Anda di halaman 1dari 18

Fourth-Century Saxons

Author(s): Philip Bartholomew


Source: Britannia, Vol. 15 (1984), pp. 169-185
Published by: Society for the Promotion of Roman Studies
Stable URL: http://www.jstor.org/stable/526590 .
Accessed: 11/08/2013 06:49
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
.
Society for the Promotion of Roman Studies is collaborating with JSTOR to digitize, preserve and extend
access to Britannia.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
Fourth-Century
S a x ons
By
PHILIP BARTHOLOMEW*
N the received
his tory
of Brita ininthe fourth
century,
a
s pecia l pla ce
is
occupied by
the
S a x ons . It is a s s erted tha t the defens ive
s ys tem
of the S a x onS hore ex is ted to
prevent
them
from
la nding;
a nd it would s eem
tha t,
unlike
a ny
of the ba rba ria ntribes withwhom Rome
ma inta ined
unea s y
rela tions
a long
the Rhine a nd the
Da nube,
unlike eventhe
Goths ,
the
pos s ibility
of their
pea ceful
s ettlement withinthe bounda ries of the mos t a cces s ible Roma n
province
could not be enterta ined. Their
a tta cks , a ppa rently,
were s o
unpredicta ble,
a nd their
na ture s o
s a va ge,
tha t inthis ca s e a lone the
policy
of the Roma na uthorities wa s one of tota l
ex clus ion. In
367, moreover,
a fter
forming
a na llia nce withthe tribes of the northa nd the
wes t,
they
overwhelmed the Roma ndefences a nd
ra va ged Brita in;
a nd this da te ha s beena ccorded
the s ta tus of a
turning-point
in
fourth-century his tory.
But a fterwa rds the S a x onS hore forts
res umed their functionof
keeping
the S a x ons out.
Fina lly,
the Roma n
ga rris on
wa s withdra wn
a nd the S hore forts
a ba ndoned;
a nd this
permits
his toria ns to
a s s ign
a n
ea rly fifth-century
da te to the
beginning
of the
Anglo-S a x on
s ettlement.
Now the
picture pres ented
here is a na ttra ctive
one,
a nd its
a ccepta nce
ha s been
genera l.
Yet the evidence onwhichit res ts is
s lighter
involume a nd more
a mbiguous
in
qua lity
tha nthe
una nimity
of s chola rs would a t firs t
s ugges t.
The writer's
purpos e
is to look
a ga in
a t the
evidence for the S a x ons '
peculia r cha ra cter,
a nd their clos e connectionwithBritis h
his tory
in
the fourth
century.
The res ults of this ex a mina tionwill be inimica l to the
a ccepted view;
but
further
ques tions
will be
thereby ra is ed,
a nd
thes e, too,
will be s eento
require
a na ns wer.
S AXONS IN THE ROMAN ARMY
But
firs t,
it is
neces s a ry
to dra w a ttentionto the evidence whichhints
tha t, throughout
the
fourth
century,
the S a x ons did
enjoy
a norma l
his tory.
It is
s lightly s urpris ing,
a fter the horror
s tories which
a ppea r
inthe
s econda ry works ,
to dis cover tha t
they provided
a unit inthe la te
Roma n
a rmy;
but inthe Notitia
Dignita tum
we find the a la
prima
S a x onumn,
inthe comma nd
of the Duke of
Phoenicia ,
s ta tioned a t Verofa bula
(Not. Dig., Or.,
x x x ii.
37). Ea rlier,
inthe
mid-fourth
century,
the
S a x ons ,
withthe
Fra nks , fought
with
Ma gnentius (Julia n,
Ora tion
i,
34D).
In
358,
Zos imus des cribes Julia n's
ca mpa igna ga ins t
a s ubdivis ionof the
S a x ons ,
who
ha d s eized Ba ta via from the S a lia n
Fra nks ; erroneous ly,
he des cribes them a s
'Qua di'
(Zos imus
iii.
6-7, es pecia lly
6.
I-2).1
After the
fighting,
recruits were enrolled inthe
a rmy
from
the
'Qua di',
the other inha bita nts of Ba ta via
(who
a re likelies t to ha ve beenS a x ons
a ls o),
a nd
the S a lia n
Fra nks ;
a nd this beca me a
genera l pra ctice,
whichrema ined inforce a t lea s t until the
ea rly fifth-century
da te of Zos imus '
s ource, Euna pius (Zos imus
iii. 8.
1).
