Source of Confusion: A graphic attempt to explain useful sources versus less useful ones, for garments

by Cynthia du Pré Argent

© 2000 Cynthia Virtue

There is confusion in the SCA (and, presumably, other historical organizations) regarding primary sources, secondary, tertiary, and so on, and not enough discussion about what constitutes a good source versus a bad source -- an axis of analysis which is very valuable to evaluating an item that has been recreated.  Some of these confusions are propagated by contest judges.  I hope to make the issue very clear: that it is far more useful to worry about quality of source, rather than its position on the primary/secondary chart.

Segments on this page:

The first issue in this puzzle is: What is the original item?  We can't have things or concepts that come from the original item without knowing what the original is.

If you are documenting painting technique -- a painting is an original.
If you are documenting costume -- a piece of period clothing is an original.
If you are documenting royal edicts -- a manuscript is an original.
If you are documenting something that is not an artifact, such as a musical performance or a battle -- the concept gets murky.  It may be that the original, by its nature, is forever lost.

Knowing what is the original, is thus, for items, relatively easy.  The trouble comes starts with the term "primary" because it is weighted with a sense that primary is automatically better, desireable, the pinnacle of sources available.  This emotional content gets in the way of clear thought.  An original item is a primary source, but it may not be the best source.

The use of primary/secondary/tertiary in archeological sources is rarely defined in the texts we historical costuming people use (such as the Museum of London series) ; it's something those in that field understand, and if you have a book entirely about primary sources, why define elementary terms?  Here is a link to a description of primary sources about a historical event -- but an event's primary source, such as a letter, is not a primary source for a garment; the garment itself is the primary source.  Failing a textbook at my elbow, the definition in the online Merriam-Webster Dictionary will have to do:

primary: 3a: DIRECT, FIRSTHAND <primary sources of information> b: not derivable from other colors, odors or tastes [...] e: belonging to the first group or order in successive divisions,  combinations or ramifications <primary nerves>
The relevant parts of the above definition are "not derivable" and "first group in successive divisions."  Using the example given of nerves, note that a secondary nerve is still a secondary nerve, even if the primary nerve is not present.   Likewise, a primary  source of information is the one that all other derivative sources proceed from.  So it is with garments; an original garment itself  is the primary source for any derivative works.  The derivative works, such as a photo, a painting, or a re-creation of the garment, are secondary sources.  They have the potential to be darn good secondary sources, but in the process of making them, information is lost, new information is added, or human interpretation may interfere with presentation.

For illustration, let us consider the situation from a less theoretical perspective.  I have, on my scanner, a common cotton sock.  The following picture of the sock is of course one generation removed from the real sock, but since the Internet won't yet let me conjure actual socks for you to paste onto your computer monitor, we'll have to pretend that the picture is a real sock, and proceed from there.

It is useful to think of the progression of sources rather like a genealogical tree; the sock begets secondary copies of itself, and those in turn beget secondary copies of themselves, i.e., tertiary copies of the original sock. The chart below may be too wide to show in your browser window; slide your width slider button around so you know what is there.

Primary  source

A sock
-A terrible sock

Secondary sources

- -

Tertiary sources

- -- --

There is a critical thing to notice in this table.  Some of the secondary sources are going to be less useful than other secondaray sources -- the bad (inept) drawing of a sock is less useful than the good (clear, high resolution, etc.) photo of a sock.

In addition, some information will be lost in each subsequent copy of the sock, even if great care is taken to replicate it as accurately as possible.  Consider a photocopy that is then photocopied through several generations.  Eventually it's a real mess, even though a photocopier reproduces in near photographic quality.  Consider also the party game "Whisper down the lane" or "Telephone."  By the end of the information transmission game, all of the content of the orginal message has been substituted for erroneous information.

Once dubious quality is put in the equation, subsequent generations will be unlikely to regain quality, so even an excellent copy of a bad drawing will itself be a bad source.    Of course, a bad copy of a bad drawing of the original sock can get worse yet!

The colors in the chart above, then, are a rough cue for research usefulness of the source.

The primary source itself, is generally unimpeachable.  But what about a terrible example of a sock, perhaps one knit by myself?  It would not be representative of the usual 20th Century sock.  A sock knit by me, although a primary source, is a poor source.  Likewise, just the cuff of even a good sock would not be enough information to reconstruct a real sock.  Thus a partial primary source is likewise suspect.

A careful person should also be aware of the bias of the source, which may present the information in such a way as to invite inaccurate conclusions.  Political campaigns are an obvious example, but bias in art or archeology is more subtle.  For example, if I'm very interested in purple Ruritainian feathers, I may not put any examples of green feathers in my webpage on them.  It would be understandable for you to assume there weren't any green ones, but that might be inaccurate.

X and Y axis illustrationTaking all these factors into consideration, you can see that the worthiness of a source is a combination of its rank on the primary/secondary/tertiary hierarchy, PLUS its quality/reliability hierarchy.  Either scale by itself is not a reliable metric in evaluating worthiness of a source, and using only one of them (most especially, using only the primary/secondary hierarchy) can be downright misleading.   Both metrics must be used, and ruling out a source solely on the basis of primary/secondary/tertiary is only half the picture.  Or half the sock.
A footnote to all this: Do not fail to evaluate synthesized works (such as my own or other people's webpages, teaching handouts, published general books on "costume" etc.) by these criteria.  How good do the sources cited in the handout appear?  Are there any medieval (or Ren.) works cited or reproduced at all?  How well are they reproduced?  (I've seen some that ask you to take everything on faith in the teacher or his drawings!)  What is the quality of the teacher's own sources?  Do you think the teacher has made good conclusions?  Such careful analysis can stop the propagation of poor information and replace it with better information.

Thanks are due to Kass McGann, who gave this an editorial eye and suggestions, and also her page on good research methods. And of course all the people over the years who have written on the various email discussion lists about source evaluation!  This article has been edited from my original, thanks to very useful commentary from Chris Laning.
All material © 2001 Cynthia Virtue Email Author with comments
Back to Virtue Ventures Main Page Back to Article Index

And lastly, for the folks so interested in the topic that they check below the link table, there is this final bit of confusion: there are other definitions not at all related to the above; one says that a primary source is an original thing or writing, and a secondary source takes that information and makes other inferences, inserts background material, and produces a somewhat popularized work as a result.  An example of this might be Janet Arnold's books Patterns of Fashion, which take primary sources (clothes and original tailor books) and produce modernly-understandable patterns, conclusions, and perspectives on their use.

In addition, if you're talking about reproductions of an original that is flat to begin with, such as a painting or a manuscript, a photo of same will lose much less information than a photo of a garment will.  This makes a photo of a manuscript a much better source for a manuscript, than a photo of a sock would make.  For such flat-to-flat reproduction, some think that "primary source once removed" would be a reasonable term; the photo isn't a primary source, but neither has it lost a lot of information.

Visit my CafePress store!