Chaîne opératoire, translated as operational sequence,
has been described as, "the different stages of tool production
from the acquisition of raw material to the final abandonment
of the desired and/or used objects. By reconstructing the operational
sequence we reveal the choices made by ... humans." (Bar-Yosef et. al. 1992,
511). Excepting that the individuals in a group have a number
of raw materials and techniques available to them; "identification
of the most frequently recurring of these choices enables the
archaeologist to characterize the technical traditions of the
social group" (ibid). Culture is expressed in these choices
that are made throughout the operational sequence. This approach
contrasts with the typological approach that concentrates on the
end product alone as opposed to the whole process of lithic exploitation.
Typology automatically produces a limited sample as only a very
small percentage of pieces are retouched. This is particularly
the case with small Norwegian sites. The two sites that will be
used as examples are Kvernepollen 9, from the Kollsnes project,
situated on the West coast (Nærøy
1994), and Farsund (Lundevågen 17 in Ballin
& Jensen 1995) situated on the southern extremity of the
west coast (see Map). Kvernepollen is dated to the early Bronze
age by typological dating because of the presence of bifacial
leaf shaped points (overflateretusjerte spisser, see Nærøy
1994, 198) and Farsund is a Mesolithic site carbon dated to
c. 7800 BP (Ballin &
Jensen 1995, 36). Both sites were totally excavated and the
analysis presented here is based on all pieces greater than 10mm.
in any dimension (i.e., omitting the splinter in Norwegian terminology).
At Kvernepollen there are 12 retouched pieces (7 bifacial
points, 3 sidescrapers,
1 endscraper, 1 backed
flake) among the 838 pieces analysed. At Farsund there are 20
retouched pieces (12 scrapers, 4 retouched bladelets, 1 retouched
blade, 1 retouched flake, 1 piercer,
1 broken tanged piece) among
the 1235 pieces analysed. These 'tools' represent entire episodes
of occupation.
Types of tools have been interpreted as being made according to
some mental template so that they were made to a preset form expressing
ethnicity. Therefore when the same types of tools are found at
different sites this represents occupations by the same culture
group. This is the basis of space-time systematics, that is the
placing of sites in chronological sequence and geographical location,
and inferring the relationships between them.
With the development of New Archaeology and the attempt to express
human behavior in terms of scientific laws, it became the fashion
to relate stone tools to environment. Stone tools became the mechanism
by which humans adapted to changing environmental conditions,
following the model of Darwinian evolutionism, and hopefully following
laws similar to genetics in animal species. Though the search
for laws of human behavior, following this model, appears to have
been abandoned, the assumed correlation between the environment
and stone tools continues within the evolutionary paradigm. However
this correlation has become increasingly difficult to sustain.
For example, there was the idea of a very rapid change from Upper
Palaeolithic industries to Mesolithic industries due to the climatic
and environmental changes during the transition to the post glacial
period. In order to adapt to these changing environmental conditions,
Mesolithic technology was adopted to facilitate hunting in more
forested environments. Though environmental adaptation plays a
part in the transition from Palaeolithic to Mesolithic it is only
one factor. The transition began before the end of the last glaciation
as microliths are found in the Magdelanian and the Azilian during
the Upper Palaeolithic, and in some areas did not occur until
after these environmental changes took place, (epi-Gravettian
in Italy). Microliths typologically and technologically indistinguishable
from North European types (as found at Star Carr, Clark
1954), are found in Howieson's Poort assemblages (Mellars
1989) at the tip of Southern Africa and dated to at least
40,000 years ago.
Also the early sites in Norway have a Mesolithic technology
and typology and yet the environment was similar to the late Upper
Palaeolithic in southern Europe, which had an Upper Palaeolithic
typology. Therefore Mesolithic tools cannot be seen simply as
an environmental adaptation.
The ecological approach has been taken a stage further in saying
that though stone tools may not be correlated with changing environmental
conditions, social structure is. This theory, propounded by people
such as Gamble (1986),
is that prior to the Upper Palaeolithic, human groups did not
have the social structure that would enable them to adapt to 'marginal'
environments. Either the dense forest of full interglacial periods
or steppe/tundra conditions of colder periods. This would mean
that no occupation of Norway took place prior to the post glacial
period. However a recent paper (Roebroeks
et al. 1992) has demonstrated that there are a number of sites
occupied in similar marginal conditions during the Middle Palaeolithic,
demonstrating that Neanderthals did indeed occupy such areas.
Of course subsequent ice action would have eradicated evidence
of such occupation in Norway. One only has to imagine the effect
of ice sheets moving over the kinds of sites that are excavated
in Norway to realize that nothing would survive such conditions,
unless there were exceptional circumstances.