It ca nbe s eenin
opera tion
in
370 when,
a fter
s us ta ining
a defea t a t the ha nds of the
genera l
S everus s omewhere
innorth-ea s t
Ga ul,
the S a x ons a re a llowed to
depa rt 'pa ctis indutiis ,
et da tis ex condicione
propos ita
iuvenibus
multis ,
ha bilibus a d militia m'
(Ammia nus
Ma rcellinus x x viii.
5. 4);
a nd the
*
I wis htotha nk Profes s or
Fergus Milla r,
Dr Gra ha m Piddock a nd Mrs HelenBrock for the
help
a nd
encoura gement
which
theyprovided during
the
completion
of this
pa per.
1
The fa ls e identifica tionis ma de
ex plicit
a t iii. 6. 1:
XceovEs
...
Kouva ous
poipa v
o
q)x v 6vra s
...
EKx TrE'PTrOU1.
169
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
I170
PHILIP BARTHOLOMEW
pres ence
of S a x ons onthe Pers ia n
ca mpa ign
of
363
ca n
perha ps
be inferred from Ammia nus
x x v. 6.
13-14,
where we rea d tha t
Jovia n,
in
res pons e
to dema nds from his
troops
tha t the
a rmy
s hould cros s the
Tigris ,
'id
impetra tur a egerrime,
ut mix ti
curm
a rctois
Germa nis ,
Ga lli a mnem
primi
omniurn
penetra rent'.
Ammia nus continues :
'Electique
s unt a d id
negotium ha biles , qui
ma x ima
pra e
ceteris
flumina
tra ns mea re in
regionibus genuinis
a
prima pueritia
s unt ins tituti'.
Ba ta via ,
where Zos imus
pla ces
the
S a x ons ,
is
certa inly
ina
region
of
ma x ima flumina ;
a nd it
ma y
be
s ugges ted
tha t S a x ons formed
pa rt
a t lea s t of Jovia n's a rctoi Germa ni.
Now a cons idera tionof the true
identity
of the
'Qua di'
who
a ppea r
inZos imus ma kes it
pos s ible
to ex tend our
knowledge
of S a x ons inthe fourth
century
intwo different directions .
Julia ntells us
tha t,
inthe
ca mpa ign
whichZos imus
des cribes ,
his
opponents
were the Cha ma vi
(Letter
to the s ena te a nd
people
of
Athens ,
280
B);
a nd it s eems bes t to a s s ume tha t the
Cha ma vi,
ina llia nce withthe
Cha uci,
who bea r a
s tronger
res embla nce to the S a x ons inra ce
a nd to the
Qua di
in
na me,
a re the
'Qua di'
of
Zos imus .2
The
Cha ma vi, too, a ppea r
inthe
Notitia
Dignita tum:
the Cohors undecima
Cha ma vorum
is s ta tioned a t
Pea mu,
in
upper Egypt,
under the comma nd of the Duke of Thebes
(Not. Dig., Or.,
x x x i.
61).
Two lines la ter there
occurs a unit of
Ala ma nni;
inbetweenthe Ala ma nni a nd the Cha ma vi we find a
s lightly
mys terious
'Cohors nona Tza norum'. Now it ha s beens howntha t the s cribes of our four ma in
ma nus cripts
of the Notitia
Dignita tum
were
copying
from a n
a rchetype
writtenina nins ula r
ha nd.3
Inthis
s cript,
'n' bea rs a
very s trong
res embla nce to
'r',
a nd 'x ' is difficult to
dis tinguis h
from either letter. The initia l 'tz' is rendered a s 'z' inone of the
ma nus cripts (V),
a nd els ewhere
inthe Notitia 'z' is a nerror for 's '
(Or.,
x li.
37).
It
ma y
be
s ugges ted tenta tively
tha t the correct
rea ding
of Not.
Dig., Or.,
x x x i. 62 is 'Cohors nona S a x onum'. As withthe firs t a la of S a x ons a t
Verofa bula ,
the na me of this
unit,
if
correctly res tored,
would
imply tha t,
inthe fourth
century,
the ea s tern
a rmy
conta ined a number of a dditiona l units of S a x ons whichthe Notitia
Dignita tum
did not mention.4
S econdly,
this clos e connectionbetweenthe Cha ma vi a nd the S a x ons
permits
us to ex tend
the involvement of the la tter inthe Roma n
a rmy
ba ckwa rds intime. Hoffma nn
s us pected
tha t
the
pres ence
of
troops
from north-wes t
Germa ny
inthe
a rmy
of the ea s t went ba ck to the
Tetra rchy;
inthe ca s e of the Cha ma vi this is confirmed
by
a
pa pyrus
dis covered a t
Pa nopolis ,
in
which,
inthe third
cons uls hip
of Cons ta ntius Chlorus a nd Ma x imia n
(which
occurred in
300oo)
reference is ma de to the
pa y
of the s oldiers of the
XcpA)pT
t-S
Xqa poocpcov
...
