The Ancient City: From Fustel de Coulanges to Max Weber and beyond

Author(s): M. I. Finley
Source: Comparative Studies in Society and History , Jul., 1977, Vol. 19, No. 3 (Jul.,
1977), pp. 305-327
Published by: Cambridge University Press

Stable URL: https://www.jstor.org/stable/177994

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access
to Comparative Studies in Society and History

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 The Ancient City: From Fustel de
Coulanges to Max Weber and Beyond
M.I. FINLEY
University of Cambridge

The Graeco-Roman world, with which I am con

of the pre- Greek Near East, was a world of c
population, always a majority, most often lived
kind, hamlets, villages, towns, not in isolated
reasonable and defensible guess that, for the b
years, more and more of the inhabitants of Eur
western Asia lived in towns, in a proportion that
United States, for example, until the Civil W
guess is possible, since statistics are lacking fo
themselves were firm in their view that civiliz
in and because of cities. Hence the growth of t
relentless accompaniment of the spread of Gr
eastward after the conquests of Alexander as f
the west from Africa to Britain with the Rom
number of towns rose into the thousands.
So self-evident did the urban underpinning of civilization seem to the
ancients that they scarcely engaged in a serious analysis of the city.
They did not even attempt a formal definition (apart from administrative
"definitions" to which I shall return briefly). Writing a glorified guide-
book of Greece late in the second century A. D., Pausanias dismissed
the claim of a little town in central Greece to city status: "no govern-
ment buildings, no theatre, no agora, no water conducted to a fountain,
and ... the people live in hovels like mountain cabins on the edge of a
ravine" (10.4.1). That at least points to a definition: a city must be more

This is a considerably revised and enlarged version of a paper I read to the annual
conference of the Urban History Group in Churchill College, Cambridge, on 7 April 1976.
For helpful criticism I am grateful to Peter Garnsey, Keith Hopkins and C. R. Whittaker,
all of whom dislike the "intellectual history" framework of the analysis.
1 This subject has not been properly investigated; as a beginning, see J. Pecirka,
"Homestead Farms in Classical and Hellenistic Hellas," in Problemes de la terre en
Grece ancienne, Finley, ed. (Paris and The Hague: Mouton, 1973), pp. 113-47; E. M.
Wightman, "The Pattern of Rural Settlement in Roman Gaul," in Aufstieg und Nieder-
gang der rbmischen Welt, ed. H. Temporini and W. Haase, vol. II 4 (Berlin and New
York: Walter De Gruyter, 1975), pp. 584-657.
305

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
306 M. I. FINLEY

than a mere conglomeration of people; there are necessary conditions

architecture and amenity, which in turn express certain social, cultu
and political conditions. Many centuries before, Aristotle had pointed
the same direction. The siting and planning of a town, he wrote in t
Politics (1330a34ff), involves four considerations: health, defence,
suitability for political activity, and beauty.
Pausanias, it will have been noticed, did not object to the pretentious
little town on the grounds of its small size. And Aristotle saw in small-
ness a virtue, even a necessary condition: Babylon, about which he must
have known very little, was for him an epithet, a symbol of elephantiasis,
hence a negation of the true city (Politics, 1265a10ff). In his day, in fact,
there was probably no town in the Graeco-Roman world with a popula-
tion greater than 125,000 or 150,000, probably not half a dozen exceeding
40,000 or 50,000 (figures that one may double if the inhabitants of the
city's agricultural hinterland are included). The trend after Aristotle was
for a substantial growth in urban population, but if Rome and possibly
Carthage finally reached perhaps half a million, the norm was even then
closer to Pompeii, with some 20,000 inhabitants at the time of its de-
struction in A.D. 79.
It will also have been noticed that neither Aristotle nor Pausanias was
concerned with the "administrative definition" of a city, though the
former was writing about the autonomous city-state, thepolis in Greek,
the latter about a tiny town in one of the provinces of the Roman empire.
Any territorial state which has a number of conglomerations within its
borders must necessarily define and distinguish among those conglom-
erations, for purposes of police, taxation, road maintenance and all the
other demands and services that social life entails. A survey of such
definitions and distinctions today alone would reveal a bewildering
variety, because these are technical matters marginal to a study of the
city, and I shall largely ignore them.
The phrase "city-state" which I just used with reference to Aristotle
is an English convention in rendering the Greek word polis. This con-
vention, like its German equivalent, Stadtstaat, was designed (I do not
know when or by whom) to get around a terminological confusion in
ancient Greek: the word polis was employed in antiquity for both
"town" in the narrow sense and "city-state" in the political sense.
When Aristotle examined the right conditions for siting a town, he wrote
polis, the word he used hundreds of times in the Politics for his main
subject, which was the city-state, not the town. He had no reason to fear
that his readers would be led astray, as modern historians allow them-
selves to be.
For Aristotle, as for Plato before him, the polis arose because of the
incapacity of the two prior forms of human association, the household

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 THE ANCIENT CITY 307

and the larger kinship grouping, to satisfy all the legitimate ne

members. Self-sufficiency, autarky, was the objective, and
structured and constituted polis should be able to attain tha
for the unavoidable lack of essential natural resources, for w
which alone) foreign trade was admissible.2 It is self-evide
tarky is a nonsensical idea for a town. Plato and Aristotle d
nonsense: they took city and hinterland, town and country, tog
unit, not as distinct variables in competition or conflict, actual
tial. Even those farmers who lived outside the town were in
thepolis. What we commonly call "class conflict" is invariab
"rich" and "poor," not between landowners and manufa
between labour and capital, or between masters and slaves.
of property and property ownership are only about land. Alt
distinguished between gentlemen-farmers, living in the town, a
ing farmers in the countryside, that was a distinction betw
leisure, who were alone capable of the good ife, and men w
for their livelihood, again not a town-country\distinction. T
farmer ranked higher on the scale than the artisan, but that wa
of morality.
The ancient city was soon to lose its autonomy. The proc
soon after Aristotle died, with the creation of the Hellenist
chies, and it was completed when the Romans embraced the
world, and much else, within their empire. Yet even then, and t
of antiquity, each city normally included a rural hinterlan
extent, often of very considerable extent, within its recognized
The city without a territory was a rare phenomenon, largely re
coastal communities of a peculiar kind. What is more import
purposes, the traditional unity of town and hinterland-poli
cal and residential-went on unchallenged. Both Hellenistic
emperors, for example, acknowledged that the hinterland w
gral part of the city for tax purposes. The same held true in th
of municipal citizenship, which retained genuine value,juridi
ically and psychologically, after the disappearance of city
It will not have escaped notice that I have so far avoided defin
I mean by a city. Neither geographers nor sociologists nor
have succeeded in agreeing on a definition. Yet we all know s
what we mean by the label, in general terms: no one will d
there was a city of Athens which was both physically and c
distinct from the city-state of Athens. The block in definition

