Perfective by Default: Aspect-Shifting Affixes in Turkish Benjamin S

Perfective by Default: Aspect-Shifting Affixes in Turkish Benjamin S

Perfective by Default: Aspect-Shifting Affixes in Turkish Benjamin S. Meriçli Rethinking the Turkish Verbal Projection ■ Broader Goal: A more intuitive approach to Turkish verbal morphology – A project at the intersection of morphology, syntax, and semantics – Reexamining the tradition of Kornfilt (1996), Erguvanlı-Taylan (2001), Göksel & Kerslake (2005), Hankamer (2013), and many more ■ A First Step: Address the potentially theory-external questions. – A preliminary matter when mapping out the combinatorics of Turkish suffixes: Which suffixes contribute which varieties of meaning? ■ Why ConCALL: How does one such problem appear through the areal lens? 2 The Meaning at Hand: Viewpoint Aspect ■ Aspect (Comrie 1976, Johanson 2000, Bhatt & Pancheva 2005, etc.) – Aktionsart – Viewpoint Aspect > hereafter, aspect ■ Narrow Goal: Why do the Turkish past tense (-DI) and indirect evidential (-mIş) morphemes convey perfective aspect after a bare verb root, but imperfective aspect when preceded by the so-called copula (-Ø- / -y-) ? T [ V ___ ] T [ COP ___ ] -mIş (EVID) aspect PERF aspect IMPF -DI (PAST) 3 Briefly: PERF vs. IMPF ■ PERF: an event is viewed “from the outside” Ali made yogurt. Event Time Reference Time or Event Time Reference Time – Göksel & Kerslake (2005): Perfective aspect presents an event as a completed whole, with an identifiable beginning and end point. ■ IMPF: an event is viewed “from the inside” Ali was making yogurt. Event Time Reference Time 4 The Turkish Verbal Spine The tip of the morphosyntactic iceberg: Hankamer (2013) 5 Basic Data: Finite -DI & -mIş ■ Indirect Evidential: -mIş Finite Post-Verbal Finite Post-Copular (1) Hande anla-(*y)-mış. (2) Hande Kars-’ta-*(y)-mış. Hande understand-(*COP)-EVID Hande Kars-LOC-*(COP)-EVID ‘Hande apparently understood.’ ‘Hande was apparently in Kars.’ ■ Past Tense: -DI Finite Post-Verbal Finite Post-Copular (3) Ali dinle-(*y)-di. (4) Ali dinle-meli-*(y)-di. Ali listen-(*COP)-PAST Ali listen-NEC-*(COP)-PAST ‘Ali listened.’ ‘Ali should have listened.’ Post-Verbal: Copula Unacceptable Elsewhere (Post-Copular): Copula Obligatory 6 Non-Finite -DI & -mIş ■ Non-Finite Post-Verbal Relative Tense Nominal Expressions (5) Hande anla-mış-Ø-tı. (6) Nasreddin Hoca-’yı iyi anla-mış biri Hande understand-PERF-COP-PAST Nasreddin Hoca-ACC well understand-PERF one ‘Hande had understood.’ ‘one who understands Nasreddin Hoca well’ (7) Hande anla-dı-y-dı. (8) *Nasreddin Hoca-’yı iyi anla-dı biri Hande understand-PERF-COP-PAST Nasreddin Hoca-ACC well understand-PERF one ‘Hande had understood.’ ■ Non-Finite Post-Copular Unacceptable (9) *Hande Kars-’ta-y-mış-tı. (10) *Kars-’ta-y-mış biri Hande Kars-LOC-COP-EVID Kars-LOC-COP-EVID one 7 ∴ Exclude Non-Finite -DI & -mIş ■ They have different interpretations than their finite counterparts. – Non-finite -mIş does not convey the core meaning of finite -mIş: EVID. – Non-finite -DI does not convey the core meaning of finite -DI: PAST. ■ They are licensed in distinct morphosyntactic environments. – Non-finite -DI and -mIş subcategorize for bare verbs to create relative tenses (e.g., past perfect). – Non-finite -mIş subcategorizes for bare verbs to create nominal expressions (e.g., APs and DPs). – In neither case can the suffixes be said to occupy T. Finite -DI and -mIş are distinct from their non-finite counterparts. In other words, a suffix and its non-finite counterpart are distinct lexical items. 