Taylor, A. (2014) ‘Dating the Dark Ages: Using Material Culture to Explore

Cultural Identities in the Fourth to Sixth Centuries AD

Rosetta 15.5: 53 – 66

http://www.rosetta.bham.ac.uk/issue15supp/taylor.pdf

53

Dating the Dark Ages: Using Material Culture to

Explore Cultural Identities in the Fourth to Sixth

Centuries AD

Abigail Taylor

The transition between Roman Britain and Anglo-Saxon England is still a

contentious topic amongst archaeologists. The Adventus Saxonum is part of

the English national mythology, the origin story of the English people which

can be traced back to the writings of Gildas1 in the fourth century and Bede’s

The Ecclesiastical History of the English People2, written in the eighth century.

My research aims to explore what Portable Antiquities Scheme (PAS) data

can reveal about the material culture of this period and what implications this

has for cultural identity in the late Roman and early Anglo-Saxon periods. It

also aims to assess for how long and to what extent British peoples remained

culturally Romano-British. When did Britain, in particular western Britain,

cease to identify with Rome and become what some archaeologists still refer

to as the ‘Dark Ages’3?

Material culture is an important source of information about past societies.

Modern anthropological studies view artefacts as part of belief systems,

economic systems, social rankings and displays of wealth4. This type of

approach to material culture can be applied to archaeological artefacts to help

research aspects of identity and ethnicity. However, researchers must be wary

of the ‘culture history’ paradigm which equates material culture assemblages

with discrete social cultures in the past5.

The PAS has become an important research tool and the source of an

extremely large body of data which can be used for analysis. In 2001, Cool

1 Gildas, De Excidio Britanniae.

2 Bede, The Ecclesiastical History of the English People.

3 Faulkner and Reece 2002; Faulkner 2004.

4 Ember and Ember 1996.

5 Jones 1997: 51; Daniel 1975: 38, 41.

54

stated that ‘Currently the entries are of variable value, and it is imperative to

improve the quality of the poorer ones’6 . The value of the PAS as a source of

data was vastly increased when, in 2003, it received a Heritage Lottery Fund

grant allowing the database and website to be improved. Daniel Pett, the IT

advisor to the PAS, stated that the critical point for research would be when

half a million objects had been recorded7. This would provide an extensive

database for research purposes. The half a million object mark has now been

surpassed, with the PAS database containing a total of 963,717 objects within

619,879 records as of 24 June 20148.

Research into material culture has sometimes omitted PAS data from the

assemblages used for study. For example, Webb’s study of personal

ornament in the Roman north excluded PAS data on the basis that they had

no context9. As demonstrated by the number of finds in the database, this

omits an extremely large body of data from any study. The issue with PAS

data is that it needs to be approached differently to excavated data, and the

biases inherent in the data must be taken into account before it can be of any

use. Robbins notes that, although many studies have now included PAS data,

few have specifically discussed the methodology of using the database and

the constraints on the data which must be considered10. Webb did, however,

compare PAS material from Cumbria to the rural assemblage of Roman finds

in the north of Britain11. Proportionally, the assemblages were found to be

similar. There are now a total of 417 research projects listed on the PAS

website12.

These research projects often study one type of find or a specific area. My

study aims to make a national assessment of PAS finds and the implications

for identity in late Roman and early post-Roman/Anglo-Saxon Britain/England.

In this sense it builds upon research into sampling bias in PAS data by

6 Cool 2001: 6.

7 Pett 2010: 2.

8 PORTABLE ANTIQUITIES SCHEME 2014a.

9 Webb 2011: 21.

10 Robbins 2013a, 2013b.

11 Webb 2011.

12 PORTABLE ANTIQUTIES SCHEME 2014b.

55

Robbins13 and the contribution of PAS material to revealing previously

unknown sites in Roman Britain by Brindle14 .

The aims of the present research are as follows:

To assess how portable antiquities can be used for research into

ethnicity and identity in the fourth to the sixth centuries.

To investigate the material culture of Britain between the fourth and

sixth centuries using the PAS.

To discuss the methodology of research using the PAS, particularly

expanding on Katherine Robbins work into the distribution of PAS finds

and focusing on periodisation and differing distributions of finds across

the period in question.

