nC* C

R"EIH||J|

FIELDIANA MAY121977

Anthropology HH * llwail UiM1 ' v

Published by Field Museum of Natural History

Volume 68. No. 1 April 28, 1977

Human Biogeography in the Solomon Islands
John Terrell

Associate Curator, Oceanic ArchaeoijOGY and Ethnology
Field Museum of Natural History

As Ernst Mayr (1969) has observed, the "richness of tropical
faunas and floras is proverbial." Although the degree of species
diversity in the tropics has at times been exaggerated, tropical bird
faunas, for example, "contain at least three times if not four or more
times as many species, as comparable temperate zone bird faunas."
It is not surprising then that a tropical island as large as New
Guinea in the southwestern Pacific has played a special role in the
refinement of evolutionary theory ( Diamond, 1971, 1973).

In similar fashion, anthropologists have long recognized that the
Melanesian islands of the Pacific, including New Guinea, are re-
markable for the extreme degree of ethnic diversity encountered on
them (fig. 1). While the magnitude of the dissimilarities among
these tropical human populations has been occasionally overstated
or misconstrued (Vayda, 1966), even casual survey of the findings
made by social anthropologists, archaeologists, physical anthro-
pologists, and linguists in Melanesia would confirm Oliver's (1962,
p. 63) assessment that no other region of the world "contains such
cultural variety as these islands."

It is not an accident that zoologists, botanists, and anthropolo-
gists have observed that one word, diversity, so aptly sums up the
character of tropical populations in general, and the island popula-
tions of Melanesia in particular. This common judgment, however,
has not led to the development of a shared set of concepts and
models, applicable at least in part both to lower organisms and to
man, to account for that diversity. Although New Guinea has been
for years "a mecca, albeit difficult of access and penetration, for
students of evolution" (Diamond, 1971, p. 69), anthropologists have

Library of Congress Catalog Card No. : 77-07336
US ISSN 0071-4739

Publication 1257 1

2 FIELDIANA: ANTHROPOLOGY, VOLUME 68

customarily shied away from evolutionary interpretations purport-
ing to explain ethnological variety. Melanesia's human diversity
has been seen primarily as testimony for diffusionist and migra-
tionist reconstructions of mankind's ancient and more recent pre-
history, and as a rich mine of evidence attesting to the extensive
variation possible in the realm of culture. The reluctance of many
anthropologists to pursue the evolutionist's approach to diversity
reflects philosophical beliefs about human nature, the indetermin-
ism of historical events, and the uniquely varying totality of each
of mankind's many differing cultures. These beliefs have been
popular within anthropology for decades (Sahlins and Service,
1960).

In recent years, anthropologists working in many areas of in-
terest and in many parts of the world have begun to reconsider if
their profession has been wise to reject social evolutionism as a form
of naive "reductionism" or as a false "biological analogy." Since the
last world war, many other sciences have experienced what the
geographer, Ian Burton, has called the "quantitative revolution":
that is, they have come to insist more than before on quantitative,
statistical methods of research, and "an attendant emphasis on the
construction and testing of theoretical models" (Burton, 1968,
p. 13).

Quantification and model-building may not inevitably lead to
evolutionism. Yet, as Leslie White ( 1960, p. vii), one of the earliest
and most vocal champions for a return to evolutionism in cultural
anthropology, has written: "The concept of evolution has proved
itself to be too fundamental and fruitful to be ignored indefinitely
by anything calling itself a science. ' '

Intellectual trends within the science of geography have always
closely paralleled those in anthropology (Mikesell, 1967). Thus,
what another geographer, Peter Ambrose, has related about the
significance of the quantitative revolution of late years in that
science should have relevance to anthropology. Perhaps the chief
issue at stake during the revolution has been the question whether
the behavior and development of human societies were predictable.
According to Ambrose ( 1969, p. 13):

The central problem is perhaps not "Is man's collective action ever pre-
dictable?" since daily life would be impossible if it were not based on all sorts
of instinctive predictions about how other people and institutions will behave.
The operational question is "Are we ever likely to know enough about the way
people behave and interact in groups, and in the environment, to enable us to

4 FIELDIANA: ANTHROPOLOGY, VOLUME 68

carry out valid predictive work rather than purely descriptive work?" Many
social scientists, including human geographers, are showing implicitly in their
research, and explicitly in their methodological statements, that they believe
the answer to the second question to be "Yes." It should be borne in mind
that prediction in this sense does not depend upon the accurate forecasting of
the actions of every individual over a given future period. What is involved is
the statement, within specified levels of probability, that a human group will
tend to behave in a certain way given a stated choice of possible alternatives.

The quantitative revolution has influenced the conduct of re-
search and discovery in many of the social sciences, especially
geography, economics, sociology, and psychology. Its impact has
been felt also in other sciences concerned with the description and
analysis of population behavior. In the past 10 to 15 years, for
example, quantitative and paradigmatic advances have trans-
formed population biology, biogeography, and ecology.

In view of this trend toward quantification and predictive model-
building, there is justification for arguing that students of anthro-
pology, particularly those of human evolution and archaeology,
should reconsider the problem of man's great diversity in Melanesia
from perspectives that are detached from the ideographic, historio-
graphic tradition of previous research in the area and which are
inspired instead by the theoretical and procedural achievements
exemplified by current work in geography, population biology,
mathematical biogeography, and ecology.

This report describes a research program initiated in 1973 at Field
Museum of Natural History under my direction entitled: "Com-
parative Human and Lower Animal Biogeography in the Solomon
Islands." This account records how "human biogeography" is being
defined during this project, surveys the kinds of data examined,
notes some of the research problems that are being raised, and re-
views two initial studies that were undertaken to ascertain how
effectively certain principles and models previously worked out in
the field of lower animal biogeography can be utilized as well to
study the human populations of the Solomons.

Definitions

In 1957 Pierre Dansereau (1957, pp. 3, 319), a botanist, defined
biogeography as "the study of the origin, distribution, adaptation,
and association of plants and animals." He (1957, p. 323) described
ecology as "the study of the reactions of plants and animals to their
immediate environment, to their habitat (and not to their geogra-
phical location)." While it may be unwise to separate ecological

TERRELL: HUMAN BIOGEOGRAPHY 5

endeavors today sharply from biogeographical ones (MacArthur,
1972), it can be seen that contemporary ecology and biogeography
do not present themselves as simply two alternative academic
disciplines with the same range of perspectives on the world. At
very least there is a difference in scope between the aims of the
biogeographer and those of the ecologist, a difference which appears
in practice, if not always in theory.

In their pioneering treatise on theoretical island biogeography,
MacArthur and Wilson ( 1967, p. 185) describe biogeography as the
study of "the distribution of species of organisms over the face of
the earth. Biogeography is concerned with the limits and geometric
structure of individual species populations and with the differences
in biotas at various points on the earth's surface. The local, ecologi-
cal distribution of species, together with such synecological features
as the structure of the food web, are treated under biogeography
only insofar as they relate to the broader aspects of distribution."

The earlier definition of biogeography given by Dansereau and
this one by MacArthur and Wilson comprise a foundation upon
which to develop a definition of the phrase "human biogeography"
used throughout this paper. Since MacArthur and Wilson, as well
as Dansereau, distinguish between ecological concerns and biogeo-
graphical ones, it is best to begin with a definition of "human
ecology," and then to observe that a similar contrast can be drawn
between that area of interest and human biogeography.

Human ecology is the study of relationships between human pop-
ulations and their immediate environment, including other species
populations (see Bates, 1953). As the geographer Barrows (1923,
p. 3) argued years ago, human ecology investigates "relationships
existing between natural environments and the distribution and
activities of man. ' '

In contrast, human biogeography, as thought of here, is the study
of the size, distribution, and population structure of, and the inter-
actions among, human populations found in similar or divergent
habitats, and of the conditions and events leading to the develop-
ment and maintenance of similarities and differences among human
populations living at various points on the earth's surface.

These two definitions imply a difference in both scope and subject
matter. While biogeographical research often includes the investi-
gation and understanding of ecological relationships at the local
level, it also incorporates analysis of the origin, development, and

6 FIELDIANA: ANTHROPOLOGY, VOLUME 68

maintenance of formal and functional similarities and differences
among and within populations. The data of human biogeography
thus include information on the history, characteristics, distribu-
tion, and relationships, in addition to the ecology, of human groups
at the regional and global scale.

To describe human biogeography, use has been made of several
concepts and terms:

population size: biologically, the number of potentially inter-
breeding individuals at a given locality (the "local population" of
Mayr, 1970, p. 82). May be more generally defined as the number
of individuals within an area which interact in a designated
fashion at a potential frequency determined to be in accord with
the purpose of the study at hand.

population distribution: the geographical occurrence of a popula-
tion in space and time.

population structure: the manner in which the components (i.e.,
individuals, neighborhoods [Rohlf and Schnell, 1971], sub-popula-
tions) of a population are inter-related in space and time; the
pattern of distribution of such components (linear, random,
clustered, etc.). In other words, the arrangement of their occur-
rence within a determined area and during a given unit of time.

population interaction: the exchange of individuals, energy,
genetic material or information among the components of a popu-
lation or among populations; the effect of one component or
population on another; the movement of one component or popu-
lation relative to others leading to a change in effective relation-
ships.

These concepts, as defined, are intimately related to each other.
The chief underlying ideas are those of number, frequency of occur-
rence in space and time (i.e., relative abundance), location, and
mutual effect. Therefore, the definition of human biogeography, and
of the components said to be its units of study (individuals and
groups or populations of individuals), is implicitly a systems
approach (von Bertalanffy, 1968) to the spatial-temporal organiza-
tion of human societies and to the evolution of man's biological and
cultural diversity. Considering the vogue popularity of "general
system theory" today, however, there is no need to emphasize this
feature of the approach being advocated.

TERRELL: HUMAN BIOGEOGRAPHY 7

There is a need to indicate how human biogeographical research
differs from customary studies long undertaken in anthropology or
geography. Further, it is necessary to demonstrate that such inves-
tigations are practical and effective.

First, the kinds of subjects and relationships subsumed under
the rubric "human biogeography" are by no means foreign to an-
thropologists and geographers. What distinguishes human bio-
geography from other related pursuits is twofold: (1) biological
populations, rather than geographical "regions" or anthropological
"cultures" and "societies," are taken as the effective unit of inter-
pretation; (2) investigation is focused on the search for regularities,
or patterns, in the characteristics, organization, and behavior of
human populations living under similar or divergent geographic
circumstances and in different parts of the world. Thus, unlike
geography or anthropology, the study of human biogeography is
specialized and comparatively narrow. At the same time, its goals
are more akin to work in zoology, biology, and ecology than much of
past and current research in anthropology and geography.

Second, the utility of work in human biogeography can be demon-
strated by examining the kinds of information about human and
lower animal populations in the Solomon Islands available for
biogeographical analyses.

