The Islamic City--Historic Myth, Islamic Essence, and Contemporary Relevance
Janet L. Abu-Lughod
International Journal of Middle East Studies, Vol. 19, No. 2. (May, 1987), pp. 155-176.
Stable URL:
http://links.jstor.org/sici?sici=0020-7438%28198705%2919%3A2%3C155%3ATICMIE%3E2.0.CO%3B2-R
International Journal of Middle East Studies is currently published by Cambridge University Press.
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/journals/cup.html.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For
more information regarding JSTOR, please contact support@jstor.org.
http://www.jstor.org
Fri May 4 04:56:37 2007
Int. J. Middle East Stud. 19 (1987), 155176 Printed in the United States of America
Janet L. Abu-Lughod
T H E ISLAMIC CITY-HISTORIC MYTH,
ISLAMIC ESSENCE, AND
CONTEMPORARY RELEVANCE
At the present time of resurgence in Islamic beliefs, the question of the Islamic
city has once again come to the fore. In many parts of the Arab world, and
especially in Saudi Arabia and the Gulf, urban planners with a new found
respect for the great achievements of the past are searching for ways to reproduce
in today's cities some of the patterns of city building that have been
identified as Islamic. They have been influenced, whether wittingly or not, by a
body of literature produced by western Orientalists purporting to describe the
essence of the Islamic city. The purpose of this article is, in Part I, to examine
and criticize some of the basic works in that tradition and then, after deconstructing
the concept of the Islamic city, to build up, in Part 11, a somewhat
different, and hopefully more dynamic and analytic model. The article ends with
a brief discussion of whether and in what ways it would be feasible or desirable
to build contemporary cities on Islamic principles.
In some ways, historiography takes the same form as the traditions of the
Prophet. The authenticity of any proposition is judged by the isnad or "chain"
by which it descended from the past. Certain chains are deemed more trustworthy
than others. One makes reference to an earlier authority in order to
substantiate a statement's authenticity or truth. The truth, therefore, is only as
good as the isnad (chain) of its "construction."
The first part of this essay is concerned with the criteria of authority, chains of
authenticity, and the construction of reality in Orientalist scholarship. Its thesis
is that the idea of the Islamic city was constructed by a series of Western
authorities who drew upon a small and eccentric sample of pre-modern Arab
cities on the eve of Westernization (domination), but more than that, drew upon
one another in an isnad of authority we intend to trace here.
One of the earliest codifications of the characteristics of the Islamic city, at
least the earliest generally cited in subsequent literature, was William Marqais'
1928 article, "L'Islamisme et la vie urbaine."' The article introduces several
themes that appear over and over again in subsequent discussions of the Islamic
city.
8 1987 Cambridge University Press 0020-7438187 $5.00 t.OO
156 Janet L. Abu-Lughod
The first is that Islam is essentially an urban religion. In support of this
contention, Marqais notes that Muhammad himself was an urbanite suspicious
of nomads, that the leadership cadres of the early Islamic proselytizers were
members of the urban bourgeoisie of the Peninsula, and that the requirement
that the Friday communal prayer be solemnized at a congregational mosque
made urban living necessary for the full Muslim life. Marqais uses an earlier link
in the chain of Orientalism when he cites Renan to legitimate his view. The
phrase he quotes is a simple allegation: "The mosque, like the synagogue and the
Church, is a thing essentially urban (citadine). Islamism [sic] is a religion of
cities." This quotation is particularly intriguing because it undermines the whole
enterprise of defining the unique character of the Islamic city; it suggests that
Islam shares with Judaism and Christianity the same quality of urbanity.
However, Marqais makes a second point as well. He notes that new cities were
often founded by new powers/dynasties in Islamdom, thus acknowledging that
Islamic civilization was not merely a set of religious beliefs and laws but also a
functioning society which was Islamic in the sense that it organized the life of
Muslims into a community not just of believers but of doers.
Finally, Marqais introduces several characteristic elements of the physical city.
Citing Ibn Khaldun, Arab geographers and legal doctrines, he reaches a definition
of the Islamic city which he contends is quintessential: a city must have a
congregational Friday mosque and it must have a marketlchief bazaar nearby.
Associated with the jfimic siiq (mosque-market) complex was a third physical
feature of Islamic cities, the public bath (hammam), of functional significance to
prepare believers for the Friday prayer. Paraphrasing Renan, however, we might
note that when the church was also the temporal power, medieval European
cities were also defined by the presence of the cathedral and the marketplace in
front of it. Thus far, therefore, we have only a very modestly etched idea of the
Islamic city, one which poorly distinguishes it from cities in other religious/
cultural contexts and one which has as yet no topography.
The ideas of William Marqais were incorporated into two articles written by
Georges Marqais, namely "L'urbanisme mu~ulman,"~and "La conception des
villes dans 1'1slam."~The former, in particular, constitutes an important link in
the chain of constructing the Islamic city. The 1940 discussion begins with a
paradox alluded to earlier by William Marqais, namely, that despite the fact that
Islam was a religion carried by nomads, it was essentially an urban religion: The
mosque created the Islamic city. He notes as well the importance of baths and of
markets in the making of the Muslim city.
It is Georges Marqais, however, who gives a morphology to the Islamic city.
He notes the differentiation between nonresidential and residential quarters and
the fact that residential quarters are often specialized by ethnicity. Finally, he
describes the physical organization of the city markets which he suggests are
ordered in a certain hierarchy which is not completely accidental. I quote at
length from this section because it is to appear again and again in subsequent
works, either in quotation marks or paraphrased.
I have said that the center was occupied by the Great Mosque, the old political center, the
religious and intellectual center of the city, where the courses were given to students from
The Islamic City 157
the various schools. Near the mosque, the religious center, we find the furnishers of sacred
items, the suq of the candlesellers, the merchants of incense and other perfumes. Near the
mosque, the intellectual center, we find also the bookstores, the bookbinders and, near the
latter, the suq of the merchants of leather and the slipper [babouchel-makers which also
use leather. This introduces us to the clothing industries and the commerce in cloth, which
occupy so large a place in the life of Islamic cities. The essential organ is a great market, a
group of markets that carry the mysterious name, Qaiqariya. The Qaiqariya . ..[is] a
secure place encircled by walls where foreign merchants, above all Christians, come to
display their cloth materials brought from all European countries. The Qaiqariya, placed
not far from the Great Mosque, as in Fez or Marrakesh, for example, is a vital center of
economic activity in the city. Beyond the commerce of textiles, of the jewellers, the
makers of hats [chechias], we find the makers of furniture and of kitchen utensils. . ..
Farther out are the blacksmiths. Approaching the gates one finds places for caravans . ..
then the sellers of provisions brought in from the countryside. ...In the quarters of the
periphery were the dyers, the tanners, and, almost outside the city, the potters.4
It is very important to note here that virtually all of the cases cited by Marqais
in his article are North African. Note the specific references to babouches
(slippers), chechias (tarbushes), Qaiqariya (cloth market)-Maghrib terms that
are not generally used in other regions. Note also the contemporary reference to
foreign (i.e., European) merchants.
These articles, then, set forth the physical characteristics of the Islamic city
primarily as they were observed in North Africa. Much less attention was paid to
the social organization of the city nor was any attempt made to explore the
underlying causes of the particular patterns found in Islamic cities. This task was
essentially left for Robert Brunschvig, in his often cited "Urbanisme mtditval et
droit mu~ulman."~
In this crucial piece, Brunschvig argues that it was customary law, applied by
judges, that over time yielded the type of physical pattern found in the cities of
Islamdom. In his analysis, Brunschvig credits, as his isnad, the work of the
German scholar, Otto ~ p i e s , ~who examined how "the rights of the neighbors"
were treated in Islam according to the Shafici religious school or at least
according to the principal Shafici doctors of law. Brunschvig criticizes Spies for
ignoring the Malikite rite, even though one of the chief works in that school (Ibn
al-Imam) had already been translated into ~rench.'(Ibn al-Imam was probably
born in the 10th century in Toledo and settled first in Cordoba and then
Qairawan.) Another source used by Brunschvig was written in the 14th century
by Ibn al-Rami, a Tunisian master-builder later elevated to Chief ~ a d i . 'Thus,
most of Brunschvig's analysis is based upon North African jurists, with some
parts of it coming from the work of Spies based upon the Shaficirite. Note that
the referent for the legal work is increasingly North African, just as the chief
referent for the physical structure is Fez. We will return to this point when we
follow another line of development in the concept of the Islamic city.
