7
Central California Archaeology:
The Concepts of Pattern and Aspect
David A. Fredrickson
(1982, revised 1984)
Introduction
HE CLASSIFICATION OF archaeological
assemblages is now and always has been one of
the most troublesome tasks for the scholar, and yet
dealing with the theoretical bases of such classifica-
don is a task that most of us tend to side-step or put off
until another day. Randy Milliken reminds me that we
tend to forget that we are anthropologists and, because
of this, we often treat archaeological materials as
objects apart from the people who were responsible for
their deposition. We seem rarely to draw upon our
anthrpological knowledge of human behavior and
cultural complexity.
The work of Willey and Phillips (1958), in my
estimation, is still the most useful exploration of con-
cepts employed in archaeological classification. I see
no problem in accepting the bulk of their basic defini-
tions, and indeed urge for the sake of more explicit
communication that these definitions be accepted by
others as well. According to Willey and Phillips
(1958:22-24), the "phase" is:
an archaeological unit possessing traits suffi-
ciently characteristic to distinguish it from all
odherunits similarlyconceived, whetherofthe
same or other cultures or civilizations, spa-
tially limited to the order of magnitude of a
locality or region and chronologically limited
to a relatively brief interval of time.
Problems do occur in operationalizing the concept.
The absence of specific criteria for defining phases is
shown by the comment that:
a phase may be anything from a thin level in
a site reflecting no more than a briefencamp-
ment to a prolonged occupation of a large
number of sites distributed over a region of
very elastic proportion (Wiley and Phillips
1958:22-24).
Further, as additional data are compiled, phases may
be subdivided into smaller units, which themselves
may be either equivalent phases or groupings of
subphases. Again there is an absence of explicit
criteria for such subdivision.
We are familiar with the construction of local
sequences, often based upon the excavation of a single
complex site or a number of related sites in a single
locality. Local sequences are most commonly the
product of a single worker, or a small group of
cooperating individuals, who have firsthand knowl-
edge of the materials which they place into the local
sequence. By contrast, the regional sequence, such as
ones in the North Coast Ranges or the upper Sacra-
mento Valley and environs, requires correlations be-
tween various local sequences. It is instructive to
76     Toward a New Taxonomic Frameworkfor Central California
again review at lengti the comments of Willey and
Phillips (1958:26-29) on this issue:
Ideally, the archaeologists of a region come
together in a harmonious session where a
careful matching of local sequences produces
a new sequence of larger scope. Actually this
happy event occurs but rarely. What more
often hapens is that phases and local se-
quences gain in scope by a sort of osmosis.
They flow outward, so to speak, often pro-
pelled by their onginators, uniting to them-
selvestheirweakercorrelatesoverawidening
circle. The process is necessarily accom-
plished by a progressive generalization of
definition until much of the original useful-
ness to research is impaired.
I provide these citations from an earlier genera-
tion to remind us that the classification of archaeo-
logical assemblages has never been an easy task and,
implicit in the Willey and Phillips analysis, there is
seldom agreement among researchers who are sepa-
rately (or even cooperatively) attempting to organize
archaeological data into meaningful spatial or tempo-
ral units. The Central Califomia sequence identified
in Bulletin 2 of the Sacramento Junior College (Lillard,
Heizer and Fenenga 1939), and later refined and
extended by Richard Beardsley (1954), is a good case
in point. Gerow (1954; Gerow with Force 1968),
Ragir (1972), and Fredrickson (1973,1974a), as well
as others from time to time, have offered critiques of
the Central California classification.
Clearly disagreements occur for a number of
different reasons. On the formal level we can see from
the Willey and Phillips definition of a phase that there
is no standard established for defining the minimum
degree of similarity between assemblages to warrant
grouping them together. There is no definition of the
nature of the differences which would warrant the
separation of assemblages into different units. If
archaeological assemblages are examined employing
the direct historical approach, with reference to known
ethnographic data, the logical problem of specialized
assemblages of the same community occurring at
different locations and otherwise incomplete assem-
blages comes to light. In fonnal tenns such assem-
blages could be grouped into separate phases repre-
senting different archaeological "cultures." It is also
evident that when we attempt to correlate local se-
quences (often incompletely and poorly defined) within
a region, the absence of adequately defined criteria
impedes the process and hinders agreement between
different workers. Even when workers accept the
same taxonomic framework, which they rarely seem
to do, questions such as the one above remain, i.e., do
contrasting assemblages represent different activities
of the same group or the basic assemblage of contrast-
ing communities? In the transition zone between
adjacent regions, regardless of how criteria are estab-
lished, it often appears arbitrary as to whether a
parficular phase is grouped with one regional se-
quence or another.