Considering the lack of correlation of the environment with stone
tools and/or social structure, the role of 'human choice' has
become more important in understanding stone age sites. One way
of studying 'human choice' is through the chaîne opératoire
approach. The operational sequence is from raw material procurement
to primary reduction techniques (the reduction of nodules to cores),
secondary reduction (the removal of blanks from cores and the
manufacture of tools with retouch), the use of tools and the discard
of the artifacts.
The essential difference between this approach and a typological
approach is that it encompasses the whole process of the life
history of the lithic material, from basic nodules to the remains
that archaeologists excavate. As Stringer and Gamble comment,
"The typology of stone tools has been largely superseded
by models of behaviour that concentrate more on the 'biography'
of the implement - how it was made, used, resharpened, recycled,
changed shape and finally thrown away." (Stringer
& Gamble 1993, 143). An extension to this operational
chain is the post-depositional disturbance of the site and even
excavation strategy, as these will have an effect on our understanding
of the human choices that were made through out the operational
sequence. Cultures, in terms of groups that were ethnically or
traditionally similar, are expressed by these choices.
Raw material procurement
Turning to Norwegian sites, raw material procurement is obviously
an important factor as a variety of raw materials were used and
often found on the same site. Not just flint
but quartz, quartzite,
rhyolite, rock
crystal, slate, etc. Some of these
non flint materials have specific sources, as in the case of rhyolite
coming from the island of Bølmo in Western Norway. In terms
of human choice, why do they choose to exploit rhyolite? This
involves traveling to the source by boat, and transporting it
as nodules or finished products throughout its distribution in
Western Norway. Is the choice of rhyolite an ethnic marker in
that a specific material was associated with a particular tribe?
Does possession of rhyolite proclaim their origin in Western Norway
as opposed to groups originating in Eastern Norway?
It could be argued that there is no choice involved, in that a
scarcity of flint determined that rhyolite would be exploited,
but why not use, for example, quartzite that was much easier to
obtain. Looking at the next link in the chaîne opératoire,
does rhyolite have any particular knapping properties that recommend
it so that the choice is technological? Does rhyolite have advantages
over other materials functionally, in that it is more efficient
than, say, quartzite for specific tasks? This would be revealed
by the uses of the tools at that phase of the operational sequence.
There are also choices in the means of raw material exploitation.
The material can be processed at source, transported as nodules,
pre-formed cores or as finished products. The two sites from Kvernepollen
and Farsund can be contrasted in this aspect. Though both are
coastal sites, exploiting similar environmental resources, there
is a difference in that pre-formed cores were transported into
Kvernepollen, whereas nodules were taken onto the site at Farsund.
This can easily be illustrated by looking at the cortical element
of the debitage (Fig. 2). The pattern from Farsund is typical
of knapping from cortical nodules. The pattern from Kvernepollen
illustrates a lack of cortex that indicates primary reduction
took place elsewhere.
Also there were 10 primary flakes (flakes with 100% cortex)
at Farsund and some of the cores are merely nodules with a few
flakes removed. The knapper appears to have been testing the nodules
for suitability and in some cases rejected them at that stage.
This would suggest that there was a local source of flint. The
pattern from Farsund is similar to the English Mesolithic site
of Raunds where cortical nodules were also being knapped (Fig.
3). It could be suggested that these different patterns are a
result of scarcity of flint in the later period when Kvernepollen
was occupied.
However, the site of Tronsetra which is dated c.7250 BP has a
similar pattern to Kvernepollen. Transporting pre-formed cores
would seem a preferable strategy at Tronsetra, which is located
in the mountains at c.800m above sea level and 100 kilometres
from the coast.
The significance is that the differences are not necessarily
related to time, but also to choice. Other sites (for example
from Songa, Telemark, in Coulson
1986) in the mountains have cortical nodules, so not taking
them to Tronsetra was from choice, not necessity.
Quartzite was extensively used at Kvernepollen (see
Fig. 5), whereas it is virtually
absent at Farsund, even though quartzite is available locally
in both areas and throughout the chronological period. Is this
from scarcity of flint or a choice? The choice being not to use
quartzite if sufficient flint is available. Alternatively, if
the social dynamics and subsistence strategy involved small highly
mobile groups (such as on hunting expeditions), the choice was
to take flint as pre-formed cores and use quartzite as required,
as opposed to occupying sites where flint was available. This
choice of raw material procurement strategy may be a cultural
marker rather than a necessity forced on the people by the availability
of flint. Sites are found were flint is brought in as nodules
and where quartzite is used (Songa, Telemark), suggesting that
different strategies of raw material procurement are chosen, rather
than simply determined by available resources.
