Archiver > GEN-MEDIEVAL > 1996-06 > 0834198614

From: Stewart Baldwin <>
Subject: Comments on "Iberian route" DFA line
Date: Sat, 8 Jun 1996 01:50:14 GMT

Rafal Prinke <> wrote:

>I thought I might send this line (which I mentioned previously)
>for possible comments. It comes from a German article by
>W. Regula, Genealogie Bd.22 H.3/4 1995
>Sorry it is partly in Polish - but hopefully it is not difficult
>to read. Regula shows two other Western descents from (9).

I made the comment "No way" in a short posting a couple of days ago,
and now that I have some time, I will back that comment up with some
reasons. One obvious observation which has been repeated many times
(and is going to get repeated again) is that if any one of these links
is invalid, then the whole chain breaks. What we have here are a
large number of unproven links, a significant number of which are very
improbable, and the number of bad links is large enough that there is
no significant chance that the line is correct. In some of these
claimed Descents From Antiquity (DFA's for short), I have seen excuses
given that look something like "Well, maybe this line isn't exactly
true, but if one of the links is wrong, then someting close to it is
true." While I am not a big fan of what you might call "discontinuous
genealogy", I am willing to accept such things tentatively if (a) the
number of generations is reasonably small, and (b) clear evidence is
given that there was an actual direct descent involved (as opposed to
a cousin relationship which might ruin a key link). I have seen too
many arguments that look like "Let's see, we know he was a member of
this family, and it looks like he might have been in this generation,
so lets assign him as a son of this guy" (translation: "Let's guess").
This problem is especially serious in the Mamikonid part of the
genealogy, but it is also a problem for other families here, like the

The alleged genealogy given below follows what I am calling the
"Iberian route" (i.e., it passes through the early kings of Iberia),
to distinguish it from other DFA attempts, such as Settipani's. I
will discuss the Iberian route in detail, and then give a less
detailed discussion of Settipani's version at the end. In order to
emphasize how many weak links there are in the chain, I have "cut"
quite a few of them, with a solid line (________) between two
generations, where the evidence linking those two generations is
completely unacceptable in my opinion. There are a number of other
places where the evidence is not as good as it could be, and I have
painted a dotted line (- - - -) between these links (to prepare them
for possible future cutting if the evidence turns out to be weaker
than that was believed). The generations without dotted or solid
lines between them appear to be OK, but I have not carefully checked
the evidence for all of them. In all cases, I have explained my
reason for a dotted or solid line (but not always in the same
paragraph). Since I am not that familiar with the Byzantine part of
the genealogy, I have left off that part, and started with
"generation" 16. I suspect that there are serious problems with some
of those links too, and Todd Farmerie has already pointed out one such


>16 Hmayeak Mamikonian.

>17 N. Mamikonian.

Pure guesswork, as is obvious from the unnamed generation. The
problem of guessing relationships based on a family name alone is
especially bad when you need to connect to a key marriage, as is the
case here. This is not like the case of the later heads of the
Mamikonid family discussed later, who can be reasonably assumed to
descend from the key marriage. However, this Hmayeak Mamikonain could
just as easily be descended from a side branch of the Mamikonids which
does not come through the key marriage at generation 19, thereby
breaking the whole chain.

Generations 18 through 24 appear to be OK, and form an important part
of many claimed DFA's.

>18 Hamazaspian Mamikonian.
>19 Sahakanoysch Gregorid.
> x Hamazasp I Mamikonian, wielki konnetabl 387-432, s.
> Artavasdes-Artaschir III (?) Mamikonian.
>20 Vasak I Wielki Gregorid, ksiaze Gruzji 373-438, + 9 VII 438.
>21 Narses I Wielki Gregorid, ksiaze Gregoridow 355-359, 367-373, *
> before 330.
> x Sandoukht Mamikonian, c. Vardan I Mamikonian.
>22 Athenogenes Gregorid.
> x Bambischu Arsakid, c. Chosroe III Arsakid.
>23 N. Arsakid.
> x Hesychius (Yousik) I Gregorid, prymas Armenii 342-348, * before
> 305.
>24 Tiridates IV Wielki Arsakid, pierwszy chrzescijanski krol Armenii
> 314, wyst. 298-330
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>25 Chosroe II Waleczny Arsakid, krol zachodniej Armenii 279/280-287,
> zamordowany 287.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>26 Tiridates II Arsakid, wyst. 216/217-252
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>27 Chosroe I Arsakid, krol Armenii 191-216/217.