Archaeologists are dealing widt the same kinds
of problems thatethnologists attempted to deal within
refining the culture area concept. Unfortunately in
archaeology, small sample sizes, the lack of compa-
rable materials, and the uneven quality of the data
make meaningful statistical correlations difficult if
not impossible, so that work analogous to that done by
Driver and Coffin (1975), and during an even earlier
generation by Klimek (1935), is not feasible in most
archaeological regions, at least in Califomia. Cri-
tiques applied to the culture area concept in general
anthropology apply many times over in archaeology.
Binford (1965) has discussed some of these
problems within the contrasting frameworks of the
nonnative and processual approaches. He rightfully
points out that culture is not shared but is a system in
which persons participate. "Individuals and social
units are articulated by means of various institutions
into broader units having different levels of corporate
inclusiveness" (Binford 1965:205). Since culture is
not a univariate phenomenon, it is methodologically
unsound to simply group assemblages together with
respect to numbers of similarities and differences. For
example, if shell beads and omaments dominate the
known assemblage, such groupings would not iden-
tify communities in the discrete sense but rather a
portion of an exchange network that may include a
wide range of separate communities. Binford
(1965:208-209) suggested that there are at least three
major types of "broad cultural alignments" that may
vary independently of one another-the tradition,
interaction sphere, and adaptive area.
For Binford, tradition refers to "demonstrable
The Concepts of Pattern and Aspect   77
continuity through time in the fonnal properties of
locally manufactured craft items," or "stylistic vari-
ability." Interaction spheres are "areal matrices of
regular and institudonally maintained intersocietal
articulation," which may crosscut both traditions and
culture areas. An adaptive area is one "which exhib-
its the common occurrence of artifacts used primarily
with coping directly with the physical environment."
Binford (1965:209) relates the adaptive area to the
culture area concept, excluding stylistic attributes
from the definition. He argues that these three spheres
should be studied as independent variables if we are to
understand more fully the operation of cultural pro-
cesses.
While Binford's critique of the nonnative ap-
proach has heuristic value and the fornulations dis-
cussed above have considerable value, the major
problems in classification of assemblages are still not
addressed. General concepts that appear to make
sense are suggested, but no clue has been offered with
respect to operationalizing them. For example, how
do we compare the stylistic traditions of one local
sequence with those of another sequence? Binford
appears to ignore the problem, and, indeed, the
"spheres" suggested by Binford are themselves mul-
tivariate phenomena. It is likely that an "interaction
sphere," for example, consists of a variable numberof
subsystems none necessarily covarying with the oth-
ers. Recent ethnographers have emphasized the vari-
ability in individual behavior within communities,
thus adding more doubt regarding the bases for taxo-
nomic grouping of archaeological materials.
Nonetheless, we archaeologists must deal with
the problem of organizing massive amounts of infor-
mation into groupings which have meaning to us,
despite apparent lack of congruence with the real
world. Inmy doctoral dissertation (Fredrickson 1973),
drawing heavily on the work of Jim Bennyhoff, I
suggested a series of concepts to deal with temporal
and spatial classification in Califomia. These con-
cepts were based upon those of Wiley and Phillips
(1958) but with some alterations that attempted to
reconcile normative and processual approaches. The
reconciliation was not explicitly done, and, upon
review with the benefit of hindsight, I did not fully
recognize the multivariate nature of culture. The
system which I proposed has been most extensively
used in the North Coast Ranges, but it has recently
gained currency throughout the state (e.g. Moratto
1984).
