Archiver > GEN-MEDIEVAL > 1995-12 > 0819328964

From: Chris Bennett <>
Subject: The Relationship of Basil I to Leo V
Date: Mon, 18 Dec 1995 15:22:44 PST

This mailing summarises issues relating to the proposal that
Basil I (Byzantine emperor from 867 to 886) was descended from
Leo V (emperor from 813 to 820), and a related issue concerning
Leo's marriage(s). It arises from an off-line discussion between
myself and Don Stone. Apart from the conclusion, it is intended
to be a neutral summary. Any inadvertent biases apparent to the
reader are mine, not Don's.

Background and Significance

The main point at issue is whether Basil I is reasonably
descended from Leo V. This has important implications for the
possibility of tracing descents to antiquity through early
Armenian nobility. The details of these possible descents are
given in Settipani's study [3]. In summary, Anna, daughter of
Leo VI (emperor from 886 to 912), very probably married the
Western emperor Louis III the Blind, and western European
descents can be traced from his son, most probably by Anna,
Charles Constantine. Leo VI's mother was Eudocia Ingerina. His
father was either her lover Michael III or her husband Basil I.

If Leo VI was the son of Michael III then a very plausible
connection to the Armenian noble family of Mamikonian exists
through Michael's mother the empress Theodora, daughter of the
turmarch Marinos, probably brother of Manuel Mamikonian.

If Leo VI was the son of Basil I then Armenian connections are
much murkier, though they undoubtedly exist. A summary of our
knowledge is given by Adontz in [1]. The patriarch Photius
prepared a genealogy, now lost, allegedly demonstrating that
Basil was descended from the Arsacid Armenian king Tiridates, the
first Christian ruler of Armenia. A summary account of Basil's
distant origins, which does not mention Photius' work, is given
by the emperor Constantine VII, son of Leo VI, who makes Basil a
scion of the Arsacid line, but this account does not seem very
consistent with known history.

It is possible that Basil may be of direct Mamikonian descent
since his grandfather Maiactes (or great-grandfather -- see point
(4) below) bears a characteristic Mamikonian name (Hmayek).
Unfortunately, there is no other indication of Maiactes'
immediate origins. Adontz suggests he was the son of a
Mamikonian prince Artavazd, son of Hmayek, but this is a guess
based on the name Hmayek and on chronological propinquity.
(Settipani [3] argues that this same prince Artavazd is likely to
be the father of Marinos (Mamikonian), maternal grandfather of
Michael III.)

If Basil was descended from Leo V, however, then a stronger case
can be made to connect Basil to the Mamikonian family, or at
least to the general complex of high Armenian nobility of the 4th-
8th centuries. Leo V is very probably a member of either the
Artsruni or the Gnuni family, and his wife Theodosia is said to
have been a Kamsarakan. Although these connections are not
traceable in detail, connections between the Mamikonians, the
Arstruni and the Kamsarakans are known. Thus, if Basil I is
descended from Leo V then a noble Armenian descent is in
principle traceable for Leo VI regardless of who his father
actually was.

The proposal that Basil I was descended from Leo V was made by
Adontz [1]. While his arguments in the same articles that Basil
was of Armenian origin have been generally accepted, this
specific thesis appears to have been ignored by most
Byzantinists, except Toumanoff, as far as I can tell from a non-
exhaustive survey. For example, Ostrogorsky's discussion of
Basil I's origins [2], which specifically cites Adontz' articles,
does not mention this thesis.

Adontz' thesis is based on Constantine VII's account of Basil I's
origins, which may be summarised as follows:

Basil's grandfather, Maiactes, of noble Armenian descent
(allegedly Arsacid - see above), married the daughter of an
Armenian resident in Constantinople called Leo, in the reign of
Constantine VI (780-797). As a child, Basil was captured with
his parents in the sack of Adrianople by the Bulgar khan Krum
(813). His parents died in captivity, but some 8-10 years after
their death Basil was released and made his way to
Constantinople, where he eventually came to the attention of the
imperial family, and finally of the emperor Michael III (842-867).