There are some disagreements about the third century Arsacid kings of
Armenia, thus the above dotted lines. Settipani has a discussion of
some of the other versions.

This link is so bad that I have cut it twice. (The link between
Chosroes and Vologases is less supicious, but that is not the
important link.) Sources of the eighth century and later state that
king Rev I of Iberia was the son of an Armenian king by the daughter
of king Pharasmanes III of Iberia. That the Armenian king in question
was Vologases is a conjecture on Toumanoff's part (see below for more
precise citation). Settipani (p. 73) suggests that Pakoros was the
Armenian king who was the father of Rev (as already pointed out by Don
Stone). So, in order to accept this link, you have to believe three
things: (1) that the *very* late Iberian sources can be trusted to
give accurate information for this period (despite evidence to the
contrary which I will give below), (2) that Toumanoff's guess (and it
is no more than that) was correct about which king, and (3) that
Chosroes of Armenia was the son of Vologases by the same mother as Rev
(therby ignoring the fact that multiple marriages were the rule rather
than the exception).

>28 N. z Iberii (Gruzji).
> x Vologases II (IV) Arsakid, krol Armenii 180-191 i arcykrol
> 191-207/208.

We are now dealing with the early kings of Iberia (in modern Georgia -
the country formed out of the former USSR, not the American state),
and this part will require a more lengthy discussion of the sources.
Whatever the immediate source of these Iberian generations might be,
it is clear that the basic secondary source for most of this
information is the article "Chronology of the Early Kings of Iberia",
by Cyril Toumanoff, which appeared in the journal "Traditio", vol. 25
(1969), pp. 1-34. (The chronology of the Iberian kings given here did
not exist until Toumanoff wrote that article.)

Toumanoff based his chronological reconstruction on several
assumptions, the most important of which was the assumption that the
native sources (all ca. 790 and later) had enough reliable information
that the chronology and genealogy could be reconstructed. No evidence
was offered as to why an eighth century source should be considered
reliable for the history of centuries earlier. It was simply assumed
that this was the case, and conclusions were drawn.

