Hartmann and Zimmermann (2007): In place — out of place? Focus strategies in Hausa #
Hartmann and Zimmermann's argument that Hausa is a counterexample to two universalist claims:
that focus is always marked, and that the position of a focus determines its pragmatic
interpretation. Focus is realized in situ or ex situ, and the paper's §3.2 matrix crosses the two
strategies with four pragmatic uses, new-information, corrective, contrastive and selective,
every cell of which is licensed (hzMatrix); strategy_does_not_determine_pragType refutes the
Meaning-Structure Mapping Hypothesis of (21), the label following Vallduví and Vilkuna, stated
as a failure of Function.FactorsThrough on the licensed utterances, so that the Hungarian and
Hausa verdicts differ on one predicate. In-situ new-information focus carries no morphosyntactic
reflex, and the §5 pilot finds no prosodic one either, so hausa_falsifies_UniversalBFR refutes
the Basic Focus Rule of Selkirk; focused subjects, by contrast, always front (§2.2.2).
Exhaustivity is not structural: the particle kawài 'only' exhaustifies over the resolved
contrast set in either position.
Implementation notes #
The §1.2 control taxonomy of uses and its factor-through theorems live in
Semantics/Focus/Control.lean, and each cell's pragmatic type is computed from its controlling
context rather than tagged. Subject foci in TAMs lacking a Relative form (future, habitual,
subjunctive) are syntactically and morphologically unmarked, which the paper analyses as
string-vacuous fronting, so the licensing predicate bans in-situ subjects unconditionally and the
future subject focus of (8) is licensed yet reflex-free. The paper counts the subjunctive with the
future and the habitual, but Newman excludes it from Rel environments, and the fragment's
licensing, which is his, rejects a fronted subjunctive (not_licensed_exSitu_subjunctive). Which
TAMs show Relative morphology is not stipulated: a licensed fronted focus changes the TAM exactly
when the TAM does not occur in Rel environments (Hausa.FocusConfig.pacTAM_ne_iff), the
completive and continuous of the paper's data. The §3.3 corpus tendencies stay in prose:
answers to wh-questions are mostly in situ (99 against 25) and selective, corrective and
contrastive foci mostly ex situ (154 against 12), but no use is categorically excluded from
either position, and only that categorical claim is a theorem. Locators follow the authors'
prepublication manuscript of the chapter.
TODO #
- The like-for-like contrast with Hungarian: kawài against the exhaustification over
Kiss1998.alternativesthat the identificational position performs. - §2.3 multiple foci: one ex-situ focus with in-situ foci (18a–c).
- §4 focus pied-piping and partial focus movement, and the final Ex-Situ Generalisation of their (47), which need a structured-meaning overlap predicate.
- §5 prosodic pilot data and §6.1 emphasis motivation are quantitative tendencies, in prose only.
References #
What is focused (§2.2.2) #
Hausa singles out subjects as what is focused (§2.2.2); everything else collapses to
nonSubject.
Instances For
Equations
- HartmannZimmermann2007.instDecidableEqFocused x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯
Equations
- One or more equations did not get rendered due to their size.
Instances For
Equations
Equations
Tagged focus utterances and Hausa licensing #
A FocusConfig tagged with its pragmatic use and focused
constituent.
- cfg : Hausa.FocusConfig
- pragType : Focus.Use
- focused : Focused
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Equations
An utterance is Hausa-licensed when it is morphosyntactically licensed
(FocusConfig.Licensed) and respects the §2.2.2 ban on in-situ subject focus. The ban is
unconditional, since in TAMs without a Relative form the fronting is merely invisible (p. 4),
not absent.
Equations
- u.IsHausaLicensed = (u.cfg.Licensed ∧ (u.focused = HartmannZimmermann2007.Focused.subject → u.cfg.strategy = Focus.Strategy.exSitu))
Instances For
Equations
- One or more equations did not get rendered due to their size.
