A Multidimensional Analysis of Pakistani English Written Discourse

The present study analyzes Pakistani English written discourse through multidimensional analysis. For this purpose, two corpora types, including Pakistani universities' newsletters and Pakistani human rights NGOs' annual reports, have been complied. It has been investigated how the language of newsletters of Pakistani universities differs from that of annual reports of Pakistani human rights NGOs on D1 and D2 of the multidimensional approach. Biber’s (1988) multidimensional analysis provides the theoretical grounding to the present study. MAT software (1.3) has been used to tag and analyze the data. Co-occurrences of linguistic features are quantitatively analyzed and then qualitatively interpreted through D1 and D2 of multidimensional approach. The results reveal that the language of this written discourse of Pakistani English is informational on D1. By comparing it with Biber's work of 1988, it has been revealed that this genre is close to the official documents on D1. The language of Pakistani universities' newsletter is more informational on this dimension than that of the genre of Pakistani NGOs annual reports as the mean scores for both are -26.79 and -25.13, respectively. The analysis of the D2 indicates that the genre of the selected written discourse is non-narrative. However, Pakistani universities' newsletters are close to broadcasts, whereas the annual reports of human rights NGOs are close to the personal letters. Moreover, the newsletters discourse is strongly non-narrative than that of the annual reports.


Introduction
This paper is an attempt to explore the linguistic features of a newly emerging variety i.e., Pakistani English. Pakistani English is a distinct and non-native variety of English (Uzair, 2011). These scholars contributed a lot in giving the name to this variety on the linguistic map. Mostly, the work has been done in phonetics, phonology, lexis, and semantics. There is a little research on the genre variation of this variety. The present study aims to contribute in the discipline of world Englishes through multidimensional approach.
CONTACT Isra Irshad at isra.irshad@uog.edu.pk Corpora of two types of texts of Pakistani English written discourse have been compiled. These include Pakistani universities' newsletters and the annual reports of Pakistani human rights NGOs. Both of these documents indicate the progress of any institution in the light of the previously planned task. Universities' newsletters and the annual reports of NGOs indicate the progress of the particular organizations' ongoing tasks. Current events (seminars, conferences, cocurricular activities, sports, convocations, etc.) held in the universities are discussed in the newsletters, and thus, the newsletters indicate how much and in which direction that institution has made progress. In the same way, the annual reports of NGOs, particularly human rights NGOs, highlight the progress in the state of human rights in a particular year. The present study analyzes these documents written in Pakistani English. Thus, this study is confined to this type of discourse written in Pakistani English. There is a little description of it in the literature as the linguistic variation of this discourse has not previously been studied.
Moreover, most of the past studies have focused on the individual linguistic features to define a genre or register. These studies have a limited scope because of reliance on unrepresentative data and individual linguistic features (Ahmad & Mahmood, 2015). There is a need to analyze the linguistic variation using multidimensional analysis, quantitative, empirical, and corpus-based. This analysis focused on the linguistic variation in five functional dimensions proposed by Biber (1988). The present study analyses Pakistani written discourse by focusing on two types of documents: Pakistani universities newsletters and annual reports of human rights NGOs on D1 and D2 of multidimensional approach. Hence, this study investigates the following research question: • How does the language of newsletters of Pakistani universities differ from that of Pakistani human rights' annual reports on D1 and D2?

Significance of the study
The study adds to the field of World Englishes as it will help us analyze the language and function of Pakistani English written discourse, particularly regarding the texts of Pakistani universities' newsletter and Pakistani annual reports of human rights NGOs.