.v
-Ea IOU.5
The identifica tionof the
Xca poS cbpot
of the
pa pyrus
withthe Cha ma vi of the Notitia
Dignita tum
is ma de certa in
by
the
occurrence,
inboth
pla ces ,
of the s a me loca tionfor the unit
(Pea mu)
a nd
the s a me number for the cohort
(the
eleventh).6
They
ca na ls o be tra ced a t a nea rlier
s ta ge
on
2
For the pres enta tionof this view, s ee the note by
Pa s choud toZos imus iii. 6
(F.
Pa s choud
(ed.),
Zos ime:
His toire
nouvelle,
tom.
2,
i
(Pa ris , 1979), 76-7).
3S ee
R. Irela nd's
cha pter,
'Tra ns mis s ionof the tex t' inM. W. C. Ha s s a ll a nd R. Irela nd
(eds .),
De rebus
bellicis
(Ox ford, 1979), es pecia lly54-5.
4If
this
hypothes is
is
gra nted,
a rea s onfor the
non-a ppea ra nce
of thes e units inthe Notitia
Dignita tum ma y
be
put
forwa rd. S t Ambros e
(Ep. 40, 23
=
PL
16,
coll.
Iio9-io)
records tha t S a x ontroops pla yed
a lea ding
pa rt
inthe
emperor
Theodos ius ' victories over
Ma gnus
Ma x imus inthe Ba lka ns in
388;
a nd Zos imus (iv. 30)
des cribes the tra ns fer of a
pa rt
of the
Egyptia na rmy
tothe Ba lka ns in
379
or
ea rly380.
Ontheir a rriva l (iv. 31. 1),
thes e menwere
a s s igned
tounits
a lrea dythere,
a nd the Roma n
s ys tem
of documenta tions eems toha ve broken
down:
pTns a Troypa qi
s E'71
Tov
Ev o7015 orpa TICOlTKoiS
a pt0pois
a va pEpoPEvcov
pvua 7rrPoVEvS .
If S a x onunits ha d formed
pa rt
of the
Egyptia nga rris on
a t a nea rlier
da te, theyma y
be a s s umed
(with
the
ex ception
of the unit
pos tula ted
for Not.
Dig., Or.,
x x x i.
62)
toha ve
pa rticipa ted
inthis movement of
troops
a nd toha ve los t their
identity
in
the
res ulting
a dminis tra tive confus ion.
,
D.
Hoffma nn,
Da s
s pditra mis che Bewegungs heer
lnd
die Notitia
Dignita tum (Duis s eldorf, 1969-70),
I, 14o
a nd
II, 48,
where the a uthor cites T. C. S kea t
(ed.), Pa pyri from Pa nopolis
inthe Ches ter Bea ttyLibra ry,
Dublin
(Dublin, 1964), 2,
11.
292
ff.
6 S kea t, op. cit.
(note 5), 151.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
FOURTH-CENTURY S AXONS
171
their
journey
to the
ea s t;
inthe
Pa negyric
to Cons ta ntius
Chlorus ,
delivered in
297,
we rea d of
their s ettlement in
Ga ul,
a nd their
s ubs equent
enlis tment: 'Ara t
ergo
nunc mihi Cha ma vus et
Fris ius ...
Quin
etia m s i a d dilectum
vocetur,
a ccurrit et
obs equiis
teritur et
tergo
cohercetur et
s ervire s e militia e nomine
gra tula tur' (Pa n.
La t. iv.
9. 3-4).
The na tura l
a s s umption
is tha t the
S a x ons of Not.
Dig., Or.,
x x x ii.
37,
the unit
pos tula ted
for x x x i.
62,
a nd the a dditiona l units
whos e ex is tence ha s been
inferred,
formed
pa rt
of the s a me movement. The
picture
which
emerges
is of a
his tory
of the S a x ons inthe fourth
century
whichis
longer,
more
ex tens ive,
a nd
more conventiona l tha ntha t whichis
us ua lly a s s igned
to them.
THE S AXON CHARACTER
The evidence for the
peculia r unplea s a ntnes s
of the S a x ons cons is ts of two
pa s s a ges
of
Ammia nus Ma rcellinus . Their
ex ceptiona l s a va gery
is a deductionfrom x x viii.
