Has English Medium Instruction Failed in Pakistan?

Abstract

Drawing on a range of sources, this paper questions whether English medium instruction, as practiced in many schools across Pakistan, has failed. To explore this question, the paper examines data collected from a number of sources and finds that neither teachers’ English language proficiency nor their professional knowledge is sufficient for them to successfully help their students to be able to learn English or learn through English. The paper also suggests that there are gaps in the educational abilities of various ethno-linguistic groups in the country and that people from different backgrounds have varied attitudes and desires regarding the medium of instruction (MOI). The data shared in this paper indicate that the current National Education Policy, including the Draft National Curriculum 2020, which insists on the use of Urdu or English as the main MOI, has been unable to provide quality and equitable education to children across the country. These findings underscore the urgency of revising the National Education Policy, providing substantial training and support to current and future staff, and developing a well-researched and clearly articulated policy on the use of language in education.

1 Introduction

This paper questions whether EMI (English medium instruction), as practiced in many schools across Pakistan, has been a failure. The sources, which include data on the MOI (medium of instruction) preferences of the general population, English language proficiency scores of teachers and students, and government documents, indicate that the current language in education practices in Pakistan are not producing the results expected and may, in fact, be a cause for why students drop out of or fail in/through education.

EMI has been a consistent, albeit controversial, part of the educational environment in Pakistan, as in many “post-colonial” countries. English was first introduced to this part of the world when the British started trading with India under the guise of the British East India Company. The use and prestige of English grew from that point on. As the Mughal Empire – and the use of Persian as the language of arts, sciences and governance – was neutralised and India became part of the British Empire, English became integrated into the legal, educational, and other systems of the country. After independence, while designating Urdu as the sole national language, Pakistan maintained English as an official language and it has continued to play a key role in the economic, educational, social and political life of some of the people of the country (Mahboob and Jain 2016; Mahboob 2019).

The choice of maintaining English in Pakistan was both a pragmatic and a political decision. It was pragmatic because it was the language used in government and higher education before independence; the language had already been developed to function in these contexts; and people were already familiar with it in those contexts. And it was political because, in the absence of another local language that served all the functions that English did, selecting another language would (and did) potentially supress other languages and alienate speakers of those languages. Thus, English, partly for political, partly for pragmatic, and partly for a lack of will and effort has remained a prominent part of the educational context of Pakistan. Today, while there are numerous Urdu and some Sindhi and Pushto medium schools in the country, English medium schools tend to have more prestige; furthermore, when it comes to higher education, other than a handful of institutions, all adopt English as the medium of education.

It needs to be noted that while EMI, especially in the context of higher education, has been on the rise in many parts of the world (see, e.g. Doiz et al. 2013; Taguchi 2014), the situation of EMI in Pakistan is different. One of the main reasons for the recent surge in EMI in different parts of the world (and specially Europe) is the adoption of policies of globalisation and internationalisation of (higher) education (e.g. Bologna process in Europe). However, in the context of Pakistan, the presence of EMI is a result of historical processes rather than deliberate decisions to globalize or internationalise its education system (although these arguments are made in support of maintaining EMI). English has been a core part of the educational and government structures of the region since before the country was established.

Another key difference is grounded in the high linguistic diversity of Pakistan. Pakistan has a linguistic diversity of 0.802 on the Greenberg index (Lewis et al. 2016). This number, which is calculated based on the population of each language as a proportion of the total population, suggests that a large number of people do not share their first or heritage language. With over 70 ethno-linguistic groups and only a handful of them used in educational contexts, many feel that the use of a few selected languages pose challenges to literacy and educational development of their children (see, for example, Rahman 1996). This adds further fuel to political conflicts, many of which are grounded in the vast differences in the socio-economic conditions between various ethno-linguistic and regional groups of people across the country. In such a context, the local debate is not simply about how EMI can be improved, but rather what language(s) should be the MOI. The orientation of the recent research on EMI can, in contexts such as Pakistan, perhaps be counterproductive because it takes EMI as a given and does not engage with broader issues of MOI which are of concern to the local populations (see also Hamid’s 2013 review of Doiz et al. 2013). To avoid this limitation, this paper will discuss EMI issues in Pakistan within a broader MOI debate. In order to do this, I will first provide a brief overview of some of the issues around MOI in Pakistan before focussing on EMI issues in more detail.

