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

This page last modified May 16, 2005

Shepherds Purse: Focusing the Search and Exploring the Limits

Return to main page

Identified patterns

Once you've established that examples of your topic are out there to find, you'll want to do two things: depth and breadth. Try to establish a significant body of data in contexts where you can find it easily, to even out the effects of individual variation and provide a context for unusual examples. Then you'll want to figure out what the limits of your topic are: when does it begin, and from what? When does it end or turn into something else? Where was it used and not used? Was something else used in its place?

Begin looking extensively or exhaustively at contexts with a known pattern of examples

You should have begun to get a sense of what the likely contexts for your object are, just by where you've found examples already. In my case, it rapidly became apparent that the major common theme was shepherds. In retrospect (especially given the name I've applied to the artifact) this seems obvious, but at the beginning it was a startling observation. Furthermore, I identified three commonly-repeating contexts in my initial examples: French/Flemish pastoral tapestries from ca. 1500, and both Annunciation and Adoration scenes from French manuscripts of the 15th century.

Identify more general types of contexts and begin looking at them

This gave me a starting point for doing some in-depth searching. I went to the library catalog and discovered where the sections on Renaissance-era tapestries and on medieval French manuscript art were. This provided me with several shelves worth of likely fodder and I quickly added more examples parallel to the ones I had already found. (I also took to browsing large art-survey books in bookstores and noting down book titles or manuscript IDs to follow up on in the library.)

Consider both physical/cultural contexts and thematic contexts

My Shepherds Purses were, so far, showing up at the intersection of several characteristics: location (France and the Low Countries), era (15-16th centuries), and occupation (shepherds). This gave me several directions to work from. Other research topics may involve different types of of factors. For example, someone researching common everyday flowers that might be known or available to medieval people will hit on the manuscript layout style that uses very realistic flowers as marginal illustrations. Someone researching medieval birds will note that in a slightly different period, marginal illustrations often involve semi-realistic birds perched in the vines and acanthus flourishes. Someone researching typical table settings for meals should discover that Last Supper depictions are a good starting place, while someone trying to trace cutting-edge changes in women's fashion may find it useful to know that Mary Magdalene was usually depicted in the height of contemporary fashion, as contrasted with the often conservative "classical" fashions of many women in religious scenes.

Begin noting examples where the context predicts your topic but it isn't found

The dog that does not bark is a clue not to be ignored. As soon as you've established a pattern for your artifact, start paying attention to where it doesn't appear. It will be tempting not to keep track of this information -- I know that I'm not as thorough about it as I should be. Making a long list of notes saying, "didn't find it here, didn't find it here, didn't find it here" feels pointless. Some of my work in this field is discussed below, but one of the things I did was to use catalogs of art available on the world wide web that were searchable by keywords. Having established that my pouch style was strongly correlated with shepherds, I fed the word "shepherd(s)" into a couple of image-related search engines.

One of them was a web site of "great art works" that was searchable by words appearing in the descriptions of the works. The contents were primarily "paintings" in the usual sense, but at least one scene from an illuminated manuscript showed up. The search on "shepherd(s)" turned up 63 works, spanning the 13th through 18th centuries, but with a peak in the 16-17th. The largest numbers are from Italy, the Low Countries, and Iberia respectively, with smaller numbers from France and central Europe (Germany and neighboring areas). Only five of the works show shepherds wearing some sort of pouch, and only two of these show an item that falls in my Shepherds Purse category.

  13th c. 14th c. 15th c. 16th c. 17th c. 18th c. total
Italy 1


1 sash-type Shepherds Purse



1 shoulder bag, flapped style

1 shoulder bag & non-shepherd type belt pouch

3 1 25
Low Countries    


1 frame and drawstring purses


1 sash-type Shepherds Purse

10   18
Iberia       2 9   11
France         2 4 6
Central Europe   1 2       3
total 1 4 11 18 24 5 64

This was a bit of an unexpected result, since I was used to seeing lots of examples in manuscript art. The catalog, as I note, was biased towards paintings, and as the above statistics show, it was biased away from French works of the pre-17th century period -- exactly the time when I was finding most of my examples. How to interpret these internal biases in the collection -- and what they mean for my artifact -- is complex. Was France not producing paintings during the relevant period? Unlikely. Was it not producing "name-brand" painters who end up in popular art collections? Possibly. Were French painters of the later medieval period not depicting shepherds? Possibly. All in all, the search was most useful in demonstrating the near absense of the motif in Italian works, although it would have been more useful if I weren't comparing the oranges of Italian paintings with the apples of French manuscripts. (In looking at collections of medieval Italian manuscript art, my impression has been that pastoral scenes simply weren't a popular genre in this medium.)

