Zenica 2015 pilot experiment double participles partial

24.8.2017

ZENICA 2015 PILOT EXPERIMENT

AGREEMENT OF DOUBLE PARTICIPLES SANDWICH & POSTVERBAL ALJOVIĆ, N., BUREKOVIĆ M., & GRUBEŠIĆ G.

EMSS ANNUAL MEETING SARAJEVO, 1-2 FEB. 2016

UNIVERSITY OF ZENICA TEAM EXPERIMENTERS: NADIRA ALJOVIĆ MELISA BUREKOVIĆ GORAN GRUBEŠIĆ DZENANA KOTORIĆ ŠEJLA KOVAČEVIĆ

THREE PARALLEL SESSIONS IN A ROW EXPERIMENT RUN ON DECEMBER 10, 2015 30 PARTICIPANTS STUDENTS OF UNZE NATIVE SPEAKERS AGE: 18-25

LARISA KRAJINIĆ

1

24.8.2017

Introduction

Results

Observations

Double participles

AJ

Different grammars

Experimental methods and materials

RT

Correlations

DOUBLE PARTICIPLES IN PLUPERFECT TENSES BCS PLUPERFECT : TWO PARTICIPLES AGREEING WITH SUBJECT IN G/Nb, AUX.VERB AGREEING WITH SUBJECT IN P/Nb SUBJECT

AUX

PART1

PART2

Djevojka je bila izgubila girlFSG is beenFSG lostFSG The girl had been lost her/the umbrella.

OBJECT

kišobran. umbrella

2

24.8.2017

PLUPERFECT SEMANTICS I - PAST BEFORE ANOTHER PAST (STATE, ACTION…) 1 A. Zašto mu nisi pokazao pismo? (Why didn't you show him the letter?) B. Nisam mogao. Već sam ga bio predao na poštu. (I couldn't. I had already mailed it.) 2 Bio sam popio rakije pa mi je vjerovatno zato bilo loše. (I'd had some brandy and that's probably why I didn't feel well.) (=Popio sam… - regular past tense) NOT (ALWAYS) NECESSARY, CAN BE REPLACED BY REGULAR PAST TENSE

"UNDO" PLUPERFECT PAST (ACTION, STATE…) UNDONE IN THE MEANTIME Examples with "undo" implications: 3. Bila se udala za direktora firme. (She married the manager of the company – but they have since divorced) 4. Bili smo kupili karte za concert. (We bought tickets for the concert – but we didn't use them (because we sold them/gave them away/the concert was cancelled/etc.)) 5. Bili su najavili prenos utakmice za 8 sati. (They had announced the broadcast of the match for 8 o'clock – but did not realize it (for some reason, known or unknown)) Compare the effect with the English pluperfect forms: 5'. I had hoped to meet them in London. …. I had expected to finish the exam by 5 o'clock.

3

24.8.2017

MORE NATURAL PLUPERFECT SENTENCES AND WORD ORDERS Pa, Ma, Ali … (But …) Pa bilo je jezero zaleđeno prošle zime! Ma bilo je jezero zaleđeno prošle zime! Ali bilo je jezero zaleđeno prošle zime! (But, the lake was indeed frozen last winter.)

• Pa bila su jezera i rijeke zaleđene prošle zime! • Ma bile su smotre i takmičenja otkazane prošle subote! • Ali bila su obavljena ispitivanja i rasprave prošlog mjeseca. • Dapače, bile su odluke i rješenja uručena svima. • Naprotiv, … • Navodno,… • Međutim, ?

SPECIAL FOCUS OF THE EXPERIMENT AGREEMENT OF THE TWO PARTICIPLES (PLUPERFECT) WITH TWO CONJOINED NOUNS (SUBJECT)

SANDWICH:

Participle1 N1 & N2 Participle2

POSTVERBAL: Participle1 Participle 2 N1 & N2

6. Jesu li [bila jezera i rijeke presušile] preko ljeta? (N-N-F-F) 7. Jesu li [bila prodana imanja i štale] na licitaciji?

(N-N-N-F)

4

24.8.2017

EXPERIMENTAL METHODS AND MATERIALS EXPERIMENT DESIGN DATA ADAPTATION PARTICIPANTS TASK

EXPERIMENT DESIGN TWO LEVEL FACTORIAL DESIGN Double participles tested in experiment appear in TWO configurations with respect to the subject (two levels): • SANDWICH: the auxiliary participle (bil-) appears before the subject, the lexical verb participle appears after the subject: Participle1

S Participle2

• POSTVERBAL: both participle appear before the subject (i.e. the subject is postverbal) Participle1 Participle 2

S

All subjects in the stimuli are conjunction phrases (ConjP) with feminine and neuter plural nouns, appearing in both orders (FN, NF).

5

24.8.2017

EXPERIMENT DESIGN PATTERNS OF AGREEMENT TESTED IN THE SANDWICH LEVEL 1.

SPLIT: The aux. participle agrees with the first noun, the lexical v. participle agrees with the second noun: Part 1 N1 N2 Part 2

2.

DOUBLE HCA: Both participles agree with the highest conjunct, i.e. the first noun in the ConjP: Part 1 N1

3.

N2 Part 2

DOUBLE LCA: Both participles agree with the lowest (last) conjunct, i.e. the second noun in the ConjP:

Part 1 N1

N2 Part 2

4. DOUBLE DEFAULT: Both participles agree in Mpl (default): Part 1 N1

N2 Part 2

EXPERIMENT DESIGN PATTERNS OF AGREEMENT TESTED IN THE SANDWICH LEVEL Split NF, FN

Prošle sedmice bila su zvona i kugle postavljene na jelku. Jesu li bile makaze i ogledala stavljena u prodaju?

Double HCA NF, FN

Kao i obično, bila su pisma i razglednice poslana na vrijeme. Jesu li bile olovke i nalivpera spremljene u torbu?

Double LCA NF, FN

Jesu li bile poglavlja i rasprave predane na čitanja? Prošli utorak, bila su skice i platna stavljena na licitaciju.

Default NF, FN

Prvobitno, bili su opravdanja i molbe prihvaćeni na sudu. Jesu li bili škole i igrališta obnovljeni preko ljeta?

6

24.8.2017

EXPERIMENT DESIGN PATTERNS OF AGREEMENT TESTED IN THE SANDWICH LEVEL

NOTE: A split pattern mixing partial and default agreement is missing from the design:

Sandwich Split-mix

ConjP

Part1

Part2

HCA- default

NF

N

default

HCA - default

FN

F

default

Part 1 N1 N2 Part 2 Prvobitno su bila opravdanja i molbe prihvaćeni na sudu. (at first are been-Npl excuses-Npl and requests-Fpl accepted-Mpl in court)

EXPERIMENT DESIGN PATTERNS OF AGREEMENT TESTED IN THE POSTVERBAL LEVEL

1.

POSTVERBAL CCA: Both participles agree with the closest (also highest) conjunct, i.e. the first noun in the ConjP: Part 1 Part 2 N1 N2

2.

POSTVERBAL LCA: Both participles agree with the last conjunct, i.e. with the second noun in the ConjP: Part 1 Part 2 N1 N2

3.

POSTVERBAL DEFAULT: Both participles agree in Mpl (default). Part 1 Part 2

N1 N2

In total: SEVEN agremeent patterns (types): 4 in the Sandwich + 3 in the Postverbal level

7

24.8.2017

EXPERIMENT DESIGN PATTERNS OF AGREEMENT TESTED IN THE POSTVERBAL LEVEL CCA NF, FN

Prošle sedmice, bila su objavljena ispitivanja i rasprave u časopisu. Jesu li bile donirane igračke i pomagala za tombolu?

LCA NF, FN

Jesu li bile naručivane piva i kafe za onim stolom? Prošlog mjeseca, bila su pronađena skice i platna u podrumu.

Default NF, FN

Unatoč zakonu, bili su izbačeni pitanja i provjere iz udžbenika. Jesu li bili korišteni olovke i nalivpera u nastavi?

EXPERIMENT DESIGN CONDITIONS AND ITEMS • Agreement type x two gender orders: N-F and F-N • This makes: 14 CONDITIONS: 8 (4x2) for the Sandwich and 6 (3x2)for the Postverbal level. • Condition x four sentences (ITEMS). Total Stimulus:14x4= 56 ITEMS

8

24.8.2017

EXPERIMENT DESIGN CONDITIONS AND ITEMS Type Sandwich Split Sandwich Split Sandwich Double HCA Sandwich Double HCA Sandwich Double LCA Sandwich Double LCA Sandwich Double Default Sandwich Double Default Postverbal CCA Postverbal LCA Postverbal CCA Postverbal LCA Postverbal Default Postverbal Default

ConjP NF FN NF FN NF FN NF FN FN FN NF NF FN NF TOTAL:

Conditions Part1 N F N F F N M M -

Part2 F N N F F N M M F N N F M M

Item Number 4 4 4 4 4 4 4 4 4 4 4 4 4 4 56

DATA ADAPTATION 1. A few lexical adaptations: sukladno > u skladu sa, bilježnice > sveske, tjedan > sedmica 2. The pluperfect is not a frequent verb tense (at least in our variant). It has the "undo" effect Prvobitno, bila je molba prihvaćena na sudu. instead of Sasvim neočekivano, bila je molba prihvaćena na sudu. 3. Word order in Sandwich (Part 1 Subject Part 2) creates a semantic effect: ‘The request had indeed been accepted ...’

9

24.8.2017

DATA ADAPTATION Part1 preceding Subject receives special QUESTIONS/SUGGESTIONS:

intonation: BILA je molba prihvaćena ...

1. Is the initial adverbial necessary? An adverbial at the beginning, with comma intonation, seems in conflict with this effect, the sentence may sound odd : ? Prvobitno, BILA je molba prihvaćena na sudu.

2. Test Sandwich patterns with predicative adjectives: e.g. Bile su kuće i polja zapušteni/a/e

?? Sasvim neočekivano, BILA je molba prihvaćena na sudu.

PARTICIPANTS • 30 students

• aged between 18-25 • 63,3% female, 36,6% male • native speakers who grew up in the Zenica region • not students of B/C/S • participated voluntarily

10

24.8.2017

TASK • the subject sees a sentence on the screen and reads it (to herself NOT OUT LOUD)

• the subject evaluates the sentence on the screen by clicking the mouse on the number from 1 to 4 (1 denotes weakly acceptable sentence and 4 strongly acceptable sentence)

• the number of sentences tested: a total of 110 sentence items (56 stimulus + 54 fillers)

RESULTS SANDWICH VS. POSTVERBAL AJ AJ Scale Combined

• SANDWICH:

1.9

• POSTVERBAL: 2.1

2.1 2.1 2.0

(+10.55%; 6.66% OF THE RANGE)

2.0

• BOTH BELOW THE SCALE MEDIAN VALUE

1.9

(2.5)

1.9 1.8 Sandwich

Postverbal AJ Scale

11

24.8.2017

RESULTS SANDWICH VS. POSTVERBAL RT RT Combined 1374.0

• SANDWICH RT (AVERAGE):

1373.7 MS

1373.5 1373.0 1372.5 1372.0

• POSTVERBAL RT (AVERAGE): 1370.8 MS

1371.5 1371.0 1370.5

(+0.21%)

1370.0 1369.5 1369.0 Sandwich

Postverbal RT

RESULTS AJ Scale Combined

• SANDWICH CONSTRUCTIONS

3.0 2.5 2.0

• SPLIT: 2.1

1.5

• DOUBLE HCA: 2.2 • DOUBLE LCA:

1.5

(NO HC AGREEMENT)

1.0 0.5 0.0 Split

Double Double HCA/NP1 LCA/NP2

Double Default

Postverbal Postverbal Postverbal CCA/NP1 LCA/NP2 Default

AJ Scale

12

24.8.2017

RESULTS DOUBLE DEFAULT AJ AJ Scale Combined 3.0

• POSTVERBAL: 1.8 2.5

Unatoč pravilniku bili su skinuti

2.0

pitanja i rasprave s dnevnog reda.

1.5

• SANDWICH: 1.8

U skladu sa zakonom bili su rješenja i

1.0

odluke objavljeni na mreži.

0.5

(BOTH WELL BELOW 2.5)

0.0 Split


Double Default


AJ Scale

RESULTS HIGHEST AND LOWEST AJS • POSTVERBAL HCA (CCA): 2.6 (THE ONLY ABOVE 2.5) Prošle sedmice bile su pale grane i gnijezda na zemlju.

AJ Scale Combined 3.0 2.5 2.0 1.5 1.0

• SANDWICH DOUBLE LCA: 1.5 U vezi sa predmetom bila su odluke i upozorenja

0.5 0.0

izrečena na sudu. AJ Scale

13

24.8.2017

RESULTS AJ Scale Combined 3.0

COMPARING:

2.5

• POSTVERBAL LCA: 1.7

2.0

(NO HC OR CC AGREEMENTS)

1.5

Jesu li bile naručivane piva i kafe za onim stolom?

1.0

• SANDWICH DOUBLE LCA: 1.5

0.5 0.0

(ONE CC AGREEMENT)

Split


Jesu li bila salate i vina poslužena u vrtu?

Double Postverbal Postverbal Postverbal Default CCA/NP1 LCA/NP2 Default AJ Scale

RESULTS HIGHEST SCORING AJs AJ Scale Combined 3.0

• 1. POSTVERBAL CCA: 2.6

2.5

• 2. SANDWICH DOUBLE HCA: 2.2

2.0

• 3. SANDWICH SPLIT: 2.1

1.5 1.0

ALL THREE INCLUDE AT LEAST

0.5

ONE PARTICIPLE – HIGHEST CONJUNCT AGREEMENT

0.0 Split

Double HCA/NP1

Double LCA/NP2

Double Default

Postverbal CCA/NP1

Postverbal LCA/NP2

Postverbal Default

AJ Scale

14

24.8.2017

RESULTS • 1. POSTVERBAL CCA (HIGHEST AJ: 3.17; AND LOWEST RT: 1234.5 MS; N N CONJP NF):

Prošle sedmice bila su objavljena ispitivanja i rasprave u časopisu. • 2. SANDWICH DOUBLE HCA (HIGHEST AJ: 2.53; THIRD LOWEST RT: 1442.3 MS; F CONJP FN F):

Prošlog vikenda bile su smotre i takmičenja zakazane u dvorani. • 3. SANDWICH SPLIT (HIGHEST AJ: 2.60; FIFTH LOWEST RT:1475.9; N CONJP NF F - QUESTION)

Jesu li bila priznanja i nagrade dodjeljene na proslavi?

RESULTS DEFAULT RT COMPARISON RT Combined • SANDWICH DEFAULT: 1368 MS • 0.5% FASTER THAN SANDWICH MEAN

1440.0 1420.0 1400.0

Jesu li bili drva i grane ispucali pod snijegom? 1380.0 1360.0

• POSTVERBAL DEFAULT: 1430 MS (+4.5%) (THE HIGHEST RT) 4.3% SLOWER THAN POSTVERBAL MEAN

Jesu li bili presušili jezera i rijeke preko ljeta?

1340.0 1320.0 1300.0 1280.0

Split


Double Postverbal Postverbal Postverbal Default CCA/NP1 LCA/NP2 Default RT

15

24.8.2017

RESULTS

AJ Scale Combined

3.0 2.5

RT AND AJ DISCREPANCY

2.0 1.5

• SPLIT SANDWICH AJ: 2.1(3RD BEST)

1.0

• SPLIT SANDWICH RT: 1420 MS (2ND HIGHEST)

0.0

AJ Scale

0.5 Split

Jesu li bila istraživanja i studije objavljene u petak?

HCA ANTICIPATED, BUT SPLIT ALSO ACCEPTED (ALBEIT SLOWER)?


Double Postverbal Postverbal Postverbal Default CCA/NP1 LCA/NP2 Default

RT Combined 1440.0 1420.0 1400.0 1380.0 1360.0 1340.0 1320.0 1300.0 1280.0

RT

Split


Double Default


16