Jump to ContentJump to Main Navigation
Mapping Israel, Mapping PalestineHow Occupied Landscapes Shape Scientific Knowledge$

Jess Bier

Print publication date: 2017

Print ISBN-13: 9780262036153

Published to MIT Press Scholarship Online: January 2018

DOI: 10.7551/mitpress/9780262036153.001.0001

Show Summary Details
Page of

PRINTED FROM MIT PRESS SCHOLARSHIP ONLINE (www.mitpress.universitypressscholarship.com). (c) Copyright The MIT Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in MITSO for personal use. Subscriber: null; date: 16 September 2021

Removing Borders, Erasing Palestinians: Israeli Population Maps after 1967

Removing Borders, Erasing Palestinians: Israeli Population Maps after 1967

(p.81) 3 Removing Borders, Erasing Palestinians: Israeli Population Maps after 1967
Mapping Israel, Mapping Palestine

Jess Bier

The MIT Press

Abstract and Keywords

Chapter 3, “Removing Borders, Erasing Palestinians”, provides an analysis of the ways that even abstract statistical facts are conditioned by the political landscapes where they are produced. It examines Israeli population maps in the years 1967-1995. After 1967, the close geographic proximity of Palestinians posed a challenge to the policy of not indicating the borders of the Palestinian Territories on Israeli maps. Roberto Bachi, the director of the Israeli population census, sought to address this challenge while also helping to turn cartography into an international science. As a result, he led the census away from mapping shaded areas of uniform population, and towards dot maps of population distribution. Such efforts served to limit the calculation methods at the census cartographers’ disposal. They also revealed that, despite repeated claims by Israeli politicians that Palestinians did not exist, in fact the Israeli cartographers’ methods were inherently shaped by the presence of large numbers of Palestinians in the region. For even though Palestinian areas were intentionally left blank, the resulting gaps actually made them show up on the map. So through their exclusion, Palestinian populations were made visible, and this is one way that the landscapes of the occupation have shaped Israeli maps.

Keywords:   Maps, Census, Population statistics, Palestine and Israel, Borders, Calculation methods, Visibility, International technoscience, Israeli occupation

Making these experiments sometimes made me feel like a kindergarten boy playing with colored papers, and afforded boundless amusement for my grandchildren.

Roberto Bachi, Graphical Rational Patterns

It was decided to carry out the enumeration from house to house under curfew. … The enumerator marked the doors of the houses enumerated with chalk to ensure an orderly and complete coverage.

CBS, West Bank of the Jordan, Gaza Strip, and Northern Sinai, Golan Heights

The Methodology of Curfew

The quotations above serve to illustrate the intimate ties between population control and the development of statistical cartography. They date from the beginning of the Israeli occupation of the West Bank and Gaza Strip in 1967. The population census described in the second quote began only a little over two months after Jordanian and Israeli troops had ceased fighting street by street through the Old City of Jerusalem. The author of the first quotation, Bachi, was the head of the Israeli CBS, which was in charge of the enumeration, and he was at the height of his career. Bachi remained a key architect of the Israeli census for over twenty years, and was praised in diverse corners of the international scientific community for his innovative work in statistical cartography and geostatistics (e.g., “In Memory of Roberto Bachi” 1996). The quotation captures Bachi’s romantic depiction of his academic research into his method of Graphical Rational Patterns (GRP), which were a new set of symbols for depicting numbers in a precise way on maps (see figure 3.1). The “colored papers” he refers to were transparent stickers or transfers that were applied when making a map, to (p.82)

Removing Borders, Erasing Palestinians: Israeli Population Maps after 1967

Figure 3.1 An example of Bachi’s (1989, 339) GRPs: A chart of the symbols for the numbers from one to one hundred.

Reprinted with permission.

prevent the cartographer from having to draw every symbol by hand. The second quotation comes from the official report of the 1967 census of the Palestinian Territories, carried out by the CBS that Bachi both founded and directed.1 It describes the curfew imposed on the Palestinians, who were forced into their homes, or the home they happened to be closest to at the moment, their doors and walls marked with chalk, so that the count might be considered “orderly and complete.”

There is a clear symbiosis between the colored papers and curfew. In the first instance, Bachi depicts himself cutting into the colored sheets, using glue to stick them into place on maps. In the second, Bachi is at the head of an army of enumerators, supported by the actual military (CBS 1967, iii), which physically marches through the streets, metaphorically cutting up groups of people and sending them back inside, fixing them in place, (p.83) marking them with chalk in order to obtain a clear picture. This confluence of seemingly innocuous snips of paper, on the one hand, and the violence of putting entire regions under house arrest, on the other hand, gives an idea of why the national census, or systematic population count, resonates across such a wide array of research areas. The census encapsulates so many major themes of contemporary social theory: the falsity of the boundary between politics and technoscience; the creep of big data and minute slicing up of individuals and communities into predetermined check boxes like “age” and “nationality.”

In many countries, the census was one of the main points of interaction between the masses and governments that sought to both discipline and depict them in ways that made sense to bureaucrats. In this context, Bachi’s seeming glibness about his paper cutouts provides a cautionary note for studies of the census that do not take into account the ordinary and extraordinary coercions of quantification. For the violence of a census under military occupation is not so far distant from the context of the everyday work of national censuses more broadly. This chapter explores these connections between classification and sorting in statistics along with their forceful implications and applications on the ground. In the process, it shows computers in a new light, offering insight into the apparently seamless digital maps that nonetheless expertly, if imperfectly, conceal their connections to occupation and war. It therefore analyzes the ways that Israeli population cartography incorporated Palestinians through the very act of erasing them.

The relationship between state control and academic research has often been conceived in terms of the interplay of representation and materiality in shaping the landscape (Harvey 2005; Low and Lawrence-Zúñiga 2003; Smith 2008). In practice, however, the focus has been primarily on representation (e.g., Mitchell 1991).2 In this chapter, instead of emphasizing how representations of space can shape the space, I will investigate the ways that the geographies that were produced through the Israeli census in turn affected subsequent depictions of the land. More specifically, the continued physical presence of subjugated populations—a presence that was acknowledged through the need to conduct a census in the first place—influenced Bachi’s theoretical work in statistical cartography. Bachi claimed his findings were scientific, and that he concentrated not on political issues but rather on obtaining an objective, quantitative view of Israel and the (p.84) Palestinian Territories. Yet even Bachi’s most abstruse equations were fully embedded in the cultural and material landscapes where he worked.

The Census as Western Science

Bachi had been a successful professor of statistics in Italy before the fascist government’s racial purity laws forced him to flee prior to World War II. Over the course of his career, Bachi was active either in government, academia, or independent research, throughout nearly every major political transition in the region in the second half of the twentieth century, from before the founding of the state of Israel in 1948 until just before the official signing of the second Oslo Accords in 1995. From the beginning, he insisted that the Israeli census be firmly rooted in Western science. The 1948 census was carried out while the war that followed the founding of Israel still raged around them. But even then Bachi contended that the census would be conducted according to rigorous and objective statistical principles (Bachi et al. 1955; Leibler and Breslau 2005). Later, as the director of the Israeli CBS, he continuously argued that military concerns should not be allowed to dominate over scientific rigor (Bachi 1981; Leibler and Breslau 2005).

Throughout this process, he looked to Europe and North America for models for his scientific work. In the planning stages of the Israeli censuses of the 1950s and 1960s, Bachi and his colleagues requested numerous census documents from countries such as Canada, France, and Spain—and coincidentally, from Iran, which at that time was still under the control of the shah, who was politically supported by governments in Europe and North America (CBS 1969b). Bachi was a main contributor to both the Statistical Atlas of Italy (Bachi 1999) and the Atlas of Israel (SOI 1970), and spoke at innumerable international conferences (e.g., Bachi 1955, 1962a, 1962b, 1974b, 1975, 1989).

