A Multidimensional Analysis of Pakistani Press Editorials.
The importance of linguistic variation has been discussed in a variety of theoretical, methodological and practical studies. Pakistani English (PE) is an emerging independent, non-native variety of English. However, currently there is a lack of a comprehensive description of its characteristics in the literature. The present research addresses this problem by investigating the linguistic variation of Pakistani English newspaper editorials (Pak. Ed.). It aims to identify if the language used in PE is a distinguished linguistic sub-register of Pakistani Newspaper English. In order to address this question, the technique of multidimensional (MD) analysis is utilized. A diverse dataset, consisting of 250 editorials from five leading newspapers, has been compiled which provides a variety and range of topics covered by different authors.
The resulting corpus is tagged for 67 linguistic features and factor analysis has been conducted, to identify major linguistic patterns of co-occurrence. Five textual dimensions are utilized to define some of the linguistic and functional characteristics of Pak. Ed. corpus. The dimensions collectively provide information content, narrative features, situation-oriented references, extent of argumentative language and abstract style. A comparison of Pakistani editorials with British editorials shows marked linguistic variation and indicates the Frontier Post to be the closest newspaper to British newspapers in terms of editorial writing. On the other hand, the Dawn displays quite distant results among all dimensions except in the narrative features. Based on these results, it is proposed that linguistic variation occurs in the editorial writings of Pakistani English on both national and international levels.
Keywords: Linguistic variation, Pakistani English, MD analysis, Newspaper Editorials, Linguistic patterns of co-occurrence.
Newspaper editorials are important media discourse and have their own distinguished linguistic features. According to Biber1 every variety or sub variety of a language or a register has systematic linguistic and functional variation, which is displayed by its specific linguistic features. Typically, the linguistic features of editorial states opinion, directs readers' attention and expresses the ideologies of a news organization. These features of editorials provide a motivation to study their linguistic structure with regards to their social, functional and cultural perspectives. Pakistani newspaper editorials are relatively not well-studied. However, they provide an interesting representation of the English language.
It is likely that Pakistani English is a distinct variety of non-native English varieties which has its own registers and sub- registers with systematic linguistic variation. This is precisely the main focus of this study as it aims to understand the register- specific features of the major sub-register of Pakistani newspaper writing, the Pakistani newspaper editorials. For simplicity, from this point onward, the abbreviation Pak. Ed. for Pakistani newspaper editorials is being used.
In Pakistan, research on generic and rhetorical structure of different newspaper editorials is growing. This has resulted in the exploration of its discourse, meta-discourse and individual linguistic features with the help of comparative analyses of content and genre. There are some drawbacks with these methodologies. Firstly, these methodologies are not based on large corpora. Secondly, these studies assess isolated linguistic features.
Thirdly, they do not provide empirical evidences through proper statistical measures. It is possible to overcome all these limitations by adopting the corpus methodology approach of register variation. There are several advantages of this methodology. It provides a comprehensive analysis through identifying the co-occurring linguistic patterns of register variation. It is conducted on larger corpora of a language variety, register and sub register.
It addresses the analysis of a set of linguistic features with the combination of statistical analysis. It provides empirical evidence of linguistic variation or differences between different varieties and registers.
Consequently, it highlights the usefulness of studying register variation with co-occurring linguistic patterns. Thus, the present research adopts the same approach to conduct a comprehensive analysis of Pakistani newspaper editorials. To study register, Biber2, 3 has given a complete methodology of multidimensional (MD) analysis of the English language, which provides an understanding of the variation among different registers and different languages. Further details of MD analysis, are given in the methodology section. The present study focuses on the register variation on internal and external levels. The internal variation highlights the linguistic variation among five leading newspaper editorials to see how far these newspapers are different or similar to the application of MD analysis.
The external variation determines the linguistic variation among Pakistani newspaper editorials with British newspaper editorials (Brit. Ed.) to see how much Pak. Ed. is different or similar to a native variety of English.
