The Influence of Cognitive and Psychological ...

3 downloads 0 Views 360KB Size Report
Topicality (Rosenbach 2002, 2008,) has been defined in terms of givenness, definiteness, ..... test to indicate the unfamiliar words on the vocabulary list. Almost ...
www.sienajournals.com

Perspectives in Language, Linguistics and Media 2 (2017) 111-148

The Influence of Cognitive and Psychological Factors on the Syntactic variation of Genitive Constructions: A comparison of English & Greek Language Shazia Akbar Ghilzai 1 ___________________________________________________________________________ Abstract The present research examined the influence of cognitive factors on the choice of genitive constructions. A much debated question is, whether the unified treatment of possessive patterns is appropriate or not. Are they subject to language specific constraints or cognitive factors? (For discussion see: Alexiadou 2005). The present research attempted to observe the influence of both language specific constraints and cognitive factors on the choice of possessive constructions. Eight types of syntactic variants of genitive constructions were used in a psycholinguistic experiment. The syntactic variants were aligned with the cognitive factors like animacy, topicality and prototypicality. The participants’ choice of genitive constructions across these eight types of genitive constructions were recorded and analyzed. Twenty monolingual native English speakers and twenty L2 Greek speakers of English language participated in the experiment. The study revealed that the control group chooses the genitives according to the set predictions. Their choice indicates the effect of cognitive factors like animacy and prototypicality. While, L2 learners deviate in their choice from native speakers and their pattern of deviation was matched with the typological difference of the L2 learner’s first language i.e. Greek language. The rating of L2 Greek learners was not based on animacy and prototypicality. This shows that there is L1 transfer that is affecting on their choice of genitive construction. There was no effect of Topicality/definiteness as the earlier studies proposed (Rosenbach 2002, 2008,). Longobardi (1996) proposed that possessive constructions do not show uniform behavior across languages with respect to definiteness. The reason is that in the present research, the experimental material was not embedded into any discourse like earlier studies and phrases were presented separately. Corresponding author: Shazia Akbar Ghilzai, Department of Linguistics, QAU. Doctoral student at USPC Paris13. Acknowledgement: the author is highly thankful to Sonja Eisenbeiss (the University of Essex, UK) for her guidance throughout the project. 1

111

www.sienajournals.com

Topicality seems to be a discourse oriented factor; therefore it couldn’t show any effect in the present research. Keywords: cognitive factors, animacy, topicality, prototypicality

Introduction Previous studies investigated a variety of factors that influence the choice between of s’ genitive and of genitive constructions. Some studies have focused on phonological, pragmatic, syntactic and morphological factors (Altenberg 1982, Jucker 1993, Hawkins 1994, Biber et al. 1999, Jahr sorheim 1980, Jespersen 1961, Quirk et al. 1985) while others have considered the cognitive and psychological factors (e.g. Taylor 1989. 1996, Heine 1997, Rosenbach 2002, 2005, 2007). The present research takes into account the later one i.e. cognitive and psychological factors (e.g. Animacy, Topicality, prototype and syntactic weight etc.) The nature of possession has been described in terms of binary factors i.e. animate or inanimate, alienable or inalienable, prototypical or non prototypical. Topicality (Rosenbach 2002, 2008,) has been defined in terms of givenness, definiteness, familiarity, identifiability and referentiality. When a speaker addresses a person, he knows what information or knowledge is already given in the addressee’s mind. In other words, the speaker knows that the addressee is already familiar with some background knowledge and information and will identify it. Identifying something means activating that entity or concept in mind i.e. memory activation. When the speaker introduces the new concepts to the addressee, he anchors the new information with the previous information. While constructing a possessive construction speaker uses a special strategy for anchoring the PM. The concept of topicality is strongly interrelated to the grammatical category of definiteness. Definiteness, in terms of conceptual properties is closely related to the familiarity, 112

www.sienajournals.com

identifiability and givenness. Definiteness is expressed by the use of definite or indefinite expressions. When a speaker uses the definite expression he assumes that the addressee is either familiar with the referent or is able to identify it. According to Gundel, Hedbergand Zacharsky (1993: 277) type of the linguistic expression referring to an entity is indicative of the assumed cognitive status of that entity or referent. They presented the following givenness hierarchy (fig. 2.3.1) to describe the shape of the NP according to the assumed cognitive status of the referent. In focus > activated > familiar > uniquely identifiable > referential > type identifiable It

that this

that N

the N

indefinite

aN

this N

this N Fig.2.3. Illustrates givenness hierarchy (Gundel, Hedbergand ,Zacharsky 1993: cited by Rosenbach 2002:52) The definiteness according to this hierarchy is defined as “A definite noun phrase is taken to be definite if its referent is at least uniquely identifiable. Which means that the addressee can identify the speakers intended referent on the basis of the nominal alone.... in the givenness hierarchy the term uniquely identifiable is associated with definite NPs” (Gundel, Hedbergand ,Zacharsky 1993: cited by Rosenbach 2002:52) Another conceptual distinction related to the term topicality is the distinction between referential and non referential expressions. Referential expressions refer to an existing referent while non referential expressions do not refer to anything; mostly they are descriptive or indicate the class membership e.g. The Lion is a dangerous animal (Rosenbach 2002). Although the NP is definite in this example but it’s descriptive not referential. Definite 113