Now, it seems to me that before a source of this type is used, there
ought to be some serious discussion about the reliability of the
source, and that if this is done, the obvious conclusion is that the
native sources are not reliabable for the period crucial to the
genealogy under discussion, and that the resulting genealogy has a two
hundred year gaping hole. Here are two very good reasons why the
native sources should not be considered reliable for the period in
(1) From time to time, contemporary sources from other countries
gave the names of various Iberian rulers, although it is not enough to
get any kind of continuous account. In such a case, a good test to
see how reliable the noncontemporary sources are is to compare them
with what the reliable contemporary sources say. When this is done,
the match is not very impressive. There are kings who appear in both
sources at roughly the same time, but most of the kings in the native
king-lists do not appear in the contemporary sources, and the
contemporary sources show quite a few kings who do not appear in the
king-lists. Toumanoff's solution to this is generally to assume that
a king had two different names, and assign the king of the
contemporary records as being the same person as somebody on the
king-list at about the right time (as with Pharnabazus and Bartom
below, and others), but there is one period for which he abandons the
native sources entirely and goes with the contemporary evidence. For
the first century and part of the second, the native sources have a
very artificial "diarchy", where the kingdom was divided between two
branches, in which the kings just happened to die and be succeeded by
their sons at the same time for several consecutive generations.
Since this was blatantly contradicted by the native sources (and is
absurd anyway), Toumanoff was forced to admit that the native sources
were not reliable for this period, and used the contemporary sources
to fill things in. However, when Toumanoff was missing a genealogical
detail that the contemporary sources did not provide (the link between
generations 31 and 32), that did not stop him from using the obviously
unreliable native sources to fill it in. Furthermore, Toumanoff
assumed that the period before the "diarchy" was reliably reported
from the native sources, and drew important conclusions from them.
Toumanoff's approach was apparently to accept every item in the native
sources which was not directly contradicted by the more reliable
contemporary sources, but it seems to me that a more reasonable
sonclusion is that the contemporary sources prove that the native
sources have little or no value for this early period, and that their
use as a source for genealogical relationships is therefore not
(2) The cultures of Iberia and Armenia were closely related, and
so a comparison with the case of Armenia is relevant. Armenia's
historical writing started much earlier than in Georgia, and much more
is known about its history (but they unfortunately don't say as much
as we would like about neighboring Georgia). Because Armenian history
is much better documented in the contemporary sources, a comparison of
the Armenian sources with the contemporary sources shows that the
Armenian sources are not reliable in detail for the second century or
earlier. (Although many events of the second century and earlier
mentioned in the native Armenian histories are clearly based on
historical events, and many of the individuals appear in contemporary
records, important details are frequently wrong.) If the native
Armenian records, with their older historical tradition, cannot be
relied on for the period in question, then it is doubtful that any
trust can be placed in the Iberian records for the same period.

Another problem with Toumanoff's article is the contortions he has to
go through to get his chronology. It is difficult to describe without
making this article much longer than it is, and it is already too long
as it is (and we aren't done yet - are you still awake?;-) ). It will
suffice to say that he is frequently forced to alter the numbers which
his sources give, in order to get rid of all of the contradictions,
and his manipulation of reign-lengths reminds me of the accountant who
is desperately trying to juggle the books before the auditors arrive.
I suggest reading his article to see what I mean. (Since the journal
"Traditio" is available even in Auburn University's mediocre library,
I assume it should be easy to find in most good university libraries.)


>29 Pharasmanes III z Iberii, krol Iberii (Gruzji) 113-185.

Attested in contemporary sources, which do not give his relationship
to any earlier or later Iberian kings.

>30 Rhadamistes z Iberii, krol Iberii 132-135.

Not mentioned in any source earlier than the eighth century. He is
called "Adam" in the native sources, but Toumanoff changed his name to
Radamistus, assuming that the name Adam was identical to the name
Radamistus which was borne by a son of Pharasmanes I in the
contemporary records.

>31 Pharasmanes II z Iberii, krol Iberii 116-132.
> x Ghadana Artaxiad, c. Vologases I (?) Artaxiad.

This supposed marriage is not mentioned in any source earlier than the
eighth century, and these sources do not give the name of Ghadana's
father (which is a guess on Toumanoff's part). Vologases was an
Arsacid, not an Artaxiad. The connection of Pharasmanes to his
alleged father comes from the "diarchy" part of the native sources,
which Toumanoff admitted was unreliable (but used anyway).

These three kings of Iberia are confirmed in contemporary records, but
the dates come from Toumanoff's number juggling, and cannot be

>32 Amazaspes I z Iberii, krol Iberii 106-116.
>33 Mithridates I z Iberii, krol Iberii 58-106.
>34 Pharasmanes I z Iberii, krol Iberii 1-58.

Pharasmanes was the brother of Mithridates, king of Armenia, but the
contemporary records do not name their parents.

>35 N. z Iberii, ksiezniczka iberyjska.
> x Kartam Kondjid.

Only mentioned in the late Iberian sources, which make husband
(Kartam) and wife (N of Iberia) first cousins *three* times removed.
(He must have been a *very* old geezer when he married that
youngster.) Yet one more indication of the unreliability of these
sources. The people in this generation probably never existed.

>36 N. Artaxiad.
> x Pharnabazes II z Iberii, krol Iberii 63-30 pne.