But before doing so, let us look at the culmination of this first isnad in the
classic 1955 article by Gustave von Grunebaum, "The Structure of the Muslim
own."^ This article represented a major step beyond his earlier discussions.
Interestingly enough, in von Grunebaum's earlier book, Medieval Islam, there is
only the briefest discussion of the role of the city in Islam. The author merely
158 Janet L. Abu-Lughod
notes that "Islam, from its very outset unfolding in an urban milieu, favored city
development."10 He further remarks that only in a city, with its Friday mosque,
its markets and, possibly, its public baths, can the duties of the religion be fully
performed. While he ostensibly quotes the Qur'an to substantiate his position,
the actual footnote to this passage (appearing on page 174) sends the reader back
to the 1928 article by William Marqais. By von Grunebaum's 1955 article,
Georges Marqais has also been integrated, as can be seen in the following lengthy
quotation which virtually reproduces the excerpt translated earlier.
Near the mosque as a religious center we will find the suppliers of the sanctuaries, the suq
of the candle merchants, the dealers in incense and other perfumes. Near the mosque as
the intellectual center we will also find the suq of the booksellers, the suq of the
bookbinders, and, as its neighbor, the suq of the leather merchants and the makers of
slippers."
Paraphrasing Marqais further, von Grunebaum simply alleges that
Adjoining this group of markets we enter the halls of the dealers in textiles, the qaisariyya,
the only section of the suqs which is regularly roofed and which can be locked and where,
therefore, precious materials other than fabrics will also be stored and exchanged. Next to
the textile trade the carpenters, locksmiths, and the producers of copper utensils will be
located; and somewhat farther from the center, the smiths. Approaching to the gates [of
the town] one will find . .. the makers of saddles. .. . Then the vendors of victuals. . . .
On the periphery of the town will be situated such industries as . .. the dyers, the tanners,
and almost outside the city limits, the potters.'2
For this entire passage and its quotations von Grunebaum naturally cites the
1940 article by Georges Marqais. But what is remarkable is the fact that he
neither comments on nor adds to this presentation. He merely incorporates the
description of Fez, Rabat, Tunis, etc., as presented by Marqais, into his generalizations
concerning the physical structure of the Islamic city.
With respect to the social organization of the Islamic city he similarly incorporates
uncritically the conclusions of earlier Orientalists. He holds with the
Marqais's that, in contrast to western medieval ones, Muslim towns lacked
municipal organization.'' Also following their lead he suggests that this lack of
municipal government was compensated for by the ethnically specialized quarters
with their own sheikhs. Further, he accepts the views of Louis ~ a s s i ~ n o n ' ~
that guild-like organizations of trades knitted the social organization together, a
view which has subsequently been rejected by 0rientalists.15 And finally, von
Grunebaum describes how Islamic law shaped the city, paraphrasing many of
Brunschvig's arguments and referring to the latter's sources which, as we have
seen, deal largely with North Africa.
As one can see from even this brief treatment, the 1955 codification of the
concept of the Islamic city by von Grunebaum patched together what Orientalists
had agreed were the chief characteristics of the Islamic city, based chiefly on the
Maghrib (although, granted, he does cite Sauvaget). Interestingly enough,
between von Grunebaum's first reference and his classic article on the Islamic
city in 1955, at least one new source had come out by another French North
Africanist, Roger LeTourneau, but this work was not cited at all, a fact which is
difficult to account for.
The Islamic City 159
In 1949, LeTourneau's first book on Fez, Fks avant le protectorat. ~ t u d e
dconomique et sociale d'une ville de I'occident mu~ulman,'~appeared. This book
was to become central to a second attempt to build up a general description of
the physical structure of the Islamic city, that of Carleton Coon. Certainly, one
of the most influential books in western literature dealing with the Middle East is
Carleton Coon's text, ~aravan."In this book, first published in 1951, the entire
chapter on "Town and City" is based on the case of Fez, for which the 1949
work of LeTourneau was the chief source. (In the 33 page chapter, 28 focus on
or draw their chief evidence from the city of Fez.) Despite this, Coon feels
confident about glossing from this case to other Islamic cities.
The Fez case is treated again in LeTourneau's Le vie quotidienne & Fks en
1900," and his summary volume, Les villes musulmanes de I'Afrique du ~ o r d , ' ~
largely generalizes from Fez. One striking impression that emerges from rereading
the corpus of LeTourneau's work is how redundant it is. Descriptions of Fez
in 1900 are projected back to Fez in the Age of the ~arinides~'and generalizations
about all cities of North Africa bear a suspicious resemblance to Fez both
in 1900 and in medieval times!
The focus on the Maghrib continues in the work of Jacques Berque, particularly
his oft-cited article, "Mtdinas, villesneuves et bidonville^."^' It is not
possible to paraphrase Professor Berque, but beneath the complex prose is a
preoccupation with many of the issues raised thus far. Thus, on pp. 12-13 he
raises the question of what makes a city in the Maghrib, and in answer refers
back to a then 30-year old source, the article by William Marqais, citing the
three desiderata: a Friday mosque, markets, and a ritual public bath! Putting
this into a typical Berquean transformation, he defines the Islamic city in
functional terms as being a place for witness and for exchange, a somewhat
different way of saying a place where one finds a mosque and a market.
If the major referent for the foregoing works on the Islamic city is North
Africa and more particularly Fez, the crucial beginning point for the next isnad
is Syria, drawing on the solid scholarly work of Jean Sauvaget on the cities of
~ l e ~ ~ o ~ These studies have had a substantial impact on the~and ~ a m a s c u s . ~ '
evolving image of the Islamic city, but the extent to which his findings (which
detail how Islamic cities evolved from Byzantine antecedents) conflict with the
image constructed on the basis of North African cities, especially Fez, has rarely
been acknowledged. Scholars seem to hold both views at the same time!24
Presumably, many of the differences between new princely towns and existing
evolving cities had disappeared in the Arab world by the 13th to 16th centuries,
the time period covered in Ira Lapidus' now classic but somewhat grandiosely
entitled Muslim Cities in the Later Middle ~ ~ e s . ~ ~Originally a dissertation at
Harvard, it was revised for publication in 1967. In the preface to this work,
Lapidus modestly demurs from testing the hypotheses of Max Weber concerning
the contrast between European and Asian cities and states instead that he will
"confine [himself] in his book to a description of several Muslim cities. The book
concentrates on Damascus and Aleppo in the period of the Mamluk Empire,
1250-1517, with supporting studies of the Mamluk capital city of Cairo." Despite
this demurrer, the opening paragraphs of the body of the text pose a general
question of how Muslim towns were governed, and the concluding paragraph of
160 Janet L. Abu-Lughod
the book says in broad fashion, "This system of relations constituted the
government of Muslim towns."26 Furthermore, the title has been given more
credence than the content. I have seen this work cited by Orientalists as proving
that Islamic cities lacked municipal governance.
