New Naratif is Fundraising!

New Naratif is a 100% ad-free, member funded, non-profit movement for Southeast Asian democracy. Membership fees fund high quality, independent journalism and create positive change in our communities. This year, we need to raise US$226,336 to continue our operations. Please consider becoming a member or donating today!

Visit our fundraising page

This article is derived from: Sandra Manickam (2014) Bridging the Race Barrier: Between “Sakai” and “Malay” in the Census Categorisations of British Malaya, Asian Studies Review, 38:3, 367-384, DOI: 10.1080/10357823.2014.928666.

New Naratif thanks the author and the publisher for permission to publish this shortened, revised version of the original.


When states engage in the construction of racial categories and ideas, they are creating new ways of seeing and possibly controlling populations. The study of census data can expose shifts in racial perceptions and classifications from one census to the next, thus uncovering the operation of both deliberate and unconscious power plays based upon race. This article uses historical census data to deconstruct colonial racial categorisation in British Malaya (the Malay Peninsula including Singapore), in order to demonstrate just how contingent and problematic the perception of a person’s race was. It was more a reflection of colonial perceptions of a person’s place in society than of readily discerned biological essences. The difficulties that the Malayan colonial authorities had in determining indigeneity reveal the pitfalls of state-driven projects of racial classification.

“What is that man?”

In Imagined Communities, Benedict Anderson’s famous study of the incipient nationalisms in Southeast Asia and the colonial institutions that fed them, he writes that “the fiction of the census is that everyone is in it, and that everyone has one – and only one – extremely clear place”.[1] Putting people in their place was indeed the aim, and the very heart of the problem of the census of aborigines in Malaya.[2]

My analysis is based on Malayan census data and associated documentation beginning with the first official census in 1871 and ending with British Malaya’s last pre-WWII census in 1931. Like all censuses, these exercises were in fact “human inventories” (to borrow a phrase from Norbert Peabody),[3] but the remarkable feature is that these inventories were taken according to race.

For colonial officials, faced with the myriad languages, cultures, religions, and skin colours of Malaya, it is easy to understand that racial categorisation had unquestionable utility. Yet, counting by race was problematic, partly because of the variety of understandings of the term “race” when applied in Malaya. The terminology used and the kind of information that was to be collected was not at all consistent: it differed for each group of people and from one census to the next. For instance, the 1901 report, written by G.T. Hare, commented that “the race or tribe of Chinese is to be determined by tongue and not by birthplace”.[4] By contrast Indians professing the Sikh religion were commonly categorised as Bengali by race, making the religious identifier critical, and linking it directly to the birthplace. ‘Race’ was not even consistently used as the master category. Alternative words, such as ‘nationality’, were often used as group terms for people, although over the years the preference for the word ‘race’ became mainstream among census superintendents.

Clearly, race encompassed more than just differences in bodies. It encompassed language, religion, lifestyle, and political status. The weight attached to each of those elements depended on the government and census bureaucracy of the day and the input of informants and respondents. In the 1921 census report, J.E. Nathan acknowledged that each racial division was not really “one race”, but rather that they were aggregated together from a longer list before the census report settled on the essential few.[5] As has been elegantly illustrated by Hirschman in a table comparing the changes and contractions in race categories throughout Malaysia’s colonial and post-colonial periods, the choice of which races to include, and how they were defined, was a political decision that changed from one year to the next in line with government ideologies and uses.[6] That the census information gathered was to serve government needs was explicitly stated in 1931 by the census’ director, C.A. Vlieland:

the word “Race” is used, for lack of a more appropriate term, to cover a complex set of ideas of which race, in the strict or scientific sense, is only one small element. It would be of little use to the administrator or the merchant to attempt a classification of the population by race in the ethnographic sense, to say nothing of the fact that any such tentative classification would be highly controversial….

In default of anything resembling a definition of the term “Race” as used in this report, perhaps the best way of conveying its meaning in a few words is to say that, in asking the question of an individual “What is your race?” the census authority is trying to obtain an answer of the same nature as we expect when we ask in ordinary non-technical conversation “What is that man?”


Apart from the influence of government outlooks in determining racial categories, his quote further indicated that the meanings of race employed by several segments of the population made their way into the census. For instance, being Muslim was a prime marker for membership in the Malay race according to many locals and officials alike. Vlieland intimated that social categories and social differentiations were most important. The question “What is that man?” could be answered in a number of ways, depending on how the individual was placed within society, who was doing the identification and being identified, and which aspects of their being were seen as most important.

