Phillip Guddemi (pguddemi@WELL.COM)
Wed, 14 Aug 1996 14:42:16 -0700

Dear Anthro-L,

I found myself reacting to the ethnicity thread, and particularly to
Kavanaugh's posting, with a series of examples. We need to look at
ethnicity over all of the data, including that of the historical record, and
I think we overreact to the politics of the present if we make our
revulsion, at journalistic uses of "tribalism" for example, the basis of our
theoretical thinking.

My anthropological fieldwork was done in Papua New Guinea. I studied in
Ama, in the Sepik region, where people who call themselves Sawiyanoo live.
There are about 500 Sawiyanoo today; their first contact with the colonial
world was during World War II, and they were "pacified" in 1964. They had a
distinctness as a group, which was based partially on being a cultic center
for initiation rituals, partly based on linguistic differences with
neighbors. Some of their neighbors were distinct not only in language but
in a variety of "cultural" ways, including ecological ones. The Sawiyanoo
live in a lowland rainforest surrounded by mountains (in which a distinct
group lives); they live in the upper reaches of streams and (unlike another
neighboring group) did not base their subsistence strategy on the use of
canoes. The language differences by the way are extreme; the Upper Sepik
area of Papua New Guinea probably has the highest linguistic diversity,
including "phylum level diversity," of any area in the world.

They had several reasons for ethnic distinctness, yet missionaries inform me
that this ethnic cohesion has increased since the first SIL mission in the
1970s. Before then, even the language was divided into regional dialects,
one missionary claims. (Yes, less than 500 people!) Thus some of the
Sawiyanoo identity may have been constructed by modern circumstances,
including the SIL desire to have a distinct language in which to translate
the Bible. (But a "northern" Sawiyanoo man informed me that some of that
Bible is difficult, or even "wrong," because of its "southern" usages,
usages which shade off into different "languages" altogether.) But this
deconstruction is only partial. A "local" identity did seem to exist before
contact. In another area of the Sepik, Simon Harrison has shown that
"local" identities existed in a sort of counterpoint (my term) with
cross-cutting "trade-friend" ties between men of different villages based on
another type of mythic clanship. It is interesting that the "local"
identities were constructed for warfare, or at times of warfare, or as an
idiom in which warfare could take place. In that area (but not in all
areas) initiations are part of the warfare complex, establishing identities
used as part of warfare. Cross-cutting "clan" identities of this sort (in
which the clan is not a subgroup, but a cross-cutting idiom of trade
relations between individuals of different groups) also exist in the
Sawiyanoo area, although there is some reason to believe that they were
elaborated since "pacification" by the Australian colonial authorities.

All of these complexities were those of ethnic groups in which the State or
nationalistic question was not present or present largely as a remote
backdrop. (I do not think we should confuse minor post-colonial changes in
local dynamics with the larger-scale national-level changes I am about to
outline.) In the national politics of Papua New Guinea today, entities like
"Sawiyanoo" do not have a real political salience. In the Sepik of the
1980s and 1990s, particularly in town settings, people often identify
themselves not according to the groups which appear in the anthropological
literature (even if the anthropologists "got the name right"), but according
to the nearest town or administrative center. A Sawiyanoo would be
"Ambunti" under this logic (a place many of them have never seen), but most
Sepiks in Wewak seem to be "Maprik" or "Wewak" or "Amboin" (names of towns
or subdistrict centers). They are of course in this setting speaking Tok
Pisin, the famous creole which was a pidgin three or more generations ago.
Earlier, during the period of labor recruitment, from the late '40s to about
1980, laborers recruited for copra processing plantations located on the
coast and islands of the New Guinea colonial entity, developed regional
"ethnic" oppositions as "Sepiks" as against "Highlands" (and the indigenous
"Islands"). Today the "tribalist" identities, i.e. those which are
problematic for the State, in Papua New Guinea are not at the "tribal" level
at all, i.e. local language (or "tok ples"). They are regional identities,
not always coterminous with provincial governments, identities which, yes,
emerged in the contemporary or "modern" context. (But the provincial
governments were first developed as a response to regional identities, at
the time of the first Bougainville crisis shortly after independence in
1975.) Traditional animosities between neighboring Sepik groups have little
salience in contemporary "ethnic-style" PNG politics, because these local
identities tend to be subsumed in regional identities (a segmentary system!).