8 Finite -DI & -mIş: Post-Verbal ■ Post-verbal -DI and -mIş incompatible with stative predicates. (11) #Türkçe iyi bil-di. (12) #Elma çok sev-miş. Turkish well know-PAST apple much love-EVID Intended: ‘She knew Turkish well.’ Intended: ‘He apparently really loved apples.’ ■ Post-verbal -DI and -mIş can coerce non-stative readings (de Swart 1998, Homer 2011). (13) Doktor ol-du. (14) Lezzetli ol-muş. doctor be-PAST tasty be-EVID ‘She became a doctor.’ ‘It (apparently) turned out to be tasty.’ #‘She was a doctor.’ #‘It was (apparently) tasty.’ Post-verbal -DI and -mIş convey perfective aspect. 9 Finite -DI & -mIş: Post-Copular ■ Post-copular -DI and -mIş compatible with stative predicates. (15) Hande Kars-’ta-y-mış. (16) Ali öğrenci-y-di. Hande Kars-LOC-COP-EVID Ali student-COP-PAST ‘Hande was apparently in Kars.’ ‘Ali was a student.’ ■ We can also construct a stative predicate from other TAM markers. (17) Türkçe iyi öğren- { iyor / ir / ecek / meli / mek-te / miş } -{ y / Ø } -DI. Turkish well learn-{ PROG / AOR / FUT / NEC / INF-LOC / PERF } -COP -PAST ‘She { was learning / would have learned / was going to learn / should have learned / was in the process of learning / had learned } Turkish well.’ Post-copular -DI and -mIş convey imperfective aspect. 10 Finite -DI & -mIş: A Final Diagnostic ■ With particular predicates, completive adverbials are known to differentiate perfective from imperfective aspect (Bhatt & Pancheva 2005). Post-Verbal Post-Copular (18) Beş sene-de hemşire ol-du-m. (19) ??Beş sene-de hemşire-y-di-m. five year-LOC nurse be-PAST-1s five year-LOC nurse-COP-PAST-1s ‘I became a nurse in five years.’ Intended: ‘I became a nurse in five years.’ ∴ Post-verbal and post-copular -DI and -mIş indeed have distinct aspectual readings. T [ V ___ ] T [ COP ___ ] -mIş (EVID) aspect PERF aspect IMPF -DI (PAST) 11 Summary and Way Forward ■ Non-finite -DI and -mIş are rightly excluded from an analysis of finite -DI and -mIş. ■ Finite post-verbal and post-copular -DI and -mIş have distinct aspectual readings. – Should a theory of morphosyntax treat a post-verbal suffix as the same lexical item as its post-copular counterpart? – Challenge: It should, and it can. ■ In theory-independent terms, this approach argues that a parsimonious account… – will treat a suffix’s non-finite and finite versions as distinct LIs. – will treat a suffix’s post-verbal and post-copular (finite) versions as the same LI. – will treat a suffix’s finite version as unspecified for aspect. 12 Finite -DI & -mIş Unspecified for Aspect ■ A parsimonious model for Turkish verbal morphology: -DI contributes only PAST whether post-verbal or post-copular (20) Lexical Item for -DI: T (PAST) ; [ { V / COP } ___ ] -mIş contributes only EVID whether post-verbal or post-copular (21) Lexical Item for -mIş: T (EVID) ; [ { V / COP } ___ ] ■ If -DI and -mIş are unspecified for aspect, then how do the different aspectual specifications (PERF vs. IMPF) arise? ■ Aspect is contributed from elsewhere in the verbal projection. 