2 Language in Education in Pakistan

In this section, we will first explore data on peoples’ preferences for MOI and then examine data that relate language heritage to education. This analysis will then be used to critically evaluate the National Education Policy (NEP). However, before we present and discuss the data or the NEP, we need to note that Pakistan does not have a documented language policy or language-in-education policy. This is not to say that there is no policy on language in Pakistan – there is; but there is no official document that specifically outlines and discusses the national language policy and its implications for education etc. (see also Mahboob and Jain 2016).

Since gaining its independence in 1947, Pakistan has followed a three-language approach: Urdu as the national language, English as the official language, and one language recognized for each province. This policy has also been adopted in education, where schools are either English medium, Urdu medium, or, in the case of some schools in Sindh and KP, use the provincial language as the MOI. According to the Ministry of Education (Pakistan 2009, p. 71), 68.3% of government schools use Urdu as the MOI; 15.5% educational institutions in Sindh use Sindhi as the MOI; 9.5% use other languages (Pushto, Balochi, Arabic etc.), and 10.4% use English as the MOI. While the report does not provide statistics about private schools, a majority of them tend to use English as the MOI, as will be noted later.

2.1 Preferences for MOI

ASER (Annual Status of Education Report), one of the few sources of educational statistics in Pakistan, provide some data relevant to the discussion on MOI in its 2012 report. Table 17.1, taken from ASER’s 2012 report, summarizes the results for participants’ preferred MOI, their home language, and the actual MOI in the schools in rural settings across various regions of Pakistan. Although there appear to be some gaps in the data included in Table 17.1, these results do suggest that a large proportion of those surveyed prefer Urdu or the home language as a MOI (the only majority support for English is in Islamabad Capital Territory). In Sindh (Sindhi), FATA (Pushto) and KP (Pushto), there is a strong preference for home language as MOI. These three regions are also the only ones where the home language is actually used as a MOI, even if in a small percentage of the schools surveyed. Khan (2015) and Rahman (1996) amongst others point out that Pushto and Sindhi speakers are very proud of their language, traditions and customs and therefore put more effort (including political pressure) in sustaining them in and through education.

Table 17.1 Preferred MOI, the home language, and the actual MOI

The data in Table 17.1 shows that there is an inverse trend in the choice of MOI between government and private schools. While a majority of government schools use Urdu as a MOI, a majority of private schools use English as the MOI. Dearden (2014) reports that private schools comprise almost 60% of all secondary schools in Pakistan. While Dearden’s report does not provide a distribution of these schools across the various regions of the country or across rural and urban areas, the figures do suggest that the use of EMI in private schools may be one factor that makes them attractive. If parents want to send their children to an English medium school, but the government schools offer only Urdu medium education, then parents may choose to send their kids to private EMI schools, if they have the means to do so.

The contrastive distribution of MOI across government and private schools, and a generally high preference for Urdu or home language as a MOI raises some interesting questions. For example, while 70% of those surveyed in Azad Jammu and Kashmir stated that they would prefer Urdu as the MOI, 68% of the private schools in the region use English as an MOI. There are many possible interpretations for such discrepancies. One possibility is that even though people may prefer one language as a MOI, they may choose to study in another because of its perceived socio-economic value.

The data from the ASER report cited here appears to be in contrast with other statistics on the preference for medium of education. For example, Mahboob (2002) reports (see Table 17.2), which corroborates findings by others such as Mansoor (2005), Irfan (2013) [see Mahboob 2017 for a comparison between these three studies], that 76% of the respondents in his study stated that English should be the MOI in primary schools; 94.4% stated that it should be the MOI in high schools and at the university level. In contrast, only 65.4% stated that Urdu should be the MOI in primary schools; 37% stated that it should be the MOI in high schools; and 26.5% stated that it should be the MOI at the university level. Of the informants who spoke a language other than Urdu as their first language, only 10% stated that their first language should be the MOI in primary schools, 4% stated that it should be the language of instruction in high schools, and none of the informants said that it should be the MOI in universities. The findings from Mahboob’s study provide a much stronger support for English in terms of preferences than the ASER report and a much weaker support for local languages. It also shows variation of attitudes towards the various languages based on the level of education, something that is missing in the ASER data. One reason for the large differences in the attitudes between the two surveys might be the source of data for these studies. While the ASER data shared here was collected in rural settings, Mahboob’s data was collected from students and faculty at a large public university in Karachi. The fact that participants in Mahboob’s study – as in most of the other surveys carried out at universities – had already made it to a university where English was the MOI may have also influenced their responses.