Literature Review
Several studies analyze the registers or genres of languages using multidimensional analysis. Early researches using this approach were confined to academic discourse. Reppen (2001) applied it to the learners' data as he collected the data from a low standard grade. Thus, by this data, he focused on the learners' writing techniques. Using this approach, Concord (1996) investigated three types of texts-textbooks of ecology courses, common composition textbook selections, and ecologicalbased professional research articles. The findings reveal that there exists a complex relationship between research writing and academic texts. This approach has been applied in the areas of fiction, prose, etc. Biber and Finegan (1994) analyzed the characteristics of the writer's style by using this approach. In the same way, Egbert (2012) analyzed the writing style of 19-century fiction by using this approach.
Similarly, this approach has been applied in the Pakistani context. Amjad and Shakir (2014) analyzed the online university prospectuses of Pakistan, India, and the UK using this approach on D1 of multidimensional model approach. It revealed that their language is informational, having prepositions and nominalization. Moreover, Indian prospectuses are more informational than those of the UK. The department section is more informational in the prospectuses than all other sections, and the section of VC message is less informational. In the same way, Nawaz and Shakir (2014) have also analyzed the same prospectuses corpora on D2-narrative or non-narrative, and it was found that their language is non-narrative. They have the features of attributive adjectives, present tense, and interactional markers. Asghar, Mahmood and Asghar (2018) have investigated linguistic variation in various Pakistani Legal English genres by using multidimensional analysis. The findings indicate that legal language is not a homogenous phenomenon, and it has a variety of linguistic features related to different legal genres; thus, it must be seen in terms of audience, purpose, goal, and context. Rashid and Mahmood (2019) have investigated the linguistic identity of the Pakistani Academic writing register of Research Articles. It is found that Pakistani academic research articles language as informational, non-persuasive, highly impersonal, explicit, and nonnarrative. Asghar, Mahmood and Asghar (2018) have investigated linguistic variation in various Pakistani Legal English genres by using multidimensional analysis. The findings indicate that legal language is not a homogenous phenomenon, and it has a variety of linguistic features related to different legal genres; thus, it must be seen in terms of audience, purpose, goal, and context. Rashid and Mahmood (2019) have investigated the linguistic identity of the Pakistani Academic writing register of Research Articles. It is found that Pakistani academic research articles language as informational, non-persuasive, highly impersonal, explicit, and non-narrative.
However, the language of Pakistani universities' newsletters and annual reports of human rights NGOs have not previously been analyzed through multidimensional analysis. The present study has selected these two types of documents from Pakistani English's written discourse and has analyzed them on D1 and D2 of this analysis.

Theoretical framework
As proposed by Biber in 1988, the multidimensional approach has provided the theoretical framework to the present study. It is a kind of methodological approach used for genre analysis and register analysis. It helps identify particular linguistic co-occurrence patterns quantitatively in a particular language (Alvi, Mehmood & Rasool, 2016). Biber (1988) analyzed the news reportage genre in his study of variation across writing and speech. He made the statistical factor analysis of 67 variables by investigating 481 texts and established the following six dimensions: 1. Involved versus Informational 2. Narrative versus Non-Narrative 3. Explicit versus Situation-Dependent 4. Overt Expression of Argumentation/ Persuasion 5. Abstract versus Non-Abstract 6. Online Informational Elaboration Biber (1993) has proposed the following characteristics of MDA: • It is a corpus-based approach.
• It is a computer-based approach as it automatically analyses linguistic features of the text.
• Its focus is on the linguistic analysis of registers, genres, texts, styles rather than on the analysis of individual linguistic features. • It combines a quantitative approach with that of a functional one. It involves the interpretation of statistical analysis in practical terms. In other words, there is an assumption that statistical co-occurrence patterns indicate the underlying communicative functions. Thus, the concept of linguistic co-occurrence is crucial in the analysis of any particular genre or text type on the linguistic dimensions. In the present study, the genre of written discourse in Pakistani English has been analyzed on D1 and D2 of this multidimensional approach. For this purpose, two types of the written documents, Pakistani universities' newsletters and annual reports of human rights NGOs in Pakistan, have been taken into account. The investigation of their cooccurrence of linguistic features helps us understand where their language falls on dimensions 1 and 2. It has also highlights their communicative functions.
D1 on multidimensional analysis deals with the involved and informational features. It posits that the negative scores indicate that the text is informational, and positive scores show that the text is involved. In the same way, D2 deals with the narrative and non-narrative functions. The negative score is linked with the non-narrative function of the text, and positive scores reflect the text's narrative function.