2 There are important nuances distinguishing Plato and Aristotle, espec

spect to internal trade: see my " Aristotle and Economic Analysis," Past an
47 (1970), 3-25, reprinted in Studies in Ancient Society, Finley, ed. (Londo
Routledge and Kegan Paul, 1974), pp. 26-52.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
308 M. I. FINLEY

the difficulties, apparently insuperable, of incorporating all the esse

variables without excluding whole periods of history in which
know cities existed, and on the other hand, of settling for a least co
denominator without lodging on a level of generality that ser
useful purpose. The more sophisticated factor analyses in contem
urban geography and sociology, with as many as one hundred
ables,3 most of which were absent from the ancient city (and th
eval and Renaissance cities as well), neatly reflect the unbridge
divide in the history of cities created by the Industrial Revolut
That is indeed the conclusion (or the assumption) of historia
sociologists specializing in the modern city, and I accept that th
right to ignore the ancient city. The reader must then be wary o
titles: the classic volume of the Chicago urban school, published
under the title, The City, is a good example. One can only wis
plead, that they have the courage of their convictions and n
impelled to make a cultural gesture to the distant past with a senten
two, or perhaps a paragraph, more often erroneous than not.
Handlin writes in introducing the volume called The Historian a
City (a title promising, even more than The City, something th
there), "The ancient world had been a world of cities, but each h
a world unto itself," he is wrong in fact and he also confuses a W
ideal type (he cites Weber at this point) with a statement of f
when Thernstrom suggests that it "may some day be possible to
a model of the process of urbanization which applies equally w
ancient Athens and contemporary Chicago," he presupposes a s
reductionism, stripping urban history down to demography an
and geographical mobility. His qualification, that it would be un
ing to "search for such regularities today," is merely a bow to d
ties in method and in the availability of information, not a recognit
the irreducible structural difference between pre-industrial and
trial cities.6
In my view, the starting-point for the historian of the ancient city must
be the attachment between hinterland and city. The geographer Strabo,
writing at the beginning of the Christian era, promised (4.1.5 and else-

3 See B. J. L. Berry, ed., City Classification Handbook (New York: Wiley, 1972). A
French inquiry managed to achieve a total of 333 variables; see Henri Lefebvre, La
Revolution urbaine (Paris: Gallimard, 1970), p. 67.
4 Current discussion of the problematique of urban culture "is concerned in fact with the
cultural system characteristic of industrial society, and, for the majority of distinctive
traits, of capitalist industrial society": M. Castells, "Structures sociales et processus
d'urbanisation: analyse comparative intersocietale," Annales (E. S. C.), XXV (1970),
1155-99, at p. 1157. Cf. the opening chapter of Lefebvre, op. cit.
5 (Cambridge, Mass., and London: Harvard University Press, 1963), p. 2.
6 "Reflections on the New Urban History," Daedalus, C (Spring 1971), 359-75,
reprinted in Historical Studies Today, F. Gilbert and S. R. Graubard, eds. (New York:
W.W. Norton, 1972), pp. 320-36, at p. 324.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 THE ANCIENT CITY 309

where) that the newly conquered western and northern barba

would become civilized as soon as they settled down to agricult
therefore to urban life. That combination is illuminating. No
author considered the relationship between the urban and the
sector in terms of the acquisition, production and exchange o
That theme is not only absent from the literature that has survive
antiquity, apart from the moral and cultural concerns I have
noted, but it continued to be incidental, at best, until the developm
the modern science of political economy. Montesquieu devo
books to commerce but saw nothing in the city as such to req
attention, nothing remotely comparable to the third book of
Smith's Wealth of Nations a generation later, with its well-know
ing:
The great commerce of every civilized society is that carried on between the inhabitants of
the town and those of the country.... We must not... imagine that the gain of the town is
the loss of the country. The gains of both are mutual and reciprocal, and the division of
labour is in this, as in all other cases, advantageous to all the different persons employed in
the various occupations into which it is subdivided.

That last point was soon challenged, for example by Marx and Engels in
The German Ideology: "The division of labour inside a nation leads at
first to the separation of industrial and commercial from agricultural
labour, and hence to the separation of town and country and a clash of
interests between them" (my italics).7 Such a disagreement is, in itself,
evidence of the arrival of the town as a subject of investigation.
My subject, however, is not the pre-industrial city in general but the
ancient city. I ask you to bear with me while I assume that the ancient
city is a distinct and distinguishable category.8 What criteria have histo-
rians or sociologists established with which to differentiate the ancient
city from cities in other eras and other societies, and then to differentiate
among the various kinds of ancient city? In purely quantitative terms,
the sad answer is: very little worthy of serious consideration. Most
historians of antiquity appear never to have asked themselves that
question; a few, in a famous polemic that began late in the last century
and continued into the first decades of our own, argued that the differ-
ences between the ancient and the modern city were merely quantita-

7 English ed. of Parts I and III, by R. Pascal (London: Lawrence & Wishart, 1938), p. 8.
The work was completed in 1846, and the fact that this part was not published in Marx's
lifetime is irrelevant to my argument.
8 The view that all pre-industrial cities, of the ancient East, classical antiquity and the
Middle Ages, resemble one another closely has been projected by G. Sjoberg, The
Preindustrial City (Glencoe, Ill.: Free Press, 1960), pp. 4-5. In his pursuit of "structural
universals," Sjoberg divides society into three types, "folk," "feudal" and "industrial-
urban" (p. 7), and asserts that in "feudal" societies (among which he includes the ancient),
"relative to the total population, urban residents are few" (p. 11). From that complex of
false starts there is no possible recovery.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 31I M. I. FINLEY

tive: smaller population, less commerce, less manufactu

toritas of Eduard Meyer, Julius Beloch and, more recentl
Rostovtzeff stilled opposition and even discussion, at least
cient historians.9
Whereas ever since Gordon Childe discovered the "urban revolu-
tion" there has been a growing and increasingly sophisticated literature
about the beginnings of urbanism in Mesoamerica, Mesopotamia and
ancient China,10 and whereas the unceasing literature from the early
nineteenth century has reached an unmanageable quantity about the
'rise of towns" (a curiously preemptive label for the rise of the medieval
city), the intervening thousand years appear as a vacuum, or perhaps I
should say a prohibited space. There is considerable publication about
what is sometimes grandiloquently called "ancient town-planning," and
no one will dispute that this is part of urban history, as are demography,
drains and sanitation.11 But a town is more than the mere arithmetical
total of layout and drains and inhabitants, and it is remarkable that the
ancient city qua city has aroused so little interest. Had it not "disap-
peared" at the end of antiquity, it would not have had to "rise" again:
that simple logic alone should have forced attention on it.
There have been exceptions, of course, and perhaps even more ap-
parent exceptions. Momigliano has recently written, "When one speaks
of the ancient city (citta) as of the society within which institutions
functioned and ideas circulated, the first modern historian who comes to
mind is Fustel de Coulanges."12 Fustel's La Cite antique was published
in 1864 and had a tremendous impact in certain circles. Writing in 1891,
W. J. Ashley pointed out that "especially in England... it fell in with all
that current of thought which was then beginning to turn into the direc-
tion of social evolution, comparative politics, and the like. For a year or

9 Thus, Mason Hammond, The City in the Ancient World (Cambridge, Mass: Harvard
University Press, 1972), carries the identification of city with city-state so far as to exclude
from his "preliminary definition" the "administrative center, however much built up, of a
state which is organized socially and politically throughout its occupied territory. The
capital of such a state is merely the nucleus of the united territory without any characteris-
tics peculiar to itself as against the rest of the state" (p. 6). Perhaps the potential reader
should also be warned that Hammond begins by saying that "the impetus of this book was
the question whether the emergence of cities in Italy resulted from a natural development
of the Indo-Europeans or whether it reflected Greek institutions planted in South Italy."
10 See, e.g., P. J. Ucko et al., eds., Man, Settlement and Urbanism (London:
Duckworth, 1972); R. McC. Adams, The Evolution of Urban Society (Chicago: Aldine,
1966); Paul Wheatley, The Pivot of the Four Quarters: A Preliminary Inquiry into the
Origin and Character of the Ancient Chinese City (Chicago and Edinburgh: Edinburgh
University Press, 1971).
11 See especially Roland Martin, L'urbanisme dans la Grece antique (2 ed., Paris: A. &
J. Picard, 1975). Cf. R. F. Wycherley, How the Greeks Built Cities (2 ed., London:
Macmillan, 1973); L. Homo, Rome imperiale et I'urbanisme dans l'antiquite (Paris: Albin
Michel, 1951).
1Z "La citth anticadi Fustel de Coulanges," Rivista storica italiana, LXXXII (1970),
81-98, reprinted in his Quinto contributo alla storia degli studi classici e del mondo antico
(2 vols., Rome: Edizioni di Storia e Letteratura, 1975), I 159-78.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 THE ANCIENT CITY 311

