Main Menu - Misc. - Clothing/Textiles - Medieval Wales - Names - Other Medieval - Publications - Harpy Publications

This page last modified May 22, 2005

Shepherds Purse: Experimentation and Presentation

Return to main page

Whether or not your main purpose in studying an artifact is to make and use simliar artifacts, this sort of "experimental archaeology" can be a vital step in understanding your topic. If you're researching it in the context of historic re-creation or re-enactment, then it's pretty much a given that your overall goal is to make something. But in the context of knowledge for its own sake, it is a valuable tool to see how your supposed materials behave -- whether they actually work the way your hypothesize they might. While the matter can be (and usually is) overdone, it is sometimes a valid criticism by "hands-on" amateurs that professional historians may stumble badly if they rely entirely on theories and ideas and don't understand the actual properties and behavior of the physical objects they're talking about. (To be fair, this primarily tends to happen when historians specializing in texts and literature branch out into material culture.)

So your next step is to try to make something that is both consistent with the appearance and behavior of your artifact in the data, and is a logically reasonable development within its historic context.

The results of my hands-on experimentation, are in the Reconstructions section.

Direct reproduction

Do you have enough information to make exact replicas of your topic?

People sometimes groan and grouse about making exact replicas of an artifact rather than being "creative", but if you do have the information to create an exact replica, you have a treasure beyond price and it shouldn't be disregarded. Going through the same processes with the same materials (and perhaps even the same tools) as were used for your model will teach you important things about why the object exists as it does. You may find your materials "wanting" to do something in a particular way that sheds light on an otherwise curious feature. Or you may find that some construction method that looked obvious and logical to you at the start turns out to have unpredictable difficulties.

In my current project, I don't have this luxury. The major stumbling block is the lack of surviving artifacts. Still, I can infer some useful information from survivals -- for example, when experimenting with the sub-types that have attached tools and objects, I can find surviving tools and whatnot that match the pictures that give me clues to what the shepherds may have been carrying. There are plenty of surviving knives and sheaths of types similar to what we see here, as well as scissors or shears. Items like the round boxes have a wider range of possibilities: leather, carved wood, bent-wood, metal perhaps. All can be tested and tried out with various possible types of contents.

Indirect reproduction

More commonly, you won't have a complete model to work from, and some level of guesswork and interpretation will be necessary to turn your research into a concrete object.

What influences might the likely precursors or materials have on the result?

Know what materials would reasonably have been available. Particularly when you are trying to reproduce "typical" objects from a culture, a construction hypothesis that involves rare substances, expensive exotic goods, or techniques known to have been first introduced in the next century will not serve the purpose. If two hypotheses exist, and one would be a natural evolution from pre-existing styles, while the other would be a complete break and innovation with respect to previous work, the first is to be preferred. Look for similar artifacts that may be "cousins" to your topic, coming from some common root. Look for evidences of the existence of techniques that might have been transferred to your topic.

For example, while I don't have any surviving netted Shepherds Purses, there are surviving fragments of medieval netting that can suggest possible knotting styles. When considering possible fastening methods for the Shepherds Purse, I can research surviving belt fittings, or examine belts in art where the fastening method is more clearly shown. I can look at how the behavior of various materials are shown in the art of my period and try to match them with how the material of the Shepherds Purses behaves. Reasonably available materials might include linen, wool, or leather, or for the netting linen or hemp. How are these materials depicted? How do they drape or fold? What colors do they appear in?

As I have discussed previously, the prevalence of white-colored Shepherds Purses in art suggests to me that the default material may have been linen or some other vegetable fiber, which could also account for some of the paler brown versions (in an unbleached form). But some of the shaped pouches in darker colors raise the possibility that leather may have been used -- as it certainly was for many other styles of pouch, of which some survive. Netting is most likely to have been made from whatever the default "string" fiber was -- almost certainly some sort of plant fiber, either linen or hemp. I discount the idea of pouches of this type being made from wool partly because of the color issue, and partly because the greater elasticity of wool seems less suited to the purpose.