Methodological Framework
Multidimensional analysis has been used as a methodological framework to analyze the data. The following steps have been taken to investigate the research as mentioned above question. MAT tagger software (1.3) has been downloaded, installed, and used as the primary research tool. Nini develops this software in 2013 according to Biber's model (1988). It only deals with the plain text. Data is annotated by using Stanford tagger. It creates the file in MAT_ file name. Then, this tagged data is analyzed in the analysis operation. Then, data is created in another folder named MAT_MAT_file name. It includes five files which are as following: These files include information about dimensions and scores of various linguistic items. These are taken to interpret the data by putting in the software for analysis.

Compilation of corpora
Corpora of the two types of documents from Pakistani English written discourse have been selected. The following illustration is helpful in this regard: The data have been collected from the internet. These files were available in pdf form and thus have been converted into plain texts. The corpus of 1 million words has been complied through convenient sampling by taking 5 lac words from the 100 newsletters (written in Pakistani English) of 20 Pakistani universities and 5 lac words from the ten annual reports of human rights NGOs from Pakistan.

Data Analysis and Discussion
The designed corpus of the selected texts has been analyzed on D1 and D2 of multidimensional approach. In this section, the multidimensional analysis of the two types of documents of Pakistani English written discourse on D1 and D2 has been discussed. According to this analysis, the positive score on D1 indicates the non-informational and involved functions. For example, conversation and prepared speeches have this function. On the other hand, the negative score on this dimension highlights the informational and non-involved function, for example, academic prose, press reportage, and official documents. Thus, the linguistic features of these positive and negative scores vary. The positive side includes present tense, first and second person pronoun, private verb etc. On the other hand, linguistic features on the negative side include prepositional phrases, nouns, and attributive adjectives. In the same way, D2 deals with the narrative and non-narrative functions. A negative score is linked with the text's non-narrative function, and positive scores reflect the narrative function of the text.
In the present study, two types of documents have been selected. These include Pakistani universities' newsletters and annual reports of human rights NGOs. The discussion begins with the analysis of these documents on D1. Firstly, both these types of texts have been analyzed on D1 separately, and then they have been compared.

Analysis of the Pakistani Universities' Newsletters on D1
The mean score of Pakistani universities' newsletters on D1 is -26.79. Thus, it is the negative score that indicates that the text is informational, and the closest genre type to it, according to Biber's multidimensional approach, is official document as the following picture is illustrating: -1.12 ______________________________________________ The following paragraph is the sample text from the Pakistani universities' newsletter having the rich informational content: The University held its 18th Convocation on 18th November 2017 at Jinnah Convention Center Islamabad. Admiral Zafar Mahmood Abbasi NI (M) was the Chief Guest of the ceremony and conferred degrees on the graduating students of Ph.D, Masters and Undergraduate programs. Addressing on occasion, Admiral Zafar Mahmood Abbasi NI (M) congratulated the graduating students on their splendid success and lauded Bahria University Islamabad Campus for providing exceptional knowledge skills to the graduated students in their respective disciplines, enabling them to lead a professionally rewarding life. He appreciated the role of Bahria University in contributing to strengthening the human resource of the country.
In the example mentioned above, the factors (noun, adjective, and preposition) are quite commonly observed. Nouns include, for example, 'university', 'convocation', 'Islamabad', 'Zafar Mahmood Abbasi' etc. Adjectives include 'Admiral', 'undergraduate', 'human' etc. Prepositions include 'at' 'of', 'on', 'to', 'in' etc. The presence of these factors is making the text informational and non-interactive. Biber (1993) posits that such texts are like prose and fiction. Biber's (1988) score for academic prose is -15, but the current genre score on D1 is -26.79, which is even farther away. The high frequency of nouns in the text signifies the integration of huge information in a text. The adjectives are used to paint the setting, theme, characters, and situation. In the example mentioned earlier, the next type of the selected documents under discussion is the corpus of annual reports of human rights NGOs in Pakistan.

