In 1893, the Viking, a replica of the Gokstad ship (ca. 900), was sailed from Norway across the Atlantic to the World Columbian Exposition in Chicago. The Viking was an accurate attempt for the time at reproducing a tenth-century ship, and its voyage was meant to recall the Viking landfall in North America four centuries before Columbus’s “discovery” of the “Indies.”

Another impudent replica of a medieval Scandinavian ship, the Saga Siglar, made a similar passage in 1992 to mark the half-millennium of Columbus’s venture. Those who witnessed either of these audacious Scandinavian salutes to the Genoese admiral would have been aware that these voyages were also tests of earlier technologies. The landfalls in 1893 and 1992 marked the conclusion of technological trials, archaeological experiments of a sort.

Experimental archaeology is the controlled replication, reproduction, or reconstruction of an artifact or a series of related artifacts and the processes involved. This strict definition implies intentionality and a program behind the act, as well as a certain critical stance towards the evidence, the experiment, and the results. Remove “controlled,” and the definition could embrace the full range of educational activities undertaken at on-site museums, the burgeoning archaeological theme parks, and by re-enactors of various levels of science and competence.

Some of their work is laudable and even useful, but all too often it descends to the level of costumed mock-Tudor “wench pinching”: pitiable, risible, and worthless. There is currently debate in archaeological circles regarding the desired degree of inclusivity of “experimental archaeology.” Rather than discuss the demerits of costumed bumbling, I would rather offer an introduction to best practice in scientifically conducted experimental archaeology, in the conviction that it can yield useful results in a variety of fields. To those primarily interested in experimental archaeology as an educational device, I simply counsel: the more rigorous the approach, the better the quality of the demonstration.

Experimental archaeology has seen its prime and earliest use outside medieval studies. For volume of work and antiquity of practice, the study of prehistoric lithics has precedence (dating at least as far as the early eighteenth century), followed by prehistoric and ancient pottery, architecture, agriculture, metallurgy, and marine archaeology, to name a few of the most active areas.

Among specific projects are Don Crabtree’s classic work on Folsom points, the trireme Olympias project, the agricultural and architectural experiments conducted at Butser Ancient Farm (http://www.butser.org.uk/iafres_hcc.html1), or the wider range of activities undertaken at the Lejre Centre in Denmark (http://www.lejre-center.dk/). Valuable work is done at both institutions, but some of their more popular activities sail perilously close to experimental archaeology as entertainment) .

Turning to medieval studies, experimental archaeology has been applied to marine archaeology, ceramics, architecture, and metallurgy, among other fields. Here one can cite the reconstructions and trials of the eleventh-century Skuldelev ships (http://www.abc.se/~m10354/uwa/skuldele.htm), or work associated with the Medieval Pottery Research Group (particularly that of R. W. Newell: http://www.medievalpottery.uk7.net/).

The general method of experimental archaeology is fairly straightforward, and can be presented under thirteen points:

1. materials used should be those available to the original craftsmen or craftswomen;

2. techniques should be those of the time, place, and activity under investigation;

3. modern technicians should not be incompetent, inexperienced, or inexpert as regards 1 and 2 above;

4. modern technicians must be fully informed of the goals of the experiment, and be sympathetic to its aims, unless the experiment design requires them to be uninformed or unsympathetic;

5. observing 1 and 2 above, modern materials and technologies should play as little a role as possible in the experiment. Where they are necessary, their use should be controlled and recorded;

6. parameters, qualifications, and limits to an experiment should be clearly formulated;

7. experiments in sequence should be consistent, where possible, for the sake of ease in making significant comparisons and observing cumulative effects;

8. experiments should be developed and run with reference to previous trials;

9. all possible ways of achieving the result, provided they are in accordance with 1 and 2 above, should be investigated or at least noted;

10. the experiment should be reproducible, where appropriate;

11. results must be stated as accurately as possible, with all necessary qualifications, chief among which is that a successful experiment provides only one possible way something may have been done;

12. the experiment must be published as fully, transparently, and rigorously as possible. Prompt publication is desirable, but ought not to be achieved through the sacrifice of rigor, transparency, or fullness; and

13. evidence from other cultures can be introduced in a controlled way when interpreting results, or even earlier, in the planning and design stages of the experiment.

It will not always be possible to satisfy fully all thirteen points, and it is best to adapt the method to the question and the nature of the materials investigated; a critical method should never be a straightjacket. An experimental archaeological approach to some aspect of the performance of medieval music or theatre will have to develop a flexible and usable requirement for reproducibility (point 10), which will be quite different from that obtaining in an experiment testing some facet of copper-alloy casting, where a more routine criterion of reproducibility can be demanded.

Points 1, 2, and 5 can be modified depending on the design of the experiment. If the goal is to evaluate some aspect of an object in use, but the method of manufacture of that object is neither crucial to the experiment nor an experimental variable, then it may not be necessary to use culturally appropriate tools and techniques in that object’s fabrication. It is important to be clear about one’s decisions and to pay particular attention to experiment design.

It is important to emphasize the limits to even the most thoughtfully designed, carefully controlled, competently executed, and responsibly reported archaeological experiment. It can rarely, if ever claim to be the discovery of the method used, rather than a possible method, even if the resulting object is identical to an extant artifact.

Then there is the problem of what might be termed the “experiential gap.” Through the benefits of a societal tradition, a thorough apprenticeship, and early and continual application, the medieval craftsperson possessed a familiarity with the materials and an ease of technique surely denied to all but the most skilled and practiced modern experimenter.

How can such a gap be bridged, and what effect does it have on the conduct of an experiment and the interpretation of its results? What about the problem of the wider context of an artifact and its use? Our conceptual world is different from that of a fifteenth-century scribe or an eleventh-century shipwright; can we truly understand an object, its manufacture, and its use without taking account of the object as symbol?

Even with full cognizance of its limitations, experimental archaeology offers research benefits that are otherwise unavailable. It allows one to falsify a theory in a very concrete fashion. Even more valuable, the actual process of making or using an object will effectively reveal what we incompletely or erroneously understand about those things. This may also indicate the sorts of evidence for which to sift.

It is not uncommon to find modern scholarly accounts of some aspect of quotidian life in the Middle Ages which seem vague or untested. Anything that could lead to less obscurity is to be welcomed. Beyond the disciplines in which it is habitually used, experimental archaeology could also benefit, for example, those of mysticism, liturgy, theatre, performance practice, rhetoric, law, communications studies, transportation, mathematics, numismatics and sigillography, oral-formulaic studies, philosophy, palaeography, and codicology.

R. A. Rosenfeld works on the technologies of communication in the Middle Ages (in lato sensu). He is based at the Pontifical Institute of Mediaeval Studies.