Table 17.2 Language and MOI preferences

The differences in the attitudes documented in the two studies above point towards a wider divide within the society. On the one hand, people who have gained status and power tend to know English and in turn support English (this includes bureaucrats, military officers, and other high level government officials who need English in order to rise to their positions); while, on the other hand, those who do not know English would like education – and success through education – to become accessible to them through Urdu or a local language as MOI. Since those who know and value English control policy decisions, they maintain policies that keep the position of English in place (as will be seen later). This perpetuates and reinforces the socio-economic status (SES) of the various communities and peoples. It is this reinforcement of the socio-economic class variations through education (and especially EMI) that has led researchers such as Khattak (2014), Rahman (2004), and Shamim (2011) to label the current educational system in Pakistan as linguistic and educational apartheid.

2.2 Language Background and Educational Poverty

Researchers, activists and policy consultants are aware of the problems of MOI in Pakistan and have argued that the current system results in unequal outcomes for students and that while students from urban middle-class settings do relatively well, children from minority, rural, and lower SES groups struggle in school. Students, especially females, those in rural settings, those from lower SES, and those whose mother tongues are not recognised in the educational system, face extreme difficulties in continuing their education. Recent data sourced from UNESCO and published in Coleman and Capstick (2012) (see Table 17.3 below) corroborate these observations.

Table 17.3 Education poverty in 17–22 year olds in Pakistan by mother tongue

Table 17.3, which only includes six major first languages, four of which (Balochi, Punjabi, Pushto, and Sindhi) have the status of ‘provincial’ languages, collected from the 17–22 year olds, lead to some very disturbing observations. The youth of Pakistan, if it does not belong to Urdu or Punjabi speaking background, has a very high probability of having no education at all, or of having spent less than 5 years in school. This includes the youth from Sindhi and Pushto speaking backgrounds, where privileged dialects of Sindhi and Pushto are used in education.

This high education poverty rate has serious consequences for the individuals concerned as well as for the economy and national development, and may result in increased crime, extremism and radicalization. In the absence of education or secure economic prospects, these large numbers of youth can become targets for criminal activity and radicalization, which can (and is) leading to national instability. The current events in Pakistan, including extremism, intolerance, violence and terrorism that are reported in the news everyday are possible outcomes of weak education in Pakistan.

While the data in Table 17.4 paint a dismal picture of educational achievement in Pakistan in general, they document how children from minority language backgrounds have a much higher ratio of educational poverty. While Table 17.4 does not provide data on gender, regional or economic disparities, the Global Education Monitoring Report website provides additional statistics on these and other disparities (but not on MOI or mother tongue). The Global Education Monitoring Report website shows that gender disparity in literacy skills is greatest in the poorest populations of the country and the least in the richest populations, regardless of the region (although there are variations across the regions). The regional disparity report shows that the highest literacy rates are in Punjab and the lowest in Balochistan and KP, with Sindh towards the lower end of the spectrum.

Table 17.4 Results from Pakistan

2.3 Pakistan National Education Policy (NEP)

In a context where the educational ‘apartheid’ is peaking and leading to national instability, it is surprising that questions about language in education policy, research into classroom practices, research on the communities and their educational experiences, are still not at top of national agenda and debate in Pakistan. There is little government action on these issues. And, instead of seriously considering alternatives such as mother-tongue based multilingual education policy (which has been adopted in many other countries, such as the Philippines, see, e.g. Cruz and Mahboob 2018) or plurilingual approaches (see, e.g. Canagarajah and Ashraf 2013), the Pakistan National Education Policy (NEP) fails to provide a vision or leadership and continues to push Urdu and English as the two main MOI.

While recognising the issue of education inequity in Sect. 3.5, Overcoming Structural Divides, the NEP continues to promote and reinforce the position of English. The underlying assumption in NEP is that structural divides can be overcome by giving all students access to English. Policy action 3, Sect. 3.5, of the NEP states:

Ministry of Education in consultation with Provincial and Area education departments, relevant professional bodies and the wider public, shall develop a comprehensive plan of action for implementing the English language policy in the shortest possible time, paying particular attention to disadvantaged groups and lagging behind regions [emphasis added]. (Pakistan 2009, p. 28)

In addition, policy action 4–8 state:

• 4. The curriculum from Class I onward shall include English (as a subject), Urdu, one regional language, mathematics along with an integrated subject.

• 5. The Provincial and Area Education Departments shall have the choice to select the medium of instruction up to Class V.