A review of some of the existing biogeographical work on lower
animals in the Solomons will point out the kinds of research ques-
tions of interest to biogeographers and some of the patterns and
processes already observed that might have relevance to the study
of human populations in the archipelago. Following this review,
various perspectives on mankind's diversity in the Solomons will
then be explored, and when appropriate, tentative biogeographical
comments will be offered, before turning to two specific research
topics.

Lower Animal Biogeography

Appropriately, the scientist who pioneered the study of natural
history in the Solomon Islands, Henry Guppy, was also the man
who gave us one of the earliest and most comprehensive surveys of
the anthropology of the archipelago (Guppy, 1887). Since Guppy *s
day substantive research in both natural history and anthropology
has been combined only rarely, although anthropologists have kept
up at least peripheral interest in flora and fauna of economic impor-
tance to the islanders, and zoologists have contributed articles, in a

8 FIELDIANA: ANTHROPOLOGY, VOLUME 68

more or less popular style, on native life and customs (Mayr, 1930-
1931; 1943; Mann, 1948). Few scholars working in these fields of
endeavor have weighed seriously the possibility that a combined
study of the biogeography of man and lower animals in the Solo-
mons might be instrumental in enhancing our understanding of
population dynamics and prehistory on these islands or among
similar socio-economic populations elsewhere in the world.

The most influential research on lower animal species in the
Solomon Islands has been that by Ernst Mayr on bird faunas (for a
summary, see Mayr, 1969). Synoptic studies by P. J. M. Greenslade
(1968a, b, 1969) on island patterns in bird and insect diversity,
however, may be used to illustrate the general relevance of lower
animal research to the anthropology and archaeology of the island-
ers. Greenslade's syntheses are particularly germane, because he
has framed his reports in the innovative theoretical tradition
associated with the late Robert Mac Arthur and Edward Wilson,
alluded to previously.

According to Greenslade, the striking feature of animal distribu-
tions in the Solomons and elsewhere in the Pacific is the occurrence
of endemic species and subspecies which are to be found on only one
or a few islands. Although, as a result, patterns of faunal similari-
ties and differences among islands are complex, he (1969, p. 274)
demonstrates that the biogeographical problems presented by re-
stricted island distributions within the Solomons "can be analysed
simply in terms of the existing islands, their areas and distances
from each other and from major sources."

His own efforts to define the principal faunal patterns in the Solo-
mons and explain their probable origin rely upon several basic
assumptions which indicate some of the fundamentals of a biogeo-
graphical approach to species diversity:

1. Species, when expanding their geographic distribution, move
most often from large islands to smaller ones, rather than the re-
verse. He justifies this assumption by noting that it may be de-
rived readily from biogeographical theory, including the work of
MacArthur and Wilson on island species equilibrium dynamics
( see below : ' ' Proximal-point analysis " ) .

2. Historically, there has been a dominant west- to-east expansion
of species out into the Pacific Basin from source areas in New
Guinea and Mainland Asia. This claim is consistent with known

TERRELL: HUMAN BIOGEOGRAPHY 9

evidence on species distribution patterns in the Pacific ( Zimmer-
man, 1942).

3. Narrow water gaps between islands are crossed more easily
than wide divides. This assumption makes common sense, and is
in accord with MacArthur- Wilson theory on island biogeography.

4. Species distribution patterns may be analyzed in terms of the
existing geographical configuration of islands in the archipelago
which probably dates from some time in the Pleistocene. Com-
parative evidence from elsewhere in the Pacific suggests that such
a time-depth is sufficient for the degree of species differentiation
observed in the Solomons to have developed. This last assumption
is also in line with current theory which emphasizes the impor-
tance of species colonization and extinction as continuing pro-
cesses in maintaining insular faunas.

Taking these assumptions, Greenslade proposes that the most
important routes of species expansion from the west down the Sol-
omons chain can be diagrammed as in Figure 2. He notes that the

BOUGAINVILLE - CHOISEUL- -SANTA ISABEL -NGGELA

:

NEW GEORGIA RUSSELL It - GUADALCANAL - SAN CRISTOBAL

GROUP

Fig. 2. Main routes and directions of faunal movement in the Solomon Islands
(after Greenslade, 1969, fig. 16).

main route from Bougainville to Nggela, although sketched by
assessing relative island areas and widths of water barriers, corre-
sponds with a possible land bridge during the low sea-levels of the
Pleistocene. He judges this former land connection, however, to be
less important to present-day faunal distributions than the narrow-
ness of the existing water gaps.

Greenslade constructs this pathway diagram to forecast the most
probable species patterns in the Solomons, in order to determine
how successfully observed distributions coincide with expectations
derived independently from his initial assumptions. There is no
reason, however, why rough modeling of this kind in biogeography
must be limited to the prediction of historical migration trends. In
fact, it need not be, as a later discussion of a procedure in human
biogeography, termed "proximal-point analysis," will demonstrate.

10 FIELDIANA: ANTHROPOLOGY, VOLUME 68

Greenslade goes on to show that for both birds and insects two
major distribution patterns dominate in the Solomon Islands. First,
in both cases there is a high incidence of island endemics on San
Cristobal, located at the most easterly extreme of the archipelago.
Second, the islands from Bougainville to Nggela postulated to be on
the main route for species expansion do, indeed, constitute a dis-
tinct group of islands with inter-related faunas. Since the excep-
tional degree of endemism on San Cristobal is also to be expected
from Greenslade's pathway diagram, both major patterns suggest
that his approach has much to recommend it.

The recognition of these and other lesser distributional patterns
in the Solomons is in itself a significant contribution to lower
animal biogeography, because it affirms that variables such as
island area, the distance between islands, and the degree of an
island's isolation and distance from potential source areas are as
important in determining patterns of species diversity as generally
assumed in theoretical island biogeography. If this is the case, then
there is reason to argue that efforts should be made to determine
how effectively such variables may be used to analyze human bio-
logical and cultural diversity in the Solomons. As Greenslade ( 1969,
p. 274) has clearly stated: "Only if these factors fail to account for
observed animal distributions is it necessary to introduce any
others."

The relevance of Greenslade's work to human biogeography,
however, does not end merely with this broad claim. Greenslade
further develops the argument that the two main distribution pat-
terns represented by San Cristobal and the Bougainville-Nggela
fauna chain can be explained best by hypothesizing a series of three
successive stages in the expansion of taxa eastward through the
Solomons, stages in what he envisions as a continuing "taxon
cycle" of species expansion, subsequent local differentiation, and
final distributional contraction ( cf. Wilson, 1961).

Greenslade's development of his concept of a Solomon Islands
taxon cycle as a logical process of expansion from source areas
farther west creates a distinct possibility that what he has to say
might be adopted in human biogeography as a first-order approxi-
mation in efforts to define processes behind human immigration,
cultural diffusion, and local island diversification in the archipelago.
Although, as he himself notes, actual species distributions are often
more complicated and less clear-cut than suggested by his taxon

TERRELL: HUMAN BIOGEOGRAPHY 11

cycle, 1 the model is especially provocative because he has been able
to correlate the stages he outlines with important ecological and
adaptive changes.

According to Greenslade ( 1969), individual species may be classi-
fied by their pattern of distribution in the Solomons into one of
these three stages:

Stage I: Initial expansion through the Solomons

Species at this stage in their history are in the process of expand-
ing their range of geographical distribution from the local source
area to the east, Bougainville Island. Also included in this stage are
species that have continuous distribution throughout the archi-
pelago, without evidence for the onset of local differentiation.

With respect to habitat preferences, species at this point in their
local history live either in a wide range of habitats, or occur at low
elevations, especially in temporary coastal environments where it is
probable that species are selected for their migratory ability, which
would preadapt them for inter-island dispersal. Faunas in such hab-
itats are also commonly of low diversity, and immigrant species
arriving in these areas are likely to face only limited competition.

Stage II: Differentiation

Species classified in this stage give evidence of having differentia-
ted into local taxa, with perhaps some suggestion of extinction on
small or isolated islands, and possibly secondary expansion as well
from large to small islands.

Stage II species in the Solomons are to be found almost exclu-
sively in the lowland rain forest. At least on larger islands in the
archipelago, the Stage II shift in habitat preference marks an in-
crease in species diversity and a probable higher incidence of com-
petition. In observed cases, there is a strong suggestion that Stage
II species have adapted themselves to specific local conditions, "es-
pecially if the habitat shift is associated with any loss of dispersal
ability and consequent break up of the population on an island into
smaller units" ( 1969, p. 278).

Stage III: Distributional contraction

Species at this stage show fragmented and discontinuous distri-
butions, and appear to be contracting their total area of occurrence.

'According to Diamond (pers. comm.), Greenslade's taxon cycle is incomplete and
many species patterns cannot be fitted to it.

12 FIELDIANA: ANTHROPOLOGY, VOLUME 68

Many Stage III species have withdrawn to mountains or, in some
instances, small or isolated islands, where there is a possibility that
competition is lower than in the rain forest.

The pertinence of this logical model to the study of human groups
is observed in the questions relating to human biogeography that
the model raises. For instance, Greenslade emphasizes by means of
the taxon cycle that colonization and local extinction are continuous
processes which serve to maintain island faunas. How legitimate is
the assumption that similarly continuous processes affect human
populations in Melanesia? Further, Greenslade also proposes that
the stages in the taxon cycle operate along predictable routes
through the Solomons. Can the paths of human migration and
inter-island affinities also be predicted in a fashion comparable to
that taken by Greenslade?

In developing the idea of a Solomons taxon cycle, Greenslade
points out the importance of factors such as dispersal ability, habi-
tat range, competition, and extinction in determining species dis-
tributions. Are immigrant human populations "pre-adapted" for
dispersal? Are they likely to show a wide habitat range? Is it
probable that migrants will settle in temporary coastal environ-
ments, where they might face less "competition" from previously-
arrived human groups? Is it possible that human populations tend
to develop through stages comparable to the three outlined for low-
er animal species? If human populations shift from coastal areas to
the lowland rain forest, do they also tend to fragment into smaller
settlement units? Is there something comparable in human biogeo-
graphy to the heightened inter- species competition in the rain
forest? How does geographic location and habitat preference affect
the likelihood of extinction among human populations?

These are only a few of the questions that could be raised through
comparative study of human and lower animal biogeography in the
Solomon Islands. Certainly questions such as these are not without
precedence in anthropology and archaeology. Nonetheless, it may
be argued that an explicitly comparative approach brings the
ecological, geographical, historical, and biological issues involved
sharply into focus. Moreover, comparative study points to possible
solutions already explored in animal biogeography, solutions which
can function at very least as a foundation upon which to develop
models and explanatory hypotheses in human biogeography.

terrell: human biogeography 13

Human Biogeography

The human diversity of the Solomon Islands can be observed from
a number of perspectives: in biology, language, geographical cir-
cumstance, social and material culture, and in prehistory. Although
the existing level of information on these perspectives is insuffi-
cient for extensive work in human biogeography, there is reason to
believe that the adoption of theoretical and procedural approaches
such as those advocated here might redress the balance con-
siderably. In the process, it seems probable that anthropological
research in the Solomons can be integrated more closely with
ecological and biological studies in the archipelago.