so, the final piece of advice which schoolmasters gave to men

going up for scholarships at the Universities was to read
antique."13 Willard Small's translation was published in t
States in 1873, and the copy in my possession, dated 1894, is
eighth edition. In the academic world, on the other hand, int
largely restricted to France among historians and apparently
lawyers in Italy.14
Now the first, and for our purposes most important, thing to
La Cite antique is that its subject is the city-state, not the t
French and Italians have not adopted the "city-state" conven
that cite (or citta), likepolis, can mean either ville, an urban cent
the words of the dictionary of the Academie, "La Constitution de
l'Etat." Fustel clearly did not mean, or concern himself with, ville. His
subject was the origin of private property, the origin of the state, and the
"revolutions" within the ancient state, and his book has a thesis,
drummed in repeatedly. I quote one typical passage:
There are three things which, from the most ancient times, we find founded and solidly
established in these Greek and Italian societies: the domestic religion; the family; and the
right of property-three things which had in the beginning a manifest relation, and which
appear to have been inseparable. The idea of private property existed in the religion itself.
Every family had its hearth and its ancestors. These gods could be adored only by the
family, and protected it alone. They were its property.15

The inextricable family-religion-property link then moved into the larger

kinship unit, the gens, and ultimately into the earliest state. Clearly the
succession, family-gens-state, was for Fustel an historical one, not
merely a conceptual one; to that extent he was following Aristotle, who,
however, never imagined ancestor worship and the cult of the fire (the
hearth) to be the fountainhead of private property. Nor did, nor could,
any ancient author have shared Fustel's addiction to the newfangled
Aryan doctrine: he included the Indians of the Rigveda and (because of a
then common error) the Etruscans along with the Greeks and the Ital-
ians in his scheme of evolution. That was the extent, and the limit, in this
book, of Fustel's famed pioneering role as a comparativist.
For a historian, like myself, who greatly admires Fustel's subsequent
work, such as his fundamental study of the late Roman colonate or his
work on medieval France and Germany, La Cite antique is not easy to

13 "The English Manor," an introductory essay to the English translation of Fustel's

The Origin of Property in Land (London 1891), p. ix. The latter was first published in the
Revue des Deux Mondes (1872), and was then reprinted in his Questions historiques, C.
Jullian, ed. (Paris: Hachette, 1893), pt. II.
14 On the latter, see the important inaugural lecture (1913) by V. Arangio-Ruiz, Le genti
e la citta, reprinted in Scritti giuridici per il Centenario della Casa Editrice Jovene
(Naples, 1954), pp. 109-58.
15 P. 78 of the American translation mentioned in the text, p. 69 of the 2nd French ed.
(Paris, 1866).

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
312 M. I. FINLEY

come to terms with. Its deployment of massive knowledge of

and Latin sources goes hand in hand with a lack of source crit
is almost incredible. Despite the deliberate refusal to mentio
modern author, the book is polemically ideological in a subtle and
complex way; so was its reception, as Ashley perceived; so, too, as
Ashley explained sadly, was the cool reception of his later, medieval
works. In the latter his wide reading emerged on every page, his way
with the sources was impeccable, the centrally creative force of religion
was allowed to drop from sight, but the insistence that as far back as we
can trace civilized societies there was private property, not communal
ownership, remained a central theme.
Yet La Cite antique was by no means without notable academic
impact in certain directions. In the first place, the book became decisive
in the development of Durkheim's ideas about religion and therefore in
its influence on the Durkheim school.16 Second, Fustel along with
Maine and Morgan, all three working independently in the halcyon days
of social evolutionism, gave kinship the central role it holds to this day in
social anthropology. And third, through Paul Guiraud and even more
Gustave Glotz, the book has left its stamp on French ancient historians.
In Glotz's classic La Cite grecque, published in 1928, which is also a
work on the city-state, not the town, the opening pages are devoted to
Fustel. "The grandiose construction of Fustel de Coulanges," he said,
"impels admiration ... Nevertheless today it is impossible to accept all
his conclusions" (a verdict echoed by Henri Berr in the introduction).
And what were Glotz's reserves? " History does not follow a rectilinear
road": in addition to the family and the city, we must consider the
individual.
"In the era when the Cite antique was published," Glotz also wrote,
"no one since the time of Montesquieu had employed (the comparative
method) with such mastery." I am unable to explain so uninformed a
judgment by a major historian; not even Glotz's overt refusal to employ
the comparative method himself is a sufficient explanation. The "com-
parative method" of La Cite antique is largely an illusion, since Fustel
claimed to be revealing a single, Aryan pattern of evolution-a typical
statement is, "The religion of the dead appears to be the oldest that has
existed among this race of men" 17-and, anyway, in the century after
Montesquieu the volume of genuine comparative studies had grown to
immense proportions. Yet, as Durkheim pointed out, by ignoring the
available ethnographic evidence, Fustel came to a false conception of
the Roman gens.18 We may nevertheless concede to Evans-Pritchard

16 See S. Lukes, Emile Durkheim (London: Allen Lane, 1973), pp. 58-63.
17 Ancient City, p. 28 English ed. (p. 20 French).
18 Preface to vol. I of L'Ann&e sociologique (1896/7).

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 THE ANCIENT CITY 313

that La Cite antique marked "the dividing-point between th

tive and dogmatic treatises of such writers as Turgot, C
Saint-Simon and Comte on the one side" and the "detailed
and "scholarly treatment" that characterize the work of Du
Hubert and Mauss.19 We may also agree that Fustel made a
able contribution in the recall from near-oblivion of the per
kinship institutions within the ancient city-state. However, t
of the city (whether town or city-state), ancient or medieval or
cannot be sufficiently analysed in terms of the cult of ancestors
of fire and the conflict within the developed state between
group and the individual.
The most notable of the theories of social evolution to have
on the basis of comparative studies, in the century between
quieu and Marx was the theory of the four stages, hunting
agricultural and commercial, through which early man evolve
proponents were in Scotland and France, and with John Mill
what Meek has now called "in effect" "a materialist con
history." In the introduction to his Observations concernin
tinction of Ranks in Society, first published in 1771, Millar li
"the causes of those peculiar systems of law and governmen
have appeared in the world" the following: "the fertility or
of the soil, the nature of its [ a country's] productions, the
labour requisite for procuring subsistence, the number of i
collected together in one community, their proficiency in ar
vantages which they enjoy for entering into mutual transacti
maintaining an intimate correspondence."20
There is no trace of the four-stages theory in La Cite antiq
only was Fustel cognizant of the theory, at least in its French fo
he himself accepted it up to a point. In the opening paragraph
Origin of Property in Land (first published in 1872), he wrote in
critics: " It is obvious that when men were still in the hunting o
stage, and had not yet arrived at the idea of agriculture, it did n
them to take each for himself a share of the land. The theory of
speak applies to settled and agricultural societies.'21 But the

19 Introduction to R. Hertz, Death and the Right Hand, R. and C. Need

(London: Cohen and West, 1960), pp. 11-12.
20 Quoted from R. L. Meek, Social Science and the Ignoble Savage (
Cambridge University Press, 1976), p. 162. Perhaps surprisingly, Werner S
already called attention to this particular passage half a century ago in a sh
"The Beginning of Sociology," and regretted the neglect of Millar's Orig
"one of the best and fullest sociologies we possess," containing the kernel of
known under "the unfortunate rubric, 'materialist conception of histo
nerungsgabe fur Max Weber, M. Palyi, ed. (2 vols., Munich and Leipzig
Humbolt, 1923), I 11, 13-14.
21 p. 1 of the English translation (see n. 13 above). In ch. 4 of this essay, a c

