EXPRESSION OF PROPER NOUNS AND PRONOUNS BY DEAF SIGNERS OF INDIAN SIGN LANGUAGE (MUMBAI-DELHI DIALECT).

Yashomathi 1 and R. Manjula 2 . 1. Lecturer in Speech Pathology, Department of Speech Language Pathology, All India Institute of Speech and Hearing Mysore – 570006. 2. Former Professor of Speech Pathology, Department of Speech Language Pathology, All India Institute of Speech and Hearing, Mysore – 570006. ...................................................................................................................... Manuscript Info Abstract ......................... ........................................................................ Manuscript History

Several other studies on ASL have recorded the role of "body shift" to indicate the third-person reference with a pointing sign directed towards the signer; the shift indicated by the change in orientation of signer"s shoulders, head and/or eyes in ASL (Lillo-Martin & Klima, 1990). On the other hand, Meier (1990) reported that head position and the direction of eye gaze in Israel Sign Language (ISL) distinguishes third-person pronoun from non-third person pronoun. MacLaughlin (1997) noted that pointing signs in Hong Kong Sign Language (HKSL) could be interpreted as a pronoun when signed alone. The pronouns in HKSL are optionally signed by using eye gaze directed at the location of referent in signing space. Liddell and Metzger (1998) reported that the use of eye gaze to express pronouns in ASL was consistently directed toward the addressee during the production of both second person pronouns and third-person pronouns and no distinction in eye gaze was found between second and third person referents.
Studies addressing the linguistic attributes of Indian Sign language (ISL) when compared to other sign languages are very scarce (Prajapati & Asai, 2010). There is need to address the sign language structure of Indian Sign language, in order to provide a base for formalizing the approaches towards teaching and learning the Indian sign language by the hearing impaired individuals as well as their communication partners. It will also facilitate and open opportunities to research and develop sign language recognition systems. This study aimed to analyze and report on expression of pronouns used by the deaf signers of Indian Sign language (ISL) using Mumbai-Delhi dialect, including the use of non-manual markers in a story narration task.