To discuss how PAS data compares to conclusions based on

traditional, excavated finds.

When using the term ethnicity it is important to define what is meant, as

definitions used can vary. Many scholars do not define the term at all which

causes complications when it cannot be known which of the varying

definitions are being used. Isajiw15 surveyed sixty-five studies of ethnicity in

the fields of sociology and anthropology and found that only thirteen included

any kind of definition of ethnicity. There are two main schools of thought

regarding ethnicity. First, that common descent is the most important factor in

defining ethnic identity and second, that it is the way people see themselves

and the way they distinguish themselves from other social groups that is the

most important factor16. This research will use the second, more flexible,

definition of ethnicity. This approach can be seen in recent interpretations of

post-Roman archaeology. For example: Carver et al’s17 report on the

13

Robbins 2013a, 2013b. 14

Brindle 2014. 15

Isajiw 1974: 111. 16

Fenton 2010: 2; Regmi 2003: 3. 17

Carver et al. 2009: 3-4.

56

excavation of the fourth to seventh century cemetery at Wasperton. They

discuss ‘Anglo-Saxon’ and ‘British’ graves as reflected by the apparent

cultural affiliations rather than fixed biological definitions. Used in this way,

ethnicity can be taken to mean the ‘social and psychological phenomena

associated with a culturally constructed group identity’18 .

Most archaeologists now accept a continuity of population from the Roman to

post-Roman period19, but there are two conflicting ideas regarding the

continuation of Roman culture. Firstly, that Roman culture survived into the

post-Roman period20. The second theory is that there was a fifth century

break with the Roman past. This idea is taken to its extreme by Faulkner and

Reece21. Faulkner states that there is no evidence for a Late Antique period in

Britain and that the fall of the Western Roman Empire led to the Dark Ages in

Britain22. Others are proponents of this view to a lesser extent23. White24 takes

the middle ground, believing that the two seemingly opposing views can be

resolved to some extent with the survival of the province of Britannia Prima

(Figure 1) into the mid-seventh century. This research will compare the

material culture from the east and west of Britain in order to test the theory put

forward by White.

18

Jones 1997: xiii. 19

White 2007: 17. 20

Dark 1994, 2000. 21

Faulkner and Reece 2002; Faulkner 2004. 22

Faulkner 2004, 23

Wickham 2009; Esmonde-Cleary 1989. 24

White 2007.

57

Figure 1. The four provinces making up the Diocese of Britannia in the Late Roman

period25.

The first step in any assessment of PAS data should be to determine the

constraints on the data and the implication of collection bias. This is noted by

Robbins in her suggested methodology for studies of portable antiquities26.

The framework for the initial analysis of my data has been based on that used

by the Viking and Anglo-Saxon Landscape and Economy (VASLE) project27,

which researched landscape use and settlement in the late Anglo-Saxon and

Viking periods. They used base maps of constraints on finds recovery (Figure

2.) such as high topography, urban areas and areas where metal detecting

may not be able to be carried out. These can be compared to the distribution

of PAS data to distinguish modern patterns of recovery from historic patterns

of use28 .

25

White 2007: 60. 26

Robbins 2013a, 2013b. 27

Naylor 2006. 28

Richards et al. 2009.

58

Figure 2. Base maps used by the VASLE project to assess factors which may bias

data recovery29.

The first step in using PAS material is to clean the data. This includes the

removal of excess rows and finds from outside the UK and ensuring that

terminology is consistent. This research uses the PAS preferred terms30.

Following this, the total number of UK finds records in the database is

523,896. This is lower than the 607,193 records in the whole dataset since it

excludes data not yet validated. Of these records 95.51% have latitude and

longitude data used to map the find spots and 95.63% have a county, unitary

authority or metropolitan borough recorded. After filling in county information

using parish or grid reference data 96.81% then had a county recorded.