…race encompassed more than just differences in bodies. It encompassed language, religion, lifestyle and political status.

When it came to enumerating indigenous people or aborigines, the enumerators appeared more at sea than when dealing with other populations because of the difficulty in locating them, in contrast to most of the population of Malaya, who were counted in dwellings. Unique bureaucratic procedures were put in place to deal with aborigines, adding complexity in the levels of reportage from the collection of data to the final publishing of the report. Special enumerators were commissioned to enter the jungle to locate and count members, or to call feasts with headmen who would invite indigenous peoples to attend and subsequently be counted. The circular ordering the commencement of the 1901 census of the FMS noted specifically that “it is highly desirable that special efforts be made to ascertain the number of Sakai and other indigenous peoples living in each state. If necessary, a few trifling presents might be offered to them in various districts, so that they may be induced to come to the nearest villages for the purpose of the census”.[8] This practice continued in 1911 when census superintendent Pountney commented that the cost of the census was very low considering that “special enumerators” had to be “despatched to visit the Sakai encampments”.[9] This method of sending people to the abodes of aborigines or organizing feasts became known as the Sakai[10] census, as opposed to the general census where enumerators went door-to-door and counted occupants of households.

Special chapters were also devoted to aborigines in the census reports of the 1900s, showing that the subject of aborigines required expert knowledge and was distinctive from the rest of the population. For the 1911, 1921 and 1931 censuses, one chapter was dedicated to aborigines, with external experts called in to contribute on census methods, to look at the census data and/or to write the report. The chapter on ‘The Sakai’ in the 1911 census, the first such chapter devoted specifically to the topic of aborigines, was written by R.J. Wilkinson, a colonial scholar-official who also studied topics such as indigenous peoples and the Malay language.[11]

The reports included details on the difficulties in locating and identifying aborigines, the methods used, and the idiosyncrasies of the entire endeavour of counting. In many instances, there were inconsistencies in definition from one census to the next and problems of enumeration that called into doubt the reliability of the count. But the unreliability of the count, such as the ways in which the count was changed, repackaged and rethought, is precisely the important aspect of the Sakai census that reflects the difficulty in categorising aborigines and, what is more, even in merely identifying who was aboriginal as opposed to Malay at certain points in time. Problems in identification arose because the assumptions held by census takers about aborigines and Malays did not fit the people who were being categorised. Due to the distinctiveness of the Sakai census, however, certain strategies were employed to capture those who were considered in-between categories, although these too occasionally offered up their own issues.

Aborigine but not Malays, Malay but not Aborigines

One of the main assumptions and guiding principles governing the taking of the census was the separation between Malays and aborigines. In the 1921 census report, the general description of the inhabitants of British Malaya started with the heading ‘The Aborigines’:

The earliest inhabitants of the Peninsula were probably the Semang, a race of Negritoes, related to the Aetas of the Philippines… Superior to them in culture are the Sakai, a race supposed, mainly on linguistic grounds, to have migrated to the Malay Peninsula from Indo China…. In Negri Sembilan and Johore are found a number of aborigines, usually called Jakun or Biduanda; it is now generally held that their real language is Malay, though not quite the Malay of the civilised Malay of the Peninsula.


The intellectual separation between Malays and aborigines built on another more fundamental assumption of the bounded and knowable categories of Malays and aborigines respectively. Despite the different peoples who comprised Malays, the colonial idea of Malays as farmers originating in kampungs on the Malay Peninsula was actively endorsed in the census and used as the basis for comparison of Malays and aborigines.

The division of aborigines into three racial types reflected the standardisation of aboriginal divisions by anthropological methods first proposed by W.W. Skeat.[13] Skeat was a prominent figure in the application of physical anthropology to the study of indigenous peoples in Malaya. He led the Cambridge Expedition to the Malay States from 1899 to 1900 that established three racial divisions of aboriginal peoples in Malaya: the “woolly-haired Negrito tribes called Semang”, “the wavy-haired tribes called Sakai” and “the straight-haired tribes called Jakun”. This racial categorisation also amounted to a ranking, with Semang considered by Skeat to be “a representative of one of the wildest races of mankind now extant”, while Jakun were placed closer to the civilised Malays through their informal designation of “savage Malays”.[14] This migration-based theory of the peopling of the Malay Peninsula is generally still held by scholars and lay people in Malaysia today and is known as the kuih lapis or layer-cake theory.

