A Manuscript Stemma: A Reprise

I sort of hate manuscript stemma. Don’t get me wrong, they have their uses, and they take some incredibly diligent and intelligent work (work that I’m really glad someone else is doing so that I don’t have to).  But stemma are also one of those devices that are occasionally put to great evil, in my book. They are sometimes used for recensional editing (the worst of evils), used to abject and even reject certain manuscripts deemed worthless based on its distance from an authoritative text, and they’re just plain mind-numbing to try to consume intellectually.

In addition to being put to evil use, stemma are also sometimes misleading, laden with jargon tending to make them indecipherable to non-experts, and just plain confusing.

Take this A MSS stemma, reproduced (absolutely without permission) from A.V.C. Schmidt’s Parallel Text Edition.


The first issue that I take with the traditional stemmatic diagram (which is not a critique of Schmidt’s stemma, or the great work done by several Langlandians to make genetic analysis possible–just a general problem with the genre) is that it has a tendency to make all the extant manuscripts look equidistant from the archetype. 

In this diagram, notice that all the sigla are lined up along the bottom of this graph, giving the perception that they’re all the same number of generations from the archetype.  Upon closer inspection, however, we can see that the manuscripts of the family have an additional generation (or two, in the case of TH2 and Ch) between themselves and the, um patriarch(?) of their family.

Second, without already being well acquainted with the terms of the manuscripts, this diagram becomes almost impossible to read.  The sigla themselves are specialized knowledge, so too are the various forms of representing parents (lowercase sigla, bolded or not), and even the notations for various versions of an authoritative A source.

In Schmidt, “A” is the authorial A-text, “(AØ)” is the possible early copy that may have circulated amongst a coterie of readers, and which is most definitely a better copy than “Ax,” the archetype to which the texts in all extant manuscripts can be traced.

And finally, the presentation of all the various MSS sigla as nothing more than x number of generations away from the archetype makes it so hard to put that information into context with temporal distance from the archetype. Thus, I have re-worked the traditional stemmatic diagram to incorporate the temporal dimension of when each manuscript was produced as well.


In addition to showing each generation of intermediate copies or lost manuscripts (represented by lowercase sigla), it also accounts for distinct periods of time and displays successive date ranges as empty nodes. The empty nodes are particularly important because they allow us to locate temporal distance between the earliest possible generation for a missing parent manuscript (represented by lowercase sigla) and the actual appearance of an extant manuscript in that family. If a node has no name or sigil attached to it, then it is simply a temporal placeholder for that MS family.

“Family” distinctions (like “r” and “m” and their children) are always located at the earliest point possible in MS descent (or the farthest left).


The earliest generation of MSS, then, comes at the 9th rung in the ladder (with A, AØ, and Ax as 1st, 2nd and 3rd).  That is the Vernon and TCC R.3.14, both made around the year 1400 (possibly earlier). Screen Shot 2014-09-22 at 6.55.58 PM

In the case of TCC R.3.14, this is at least five generations from the Ax archetype, so in order to allow both MSS appearing at ca. 1400 to appear parallel on this chart, I have inserted empty nodes on all other trees as a placeholder for the time in which such a generation may have passed.  It’s important to note that an empty (unlabeled) node does not represent a definitive generation between the lost exemplar (“v” for example) and the actual manuscript (V). It is instead, a placeholder pointing to a lapse of time between the ancestor MS and the appearance of the extant manuscript.

The reason I find this necessary is because the original stemma may misleadingly make it look like D and U (for instance) have a parallel (if not simultaneous) emergence because of how many generations they are away from the parent MS.

Screen Shot 2014-09-22 at 7.03.21 PM

Moreover, one may be led to believe, from the stemma that the “m” family of manuscripts are “closer” to the archetype than the “r” family because they are only a third generation removed from Ax, but if we pay attention to the dates, T (TCC R.3.14), which is actually the farthest removed genetically is one of the earliest manuscripts with significant A-text. 

This is an image of all the lost MSS and their genetic relationship to the archetype.

The large stemma above thus compensates for this misrepresentation by making use of the empty nodes.

To see it up close and personal, I’ve also built an interactive stemma that allows you to explore the genetic relationships one node at a time.  It is essentially the same diagram, just interactive.


When you explore these generations along a single branch, one might find new relationships and insights.

For example, this stemma includes all manuscripts that contain large A-portions (sort of…it doesn’t include the BmBoCot MSS that all have a bit of A slipped in between their B beginnings and C endings, or any A interpolations). That means it includes

  • A-only manuscripts
  • A/C splices–A texts or A beginnings with C endings added on
  • the BA hybrid–a B beginning on an A text
  • and Z, which isn’t properly an A-text, but is included in this stemmatic because of the A ending tacked onto the Z Visio before it’s completed with a C ending

If we change our focus a bit, we could compare where the AC hybrids sit in comparison to the A-only (and BA) texts are.


What’s interesting, at least to me, is that both the ACs and the A-only MSS have one ca. 1400 MS, but otherwise the As tend to be later than the ACs.  Also interesting is that the entire t-subgroup consists only of AC MSS, all of comparatively earlier dates. Plus, the only families to be exclusively A-only are in the e and m subgroups of the m family, with some of the later and latest manuscripts.  

What all of it means I can only begin to wonder.  Instead, I just organize the data and leave you to ponder from there!


One thought on “A Manuscript Stemma: A Reprise”

Please do collaborate!

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s