Notes on the name Tintagel

The local form of Tintagel - Dundadgel - is something I have only found in Ekwall and in a couple of other old sources.  To find out if such a spelling for Tintagel could indicate that the site is a relocation of the Welsh Caer Dathal, which was associated with Uther in Welsh tradition, I asked Andrew Climo of The University of Exeter about the place-name.  His response was quite thorough, and I have included it in totality below:

Summary

Toponymists have struggled with this name for decades and place name evidence for Tintagel is sparse. One way forward is to go with the attested form Tintaieol as a basis, as it may preserve the Cornish pronunciation from c.1200. Toponymists have glossed over the matter of the <o/e> alternation, but this is adequately borne out by the examples provided in the texts. Unfortunately, this makes Weatherhill’s highly attractive Tente D’Agel ‘Devil’s Stronghold’ look unlikely. It is reasonable to substitute initial <d> and assume that ‘Din’, fortress, is the prototheme on the basis of consonance. This renders Dintajeol. One could then leave open the matter of whether *Tajeol was a personal name, refers to the Devil (C. ‘Dyawl’), or something different entirely. This way forward accounts for all morphological features seen within the texts and provides a usable form.

Sources and Hypotheses

Sources (in Padel, 1988; Weatherhill, 2005). The following apply to Tintagel Head and/or Castle:

·        Padel: Tintagol c.1137; Tyntagel 1208

·        Weatherhill: Tintagol c.1145

As both point out the settlement of Tintagel was Trewarvene 1259; and Trevenna c.1870 (Padel). Weatherhill thinks the name is Norman French Tente d’Agel ‘Devil’s Stronhold’, cf. Tintagau (Sark).

Ekwall’s assumes a <dg> spelling, perhaps based on the notion that Middle English <g> is [dʒ]. The spelling <dg> appears speculative although was used in Middle Cornish from time to time although the Norman French usage of <g> was also common and <gg> is also found. Neo-Cornish uses the graph <j> to represent [dʒ], and native Cornish names used would have derived from either <s> or <d> in Old Cornish.

Phonology

There is a need for additional forms to make a clearer determination on linguistic grounds, but Layamon’s Brut (in Project Gutenburg), which is in Middle English, shows Tintateol, presumably a transcription error from Tintaieol, which also occurs in that text, as well as the form Tintageol, which suggests that <g> really was pronounced [dʒ] ~ <j>.

The initial <t> is easily explained as a substitution for <d> ~ [d], and as pointed out by Padel (1985, p.85) is found in names such as Tenby, Tintern. Initial fortis (hardening of d à t or g à k) is not much of a stretch, particularly resulting from consonance (so the consonants at the front, middle or end then agree).

If Weatherhill is correct that the name is NF and one can assume <g> ~ <j> ~ [dʒ], then Tente d’Agel could be rendered in Cornish as Tentajel. The alternation of spellings <e/i> is common in C. and whether it is of any consequence or not is a question for another day and still occupies C. users of different persuasions.

On the other hand, if Padel is correct and it is C., then it would be spelt in MC something like Dintajel, Dintajol or Dintajeol. Din is the m. noun ‘fortress’ so *tajel/tajol/tajeol is a qualifying name, noun or adjective, presumably in MC. However, the suggestion that *tajel < *tagell ‘neck’, ‘noose’ or ‘constriction’ is problematic: Whilst it makes topographically tempting, it would require a NF qualifying noun added to a C. prototheme and then a series of changes to occur to provide the present spelling (Din > Din-tagell > Dintajeol > Tintagel). This seems a bit of a stretch. The other issue is that there is a vowel alternation <o/e> or possibly <eo>, which should not be casually dismissed (neither Padel nor Weatherhill address this).

Morpheme Splitting

There are several ways that tajeol might be split: (i) ta-, (ii) taj- or (iii) taje-. The first way of splitting might suggest the preposition -to or -de ‘unto’, which typically requires a personal name following. ‘Devil’ was diavol (Voc. Cor.), so the hypothesis of ‘Devil’s Fortress’ could conceivably work with a little forcing: *Dintodiavol > *Dintodyawl > *Dintojawl. Second and third way of splitting would suggest an OC root *tadi- or *tado-, and potentially an instrumental suffix -illo: *tadoel > *tajeol. As a parallel, one thinks of the name *teuto-uualos ‘Warrior of the People’ > *Tudwal.