Human biological diversity

It has long been recognized that the islanders living in the Solo-
mons differ among themselves in their physical appearance. The
northern islands, from Nissan some 40 miles north of Buka on down
to the New Georgia area, are commonly known as the "black spot"
of the Pacific in which the skin color of the islanders approaches
that of the darkest African populations (Oliver, 1949, p. 13). It has
also long been claimed that the density of skin pigmentation
steadily declines as one travels south of this region along the Solo-
mons chain (Firth, 1944, p. 632). In addition, the small, isolated
islands of Rennell, Bellona, Ontong Java, and Sikaiana are some of
the famous enclaves of Polynesian-speakers in island Melanesia.
These islanders appear physically unlike their Melanesian neigh-
bors and resemble, instead, the island populations farther east with
whom they appear to be tied linguistically (Simmons and Gajdusek,
1966; Willis and Booth, 1968; Booth and Vines, 1969).

Since World War II the pace of modern biological research on the
genetics of the Solomon Island populations has increased rapidly.
While today genetic research is far from comprehensive, a major
program carried out since 1966 by teams from Harvard University
working on Bougainville and Malaita has greatly expanded our
knowledge of at least these islanders (Damon, 1973). Perhaps the
most significant finding to date for human biogeography is the
recognition of a possible cline in the distribution of the Inv x allotype
of human immunoglobulin G, with a decrease in frequency from
Bougainville to Malaita (Steinberg et al, 1972). Such a pattern sug-
gests that the population structure of the chain of larger islands
running along the eastern side of the archipelago may prove, in
general, to have a clinal pattern (Mayr, 1970, pp. 215-217), a struc-

14 FIELDIANA: ANTHROPOLOGY, VOLUME 68

ture in conformity with the more crudely recognized pattern in skin
color (above), and in accord with what might be expected on the
basis of an elementary simulated network model of inter-island re-
lationships (pp. 34-39). The pattern also agrees with what is known
about the biogeography of variation in some species of lower
animals, as noted previously.

Without further elaboration, continuing research on the biology
and genetic structure of the Solomon Islanders should supply direct
and useful information pertinent to human biogeography.

Linguistic diversity

In a recent study of the available linguistic information on the
New Hebrides and the Solomons, George Grace (1971, p. 341) ob-
serves, as indeed others have before him, that both of these island
territories "present the characteristic Melanesian picture of great
linguistic diversity within a restricted area." Because knowledge of
the languages of the Solomons is rudimentary, only a rough esti-
mate can be given for the total number of mutually unintelligible
tongues spoken there. The most extensive and complete listing
(Capell, 1962) records the existence of nearly 80 Melanesian (i.e.,
Austronesian), Papuan (i.e., "Non-Austronesian"), and Polynesian
languages in the British Solomons alone. If to this number the lan-
guages of the Papua-New Guinea Solomons are added (Allen and
Hurd, n.d.), the total count runs up to over 100 separate tongues.
Moreover, from the limited information available it is certain that if
dialectal diversity were included, the figure might be several times
this estimate.

Even though little is known about most Solomon Island lan-
guages, it is possible to arrive at a few general statements about
them (Grace, 1971). First, Polynesian languages are found on a
small number of the more remote islands. These so-called "Poly-
nesian Outliers" are commonly interpreted today to be linguistic
colonies derived from farther east: that is, foreign speech communi-
ties that have managed to survive perhaps because of their isolation
from even their nearest Melanesian-speaking neighbors.

Second, although Papuan languages in the Solomons are similarly
few in number, their occurrence so far from the New Guinea main-
land, where most Pacific Island languages classified rather ambigu-
ously as Non-Austronesian are to be found, is something of an
enigma. It is a moot question, for instance, whether or not they
represent ancient populations that were formerly more widely-

TERRELL: HUMAN BIOGEOGRAPHY 15

distributed and which have been replaced on most islands by later-
arriving Austronesian-speakers (a pattern which would be com-
parable, in other words, to the distributions of those lower animal
species in the Solomons that have reached Greenslade's Stage III
in his postulated taxon expansion cycle). An alternative interpreta-
tion, however, might be that they are highly divergent but locally-
evolved linguistic isolates that may be as little related to each other
as they are to the dominating Austronesian language populations.

Linguists appear to favor the first solution to the problem pre-
sented by the Papuan languages. Yet they do not always agree on
why Papuan-speaking communities have survived in the Solomons,
precisely how and why they were replaced by Austronesian-spea-
kers on the islands they may have once occupied (assuming there
was in fact population replacement rather than language extinc-
tion), and why they occur not only in areas which could be described
as "mountain refuges," but also on several of the smaller islands in
the archipelago.

A major linguistic phylum of 12 Papuan languages and sub-
languages exists only on Bougainville at the northern extreme of
the Solomons (Allen and Hurd, n.d.). Elsewhere, languages also
classified as Non-Austronesian appear on Vella Lavella, Rendova,
New Georgia, Russell Islands, Savo Island, and in the Reef Islands
— Santa Cruz group beyond the southeastern end of the chain
(Wurm, 1971, pp. 631-632, 646-647).

The linguist Stephen Wurm (1973, pp. 2, 10-11) claims that these
Papuan languages should be grouped with all the Non-Austronesian
languages occurring in island Melanesia east of the main island of
New Guinea. He believes that this "East Papuan Phylum" may be
taken as evidence for inter-related prehistoric migrations by a "lan-
guage group" originally in possession of the southeastern reaches of
New Guinea Island.

Based on her own field research, Evelyn Todd ( 1973) of Trent Uni-
versity has proposed tentatively that at least the Savosavo (Savo
Island), Lavukaleve (Russell Islands), Baniata (Rendova Island),
and Bilua (Vella Lavella) Papuan languages can, in fact, be grouped
together into a single family which she has termed the "Solomon
Language Family." Building on a recent suggestion by Greenberg
(Greenberg, 1971, pp. 816-819), Todd also proposes that the Yele
language of Rossel Island in the Louisiade Archipelago near south-

16 FIELDIANA: ANTHROPOLOGY, VOLUME 68

eastern New Guinea Island should be included in the posited Solo-
mon Family of Non-Austronesian languages.

Unlike Wurm, she has not explicitly drawn migrational infer-
ences from this latter proposition. Indeed, again on evidence from
her own original field work, Todd questions Wurm's claim that
there is a demonstrable connection between at least one of the Non-
Austronesian languages of Bougainville Island (the Nasioi lan-
guage of the east coast), the Baining language of New Britain, and
what she recognizes as the Solomon Family (including Yele).

Clearer knowledge of the development, maintenance, and affini-
ties of the Non-Austronesian languages of the Solomons should
provide useful information on human biogeography in the archi-
pelago. While it may be possible to argue in some cases that the
divergence of these languages from the rest of the languages in the
Solomons has been maintained (and at least to a degree fostered) by
their spatial or topographic isolation (e.g., the Bougainville
Phylum), in other cases these linguistic populations occur in geo-
graphic locations where such an explanation would be difficult to
entertain (e.g., Savosavo ) .

Third, apart from the relatively minor groups of Polynesian and
Non-Austronesian languages, the majority of Solomon Island
tongues are classified by linguists as "Melanesian Austronesian."
Unfortunately, this allocation is less helpful than it might be, be-
cause the recognition of the class reflects as much the history of
Oceanic linguistics as any sound linguistic reasoning (Grace, 1971,
pp. 343-345). This observation underscores the point that the
human biogeographer is just as dependent on the accuracy of funda-
mental taxonomic work as his botanical and zoological counter-
parts. There is little evidence to claim that the Melanesian
languages of the Solomons, taken together, constitute a single
genetic grouping within the vast assemblage of Pacific Austro-
nesian languages (Grace, 1971, p. 347; Pawley, 1972, pp. 137-138).

Specifically, there are linguists who maintain that the languages
spoken in the southeastern Solomons most closely resemble
languages occurring in Polynesia, Fiji, Rotuma, and in certain areas
of the New Hebrides. Moreoever, they would exclude the languages
of the western (British) Solomons (and probably also those of the
northern Solomons, as noted later) from any grouping that incor-
porates the southeastern Solomons (Pawley, 1972, pp. 98-110).
Briefly stated, the argument often put forward is: the languages of

TERRELL: HUMAN BIOGEOGRAPHY 17

the western and northern Solomons seem to be "idiosyncratic" or
"less Austronesian" in character (Wurm, 1967, p. 33; Capell, 1969,
pp. 147); those of the southeastern Solomons display a "more sub-
stantial Austronesian component" or are more "typically" Austro-
nesian in form (Grace, 1971, pp. 345-349).

George Grace (1971, p. 349) has commented that the historical
significance of this apparent difference between idiosyncratic and
typical Austronesian languages has been variously interpreted and
remains uncertain. According to some scholars, the contrast may
reflect, in part, a difference in time of divergence (Grace, 1971, p.
346). Grace himself is one of a growing number of lignuists who
would interpret Polynesian, a case in point, as a "relatively recent
offshoot" from certain ancestral languages commonly presumed to
have been spoken somewhere in the southeastern Solomons ( Paw-
ley, 1972, p. 141).

In spite of the limitations and uncertainties of current linguistic
evidence about the Solomon Islands, can such information contri-
bute to human biogeography? A number of generalizations may be
made which have biogeographical significance: ( 1 ) The total count
of languages in the Solomon Islands is "impressively large" and
"there is still no language with more than a few thousand speakers"
(Grace, 1971, p. 350). Inevitable questions arising from these state-
ments are: What conditions serve to maintain such small speech
populations? Has so large a number of distinct, mutually unin-
telligible languages in so small a region evolved locally? How exten-
sively must one posit "language group migrations" to explain ob-
served affinities between separate tongues? How closely do distinct
languages correlate with contrasting ecological environments?
Alternatively, how far do the conditions of life regulating popula-
tion interaction determine the maximum size that a given popula-
tion may attain while continuing to preserve a more or less homog-
eneous language or set of mutually intelligible dialects?

(2) A small number of remote speech communities (i.e., the Poly-
nesian Outliers) appear intrusive. Their occurrences give rise to the
likelihood that these populations may be usefully examined to ex-
plore the kinds of events and circumstances that contribute to the
colonization, survival, or extinction of divergent human populations
within an island world.

(3) The linguistic diversity of the Solomons displays a degree of
spatial patterning wherein the western (and, it is proposed here, the

18 FIELDIANA: ANTHROPOLOGY, VOLUME 68

northern) populations may be contrasted with those in the south-
eastern limits of the archipelago. How closely does this linguistic
variation follow a clinal pattern?