In this essay, I focus on the concepts of "pattem,"
"aspect," and "period," all of which were first intro-
duced to northem California archaeology in 1973
(Fredrickson 1973). At that time, I (Fredrickson
1973:118-19) wrote that a pattern was characterized
by: "(a) similar technological skills and devices (spe-
cific cultural items); (b) similar economic modes
(production, distribution, consumption), including par-
ticipation in trade networks and practices surrounding
wealth (often inferential); and (c) similar mortuary
and ceremonial practices." I stated that local varia-
tion in a pattem may sometimes be extreme depend-
ing upon "(a) abundance and nature of specific envi-
ronmental resources; (b) regional specialization and
elaborations; (c) degree of cultural and geographic
marginality; (d) influence of neighboring pattems."
I originally conceived of both pattern and aspect
as representing different levels of interaction spheres.
The aspect represented direct community interaction,
and the pattem represented more widespread inter-
action, perhaps comparable to what might be called
an "effective interaction sphere," as contrasted with
the "minimum interaction sphere" represented by the
aspect and the "maximum interaction sphere," repre-
sented, for example, by trade horizons, which may cut
across the boundaries of adjacent archaeological ar-
eas. While I still hold this view, I believe we can
further refine the concepts, borrowing from Binford.
At this time, I offer a refinement in the definition
of pattern, more in keeping with the "adaptive sphere"
of Binford. I retain criteria of similar technological
skill and devices but reject as a necessary criteria the
similarity of mortuary and ceremonial practces.
This is not to say, however, that such linkages do not
occur fiom time to time. The concept of economic
modes which I employed in 1973 is far too broad and,
in fact, represents a series of distinctly separate activi-
ties and processes. I would retain similarity with
respect to the means of production (and probably
consumption), but distribution, especially as it relates
to exchange systems and wealth practices, must be
kept apart from the pattern concept. Although I keep
78      Toward a New Taxonomic Frameworkfor Central California
exchange outside of the pattem concept, I think it
likely that there is a level of exchange, or interaction,
which is represented within a pattem. I see no need for
any significant change in the qualifications regarding
local and regional variation with a pattem.
North coastal California pattems identified to
date include Gunther (of Northwest Califomia) and
Augustine (of Central California), both of the Emer-
gent period; Berkeley (of Central Califomia, emanat-
ing from the Bay Area and with historical origins
possibly from the south coast); Borax Lake; and the
still-provisional Post (see Fredrickson 1984; Fredrick-
son and White 1988). Discussion is still underway
regarding the nature and definition of assemblages
that appear to represent one or more patterns interme-
diate between Borax Lake and Gunther and Borax
Lake and Augustine. Each pattem can be conceived
as representing an adaptive sphere strongly marked
by the common occurrence of artifacts used primarily
in coping directly with the physical environment. We
should recall that although culture is indeed a multi-
variate phenomenon, independent variables may in-
deed(butnotnecssarily)covarysimultaneously. Thus,
we do at times observe burial modes changing simul-
taneously with extractive technology. The pattem
concept can be seen to represent a "lumper's delight."
The phase and subphase, by contrast, are each a
"splitter's delight," and, since each is ideally based
upon detailed analyses of stylistic elements, each
representsanequivalentofBinford's'taditionsphere."
Having now borrowed from Binford, we are still
left with the problem of dealing with variation in the
pattern, or adaptive sphere. Such variation can be
identified on the basis of stylistic elements (Binford's
traditions). However, phases and subphases represent
the smallest archaeological units that can be identified
and may, if we are fortunate, actually represent the
equivalent of a community. Following Willey and
Phillips (1958), the geographic space through which
a given phase or subphase can be traced would be a
locality. In practice, however, we can rarely distin-
guish one locality (in the sense of a local tribelet
territory) from another. Forpurposes here-acknowl-
edging Jim Bennyhoff as the originator of the idea-
I refer to the space encompassed by a phase or subphase
that canbe traced and canbe referred to as a "distnrct."
As a working hypothesis, Bennyhoff (see Bennyhoff
1977), has proposed that districts equate with linguis-
tic communities. While ideally every politically au-
tonomous community should be represented in the
archaeological record by a distinct sequence of phases,
in practice-based on ethnographic and protohistoric
parallels-the district is most commonly made up of
a series of interacting communities, communities that
are probably linked through internarriage and kin-
ship. If we examine Yellen's (1 977) data from Africa,
and data recently generated by Randy Milliken from
Mission Registers, such communities may be actively
and constantly exchanging personnel.