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
314 M. I. FINLEY

parted radically, as he departed from Aristotle, replacing the mode of

subsistence by religion as the focus of attention and the key to the
formation and change of institutions. Ashley correctly observed that
even in his work on the colonate Fustel failed to take proper account of
"the economic as well as the constitutional or legal."22
So far as I know, the first man to insist on, and to formulate, an
"economic theory of town formation (Stddtebildung)," of "the neces-
sary relationship between the phenomenon of the town and the prevail-
ing economic system," was Werner Sombart in Der moderne
Kapitalismus, originally published in Leipzig in 1902.23 In that work he
presented a series of models, starting with the obvious working defini-
tion, "A town is a settlement of men who rely for their maintenance on
the products of foreign (or alien) agricultural labour."24 In the second
edition, fourteen years later, he introduced a slight modification by
adding the admittedly vague word "larger"-"a larger settlement.' 25
This definition, he explained, was designed to exclude the Landstidte of
the Middle Ages, in which the majority of the inhabitants exploited the
land themselves, as well as the "giant cities" of the ancient Near East,
of ancient India or of the type represented by Teheran today. That he did
not specify the cities of Graeco-Roman antiquity, or at least some of
them, can be explained by his concentration on his subject, the rise of
modern capitalism and therefore the rise of the city in the Middle Ages.
And the key idea in his definition of a town goes back to Adam Smith-
Sombart placed at the head of this section the same passage from Book
III of the Wealth of Nations that I quoted earlier, and he said explicitly
that his models were "'variations on a theme,' a theme formulated in
Adam Smith's words."26
In the long and historiographically fecund period between Smith and
Sombart there had of course been massive research into, and publication
about, towns. But the interest, insofar as it was more than mere local
history in the antiquarian sense, had always been in the evolution from
feudalism to capitalism, in the rise of the medieval city, in the Renais-
sance city, and in the subsequent modern developments. Occasional
remarks about the ancient city can be found, some of them penetrating,

de Laveleye, De la propriete et sesformes primitives (Paris, 1874), Fustel demonstrated

his ability to deal with ethnographic data when pressed. This chapter is actually entitled
"Of the Comparative Method."
22 Op. cit., pp. xlii-xliii.
23 My quotations appear on pp. 191 and 194 of vol. II.
24Ibid., p. 191.
25 1128. The second edition was a radically rewritten, restructured and enlarged work,
but the chapter on the city was not significantly altered in substance. All subsequent
editions of the original two-volume core of Der moderne Kapitalismus were merely
photographic reprints of the second.
26 1st ed., 1I 194.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 THE ANCIENT CITY 315

from Adam Smith on (and David Hume, too, one should alw
member), but they were by the way, incidental to the subject at
never elaborated. It would repay the effort to collect and exa
remarks, but I can stop briefly only for one man, Karl Biic
In 1893 Biicher, who had already produced a remarkab
statistical" study of the city of Frankfurt in the fourteenth an
centuries, published his Die Entstehung der Volkswirtschaf
of the National Economy), in which, building on an idea of R
he extended the old four-stage evolutionary theory by sugge
further stages in the history of the last, the commercial on
called closed household-economy, city-economy and national econ-
omy.27 This was the book that sparked off the dispute with the ancient
historians, now commonly referred to as the Biicher-Meyer con-
troversy, which was "won" by the latter, to their own satisfaction, as I
have already indicated.28
The year of Biicher's Entstehung, 1893, was also the year of the first
of Henri Pirenne's three famous articles in the Revue historique on
"The Origin of the Urban Constitutions of the Middle Ages," in which
he formulated the leading ideas that were to preoccupy him for so much
of his life.29 The rise of the medieval city, he insisted time and again, was
in the first instance "the product of certain economic and social
causes."30 These "economic and social causes," alas, turn out to be
nothing more than a mysterious "natural" process set in train by mer-
chants, and Pirenne quickly slid back to the very stress on jurisdiction
and Verfassungsgeschichte he had so powerfully condemned in others.
Apart from banalities about the "sterility" of the town, there is nothing
that rises above the purely descriptive level, intelligent, learned and
invaluable as he certainly was on that level. He admired Bicher's book
on Frankfurt, but in the latter's theoretical work, Pirenne warned his
students in lectures, he "was too much the economist and not enough
the historian, . . . his theories on economic development, however
stimulating, were not related to historical evidence."31 Only once, to my
knowledge, did Pirenne deign to discuss and dispute Biicher and Som-
bart, in a paper which I can best summarize as a medievalist's echo of

27 Biicher had published an earlier version of his theory in an obscure journal as far back
as 1876, but it received no attention until the appearance of Die Entstehung; see G. v.
Below, " Ueber Theorien der wirtschaftlichen Entwicklung der Volker...," Historische
Zeitschrift, LXXXVI (1901), 1-77, at p. 8.
28 See E. Will, "Trois quarts de siecle de recherches sur l'6conomie grecque antique,"
Annales (E.S.C.), IX (1954), 7-22; M. I. Finley, "Classical Greece," Proceedings of the
2nd International Conference of Economic History, Aix-en-Provence 1962, vol. I (Paris
and The Hague: Mouton, 1965), pp. 11-35.
29 The three articles are reprinted in the posthumous 2-volume Pirenne collection, Les
villes et les institutions urbaines (Paris and Brussels: Felix Alcan, 1939), I 1-110.
30 Ibid., p. 32.
31 Bryce Lyon, Henri Pirenne (Ghent: E. Story-Scientia, 1974), p. 146.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
316 M. I. FINLEY

the arguments of the "modernizing" ancient historians, conclud

with the latter, that the difference between modern capitalism a
"capitalism" that began in the twelfth century was "only a differ
quantity, not a difference of quality, a simple difference of intensity
difference of nature."32 Pirenne later heard, we are told, that Weber
surprisingly (if true), "caustically referred to him as that Belgian
evalist who knew no medieval economic and social history."33
Weber elsewhere protested that the historians had misunderst
Biicher's avowedly "ideal type" approach,34 but historians, whet
antiquity or of any other era, are customarily either allergic or
deaf to ideal types. Thus, the distinguished medievalist Geor
Below, more sympathetic than most to Biicher's contribution, nev
less concluded that the enterprise was doomed from the start bec
the concern for "norms": "It is precisely the deviations whi
interesting, or at least no less important than the rule.' 35 Eduard Me
was less comprehending, and Bucher had refused an invitation fro
editor of the Jahrbiicher fir Nationalokonomik und Statistik to r
Meyer on the ground, in his own words, that Meyer revealed "so
understanding of the economically essential."36 A few years lat
could not resist, and in a long essay, full of erudition and w
examined in detail the Athenian evidence adduced by Meyer and
loch, and he left their conclusions in tatters.37
Biicher, in short, knew perfectly well that the closed househol
not the sole or universal economic formation in Graeco-Roman an-
tiquity. That he did not discuss Graeco-Roman towns at any length is
another matter-his chapters on Stadtwirtschaft deal with the Middle
Ages-but he incorporated the ancient city into his evolutionary schema
by stressing the shift in town-country relations: "The Greek and Roman
town-dweller was a possessor and exploiter of the land, even if he
allowed the labour to be performed by slaves or tenants .... That the
inhabitants of our medieval towns were not..... Town and country had

