Welcome Reader!

Lahore University of Management Sciences is Pakistan's most prominent private sector research university. Every quarter, LUMS students produce hunderds if not thousands of research papers. Many of those papers are well-researched and well-written; some are rare unadmired gems.

The LUMS STUDENT JOURNAL aims to bring together all those research papers. Soon you will see just how massive the students' research output is!

The internet edition publishes each and every paper that it receives. However, we hope that once we start receiving a good number of high-quality papers, we will also publish a print-based quarterly edition which would be peer-reviewed.

Defining Ethnicity: An Infinitely Malleable and Paradoxical Concept


Course: Anthropological Perspectives on Ethnicity and Nationalism

Spring 2007


Umer Gilani





In the introduction to his famous work, Barth remarks, in a tone that sounds rather smug: "I found it possible, by explaining the circumstance in the north, to make Southern Pathans agree that these( Pathans in Swat) were indeed Pathan too…"[1] Barth might have been pleasantly surprised at this feat of his, but another scholar of ethnicity was in for a bigger shock when she conducted the following experiment on her students, at a small private university in Pakistan.


Having exposed them to the definitional frameworks concerning ethnicity, presented by eminent theorists like, Smith and Barth, she asked them to consider the question of whether members of their university constituted a separate ethnic group. To her surprise, she found that all students – around 30 of them – argued fairly consistently that on the definitions provided, their campus community called 'Lumunites' had as much right to be considered an ethnic group as other more recognized ethnic groups. They conceded that in some respects Lumunites did fall short of the criteria set for constituting a separate ethnic group, but then so did most other human collectivities generally considered distinct ethnic groups.


As for the shocked scholar of ethnicity, she later claimed that this activity had been planned to expose a problem inherent in the traditional concept of ethnicity. Common sense dictates that a university is not an ethnic group; indeed, being small, modern and functional, it is far from it. And yet, the problem is that if someone tries earnestly and diligently enough, just about any other collectivity, even a university, can be nomologically fitted under the label of 'ethnic group'. The instructor's ex post facto remark might have been a cover up.





Regardless of intentions behind it, this experiment badly exposed the severe short-comings of ethnicity as a conceptual tool and an ontological category. This paper argues that the experiment described above can be used to demonstrate the poverty of ethnicity as a concept; also this experiment exposes a paradox inherent to the concept of ethnicity: even though the concept stems directly out of common sense, it ends up defying it. The implications for ethnicity, therefore, are adverse. If the argument presented in this paper are accepted, it would only be natural to conclude that the use of the concept of ethnicity, even as functional term, should be discontinued in academic discourse, because this infinitely malleable concept 'obfuscates more than it illuminates'.


The paper begins by presenting the concept of ethnicity as defined by various eminent theorist of the field, selecting Smith's definition from amongst them. Special attention is paid to process whereby Smith arrives at his definition. Then, characteristics of the community of Lumunites are described and the question of whether, under the criteria laid down earlier, Lumunites are eligible to be considered an ethnic group, is considered. Subsequently, the same question is explored with regards to Pakhtuns, a generally recognized ethnic group. Finally, the findings of these two lines of enquiry are juxtaposed, whereby the contradictions inherent to the concept of ethnicity are exposed. It is argued that the ramifications severely damage the legitimacy of the concept of ethnicity, which should, therefore, give way to most robust analytical categories.


What makes an Ethnic Group?


Various theorists have come up with definitions of ethnic groups. A review of the literature available on this question shows that, apparently, in recent years, the question that is troubling the mind of the author has been troubling others too. Consequently, there is burgeoning literature where modified and re-modified definitions of ethnic groups are being presented.


Gil-White(2005)[2] presents a sampling of that literature: 'We shall call 'ethnic groups' those human groups that entertain a subjective belief in their common descent—because of similarities of physical type or of customs or both, or because of memories of colonization and emigration—and in such a way that this belief is important for the continuation of non-kinship communal relationships . . . regardless of whether an objective blood relationship exists or not.[3]


'The term ethnic group is generally understood in the anthropological literature (cf. e.g. Narroll 1964) to designate a population which: (a) is largely biologically self-perpetuating (b) shares fundamental cultural values . . . ; (c) has a membership

which identifies itself . . . ' [4]


'An ethnic group is a self-perceived inclusion of those who hold in common a set of traditions not shared by others with whom they are in contact. Such traditions typically include "folk" religious beliefs and practices, language, and common ancestry or place of origin . . .which includes some concept of an unbroken biological-genetic generational continuity. . . Endogamy is usual . . . '[5]


 '. . . should members subjectively assume the existence of such 'mythical' [primordial] bases, the salient condition of ethnicity is met'[6]


'The term bangsa in Malay is the equivalent to our "ethnicity". It conveys the double ideas of people sharing both a common origin and a common culture. Etymologically it is derived from the Sanskrit vamsa, 'line of descent'. Emically, it has a primordial quality,

for it implies that the cultural traits are inalienably and inextricably associated

with a particular people, that is, carried by a community whose ultimate

unity derives from a single origin. (. . . ) While the concept of bangsa . . . has overtones

of shared culture, this is secondary . . . to the solidarity acquired though

common descent or origin . . . '[7]


'The reference to origin is . . . the primary source of ethnicity which makes a socio-cultural boundary into an ethnic boundary. . . . ethnic identity can best be defined as a feeling of belonging and continuity-in-being (staying the same person(s) through time)

resulting from an act of self-ascription, and/or ascription by others, to a group

of people who claim both common ancestry and a common cultural tradition'[8]


This literature is too vast and dynamic to be grappled with, in toto, in this paper. Instead, he we will review only one traditional understanding of this concept most prevalent in the anthropological literature concerning this topic – the definition provided by Smith. Gil-White argues, however, that all of the definitions given above share essential features with Smith's classification. It may be noted here that Barth does not come up any original definition of ethnic group. Rather he borrows Norrell's(1964) scheme and adds a relational and deeply subjective touch to that early schema.





Smith's Definition of Ethnie


Smith states "[E]thnie (ethnic communities) may now be defined as named human populations with shared ancestry myths, histories and cultures, having an association with a specific territory and a sense of solidarity."[9] Six criteria, or as he prefers to have them called, dimensions of an ethnie, ethnic group or community are described her.[10] It will be useful to enumerate and clarify these criteria before proceeding further:


  1. Named human population: the name could be self-ascriptive or ascribed by others.
  2. Association with a specific territory: this need not be a territory that the group members own or actually reside in; 'association with it may be just a potent memory.' And 'A land of dreams is actually more significant than any actual terrain'
  3. Shared culture: religion and language fare highly but so do other forms of cultural values which also express themselves publicly.
  4. A sense of solidarity: often finds institutional philanthropic expression.
  5. Shared history: shared memories
  6. Shared ancestry myths,


Methodological critique


Paramount, and often ignored, is the process whereby Smith reaches this definition of ethnicity. I will try to analyze that process here, backing it up an epistemological critique. Effectively, Smith's methodology is three-pronged.


  1. Using his intuition, he posits a question : 'Is Characteristic X is common to all ethnic groups"
  2. To answer this he must identity ethnic groups. To do that, he must have a definition of ethnic groups, which as yet he doesn't, given that he is still looking for a yardstick whereby ethnic groups could be told apart from other groups.  Therefore, he puts together all groups which, under common sense, appear as ethnic groups to him. He does not explicitly state this, and might not confess to it, but he cannot proceed without making this step.
  3. He tests his thesis against the data set he has selected (on the basis of sheer common sense) and states the result. "Are there any unnamed ethnie… I do not know of any."[11]


Smith's account gives us the impression that they are pointers to a definite reality which lies out there in the real world. The more closely the pointers correspond to the facts on ground, the better the definition. This, however, is not the case. Epistemologically, his methodology is not truly inductive and cannot yield any new knowledge about facts of the world. Instead, it only serves to enunciate the author's intuitive feelings. When the author tries to see whether his proposition corresponds to fact, the set of facts available to him is one which he himself had selected. And more likely than not, the same biases which made him come up with the thesis also affected his selection of data. Therefore, it is all too likely that once he states a thesis, his facts will corroborate it. So if all ethnic groups have a label because the author can't think of any that don't, this could be so simply because the author refused to pick nameless groups in his initial undeclared selection of ethnic groups.


In short, definitions do not reveal any new knowledge about the reality being studied; they are tautological. They are, however, a systematic expression of common sense. The criteria for judging constructs like definitions are laid later in the paper.


