225
OVERVIEW OF GREAT BASIN OBSIDIAN SITUIES
Clement W. Meighan
The papers in this volume represent a considerable advance in both the amount
-of interest and the sophistication of the reports on obsidian studies. I
recollect a national meeting only about 15 years ago at which a synposium on this
subject was able to attract only five or six people -- the speakers were there
but the audience was not. In contrast, the synposium reported here had a large
and faithful audience which sat through an entire day of papers relating to
obsidian studies. Archaeologists have clearly recognized by now the significance
and the multiple applications of these studies to archaeological problems.
Particularly in the Great Basin, there is greater appreciation of what these
studies can do for archaeological understanding, since the Great Basin is charac-
terized by many shallow (often surface only) archaeological sites, and often the
ccnxnest cultural material is obsidian artifacts and chipping waste, due to the
ultiplicity of obsidian sources in and around the Basin. With so many sites
lacking samples datable by radiocarbon, and lacking physical stratigraphy,
obsidian dating is often the best source of information on the age of the site,
and obsidian sourcing is often the best source of information on trade relation-
ships. Hence the relatively greater interest in obsidian studies from those who
work in the Great Basin.
There are four papers here dealing with the important issue of determining
the source from which obsidian came. In spite of the increasing attention to
this problem, there are still numerous practical and intellectual matters to
resolve. Hghes points out the difficulties in determining the sources of obsi-
dian (through chemical analyses of varying kinds), and cautions that incorrect
assignment to source may go undetected and unchallenged, leading to misinterpre-
tation of trade routes. This may also lead to confusion in applying obsidian
hydration dating, because it is becoming increasingly clear that different
obsidian sources show formation of hydration bands at quite different rates.
Hampel also stresses the need to adhere to specific basic procedures in
x-ray fluorescence (XRF) studies of obsidian sources, while pointing to refine-
ments of the method. Nelson applies XRF methods and identifies the chemistry
of many obsidian sources in the Great Basin. This is particularly valuable
since mnst of the earlier source identifications were made on obsidian from
California, at one edge of the Great  asin. (Xprative data for Nevada,
Oregon, Utah, and Idaho greatly expand knowledge of Great Basin sources and
provide a much firmer basis for future studies of contact and obsidian trade.
Nelson observes that prehistoric people usually got their obsidian from the
closest available source but that there are numerous instances where this
practice was not. followed -- hence the basis for study, and hopefully explan-
ation, of long-distance trade in the past. That long-distance trade was very
important can be seen in the great amounts of obsidian recovered in areas where
there are few or no sources of this material, such as coastal southern
California.
226
The possibility of avoiding the cost and difficulty of laboratory deter-
minations of obsidian sources is considered by Bettinger, Delacorte, and
Jackson, who have sorted obsidian on the basis of visual characteristics with
reasonable success. This procedure is not without some error, but neither are
XRF or other procedures. Obviously an in-depth study requires some chemical
analyses to identify source, but once the sources are known the visual charac-
teristics of the obsidian can be carefully tabulated and observed and it then
becomes possible to recognize pieces from that source without a full chemical
analysis of each piece. This is important because there is no possibility that
chemical analyses can be performed on the many thousands of individual specimens
that can be collected from even a single site. A judicious combination of
chemistry and observation of known traits is probably the only practical way of
dealing with large collections.
It may be noted that other "short-cut" methods are applicable as well.
Some obsidian sources have aberrantly high or low amounts of specific trace
elnemnts, and it will be possible to get a pretty good idea of the source by
testing merely for those elements. This requires mnre thorough knowledge of
the chemistry of all the sources than is now available, but as the data base
improves it will certainly become possible to identify some obsidian by looking
for distinctive marker elements without the need for more detailed analysis.
It has also been pointed out by Jonathon Ericson that in some cases the hydration
rate for specific obsidian sources is very distinctive (exceptionally fast or
exceptionally slow) and that in some archaeological contexts it is possible to
know the source by knowing the applicable hydration rate.