13 Imperfective by Copula ■ As we saw above, post-copular -DI and -mIş are compatible with stative predicates and degraded with completive adverbials. – We took this as evidence that post-copular -DI and -mIş are IMPF. ■ However, say we attribute IMPF aspect to the copula itself. – Then post-copular -DI and -mIş can remain unspecified for aspect. ■ Historical Support: The copula in question -y-/-Ø- descends historically from the now seldom used bound copula i- (see Bybee & Dahl 1989, Göksel & Kerslake 2005). (22) Hande Kars-’ta i-miş. (23) Ali öğrenci i-di. Hande Kars-LOC be-EVID Ali student be-PAST ‘Hande was apparently in Kars.’ ‘Ali was a student.’ 14 Perfective by Default ■ Similarly, post-verbal -DI and -mIş are infelicitous with stative predicates but fine with completive adverbials. – We took this as evidence that post-verbal -DI and -mIş are PERF. T ■ However, say we propose PERF as the aspectual elsewhere case. (COP) – Then post-verbal -DI and -mIş can remain unspecified for aspect. V (ASP) ■ For clauses otherwise unmarked for aspect, the clause defaults to PERF. – i.e., bare verbs ({ ASP / COP } = Ø) are perfective ■ Evidence from Imperatives: Imperatives, which often take the form of a bare verb, bear perfective aspect as a cross-linguistic default (Kaufmann 2012). 15 PERF & IMPF: Composition ■ What results is a fundamental morphosyntactic/semantic opposition. – Unmarked aspect: PERF – Marked aspect: IMPF T (COP) ■ We assume that the verbal domain denotes an event predicate. V (ASP) (24) [[ V ]] = λe.P(e) ■ COP now becomes the domain of viewpoint aspect (PERF vs. IMPF), with ASP reserved for other TAM information. (25) [T -DI / -mIş [COP Viewpoint Aspect [ASP Other TAM Content [V Event Predicate ]]]] ■ Recall that EVID, though not a tense marker, is also hosted in T. 16 PERF & IMPF: Working Denotations ■ If COP is empty, the clause defaults to [[ PERF ]]. When the copula is present, COP contributes [[ IMPF ]]. ■ Adapted from Bhatt & Pancheva (2005): (26) [[ PERF ]] = λP . λt . ∃e [ τ(e) ⊆ τ(r) ∧ P(e) ] T (27) [[ IMPF ]] = λP . λt . ∃e [ τ(r) ⊂ τ(e) ∧ P(e) ] (COP) e : an event V (ASP) τ(e) : event time; the runtime of the event τ(r) : reference time; time during which proposition at issue is taken to be true P(e) : a predicate P 17 Summary of Proposal ■ Theoretical Parsimony: Even before assembling a formal account, we can aim for simplicity by minimizing its moving parts. ■ This account does so by – distinguishing elements by their syntax & semantics, not their phonology, and – uniting post-verbal and post-copular -DI and -mIş. ■ While the account generally follows in the tradition of Kornfilt (1996) through Hankamer (2013)… – it presents a generic roadmap for any theory aiming to tackle the problem. – In other words: we are one step up from a descriptive grammar. 18 Looking Forward ■ Historical Projects: Semantic & Syntactic Change – Model the development of -y-/-Ø- from the bound copula i-. – Investigate development of non-finite -mIş. ■ Areal Projects: Variation & Areal Diffusion – Examine similar phenomena in other Turkic languages. – Compare to genetically unrelated but co-located languages. ■ Regarding Default Perfective Aspect: – Show whether this default aspect can be entailed or implicated in the sense of Bohnemeyer & Swift (2004) -- i.e., a proper formal explanation. – Conduct a thorough study of the aspectual properties of imperatives and other “bare” verb forms -- e.g., bil, bil-se, bil-sin.

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    22 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us