In short, just as we find the first isnad to be based chiefly on French North
African sources/studies, particularly focusing on the city of Fez, so we find a
second isnad based upon the Syrian cities of Aleppo and Damascus, as studied
physically by Sauvaget and sociopolitically by Lapidus. In each case, a very
tentative set of place-specific comments and descriptions appears. These enter
the literature and take on the quality of abstractions. With each telling, the tale
of authority grows broader in its application. Forgotten is the fact that only a
handful of cities are actually described. Forgotten is the fact that only certain
legal codes-on which the Islamic form of city is presumed to be based-have
been studied. Forgotten is the fact that Islamic cities have evolved over time and
that the sociopolitical system in Damascus and Aleppo in the 14th century under
Mamluk rule cannot possibly provide a convincing description of how Islamic
cities sui generis were governed!
My own book on Cairo fell into the trap set by the Orientalists by accepting
many of the earlier authorities about the nature of the Islamic city." The edifice
they had built over the years seemed to me a strong and substantial one. Only
gradually did it become clear how much a conspiracy of copying and glossing
had yielded this optical illusion. But criticisms were already beginning to appear.
In this connection we must turn to the 1970 volume on The Islamic City edited
by Hourani and This book marks the beginning of some reasonable
questioning of past work. First, Hourani's introduction points out that North
Africa provided too many of the cases from which generalizations had been
drawn. Second, the articles by Stern and Cahen are virulent refutations of
traditional work on guilds, which they suggest were not found in Islamic-cities.
And third, articles by Saleh Ahmad El-Ali and J. Lassner on the Round City of
Baghdad suggest that at least one planned new city in Islam in no way resembled
the stereotype.
Since that time, we have added a number of new case studies on Arab cities,
and a few more books and many articles have been published with the termIslamic
City-somewhere in the title.29But we seem to be no clearer on the issue,
well stated in Dale Eickelman's article, "Is There an Islamic City?" but only
tentatively answered by Eickelman and Kenneth ~ r o w n . ~ '
The fact is that most studies still focus on a single case and try to generalize,
rather than start with the more fundamental question: Why would one expect
Islamic cities to be similar and in what ways? Part I1 of this paper addresses this
question.
Before this, however, a brief biographical note. I went to India in the late
seventies, looking for (and hoping not to find) the Islamic city. To my surprise, it
was easy to distinguish Muslim from Hindu urban quarters. Several cues seemed
to trigger a subliminal response. First, the ratio of males to females on the street
and in public places was higher in Muslim than in Hindu areas. Second, butcher
shops and trades related to hides and animal products were located almost
The Islamic City 161
exclusively in Muslim areas. And third, it seemed that the decibel level of sounds
was higher and more animated in Muslim quarters, of which the call to prayer
was only one of the added elements.
These semiotics were chiefly in public space; it was harder to penetrate semiprivate
and private space, except in several Hindu areas to which I had personal
entree. The quarter, in the sense of a semi-private lane or courtyard apartment
house, seemed as typical of poor Hindu areas as of Muslim areas, but circumspect
behavior outside the private living quarters seemed less in Hindu than in
Muslim areas. One explanation for this might be differences in the rules of
veiling. Hindu women veil primarily before close relatives (especially fathers-inlaw)
whereas Muslim women veil chiefly from outsiders. Given this, one would
expect Muslims to make a greater distinction between in and out of the house. In
brief, it seemed that social patterns of gender segregation and social patterns of
proscribed foods were the chief religion-linked variables distinguishing Muslim
from Hindu quarters in mixed cities.
This, of course, was not a true test of the case, so I also tried to compare the
urban patterns of Muslim-origin cities with those of Hindu-(or Vedic-) origin
cities. Here again it was possible to make the distinction. Cities originally
occupied by Muslims or substantially expanded by them had far more convoluted
street patterns than cities founded by Vedic/ Hindu populations. The
latter were arranged more regularly, had straight streets unencroached by structures,
and achieved privacy largely through court-houses rather than alleys and
dead-ends. On the basis of this, it seemed that the most probable cause of these
differences might be the nature of the law of real property, rather than religion
per se.
The next logical test was to examine the literature about Muslim cities elsewhere.
With reference to Africa south of the Sahara, scholars thought there were
significant differences between Islamic versus non-Islamic cities. African cities
inhabited by Muslim populations were said to contain complex and narrow
street systems, courtyards, and the spatial segregation of males and females.
And finally, informants suggest that cities in Muslim areas of Asia (Indonesia
and China, for example) exhibit distinctive street patterns, noise levels, and a
sense of Islam. By the time one comes to these cases one is sure that if there is
something Islamic a out cities, it must be more than simple architectural patCP
> -,
terns and designs, since the architectural vernaculars become increasingly distant
from Damascus and Fez.
What, then, can it be that is distinctive?
It is easy to say what it is not. It is not a form, per se. The careful reader will
have noted that the Orientalists' discussions of the Islamic city focus on a unique
conjunction of forces that created a few cities they take to be prototypical. The
forms of these cities at certain points in time are taken as ideal types and are
further abstracted to obtain a final ideal which is created out of congruent forms.
This, they suggest, was (is?) the Islamic city. Not only scholars but present
practicing Muslims looking to build new Islamic cities have accepted this
approach and are trying to find in the planning and architectural repertoire of
the past the tricks and techniques that will reproduce it.31But these approaches-
162 Janet L. Abu-Lughod
scholarly or practical-miss the point. Cities are the products of many forces,
and the forms that evolve in response to these forces are unique to the combination
of those forces. A city at one point in time is a still photograph of a complex
system of building and destroying, of organizing and reorganizing, etc. In short,
the intellectual question we need to ask ourselves is: Out of what forces were the
prototypical Islamic cities created?
PART 11-WHAT CREATED THE "TRADITIONAL" CITY IN THE ISLAMIC WORLD
A modest list of the forces that created the traditional Islamic city would
include: a terrainlclimate; a technology of production, distribution and transportation;
a system of social organization; and a legal/political system which, in
Islamic places and times, could vary considerably.
It is exceedingly hard to unpack this complex bundle to determine the extent
to which Islam influenced any one of them at any point in time. We must dismiss
terrain entirely as being Islamic, even though the Arab region in which this genre
of city building was developed had a characteristic climatelterrain as well as a
historic inheritance which encouraged common solutions. We must also dismiss
technology as being Islamic, for there is nothing religious about pack animals,
handicraft production, small-scale market arrangements, etc., any more than
there is something religious about terrain and climate. Once one eliminates the
influence of these factors, one is left with exploring the social, political, and legal
characteristics of Islam that shaped, but did not determine, the processes whereby
Islamic cities were formed, transformed, and transformed again. These are the
characteristics we explore in the rest of this article.j2
Let us begin with an extreme statement: namely, that the division of the
Middle Eastern-North African city into a nested set of territories with clear
markers and defended borders was not Islamic per se but reflective of a social
order that had much in common with other societies based upon the family writ
large (tribalism, clans, and ethnicity). Fluctuations in the strength of the markers
and the degree to which boundaries were defended were contingent more upon
the state of law and order than on shifts in religious ideology. This is a clear
indication that religion was not the determining variable.
However, to say that Islam was not the only cause of urban form is not to say
that it was unimportant. On the contrary, it was a crucial contributing factor in
shaping cities within its realms. It contributed in several important ways.
First, it made rough juridical distinctions among population classes on the
basis of their relation to the Umma (community of believers) and thus the State.
These distinctions were thus available in the repertoire of territoriality and could
be translated into spatial segregation under certain conditions. Furthermore, the
frequent inability of the state to transcend communal organizations and the
laissez-faire attitude of the state toward civil society33left important functions to
other units of social organization which strengthened them. Since many of these
functions were vicinal ones (maintaining streets and utilities, guarding turf,
providing lighting, supervising and sanctioning behavior, etc.), and since many
The Islamic City 163
vicinal units were composed of socially related people, what we would call the
neighborhood became a crucial building block of cities in the Arab world during
medieval and even later times. It was therefore available in the repertoire of
social responses, sometimes employed by the residents themselves when state
power broke down, sometimes employed by the state as a way to control its
subjects. Whether it was a mechanism for defense or a modality for control
depended upon factors which were certainly not Islamic.
Second, by encouraging gender segregation, Islam created a set of architectural
and spatial imperatives that had only a limited range of solutions. What
Islam required was some way of dividing functions and places on the basis of
gender and then of creating a visual screen between them. This structuring of
space was different from what would have prevailed had freer mixing of males
and females been the pattern. Such spatial divisions were a functional supplement
to alternative patterns of person-marking which were also used but often
not fully satisfactory. Semiotics of space in the Islamic city gave warnings and
helped persons perform their required duties while still observing avoidance
norms.
And finally one returns to the system of property laws that governed rights
and obligations vis a vis both other property owners and the state. Such customary
laws and precedents set in motion a process whereby a pattern of space
was continually reproduced. Of primary importance were the pre-existing rights
of individual or collective users of land and immoveable property. Of secondary
importance were the rights and responsibilities of proximate neighbors, followed
by those of more distant ones. Then, finally, as a residual, there was the right of
the collectivity or larger administrative unit. Under such circumstances, access to
entrances took priority over major thoroughfares or the reservation of land for
public purposes.
I would like here to speculate on these three themes, drawing upon evidence
from both medieval and modern time periods. Even though this is hardly a
scientific approach, on occasion scholars must be willing to engage in free
association to generate hypotheses which might be fruitfully explored, even
though and perhaps especially because such hypotheses step beyond what is
known.
Territoriality in the Arabo-Islamic City
One of the most striking features of the cities of the Middle East and North
Africa, certainly during medieval times but to some extent persisting feebly to
this day in the older residential quarters, is its subdivision into smaller quarters
whose approximate boundaries remain relatively constant over time and whose
names continue to be employed as important referential terms, even when they
do not appear on modern markers of street names, etc.
In contrast, in his study of neighborhoods in Chicago, Albert Hunter found
that territorial names seldom persisted for even a generation, much less a
hundred years, and that there was widespread disagreement both as to the
164 Janet L. Abu-Lughod
recognizability of many neighborhood designations and as to their geographic
extent and borders.34To some extent, the contrast between the persisting boundaries
and nomenclature in Arabo-Islamic cities and the unclear and changing
character of Chicago neighborhoods can be attributed to spatial design and
markers.
Kevin Lynch, in his brilliant book on urban form, The Image of the
tried to probe, through mental mapping, the psychology of spatial borders and
spatial concepts; he concluded, interestingly enough, that some quarters and
cities are more imageable than others. There is no doubt in my mind that the
historic quarters of Arab cities were built to be imageable in a way that gridironplanned
Chicago was not.
The design of the Arabo-Islamic city, with its convoluted paths, was intended
to subdivide space into relatively permanent quarters. But a recognition of this
fact simply begs the question. Medieval European towns were equally devoid of
right angles and through streets; they also were subdivided into potentially
organizable subpockets. Nevertheless, they are quite different in physical pattern
and were quite different in social organization from the medieval Arab
One must look for more than design to explain the signification of turf in the
Arabo-Islamic city. One must look for a common creator of boundaries. There
may be an overarching Islamic reason for the latter. Islamic property laws about
differential responsibility to neighbors and control over accessways to dwellings
may have been partially responsible for the typical cellular pattern found in
residential quarters of medieval Islamic cities.37However, I see Islamic law as an
adaptive mechanism for helping the society to achieve its goals, rather than as a
deus ex machina determining them. Therefore, we must go behind the issue of
how imageable cells of residence were created and maintained to explore why
they were so typically the building block of urban society. This brings us, then,
to the three hypotheses I would like to advance.
Turf and Juridical Classes
States that make juridical distinctions among residents lay the foundation for
what can evolve into a system of spatial segregation. It is a necessary but,
however, not a sufficient cause of residential apartheid. That is because social
distance and physical distance are not necessarily the same thing. By social
distance we mean the degree to which open egalitarian interaction is blocked, or
rather the amount of work which must be done by two parties to overcome
social barriers to intimacy. By physical distance we mean the degree to which
physical contact is blocked by space or rather the amount of work which must be
done to overcome spatial barriers to face-to-face contact. Clearly, we can all
think of cases in which maximum social distance can coexist with minimal
spatial segregation (master-valet, master-house slave) and, conversely, other
cases where minimum social distance can be sustained over great physical distance
(loved ones in other countries). Indeed, it is generally when lines of social
distance become less marked that physical distance is intensified.
The Islamic City 165
Medieval Islamic cities certainly did maintain the distinctions between juridical
classes through social distance (as evidenced by sumptuary regulations, the
semiotics of clothes, body postures, etc.) but spatial distance was not always a
mechanism for maintaining social distance. On occasion, however, particularly
during periods of tension, physical segregation was employed to intensify the
social boundary markers.
For example, a juridically different category in most Middle Eastern theocracies
consisted of the dhimmis (Christians, Jews, and Zoroastrians enjoying a
special status of protection) who, where relevant, might be subclassified by
specific faith or place of origin. Always there were rules governing their behavior
and regulations concerning their collective responsibilities to the state. Often
there were specific restrictions on the occupations open to them. Occasionally,
there were rules specifying consumption patterns (whether they could ride horses
or were confined to mules, and the like) and even permitted dress. But for the
most part these regulations do not seem to have been viewed as oppressive.
(When they became so, as for example during the reign of al-Hakim in Cairo,
they occasioned great comment about the sanity of the ruler, indicating how
great a deviation from normal they were.)
Residential segregation, however, was not invariable and was seldom involuntary.
Voluntary concentrations are noted over and over again in urban histories,
either in relation to certain economic functions (Coptic quarters near ports in
Cairo) or to certain political advantages (Jewish quarters near the palace of the
ruler). Such concentrations facilitated the exercise of self-rule in matters of
personal status and helped, in the proximity-based city of the time, to gather the
density required to support common special services and institutions. These
common services and institutions, in turn, created markers for quarters which
indicated to outsiders who was supposed to live there and indicated to insiders
that they belonged there.
But there were certain times and places where the potential for physical
segregation was translated into the creation of absolute segregation by juridical
status. In Moroccan cities, for example, Jews were not segregated into ghettos
until the 19th century. The promulgation of a new government regulation was
rationalized as being required for the safety of the Jewish population and it
was not enforced for long. The fact that it occasioned comment and required
explanation is, in itself, evidence that it was unusual to require what was often
chosen.
Clearly, the extent to which territorial distinctions were based upon juridical
status varied from one part of the Arabo-Islamic world to another depending in
part upon the statistical representation of dhimmis of various kinds. On the
Peninsula this situation did not arise until the oil boom. Similarly, North Africa
after the Almohads had no important Christian minority until the appearance
of Europeans, and therefore the terms Christian and European were often
interchangeable. A Western quarter was by definition a Christian quarter and
vice versa.38This was not true for Egypt or the Fertile Crescent, where indigenous
Christian minorities persisted throughout the Islamic period and continued
166 Janet L. Abu-Lughod
to play an integral role in society. Jewish communities were found scattered
throughout the Arab world, although usually concentrated within only certain
parts of countries. Such concentrations were often linked to political conditions,
as when Jews of Morocco moved to Mogador at the behest of the Sultan to
handle the trade in his new port, or when Jews concentrated in the capital city to
take advantage of royal protection.
The events in Lebanon over the past 10 years have made me think a great deal
about the conditions under which juridical status differences are likely to be
translated into spatial segregation. It is clear that 10 years ago Beirut had the
nuclei of predominantly Christian and predominantly Muslim quarters which
were the result of long-term voluntary sifting and sorting, but such modest
segregation was far from complete. It is interesting to review some of the social
science literature about Beirut at that period. Samir Khalaf and Per Kongstad's
study, Hamra of ~ e i r u t , ~ ~indicated enormous mixing of persons of various
religious persuasions as late as the 1970s in this section of what was later to be
called Muslim West Beirut. Similarly, Fuad Khuri's anthropological study of the
peaceful relations between a predominantly Muslim suburb and its adjacent
neighbor, a predominantly Christian suburb, stressed the intermixing of people
and their co-e~istence.~'And yet, the civil war that erupted in 1975 began at the
border between Fuad Khuri's two suburbs, and Hamra was included in the socalled
Muslim quarter of West Beirut in the civil war and then the siege. It was
not segregation that gave rise to intercommunal tensions; rather, it was tensions
that gave rise to labels which were then partially translated into reality, as
population resifted and resorted itself to share a common fate. To the very end
of the Israeli siege of Beirut, so-called Muslim West Beirut was never exclusively
Muslim, as eyewitness accounts confirm.
Thinking over the present experience makes it possible to read back into
history to try to fathom under what circumstances juridical status might have
changed from a potential to a real cause of ethnic/religious segregation. It is
clear that at times of outside threat when the state was weak, fragmentation was
likely along the fault lines of society, which in Islamic states tend to lie along
juridical status cleavages. When loyalty to the state is crucial, when defense
becomes paramount, and when the state itself is under attack, one can expect
segregation-either to protect or to control.
Economic threats may be as significant as military ones. The sectarian cleavages
of Lebanese society were strongest in the late 19th century and appear to
have been related to western commercial incursion^.^' The incursions of a commercial
Europe into Moroccan economic life at the beginning of the 19th century
seem to account for both the institution of Jewish ghettos in Fez and Rabat and
the creation of Mogador as a largely Jewish commercial port.
The present experience also helps us to fathom how the transition might have
taken place. Designation of a specific area as ethno-religious turf helps both to
drive out others and to gather in persons of similar status. If, in addition, there
are decrees or laws promulgated or if there are sanctions imposed, the segregation
is intensified. (For example, after the ghetto in Rabat was established in
1807, guards were posted at the single entrance to the quarter to supervise entry
The Islamic City 167
and exit. The dual function of defense of inmates and control over them is
clearly evident in the semiotics of space.)
Gender Segregation and the Arabo-Islamic City Form
The creation of male and female turf is perhaps the most important element of
the structure of the city contributed by Islam. It is important to remember,
however, that the rules of turf were not only to establish physically distinctive
regions; more important, they were to establish visually distinctive or insulated
regions. The object was not only to prevent physical contact but to protect visual
privacy. Line-of-sight distance, rather than physical distance, was the object of
urban design. Thus, Islamic law regulated the placement of windows, the heights
of adjacent buildings and the mutual responsibilities of neighbors toward one
another so as to guard visual privacy. Architecture assisted this process. Not
only the devices of mashribiyya (lattice wood) screening but the layout of houses
and even of quarters created the strangely asymmetrical reality that women
could see men but men could not see women, except those in certain relationships
with them. Here is Lane-Poole's description of an Egyptian upper-class
house in the 19th century:
[As one enters the house there] is a passage, which bends sharply after the first yard or
two, and bars any view into the interior from the open door. At the end of this passage we
emerge into an open court. ... Here is no sign of life; the doors are jealously closed, the
windows shrouded. ...We shall see nothing of the domestic life of the inhabitants; for
the women's apartments are carefully shut off from the court. ... The lower rooms,
opening directly off the court are those into which a man may walk with impunity and no
risk of meeting any of the women.. .[another] door opens out of the court into the
staircase leading to the harim rooms, and here no man but the master of the house dare
penetrate. . . . When a man returns there he is in the bosom of his family, and it would
need a very urgent affair to induce the doorkeeper to summon him down to anyone who
called to see him.42
But one need not take the testimony of a foreign observer to substantiate the
universality of some of the principles. In the aqsrfr(castles) of southern Morocco,
one can find the same bent entrances designed to create a visual blindspot. And
urban building regulations were replete with requirements that the doors of
buildings occupying opposite sides of the street must not face one another,
another mechanism of visual control.
Similarly, within the dwelling the ideal was to segregate public from private
space so that males could circulate without interfering with the movement and
activities of females. Obviously, this was possible only, if at all, for the very
wealthy, such as the household home that was described above. Compare LanePoole's
account with the following description of elite housing in Jidda, Saudi
Arabia, in the 1960s.
A typical old house includes a wing for the men, usually on the first floor, with an access
to the garden and to the ground floor balconies. In this part of the house the head of the
household has his sleeping room, study, guest rooms, and sitting rooms. The second and
third floors belong to the women; one is for daily living and another for receiving guests.43
168 Janet L. Abu-Lughod
For the poor, no such absolute segregation was possible. Rather, signs and codes
helped regulate spatial symbiosis, often by rules that governed timing.
The most obvious semiotic of sexual segregation in the Islamic city was the
sign used in front of the public bath to indicate ladies' day. Subtler signs
governed other divisions of time and/or space, however. Take, for example, the
zone just outside the houses that share a common accessway in a dead-end
quarter. These are found in most parts of the Arab world, from the Fertile
Crescent to Morocco. I have elsewhere termed this space semi-private space, a
third category between public and private which is found infrequently in sexintegrated
societies but is often found in sex-segregated societies.44
The fact is that the ideal of separation between the sexes is best achieved by
the wealthy who can afford to duplicate space and can afford the servant or slave
girls who were never guarded from male sight or contact. Most poor women
were less able to meet the ideal. For them, the family writ large permitted the
doing of tasks as well as the protection from strange males because the local
neighborhood was an extension of the home and therefore the family. The blind
alley or dead-end court street was such a device for achieving this compromise
between the exigencies of life and the directives of female seclusion. Nawal
Nadim has written very sensitively on this subject in her contemporary anthropological
study of life in a poor hdra (neighborhood) passageway in Cairo.
A large number of activities take place in the harah passage which in other parts of Cairo,
or even during different historic stages of the harah, would be restricted to the physical
setting of the dwelling. . . .[Tlhe manner and form of familiarity with which various
intimate activities are carried out in the harah passage make it evident that the alley is
actually considered by both sexes to be a private domain. Members of the two sexes in
the harah treat each other with familiarity similar to that existing among members of the
same family. Even outside the harah, any male resident is responsible for protecting any
female member of his harah. He is further responsible for what she does, and he has the
right to interfere in her activities if he finds them inappropriate.45
As one can see from this discussion, the family is simply written larger when it is
impossible to achieve the physical and visual separation required between
strangers. I am struck with the similarity of these rules in semi-private urban
space to those that apply within an encampment of sedentarized bedouins, as
described by Lila ~ b u - ~ u ~ h o d . ~ ~
It is clear that when densities are high and houses too small to contain the
manifold activities women are supposed to do in them, the spillover space
becomes appropriated as semi-private space and co-residents who might inadvertently
have visual access are appropriated into a fictive kinship relationship
to neutralize danger. Dress is an important part of the semiotics of space. As
Nadim notes, "clothing which is acceptable for a woman within the lodging is
also acceptable in the harah."47 Nor is it only in Egypt that such adaptations
take place. Elizabeth Fernea's descriptions of her Street in ~ a r r a k e c h ~ ~indicate
that when women ran next door within the enclosed portion of their street, they
did not cover themselves as fully as they would have, had they been going into
public space. It was all in the family. Clearly, then, one of the reasons why the
The Islamic City 169
older pattern of city building has been maintained in many sections of AraboIslamic
cities, even today, is that it is still well adapted to the complex demands
for visual privacy for females.
I am often struck, as I wander around Arab cities, with how easy it is to tell
whether I am in public space or have blundered into semi-private space. I have
often tried to identify the markers that indicate this. A sudden narrowing of the
path, particularly if that narrowing has been exaggerated by the implanting of
low stone posts or even a pile of bricks, is a sign of the shift, especially when the
road widens again soon afterwards. But even when the spatial semiotics are
absent, the personal ones are present. There is the questioning look or the
approach of someone wanting to help but clearly also wanting to know.