(4) The divergence of the languages of the southeastern Solo-
mons may reflect a difference in time of divergence. Specifically,
those islands farthest from the Asiatic Mainland (presumably the
"homeland" or original source area for the islanders) and nearest
the scattered Polynesian-speaking islands of the central and eastern
Pacific are also most like the Austronesian languages of Polynesia
and the intervening New Hebrides. Can the linguistic evidence be
employed to define probable patterns of advancing island coloniza-
tion? Or is it necessary to propose a more complex and extensive
pattern of human migrations out of New Guinea and perhaps South-
east Asia to account for observed affinities? With regard to this
last possibility, some linguists would incorporate Austronesian lan-
guages found in the Bismarck Archipelago and on the Papuan Coast
of southern New Guinea within the so-called "typical" class of Aus-
tronesian languages, while at the same time they have proposed
excluding others occurring in the southern New Hebrides (Grace,
1971, p. 348; Pawley, 1972, p. 136).

In summary, close attention to linguistic diversity in the Solo-
mons archipelago for evidence on migration (i.e. "colonization" in
the biogeographical sense) and also on spatial-temporal differentia-
tion (local linguistic evolution) should contribute much in future
years to understanding the population structure, history, and net-
works of population interaction within and beyond the Solomons. A
few examples supporting this observation will be given later in
detail. At this point there is some reason to think linguistic diver-
sity in this part of the world conforms relatively well to the kind of
area/species diversity pattern often encountered on islands in the
study of lower animal biogeography. Further, use of an elementary
network model of the Solomons, alluded to previously, leads to the
"prediction" that the linguistic populations living in the western
and northern Solomons ought to be, as they indeed appear to be,
observably different from those residing on the eastern and south-
eastern islands of the archipelago.

Geographical Diversity

To date the most comprehensive and ambitious attempt to deter-
mine how extensively man/land relationships in the Solomons have
contributed to the evolution and maintenance of biological and cul-

TERRELL: HUMAN BIOGEOGRAPHY 19

tural diversity in the region would appear to be my (Terrell, 1976;
summarized in Terrell, in press) own research on the peoples of
Buka and Bougainville islands in the Papua New Guinea Solomons.
The scheme of "geographic systems" devised for those islands is
sufficiently analytical that it might be generalized as a model for
work elsewhere. Currently, such an extension is planned under the
aegis of the human biogeography research project at Field Museum.

Ethnological Diversity

Douglas Oliver (1962, p. 225) has written that "nothing short of
an encyclopedia could fully describe the great variety of social forms
and religious practices of all the Solomons." Indeed, "marked tribal
differences . . . spell despair for anyone seeking a simple formula for
understanding the Solomonese as an entity." It may not have been
despair that led to the decline of comparative ethnology in the Solo-
mons and elsewhere after its heyday at the turn of the century. Yet
there is no doubt that few contemporary anthropologists have
attempted to match the encyclopedic breadth of the pioneering
ethnologists who wrote about island Melanesia (Guppy, 1887;
Codrington, 1891; Rivers, 1914). For the most part, recent studies
in Solomon Island comparative ethnology have been concerned only
with selected aspects of material culture (e.g., Speiser, 1942a, b;
Anell, 1955). Integrating formulas, simple or otherwise, have been
remarkable for their rarity.

Nevertheless, there may prove to be a few useful formulas to dis-
till from the seeming welter of ethnographic details, provided that
the apparatus to do so is properly set up. For example, Blackwood
(1931, p. 199) writes: "Buka and Bougainville are the two most
northerly islands of the Solomon Group, but from the ethnological
standpoint they differ from the better-known British Solomons in
many important respects." Similarly, and more recently, Cranstone
(1961, p. 47) at the British Museum observed: "Culturally and
artistically the Solomon Islands can be divided into two main areas,
each of which can be sub-divided: Bougainville, Buka and some
smaller islands in the north-west, and the rest of the group."

Although impressionistic, statements such as these by Black-
wood and Cranstone, taken in combination with more localized
descriptions of specific islands or island groups (for example,
Oliver, 1949, pp. 12-13), give the clear impression that the formula
of spatial patterning could be applied successfully to the peoples
and cultures of the archipelago. Indeed, when an explicitly popula-

20 FIELDIANA: ANTHROPOLOGY, VOLUME 68

tionist approach is used to analyze and comprehend ethnological
diversity in the Solomons, the results are illuminating (Terrell and
Irwin, 1972).

Without question, the best method to study the ethnological
diversity of the Solomon Islands on an extensive scale is to observe
directly what the islanders are doing "on the ground" (Leach, 1965,
p. xiii). Until such field studies in human biogeography can be
carried out, however, efforts at Field Museum will focus on renewed
examination of items of Solomon Island material culture held
among the extensive collections from Melanesia housed at the
Museum.

Henry Cummings' pilot study of compound (bonito) fishhooks
from the Solomons, published as the first issue of Solomon Island
Studies in Human Biogeography (Cummings, 1973), succinctly
demonstrates the feasibility and effectiveness of comparative
work with museum collections. In spite of deficiencies in the
provenance data available to him, Cummings has sketched a plausi-
ble geographic pattern that prevails over the diversity observable in
hook types. On this evidence he has been able to suggest likely
inter-island connections and disjunctures.

The principal advantage of such museum studies arises from the
hypotheses about population structure and interaction that may be
framed as a result of such work. Further testing of these hy-
potheses, of course, will depend ultimately on additional research
at museums and on biogeographically-designed fieldwork.

Prehistoric Diversity

While a certain amount of research that might be labelled archae-
ological in character was carried out in the Solomons in the years
prior to the Second World War, planned surveys and controlled ex-
cavations were not attempted until 1964-1966. William Davenport
(1972), a social anthropologist at the University of Pennsylvania,
excavated at archaeological sites on Santa Ana (Owa Raha) and
Santa Catalina (Owa Riki) islands south of San Cristobal while he
was in the area as an ethnologist. Also in 1964, Keio University in
Japan sent an archaeological and ethnological expedition to the
Solomons to survey parts of Choiseul, New Georgia, Simbo,
Ganongga, Vella Lavella, and nearby islets. Masashi Chikamori
conducted excavations on Vella Lavella (Chikamori, 1967).

TERRELL: HUMAN BIOGEOGRAPHY 21

Since these first explorations, work has been done on Buka
(Specht, 1969), Bougainville (by this author), the Shortlands
(Irwin, 1972, 1973), Bellona (Poulsen, 1972), Kolombangara
(Kirsch, pers. comm.), Guadalcanal (Davenport, 1968), and in the
area of the southeastern Solomons and Santa Cruz Islands (Green,
1973).

The increasing interest in the prehistory of the Pacific is reflected
in the rapid growth in the pace of field research in the archaeology of
the Solomons since 1964. There is reason to believe, nonetheless,
that the aims and research designs of only some of these field pro-
jects have been such that their results may be readily and exten-
sively utilized to explore broader questions pertinent to the bio-
geography of man in the archipelago. While this observation could
be argued, it may be enough to claim that the surveys and excava-
tions conducted on Bougainville and in the Shortland Islands were
undertaken to explore the prehistory of those islands in ways which
may now be defined succinctly as biogeographical in intent (Irwin,
1972, 1973; Terrell and Irwin, 1972; Terrell, in press). In addition,
the extensive research project organized by Roger Green of
Auckland University in the southeastern Solomons, while designed
chiefly to pursue issues in culture history concerning the origins of
the Polynesian peoples, promises to supply information bearing
upon human biogeography.

There may be inevitable difficulties which will have to be faced
during any attempts to interpret archaeological evidence in ways
commensurate with procedures, concepts, and models used in bio-
geography. Such difficulties may include the fragmentary and
selective nature of the archaeological record, chronological impre-
cision, conflicting interpretations of the same evidence, and the like.
Nonetheless, efforts toward this end will be crucial in human bio-
geographical research in the Solomon Islands. Just as paleontolo-
gical studies are fundamental to work in plant and animal
biogeography, archaeology is the most reliable means by which
explanatory models and hypotheses purporting to explain human
diversity and human biogeographical patterns in the Solomons can
be given the test of time.

Analysis and Synthesis in Human Biogeography

In a popular article on the Solomons, written some years ago,
Ernst Mayr (1943, p. 30) asserted: "There is probably no island

22 FIELDIANA: ANTHROPOLOGY, VOLUME 68

group in the world that surpasses the Solomon Islands for variety.
Only one general statement can be made about them and that is
that one cannot generalize." Although his explicit reference is to
their geologic diversity and ethnic complexity, the hyperbole in his
claim should not detract from its broader relevance. The brief sur-
vey of lower animal and human diversity in the archipelago just
presented suggests that these islands offer a great deal of grist for
the biogeographer's mill.

The Solomon Islands research program, commenced at Field
Museum in 1973, has these objectives:

1. As MacArthur (1972, p. 77) has insisted: "The concept of pat-
tern or regularity is central to science." The first and fundamental
objective of the Solomons program is to determine how successfully
one can recognize patterns in human biological and cultural diver-
sity repeated in space from one locale or area to another, and re-
peated, as well, in data reflecting differing aspects and characteris-
tics of human settlement and adaptation. Prime emphasis, in short,
is on phenomenology and pattern recognition.

2. Phenomenology denotes classification along with description
and implies a low order of synthesis. Pattern recognition, however,
signifies something more. "Pattern implies some sort of repeti-
tion," as MacArthur has also written. "The existence of the repeti-
tion means some prediction is possible" (MacArthur, 1972, p. 77).
The second objective of the Solomons Program aims at a higher
level of synthesis than the first: to discover how effectively simple
models, either graphical or algebraic, may be developed to predict
geographic pattern variation among tropical populations of subsis-
tence horticulturalists.

3. No serious anthropologist would contend the assertion that
man's capacity for culture and complex decision-making places man
apart from other species populations. Yet few anthropologists
would be brave enough to say where the threshold lies between man-
as-biological-organism and man, the bearer of culture, man the
decision-maker. For this reason, one goal of the Solomons project is
to discern how legitimately models constructed solely with animal
populations in mind may be applied to human populations as well.
The operative questions being asked are: What factors, if any, must
be introduced to model human populations that are not considered
relevant to other species populations? Under what circumstances
are they required? Are they universally needed whenever human

TERRELL: HUMAN BIOGEOGRAPHY 23

populations are being studied? Or do they apply only in certain
situations?

These are the three principal aims of the comparative biogeogra-
phy program. Before describing two examples of research which has
been done to further these aims, however, a few additional points
must be emphasized. First, a philosophical comment. For the most
part, the problems being pursued during the current program are
ones suggested by evidence from the Solomon Islands. Nonethe-
less, sound research proceeds in two complementary directions:
from evidence to synthesis, and from model-building to the search
for supporting and refuting data. None of the three objectives out-
lined here, therefore, is held to be of greater weight than the others.

