SOME OF ROWE'S CONTRIBUTIONS IN THE FIELD OF ANDEAN CULTURE HISTORY

Dorothy Menzel

University of California, Berkeley

John H. Rowe has brought fundamental changes to the field of
Andean archaeology during the last twenty years. Since all of these

changes involve new or different ideas concerning organization, methods

of classification, theories of human behavior, the nature of justifiable
inference and fruitful scholarly procedure, most of his contributions

are as pertinent to all archaeological studies as they are to the Andean
field. Many of his innovations are the subjects of considerable contro-
versy because they run counter to some old established theories and
some new methods.

By 1950 Rowe had come to the conclusion that many of the gen-

eralizations concerning Andean prehistory were not supported by existing
evidence. This became increasingly obvious as more data were coming in.
It was evident to Rowe that most of the assumptions on which the gener-
alizations were based could not be taken for granted and needed to be

questioned, or at least justified. Even the chronological framework in
terms of which Andean archaeology was discussed suffered from implicit

unquestioned assumptions. One of the most important services performed
by Rowe has been to simplify the chronological framework by ridding it

of the unnecessary and unproved preconceptions and so changing it into a

viable tool for new discoveries and all kinds of inferences. He developed
his ideas concerning such a relative chronology in three publications
("Cultural unity and diversification in Peruvian archaeology," 1960;

"Tiempo, estilo y proceso cultural en la arqueologfa peruana," 1960; and
"Stages and periods in archaeological interpretation," 1962, 1967).

The essence of Rowe's revision is to make is possible to talk

about contemporaneity and relative time differences of different cultures

100

or culture phases without having to confuse these discussions with infer-

ences concerning cultural development. The widely used system of relative
chronology in Andean archaeology since the 1940's has been to attach to
the archaeological divisions purporting to deal with relative time the

anticipated conclusion that these periods represented stages of cultural
progress which were equally applicable to all areas. Rowe argued that

such a conclusion could not be assumed at the outset, but would have to

be proved on the basis of independent evidence.   By combining the conclu-
sion with the chronological framework it was impossible to reach any
other conclusion, or to formulate alternative 'hypotheses.

It will be apparent from the foregoing that Rowe's suggestion
for a different chronological framework ultimately involves fundamental
problems concerning theories of cultural development, scientific logic

and the uses of evidence. The far reaching theoretical implications of

the nature of the chronological framework used is expressed most clearly

in Rowe's article of 1962 on "Stages and periods in archaeological inter-
pretation." In it he pointed out that as soon as the chronological

framework is confined simply to time and the evidence is considered on
itx merits, it becomes possible to see all kinds of data leading to al-
ternative hypotheses concerning Andean cultural development.

Rowe pointed out that the idea of a simple sequence system was
not new.  However, he has added a significant innovation by suggesting

that the chronology should be based on a single master sequence consisting
of a well controlled archaeological chronology in one small area, in this

case the coastal valley of Ica, Peru. This suggestion was made to produce
greater precision in crossdating and in discussion concerning relative

time over large areas. The use of a master sequence forces the archaeol-
ogist to be very precise in statements concerning the contemporaneity or
sequence of different cultures, culture phases and cultural events, and
to justify these statements by means of independent evidence.

101

An interest in reaching much greater precision in discussions

of relative time is another one of Rowe's contributions that has exerted
great influence on Andean studies. Lack of precision concerning contem-
poraneity or sequence of cultural events makes it impossible to make any
but the most general inferences concerning culture change, which usually
lead to mistaken conclusions and faulty generalizations. This lack of

precision in crossdating is one of the factors that has made it possible
for archaeologists to persist in clinging to a chronological scheme with
built-in conclusions.

Rowe's search for greater precision also has led him to seek

better methods for developing shorter and more precisely defined periods
of relative time. His training as a historian has made him particularly
appreciative of the fact that historical processes cannot be understood

if the precise order and points of origin in which particular events take
place are not known. Although the ultimate goal of absolute synchroniza-

tion can perhaps never be reached, Rowe has been instrumental in developing
methods that have succeeded in reducing archaeological time intervals in
which changes can be perceived from something on the order of 200 to 1000
years to as little as 25 to 50 years. It is particularly significant

that this refinement has been achieved through the use of archaeological
evidence alone, without the aid of absolute written dates such as might
be found inscribed on buildings, coins or in documents, or of written
records of any kind. As Hammel points out, Rowe has achieved this in-
creased accuracy by employing methods of style study and the serial

arrangement of stylistic changes resembling some of the techniques used
in the study of classic Greek pottery and the techniques of seriation

first proposed by John Evans in 1849. The further development of these
techniques has led to a great refinement in reliability and precision of

dating in the Andean area. The method is capable of verification through
independent evidence, and is applicable to most aspects of culture, in-
cluding most material remains. Its effect on the possibilities of

102

inference concerning culture history and cultural processes cannot be

overestimated. Rowe has outlined the method in his article on "Stratig-
raphy and s-eriation" (1961).