Technology
Technology is divided into primary reduction, secondary reduction
and typology. Primary reduction techniques are concerned with
the reduction of nodules to cores, the kind of cores produced
and the technology involved, for example the use of anvil technique
to produce bi-polar cores. Secondary reduction techniques are
those involved in producing blanks from cores and include such
aspects as blade and flake technology, use of hard or soft hammer
and micro-burin technique for the production of microlith blanks.
Typology is concerned with tool production and the techniques
of retouch, including such aspects as pressure flaking, burin
technique and the differences in retouch; both of placement (direct,
inverse, etc.), and type (abrupt, invasive, etc.). Figure 4 illustrates
this reduction sequence. In the example, primary reduction is
by direct percussion to
remove the cortex and shape the core. Secondary reduction is by
punch technique to produce
blades. Tool production is by pressure
flaking to produce, for example, a bifacial
lanceolate point (see Helskog
et al. 1976, 33).
For a comparison of technologies, between Farsund and Kvernepollen,
the discussion will be limited to flint because the presence of
a significant amount of quartzite at Kvernepollen (17%) as opposed
to its absence at Farsund (Fig. 5), effects the overall technology.
The technology used with quartzite is effected by the nature of
the raw material so that comparison between technologies on different
raw materials is less reflective of cultural choices. For example,
it is difficult to use blade technology with quartzite, so blade
technology is less likely to be 'chosen'.
The technologies used are similar a with a slight preference for
blade technology at Farsund (Fig. 6) and both sites have 2 crested bladelets (ryggflekke)
demonstrating the deliberate preparation of cores for bladelet
production. Carefully made conical and cylindrical
bladelet cores were found at Farsund whereas only core fragments
were found at Kvernepollen (complete cores perhaps being removed
for later use).
Blade technology simply means the deliberate production of
blades. A 'technological' blade must be at least twice as long
as it is wide (i.e., the definition of a blade blank). In addition
it must have parallel sides and parallel dorsal ridges and (if
the platform is intact) a prepared platform. The parallel dorsal
ridges mean that it is within the process of reduction of a number
of blades rather than a being a 'one off'. Broken pieces, if they
have the above technological criteria but are not twice as long
as they are wide, are considered as broken 'technological blades'.
Flake technology can produce blanks with a length:breadth ratio
>2, and therefore are blade blanks, but if they do not have
the above technological features they are a product of flake technology
even though they are blade blanks. This avoids using such cumbersome
terms as blade-like flakes. Such a piece would be a blade blank
made with flake technology. These criteria are applied by the
use of an expert system (Grace
1993), so that the amount of blade technology at different
sites is directly comparable, and not influenced by any a priori
expectations of what kind of technology should be found at a site
of a particular period. Blade production can be carried out in
a number of ways; by direct percussion with hard
or soft hammer, indirect
percussion, punch technique (see Fig. 4), etc. Often a combination of these techniques
is used to accommodate the vagaries of individual stone nodules.
Technological comparison is limited to secondary reduction techniques
because of the lack of complete cores at Kvernepollen and the
small numbers of typological tools at both sites (Kvernepollen
n =12, Farsund n = 20).
Considering blank types, more blades/bladelets were produced at
the Mesolithic site of Farsund as might be expected (Fig. 7).
This pattern continues when considering which blanks were chosen
for use (Fig. 8).
The choice here is that though both groups possess the same
technological capability, the knappers at Kvernepollen choose
to produce proportionately more flakes and show an even more marked
preference in choosing flakes for use. Perhaps this sequence is
merely reflecting the time difference between a Mesolithic site
and a Bronze age site following the general trend to produce larger
flake tools in the Bronze age, rather than the production and
use of blades/bladelets in the Mesolithic.
However, the Raunds Mesolithic site has a very similar technological
configuration to Kvernepollen (Fig. 6),
but the Raunds Mesolithic site uses the same technology to produce
more bladelets as with Farsund (Fig. 7).
When choosing which blanks to use the people at Raunds chose blades
(as opposed to bladelets) and flakes, that is, the larger blanks
in a similar way to the people at Kvernepollen, but different
to the people at Farsund who choose relatively more bladelets
as opposed to flakes to use (Fig. 8).
So though all three sites have similar technologies they use those
technologies in different ways to produce different blanks and
then choose different blanks to use. These choices are not related
to chronological period.