One of the more useful things that this survey helped confirm for me was that the 17th century was pretty much the later limit of my artifact in art -- and actually looking at the works from the 17-18th centuries helped explain it: shepherds, whether in religious scenes or in secular pastoral scenes (which become even more common) begin being depicted with classical draperies rather than ordinary clothing. They have become an excuse for an artistic exercise -- "the shepherds in Arcadia" and so forth -- rather than a reflection of the everyday surroundings of the artist. This stylistic shift doesn't entirely preclude the appearance of pouches -- the Bronzino adoration (ca. 1535) combines a classically (un)draped shepherd with a shoulder bag at least reminiscent of the flapped crescent style of Shepherds Purse. But at the same time, this survey introduced a new question: to what extent was the medium interacting with culture and subject matter? My impression was that Shepherds Purses were statistically more common in manuscript Annunciation scenes than in paintings of the same genre -- but did this mean that book artists and easel artists were recording different things or following different conventions? Or was the difference more geographical, with the Italy/France split being more relevant than the difference in medium? But then, what about the Low Countries, where manuscript and tapestry art appeared to show the artifact more commonly than paintings did?

I did a similar systematic search using the Google "image search" function with the keyword "shepherd", looking at the first 20 screens of results (this would, I believe, be 200 total images), and recording only pre-contemporary works of art, of which there were ten: seven Annunciations, two Adorations, and one generic Pastoral, covering manuscript art, woodcuts, etchings, and paintings -- some that I had already encountered, but most new to me. Of these, five included a Shepherds Purse (all of the crescent-shaped style, and all in Annunciation scenes from manuscripts of around the 15th century and French or in a very similar style to the French style). Three involved shepherds with other types of belt pouches (e.g., kidney-shaped pouches), and the other two sets of shepherds had shoulder bags of some time (a "pilgrim" style of deep rectangular bag in one case, and in the other apparently the "doubled-ended" style typically associated with field workers).

So the Google survey contradicted the art-catalog survey and agreed with several observations from my initial research:

All of this will be important in the "artifact or artistic motif" debate.

Extrapolated patterns

Having developed a significant body of data in a focal context, the next step is to systematically explore the limits of that context.

Vary a single factor in the established contexts and explore

As in any scientific exploration, the easiest and safest way to explore a phenomenon is to change one variable at a time and see what happens. As previously mentioned, my "core data" involved a number of parameters: occupation, genre of scene, type of medium, time-period, geographic location.

Since most of my general reference works covered a particular medium, geographic region, and general time-period, the easiest check to do first was to take works in which shepherds wearing Shepherds Purses occurred, and look exhaustively at what non-shepherds were wearing. This confirmed what my anecdotal observation had been: with extremely rare exceptions, the Shepherds Purse style was associated with the one particular occupation -- but was not restricted to only particular genre scenes.

Since the bulk of my examples came from religious manuscripts, especially Annunciation and Adoration scenes, that was my fixed context when looking in other geographic regions. (This particular expansion is tricky because manuscripts traveled, and the eventual location and associations of an object might be very different from the context of its creation.) This search turned up several types of results. In a few regions (e.g. England and Germany) I found manuscripts with "realistic" shepherds that sometimes wore Shepherds Purses. In other regions (e.g. Italy) the same types of genre scenes occurred, but the shepherds were portrayed wearing very stylized "classical" garments (and no Shepherds Purses), such that there seemed no basis for assuming that they represented the appearance of actual contemporary shepherds and their accoutrements. In other regions figures might be portrayed in a very stylized manner, not in a self-consciously "classical" style, but in a way that raised similar doubts that the art reflected contemporary dress. (For a more familiar comparison, imagine trying to decipher ordinary early-medieval Irish clothing on the basis of the figures shown in the Book of Kells and similar works.)

This single-variable approach began to falter when I wanted to look at geographic regions where the same tradition of manuscript art was not available. Here, I went back to my "general art survey" approach, but in collections with the geographic or cultural focus I was interested in. So, for example, I wanted to see if the motif apeared in the Balkan region -- I located the library section were extensive surveys of medieval art from that region were filed and started browsing. I found no Shepherds Purses -- but then, I found next to no shepherds. The concerns of the artists in that time and place were more on "focal" religious figures (the "names", as it were -- the Holy Family, saints, major Biblical figures) and not on "everyday" figures in genre scenes. So there was, in fact, no clear evidence as to whether Shepherds Purses were or were not used: I had no positive evidence that they were, but conversely, I didn't have the genre context in which to observe their clear absence either.

This raises an important point to remember: medieval art (and, for that matter, surviving medieval artifacts) is not a random "slice-of-life". It's not a photograph where everyday life is accidentally included in the background. In particular, the topics of medieval art were typically those that some (rich, and therefore usually noble) person commissioned to have included. And the concerns of the wealthy, noble patrons are not necessarily going to focus on the occupations and dress of lower class laborers. So why are shepherds found at all? In some cultures, the Nativity genre focuses exclusively on the Holy Family, in others, the focus expands to include the Adoration of the Magi -- members of the patrons' own class. There are many participants and events from the Gospels that could have become standard genre scenes for a prayer-book, and the choice of the shepherds' Annunciation is not necessarily a natural or predictable choice. Part of the explanation for this focus, no doubt, is the key image of the shepherd as a Judeo-Christian religious metaphor. "The Lord is my shepherd", and both John the Baptist and Christ are commonly given the attributes of a shepherd (John, deriving from his purported occupation, Christ from a more metaphoric motivation). For similar reasons, images of carpenters are disproportionately represented in medieval art.

The genre of calendar illustrations is, perhaps, less curious. The inclusion of a calendar at all primarily has the purpose of listing feasts and days for which special prayers might be appropriate. When considering typical seasonal activities to illustrate the months, even for a nobleman, the salient activities will relate to the agricultural year: planting, harvesting, making hay, slaughtering animals in the fall. In some cases, upper-class activities such as hunting may be included as appropriate, but usually the agricultural cycle will be included as background. The other very popular source of calendar illustrations are zodiac motifs.

Even in the medieval manuscripts, lower class figures such as shepherds may be "cleaned up" for their portrayal in art (although it is also common to show them ragged and dirty). But towards the beginning of the 16th century, we get the romanticization of the lower classes as carefree Arcadian figures (note also the rise of the pastorale genre of poetry). We find scenes where working shepherds are "gussied up" in expensive fabrics and high-fashion styles, as well as scenes where upper class people (in clothing appropriate to their station) are playing at being shepherds and shepherdesses. This is not a trend that applies to all types of working-class occupation. We don't find nobles playing at being millers and bakers, or cavorting merrily behind the plough. Somewhere along the way, in certain European cultures, shepherds had acquired a special symbolic status that, among other things, led to them being portrayed (at varying levels of realism) relatively commonly in art.

But these artistic themes and motifs are often highly culture-specific. And so, even if the art of all cultures accurately portrays motif X in that culture, relevant art using that motif will not necessarily be equally produced in all cultures. (And, as has been noted, cultures may vary considerably in how realistic their portrayal of a particular motif may be.)

Look for the "edges" of the phenomenon, both in space and time

I have not entirely accomplished this step to my own satisfaction. In terms of the core Shepherds Purse style, my geographic scope extends from England in the north and west, to Germany in the east, to Iberia (Spain and Portugal) in the south and west, with the strongest representation being in France and the Low Countries, at the center of that circle. I've been told that there are examples in Danish church wall paintings, but so far the surveys of this medium haven't turned up any examples (although shepherds do appear). My experience so far has been that the Slavic areas of eastern Europe have a very low incidence of shepherds in art. Shepherds abound in Italian art, but the classicizing trend begins early there -- significantly overlapping the era when Shepherds Purses are found elsewhere. Spanish examples are all relatively marginal in terms of the Shepherds Purse style-group -- there are clear similarities, but the overall effect is shifted.

Similarly, I am not comfortable with establishing clear temporal boundaries for the supposed use of the artifact, although there are some clear limits to its appearance in art. At the early end, the style emerges when the clothing depictions become less loose and blousey -- that is, in the era prior to the first examples of the Shepherds Purse, it is not clear that a waist-bound Shepherds Purse would normally be visible anyway. So does it only begin to be shown in art when the clothing changes such that it would be seen? Or does it only begin to be shown because it is a newly developing style?

At the later end of the scale, the reason for the Shepherds Purse disappearing from art is clearer: the pastoral genre (whether religious or secular) shifted strongly to classical draperies rather than realistic "photographic" depictions, and changes in popular styles of religious art led away from the previous traditions in personal prayer-books (to say nothing of the general effect of the printing press on the entire field of manuscript art). All of these changes led to a closing, or at least a narrowing, of the window of opportunity for the realistic depiction of shepherds in art. So with the close of the 16th century, the Shepherds Purse motif more or less disappears from the page and the canvas -- but we are left ignorant as to whether it continued in use as an everyday artifact or, indeed, whether it was still in use at that point as an everyday artifact.

Look for "holes" in the phenomenon

When laid out in an organized fashion (as in the date & location chart), a number of weak points in the data show up. I have relatively few works created in England (although a number of the French products ended up there), and I should explore this more. In the mid 16th century, the French and Low Countries material shifts almost entirely to tapestries, and it would be useful to fill out the information in different media. These are "holes" in the sense of "things I haven't researched specifically enough".

The other kind of "hole" to notice are those where I've looked at large amounts of art -- where I might reasonably expect the artifact to appear -- and yet it doesn't. These would be "holes" like Italy, if one expects to find strongly similar material culture between France and Italy. But, in general, there don't seem to be any major discontinuities within the established scope.

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.