Bachi’s repeated attempts to present his research as objective and neutral are evidenced by the fact that under his leadership, the CBS fought to compile data “solely according to professional considerations, without interference from political quarters” (Schmelz and Gad 1986, xii). Yet Bachi’s more politicized work counting Palestinians nonetheless exerted a strong influence on his abstract research, and did so in ways that were not wholly incompatible with his aim of developing statistical cartography as a science. As a result, Bachi’s work was both thoroughly scientific and (p.85) thoroughly colonial at the same time. It therefore demonstrates how empirical science and settler-colonial practices can coincide.

Indeed, Bachi’s academic context was also related to his governmental efforts. In addition to running the census, Bachi founded the Department of Statistics at the Hebrew University of Israel, thereby further linking academic statistics to Israeli state objectives. His two major academic books were published in tandem with census milestones: the first, in 1968, immediately after the census of the Occupied Territories discussed above, and the second, after his death in 1995, following the earliest census in Israel to fully incorporate quantitative digital mapmaking of the kind that Bachi had long advocated (Bachi 1999, ix; “In Memory of Roberto Bachi” 1996, 13–14; Schmelz and Gad 1986, xi–xiv). As I examine in detail below, both books address key issues of the census and develop methodologies first implemented in the enumeration of the OPT. As such, the theoretical, academic, and governmental areas of Bachi’s professional career were thoroughly interlinked, and mutually reinforced one another.

The relationship between Bachi’s governmental and theoretical work was also affected by the advent of digital cartographic technology. Throughout his career, Bachi took part in a technological transition in cartography from a professionalized trade that relied on hand-drafting skills into a quantified, mechanized science that became heavily dependent on computers. Bachi himself was trained within a modernist tradition of statistical cartography in Italy. He was instrumental in bringing the paradigm to the nascent Israeli state, where it was further developed in part due to his personal efforts (Schmelz and Gad 1986). Additionally, later methods enabled the detailed management of piecemeal territories—a type of complexity that is considered a hallmark of postmodern or late modern digital governance.

Late Modernism and the Influence of Landscape

Computers are often presented as a break or revolution in the history of technology, an abrupt disjuncture between the modern and postmodern. However, Bachi’s maps serve to demonstrate that this is not an all-or-nothing transition. Rather, Bachi chose to further key modernist goals in formulating cartographic knowledge, such as accuracy and the continuity of borders, and this lent continuity to the maps he made both before and after computers became commonplace. He also actively rejected other goals (p.86) considered to be hallmarks of modern statistics, like completeness and the consistency of space. So his later studies arguably are evidence of a new form of modernism that challenged the fundamental goals of completeness and consistency, but nevertheless it is one that still would be accepted and recognized among international scientific communities.

Moreover, Bachi’s research illustrates how in the context of modernization processes, the landscapes that are produced can transform the goals of modernization. The efforts to enforce a single set of national borders for Israel instead spiraled into the construction of tightly linked networks of bounded settlements in the West Bank. In time, the very process of determining their borders would transform the aims and purpose of Bachi’s theoretical work. Bachi largely cooperated with broader efforts to literally and figuratively expunge Palestinians from the land. Yet he could not extricate himself from the social and material connections afforded by his physical proximity to Palestinian communities. As a result, years of occupation ultimately fed back into his statistical cartography, shaping both Bachi’s research trajectory and, through him, the very fabric of the digital canvas on which population statistics in the region continue to be mapped.

In what follows, I analyze key cartographic methodologies that Bachi developed, with a focus on his two major theoretical books on statistical visualization that appeared in 1968 and 1999—the latter posthumously from work completed in 1995. I show how each book was influenced, even at its most quantitative and theoretical levels, by the very presence of Palestinian populations in the landscape—an influence felt in terms of the concerns that arose while counting populations in the West Bank. It also was inflected through central political and demographic arguments of each book’s respective era: in 1968, through the charged claim that Palestinians did not exist, and second, in 1995, by way of the ongoing debates that came to be known as the demographic war. I first introduce each book and the census that predated it (in 1967 and 1995, respectively) in the context of their related political debates. I then delineate the development of Bachi’s theoretical claims in relation to the relevant census methodologies. I proceed by examining each book in the context of Bachi’s work as a whole, pinpointing his simultaneous retention of modernist cartographic ideals such as accuracy and continuity, and rejection of others, like completeness and consistency (Haraway 1988; Livingstone 1992; Monmonier 1991). In the conclusion to this chapter, I return to the role of computerization in (p.87) Bachi’s efforts to innovate while maintaining international scientific standards.

By analyzing Bachi’s research in the theory of statistical maps in 1967 and 1995, this chapter encompasses the entire period of direct and full Israeli control of the West Bank and Gaza Strip, until the beginning of limited management on the part of the PA. This period also coincides with the early history of digital cartography. During these decades, even as Bachi sought to cement the transformation of Israeli statistical cartography into an international science, he himself was thoroughly situated in local and regional as well as international landscapes.

The Critical Cartography of Census History

National censuses are a prime means by which state bureaucrats collect information about the people under their jurisdiction. The census has played a central role in linking academic knowledge to the management of human mobility in Palestine and Israel (Zureik 2001), and cartography is crucial to the use of the census in surveillance and control (Crampton 2003).3 Foucault’s call to study geographies of power has sparked a body of research into the ways that colonial geographies were integral to modernity’s core (Stoler 1995). Yet this work has only just begun to be applied to the critique of quantitative geographic knowledge (Crampton and Elden 2007). While the central role of census mapmaking is often acknowledged in the academic literature (Anderson 2006; Crampton 2003; Edney 1997), statistical cartography’s technical aspects have less often been treated in detail (exceptions include Hannah 2001, 2009; Mood 1946; Pavlovskaya and Bier 2012).

Even so, the census doesn’t simply take place within predetermined national borders. Instead, it forms part of the labor necessary to constitute the nation on the ground (Anderson 2006; Leibler 2004, 2007; Leibler and Breslau 2005; Zureik 2001). In studies of the census in postcolonial contexts, modern conceptions of space have taken more of a pivotal role due to the well-documented relationship between cartography and imperialism (Cosgrove 2008; Edney 1997; Godlewska and Smith 1994; Gregory 1994; Harley 1989; Kalpagam 2000; Stone 1988; Turnbull 2000; Winichakul 1994) as well as the links between the spread of technologies and bureaucratic systems for managing populations (p.88) (Bektas 2000; Feldman 2008; Hull 2003; Leibler and Breslau 2005; Mrázek 2002; Zureik, Lyon, and Abu-Laban 2011). But to date less critical attention has been paid to methods for conceptualizing census boundaries and the spatial epistemologies that are implicit in such definitions. There are notable exceptions, including the literature on gerrymandering in statistical districts more broadly (Bunge 1966; Cranor, Crawley, and Scheele 1989; Mood 1946; Sauer 1918; Sherstyuk 1998). In addition, Elia Zureik (2001, 227) has analyzed how borders have been shaped by “a Palestinian–Israeli dialectic of state construction” framed by asymmetrical power relations—although Zureik emphasizes continuities in contestation over the transformations in maps.4

International Hierarchies of Scientific Knowledge

Just as those in power have attempted to define and manage populations through the census, their own subjectivities have also been influenced by the geographic contexts that they operate within. Bachi certainly served as a privileged denizen of dominant statistics in Israel, but his international role was complex. He benefited from international hierarchies because unlike Palestinian statisticians, he came from a nation that was widely acknowledged as a nation. As discussed in chapter 2, Israel’s dominant culture was increasingly viewed, within a global context of racism and Orientalism, as racially white and culturally European. But as a Jewish scientist from a small nation, Bachi was constrained by these same hierarchies, which favored large and established states, and also operated in a global context of anti-Semitism.5 Thus, international asymmetries would intricately shape the ways that Bachi could innovate while still being accepted as a scientist. Such recognition was crucial to participating in international debates as well as being viewed as a theorist whose methods were credible and whose findings could be accepted as objectively true.6 With this in mind, Bachi’s insistence that the census be scientific—namely, that it be conducted not according to political dictates but rather in line with the requirements of rationality (Leibler and Breslau 2005)—was not only the outcome of personal preference or beliefs. Instead, it can also be viewed as part of an attempt to fit within an international academic hierarchy whose members often had rigid, if implicit, notions of belonging (Porter and Ross 2003; Wagner 2001).