It appears to me that the differences between the kind of
stylistic classification proposed by Rowe, and others in use among

archaeologists, are not generally very clearly understood. Since these
differences are important, I should like to make an attempt here at ex-
pressing them.

The most popular method currently in use among American archeo-
ologists is the classification by "types," usually applied to pottery

fragments. In this method, certain features that are observed to occur

commonly together on fragments of pottery are considered to form a "type.
The types are immutable, and make a classificatory "weld" of several in-

dependent features. This system is comparable, let us say, to the classi
fication of human "races" in physical anthropology commonly used some 20
to 30 years ago. Types thus become rigid, unwieldy classificatory units
which more often obscure meaningful stylistic patterns than they reveal
them. A factor adding to the difficulties in this taxonomic system is
that the descriptive definition of the type characteristics also some-

times leads to semantic transferences, in which, for example, all frag-
ments with "black and white" decoration or other simple classificatory
elements have been lumped regardless of other differences among them.

Statistical counting of such types usually follows their definition, and
inferences concerning chronological contrasts are based on relative per-
centages rather than on the stylistic evidence directly.

The classificatory system proposed by Rowe is a much more

flexible one in which smaller independent units, i.e., "features," form

the basic elements. Features can only be defined by inductive reasoning,
and there can be no universal rule about what constitutes a useful

feature for classificatory purposes. The co-occurrence of features on
the same object is considered as a separate problem reflecting patterns

103

and compositions created by the makers of the objects at any given time.
The features are thus in some ways comparable to sound combinations in

a language, of the kind observed particularly in the study of changes in
linguistic traditions. As in the corresponding linguistic studies,

features are especially useful in tracing stylistic continuity and change
over spans of time. The observation of the combination of features into
stylistic compositions and themes by the creators of the objects at any
given point of time, on the other hand, is in some ways comparable to

the discovery of grammatical patterns and meaning in a particular lan-

guage, so in the study of style it is necessary to discover through in-
ductive reasoning and observation the patterns of features that were

meaningful to the creators of the objects. Rowe frequently points out

that a unit of contemporaneity in style, like a language, can and has to

be understood in the makers' terms, in such a way that the rules governing
the compositions and the meanings expressed through them are revealed.

Stylistic "grammars," like linguistic ones, are subject to different kinds
of regularities and patterns of change than the features of which they

are composed. Since the meanings understood by the makers are important
guides to the discovery of patterns of composition and change in a sty-
listic tradition, they must be constantly considered in style studies.

For the same reason it is necessary that the remains of a style be viewed
in terms of complete compositions, that is, whole objects, not just

fragments. In a procedure such as this statistical counting is usually
irrelevant, although it can sometimes be used in statements concerning

the relative frequency of features or certain combinations of features of
a style at any given point in time.

In the light of this discussion it can perhaps be more readily
appreciated why the establishment of exact units of contemporaneity is

all-important. In the absence of such units stylistic regularities can
be only poorly understood, or they cannot be understood at all. In part
evidence for contemporaneity or the lack of it can be discovered through

104

style studies alone, through the observation of co-occurrences of dif-
ferent features on the same objects. In part, however, such evidence
has to be collected through the very careful observation of all kinds
of archaeological associations, such as burials, caches, deposition

strata and the many other kinds of associations observable among remains
of human activity (cf. Rowe's articles on "Stratigraphy and seriation,"
1961, and "Worsaae's Law and the use of grave lots for archaeological

dating," 1962). It has been one of Rowe's great objects to help intro-
duce more exact standards in the observation of associations in Andean

archaeology, an endeavor in which he follows the remarkable pioneer lead
of Junius B. Bird.

Rowe views the discovery of more exact units of contemporaneity
as the first task in style studies, as in all archaeological studies.

However, the methods of style studies that he uses reveal far more than
simple chronological units and cannot, in fact, be discovered without

taking into account other kinds of patterns of human thought and behavior,
such as those revealing differneces of rank or occupation of the owners
of different objects, patterns of imitation and prestige, of abrupt or

gradual changes, of archaizing and nativistic revivals of various kinds,
and patterns of deliberate rejection and purge of style elements. Poli-
tical power, nationalism, or religious sentiments of peoples can be per-
ceived through style studies alone under certain circumstances, and

even the degree of remembrance of oral traditions can sometimes be in-

ferred. (For an example of how the latter can be discovered, see Patri-
cia J. Lyon 1966).

Style studies thus can be one of the most effective means

leading to the discovery of patterns of human behavior and the reconstruc-
tion of particular events in the culture history of a people. They are
very far from being confined to the study of material culture or simple
descriptive units serving only as tools for making temporal contrasts.