Pharnabazus is attested by contemporary sources. However, the king of
Iberia during this period is called Bartom in the Iberian sources,
which don't mention a Pharnabazus reigning at this time. Toumanoff
"solves" this by making Pharnabazus and Bartom the same person,
without any other reason than the fact that his number juggling had
the traditional king Bartom reigning at the time that the contemporary
records have Pharnabazus. The claim that Bartom married an Armenian
princess is also from the late Iberian sources, and the claimed
identity of her unnamed father is the usual "let's pick somebody who
is about the right time" guess.

The next three generations might be OK, but there does not seem to be
complete agreement on early Artaxiad genealogy.

>37 Tigranes II Wielki (?) Artaxiad, krol Armenii 95-56 pne.
> x Kleopatra z Pontu, c. Mithridates VI Eupator z Pontu.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>38 Tigranes I Artaxiad, krol Armenii 123-93 pne.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>39 Artaxiades Artaxiad, seleukidzki strateg Armenii p. 200 pne i krol
> Armenii 188-p.161 pne., zapewne syn Zachariasza

Although the father of Artaxias was named Zariadres (as confirmed by
an inscription), it is very doubtful that his father was the same
Zariadres who ruled Sophene as a contemporary of Artaxias. The
otherwise unknown Zariadres who was father of Artaxias was probably a
prince of the Orontid dynasty, but attempting to assign a father for
him would be pure guesswork. Note that Settipani (who is not shy
about making conjectures) does not even attempt to assign a father to
Zariadres (Settipani, p. 100).

>40 Zariades Orontid (?), strateg Sophene p.200-190 pne i krol Sophene
> 190- pne (?).

>41 Xerxes I Orontid, krol Armenii.
> x before 212, Antiochis Seleukid, c. Seleukos II Kallinikos
> Seleukid i Laokide Seleukid.

No primary sources give the parentage of Xerxes. Several guesses have
been made as to his parentage, of which this is one. (Settipani p.
100 makes the same guess, with a dotted line on his chart to indicate
a conjecture.)

The following generations come from the inscription at Nimrud-dagh, in
which king Antiochus I of Commagene (1st century BC) mentioned a
number of his ancestors. Since the source is noncontemporary by 200
years, and the inscription has been interpreted in different ways,
some caution is required with regard to these generations.

>42 Arsames I Orontid, krol Armenii p.260-p.228 pne.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>43 Samos I Orontid, krol Armenii c260- pne (?).
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>44 Orontes III-II (?) Orontid, krol Armenii c317-c260 pne.

>45 Mithranes I Orontid, krol Armenii 331-c317 pne, perski satrapa
> Lydii i Mysii i satrapa macedonski.

The inscription at Nimrud-dagh gives ----anes, with several letters
missing at the beginning. Based on the last four letters of the name
only, ----anes has been identified with a certain Mithranes who
appears in the contemporary records, but I am not convinced by this
rush to identify ----anes with the most convenient person from other
sources whose name ends in the right letters. There is no reason to
assume that everyone in the inscription appeared in other sources.

>46 Orontes II-I Orontid, perski satrapa Armenii c361-331 pne.

See Settipani, who puts a dotted line on his chart (p. 100) between
this generation and Rhodogune.

These generations are OK.

>47 Rhodogune Achaemenid.
> x ok. 401, Orontes I Orontid, perski satrapa wschodniej Armenii
> 401-344 pne, + 344.
>48 Artaxerxes II Achaemenid, arcykrol Persii 404-358, + 358.
> ur. 452 pne
> x Stateira Hydarnid, c. Hydarnes III Hydarnid.
>49 Ochos (Darius II Nothos) Achaemenid, arcykrol Persii 424 pne, +
> 404.
>50 Artaxerxes I Achaemenid, arcykrol Persii 465 pne, + 424.
> x (1), N.N..
> x (2), N.N..
> x (3), Kosmartydene z Babilonii.
>51 Xerxes I Achaemenid, arcykrol Persii 485 pne, zamordowany 465.
> ur. 521/520 pne
>52 Dariusz I Wielki Achaemenid, arcykrol Persii 522 pne.

