constructive typology

Wed, 25 Jan 1995 12:50:10 +0600
Nikolai S. Rozov (rozov@adm.nsu.nsk.su)

Let me realize my promise according w-s typology. In fact
my task occurred to be more wide because the problems of
differentiation between w-ss and non-w-ss are not less
significant than the problems of differentiation between types
of w-ss themselves.
That's why I wish to present now some primary ideas on

CONSTRUCTIVE TYPOLOGY OF SOCIETAL SYSTEMS

What preliminary requirements to our typology should we
set?
R1. The typology of s-systems must include all commonly
used concepts (i.e. w-systems, w-economies, w-empires,
civilizations, societies) as units of analysis for diverse
social disciplines. The taxons in typology should not differ
essentially in volume and meaning from their correspondent
concepts. (The task is not a terminological revolution but
conceptual clarification and systematization.)
R2. It should help us to divide types of societal systems
(s-systems) with different essential features and to integrate
s-systems without such essential difference into one type
(even if the last differ very much in volume, geographical
location and historical period).
R3. The criteria of "essential" must be flexible because
of development of our thought, research interests and values.
R4. The typology should be provided with the criteria and
methods for empirical identification of s-systems.

Typology of s-systems is not a new task. We can remember
many taxonomies since Plato's and Aristotle's descriptions
of state political forms. All traditional taxonomies (which I
know) begin from the apriori existence of s-systems ("polices,"
states, empires, nations, societies, cultures, civilizations,
etc.) The w-s paradigm deviated from this tradition.
I. Wallerstein (from my viewpoint) deals with w-ss
(w-empires and especially w-economies) not as with empirically
presented phenomena but as with theoretically reconstructed
essences. In his analysis of "basic logic" he focused on the
economical and political c o n n e c t i o n s between parts
of w-ss.
This line was developed in the prospective idea of nested
networks by C.Chase-Dunn & T.Hall, 1993 (it's a pity I have
only the brief outline of this idea in Chris C-D and Peter
Grimes's post to wsn).
My suggestion is to make this approach more clear and
systematic for myself (surely it's only my responsibility for
all mistakes).

The idea is to begin from "parts" or "loci" and then
varying the sorts of connections to c o n s t r u c t
conceptually the ideal types (M.Weber) of different s-systems
(including w-systems, oicumenas, civilizations and societies).
The typology using approach of this kind I name
constructive typology.

I expect the most probable question: "You reject taking
apriori societies and w-ss, but why on Earth do you take
apriori your loci?"
I agree that "locus" is not less theoretical abstraction
than "society,""w-s," "civilization" or any other s-system.
Nevertheless I have at least two arguments for beginning
namely from loci to s-systems.
First, loci are more close to empirical (historical,
archeological, economical, cultural) analysis. Two researchers
using empirical data can gain consensus according features and
connections of one province much sooner than in discussion of
the features, borders and connections of the whole
encompassing s-system (society, civilization, w-system).
Second, "locus" is an artificial construct which can be
used as an "elastic concept": in various research situations
one can change flexibly the meaning, volume, borders of his
"loci" (surely announcing it explicitly).

Now I will present some concepts which have "Pickwick's
meaning," i.e. they pretend only to demonstrate the very logic
of constructive approach. Probably you will find some blanks
and defects and I will be grateful for any comments. At the
same time I ask to focus mainly on the general principle and
possibilities of the constructive approach.

LOCI - inhabited geographical places which can be
considered as homogenous relatively to connections with other
places.
It is possible if necessary to differ:
i) mini-locus, f.e. small village of group of family farms
using the same trade, social, transport and other networks,
ii) midi-locus (a small province, a "nom" in Shumer, a
"police" in Greece, a "uezd" or a "rayon" in Russia, a county
in US) usually including a town with its countryside,
iii) mega-locus (large province, "gubernia" or "oblast" in
Russia, land in Germany, state in US) including large city
(directly connected to other large cities) with its partly
urban and partly rural surrounding.

(Interloci) CONNECTIONS - any kinds of regular
economical, political, demographical, social, cultural,
technological and other relations between loci.

I suggest to differ four main types of CONNECTIONS:

MATERIAL and INFORMATIONAL TORRENTS: movement of
goods and raw materials, texts and other cultural patterns,
finances;

INTERACTIONS: military, economic, political,
sociocultural, and MIGRATIONS,

COMMUNITY OF CULTURAL GENESIS including common
features and elements in language, morals, religion, art, also
in technological, economic, political, legal patterns, etc.

COMMUNITY OF POLITICAL AND LEGAL ORDER
subordination to common power agents and set of laws and rules.

Connections as a rule are not equal but oriented
according to active or passive role of each locus in
organizing, realizing, supporting this connection. It is the
main way for constructing core-periphery relations. This
aspect is well known and I will not explicate it in the
following concepts.

Any inhabited place (locus or stable group of
interconnected loci) whose interior connections have essential
difference from exterior connections can be considered as
SOCIETAL SYSTEM (S-SYSTEM).
As you can see s-system is also an elastic concept which
depends directly from our current criteria of "essential
difference." "Basic logic" of system (I.Wallerstein) is very
fruitful but probably not unique criterion for "essential
difference."