Institutions have been retained from earlier periods, one of which is the
nadorgi (from nazara, "to sight") who, in Nadim's Cairo hara, was responsible
not only for overseeing proper behavior between male and female hara residents
but also for spotting strangers. As she describes it, the
nadorgi . . . is responsible for keeping an eye on those entering the harah and detecting
their movements. He is usually someone whose shop or house is close to the entrance to
the harah where he remains most of the time. Besides being a source of information
concerning external movement into the harah, the nadorgi can provide equally valuable
insights into the internal movements of the various harah residents. . ..Whenever illegal
activities occur in the harah, and in most cases this is the smoking or trading of hashish,
the role of the nadorgi becomes vital since he quickly warns of the entrance of outsiders
into the harah. The nadorgi will approach the outsider under the pretense of wishing to
help him find whomever he wants. This tactic serves two purposes: first, it detains the
intruder and secondly, it provides the nadorgi with information about the outsider's
destination and contact.49
We shall return to this point when we investigate the other function of the
neighborhood in the Arabo-Islamic city, namely, defensive space.
The Neighborhood as a Key Element in Civil Society and the State
The final way in which Islam shaped the traditional Arabo-Islamic city was
through neglect, ironic as that may seem. By failing to concern itself with
matters of day-to-day maintenance Islamic states often encouraged the vitality of
other sub-state functional units. One of these was definitely the residential
neighborhood.
The rather more rigid segregation between commercial and residential quarters
in the classic Islamic city has been attributed to the need to separate private (i.e.,
female) from public (i.e., male) space. Whatever the cause, such segregation
certainly did have important effects. It left to the residential areas a large
measure of autonomy, since many of the public functionaries (the supervisors of
the market places or the supervisors of public morals) operated largely in the
commercial sections of the city. Neighborhoods handled many of their internal
functions on a more ad hoc basis, being unable to afford more commercialized
services. For example, in the aswaq (markets), adjacent merchants might hire a
170 Janet L. Abu-Lughod
guard, but in the neighborhoods more informal arrangements were likely. Having
functions that could only be performed on a neighborhood-wide basis certainly
strengthened the cooperation within the district. It must also be acknowledged,
however, that before the modern era which emphasized the separation of population
on the basis of class, the neighborhood often was composed of a leading
family-household, surrounded by poorer families which had a clientele relationship
to the dominant household. In such cases, what we think of as municipal
services were actually provided by the dominant household.
A second factor that strengthened the neighborhood was its role as protector.
I would like to explore the issue of turf and defended neighborhoods because I
find the literature produced by Orientalists on the role of dhuciir (militant),
futuwwa (chivalrous society), etc., in the medieval Islamic city highly deficient in
sociological sense. I have been struck over and over again with the fact that the
traditional Arabo-Islamic city was designed to maximize what Newman has
termed defensible space.
In the introduction to his book, Defensible Space: Crime Prevention through
Urban Design, Oscar Newman writes that he is trying to find an architectural
solution to the rising disorder in American cities. He claims:
Architectural design can make evident by the physical layout that an area is the shared
extension of the private realms of a group of individuals. For one group to be able to set
the norms of behavior and the nature of activity possible within a particular place, it is
necessary that it have clear, unquestionable control over what can occur there. Design can
make it possible for both inhabitant and stranger to perceive that an area is under the
undisputed influence of a particular group, that they dictate the activity taking place
within it, and who its users are to be. This can be made so clearly evident that residents
will not only feel confident, but that it is incumbent upon them to question the comings
and goings of people to ensure the continued safety of the defined areas. Any intruder will
be made to anticipate that his presence will be under question and open to challenge; so
much so that a criminalcan be deterred from even contemplating entry.
Defensible space is a model for residentialenvironmentswhich inhibit crime by creating
the physical expressionof a social fabric that defends itself.50
Certainly, what Newman has just described is the Arabo-Islamic semi-private
quarter par excellence. But is the picture as benign as he has drawn it? Yes, the
neighborhood defends itself-but perhaps it is defending its criminal activities or
its warfare with the rest of society. Two non-benign parallels present themselves.
These are boys' gangs and the militia turf in embattled Beirut. Both offer a
seamier view of the defensible space advocated by Newman.
Boys' gangs were certainly a feature of Arabo-Islamic cities in the past and
continue to be present today. Futuwwa, dhucar, etc. may be given exotic
meaning and significance by Orientalists, but to a sociologist they have a most
familiar mien. We recognize the organization of local young males for the
defense of their quarter-whether such organizations are called Latin Kings or
Awliid al-Hiira. We recognize the gang leader, whether he is called Jeff Fort or
Muhammed ibn 'Ali, and we recognize his role, whether he is called Doc or
The Islamic City 171
Zacim (Boss). Even the so-called codes of chivalry are to be found in boys'
gangs.
Urban sociologist, Gerald Suttles, predated Newman in his conceptualization
of a "defended neighborhood" which he defined as a "residential group which
seals itself off through the efforts of delinquent gangs, by restrictive covenants,
by sharp boundaries, or by a forbidding reputation."51He goes on to specify the
conditions under which defended neighborhoods become important in cities (and
here he is discussing places like Chicago, not Cairo, and yet the applicability is
obvious):
Granted the inability of formal procedures of social control to detect and forestall all or
even most forms of urban disorder, some additional mechanisms seem necessary for the
maintenance of order. Among the available mechanisms, a set of rules governing and
restricting spatial movement seems a likely and highly effective means of preserving order.
Such a set of rules has some fairly obvious advantages: it segregates groups that are
otherwise likely to come into conflict; it restricts the range of association and decreases
anonymity; it thrusts people together into a common network of social relations that
overlap rather than diverge from one another.52
Residents are particularly likely to intensify their defense when the order in the
outside society becomes weakened.
Historically, in Arabo-Islamic cities, the neighborhood has been in dialectical
process with the external society. When central power was strong and when the
city-wide hierarchical structure was working smoothly, agents of the central
administration operated within the neighborhoods to provide information to the
center and ensure conformance with central directives. This was certainly the
case with the sheikh of the hara in Cairo at certain points, when he was
essentially an informer for the police as well as for the muhtasib (inspector of
morals). He acted, in his capacity as real estate expert, to "steer" or supervise
who should have access to vacant dwelling units in the quarter. Sometimes, the
neighborhood was an administrative subset of the state.
More often, however, the quarter played the opposite role, that of a defended
neighborhood, particularly when chaos reigned. One reads, in the historical
accounts, of civil strife/invasions/street battles, the recurring phrase, "and people
closed the gates to their harat." Alternatively, to gain control over the city,
conquerors always had to destroy the gates to the harat, as Napoleon's forces did
when they invaded Cairo.
One has only to think of contemporary Beirut to have these phrases take on
fuller meaning. During the height of disorder, virtually every block belonged to a
different group or faction. Checkpoints blocked entry and exit to these defended
territories. Often, barricades were constructed at the boundaries. The opposite
side of defense was also evident. During the Israeli siege, neighborhood assistance
was organized by block committees which allocated vacant apartments,
oversaw the rationing of water use, and distributed food and medical relief as
needed. One cannot resist reading back into history to evaluate some of the roles
neighborhoods formerly played in the Arabo-Islamic city.53
172 Janet L. Abu-Lughod
CONCLUSIONS AND A NOTE OF WARNING
In the first part of this paper we attempted to deconstruct Orientalist thinking
about the Islamic city by showing not only that the idea itself was "created" on
the basis of too few cases but, even worse, was a model of outcomes rather than
one of processes. By that I mean that the goal was to generalize about a specific
form of city at one long historic moment without unpacking the various causes
of that particular outcome. That form was then equated with the Islamic city,
regardless of whether there was anything particularly Islamic about the causes.