The conclusion is pretty obvious. The "Iberian route" is severely
flawed as an attempted DFA. Settipani apparently has the same
opinion, because his book on DFA's does not attempt to follow this
path. What about other possible DFA's? Well, I think that depends on
what you mean by a DFA. Does "antiquity" mean just ancient times, so
that getting back to 476 AD (the artificially chosen end of ancient
times) is good enough? I think most would say that it is not early
enough. Would getting back to the Seleucid dynasty be good enough?
(That would be good enough for me.) Or, will we be satisfied with
nothing less than the earliest recorded Pharaohs of Egypt and kings in
Sumeria. [I think I'll try to get my line traced back to "Lucy." ;-)]

The modern end is also important. Who is descended from antiquity?

(1) Is it good enough to have some individual living today who has a
validly documented line from antiquity? -or-

(2) Do we demand that we can get such a descent for an early Western
European ruler, so that many of the readers of this group can take
advantage of it? -or-

(3) Do we want a validly documented descent from _me_ back to
antiquity? [Clearly much more important than (1) or (2) ;-)]

Since the number of people interested in number (3) is relatively
small, and number (2) involves additional unsolved problems with the
Byzantines and/or Crusades that are too late to be directly related to
the problem at hand, I will address only number (1).

Samuel (d. 772), prince of the Mamikonids, had a daughter who married
prince Smbat of the Bagratids (d. 772), and had Ashot, prince of
Armenia, whose son Smbat the Confessor, prince of Armenia (father of
Ashot I, king of Armenia) had a daughter who married Bagrat I (d. 872)
of Georgia, ancestor of the later kings of Georgia, who ruled until
the nineteenth century, with descendants surviving today. Since
Samuel was the head of the Mamikonid family, there can be little doubt
that he was descended from the key marriage at generation 19, as
prince Hmayeak (brother of the famous Vardan and son of Hamazasp I and
Sahakanoysh) was the ancestor of the Mamikonid princes who followed
Hamazasp I and Vardan. However, the exact line of descent is still
unproven, as Settipani's tree involves several conjectures. It is not
clear whether or not enough evidence survives to completely nail down
the Mamikonid genealogy. Nevertheless, this gives a plausible line of
descent back to the Arsacid kings of Armenia. The Arsacids suffer
from the same problem as the Mamikonids, i.e., the contemporary
sources do not provide enough data to reconstruct their genealogy
without a significant amount of conjecture. If you accept one or
another of the conjectured trees for the main Arsacid line, this gets
you back to the 3rd century BC, which is pretty good. The big problem
is trying to get any further than back than the Arsacids with an
outline that is at least plausible, and that is where the big problems
come in, i.e., trying to find an Arsacid marriage which will take you
back to one of the earlier dynasties. The "Iberian route" is one of
these attempts, but based too much on late sources and guesswork to
take seriously.

What about other possibilities. The other main one which has been
suggested has been through the kings of Commagene. The kings of
Commagene are a very desirable dynasty to trace from, as they have a
certain descent from the Seleucid kings of Syria and a probable
descent from the Achaemenid kings of Persia. Unfortunately, the
attempts of Settipani to get back from the Arsacids to the kings of
Commagene are not convincing. The most obvious attempt, through the
marriage of the Parthian king Orodes to the daughter of king Antiochos
I of Commagene, is no good, because the later Parthian kings are known
to not descend from that marriage. Settipani's attempt tries a
different route, the key generations of which are given here, starting
with the earliest "generation." Settipani marks ALL of the links
below with dotted lines, indicating that thay are all unproven.