It has been assumed that technologies developed in chronological
sequence. In the post glacial period the Mesolithic is considered
synonymous with bladelet production, followed by flake production
in the Neolithic and Bronze ages. The example of Kvernepollen
demonstrates that blade production was also carried out in the
Bronze Age as well as in the Mesolithic. Also flake technology
is used throughout the post glacial period. For the three sites
mentioned here, the relative amount of flake technology remains
consistent (Farsund 30%, Kvernepollen 30%, Raunds 31%). There
may be a general trend towards flake production but a simple linear
technological sequence is not the case.
Function
The next stage in the operational sequence is site function, including
the specific use of tools, both of motion and worked materials
(scraping bone, whittling wood, etc.), the identification of activities
(hunting, hide processing, etc.), and the interpretation of site
type (hunting camp, home base, etc.).
At Kvernepollen, 21 pieces had recognisable use-wear and Farsund
had 48. In terms of site use it is interesting that there is a
difference between Farsund and Kvernepollen even though both sites
are coastal and exploiting similar natural resources. Kvernepollen
has a limited range of activities (Fig. 9), consisting of processing
wood and fish. It is suggested that the projectile points were
not used in conjunction with the site (i.e., it is not a kill
site), because of the distribution of the projectile points (see
below). Farsund has a wide range of activities including processing
wood and fish but also working bone, antler, hide and one case
of scraping shell. This produces a different functional configuration
(see Grace 1990), to
Kvernepollen (Fig. 10).
Farsund is interpreted as a temporary home base because of
the representation of a spread of activities. From this and the
size of the site it was probably occupied by an extended family
group. Kvernepollen represents a small group, possibly only one
or two individuals, occupying the site for a short period during
a hunting expedition.
Discard
The final stage is the discard of the material seen in knapping
concentrations and clearance of areas with accompanying 'dump'
areas. Curation of tools would be included at this stage in that
the removal of material from the site, for use elsewhere, constitutes
a form of discard.
By incorporating use-wear analysis into the chaîne opératoire,
activity areas can be located from the discard of used tools,
These activity areas can be used to interpret the use of space
on the site which can indicate such things as social differentiation
within the group. For example, the location of specific activity
areas is a prerequisite for the assignment of gender roles. With
undisturbed sites such activity areas can be isolated. The early
Mesolithic site of Three Ways Wharf in England has a concentration
of tools used for adzing/chopping wood (Fig. 11), and can only
be an activity area as the pieces used for adzing/chopping wood
are of various types consisting of; 1 broken
ax, 2 axe/adze re-sharpening flakes, 8 flake
end scrapers and 1 blade end scraper.
Other tools of these types are distributed throughout the site
and either have other functions or are unused, so that the concentration
is only related to the activity of adzing and chopping wood. The
adzing/chopping wood concentration is clearly separate when seen
in contrast to the distribution of all used pieces from the site
(Fig. 12 and see Lewis,
in press).
The spatial distribution of material at Kvernepollen shows
clear concentrations of knapping debris with a separation between
the quartzite (Fig. 13) and flint (Fig. 14), possibly representing
separate knapping episodes.
These concentrations imply that no clearance has taken place.
The Farsund site has no such concentrations (Fig. 15), which could
be due either to post depositional movement or by being spread
about by prehistoric activity, which would imply longer occupation
than the 'undisturbed' concentrations at Kvernepollen.
The most striking aspect of the distribution of the used pieces
at Kvernepollen is that the majority of the projectile points
are outside the main concentration, some being several metres
away (Fig. 16).
The 'discarded' nature of the projectile point distribution,
with the absence of butchery, suggests re-tooling. That is, replacement
of projectile points and the repair and/or manufacture of arrow
shafts. The presence of considerable flint knapping perhaps represents
the manufacture of new projectile points in flint that were then
taken away as new arrows from the site. This would partially explain
the amount of knapping and the lack of used pieces. The remaining
used pieces are near the centre of the concentration adjacent
to the hearth and probably constitute a single activity area.
At Farsund there is no significant difference between the distribution
of the used pieces and the distribution of all lithics (Fig. 17).
Conclusions
The 'chaîne opératoire' approach provides a more
complete picture of differences and similarities among human groups,
rather than concentrating on single elements. With these sites
typological analysis results in a very small data base, technology
shows no significant difference, spatial analysis is difficult
to compare because of the lack of patterns at Farsund. However
by combining all these elements within the framework of the operational
sequence a greater understanding of the cultural differences can
be sought.
Figure 18 summarizes the 'chaîne opératoire' approach.
The four basic links are raw material procurement, technology
(separated into primary and secondary reduction and typology),
use and discard.