32 Pirenne, "Les p6riodes de l'histoire sociale du capitalisme," Bulletin de la Classe des

Lettres, Acad6mie Royale de Belgique (1914), 258-99, at p. 264. An English translation
appeared in the American Historical Review, XIX (1914), 493-515, but with much of the
annotation omitted.
33 Lyon, op. cit., p. 199. Weber was hardly noticed in the paper, and Lyon himself
manages to confuse Biicher, Weber and Marx (e.g. p. 176).
34 "Agrarverhaltnisse des Altertums," Gesammelte Aufsiitze zur Sozial- und
Wirtschaftsgeschichte (Tiibingen: J. C. B. Mohr (Paul Siebeck), 1924), pp. 1-288, at pp.
7-8 (originally published in the 3rd ed. of the Handwirterbuch der Staatswissenschaften,
1909). This protest is not in the short earlier (1897) version.
35 Op. cit., p. 33. See also his review-article on the 1st ed. of Sombart, Der moderne
Kapitalismus, in Historische Zeitschrift, XCI (1903), 432-85.
36 Biicher, Beitriige zur Wirtschaftsgeschichte (Tiibingen, 1922), p. 3.
37 "Zur griechischen Wirtschaftsgeschichte," Festgabe fiir A. Schdffle (Tiibingen,
1901), enlarged and reprinted as the 100-page first chapter of his Beitriige, a completely
neglected work.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 THE ANCIENT CITY 317

separated in economic function." The medieval town "was not a mere

centre of consumption, as were the towns of the Greeks and Ro-
mans."38 Sombart then elaborated and refined the notion: "By a con-
sumption city I mean one which pays for its maintenance (Lebensun-
terhalt) ... not with its own products, because it does not need to. It
derives its maintenance rather on the basis of a legal claim (Rechtstitel),
such as taxes or rents, without having to deliver return values." He then
added a qualification: "The original, primary city creators were con-
sumers, the derived secondary (tertiary, etc.) creators were produc-
ers," and the latter were a dependent element, "whose existence was
determined by the share of the consumption fund allowed to them by the
consumption class."39
And that brings us at last to Max Weber. The intellectual relationship
between Weber and Sombart was a very close one: they were joint
co-editors of the revitalized Archiv fur Sozialwissenschaft und Sozial-
politik for one thing.40 Biicher was not a member of the Weber circle, but
Weber's Agrarverhiltnisse opens with a powerful defence, though not
an unqualified acceptance, of Biicher's Entstehung der Volks-
wirtschaft.41 My concern with showing that Weber's infinitely better
known work on the city had important forerunners and, in a sense,
co-workers has more than antiquarian interest for my subject. We need
Sombart and Biicher to help us fill out the picture, for Weber's one
sustained analysis of the city is a posthumous, unannotated, book-length
essay, subsequently included within a context that is often overlooked,
in his Wirtschaft und Gesellschaft. The latter is itself not only a post-
humous work on which he had been working for more than a decade (and
therefore with changing styles and aims) but one left by Weber in such a
state that not even the sequence of the sections was indicated.42 And, it
should be added, Weber's style in his later works, like his thought-

38 K. Biicher, Die Entstehung des Volkswirtschafts (5th ed., Tiibingen, 1906), pp. 370-1
(cf. pp. 441-4). The quotation in my text does not appear in the English translation, made
from the 3rd ed., by S. M. Wickett under the grossly misleading title, Industrial Evolution
(London and New York, 1901). My other reference, however, will be found in the latter,
pp. 371-4.
39 Op. cit., I 142-3. In the first edition there is only a fleeting hint of the concept: II
222-3.
40 See, e.g., Weber's references to Sombart in The Protestent Ethic, the references in
Marianne Weber, Max Weber. Ein Lebensbild (Heidelberg: Lambert Schneider, 1950),
and Sombart's introduction to the 2nd ed. of Der moderne Kapitalismus.
41 Biicher's importance for Weber is still more evident and more explicit in the second
chapter of Wirtschaft und Gesellschaft, "Soziologische Grundkategorien des Wirtschaf-
tens." I shall cite this work in the 4th ed. by J. Winckelmann (2 vols., Tuibingen: J. C. B.
Mohr (Paul Siebeck), 1956).
42 "Der Stadt" first appeared in Archivfiir Sozialwissenschaft und Sozialpolitik, XLVII
(1921), 621-772. In the 4th ed. of Wirtschaft und Gesellschaft it is reprinted, II 735-822,
and I shall cite that edition. On the "three layers" within the work, see W. J. Mommsen,
The Age of Bureaucracy: Perspectives on the Political Sociology of Max Weber (Oxford:
Basil Blackwell, 1974), pp. 15-17.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
3I8 M. I. FINLEY

processes, was extraordinarily dense and complex; in the two works

that concern me, so much so that the available English translations are
unreliable at best, and at worst blundering.
Weber was of course the most profoundly historical of sociologists
He began his career as a legal historian, interested particularly in tw
large subjects, the history of the organization of land-exploitation (wit
its political and social implications or consequences) and the develop-
ment of commercial practices and institutions. In that first period he
wrote his Romische Agrargeschichte (1891), a brilliant piece of histori
cal research still within the recognizable framework of an establishe
academic discipline. Thereafter his only substantial work on antiquit
was a tour deforce, a full-length book written in four months in 1908 and
published the following year in the encyclopaedia which commissioned
it and which is responsible for the misleading title, Die Agrar-
verhiltnisse des Altertums (even worse in the English title chosen for the
translation that has just appeared: The Agrarian Sociology of Ancient
Civilizations). His widow characterized it, not inaccurately, as "a sort
of sociology of antiquity" prefaced by "an economic theory of the worl
of ancient states,"43 among which he included not only the Greek and
the Roman but also the Near Eastern (Egypt, Mesopotamia and
Judaea). For all Weber's concern with the dynamics of social institu-
tions and social-cultural interrelations, the Agrarverhiiltnisse is not a
history, whether of ancient agriculture or of ancient society. Weber had
abandoned the writing of history. Still less historiographical is the
somewhat later "book" on the city, though the data about antiquity are
largely taken over from the Agrarverhaltnisse. It is not insignificant that
each section of the later study begins with either general concepts or
with medieval material before the ancient world is introduced for pur-
poses of clarification or contrast.
In sum, Weber never published a study of the ancient city, and his
views on the subject, as on other aspects of the ancient world, must be
elicited, with effort (including what amounts to decoding), from his
whole oeuvre, not merely from the writings overtly dealing with an-
tiquity, with constant alertness to changed nuances in his thinking.44
Some of the pivotal concepts have an obviously close kinship with those
of Biicher and Sombart. He, too, began with an economic definition,
which turns out to be a refined and elaborated statement of Sombart's: a
town is a place in which "the resident population satisfies an econom-