1. Antiochos I, king of Commagene, d. 36 BC
2. daughter, md. Artavazd, king of Media Atropatene, d. 20 BC
3. [Darius] (name, existence, and relationship all conjectured)
4. Vonones II, king of Parthia, d. 51 AD

Although an exact line of descent is still unproven, the descent of
the Arsacid kings of Armenia from Vonones is at least a reasonable
possibility. The descent of Antiochos of Commagene from the Seleucids
is certain, which takes his ancestry at least back to the 4th century
BC, and his descent from the Achaemenids of Persia, though having weak
links, is at least probable. Thus, the above four generation form the
key to Settipani's attempts to get a line from antiquity earlier than
the Arsacids. This material appears on pages 84-94 of Settipani's
book. The main points of Settipani's arguments are as follows:

(1) King Artabanos II of Parthia is said by Tacitus to have been an
Arsacid through his mother, and to have ruled Media Atropatene before
becoming king of Parthia. Tacitus also states that Vonones II ruled
Media Atropatene before his short reign as king of Parthia. When
Vonones died, his son Vologases succeeded to Parthia, and his son
Pakoros to Media Atropatene.

(2) Various pieces of evidence suggest that Artabanos II and Vonones
II were brothers, a conclusion which I find reasonable, though still

(3) Strabo, in his Geography, 11, 13, 1, describing Media, has a
passage which is important enough to quote in full, from the
translation given in the Loeb Classical Library edition:

..."The other part is Atropatian Media, which got its name from the
commander Atropates, who prevented also this country, which was a part
of Greater Media, from becoming subject to the Macedonians.
Furthermore, after he was proclaimed king, he organised this country
into a separate state by itself, and his succession of descendants is
preserved to this day, and his successors have contracted marriages
with the kings of the Armenians and Syrians, and, in later times, with
the kings of the Parthians."

That is pretty much it as far as evidence is concerned. The Armenian
marriage is known, for king Mithridates of Media (d. ca. 67 BC) was
married to a daughter of Tigranes the Great, king of Armenia. It is
reasonable to argue that the Parthian marriage is what resulted in
Artabanos and Vonones attaining the Parthian throne (so the later
"Arsacids" were apparently only Arsacids in the female line). There
is no evidence for the Syrian marriage other than Strabo, and
Settipani suggests that the Syrian marriage was between king Artavazd
of Media Atropatene and a daughter of king Antiochos I of Commagene
(located in Syria), and that they were the parents of a son who never
appears in any known records, and whom he supplies with the
conjectural name of Darius (because Artabanos II had a son of that
name, and a previous king of Media Atropatene also ahd that name).

So, in order to accept Settipani's account of these generations, we
have to assume that each of the following guesses made by Settipani
was correct:

(a) The guess that the marriage contract referred to a prince of
Atropatene marrying a princess of Syria, and not the other way around.

(b) The guess that "Syria" really means Commagene.

(c) The guess that Antiochos I was the father of the "Syrian"

(d) The guess that Artavazd of Atropatene was the groom. (Three
other kings of Atropatene are known between Mithridates and Artabanos,
any one of whom could just as likely be the groom.)

(e) Assuming that all of the above guesses are correct, we must also
believe the guess that the father of Vonones was a child of the above
marriage. (Of course, it is irrelevant whether or not Settipani's
guess about this father's name is correct.)

This is an awful lot of guesses piled on top of each other. It is
quite probable that Artabanos and Vonones were the sons of a prince
(or king?) of Atropatene by an Arsacid princess, but in my opinion it
is ridiculous to conclude on the basis of Strabo's inconclusive
remarks that the father of Artabanos and Vonones was the son of a
princess of Commagene. Also, it is rather disturbng that among all of
the possible ways in which Strabo's vague statement could be
interpreted, the one is chosen which gives the desired conclusion. It
is not a proper use of evidence to take the most desirable possibility
among numerous alternatives. I see no reasonable probability here
that this crucial part of Settipani's line is correct.

For those of you who made it this far, thanks for taking the time to
read these remarks. Any comments? Oops. Wait just a minute. Let me
crawl into this suit of armor here. OK, NOW I'm ready for comments.

Stewart Baldwin

This thread: