Then too. . . Benjamin Slade and Aniko Csirmaz Dept. of Linguistics - - PowerPoint PPT Presentation

Then too. . .

Benjamin Slade and Aniko Csirmaz

• Dept. of Linguistics

University of Utah

Meaning in Flux 14 October 2017

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 1 / 58

Aspectual adverbs in Hindi & Nepali

1

Aspectual adverbs in Hindi & Nepali

2

Formally defining Hindi phir, Nepali pheri

3

Historical developments involving aspectual adverbs

4

Additive particles & phir/pheri

5

Crosslinguistic comparisons - Hungarian &c.

6

Conclusion & refs

7

Additional materials Additional Hungarian diachronic & synchronic data More on other treatments

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 1 / 58

Aspectual adverbs in Hindi & Nepali

Overview

Hindi phir, Nepali pheri are ambiguous between “again” and “then” (ordering)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 2 / 58

Aspectual adverbs in Hindi & Nepali

Overview

Hindi phir, Nepali pheri are ambiguous between “again” and “then” (ordering)

developed from an earlier form meaning “again” (< verb “turn; return”)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 2 / 58

Aspectual adverbs in Hindi & Nepali

Overview

Hindi phir, Nepali pheri are ambiguous between “again” and “then” (ordering)

developed from an earlier form meaning “again” (< verb “turn; return”)

→ Propose a template definition shared by these two meanings (and others) : explains observed ambiguities & change

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 2 / 58

Aspectual adverbs in Hindi & Nepali

Overview

Hindi phir, Nepali pheri are ambiguous between “again” and “then” (ordering)

developed from an earlier form meaning “again” (< verb “turn; return”)

→ Propose a template definition shared by these two meanings (and others) : explains observed ambiguities & change Hindi & Nepali have a morphologically complex form formed from an aspectual adverb combined with an additive particle: Hindi phir bh¯ ı, Nepali pheri pani

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 2 / 58

Aspectual adverbs in Hindi & Nepali

Overview

Hindi phir, Nepali pheri are ambiguous between “again” and “then” (ordering)

developed from an earlier form meaning “again” (< verb “turn; return”)

→ Propose a template definition shared by these two meanings (and others) : explains observed ambiguities & change Hindi & Nepali have a morphologically complex form formed from an aspectual adverb combined with an additive particle: Hindi phir bh¯ ı, Nepali pheri pani

unrelated Hungarian shares a structurally equivalent parallel m´ egis

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 2 / 58

Aspectual adverbs in Hindi & Nepali

Overview

Hindi phir, Nepali pheri are ambiguous between “again” and “then” (ordering)

developed from an earlier form meaning “again” (< verb “turn; return”)

→ Propose a template definition shared by these two meanings (and others) : explains observed ambiguities & change Hindi & Nepali have a morphologically complex form formed from an aspectual adverb combined with an additive particle: Hindi phir bh¯ ı, Nepali pheri pani

unrelated Hungarian shares a structurally equivalent parallel m´ egis

discuss historical developments involving these aspectual adverbs and ambiguity/underspecification

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 2 / 58

Aspectual adverbs in Hindi & Nepali

Hindi phir, Nepali pheri

Hindi (1) R¯ am Ram phir then/again so sleep gay¯ a. went “Ram slept then/again.” Nepali (2) R¯ am Ram pheri then/again sutyo. slept “Ram slept then/again.”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 3 / 58

Aspectual adverbs in Hindi & Nepali

Hindi phir bh¯ ı, Nepali pheri pani

Hindi (3) Shy¯ am Shyam gun .d .¯ a villain hai; is, phir then/again bh¯ ı too mer¯ a my dost friend hai. is “Shyam’s a villain; stillconc he’s my friend.” Nepali (4) Pheri then/again pani too timro your tasbir image ˜ ¯ akha.m¯ a eye.in ¯ ae.rahancha. come.continues “Stillconc your image keeps coming into my eyes.”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 4 / 58

Aspectual adverbs in Hindi & Nepali

Some Adverbials in Hindi & Nepali

Hindi Nepali Meaning tab tab(a) “then (at that time)” taile phir pheri “then (after that)” tab(a) pachi phir pheri “again” (repetitive, restituitive) phir se (dob¯ ar¯ a) abh¯ ı bh¯ ı aile samma temporal/continuity ab tak ajha(i) (pani) “still” aile pani phir bh¯ ı pheri pani concessive/adversative tai pani “still” tarai pani ra pani bh¯ ı pani additive & scalar(-additive) particle “also”/“even”

Table: Hindi & Nepali “pieces”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 5 / 58

Formally defining Hindi phir, Nepali pheri

1

Aspectual adverbs in Hindi & Nepali

2

Formally defining Hindi phir, Nepali pheri

3

Historical developments involving aspectual adverbs

4

Additive particles & phir/pheri

5

Crosslinguistic comparisons - Hungarian &c.

6

Conclusion & refs

7

Additional materials Additional Hungarian diachronic & synchronic data More on other treatments

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 5 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then”

Hindi (5) R¯ am-ne Ram-erg kh¯ an¯ a food kh¯ a eat liy¯ a. took. Vah He/she phir then so sleep gay¯ a. went “Ram ate food. Then he slept.” Nepali (6) R¯ am-le Ram-erg bh¯ at food kh¯ a-i-diyo. eat-abs-gave. U He/she pheri then sutyo. slept “Ram ate food. Then he slept.”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 6 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then” defined

phir “then”: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

phir/pheri combines with an ordered scale of times T, a time t, an eventuality e, and a saturated predicate P [P inherits the specifications

• f saturated predicate, so might appear in fuller form as e.g.

sleep(e∗, t∗, John) & in(e∗, kitchen)]

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 7 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then” defined

phir “then”: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

asserts an eventuality (“P”) which occurs at time t

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 8 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then” defined

phir “then”: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

asserts an eventuality (“P”) which occurs at time t presupposes a (prior) eventuality (“Q”) which occurs at time t*, where:

