A Multidimensional Analysis of Pakistani Legal English

The present study aims to investigate linguistic variation among genres of Pakistani Legal English by applying multidimensional analysis. In a legal context, language performs different functions. This results in a variety of textual categories on the basis of purpose of communication and linguistic properties. In order to recognize the linguistic properties of any individual genre, a comparative study of genre categories is essential. The study has been conducted on the sample of eight Pakistani Legal genres based on around two million words. Data have been analyzed by applying Biber’s (1988) model of Multidimensional analysis. Findings reveal variation in linguistic patterns. All categories have been found significantly different along each dimension. It indicates that legal language is not a homogeneous phenomenon. It has a variety of linguistic features associated with different legal genres, so it must be viewed in terms of goal, purpose, audience and context (variable which affect the language choice.).


Introduction
Language, as a source of communication, is influenced by place, time, subject matter as well as situation of communication.It leads to variety of language choices and styles which provide a constant basis for research into different fields of language use i.e. medical, law etc.Variety of language activities are performed in legal contexts, for example, statutes, contracts between social members, appeals etc. Diversification of legal activities results into linguistic diversity.
Legal language which consists of a variety of legal discourses built around multiple written and oral genres, is classified by Danet (1985).This classification is based on two criteria: (1) the means of language use i.e. written or spoken.and (2) the level of formality, for example frozen , formal , consultative and casual etc. Documents like wills and contracts are in frozen written style, while statutes and briefs are examples of formal written style.Frozen spoken includes witnesses' oaths and marriage ceremonies whereas lawyer's questioning in court from witness is a type of formal spoken language.So far as casual spoken style is concerned, it involves lobby conferences and conversation between lawyers.
Variation in Legal language depends not only on the mode of language use (speech or writing), situation in which language is produced, participants, relationship between the participants, the purpose of communication, but also in its language patterns which can be comprehended if different genre categories are studied in comparison with reference to language characteristics.Though legal discourse is a comprehensive term which involves a variety of legal discourses in both written and oral mode resulting into different text types (genre) (Maley, 1994;Trosborg, 1995 andMattila, 2006), but previous studies on legal English in Pakistani context have dealt it as a homogenous phenomenon (Khan & Khan, 2015;Ahmad, 2005).Register studies have emphasized that distribution of linguistic features is different across genres/registers.So, comparative analysis of these textual categories is required to comprehend linguistic properties of different genres.As no study in Pakistani legal context has been conducted in this regard, this study aims to analyze linguistic variation across legal genres in Pakistani context through multidimensional analysis.

Research Question 1).
To what extent do textual categories in Pakistani legal English differ across textual dimensions of 1988 model of variation?

Significance of the Study
As the study is aimed to identify linguistic variation among genres of Pakistani legal English, it will provide a comprehensive picture of Pakistani Legal English.The research will contribute to existing research on language variation in and across culture.The study will also be a significant addition in existing corpus based studies.Researchers working on discourse analysis may get help from the findings of the study.It will provide methodological framework for researchers.It will be a reference point for comparison with any special or nonspecial registers on the basis of results of 1988 MD.The findings of the study will be beneficial for ESP teachers to help them focusing on linguistic forms and their functions in particular genre.The findings of the study will be helpful for syllabus designers in order to decide content while dwelling upon linguistic features and their functions identified in this work.
The current research will prove to be of significant value to different researchers working on the registers of Pakistani English and the people working in establishing the identity of Pakistani English as a distinct non-native variety with its independent registers.Besides, it will be helpful for researchers conducting cross linguistic studies.