These notions indicate one additional way that Bachi’s work was constrained by international hierarchies. Bachi appears to have embraced the (p.89) dominant paradigms in demographics wholeheartedly. Still, his attempts to demonstrate his scientific legitimacy are also the logical outcome of the widespread belief that there is only one true way to do science (Smith and Marx 1984; Wyatt 2008b). It has long been held that the power of science and technology is due precisely to the fact that scientific discoveries are true everywhere that they apply, independent of social or cultural influence. Yet this view masks the incredible local variation in science and its findings. Furthermore, it has important implications in terms of the international landscapes of science and technology, because it includes an assumption that the “right” way to do technoscience is that practiced in Europe and North America.

There is an implicit logic to this rather-illogical assumption, and it reads as follows: if there is one right way to do technoscience, and that right version is the one practiced in Europe and North America, then developments that do not conform to Western technoscience are unscientific by definition. For this reason, it would be imperative for Bachi and other Israeli academics to demonstrate their willingness to conform to dominant conceptions of science and technology in order for it to be recognized internationally that their efforts indeed were scientific. Even as he used methods that were internationally recognized to assist in controlling Palestinian populations, then, the fact that Bachi was not working solely in Europe and North America would have set limits on the scope of his potential innovations.

The Quantitative Revolution: Geography as a Statistical Science

This raises the question: what was this interdisciplinary paradigm to which Bachi strove to adhere? For starters, it was increasingly mathematical and digital. Bachi sought to computerize his cartographic methods early on, starting in the 1950s. His role as an early adapter of digital mapmaking involved countering the historic perception of the discipline. At the turn of the twentieth century, geography, including the subdiscipline of cartography, had been considered a field comparable to history. While attempts previously had been made to establish it as a science (Godlewska 1999), it was widely believed that geographers gave descriptive, analytic accounts of particular landscapes (Livingstone 1992).

This changed after World War II, as a new generation of geographers worked to transform geography into a statistical science—a movement that (p.90) became known as the “quantitative revolution” (Barnes 2001). Due to their efforts, by the 1960s mapmaking was no longer viewed internationally as a descriptive spatial record of territories. Instead, as noted in chapter 2, it was seen as a scientific method for using a Cartesian grid to display variations in statistical quantities (Godlewska 1999; Livingstone 1992).7 So by the 1960s, although Bachi himself was quite sensitive to the specifics of geographic mapmaking, he tellingly uses the words graph and map almost interchangeably. This is in keeping with the conventions of quantitative cartography, where maps are merely one type of a broader category that includes graphs of mathematical functions.

As a contributor to this international quantitative revolution, Bachi made seamless transitions back and forth between population statistics and geostatistical theory. Thus he put mathematics to work, via technology, in the name of modern empirical science. In Bachi’s (1968, 1–2) case, his stress on clarity and transparency is apparent in his exhortation that GRPs would make graphing “quick and easy,” and would avoid types of mapmaking that are “inaccurate or even misleading.” By framing his efforts in this way, Bachi helped early on to bring the insights of geography’s quantitative revolution to Israel, and serve as a link between Israel and international networks of technoscience. But since he was an innovator as well as transmitter of this new methodology, he challenged the assumption that the quantitative revolution, or indeed the concomitant development of GIS mapmaking, took place primarily in Europe and North America. For comparison, although the US census had been developing computer cartography for the purposes of census enumeration since the late 1960s, the first GIS files were only created for the 1990 census, a mere five years before Israel fully computerized its population counts.8 As early as the mid-1950s, while the quantitative revolution was just starting up in Europe and the United States, Bachi (1956) was already pursuing similar avenues of research in Israel.

The transformations in Bachi’s maps also illustrate how the quantitative revolution was not wholly a revolution. Instead, there were significant consistencies in cartographic methods before and after the advent of computers. Bachi’s work highlights the specifics of this process, for although he might not have been able to challenge the paradigm wholesale, he nonetheless was able to innovate within its bounds. He navigated the alternating process of change and conformity to international dictates. Bachi (p.91) adapted existing statistical methods so that they might become practically useful in the contexts, like the census, in which he worked. This process would transform Bachi as well. Through the ultimate ambivalence in his depictions of statistics for Palestinians, the Palestinian struggle for selfdetermination would also leave its mark on his research and cartography more broadly.

Locating Existence under Occupation, 1968

The outcome of a conflict between two parties usually doesn’t hinge on a debate over whether or not one of them exists. But this is precisely the way that some Zionist groups have attempted to frame debates over the right of Palestinians to live in and enter Israel and the Occupied Territories. Prime Minister Golda Meir’s famous proclamation in 1969 that, “There never was such a thing as Palestinians. … They did not exist” (“Golda Meir Scorns Soviets” 1969), has been reiterated over countless election cycles both within Israel and abroad.9 Meir’s and others’ challenges to Palestinian identity have been skillfully rebutted at length (Bishara 2003; Kanaaneh and Nusair 2010; Khalidi 2006, 2010; Massad 2001, 2003, 2006; Seikaly 1995). Yet her comments deserve further scrutiny for how they relate to contemporary debates.

In the longer piece from which the quotation is taken, Meir bases her argument on principles of statehood, claiming, “When was there an independent Palestinian people with a Palestinian state?” By suggesting that Palestinians don’t exist because they have been prevented from building a nation, Meir in effect uses the language of social constructivism against itself. As suggested by the title of Imagined Communities, however, Anderson’s (2006) foundational book on the nation, although imagination might be necessary for the formation of a nation, formal legal recognition is not. So, few contemporary researchers would join her in claiming that nationhood is what brings those people into existence.10

Nonetheless, Meir’s assertion has long been used for instrumental reasons, and its power in this respect shows the political danger of severing the social from the material, as noted in chapter 2. Given that the social has been the focus of much recent work, in this chapter I concentrate strategically on the more material aspects of the Palestinian presence in the land and their influence. So instead of emphasizing how the social shapes the (p.92) material, this chapter shows how material landscapes, which themselves are imbued with the social, also feed back into social and scientific knowledge. Indeed, through Bachi, they would come to have a decisive effect on Israeli population maps.

Palestinian Existence and the Israeli Census

Irrespective of any claims to Palestinian nonexistence, from the perspective of the Israeli military administrators who took over the Palestinian Territories in 1967, the Palestinians existed and needed to be counted. The first steps were to devise population categories, count people according to them, and note the precise locations of every individual. This is precisely what the CBS set out to do. But the population count was not only aimed at the surveillance of those in the OPT. If the push to conduct a census after the 1967 war demonstrates a tacit acknowledgment of Palestinians’ existence, the census also aimed to limit the number of Palestinians who were allowed to remain. As Anat Leibler has shown (2004, 2007; Leibler and Breslau 2005), one of the primary motivations for conducting the census so quickly was to prevent those who had fled during the conflict from returning. The census was the basis for issuing identity cards that allowed their bearers to reside permanently in the Palestinian Territories, if not to become citizens. So by performing the census early, the administrators prevented those who were away from gaining the right to come back.