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 8 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then” defined

phir “then”: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

asserts an eventuality (“P”) which occurs at time t presupposes a (prior) eventuality (“Q”) which occurs at time t*, where:

t∗ ≺ t

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 8 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then” defined

phir “then”: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

asserts an eventuality (“P”) which occurs at time t presupposes a (prior) eventuality (“Q”) which occurs at time t*, where:

t∗ ≺ t the earlier presupposed eventuality (“Q”) is one of the focus alternatives to the asserted eventuality (“P”)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 8 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then” defined

phir “then”: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

asserts an eventuality (“P”) which occurs at time t presupposes a (prior) eventuality (“Q”) which occurs at time t*, where:

t∗ ≺ t the earlier presupposed eventuality (“Q”) is one of the focus alternatives to the asserted eventuality (“P”) The focus alternatives are determined based on what is focussed (see Rooth 1985, 1992)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 8 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then” defined

phir “then”: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

E.g. for: “John woke up. Then he [had a shower]F”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 9 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then” defined

phir “then”: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

E.g. for: “John woke up. Then he [had a shower]F”

Focus alternatives to focussed constituent [had a shower] include: “fell asleep, got dressed, ran a mile, left home, . . . ”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 9 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then” defined

phir “then”: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

E.g. for: “John woke up. Then he [had a shower]F”

Focus alternatives to focussed constituent [had a shower] include: “fell asleep, got dressed, ran a mile, left home, . . . ” So the focus alternatives to the clause will be, respectively: “John fell asleep, John got dressed, John ran a mile, John left home, . . . ”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 9 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then” defined

phir “then”: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

E.g. for: “John woke up. Then he [had a shower]F”

Focus alternatives to focussed constituent [had a shower] include: “fell asleep, got dressed, ran a mile, left home, . . . ” So the focus alternatives to the clause will be, respectively: “John fell asleep, John got dressed, John ran a mile, John left home, . . . ” (Where the focus was on [John], the focus alternatives might include “Mary”, “Bill”, “Sally” . . . , and so the focus alternatives to the clause wold be “Mary woke”, “Bill woke up” . . . )

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 9 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then” defined

phir “then”: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

E.g. for: “John woke up. Then he [had a shower]F”

Focus alternatives to focussed constituent [had a shower] include: “fell asleep, got dressed, ran a mile, left home, . . . ” So the focus alternatives to the clause will be, respectively: “John fell asleep, John got dressed, John ran a mile, John left home, . . . ” (Where the focus was on [John], the focus alternatives might include “Mary”, “Bill”, “Sally” . . . , and so the focus alternatives to the clause wold be “Mary woke”, “Bill woke up” . . . )

So the discourse “John woke up. Then he [had a shower]F” will be felicitous since “John woke up” is a possible focus alternative to “John [had a shower]F” and the former temporally proceeds the later.

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 9 / 58

Formally defining Hindi phir, Nepali pheri

(Ordering) phir/pheri “then” tree

vP then vP subevent vP subject v VP subevent VP V

• bject

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 10 / 58

Formally defining Hindi phir, Nepali pheri

(Repetitive) phir/pheri “again”

Hindi (7) R¯ am-ne Ram-erg bat .an-ko button-dat dab¯ a push diy¯ a. gave. Kuch Something nah˜ ¯ ı not hu¯ a. became. Us-ne he/she-erg phir (se) again bat .an-ko button-dat dab¯ a push diy¯ a. gave. “Ram pushed the button. Nothing happened. He pushed the button again.” Nepali (8) Birendra-le Birendra-erg bat .an-l¯ a¯ ı button-dat dab¯ a-i-diyo. push-abs-gave. Ke What pani also/even bhaena. became.neg. Us-le he/she-erg pheri again bat .an-l¯ a¯ ı button-dat dab¯ a-i-diyo. push-abs-gave. “Birendra pushed the button. Nothing happened. He pushed the button again.”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 11 / 58

Formally defining Hindi phir, Nepali pheri

Repetitives

phir (se), pheri: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

Identical to temporal ordering “then”, except that the time specification (rather than some other constituent) is under focus (may be unpronounced), and so the alternatives will vary in terms of temporal specification

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 12 / 58

Formally defining Hindi phir, Nepali pheri

Repetitives

phir (se), pheri: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

Identical to temporal ordering “then”, except that the time specification (rather than some other constituent) is under focus (may be unpronounced), and so the alternatives will vary in terms of temporal specification e.g. FAs to “John woke up again ([at 6am]F)” might include {John woke at t1, John woke at t2, John woke at t3, . . . }

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 12 / 58

Formally defining Hindi phir, Nepali pheri

Repetitives

phir (se), pheri: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≺ t & t, t∗ ∈ T

     .P(e, t, . . . )

Identical to temporal ordering “then”, except that the time specification (rather than some other constituent) is under focus (may be unpronounced), and so the alternatives will vary in terms of temporal specification e.g. FAs to “John woke up again ([at 6am]F)” might include {John woke at t1, John woke at t2, John woke at t3, . . . } so “John woke up at 4am. He woke up again [at 6am]F” is felicitous since the earlier sentence is a focus alternative of the latter and

• ccurs at an prior time

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 12 / 58

Formally defining Hindi phir, Nepali pheri

(Repetitive) phir/pheri “again” trees 1

Again, repetitive vP again vP subevent vP subject v VP subevent VP V

• bject

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 13 / 58

Formally defining Hindi phir, Nepali pheri

(Repetitive) phir/pheri “again” trees 2

Again, restitutive vP subevent vP subject v VP again VP subevent VP V

• bject

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 14 / 58

Formally defining Hindi phir, Nepali pheri

Connections between again and then

As defined, again and then are largely identical, only differing in where focus is placed (and so what the focus alternatives end up being)