As we view the archaeological data from the
Califomia culture area, the major exchange networks
themselves allow the suggestion that we are in need of
a taxonomic grouping to deal with variability not
expressed within the pattern and aspect concepts. We
can observe a level of assemblage similarity that
appears to fall between these two concepts as they
have been applied to date. Conceiving of the aspect as
a "minimum interaction sphere," as noted above, has
limited our synthesizing ability. It may be better
conceived of as an "internediate interaction sphere."
In 1973 I defined an aspect as being equivalent at any
one time to a phase, and over a longer period of time
by a sequence of phases. With the aspect expanded to
the intennediate level, at any given time slot, it can be
made up of perhaps several synchronous but some-
what contrasting phases. A specific phase, then, is
situated within a specific locality. Several similar
phases in adjoining localities would be grouped into
an aspect, with the localities grouped into a district.
To facilitate communication prior to definition
of specific phases and aspects, I further suggest a
concept that I refer to as a "style-area" be adopted to
deal with this mid-range variability. How the concept
is to be employed to distinguish the multivariate
natureof archaeological materials remains tobe worked
out. I can conceive of the concept being employed in
each of two major ways: (1) by a qualifier expressing
geographic spread, best applied to assemblage simi-
larities on a level intennediate between pattem and
aspect (but not necessarily limited to this level), and
(2) by a stylistic qualifier representing the geographic
distribution of a particular stylistic element. For
example, the "Tehama" materials discussed by Elaine
Sundahl and similar materials discussed by Bill Dryer
The Concepts of Pattern and Aspect    79
might be grouped together into a "Southem Cascade
style-area." On the other hand, the geographic area in
which the indented base, wide-stemmed projectile
point (spear point or knife) is found could be referred
to as the "indented base, wide-stemmed point style
area," or more simply, the "wide-stem style-area."
This practice could reduce our tendency to equate the
wide-stem point necessarily with the Borax Lake
pattem. Further discussion regarding this concept, or
an alternative method of referring to the middle level
of cultural grouping, could prove productive.
I do suggest that we pay more heed to the multi-
variate nature of culture, adding such focus to the
synthesizing tasks in which we are already involved.
Although I have pointed out the lack of satisfactory
criteria for defining the several taxonomic units, I
have not attempted to deal with the problem. Over-
coming the problem may well take us back to an "'old-
fashioned" approach, that of plotting geographic and
temporal distributions of discrete artifact types and
subtypes, rather than focusing all our energies upon
the assemblage alone. We might call this the testing
of the multivariate hypothesis. Such an approach may
help us refrain from reifying our constructs, and of
assuming similarities when we apply the same name
to assemblages from different locations. Our state of
knowledge should now encourage us to entertain
multiple working hypotheses. I cannot overempha-
size my perception that we are at an ideal place in data
collection and analysis. We may now take a step back
and be critical (and subsequently creative) regarding
methodological issues. Our comparative skills and
methods, for sure, need honing.
I have recently been made aware of yet another
dimension of assemblage distribution which is not
encompassed by the taxonomic system as originally
conceived or as it might be modified according to the
suggestions above. Greg White has infonned me of
the apparent coexistence of contemporaneous but
contrasting assemblages within the same locality.
What appears to be persistence of the milling stone
assemblage in Lake County is found cotenninously in
localities containing Houx assemblages. Brian
Wickstrom has noted the identical phenomenon in the
Santa Rosa locality, with a milling stone assemblage
co-occurring with Berkeley assemblages. Both of
these examples could be similar to the "Tehama Pat-
tem" as described by Clewett and Sundahl (1990).
It may be that archaeology must lose the unity of
the Califomia culture area as defined on the basis of
ethnographic traits. In that event, Augustine and
Shasta would become synchronous pattems and
Tehama, then, would become an aspect of Shasta or
Augustine, or even be elevated to the pattem level
itself. In the North Coast Ranges, Houx would be-
come a pattem separate from Berkeley. The major
problem with this approach is that it encourages
classificatory fragmentation, but perhaps that cost is
in keeping with the benefits that might be accrued.
As for the second problem, we know on the basis
of ethnographic data that contrasting cultures can
indeed occupy different niches within the same habi-
tat (i.e. within the same or overlapping geographic
ranges). It is clear to me that the Central Califomia
taxonomy as I have discussed it is inadequate in
treating this circumstance.