Literature Review
Legal language involves the type of language used in a legal setting.Tiersma (1999) defined legal language, "as the distinct manner of speaking and writing that has been developed by just about any legal system throughout the world".Danet (1985) asserted, "While language is central to all human affairs, it is particularly critical in the law.Physicians work with physical substances and entities; in contrast, the work of lawyers and judge is symbolic and abstract.In a most basic sense, law would not exist without language" (p.273).
Legal English involves diverse oral and written genres based on their purpose of communication, the situation where they are used, participants, their relationships and the background knowledge of those who are engaged in activity of language use (Bhatia, 1987).Tiersma (1999) acknowledged variation in legal language in his book "Legal Language" in these words, "It should be evident by now that there is great variation in legal language, depending on geographical location, degree of formality, speaking versus writing, and related factors.The language and style of lawyers also differ substantially from one genre of writing to another".So, legal language is divided into different genres depending upon its use in different legal contexts i.e. different legislations, judgements, contracts, articles, textbooks, ordinances, appeals, wills, statutes, lawyers' counseling, witness examination, law statements produced by media reporters.This long list points out the multiplicity of legal discourse.Each category has distinctive terminology.Maley (1994) described that "There is not one legal discourse but a set of related legal discourses.Each has a characteristic flavour but each differs according to the situation in which it is used" (p.13).Having acknowledged that legal discourse is manipulated in different legal contexts, Maley (1994) proposed different categories such as "judicial discourse, courtroom discourse, the language of legal documents (contracts, regulations, deeds, wills, statutes) the discourse of legal consultation" (p.13).Tiersma (1999) asserted this idea by saying: Clearly, legal language is not monolithic.Even if we limit ourselves to the written variety, there is substantial variation among different genres of documents.Generally speaking, operative documents have by far the most legalese, as compared to persuasive and expository documents (p.141).
In spite of the fact that legal language is diversified, many of the researchers have dealt it as monolithic phenomenon.A lot of work done in legal discourse focused on different facet of legal textuality (Palmer & Pearce, 1983;Danet, 1983;Maley, 1994;Trosborg, 1995;Tiersma, 1999).Bhatia (1993) identified structural and lexico-grammatical features in different genres of legal discourse.Lundquist (1995) studied the functions of indefinite noun phrases in different legal categories i.e. textbooks, judgments, laws and legal articles.This study highlighted characteristics of different textual categories in regards to the use of indefinite noun phrases.Linguistic and pragmatic properties in English legal discourse were examined by Gotti (2001Gotti ( , 2008) ) whereas areas of semiotics, pragmatics and forensic linguistics were investigated by Jackson (1995), Trosborg (1997) and Gibbons (2004).
These studies highlighted the importance of language variations reflected in different genre categories where law and language interface but their focus is single linguistic feature.Single linguistic feature does not help to understand any distinct variety in depth.Detailed picture of co-occurrence of linguistic patterns manifested in any category is essential.For this purpose, multidimensional analysis is a comprehensive methodology.Brown and Fraser (1979) asserted the importance of co-occurring linguistic features in following words: It is often difficult or even misleading to concentrate on specific, isolated [linguistic] markers without taking into account systematic variations which involve co-occurrence of sets of markers.A reasonable assumption is that socially significant linguistic variations occur normally as varieties or styles, and it is on those varieties that we should focus.(pp. 38-39) In previous literature, two studies in American and Turkish contexts (i.e.Patterns of Linguistic Vaiation in Americann English by Roszkowski (2011) and A Comparative Register Perspective on Turkish Legislative Languag by Ozyildirim ( 2011)) focussed on linguistic variation in legal genres by applying multidimensional analysis.They identified markers of language use and language structures that vary from those found in other communication situations.So far as Pakistan is concerned, register studies other than legal register in English were conducted using Multidimensional analysis, for example a researcher (Shakir, 2013) identified linguistic variation across print advertisements in Pakistani media.In a study, linguistic variations across press reportage in Pakistani print media were investigated (Ahmad, 2015).In another study (Ali, 2016), linguistic patterns in Pakistani English Fiction were analyzed.Alvi (2016) studied linguistic variations in Pakistani press Editorials.But no study has yet been conducted to explain the way genres of Pakistani legal English are distinguished from one another based on sets of co-occurring linguistic features.So the present research focuses this aspect of Pakistani Legal English.

Research Methodology
This section describes the research methods applied to conduct this study.It includes nature of study, corpus design and data analysis procedure.This study is a blend of quantitative and qualitative methods.It provides a comprehensive linguistic description of Pakistani legal genres and calculates linguistic dimensions on the basis of co-occurring linguistic features.

Corpus
Corpus of Pakistani legal English has been chosen for this study.It covers the period of 2007-2014 and includes texts or categories of legal genres.The collection of texts comprises eight legal genres i.e. constitutions, directives, acts, articles, legal decisions, ordinances, legal reports and rules and regulations.The word range contains around two million words.The Pakistani Law Corpus has been borrowed from department of English linguistics Government College University Faisalabad.The legal genres related to written activities in Pakistani legal culture have been focused.All of them have been selected depending upon significance, accessibility and readerships.

