An Essay on Investigating Factors Influencing Comprehensibility of World Englishes by Critically Evaluating Studies in the Domain British Journal of Arts and Humanities

This paper uncovers the factors that influence comprehensibility – a construct predominantly the focus of Applied Linguistics research due to the dual role that speakers and listeners play in interpreting meaning. The essay discusses the factors highlighted by Smith & Nelson, (1985) outlined in their seminal paper. The critical evaluation of the research published in the domain extrapolates; comprehensibility between interlocutors – NS-NNS and NNS-NNS is not impeded by the accentedness of NNS ’ s, but predominantly by the lack of familiarity with it. NNS ’ s comprehensibility for both native listeners/raters and non-native listeners/raters improved with familiarity with the interlocutor or with the variety of World English (WE). Research also emphasized the factors: the interlocutor ’ s proficiency, L1 interference in L2 utterance, attitude towards WE ’ s, listeners/raters background, and effort to communicate greatly influenced comprehensibility . Awareness of the NNS ’ s accent made the listeners/raters overlook other errors of accent either segmental, supra segmental and even lexico-grammatical errors (Webb et al ., 2016). In case of NS-NNS communication, it is an unrealistic expectation of the NNS to achieve native like accentedness to make the comprehensible. In high stakes tests like TOEFL and IELTS, the raters could improve L2 comprehensibility by gaining exposure to World English ’ s. Teachers are advised to include pronunciation in their syllabi. In particular, they could stress on teaching those segmental and supra segmental features which are characteristic of learners L1 that influence erroneous L2 utterances.


Factors Affecting L2 Comprehensibility
In their summary, Smith & Nelson, (1985) have raised certain future research implication questions that highlight factors which can govern comprehensibility of interlocutors: 1) English proficiency of both speakers/ listeners: proficiency of the interlocutors improves comprehensibility.

2) Topic difficulty for both speakers/listeners:
mitigates comprehensibility of L2 utterance. 3) Communicative setting or location of the discourse e.g. noisy or quiet. 4) Familiarity of interlocutors: either at an individual level or with the variety of WE improves comprehensibility. 5) Interlocutors attitudes towards different varieties of WEs. If the native/non-native listeners and raters possess reproving attitude towards other varieties of English, comprehensibility of L2 utterances is mitigated. 6) Comprehensibility of NNS to NNS with different L1s. Non-native speakers' comprehensibility increases for Non-native listeners/raters with different first language if there is a similarity in accents or if they are already exposed to it. 7) Comprehensibility of NSs to NNSs. Awareness of native accents aids NNSs to improve the incomprehensible segments of their utterances. 8) Interlocutor's effort to communicate: greatly influences their comprehensibility.
To validate the mentioned factors, studies addressing these factors for efficient comprehensibility between interlocutors will be critically analysed henceforth.

Listener's/ Speakers Proficiency
Isaacs & Trofimovich, (2012) investigated the effect of experience of raters on their comprehensibility ratings to feature them in high stakes rating scale guidelines. The researchers also probed the linguistic aspects that efficiently differentiated between learners at the low, intermediate and high levels of L2 comprehensibility.
An eight framed picture story was narrated in English in a quiet office to avoid any outside noise distraction. Later, quantitative analysis of the speech of 40 French learners of English was rated by 60 inexperienced NSs for comprehensibility ratings. Linguistic features were slotted under four main categories:  phonology, grammatical accuracy, lexical richness,  and story cohesion. To authenticate the novice raters  markings, a coding scheme was invented by three  experienced ESL teachers (T1, T2 and T3).
Each of the 40 Quebecois Francophone speakers (13 male, 27 females) belonged to varied age groups (Mage = 35.6, range = 28-61). All of the participants had already participated in L2 phonological learning (Trofimovich et al., 2007). Each had similar exposure to L2 -45 minutes/week ESL classes in primary schools with 3hr/week of subsequent ESL instruction. As L2 speakers were of different age range, memory of the English learnt at school level could be of varied depending upon the time that had passed.
Four categories of comprehensibility measures were short listed: phonology comprising of segmental (vowels and consonants sounds) and supra segmental measuring (prosodic features); fluency involving temporal measures (speech rate) and frequency counts of pauses; linguistic resourcesgrammatical and lexical measures. The last category was discourse: capturing speaker's story telling strategies.19 linguistic features were emphasized as framework of scale, but for convenience of teachers and raters, 5 measures were short-listed.
After grammatical errors, vocabulary and fluency were most commented upon by teachers. It was observed that the higher the comprehensibility of the L2 speaker, the better the production of fluent utterances. Moreover, proficient speakers possessed sufficient vocabulary to narrate the story, but the converse was true for the medium and less proficient L2 speakers.
The researchers overcame most confounding variables through triangulation using mixed method analysis. The diverse age group of participant's ensured heterogeneity in the proficiency levels of the Francophone's, but participants self-reported their proficiency levels. Their professional engagements, which were not specified, may have required them to speak with NS or other NNS. The limitations of the study were clearly outlined. Since only Francophone's participants were selected, the generalisability to other L1 speakers becomes inapt. Also, participant's speech was tested through formulation of a picture story which was considered ecological, but spontaneity of speech produced during a direct discourse between NS-NNS may be more indicative of a natural setting and may have produced different results of comprehensibility as a construct.
One of the three teachers (T3) pointed out the segmental features while the rest (T1 & T2) commented only on the grammatical errors which might be due to their ESL background. Also, T2's comments on lowcomprehensibility of L2 learners were about the lack of teachers' familiarity with the L1 accent, the context, and contents of the picture story that could support in understanding their speech. Moreover, since the Francophone's had already participated in a previous study, they were familiar with the researchers symptomatic of the individual familiarity factor pointed out by Smith & Nelson, (1985).
Nevertheless, Isaacs & Trofimovich, (2012) were able to confirm some of the factors influencing comprehensibility mentioned by Smith & Nelson (1985) i.e. its dependence on the listeners/speakers proficiency not only lexicogrammatical but segmental and the requirement of a quiet settings. The study also suggested the individual familiarity factor in assisting L2 comprehensibility.

