hrj

Parry, T. 1929. Peniarth 49. Gwasg Prifysgol Cymru, Caerdydd.

I confess that the likelihood of me doing language-geeky things with Welsh texts tends to decrease as time goes by. Not from lack of interest, but from a realization of time-prioritization in my life and the fact that other things give me a better pay-off these days. But that doesn't stop me from stocking my library with books that would be very useful in the alternate universe in which I continued to make historic Welsh linguistics my primary focus.

Peniarth 49 is a manuscript of poems by Dafydd ap Gwilym, copied/written/transcribed by the scholar John Davies right around 1600. But my particular interest isn't so much for the literary material as for the reflection of the development of Welsh orthography at that date. The evolution of spelling can be a surprisingly difficult topic to study. Individual edited texts tend to focus their discussions on specific narrow features that provide evidence for chronology or reflect particular regional differences, but reverse-engineering those bits into a more general understanding of the range of typical orthographies for a specific time and place is frustrating. And, in particular, coming up with a general understanding of the range of spellings current in Welsh ca. 1600 is harder than you might think (just in case one were, say, interested in creating name consistent with that period).

The usefulness of a text such as this for studying spelling is that the material falls in an identifiable bracket. Unlike genealogical material, romances, or historic chronicles, we have a clear composition date (the dates of the poet) and the text will not reflect an older orthographic tradition than that date. And as a later terminus, we have the date at which the specific manuscript was created. In this case, that still leaves us well over a 200 year span of time, with any particular feature potentially reflecting any time during that span, depending on whether the text in question was faithfully copied from manuscript, transcribed from oral performance, deliberately updated in spelling by the writer, or any combination of these paths. But in a sense this is a feature, not a bug, as it gives us a sense of the range of spellings that one single author, creating a written collection with thematic unity, considered to be unsurprising to appear in combination.

So, for example, we can trace the different treatments of the sound rendered in Modern Welsh as "rh" which appears here in the most conservative form of plain "r", in the briefly experimental "rr", as well as the eventual winner "rh". Similarly, the establishment of "f" to represent the sound [v] in all locations (even the laggard word-initial position) can be traced through the three segments of the manuscript, written by the same man at different times.

Perhaps most interesting as a spotlight example are the variants of the poet's name throughout the manuscript, thanks to the common formula ending each poem with "so-and-so sang this". The name David is remarkably resistant to following the usual rules of Welsh orthography in many pre-1600 context, due to the combined influence of the early medieval Welsh spelling being identical to the Latin (David) and likely encouraging a retention of the Latin form, plus the orthographic influence of English on the legal records that are our most prolific source of personal name data for 15-16th century Wales. Yet here, apparently relatively free of those normalizing influences we have the following. I've included every single mention of the name, in the order in which they appear in the manuscript -- and thus in the order they were recorded -- to show both the variability, the patterns, and the shift from one default to another over the course of the book's completion. The manuscript has three "hands", which current theory holds represent the same writer at three different periods of work (in which he was copying poems from different manuscript sources that had come into his hand). Whether the different levels of consistency represent his originals or a progressive tendency to use a standard form is an open question. But the most prominent differences between group 1 and group 2 (the use of "dd" versus "Da" as the shortest abbreviation; and the use of "v" versus "f" in the fully spelled out first name) are consistent with an earlier and a later spelling tradition, while the shift from "Davyt" to "Davydd" in the first group suggests the spellings may come from two different spelling traditions within his source for that portion.

The first source

dd ap gwilym (3 times)
Dauit ap glm
Davyt ap glm (3 times)
Davyt ap gwilym
dd ap glm
Davyt ap glim
dd ap glim
Davyt ap glim (2 times)
dd ap glm
dd ap glim (5 times)
Davyt ap glim (2 times)
Dauydd ap glm
Dd ap glm
dd ap glim
Davydd ap glim (3 times)
davydd ap gwilym
davydd ap glim
davydd ap glim
dd ap glm
dd ap glim (2 times)
davydd ap glm
dd ap glim
davyd ap glm
Dauyt ap glm
dauid ap glim
davydd ap glym
dauyd ap glim
Da ap glm

The second group

Da Glm
Dauydd ap Glm
Da ap Glm
Dafydd ap Glm
Da ap Glm (8 times)
D ap Glim
Da ap Glm (14 times)
Dafydd ap Glm
Dd ap Glm
Da ap Glm (7 times)

The third group

Da ap Glm (36 times)
D ap Glm
Da ap Glm (16 times)