Method:-
Participants:-Ten fluent Deaf Signers (5 Males & 5 Females) using Indian Sign Language (ISL) signs of Mumbai-Delhi dialect were included. The participants were in the age range of 18-40 years. Informed written consent was obtained from each participant prior to the recording of their samples. The participants met the following criteria: They were pre-lingually deaf born to hearing parents. They were proficient in using ISL in day to day conversation. They were using ISL for a minimum of 10 years, as ascertained using a checklist prepared for the purpose by the investigator.
They interacted with their parents, family members and other communication partners using a combination of sign language and few verbal expressions for day to day communication.
They had a minimum educational qualification of secondary education in a school for Persons with Hearing Impairment with the medium of instruction for curricular work being English.
They did not have any other disabilities in motor, visual and neurological domains (as ruled out using Quick neurological screening test (QNST) by Mutti, Sterling and Spalding, 1978), and cognitive deficit (as ruled out using Mini-Mental State Examination (MMSE) by Folstein, Folstein, and McHugh, 1975).
All the participants used Mumbai-Delhi dialect of Indian Sign Language (ISL) as ascertained through information collected in the demographic data sheet by the investigator.
Task and procedure:-A story on the topic "The Snake and the Mongoose" was selected and depicted as sequential picture event on 6 cards with the help of an artist. This story was selected as it provided opportunities to use pronouns repeatedly when the story was narrated. The pictures depicting the story sequence were verified for disambiguity and order of continuity in the sequence by two qualified speech-language pathologists on a three-point rating scale. Video recording of each participant at a time was carried out at the educational centre or work place of the participant by the principal investigator. The video recording of the signed expression of the participants for narration of story was carried out in a room which was well illuminated with minimal visual and sound distraction. The sequential story cards were initially given to the participant (from whom the data was collected individually) and was asked to understand the concept represented in each card and the summary of the story. The participant was then asked to narrate the story using sign language as naturally as possible and the signed expressions of the participants were video-recorded by 1416 the principal investigator using a hand held Sony Digital Video Camera Recorder (Model no: DCR-SR47E) with the capacity of 800K pixel, 60*optical zooming, and storage capacity of 45 hours of video footage. The camera was mounted in front of the participant on a tripod stand at a distance of minimum 5 to 7 feet and zoomed to capture the sign space from head to trunk and arms extended on both the sides of the participant. Each participant was given two trials each and the same was recorded.
Analyses:-Of the two video samples recorded, the sample which had less visual distracters/repetitions/good visibility of signing was considered for the analyses. The sample was fed into a computer. ELAN (EUDICO Linguistic Annotator) software (Hellwig & Van Uytvanck, 2007) was used for coding the signed expressions. A set of annotations sharing the same characteristics can be coded on a single tier using ELAN software (Johnston & Crasborn, 2006). Few other investigators have provided annotation guidelines for coding sign language video samples (Crasborn et. al., 2007;Johnston & Crasborn, 2006;Johnston & Schembri, 2006;Neidle, 2002;Zwitserlood, Ozyurek & Perniss, 2008). Based on these guidelines of annotations and transcription conventions, the signed expression of each participant was coded by the principal investigator. The coding of the signed expressions of the participants was carried out to record the use of (a) different pronoun types, (b) strategies used in the expression of the pronouns and (c) the use of non-manual markers to sign various pronoun types.
The data was analyzed and coded under three tiers in ELAN software as follows:-1 st tier -Sign Gloss: The data was transcribed and annotated for expression of lexical signs i.e., glossing for each sign of each participant was carried out by selecting the respective portion in video sample under this tier. The glosses in this tier were transcribed in upper case as per the guidelines.
2 nd tier -Identification of types of pronouns: In this tier, data was checked for the presence of pronouns in signed expression. Instances where pronouns were expressed by the participant were noted and coded for types of pronouns such as personal pronouns, demonstrative pronouns, possessive pronouns etc.
3 rd tier -Identification of non-manual markers used in the expression of pronouns: In this tier, the data was coded for use of different non-manual markers in signed expression of pronouns. This was carried out as per the operational definition of Non-Manual Markers defined by the investigators (provided in Annexure 3). The function of each of the 3 tiers as defined in this study is presented in Annexure 1.
The annotations used to code different types of pronouns and non manual markers expressed by participants are shown in Table 1 and 2 respectively. IX-2P (pe pr) 2 nd Person (personal pronoun) 3.
Dem Demonstrative Pronoun Reliability check:-10% of the video samples of each participant were subjected to reliability check. Another coder/judge with post graduate degree in speech-language pathology and knowledge of sign language, having normal hearing and vision, and familiar with the operation of a computer was recruited for the reliability check. The 2 nd coder was trained by the first coder (principal investigator) using samples of deaf signers that were not included in this study, on the following aspects: 1) identification and selection of signed expressions 2) referring to operational keys and annotating data on ELAN software and 3) coding the pronoun types as per the operational definition of the pronoun types (Annexure 2) and reviewing the annotated samples to confirm the expression of pronouns as per the definitions. Once the data was coded by the 2 nd coder, the inter and intra judge reliability was established.
Inter judge reliability: This was established between the two coders, based on item by item verification for agreement/disagreement in various domains of the data. In instances of disagreement between coders for any of the items, analyses were repeated by coders independently and parameters mutually agreed upon by two coders were noted as final. The reliability coefficient was calculated Intra-judge reliability: Reanalyses of 10% of the data per participant was carried out within a month from the day of first analyses by the 1 st coder (investigator) and the 2 nd coder also in order to check for intra judge reliability. In instances of disagreement between the two instances of coding, it was resolved by reviewing and confirming the parameter analyzed. The reliability coefficient for the samples analyzed in two instances by investigator and other coder was determined. The raw data obtained by the coders was tabulated and subjected to statistical analysis.

Results and Discussion:-
The coded data was used to tabulate and record frequency of occurrence of different pronoun types, strategies used to express pronouns and use of non-manual markers/facial expressions by participants to express the pronoun types. The signed expressions of the story narrated by most participants majorly included use of proper nouns (woman, snake, mongoose, baby, and cradle) that occurred in the story and several pronoun types such as demonstrative pronouns (here, there, that, this), 1 st person pronoun (I, me), 2 nd person pronoun (you, she), 3 rd person pronoun (it) and 1 st person possessive pronoun (my, mine) that substituted the proper nouns.
The proper nouns were expressed by participants using following strategies: a) Signing the noun signs/lexical signs and b) Finger spelling the nouns. The expression of pronoun types involved use of strategies such as a) manual pointing/indexing alone and b) manual pointing/indexing used along with non-manual markers to mark pronoun types. The frequencies of occurrence of the features under study were calculated as follows: The The expression of various features studied as computed by the ELAN software were noted and subjected to statistical analysis. The data was subjected to statistical analysis using SPSS (version 16). Shapiro-Wilk test of normality revealed that the data was not normally distributed (p > 0.05). Hence non-parametric tests were used to find significant difference if any in the frequency of occurrence of various pronoun types and proper nouns. 1418