43 Marianne Weber, op. cit., p. 375. The 1897 version does not refute my remarks.
44 Alfred Heuss in the opening pages of his centenary article, "Max Webers Bedeutung
fur die Geschichte des griechisch-r6mischen Altertums," Historische Zeitschrift, CCI
(1965), 529-56. Heuss' account would have been more complete, though perhaps only a
little less gloomy, had he been less parochial and looked outside Germany.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 THE ANCIENT CITY 319

ically essential portion of their daily needs in the local market, and
an essential part by means of products which the residents a
inhabitants of the immediate vicinity have produced or other
quired for sale in the market." When the large consumers deri
income in one form or another as rentiers, the city is a consumer-
in antiquity. For, "if today we rightly conceive of the typical t
as one who does not secure his sustenance from his own land, o
the opposite was true of the mass of the typical cities (poleis
tiquity."45
Two words in that series of quotations require the closest attention:
"originally" and "typical." Originally the ancient city arose round the
town-dwelling large landowners, but as it grew, more and more of its
inhabitants were neither large nor small landowners. Yet it remained a
consumer-city: even in its late, "democratic" phase the social conflicts
within the ancient city were sparked by the demands of "essentially
debtor interests. Therefore consumer interests," unlike the "manufac-
turing" interests underlying the parallel conflicts in the medieval city.
In order to explain that fundamental difference in development, an
independent variable has to be introduced into the analysis, namely,
slavery.46 The widespread use of slaves in agriculture and manufacture
severely restricted the scope for free labour and blocked expansion of
the market, especially of the market for mass-consumed manufactures.
It also hindered, and effectively prevented, increasing rationalization of
production: given the uncertainty of the market and the fluctuating costs
of slaves (for both procurement and maintenance), the slaveowner had
to be free to dispose of a portion of his slave force at a moment's notice,
or to exploit them in ways other than direct employment in production.
Extensive division of labour and other forms of rationalization would
have destroyed the owner's flexibility. In sum, the slaveowner of an-
tiquity, like the landowner and the "money-owner," was a rentier, not
an entrepreneur.47 The contrast with the development of manufacture in
the Middle Ages is self-evident.
From these distinctions there flowed equally sharp differences in
policy, and now a new variable must be introduced. In the opening
section of the later work, Weber began with the "economic" definition
of the city, as I have mentioned, but he quickly went on to indicate that it
was not a complete definition. "The mere fact of a residential agglom-
eration of traders and manufacturing interests and the regular satisfac-
tion of daily needs in the market do not of themselves exhaust the
concept 'city.'" It is also "an economy-regulating association" encom-

45 Wirtschaft und Gesellschaft, II 736-9.

46 Ibid., II 805-9; cf. Agrarverhiltnisse, especially pp. 139-46, 256-7.
47 Agrarverhaltnisse, pp. 143-4.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
320 M. I. FINLEY

passing "characteristic objects of economic-policy regulation on b

of the association and a matrix of characteristic measures.' 48 The focus
had shifted from that of the Agrarverhiiltnisse, though much of the later
work can be discerned in the earlier one.
Bluntly, and therefore over-sharply, stated, policy and political au-
thority moved to the centre. When "The City" reappears in Wirtschaft
und Gesellschaft, it has a longer title, "Non-legitimate Domination
(Typology of Cities)," and it is but one section of a much larger one, on
Herrschaft (Domination), which includes, among others, bureaucracy
and charisma.49 As early as 1895, in his Freiburg Inaugural Lecture, he
had argued that the preservation and growth of the nation-state overrode
all other considerations and interests.50 Although this strong nationalist
stance and its concomitant political stress were less visible in the histori-
cal writings of the succeeding years, they were never absent (as we shall
see shortly). They reemerged in full force in the final decade of his life,
both in his political activity and in his theoretical work.51 In Wirtschaft
und Gesellschaft, with its two fundamental themes, rationality and
domination, he sealed the "fateful connection between industrializa-
tion, capitalism and self-preservation.' 52
And at last we turn to the second of the two words I said we must
attend to with care, "typical." Of course Weber knew that cities sur-
vived for centuries under the Roman Empire, though they had lost all
capacity for "economic-policy regulation;" that cities in fact prolifer-
ated in that era and sprang up in new territories, under direct stimulus
and sometimes compulsion from the central authority. But his "typology
of cities"-the subtitle of the later work-was intended, and can only be
understood, as a typology of "ideal-type cities." As he himself wrote,
"In the reality the types were everywhere fluid among themselves.
That, however, is true for all sociological phenomena and should not
prevent the establishment of the predominantly typical."53 Hence his
frequent employment of inverted commas, especially in the Agrar-
verhiiltnisse, for such terms as "feudal" and "capitalistic" (usually as

48 Wirtschaft und Gesellschaft, II 739.

49 That this was Weber's own schema is shown by the most recent editors: see J.
Winckelmann in his introduction to the 4th ed. of Wirtschaft und Gesellschaft, I xi-xii; cf.
G. Roth in his introduction to the 3-vol. English translation, Roth and C. Wittich, eds.
(New York: Bedminster Press, 1968), I lxxvii n. 87, xci-xciv.
50 The lecture is reprinted as the first essay in his Gesammelte Politische Schriften, J.
Winckelmann, ed. (3rd ed., Tiibingen: J. C. B. Mohr (Paul Siebeck), 1971).
51 See W. J. Mommsen, Max Weber und die deutsche Politik 1890-1920 (Tiibingen: J.
C. B. Mohr (Paul Siebeck). 1959); briefly his Age of Bureaucracy, ch. 2, with good
bibliography.
52 H. Marcuse, "Individualism and Capitalism in the Work of Max Weber," Nega-
tions, J. J. Shapiro, trans. (London: Allen Lane, 1968), ch. 3, at pp. 201-3; cf. J.
Habermas, "Technology and Science as Ideology," Toward a Rational Society, J. J.
Shapiro, trans. (London: Heinemann, 1971), ch. 6.
53 Wirtschaft und Gesellschaft, II 782.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 THE ANCIENT CITY 321

adjectives rather than as nouns in these critical instances), a f

of what he, with equal frequency, calls Ansitze (preliminar
indication of fluidity, of the genesis within one type of elemen
teristic of another type. Rarely, if ever, does he dodge the o
explain the failure (when that was the case) of Ansiitze to
eventually to dominate.
Thus, the final section of the Agrarverhiiltnisse attempts
why the Roman Empire and the pax Romana destroyed, rat
nourished, the Ansitze of capitalism he had detected in
town. The argument is dense, but it may reasonably be sum
this way. Thepax Romana put an end to extensive territorial
and to the accumulation of booty, including vast quantities
booty, both essential processes for the growth of wealth in
Roman economy. Previous expansion had brought into the
the first time large tracts of inland territory, away from t
therefore with inadequate access to routes of trade and com
The consequent tendency on inland estates was for rural se
around a villa where the basic requirements of mass consum
produced, thereby "disarming" the town by reducing its op
for gainful activity. The decisive blow fell in the political
absolute monarchy replaced city administration by the "dy
fessional army and bureaucracy," ending in a "liturgy-stat
relying on compulsory services). "Since the capitalism of ant
politically anchored and depended on private exploitation o
relations of domination in an expanding city-state, it came to a
with the disappearance of this source of capital-formation
bureaucratic system killed the political initiative of its subjects
the economic initiative, for which the appropriate opportu
lacking." And then the despairing coda: "Every bureaucracy
tendency to accomplish the same effect by self-expan
sichgreifen). Ours too."54
For historians who are allergic to ideal types, there is noth
discuss; there are no propositions deserving examination an
One can find sufficient comfort and refuge in the "discov
Weber's knowledge of the Greek world was very much less
and accurate than of the Roman;55 in the demonstration tha
now be shown to be wrong when he called the Roman equi
national capitalist class."56 One can (legitimately) challen
conception of the feudal and capitalist elements in antiquit

