Documents, Historiography and Material Culture in
Historical Archaeology

James Deetz and Patricia E. Scott

Consideration of context, local, regional, national or even international, is critical to
explanation in historical archaeology. As archaeologists we can and should use site specific
documentary material to construct the lifeways represented by the archaeological data from
sites under investigation, but all too often such documentary data are simply combined with
the site specific archaeological data and little is done to establish the broader context in which
both records were formed. We are all familiar with situations where the documentary record is
thin or non-existent, as is the case in some Virginia counties which lost their records in the
Civil War, or where for some reason the occupants of a site left no written trace of their
passing, slaves being among the most common example. Yet the people who occupied sites
which have little documentary evidence are not lost to history, for they did leave a record, in
things rather than words, which can be understood if one only considers the broader history of
the region, or even nation, in which they lived. People did not live in complete isolation, but
were subject to forces and events which took place far beyond their immediate world, and the
recollection of this fact enables us to look for explanations which transcend confirmation of
the site specific documentary data.

As a modest example, consider for a moment wine bottle seals which bear the initials
of their owner. We have all attended illustrated presentations where a slide of such a seal is
shown, and the point made that this is evidence that indeed the site where it was found was
occupied by the individual whose initials were molded in glass, although the documentary
record had already amply demonstrated that fact. Little new is provided by such an artifact in
this case. On the other hand, a wine bottle seal can add to our knowledge of not only the site
which produced it, but also of events in the larger world of which the site was a part. Consider

a seal marked "John Hood 1751," recovered from a trash pit on the largest site at Flowerdew

Hundred Plantation in Virginia. Hood was known to have been one of five identified owners
of half-acre house lots in what was known as Powhatantown, located somewhere on the
Flowerdew Hundred property. This town was one of a number of communities which were
planned by the General Assembly to move trade from the private plantations to public centers.
While the plan seems not to have been fully realized, something was established at Flowerdew
Hundred, and the John Hood seal would seem to indicate that it occurred at the site where it
was found. Since the site is the largest by far on the property-at least over five acres in
extent-and the dates based on the artifacts match the known facts, the discovery of the Hood
seal has made it quite probable that the location of Powhatantown has been established. We
know very little about how such communities were laid out, and further archaeology at this
location is likely to be very informative on the matter. So it is that the John Hood seal

James Deetz and Patricia E. Scott, Department of Anthropology, University of Virginia, Charlottesville,
VA 22903

Documents, Historiography & Material Culture

connected a specific site to the implementation of government policy, one that applied over the
entire colony.

This example is just that, and nothing too profound is intended in offering it, but it
does serve to demonstrate that the same kind of object can function in different ways
depending on how one goes about using it. The connection of site specific data of both types
is, in a sense, a "horizontal" operation, relating two bodies of material which reside at the same
level of specificity. Relating either to the larger context in which it is situated is by contrast
"vertical", in that it involves moving up to a higher level of generality. In this respect, it is
rather like, if not identical to, what is referred to as "Middle Range Theory" in archaeology.
We are not suggesting that this kind of thinking about data is never done, but rather that it
could be done more frequently, and indeed, given a choice between "horizontal" linkages and
"vertical" ones, one should opt for the latter, for it is at the more general level that one can
begin to speak of culture, that outmoded but still useful concept that we feel should shape and
guide our explanation of the material with which we work.