Second, the relevance of models, concepts, and procedures taken
from other disciplines, notably population biology, locational
geography, and theoretical biogeography, is considered one of the
objects of study, not a foregone conclusion. There is good reason for
this position. As the geographers Chorley and Haggett (1969, pp.
21-26) have insisted, a model is a simplification of reality and must
not be confused with reality itself. The biologist Levins (1966, p.
430) has phrased the case in unmistakable terms: "all models leave
out a lot and are in that sense false, incomplete, inadequate." It
would be naive to hope that a given model, concocted in one disci-
pline, could be carried over, wholly and without modification, to
serve double-duty in another, to handle problems and assumptions
perhaps quite removed from those for which it was initially con-
ceived. Indeed, Levins has argued that even a single problem within
a single field of study is wisely attacked, not with one, but with
several alternative models, each grounded on its own set of simpli-
fications. "Then, if these models, despite their different assump-
tions, lead to similar results we have what we can call a robust
theorem which is relatively free of the details of the model. Hence
our truth is the intersection of independent lies" ( 1966, p. 423).

Levins' conclusion might be taken to infer that what goes on in
one discipline could scarcely have relevance to another. So pessi-
mistic a claim, however, need not be made, for as he himself has ob-
served: "The validation of a model is not that it is 'true' but that it
generates good testable hypotheses relevant to important prob-
lems" (1966, p. 430). In other words, relevance must not be con-
fused with truth. In a sense, even a "false" hypothesis is better than

24 FIELDIANA: ANTHROPOLOGY, VOLUME 68

no hypothesis at all, because hypotheses frame and direct useful
research.

Scott Boorman (1972, p. 393) in a review of MacArthur's Geo-
graphical Ecology, has argued the case for the relevancy of biologi-
cal models to the social sciences. In his opinion, modeling efforts in
contemporary mathematical social science show signs of "incipient
arteriosclerosis." Social scientists would benefit from knowing
more about model-building in ecology and related disciplines today
because these fields, in contrast, "are showing enviable ingenuity
in attacking some of the same theoretical problems that social
science faces." Boorman is cautious in reaching this observation.
He acknowledges that human populations are cultural and can be
more intricately organized than animal populations. Moreoever,
decision-making may affect the behavior of human populations in
unexpected ways. Thus he concludes: "Work in biological ecology
and in population biology as a whole can at most stimulate human
social science; it can neither subsume nor be subsumed by it. "

What Boorman seems to mean, however, by "social science" is
modern economics and the sociology of complex urban societies.
Perhaps biological model-building can be merely provocative and
stimulating in sociology and economics. Yet common sense sug-
gests that anthropologists, archaeologists, and other social scien-
tists working with human populations of lesser size and complexity
than modern urban centers and political states might find more
than enviable ingenuity in the work of theorists and phenomenolo-
gists such as MacArthur, Diamond, Wilson, and Levins. Popula-
tion biologists are not unaware that this likelihood exists. Jared
Diamond (1973, p. 768), for instance, has proposed that the models
developed by MacArthur and Wilson might be useful in studying
human evolution and could help account for striking parallels be-
tween human and bird distributions in New Guinea.

The significance of biological models for social science is not
something to be assumed or rejected out of hand. Rather, their
fitness in social science needs to be determined by setting data
against model and observing what persists. Two research examples
may serve to give substance to this abstract argument. The prob-
lem in each is different. The first explores the question of whether
a fundamental kind of biogeographical patterning commonly ob-
served for species diversity among other animals in the tropics has
any parallel in human diversity in the Solomons: specifically, in

TERRELL: HUMAN BIOGEOGRAPHY 25

linguistic diversity. The second asks whether a few elementary
measures often adopted as simplifications in biological and geo-
graphical model-building, have some predictive rigor, as well, in
modeling interaction patterns among human populations living in
the Solomons archipelago. While these examples are basic and
relatively unsophisticated, the results obtained have proved to be
intriguingly useful.

Area-Diversity Patterns:

A dictionary definition of the word area is "the amount of surface
contained between given limits." Measuring the amount of land sur-
face in some biotic region or sampling locality is clearly important
in determining species population densities and other indices of
abundance. But does the concept of area itself have much theore-
tical value? The answer would have to be affirmative, at least in the
sense that measures of area constitute a useful simplification of
complex reality.

One function of area determinations in theoretical population
biology has been described succinctly by MacArthur and Wilson
(1967, p. 8):

Theories, like islands are often reached by stepping stones. The "species-
area" curves are such stepping stones. Our ultimate theory of species diver-
sity may not mention area, because area seldom exerts a direct effect on a
species' presence. More often area allows a large enough sample of habitats,
which in turn control species occurrence. However, in the absence of good in-
formation on diversity of habitats, we first turn to island areas.

Figure 3a, from Greenslade's (1968b, fig. 2) paper on island pat-
terns in Solomon Islands bird fauna, shows graphically how useful a
measure area can be in biogeography. When the number of land and
freshwater bird species encountered on each of the main islands in
the archipelago is plotted against the log of the land area of each, a
positive correlation between area and species diversity results. A
biogeographer would not be too far off the mark if he tried to predict
the number of species on a given island in the Solomons knowing
only the line of regression and the island's area (Diamond and Mayr,
1976).

This kind of positive relationship between species diversity and
land area is often observed in biogeography. In the words of Mac-
Arthur and Wilson ( 1967, p. 8): "There exists within a given region
of relatively uniform climate an orderly relation between the size of
a sample area and the number of species found in that area." These

• •

r

o

CM

— «M

sawnoNvi do aaawnN

CO CO

-J o>

o o

CO <

sg

< in
UJ m

< o

< oc

— ' Li.

CO
- CO

o <

9S

PQ

QNV1SI U3d SdlOBdS dO U38NDN

a

a

■n

OD

4a

P.

<C

C

h

o,

o

3-

£

■

-C

.fa "w

s

2 o

C B

C8 So

O «

S o

■S b

s s

c E

C e

CB T3

S 5

a> cs

"o i

Q —

J5 ^

"* 5

•-H CO

1 *

II
S J

— v
6 ^

4

26

TERRELL: HUMAN BIOGEOGRAPHY 27

authors have demonstrated that this relationship can be treated
mathematically and may be used to develop a number of productive
models in island biogeography referring to biotic equilibrium and
the dynamics of species immigration and extinction on islands of
differing size and distance from each other or from mainland areas.
Moreover, Simberloff and Wilson (1969, 1970) and Diamond (1969,
1971, 1973) have reported field evidence confirming the general
effectiveness of the MacArthur- Wilson models.

An obvious question, therefore, arises in human biogeography:
does human diversity also correlate in a positive way with island
area in the Solomons? As Figure 3b illustrates, the answer seems to
be "yes." Before examining this relationship more closely, however,
a few preliminary remarks must be considered.

First is the question of taxonomy. In lower animal biogeography,
"number of species" refers to the total count of recognizably dis-
tinct species in a designated faunal set recorded for a given local-
ity. In considering diversity in man, however, "number" can never
refer to different species of men, but only to some taxonomic
measure of observed qualitative or quantitative differences between
populations within a single species, Homo sapiens. This distinction
between species diversity in biogeography and population diversity
in human biogeography is unquestionably important to the biol-
ogist and the evolutionist. Yet the distinction is, in a real sense,
only one of degree. Raising it here merely helps underscore the fun-
damental role played by taxonomic classifications in composing
area-diversity curves in either case.

Second, common sense dictates that the degree of human diver-
sity observed on a given island in the Solomons should vary accord-
ing to the kind of biological or cultural trait examined. This obser-
vation is less important than it might seem to be. Similar variation
also appears in lower animal biogeography, depending on the kinds
of species enumerated. In both human and animal biogeography,
therefore, it will not do to count apples in one area and oranges in
another and then expect an orderly relation to appear. Similarly,
the unit of measure used to determine area cannot be changed at
will without disrupting the comparative results.

Returning to Figure 3b, note that the number of reported lan-
guages spoken on 24 islands in the Solomons have been plotted
against log-island area. The language counts are taken chiefly from
the lists provided by Capell (1962) and Allen and Hurd (n.d.). In

28 FIELDIANA: ANTHROPOLOGY, VOLUME 68

view of earlier comments regarding the limitations of current lin-
guistic evidence, these figures undoubtedly vary greatly in their
accuracy. The island areas are those given by Greenslade.

If these likely sources of error are ignored, what observations can
be made about the area-diversity curve for Solomon Island lan-
guages? Unlike the case in bird faunas, the relation between island
area and the number of languages is not clearly linear. There is a
break in the curve roughly after the log for an area of 100 sq. miles.
Only single languages are spoken on islands below that value.
Thereafter, however, the number of languages seems to increase
rather orderly: one for about every 190 sq. mile increment in island
area.

Forty years ago Ludwik Krzywicki (1934, pp. 20-21) suggested
that primitive man, when he left behind him the stage of nomadic
hunting and gathering and settled down to cultivate the ground, not
only increased in numbers, but also experienced a profound change
in the fabric of his social life: "we should call attention to the fact
that settled life in general at first weakens the intensity of social
cohesion. The ethnic unit is larger: a greater number of persons
speak the same language, which is divided only into dialects differ-
ing but little from each other. But in the very nature of primitive
settling there are inherent factors which counteract the growth of
mutual cohesion."

According to Krzywicki's (1934, pp. 50-51) view of man's econ-
omic and social evolution, primitive agriculturalists, because they
are bound to the soil they cultivate, find themselves living apart in
"very small groups (or communities) — parochial groups, to borrow
a term from mediaeval examples — with similarly parochial views. ' '
While they may have a sense of broad ethnic identity, each settle-
ment leads its own life as an autonomous village community: a
separate, independent social entity, often indifferent to the fate of
its nearest neighbors and sometimes even hostile to them.

Recently and independently, Jon Friedlaender (1971) and I
arrived at conclusions with regard to the village subsistence culti-
vators of Bougainville Island which are in essence those proposed
earlier by Krzywicki. Our field data, in conjunction with previous
ethnographic testimony, indicate in particular that these Solomon
islanders are remarkably sedentary. For example, Friedlaender 's
information on current marriage migration patterns for 18 villages
in central and southern Bougainville reveals that the great majority

TERRELL: HUMAN BIOGEOGRAPHY 29

of married adults (86 per cent) live within only a kilometer or two of
their birthplaces, and very few (5 per cent) live more than 10 km.
distant ( Friedlaender, 1975). Both Friedlaender and I have noted
that a distance on this last order of magnitude well approximates
the average geographical radius of the numerous languages spoken
on the island. Both have inferred, not unexpectedly, that this
linguistic pattern reflects the highly sedentary life of the islanders
and, more fundamentally, their extremely localized ecological
dynamics ( Friedlaender, 1975; Terrell, 1976, in press).