54 Agrarverhaltnisse, pp. 271-8.

55 G.E.M. de Ste. Croix, "Karl Marx and the History of Classical Antiqui
usa, VIII (1975), 7-41, at pp. 19-20.
56 Wirtschaft und Gesellschaft, II 818.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
322 M. I. FINLEY

political definition of the city. But when the demolition has

pleted, the phenomena have not stolen away silently. It stil
true, and needing an explanation, that the peasant was a
element in the ancient city, but not in the medieval; that the gu
integral element in the medieval city, but not in the ancien
may be allowed to repeat what I wrote recently on the secon
seems commonly overlooked that the excavators of Tarsus
no Cloth Hall, that all ancient cities lacked the Guildhalls a
which, next to the cathedrals, are to this day the architectur
the great medieval cities of Italy, France, Flanders, the Hansa
England. Contrast the Athenian Agora with the Grande Pla
sels."57 Furthermore, it still remains true, and needs an ex
that ancient urbanism decayed so badly as to require a seco
towns" in the Middle Ages. If Weber does not offer satisfact
nations, even partial ones, where do we turn?
To Karl Marx perhaps? Marx was the spectre haunting Web
course Sombart) throughout his life, much more so than mi
ferred from the rare and even crude comments on Marx and Marxism to
be found in Weber's writings.58 I have no intention of entering into the
subject, except to note that it is more complex than some current,
oversimplified and dogmatic accounts suggest. Merely to dismiss Weber
as an "idealist," whose emphasis on "spirit" and trade led him to see
"capitalism" where it never existed, is a caricature, an idle playing
about with words. In his 1857 notebooks, Marx wrote about "the
civilizing influence of external trade," though at first only a "passive
trade,"59 in a passage that cannot but remind us sharply of Weber's
thesis that the archaic shift from passive to active trade was the first step
leading to the gulf between the western and the eastern city. For Marx
(and Engels), there was never any doubt that "commercial capital,"
"commercial cities" and even "commercial peoples" (Phoenicians and
Carthaginians) were widespread ancient phenomena, and that in some
instances, ancient Corinth for example, trade led to a highly developed
manufacture.60

Weber, like Marx, had at the centre of his interests the phenomenon of

57 The Ancient Economy (Berkeley and Los Angeles: University of California Press,
1973), p. 137.
58 See H. J. Kocka, "Karl Marx und Max Weber. Ein methodologischer Vergleich,"
Zeitschriftfur die gesamte Staatswissenschaft, CXXII (1966), 328-57, at pp. 329-35. An
excellent starting-point on Marx and Weber, with good bibliography, is provided by
Mommsen, Age of Bureaucracy, ch. 3.
59 Grundrisse, M. Nicolaus, trans. (Penguin Books, 1973), p. 256.
60 The texts are conveniently assembled by E. Ch. Welskopf, Die Produk-
tionsverhiltnisse im alten Orient und in der griechisch-romischen Antike (Berlin:
Akademie Verlag, 1957), ch. 10.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 THE ANCIENT CITY 323

capitalism.61 That the two analyses ultimately diverged sha

point of conflict, is undeniable (quite apart from the violent
ment over political action and future goals). Marx's theorie
solutely unpalatable" to Weber "as ontological proposition
other hand, he saw "Marx's interpretation of history, in ter
various forms of production, as a most useful hypothesis wh
help to gain important insights into the development of mod
trial society."62 In consequence, for the pre-industrial ages, a
ancient city in particular, there was a large area of overlap a
ment between them.
Marx, of course, never made a systematic inquiry into the ancient
world in general, or the ancient city in particular. On the latter, his few
scattered remarks all stem from the proposition I quoted earlier from
The German Ideology, repeated in the first volume of Capital:
The foundation of every division of labour which has attained a certain degree of develop-
ment and has been brought about by the exchange of commodities, is the separation of
town from country. One might well say that the whole economic history of society is
summed up in the movement of this antithesis. However, for the moment we shall not go
into this.63

Not only "for the moment," I may add: in the whole Marxian corpus no
more is to be found about the ancient town than an occasional statement,
propositions about ideal types which are often more or less Weberian in
substance.64 Thus, we read in the Grundrisse: "In the world of an-
tiquity, the city with its territory is the economic totality. . . . Urban
citizenship resolves itself economically into the simple form that the
agriculturist is a resident of the city."65
This is not the place for an extended analysis of the parallels (or the
divergences), but two further examples may be useful. "The modern
proletariat, as a class, was absent. For ancient culture either rested on
slavery at its centre of gravity (as in late Republican Rome), or, where
'free' labour in a private-law sense predominated (in the Hellenistic
world and in the Roman Empire), it was still permeated by slavery to a
degree that never existed in medieval Europe." That is Weber,66 but few
Marxist historians could reasonably disagree, except perhaps to transfer
the first two centuries of the Roman Empire in the west to the first of the

61 On the centrality of capitalism in Weber's work, see G. Abramowski, Das Ge-

schichtsbild Max Webers (Stuttgart: Ernst Klett Verlag, 1966).
62 Mommsen, Age of Bureaucracy, pp. 50-1.
63 I quote from the Penguin translation by Ben Fowkes (1976), p. 472.
64 I write "ideal types" deliberately. On important similarities in the approach of both
Marx and Weber, see R. Ashcraft, "Marx and Weber on Liberalism as Bourgeois
Ideology," Comparative Studies in Society and History, XIV (1972), 130-68.
65 Grundrisse, p. 484.
66 Agrarverhaltnisse, p. 6.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
324 M. I. FINLEY

alternatives. "Military power was more closely locked to

growth than in perhaps any other mode of production, befo
because the main single origin of slave-labour was normally
prisoners of war, while the raising of free urban troops for war
on the maintenance of production at home by slaves." That
Anderson, in a recent, subtle Marxist account,67 and the Weberian
parallel is evident from the summary I have already given of Weber's
view on the impact of the pax Romana.
Suppose one accepts that these propositions, and others I have drawn
from my survey of the history of theories of the ancient town, are, at the
least, interesting enough to be pursued by a detailed examination of the
available data, literary, epigraphical, archaeological. What are the im-
plications for further historical inquiry? Not even the most sociologi-
cally minded historian is willing to stop with the formulation of ideal
types. The variations within each type, the changes and developments,
the implications over the whole range of human thinking and acting
require detailed, concrete exposition-an exposition which would at the
same time be a test of the ideal type.68 Such an account does not yet exist
of the ancient town. There are, to be sure, a growing number of "his-
tories" of individual towns, Greek and Roman, from the archaic age to
the end of antiquity. With scarcely an exception, however, they lack a
conceptual focus or scheme: everything known about the place under
examination appears to have equal claim-architecture, religion and
philosophy, trade and coinage, administration and "international rela-
tions." The city qua city is flooded out. The approach is usually de-
scriptive and positivistic, "collecting evidence and interrogating it with
an open mind":69 the unexpressed assumptions about the economy are
usually "modernizing." I do not underestimate the contribution to
knowledge that has been made by these studies, nor the difficulties
inherent in the attempt, nor such conceptual advances as there have
been in the past decade or two.70 However, it is the case that the
considerations I have raised, the issues presented by Marx, Biicher,
Sombart and Weber, are peripheral, at best, in current study of the
ancient town.71

67 Passages from Antiquity to Feudalism (London: NLB, 1974), p. 28.

68 I trust that it is obvious that this approach to ideal types is fundamentally different
from von Below's, quoted above at n. 35.
69 M. W. Frederiksen, reviewing my Ancient Economy in Journal of Roman Studies,
LXV (1975), 170-1.
70 For one region, see briefly E. Fr6zouls, "Etudes et recherches sur les villes en
Gaule," Quaderno no. 158 of the Accademia Nazionale dei Lincei (1973), 153-66.
71 In one field, the Greek "colonies" of southern Italy and Sicily, the persistent efforts
of E. Lepore to introduce a proper conceptual approach should be noted: "Strutture della
colonizzazione focea in Occidente," Parola del Passato, XXV (1970), 19-54; "Napoli
Greco-Romana. La vita politica e sociale," in Storia di Napoli, I (Naples 1968), pp.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 THE ANCIENT CITY 325