Are Lumunites an Ethnic Group?


The Lahore University of Management Sciences is an elite private university located at the outskirts of Lahore. It has a student population of around 2000, most of whom are undergraduates, who spend an average of four years the institution. Around half the students come from outside the city of Lahore and reside on campus. Thousands of alumni, many of whom regularly return to the campus, are also a part of the community of Lumunites.


1. All students, in particular, and all members of the university community, can be referred to using the label "Lumunites". In fact, this common label has been gaining in popularity over the years, as evidenced by student publications and mail records.[12] This label is not just self-ascriptive; various businesses now advertise their products, referring to their target readership as 'Lumunites'. That the campus community is a named human population is beyond doubt.   


2. At first, an observer might reckon that coming from different cities, there is no specific territory with which all Lumunites have an association. But one thought at the lush green campus with its rolling gardens and elegant red-brick buildings should remove this misconception. In fact, the recent rise in alumni's investment and involvement in campus development, and massive participation by students in protests against projects that are perceived as threatening to the campus environment, reaffirm our conviction that Lumunites do associate with a specific territory – their university campus. The territory with which Lumunites associate, might be transformed with the passage of time, as new structures are erected and older ones are raised. However, there is sufficient continuity and the rest is made up by imagination which has big role to play in identity formation.


3. Many Lumunites believe that LUMS is culturally very diverse. An objective observer, however, may not corroborate this finding. It appears that Lumunites are generally all affluent, Westernized and modern in their outlook. Various interviews with people who do not conform to these cultural norms find themselves ill at ease here. This could not have been a case if Lumunites did not a shared culture. All Lumunites can communicate in English and an overwhelming majority is, at least nominally, Muslim. However, even if Lumunites' self-perception about their cultural diversity is true, it can still be said that diversity forms their shared culture.


4. Like all other ethnic groups, Lumunites' sense of solidarity ebbs and flows, reaching a peak when cofronted by the other. To corroborate the this claim, one has only to see the crowds that gather to cheer the LUMS sports teams in their fixtures against 'other teams'. Lumunites try to help each other in things as small as cheating on tests, and giving a ride to getting jobs and admissions. Becoming a part of the community of Lumunites is advertised as one of the major assets that one acquires through the LUMS experience.[13] Alumni, now far away from their homeland, donate generously to support needy Lumunites under the National Outreach Programme, whose members are expected to comprise 10% of all students in a year's time. At least one auditorium was constructed with their money. 


5. The history of LUMS may be short, but it is eventful enough to provide anecdotes, tales and sagas which lumunites can share with each other, during small talk. Every year as freshmen arrive or when new members are inducted into student societies, or when farewell is bade to graduating classes, much attention is paid to oral transmission of history from one 'generation' to the next. Perhaps the most import way in which Lumunites develop a sense of shared history is after they graduate. Lumunites often recollect and share memories of the 'wonderful college days'. This history becomes a source of identity for them and a discussing it connects them with other Lumunites.


6. Perhaps the only criteria in Smith's definition that Lumunites have some difficulty in, is having a shared myth of descent. Deeper analysis, however, shows that even this element can be found amongst Lumunites. First of all, Smith's idea of a common myth of descent does not just refer to myths about how people believe themselves to have been born. Instead it refers, in general, to myths about the process all group members ended up together. And in that broader sense, Lumunites hold this belief, almost universally, that they have landed in LUMS because of being the most intelligent and gifted people in the country. This myth is systematically propagated at the 'Orientation Day' for all incoming members and is thereafter kept afloat by careful manipulation of admission policy.


It may be contended that this myth is close to the facts on ground. To this it may be replied that this contention is irrelevant because a myth does not have to be necessarily false; it is enough that it is popularly upheld without ample demonstration. Moreover, Lumunites' myth might actually be a falsehood. Given the narrow socio-economic strata from which Lumunites are all drawn, universal comparisons are impossible. And if this is so, the statement about Lumunites being the cream of the cream remains, at best, an unsubstantiated hypothesis. Therefore, it can be referred to as Lumunites' shared myth of descent.