Art suggests that fabric sash-belts may have been in use as belts significantly before their appearance as what appear to be carrying devices, and this evolutionary path has a clear motivation. Depictions of pieces of unshaped fabric being used to bundle up goods for carrying also supply a possible precursor. From this, experiments on how an ordinary piece of fabric might securely hold small objects in a belt-like fashion may lead to promising results. From the other side, there are prior examples of the "pilgrim" style of shoulder bag -- a large flapped bag on a long strap -- and this could be experimented with to see how it would need to be altered to be worn conveniently at the waist.

What materials and processes produce an effect that "looks like the pictures"?

This is always a tricky question, because artists often paint to an idealized effect that disregards typical behavior. If you look at 10-11th c. English painting, you find garments being depicted as if there were very light and drapey, falling naturally into many small close parallel folds. Move on a century or two, and the same types of garments are depicted with smoother, simpler draping. How much of the difference is a change in the garments themselves, and how much is a change in what the artist thought they ought to look like? One of the clearest ways to see the effect of artistic ideals on how artifacts are portrayed is to look at books on historic costuming, or artists' depictions of existing works of art made in various centuries. Take depictions of the same style of dress, or copies of the same original work, and note how they have been interpreted. Look not only at the clothing, but at the bodies wearing them. For example, you can recognize a Victorian-era illustration even when the artist is copying a medieval effigy, by the way the artist interprets and modifies the subject.

So beware of a hyper-reliance on making your artifact "look like the picture", while still using your data as a goal. Look at how other, known objects are depicted for which you have objective referents (e.g., surviving artifacts) and develop a mental "filter" for those elements you think likely to be skewed. Are all the textiles depicted as heavy and stiff? Well, then, maybe your textile pouch isn't really as stiff as it is shown. Are people's clothes shown with lots of close, small folds? Then maybe your pouch isn't quite as finely gathered as the artist has made it out. Are other objects in the scene shown in relatively schematic form -- omitting expected details? Then maybe the lack of detail in the depiction of the pouch doesn't necessarily imply as simple a construction as it seems.

But assume that the depictions are meaningful in reasonable ways. For example, if the main part of the pouch is shown as being gathered into a strap, then that main part is more likely to be a textile than, for example, leather. If a pair of knotted ties are shown on the outside of a flap, then assume that they are probably attached to something underneath so that the knot holds two parts together.

What materials and processes produce an effect that plausibly functions like the expected goal?

This is very similar to the preceding question, but requires some conclusions about what the function is. What are these pouches likely to be carrying? What materials and structures would work best to carry those items? On the most basic level, we assume that these are pouches, and that one must be able to have access to the interiors, so even if "the picture" doesn't show a method of access (as with some of the very schematic crescent-shaped pouches), we can assume that some method of access will be necessary. On another angle, the netted bags are primarily shown with the meshes set lozenge-wise to the waist, but a small number show the meshes square-wise. Are these both plausible alternatives? Make (or find) netted bags made both ways and play with them. The lozenge-wise mesh expands easily to form a wider (but shorter) bag, making it easier to get things in and out, while the square-wise mesh is not able to expand across the width of the bag. My interpretation from this is that the greater prevalence of the lozenge-wise depiction actually reflects the normal form of the bag, and that the square-wise examples may reflect an artist who is less familiar with the materials and is thinking only in terms of a visual pattern.

A great deal of the functional questions will be answered simply by playing with various materials and seeing how they behave.