For again the time specification is focussed; for then some phrasal element is focussed so again ends up presupposing the prior occurrence of the eventuality (at an earlier time) and then presupposes the prior occurrence of a partially similar eventuality (at an earlier time)

In Hindi & Nepali, phir pheri appear to be underspecified in terms of necessary focussed constituent, i.e. either a non-temporal phrasal subcomponent or the time specification (possibly covert/unpronounced)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 15 / 58

Formally defining Hindi phir, Nepali pheri

Other underspecified aspectual adverbs: Sanskrit p´ unar

• Skt. p´

unar polysemous consistent with templatic analysis: (9) hitv´ ¯ aya abandon.conv avady´ am imperfection.acc p´ unar again ´ astam home.acc ´ ¯ a to ihi come.impv “Having cast off imperfection, come home again.” [R . gveda X.14.8c] (10) Punar again/back ¯ agamya come.conv nis .adh¯ an Nishadha nale Nala.loc sarva ˙ m all.acc nyavedayat know.caus.impf.past.3sg “Having returned back to Nishadha, [the goose] made all known to Nala.” [Nala I.32] (11) ´ sr .n .u listen.impv punah . again/back “Listen still/further!” (12) Punar again ca and raman .¯ ıyes .u pleasant.loc.pl vanes .u forests.loc.pl upavanes .u groves.loc.pl ca and Damayanty¯ a Damayanti.instr saha with Nalo Nala.nom vijah¯ ara wandering-for-pleasure amaropamah .. like-immortals. “Moreover, Nala with Damayanti, like immortals, was always wandering about for pleasure in pleasant forests and groves.” [Nala V.44] (13) arthaih . wealth.instr sa ˙ mcayav¯ an accumulated-wealth-person arth¯ an wealth.acc pr¯ apnoti

• btains

kiyad little adbhutam, wonder, may¯ a me.instr punar still vin¯ a without eva emph artha ˙ m wealth laks .m¯ ıh . Laxmi ¯ as¯ adit¯ a sit.caus.impf.fem pur¯ a formerly “It is little wonder that someone who has inherited wealth should obtain wealth by using wealth; still I achieved prosperity long ago without any wealth to start with.” [Kath¯ asarits¯ agara XXIII.1–2]

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 16 / 58

Formally defining Hindi phir, Nepali pheri

Generalised template for aspectual adverbs

Template λSλxλeλP : ∃x∗ ∃e∗ ∃Q

    

Q(e∗, x∗, . . . ) & Q(e∗, x∗, . . . ) ∈ FA(P(e, x, . . . )) & x∗R x & x, x∗ ∈ S

     .P(e, x, . . . )

P, Q are (saturated) predicates x, x∗ are scalar entities (times, degrees, &c.) e, e∗ are eventuality variables R is a relation (e.g. ≺, ≻, ∝, &c.) S is a scale FA is a set of focus alternatives to P(e, x) which differ in terms of variation of elements under focus, which may include times, degrees,

• r subconstituents &c. (assuming ‘transparency’ of the event variable)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 17 / 58

Formally defining Hindi phir, Nepali pheri

Generalised template for aspectual adverbs

Template λSλxλeλP : ∃x∗ ∃e∗ ∃Q

    

Q(e∗, x∗, . . . ) & Q(e∗, x∗, . . . ) ∈ FA(P(e, x, . . . )) & x∗R x & x, x∗ ∈ S

     .P(e, x, . . . )

P, Q are (saturated) predicates x, x∗ are scalar entities (times, degrees, &c.) e, e∗ are eventuality variables R is a relation (e.g. ≺, ≻, ∝, &c.) S is a scale FA is a set of focus alternatives to P(e, x) which differ in terms of variation of elements under focus, which may include times, degrees,

• r subconstituents &c.

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 18 / 58

Formally defining Hindi phir, Nepali pheri

Utility of templatic definition

not only is this sort of template useful in terms of describing adverbs which are ambiguous in a manner consistent with underspecification (of certain parameters) but this also helps to make sense of historical developments

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 19 / 58

Historical developments involving aspectual adverbs

1

Aspectual adverbs in Hindi & Nepali

2

Formally defining Hindi phir, Nepali pheri

3

Historical developments involving aspectual adverbs

4

Additive particles & phir/pheri

5

Crosslinguistic comparisons - Hungarian &c.

6

Conclusion & refs

7

Additional materials Additional Hungarian diachronic & synchronic data More on other treatments

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 19 / 58

Historical developments involving aspectual adverbs

Historical developments

HindiHindi phir “then, again” is related to Hindi phirn¯ a “to turn”, which derives from a reconstructed Old Indo-Aryan *phirati “moves, wanders, turns”, cp. Prakrit phira¨ ı “goes, returns” (Turner 1966: #9078) Kutchi Gujarati (Patel-Grosz & Beck 2014) Pacho “again (repet. & restit.) & back” < OIA. *pa´ sca- “hinder part” (much like English back) [Turner 1966: #7990]

• cp. Hindi v¯

apas “back” (no repetitive senses (yet)), loanword from Persian, with the p¯ as part being cognate with *pa´ sca- [Platts 1884:1171]

• Cp. English again

Again originally meant “back, in the opposite direction” OE ongean: “He sceaf þa mid Dam scylde, Dæt se sceaft tobærst, and þæt spere sprengde, þæt hit sprang ongean.” [“He shoved then with shield so the shaft burst — the spear broke and sprang back.”](Battle of Maldon 137)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 20 / 58

Historical developments involving aspectual adverbs

Old English

Old English eft Similar “polysemy” is also found in earlier English eft: (14) Efterward afterward me

• ne

ssel shall þerne this mete food eft again chyewe chew ase as þe the

• xe
• x

þet...

• that. . .