Frequency of occurrence of proper nouns and pronoun types
The frequency of occurrence of proper nouns and different pronoun types is represented in table 3.   pronouns (1 st person) required the participant to establish a loci in space, whereas use of proper nouns by the deaf signers involved signing the noun signs or using fingerspelling, which could probably be the reason for more frequent use of proper nouns when compared to the pronouns.
There was a significant difference in the frequency of occurrence of demonstrative pronouns demonstrative pronouns and 1 st person pronouns (Z = -2.666, p < 0.01), demonstrative pronouns and 2 nd person pronouns (Z = -2.666, p < 0.01), demonstrative pronouns and 3 rd person pronouns (Z = -2.803, p < 0.01), and demonstrative pronouns and 1 st person possessive pronouns (Z = -2.803, p < 0.01). Amongst the pronoun categories, the demonstrative pronouns occurred more frequently compared to 1 st, 2 nd and 3 rd person pronouns and 1 st person possessive pronouns. However, difference between the frequencies of 1 st , 2 nd , 3 rd person pronouns and 1 st person possessive pronoun were not significant.
Overall, the participants used proper nouns and different types of pronouns in their signed expression of the story. Amongst the pronoun types that represented nouns, demonstrative pronouns were used more frequently compared to that of other pronoun types (1 st , 2 nd , 3 rd person pronouns and 1 st possessive pronoun).