Do any of your steps involve teleological thinking? (I.e., doing something illogical in order to get the desired result)

This is one of the commonest pitfalls that I see among amateur researchers. It stems from beginning with a result and then trying to establish a historic basis for it. It may manifest as an interest in a particular modern craft or technique, followed by a desire to establish a medieval history for this craft. (E.g., "I like to crochet -- crocheting seems fairly simple and obvious, so people must have been doing it since ancient times -- what evidence can I find that could be interpreted as crochet?") It may manifest as an esthetic appreciation for a particular effect or form, followed by a desire to create that effect or form in a historic context. (E.g., "I really like how I look in tight-fitting clothes -- surely Viking women also liked how they looked in tight-fitting clothes -- what evidence can I find for Viking styles that would show off my figure?")

While a certain amount of this hazard can be avoided when your initial research-stimulus is taken from historic sources, rather than from the modern environment, it is difficult to entirely avoid thinking in terms of modern parallels and then projecting those modern parallels back on your historic artifact. In the Shepherds Purse case, one of my initial reactions was, "Cute -- a medieval fanny-pack!" and, as mentioned previously, another situation that drew my attention back to the topic was the modern Rennaisance Faire practice of hanging all manner of useful implements, especially drinking vessels, from the belt. So I need to look back at my interpretations and ask, "How has my experience of modern fanny-packs influenced my understanding of the Shepherds Purse, both in terms of purpose and construction?" and secondly, "Do I have an underlying desire either to prove or disprove that carrying implements hung off a belt or pouch is a typical medieval practice?"

It's certainly possible that my own preferences in how to carry stuff around have shaped my beliefs about whether particular methods are more or less convenient. I personally find shoulder bags inconvenient and awkward, espcially if I'm trying to be active. If they hang off the near shoulder, they're constantly in danger of falling off; if they hang off the opposite shoulder, the strap is uncomfortable across my chest. And yet, demonstrably, many modern people don't find them inconvenient or uncomfortable, so there's no reason to assume that medieval people would share my preferences rather than the others. While the specific construction methods used for modern fanny-packs are clearly irrelevant to a medieval artifact, is it possible that the unitary nature of it may influence my expectations of the Shepherds Purse? Am I designing one-piece objects rather than multi-part ones for modern reasons? For example, could the netted-crescent style be composed of a net with a separate belt threaded through it, rather than a netted bag that extends into straps only at the ends? (The way the ends of the "pouch section" taper argues agaisnt this, but the question needs to be asked.)

On the attached object question, I'm not even sure myself. I find the idea of an easy way of carrying small tools and containers both intriguing and appealing -- and it's something that I'm strongly tempted to experiment with as part of a medieval outfit, even though I don't portray a shepherd. On the other hand, I don't think I feel a need to argue historic validity for a more general version of this practice, rather than taking a purely descriptive view of how the practice appears in the data. The danger is likely to come in communicating to others the difference between what I understand to be the historic parameters, and how I might experiment with the idea outside those historic parameters.

What other contemporary artifacts or practices might reasonably be expected to parallel your topic?

This is a re-stating of topics that have been mentioned above. What other artifacts might have been made for similar purposes or by the same craftsmen? What different artifacts may have been made by the same techniques? What kind of need or problem does your artifact solve? Who else might have had that need, and how did they solve it?

Healthy skepticism

Don't get too invested in your own interpretation -- keep track of what you know versus what you surmise. Be skeptical not only of your own motives, but of the motives of the people who produced your data (especially in the case of art).

What other reasonable explanations might there be for the various features of your data?

This repeats a theme seen previously. Don't mistake your first guesses for the only possible ones. It's even possible that more than one version of your artifact existed historically, and that it could be created in more than one way. Once you start developing models and hypotheses, avoid the temptation to shoehorn the rest of the data into them. If aspects of your data don't fit your explanations, acknowledge that. Maybe you have reasons for discounting their importance, but don't gloss over it or deny it.

What ulterior motives might there have been in the representations in your data?

It's no secret: medieval artists virtually always had an agenda. Often they had more than one agenda. If you're lucky, their agendas were unrelated to the features of their work that you're trying to study. St. Lawrence was martyred on a gridiron -- images of St. Lawrence are not a good place to look for unbiased information about methods of torture or the lives of early Christians. They may, however, be a very useful place to research the physical nature of medieval gridirons. When the pastoral life became romanticized and the subject of upper-class fantasy, many aspects of a shepherd's life were sanitized and gussied up. The shepherds that appear in the early 16th century pastoral tapestries are certainly better dressed, better equipped, and more idle than working shepherds are ever likely to have been. The feast laid out in Repas Champetre, complete with spread cloths, platters of roasted meat, and fine metal salt-cellars would create a poor understanding of the daily diet of shepherds (although it might be a reasonable model for an upper-class picnic), compared for example to the meal seen in the Hardouyn Hours. Does this mean that I should distrust these fanciful pastoral scenes as data? Not necessarily -- for example, it may mean simply that I should distrust the quality and quantity of features in the data.

To what extent does your data reflect unprocessed raw "recording" and to what extent does it reflect existing stylistic traditions?

This is a key question. We know that medieval artists used pattern books (some survive, showing individual sketched motifs to copy), and we know that workshops could produce multiple versions either of a book or a painting that clearly echo each other. For that matter, one can sometimes find manuscript illustrations that are clearly copies of another (or of a common source) centuries earlier. So how confident can we be that the items depicted in a work represent actual contemporary artifacts? With caution, naturally. How "realistic" is the style of the work? Certain levels of detail are difficult to maintain if the artist isn't familiar with the original. Can you identify known precursors to a depiction that would undermine its contemporary value? Can you evaluate other artifacts in the work in terms of the artist's expected surroundings -- e.g. durable goods such as pottery or jewelry where actual examples may survive? Does your artifact change in its representation over time? Does it gradually become more stylized, or does it change in unpredictable ways? One of the strongest arguments that the Shepherds Purse represents a group of actual contemporary artifacts is the range of variation found in depictions and the way the change in sub-styles over time seems unlikely to reflect a loss of touch with actual artifacts.

What modern beliefs, knowledge, or agendas have you brought to your study that you may be projecting onto your data?

This is another way to approach the question: "what are your prejudices?" I will confess to an underlying desire to determine that the Shepherds Purse is a "real" artifact. I may have been more eager to identify arguments for this conclusion than those against it. In thinking about the question "why shepherds?"I have been working primarily on casually accumulated knowledge of the medieval occupation, and I really need to double-check this "knowledge" with proper source materials. I am working from assumptions about the available source materials and the probable effects of class that are likely to be based more on the situation in Britain (my primary focus of study) than the continent (where the epicenter of my artifact appears to lie). I have only a vague general familiarity with the rise of romantic pastoralism, and this is another field where I really need to improve my specific background knowledge. My background in formal art history is minimal, and every time I start talking about anything touching on this specialty, I feel like a fraud. If I ever turn this research project into a scholarly presentation, I'll probably need to collaborate with at least one other person whose specialties complement mine.


Given your data, but not your analysis, what ideas do other people come up with?

This is always a good double-check on your reasoning and conclusions. We all have our blind spots and unexamined assumptions. You may not agree with the interpretations other people come up with -- they may differ due to validly different interpretations of the evidence -- but you'll be better off for having considered them.


After you've done all this work, presumably you're going to share the results with other people in some way ... but how? Part of the "how" will depend on your intended audience, but part will depend on what you're comfortable with. Being good at research doesn't necessarily mean being good at writing articles or being good at giving lectures. On the other hand, if you've got a subject that you're passionately intersted in (and very knowledgeable about), that can be a good way to slide into types of presentations that you need more practice on.

Who is your target audience?