As a result, the census could be said to have two potentially conflicting priorities: on the one hand, to gain an accurate count of the populations now under Israeli control; on the other hand, to exclude as many people as possible in order for fewer Palestinians to be able to claim residency. Since the population of Israel was produced in part through the census, it was crucial for census takers to preemptively make as many Palestinians as possible uncountable, and therefore invisible and, for national purposes, nonexistent. For as John Law, Evelyn Ruppert, and Mike Savage (2011, 8) have pointed out, as far as any particular method is concerned, “that which is invisible for all intents and purposes doesn’t exist.” So it was in the best interest of the census takers both to rigorously count Palestinians who were there and shape the population by excluding Palestinians before anyone was ever counted.

In this context, the census authorities were especially wary of being charged with undercounting. They consequently stress the extreme lengths (p.93) to which census enumerators went to obtain precise enumerations. Their report notes that “despite the use of special vehicles (and even donkeys), the enumerators could not reach isolated houses or distant localities (especially nomads’ tents), because of difficulties of access or danger of mines” (CBS 1967, xxx–xxxii). In so doing, they highlight some of the ways that as a result of enduring political realities, much of the West Bank became practically inaccessible to the Israelis in the immediate aftermath of the war. So just as the very presence of Palestinians made the 1967 counting necessary, in the eyes of Israeli officials, the ongoing geographic impacts of 1948 and 1967 circumscribed their ability to conduct that census. The census in turn would affect the types of maps that Bachi used and advocated in his 1968 book. His theoretical academic work is thus emblematic of the census maps that had tremendous practical power over Palestinians’ lives.

Bachi played a key role in the way the 1967 census was conducted, and his name appears on the official report both as the “director of the census” and “government statistician” (CBS 1967). In the process, however, he contradicted one of his own judgments from 1948. Back then, against members of the Israeli military who had wanted to count Palestinians with different methods from those used by Israelis, Bachi argued forcefully that the methodology should remain the same for both groups. He claimed that only with consistent methods would the results be seen to be statistically rigorous (Leibler and Breslau 2005). The 1948 census reports also described the hesitancy to enumerate Jews via a curfew, given that this method had been widely used by the British during their occupation after World War I, and thereby would have brought back traumatic memories (Bachi et al. 1955). Yet by 1967, under Bachi’s direction, the census did precisely this for Palestinians—a process undertaken for the OPT alone. Bachi oversaw the one-day curfew with the stated reasoning that indiscriminately confining people to their homes would improve the chances of counting as many people as possible.

So the curfew was justified using arguments for accuracy, but it had its own consequences in terms of the accuracy of the statistics. The curfew indeed excluded refugees who were missing, away from home, or homeless due to the war. It also, by the census takers’ own omission, had the effect of creating “differences between the locality in which [inhabitants] were registered and the permanent places of residence” (CBS 1967, xxx–xxxii). People were counted according to where they happened to be under the curfew, (p.94) not necessarily where they primarily lived. The resulting census data for the Palestinian Territories would have presented a logical conundrum for Bachi. As someone who considered himself a scientist, to Bachi the Palestinian census data would have seemed less than rigorous—seeing as how they were collected under restrictive conditions that made them only partially comparable to census data for the fully annexed areas of the state of Israel. Yet also as a scientist, he could not completely ignore that the data existed—for example, by placing labels that read “no data available” on relevant areas of his maps of Israel. After all, he spearheaded the operation that collected the data in the first place. So on scientific maps, the question for Bachi would have become: how is it possible to best map Israel while neither denying that data exist for Palestinians, nor actually including that (only semirigorous) census data on the map?

The presence of the Green Line boundary between Israel and those Palestinian territories that were newly occupied compounded the difficulty of how and to what extent to acknowledge the existence of data for Palestinians. As recounted in more detail in chapter 4, the Green Line was first drawn during the cease-fire agreements of 1949, and in places between 1949 and 1967, it served as a highly militarized border between central Israel and the (Jordanian-controlled) West Bank. While Jerusalem was rigidly divided, in some areas the separation was inconsistent, and border zones, including those where homes had been demolished in the process of marking the border, were purposefully filled in with Israeli settlements. As a result, by the late 1950s, Hadawi (1957, 1) could claim that in many areas, “the Armistice Lines have not until this date been demarcated on the ground.” As discussed in chapter 1, from 1967 onward even the official policy of enforcing the border was reversed. The agencies of the state of Israel claimed that the Green Line was no longer a valid international border. Instead of referring to the military occupation of the Palestinian Territories, academics and government agents euphemistically proposed that Israel and the OPT had been “reunited.” In keeping with this claim of reunion, the Green Line no longer appeared on most state maps (Benvenisti 1984; Gorenberg 2012; Shehadeh 2007, 178).11

The official omission of the Green Line border after 1967 was in contrast to the situation at ground level, given that the Palestinian Territories occupied in 1967 have not been formally annexed by the Israeli state—with the notable exception of East Jerusalem. The administrative, military, (p.95) and economic policy discrepancies between annexed Israel and the un-annexed OPT were and are extreme (Benvenisti 2000). The separate-and-unequal policies of the 1967 census is emblematic of these injustices.12 As a result of the Palestinians’ existence and ongoing struggle, for most census purposes, it was the Green Line, not the official state line, that served as a boundary between Israel and the Palestinian Territories occupied in 1967. This circumscribed the collection of data for the majority of the Israeli population.

The resulting inconsistencies in mapping the Green Line are borne out in Bachi’s (1968) Graphical Rational Patterns. The aim of his book is to introduce the eponymous GRPs as a standardized set of symbols for displaying numbers on maps. Maps often indicate numerical values with circles of varying sizes, but the different size of the circles can be difficult to judge visually. By constructing his elaborate symbology of GRPs, Bachi hoped that anyone familiar with GRPs would be able to tell the exact numbers the cartographer sought to display (see figure 3.1). Through GRPs, though, Bachi also started to move away from rigid notions of consistency, which would have required him to display all internal borders including the Green Line, in favor of greater accuracy for a few selected points. Perhaps not surprisingly, the points he selected to display most often represented Jewish Israelis.

In fact, despite directing and overseeing the census of the OPT, Bachi rarely, if ever, used the data from that census in his work. To omit the data without appearing to be unscientific, Bachi increasingly began using one type of map from among two commonly available options. As is evident in his 1968 book, for mapping census data collected from 1967 onward, Bachi started to favor graduated circles more and more. He became increasingly critical of the shaded area technique called choropleth, which was more commonly used at the time. Choropleths are maps in which districts, states, or regions are progressively shaded darker or lighter in order to represent increased percentages of some particular characteristic, such as the average number of people in each household for specific areas (see figure 3.2a). Choropleths are convenient because unlike graduated circle maps, they allow cartographers to indicate statistical data without an abundance of different symbols. At the same time, they can be misleading, and there are several problems that arise. For instance, because different regions are shaded in, those with larger geographic areas often stand out as (p.96) more significant than they would otherwise appear to be, if they were judged by the statistics alone—and this is only one of multiple layers of complexity.

In contrast to choropleths, graduated circle maps (see figure 3.3) are those, as described earlier, that use shapes of different sizes, generally circles, to represent a particular statistic. The difference between choropleths and graduated circles are not obvious at first, and both are widely used. But graduated circle maps have one advantage that relates to Bachi’s research: they allow for the omission of certain boundaries, including the Green Line. For instead of shading an entire area or subregion, the graduated circle is located with respect to one single point on the map. This allows for greater precision in depicting—or choosing arbitrarily, since a random point could be used—the precise locations of populations within national boundaries. The result, however, is that in graduated circle maps, the actual boundaries of those districts do not have to be included on the map. In contrast, on choropleths, if each district is shaded in, then the edges of those shaded regions—their boundaries—are already implicitly indicated (figure 3.2a).