5. 1-7.
This
pa s s a ge
des cribes a S a x onincurs ioninto north-ea s t Ga ul in
370.
It wa s met
by Na nnenus ;
but
he found hims elf confronted
by
a
people
'a d mortem des tina ta e'
(x x viii. 5. 2),
a
phra s e
which
the Loeb
tra ns la tor,
J. C.
Rolfe,
renders
by
'res olved to
fight
to the
dea th'.7
After
being
defea ted ina
s kirmis h,
Na nnenus
a rra nged
for the
emperor
to
repla ce
him
by
S everus . When
the
la tter,
withhis
reinforcements ,
ca me into conta ct withthe
S a x ons , they ca pitula ted
a nd the
truce
(to
whichreference ha s beenma de
a bove)
wa s
a rra nged. S everus , however,
ordered his
mento a mbus hthe S a x ons ontheir
wa y
home. This
pla nnea rly
ended in
dis a s ter,
whena
prema ture
Roma na tta ck encountered
unex pected res is ta nce;
but a s econd unit
intervened,
a nd,
a fter a fierce
contes t,
the Roma ns 'cla us os hos tes eductis
gla diis
obtrunca ba nt:.
nec
quis qua m
eorum
genita les
revis ere
potuit la res ,
ne uno
quidem
ca edibus
concorpora lium s uperes s e
permnis s o'.
Ammia nus concludes
by offering
wha t
a ppea rs
to be a
jus tifica tion
for this ma s s a cre:
'Ac licet
ius tus quida mn
a rbiter
rerum,fa ctumn
incus a bitperfidum
et
deforme, pens a to
ta men
negotio,
non
feret indigne,
ma nurm
la tronum
ex itia lemn
ta ndem
copia
da ta
cons umpta m' (x x viii. 5. 7).
Now this fina l s entence
might
be ta kento
imply
tha t the S a x ons were s o
da ngerous
a foe
tha t
a ny
mea s ures ta ken
a ga ins t
them were
jus tifia ble, provided they
were
s ucces s ful;
a nd this
view is reinforced
by
Rolfe's
interpreta tion
of 'a d mortem des tina ta e'. But it is
ex tremely
difficult to reconcile this withthe fa ct
tha t,
a s s oona s the S a x ons ca me fa ce to fa ce with
S everus '
a rmy, they ga ve wa y. Further,
Rolfe's
rendering
of 'des tina ta e' a s 'res olved' is
very
s tra ined.
Des pite
the
pa ra llel
he
cites ,
the us ua l
mea ning
of 'des tina tus ' inAmmia nus is
'ma rked out'.8 The
pres ent pa s s a ge
is
quite na tura lly interpreted
a s
mea ning
tha t the S a x on
ra iding-pa rty
wa s ma rked out for
dea th;
Ammia nus will
s imply
be
a nticipa ting, by
this
phra s e,
the ca la mitous end to their
ex pedition.
And if this is
s o,
the
only
evidence inthis
pa s s a ge
for the
unique qua lity
of the S a x ons a s enemies will cons is t of Ammia nus '
s umming-up.
The interva l which
s epa ra tes
the
compos ition
of Ammia nus '
His tory
from the events which
it des cribes is not
pa rticula rly long;
a nd one s hould never los e
s ight
of the fa ct tha t he wa s
writing
for the benefit of a
public
to whom the ma in
milita ry
lea ders inhis na rra tive were
fa milia r
figures .
Their
cha ra cters ,
a nd their rea l
a chievements ,
were
well-known;
a nd it
ma y
be
s uppos ed
tha t
they
a ll
occupied pla ces
ona
genera lly a ccepted
s ca le of
popula rity.
The
ex is tence of this common
knowledge
a nd thes e common
prejudices
ma de it
pos s ible
for
Ammia nus to
indulge
ins a rca s m. At
times ,
he is
quite
bruta l: for
ex a mple,
when
des cribing
a
victory
over the
Juthungi
in
357,
he
gives
s cornful
empha s is
to the rumour tha t
Nevitta ,
who
wa s
widely dis liked,
ha d
a ctua lly
been
pres ent
a nd ha d
a cquitted
hims elf
credita bly.9
His
7Ammia nus Ma rcellinus ,
witha n
Englis h
tra ns la tion
by
J. C. Rolfe
(Ca mbridge, Ma s s ., 1935-40).
This
tra ns la tion
a ppea rs
on
pa ge
I6I
of vol.
3.