As described earlier, newspaper research in Pakistan, particularly on editorial writing, has increased. Most of the studies are based on the generic and rhetorical structure of different texts. These studies depict some social, cultural and political aspects in editorial writing. In the past decades, the methodologies have also been limited to discourse and rhetorical analysis, which could be applicable to certain text limits. Shabbir4 has done a comparative discourse analysis of the editorials of the Hindustan Times and The Dawn during post-nuclear test scenarios. The study covered editorials with special reference to 'Nuclearization in South Asia' and concluded that the media discourse aggravated the political issues to formulate the public opinion on any foreign policy. Razi5 has done a qualitative research on two leading English dailies of Karachi with interviews and direct observation of the text through content analysis.
The content analysis was done through analyzing the narration, arguments, word choice, objective style of representation, topic choice and the type of readership. The analysis showed that Pakistani English Press is playing the credible role for decision makers with the help of editorials. But sometimes the inappropriate word choice appeared in editorials showed the misuse of freedom. Rehman and Eijaz6 conducted a research on two popular Urdu and English dailies to see the difference in portraying a social issue of Lal Masjid. The findings showed that both dailies have the tendency of showing conflict journalism instead of showing the characteristics of peace. There is a significant difference in media coverage. The English daily showed binary opposition and conflicted religious element, whereas Urdu daily gave overall low coverage to the issue.
Khan and Safdar7 conducted a comparative study on one English Pakistani daily and the other Urdu Pakistani Daily to see the US portrayal after 9/11 incident through media conformity Theory. The theory argues that the mass media generally conform the foreign policy of governments. With the help of content analysis, the findings did not support the media conformity theory in this case when US appeared negative in both dailies after 9/11. Shabbir, Et al8 has conducted a comparative analysis of editorials in two leading English dailies with the help of content analysis. The aim of the study was to examine the editorial coverage towards the government policy. The findings showed the both dailies gave coverage and at times criticized the government policies.
It is clear that most of the researches based on editorials have addressed the communicative and political issues. However, still there is a need for studying different genres of newspaper writing in the detailed grammatical description. Therefore, the present study is conducted to analyze a particular genre with comprehensive linguistic description. Another important point is, PE is being studied largely to find out its individual linguistic features, but there is a dearth of studies to find out its own distinguished features for its register exploration. Pakistani Newspaper Editorials (Pak. Ed.) as a sub register of PE variety are also needed to be studied with this perspective by applying MD Analysis.
Collection of Data and Corpus Compilation
In the following, the choice of dataset is described to construct a proper corpus. The corpus was compiled with the goal of selecting latest newspaper editorials from five leading newspapers i.e., The Dawn, The Daily Times, The Nation, The News International and The Frontier Post. They represent all major cities of Pakistan; Lahore, Karachi, Islamabad, Quetta and Peshawar. Editorials written by top writers were selected keeping in view the variety and wide range of topics. The word range of editorials contained maximum 1000 words per editorial. The corpus contained 250 editorials each with text from a single editorial and varying word lengths ranging from 600 to 1000 words. On average, each editorial contained 800 words. The corpus was retrieved from the websites of all these newspapers. Editorials were gathered on three bases. Firstly, 10 editorials were selected from each author's writing.
Secondly, top opted authors were selected who have written frequently in their particular newspaper. Thirdly, a wide range of topics based on social, economic and political issues were included. Once the editorial was selected, its contents were downloaded and converted into text files. With the exception of the material written in the body of an editorial, all other information was excluded from the corpus. 50 articles were selected from each newspaper. Thus the compiled corpus contained 250 editorials. Afterwards, the corpus was automatically tagged and analyzed through MAT tagger. The tagger identified the co-occurrence of 67 linguistic features on five textual dimensions and provided the mean dimensional scores. As the present study is mainly concerned with the comparison of Pak. Ed. with Brit. Ed. on the basis of mean dimensional scores, therefore the same scores of Brit. Ed. were taken from MD Analysis.9
The MD approach for comprehensive analysis of register variation was first used in an investigation of the register variation in English and later in other languages, for cross cultural perspective.10 Moreover, it is used to identify the salient linguistic co-occurrence patterns in a language quantitatively. The co- occurrence patterns are analyzed as underlying dimensions i.e., involved versus informational production, narrative versus non- narrative concerns, and the degree of referential elaboration, which can be persuasive or argumentative, and abstract versus non- abstract style. In this study, Pakistani newspaper editorials are compared with British newspaper editorials on the basis of their dimensional scores calculated through MAT tagger, a proper tagging program. It replicates the analysis of Biber11 MD analysis is useful for studying linguistic variation due to certain reasons.