In fact, the system of style analysis proposed by Rowe cannot be done by

105

description alone. It requires simultaneously the exposition of rules
of human behavior and the explanation of the processes of change. In-
ference and classification thus must go hand in hand and complement

each other. It is impossible to classify first and make inferences later.
Equally, it is useless to attempt a single classification for all pur-
poses. Different features and patterns of features reveal answers to

different questions being asked. On the other hand,, the questions them-
selves are suggested in part by the kinds of patterns that are revealed
in the process of study.

Rowe has also inculcated in his students, chiefly by example,

the need for basing inferences closely on the evidence, and for formulating
alternative hypotheses as a guard against entrapment in a ruling theory.

He has encouraged his students tp read a classic article on this subject,
written in 1897 by Thomas Chrowder Chamberlin, entitled "The method of
multiple working hypotheses." Furthermore, Rowe has pointed out that

archaeological inference concerning human behavior is not possible with-
out relating the archaeological observations to observations of living
peoples. He has often stated that an archaeologist must also be an
ethnologist, for this reason.  Archaeological objects merely reflect

human thought and behavior, and must be interpreted in terms of it; they
are not objects of study for their own sake, and archaeology is only oc-
casionally the study of material culture. One of the most useful publi-
cations by Rowe in this regard is his article of 1962 entitled "A social
theory of cultural change," in which he discusses cultural behavior con-
cerning imitation and prestige, based on the observations of the French
sociologist Gabriel Tarde. Rowe's contribution is to have recognized
its value and its relevance to archaeological analysis, and to have

brought these concepts to the attention of archaeologists, who, after
all, deal with cultural change far more than either sociologists or
ethnologists are able to do.

106

Rowe's contributions to the great expansion of the possibil-
ities of inference in the Andean field have proceeded along several

different lines. Perhaps his greatest contribution lies in the area of

historical research based on documentary evidence. Several of his studies
of Inca culture at the time of the Spanish conquest, Inca traditions in
colonial and modern times, historical dating, and the Inca language have
become classics in their field (cf. "Absolute chronology in the Andean
area," 1945; "Inca culture at the time of the Spanish conquest," 1946;
"The distribution of Indians and Indian languages in Peru," 1947; "The

kingdom of Chimor," 1948; "Sound patterns in three Inca dialects," 1950;
"Colonial portraits of Inca nobles," 1951; "Eleven Inca prayers from the
Zithuwa ritual," 1953; "El movimiento nacional inca del siglo XVIII,"

1955; "The Incas under Spanish colonial institutions," 1957; "Inca re-
ligion," 1958; "The age grades of the Inca census," 1959; "The origins
of creator worship among the Incas," 1960; "The chronology of Inca

wooden cups," 1961; "Un memorial de gobierno de los incas del ano 1551,"
1966; "What kind of settlement was Inca Cuzco?," 1967; and others).

These historical studies form the necessary starting point for fruitful
inquiry into the Andean past. Only one other anthropologist, John V.
Murra, has up to now made comparable contributions along this line of
inquiry. Rowe's historical analyses have greatly expanded the possi-

bilities of interpretation of the archaeological record, especially in
areas such as religion, techniques and evidence of conquest, and art

history. Rowe has also done pioneer work in the archaeology of Cuzco,

the site of the capital of the Inca empire, and thus the area where the
archaeological remains form the major connecting point with the documen-
tary data ("An introduction to the archaeology of Cuzco," 1944).

In addition to his historical researches and work on the

archaeology of Cuzco, Rowe had made important contributions in chrono-
logical studies of more ancient remains, and in the interpretation of

art history and ancient religion from archaeological remains. Particularly,

107

he has contributed to the study of the Chavln culture and the area of

its influence. He was the first to show evidence for chronological dif-
ferences in the art style of Chavln itself, as well as in the Chavin

style pottery of the north coast of Peru (cf. "A seriation of Cupisnique
stirrup spouts," ms.; "Chavln art; an inquiry into its form and meaning,"
1962; "Form and meaning in Chav{n art," 1967).   His work on Chavfn art

is also of major significance in another respect. Rowe used the curious
and unnatural patterns of associated representational features to argue
that they represent an elaborate symbolism which depicts a rich mythical
pantheon. Hammel points to other important features of this study.

Rowe has been deeply involved in all aspects of the chronology
and culture history of the Ica valley and neighboring valleys with re-

lated traditions, since the Ica sequence is used as the master sequence
for the relative chronology and therefore requires special attention.

