Home | Main Menu | Castle Index | Historical Essays | Recommended Castles | What's New | Links

Castles Are Rubbish

copyright © by Daniel Mersey

Using Material Evidence From Excavations
to Give an Overall View of Daily Life

The title of this article is not my poisoned little view on medieval architecture - it is a comment that nicely sums up the last use of the majority of recovered artefacts at castle sites. As we shall see further on in this piece, most of the valuable items in a castle were removed upon its' abandonment, leaving the site as a glorified rubbish tip - only unwanted, broken and lost items were left behind for the modern archaeologist to recover.

Below right: a 1996 view showing the excavation of the North Range at Dolforwyn Castle.
Photograph copyright © Daniel Mersey

In archaeological investigation, the assessment of material evidence is one of the key factors in providing our ideas of the past; despite the fact that the archaeology of the castle is a study within its own right, it must recognise this method of research just as in any other archaeological period.

Kenyon (1990) outlines the more usual aspects of daily life which can be recorded through archaeology at castle sites. His examples are not exhaustive, as he realises, but they are the most common kinds of evidence: castle building, militaria, household objects, costume, pastimes, food and diet. This list is little different to that expected from a site of any period - but by realising each period's artefact advantages and limitations, it is possible to look at each in a unique way.

Perhaps one of the most important advantages that the medieval archaeologist has when considering material evidence is the fact that his sites are only between 400-1000 years old; this means that a much higher standard of preservation is likely to occur than on sites from an earlier period. It is perhaps the quality of remaining evidence that is important here - objects are less corroded the more recent that they are (as a general rule), meaning that a more detailed identification can be made. Similarly, objects which one would expect to be absent on chronologically earlier sites may still be present (such as nails surviving intact but corroded, as opposed to surviving only as a metallic stain).

Linked to this idea is the fact that due to the more recent dates of castle sites, objects recovered can usually be dated to a closer timescale (of say, thirty years) than in earlier periods. For example, more precise styles and typologies of pottery are known for the medieval period than earlier ones, so closer dating can be given. Artefact dating, however, is also one of castle archaeology's major limitations - the fact that artefacts may only be able to be dated to within perhaps thirty years is a problem when attempting to give precise enough dates to be of relevance within the castle's history. This is because the investigation of castles is dealing with a period closer to the modern period than most within archaeology, and therefore more is known of the medieval period.. The changes that could take place within a well documented thirty year period are considerable, and bearing in mind that key dates within a castles history may already be documented, a thirty year span is not always acceptable for defining a chronology.

Below right: south entrance of Dolforwyn Castle under excavation, 1994.
Photograph copyright © Daniel Mersey

The availability of documentary evidence does not prevent the study of castles from an archaeological point of view, though. Documents are very useful to build a basic chronology from, and can add details which cannot be learnt through excavation alone; for example, the Welsh castles of Edward I are well documented, and not only can we tell their building history, but the expenses, architects and even from where the labourers came is all known (Taylor, 1963). Other documentation may or may not be able to be proved through excavation - more often than not, as with the vast majority of archaeology, a straight forward "true or false" answer may be hard to decide, but excavation may produce evidence to support a documented date (as long as the dating problems discussed above are borne in mind). Further documentation may list the buildings present within a castle (such as the 1321 survey of Dolforwyn), and the material evidence may be able to suggest the exact location of these (such as with the kitchens at Dolforwyn).

The material evidence present may even allow us to designate a certain usage to a room without any documentary evidence: a large bone assemblage may indicate the castle's kitchen, as can hearths, and possibly wells (such as at Middleham, North Yorkshire). These last two items indicate to us that not only household objects, but also structural remains can be used as evidence. We can also usually judge the social standing of the occupant in residential rooms by such methods - the size of the fireplace, quality of the decor (either stonework, or even painted plaster, where recovered), and the amount of access can indicate this also (higher quality rooms usually had more limited access, often through the hall only - Goodrich being such an example). Despite Kenyon's claims (1990), militaria is less often found than one may imagine for a castle - as mentioned in my introduction, the most expensive and useful items inside the castle would be removed when it went out of use, and this would include most military equipment. Occasional finds might include arrowheads too blunt to reuse, crossbow mechanisms too cracked to repair, or swords snapped and discarded.

The removal of items of use and items of value when a castle was abandoned is a further limitation to castle studies; unless the castle was left littered and dirty whilst in use (conceivable, especially in courtyard and service areas), the evidence uncovered by excavation produces only unwanted, and quite often later material - as opposed to a good sample of objects spanning the castle's history (unlike other period's sites, where unwanted but more diversely dated material can be found). Exceptions to this can, of course, be found, notably in waterlogged deposits and environmental evidence: research into the waterlogged drains at Barnard Castle in northern England recovered a wide variety of evidence spanning a wide period of time (Kenyon, 1990: 180).

Linked to the problems of data recovery, we must also remember that a ruined castles' stratigraphy (the idea that a layer of soil underneath a higher one must be from an earlier period) is not straightforward; the layers of deposit are most often tumble (i.e. they have fallen from a wall, upper floor, or similar higher level), and this means that to some extent, the stratigraphy of the castle is reversed, or in the worst cases entirely scrambled (a wall may fall all at once, spreading several vertical layers as one horizontal layer, or it may fall piecemeal, disrupting stratigraphy completely). Added to the problem of tumbling masonry is the fact that tumble is often quite vacuous, allowing smaller objects to work their way to a much lower level; from my own experience, I know all too well that post-medieval artefacts can slip down into medieval layers - a problem not helped when robber trenches and earlier excavations have taken place...

Below right: The vaulted cellar of Dolforwyn Castle viewed from the north, 1997.
Photograph copyright © Daniel Mersey

A further problem regarding the actual structures on a castle site is that when later buildings have been added, altering or destroying the original features. A very good example of this is the motte at Caernarvon, incorporated into the Edwardian castle, and completely removed during the 1880s. Non-intrusive research techniques (i.e. geo-physics) can be of help here, as can architectural historians in cases where the buildings are still standing. The problem of multi-period buildings is, however, less of an obstacle with regard to the standing remains and more of an obstacle to deciding what lies under them, or what they may have replaced. Documentation may again be able to help the archaeologist with this.

Finally, the area covered by excavation, and therefore the limit of material evidence recoverable can become a limitation; as we have noted, the artefacts recovered within a castle's walls may not represent the occupation period fully unless the site was unclean; a higher level of material culture can be gained from surrounding ditches and moats - the most convenient rubbish tip from the castle's walls. It is therefore useful to examine these, in addition to the more impressive castle interior if excavation time and funds allow, to gain a better picture of daily life within a castle.

Overall, it can be seen that castle archaeology has it's own unique problems and research methods; the fact that the term "castle" can incorporate anything ranging from a small eroded motte, up to the largest and latest fortifications taking the form of residences as well as defences, makes these factors more diverse - and some would clearly be of more importance than others, depending upon the site (such as the dating of pottery found on a post-Conquest short occupancy motte, to that found at a site spanning several centuries, such as the tower of London).

Daniel Mersey



Cathcart-King, DJ 1988, The Castle in England and Wales
Kenyon, JR 1990, Medieval Fortifications
McNeill, T 1992, English Heritage Book of Castles
Taylor, AJ 1963, The Welsh Castles of Edward I


Additional articles by Dan Mersey

Home | Main Menu | Castle Index | Historical Essays | Recommended Castles | What's New | Links

Copyright © 2009 by Daniel Mersey and the Castles of Wales Website