The reason it is important to criticize this approach is that in a number of
Arab countries today planners are trying to recreate Islamic cities-but by
means which are terribly inappropriate because they focus on outcomes, rather
than processes. Such planners hope, by edict and ordinance, to preserve and to
build anew cities on an Islamic pattern. It should be clear by now, however, why
this approach is likely to fail.
Cities are processes, not products. The three Islamic elements that set in
motion the processes that give rise to Islamic cities were: a distinction between
the members of the Umma and outsiders, which led to juridical and spatial
distinction by neighborhoods; the segregation of the sexes which gave rise to a
particular solution to the question of spatial organization; and a legal system
which, rather than imposing general regulations over land uses of various types
in various places, left to the litigation of neighbors the detailed adjudication of
mutual rights over space and use. These three factors were Islamic, per se.
However, in addition, the historic cities that developed in Arabo-Islamic lands
in premodern times were deeply influenced by such non-Islamic factors as
climate, terrain, technologies of construction, circulation, and production, as well
as political variables such as the relation between rulers and the ruled, the general
level of intercommunal strife, and fluctuations in the degree of internal and external
security. Furthermore, the nature of any Islamic city at any point in time
was the result not only of the contemporaneous nature of these variables but the
inherited forms which took shape under earlier and different circumstances.
It is clear, then, that one does not have the capacity to recreate Islamic cities
by edict. One has only the capacity to create conditions that might set in motion
processes which, in the past, generated the forms of the traditional city in the
Arabo-Islamic world. But it must be recognized that one cannot do that without
being willing to live with the three conditions mentioned above, namely: (I) juridical
distinctions between Muslims and/or citizens and outsiders; (2) segregation
by gender and a virtually complete division of labor according to it; and (3) a
fully decentralized and ex post facto system of land use and governmental
regulation over space. In today's world, these three are considered retrogressive.
First, modern states accord basic rights and responsibilities in an egalitarian
manner-at least in theory. Where distinctions are made on the basis of ascribed
status they have attracted the criticism of the world. Hence, states such as South
Africa which make sharp juridical distinctions on the basis of race, or Israel and
certain Muslim theocracies that make similar distinctions on the basis of religion,
are accused of using racist criteria. And contemporary societies that are orga-
The Islamic City 173
nized on the basis of communal affiliations run the risk of fragmentation and
strife, as has occurred in Lebanon.
Second, throughout the world there has been a trend toward increased equality
between the sexes. Integration, not segregation, has been the ideal toward which
many cultures are moving.
Finally, modern municipal governments stress the provision of community
facilities through a centralized system and stress the establishment of laws that
apply to whole classes of places and uses, that is, zoning laws, building codes,
street alignments, subdivision regulations, etc. Such regulations, as we have tried
to show, are the antithesis of the assumptions and mechanisms of property law
under Islamic legal approaches.
Therefore, none of the conditions still exist which would permit us to reconstruct
Islamic cities by design. Only a view of the Islamic city such as that held
by earlier Orientalists would allow one to even entertain such a notion.
That is not to say, however, that we could not build better cities in the
contemporary Arabo-Islamic world if we paid closer attention to some of the
true achievements of the past and if we learned from them. The historic Islamic
city often achieved community, privacy and beauty. It would be wise to seek
these same goals, even though the old means are no longer available. But since
cities are living processes rather than formalistic shells for living, they cannot be
built by us. We can only encourage them to grow in the desired direction. Can
we nurture neighborhoods that are supportive but not defensive? Can we foster
privacy not for women alone but for households? Can we guard the rights of
neighbors while still applying laws consistently? That is the task Arab city
planners must set for themselves.
NORTHWESTERN UNIVERSITY
ACKNOWLEDGMENTS
Janet Abu-Lughod is Professor of Sociology, Geography, and Urban Affairs
and Policy Research at Northwestern University. The support of the Center for
Urban Affairs and Policy Research is gratefully acknowledged.
Part I of this article was presented initially at the November 1985 meetings of
the Middle East Studies Association. A preliminary version of Part I1 was
presented to a May 1984 colloquium sponsored by the Middle East Center and
the Graduate School of Architecture and Planning at U.C.L.A.
NOTES
'w. Marqais, "L'Islamisme et la vie urbaine," L'Academie des Inscriptions et Belles-Gttres,
Comptes Rendus (Paris: January-March 1928), pp. 86-100.
'G. Marqais, "L'urbanisme musulman," 5e Congr2s de la Fideration des Sociitis Savanres de
I'Afrique du Nord (Algiers, 1940), which was reprinted on pp. 219-31 of the collection of his works,
Melanges d'hisroire et d'archiologie de I'occident musulman, tome I, articles et confkrences de
Georges Marqais (Alger, 1957).
174 Janet L. Abu-Lughod
3 ~ .Marqais, "La conception des villes dans I'Islam," Revue d'Alger, 2 (1945), 517-33.
4 ~ .Marqais, "L'urbanisme musulman" as reprinted in MPlanges. . . , my translation from
pp. 230-3 1.
5 ~ . "Urbanisme mkdikval et droit musulman," Revue des ~ t u d e sIslamiques, 15Brunschvig,
(1947), 127-55. I should point out here that the hierarchical arrangement of trades within the suq is
ignored and Orientalists failed to see these as "economically determined."
'0tto Spies, "Islamisches Nachbarrecht nach schafaitischer Lehre," Zeitschrift fur vergleichende
Rechtswissenschaft, 42 (1927), 393-421.
7 ~ e eIbn al-Imam, translated by Barbier under the title, "Les droits et obligations entre propriCtaires
d'hbritages voisins," Revue AIgPrienne et Tunisienne de Ligislation et de Jurisprudence, 1 and
2 (1900-1901).
'see Kitiib al-iCliinbi-ahkiim al-bunyiin.
'G. von Grunebaum, "The Structure of the Muslim Town," Islam: Essays in the Nature and
Growth of a Cultural Tradition, Memoir No. 81, The American Anthropological Association (Ann
Arbor, 1955). Reissued in second edition in London in 1961.
10
G. von Grunebaum, Medieval Islam (Chicago, 1946). Quotation taken from p. 173 of the second
edition, 1954.
"G. von Grunebaum, "The Structure of the Muslim Town," 1961 edition, p. 146.
I21bid.,pp. 146-47.
"shades of Max Weber's The City, which will not surface until the work of Ira Lapidus, Muslim
Cities in the Later Middle Ages, to be covered later.
I4see Louis Massignon, "Les corps de mCtiers et la citC Islamique," Revue Internationale de
Sociologie, 28 (1920), pp. 473-87.
" ~ l b e r tHourani and S. M. Stern, eds., The Islamic City (Philadelphia, 1970), based on an Oxford
University conference held in 1965, to be discussed below.
16
Roger LeTourneau, Ft?savant le protectorat. ~ t u d eiconomique et sociale d'une ville de Ibccident
musulman (Casablanca, 1949).
I7c.Coon, Caravan (New York, 1951). The chapter on "Town and City" appears on pp. 226-59.
I 8 ~ o g e rLeTourneau, La vie quotidienne a F6s en 1900 (Paris, 1965).
" ~ o g e rLeTourneau, Les villes musulmanes de I'Afrique du Nord (Alger, 1957).
?0
Roger LeTourneau, Fez in the Age of the Marinides, Besse Clemen, trans. (Norman, Oklahoma,
1961).
2 1
Jacques Berque, "Medinas, villesneuves et bidonvilles," Les Cahiers de Tunisie (1958), pp. 5-42.
22
Jean Sauvaget, Al6p: Essai sur le diveloppement d'une grande ville Syrienne, des origines au
milieu du XIXe sit?cle(Paris, 1941). In two volumes.
23
Jean Sauvaget, "Esquisses d'une histoire de la ville de Damas," Revue des Etudes Islamiques, 8
(1934), 421-80.