“Afterward one shall chew this food again like the ox

• that. . . (CMAYENBI111.2146 — from Gergel et al. 2016)

[repetitive reading] (15) þone that.acc mon man eft afterwards

• n

in Cent Kent forbærnde. burned. “That man was afterwards burned in Kent.” [Anglo-Saxon Chron.

• ann. 685 (Parker MS.) ]

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 21 / 58

Historical developments involving aspectual adverbs

Historical patterns

a number of recurrent developments

development of ’back’ > ’again’, ’back’ (> ’again’)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 22 / 58

Historical developments involving aspectual adverbs

Historical patterns

a number of recurrent developments

development of ’back’ > ’again’, ’back’ (> ’again’) development of ’again’ > ’then’, ’again’

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 22 / 58

Historical developments involving aspectual adverbs

Historical patterns

a number of recurrent developments

development of ’back’ > ’again’, ’back’ (> ’again’) development of ’again’ > ’then’, ’again’ development of ’hinder parts’ > ’back’

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 22 / 58

Additive particles & phir/pheri

1

Aspectual adverbs in Hindi & Nepali

2

Formally defining Hindi phir, Nepali pheri

3

Historical developments involving aspectual adverbs

4

Additive particles & phir/pheri

5

Crosslinguistic comparisons - Hungarian &c.

6

Conclusion & refs

7

Additional materials Additional Hungarian diachronic & synchronic data More on other treatments

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 22 / 58

Additive particles & phir/pheri

Concessive/adversative “still”

Recall from above: for the “concessive still” sense, both Hindi & Nepali combine the temporal phir/pheri with a particle, Hindi bh¯ ı, Nepali pani Hindi (16) Shy¯ am Shyam gun .d .¯ a villain hai; is, phir then/again bh¯ ı too mer¯ a my dost friend hai. is “Shyam’s a villain; stillconc he’s my friend.” Nepali (17) Pheri then/again pani too timro your tasbir image ˜ ¯ akha.m¯ a eye.in ¯ ae.rahancha. come.continues “Stilltemp your image keeps coming into my eyes.”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 23 / 58

Additive particles & phir/pheri

Additive particles in Hindi and Nepali

The particles which show up in the “concessive still” in Hindi & Nepali, bh¯ ı & pani, are members of the µ-type (< Japanese mo) which appear in universal & conjunctive environments (see Szabolcsi 2010,2015, Slade 2011, Mitrovi´ c 2014, amongst other; cf. Reichenbach 1947) Japanese Dravidian Sinhala Nepali Hindi Hungarian µ-series mo um t pani bh¯ ı is, mind

Table: µ series in select languages

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 24 / 58

Additive particles & phir/pheri

Examples of bh¯ ı and pani as additive particles

Hindi (18) R¯ am Ram bh¯ ı also/even mar die gay¯ a went “Ram died too.”/“Even Ram died.” (19) R¯ am Ram bh¯ ı also Shy¯ am Shyam bh¯ ı also . . . . . . “Both Ram and Shyam” Nepali (20) R¯ am Ram pani even/also maryo died “Ram died too.”/“Even Ram died.” (21) R¯ am Ram pani also Shy¯ am Shyam pani also . . . . . . “Both Ram and Shyam”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 25 / 58

Additive particles & phir/pheri

Examples of bh¯ ı and pani as additive particles (cont.)

Hindi (22) Ko¯ ı Someone bh¯ ı also/even nah˜ ¯ ı not ¯ ay¯ a. came. “No-one came.” (23) Jo rel.pro bh¯ ı also/even lar .k¯ ı girl vah˜ ¯ a there khar .¯ ı standing hai, is, vah he/she mer¯ ı my dost friend hog¯ ı be.fut “Whichever girl is standing there will be my friend.” Nepali (24) Ma I kahile sometimes pani even/also raksi alcohol pi˜ udina drink.neg “I never drink alcohol.” (25) Jo who.rel.pro ¯ ae came pani even/also huncha is “Whoever comes, it is all right.”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 26 / 58

Additive particles & phir/pheri

‘Concessive/adversative’ phir bh¯ ı/pheri pani

Concessive

λSλwλeλP: ∃e∗ ∃Q ∃e′ ∃R

∃W cg ⊆W

        

R(e′,w,... ) & Q(e∗,... )∈FA(P(e,... )) &

Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }) < Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg }) Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }), Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg })∈S

        

.P(e,w,... ) W cg is the set of world consistent with the common ground because verum is focussed, FA(P(e)) = {P(e), ¬ P(e)} Λ(w′) = likelihood of w′ Σ({Λ(w′)| . . . }) is the aggregate of the likelihood of every world in a particular set. Thus both the number of worlds in the set and the individual likelihood of each particular world affects the result. S is an ordering of real numbers So here the overall likeliness of the worlds in which both the presupposed ‘frame-setting’ eventuality and the eventuality in question both occur is lower than the overall likeliness

• f the worlds in which the ‘frame-setting’ eventuality occurs but the eventuality in

question does not FAs = {p,¬p}

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 27 / 58

Additive particles & phir/pheri

‘Concessive/adversative’ phir bh¯ ı/pheri pani

Concessive

λSλwλeλP: ∃e∗ ∃Q ∃e′ ∃R

∃W cg ⊆W

        

R(e′,w,... ) & Q(e∗,... )∈FA(P(e,... )) &

Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }) < Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg }) Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }), Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg })∈S

        

.P(e,w,... )

involves extra components beyond the basic template, a “framing context” (“R”)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 28 / 58

Additive particles & phir/pheri

‘Concessive/adversative’ phir bh¯ ı/pheri pani

Concessive

λSλwλeλP: ∃e∗ ∃Q ∃e′ ∃R

∃W cg ⊆W

        

R(e′,w,... ) & Q(e∗,... )∈FA(P(e,... )) &

Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }) < Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg }) Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }), Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg })∈S

        

.P(e,w,... )

involves extra components beyond the basic template, a “framing context” (“R”) the focus is obligatorily on verum, so the only relevant alternative (“Q”) is the negation of the asserted clause (“P”)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 28 / 58

Additive particles & phir/pheri

‘Concessive/adversative’ phir bh¯ ı/pheri pani

Concessive

λSλwλeλP: ∃e∗ ∃Q ∃e′ ∃R

∃W cg ⊆W

        

R(e′,w,... ) & Q(e∗,... )∈FA(P(e,... )) &

Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }) < Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg }) Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }), Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg })∈S

        

.P(e,w,... )

involves extra components beyond the basic template, a “framing context” (“R”) the focus is obligatorily on verum, so the only relevant alternative (“Q”) is the negation of the asserted clause (“P”) presupposes that the summed likelihood (over all worlds consistent with the common ground) of the worlds in which “R” eventuality occurs and the “P” eventuality occurs is less than the summed likelihood of the world in which the “R” eventuality occurs and the “Q” eventuality occurs

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 28 / 58

Additive particles & phir/pheri

‘Concessive/adversative’ phir bh¯ ı/pheri pani

Concessive

λSλwλeλP: ∃e∗ ∃Q ∃e′ ∃R

∃W cg ⊆W

        

R(e′,w,... ) & Q(e∗,... )∈FA(P(e,... )) &

Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }) < Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg }) Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }), Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg })∈S

        

.P(e,w,... )

involves extra components beyond the basic template, a “framing context” (“R”) the focus is obligatorily on verum, so the only relevant alternative (“Q”) is the negation of the asserted clause (“P”) presupposes that the summed likelihood (over all worlds consistent with the common ground) of the worlds in which “R” eventuality occurs and the “P” eventuality occurs is less than the summed likelihood of the world in which the “R” eventuality occurs and the “Q” eventuality occurs where the only relevant possible “Q” is equivalent to “¬P”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 28 / 58

Additive particles & phir/pheri

‘Concessive/adversative’ phir bh¯ ı/pheri pani

Concessive

λSλwλeλP: ∃e∗ ∃Q ∃e′ ∃R

∃W cg ⊆W

        

R(e′,w,... ) & Q(e∗,... )∈FA(P(e,... )) &

Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }) < Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg }) Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }), Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg })∈S

        

.P(e,w,... )

involves extra components beyond the basic template, a “framing context” (“R”) the focus is obligatorily on verum, so the only relevant alternative (“Q”) is the negation of the asserted clause (“P”) presupposes that the summed likelihood (over all worlds consistent with the common ground) of the worlds in which “R” eventuality occurs and the “P” eventuality occurs is less than the summed likelihood of the world in which the “R” eventuality occurs and the “Q” eventuality occurs where the only relevant possible “Q” is equivalent to “¬P” i.e. “Q” & “P” is less likely than “Q” & “¬P”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 28 / 58

Additive particles & phir/pheri

‘Concessive/adversative’ phir bh¯ ı/pheri pani

Concessive

λSλwλeλP: ∃e∗ ∃Q ∃e′ ∃R

∃W cg ⊆W

        

R(e′,w,... ) & Q(e∗,... )∈FA(P(e,... )) &

Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }) < Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg }) Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }), Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg })∈S

        

.P(e,w,... )

e.g. “John is irritating; still he’s my friend.”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 29 / 58

Additive particles & phir/pheri

‘Concessive/adversative’ phir bh¯ ı/pheri pani

Concessive

λSλwλeλP: ∃e∗ ∃Q ∃e′ ∃R

∃W cg ⊆W

        

R(e′,w,... ) & Q(e∗,... )∈FA(P(e,... )) &

Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }) < Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg }) Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }), Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg })∈S

        

.P(e,w,... )

e.g. “John is irritating; still he’s my friend.” asserts both “John is irritating” and “John is my friend”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 29 / 58

Additive particles & phir/pheri

‘Concessive/adversative’ phir bh¯ ı/pheri pani

Concessive

λSλwλeλP: ∃e∗ ∃Q ∃e′ ∃R

∃W cg ⊆W

        

R(e′,w,... ) & Q(e∗,... )∈FA(P(e,... )) &

Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }) < Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg }) Σ({Λ(w′)|R(e∗,w′)∧P(e,w′)∧w′∈W cg }), Σ({Λ(w′′)|R(e∗,w′′)∧Q(e,w′′)∧w′′∈W cg })∈S

        

.P(e,w,... )

e.g. “John is irritating; still he’s my friend.” asserts both “John is irritating” and “John is my friend” presupposes that John being irritating and being my friend is less likely than John being irritating and not being my friend

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 29 / 58

Additive particles & phir/pheri

Concessive tree

AspP Concessive.still AspP event AspP Perf/Imperf vP subevent vP subject v VP subevent VP V

• bject

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 30 / 58

Additive particles & phir/pheri

Concessive

This sense is rather different from the “templatic” phir/pheri Unsurprising given the additional element bh¯ ı, pani The additive bh¯ ı/pani seems to correlate with the additional presupposition of a “framing” eventuality Also note the complexity of scalar elements (summation of likelihood

• ver sets of worlds)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 31 / 58

Crosslinguistic comparisons - Hungarian &c.

1

Aspectual adverbs in Hindi & Nepali

2

Formally defining Hindi phir, Nepali pheri

3

Historical developments involving aspectual adverbs

4

Additive particles & phir/pheri

5

Crosslinguistic comparisons - Hungarian &c.

6

Conclusion & refs

7

Additional materials Additional Hungarian diachronic & synchronic data More on other treatments

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 31 / 58

Crosslinguistic comparisons - Hungarian &c.

Comparison of (a subset of) Hungarian, Hindi, Nepali temporal adverbials

Hungarian Hindi Nepali Meaning megint phir pheri repetitive, ism´ et phir se restituitive m´ eg (mindig) abh¯ ı bh¯ ı aile samma temporal/continuity ab tak ajha(i) (pani) “still” aile pani m´ egis phir bh¯ ı pheri pani concessive/adversative akkor is tai pani “still” is bh¯ ı pani additive particle “also”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 32 / 58

Crosslinguistic comparisons - Hungarian &c.