Once the data have been cleaned, the find spots can then be mapped and the

distribution analysed. The overall distribution patterns of PAS data (Figure 3)

shows that there are, in general, more finds in the east than the west. This

distribution (shown with a red line) matches Cyril Fox’s division into the

29

Richards et al. 2009. 30

PORTABLE ANTIQUTIES SCHEME 2013

59

Highland Zone, characterised by cultural continuity, to the west and the

Lowland Zone, characterised by cultural differences and replacement, to the

east31. Other constraints on finds recovery are demonstrated by areas in the

east with relatively low densities of finds. Coastal areas such as The Wash,

urban centres such as London and heavily forested areas such as The Weald

have fewer finds than surrounding areas. This distribution can then be

compared to period specific datasets to determine whether the patterns seen

are due to modern finds recovery and reporting or actual historic patterns of

use.

Figure 3. The distribution of the entire dataset plotted onto a topographic base map.

Red lines show the boundary between the Highland Zone to the west and Lowland

31

Fox 1933.

60

Zone to the east32.

The Roman finds (Figure 4) follow largely the same general distribution

pattern as the entire dataset. However, the Early Medieval finds (Figure 5)

diverge from this pattern somewhat. The division between Highland Zone and

Lowland Zone is still visible (marked by the red line), however, there is also a

clear divide between east and west, much more along the lines of the

assumed division between Anglo-Saxon England to the east and western

Britain (marked by the green line). This raises important research questions

regarding the reason for this difference. It could be that this division into

eastern England and western Britain are down to a general lack of finds from

the west. However, the difference in distribution between the Early Medieval

and finds from other periods still needs explaining. It is possible that dating is

a key factor. Due to the nature of PAS finds, they are often only dated by

typology. If Roman culture continued in the west of Britain then the material

culture used may remain largely the same as that used in the late Roman

period. With the collapse of the Roman economy people may also have kept

artefacts for much longer than in the past, leading to post-Roman finds looking

the same as late Roman finds. Such finds are still likely to be dated as Roman

(AD 43-410) rather than Early Medieval (410-1066).

32

Map produced in ArcMap using data from the PAS.

61

Figure 4. Roman finds distribution plotted onto a topographic base map. Red lines

show the boundary between the Highland Zone to the west and Lowland Zone to the

east33.

33

Map produced in ArcMap using data from the PAS.

62

Figure 5. Distribution map of Early Medieval PAS finds showing the boundary

between the Highland and Lowland Zones (red) and the boundary between high and

low densities of Early Medieval finds (green)34.

In conclusion, the PAS is an important research tool, possibly the most

important new body of archaeological data in the last thirty years. It represents

large number of finds that would otherwise go unstudied. However, it is

necessary to understand the constraints on the data and the bias inherent in

them.

34

Map produced in ArcMap using data from the PAS.

63

Initial analysis of the distribution of PAS material demonstrates that there is a

clear east-west divide visible in all PAS data. It must be considered that, to

some extent, any divide in Anglo-Saxon artefacts may be due to this. There is

also a divergence in distribution patterns between the Early Medieval period

and the entire dataset. This suggests that this may reflect actual, historic

patterns of use but may also be down to issues with dating PAS material.

List of images:

Figure 1. The four provinces making up the Diocese of Britannia in the Late

Roman period.

Figure 2. Base maps used by the VASLE project to assess factors which may

bias data recovery.

Figure 3. The distribution of the entire dataset plotted onto a topographic base

map. Red lines show the boundary between the Highland Zone to the west

and Lowland Zone to the east.

Figure 4. Roman finds distribution plotted onto a topographic base map. Red

lines show the boundary between the Highland Zone to the west and Lowland

Zone to the east.

Figure 5. Distribution map of Early Medieval PAS finds showing the boundary

between the Highland and Lowland Zones (red) and the boundary between

high and low densities of Early Medieval finds (green).

64

Bibliography:

Bede, The Ecclesiastical History of the English People, trans. J. McClure and

R. Collins [Oxford World’s Classics] (Oxford 1994).

Brindle, T. 2014. The Portable Antiquities Scheme and Roman Britain.

London: British Museum Press.

Carver, M., Hills, C. and Scheschkewitz, J. 2009. Wasperton: A Roman,

British and Anglo-Saxon Community in Central England. Woodbridge: Boydell

Press.

Cool, H. 2001, ‘The Iron Age and Roman finds in the Portable Antiquities

database’, in G. Chitty (ed.) Review of the Portable Antiquities Pilot Scheme.