A much more substantial example of how such connections can be made is provided
by a series of sites at Flowerdew Hundred. While the results of this analysis have been
published (Deetz 1988, 1993) it is still useful to emphasize just how they were arrived at. It is
also important to note that these sites have very sparse documentation, but yet, taken together
in the context of Chesapeake history, they can inform us about their anonymous former
inhabitants and the way they were subject to the forces of events and circumstances. The
eighteen seventeenth and earlier eighteenth-century sites in the Flowerdew Hundred
bottomlands can be grouped into three sets, based on shared Harrington-type histograms.
These sets are clear and unambiguous, and the earliest and latest contrast with the intermediate
one in having a different spatial distribution pattern, so that settlement pattern further validates
the sets. The earliest group of sites, seven in number, show settlement fairly evenly spaced out
along the bottomlands. All were occupied during the second quarter of the seventeenth
century, and all were abandoned at roughly the same time, shortly after mid-century. The six
sites in the second group are located near the river at the northern and southern ends of the
bottoms, and were occupied for a longer time, overlapping the sites of both the first and third
group. The third group of sites, five in number, show a spatial distribution more like those of
the first group, and, like them, a briefer period of occupation. Even if there were ample
documentary material for these sites, it is doubtful that an explanation of this pattern in time
and space would have been arrived at without placing them in a wider context. With the little
documentation available, it is virtually certain that any explanation would not have been
forthcoming without considering the history of the Chesapeake as a region. The dates of site
groups one and three match two important events in Chesapeake history. The abandonment of
the first group of sites occurred at a time of falling tobacco prices and it seemed as though
there might be a connection between the two. The build up of occupation in the third period
coincides with the time when the importation of African slaves increased sharply. These two
causes were, thus, tentatively advanced to explain the two groups of sites, and formed the basis
for looking at the archaeological record for further support and substantiation. Such was
forthcoming in both instances. Group one sites that have been excavated all show earthfast
construction, a building type amply demonstrated to have a strong association with tobacco
monoculture (Carson et. al., 1981). Every one of the five sites of group three has produced

Deetz & Scott

III

Kroeber Anthropological Society Papers

examples of Colono ware. While there is still some controversy surrounding the identity of
makers of this pottery, there is general agreement as to who used it, and those people were
slaves of African origin. The intermediate set of sites, showing a longer occupation, seemed
different from their contemporary sites of group one in that they suggest a more permanent
mode of settlement, not tied closely to the production of tobacco. Two of the six sites have
been excavated, and in each case, a very different type of building was encountered, as well as
some evidence of manufacturing activities. While a causal relationship had been established
between the site specific data and regional history, what could not be determined was the
degree to which the people at Flowerdew might have shared in this pattern with other
communities in the region. Studies of pipe stem histograms at Martin's Hundred, Virginia
(Edwards and Brown, 1993), have shown rather convincingly that they did indeed share the
pattern, and that what is seen in the two places may well be a regional pattern of settlement
and demographic change.

An even broader historical perspective and context was necessary in attempting to
explain the contents of an ice house at Flowerdew Hundred, abandoned and filled with refuse
sometime in the late 1820's. The explanation borders on speculation, but has considerable
merit nonetheless. The ice house is but one of a number of similar features encountered along
the eastern seaboard from New Hampshire to Georgia. All are characterized by rich fill
comprising large numbers of fully restorable items, usable tools, building debris and faunal
materials showing varying degrees of weathering. Certainly site specific explanations are not
powerful enough to account for these features; they all appear to date to almost the same time,
and do not occur in the archaeological record from dates earlier or later than that time. They
indicate that at least some people were disposing of quantities of serviceable items, remodeling
their houses and cleaning up the house lots at about the same time. Seeking an explanation for
this phenomenon certainly required moving beyond site specific archival data, although it has
been suggested that some major "life event" such as marriage, death or property transfer might
have been the causal agent. However, such an explanation cannot account for the near
contemporaneity of such dumping events in a number of places at widely scattered locations.
Earlier attempts at explanation were little more successful when a regional or even national
scale was used, and it was not until the context was broadened to one that involved comparing
the history of two English frontiers that any convincing explanation, however speculative,
could be suggested. A dump sharing all of the features with those in America was excavated in
South Africa at the village of Salem, situated some fifteen miles from Grahamstown, capital of
the eastern Cape, on the nineteenth-century English frontier. The date of the deposit is circa
1870, and there are definite indications that others are awaiting discovery, dating to about the
same time. The question that was asked of these data was how the 1820's and 1870's might
have been significant dates in America and South Africa respectively.