Comparison of (a subset of) Hungarian, Hindi, Nepali temporal adverbials

Hungarian Hindi Nepali Meaning megint phir pheri repetitive, ism´ et phir se restituitive m´ eg (mindig) abh¯ ı bh¯ ı aile samma temporal/continuity ab tak ajha(i) (pani) “still” aile pani m´ egis phir bh¯ ı pheri pani concessive/adversative akkor is tai pani “still” is bh¯ ı pani additive particle “also”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 32 / 58

Crosslinguistic comparisons - Hungarian &c.

Repetitives in Hungarian

The Hungarian repetitives are etymologically connected to m´ eg via meg: (26) Feri F-nom megint again / / ism´ et again ivott drank egy

• ne

p´ alink´ at. palinka-acc ‘Feri drank a palinka again.’

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 33 / 58

Crosslinguistic comparisons - Hungarian &c.

Concessive m´ egis

Concessive interpretation (27) B´ ar though fogy´

• k´

ur´ azott, diet.V.past,3sg Feri F-nom m´ egis still evett ate zs´ ıroskenyeret. lard-adj.bread-acc ‘Even though he was on a diet, Feri still ate some bread with lard.’

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 34 / 58

Crosslinguistic comparisons - Hungarian &c.

Additives in Hungarian

Like Hindi phir bh¯ ı, Nepali pheri pani, Hungarian m´ egis contains an additive particle (28) J´ anos J. is also j¨

• tt.

came. “John came too.” (29) J´ anos J. is also (´ es) (and) Mari M. is also . . . . . . “Both John and Mary”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 35 / 58

Crosslinguistic comparisons - Hungarian &c.

Scalar additive

Scalar component from ‘m´ eg’, additive from ‘is’. (30) Mindenki everyone-nom zs´ ıroskenyeret lard-adj.bread-acc k´ ert. asked. M´ eg still Feri F-nom *(is) too zs´ ıroskenyeret lard-adj.bread-acc k´ ert. asked ‘Everyone asked for some bread with lard. Even Feri asked for some bread with lard.’

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 36 / 58

Crosslinguistic comparisons - Hungarian &c.

Another ordering relation in Hungarian

Hungarian m´ eg has (amongst other uses) a temporal ordering use, which interestingly is the inverse of the relation in the Indo-Aryan

• rdering phir, pheri

the asserted eventuality precedes (rather than follows) some presupposed eventuality (31) M´ eg still ´ ep´ ıtett built egy

• ne

h´ azat house-acc (miel˝

• tt

before meg meg halt) died ‘He built a house (before he died)’

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 37 / 58

Crosslinguistic comparisons - Hungarian &c.

Hungarian inverse ordering

m´ eg “(inverse) then; before that”: λTλtλeλP : ∃t∗ ∃e∗ ∃Q

    

Q(e∗, t∗, . . . ) & Q(e∗, t∗, . . . ) ∈ FA(P(e, t, . . . )) & t∗ ≻ t & t, t∗ ∈ T

     .P(e, t, . . . )

FAs, e.g. = {He built a cabin, He sneezed, He walked about, He

• died. . . }

identical to the “ordering then” sense of Hindi phir, Nepali pheri, except that the temporal ordering required is reversed

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 38 / 58

Crosslinguistic comparisons - Hungarian &c.

Comparison of Hungarian, Hindi, Nepali, German

Hungarian Hindi Nepali German Meaning akkor tab tab(a) da “then (at that time)” taile az (u)t´ an phir pheri dann “then (after that)” tab(a) pachi m´ eg — — — “before that” megint phir pheri wieder repetitive, ism´ et phir se restituitive ´ ujra (dob¯ ar¯ a) . . . m´ eg (mindig) abh¯ ı bh¯ ı aile samma noch temporal/continuity ab tak ajha(i) (pani) “still” aile pani m´ egis phir bh¯ ı pheri pani noch (denn?) concessive/adversative akkor is tai pani “still” tarai pani ra pani m´ eg . . . is bh¯ ı pani sogar scalar(-additive) particle “even” is bh¯ ı pani auch additive particle “also”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 39 / 58

Crosslinguistic comparisons - Hungarian &c.

Further echoes of additives:

Even in German, the form noch “still” historically contains an additive particle (as Hungarian is, Hindi bh¯ ı, Nepali pani):

German noch < PGmc. *nuh < PIE *n¯ u- “now” plus the PIE additive particle *-kwe (Pokorny 1959)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 40 / 58

Conclusion & refs

1

Aspectual adverbs in Hindi & Nepali

2

Formally defining Hindi phir, Nepali pheri

3

Historical developments involving aspectual adverbs

4

Additive particles & phir/pheri

5

Crosslinguistic comparisons - Hungarian &c.

6

Conclusion & refs

7

Additional materials Additional Hungarian diachronic & synchronic data More on other treatments

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 40 / 58

Conclusion & refs

Aspectual Adverbials - Different Scales, Relations, Foci

number of aspectual adverbs which are ambiguous underspecified, or etymologically/diachronically connected to other aspectual adverbs (or morphological composite) vary along particular dimensions (template):

Item Scale Relation Focus Identity of Scale/Focus “temporal” still time ∝ time yes “temporal additive” time ≺ time yes still “marginality” still degrees ≻ individuals no “concessive” still likelihood < verum no again time ≺ time yes then time ≺ non-time element no

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 41 / 58

Conclusion & refs

Summary

templatic relation between various senses of temporal/relational adverbs

• vertly manifested in morphology in some languages

can explain historical relationships (e.g. in Hungarian), or morphologically-connected paradigms of adverbs compositionality of composite (concessive) adverbs remains to be worked out in detail