The sorts of interior connections are the basis for setting the types of
s-systems. We will use the idealization that within one
s-system its loci are connected only with one sort of
connections. Such (surely non-existing) s-systems are ideal
types. Identification and research of definite real s-system
consist of comparing it with correspondent ideal types.

Ideal types of S-SYSTEMS:

NETWORK OF TORRENTS - s-system whose loci are connected
by material and informational torrents.
OICUMENA (MILITARY/POLITICAL O., TRADE O., ECONOMIC O.,
CULTURAL O.) - s-system whose loci are connected with
correspondent interactions.
CIVILIZATION - s-system whose loci have community of
cultural genesis.
SOCIETY - s-system whose loci have community of political
and legal order.

What are the relations between these types?

NETWORKS OF TORRENTS always encompass OICUMENE.
OICUMENA frequently includes parts of different
CIVILIZATIONS.
The spread of CIVILIZATION can lead to several new
OICUMENE.
OICUMENE and CIVILIZATIONS usually encompass SOCIETIES.
Some SOCIETIES can include parts of different
CIVILIZATIONS and take part in different OICUMENE.

Well, but where is our goal concept - world-system?
I suggest to name as w-ss both oicumene and networks of
torrents. The very meaning of the term "w-s" tells that this
system is (or seems to its people) like the whole world.
So we can consider OICUMENA (inhabited land) as the
"visible" or "subjectively the whole world."
At the same time we can study by objective methods that
the torrents of goods, texts, other patterns were much wider
than the space of interactions (oicumena). So we can reveal
"objectively the whole world"- NETWORK OF TORRENTS.
Both have right to be named world-systems. The division
of labor in research practice maintains this position:

Macroeconomics, World Economics deal with NETWORKs OF
TORRENTS.
International Relations, Political History, Geopolitics,
(Macrosociology?), etc. deal with diverse OICUMENE.
Culturology and civilization approach deal with
CIVILIZATIONS.
Traditional sociology deals with SOCIETY.

Namely in two former cases World-System approach occurred to be
the most efficient. Nevertheless two last units of analysis
should not be neglected. The constructive typology suggests
the conceptual bridge between all these disciplinary
traditions.

The elastic concepts of LOCUS and S-SYSTEM allow in
principle to fulfill R2-3. I mean that the aerials of w-ss,
civilizations, societies are not absolute but depend of our
current research conventions of what criteria of "homogenous"
or "differences of interior/exterior connections" are
essential.

What about other well-known or discussed in wsn terms
(see R1)?

"WORLD ECONOMY" can be TRADE OICUMENA but it also can
occur ECONOMIC OICUMENA if the econ. connections are more
strong (investments, common projects, etc.).

"WORLD EMPIRE" usually is POLITICAL/ECONOMIC OICUMENA
with strong regular uni-center core-periphery connections.
Some world empires can occur or develop into SOCIETIES (China?
Russia?).

"Interactive zones," "interaction networks," "political
military interactions" (R. Dunn, Wilkinson, C.C-D., T.Hall) can
be considered as MILITARY/POLITICAL OICUMENE.

"DISCOURSE-BASED W-SS" (Voll) seem to occur sometimes
CIVILIZATIONS, sometimes CULTURAL OICUMENE (the difference is
not trivial but it is a special topic). "Symbolic networks"
(E.Ermolaeva) can be considered, as I suppose, also as
cultural oicumene.

MINI-SYSTEMS (I.Wallerstein) and STATELESS W-SS
(C.Chase-Dann) can be isolated LOCI (as autonomous s-systems)
or pheripherical terminals of TRADE OICUMENE and encompassing
NETWORK OF TORRENTS.

"W-s" from China to Europe in 13 century (Abu-Lughod) can
be considered as a NETWORK OF TORRENTS of goods. This network
consisted of several overlapping TRADE OICUMENE (I support
here the position of I.Wallerstein in his resume to A-L's
book). "Prestige goods networks" I consider as the same type.

I don't think that we really can study the modern World
System (MWS) as one oicumena of interactions (even trade
oicumena). I suggest to take more modest task: to consider MWS
only as a NETWORK OF TORRENTS encompassing rather many
ECONOMIC/POLITICAL/CULTURAL OICUMENE, POLITICAL/MILITARY
OICUMENE and also CIVILIZATIONS which did not disappear and
don't go to die at all!
Surely informational (cultural, financial) torrents in
MWS are not less but maybe more significant than material
torrents.

What about R4 - the criteria and methods for empirical
verification? The basic logic of these criteria and methods is
very simple: empirically discovered presence/absence of signs
of torrents, interactions (of diverse sorts), common cultural
genesis and common political/legal order between definite
LOCI. The whole picture of these interloci connections allows
to make decision about type, composition, areal of the
s-system (group of s-systems).

For me the most interesting are the evolutionary trends
(patterns? laws?) of historical transformation of s-systems of
diverse types and application of this knowledge to the
problems of probable future transformations of MWS and its
parts.
Sorry for too long posting and for my English,
I am looking forward for comments.

Nikolai S. Rozov
rozov@adm.nsu.nsk.su