www.sienajournals.com

Referential expressions are mostly personal pronouns, proper nouns or noun phrases marked by definite article, demonstrative or possessive pronoun. Personal pronouns or proper nouns are inherently definite. Indefinite expressions NPs are marked by a, one, any etc. e.g. someone, somebody, anybody, nobody etc. Both the definite and indefinite article can be used to a specific referential referent and in a generic non referential sense. The referential indefinites and generics are very close to each other, that’s why it’s very difficult to distinguish whether a definite article is used referentially or generically. According to Rosenbach (2002) it’s very important to distinguish between the referential and non referential expressions for the investigation of English genitive constructions for two reasons: 1- It accounts for the inherent ambiguity of s-genitive containing an indefinite possessor which can be often interpreted as both specifying and classifying genitive. 2- Certain problems for the empirical investigation many arise due to this ambiguity. Therefore, only specifying genitives with a referential possessor should be used as potential choice contexts. Animacy Animacy is used to distinguish between the animate referents (living creatures) and the inanimate referents (non-living things). Animacy has been isolated as an independent and dominant factor that causes the variation between the s-genitive and of-genitive constructions by Rosenbach 2002, 2005, 2008) and O’Conner et al. (2004). Many research scholars have proposed different scales of animacy e.g. animacy hierarchy proposed by Silverstein (1976, cf. Rosenbach 2002: 42) that incorporates aspects of person, nominal expression types, topicality and animacy etc. 1st person > 2nd person > 3rd person > pronoun > proper name > human > animate > inanimate The animacy hierarchy constitutes several independent hierarchies e.g. 114

www.sienajournals.com

a) b) c) d)

1st person > 2nd person > 3rd person pronoun > proper noun > common noun animate > inanimate human > non human

A person hierarchy NP-type hierarchy Animacy hierarchy Humanness hierarchy

Rosenbach (2008) suggested a scale (fig. 2.4) with graduated animacy values derived from various empirical studies. animate Human >

inanimate

animal N >

collective N >

Temporal N >

Locative N >

the dog’s >

he company’s

Monday’s

London’s

collar

director

mail

Suburbs

Common inanimate N

The building’s door

The boy’s bike >

Fig: 2.4 Animacy scale shows the decline of animacy from left to right (Rosenbach 2008:153) The present research explores only the two extreme points of animacy scale i.e. the humans and the common inanimate things. Animacy is related to topicality and contributes to the topic worthiness of possessor. It is very difficult to distinguish in pronominal possessors whether it is the noun possessor that causes the choice of s-genitive or whether it is the humanness property of the possessor that results in the selection of s-genitive (for more detail see Rosenbach 2008:156). Therefore, the present study focuses only on the genitive constructions with definite and indefinite possessors. Pronominal possessors are not used. Rosenbach (2005) provided evidence both from a corpus analysis and an experimental study that animacy is a processing factor that influences the grammatical variation and it dominates the syntactic weight. In an experimental study, subjected were presented with a questionnaire based on a short text passage that provided context for both s-genitive and of-genitive constructions. There were four conditions: two animate and two inanimate. Two neutral

115

www.sienajournals.com

conditions differ in Animacy but not in weight. Figure shows the results of the experiment from 39 monolingual native speakers of American English.

Fig: 2.5. Shows the relative frequency of s-genitive and of-genitive according to the four conditions. (Experimental results of Rosenbach study on animacy and weight (2005: 620). The results clearly show that the s-genitive is more frequent in the animate neutral condition than in the inanimate neutral condition. This shows that there are animacy effects that are not due to syntactic weight. Comparing the animate neutral and animate long short condition, s genitive is more frequent in neutral than in long short condition. This shows the weight effects that are not due to animacy. For inanimate possessor the frequency of s-genitive is higher for neutral for short possessor as compared to the neutral condition. Thus the results clearly show that the animacy and weight are independent factors and both influence the choice of possessive construction. On the whole apart from the weight, it is clear that Native speakers of English prefer s-genitive for animate possessor while of-genitive for inanimate possessors. In the present experiment the syntactic weight of the both the PR and PM is kept constant. The present research takes into account only the animacy factor and investigates whether the animacy affects the choice of Possessive construction in other languages as well or whether the typological differences affect the choice of Possessive construction.

116

www.sienajournals.com

Prototype A prototype is an original type, form or example of something that serves as a typical example or standard for other things of the same category. In other words prototypes are typical instances of a category that represents the other members of that particular category. Some members of a category are more central than others. Prototypes are in fact set of concepts that are related to fuzziness in semantics. According to Eleanor Rosch (1975) natural categories are inconsistent to the status of their constituent members and are not characterize able due to their fuzzy boundaries and hence they are graded. It has generally been observed in response time experiments that the quarries that involve prototypical members elicited a faster response time than the non prototypical members. In terms of possession prototype denotes the type of relation between the PR and the PM. There are many different type of possessive relations Between the PR and the PM e.g. body part relations, kinship relation, ownership etc. These can be described in terms of concreteness vs. abstractness. The more a relation is concrete the more it is prototypical and the more a relation is abstract the more likely it is to be non prototypical e.g. body parts and kinship relations are prototypical, for ownership concrete relations are prototypical but abstract relations are non prototypical e.g. the strings of guitar is prototypical but the sound of guitar is non prototypical. Likewise if a relation expresses the mental or psychological state of the PR it’s non prototypical e.g. the woman’s surprise, the child’s joy etc. The abstract relation of PR and PM in inanimate things e.g. the style of the shirt, the pattern of the scarf, the falvour of soup etc. are non-prototypical because the nature of relation is abstract. When the relation is between concrete things then it is prototypical. For instance in the heel of the shoe, the key 117

www.sienajournals.com

board of the laptop both the PR and the PM are concrete so the relation is prototypical. In an experiment on English genitive variation Rosenbach (2002) found that s-genitive is the preferred for the prototypical possessive constructions while for non prototypical possessive constructions the of-genitiveis preferred by the Native English speakers. To sum up the discussion the previous research suggests the following factor hierarchy for the choice of sgenitive: Animacy > topicality > possessive relationship. Keeping in view the above discussion the following predictions have been made about the choice of English genitives by the Native speakers. PREDICTIONS s-genitive

of-genitive

[+animate] possessor

[- animate] possessor

Preference for animate possessors

Preference for inanimate possessors

[+topical] possessor

[-topical] possessor

Preference

for

definite

topical

Preference for indefinite or non

possessors

topical possessor

[+prototypical]

[-prototypical]

Preference for prototypical Possessive Preference relation

for

Non-prototypical

Possessive relation

The present research investigates whether the L1 learners choose the genitives according to the set predictions. Whether L2 learners make the same choice like L1 or they deviate from the set predictions. If the L2 learners deviate in their choice from native speakers, then their pattern of deviations needs to be matched with the typological difference of the L2 learner’s first language. If there is same pattern as predicted by the typological difference of the L2

118

www.sienajournals.com

learner’s first language then it means there is L1 transfer that is affecting on their choice of possessive construction. The typological differences between English and Greek language are given below. Typological differences between English and Greek Possessive Constructions English uses a possessive clitic’s, a preposition of and pronominal form my, your to encode the possession. Semantically, a possessive relationship is used to refer to kinship relations (my brother) body parts (the girl’s eyes) and legal ownership (jean’s shirt). S-genitive is used with animate nouns while “of” is used with inanimate nouns. The present study will focus only on two types of possessive forms s-genitive vs. of genitive constructions. The nature of genitive-s is controversial and much debated issue. Some researchers view it as inflectional suffix (phrasal affix see Zwicky 1987, Lapointe 1990, Miller 1992, Longobardi 1996), others view it as a clitic (Carstairs 1987, Alexadiou 2005). While still there are some other researchers who treat it both as inflectional affix and as a clitic (see Rosenbach 2004, Plank 1995)

Greek language has both orders for possessive constructions i.e. Possessor> Possessum and Possessum > Possessor. Both the Possessor and the Possessum are preceded by a definite article. e.g. a.

το σπιτι της Λουκιας To

spiti

tis

loukias

(PM > PR)

Loucy’s house b. της tis

Λουκιας

το

σπιτι

loukias

to

spiti

Lucy’s house

(PR > PM) (David Halton, 1997)

119

www.sienajournals.com

In the example a. the Possessum is preceded by the Possessor. της is a genitive case used to express the possessive relation for singular feminine nouns. In example b. Possessor is preceded by the Possessum but the same genitive case της is expressing the possessive relationship.

c. Pir Took

to

vivlio tu

Niku

ACC

book

Nikos- GEN

GEN

(Possessum > Possessor)

i took Niko’s book

d. Pira Took

(Theodore Marinis, 2002)

tu

Nikou

to

vivilo (Possessor> Possessum)

the-GEN

Nikos

the-ACC

book

I took Nikos’s book

(Theodore Marinis, 2002)

In example b. the Possessor follows the Possessum, while in example c. the Possessum follows the Possessor. In order to understand the possessive expression in Greek, it’s important to understand the Greek articles, genders, cases and numbers (Table: 1). Greek has both definite and indefinite articles. They are inflected by gender and case. The definite article is also inflected for number. The Greek definite article agrees with the noun it modifies. The Genitive case is used to express the possession. In order to express the possession του is used for singular masculine nouns, της is used with singular feminine noun, του is used for neuter. For the plural possessive construction, again the genitive case is used for three genders. The definite article των is used with plural masculine, feminine and neuter nouns. Definite Article

singular Feminine

plural Neuter

Masculine

Feminine

Neuter

120

www.sienajournals.com

Masculine ο

η

i

το

to



i



i

τα

ta

του

to

της

tis

του

to

των

ton

των

ton

των

ton

τον

ton

την

tin

to

τους

tus

τις

tis

τα

ta

Nominative O case Genitive case Accusative

το

case

Table1: Greek definite articles, cases and Gender system. As in English the article a, an is used for singular, the same way Greek indefinite article is used for singular nouns. It denotes the same meaning as numeral one. Indefinite article is not used with the plural nouns. Genitive case is used to express the possessive relations (Table: 2). For the indefinite possessive relation

ενος

is used with indefinite masculine singular nouns,

µιας is

used with

indefinite singular feminine nouns, while ενος is used with indefinite neuter nouns.

indefinite Articles

singular Masculine

Feminine

Neuter

Nominative case

ενα

enas

µια

mia

ενα

ena

Genitive case

ενος

enos

µιας

mias

ενος

enos

Accusative case

ενα(ν)

ena(n)

µια

mia

ενα

ena

Table 2: shows the Greek indefinite articles, cases and Gender system.

121

www.sienajournals.com

The present research deals with the genitive case only. That’s why only the genitive case is mentioned here.

PR > PM order in Greek This type of construction and the order of possession are grammatical in Greek language but it’s dispreferred by the native speakers to such an extent that in spoken language this order is almost nonexistent. While the PM>PR order is highly preferred by the Greek native speakers.

1. a.

Μιας

σουπας

η

γευση

Mias

soupas

i

gefsi

GEN-FEM

PR-GEN

NOM-FEM

PM-NOM

a

soup’s

the

flavour

A soup’s flavour

Example 1.a. is an inanimate non prototypical possessive construction, Following the PR > PM ORDER. An indefinite FEM-SG GEN case µιας precedes the PR. In order to express the possessive relationship PR is marked morphologically by a FEM-SG-GEN morpheme ας and assigns the genitive case to the PR. Unlike the English possessive-s in Greek genitive case is assigned by the Noun. The FEM-SG NOM definite article η precedes the PM.

1. b.

Του

αυτοκινητου η

οψη

Tou

aftokinitou

i

opsi

GEN

PR-GEN

NOM-FEM

PM-NOM

The

car’s

the

front

the car’s front

122

www.sienajournals.com

Example 1.b. is an inanimate prototypical possessive relation in nature. It’s following the PR>PM order which is very rare in Greek spoken language as explained earlier. The article του receives NEUT-SG GEN case from PR noun and the morpheme ου at the ending of the PR αυτοκινητου (car) assigns the possessive case to the PR. The PM (N) assigns the definite FEM-SG NOM case to the preceding determiner η. Here in this example both the Possessor and the Possessum have the same ending like their preceding articles. But this is not the case with all the nouns. As mentioned earlier morphological ending can be different from the preceding determiners. 1. c.

Ενος

γιου

το

χερι

enos

giou

to

cheri

GEN-MAS

PR

NOM-NEUT

PM

A

boy’s

the

face

A boy’s face

Example 1.c. is an animate prototypical possessive construction, following a PR>PM order. While an indefinite MAS-SG GEN case ενος is preceding the PR and the PR is marked by GEN-MAS morpheme ου to express the possessive relationship and assigns a GEN-MAS case to the preceding article. While PM is assigning the NEUT-SG NOM case to the preceding definite determiner το.

1. d.

Του

κοριτσιου

το

μελλον

tou

koritsiou

to

mellon

GEN-NEUT

PR-GEN

NOM-NEUT

PM-NOM

The

girl’s

the

future

The girl’s future 123

www.sienajournals.com

The Example 3.b. is an animate Non-Prototypical Possessive construction, following the PR>PM order. An indefinite NEUT-SG GEN article is preceding the PR, and the PR is marked with a NEUT-SG- GEN morpheme ιου. It seems strange here that the PR is marked with a Neuter morpheme. The reason is that Girl is Neuter in Girl is neuter in Greek language while sister and women are Feminine.

P M > PR Order in Greek 2. a

Το

σφυριγμα

του

βραστηρα

To

sfyrigma

tou

vrastira

The

whistle

of the

kettle

NOM-NEUT

PM-NOM

MAS-GEN

PR-GEN

The whistle of the kettle

The example 1.a. is inanimate non-Prototypical possessive construction, following the order PM > PR. A 2 NEUT-SG Nominative case το precedes the PM σφυριγμα (whistle) and agrees in number gender and case with the PM. A definite FEM- SG Genitive case του precedes the PR βραστηρα (kettle) and it also agrees in number, gender and case with the PR. The PR is morphologically marked to express the possessive relation. The morpheme α at the end of the word Βραστηρα (kettle) assigns a singular masculine genitive case to the Possessor.

2. b.

Η

ειδοδος

ενος

κτιριου

i

eisodos

enos

ktiriou

The

entrance

of a

building

2

NOM- Nominative case, GEN-Genitive case, PR-Possessor, PM-Possessum, FEM-feminine, NEUT- Neuter, MAS-Masculine, SG-Singular.

124

www.sienajournals.com

NOM

PM-NOM

GEN

PR-GEN

Example 2.b. is an inanimate prototypical possessive construction, following the order PM > PR. A definite FEM-SG NOM case η is modifying the PM and an indefinite

NEUT-SG

GEN ενας case is used with the PR κτιριου and the PR is morphologically marked by a NEUT-SG-GEN morpheme ου for expressing the possessive relationship. 2. c.

Η

οργη

της

αδελφης

i

orgi

tis

adelfis

The

anger of the

sister

NOM-FEM

PM

PR

GEN-FEM

Example 2.c. is an animate non prototypical possessive construction. A definite FEM-SG article η

is modifying the PM and a definite FEM-SG GEN case τις preceding the PR αδελφης and the PR is marked by a FEM-SG-GEN morpheme ης to express the possessive relationship. 2. d.

Το

ταξιδι

μιας

γυναικας

To

taxidi

mias

gynaikas

The

trip

of a

woman

NOM-NEUT

PM-NOM

GEN-FEM

PR-GEN

The trip of a woman In example 2.d. is following the PM>PR order. A definite article τν receives NEUR-SG NOM case from PM while the PR assigns FEM-SG GEN case µιας to an indefinite article. the ending morpheme ας on PR assigns a FEM-SG GEN case to the PR. 2. e.

Το

πηγουνι

του

μωρου

To

pi̱gouni

tou

mo̱rou

125

www.sienajournals.com

The

chin

of the

baby

NOM-NEUT

PM-NOM

GEN-NEUT

PR-GEN

The chin of the baby

Example 4.a. is an Animate and prototypical possessive constructions, following the order PM> PR. A definite article το is modifying the PM and receiving the NEUT-SG NOM case from PM. While the article του preceding PR receives a MAS-SG GEN case from the following PR. The PR is marked by a Neuter GEN morpheme ου. As the baby or child are considered Neuter in Greek language. Note here the ending of both the PR and the preceding article are same. It is mostly the case when the PR or PM is preceding by a definite article. But some nouns with definite article have different ending as well. Summing up the above discussion, Greek language has both PR>PM order and PM>PR order. Both orders are grammatical. But as discussed above the former order is very less common in Greek spoken language to such an extent that this order is almost nonexistent in Greek spoken language. While the later order is most frequent order is 99% preferred by the Greek native speakers. Hence, it is predicted that if L2 speakers transfer the L1 properties, then the L2 Greek learners of English will obviously prefer the PM > PR that is the Prepositional Phrase counterpart in English. On the basis of the above discussion and examples we can see the difference between the English possessive constructions and Greek possessive constructions. All of the Greek possessive constructions discussed above are accompanied by an English counterpart that are somehow similar semantically and following the same order of possession but they are very different syntactically. In English the possessive s’ assigns genitive case to the Possessor. While in Greek the Possessors are marked morphologically for the expression of possession 126

www.sienajournals.com

and assign the case to the preceding articles. Both the Possessor and the Possessum agree in number, gender and case with the preceding articles. The most important point to note here is that unlike English, in Greek language there is no distinction in possessive relations on the basis of animacy and prototypicality. Whether the possessive constructions are animate or inanimate, prototypical or non-prototypical they are constructed in the same way. There is no difference between these four types of possessive relations in Greek syntactically and semantically. Therefore it isn’t unreasonable to predict that if L2 speakers transfer the L1 then in the rating task they will not distinguish the animacy and prototypically. Table.1. summarizes the characteristics of possessive constructions and the typological difference between English and Greek Possessive constructions. Characteristics of Possessive Constructions English

Greek

PR>PM

PR>PM

PM>PR

PM>PR

DEF/INDEF

DEF/INDEF

No Agreement with the PR Or

Agree in NUM, GENDER & Case

PM

with the PR Or PM

Gender

N/A

M/F distinction marked morphologically

Animacy

Animate Inanimate distinction

N/A

Order

Articles

Prototype , N/A

Prototype Non-prototype distinction Possessive S / of

Tou/tis etc.

markers

127

www.sienajournals.com

Table.3. Shows characteristics of possessive constructions and the typological difference between English and Greek Possessive constructions

Experimental Material The materials for the rating task: short genitive phrases consisted of a PR 1-2 syllables in length and a PM not more than three syllables. All the PR and PM were singular count nouns. The experimental material was constructed by using 40 animate and 40 inanimate PR phrases. Both animate and inanimate PR phrases were further followed by 20 prototypical and 20 non-prototypical PM phrases. That resulted in four types of phrases i.e. animate prototypical, animate non prototypical, inanimate prototypical, and inanimate non prototypical. These four types of phrases were incorporated with definite and indefinite articles that produced eight conditions i.e. animate definite prototypical(+a+t+p), animate indefinite prototypical(+a-t+p), animate definite non-prototypical(+a+t-p), animate indefinite non-prototypical(+a-t-p), inanimate definite prototypical(-a+t+p), inanimate indefinite prototypical(-a-t+p), inanimate definite non prototypical(-a+t-p), inanimate indefinite nonprototypical(-a-t-p) . All of the eight conditions were once merged with s-genitive and once with of-genitive. The 8 conditions resulted in 10 items per condition that means 80 experimental items that constituted 10 items for each 8 condition. 30 filler items were used to distract the participants. Word order, mass count, and prepositional phrases were used

as

fillers. The material was then equally distributed into two lists using the Latin square design. The two lists were randomized, so that one condition mix together properly and may not occur simultaneously. The lists were checked for semantic priming effects as well. The lists were presented to the participants using the DMDX (reaction time experiment software) for rating task.

128

www.sienajournals.com

Participants A group of 20 L1 Greek speakers learning English as L2 performed in the experiment. The experiment was conducted in a quiet room. A group of 20 monolingual native English speakers served as a control group. All participants had a corrected or corrected to normal vision. The participants‟ proficiency level was assessed by using the grammatical portion of a short placement test (oxford placement test by Allen, 1992). Majority of the participants got 6.5 OPT band. In order to check whether the participants were familiar with the vocabulary used in constructing the experimental material, the participants had taken a short vocabulary test to indicate the unfamiliar words on the vocabulary list. Almost all of the participants were familiar with the vocabulary used in the experiment. Method The experiment was carried out in a psycholinguistics lab. All participants were tested individually in a quiet atmosphere. An experiment information sheet was given to the participants. Experiment method was explained to the participants both in written form and orally. DMDX was used to present the material and for recording the participants choice and reaction times. The experimental phrases were presented in black letters on a white back ground. Phrases appear in a pair wise fashion (containing one s-genitive and one of-genitive) in the middle of the screen for 2500ms followed by hash marks in the middle of the screen and a 3500 time out. The experiment started with a practice trail to familiarize the participants with the experimental procedure. The participants had to indicate their preference by rating the

129

www.sienajournals.com

phrases as quickly as possible by pressing the buttons for the first or second phrase on a dual Analoge pad.

Results Comparison of L2 Greek speakers with Native English Control Group

Descriptive statistics English s-genitive

Conditions

Greek

of-genitive

s-genitive

of-genitive

Mean

SD

Mean

SD

Mean

SD

Mean

SD

89

0.58

11

0.31

50

11.08

50

14.76

84

0.72

16

0.22

53

8.67

46

10.57

72

1.53

28

1.31

61

12.42

38

10.65

64

2.04

36

2.15

41

10.73

58

14.23

47

1.89

53

1.24

31

13.87

68

15.67

26

1.63

74

1.54

31

11.46

68

10.68

15

0.45

85

0.56

25

15.73

75

11.58

animate definite prototypical +a/+t/+p animate definite non-prototypical +a/+t/-p animate indefinite prototypical +a/-t/+p animate indefinite non-prototypical +a/-t/-p Inanimate definite prototypical -a/+t/+p Inanimate definite non prototypical -a/+t/-p Inanimate indefinite prototypical -a/-t/+p

130

www.sienajournals.com

Inanimate indefinite non prototypical 12

0.21

88

0.34

25

10.58

75

13.46

-a/-t/-p Table 4. Shows the mean rating and standard deviations of Control group and L2 Greek speakers.

The descriptive statistics show that the control group rated high number of s-genitive in the first four animate experimental conditions (M = 89, 84, 72, 64 respectively) and it dropped down in the fifth condition (M= 47) which is also animate but non-prototypical. This sudden decline in rating of s-genitive at the fourth condition could be the effect of prototype. Conversely, in the last four inanimate experimental conditions the control group rated the less number of s-genitive (M = 47, 26, 15, & 12 respectively) and they preferred the of-genitive more in these conditions. On the other hand, the Greek L2 speakers rated less number of sgenitive in the first three conditions (M= 50, 53, 61 respectively) as compared to the control group which shows no effect of animacy. The rating declined in the fourth condition (M = 41). In the last four in animate conditions they rated higher number of s-genitive (M = 31, 31, 25, 25 respectively) than the control group. It shows that there was no effect of animacy on the choice of s-genitive by Greek L2 speakers. Comparison of the rating of of-genitive reveal that in the first three animate conditions native group rated less number of of-genitive (M = 11, 16 & 28 respectively) which is an effect of animacy. However, the Greek L2 speakers rated higher number of of-genitives in the first three conditions (M = 50, 46, 38) than the control group. The rating of of-genitive increases dramatically at the fourth condition by both control group and L2 group (M = 36, 58 respectively). While, in the last four inanimate conditions the control group preferred greater number of of-genitive (M = 53, 74, 85 & 88 respectively) which is again indicates the effect of animacy effect. The Greek L2 group also preferred of-genitive in the last four inanimate conditions (M= 68, 68, 75& 75 respectively). Overall the rating of control group shows the effect of animacy and prototypicality while the L2 Greek learners didn’t show any effect of animacy and prototypicality. 131

www.sienajournals.com

Comparison of L2 Greek Speakers with Control Group 100 90 80 70 60 50 40 30 20 10 0 +a/+t/+p animate definite prototypical

+a/+t/-p

+a/-t/+p

+a/-t/-p

-a/+t/+p

-a/+t/-p

-a/-t/+p

-a/-t/-p

animate animate animate inanimate inanimate inanimate inanimate definite indefinite indefinite definite definite non indefinite indefinite nonprototypical nonprototypical prototypical prototypical nonprototypical prototypical prototypical

English s-genitive

English of-genitive

Greek s-genitive

Greek of-genitive

Graph1. Shows the the mean rating of s-genitive and of- genitive by the control group and L2 Greek speakers.

The blue line shows rating of s-genitive by control group and rust line shows rating of sgenitive by L2 Greek speakers. In the first four animate conditions the control group preferred the s-genitive while the Greek L2 speakers rated less s-genitive as compared to the control group. After the fourth condition the rating of s-genitive dropped suddenly and then gradually declined in the last four inanimate conditions. The yellow line shows the rating of of-genitive by control group while the grey line shows the rating of of-genitive by L2 Greek speakers. The control group rated less number of of-genitives in the first four animate conditions. The rating of of-genitive suddenly started increasing at the fourth condition and it gradually increased till the last inanimate condition. It shows that control group prefers ofgenitive for the inanimate condition. On the other hand the Greek L2 learners rated higher number of of-genitives as compared to the control group. It shows that animacy and prototypicality has no effect on Greek L2 speakers’ choice of genitive construction. The

132

www.sienajournals.com

rating of of-genitive increases gradually in the last four inanimate conditions. The syntactic weight and length of the experimental material was kept constant. Rating Task Results Descriptive Statistics for Participant Analysis

Descriptive Statistics

L1

Mean

1

86.6667

2

50.0000

1

81.6667

2

61.6667

1

88.3333

2

53.3333

1

88.3333

2

41.6667

1

48.3333

2

31.6667

1

48.3333

2

23.3333

1

45.0000

2

23.3333

1

63.3333

2

25.0000

ANI_DEF_PROTO

ANI_IDEF_PROTO

ANI_DEF_NPROTO

ANI_IDEF_NPROTO

Std. Deviation

12.11060 22.80351 24.83277 27.14160 11.69045 28.75181 7.52773 28.57738 17.22401

N 20 20 20 20 20 20 20 20 20

INANI_DEF_PROTO

INANI_IDEF_PROTO

INANI_DEF_NPROTO

INANI_IDEF_NPROTO

7.52773 17.22401 17.51190 22.58318 17.51190 33.86247 17.60682

20 20 20 20 20 20 20

133

www.sienajournals.com

Descriptive Statistics for Item Analysis

Descriptive Statistics animacy

Prototype NPROTO

INANI

PROTO Total NPROTO

English_def

ANI

PROTO Total NPROTO

Total

PROTO Total NPROTO

INANI

PROTO Total NPROTO

English_indef

ANI

PROTO Total NPROTO

Total

PROTO Total NPROTO

INANI

PROTO Total NPROTO

Greek_def

ANI

PROTO Total NPROTO

Total

PROTO Total NPROTO

Greek_indef

INANI

PROTO Total

Mean 45.0000 39.9850 42.4925 88.3350 86.6800 87.5075 66.6675 63.3325 65.0000 43.3400 48.3350 45.8375 88.3400 81.6800 85.0100 65.8400 65.0075 65.4238 23.3350 31.6600 27.4975 53.3250 50.0000 51.6625 38.3300 40.8300 39.5800 25.0000 24.9900

Std. Deviation

N

34.67646

20

31.72212

20

32.90146

40

22.36422

20

16.73740

20

19.51532

40

36.20770

40

34.43555

40

35.14848

80

37.62823

20

36.64362

20

36.74710

40

24.83254

20

20.15264

20

22.57549

40

38.85154

40

33.72121

40

36.14839

80

28.82759

20

25.31626

20

27.10850

40

33.16978

20

35.05385

20

33.72654

40

34.22700

40

31.57717

40

32.74385

80

28.36686

20

23.88203

20

25.88217

40

134

www.sienajournals.com

24.9950 NPROTO ANI

PROTO Total NPROTO

Total

PROTO Total

41.6600 61.6850 51.6725 33.3300 43.3375 38.3338

33.99347

20

16.33228

20

28.20879

40

32.03365

40

27.44226

40

30.06176

80

Preference results In order to see the reliable difference in the preference rating task a repeated measures ANOVA was carried out. The results of the ANOVA are as follows: 3 Main effect of L1group Both Participant and item analysis showed significant effect of L1. F1 (1, 38) = 4.652 p < .05 significant F2 (4, 76) = 29.185 p .05

Effect of Prototype F2 (1, 76) = 3.258

p= .075 p > .05

3

Note that the ANOVA results are for the s-genitive construction.

135

www.sienajournals.com

Effect of definiteness The effect of definiteness is also not significant in both participant and item analysis. Effect of definiteness F1 (1, 38) = .016

p = .897 p >.05

Effect of Definiteness F2 (1, 76) =.006

p=.937 p>.05

Interactions between variables and groups The interaction between L1 and animacy is only significant in item analysis but it’s not significant in participant analysis. The reason why this interaction is significant only in item analysis could be the greater number of items than participants.

Interaction between L1 & animacy F1 (4, 25) = .525 p > .05 F2 4, 76) = 4.612 p= .001 p < .05

Interaction between L1 & Prototype The interaction between L1 and prototype was not significant in both participant and item analysis. F1 (1, 38) = 1.028 P > .05 F2 (3, 76) = .777 p = .541 p >.05

Interaction between L1 &definiteness The interaction between L1 and definiteness was also not significant in both participant and item analysis. F1 (1, 38) = .274 p > .05 F2 (3, 76) = .158 p= .955 p> .05

136

www.sienajournals.com



Over all there is a highly significant effect of L1 and animacy. But the interaction between L1 and animacy is only significant in item analysis but not in participant analysis, this could be due to the larger number of items than participants.



The effect of prototypicality is marginally significant. This might become significant with larger sample size. The interaction between L1 and prototypicality is not significant in both participant and item analysis but if the p-values of prototypicality are compared with p-values of definiteness prototypicality is higher in the significance hierarchy than definiteness.



The effect of definiteness is not significant. Definiteness does not seem to play a role. This might be due to the task – presenting phrases in isolation, not in discourse context as presented by Rosenbach (2002). Topicality is a discourse feature. Thus, if the same items are embedded in a different task like in a discourse context one may get significant results for definiteness as well.



There were less filler items in the experiment; this may give participants chance to develop strategies for animacy. One can use more filler items in a follow up study. A follow up study can also be done with only definite items rather than having both the definite and indefinite items.

Conclusion Overall, the graph shows that the control group ratings and preferences are influenced by cognitive factors and these factors determine the choice of genitive construction for Native English Speakers. But for the non-native speakers like Greek whose first language lack the animacy and prototypicality characteristics in possessive constructions, it doesn’t influence on their choice of L2 genitive constructions. The difference between the ratings of these two groups indicates that L1 speakers’ choice of genitive construction is governed by the cognitive factors as predicted while the L2 speaker’s choice is matched to the patterns predicted by the typological difference of the L2 learner’s first language. It means there is L1 transfer that is affecting on their choice of possessive construction. The results are in line 137

www.sienajournals.com

with S.A., Ghilzai (2014). She found influence of L1 on German native speakers’ preference of genitive constructions. A comparison of the effect of animacy and prototypicality on the syntactic variation of genitive constructions is given below: The effect of animacy on s-genitive rating of control group Animate-definite-prototypical > Inanimate-definite-prototypical

(89 > 47)

Animate-definite-non-prototypical > Inanimate-definite-Non-prototypical

(84 > 26)

Animate –Indefinite-Prototypical > Inanimate Indefinite prototypical

(72 > 15)

Animate-Indefinite-Non-prototypical > Inanimate Indefinite - Non-prototypical (64 >12)

The descriptive statistics show that the mean ratings of s-genitive for the animate conditions are greater than the inanimate conditions.

The effect of animacy on of-genitive rating of control group Animate-definite-prototypical < Inanimate-definite-prototypical

(11 < 53)

Animate-definite-non-prototypical < Inanimate-definite-Non-prototypical

(16 < 74)

Animate –Indefinite-Prototypical < Inanimate Indefinite prototypical

(28 < 85)

Animate-Indefinite-Non-prototypical < Inanimate Indefinite - Non-prototypical (36 < 88)

The descriptive statistics show that the mean ratings of of-genitive for the animate conditions are less than the inanimate conditions. The effect of prototypicality on s-genitive rating of control group Animate-definite-prototypical > Animate-definite-non-prototypical

(89 > 84)

138

www.sienajournals.com

Animate –Indefinite-Prototypical > Animate-Indefinite-Non-prototypical

(72 > 64)

Inanimate-definite-prototypical > Inanimate-definite-Non-prototypical

(47> 26)

Inanimate Indefinite prototypical > Inanimate Indefinite - Non-prototypical

(15> 12)

The descriptive statistics show that the mean ratings of s-genitive for the prototypical conditions are higher than the non-prototypical conditions. The effect of prototypicality on of-genitive rating of control group Animate-definite-prototypical < Animate-definite-non-prototypical

(11