Frequency of occurrence of demonstrative pronouns
It was observed that the participants used demonstrative pronouns such as "this", "that", "here", "there" more frequently compared to person pronouns and possessive pronouns while narrating the story. Diessel (1999) reported that demonstrative pronouns are the very first words that typical children learn. Wu (2004) also reported that demonstrative pronouns are frequently used by typical adults in conversation. This seems to be true amongst the deaf signers also as seen in the results of this study. The demonstrative pronouns were expressed using the following strategies by the participants: Using manual pointing signs/Indexing alone: this involved the signers extending their index finger towards the referent in signing space to establish reference/loci in space.
Using manual pointing/indexing accompanied by various combinations of non-manual markers such as Eye gaze, raising of eyebrows, body leaning forward in the direction towards the loci, head movements such as head bent down towards the loci, head turn towards the loci, head movements such as head bent down towards the loci, head turn towards the loci /referent etc in conjunction with pointing sign to mark the demonstrative pronouns.
Similar findings have also been reported in sign languages such as ASL ( (Meier, 1990), where eye gaze, head turn, body shifts, were observed in addition to manual pointing to mark pronouns.
Several groups of strategies were used by the deaf signers of Mumbai-Delhi dialect of ISL to express the demonstrative pronouns during the story narration task and they are as follows: Group 1: Pointing to space, eye gaze, head bent down and Body leaned forward Group 2: Pointing to space, eye gaze, and head bent down Group 3: Pointing to space, eye gaze Group 4: Pointing to space, eye gaze and heard turn towards the referent Group 5: Pointing to space Group 6: Pointing to space, eye gaze, and eyebrows-raised Shapiro-Wilk test for normality of distribution revealed that the data was not normally distributed (p > 0.05). Hence non-parametric tests were administered to verify for significant difference if any in the frequency of occurrence of combination of strategies used in the expression of demonstrative pronouns. Friedman test revealed a significant difference in frequencies of occurrences [  (5) = 13.612, p < 0.05] of combination of strategies used to express the demonstrative pronoun.
Several combinations of non-manual markers along with manual pointing were used by participants to indicate demonstrative pronouns in story narration task and the same is represented in Figure 2. A pair -wise comparisons of frequencies of occurrence of combination of strategies was carried out using Wilcoxon Signed Rank test. The results revealed a significant difference in frequency of occurrence (Z = 2.521, p < 0.01) between combination 1 and combination 4. A significant difference in frequency of occurrence between the combination 1 versus combination 6 (Z = -2.547, p < 0.01), and combination 3 versus combination6 (Z = 1.992, p < 0.01) was also observed.
The predominant strategy used to sign demonstrative pronouns by participants in story narration task involved a) manual pointing sign accompanied by non-manual markers such as Eye gaze, Head bent down, Body leaned forward toward the referent (combination 1) followed by a combination of b) manual pointing sign and Eye gaze (combination 3) when compared to other combination of strategies ( Figure 2). Observation of the data also revealed that although various combinations of non-manual markers and pointing were used to express the demonstrative pronouns by participants, manual pointing sign and eye gaze were invariably used as a common strategy by the participants.
Qualitative observation of the data revealed that the proximal demonstrative pronouns ("this" & "here") were differentiated from that of distal demonstrative pronouns ("that" & "there") based on the distance of locus being established by the signer in signing space. The demonstrative pronouns "this" & "here", were expressed with manual pointing sign in space that was proximal to the signer"s body whereas, for pronouns "that" & "there", manual pointing signs were far away from the signer"s body. This observation is similar to the report by Lyons (1979) for American Sign Language (ASL).
The phenomenon of establishing loci in space through use of manual pointing sign and use of non-manual markers to express the pronouns in sign language is known as pronominalization. Use of "pronominalization" to express pronouns have also been documented in several sign languages such as American Sign Language (ASL) (Cormier, 2005(Cormier, , 2007Hatzopoulou, 2008;McBurney, 2002), British Sign Language (BSL) (Kyle & Woll, 1985;Meier, 1990), Israel Sign Language (ISL) (Meier, 2003), Sign Language of Netherlands, (SLN) (Bos, 1995), Swedish Sign Language (SSL) (Bergman, 1980), and Brazilian Sign Language (LSB) (Berenz, 1996). As reported by several researchers, pointing is also used in spoken language as a co-speech gesture (Clark, 2003;Haviland, 2000b) and it is used to indicate the direction, location, objects and/or people (Kita, 2003b). The sign languages also makes use of pointing gesture to mark person and number inflections (Zeshan et al., 2004). In this study, the deaf signers of Mumbai-Delhi dialect also relied on manual pointing sign to establish the pronominal reference in space. They also used non-manual markers such as raised eyebrows, body leaned forward (towards the referent), head bent down (towards the referent), head turn (towards the referent) in addition to pointing sign. This can be reasoned based on the fact that these were used by the deaf signers to augment semantic meaning for the expression of the demonstrative pronoun.

Frequency of occurrence of 1 st person pronouns
The It was observed that pointing to oneself was the predominant strategy used to indicate 1 st person pronoun. Various combinations of non-manual markers and pointing were also used to express the 1 st person pronouns by the participants. This confirms the use of "pronominalization" phenomenon by the participants. Similar observations have also been reported in sign languages such as American Sign Language (ASL) (Friedman, 1975;Padden, 1983), British Sign language (BSL) (Kyle & Woll, 1985;Meier, 1990), Swedish Sign language (Bergman, 1980), Brazilian Sign Language (LSB) (Berenz, 2002) and Croatian Sign Language (HZJ) (Alibasic Ciciliani & Wilbur 2006). It was also observed that the direction of pointing sign in the expression of 1 st person pronoun differed from that of pointing sign used in the expression of demonstrative pronouns. The 1 st person pronoun was expressed with index finger pointing and moving towards the signer"s chest. On the contrary, demonstrative pronouns were expressed with index finger moving away from the signer to a locus in signing space. Thus, in Mumbai-Delhi dialect of ISL, the deaf signers used "pointing to oneself" as a main strategy to express 1 st person pronouns