Data Analysis
The current study examines differences and similarities among Pakistani legal genres by applying model of MD analysis proposed by Biber (1988) in his work "Variation across Speech and Writing".This model analyzes new discourse categories with respect to formerly determined dimensions (Biber, 1988) and is generally known as the 1988 MD model.It includes five dimensions i.e. "Involved vs. Informational Production", "Narrative vs. Non-narrative Concerns", "Explicit vs. Situation Dependent Reference", "Overt Expression of Persuasion", "Abstract vs. Non-Abstract Information".This model of variation is based on 67 linguistic variables classified into 16 main categories on the basis of grammar and function (Detail of the linguistic features underlying these categories can be seen in Biber (1988, pp. 223-245).
Analysis is based on following steps:

Tagging Lexical and Syntactic Features (Computational Identification of Linguistic Features) and Computing Frequencies of Linguistic Features
Data have been tagged through Biber's tagger.All linguistic features related to textual diemsnions of 1988 model of variation have been identified through this tagger.After tagging, frequencies of these features have been computed.

Normalizing Frequencies
Next to the computation of raw frequencies of linguistic features, is the step of normalizing frequencies.Frequencies of linguistic features have been normalized out of 1000 words in order to avoid error due to varying    The State Pattis and whether th provided.

Analys
In order to applied.Le  1 presents the results of comparison of legal genres (Constitutions, Directives, Acts, Articles, Legal Decisions, Ordinance, Legal Reports, Rules and Regulations) along dimension one: Involved vs. Informational Production.The results reveal that there is a highly significant difference (F= 14.95, P<0.01) among categories on this dimension.Though all of them are informational (as is indicated in ch.4 section 1), yet there is a difference in density of information packed in these categories.Note.NS = Non-significant (P>0.05);* = Significant (P<0.05);** = highly significant (P<0.01).
Table 2 shows the result of comparison among legal genres along dimension 2 labelled as "Narrative vs. Non-Narrative Concerns".Results reveal that all legal genres (Constitutions, Directives, Acts, Articles, Legal Decisions, Ordinance, Legal Reports, Rules and Regulations) significantly differ from one another (F=76.80,P<0.01)).The significant F score indicates that means of categories are not equal and there is greater variation among groups than within groups.Note.NS = Non-significant (P>0.05);* = Significant (P<0.05);** highly significant (P<0.01).
Table 4 displays the results of ANOVA, which has been conducted to find out differences among legal genres (Constitutions, Directives, Acts, Articles, Legal Decisions, Ordinance, Legal Reports, Rules and Regulations) along dimension four labelled as "Overt Expression of Persuasion".The results reveal highly significant variability (F=51.70,P<0.01) among legal genres along this dimension.As the F score is high, it shows that difference between legal genres is greater than within genres.Note.NS = Non-significant (P>0.05);* = Significant (P<0.05);** = highly significant (P<0.01).
Table 5 shows the comparison of legal genres (Constitutions, Directives, Acts, Articles, Legal Decisions, Ordinance, Legal Reports, Rules and Regulations) along dimension 5 labeled as "Impersonal vs. Non-impersonal Style".The results show statistically highly significant difference (F=41.18,P<0.01) among legal genres.The following section presents multiple comparisons among legal genres to indicate the point of difference in them.