All of these short-cut methods involve extrapolation from knowing part of
the data to drawing conclusions about other parts of the data, a commDn-sense
procedure which often works but is based on assumptions. All of us making use
of such methods are obligated to make clear in our reports what statements are
verified objectively and what statements are in fact assumptions or extrapola-
tions from partial data.
These various studies of obsidian sources are aided by the apparent tendency
of obsidian users to collect their obsidian from small areas, returning to the
same spots (possible to the same ledge or outcrop) over many visits. This
behavioral pattern tended to collect and distribute morre uniform kinds of obsi-
dian than would be the case if obsidian were gathered at random from every
possible location on a given obsidian flow. In other words, the total chemical
variability of an obsidian flow does not seem to be strongly reflected in the
observed obsidian found in archaeological contexts. This fortunate circumstance
of patterned collecting activity aids both chemical and visual identifications
of source.
Three papers deal with obsidian hydration analysis. R.J. Jackson points out
the methodological problems of hydration studies and makes the same plea for
standardized and replicable procedures voiced by the sourcing specialists with
respect to their methodology. Jackson provides interesting comparisons between
obsidian hydration readings obtained on the same specimens by different labora-
tories. At least two other studies of this kind are in press or underway, and
they provide objective evidence on replicability, margins of uncertainty, and
227
other problems with making the slides and measuring the obsidian hydration bands.
Cross-checking between laboratories becomes increasingly important as the number
of technicians increases and differing optical and slide-preparation techniques
are used. Some of the observed variability between laboratories may be due to
masurLng the obsidian hydration band at different locations, but whatever the
CseUM we should all have a little humility about the accuracy of our hydration
IMasuremets.
I would like to add to Jackson's observations of variability my own plea for
a) systematic publication of obsidian hydration readings;, and b) systematic
cataloging and preservation of the published slides. It is important to know
who the technician was, how the work was done, and where the original slide can
be obtained for checking and verification of the reading. It is equally impor-
tant that the original data be made available; it is impossible to know what
sites have been studied for obsidian hydration since the majority of the data
are sequestered and are either never made public, or appear as an appendix to a
site report many years after the readings were taken. It is my impression that
the researchers on obsidian sources have done a much better job of publishing
their basic data and making it available for comparative study. Only two of the
existing obsidian hydration laboratories make systematic efforts to publish their
readings, and only UCLA publishes all of its obsidian hydration readings and
makes all of its slides available to scholars for re-examination when needed.
The two other articles in this section of the volume (Davis; Termbour and
Friedman) deal with the important matter of the temperature variable in the rate
of obsidian hydration. Davis presents a most ingenious way of determining
effective hydration temperature in the past. Trembour and Friedman stress the
importance of temperature in their calculations of hydration rate, with snall
differences in temperature having a large effect on the rate. They also stress
the necessity of sorting out surface from buried pieces and the difficulties
arising from obsidian exposed for even a short time to high temperatures (such
as a cooking fire or brush fire). This is a formidable set of difficulties in
archaeological collections. Not only does the temperature vary seasonally, but
the period of hydration may span warmer and cooler climatic cycles.  Any obsi-
dian lying in chapparal areas has a high probability of exposure to brush-fires
every 50 years or so if not more often.  Pieces once on the surface get interred
by ground squirrels, while the busy rodents are also bringing up buried pieces
and depositing them in the surface sunshine. Finding the temperature in order
to work out the hydration rate involves a number of unverifiable assumptions and
may lead to the same kinds of difficulties encountered with mino-acid dating,
in which incorrect assumptions about past temperatures lead to major dating
errors. The system can be scientifically correct and internally consistent but
still be wrong as to the age of the specins due to unknown factors.
Because of these problems, working it the other way around -- that is, using
hydration to determine effective annual temperature, seems to me a more promis-
ing approach. In many cases such findings would be immensely valuable, for the
paleotemperature data would be indicative of other environmental variables
relevant to the interpretation of archaeological sites.