One strong possibility that would take into account the material in both places is that
each date is some fifty years-two generations-following a kind of separation between
England and her colonies on the two continents. The separation was political in America, the
outcome of the Revolutionary War. In South Africa, while still affiliated politically with
England, the settlers, who arrived as a group of some 5,000 in 1820, were far more isolated
and cut off from the Mother Country than were the Americans, even after the Revolution. In
both places, there would have been a group of people reaching maturity who were never

112

No. 79

Documents, Historiography & Material Culture

English in the stricter sense, but rather the carriers of a culture combining elements of both
England and the colony of which they were a part. If this explanation is at all tenable, what the
two dates indicate is that time when the colony ceased being a simple extension of England,
and became a distinctive colonial culture in its own right.

So it is that a single South African refuse deposit, with a very close resemblance to a
number in America, permits a tentative explanation of events in both places. But given that,
the far more persuasive American evidence also lends support to the explanation as it applies
to South Africa. It remains then to look at other aspects of South African archaeology and
material culture to see if they too are consistent with the explanation advanced. Excavations in
the front yard of the lot on which the Salem dump was located provided clear evidence that
there had been significant landscape modification in circa 1870, in the form of raising the level
of the ground evenly across the entire lot by about a foot. Furthermore, at about the same time,
a major addition was made to the house in the form of a two-thirds I house, almost certainly
producing some of the structural debris-mortar, plaster, roofing material-in the dump fill.
At the Observatory site (a Cape Regency building remodeled to an elaborate multistoried
store/dwelling house) in downtown Grahamstown, excavations in the rear lot produced
evidence of both landscaping and remodeling of the building, also around 1870, although no
massive refuse deposit was encountered.

The slow process of Africanization of the British settlers can also be observed in
other material culture categories. While occasional hall and parlor houses are to be seen in the
area which were built in the early years of settlement, the most common form to develop was a
central hall I house as seen from the exterior, but one which has had one of the hall walls
omitted within, giving a hall and parlor type of spatial organization. In Grahamstown, this
unexpected modification of the symmetrical, ordered Georgian interior appears to have
continued in a number of houses until the 1860's to 1870's when a shift takes place and houses
are remodeled with cast-iron verandahs and elaborate window moldings, and new ones built in
various Victorian styles which included a central hall.

Another pointer to the process of change is seen in the modification of interior
domestic spaces (Scott, 1993). The majority of the settlers were from a working class
background, and in England their houses were characterized by the presence of a living
room/parlor-kitchen which for generations had been the hub of family life. On the English
Cape frontier, however, by the mid-century, a very different pattern of room usage was
emerging. Although the hall and parlor layout was retained behind the Georgian facade, the
kitchen was separated from the living space of the family, either by a partition wall being
erected, use of one of the other rooms, or a separate room being added to the building. This
can probably be attributed to a combination of two factors. First, with a readily available
supply of cheap labor in the form of indigenous people, the formerly working class settlers
became more mobile socially, creating interior spaces more similar, at least superficially, to
those of the English middle class as they were organized from at least the late eighteenth
century. The emerging kitchen on the British frontier in South Africa differed, however, from
its English counterpart in being very sparsely furnished, as shown by probate inventories.
Whether this difference is based in economic factors, or, as is more likely, attitudes which
developed toward a servant class from a completely different cultural background, is not clear.
If the latter, this is yet one more difference accountable to the settlers accommodating to a

Deetz & Scott

113

Kroeber Anthropological Society Papers

different cultural environment. A second development with regard to the spatial organization
of the English settlers on the eastern Cape frontier which is of relevance is that a pattern of
separate dining and parlor room usage was not emerging in the increasingly gentrified
households in Grahamstown. This non-development suggests that something else was going
on in the minds of these first generation English settlers, and that their changed pattern of
room usage, from living room/parlor-kitchen to parlor-dining room, reflected a continued
adherence to the corporate lifestyle which they had known for generations past in England.
Although the industrial revolution which, through a growing export market, was bringing to
the Cape furnishings and style guides which were fast gentrifying urban England, the English
colonists, despite their upward social mobility, were selecting a corporate, as opposed to
individual, use of interior spaces.