However the sequence includes feed-back loops in that, for
example, the intended use of tools will affect the choice of technology
and raw material. If the intent was to make projectile points
for hunting this could influence the choice of technology. For
example, tanged A points (see
Helskog et al. 1976,
26), are made on blades, so that blade technology would have been
chosen to produce the suitable blanks. Also function need not
be limited to utilitarian use. If the intention was to make a
flint copy of a bronze dagger
for a status or ritual function, the procurement of raw material
might be by trade, in order to obtain large pieces of good quality
flint with which to make such a dagger. Thus the intended function
of the tool can influence technology and the method of raw material
procurement.
The dotted lines in Figure 16 indicate the kind of interpretations
that can be made from the various elements of the operational
sequence. Figure 19 indicates and compares the interpretations
from Kvernepollen and Farsund.
These differences may reflect different social structures and
subsistence strategies employed during different periods. If one
assumes a larger population in the Bronze age then a pattern emerges
of larger base sites from which small groups (or individuals)
go out from on hunting/fishing expeditions, as opposed to small
family groups moving around together. These different strategies
are chosen as the preferred means of exploiting similar environmental
resources, rather than being adaptations to different environments.
Basing ethnic or chronological divisions on typology and/or technology
alone is a crude devise that ignores much of the evidence of choice
that reflects the social structure of human groups. Archaeological
sites are the product of dynamic interaction between individuals
within the social group, rather than static structures to be simply
classified by typological lists or by measurement of debitage.
This dynamic interaction can be studied with the chaîne
opératoire approach that allows for a greater understanding
of the complex human behaviour that lies behind the archaeological
data.
The analysis presented in this paper is the product of the author's
own research and does not necessarily agree with the excavators
of the sites. The methods used for the lithic analysis are explained
fully in Grace 1989
and Grace 1993.
Ballin, T. B. & Jensen, O. L.
1995
Farsundprosjektet- stenalderbopladser på Lista. Varia 29.
Universitetets Oldsaksamling
Bar-Yosef, O., Vandermeersch, B., Arensburg,
B., Belfer-Cohen, A., Goldberg, P., Laville, H., Meignen, L.,
Rak, Y., Speth, J. D., Tchernov, E., Tillier, A-M., and Weiner,
S. 1992. The Excavations in Kebara Cave, Mt. Carmel.
Current Anthropology, vol. 33, no. 5, 497-550.
Clark, J. G .D. 1954
Excavations at Star Carr. Cambridge University Press. Cambridge
Coulson, S. 1986
Refitted flint nodules from Songa, Telemark. Universitetets Oldsaksamling
Årbok 1984/1985. Oslo 1986
Gamble, C. 1986
The Palaeolithic settlement of Europe. Cambridge. Cambridge University
Press. 1986
Grace, R. 1989.
Interpreting the Function of Stone Tools: The quantification and
computerisation of microwear analysis. B.A.R. International series
474.
Grace, R., 1990, The limitations and
applications of functional analysis.
in The Interpretative Possibilities of Microwear Studies. Proceedings
of the International Conference on Lithic Use-wear Analysis, 15th-17th
February 1989, (eds. Gräslund, B., Knutsson, H., Knutsson,
K., and Taffinder, J.) Aun, 14, 9-14.
Grace,R. 1993.
The use of expert systems in lithic analysis. Traces et fonction:
les gestes retrouvés' Eraul 50, vol. 2, 389-400. Liege
1993
Helskog, K., Indrelid, S., and Mikkelsen,
E. 1976.
Morfologisk klassifisering av slåtte steinartefakter. Særtykk
fra Universitetets Oldsaksamling årbok 1972-1974 Helskog
et al. 1976
Lewis, J.S.C. in press.
A late glacial and early post glacial site at Three Ways Wharf,
Uxbridge, Middx, England.
Mellars, P. 1989
Major Issues in the Emergence of Modern Humans. Current Anthropology
v.30 no. 3 June 1989 pp. 349-385
Nærøy, A. J. 1994
Troll-Prosjektet: Arkeologiske undersøkelser på Kollsnes,
Øygarden K,
Hordaland, 1989-1992. Arkeologiske Rapporter 19. Arkeologisk Institutt,
Universitetet i Bergen
Roebroeks, W., Conard, N. J., and van
Kolfschoten, T. 1992
Dense Forests, Cold Steppes, and Palaeolithic Settlement of Northern
Europe.
Current Anthropology vol. 33, no. 5, December 1992
Stringer, C. & Gamble, C. 1993
In Search of the Neanderthals. Thames and Hudson.