Still, I think it would be a mistake to deny the existence of some
"ethnicity" in the precontact Papua New Guinea setting. However cross-cut
and localized, I think the ethnography shows that group identities largely
isomorphic with language did exist in the precontact context. These
identities were reinforced by certain institutions and relationships, and
(partially) subverted by others.

Another area of the world which I have researched is the Balkans, although I
have not yet done fieldwork there. (I have worked a little with an
organized society of immigrants to the U.S. from a particular Balkan group.)
The ethnicity situation which is most interesting to me (for definitional
purposes) is that of Macedonia prior to the Balkan Wars as discussed in a
book by Brailsford written at that time. (Macedonia for Brailsford's
purpose constitutes a part of the area remaining within the European portion
of the Ottoman Empire before 1912; my invoking the geographic area as he
used it is not meant to imply anything regarding current contemporary
claims.) It seems that, as with many areas of peasant Europe, the national
or even "ethnic" identity of a great number of individuals in this region
was attenuated. They did have a religious identity, as Christian or Muslim;
and in the Ottoman Empire this religious identity determined the laws under
which they would live (this was the famous "millet" system). For most of
the Christian peasants "ethnic" identity seemed far less evident than for
the precontact western Sepik of New Guinea (if this is apples and oranges,
please show me!). As with prerevolutionary France (I don't know where I get
this from) and other areas of Europe, the identity of at least peasant
populations was generally given in terms of the largest town or a very small
region. These identities were not oppositional, although I suppose there
were stereotypes. In Macedonia the Christian peasant population mostly
spoke a series of Slavic "dialects" which varied according to a continuum of
location, in some areas closer to "Serbian" and in other areas closer to
"Bulgarian" (these being newly-constructed "official" languages of
newly-constructed European nation-states, languages whose "dialects" formed
continuums within each nation). In the late 19th Century there was a
competition among neighboring nation-states for these peasants, to get them
to identify with one or another national identity. Greece, Bulgaria, and
Serbia each financed schools and churches in order to recruit or entice
peasants to identify with one or another group. The Greek claim was based
on religion (many people regardless of language thought of themselves as
"Greek" based on their identity with the "Greek" church and the Byzantine
heritage). The Bulgarian and Serbian claims were based largely on language.
There were several examples of villages "negotiating to get the best deal"
in this courtship process. The armed bands emerged a little later.

Eventually the boundaries between nations were set in the Balkan Wars of
1912 and 1913, and the local peasants conformed their identities or their
residences (voluntarily and otherwise) to suit their new national locations.
[In the ensuing decades the area which came under Serbian control developed
a nationalistic identity (or "national consciousness," to use a more emic
term) in opposition first to Serbia and then to Bulgaria (which many
considered a liberator State when it first occupied the area during World
War II). This new "national consciousness" calls itself the Macedonian
national identity, which of course can trace its newly-systematized national
language back to the 6th Century, since a Slavic language with local
characteristics has been spoken since that time, and which of course can
identify itself with various medieval States which quite possibly had been
peopled and/or ruled by people from that region and linguistic background.
The nationalist identity claims of Macedonia have, in other words, been
woven on the Benedict Anderson loom according to the Balkan and European
patterns, and the design similarities are evident.] Yet what is striking is
that, at least from Brailsford's account, the "pre-nationalistic" identities
of many of those in the region, during the late Ottoman times at least, were
fluid and even ill-defined, not "primordial" at all, and very localized.
[And Danforth in a recent book brings out that people, immigrants to
Australia, who are claimed by Macedonian nationalists as having the
characteristics of "Macedonian identity," but who are uncomfortable with
that national identification, today use a term like "local people" -- in
their Slavic language -- to define themselves.] Here we seem to have the
ultimate playing-field of the so-called "imagining" into being of national