Benjamin, however, suggests that in situ differentiation could equally explain the different cultures and the diverse physical features of people on the Malay Peninsula.[15] Most recently, Scott has argued that “the invention of the tribe is best understood as a political project”.[16] Approaching indigenous people from this angle allows us to ask what political projects and which parties were involved in delimiting the group(s). For the case of Malaya, the political projects that facilitated the emergence of indigenous people as different from Malays (and Malays as different from indigenous people) were anthropological applications of race which, in tandem with (or possibly in subservience to) the governmental imperatives of having a stable Malay population, constructed racial boundaries through the census.

The layer-cake theory, however, assumed a biological and cultural separation between people known as “Malays” and “aborigines” respectively, and this thinking was supported by cutting-edge anthropological research of the day. Despite this assumption, observers in Malaya had long noted the trend of some people identified as aborigines who displayed the culture, language and look of stereotypical notions of Malays.[17] These aborigines were commonly known by the phrase “tame Sakai” or “tame aborigines”, phrases that appear in informal colonial sources and some scholarly works on Malaya.[18] This phrase, however, suggests movement between races, contradicting the assumption of static racial essences and the term’s codification in the census.

Despite the different peoples who comprised Malays, the colonial idea of Malays as farmers originating in kampungs on the Malay Peninsula was actively endorsed in the census and used as the basis for comparison of Malays and aborigines.

The “tame” and “wild” concepts as applied to aborigines in Malaya have their particular origins in the Malay terms jinak and liar.[19] Peabody, in discussing the case of late pre-colonial and early colonial India, suggests that “colonial discourses often built upon indigenous ones in ways that inflected local politics about which the British initially were only dimly aware”.[20] Similarly, scholars of Malaya in the late nineteenth century borrowed phrases used by their Malay-speaking informants concerning Orang Sakai liar and Orang Sakai jinak (the wild and tame Orang Sakai).[21] In these writings, the Malay terms were freely translated into the English for tame and wild, since there already existed parallels between the connotations surrounding the terms in both Malay and English. The entries in dictionaries of the Malay language give us further hints of the meanings of jinak and liar. From the nineteenth century, jinak connoted tameness, the state of being domesticated, and meek. The additional definitions “docile, tractable” and “easy or reconciled to one’s position” show that acceptance of a hierarchical system, a larger society or state perhaps, was part and parcel of tameness.[22] To lie outside of that scheme was thus to be wild.[23]

Tellingly, Crawfurd’s 1852 dictionary translated jinak as “Tame, not wild”, defining tameness in opposition to wildness but also suggesting an intrinsic connection between the two terms such that the boundary between tame and wild was not always so clear. Tameness in the Malay Peninsula was understood as having the characteristics of being Malay and no longer wholly aboriginal, while wildness had those of non-Malay forest- or sea-dwelling people. Therefore, tameness could connote a range of states from having once been wild but no longer being so (to the extent of being identified as Malay) to not-quite aboriginal. It is this last characteristic of tameness, that of movement from one category of behaviour (wildness) to another (no longer wild, or tame), that defies rigid racial classification which otherwise assumed that membership in one race meant a different biological and cultural history. Thus, tameness and its implications of movement between these stereotypes further confounded the rigid separation mentioned before between Malays who were the most developed of the Malay Peninsula’s natives and the less-civilised aborigines who were supposed to have arrived earlier.

The term “tame” or “settled” Sakai was gradually introduced into the census precisely to categorise people who were otherwise in-between categories and who might possibly be moving towards becoming “Malay”. In general, the names given to aborigines and Malays in the census prior to the introduction of the tame Sakai category in 1911 did not suggest movement between races. Rather, census categorisation reflected the changing understandings of aborigines’ links to other groups. For instance, aborigines were enumerated since 1871, when a group of “Mantra” were counted in the Straits Settlements.[24] “Mantra” was then taken away and the population once included under that term was now generally called “Aborigines of the Peninsula”.[25] Similarly for Malays, the groups that comprised that category may have shifted, but not due to changes from one race to another. Instead, what changed were the assumptions about the connections between groups. Malays in Kedah, for example, were once counted as foreign because they were under Siamese, rather than British, control. Once Kedah became a British protectorate, they were reclassified as local Malays.[26] In both these instances, what changed throughout the years were the associations surrounding aboriginal groups and Malays in Kedah.