Analysis of Annual reports of Human rights NGOs in Pakistan on D1
The mean score of annual reports of human rights NGOs in Pakistan on D1 is -25.13. Thus, it is the negative score, and the closest genre type to it is official documents, as the following picture is illustrating: The following paragraph is the sample text from the annual reports of human rights NGOs in Pakistan having the rich informational content: The sources, where not quoted in the text, were HRCP surveys, fact-finding reports, and communications from its correspondents and private citizens; official gazettes, financial and legal documents and other public releases and statements; reports in the national and regional press; and publications of international agencies like the UNDP, ILO, WHO, UNICEF and the World Bank. Considering the limitation of the official reports, press accounts, and sample surveys conducted by NGOs, figures and assessments offered here may not always represent the full or the exact picture. They should be taken as a reflection of the trend during the year.
In the above mentioned example, the factors (noun, adjective and preposition) are quite commonly observed. The linguistic features for newsletters on D1 include noun, adjective, preposition etc. Nouns include, for example, 'sources', 'surveys', 'reports', 'communications', 'correspondents', 'citizens', 'gazettes', 'documents' etc. Adjectives include 'HRCP', 'private', 'official', 'public', 'national', 'regional', 'international' etc. Prepositions include 'in', 'of', 'during' etc. The following table indicates the scores for these features: The presence of these factors is making the text informational and non-interactive, like prose and fiction. Biber's (1988) score for academic prose is -15, but the current genre score on D1 is farther away from it. The high frequency of nouns in the text signifies the integration of massive information in a text. The adjectives are used to paint the setting, theme, characters, and situation. Next, the comparison of both these types of documents has been made with each other and Biber's genres (1988) on different dimensions.

Comparison of both types of documents with each other and with those of Biber' genres (1988) on D1
The discussion as mentioned above indicates that two types of documents, Pakistani universities' newsletters and Annual reports of human rights NGOs in Pakistan, have been selected. The mean score for the newsletters is -26.79; whereas the annual reports' mean score is -25.13. There is a slight difference in the mean scores between these two documents on D1. The production of newsletters is more informational than that of annual reports. The following graphical representation highlights the comparison between these two types of documents:

Annual reports of Human rights NGOs in Pakistan
There is a need to discuss how two types of documents differ from those of 6 types of texts investigated by Biber in 1988. Thus, the following illustration is helpful in this regard:  Biber (1988)

Analysis at Dimensions 2
The present study also analyzes the text at dimension 2 of multidimensional approach. This dimension talks about the narrative or non-narrative dimensions of the text. Negative scores in D2 indicate that the text is non-narrative, whereas the positive scores indicate that the text is narrative. Linguistics features/factors in both of these scores can be observed with the help of the following illustration: In the present study, both types of documents have also been analyzed on D2 one by one and then compared. The discussion begins with a D2 analysis of universities' newsletters.

Analysis of Pakistani Universities' Newsletters on D2
The mean score of Pakistani universities' newsletters on D2 is -3.50, which is negative and indicates that the text is non-narrative. The closest genre type to it, according to Biber's work of 1988, is broadcast as the following illustration is indicating: The negative scored linguistic features for the newsletters on D2 include attributive adjective, present tense verb, and the following table indicates the scores for these features: Achieving a landmark success, Pakistan's most prestigious educational institute Punjab University in the recent survey by the Higher Education Commission of Pakistan, has improved its ranking. Now it stands 4,h with 352 research publications in impact factor journals in the year 2010. The credit goes to the current administration's focus on research, as the research budget has been increased from Rs 4 million to 70 million, while Vice-Chancellor Dr Mujahid Kamran has vowed to increase it 100 million if necessary. Since Prof. Dr Mujahid Kamran has taken charge of the alma matter in January 2008, the number of research publications of the varsity has more than doubled. This indicator speaks of academic progress and the honorable Vice Chancellor's measures for promoting research culture in the university. (Punjab University Newsletters) In the example mentioned earlier, the factors (attributive adjectives and present tense verbs) are observed. Attributive adjectives include, for example, 'landmark', 'prestigious', 'honorable' etc. and present tense verbs include 'speaks', 'stands', 'goes' etc. The presence of these factors is making the text non-narrative in nature. The attributive adjective is used to explain nominal information, whereas the present tense verb is used to deal with the present matters.