His research has encompassed the entire time span of recorded Ica culture
history from preceramic times to the present. In this work he has col-
laborated with many different individuals, from the native Iqueno his-
torian Alberto Casavilca Curaca who was a collector of invaluable

documents of the early Colonial period and who discovered the first pre-
ceramic site at Ica, to Gabriel Escobar and Eugene A. Hammel, who did

field work in modern communities of the native people of Ica (cf. Hammel
1969). Together with Patricia J. Lyon, Rowe discovered one of the

earliest ceramic sites on the coast of Peru at Ica, and has made an in-
tensive analysis, as yet unpublished, relating it to a large site with
the same kinds of remains in the nearby Acarl valley. An analysis of
the textile remains of the Acarl site has been published by Ann H.

Gayton (1967). Rowe's research contributions have covered the entire
subsequent time span of Ica history, including the Colonial period.

His work on the colonial records of Ica and neighboring areas has been
of particular importance. Although much of the work on Ica remains un-
published, several reports by Rowe, with or without the collaboration

108

of associates, have appeared or are in preparation ("Archaeological ex-
plorations in southern Peru, 1954-55," 1956; "La seriacion cronologica

de la ceramica de Paracas elaborada por Lawrence E. Dawson," 1958; "Nuevos
datos relativos a la cronologla del estilo Nasca," 1960; "La arqueologla
de Ica," 1962; "The Paracas pottery of Ica," 1964; "The role of Chincha

in late pre-Spanish Peru," 1966). Additional reports have been published
by his individual associates.

Beyond these research projects on the archaeology of Cuzco,
Chavin and Ica, there are a great many others covering all aspects of

Andean prehistory in which Rowe has collaborated in some form. Another
major contribution by Rowe has been to test the prevailing theories that
Andean culture history proceeded in progressive evolutionary stages, by
posing questions that could be answered on the basis of the evidence.
His article on "Urban settlements in ancient Peru," 1963, 1967, is an
example of such a test, as Hammel points out.

In a broader sense, Rowe's contributions to the study of Andean
culture history and culture history in general are based on the premise
that useful working hypotheses concerning processes of culture change

can be formulated only through the use of specific and detailed histori-

cal and archaeological (i.e., diachronic) data, since such data constitute
the only legitimate evidence in this case, in accordance with the most

basic principles governing any kind of scientific and scholarly inquiry.
The only alternative procedure is to formulate theories concerning cul-

tural development at the outset, basing them on preconceptions concerning
the nature of culture and culture change, and supporting them by some
diachronic data which are selected or interpreted to prove the theory.
Such a procedure is self-defeating, for it can never lead to the dis-
covery of anything that is not already assumed.

The foregoing does not by any means cover all of Rowe's con-

tributions to the field of Andean studies, only some that I consider to
be particularly far reaching in their effects. However, Rowe has also

109

performed a variety of essential services to the field. For example, he
has written a biography of Max Uhle, the principal founder of Andean

archaeology, with a discussion of his contributions (1954), he has made
a survey and analysis of the problems posed by conflicting radiocarbon

dates in the Andean area (1966), and he has organized an efficient system
of site designations (cf. "Indicaciones para la utilizacion de la clave

de abreviaturas y cifras usadas en el sistema Rowe...," 1965, 1966).   He
makes field trips to Peru and, more recently, Bolivia regularly in order
to pursue new field research, to keep in touch with ongoing field work
by other researchers, and to keep abreast of publications generally

available only locally. Another purpose of these trips is to initiate

new and promising students into Andean studies of all kinds, to keep in
touch and exchange information with Peruvian and Bolivian scholars, and
to be of as much assistance as possible to all individuals endeavoring
to make contributions in this field.

It was part of Rowe's conception of service of this kind that
he organized the Institute of Andean Studies in 1960. The principal
purpose of this organization has been the publication of an annual

journal on Andean archaeology, entitled Nawpa Pacha, which Rowe edits.
However, the Institute also sponsors field projects and research. The
publication program, begun in 1963, has filled a particularly pressing

need, for there was no other journal of this kind in existence in North
America, and the publication pressure on existing journals on general

anthropology and archaeology was too great to allow for the more ample
specialized publication requirements brought on by the active research
programs in the Andean field. The annual meetings of the Institute of
Andean Studies have become an important meeting ground for a growing
number of Andeanists, and the free and stimulating exchange of infor-
mation during these meetings has been particularly helpful in leading
to further research and fruitful collaborations.

110

REFERENCES

Chamberlin, Thomas C.

1897    The method of multiple working hypotheses.   Journal of

Geology 5:837-848.
Gayton, Ann H.

1967    Textiles from Hacha, Peru.   Sawpa Pacha 5:1-13.   Berkeley.
Hammel, Eugene A.

1969    Power in Ina: the structural history of a Peruvian community.

Boston: Little, Brown and Co.   (In Press)
Lyon, Patricia J.

1966    Innovation through archaism: the origins of the Ica pottery

style. Nawpa Pacha 4. Berkeley.

[EDITOR'S NOTE: For reference to publications by John H. Rowe, please
see bibliography of his works in this volume.]

111