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 42 / 58

Conclusion & refs

References

Beck, S. 2005. There and back again: a semantic analysis, JoS 22:3–51 Beck, S. 2016. Discourse related readings of scalar particles. SALT26 Benk˝

• , L. 1991. A magyar nyelv t¨
• rt´

eneti nyelvtana. Budapest: Akad´ emiai Kiad´

• Gergel, R., Bl¨

umel, A., & Kopf, M. 2016. . . . Notes from a Dying Adverb. PENN39 Horn, L. 1969. A presuppositional analysis of only and even, CLS 5:98–107 Ippolito, M. 2007. On the meaning of some focus-sensitive particles. NLS 15(1) Kayne, R. 2016. The unicity of there and the definiteness effect. Ms., New York Uni. Klein, W. 2007. About the German particles schon & noch. Ms., Max Planck Krifka, M. 1998. Additive particles under stress, SALT VIII, 92–110 Michaelis, L. 1993. ‘Continuity’ within 3 scalar models: . . . polysemy of . . . still. JoS 10 Mitrovi´ c, M. 2014. Morphosyntactic atoms. . . Uni. of Cambridge PhD Patel-Grosz, P. & Beck, S. 2014. Revisiting again . . . Gujarati. Sinn & Bedeutung 18 Platts, J.T. 1884. A dictionary of Urdu, classical Hindi, & English. London: W.H. Allen Pokorny, J. 1959. Indogermanisches etymologisches W¨

• rterbuch, Bern: Francke

Reichenbach, H. 1947. Elements of symbolic logic. New York: Macmillan Rooth, M. 1985. Association with Focus, UMass PhD Rooth, M. 1992. Theory of focus interpretation, NLS 1:75–116 Slade, B. 2011. Formal and philological inquiries. . . . Urbana: UIll PhD Szabolcsi, A. 2010. Quantification, Cambridge: CUP Szabolcsi, A. 2015. What do quantifier particles do? L & P 38:159–204 Turner, R.L. 1962–6. A compar. dict. of Indo-Aryan langs. London: Routlege Zaicz, G. (ed) 2006. Etimol´

• giai sz´
• t´

ar [Etymological dictionary]. Budapest: Tinta Kiad´

• Zimmermann, M. 2017. Scalar particles . . . in . . . Vietnamese. IATL2015

Hungarian Generative Diachronic Syntax (http://corpus.nytud.hu/hgds-dev/en-intro.html) B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 43 / 58

Additional materials

1

Aspectual adverbs in Hindi & Nepali

2

Formally defining Hindi phir, Nepali pheri

3

Historical developments involving aspectual adverbs

4

Additive particles & phir/pheri

5

Crosslinguistic comparisons - Hungarian &c.

6

Conclusion & refs

7

Additional materials Additional Hungarian diachronic & synchronic data More on other treatments

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 43 / 58

Additional materials Additional Hungarian diachronic & synchronic data

Origins of Repetitives in Hungarian

Hungarian m´ eg appears in a number of different senses (temporal “still”, additive “still”, &c.) m´ eg derives from the particle meg (Zaicz 2006), which could mean “again”: (32) (M¨ uncheni k´

• dex, 1466)

´ es and t˝

• n

did meg again ´ ugy like.that “and he did again like that” (?) Both megint, ism´ et are derived from meg (with an additional additive for ism´ et) Meg originally had the interpretation of ‘back’ (Zaicz 2006) (33) ‘Back’ (M¨ uncheni k´

• dex, 1466)

ne not menn´ enek went.3pl meg meg Her´

• deshez,

Herod-to m´ as different ´ uton road-on fordul´ anak turned.3pl meg meg ˝

• they
• rsz´

agukba country.pos.3pl-to “They didn’t return to Herod, they returned on a different road to their country.”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 44 / 58

Additional materials Additional Hungarian diachronic & synchronic data

meg as repetitive

(34) (M¨ uncheni k´

• dex, 1466)

´ es and t˝

• n

did meg again ´ ugy like.that “and he did again like that” (?) (35) (Szabolcs Viadala 1476) nagy big b´ us sad jonhhal ?-with meg-visszat´ er´ enek again-back.returned “they returned again with big sad ?” (36) esm´ eg (Jakob k´

• dex, 1440)

adjad give.imp esmeg (=ism´ et) again ennekem me “give it to me again”

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 45 / 58

Additional materials Additional Hungarian diachronic & synchronic data

Other repetitives

(37) ism´ et (M¨ uncheni k´

• dex, 1466)

ism´ et ism´ et mondom say.1sg tinektek you.pl-to ‘I say to you again’ (38) megint (Jord´ anszky k´

• dex, 1516-19)

´ es and ki who szent, saint szenteltess´ ek blessed.imp megint megint ‘and that who is a saint should be blessed again’

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 46 / 58

Additional materials Additional Hungarian diachronic & synchronic data

M´ eg as still

(39) Temporal m´ eg (B´ ecsi k´

• dex, 1430-60)

M´ eg m´ eg negyven forty napok days vannak are ‘There are still forty days’ (40) Concessive m´ egis (Bod k´

• dex, first half of 16th c)

de but m´ egis m´ egis t¨

• bbet

more-acc akar want vala had b´ ırnia have.inf ‘but still he wanted to have more’

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 47 / 58

Additional materials Additional Hungarian diachronic & synchronic data

Additives

(41) Additive (Birk k´

• dex, 1474)

˝ O he rajta

• n

is is k¨

• ny¨
• r¨

ult¨

• k

mercy.V.2pl ‘You(pl) have mercy on him as well’ (42) Scalar additive (M¨ uncheni k´

• dex, 1466)

m´ eg m´ eg ti you.pl is is ´ ertelem reason n´ elk¨ ul without vagytok-e are.2pl Q ‘Whether even you are without reason’ (43) Additive (Birk k´