24
Typical of this is Xavier de Planhol's The World of Islam (Ithaca, 1959). Following an earlier
distinction made by the Mar~aises,de Planhol rigorously contrasts new towns (princely and otherwise)
founded by Islamic dynasties, more likely to be found in North Africa, with older towns of
RomanIByzantine base, upon which Islamic characteristics were superimposed, more likely to be
found in the Fertile Crescent. The question of which of these is the true city is never resolved.
25
See Ira Lapidus, Muslim Cities in the Later Middle Ages (Cambridge, Mass., 1967), reissued in a
student edition by Cambridge University Press, 1984. The revised preface tries to place his cases in
context.
26 .F~rstquotation comes from p. vii of the 1967 preface. The second quotation has been taken from
p. 191. In all fairness, the sentence that follows the one we have quoted does limit the reference to
Mamluk cities.
27
Janet Abu-Lughod, Cairo: I001 Years of the City Victorious (Princeton, 1971), although the
writing had been completed by 1967. Among the sources I depended upon were those incorporated
into H. A. R. Gibb and Harold Bowen, Islamic Society and the West, Vol. I, Part 1 (London, 1950),
especially Chapter VI, "The City: Industry and Commerce," pp. 276-313, which draws most of its
examples from late Ottoman Cairo.
28
Albert Hourani and S. M. Stern, eds., The Islamic City, cited above.
The Islamic City 175
29
See inter alia, my "Comments on the Form of Cities: Lessons from the Islamic City," Janus:
Essays in Ancient and Medieval Studies, L. Orlin, ed. (Ann Arbor, Michigan, 1975), pp. 119-30;
Paul Wheatley, "Levels of Space Awareness in the Traditional Islamic City," Ekistics, 253 (March
1977); R. B. Serjeant, ed., The Islamic City [on Sanca], (Paris, 1980). See also L. Carl Brown, ed.,
From Medina to Metropolis (Princeton, 1973). Nor is it only Westerners who are seeking the key to
the Islamic city. Throughout the Arab world there are scholars seeking precedents for a new/old
form of city building in tune with the culture and with Islamic values. See, for example, Islamic
Architecture and Urbanism, edited by Ayden Germen, and published by King Faisal University
(Dammam, Saudi Arabia, 1983). By far the best of these has just been published. Although it is now
too late to integrate this work in the present article, written before I received my copy, I strongly
recommend it to anyone thinking or writing about Islamic cities. See Besim Selim Hakim, ArabicIslamic
Cities; Building and Planning Principles (London, 1986).
'O~aleEickelman, "Is There an Islamic City?," International Journal of Middle East Studies, 5
(1974), 274-94. As revealed by the subtitle, we have again been promised too much since it describes
only one Moroccan town. However, rather than attempting to generalize, the article actually tries to
test some of the propositions about the Islamic city against the reality of one contemporary
Moroccan town, a significant methodological breakthrough. The book whose discussion of the
Islamic city makes most sense to me is one by another anthropologist studying a different Moroccan
town. See especially the introductory chapter of Kenneth Brown, People of SalP: Tradition and
Change in a Moroccan City. 1830-1930 (Cambridge, Mass., 1976).
3 1 ~ a n yof my reservations about the Islamic city have been triggered by discussions and conferences
with Muslims from the Arabian peninsula where, at present, the most serious and sincere
attempts are being made to devise an operational definition of the Islamic city in order to build
contemporary ones. Several students from that area, most recently at Northwestern University, have
forced me to think about this topic, if only because they were charged with studying how to do it. I
acknowledge their contribution here.
32
See, for example, Janet Abu-Lughod, "Preserving the Living Heritage of Islamic Cities," in
R. Holod, ed., Toward an Architecture in the Spirit of Islam (Philadelphia: Aga Khan Foundation,
1978), pp. 27-36; and "Contemporary Relevance of Islamic Urban Planning Principles," in Ayden
Germen, ed., Islamic Architecture and Urbanism, 1983, cited earlier. I take these points further in a
paper on the "Semiotics of Space in Islamic Cities," delivered May, 1984 to a conference on the
subject held at UCLA. The section that follows draws heavily on that unpublished paper.
33
I would not necessarily attribute this to Islam, but I would note that such a pattern was all too
often associated with its polities, for whatever reason.
34
Albert Hunter, Symbolic Communities: The Persistence and Change of Chicago's Local Communities
(Chicago, 1974).
35
(Cambridge, Mass., 1960).
36 .G~deonSjoberg has argued, in his The Preindustrial City (Glencoe, 1960), that eastern and
western versions were quite similar, due to their common levels of technology. However, this ignores
the fact that tribalism-ethnicity was seldom the organizing principle of spatial structure in western
medieval towns, nor was there as much separation between residence and business as there was in
cities where gender segregation was the rule (i.e., in ancient Greece and in Islamdom).
3 7 ~ e s i mS. Hakim, in his Arabic-Islamic Cities, argues this in convincing fashion, but he draws his
empirical evidence almost exclusively from Tunis. I doubt whether Islamic law can be the sole
explanation, for the cellular structure of communities in that climate and culture region long predates
the appearance of Islam, and the alternative to the "hara, mahalla, huma," etc. system found
throughout the Fertile Crescent and North Africa, namely, the tall qasr or apartment building found
in Yemen, southern Morocco, and Saudi Arabia, is equally Islamic, in being indigenously developed
and suited to the cultural and geopolitical climate.
38
See, for example, Henry Munson, The House of Si Abd Allah (New Haven, 1983). 

39
Samir Khalaf and Per Kongstad, Hamra of Beirut (Leiden, 1973). 

40
Fuad Khuri, From Village to Suburb: Order and Change in Greater Beirut (Chicago, 1975). 

41
See, for example, Leila Fawaz's study, Merchants and Migrants in Nineteenth Century Beirut 

(Cambridge, Mass., 1983).
176 Janet L. Abu-Lughod
42
Stanley Lane-Poole, The Story of Cairo (London, 1902), pp. 12-17.
43
See Chapter two of Soraya Altorki's Women in Domestic Groups: Ideology and Behavior of
Jiddah Elites (New York, 1986).
44
See, for example, my article, "Contemporary Relevance of Islamic Planning Principles," Ekistics,
47 (Jan.-Feb. 1980), 6-10. One should note, however, that this "privatization of public space" is not
exclusively a phenomenon of the Arabo-Islamic or Middle Eastern city. In fact, I have taken the
term from Lyn H. Lofland, A World of Strangers: Order and Action in Urban Public Space (New
York, 1973), who uses it to describe how Americans develop proprietary interests in public space.
45
Quoted from p. 174 of Nawal Nadim, "The Relationship between the Sexes in a Harah of Cairo,"
Doctoral Dissertation, Department of Anthropology, Indiana University (Bloomington, 1975).
46
See Chapter one of Veiled Sentiments: Honor and Poetry in a Bedouin Society (Berkeley, 1986).
47
Nawal Nadim, "The Relationship between the Sexes," p. 180.
48
Elizabeth W. Fernea, A Street in Marrakech (New York, 1975).
49
Nawal Nadim, "The Relationship between the Sexes," pp. 187-88.
50
Oscar Newman, Defensible Space: Crime Prevention through Urban Design (New York, 1972),
pp. 2-3.
" ~ e r a l dSuttles, The Social Construction of Community (Chicago, 1972), p. 21.
'?bid., pp. 31-32.
53
Comparative studies are always valuable because they prevent us from jumping prematurely to
the conclusion that our case is unique. In this connection it is interesting to note the parallels from an
entirely different case, Santa Domingo, during the 1965 revolution. The social organization within a
defended neighborhood is graphically portrayed by participant observer sociologist, JosC Moreno.
See his Barrios in Arms (Pittsburgh, 1970).