In spite of all this, in this region there were groups, some defined by
religion but others by ecology, which _did_ have a pre-existing "bounded"
identity during the otherwise somewhat "pre-ethnic" period of Macedonia.
One of these groups was the Vlachs or Aromanians, who spoke a Romance
language (in the area since the 900s), and were mostly merchants and
transhumant shepherds. They were themselves courted by the Greek national
cause (because they were Orthodox and admired Byzantino-Greek culture) and
by the Romanian state (because the languages are very close). The ensuing
"nationalist" factional opposition still has repercussions among Aromanians,
and Aromanian-Americans, for example the group in Connecticut which I have
visited. But what is interesting to me is that the pre-existing Aromanian
ethnicity is an "ethnicity" which was fairly well-defined, as against a
perhaps less "ethnically" self-identified set of majority populations in the
southern Balkans. The Vlach identity was defined not so much by geography
(although certain mountain villages were nearly all Aromanian), but by
ecology/occupation and language, and a high degree of endogamy, and perhaps
(relying on Muriel Schein for the Greek context) by a certain amount of
stigma or sterotype from the "majority" population. (The Vlachs/Aromanians
by the way are notable for their bilingualism and their assimilation into
encompassing populations, although they sometimes maintain their ethnic
identity among themselves while self-presenting to outsiders as members of
majority or regnant or national ethnicities.)

Think of the "ethnic" existence of the Jewish population in Europe over the
centuries. Might they have had more of an "ethnic" existence than some of
the "local peasant groups" of Europe in earlier centuries?

And Armenians (about whom this thread began) seem to have been, somewhat
like the Jews, a "cosmopolitan" and largely expatriated merchant group
within the Byzantine and Eastern Christian world for centuries. Many
Armenians in Constantinople or Kievan Rus in the 10th Century might not have
felt they were "gonna go back" to their original homeland. The Aromanians
(whose "original homeland" is unclear) eventually (during the Ottoman
period) developed a component which was similarly mercantile and
expatriated, sometimes quite wealthy, ranging all the way to Belgrade and

And certain areas of Europe seem to have had more "ethnicity" all along.
Think of the south part of the island of Great Britain, which was
"ethnically" divided into Welsh, Cornish, Saxon, and Norman populations,
linguistically defined, hierarchically ordered, and mutually stereotyping.
A division going back much further in time than some of the loudest
divisions in the Balkans -- and going back further than the nation-state as
we know it? or setting the pattern for the nation-state and ethnicities as
we know them?

So what is my point? I think that I am getting tired of the emerging dogma
(which I guess I don't entirely understand) that ethnicity is entirely to be
understood as deriving from nationalism which in turn is mostly a modern
invention of the nation-state. I want to show that the truth in that is
part of a truth which is far more complex than that. I feel that we are
"flattening the world" by assuming that modernity is its fount and
wellspring, and that dissenting voices are somehow becoming unwelcome. We
need to maintain our own identity as anthropologists, by bringing all times
and places into discourses which are becoming parochial in terms of present
times and privileged places. We used to identify ourselves with the
marginal and the obscure, but as we do this less I think we lose our
perspective and end up identifying with "metropolitan" processes and thinking.

Also, by looking at "nation-building" disputes, we forget how ethnic groups
are not always forged by State processes. Sometimes home-grown or
garden-variety discrimination, practiced by non-hegemonic folk, can do as
much to form or differentiate "ethnicities" (and even "ethnic hatreds" if
you like) as can nationalist intellectuals. And of course people, sometimes
people who are "dissed" by encompassing populations, often develop
"clannishness" (this is a chicken-and-egg or positive feedback cycle in
which looking for origins is, firstly, difficult, and secondly, part of the
positive feedback cycle in and of itself). I think that ethnicity,
"groupishness," should be seen as part of cybernetic or schismogenetic
feedback relationships, and that yes, it has been with us since the
beginning of human time.

But what people have been doing with their "groupishness" (or that of
others, or that of Others) has varied widely and been extremely specific to
time and place, since that beginning of human time. Nationalism is an
increasing part of what people are doing about their "groupishness" today;
it seems that it does not bring that "groupishness" into being. It just
creates a new mold into which the human propensity for group definition is

Sincerely, Phillip Guddemi