decomposition.applyDeletion m).isPresent

Kratzer's SOT deletion and [Ogi96]'s zero-tense binding build the same embedded frame: deletion yields exactly the simultaneousFrame whose embedded event time is the matrix event time, and that frame is PRESENT relative to the shifted perspective. The two accounts provably agree on the core past-under-past simultaneous cell; they differ in mechanism (deletion of a genuine PAST vs a bound zero PRESENT — see Tense.SOT.Ambiguity.PastReading for the typed mechanism-level divergence).

Cross-paper bridge theorems (Phase F)

The contrast theorems with Ogihara, Sharvit, von Stechow, Klecha are intentionally not yet landed; substrate is ready (applyDeletion, sotDeletionApplicable, the kratzer-named lexical entries are all exported from Tense/SOT/Decomposition.lean).