Fortunately, the archaeologist has additional data which provide a check on
both dating and temperature results. For chronology, these data include other
228
chronological methods such as radiocarbon, stratigraphic placement, and "time
markers" of known age in the artifact assemblage. Obsidian hydration dates are
tested against these empirical data, and when there is serious lack of agreement,
re-evaluation of the results is required. With regard to conclusions about
ancient temperatures, many archaeological sites contain an abundance of climatic
indicators in the form of pollen, floral, and faunal remains. Plant and animal
remains reflect the climate and temperature of the time when the archaeological
site was in use, so they can provide a separate line of evidence to evaluate
paleotemperature calculations.
The final section includes four papers which apply the methodology to vary-
ing problems and use the source and hydration findings to move forward to
archaeological explanations.
T.L. Jackson assesses aboriginal obsidian production from two eastern
California sources, Bodie Hills and Casa Diablo. In a detailed review of the
several exploratory studies done on obsidian hydration from these sources,
including his own large sample of 800 pieces from Casa Diablo, it is made clear
that "production" curves are only as believable as the supporting chronological
evidence. For these two sources, understanding of the chronology (obsidian
hydration rate) is still elusive, although the serious consideration given by
Jackson should get us closer to understanding the hydration rate(s).
Bouey and 3asgall look at trade relationships between the Great Basin and
California, based on hydration dating and the model of "economic articulation"
which proposes that a change in production and/or exchange in one area will
have an observable repercussion in the other. As these authors point out, any
general explanations applicable to such a large area are tentative because of
the many scholars involved in collection of the data, plus differing interpre-
tations of obsidian hydration rates. Nonetheless, an effort can be made to
look at political and economic processes which might explain the observed-pat-
tern of obsidian distribution (in space and time). They conclude that Central
California was a major consumer area, that the obsidian trade was linked, at
least initially, to a demand for status goods, and that there was a production
peak between 3000-1600 B.P. Other sub-regions are evaluated and some alternative
models are also presented, including population replacement in local regions.
In examining the structure of obsidian exchange, Bouey and Basgall are get-
ting a great deal of historical explanation out of trade in a single raw
material: obsidian. Their suggestions go far beyond anything dreamed of by
most of the diggers of an earlier era, who never anticipated that general con-
clusions of broad significance could come from their arrowheads and chipping
waste. On one hand, it is encouraging to find that the dating and sourcing of
obsidian can contribute to new ideas and that archaeologists are able to advance
their field of study to new levels of sophistication. On the other hand, there
is also the moral that careful field work, documentation, and preservation of
the obsidian collections are increasingly essential. Shortcomings in the tradi-
tional archaeology will abort formulations of the new archaeologv. and the two
need to work hand in hand if the discipline is to move ahead intellectually.
229
The obsidian hydration paper by R.J. Jackson points out the value of using
obsidian hydration for relative dating, and reminds us that obsidian hydration
studies have important values whether or not they can be used for absolute dating.
However, he cannot resist providing another suggested absolute chronology for
Casa Diablo obsidian. This is worth brief general comment:
First, the four Casa Diablo hydration rates in Table 2 differ, but not sub-
stantially except for readings under 2 or over 10 microns, sizes that are quite
rare in known archaeological collections of Casa Diablo obsidian. I would con-
clude that all of the proposed rates in Table 2 give reasonable age determina-
tions, even though they were worked out in different ways by different people.
It is distressing that we cannot support our formulae beyond question, but it
should be encouraging that similar answers are determined independently. In the
mid-range, for example, the four rates are within a span of 211 years for 3
microns of hydration, with a span of 156 years for 6 mircrons, and within 610
years for 9 microns. This strikes me as a    ng consistency, the greatest vari-
ability being ? 12-20% of the true age, assuming the true age to be somewhere in
the time range expressed by the four rates. This is certainly far better absolute
dating than has been available for most Great Basin sites.