• dex, 1474)

ne not csak

• nly

sz´ atok mouth.poss.2pl ´ etket, food-acc de but meg meg f¨ uletek ear.poss.2pl is too bevegye in.take.imp hallgassa hear.imp istennek god-dat ig´ ej´ et word.poss-acc ‘Let not only your mouth take in food, but also your ears take in and hear God’s word’

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 48 / 58

Additional materials Additional Hungarian diachronic & synchronic data

Other stills in Hungarian

comparative (44) Ez this egy

• ne

nagy big labda. ball (Az that nagyobb.) bigger ´ Es and az that m´ eg still nagyobb. bigger ‘This is a big ball. (That one is bigger.) And that one is still bigger.’ S-marginality [standard-marginality] (vs. “comparative” C-marginality) (45) Sopron Sopron-nom m´ eg still Magyarorsz´ agon Hungary-on van. is ‘Sopron is still in Hungary.’ (‘true’ marginality) (46) Ez this a the ruha dress dr´ aga. expensive Az that a the ruha dress is too / / #m´ eg still dr´ aga. expensive ‘This dress is expensive. That dress is expensive too / still expensive.’ (comparative marginality, only ok if temporal m´ eg)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 49 / 58

Additional materials Additional Hungarian diachronic & synchronic data

Additive m´ eg

Evidence for distinct structural relations (47) J. J. ivott drank egy

• ne

s¨

• rt.

beer-acc. ??M´ eg Still ivott drank egy

• ne

s¨

• rt

beer-acc / / egyet

• ne-acc

(miel˝

• tt

before el away indult). left ‘J. drank a beer. He drank another beer / one before he left.’ (48) J. J-nom ivott drank egy

• ne

s¨

• rt.

beer-acc M´ eg still ivott drank egy

• ne

k´

• l´

at coke-acc (miel˝

• tt

before el away indult). left ‘J. drank a beer. He drank a coke before he left.’

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 50 / 58

Additional materials More on other treatments

• Misc. Remarks on Previous Accounts

Michaelis (1993)

Examines three senses of English still with reference to diachronic development (temporal, marginality, concessive) Posits for each of the 3 that still denotes the existence of effectively identical elements at two contiguous scalar loci The more advanced locus is asserted, the less advanced presupposed Scales may differ (times, worlds, rankings along property scale) Nb: problematic claim about homogeneity of contiguity requirement

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 51 / 58

Additional materials More on other treatments

• Misc. Remarks on Previous Accounts (cont.)

Ippolito (2007)

Examines English still and already, discussing aspectual/temporal, marginality, concessive uses (relating these to additive particles, scalar particles, and exclusive particles), also investigating again Also notes focussing of time variable for (temporal) still & again Complex interactions of adverbials with aspectual heads In some cases definitions perhaps overspecified

Concessive still: requires worlds considered to be maximally similar to evaluation world Argues that concessive involves presupposition that the set of worlds in which the framing proposition (John studied all night) and the proposition at issue ((still) he failed the exam) are both true are less likely the worlds in which the framing proposition is not true but the proposition at issue is true - which seems incorrect

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 52 / 58

Additional materials More on other treatments

• Misc. Remarks on Previous Accounts (cont.)

Beck (2016)

Focuses on German noch & English still, in various senses, including temporal, marginal, and various “discourse”-related (though not concessive) Also posits common core/template, but which like Michaelis’s problematically posits contiguity (=abutment) for all senses Also explores implicatures

Beck suggests that temporal noch/still carries implicature that P will hold in future (explain oddity of John’s still dead) In actuality, implicature seems to involve P being true in some accessible world (John is still annoying)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 53 / 58

Additional materials More on other treatments

Vietnamese

(49) Tˆ an Tan.top,1 th` ı prt thˆ a ˙ m ch´ ı scal 1.7m 1.7m c˜ ung add nhay qua jump ‘Tan can even jump 1.7m.’ [Zimmermann 2017: 141] thˆ a ˙ m ch´ ı. . . c˜ ung is reminiscent of the m´

• eg. . . is of Hungarian

Zimmerman (2017) argues that thˆ a ˙ m ch´ ı is a scalar(-additive) particle and c˜ ung is an additive However, the contexts for the Vietnamese examples seem to largely involve a contrastive topic in addition to the scalar-additive and thus apparently differ from Hungarian Further, Zimmermann (2017: 140) notes that c˜ ung sometimes also appears to bear a scalar reading even without the scalar thˆ a ˙ m ch´ ı

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 54 / 58

Additional materials More on other treatments

Remaining problems

more fully work out syntax and compositional account e.g. explain precise contribution of additives in forming “concessive still”

for bh¯ ı/pani, ambiguous between additive, and scalar-additive, possible explanation is clearer (scalar additives typical rank according to likelihood) but Hungarian is seems to be a plain additive (unless it was similarly ambiguous at an earlier stage)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 55 / 58

Additional materials More on other treatments

Other accounts

Michaelis 1993, Ippolito 2007, Beck 2016 on various senses of still (among others) Focus of papers: no morphological relevance, no templatic definition Concern (Michaelis 1993): aspectual restrictions Morphological facts (also later), role of additive particles

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 56 / 58

Additional materials More on other treatments

Other accounts: abutment

Abutment in temporal interpretation For others, abutment survives in other interpretations, e.g. the marginality sense No necessary abutment relation (e.g. comparative, C-marginality, perhaps also just an implicature in S-marginality)

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 57 / 58

Additional materials More on other treatments

Other accounts: temporal implicature

Beck 2016: temporal implicature with temporal still It’s still raining implicates that it won’t be raining in the future (Beck) This dress is still expensive: no necessary entailment, presupposition

• r entailment that it will be cheaper at a later time

B Slade & A Csirmaz (Uni. of Utah) Then too. . . Meaning in Flux / Yale 58 / 58