Furthermore, Lumunites longingly remember their Golden Age when, in the university's early years, the community was less populous and their territory was vast and green and the atmosphere was libertine. Given the rise in the quality and quantity of human and financial capital in later years, this conclusion may not be borne out by most objective scales. This too, therefore, can be considered a part of Lumunites' baggage of myth which brings them together as group.


In sum, Smith's criteria can be legitimately interpreted so as to declare Lumunites as an ethnic group.



Are the Pakhtun an Ethnic Group?


1.      The Pakhtun are one of the most highly recognized of all ethnic groups. Pakhtun is the collective name they use for themselves, while others refer to them variously as Pathan, Pakhtun or Pashtun. That the Pakhtun community is a named human population is beyond doubt.


  1. The Pakhtun associate with the area of land that lies in one of Pakistan's provinces, NWFP and north-eastern and southern parts of Afghanistan. Within the Pakhtun, however, are various groups each of which identifies with its own territory.


  1. Almost all Pakhtun can speak Pashto language. Also, the Pakhtun have a complete code which sets their cultural norms. It is believed to be in conformance with Islam and thus divinely sanctioned. The culture of the Pakhtun is fairly distinct. Not only are they highly self-conscious[14] of it but other also identify them with it. Barth argues that there is a strong trend of deviating from this culture and seeking emancipation from it, when this suits someone's interests. He also analyzes situations where this can easily be done and where it cannot be, concluding that cultural standards themselves might sometime be changed in some cases. Whether this common culture evolves continuously or discretely, it is beyond doubt that at each moment there is some culture that Pakhtuns consider common to their entire group.  


  1. Pakhtun have earned fame as long-distance traders and transporters, thanks to their communal solidarity.


  1. Pakhtun have a strong sense of shared history. They look back to Ahmad Shah's time as their Golden Age. They vividly recall heroes like Baba Khushhaal Khan who resisted foreign onslaughts. They are proud of their reputation for valiantly fighting, first against and then for the British.


  1. All Pakhtun are believed to have 'a common ancestor; who lived 20-25 generations ago according to accepted genealogies'. 'The putative ancestor Qais lived at the time of the Prophet. He sought the Prophet out in Medina, embraced the faith, and was given the name of Abd-ur-Rashid.'[15] From a strictly patrilineal descent is traced.


It can, therefore, be concluded that the Pakhtun possess all six of the characteristics which, according to Smith, make them a distinct ethnic group.



What good are concepts?


I subscribe to an instrumentalist view with regard to the nature of scientific constructs.

I do not think if the worth of constructs lies in their ability to 'cut nature at the joints'; it lies rather in their academic utility. What good are constructs except as tools that help us predict and grapple with realities of the world? Therefore, I will not judge the worth of Smith's definition by looking at whether it corresponds to facts( I have no way of doing so, in any case!); rather I will judge it with regard to whether it yields any meaning or utility; this is whether it helps in predicting behaivour and framing adequate responses.


Conclusion: The infinitely malleable and paradoxical nature of ethnicity


Juxtaposing the findings from our two lines of inquiry, we are confronted with a problematic situation: either Luminites and Pukhtuns are both ethnic groups or none is. If one interprets Smith's criteria conservatively, Luminites obviously don't pass as an ethnic group but even Pakhtuns don't qualify. This is so, because it may be said that not every Pukhtun subscribes to Pashtunwali as his/her cultural norm anymore, or that the diasporic Pashtun no more associate, even in their imagination, with any specific territory nor remember their history nor believe in their myths of descent. However, a category so narrowly defined becomes meaningless for most purposes ranging from policing the roads and gathering census-data to allocating quotas in an economic redistribution policy.  Conversely, if one interprets the criteria loosely, the Pakhtun would pass for an ethnic group, but so would Lumunites and countless others. This time round, the category compasses all and sundry and becomes too broad and diverse to be meaningful for most purposes, for instance, the above-mentioned functions.


Additionally, this second situation confronts us with a paradox. This is so because earlier we established that Smith's definition represent no more than a systematic expression of common sense. Now we see that the framework built up on those definitions yield the highly non-sensical conclusion that Lumunites are as much an ethnic group as Pakhtuns. A system built using common sense as its axiom yields conclusions that defy common sense. The implications for ethnicity are damaging.