This line of reasoning may be taken to suggest the factors respon-
sible for the language area-diversity curve depicted in Figure 3b for
the Solomons in general. They are: (1) distance and (2) population
size and settlement density. Under traditional socio-economic con-
ditions in the Solomon Islands, it seems that a language population
may attain only a certain maximum spatial dispersion without
differentiating into mutually unintelligible daughter languages.
Judging from the curve in Figure 3b, it may be inferred that dis-
tance begins to affect intelligibility somewhere in the areal range be-
tween 100-200 sq. miles. Once this distance limit is exceeded, the
frequency of interaction between settlements, or to use Krzywicki's
terms the "intensity of social cohesion," reaches the point where a
more or less homogeneous language, or set of mutually intelligible
dialects, can no longer be maintained and language divergence be-
gins to set in.

A transcription into graphical form of these likely factors con-
tributing to linguistic divergence between speech communities is
shown in Figure 4a. The graph is MacArthur's and Wilson's
familiar equilibrium model predicting the number of species in the
biota on a given island as a function of species immigration from a
source region and local species extinction. In the present instance,
however, the rate function originally called "immigration of new
species" has been interpreted as "borrowing of new words" from
some other speech community. Similarly, the increasing function
originally termed "species extinction" is here labelled "loss of
shared words." The intelligibility between two communities in
speech behavior, expressed as the number of shared or "cognate"
words used in common, is illustrated as the equilibrium value (c) at
the intersection of these two functions.

This simple representation takes into account neither variation in
the effect of distance on social interaction among communities, nor
variation in the modes of interaction employed. Moreover, this

30 FIELDIANA: ANTHROPOLOGY, VOLUME 68

figure does not reflect the likelihood that population size, particular-
ly in terms of the number of settlements for a given land area, may
affect the cohesiveness of a language population. A more realistic
model, therefore, would be one offering a range of intelligibilities
or cognate percentages, depending on varying kinds of interactive
mobility and alternative settlement densities. Figure 4b presents
such a model, one which is also a direct borrowing from MacArthur
and Wilson (1963).

The example of one Solomon Islands language which has an un-
commonly wide area of occurrence will illustrate how distance and
settlement density influence the territorial dimensions of a lan-
guage population. The largest single language group on Buka
Island and the northern end of Bougainville is the Halia language
(some 10,000 speakers). Unlike the restricted spatial pattern ob-
served for the other languages spoken on Buka and Bougainville,
Halia is spoken along a narrow strip of coast on both islands, for a
distance of some 45 miles. Yet, in spite of the fact that Halia has
apparently been in use for several thousands of years in its present
location, it has only differentiated into a chain of four dialects which
are less like each other the farther apart they are along the coast:

Haku Dialect (N.W. Buka)
78% Halia Dialect ( N.E. Buka)

73% 84% Hangan Dialect ( S.E . Buka )

57% 75% 79% Silao "Sub-language" Dialect (N.E.

Bougainville Island)

[data: similarity figures from Allen and Hurd, n.d. ]

Comparative study of the geography of all the languages of Bou-
gainville and Buka (Terrell, 1976, chap. 2, appendix II) suggests
that several specific circumstances have maintained Halia as a
chain of mutually-intelligible dialects, rather than as a set of histori-
cally-related languages. These local conditions are: ( 1 ) the coastal
location of the language communities and the reported ease of travel
by sea and over the level coastal terrain; (2) the high density of
population (the highest on these islands: ca. 160 people per square
mile); and (3) the corresponding high number of village settlements
(over 30 at the present time) situated in close proximity to each
other along the coast. Elsewhere on Buka and Bougainville such a
combination of circumstances does not pertain. In spite of sizable
populations in some instances, the maximum distance between the

RATE

B

FJG.4 «l Ax

R OF COGNATE

« kl

N
d

n

32 FIELDIANA: ANTHROPOLOGY, VOLUME 68

Table 1

Island

Area 4

Expected

Estimated

Deviation

(

sq. miles)

Languages

Languages

Rendova

147

1.55

2

0.45

Vanugu

210

1.88

3

1.12

San Cristobal

1193

7.00

6

1.00

Choiseul'

1145

6.75

6

0.75

Isabel

1581

9.02

8

1.02

Malaita 2

1663

9.45

8

1.45

New Georgia

789

4.89

7

2.11

Guadalcanal

2039

11.40

14

2.60

Bougainville 3

3317

18.06

18

0.06

'Scheffler, 1965

2 Ross, 1973

3 Allen and Hurd, n.d

"Diamond and Mayr,

1976

two most separated villages in each of the remaining languages
averages only 20 miles.

The exceptional status of Halia, therefore, may be cited not only
to define the likely manner in which particular local circumstances
affect the geographic range of a language in the Solomons, but also
to propose why it should still be possible to predict rather closely
(table 1) the number of languages spoken on a given island from
knowledge only of the area of the island in question and the cal-
culated line of regression for the archipelago as a whole (fig. 5).
Apparently the ecology of human settlement in the Solomons is
sufficiently commensurate from island to island, and geographic
variations are sufficiently comparable, regardless of the island, that
an orderly relation between language diversity and island area has,
nonetheless, been achieved.

While the area-diversity curve recognized for the Solomon Islands
may be altered with the acquisition of more reliable data on the lan-
guages of the region and island areas, it would be surprising if the
broad pattern discerned were wholly discredited, in view of the size
of the island sample and the high correlation value obtained (r =
0.9787). '

The area-diversity pattern is not trivial. It affirms that human
biogeographical research in the Solomons will not be fruitless. It

'Accepting the data given by Hackman (1968), a similar correlation is obtained
(N = 32, r = 0.9661, m = 0.0059, b = 0.8540).

TERRELL. HUMAN BIOGEOGRAPHY

33

1000 2000

AREA (SQ. MILES)

Fie. 5. Area-language curve for the Solomon Islands.

shows a correspondence between human and lower animal popula-
tions few might have anticipated. Equally important, the area-
diversity curve raises significant questions about linguistic evolu-
tion and the geographic behavior of human populations practicing
subsistence cultivation, questions which, there is reason to believe,
demand closer attention than has been customary in Pacific Island

34 FIELDIANA: ANTHROPOLOGY, VOLUME 68

historical linguistics. Even granting probable differences in the
time-depth of first human settlement in Melanesia and Polynesia,
for example, it may not be as remarkable as often thought that the
large islands of Melanesia are more linguistically diverse than the
small island groups in Polynesia.

Proximal-Point Analysis:

The second example of current efforts to develop predictive mod-
els defining the probable behavior of human populations in the Solo-
mon Islands, utilizing principles already explored in biogeography
and here also modern locational geography, is illustrated in Figure
6. The network graph shown superimposed over a conventional map
of the Solomons depicts the most likely directions of inter-island
movement if only two variables, distance and area, are assumed to
govern frequency of interaction. The procedures followed to con-
struct this network model can be simply described.

Perhaps the best known mathematical model in geography is the
"energy of interchange" or "gravity" model, associated closely with
the work of John Stewart, an astronomer, and G. K. Zipf, pioneers
of contemporary quantitative geography (Stewart, 1947; Stewart
and Warntz, 1968). As Gerald Carrothers (1969, p. 226) has de-
scribed this model:

In general terms, the gravity concept of human interaction postulates that
an attracting force of interaction between two areas of human activity is
created by the population masses of the two areas, and a friction against
interaction is caused by the intervening space over which the interaction
must take place. That is, interaction between two centres of population con-
centration varies directly with some function of the population size of the two
centres and inversely with some function of the distance between them.

Phrased in the most elementary mathematical fashion the fre-
quency of interchange or interaction {la) between two populations
located at points (i) and (/) may be predicted from the size of each
population {P, Pj) and the distance separating them (dy), raised to
an exponent (6) to allow for the likelihood that the effects of dis-
tance will vary from one kind of situation to another and with the
methods of interaction and movement employed ( Harvey, 1969, pp.
559-560):

d\j

Carrothers (1969, pp. 236-237) has remarked upon the major dif-
ference between this model and that expounded in Newtonian phys-

c

J3

-1 Jg

&

t

-i s.

e

a

CO

c

« c

CO '2
M

35

36 FIELDIANA: ANTHROPOLOGY, VOLUME 68

ics. Neither formulation purports to forecast the behavior of indi-
viduals, be they molecules or persons. Both only attempt predic-
tions about large, aggregate numbers of individuals. However,
people make decisions regarding their behavior; molecules do not.
There must be, therefore, a threshold "where the power of individual
decision-making critically affects the results" obtained by the geo-
graphic analogue. While this limitation may be significant (Sack,
1971), one way around it is to use the model and then judge the
effectiveness of the results. This pragmatic approach is taken here.

MacArthur and Wilson (1967, pp. 19-41), in their theoretical
investigations on island species diversity, make use of variables
akin to those contained in the gravity model. In their usage, how-
ever, island area assumes the kind of role played by population size
in the latter. Their aim is to develop an equilibrium model predict-
ing, not population interaction, but rather the dynamics of species
immigration and extinction. In spite of these differences, the paral-
lels between their approach and interaction models are brought out,
for example, when they comment that their model "leads to the
prediction that the logarithm of the number of species should in-
crease with area more rapidly on distant islands than on near
islands, and should decrease more rapidly with distance on small
islands than on large islands" (MacArthur and Wilson, 1967, p. 66).

Possibly the main difference between the basic equilibrium model
created by MacArthur and Wilson (which, as noted earlier, defines
the number of species on an island as a balance between the rate of
immigration of new species to the island against the rate of extinc-
tion of species from the island) and the gravity model used in geog-
raphy lies in the fact that the equilibrium model is less concerned
with actual frequencies or rates of extinction and immigration
than with the relation or ratio between these two processes. As it
will soon become apparent, the island model described here for the
Solomons is also one dealing more with ratios than with calculated
interaction frequencies, although the assumptions on which it is
grounded are those common to both the MacArthur- Wilson and
gravity models.

The elementary network graph appearing in Figure 6 represents
one attempt to see what kinds of interaction patterns should be
expected among the Solomon Islands if measures as fundamental as
distance and area strongly predict the spatial movements of the
islanders. The procedures followed to sketch the network shown

TERRELL: HUMAN BIOGEOGRAPHY 37

require nothing more elaborate than a pair-of-compasses, a ruler,
and a pencil.

In constructing the model it was necessary, first of all, to simplify
the complexities recorded on geographic maps of the archipelago.
Intuitively, the decision was made to let dots, or single points,
represent small islands. Such a reduction will hardly do in the case
of large islands, however, where it would lead to the loss of all di-
mensional information. Because the main islands in the Solomons
are remarkably elongate, due to their underlying physical geology,
large islands were reduced in the figure to more or less straight
lines, defined by three points: a mid-point and two end-points at the
farthest extremes roughly located on a line through the mid-point.