Controlling contexts of the §3.2 matrix #
Every cell's mini-scenario has one shape: a two-point answer domain —
the mentioned answer ans vs the contextual alternative alt — whose
content (fish vs other dishes, fifteen vs twenty Naira, behind vs in
front, …) lives in the cell's data row. The antecedent is the canonical
Use.model of the row's context shape; the answer is composed by the
engine (pairAnswer); ctx_resolves is the uniform full-resolution
fact. Ex-situ variants are interpreted at the base structure (fronting
is A'-movement for the paper, and alternatives do not distribute
through abstraction: PredAbs WithAlternatives = ⟨none⟩).
The two-point answer domain of a cell's mini-scenario.
Instances For
The canonical antecedent of a use over the two-point scenario.
Equations
Instances For
The composed answer of the two-point scenario.
Equations
Instances For
Every cell's context fully resolves against the composed answer — all squiggle clauses, plus the correction clause for the corrective cells. One semantics, four pragmatic uses.
Exhaustive focus (§3.2, (32)) #
Exhaustivity is not structurally encoded: it is induced by focus particles (kawài 'only'; nee/cee per the paper's fn. 3) over the resolved contrast set, in either strategy — (32a/b) attest in-situ and ex-situ only BOOKS alike.
The exhaustified answer is the exclusion asserted by only over the scenario's resolved contrast set.
Equations
Instances For
The exhaustified answer computes to the bare true answer, uniformly across the four uses. Exhaustification consumes the resolved contrast set and prejacent, never the strategy, which is the §3.2 point that exhaustive readings are available in both positions.
The 8-cell empirical matrix (§3.2) #
Each cell's pragmatic type is computed from its controlling context:
the constructors take a Focus.Antecedent, not a tag.
Equations
- HartmannZimmermann2007.mkExSituUtt tam pacTAM cell g hasStab ctl foc = { cfg := Hausa.mkExSitu tam pacTAM cell (some g) hasStab, pragType := ctl.use, focused := foc }
Instances For
Equations
- HartmannZimmermann2007.mkInSituUtt tam cell g ctl foc hasStab = { cfg := Hausa.mkInSitu tam cell (some g) hasStab, pragType := ctl.use, focused := foc }
Instances For
The first person singular PAC cell.
Instances For
The third person masculine singular PAC cell.
Equations
Instances For
The third person feminine singular PAC cell.
Equations
Instances For
The impersonal PAC cell.
Equations
Instances For
Ex-situ new-information focus ((22), Examples.ex22).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Ex-situ corrective focus on a feminine subject ((24),
Examples.ex24).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Ex-situ selective focus, no stabilizer ((29), Examples.ex29).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Ex-situ contrastive focus, no stabilizer, with the impersonal akèe ((27),
Examples.ex27).
Equations
- One or more equations did not get rendered due to their size.
Instances For
In-situ new-information focus ((23), Examples.ex23).
Equations
Instances For
In-situ corrective focus with sentence-final nèe ((25),
Examples.ex25).
Equations
- One or more equations did not get rendered due to their size.
Instances For
In-situ selective focus ((30), Examples.ex30).
Equations
Instances For
In-situ contrastive focus ((26), Examples.ex26).
Equations
Instances For
The 8-cell matrix of §3.2 crosses both strategies with all four pragmatic types.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Every cell of the §3.2 matrix is Hausa-licensed.
Strategy does not determine pragmatic type (§3.2) #
On Hausa-licensed utterances pragType does not factor through strategy, since (22) and
(24) are both ex-situ with distinct pragmatic types.
The refutation persists when restricted to in-situ utterances, by (23) against (25).
Subject-focus generalization (§2.2.2) #
Focused subjects front (§2.2.2); unpacks the second conjunct of
IsHausaLicensed.
The starred in-situ subject focus ((17 A2), Examples.ex17a2).
Equations
- One or more equations did not get rendered due to their size.
Instances For
The grammatical ex-situ subject focus ((17 A1), Examples.ex17a1).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Subject focus in the future, which occurs in Rel environments unchanged ((8),
Examples.ex8), is string-vacuous fronting with no overt reflex
(exSitu_subject_future_no_reflex).
