Pictish succession - part 2 of 3
Stewart Baldwin
-----------------
2. THE EVIDENCE FOR A RELATIONSHIP BETWEEN CERTAIN PICTISH AND
NON-PICTISH KINGS IN THE SEVENTH CENTURY
In the absence of direct genealogical evidence, if one wishes to
assert that an inheritance must have come through a female link, then
the most straightforward way of doing that is to show that the
inheritance did not come from the direct paternal line. Thus,
evidence that the paternal ancestry of some Pictish kings went back to
non-Pictish dynasties has been advanced as evidence that their title
to rule was through their mother.
Woolf states [p. 158] that "In the period before ca 720 there were
only two Pictish kings who would seem to have fathers who were kings
outside of Pictland, Talorcan son of Eanfrith and Bridei son of Bile:
the former, possibly a son of Eanfrith king of Bernicia (633), and the
latter the son of a king of Dumbarton." This statement is misleading.
There are at least two other non-Pictish kings during that period
(Maelgwn, king of Gwynedd, and Domnall, king of Dál Riata, of whom the
former is not mentioned by Woolf) who have the same name as fathers of
certain Pictish kings, and who would make chronologically feasible
fathers for those kings. Although Woolf argues correctly that there
is no proof that the Donuel who was father of two Pictish kings was
the same person as Domnall of Dál Riata (and the same applies to the
case of Maelgwn which he did not mention), it is also the case that
there is no good reason to regard these possible identifications as
false. The evidence is simply ambiguous. Also, it can be noted that
the restriction to "kings" outside of Pictland leaves out another
non-Pictish father of Pictish kings for which Woolf had argued before
(the case of the sons of Derile, already discussed above). Kings or
not, the more non-Pictish fathers of Pictish kings that you have, the
weaker the argument for patrilineal succession becomes.
Much of the argument has centered around the Pictish kings Talorcan
son of Eanfrith and Brude son of Bile, for whom the evidence, as
interpreted by those arguing in favor of Pictish matriliny, appears to
show matrineal succession in action. Thus, the Pictish king Talorcan
(658-662) was apparently son of Eanfrith, king of Bernicia (633-4),
who had been in exile among the Picts after the death of his father
Aethelfrith. This means that Talorcan was the first cousin of king
Ecgfrith of Northumbria, who was defeated in the year 685 by the
Pictish king Brude, son of Bile (Beli), who in turn was referred to as
a "fratruelis" of Ecgfrith by the ninth century author of Historia
Brittonum [HB].
The following table illustrates the basic known information (but
without the conjectural links that are often added to fill in the
picture). As usual, it must be displayed in a fixed-width font (such
as Courier) in order for the spacing to look right, and a Line length
of at least seventy characters should be used.
Aethelfrith, d. 616
king of Bernicia
________|_________
| |
Oswy, d. 670 Eanfrith, d. 634
king of king of
Northumbria Bernicia
| |
Ecgfrith, Talorcan, d. 662
d. 685, king of the Picts Bile
king of |
Northumbria, Brude, d. 693
defeated in king of the Picts
battle by Brude,
"fratruelem suum"
according to HB
The exact meaning of the term "fratruelis" is part of the controversy
as to how this information should be interpreted. The usual meanings
of the word would imply that the individuals were first cousins,
either the sons of two brothers, or the sons of two sisters.
Proponents of Pictish matriliny have generally argued that the term
should be accepted in a looser sense of close relative. Given that
interpretation, along with the probability that Brude's father was a
king of Strathclyde, the most natural interpretation of the evidence
then becomes that Brude's claim to the Pictish throne was as a son of
Talorcan's sister. When that is added to the high probability that
Talorcan (with his very Pictish name and his title as king of the
Picts) was the son of a Pictish princess (presumably married by
Eanfrith when he was in exile among the Picts), the appearance of
matrilineal succession becomes striking. As Sellar stated:
"Conjecture it may be, but the thesis of matrilineal succession is so
neat that it is difficult to believe that it is mistaken. Any other
explanation of the facts which survive would seem to involve the
unnecessary and improbable multiplication of hypotheses." [Sellar
(1985), 38]
Smyth attempted to explain the term "fratruelis" by speculating that
Aethelfrith has married a member of the Strathclyde dynasty, a
conjecture which has far less probability than the scenario outlined
above. Woolf takes a different approach. He notes that the above
scenario requires a looser definition of the term "fratruelis" than
the strictest definitions of the term, and suggests that term ought to
be interpreted in the strict sense. The problem is that neither of
the two alternatives of the strict form of the definition looks very
likely (which is why it has often been suggested that a looser
definition was meant). Defining "fratruelis" to mean "sons of two
brothers" would require Bile to be a brother of Oswy, a very unlikely
relationship, and defining it to mean "sons of two sisters" would mean
that Brude and Ecgfrith would have to have the same maternal
grandparents. Now, the maternal grandfather of Ecgfrith is known to
be the famous Edwin of Deira, king of Northumbria, so this
interpretation would require Brude's mother to be a daughter of Edwin,
a relationship for which there is no other evidence.
The important point, however, is that either one of these strict
interpretations would require Brude's claim to the Pictish throne to
come through his father Bile. Woolf deals with this by providing a
conjectural patrilineal pedigree for several seventh century Pictish
kings. He accepts that Bile was the same person as the "Beli map
Neithon map Guipno ..." who appears in the Harleian genealogies of the
Strathclyde dynasty [EWGT, 10], and follows Molly Miller's suggestion
[Miller (1978), 54] that Neithon son of Guipno (Gwyddno in later
sources) may have been the same as the "Nectan nepos Uerb" of the
Pictish king-list (and in giving this kings death date as 621, he is
apparently also identifying him with the Nechtan mac Canonn whose
obituary appears under that year in AU). He also expands on one of
Miller's comments [Miller (1978), 54-5] to suggest that Guipno/Gwyddno
was the same person as the Uuid who appears as a father of three
seventh century Pictish kings Gartnait, Bridei (Brude), and Talorc to
get a conjectural table of the following form [same comments about
font and spacing apply]:
Uuid/Gwyddno
_______________|_______________________
| | | |
Nechtan Gartnait Brude Talorc
king of king of king of king of
the Picts the Picts the Picts the Picts
|
Bile daughter of one
| of the above kings = Eanfrith
|_____________ |
| | |
Owain Brude Talorcan
king of king of king of
Strathclyde the Picts the Picts
[Note: I am using the Pictish names in the form given in Miller
(1978), i.e., "Brude" instead of "Bridei", etc.]
Woolf acknowledges that such a table must remain hypothetical, and
then states that "It may be that rather than seeing Bridei as the son
of a Briton inheriting the Pictish kingdom we should see the second
dynasty of Dumbarton as Pictish in origin." Of course, given the
fragmentary nature of the genealogical evidence for Pictish kings, it
is not surprising that a hypothetical table can be constructed that is
consistent with patrilineal inheritance, but there are problems that
go beyond this speculation. In particular, the suggestion that the
so-called "second" dynasty of Strathclyde (which is in fact a cadet
branch of the "first" dynasty in the earliest genealogical
manuscripts) was of Pictish origin has every appearance of being
motivated by the necessity of making Bile a Pict (in order to deny
that Brude's claim to the Pictish throne was through his mother)
rather than being supported by any significant evidence. Second, as
pointed out by Miller (1978), it is quite cosistent with matrilinear
succession that a king could be the paternal grandson of a previous
king, and her conjectural tables show this possibility. The fact
remains that at least two Pictish kings (Talorcan son of Eanfrith and
Brude son of Bile), plus other examples which are at least plausible,
appear to have been the sons of non-Picts, and that the simplest
explanation for this is that their claim to the throne came through
their mothers.
---------------
end part 2 of 3
Stewart Baldwin
succession, as outlined below. Although my error does not affect the
argument, I wanted to get it corrected anyway. The relevant part
read:
> ... Defining "fratruelis" to mean "sons of two
>brothers" would require Bile to be a brother of Oswy, a very unlikely
>relationship, and defining it to mean "sons of two sisters" would mean
>that Brude and Ecgfrith would have to have the same maternal
>grandparents. Now, the maternal grandfather of Ecgfrith is known to
>be the famous Edwin of Deira, king of Northumbria, so this
>interpretation would require Brude's mother to be a daughter of Edwin,
>a relationship for which there is no other evidence.
>
>The important point, however, is that either one of these strict
>interpretations would require Brude's claim to the Pictish throne to
>come through his father Bile.
The point is that it is not true that the first of these alternatives
would require Brude's claim to the Pictish throne to come through his
father. Since I think that those on both sides of the argument would
agree that Bile and Oswy were not brothers, this does not affect the
argument in any way. However, in order for the last sentence quoted
above to be correct, it has to be changed to:
"The important point, however, is that the second of these strict
interpretations would require Brude's claim to the Pictish throne to
come through his father Bile."
Stewart Baldwin