Assuming Jackson's rate to be the most valid, the time intervals between
each micron of hydration (omitting the small bands) are:
Micron 2-3                        520 years
Micron 3-4                        613 years
Micron 4-5                        689 years
Micron 5-6                        759 years
Micron 6-7                        822 years
Micron 7-8                        879 years
Micron 9-8                        934 years
Micron 9-10                       984 years
It is extremely rare to have an archaeological context showing 2-10 microns
of hydration, and often the range of hydration is only 3 or 4 microns in a given
site or archaeological horizon. I have pointed out elsewhere (Meighan 1983) how
this narrow range in a given context makes it impossible to verify an emirically-
determined rate formule - there is just not enough variability in most archaeo-
logical collections to test the formula. Hence alOmst any formula can be made to
yield acceptable answers. Assure, for example, a collection of Casa Diablo
obsidian with a range of 7-10 microns of hydration. With a couple of radiocarbon
dates, one might well propose a simple linear rate of 700 years per micron
(average of Jackson's intervals for 7-10 microns). Such a rate would differ
from Jackson's numbers by as little as 19 years and as much as 537 years (the
latter in a Jackson age of 6837 years, or 8%. of the age). The 700-year rate
would not work at all for hydration bands of significantly larger or smaller
size, but in our hypothetical collection it would be as close to the truth as
any other rate, and it would be impossible to dispute by any archaeological
evidence (including radiocarbon dating).
The rough and ready linear rates used for determining ages of particular
assemblages are irritating to the purists, lack elegance, and are "wrong" in the
230
laboratory or theoretical sense. But they work with the real data the
archaeologist has at his disposal, and they do provide age determinations as
close to the truth as we can measure.
When dealing with collections from a wide area and many sites, as Jackson
does in this paper, the rough and ready methods can lead to some big mistakes.
Jackson's relative dating chart (Figure 3) shows hydration measurements of 1-10
microns; here he avoids absolute dates but does provide the hydration sequence
for six named Great Basin cultures. Since most of these cultures are defined
primarily on the basis of surface finds, the objective sequencing is no mean
accomplishment and provides a good basis for further analysis.
In the final paper, Tuohy applies obsidian dating to 38 "early rrmn" points
from Nevada, including examples of Clovis, Elko, and other types associated
with earlier cultures. In this complex subject, Tuohy confronts the morass of
typological confusion, varying obsidian sources, and related problems which
make it difficult to go beyond tentative conclusions.
One situation observed here by Tuohy (and elsewhere by others) poses a con-
siderable problem to users of obsidian hydration data. This is the confusion
caused by abrasion or erosion of the surface of obsidian pieces. Exposure of
obsidian to desert winds full of sand and silt can sand-blast the surface and
alter the obsidian hydration band. This abrasion can sometimes be detected
microscopically, and it is no problem if all the hydration layer is reamved.
But there is good evidence that abrasion can sometimes remove part of the hydra-
tion band and leave a fraction of it intact to be measured by the archaeologist.
This shows most clearly when the surface of a specimen has a small hydration
layer while a much larger layer is visible in cracks protected from surface
abrasion. There does not appear to be any simple solution for this problem, but
a large enough sample will usually detect anomalous readings and raise the
suspicion that something has happened to the surface of the piece being studied.
On the whole, the papers in this volume represent a lot of brain power.
In my view, Great Basin archaeology is put forward by these studies, which
reflect the ever-present problem of getting more information out of the fragmen-
tary evidence of past activity. It is a contribution to raise an intelligent
question, and the authors of these papers have laid out a challenging agenda of
relevant issues to be explored. There is, of course, much more to Great Basin
archaeology than obsidian studies, but it is remarkable how studies of this
ubiquitous raw material can throw light on multiple lines of investigation:
chronology, trade and even social organization and social classes. The editor
of this volume deserves much credit for organizing and presenting this set of
papers; they will be essential references for future workers.
References
Meighan, C.W.
1983     Obsidian dating in California: theory and practice. American
Antiquity 48(3): 600-609.