The changes in the enumeration of aborigines that occurred in the 1901 and 1911 censuses, however, signalled an acknowledgment of the movement between races that were still seen as separate. One of the main indications of such a movement was the appearance of aborigines who were not counted in the forested areas of the peninsula or as nomads. As mentioned previously, the knowledge of aborigines who lived in remote places that were difficult to reach by enumerators brought about the group method of enumeration in 1901 whereby headmen and influential members of society who had contacts with aborigines were asked to invite them to a feast. This effectively created a special census of aborigines separate from the main count. Along with the intended effect of a more comprehensive count of Malaya’s population, however, was the unintended effect of highlighting that there were people who were identified by them- selves or others as aborigines, but who lived in the general population. In the summary table of the distribution of the main races, “aborigines” came to a total of 18,574 people, of whom 1,397 “Sakai” were enumerated as part of the general population.[27] This indicated that they were not living in remote places at the time of enumeration, that they had occupations like other “ordinary” inhabitants, and that they were living among people in one of the other major racial categories. Besides noting the number of aborigines counted outside the Sakai census, they were otherwise not separated from the general Sakai category and no other information was provided on this group.

This was not the case in 1911, when tame Sakai appeared in the report after officials had given further thought to the matter of aboriginal classification. In this year, only 941 people were counted as aborigines in the general population. Even though this was less than the number counted in 1901, they were now given a separate section in the tables on aborigines and a separate name: tame Sakai. Wilkinson, who wrote the chapter on aborigines for that year, stated that “including ‘tame’ Sakai the complete total [number of aborigines] is 27,218”.[28] This in-between category came about because stereotypical elements of Malays and Sakai/aborigines were not fulfilled on all counts. In this instance, the place of enumeration was the determining characteristic of tameness. In fact, the place of enumeration was already a factor in the differentiation of aborigines in 1901 even though no term was employed then to describe this segment.

Approaching the occurrence of tame Sakai from this aspect, the term could be considered merely as a filler, a term used to capture aborigines who were still aborigines even though they were counted in the regular census. Yet additional notes about tame Sakai show that the term captured the dilution of the concept of aboriginality and supposedly identified the process of becoming something else, namely Malay. A few crucial aspects indicated a movement into Malayness, such as inter-marriage and lifestyle attributes like dress and living circumstances. Wilkinson explained that tame Sakai were “people of aboriginal descent [who] were included in the regular census schedules through their marrying or settling down among the civilised peoples of the country”.[29] Although he was not specific about which “civilised peoples” aborigines married or lived with, elsewhere he made it clear that he had Malays in mind. Wilkinson explained that Malays had intermarried with aborigines but had “failed to absorb [them] completely”.[30] Marriage to Malays meant aborigines were on their way to becoming absorbed into a more civilised race, but not quite yet if they could still be identified (or identified themselves) as aborigines. In this case, the intermediate categorisation was brought about by the actual cohabitation of people known to be Malay and aboriginal respectively.

Inter-marriage in itself was not, however, a prerequisite for tameness. Practising a culture that was not stereotypically aboriginal (which was frequently thought of as a nomadic existence living solely in the forest or at sea) was also sufficient to spark ideas of tameness in the minds of enumerators and report writers. Wilkinson commented on the difference between “tame” and “wild” Sakai:

It may, however, be surmised that the “tame” aborigines differ from the “wild” only in the fact that they have discarded their old communal houses and the use of the bow and are losing other racial traits such as the making of bark-cloth and the painting and tattooing of the face. Briefly, they are becoming sophisticated.


There was a further indication of the ideological overlap between lifestyle and tameness in the 1921 census report on aborigines. The writer of the 1921 chapter on ‘The Aboriginal Races’, Winstedt, changed the earlier 1911 “tame Sakai” category to “settled aborigines” and the general aboriginal population (who were not tame/settled) to “nomads”.[32] This change was maintained in the next census as well. Winstedt wrote:

it has been apparent to everyone who has come into contact with the wild tribes during the last few years that in many districts they are tending to settle down among Malays, and that, even where they are kept apart, they tend to lose their nomadic habits and form semi-permanent settlements.