After these points had been determined for all islands, three lines
were drawn out from each to the three other points found to be
closest or most "proximal" to each point. This geometric procedure
graphically expresses the assumption that, since these points are
closest, they are also the ones most likely to be contacted from each
point of origin, if all of them were in truth human settlements. The
final network of points and lines depicted in Figure 6 portrays sche-
matically the most probable directions of inter-island trade, travel,
inter-marriage, migration, and the like, provided that measures of
area and distance accurately predict the most probable paths of
interaction, and also that the simplifications introduced have not
critically distorted either geographic reality or those measures. One
further assumption is hidden in the network which must be speci-
fied if the diagram is to be interpreted correctly.

The procedure of drawing three lines out from every single point
reflects the assumption that village settlements in the Solomons all
enjoy an equal status and that none performs centralized, special-
ized functions for its surrounding neighbors. For the most part, this
assumption seems reasonable as a characterization of the egalitar-
ian societies found in Melanesia at the time of first European con-
tact. In most areas of the world today, even in Melanesia, it would
be necessary to weight the points in Figure 6 according to some
formula predicting the position that a given hypothetical settlement
is likely to hold in a non-equivalent, hierarchically ordered network
system of social, religious, political, and economic interactions:
predictions of the sort attempted in modern locational geography
by Central Place theory and the like ( Haggett, 1966, pp. 114-152).

38 FIELDIANA: ANTHROPOLOGY, VOLUME 68

If, nonetheless, the assumption is accepted for the Solomons that
all points in the network are equivalent, then the model may be read
directly to reach predictions concerning: ( 1 ) the relative frequency
or ratio of interaction among selected points in the hypothetical
system ( by comparing the lengths of the lines drawn and consider-
ing the distance exponent to be invariant), (2) the efficiency of inter-
action among any set of points in the system (by constructing dis-
tance matrices, see: Harary et al., 1965), (3) the degree of connectiv-
ity in the system taken in whole or in part ( Haggett, 1966, pp. 70-72,
238-239), and so on.

In short, the network model leads to hypotheses about, first, how
island interaction is patterned in the archipelago, and second, the
kinds of pattern variation to be expected in any cultural or biologi-
cal trait of the islanders believed sensitive to change in interaction
frequencies. For example, since languages are usually thought to
diverge under conditions of isolation, the model may be interpreted
to suggest that the northern Solomons (Buka and Bougainville) and
the western Solomons (Vella Lavella-New Georgia) are probably
dissimilar, linguistically, from the highly interconnected southeast-
ern Solomons: a prediction in apparent agreement with existing
language evidence noted earlier.

As this one example illustrates, hypotheses from the model can be
tested. Predictions about island interaction patterns may be as-
sessed by determining what actually happens during trade and
travel in the Solomons through field observation and study of the
ethnohistorical literature. Predictions about pattern variation can
be tested by phenomenological study of trait diversity.

Tests of both sorts are in progress at Field Museum. For instance,
detailed ethnohistorical and archaeological information has been
gathered on trading systems and interaction spheres in the northern
Solomons: on the one hand between Buka, Nissan, and New Ireland
(Kaplan, 1976) and on the other, between southern Bougainville, the
Mono-Shortland Islands, and Choiseul (Terrell and Irwin, 1972). In
both areas, the correspondence between the network model and
attested reality is, as Kaplan has previously remarked, strikingly
close. Similarly, what Cummings (1973) has written about geo-
graphic variation in compound fish-hook types throughout the
Solomons matches well with what one would predict from the
model. Further, considering what Roger Green (1973) has written
about trade and cultural affinities in the southeastern Solomons, it
is likely that the model may hold for that region as well.

TERRELL: HUMAN BIOGEOGRAPHY 39

If proximal-point analysis proves to be as predictive as available
evidence suggests, there is good reason to suspect that much of the
bewildering array of diversity in the Solomon Islands may be explic-
able with the aid of a very few basic propositions about the spatial
organization of interactions among the island populations. The
generation of diversity through environmental adaptation may be
of limited importance: a conclusion which might be expected consid-
ering how little the environment seems to vary from one part of the
island chain to the other.

However accurate the network model may ultimately appear to
be, it has one unquestionable virtue: the hypotheses it generates
orient research in explicit directions. In fact, it will be the excep-
tions to the model that will stand out with great clarity as prime
areas for study.

In summary, modeling procedures, such as elementary proximal-
point analysis, hold out promise of being able to resolve problems of
long-standing interest referring to the probable causes of the un-
common degree of biological and cultural diversity found among the
Solomon Islanders. In the future it would be possible, of course, to
generate with the aid of computer techniques more complex models
than that presented in Figure 6. Thus far, no occasion has arisen
when it seemed necessary to abandon pencil and paper and do away
with the simplifying steps required by the manual technique out-
lined above. It is rare, however, that old puzzles are solved without
revealing new problems previously obscured by the old ones. Need-
less to say, new problems usually demand new techniques for their
solution.

Retrospect and Prospects

Human biogeography has been defined as the study of human
populations, their characteristics, organization, and behavior, and
of the conditions and kinds of events that lead to the growth and
maintenance of similarities and differences among human popula-
tions at various points on the earth's surface and at various times in
man's history. What distinguishes human biogeography from other
related, older disciplines, such as social anthropology and cultural
geography, is at least twofold. Biological populations, rather than
geographical regions or the cultures and societies of the anthropolo-
gist, are considered the fundamental unit of interpretation. Re-
search is focused on discovering how human populations are distrib-

40 FIELDIANA: ANTHROPOLOGY, VOLUME 68

uted over the face of the earth, how they are affected by their size,
growth, geographic location, and habitat, and how their relations,
or interactions, with the natural world and with other human
populations are influenced and structured by the conditions under
which they live.

The argument has been made that evidence suitable for the study
of the biogeography of man may be found not only by analyzing the
distributions of human similarities and differences, but also through
comparative research on other animal species. Indeed, it has been
suggested that contributions by ecologists, theoretical biologists,
and biogeographers should prove extremely helpful in developing
suitable procedures, concepts, and models to describe and explain
the characteristics and behavior of human groups.

In support of these propositions, two examples of research on the
human biogeography of the Solomon Islands in the southwest Pacif-
ic were described. In the first, it was observed that there appears to
be a positive correlation between island area and the number of
languages spoken on an island, a relation so direct, in fact, that it is
possible to predict the number of languages on any given island in
the Solomons knowing only its land area and the calculated regres-
sion for the archipelago as a whole. This relation closely parallels
similar ones reported for other animal species in the Solomons, and
reveals a correspondence between human and animal populations
few anthropologists might have anticipated.

In the second example, two fundamental variables, area and dis-
tance, often used to develop spatial models in economic geography
and theoretical island biogeography, were adopted to construct a
simple network model predicting patterns of interaction among the
islands in the Solomons chain. Following a procedure developed
here called "proximal-point analysis," it was shown that the result-
ing model successfully anticipated patterns of trade, inter-island
voyaging and cultural diversity.

Although much information on human diversity in the Solomons
is needed before great advances can be made in the biogeography of
the islanders, results obtained thus far hold out promise that work
along the lines sketched in this survey will be practical and illumi-
nating. It is likely that comparative research on the ecology and
biogeography of both human and other animal populations will
resolve one of the great riddles of the Pacific: the extraordinary
biological and cultural diversity of the island world of Melanesia.

terrell: human biogeography 41

Acknowledgements

I thank Jared Diamond, Susan Kaplan, Joel Fagan, Marty Zele-
nietz, and Bill Fitzhugh for their help and encouragement. I thank
also Marilyn Miller for her editorial labors on the original mimeo-
graphed manuscript entitled "Comparative Study of Human and
Lower Animal Biogeography in the Solomon Islands" (1974, Solo-
mon Island Studies in Human Biogeography, no. 3) which was pre-
pared for the Smithsonian Conference on the Application of Models
in Theoretical Biology and Biogeography to Archaeology and
Anthropology (Washington, 1974). That inter-disciplinary confer-
ence was organized by Bill Fitzhugh and I and was supported by
grants from the Wenner-Gren Foundation for Anthropological
Research, Inc. and the Smithsonian Institution, which are grate-
fully acknowledged.

REFERENCES

Allen, Jerry and Conrad Hurd
n.d. Languages of the Bougainville District. Summer Institute of Linguistics,
Papua and New Guinea.

Ambrose, Peter
1969. Analytical human geography. American Elsevier Publishing Co., N.Y.

An ell, Bengt
1955. Contribution the the history of fishing in the Southern Seas. Stud. Eth-
nogr. Upsaliensia, IX.

Barrows, Harlan H.
1923. Geography as human ecology. Ann. Assn. Amer. Geogr., 13, pp. 1-14.

Bates, Marston
1953. Human ecology, pp. 700-713. In: Kroeber. A. L. ,ed., Anthropology today.
University of Chicago Press, Chicago.

Berry, Brian J. L. and Duane F. Marble, eds.

1968. Spatial analysis. Prentice-Hall, Englewood Cliffs. N.J.
Blackwood, Beatrice

1931. Report on fieldwork in Buka and Bougainville. Oceania, 2. pp. 199-219.

Boorman, Scott A.
1972. Review of: Robert H. Mac Arthur, Geographical ecology. Science, 178. pp.
391-394.

Booth, P. B. and A. P. Vines

1969. Nuguria atolls (Fead Islands), Bougainville District, New Guinea: blood
group and other genetic data. Archaeol. Phys. Anthropol. Oceania, 4, pp. 120-
122.

42 FIELDIANA: ANTHROPOLOGY, VOLUME 68

Burton, Ian

1968. The quantitative revolution and theoretical geography, pp. 13-23. In: Berry,
Brian J. L. and Duane F. Marble, eds., Spatial analysis, Prentice-Hall, Engle-
wood Cliffs, N.J.

Capell, A.
1962. A linguistic survey of the south-western Pacific. South Pacific Commission,
Noumea, New Caledonia, Technical Paper no. 136.

1969. A survey of New Guinea languages. Sydney University Press, Sydney.
Carrothers, Gerald A. P.

1969. An historical review of the gravity and potential concepts of human interac-
tion, pp. 226-242. In: Ambrose, Peter, ed., Analytical human geography,
American Elsevier Publishing Co., N.Y.

Chikamori, Masashi

1967. Preliminary report on the archaeological researches in western Solomon
Islands. In: Chikamori, Masashi and Seiji Itok, Eiryo Solomon Shoto ni Okeru
Kokogakuteki, Minzokugakuteki Choza Ryakuho [Preliminary Reports on
Archaeological and Ethnological Fieldwork in British Western Solomon
Islands]. Published privately. Printed by Kaimeido, Tokyo.

Chorley, Richard J. and Peter Haggett

1967. Models in geography. Methuen, London.

1969. Physical and information models in geography (Parts I, II, and V of Models
in geography). Methuen, London.

CODRINGTON, ROBERT H.

1891. The Melanesians. Clarendon Press, Oxford.
Cranstone, B. A. L.

1961. Melanesia, a short ethnography. Trustees of the British Museum, London.
Cummings, Henry

1973. The distribution of compound fishhook types as a gauge of population
interaction in the Solomon Islands. Solomon Island Stud. Hum. Biogeogr.
( Mineographed, Field Museum of Natural History, Chicago), 1.