Bachi’s 1968 Book: From Choropleths to Graduated Circles

The trend toward graduated circles and omitting the Green Line can be seen progressively over the course of Bachi’s work. The standard practice for graduated circle maps is still to indicate boundaries, even though it’s not practically necessary. Yet as Bachi moved increasingly toward graduated circles, the Green Line appears less frequently on his maps. In Bachi’s (1968, 196–197, 227, plates viii and x) Graphical Rational Patterns book in particular, several maps of other regions or earlier data indicate the internal boundaries of the area being mapped. By contrast, nearly every single map of the region that depicts data collected in 1967 or later uses graduated symbols as opposed to choropleths (e.g., Bachi 1999). Similarly, prior to 1967, his graduated circle maps rigorously displayed the Green Line (e.g., Bachi 1962b). Then, in the first years following 1967, the boundary was irregularly omitted and displayed, seemingly without respect to whether the data displayed included those for the West Bank or not (Bachi 1974a), and over time it was dropped entirely.

Bachi’s increasingly frequent omission of the Green Line was coupled with the disappearance of maps that show broader views of the region, and (p.97) an increasing substitution of maps of other countries for those of Palestine and Israel. Throughout the 1970s, Bachi moved toward only mapping Israel and Palestine at a finer scale, with a focus on Jerusalem. During this period, he also expanded his use of choropleths of Italy and the United States to substitute for the absent maps of broader Israel and Palestine (e.g., Bachi 1975). But for these maps, too, he eventually came to only use graduated circles, with the result that even the internal borders of Italy and the United States would come to be alternately included and excluded from them (e.g., Bachi 1999).

In his 1968 book, Bachi still advocates the use of choropleths in combination with GRPs. He describes several methods of doing so, thereby showing that choropleths were not technically or rationally incompatible with his GRP (Bachi 1968, 198–216). Bachi nevertheless criticizes choropleths in the same work. He claims that when using choropleths to represent percentages of population by subregions or provinces, “the distortion resulting from this method may be extremely dangerous” (ibid., 205). Later, comparing a choropleth and graduated circle map with GRPs for the same areas, he notes that “the visual impression obtained by the two graphs [maps] is completely different and almost opposite.” The graduated circle map “enables us to receive the correct impression” while the choropleth “may thus fail almost completely to convey an accurate view of the distribution under survey; Moreover, comparison of data for each region and of the national average may create an impression of discrepancy.” He argues that this discrepancy is not present in graduated circle maps that use GRPs (ibid., 210–213, emphasis added). The practical result of these convictions was that Bachi’s maps were both increasingly accurate, by pinpointing values via graduated circles and GRP symbols, yet increasingly inconsistent in terms of his depiction of the Green Line.

In his 1999 book, the Green Line is absent from every one of Bachi’s relevant maps. This includes both his multiplicitous maps of the Jerusalem metropolitan area (which straddles the Green Line) as well as his sparse maps of Palestine and Israel more broadly (Bachi 1999, 46, 104, 149, plates 2 and 3). But even when the Green Line is not indicated, a ghost of the political border can be seen on maps where areas of high Palestinian concentration, depicted as blank areas, in effect sketch out the boundary as a type of palimpsest (see figure 3.3). Choropleths were not entirely absent from Israeli census publications after 1967. Although the Green Line was (p.98)

Removing Borders, Erasing Palestinians: Israeli Population Maps after 1967Removing Borders, Erasing Palestinians: Israeli Population Maps after 1967

Figures 3.2a and 3.2b At left is a sample choropleth map from the 1961 Israeli census. At right is a map that shows the extent of the development of an Israeli national GIS database as of 1996. In figure 3.2a (CBS 1963), the shaded areas indicate the average number of people in each household in that subdistrict, called a “natural region,” as defined by the CBS. Because the map was made prior to 1967, data for the West Bank and Gaza Strip are not included.

Notice how the use of shading requires the cartographer to display some internal boundaries—at the very least, between the areas that are different shades. Here the Green Line is the boundary of the West Bank and Gaza Strip, indicated in black. Yet even if the Green Line were not indicated, the boundaries of the Israeli districts shown fall along it. So simply shading those districts gives an impression of the Green Line. Figure 3.2b demonstrates how, in part but not solely due to Bachi’s legacy, GIS data in Israel were developed using an overhead view that seamlessly incorporated all of the Palestinian Territories (Peled 1996). This was likely accomplished precisely with the expectation that Israel would lose the ability to collect data on the ground throughout the entire Palestinian Territories due to the handover of limited sovereignty to the Palestinians following the Oslo Accords. It contrasts, but also coincides, with the shift to data that included only Israeli areas within the Palestinian Territories rather than the Palestinian Territories as a whole (figure 3.4).

Reprinted with permission. In figure 3.2b, the shaded regions indicate those areas where digital map data existed or was in progress. The SOI appears to have prioritized the West Bank and Gaza Strip, which both fall under the “Mapped” category—namely, those areas that were converted first to GIS from paper maps.

(p.99) (p.100)

Removing Borders, Erasing Palestinians: Israeli Population Maps after 1967

Figure 3.3 A later map of Bachi’s, comparing graduated circles (left) to GRPs (right). The Green Line and any indication of the PA are omitted. Yet even though Palestinian cities and towns are not shown, a blurry outline of the West Bank can still be seen. This is because the omitted Palestinian towns are concentrated in the West Bank, given that many Palestinians are not permitted to live in central Israel (see Bachi 1999, 165).

Reprinted with the permission of Springer and Kluwer Academic /Plenum Publishers, New Methods of Geostatistical Analysis and Graphical Presentation: Distributions of Populations over Territories, chap. 6, “Graphical Rational Patterns,” 165, Roberto Bachi © 1999.

(p.101) often left off its population maps, the CBS did continue to use choropleths sparingly on technical maps that depicted census boundaries. These were intended primarily for internal use, and were aimed at Israeli and/or Jewish statisticians (Bachi 1974a; CBS 1969b, 1985). Yet unlike his earlier work (Bachi et al. 1955), by the late 1970s Bachi was no longer routinely using CBS choropleths, even though technically such maps were still available.

Over time, as Bachi came to avoid indicating the Green Line on his statistical maps, this also limited his options in terms of data visualization and display. As discussed above, he justified these limitations on scientific grounds, while perhaps showing some uneasiness. The omissions, however, conveniently fit the practice among state Israeli cartographers of no longer depicting that border. So instead of having two common options available to him for mapmaking, both choropleth and graduated circle maps, Bachi was restricted to using only graduated circle maps. This restriction had significant implications because it dictated which statistics might be visualized. Moreover, it affected the entire course of Bachi’s subsequent research. It is perhaps then not surprising that his GRP symbology consists precisely of an innovation almost exclusively in the area of graduated circle cartography.

If by creating GRPs, Bachi was attempting to construct a more refined system of statistical visualization, then it was obtained at the price of the consistency in the representation of regional territorial borders. In addition to addressing contemporary concerns with the visualization of statistical data, by grappling with the Green Line, Bachi’s work is representative of how the existence of the Palestinians and the occupation more broadly shaped the key methodological issues that Israeli cartographers faced in the unique post-1967 context. But Bachi’s efforts with graduated circles and points didn’t require him to give up mapping borders altogether. Instead, by 1999, amid Israeli settlement construction in the context of the ongoing demographic war, Bachi would begin using points to map entirely new boundaries of his own.

The War of Scattered Boundaries, 1999

If the 1967 census represented tacit acknowledgment of the existence of significant populations in the newly occupied territories, the 1995 census (p.102) showed the persistence of the demographic war—a competition between groups, whose members are encouraged to bear ever more children in order to secure greater political power (Courbage 1999; Kanaaneh 2002). It is the obstetric expression of the adage “strength in numbers.” On the Israeli side, its proponents peddle fears that Palestinians outnumber Israelis. They compile evidence that European-origin Israelis are a minority in Palestine and Israel, which frightens those concerned with maintaining Israel’s status as a Europeanized nation. They thereby highlight the currents of racism and discourses of racial purity in the conflict (Massad 2003).