The change in terminology can be attributed to prevailing notions about untouched aborigines being originally nomadic while aborigines who maintained some sort of permanent settlements were assumed to have adapted to this apparently civilised lifestyle through interaction with Malays. Different cultural practices were placed on a gradation from wild to civilised, and were seen to be the essential state of certain populations and not others. Winstedt assumed that the change in way of life and location (settling down among Malays, no longer practising a nomadic lifestyle) was indicative of tameness; he also noted the “merging of the aborigines into the Muhammadan Malay”.[35] In line with Winstedt’s understanding of the differences within the aboriginal population, the census data for 1911 and 1901 were reproduced in the 1921 report with the division between settled and nomadic aborigines in mind.[36]

The census process itself also created irregularities in the codification of the racial terms, due to the various stages through which the census data had to travel before appearing in the report. This aspect of the census process is yet another striking indictment of how the process of racially categorizing people was at odds with the idea of race as an essential and obvious aspect of being. The actual categorisation of aborigines was complicated by the distance between the rationale behind census questions, the organisation of the census on the ground by district officers who themselves relegated the task of enumeration to certain locals (most commonly mentioned as Malays), and the eventual collection and interpretation of data that resulted in a report in the census. Within this scheme, there was plenty of room for differing interpretations and standards for judging racial affiliation, as well as opportunity for external experts to impose an overarching racial scheme on the schizophrenic data.

One set of irregularities was the categorisation of aborigines at the level of the report, which could differ substantially from the data collected by enumerators. For the 1911 census, Wilkinson had engineered the census questions by compiling a list of words in Jawi (Malay in Arabic script) to be given to enumerators. These enumerators were to ask aborigines for the equivalent words in their language. Based on the list of words compiled, Wilkinson then divided the aboriginal population according to seven racial divisions: Negritoes, Northern Sakai, Central Sakai, Besisi, Jakun, Benum Aborigines, and Malay-speaking aborigines. This division was carried out despite his initial caveat that the tribes were difficult to locate and classify, and that they differed from one another in custom and belief, and were very often of “mixed race”. Divisions were further supplemented with bodily descriptions such as facial features, skin colour and hair type.[37] Yet, when reading the specific reports by enumerators, it was clear that these divisions were imposed on the data afterwards and that supervisors and enumerators merely counted people as Sakai. For instance, Wilkinson separated the people enumerated in the district of Kuala Kangsar, Perak into three aboriginal races based on language (Negritoes, Northern Sakai and Central Sakai), whereas the extract of the enumerator’s report on Kuala Kangsar mentioned that the census was conducted with “the services of a Malay who talked some Sakai” and it was not specified which aboriginal languages he spoke. Extracts from other districts confirm that the term Sakai was used throughout to refer to all aborigines and not the specific divisions of Wilkinson.[38]

It was also frequently the case that some people considered “Malay” or “Other” by enumerators might have been considered “aborigines” by report writers. In 1921, Winstedt noted in hindsight that aborigines were poorly enumerated in the UFM in 1911. He said that “in 7 out of 9 districts of Trengganu the Malay officers in charge reported that no aborigines could be found, an almost incredible statement in view of their large number in Ulu Kelantan”.[39] The speculation about aborigines who were enumerated as Malay extended to the 1921 and 1931 censuses. In 1921, the situation in Negeri Sembilan was particularly fraught with such issues. Winstedt said that the decline in the enumeration of aborigines there was most likely due to their being counted as Malays if they were Muslim, and not necessarily due to an actual decline.[40] In 1931 Vlieland noted the same situation:

It is … certain that the conversion of the aborigines from their characteristic nomadic habit of life has been far more considerable than these [settled and nomad] figures suggest, and they have, of recent years, become progressively more assimilated with the Malays. Where no assimilation, or intermarriage has taken place, the aborigines still tend to lose their nomadic habits and form permanent or semi-permanent settlements marked by coconut and other plantations. It is fairly certain that many of these settled aborigines would not be detected as such by, and often would not admit their true descent to, any enumerator not well acquainted with them. Further, any aborigine who had been converted to Islam would be returned as “Malay”.