It provides the comprehensive description as it is based on a cluster of linguistic features instead of an individual feature. It is used to investigate a larger corpus of naturally occurring discourse with computational techniques. Its cumulative results are applicable on additional corpora. It investigates the full range of spoken and written registers in the English language as well as other languages on the same basis.
There are two types of MD analyses, known as the old MD analysis (which is abbreviated as 88 MD analysis ) and the new MD analysis. MAT tagger automatically provides the reliable results of MD analysis 1988 as the sets of co-occurring linguistic features are already identified with five major textual dimensions.
Textual Dimensions of Variation (Biber, 1988)
The following analysis is based on five dimensions of variation as described in Biber's Variation across Speech and Writing.12 The dimensions are based on the sets of co-occurring linguistic features that share a communicative function. A correlation between linguistic features can be either positive or negative. A large positive correlation means that the features in question systematically occur together, while a large negative correlation means that those features systematically co-vary in a complementary way. The communicative functions for each dimension are then assessed and used to interpret results. The following dimensions are considered:
Dimension 1: Involved vs. Informational Production Dimension 2: Narrative vs. Non-narrative Concerns Dimension 3: Explicit vs. Situation Dependent Reference Dimension 4: Overt Expression of Persuasion/Argumentation Dimension 5: Abstract vs. Non-abstract Style
Comparison of Pakistani English Newspaper Editorials and Biber's Model of Linguistic Variation on Dimensional mean score The following table shows the average dimensional score on each textual dimension.
Table 1: Comparison of Pak. Ed. with Brit. Ed. of Dimensional Scores through Multidimensional Analysis (1988)
No.###Textual Dimensions of Linguistic Variation###Pak. Ed.###Brit. Ed
D1###Involved vs. Informational Discourse###-14.08###-10.0
D2###Narrative vs. Non-narrative Concerns###-0.84###-0.8
D3###Explicit vs. Situation Dependent Reference###5.51###1.9
D4###Overt Expression of###0.62###3.1
D5###Abstract vs. Non-abstract Information###2.20###0.3
The comparison between Pak. Ed. and Brit. Ed. shows clearly that the variation exists on D1, D3, D4 and D5. Although it is found that there appears a similarity on D2 dimensional scores. As far as the comparison among five newspapers is concerned, the table 2 is given to describe it in detail.
Table 2: Newspapers Dimensional Score through Multidimensional Analysis
D2###Narrative vs. Non-###-0.75###-1.22###-1.01###-0.90###-0.59
D4###Overt Expression of###0.67###-0.15###0.23###0.61###1.19
D5###Abstract vs. Non-###2.05###2.92###1.82###3.03###1.70
On the basis of the data of Table 1 and Table 2, the following description presents the linguistic variation among Pak. Ed. and Brit. Ed. as well as among five leading newspapers.
Dimension 1: Involved versus Informational Production According to 88 MD analyses, D1 on its positive side shows the involved and non-informational functions, for example the registers of face to face conversations, telephone conversations and public conversations and on its negative pole, this dimension has informational production such as official documents, academic prose, press reportage etc. The linguistic features on the positive pole represent interactive and casual discourse through using present tense, private verbs, 1st and 2nd person pronouns etc. On the negative pole features, such as nouns, prepositional phrases and attributive adjectives are marked. The mean score of the Pak. Ed. corpus, -14.08 displays a different degree of informational production from Brit. Ed. It shows that Pak. Ed. is more informational than Brit. Ed.
The distinctive characteristics of this text type are a marked informational focus, where the narration is mainly used as a part of the expository information being conveyed. It shows that the nouns, prepositions, and attributive adjectives are being used in an excessive manner. However, the comparison among newspaper shows that the closest score is the score of The Frontier Post (at -12.52), compared to the Brit. Ed. (at -10.0). The Dawn and The News International have much higher scores, -15.55 and -15.38, respectively. The Daily Times and The Nation are also more informational than Brit. Ed. with the scores of -14.98 and -13.50, respectively. It is noted that this result in not surprising as editorials are written to convince the audience.13
They have the objective of providing information according to their background knowledge. Pak. Ed. seems to have more technical information through a high content of nouns and other informational features. As example, Table 3 shows a text sample from Pak. Ed.