Bachi himself was a central actor in demographic debates and is credited as being one of the earliest scholars to warn of (allegedly) impending Jewish population decline around the world (“In Memory of Roberto Bachi” 1996; Schmelz and Gad 1986). A consideration of Bachi’s later work, in light of the 1995 census’s attempts to quantify the outcome of almost thirty years of demographic struggles, provides a fitting counterpart to the analysis of the claim of Palestinian nonexistence that was explored above. From one perspective, the demographic war is the flip side of the assertion that Palestinians did not exist. While the existence claims mistakenly suggest that there were no groups present that (legitimately) identified as Palestinian, in contrast, the rhetoric of the demographic war argues that Palestinian numbers are so high that they threaten to “engulf” (non-Palestinian) Israelis. The current prime minister, Benjamin Netanyahu, has even referred to a “demographic bomb” (Munayyer 2012). Seen from another perspective, however, the demographic war and existence claims have much in common. For in statistical terms, the question of whether a group exists is often translated to a question of whether that group has a high enough population to be considered statistically significant. And the census is precisely the mechanism to determine which groups have numerically significant populations. So both the question of whether a group exists, along with the question of which group has or will have the largest population are determined by the census.

Israeli Settlements in the 1995 Census

As with the existence claims, then, what is striking about the demographic debates are the interlinkages between governance methods, social discourse, and the material bodies of a population. For although Meir’s statement that Palestinians didn’t exist, as noted earlier, was framed in historical (p.103) terms it could also be considered an exercise in prophesy. It is only possible to claim that Palestinians didn’t exist in a situation where many Palestinians had been violently expunged from the landscape. To turn it around, Meir’s claim about Palestinians nonexistence in the past was more a prediction that she would not let them build a nation of Palestine in the future. Despite this, on Bachi’s later graduated circle maps, the erased Palestinian populations are so numerous that they show up precisely through their absence, as a white, blank area. For example, in figure 3.3, the Green Line and any indication of the PA are omitted. Even though Palestinian cities and towns are not shown, a blurry outline of the West Bank can still be seen. This is because the omitted Palestinian towns are concentrated in the West Bank, given that many Palestinians are not permitted to live in central Israel.

So after World War II, there were organized and concerted efforts to manufacture the “fact on the ground” that Palestinians weren’t a numerically significant population by preventing those who were pushed out from returning (Leibler and Breslau 2005; Morris 2004), and attempting to ensure that they would not continue to exist in years to come. Yet Meir’s claim was bolstered by further action. Zionist leaders worked to bring Jewish families to Israel, in part to counter any potential claims that Jews themselves did not comprise a large population in the region. They encouraged emigration from around the world. They also transported large numbers of Jewish families to Israel from elsewhere in the Middle East in the 1950s, in the context of the broader regional protests over the founding of Israel, and eastern Europe in the 1990s, following the fall of the USSR (Meir-Glitzenstein 2011).

After 1967, this was coupled with coordinated attempts to bolster the Jewish population in the Palestinian Territories through the settlement movement, which was inspired by fears that the significant Palestinian population of the OPT—including large numbers that had been made refugees in 1948 and had subsequently resettled in the territories—would result in Israel losing its hold on the region. As a response, Israeli settlers began moving to segregated enclaves in East Jerusalem and the Palestinian Territories. The aim was to complicate any future geographic division as well as obfuscate what until 1967 had been a starkly separated territory (at least with respect to central Israel) by building a small and diffuse but tightly linked network of Israeli settlements (Weizman 2007; see chapter 5).

(p.104) As a reaction to fears of Palestinian expansion, the settlements therefore also represent a tacit acknowledgment of Palestinians’ existence. Nonetheless, although the settlements were created as an effort to make the material world conform to claims made about it—and specifically to confound traditional methods of drawing boundaries between groups—the resulting changes to the population landscapes would influence maps of Israel. In the context of the ongoing drive toward the West Bank, the CBS and Bachi began to develop methods that would allow them to draw borders around complex, small population clusters such as the settlements. Likewise, Bachi’s innovations in his later work lay precisely in his ability to separate what might otherwise have been inseparable: to draw out and count the intentionally imbricated Israeli settlers from their distributed points across the West Bank.

Yet it was not a foregone conclusion that the settlement populations would be counted in the Israeli census, given that they were located on disputed territory where, as we have seen, Palestinians were separately enumerated. The need for the CBS to give the state legitimacy by demonstrating the numerical strength of Israelis, however, dovetailed (perhaps unsurprisingly) with state efforts to claim the settlements as being an integral part of Israel. In contrast, the publicly available Israeli counts of Palestinian populations were left deliberately vague. After 1967, the CBS did not attempt to publicly count Palestinians in the OPT again—that is, except for those residing in areas annexed to Israel, such as East Jerusalem. The CBS instead relied on estimates of population growth (e.g., Ministry of Defense 1973, 98). Moreover, after the 1994 signing of the Oslo Accords, Palestinians gained some limited sovereignty, and the duty of counting Palestinians was transferred to the newly formed PA. For this reason alone, the 1995 Israeli census was unique. It was the first census after 1967 that was conducted with the expectation that the Palestinians would also conduct their own census, focusing on Palestinian-controlled areas of the OPT as well as East Jerusalem at roughly the same time. The Palestinian census was in fact completed in 1997 (PCBS 2008).13

In the context of the demographic war and new Palestinian census, the 1995 census marked a push forward for increased accuracy on the part of the CBS methodology. Although the CBS conducted censuses roughly every decade after 1967, the 1995 census was a methodological watershed of sorts. For starters, it was the first census to comprehensively use GIS (p.105) mapmaking in an attempt to record the precise location of “every” dwelling—and thereby, in theory, every person in Israel (Lasman 1997). In this respect, it built on the groundwork of the SOI, which had started converting its paper maps to a digital GIS framework in the late 1980s (Peled 1996). Moreover, instead of incorporating digital maps on the side, the census shifted wholesale toward conducting the census through the use of GIS. Districts were redrawn, and the counting process itself was mechanized, as GIS was used to determine and print maps for each enumerator to follow while conducting the count. In addition, a national geographic database was developed to store census data for the foreseeable future (Bahat 1997; Barak 1997; Ben-Moshe 1997a, 1997b, 1997c; Blum 1997; Calvo 1997; Kagan 1997; Lasman 1997; Peled 1996; Stier 1997).

Distinct parallels can be drawn from the methodology of the 1995 census and Bachi’s later work. Bachi himself did not have a direct hand in the run-up to the 1995 census because he had retired from the CBS, although he remained professionally active until he passed away in 1995 (“In Memory of Roberto Bachi” 1996). But in many ways the 1995 census represents the outcome of the methodology that Bachi developed over the course of his career. Bachi was influential in all the major developments that culminated in the 1995 census. These included the early turn to computerization, emphasis on meeting international scientific standards, and efforts to rely on direct counts versus smaller statistical samples to pinpoint exact locations of individuals and form a “snapshot” of the population (Leibler 2004, 2007).14 Indeed, much of the material for Bachi’s 1999 book was written by Bachi himself precisely during the years that the 1995 census was in development.