Equations
- One or more equations did not get rendered due to their size.
Instances For
No fronted subjunctive is licensed, since the subjunctive has no Rel counterpart.
Universalist Basic Focus Rule (§5, §6.2) #
The overt reflexes of a focus utterance in the shared Reflex vocabulary are non-vacuous
fronting (subjects front string-vacuously), the Rel form of the PAC where fronting changes the TAM,
and the stabilizer.
Equations
- One or more equations did not get rendered due to their size.
Instances For
A focus utterance has a morphosyntactic reflex when some reflex lies outside the phonological channel.
Equations
- u.HasMorphosyntacticReflex = ∃ ρ ∈ u.reflexes, ρ.modality.channel ≠ Reflex.Channel.phonological
Instances For
Equations
- One or more equations did not get rendered due to their size.
The universalist claim — [Sel95]'s Basic Focus Rule and its tradition (§5, §6.2) — that every focused utterance carries some structural reflex, weakened to morphosyntax and quantified over Hausa utterances only.
Equations
Instances For
(23) is licensed and reflex-free; the §5 pilot finds no prosodic reflex either.
The subject-side counterexample, (8).
Every Hausa focus reflex is morphosyntactic, so a morphosyntactic reflex is just an overt one.
Hausa refutes the universalist claim that every (licensed) focus
receives an overt reflex — the same shape Tangale refutes in
HartmannZimmermann2004.lean.
Polar tone of nē/cē (§2.1) #
The rows' tone labels.
Equations
- HartmannZimmermann2007.toneTable = [("H", Tone.TRN.H), ("L", Tone.TRN.L)]
Instances For
In (3a) and (3b) the stabilizer's tone is polar to the host's final tone, high after Kandè and low after Kiifii.
Data linkage #
Each cell's tags are pinned to the paperFeatures of its
Data.Examples.HartmannZimmermann2007 row.
Equations
Instances For
Equations
- HartmannZimmermann2007.pragLabel Focus.Use.newInfo = "newInfo"
- HartmannZimmermann2007.pragLabel Focus.Use.corrective = "corrective"
- HartmannZimmermann2007.pragLabel Focus.Use.selective = "selective"
- HartmannZimmermann2007.pragLabel Focus.Use.contrastive = "contrastive"
Instances For
Equations
Instances For
Equations
- HartmannZimmermann2007.stabLabel c = match c.stab? with | some p => if p = Hausa.ce then "cee" else "nee" | none => "none"
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Every cell's tags agree with its row's paperFeatures.
(8) carries no pragmatic tag; its other tags agree with its cell.
Equations
- HartmannZimmermann2007.tamLabel Hausa.TAM.completive = "completive"
- HartmannZimmermann2007.tamLabel Hausa.TAM.preterite = "preterite"
- HartmannZimmermann2007.tamLabel Hausa.TAM.continuous = "continuous"
- HartmannZimmermann2007.tamLabel Hausa.TAM.relContinuous1 = "relContinuous1"
- HartmannZimmermann2007.tamLabel Hausa.TAM.relContinuous2 = "relContinuous2"
- HartmannZimmermann2007.tamLabel Hausa.TAM.future = "future"
- HartmannZimmermann2007.tamLabel Hausa.TAM.allative = "allative"
- HartmannZimmermann2007.tamLabel Hausa.TAM.potential = "potential"
- HartmannZimmermann2007.tamLabel Hausa.TAM.rhetorical = "rhetorical"
- HartmannZimmermann2007.tamLabel Hausa.TAM.habitual = "habitual"
- HartmannZimmermann2007.tamLabel Hausa.TAM.subjunctive = "subjunctive"
Instances For
The rows that tag a TAM tag the clause's TAM, not its PAC's, as (17 A1) is continuous with a Rel-continuous PAC.
The row the paper stars is the cell the licensing predicate rejects.