Table 3. A sample text from Pak. Ed. corpus to illustrate the rich informational content of the text.
File###Dimension 1 score###Text Sample
DPNE 3###-14.08###The House of Commons speaker
###enjoys no such power. He can ask the
###MP to withdraw his remarks and if he
###fails to do so he can suspend him. This
###is a purely Indian innovation. Articles
###105(3) and 194(3) empower,
###respectively, parliament and the state
###legislatures define their powers,
###privileges and immunities "by law".
Dimension 2: Narrative versus Non-narrative Concerns
According to the 88 MD analysis, D2 has the features such as past tense, perfect aspect verbs, third person personal pronouns, public verbs, and present participial subordinate clauses that show a narrative function in the discourse. There are only two features with slight negative loading on D2 i.e. present tense and attributive adjectives that show the element of non-narrative discourse. The texts with a high positive D2 score have a narrative focus (e.g., a novel) while texts with a low D2 score such as press reviews have non-narrative focus. The mean D2 score of the Pak. Ed. corpus is - 0.8, which is low but almost equal to British register, which is -0.5. It shows it similar to the British newspaper editorials. Pakistani newspaper editorials are also non-narrative. Both registers usually have a present tense to explain situations and events. This indicates that both registers have non-native style of writing.
This also shows that the negative loading features of present tense and attributive adjectives are being used. The writing style is descriptive, instead of being event oriented. However, all of the five newspapers are found to be non-narrative in style on different grounds. The Dawn and The Frontier Post have the closest score, - 0.7 and -0.5, respectively, compared to Brit. Ed. score of -0.8, while The News International and The Daily Times have relatively high scores, -1.22 and -1.01, respectively. According to Biber14 non-narrative writing's prominent linguistic feature is the present tense to describe actions. It provides simple and concise information. Overall, five Pakistani newspaper editorials are non- narrative. This is illustrated by an example in Table 4.
Table 4: A sample text from Pak. Ed. to illustrate non-narrative context of the text.
File###Dimension 2 score###Text Sample
DPNE###-0.84###In 1958, the Supreme Court ruled that
196###"the House of Commons had at the
###commencement of our constitution the
###power or privileges of prohibiting the
###publication of even a true and faithful
###report" of the proceedings and also that
###the privileges prevail over the
###fundamental right to freedom of speech.
Dimension 3: Explicit versus Situation-Dependent Reference According to 88 MD analysis, D3 shows the differentiation between two kinds of references, i.e., elaborated and situation dependent reference. The positive loading or high D3 score tends to represent a high frequency of three relative constructions, which include wh-relative clauses on object position, wh-relative clauses on subject position and pied-piping construction. The other co- occurring patterns of this dimension are phrasal coordination and nominalizations. According to the D3 mean scores, Pak. Ed. corpus shows an elaborated and explicit identification of referents. This indicates the writers used more wh-relative clauses on an object and subject position and pied-piping constructions. It also shows the informational manner of nominalizations and phrasal coordination.
However, in British editorials, D3 has a low score, which shows that its register is relatively close to the biographies and spontaneous speech. This means that these texts are situation- dependent and are loaded with the features of many adverbials used for deixis, as mentioned above. Overall, the corpus of Pak. Ed. does not share similarity with Brit. Ed. However, the comparison among all newspapers shows the Frontier Post to be the closest newspaper, with the score of 4.11, compared to Brit. Ed. score of 1.3. The rest of the newspapers, have quite different scores, as shown in Table 1 and 2.
Table 5. A sample text illustrating explicit style in Pak. Ed. corpus.