Adontz' proposal is that the Armenian Leo whose daughter married
Maiactes is the later Leo V, the Armenian. It should be said
that Adontz presents this tentatively, as an "audacious"
hypothesis which would nevertheless explain some apparently
curious facts, notably Constantine's silence regarding the names
of Basil's parents. The following list summarises the arguments
for and against the proposal. Most of the points below may be
found in the references; some of them are mine.

Summary of arguments favouring Adontz' proposal

1) Many names are shared between Leo V's and Basil I's families
(Leo, Smbat, Bardas, Basil).

Counter: These names are not particularly unusual Armenian
names. There are an equal number of other names which are not
shared. In Leo's family, we have Gregory and Theodosius; in
Basil's we have Marianos, Alexander and Stephen.

2) Maiactes and Leo V were of similar social standing, being sons
of Armenian nobility.

Counter: we know of other Leos of Armenian origin in this
period, e.g. Leo Skleros (see point (6) below). Many Armenians
came to Constantinople in the 780s.

Summary of arguments against Adontz' proposal

3) Constantine VII does not say that the Armenian Leo was Leo V
the Armenian.

Counter: Leo V was an iconoclast, so it was not politically
desirable for Constantine to draw attention to him too
explicitly. It is notable that Constantine does not name Basil's
father (whose name is not in fact known - Adontz suggests
Constantine, the name of Basil's eldest son, on the principle of
papponymy), and his mother's name Pancalo is only mentioned in
passing, in a completely unrelated context. Adontz' inference is
that Constantine VII had some reason not to mention their names.

4) Leo V's eldest son Smbat was still a boy on Leo's accession in
813, since Michael I Rhangabe (reigned 811-813) had become his
godfather some time between 802 and 811. Since Smbat took part
in ceremonial functions in 815, he was probably born c803,
shortly after the failure of the rebellion of Bardanes Turcas.
Leo had three younger sons after Smbat. It is thus very unlikely
that Leo had a marriageable daughter in the period 780-797.

Counter: Constantine VII has certainly got his family chronology
wrong. By his story, Basil I would be about 30 years older than
Michael III (born 840), but this seems very unlikely. There are
several indications of this. Most significantly, before
assassinating him, Basil was very close to Michael, e.g. engaging
in many escapades with him, and marrying Michael's mistress
Eudocia Ingerina to provide her with respectability, while taking
as mistress Michael's sister Thecla. Such a close relationship
indicates that their ages were actually very close. It is most
likely that Basil was born only a few years before Michael.
Since Constantine VII was retailing his own immediate family
history, however, his account is almost certainly factually
based. It makes more sense to suppose that it was Basil's father
who was swept up in the sack of Adrianople as a child, not Basil
himself, and that his grandmother was not married in the reign of
Constantine VI but was born in that reign, probably late. A
birth date of c795 would make her only 7 years older than Smbat
and would allow for a marriage to Basil's grandfather c810-812,
i.e. just before the Bulgar war; Basil's father could have been
born shortly thereafter, and would thus be about 20-25 years old
at Basil's birth.

Counter-counter: We don't know the birthdates of Leo V's younger
sons, but it seems reasonable to suppose that the youngest was at
least a toddler of 2 or 3 when Leo was assassinated in 820. This
gives a maximum age spread of about 14 years over 4 sons -- at
most 4.5 years between successive births. The proposed 7 year
difference between the hypothetical daughter and Smbat is nearly
twice this number.

5) Treadgold [4] proposes that Leo V married "Barca", a daughter
of Bardanes Turcas, strategus of the Anatolics, whom Leo started
serving under in 802. If this is correct, then (barring yet a
third, unknown, earlier marriage) Smbat was not only his oldest
son but was his oldest child.

Counter: Treadgold's case has been strongly criticised by Turner
[5]. In fact, Leo V's only known wife is his empress Theodosia.
The date of the marriage to Theodosia is unknown.