Ceramics provide another indicator of changes which were taking place on the
English frontier in the eastern Cape as one culture encountered another and became
increasingly distinct from its origins. The majority of the ceramics exported to the Cape and to
America were made in the English Staffordshire potteries. A preliminary comparison between
early nineteenth-century ceramic samples excavated in Grahamstown and Salem with those
found in America has shown that in South Africa, apart from a very early 1820's use of blue
and white wares, colorful wares predominated. This is in sharp contrast to the situation in
America where white molded wares become more common as the century goes on. The
colorful selection shows a variegated palette which is consistent with the corporate, shared
lifestyle reflected in the spatial modifications seen in the vernacular housing of the English
Cape frontier. Gravestones are equally diverse in material used, with slates in a variety of
hues. Certainly by mid-century, a very distinctive frontier culture was beginning to emerge,
which in its material aspects was very different from its American counterpart, as well as from
its English antecedent.

So it is that in a number of material culture categories we can see that the society of
the eastern Cape frontier became a very distinctive one, markedly different from both its
American counterpart and the parent society of Britain. The explanation put forth for this
difference derives from data that far transcend the site specific, and indeed, numbers of sites
had to be considered in terms of shared characteristics as they relate to broader aspects of
regional, national and even comparative international historiography.

It might be useful to coin a new phrase at this point, the "material event." Material
events can be of differing degrees of significance, and their significance is in turn a function of
the extent that the event was more than idiosyncratic. For example, if John Goddard broke his
chamber pot (an historical event), its presence in the refuse pit where it was thrown (a material
event), is of little or no significance in our construction of history through the use of
archaeological data. But when the planters of the Chesapeake embraced tobacco as their main
cash crop, that historical event produced a material event of major significance, the
construction and subsequent abandonment of hundreds of earthfast buildings throughout the
region. This relationship between historical and material events, occurring over an entire
region, allows us to say some things about Chesapeake history in the seventeenth century
which could not have been said in the absence of this relationship. Over-reliance on site
specific documentary data puts us at risk of dealing with material events of inconsequential

No. 79

114

Documents, Historiography & Material Culture

significance since these may well have been the product of historical events unique to a single
site and thus not reflecting in any way the larger course of events in which they took place.

Other examples of significant material events come easily to mind. The striking
similarity in layout of early seventeenth-century English frontier settlements in Ulster and
America speaks to shared concerns both military and commercial. The appearance of Colono
Ware over the American South in the later seventeenth century is a material event that can be
explained in terms of the historical event of changing social relationships between master and
slave. The difference in timing of stylistic replacement in New England gravestones allows us
to pinpoint with greater accuracy when and to what degree various communities were affected
by the Great Awakening. Other examples, discussed above, include the shared settlement
pattern suggested by pipe stem histograms at both Flowerdew and Martin's Hundreds and
mass dumping and remodeling.

In a sense, an archaeological site, or an old house, is simply a collection of material
events, some of which are seen again and again at other locations, and others which are not.
And while we should document our sites to the fullest extent possible, this is a means to an
end, and not an end in itself. Only when we distinguish between those material events which
are highly significant by virtue of being encountered over a wide area and thus the result of
larger scale social forces, and those which attest only to happenings in one place, can we begin
to make the most out of the true potential of the archaeological record.

References

Carson, Cary, Norman F. Barka, William M. Kelso, Garry Wheeler Stone, and Dell Upton

1981  Impermanent Architecture in the Southern American Colonies. Winterthur Portfolio

16:135-96.
Deetz, James

1988 American Historical Archaeology: Methods and Results. Science 239(22 January):362-7.

1993 Flowerdew Hundred: The Archaeology of a Virginia Plantation, 1619-1864. Charlottesville:

University Press of Virginia.

Edwards, Andrew C. and Marley R. Brown, III

1993 Seventeenth-Century Chesapeake Settlement Patterns: A Current Perspective from Tidewater

Virginia. In The Archaeology of ) 7th-Century Virginia. pp. 285-310. Theodore R. Reinhart
and Dennis J. Pogue, editors. Richmond: Archaeological Society of Virginia, Special
Publication No. 30.
Scott, Patricia E.

1993 Discovering British Culture Transformation in Grahamstown, 1820-1882. In A World Too

Wide: Essays on Libraries and Other Themes in Honour of Reuben Musiker. pp. 257-72.
Joseph Sherman, editor. Johannesburg: University of the Witwatersrand Library.

Deetz & Scott