While numbers and counting were “part of the enterprise of translating the colonial experience into terms graspable in the metropolis”, and locally too, there seemed to be a continual distrust of the census even as the figures for tame and wild (or settled and nomadic) were presented.[42] It boiled down to who did the questioning, who answered, and which yardstick was used when determining categories that did not have definite boundaries themselves. Both Winstedt and Vlieland noted the key roles played by enumerators in the process of identifying aborigines, and frequently argued that the census of aborigines was only as good as the enumerators tasked with the “correct” identification of people’s race.

There were even irregularities in the application of the in-between category of tame Sakai. The constant questioning surrounding the census data and methods of classification from the ranks of report writers themselves illustrates the fundamental problematic of the whole endeavour, which was the difficulty of counting people as aboriginal, Malay or in-between when such identification was always a personal and an idiosyncratic decision. The extracts from the reports by assistant superintendents of the census show differing understandings of aborigines from Wilkinson’s, particularly in relation to who was tame and who was wild. Determining language use and religious observation were particularly important to the judgment of tameness.

In today’s political, social and economic climate, fixed ideas of race are ingrained in forms of law, governmental and economic practices, and personal identities. It then becomes impossible to imagine a time when what is thought of today as indigenous and Malay, or Chinese and Indian, was not the same as what was in the past.

From the privileged vantage point of the present, many of the observations on the census of aborigines have far-reaching implications, not only for the indigenous community, but for anyone subject to racial enumeration and thus to government policies linked to race. Similar judgments were exercised on many people in Malaysia’s past and present in order to produce a count of races and an ideologically-laden report of the social situation.[43] As mentioned by Appadurai in relation to British census activities in colonial India, the meaning and import of numbers were “often either non-existent or self-fulfilling.[44] Such a complex reality was one that was not adequately captured by racial thinking in terms of separate waves of people and essences. Although there are no sources that detail the specific involvement of locals, and the responses of people classified as Sakai, Malay or in-between to these classifications, the Sakai census and report were unique in that the documents do show certain local traces in the terminology and its application, and not just an imposition of foreign discourses.[45] This oblique participation in the conceptualisation of the count, though not concrete or always highlighted in the reports themselves, is nevertheless present and important to note in any historical reconstruction of ideas on aborigines.


In 2008, David Lim wrote that Malaysians today hold a passion for race from which it is difficult to break free.[46] The colonial census reporters could similarly be said to have been in the throes of racial passion. They were influenced by a scholarly tradition that supported the assumption of separate races, and by a government that held deeply rooted ideas about the ideal roles and cultural practices of segments of Malaya’s population, all of which fuelled the desire to count races. But there were several facets to their belief in race, some of which were incongruous with others. The idea that one could change from one race to another was one such facet that seemed to go against the overarching principle of separate races. Yet this aspect was included in the reports on aborigines since it became a more regular feature in the enumeration procedures to come across such individuals. In this instance, the census underlined the importance of tracking a changing present, even though that present did not always agree with pre-formed underlying assumptions. Despite the principle of separate races of aborigines and Malays expressed at the meta-levels of the census, this principle was not always in operation in the Sakai census and reports.

The theory of separate bodies, histories and migrations into the peninsula (according to the layer-cake theory) was undone by the complexity of ideas used to gain a more comprehensive categorisation of aborigines. Categorising some indigenous people in the general population as tame raised more questions and uncertainties, ranging from which characteristics indicated tameness to the application of tameness at different levels of the census operation. Going back to the main categories of Sakai/aborigines and Malay, issues in the tame category conversely brought to light variation in opinion on who was aboriginal or Malay, with speculation by report writers on the under- or over-reporting of aborigines. While tame Sakai undid the narrative of separate races, the questioning of tame attributes compromised the naturalness of racial identities as a whole and the objectivity of censuses that count by race.

Highlighting these aspects of racial identification and census operations is important when questioning the objectivity of racial categories in colonial and present times. That people were counted, their numbers noted and presented in tables, often holds the sacred power of demanding to be taken at face value for what it purports to show about the population. Questioning how the categories and numbers came to be hopefully offers a powerful antidote to the aura of essentialised race categories and the policies past and present that are based on such data.