The implications of this experiment are similarly devastating for those theorists, including Smith himself, who base the concept of nationalism upon that of ethnicity. Some of these theorists claim that each ethnic group comprises a nation, and once a group comprises a nation, it get the right to self-determination under international law. However, if the foundational concept of ethnicity is found to be infinitely malleable and paradoxical, then it can provide no more than shaky foundations for nationalism, and ultimately for self-determination.




Around the middle of the 20th century, scholars began to suggest that the concept of race was seriously flawed, scientifically useless and should thus be discarded. The publication of the UNESCO statement titled "The Race Question" [16] in 1950 galvanized this movement. The stirring words of this document should leave us in no doubt as to the importance of correct, and not arbitrary, categorization of human collectivities. It reminds us also of the horrors that ensued from a misplaced emphasis on the badly defined category of race, under the Nazis.  


Soon after 1950, various states reformed their policies replacing the concept of race with ethnicity, thus remedying much injustice. In the United States for instance, the courts apply strict scrutiny in case of racial discrimination,[17] while the provisions for discrimination along the lines of ethnicity are less stringent, under the belief that the latter concept is less liable to abuse. It appears now, particularly in the light on our analysis in this paper, that as a concept or as an analytical category, ethnicity is no good either. Being infinitely malleable and paradoxical, it does not help further our understanding of the behaviour of people mush. It can be, and often has been, used to discriminate arbitrarily against people. It may, therefore, be time to replace it with some more meaningful, concrete and consistent category. If the purpose of social science scholarship is to refine our tools of understanding the social world and grappling with it, then a better construct than ethnie needs to be employed. If nothing else, at least a better definition of ethnicity should be arrived at. The effort of Gil-White represent a positive step toward clearing up the mess. However, the so the search for better, less arbitrary conceptual tools must go on.

[1] Barth, F. Introduction. Ethnic Groups and Boundaries. p. 5

[2] Gil-White, F. J The Study of Ethnicity and Nationalism Needs Better Categories: Clearing up the Confusions that Result from Blurring Analytic and Lay Concepts. Journal of Bioeconomics (2005) 7:233 250–251


[3] Weber, Max. (ed). 1968. Economy and society: an outline of an interpretive sociology. Vol. 1. University

of California Press, Berkeley. p. 389


[4] Barth, Fredrik. 1969. Ethnic groups and boundaries: the social organization of cultural differences.

Little Brown & Co., Boston.


[5] De Vos, G. 1975. Ethnic pluralism: Conflict and accommodation. Pp. 5–41 in G. De Vos &

R. Romanucci (ed.) Ethnic Identity: Cultural continuities and change, Mayfield, Palo Alto. p. 9


[6] Patterson, Orlando. 1975. Context and choice in ethnic allegiance: a theoretical framework and a Caribbean case study. Pp. 309 in N. Glazer & D.P. Moynihan (ed.) Ethnicity: Theory and Experience,

Harvard University Press, Cambridge, MA.


[7] Nagata, Judith. 1981. In defense of ethnic boundaries: the changing myths and charters of Malay

identity. Pp. 98-99 in C. F. Keyes (ed.) Ethnic Change, University of Washington Press, Seattle &



[8] Roosens, Eugeen. 1994. The primordial nature of origins in migrant ethnicity. Pp. 83-84 in H. Vermeulen & C. Govers (ed.) The Anthropology of Ethnicity: Beyond "Ethnic Groups and Boundaries". Het Spinhuis, Amsterdam.


[9] Smith, A.D.  Foundations of Ethnic Community in Smith, A.D. The Ethnic Origins of Nation, London: Blackwell. p. 32

[10] Despite Gil-Wite's rather convincing protestations for differentiating these terms, in this paper these terms are used interchangeably. They were used thus, during the above-mentioned experiment. Consistency, therefore, requires that practice to be continued through this discussion.

[11] Smith. A. D. ()1986) ibid, p. 23

[12] The Student Council of LUMS has made it a consistent practice to address all it public mails to "Lumunites", thus invoking the common identity of its referents.

[14] Barth, F. Pathan Identity and its Maintenance.


[15] Ibid.

[17] The Court declared: "…disadvantaging racial classifications are ordinarily 'suspect', must be subjected to 'the most rigid scrutiny,' and bear a 'very heavy burden of justification.' in Brown v Board of Education. 347 U.S 483. 74 Ct. 686, 98 L.Ed. 873 (1954)