• 6. English shall be employed as the medium of instruction for sciences and mathematics from class IV onwards.

• 7. For 5 years Provinces shall have the option to teach mathematics and science in English or Urdu/official regional language, but after 5 years the teaching of these subjects shall be in English only.

• 8. Opportunities shall be provided to children from low socio-economic strata to learn English language. (Pakistan 2009, p. 28)

The policy promotes the adoption of English, first as a subject, then as a MOI for mathematics and science. It also states that English will be the mandatory language for teaching science and mathematics across all grade levels after 2014. The policy gives some recognition to ‘official regional languages’, but none to non-official regional languages. Some (e.g. Habib 2013) argue that these policy decisions are based on parents’ demands and the assumption that students need to learn English and learn about science and mathematics through English because English is the language of knowledge-production in these fields. However, initial reports from the field are quite negative. For example, Bari (2013), writing in the Daily Dawn, notes that teachers are used to teaching in Urdu or a local language and are unable to teach their subjects in English, “let alone helping students learn English”.

The Coleman and Capstick (2012) report along with a number of other academics (e.g., Khattak 2014; Manan et al. 2015; Khan 2015; Rahman 2010) have also questioned these policies and argue that such policies may further disadvantage students from non-elite backgrounds because they do not have appropriate English language skills (or, for that matter, Urdu language skills) to study mathematics and science through the medium of English or Urdu. Regardless of the criticisms, the government is slow to make any changes. One reason for this may be that the government officials, bureaucrats and other high-powered entities believe that English can give access to others in the same way as it gives them a position of power. On the other hand, people who are struggling feel that English is keeping them away from access to education and socio-economic stability and thus state that they want access to education in Urdu or a local language, while still sending their children to non-elite private EMI schools (if they can afford it) in the hope that their children can learn English and do better (as noted in our discussion of the ASER 2012 report earlier). These beliefs and practices contribute to the maintenance of English language hegemony and perpetuate socio-economic class variations.

Based on the data and policies reviewed above, it appears that the NEP and current government policies reflect a gap in understanding of issues between what experts and policy makers believe should happen (or what they would like to happen) and what is actually happening in reality. The problems in the conceptualisation and implementation of educational policies and a lack of a clearly formulated language-in-education policy have created a situation where one can ask if EMI educational institutions fail their students, parents, communities and the country? We will look into this in more detail in the next section.

3 Evaluating the Success of EMI in Pakistan

In the previous section, we considered how the MOI issues in Pakistan have resulted in unequal distribution of access to education across various ethno-linguistic and regional communities, which may impact the country’s economic and political stability. Regardless of the issues with current practices, the government (via the NEP) continues to push for Urdu and English MOI. We will now focus our attention to issues of EMI in the context of schools and will first look at data that document teachers’ language and professional knowledge and then data that evaluate students’ language abilities. The results from this analysis show that a large number of teachers (including English language teachers) have low English language proficiency and have very low declarative knowledge about their profession. In addition, reports indicate that students’ English language abilities are also quite weak. These results further lead to the question of whether EMI in Pakistan has failed.

3.1 Teachers’ Language Proficiency and Professional Knowledge

In a TIRF funded project, Kamhi-Stein and Mahboob (2011) gathered data on teachers’ language proficiency and their professional knowledge about teaching and learning from Pakistan, Argentina, and South Korea. The data for this project was collected using a battery of tests and observations, including a Declarative Knowledge Test (which included items on both language and professional knowledge), Cambridge Quick Place Test, an essay task, and a summary writing task. In addition, teachers were asked questions about their beliefs about language use in class and were also observed teaching. In Pakistan, the data was collected from English language teachers in three schools (two teachers in each school): one government Urdu-medium school (UMI), one (non-elite) private English-medium school, and one private elite-English medium school. Some of the data from Pakistan is given in Table 17.4 below and the averages for the three countries in the study are given in Table 17.5.

Table 17.5 Average scores for the three countries

The results from Pakistan showed that there were differences between teachers’ language proficiency between the elite EMI schools and the other two schools; and between the UMI school and non-elite EMI school. The teachers in the elite EMI had the strongest language skills amongst the three schools and the teachers in the UMI school had better English language proficiency than the non-elite EMI school. Furthermore, one of the two teachers in the UMI school had noticeably higher language proficiency than the other. Of the six teachers in the study, only two (both from the elite EMI school) used mostly English in their classes. The English teachers in the non-elite EMI school were observed using the least English in their classes.