What is evident from such historicising is that categories such as race become overlaid with several different meanings and uses so that it becomes hard to separate the different strands that make up the idea. Indeed, what can be seen from the case above is that several related bodies of knowledge all had to work to maintain the very specific meaning of “Sakai” and “Malay” found in the census in the face of divergent understandings. In today’s political, social and economic climate, fixed ideas of race are ingrained in forms of law, governmental and economic practices, and personal identities. It then becomes impossible to imagine a time when what is thought of today as indigenous and Malay, or Chinese and Indian, was not the same as what was in the past. History, however, has shown us that things were not always like this, and most likely, they will not remain the same.


[1]Benedict Anderson, Imagined communities: Reflections on the origin and spread of nationalism (London: Verso, 1991), p. 166.
[2]Throughout this article, I use the terms “indigenous” and “aboriginal” interchangeably. In many historical sources, terms ‘aborigine’ and ‘aboriginal’ are employed as neutral descriptions, and I retain that usage. I use both terms in the uncapitalized form without intending any disrespect to indigenous communities present or past.
[3]Norbert Peabody, Cents sense, census: Human inventories in late precolonial and early colonial India. Comparative Studies in Society and History 43, (2001), pp. 819–50.
[4]George Thomson Hare, Federated Malay States, census of the population, 1901 (Kuala Lumpur: Government Printers, 1902), p. 16.
[5]J.E. Nathan, The census of British Malaya, 1921 (London: Dunstable and Watford, 1922), p. 70.
[6]Charles Hirschman, The meaning and measurement of ethnicity in Malaysia: An analysis of census classifications. The Journal of Asian Studies 46(3) (1987), pp. 571–76.
[7]C.A. Vlieland, British Malaya (the colony of the Straits Settlements and the Malay States under British protection, namely the Federated States of Perak, Selangor, Negri Sembilan and Pahang and the States of Johore, Kedah, Kelantan, Trengganu, Perlis and Brunei), a report on the 1931 census and on certain problems of vital statistics (London: Crown Agents for the Colonies, 1932), pp. 73–74.
[8] Hare, Federated Malay States, 1901, pp. lxxxvi–lxxxvii.
[9] A.M. Pountney, The census of the Federated Malay States, 1911: Review of the census operations and results including tables exhibiting the population by sex, age, race, birthplace, religion, and occupation (London: Darling, 1911), p. 91.
[10]The term “Sakai” became widely employed as a synonym for aborigines in English and Malay in the colonial period from the late 19th to the 20th century. The term is and was generally felt by many indigenous groups to be derogatory. Today, Sakai has largely been replaced by the term Senoi, which, along with Negrito and Proto-Malay, make up the tripartite division of indigenous peoples in Peninsular Malaysia.
[11]Pountney, The census of the Federated Malay States, 1911, p. 68; R.J. Wilkinson, Papers on Malay subjects, selected and introduced by P.L. Burns (Kuala Lumpur: Oxford University Press, 1971).
[12] Nathan, The census of British Malaya, 1921, p. 1.
[13] Sandra Khor Manickam, Situated thinking: Or how the science of race was socialised in British Malaya. Journal of Pacific History 47(3) (2012), p. 297.
[14]W.W. Skeat, (1902) The wild tribes of the Malay Peninsula. Journal of the Anthropological Institute of Great Britain and Ireland 32 (1902), pp. 124–27.
[15]Geoffrey Benjamin, On being tribal in the Malay world, in Geoffrey Benjamin and Cynthia Chou (eds), Tribal communities in the Malay world: Historical, cultural and social perspectives, pp. 7–76 (Singapore: ISEAS, 2002); Alice M. Nah, Recognizing indigenous identity in postcolonial Malaysian law: Rights and realities for the Orang Asli (aborigines) of Peninsular Malaysia. Bijdragen tot de Taal-, Land- en Volkenkunde 164(2/3) (2008), pp. 212–37.
[16]James C. Scott, The art of not being governed: An anarchist history of upland Southeast Asia (Singapore: NUS Press, 2010), pp. 256-59.
[17]Manickam, Situated thinking, pp. 291–92.