Counter-counter: Let us grant for the sake of argument that
"Barca" was a nickname for Theodosia. She was also the daughter
of a prominent man, Arsaber the quaestor, chief legal official of
the empire. Her marriage to Leo is also unlikely to have
occurred before Leo became noticeable himself, i.e. until after
his promotion by Turcas in 802 at the earliest, and probably
before Arsaber's rebellion in 808. That is, it doesn't matter
whether Treadgold is right - what he highlights is that both
candidates for Leo's wife before 813 have fathers who were too
prominent for a marriage before 802.


This completes the summary of the arguments for and against
Adontz' thesis. Although the Treadgold thesis of Leo V's
multiple marriages is only tangentially relevant to the main
point at issue, it has some interest. What follows summarises
the arguments for and against this thesis.

Summary of arguments favouring Treadgold's proposal

6) There is a well-reported story of a prophecy that Michael II
and another officer who served under a strategus of the Anatolics
would become emperors, whence the strategus married his two
daughters to the two officers. Since only successful prophecies
are worth recording, the prophecy must be true. Michael II
married Thecla, the daughter of Bardanes Turcas, strategus of the
Anatolics. Leo V is the only other future emperor who served
under Turcas, therefore he must have married the other daughter.

Counter: The story is evidently a folk-legend, and is only
evidence that Leo V was popularly believed to have married a
daughter of Turcas. Other explanations are possible, for example
that Leo V has been confused with some other Leo, less well-
known, who married the other daughter of Bardanes Turcas. Turner
[5] points out that an anonymous daughter of Turcas is associated
with the saint Theodore Stylites, and argues that she may
reasonably be identified with his patron Irene, who married an
Armenian Leo, strategus of the Peloponessus, probably to be
identified with Leo Skleros, strategus of the Peloponessus in 811.

7) Leo's wife was called "Barca" by the empress Procopia in 813.

Counter: "Barca" is not a name, but an insulting nickname,
roughly translatable as "waddler" "fatso" or "fat sow".
Procopia, whose husband was being deposed at the time, was not
inclined to be kind to the incoming empress.

Summary of arguments against Treadgold's proposal

8) Leo V is named as "exadelphos" of Bardanes Turcas' son
Bryenios. The term normally means cousin or nephew. It is
unlikely that Leo would have married a cousin.

Counter: Treadgold argues that "brother-in-law" better fits the
context in this case.

Counter-counter: Treadgold does not explain why, and adduces no
other examples of this usage.

9) Only Theodosia is named as an empress in subsequent lists of
empresses, but "Barca" was Leo's wife at the moment of his

Counter: Leo could have divorced Barca in a form of disgrace
(e.g. divorce for adultery) that permitted her exclusion from
official lists of empresses.

Counter-counter: There is total theological silence on the
divorce of an empress, which is very unlikely.

10) Theodosia was the daughter of Arsaber, who rebelled in 808
and failed in a bid for the throne. The marriage of Leo to
Theodosia therefore almost certainly occurred before 808.

Counter: Once Leo was emperor and divorced he could marry
whomever he wanted.


I entered this discussion thinking that Adontz' thesis, which I
had first encountered in the late 1970s and had not re-examined
for many years, was very unlikely. I also thought that
Treadgold's thesis was very plausible, which reinforced my doubts
about Adontz. After this review, I now think that Treadgold is
probably wrong, but I still think that the case for Adontz'
thesis does not hold up. What do others think? I welcome
comments, particularly from any Byzantinists or Armenianists who
may be monitoring this group.


[1] N. Adontz: L'age et l'origine de l'empereur Basile I (867-
886) [Byzantion 8:2 (1933) 475 and 9:1 (1934) 223]

[2] G. Ostrogorsky: History of the Byzantine State, tr Rutgers

[3] C. Settipani: Nos ancetres de l'antiquite, Paris 1991.

[4] W. Treadgold: The Byzantine Revival 780-842, Stanford 1988.

[5] D. Turner: The Origins and Accession of Leo V (813-820)
[Jahrbuch der Osterreichischen Byzantinistik 40 (1990) 171]

This thread: