ComparativealternationinMauritian:Using experimental methods forunderstudied

languages

AMLaP Asia – University ofHyderabad(India)02/02/2018

S.Hassamal (MIE,LLF)&B.Hemforth (LLF,CNRS)

Outline

1. Introduction2. Comparativeadverbs inMauritian3. Superiority comparatives

3.1. Experimentalstudy inMauritian3.2. Experimentalstudy inFrench3.3. Conclusions

4. Conclusions

1

IntroductionExperimentallinguisticsforCreolelanguages

• Wewillshowinthispresentation– inwhichwayexperimental linguisticsiscentraltostudylinguistichypotheses

– inparticularforlanguageswherelargecorporaareunavailable

– andwhereparadigmshavetobeadaptedtotheculturalenvironment

2

IntroductionThephenomenon:Degree adverbs

completely,too,very…Semantics :• modify avalueonascale

• anintensity scale forproperties or• aquantity scale forobjects andevents (Kennedy&McNally 2005)

Syntax :• modify any ofthemajorcategories

• nouns,verbs,adjectives,adverbs (Abeillé&Godard2003).

• troppeur(too much fear),tropboire(too muchdrinking),tropgrand(too big),tropvite(too quickly)

3

IntroductionMauritian:aquickoverview

• French-basedcreole,with 90%ofthelexiconinherited from French.

• Recent standardisedwritten form (Hookoomsing 2004)

• Fewwritten corpora (Lalit,Virahsawmy...)

4

Comparativeadverbs inMauritian

Comparative adverbs belong to the larger class ofdegree adverbs, comparing two values on one or moredegree scales.

We will focus on:• Inequality adverbs, and particularly

- Superiority comparative: pli / plis ‘more’

5

Researchquestion

• Two questions are motivating this research:–What are the constraints for superioritycomparatives inMauritian?

– Are these constraints different from thedistribution in French?

6

Superiority comparativesinMauritian

Table1:DistributionofpliandplisinMauritian (from corpora)

7

Experimental study inMauritian

QuestionàWhich factors determine the distribution of pli and plis in

Mauritian?

Adaptation of the experimental paradigm to Mauritian:• essentially a spoken language with a recent standardised

written form• there is a difference in the written form : pli and plis

(contrary to French where it is only phonological: /ply/ and/plys/)

à Production task with visual and audio stimuli

8

Experimental study inMauritian:Participants

• 26nativeMauritian speakers(tested inMauritius).• 17women&9men• Aged between24and66years old• Fromdifferent regionsoftheisland andwith different

levels ofeducation• They allfilled anadapted versionofthePennState

University LanguageHistoryQuestionnaire• Theywerepaid Rs.100(approximately 2.5€)

9

Experimental study inMauritian:Materials7 conditions:

• Nouns:• associatedwithanintensityscale(fin ‘hunger’).• associatedwithaquantityscale(kilomet ‘kilometre’)

• Verbs:• associatedwithanintensityscale(grosi ‘tobecomefat’)• associatedwithaquantityscale(alsinema ‘togotothecinema’)

• Adjectives:• associatedwithanintensityscale(long ‘tall’)• associatedwithaquantityscale(absan ‘absent’)

• Adverbs(vit ‘fast’)(nosemanticdistinction)

28experimental items(4itemsofeach condition)and28fillersDuration:approx.30minutes

10

Experimental study inMauritian:Procedure

We used E-prime tosetuptheexperimentPresentation was randomised indvidually foreach participant.

Unfolding of the experiment …

Step1: The participant sees a pair of images and hears a description (8a).

Step 2: The participant presses any key to proceed to the next screen. Theparticipant sees the same pair of images, with another voice message (thistime incomplete) (8b).

(8) a. Rita ek Sunil kondir vit.‘Rita and Sunil drive fast.’

b. Me Sunil kondir ___ vit ki Rita.‘But Sunil drives ___ fast than Rita.’

11

Experimental study inMauritian:Materials

Fillers had the same format:

(9) a. Rita so labous inn brile.‘Rita burnt her mouth.’

b. Li’nn bril so labous parski lasoup la ____ so.‘She burnt her mouth because the soup was ___ hot.’

Participant’s task:• repeat the second (incomplete) descriptions (8b, 9b) and complete the short

sentence by filling in the blank from a list of degree words:

pli, plis, mwin, mwins, tro, mari, osi , otan

The participants were recorded using Audacity.

12

(8) Rita ek Sunil kondir vit.‘Rita and Sunil drive fast.’

13

Me Sunil kondir ___ vit ki Rita.‘But Sunil drives ___ fast than Rita.’

14

Experimental study Mauritian:Results

15

Production of plis per condition:

Experimental study inMauritian:Results

Logistic regressions for choices of pli and plis show that:• Overall plis is significantly more frequent for nouns and verbs than for

adjectives and adverbs• Overall, plis is significantly more frequent for a quantity scale than for

an intensity scale.• Syntactic and semantic factors interact signifcantly:

• While there is no significant difference for verbs per semantic scale, thereis such a difference for adjectives (with a marginal difference for nouns).• Pli prefers adjectives that are associated with an intensity scale• Plis prefers adjectives that are associated with a quantity scale

16

Experimental study inFrench

Question:àDoes this alternation come from French?

Spoken French :-3 forms for plus : /ply/ , /plys/ and /plyz/ (liaison) (Léon, Bonami &Delais Roussarie)

17

Experimental study inFrench:Participants

• 29Frenchspeakers• 18women&11men• Aged between 21and62years old• AllliveinParisandwere recruited from the

RISC• They allcompletedalanguage history

questionnaire• Theywere paid 4€

18

Experimental study inFrench:Materials

7conditions:• Nouns

• associatedwithanintensityscale(faim‘hunger’).• associatedwithaquantityscale(kilomètre ‘kilometre’)

• Verbs• associatedwithanintensityscale(grossir‘to becomefat’)• associatedwithaquantityscale(aller aucinéma‘to gotothecinema’)

• Adjectives• associatedwithanintensityscale(grand‘tall’)• associatedwithaquantityscale(absent‘absent’)

• Adverbs (vite ‘fast’)(We donotdifferentiate between semantic typesofadverbs)

28experimental items(4itemsofeach condition)and28fillersDuration:approx.20minutes

19

Experimental study inFrench:Procedure

We used E-prime to set up the experimentPresentation was randomised indvidually foreach participant.

Unfolding of the experiment …

Step1: The participant sees a pair of images and hears a description (10a).

Step 2: The participant presses any key to proceed to the next screen. Theparticipant sees the same pair of images, with another voice message (this timeincomplete) (10b).

(10) a. Marie et Pierre conduisent vite.‘Marie and Pierre drive fast.’

b. Mais Pierre conduit ____ vite que Marie.‘But Pierre drives ____ fast than Marie.’

20

(10) Marie et Pierre conduisent vite.‘Marie and Pierre drive fast.’

21

Mais Pierre conduit ___ vite que Marie.‘But Pierre drives ___ fast than Marie.’

22

Experimental study inFrench:Results

Production of /plys/ per condition:

23

Experimental study inFrench:Results

Logistic regressions for choices of /ply/ and /plys/ show that:

• Overall /plys/ is significantly more frequent for a quantity scalethan for an intensity scale.

• Syntactic and semantic factors interact significantly:• While there is no significant difference for verbs for semantic scale,

there is such a difference for adjectives.

• Results from this study are corroborated by a reading aloudstudy.

24

Conclusions:ComparingMauritianandFrench• MauritianandFrenchshowverysimilarpatterns• However,numerically,thesemanticfactorseemsstrongerinMauritian

• Twowayinteractionsarenotsignificantforsemantics*language neitherfornounsnorforadjectives

• WehadlessandmoreheterogeneousMauritianparticipants• Currentstudies lookintointerindividual differences:Thesemanticfactors

seemtobestrongerdependingon• Education(lessstrongforhighereducation)• Andlinguisticbackground(strongerforparticipantswhospeakFrenchless

frequently)

25

Mauritian French

Conclusions

• Newfieldwork methods• Mauritian :newmethodology forexperimental syntax and

semantics• visual andoralstimuli&visual andoraldistractors

• French:samemethods can be applied.

à Isthere aFrenchinfluenceonMauritian comparativealternation?

Probably:Similar factors areatwork forcomparativealternations.

But:Interindividual variationmay play arole aswell forMauritian• Language background• Education

• Moredetailed studies areplanned andcurrently running

26

Thank you !

27