[18]John Crawfurd, A descriptive dictionary of the Indian islands and adjacent countries (London: Bradbury and Evans, 1856); N. Miklouho-Maclay, Ethnological excursions in the Malay Peninsula – November 1874 to October 1875 (preliminary communication). Journal of the Straits Branch of the Royal Asiatic Society 1 (1878), pp. 205–21; W.E. Maxwell, The aboriginal tribes of Perak. Journal of the Straits Branch of the Royal Asiatic Society 4 (1879), pp. 46–50.
[19] R.K. Dentan, The persistence of received truth: How the Malaysian ruling class constructs Orang Asli, in R.L. Winzeler (ed.), Indigenous peoples and the state: Politics, land, and ethnicity in the Malayan Peninsula and Borneo, pp. 98–134 (New Haven: Yale University Southeast Asia Studies, 1997), pp. 109–11.
[20]Norbert Peabody, Cents, sense, census: Human inventories in late precolonial and early colonial India. Comparative Studies in Society and History 43 (2001), p. 819.
[21]Maxwell, The aboriginal tribes of Perak; Miklouho-Maclay, Ethnological excursions in the Malay Peninsula.
[22]William Marsden, A dictionary of the Malayan language, in two parts, Malayan and English and English and Malayan (London: Cox and Baylis, 1812); John Crawfurd, A grammar and dictionary of the Malay language with a preliminary dissertation (London: Smith Elder, 1852).
[23]The lack of local voices on this matter is unfortunately not unusual given that available historical documents pertaining to the census are heavily biased towards printed British colonial records and other government-linked publications in English.
[24]J.F.A. McNair and A. Knight, Straits Settlements. Census. Reports and returns. 1871 (Singapore: Straits Settlements Government Press, 1872), pp. 6, 32.
[25]S. Dunlop, W.A. Pickering, V. Cousins, H. Hewetson, A. Knight and A.P. Talbot, Straits Settlements. Population (according to the census taken in 1881) (Singapore: Straits Settlements Government Press, 1882), p. 3.
[26]Charles Hirschman, The making of race in colonial Malaya: Political economy and racial ideology. Sociological Forum 1(2) 1986, p. 30.
[27]Hare, Federated Malay States, 1901, p. 30.
[28]Pountney, The census of the Federated Malay States, 1911, p. 68.
[29]Ibid., pp. 69, 72–74.
[30]R.J. Wilkinson, Papers on Malay subjects: A history of the Peninsular Malays with chapters on Perak & Selangor, 2nd edition (Singapore: Kelly & Walsh Ltd., 1920), p. 1.
[31]Ibid., p. 18.
[32]Nathan, The census of British Malaya, 1921, p. 125.
[33]Vlieland, British Malaya, pp. 101–104.
[34]Nathan, The census of British Malaya, 1921, pp. 126–27.
[35]Ibid., p. 124.
[36]Ibid., pp. 126–27.
[37]Pountney, The census of the Federated Malay States, 1911, pp. 69–70. It is not known how Wilkinson identified the bodily features of the individuals enumerated, since he did not carry out the collection of the word lists himself. It is likely that the description, like many others of aborigines, was based on prior knowledge production in general and not only or primarily on the individuals counted.
[38]Ibid., pp. 70–72, 172.
[39]Nathan, The census of British Malaya, 1921, p. 124.
[40]Richard Olaf Winstedt, ed. Malaya: The Straits Settlements and the Federated and Unfederated Malay States (London: Constable, 1923), pp. 124–25.
[41]Vlieland, British Malaya, p. 101.
[42]Arjun Appadurai, Number in the colonial imagination, in Carol A. Breckenridge and Peter van der Veer (eds), Orientalism and the postcolonial predicament, pp. 314–40 (Pennsylvania: University of Pennsylvania Press, 1993), p. 326; italics in original.
[43]Joel Kahn, Other Malays: Nationalism and cosmopolitanism in the modern Malay world (Singapore: Singapore University Press, 2006), Chapter 2.
[44]Appadurai, Number in the colonial imagination, p. 317.
[45]The lack of local voices on this matter is unfortunately not unusual given that available historical documents pertaining to the census are heavily biased towards printed British colonial records and other government-linked publications in English.
[46]Lim, David C.L. Introduction, in David C.L. Lim (ed.), Social sciences in Asia, volume 19: Overcoming passion for race in Malaysia cultural studies, pp. 1–12 (Leiden: Brill, 2008).

Nadhir Nor

An illustrator from Selangor, Malaysia. Nadhir Nor's art explores the otherworldly and the familiar. From mythical creatures intermingling with modern mundanity to tall tales of ancient temples and digital deities, there's always a sense of wonder to be explored and foraged from his works. Find out more about him at