Bachi’s 1999 Book: Convex Hulls from Point Features

In the context of the demographic war, the CBS wanted to count as many Israeli Jews as possible. So the question would have become not whether but instead how to include the settlements in the census. In the end, an attempt was made to seamlessly incorporate the post-1967 settlements into the new computerized regions. In the 1995 census, each settlement is defined as one or more census tracks, and they appear throughout the CBS’s publications, including lists of “towns in Israel” and maps of the “urban areas of Israel” (CBS 2000b), without reference to their unique status as settlements within the Palestinian Territories (figure 3.4). Similarly, the (p.106)

Removing Borders, Erasing Palestinians: Israeli Population Maps after 1967

Figure 3.4 An annotated screenshot of Jerusalem (center) from the CBS online interactive map of 1995 Israeli census data. This map was made in the context of the Oslo Accords, and the Green Line is drawn clearly and prominently through the center of the map. It surrounds the (Israeli-defined) Jerusalem metropolitan area and divides the annexed areas of Israel on the left from the West Bank on the right. Yet the West Bank is referred to using the adapted biblical name of “Judea and Samaria”—a use that is part of a political strategy to avoid reference to the Palestinian West Bank. The map shows the borders of statistical areas, revealing how Israeli settlements in the West Bank (the smaller pockmarked areas at right) are incorporated as Israeli municipalities for the purposes of the census. The largest of these areas, just to the east (right) of Jerusalem, is the settlement of Ma’ale Adumim. The attempt to incorporate Israeli West Bank settlements into the census has meant giving up on the contiguity of Israeli territory. Here, Israeli territory includes all the complex areas toward the left as well as the settlements toward the right, which are depicted as numerous islands of Israeli sovereignty within the West Bank. The settlements appear in far fewer numbers than their present levels, however, because the map, made in the mid-1990s, does not show more recent expansion.

Annotations by the author (CBS 2000a).

(p.107) properties of Israeli settlements in the West Bank are registered with the Israel Land Authority rather than Israeli administrators of the Palestinian Territories. This helps to legitimize their status, legally and economically, as “Israeli” lands (Shehadeh 2007, 83). In Hebrew, the word settlement (yishuv) is most frequently used in its general sense to refer to all places of steady human habitation. So throughout the CBS and other Israeli state maps at the time, no internal borders are shown, no Palestinian towns in the OPT are shown, and there is no distinction between Israeli cities in Israel (which are also referred to as settlements in the more general sense) and the Israeli settlements in the Palestinian Territories.

The 1995 census therefore indicates an ongoing shift in the conceptions of sovereignty at work, from one that seamlessly incorporated all of the West Bank (as suggested by figure 3.2b) to one that omitted Palestinian-controlled areas but nonetheless selectively retained the Israeli settlements (figure 3.4).15 Whereas Palestinian populations were treated entirely separately by the census in 1967, using distinct definitions and methods of enumeration, in 1995 the groups of Israelis in the same land are fully incorporated, despite the fact that they lie beyond the Green Line. Furthermore, while the Green Line was not often displayed explicitly, the census was still actively involved in drawing implicit and explicit boundary lines. For the settlement boundaries were being redeveloped and reinstated with increasing sophistication. The definition of these settlement boundaries was a lengthy and detailed process (Calvo 1997) that required new methods of the sort that Bachi just happened to be developing at the time.

In his 1999 book, the culmination of his life’s work, Bachi avoids choropleths altogether. But this does not mean that he didn’t emphasize boundaries. Yet rather than focusing on political boundaries, Bachi centers his arguments on methods for drawing complex statistical borders around existing small population groups. These are precisely the types of methods that would have perfectly fit the needs of the Israeli census because they would allow the census to use statistics to contend that the settlements fit within the “statistical area” defined by the Israeli population. However, this effaced the process whereby—instead of producing statistics in order to count existing population distributions—the settlements were put in place precisely so that they could be counted.

One of Bachi’s methods is particularly useful in this context: his research on convex hulls, a type of mathematical set. The problem the census faced (p.108) was how to define the national area in light of population groups, like the settlements, that were not geographically contiguous with the main area of Israel. Convex hulls represent one possible solution. The convex hull is a mathematical term for a boundary that fits tightly around all the points of a specific set. If there are a finite number of points in that set, then forming the convex hull is like stretching a rubber band around the outermost points (figure 3.5).

Finding a convex hull would be especially useful, for example, in determining the areas that are defined by close networks of Palestinian towns—and thereby determining the leftover packets of space that might be settled by Israelis. It also parallels the redrawing of political boundaries after 1967. Jewish settlements were founded in key strategic spots, and then political boundaries were redrawn around those points in a sort of convex hull. This was justified by noting that since Jews now lived there, the areas were to be incorporated into the Israeli state. Most notably, such borders include the expanded boundary of “Greater Jerusalem” that accompanied the Israeli annexation of East Jerusalem and large areas of the West Bank that immediately surround the city. Analyzing redistricting in relation to the demographic war, Meron Benvenisti (1999) has claimed that “the annexation boundaries did not determine the city’s demographic ratio. Rather, the ‘optimal demographic ratio’ has created the city’s boundaries” (see also Zureik 2001).

Yet in time, the material efforts of those redrawn borders would shape the social conceptions that had inspired the borders in the first place. For example, instead of using one solid political boundary, whether for nations or cities, the use of convex hulls provides a more textured and complex rendering of scattered populations. This rendering was made possible in part through the settlement project. For the convex hull is also useful for estimating the geographic size of individual settlements, then statistically appending their areas to the total area of Israel. Indeed, although settlements have political boundaries, the inhabited areas often expand quickly—a process that involves claiming outside lands for defense walls and other security structures (Weizman 2007). Thus the populated area of any particular settlement, which is of interest to a census concerned with mapping every habitable building, not infrequently falls outside the municipal area, and attempts to deal with this situation would mold settlement cartography. (p.109)

Removing Borders, Erasing Palestinians: Israeli Population Maps after 1967

Figure 3.5 A map of Tel Aviv hotels (points) with an enclosing convex hull (line). This figure demonstrates how point or circle maps can obviate the need to show political boundaries. None are given, and the image has been made at a high level of abstraction. In addition, it illustrates how Bachi begins with data and then constructs statistical boundaries—in this case, a convex hull consisting of the straight line that connects the most distant points.

The Palestinian city of Jaffa, which falls within the broader Tel Aviv–Jaffa municipal boundary, is cut off from the map. This is accomplished through the act of drawing the convex hull, thereby transferring the background area of the hotels from the Tel Aviv municipal (political) boundary to the convex hull (statistical) one. The resulting boundary is entirely dependent on the accuracy of the statistical data set that is used, which appears to have only included Israeli-owned hotels. Bachi (1981, 1018; 1999, 91, 94) defines the convex hull as the “gross [territorial] range,” yet here it serves precisely to exclude Palestinian residents. For an analysis of how Tel Aviv was defined and built through a contrast with—as well as the obliteration of—Jaffa, see Rotbard 2015.

Figure reprinted with permission.

(p.110) Bachi’s 1968 book is evidence of how, in his early work, he started omitting internal boundaries and moving from areas to points. In contrast, by 1995, Bachi had largely abandoned his attempts to rectify GRPs with choropleth maps. He turned instead to drawing boundaries around statistical data, beginning to constitute new borders based only on those points that he selected as being significant to the census. In so doing, Bachi again adapted and innovated on international statistics in order to meet the strategic political goals of the Israeli state. In the process, he rejected the traditional focus on the consistency of national territory in order to retain the continuity of borders around the fragmented territories, including the settlements, claimed by the state. His methods allowed for an enlargement of the total area of Israel that is inhabited by Israelis, and they did so in a way that naturalizes the settlements as a part of Israel.

In addition, as noted above, when the total area of a particular settlement is determined based on the furthest points of habitation, these can be far larger than its official municipal boundaries, which might take longer to catch up to the pace of construction. Any private Palestinian areas that happened to fall between two outlying points would simply be incorporated into Israeli territory through the use of the convex hull—that is, by drawing an enclosed line around those (newly) Israeli points and calling everything in between “Israeli.” In combination with GIS software, the geographic database that was developed in the course of the 1995 census enabled the cartographers to calculate convex hulls for hundreds of thousands of individual points at ever finer scales, including individual buildings, thereby defining and quantifying a set of ever more multiplicitous borders. The convex hull method sacrificed national contiguity in order to maintain strict and increasingly numerous boundaries. Yet the fact that there were more borders did not mean that they became more open or less guarded; quite the contrary.