In addition, Table 17.5 shows that the average scores for Pakistan were markedly lower than those for Argentina and South Korea. And that even the most proficient of teachers in the elite EMI school had lower language scores than the average scores of the two other countries. This was a surprising finding as English has a long history in Pakistan and is an official language of the country; the other two countries in the study belong to the Kacrhuvian expanding circle and started using English in educational contexts relatively recently. However, both South Korea and Argentina do have stronger economies than Pakistan and invest much more of their GNP on education in their local languages than Pakistan. A 2010 UNESCO fact sheet states that Pakistan has some of the worst education indicators in the world and that the funding on education was further reduced from 2.6% of GDP in 1999 to 2.3% of GDP in 2010 and currently sits at around 2.9% of GDP. This low investment in education may be one reason for Pakistan’s poor performance in education.

Table 17.4 also shows that none of the six teachers observed were able to secure even 50% in the pedagogy part of the DKT; and two of them got only one item right on the test. While the elite-EMI teachers did better than the other two schools, their scores were again considerably lower than the average scores for the two other countries in the study. One of the teachers at the UMI school also scored as high as the teachers in the elite EMI school. This set of findings is just as alarming as the one for language proficiency. Of the participants in this study, three had professional qualifications: Butt had an M.Ed.; Ghaus had a B.Ed.; and Fatima held a diploma in teaching. However, regardless of these qualifications and years of professional experience, none of these teachers scored over 50% in the DKT-pedagogy; this result, in turn, questions the quality of the professional preparation of these teachers.

To study the relationship between language proficiency and professional preparation, Kamhi-Stein and Mahboob (2011) adopted Pasternak and Bailey’s (2004) framework. Central to their framework are two notions. First, language proficiency and professional development need to be perceived as continua. Second, “there are different degrees of proficiency: being proficient is a continuum, rather than an either-or proposition” (Pasternak and Bailey 2004, p. 163). Figure 17.1 below presents the framework.

Fig. 17.1

Pasternak and Bailey’s continua of target language proficiency and professional preparation

As explained in Pasternak and Bailey (2004), Quadrant 1 reflects those teachers who are both proficient in the TL and professionally prepared. Quadrant 4 reflects those teachers who are not proficient or professionally prepared. Quadrant 2 reflects teachers who are professionally prepared but are not proficient in the TL and Quadrant 3 reflects teachers who are proficient in the TL but are not professionally prepared.

The results from the Pakistani study showed that all six of the teachers fall in quadrant 4 – they are neither proficient in the language nor do they have appropriate professional knowledge. These findings, from English teachers, reflect a dismal state of education across the board – including in at least some of the elite EMI schools. One question that emerges here is: with such low language proficiency and poor professional knowledge, how successful can these teachers be in helping their students develop sufficient English language proficiency to study all the other subjects through English?

A more recent study, the PEELI Report (British Council 2013), corroborate the findings reported above and extends them to look at not just English teachers, but teachers of all subjects. The PEELI Report entitled ‘Can English Medium Education Work in Pakistan: Lessons from Punjab’ is an evaluation of the impact of the Punjab Government’s implementation of the NEP and teaching mathematics and science through English in all schools from grade 1 onwards. In carrying out this review, the British Council collected language proficiency data from 1720 teachers in the province of Punjab using their Aptis test. The results of the test are presented using CEFR levels. Table 17.6 below summarises their findings of the differences between private school and government school teachers. It needs to be noted that the UNESCO reports, discussed in an earlier section, document that Punjab has the strongest education indicators of the four provinces in Pakistan. Thus, based on the results and discussion below, we can argue that the situation is perhaps even worse in the other three provinces.

Table 17.6 Teachers’ language proficiency based on CEFR levels in government and private schools

Table 17.6 shows that more than 56% of all teachers surveyed have no English language competency at all; of these, a higher proportion of teachers in private schools have no English language proficiency as compared to government schools. On the other end, only 2% of the government and 3% of the private school teachers have B1 (intermediate) or above level of language proficiency. These results provide further support to the Kamhi-Stein and Mahboob’s (2011) study and shows that school teachers in Pakistan generally have very low English language proficiency.