Analysis of Annual reports of Human rights NGOs in Pakistan on D2
The mean score of annual reports of human rights NGOs on D2 is -0.13, which is negative and indicates that the text is non-narrative. The closest genre type to it, according to Biber's work of 1988, is a personal letter, as the following illustration is indicating: The following paragraph is the sample text from the annual reports of human rights NGOs in Pakistan. It has very few linguistic contents of attributive adjective and present tense verb, making the text non-narrative with the mean score of -0.13, which is nearest to 0 score. The following example is taken from one of the annual reports of human rights NGOs in Pakistan having the linguistic features of attributive adjective and present tense verb: Considering the limitation of the official reports, press accounts, and sample surveys conducted by NGOs, figures and assessments offered here not always represent the full or the exact picture. They should be taken as a reflection of the trend during the year. (2010: 1).... Ratification by Pakistan of all core international human rights treaties was among the positive highlights of the year, although the benefits were not immediately visible. The factors include attributive adjective ('positive') and present tense verb ('represents') are rarely observed in the above example. The scores for these linguistic features are as following: The presence of these factors are making the text non-narrative in nature. However, these features are rarely found in the text, as the scores as mentioned above for attributive adjective (-0.13) and present tense verb (0.88) indicate. Next, the comparison of both these types of documents has been made with each other and Biber's genres (1988) on dimension D2.

Comparison of both types of documents with each other and with those of Biber' genres (1988) on D2
As mentioned earlier, the discussion indicates that the mean score for the newsletters on D2 is -3.50, whereas the mean score for the annual reports is -0.13. Though both the text lie in the nonnarrative category, the text of the newsletters is strongly non-narrative than that of annual reports of human rights NGOs, as the mean scores for both on D2 indicate. The following graphical representation highlights the comparison between these two types of documents: Further, it has been analyzed where both of these types of documents lie on D2 concerning other prescribed texts by Biber (1988). Thus, the following picture will help us to understand this: Figure 4. Comparison of the selected documents with those of the six genres studied by Biber (1988) Thus, the above-mentioned discussion illustrates that Pakistani universities newsletters are close to the broadcast genre, whereas the annual reports of human rights NGOs are close to the personal letters.

Findings and Conclusion
The present study has focused on the language of written discourse of Pakistani English. It includes universities' newsletters and annual reports of human rights NGOs. Both types of documents indicate the progress of the institution. It highlights that their language is informational on D1. By comparing it with Biber's work of 1988, it has been revealed that this genre is close to the official documents. Moreover, the language of Pakistani universities' newsletter is more informational on D1 than those of Pakistani NGOs' annual reports as the mean for both is -26.79 and -25.13, respectively. The analysis of the D2 indicates that the genre of the selected written discourse is non-narrative. However, Pakistani universities' newsletters are close to broadcasts, whereas the annual reports of human rights NGOs are close to the personal letters. Moreover, the newsletters discourse is strongly non -narrative than that of the annual reports.
Furthermore, the study has much significance for the prospective researchers generally working in linguistics, specifically in the register analysis. It will allow them to analyze these documents on other dimensions (3, 4, 5, 6) of multidimensional analysis for a better and comprehensive understanding of the linguistic variations in the Pakistani English written discourse, particularly regarding the genres of Pakistani universities newsletter and annual reports of human rights NGOs in Pakistan. Moreover, it will also guide the writers of universities' newsletters and those of human rights NGOs' annual reports about using certain linguistic features.