In the end, I believe that the history of individual ancient towns i

de sac, given the limits of the available (and potential) document
the unalterable condition of the study of ancient history. It is not w
perverse to see an advantage in the weakness. There is mounting
cism of contemporary urban history for allowing the deluge of
obscure the questions being asked and their purpose,72 a dange
ancient urban historian is happily safe from. But what questions
wish to ask about the ancient city, whether they can be answe
satisfactorily or not? That is the first thing to be clear about, befor
evidence is collected, let alone interrogated. If my evaluation o
current situation is a bleak one, that is not because I dislike the ques
that are being asked but because I usually fail to discover any qu
at all, other than antiquarian ones-how big? how many? what m
ments? how much trade? which products?
To understand the place of the town as a pivotal institution i
Graeco-Roman world and its development, the starting-poin
surely be two facts. First, the Graeco-Roman world was more ur-
banized than any other society before the modern era. Second, the
city-state, the closely interlocked town-country unit, remained the basic
module even after the state component in city-state had lost its strict
original meaning. Did it also remain a "consumer-city"?
That there were such consumer-cities throughout antiquity is indis-
putable. In 385 B.C. Sparta defeated Mantinea in Arcadia and laid down
as a peace condition that the town be razed and the people returned to
the four villages in which they had once lived. "At first they were
discontented," comments Xenophon (Hellenica 5.2.7), "because they
had to demolish their existing houses and build new ones. But when the
property owners were living near the estates they possessed around the
villages, and they had an aristocracy and were rid of the burdensome
demagogues, they were pleased with the state of affairs." Xenophon's
political comments are irrelevant for my purposes; the viability of the
Spartan demands is what matters. And when the town of Mantinea was
eventually restored, it continued for centuries as a centre for landowning
residents, as it had been when the Spartans destroyed it.73
How typical was Mantinea? Capua, Cicero tells us (On the Agrarian
Law, 2.88), was preserved by the victorious Romans in the interest of

141-371; "Per un fenomenologia storica del rapporto citta-territorio in Magna Grecia," in

La citta e il suo territotio [Atti del 7? Convegno di Studi sulla Magna Grecia (Naples,
1968)], pp. 29-66.
72 See e.g. R. R. Alford, "Critical Evaluation of the Principles of City Classification,"
in Berry, ed., City Classification Handbook, ch. 11; M. H. Frisch, "L'histoire urbaine
americaine: r6flexions sur les tendances r6centes," Annales (E.S.C.), XXV (1970), 880-
96.
73 For the evidence, see U. Kahrstedt, Das wirtschaftliche Gesicht Griechenlands in
der Kaiserzeit (Bern: Francke Verlag, 1954), pp. 132-6.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
326 M. I. FINLEY

the Campanian farmers, among other things so that, "wearied by the

cultivation of the lands, they would have the use of the homes in the
town." The steady urban growth in central and northern Italy during the
late Republic produced towns of the same kind.74 So did the "Romani-
zation" of the Danubian region incorporated into the province of Pan-
nonia under the Empire.75 Rome itself was of course the prototype of a
consumer-city, as it has been throughout its history. Antioch, too, the
fourth city of the empire: for the fourth century the urban population has
been estimated at between 150,000 and 300,000; its extensive territory
was at least 300 times as great as the area within the town walls, and the
foundation of its wealth was in the land and in its leading place within the
imperial administrative system.76 The extra-urban districts were thick
with villages, each with its own local production and distribution
through rural fairs. In consequence, Libanius explains (Orations,
11.230), the villagers "have little need for the town, thanks to the
exchange among themselves."
Present-day overtones of the word "consumer" should not be al-
lowed to intrude and mislead. No one is suggesting that the urban lower
classes were a host of beggars and pensioners, though it has become a
favourite scholarly pastime to "disprove" that contention for the city of
Rome; though, too, the extent of beggary, unemployment and famine is
not to be underestimated. The issue implicit in the notion of a
consumer-city is whether and how far the economy and the power
relations within the town rested on wealth generated by rents and taxes
flowing to, and circulating among, town-dwellers.77 Even the quintes-
sential consumer-city, Rome, required innumerable craftsmen and
shopkeepers for intra-urban production and circulation. In so far as they
were engaged in "petty commodity production," the production by
independent craftsmen of goods retailed for local consumption, they do
not invalidate the notion of a consumer-city.
It is also not suggested that the examples I have given-a handful from
many available instances-were all identical towns. If it is the case that
they were all in some respects consumer-cities, the next step in the
inquiry is to examine the variations of (or from) the ideal type, to
establish a typology of ancient towns. Consider Cyzicus on the Sea of

74 H. Galsterer, Herrschaft und Verwaltung im republikanischen Italien [Miinchener

Beitrage zur Papyrusforschung ..., no. 68 (Munich: Beck, 1976)], pt. I; see E. Gabba,
"Urbanizzazione e rinovamenti urbanistici nell' Italia centro-meridionale del I. sec.
a.C.," Studi classici e orientali, XXI (1972), 73-112.
75 P. Oliva, Pannonia and the Onset of Crisis in the Roman Empire (Prague: Academy
of Science, 1962), pp. 236-42.
76 These figures are taken from the best modern account of the city in the later Empire,
J. H. W. G. Liebeschuetz, Antioch (Oxford: Clarendon Press, 1972), ch. 2.
77 See myAncient Economy, ch. 5; A. H. M. Jones, The Roman Economy, P. A. Brunt,
ed. (Oxford: Basil Blackwell, 1974), ch. 1-2.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms
 THE ANCIENT CITY 327

Marmara, a harbour-town identified by historians as "a great

house for the trade of the Euxine (Black Sea),"78 famous for
circulated "white gold" (electrum) coins. In 319 B.C., in the
the wars among Alexander's successors, it was attacked by th
Hellespontine Phrygia in a surprise move that found the to
pared, with only a few people within the walls while the majo
out in the fields. There is no ground for disbelieving the
Diodorus (18.51.1-2) on this score. Where, then, do we locate
in a typology? Unless we are content with the familiar and m
serial formulation-"the economic life" of Noricum "depended on
agricultural production, pastoralism, mining, industry-above all iron-
smelting and metal-working-and trade"79-proper factor analysis is
essential. The factors may not often coincide with the modern ones and
the opportunities for genuinely quantitative and dynamic analysis are
few and frustrating; the procedure is nevertheless unavoidable.
It is not my intention in this essay to enumerate the variables or to
formulate a typology. Much of what I should include is anyway implicit
(and sometimes explicit) in what I have already said-the extent (and in
rare cases, the absence) of agricultural territory appertaining to the
town; the size ofthe town and of its population; access to waterways; the
extent and "location" of the slave-labour force; self-sufficiency on the
large estates; peace or war; the changing role of the state with the
development of the large territorial empires. This is not an exhaustive
list, but it will suffice for present purposes. It points back to the ques-
tions which distinguish history from antiquarianism.
I have come to the end, still referring to the ancient city. Is it a
defensible category? Mere chronology is no argument in its favour, nor
is the undeniable variety among ancient towns an argument in its dis-
favour. My defence is a simple one. The city does not exist in isolation: it
is an integral part of a larger social structure, in the Graeco-Roman
world a pivotal institution. Unless and until the kind of concrete investi-
gation I have suggested demonstrates that, allowing for exceptions,
Graeco-Roman towns did not all have common factors of sufficient
weight to warrant both their inclusion in a single category and the
differentiation from both the oriental and the medieval town, I hold it to
be methodologically correct to retain the ancient city as a type. And
"type" has crept back, as my final word.

78 David Magie, Roman Rule in Asia Minor (2 vols., Princeton: Princeton University
Press, 1950), I 81.
79 G. Alfoldy, Noricum, A. Birley, trans. (London: Routledge & Kegan Paul, 1974), p
43.

This content downloaded from

147.91.1.43 on Thu, 07 Oct 2021 10:54:12 UTC
All use subject to https://about.jstor.org/terms