NNSs L1 Interference:
Crowther, Trofimovich, Saito, & Isaacs, (2015) investigated the factor of L1 interference on listener's judgements of comprehensibility and accentedness in L2 speech. 45 tertiary-adult speakers from three L1 back-grounds Chinese, Hindi and Farsi performed the same picture narrative task used in Isaacs & Trofimovich, (2012). They were rated by 10 NS and the findings were relative to the speakers L1. Chinese speaker's comprehensibility was affected by pronunciation variable (segmental errors). Hindi speaker's speech was affected by lexicogrammar variables where as Farsi speakers revealed no specific variable association. Tests of interactions revealed significant effect sizes of each L1 group's comprehensibility with Hindi (p< 0.0001) and Farsi (p<0.0001) being the more comprehensible while the Chinese to be less comprehensible and most accented group than the other two groups. To overcome the lack of familiarity factor with L2 accents, the10 NS raters selected were familiar with the respective Asian accents. The NS raters easily comprehended the Farsi speakers (variable association) narration because they studied in the same university thus familiar with their individual accents. Even thought he familiarity factor was cont-rolled considerably, the lack of variable association found for Farsi speakers could have been more evident if the researchers used a more nuanced approach to fine tune the sliding test scales. While it was not a direct interaction, the researchers were able to overcome most of the confounding variables in a controlled environment. The comprehensibility of NS listeners judgement of NNS speech was affected by NNS'sL1 interference.

Familiarity & Rater's Background
Saito & Shintani, (2016) investigated the perception of L2 accentedness by two groups of native speakers monolingual Canadians and multilingual Singaporeans. In an earlier study, Saito & Shintani, (2016) deciphered the correlation between the raters background to the linguistic features that are required by NNS to make them successfully comprehensible.
In the precursor study, they listened to 50 Japanese learners' spontaneous speech samples and rated them on a 9 point scale (1= easy to understand and 9 = difficult to understand). Through a paired t-test, it was revealed that the Singaporean raters were more lenient in assigning comprehensibility scores (M= 4.0, SD=1.5) as compared to the Canadian raters (M= 4.7, SD=1.5) of the speech samples. A multiple regression analysis revealed that the Canadian Raters focused more on pronunciation, fluency, vocabulary, and grammar, while the Singaporean raters focused more on the lexicogrammatical aspects of speech for comprehensibility judgements.
To investigate the L2 accentedness rate (the measure of closeness to native like accent) the researchers compared the accentedness scores with those of L2 comprehensibility judgements scores of the precursor study.
A picture description task was employed in which the speakers had to describe 7 pictures within 5s preparation time. To ensure spontaneity, three cue words were provided. The first four pictures were given for practice while the last three were tested as speech samples. The Japanese learners were of different proficiency levels owing to the difference in age and length of stay in Canada ranging from 6months to 11 years. The longer the stay the more probability of native like accentedness (Fuentes, 2021).
Out of the 10 native Canadian monolingual raters, 3 had beginner or intermediate knowledge of French. Ten Singaporean multilingual raters also judged the speech. While the Singaporean spoke predominantly English at home/school/social settings (68 -82.5%), but in daily communication, they also used Mandarin, Malay, and Tamil (11-21%). The findings revealed that the monolinguals assigned higher and stricter ratings than the multi-lingual of L2 accentedness but both highlighted phonological aspects of the NNSs speech.
Most of the conflated variables were overcome by Saito & Shintani, (2016) by employing experienced coders to compare the ratings of the monolinguals and multilinguals. These coders were trained in phonology. Nonetheless, 3 monolinguals knew French which may weaken the trait of being monolingual, but the rater's attitudes towards accentedness may have prompted the 3 French speaking monolinguals to rate the speech strictly. The ecological validity was limited because the element of spontaneity was mitigated by the rehearsals before the actual task. In an actual communicative setting, an interaction between NS (Monolinguals and Bilinguals) and NNS through dialogue may have reproduced a more natural setting for judging comparison of linguistic background comprehensibility variability.
However, it is evident from the results of both the precursor study and the later one by Saito & Shintani, Unquestionably, comprehensibility as a construct involves two stakeholders; listeners and speakers either in NS -NNS or NNS -NNS communicative settings and both interlocutors should strive to overcome the factors outlined above. Teachers can play an instrumental role in improving L2 comprehensibility by increasing the proficiency in all the four categories of speech: phonology, fluency, lexicogrammar, and discourse.

ACKNOWLEDGEMENT:
Special thanks to Dr Heath Rose, Associate Professor of Applied Linguistics, Department of Education at the University of Oxford, without his mentorship and guidance this essay could not have been possible.

CONFLICTS OF INTEREST:
There are no potential conflicts of interest in the research analysis, writing, and publishing of this essay.

REFERENCES:
1) Crowther, D., Trofimovich, P., Saito The same French L2 participants and the same picture story used in Isaacs & Trofimovich, (2012) were employed. Instead of first30sthe entire narration was recorded (100 words minimum) which increased the difficulty level for both the interlocutors.
Since the raters were bilingual Canadian NSs with French as their L2, familiarity withL1 accent assisted comprehensibility.