Carnforth: Hawkeshead Archaeology and Conservation.

Daniel, G. 1975. A Hundred and Fifty Years of Archaeology. Trowbridge and

Esher: Redwood Burn Limited.

Dark, K. R. 1994. Civitas to kingdom. British political continuity 300-800.

Leicester: Continuum International Publishing

Dark, K. R. 2000. Britain and the end of the Roman Empire. Stroud: The

History Press.

Ember, C. R. and Ember, M. 1996. Cultural Anthropology. 8th Edn. Upper

Saddle River, NJ: Pearson.

Esmonde-Cleary. A. S. 1989. The Ending of Roman Britain. London and New

York, NY: Routledge.

Faulkner, N. 2004. ‘The Case for the Dark Ages’, in R. Collins and J. Gerrard

(eds.), Debating Late Antiquity AD300-700, BAR Brit Ser 365, Oxford:

Archaeopress, 5-12.

65

Faulkner, N. and Reece, R. 2002. ‘The Debate About the End: A Review of

Evidence and Methods’, Archaeol J 159, 59-76.

Fenton, S. 2010. Ethnicity. 2nd Edn. Cambridge: Polity.

Fox, C. 1933. The Personality of Britain: Its Influence of Inhabitant and

Invader in Prehistoric and Early Historic Times. 2nd Edn. Cardiff: National

Museum of Wales.

Gildas, De Excidio Britanniae, The Ruin of Britain, and other works, trans. M.

Winterbottom [Arthurian Period Sources], 7 (London 1978).

Isajiw, W. W. 1974. ‘Definitions of Ethnicity’, Ethnicity 1, 111-124.

Jones, S. 1997. The Archaeology of Ethnicity: Constructing Identities in the

Past and Present. London and New York, NY: Routledge.

Naylor, J. 2006, ‘Viking and Anglo-Saxon Landscape and Economy’,

http://www.york.ac.uk/archaeology/vasle/index.html (accessed 30 July 13).

Pett, D. 2010. ‘The Portable Antiquities Scheme’s Database: its development

for

research since 1998’, in S. Worrell, G. Egan, J. Naylor, K. Leahy and M.

Lewis (eds.). A Decade of Discovery: Proceedings of the Portable Antiquities

Scheme Conference 2007. BAR British Series 520, Oxford: Archaeopress,

156-166.

PORTABLE ANTIQUITIES SCHEME. 2013, ‘Controlled Terminologies’,

http://finds.org.uk/database/terminology/ (accessed 03 Jul 13).

PORTABLE ANTIQUITIES SCHEME. 2014a, ‘Database’,

http://finds.org.uk/database (accessed 24 Jun 14).

PORTABLE ANTIQUITIES SCHEME. 2014b, ‘Research based on Scheme

data’, http://finds.org.uk/research (accessed 24 Jun 14).

Regmi, R. 2003. ‘Ethnicity and Identity’, Occas Pap Sociol and Anthropol 8, 1-

11.

66

Richards, J. D., Naylor, J. and Holas-Clark, C. 2009. ‘Anglo-Saxon Landscape

and Economy: using portable antiquities to study Anglo-Saxon and Viking Age

England’, Internet Archaeology, 25, http://dx.doi.org/10.11141/ia.25.2

(accessed 20 Oct 13).

Robbins, K. 2013a. ‘Balancing the Scales: Exploring the Variable Effects of

Collection Bias on Data Collected by the Portable Antiquities Scheme’,

Landscapes 14, 54-72.

Robbins, K. 2013b. From past to present: understanding the impact of

sampling bias on data recorded by the portable antiquities scheme. Ph.D.

Thesis. University of Southampton: UK. Available online at

http://eprints.soton.ac.uk/id/eprint/360475

Webb, T. 2011. Personal Ornamentation as an Indicator of Cultural Diversity

in the Roman North. Oxford. BAR Brit Ser 547, Oxford: Archaeopress.

White, R. 2007. Britannia Prima: Britain’s Last Roman Province. Stroud:

Tempus.

Wickham, C. 2009. The Inheritance of Rome: A History of Europe from 400 to

1000. London: Penguin.