Multiple Comparisons
ANOVA results have identified statistically significant difference among categories of legal genres along each dimension, but have not pointed out the differences among genres on an individual basis.For this purpose, Multiple Comparisons have been applied.
Following table demonstrates the difference in legal genres on an individual basis.Table 6 shows comparison of each legal genre with other legal genres on individual basis.It reveals that constitutions are not significantly different from articles, legal decisions, ordinances, legal reports and rules and regulations whereas they vary significantly from other legal genres i.e. directives and acts.
Directives are found statistically significantly different from all other legal genres.In case of acts, significant difference has been observed among this genre and constitutions, directives, legal decisions, legal reports and rules and regulations.Besides this, articles have been identified statistically significantly different from other genres i.e. directives, legal decisions, legal reports and rules and regulations.Legal decisions are significantly different with directives, acts and articles.Ordinances' difference with directives and legal reports is statistically significant.As regards legal reports, they are significantly different from directives, acts, articles and ordinances whereas genre of rules and regulations varies significantly from directives, acts and articles.Table 7 indicates that constitutions, acts and articles are significantly different from directives and legal decisions whereas their difference with each other and other legal genres is statistically non-significant.Directives and legal decisions' means do not share similar letters with other legal genres which indicates statistically non-dignificant difference among them and other legal genres i.e. constitutions, acts, articles, ordinances, legal reports and rules and regulations.Articles have been found significantly different from directives, legal decisions, ordinances and rules and regulations.So far as ordinances and rules and regulations are concerned, they reveal significant difference from directives, articles and legal decisions.Table 8 demonstrates the results of multiple comparison on dimension 3. Constitutions differ significantly from acts, ordinances and rules and regulations whereas directives vary significantly from acts, articles, ordinances, legal reports and rules and regulations.Acts and rules and regulations have been noted different from constitutions, directives, articles, legal decisions, ordinances and legal reports and this difference is statistically significant.Articles and legal reports are statistically significantly different from directives, acts, legal decisions, ordinances and rules and regulations.In case of legal decisions, significant difference has been shown between this genre and other legal genres i.e. acts, articles, ordinances, legal reports and rules and regulations.Ordinances have also been found significantly different from constitutions, directives, acts, articles, legal decisions, legal reports and rules and regulations.Results of multiples comparison on dimension 4 exhibit significant difference of constitutions and ordinances with legal decisions and rules and regulations.Directives, articles and legal reports are significantly different from acts, legal decisions and rules and regulations.Acts vary significantly from directives, articles, legal decisions, legal reports whereas legal decisions' difference with constitutions, directives, acts, articles, ordinances, legal reports and rules and regulations is statistically significant.A significant variation has also been observed among rules and regulations and constitutions, directives, articles, legal decisions, ordinances and legal reports.Table 10 presents results of multiple comparison on dimension 5.It shows significant difference among constitutions and two other legal genres (directive and legal decisions).Directives and legal decisions have been found significantly different from all other legal genres mentioned above.Acts and rules and regulations have also been noted different from directives, articles and legal decisions and this difference is statistically significant.As regards articles, significant variation has been observed among this genre and sirectives, acts, legal decisions, legal reports and rules and regulations.Ordinances differ significantly from legal reports, legal decisions and directives while legal reports have shown significant difference from directives, articles, legal decisions and ordinances.
Tables 6 to 10 have illustrated the results of multiple comparisons of legal genres along five dimensions.The comparison shows that the most of the legal genres vary from one another along each dimension which supports the idea of variation within register.
It shows that legal discourse is not a monolithic phenomenon; differences among legal genres depend upon their context of use.

Conclusions
The findings of the study lead to following conclusions: From the comparison of legal genres along 1988 model of variation, it turns out that all genres are informational though significant variation in informational level exists.Almost all of them are non-narrative.Only directives and legal decisions have least concern for narrative orientations.All of them have the tendency towards elaboration and explicitness and have a little tendency towards situation dependent discourse.Besides, most of them are marked with least persuasion and all of them are impersonal and objective in the information they impart.However, a considerable variation is found among legal genres in the degree to which they have these characteristics.The distribution of the mean scores of legal genres on each dimension shows remarkable differences within legal discourse.In point of fact, regular patterns can be observed in dimension1, 3 and dimension5 where all legal genres fall on the same pole either positive or negative.Dimension 2 and dimension4 split legal genres in both of the poles.On dimension 2, directives have positive scores, but scores of legal decisions have around zero mean score and all other genres have negative scores thus non-narrative.Dimension 4 shows, acts and rules and regulations clearly on the positive pole while other genres have negative weight.This variation among legal genres is described by Mattila (2006), there are different subgenres within legal language and each has particular characteristics.This indicates different purpose of their communication and conforms to the notion of register studies that lexico-grammatical features are distributed differently in different genres or textual categories.Reason for this difference is stated by Lemke (1995), with the difference in the field of activity, frequencies of lexico-grammatical patterns change which indicate changes in communicative purpose.
This study concludes that term legal discourse is a misnomer.In fact, there are legal discourses found across different legal genres.

Table 2 .
Comparison of legal genres along dimensions 2 of 88 model of variation

Table 3 .
Comparison of legal genres along dimensions 3 of 88 model of variation

Table 4 .
Comparison of legal genres along dimensions 4 of 88 model of variation

Table 5 .
Comparison of Legal Genres along Dimensions 5 of 88 Model of Variation

Table 6 .
Comparison of Mean±SE along dimension 1 Means sharing similar letters are statistically non-significant (P>0.05).

Table 7 .
Comparison of Mean±SE along dimension 2

Table 8 .
Comparison of Mean±SE along dimension 3

Table 9 .
Comparison of Mean±SE along dimension 4

Table 10 .
Comparison of Mean±SE along dimension 5 Means sharing similar letters are statistically non-significant (P>0.05).