File###Dimension 3 score###Text Sample
DPNE###5.51###INDIAN Prime Minister Narendra
39###Modi's stopover in Lahore on
###Christmas Day seemed to have caught
###the media totally unawares and the top
###guns belonging to the reputedly well-
###informed anchor and commentator
###community didn't seem to like the way
Dimension 4: Overt Expression of Persuasion
According to 88 MD analysis, D4 has only features of positive loadings i.e. prediction modals, necessary modals, possible modals, persuasive verbs, conditional subordinates and split auxiliaries. All these features mark persuasion. The low D4 scores generally do not indicate an overt or covert expression of persuasion in the text.. Consequently, the low D4 score of Pak. Ed. corpus shows that it has low elements of overt expressions of persuasion. In the case of all these five newspapers, the results indicate that the Pak. Ed. corpus does not seem to convince their audience. They provide a subjective analysis of events and situations. It is interesting to note that, among all newspapers, The Frontier Post has a score of 1.19, indicating relatively higher elements of persuasion. The rest of the newspapers have scores below 0, as shown in Table 2. Among all these four newspapers, The News International has the least score, -0.15.
This indicates that the writers of this newspaper have a higher objective and the least argumentative style of writing. This is demonstrated in Table 6.
Table 6. A sample text illustrating less expressions of argumentation in text of Pak. Ed. corpus.
File###Dimension 4 scores###Text Sample
DPNE###0.62###CHINA cannot afford a
208###neighborhood in chaos because that
###would greatly affect its geo-
###economic engagements and
###priorities in the region. China is also
###worried that if any of its neighbors is
###caught in turmoil, it could become
###China's strategic, political and
Dimension 5: Abstract versus Non-abstract Information
According to 88 MD analysis, a high D5 shows the level of abstractness in a text. It represents the informational discourse that is formal and technical. The positive features on this dimension are by-passives and agent-less passives, adverbial subordinates, conjuncts, past participial WHIS deletions and predicative adjectives. The negative loading features on this dimension are a type/token ratio that shows the low lexical variety using a small set of technical vocabulary for avoiding ambiguity. Low D5 score genres include fiction and face-to-face conversations. This dimension is also called impersonal vs. non-impersonal style. The Pak. Ed. corpus has the mean score of 2.21, which shows to have abstract style by using the following linguistic features in much more quantity, i.e., agent-less passives, by passives and adverbial subordination, adverbial past participle clauses.
In general, these passive constructions have a formal style which is found in academic prose while Brit. Ed having the score of 0.3. It seems that British editorials are written in non-abstract or non-technical manner. Brit. Ed. score on this dimension indicates that this register is having the quality of fiction and face to face conversation. However, the corpus of Pak. Ed does not tend to have this style on average. Overall, The Frontier post's score 1.70 is closest to the Brit. Ed score 0.3. The other closer newspaper is the Daily Times with a score of 1.82. The far distant newspaper is the Nation having the score of 3.03. The Dawn and the News International are having the scores 2.05 and 2.92 respectively. Overall the corpus of Pak. Ed. seems to have an abstract style of editorial writing. For example,
Table 7. A sample text illustrating abstract style of writing in text of Pak. Ed. corpus.
File###Dimension 5 score###Text Sample
DPNE 63###2.20###FOR a city that prides itself on the
###wit and subtle humor of its
###inhabitants, Lyallpur, also known
###among newcomers as Faisalabad,
###should have been worried a long time
###ago. For a party that must always
###propagate discipline and single-
###minded advance, it is intriguing how
###the PML-N could allow this fracas
###involving family and friends in the
###name of grass-roots democracy.
The graphical representation of linguistic variation among the corpus of Pak. Ed. and Brit. Ed. as well as among five leading newspapers is also being displayed for showing the differences and/or similarities.
In this study, MD analysis is used in the Pakistani editorials corpus to highlight its linguistic and functional characteristics on the basis of five textual dimensions. The corpus shows high levels of heterogeneity and information content. In particular, the average score of D1 shows that Pakistani editorials are highly informational as compared to British editorials. This is highlighted by the observation that British editorials contain moderate use of nouns, prepositions and attributive adjectives while Pakistani editorials make extensive use of these linguistic features to convey information. The D2 score shows that the Pak. Ed. corpus contains non-narrative content and the variation is similar to what is noted in Biber's measurements.15 The D3 score of the corpus of Pak. Ed. shows greater textual variation than Brit. Ed.