Frequency of occurrence of 2 nd person pronouns
The frequency of occurrence of 2 nd person pronouns was less when compared to frequency of occurrence of proper nouns and demonstrative pronouns. Few instances of 2 nd person pronoun form such as "You" were observed during narration of the story. It was observed that pronominal reference equivalent to 2 nd person pronouns were carried out using few strategies such as pointing to space and use of non-manual markers. Various combinations characterizing pronominalization such as manual pointing to space accompanied by non-manual markers such as eye gaze, head bent down, body leaned forward towards referent were used and the mean frequency of occurrence of these is shown in figure 3. Similar observations are reported in several sign languages such as American Sign Language (ASL) (Berenz, 2002;Cormier, 2005Cormier, , 2007Hatzopoulou, 2008;McBurney, 2002), British Sign Language (BSL) (Kyle & Woll, 1985;Meier, 1990), Israel Sign Language (ISL) (Meier, 2003), Sign Language of Netherlands, (SLN) (Bos, 1995), Swedish Sign Language (SSL) (Bergman, 1980), and Brazilian Sign Language (LSB) (Berenz, 1996). The group of strategies that were used by deaf signers of Mumbai-Delhi dialect of ISL to express the 2 nd person pronouns during the story narration task are as follows: Combination1: Pointing to space, head bent down and body leaned forward

Frequency of occurrence of 3 rd person pronouns
The participants used 3 rd person pronouns such as "It", "He" and "she" during the narration task. The 3 rd person pronoun form "it" was expressed by pointing towards the referent in space (Psp); the signed expression of "He" involved signing the noun "Boy"/ "Girl" (as per the situation) followed by pointing towards the referent in space (Psp). These 3 rd person pronouns (it, He & she) were also signed using non-manual markers such as eye gaze, head bent down, body leaned forward towards the referent, and head turn towards the referent. Similar findings have also been reported in several sign languages such as American Sign Language (ASL) (Berenz, 2002;Cormier, 2005Cormier, , 2007Hatzopoulou, 2008;McBurney, 2002), British Sign Language (BSL) (Kyle & Woll, 1985;Meier, 1990), Israel Sign Language (ISL) (Meier, 2003), Sign Language of Netherlands, (SLN) (Bos, 1995), Swedish Sign Language (SSL) (Bergman, 1980), and Brazilian Sign Language (LSB) (Berenz, 1996). The following combinations of strategies were observed in the signed expression of the deaf signers to express the 3 rd person pronouns and the same is represented in figure 5. The use of non-manual markers in association with the manual pointing sign thereby provided additional semantic cue to the pronoun and the use of these non-manual cues in addtion to manual pointing have also been documented in other sign languages such as ASL ( (Berenz, 2002) and Croatian Sign Language (HZJ) (Alibasic Ciciliani & Wilbur, 2006). The frequency of occurrence of various combinations of non-manual markers in association with manual pointing used by the deaf signers in the narration of the story to mark the 3 rd person pronouns were compared.
The data of 3 rd person pronouns was subjected to Friedman"s test and the results did not reveal any significant difference in frequencies of occurrence [  (1) = 4.000, p > 0.05] between combination of strategies used to express the 3 rd person pronouns. In other words, all these strategies were used equivalently by the deaf signers to refer to 3 rd person pronouns. Zeshan et al., (2004) reported that in most sign languages spatial inflections to mark person and number inflections are used; i.e., the pronouns are indicated by pointing to different locations in space to differentiate across person pronoun (first, second and third person pronouns) and possessive pronouns (first, second and third possessive pronouns).

Summary and Conclusions:-
The study aimed to analyze the expression of pronouns by deaf signers using Mumbai-Delhi dialect of Indian Sign language in a story narration task. The results revealed that the signed expression of the deaf signers in the story narration task involved use of proper nouns and different pronoun types such as demonstrative pronouns, 1 st person pronoun, 2 nd person pronoun, 3 rd person pronoun and 1 st person possessive pronoun. The proper nouns were expressed by the deaf signers using finger spelling of the noun sign or signing the noun sign. The expression of demonstrative pronoun by the deaf signers involved making a pointing sign in space accompanied by combinations of various non-manual markers such as eye gaze, raising of eye brows , body leaned forward in the direction of the referent, head movements such as head bent down, head turn toward the referent, head tilt to the side etc. The expression of the 1 st person pronoun and 1 st person possessive pronoun also involved manual pointing but, the pointing sign was directed towards the signers" chest accompanied by non-manual markers. The 2 nd person pronoun and 3 rd person pronoun also involved making a pointing sign in space accompanied by combinations of various nonmanual markers. The expression of pronouns thus confirms the use of "pronominalization" as a phenomenon in Mumbai-Delhi dialect of Indian Sign language (ISL).