In addition, the PEELI Report also provides a summary of findings based on the MOI that teachers work in. Some of these findings are included in Table 17.7 below which shows that while the highest percentage of teachers with B1 or higher English language proficiency teach in EMI schools, they only comprise 6% of teachers in these schools. 44% of teachers in EMI schools have no proficiency in the English language at all and 50% have basic proficiency (A1 & A2). The state of things in other schools, including bi/multilingual schools is still worse; with no teachers in mother tongue schools surveyed reaching a pre-intermediate level of English language proficiency (B1).

Table 17.7 Teachers’ language proficiency based on CEFR levels across various MOI schools

3.2 Evaluation of Students’ Language Abilities

So far we have looked at two studies that report on teachers’ language proficiency. All data reviewed shows that teachers (including English language teachers) have low English language proficiency and weak professional knowledge. A separate data source, the ASER Report (2013), provides us snippets into students’ English language ability. Figure 17.2a provides a summary of students’ English language ability across the rural parts of the country and Fig. 17.2b provides comparable statistics for the urban parts of the country. Both these figures provide comparative data over a period of 3 years (2011–2013) and use the same criterion for assessment: “Ask her/him to read the 4 sentences. If s/he reads all 4 correctly, then mark her/him at the ‘sentence level’” (ASER 2013, p. 41). ‘Sentence level’ is the highest level of English language ability as measured by ASER and corresponds to Grade 1 & 2 Competencies, Standards, and Benchmarks of the National Curriculum for English Language (2006). While this criterion may not meet the requirements of standardized testing, the data does provide a general impression of students’ language ability. Before we look at the data, we need to note one other thing: the figures provided here are based on cross-sectional data and therefore do not show a language development trajectory.

Fig. 17.2a

Students’ English language ability in rural areas. (ASER Report 2013)

Fig. 17.2b

Students’ English language ability in urban areas. (ASER Report 2013)

The data included in Figs. 17.2a and 17.2b help us make a few observations. First, in general, there is little change in any one class across the 3 years reported here. This suggests that there is little cohort effect on these statistics, with one exception. A slightly larger number of students in the urban areas in 2011 seem to have been able to read four sentences correctly than those surveyed in 2012 or 2013. The ASER Report does not provide any additional data or explanation for this slight anomaly.

The two figures also show that a slightly higher number of students in urban areas are able to meet the test target in each grade level. However, what is perhaps more revealing is that by grade 6 the difference between rural and urban settings is greatly reduced. The 2013 figures show that 13% more of grade 3 students in urban settings were able to read four sentences than those in grade 3 students in rural settings. This gap widened to 16% in grade 4 and 5, but then drastically reduced to only 6% in grade 6. Again, the ASER Report does not provide any additional data or explanation for these changes. If these figures are to be relied upon, they suggest that while students come into the rural and urban school system with different ability levels, this difference is minimized within a few years. Further investigation into this may provide useful insights that can help improve the quality of schooling across both rural and urban settings.

This low student performance on one basic matrix in the ASER report across urban and rural areas over 3 years poses serious questions about the success of English language education, including EMI, in the country. The results, unfortunately, suggest that the current schooling system is not working. As these students continue their educational journeys and scrape through to higher education (if they are able to somehow manage that), they will face enormous troubles there, as most of the institutions of higher education use EMI (see Mahboob 2017).

4 Concluding Remarks

The data presented in this paper from a number of sources to indicate that neither the current level of English language proficiency of the teachers nor their professional knowledge is sufficient for them to successfully help their students to be able to learn English or learn through English. Other data presented here shows that there are gaps in the educational abilities of various ethno-linguistic groups and that people from different backgrounds have varied attitudes and desires regarding the MOI. The analysis of this data show that the current NEP, which insists on the use of Urdu or English as the main MOI, has not fully succeeded in providing quality and equitable education to children across the country. These findings underscore the urgency of revising the NEP, providing substantial training and support to current and future staff, and developing a well-researched and clearly articulated policy on the use and support of local languages in education.

This paper provides evidence that shows just how problematic things really are. A lack of will or action by the government will continue to divide the people of the country and increase the gap between a majority of semi- or illiterate people with poor economic prospects and a minority of rich and powerful people with access to world-class resources and education. These divisions can give further opportunity to extremist and criminal elements to attract the poorly educated and ill/semi-literate people to fundamentalist, radicalised and criminalised positions and endanger the state.

Finally, this paper documents that in the context of Pakistan, and perhaps other post-colonial nations, debates and discussions about EMI need to be couched within broader discussions of MOI. Without doing so, we may be inadvertently supporting positions and policies that support EMI at the cost of local languages and the people who speak these languages.