This is supported by the result that Pak. Ed. corpus seems to make extensive use of relative constructions for example, Pak. Ed. contain greater use of wh-relative constructions, nominalizations and phrasal coordination than Brit. Ed. These features show information, explicit and elaborated discourse which is typically found in academic prose. On the contrary, Brit. Ed. show situation dependent discourse and it seems to make use of time and place adverbs sufficiently to support the description of current events and situations. The D4 score of Pak. Ed. corpus shows that the most of the editorials lack the argumentative discourse while Brit. Ed. corpus shows high content of argumentation or persuasive expressions. This is surprising as it is expected for editorial writing to be persuasive. Typically, editorials state opinion, which are clearly stated with specific linguistic features, such as frequent modals (should, etc.), high concentration of conditionals.16
It is mentioned that Pak. Ed. either seems to be lacking the argumentative discourse or is using the covert way of argumentation and persuasion. High D5 scores show that the register of Pak. Ed. has a level of abstraction that uses the linguistic features of by-passive, agent-less passives, adverbial subordinates, conjuncts, past participles, WHIS deletions and predicative adjectives. This result shows an informative discourse that is formal and technical. On the other hand, Brit. Ed. seems to have low D5 score, which demonstrates resemblance with fiction and face to face conversation.
It is also noted that the Frontier Post's editorials are the closest to British editorials among all five Pakistani newspapers on the basis of five textual dimensions such as information content, narrative features, situation oriented references, extent of argumentative language and abstract style while the Dawn displays quite distant results with British editorials among all dimensions except in the narrative features. The present study highlights the usefulness of the MD analysis to characterize linguistic features of Pakistani newspaper editorials. Based on the results, a few extensions are proposed. The present research is based on the 88 MD analyses of already identified textual dimensions with the goal of comparison with British editorials. It is suggested that further application of MD analysis will identify new dimensions and variation in the linguistic features of Pakistani newspaper editorials.
It is expected that the further analysis will strengthen the claim that Pakistani editorials are the sub-register of Pakistani Newspaper writing with its own distinguished linguistic features. It will also qualify the argument that Pakistani English (PE) is evolving as an independent, non-native variety which stimulates further investigation.
Notes and References
1 Douglas Biber, Variation across Speech and Writing (London: Cambridge University Press, 1991).
3 Douglas Biber, "An analytical framework for register studies". In Douglas Biber and Edward Finegan, ed. Sociolinguistic Perspectives on Register. (London: Oxford University Press on Demand, 1994), 31-56.
4 Ghulam Shabir, Abdul Wajid Khan, Malik Adnan, and Ghulam Safdar, "A Comparative Analysis of the Editorials of the Nation and the News: The Case Study of Pak-India Relations Issues (2008-2010)" Journal of Political Studies 21, no. 1 (2014): 41-59.
5 Samreen Razi, "Editorials in the Leading English Press in Karachi: An Analysis of Their Representation", Academic Research International 5, no. 5 (2014): 236.
6 Bushra Hameedur Rahman and Abida Eijaz, "Pakistani Media as an Agent of Conflict or Conflict Resolution: A Case of Lal Masjid in Urdu and English Dailies", Pakistan Vision 15, no. 2 (2014): 238.
7 Muhammad Ashraf Khan and A. Safdar, "Image of US in Pakistani elite newspaper editorials after 9/11 incident: A comparative study of The Dawn and Nawa-i-Waqt with special regard to media conformity theory", Pakistan Journal of Social Sciences 30, no. 2 (2010): 325-339.
8 Ghulam Shabir, et.al, "A Comparative Analysis of the Editorials of the Nation and the News", op. cit.
9 Douglas Biber, Variation across Speech and Writing, op. cit.
10 Douglas Biber, "An analytical framework for register studies", op. cit.
11 Douglas Biber, Variation across Speech and Writing, loc. cit.
13 Douglas Biber and Susan Conrad Register, Genre, and Style (n.p.: Cambridge University Press, 2009).
14 Douglas Biber, Variation across Speech and Writing, loc. cit.
16 Douglas Biber and Susan Conrad Register, Genre, and Style op. cit.
|Printer friendly Cite/link Email Feedback|
|Author:||Alvi, Urooj; Mehmood, Muhammad Asim; Rasool, Shafqat|
|Date:||Sep 30, 2016|
|Previous Article:||Effects of Terrorism on Secondary School Students in Khyber Pakhtunkhwa.|
|Next Article:||Effect of Governmental Notification of Up-gradation on Job Satisfaction of Teachers of Children with Mental Retardation in Punjab.|