Damon, Albert, ed.
1973. Some genetic traits in Solomon Island populations. Amer. Jour. Phys.
Anthropol., 39, pp. 169-203.

Dansereau, Pierre
1957. Biogeography. An ecological perspective. The Ronald Press Co., N.Y.

Davenport, William H.

1968. Anthropology in the British Solomon Islands. Expedition, 11, pp. 31-34.
1972. Preliminary excavations on Santa Ana Island, Eastern Solomon Islands.

Archaeol. Phys. Anthropol. Oceania, 7, pp. 165-183.

Diamond, Jared M.

1969. Avifaunal equilibria and species turnover rates on the Channel Islands of
California. Proc. Nat. Acad. Sci., 64, pp. 57-63.

TERRELL: HUMAN BIOGEOGRAPHY 43

1971. Birds of the Karimui Basin, New Guinea. Nat. Geogr. Soc. Res. Repts..

1965 Projects, pp. 69-74.
1973. Distributional ecology of New Guinea birds. Science, 1 79. pp. 759-769.
Diamond. Jared M. and Ernst Mayr
1976. Species-area relation for birds of the Solomon Archipelago. Proc. Nat. Acad.
Sci.,73, pp. 262-266.

Firth. Raymond, ed.

1944. Pacific Islands. Naval Intelligence Division of the Admiralty. Great Britain.
Geographical Handbook Series, III.

Friedlaender. Jonathan S.

1971. Isolation by distance in Bougainville. Proc. Nat. Acad. Sci. 68. pp. 704-707.
1975. Patterns of human variation: the demography, genetics and phenetics of
Bougainville islanders. Harvard University Press. Cambridge, Mass.

Grace. George W.
1971. Languages of the New Hebrides and Solomon Islands, pp. 341-358. In:
Sebeok, Thomas A. ed., Current trends in linguistics. 8.

Green. Roger C.

1973. Introduction to: Sydney M. Mead, Material culture and art in tin- St;ir
Harbour region. Eastern Solomon Islands. Royal Ontario Museum: Ethno-
graphy Monograph, 1.

Greenberg. Joseph H.

1971. The Indo-pacific hypothesis, pp. 807-871. In: Sebeok. Thomas A., ed.. Cur-
rent trends in linguistics, 8.

Greenslade. P. J. M.
1968a. The distribution of some insects of the Solomon Islands. Proc. Linnean

Soc. London, 179, pp. 189-196.
1968b. Island patterns in the Solomon Islands bird fauna. Evolution, 22, pp. 751-

761.
1969. Land fauna: insect distribution patterns in the Solomon Islands. Phil.

Trans. Royal Soc.. B255. pp. 271-284.

Guppy. Henry B.
1887. The Solomon Islands and their natives. Swan Sonnenschein. Lowry.
London.

Hackman. B. D.
1968. A guide to the spelling and pronunciation of place names in the British
Solomon Islands Protectorate. Government Printing Office. Honiara. B.S.I. P.

Haggett. Peter
1966. Locational analysis in human geography. St. Martin's Press. N.Y.

Harary. Frank. Robert J. Norman, and Dorwin Cartw right
1965. Structural models: an introduction to the theory of directed graphs. John
Wiley and Sons, N.Y.

44 FIELDIANA: ANTHROPOLOGY, VOLUME 68

Harvey, D.
1969. Models of the evolution of spatial patterns in human geography, pp. 549-
608. In: Chorley, Richard J. and Peter Haggett, eds., Integrated models in
geography. Methuen, London.

Howells, William

1959. Mankind in the making. Doubleday and Co., Garden City, N.Y.
Irwin, Geoffrey J.

1972. An archaeological survey in the Shortland Islands: B.S.I. P. (Unpublished
M.A. thesis, University of Auckland).

1973. Man-land relationships in Melanesia: an investigation of prehistoric settle-
ment in the islands of the Bougainville Strait. Archaeol. Phys. Anthropol.
Oceania, 8, pp. 226-252.

Kaplan, Susan
1976. Ethnological and biogeographical significance of pottery sherds from Nis-
san Island, Papua New Guinea. Fieldiana: Anthropol., 66, no. 3.

Krzywicki, Ludwik
1934. Primitive society and its vital statistics. Macmillan, London.

Leach, E. R.

1965. Political systems of highland Burma. Beacon Press, Boston.

Levins, Richard

1966. The strategy of model building in population biology. Amer. Sci., 54, pp.
421-431.

MacArthur, Robert H.
1972. Geographical ecology. Harper and Row, N.Y.

MacArthur, Robert H. and Edward O. Wilson
1963. An equilibrium theory of insular zoogeography. Evolution, 17, pp. 373-387.

1967. The theory of island biogeography. Princeton University Press, Princeton,
N.J.

Mann, William M.
1948. Ant hill odyssey. Little, Brown and Co., Boston.

Mayr, Ernst

1930-1931. Worter der Nissan-sprache. Z. Eingeborenen Sprachen, 21, pp. 252-

256.
1943. A journey to the Solomons. Nat. Hist. 52, pp. 30-36, 48.

1969. Bird speciation in the tropics. Biol. Jour. Linnean Soc, 1, pp. 1-17.

1970. Populations, species, and evolution. Belknap Press of Harvard University
Press, Cambridge, Mass.

Mikesell, Marvin W.
1967. Geographic perspectives in anthropology. Ann. Assn. Amer. Geogr., 57, pp.
617-634.

TERRELL: HUMAN BIOGEOGRAPHY 45

Oliver. Douglas L.

1949. The Peabody Museum expedition to Bougainville, Solomon Islands, 1938-

39. Papers Peabody Mus. Amer. Archaeol. Ethnol., Harvard Univ., 29. no. 1.
1962. The Pacific Islands. Harvard University Press, Cambridge, Mass.

Pawley. Andrew
1972. On the internal relationships of Eastern Oceanic languages, pp. 1-160. In:
Green, R. C. and M. Kelly, eds., Studies in Oceanic culture history, 3; Pac.
Anthropol. Rec., 13.

Poulsen. J. I.

1972. Outlier archaeology: Bellona. A preliminary report on field work and radio-
carbon dates. Archaeol. Phys. Anthropol. Oceania, 7, pp. 184-214.

Rivers. W. H. R.
1914. The history of Melanesian society, 2 vols. Cambridge University Press,
Cambridge.

Rohi.f. James F. and Gary D. Schnell
1971. An investigation of the isolation-by-distance model. Amer. Nat., 105, pp.
295-324.

Ross. Harold M.

1973. Baegu: social and ecological organization in Malaita, Solomon Islands.
111. Stud. Anthropol., 8.

Sack. Robert D.
1971. Geographic location and human behavior. Unpublished Ph.D. thesis,
Univ. of Minnesota.

Sahlins. Marshall D. and Elman R. Service, eds.
1960. Evolution and culture. University of Michigan Press, Ann Arbor.

SCHEFFLER. HAROLD W.

1965. Choiseul Island social structure. University of California Press, Berkeley.
Simberloff. Daniel S. and Edward O. Wilson

1969. Experimental zoogeography of islands: the colonization of empty islands.
Ecology, 50, pp. 278-296.

1970. Experimental zoogeography of islands: a two-year record of colonization.
Ecology. 51. pp. 934-937.

Simmons. R. T. and D. C. Gajdusek

1966. A blood group genetic survey of children of Bellona and Rennell islands
(B.S.I. P. I and certain northern New Hebridean islands. Archaeol. Phys.
Anthropol. Oceania, 1, pp. 155-174.

Specht. Jim R.
1969. Prehistoric and modern pottery industries of Buka Island, T.P.N. G. Un-
published Ph.D. thesis, Australian National Univ.

Speiser. Felix

1942a. Ueber schutzwaffen in Melanesien. Int. Arch. Ethnogr.. 40, pp. 81-117.

46 FIELDIANA: ANTHROPOLOGY, VOLUME 68

1942b. Ueber totenbestattungen in insel Melanesien. Int. Arch. Ethnogr., 40,
pp. 125-174.

Steinberg, Arthur G., Albert Damon, and Jerry Bloom
1972. Gammaglobulin allotypes of Melanesians from Malaita and Bougainville,
Solomon Islands. Amer. Jour. Phys. Anthropol., 36, pp. 77-84.

Stewart, John Q.
1947. Empirical mathematical rules concerning the distribution and equilibrium of
population. Geogr. Rev., 37, pp. 461-485.

Stewart, John Q. and William Warntz
1968. Physics of population distribution, pp. 130-146. In: Berry, Brain J. L. and
Duane F. Marble, eds., Spatial analysis, Prentice-Hall, Englewood Cliffs, N.J.

Terrell, John E.

1976. Perspectives on the prehistory of Bougainville Island, Papua New Guinea:
a study in the human biogeography of the southwestern Pacific. Unpublished
Ph.D. dissertation, Harvard University.
In press. Geographic systems and human diversity in the North Solomons. World
Archaeology, 9, no. 1.
Terrell, John E. and G. J. Irwin

1972. History and tradition in the northern Solomons: an analytical study of the
Torau migration to southern Bougainville in the 1860's. Jour. Polynesian Soc.,
81, pp. 317-349.

Todd, Evelyn M.

1973. Savosavo and the Solomon language family. Dept. of Anthropology, Trent
University, Peterborough, Ontario, (mimeographed).

Vayda, Andrew P.

1966. Diversity and uniformity in New Guinea. Acta Ethnogr. Acad. Sci. Hung.,
15, pp. 293-300.

von Bertalanffy, Ludwig

1968. General system theory. George Braziller, N.Y.

White, Leslie

1960. Introduction to: Sahlins, Marshall D. and Elman R. Service, eds., Evolution
and culture. University of Michigan Press, Ann Arbor.

Willis, M. F. and P. B. Booth
1968. Takuu and Nukumanu atolls, Bougainville District, Territory of Papua and
New Guinea. Archaeol. Phys. Anthropol. Oceania, 3, pp. 55-63.

Wilson, Edward O.

1961. The nature of the taxon cycle in the Melanesian ant fauna. Amer. Nat., 95,
pp. 169-193.

Wurm, Stephen A.

1967. Linguistics and the prehistory of the south-western Pacific. Jour. Pac. Hist.
2, pp. 25-38.

TERRELL: HUMAN BIOGEOGRAPHY 47

1971. The Papuan linguistic situation, pp. 541-657. In: Sebeok, Thomas A. ed..
Current trends in linguistics, 8.

1973. The emerging linguistic picture of the south-western Pacific — and linguis-
tic prehistory. (Xerographed manuscript No. 1589) IXth Int. Congr. Anthropol.
Ethnol. Sci., Chicago. U.S.A.

Zimmerman. ElwoodC.
1942. Distribution and origin of some eastern Oceanic insects. Amer. Nat.. 76.
pp. 280-307.