Don't assume a one-size-fits-all approach. For any given presentation, have a sense of who you're talking to (even if others may make use of it as well). Are you going to be aiming at people who are primarily interested in understanding the original data, or people who are primarily interested in making their own artifacts? Will they have a good general background on the historic setting or will you need to set up some context first? Or maybe you'll be talking to relatively knowledgeable people who are interested in details and nuances. Consider what your relationship is to your audience. Are you talking to people who are more or less your peers: similar background and expectations? Are you going to be in the position of an expert? (That puts an extra burden of responsibility on you, if people are likely to take your every hypothesis as "fact".) Are you going to be an amateur talking to professionals? (Don't pretend to any more expertise than you have -- be upfront about what your background is.) Test your presentation on people who will give you honest feedback. Is your language too technical? Not technical enough? Is your approach too detailed? Too superficial?

What use will they make of your information?

What will your audience want to take away with them from your presentation? Are they interested in getting general background knowledge -- helping build of a historic context of a particular time and place? Are they interested in getting a general feel for the styles and structures of your topic, or do they want the specific information necessary to make a particular artifact? Or maybe they simply want to be able to recognize historic styles when they run into them. Don't try to do too many different things at once -- you probably can't do both a general survey and a practical hands-on presentation in the same session unless both are fairly simple. If you're teaching people how to make an artifact, make sure you give them its historic background -- especially make sure that they know how it relates to historic artifacts. Have you simplified features? Made substitutions? Especially in a historic re-creation context, people will tend to assume that anything you teach them is historically accurate unless you make it very clear it isn't.

Do you have multiple audiences for different aspects?

This is a very real possibility -- and it can help stave off the frustration of having a topic that isn't an exact match for any one group. It's possible to take one research project and do a general survey and hands-on workshop for re-creationists, a paper on some focused topic for academia, an extensive catalog of data on a web site, and maybe two or three articles of varying depth and length on various aspects for various venues.

What types of presentations or descriptions will work best for your audience and purpose?

If you've got a topic that depends heavily on visuals -- especially with lots of color -- consider putting together a set of slides of your material. If you can present your material sufficiently in black and white, then handouts may be more efficient. (I've been doing some experimenting with computer presentations, e.g., using Power Point, but the cost of the projection equipment is a bar in most contexts.) If your presentation is more informal, your handouts may only need to include outlines and figures and room to take notes. If your audience wants to walk away with a reference for future use, then you may want something more detailed with extensive references. Have a realistic sense of your ability to present the material when lecturing. Will you need a fairly detailed "script" to make sure you remember everything, or are you comfortable with a more off-the-cuff style? Do you need to be very precise about how long your presentation is, or do you have a lot of leeway? How much time do you want to allow for informal Q&A? If you're creating a written presentation (including illustrations), what sorts of constraints do you have on length and format? Are you writing for a publication that expects a particular writing style? If you're writing directions for creating artifacts, make sure you beta-test them on people with a variety of backgrounds to make sure they reliably produce the desired result. If you're dealing with formal publication (as opposed to class handouts) keep in mind copyright concerns for any material you take from elsewhere. (This is an issue that seriously affected the form of this present article.)

For the current project, I've used a variety of approaches: lecture combined with skeleton-handout, electronic "handout" with extensive graphics and independent text. Aspects of this project would also work as a slide-show lecture or a computer-based visual presentation. In theory, this project could turn into an academic lecture, and I may try putting together a very brief overview article with one or two construction patterns.

What type of distribution system will work best?

If you're producing articles or publications, you will either be doing it in the context of an existing publication (in which case, distribution is not your direct concern) or you'll be handling the distribution yourself. Do you want to do hard-copy publication or would electronic publication work better? Do you want to control distribution (e.g., only make it available in connection with lectures or classes) or do you want to broadcast it as widely as possible? Are you producing something that is commerically viable, and can you find existing vendors who might be interested in carrying it? Or is it something that doesn't really function beyond an "informal handout" level?

Previous --------------------------------------------- Main --------------------------------------------- Next

This site belongs to Heather Rose Jones. Contact me regarding anything beyond personal, individual use of this material.

Unless otherwise noted, all contents are copyright by Heather Rose Jones, all rights reserved.