Conclusion: The Continuity of Computer Cartography

Population statistics, like those developed in the national census, form one of the core connections between the state and local communities, and they are central to everyday governance practices. As one of the primary means through which population statistics are communicated, census cartography fundamentally influences both national and international public (p.111) imaginations of those governance practices as well as society itself. Yet cartographic methods are also shaped in local social and geographical landscapes. Bachi’s attempts to make graphing and statistical mapmaking more readily available served to place him in the international statistical vanguard, but his work continued to be informed by the political realities of the Israeli occupation.

This was not obviated by the spread of computers and their allegedly universal rationality. In the end, instead of representing a radical departure from previous cartographic methods (November, Camacho-Hübner, and Latour 2010), computers and GIS proved useful in legitimizing Bachi’s political position. While the public use of his GRP symbols was fast outpaced by computer graphics capabilities—even to the extent that they appear less frequently in Bachi’s (1999) own later research—his contributions to core concerns of geographic and demographic statistics, and their application in statistical cartography, continue to be influential (Louder, Bisson, and La Rochelle 1974; Shea and McMaster 1989; Wegman and Carr 1993).

Contrary to popular rhetoric at the time, the advent of computers did not unilaterally involve further abstraction from local landscapes. Indeed, computers represented the triumph of rather than a challenge to the theoretical foundations of Bachi’s work, including his efforts in local and regional landscapes. They also allowed for the incorporation of the physical existence of Palestinians and the Israeli settlements, which in turn influenced the main directions of population cartography. In 1967, the very presence of Palestinians on the newly occupied areas meant that the census cartographers, including Bachi, had to rush to count and manage them. The policy of omitting the Green Line, in the context of the existence of large numbers of Palestinians in the landscape, presented a challenge to Bachi’s attempt to construct a cartographic image of the homogeneous nation of Israel. As a result, he began phasing out his use of shaded choropleths instead of exclusively using graduated circles in combination with his GRP symbols. In 1999, the ongoing conflict along with the settler movement that was a response to the continued Palestinian presence meant that the census now had to include disparate and noncontiguous settlements in tabulations of Jewish populations. Consequently, Bachi would further innovate on his use of dots by adapting convex hulls for geostatistical use. He thereby moved from developing ways of mapping that obviated the (p.112) need for borders and toward finding ways to draw continuous statistical boundaries around noncontiguous political spaces.

By omitting internal boundaries altogether, Bachi could include Israeli data without making plain that such data for the Palestinian Territories were either not being collected or, in the case of the census, not being indicated on the map. He then helped to actively naturalize the presence of Israeli settlements in the landscape by redrawing boundaries around the area of Jewish habitation—an area that by design, were spread across most of the Palestinian Territories. As such, both GRPs and convex hulls were useful in constructing facts that were empirical and political. Yet in order to reconstruct a total national area that includes the settlements, the 1995 census first had to extricate the settlements from their surroundings—that is, precisely the numerous Palestinian towns of the West Bank and Gaza Strip.

Bachi’s research is therefore both fully internationally scientific and thoroughly influenced by local political contexts. Despite his best efforts to position himself otherwise, Bachi’s (1980) work even at the most abstract levels was shaped by the landscape of Palestine and Israel. At the same time, his choices were circumscribed by the hierarchies of international technoscience—whereby his legitimacy depended on his ability to present his findings as objective and exact. Computers lent credence to this goal. But the tension between maintaining supposed empirical rigor, on the one hand, and adapting his methodology to the people and landscapes so that they might further the practical goals of the Israeli government, on the other hand, is one that continued through the adoption of GIS in the region. These tensions played out in different ways in the nascent PA, which is the subject of the next chapter.


(1.) For a digital version of the 1967 census data for the Palestinian Territories, see Perelmann 2012.

(2.) On representation and maps, see chapter 2.

(3.) A census also figures prominently in the biblical story of the birth of Jesus, where it is given as the reason why Jesus ended up being born in a stable. Mary and Joseph are said to have traveled to Bethlehem to be counted in a Roman census, and the reason they couldn’t find a place at an inn was because so many others were in town for the census as well.

(4.) Zureik’s work also contributes to the large body of literature on Palestinian resistance to these asymmetrical relations, which I discuss in chapter 6.

(5.) The persistence of anti-Semitism, or racist prejudice against Jews, is well documented in a variety of contemporary contexts. In recent years, however, there has been a campaign on the part of advocates of Israel to conflate any criticism of Israel or Zionism with anti-Semitism. In effect, it involves accusing someone of antiSemitism any time they draw attention to an oppressive policy enacted by the state of Israel. Pro-Israeli settler movements use this tactic in an attempt to close down dialogue and deflect attention away from the ongoing racism directed against Palestinians, who are systematically oppressed under the Israeli occupation. Although largely unsuccessful, such strategically misguided instances of “crying wolf” do have the unfortunate side effect of hampering acknowledgment of instances of anti-Semitism when they do occur. On racism against Palestinians, see Joseph Andoni Massad (2003, 2006), who provocatively argues that anti-Palestinian racism is a form of anti-Semitism.

(6.) In chapters 4 and 5, I explore the implications of knowledge hierarchies for Palestinian cartographers.

(p.234) (7.) Anne Godlewska (1999) has pointed out how in the Middle Ages, geography was also viewed as a science, and that it was founded on accurate descriptions rather than hypothesis testing. It lost this status as a science during the Renaissance, only to regain it in the twentieth century. In addition, Godlewska convincingly investigates exceptions to the broad trends described here—the explorer and cartographer Alexander von Humboldt foremost among them.

(8.) For more on the early history of GIS use in the census, see the GIS History Project 1997.

(9.) This position is still put forth, as can be seen from the claims of several candidates in the 2012 US Republican primary and numerous online articles, such as Joseph Farah (2002). Also, the online “Map of Palestine” (2015) directs viewers to a page that says, “The map you requested could not be found. It is possible that the address is incorrect, or that THE COUNTRY DOES NOT EXIST!”

(10.) Meir’s words also contrast with a well-known quote from Angela Davis (1971): “Human beings cannot be willed and molded into nonexistence.” Davis was writing from prison just two years after Meir spoke, albeit in the different context of the US Black Panther struggle. Viewed one way, Davis’s quote seems to insist on an essentialist materialism where physical presence cannot be manipulated or constructed through social means. Yet her statement instead could be read pragmatically—that is, to suggest that human beings will not often allow themselves to be willed into nonexistence. In this respect, Davis’s words highlight how attempts like Meir’s can actually be counterproductive, because they frequently generate the types of resistance they seek to foreclose.

(11.) In a related effort, PLO and other Palestinian maps did not show the Green Line even before 1967 (e.g., Hadawi 1957) although as with Israeli maps, this practice has mitigated somewhat over time.

(12.) This is epitomized in the disastrous category of present-absentees, as noted in chapter 2.

(13.) In chapter 4, I focus on the Palestinian state mapmaking efforts that began during this period.

(14.) The snapshot metaphor was built on the census methods under the British occupation. For example, Hadawi (1957, 8) argues that the 1931 census “enumerated all the persons present in Palestine at midnight on November 18th, 1931, irrespective of whether they were residents of the country or not.” Similarly, according to Hadawi, residents who were not in Palestine at the time were not enumerated, and this held for further estimates, based on the census, right up to 1946.

(15.) More recently, Google Street View, which is described in chapter 1, has begun to collect data for the Israeli settlements (Sheizaf 2012). Street View has not been extended to the Palestinian Territories more broadly at the time of writing (see figure 4.5b).