8 This ,
a nd the obs erva tions onAmmia nus '
s tyle
which
follow,
ha ve beenma de
pos s ible by
cons ulta tion of
G. J. D. E. Archbold's
microfiche,
Concorda nce tothe
His toryofAmmia nus
Ma rcellinus
(Toronto, 198I).
'
'Huic pugna e Nevitta , pos tea cons ul, eques tris pra epos itus turma e,
et
a dfuis s e
et
fortiter fecis s e firma tur'
(x vii.
6.
3).
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
172
PHILIP BARTHOLOMEW
trea tment of
Frigeridus ,
one of Va lens '
genera ls
in
Thra ce,
is
s lightly
more
gentle:
his relucta nce
to
enga ge
the Gothic
a rmy
is
repres ented
a s
ca ution,
but it is
impos s ible
to mis s the
irony
with
whichAmmia nus a s cribes his fa ilure to
pa rticipa te
inone of the ba ttles of the
Adria nople
ca mpa ign
to a n
a lleged
a tta ck of
gout.10
Now the
pres enta tion
of S everus is more res tra ined
a ga in:
but the
implica tion
of the
wa y
his
pa rt
inthe Ala ma nnic
ca mpa ign
of
371
is des cribed
is tha t he wa s
thought
to be a cowa rd:
'S everus , qui pedes trem
cura ba t
ex ercitum, perpens a
militum
pa ucita te,
territus
s tetit,
timens ne res is tere
nequiens ,
inruentium
opprimeretur
hos tilium
a gminum
mole'
(x x ix .
4-
3). Nothing
could be more
cowa rdly
tha nS everus ' ma s s a cre of the
S a x ons :
pres uma bly
he s a w it a s a
wa y
for a
s enior,
but
ineffective,
Roma nofficer to s core a n
ea s y milita ry
s ucces s . Ammia nus '
public
knew
this ; they
a ls o knew S everus '
reputa tion.
In
tha t
ca s e,
the
'jus tifica tion'
whichAmmia nus
gives
for S everus ' a ctions mus t be s eena s a
s a rca s tic
a ttempt
to ridicule a defence which
S everus ,
in
keeping
withhis
cha ra cter,
would ha ve
offered for the ma s s a cre. It is not to be ta ken
litera lly:
the 'ius tus a rbiter rerum' is a
litera ry
device round whichthe s a rca s m
revolves ,
a nd Ammia nus does indeed mea ntha t the ma s s a cre
wa s
'perfidumin
et
deforme'.11
There is no
s ugges tion
here tha t
s omething
inthe na ture of the
S a x ons
jus tified
the conduct of S everus .
The
unpredicta bility
of the S a x ons is derived from Ammia nus x x viii. 2. 12. We a re told tha t
'pra e
ceteris hos tibus S a x ones timentur
ut
repentini',
a nd it is s ta ted of the tribe withwhom
they
a re
being compa red
inthis
pa s s a ge,
'nec
quis qua m
a dventum eorum ca vere
potera t inopinum'.
But this tribe is the
Ma ra tocupreni;
their
s phere
of
opera tions
wa s centred on
Apa mea
in
S yria .
Inthe entire na rra tive of
Ammia nus ,
the S a x ons ha ve ma de
only
two
previous
a ppea ra nces ;
to the
rea der,
the
citing
of a notherwis e
ins ignifica nt group
to illus tra te a
point
in
ba rba ria n
milita ry
ta ctics comes a s a cons idera ble
s urpris e.
And the
s ingle ma nus cript upon
which,
withthe
ex ception
of four s hort
pa s s a ges ,
our entire
knowledge
of the tex t of Ammia nus
is
ultima tely dependent,
does not conta inthe word 'S a x ones '. This is the
ninth-century
Codex
Fuldens is ,
now Va t. La t.
1873,
a nd referred to a s 'V'. Like the
ma nus cripts
of the Notitia
Dignita tum
it
belongs
to a tra ditionwhichha d
pa s s ed through
a nins ula r
pha s e;
a nd it
pres ents
ma ny
difficulties . At this
point
inthe
ma nus cript
we find the
mea ningles s
's a x a ves tumentur ut
repentini'.
The tex t whichis
printed by
Cla rk inthe s ta nda rd editionof Ammia nus '2 a nd
by
a ll
other moderneditors is tha t which
a ppea rs
inthe editionof
Ghelen, publis hed
a t Ba s el in
July
1533. Now,
in
prepa ring
Books x iv-x x vi of his
edition,
Ghelenrelied
predomina tely
onthe
printed
editionof Petrus
Ca s tellus ,
whichha d
a ppea red
a t
Bologna
in
1517.
But for the
rema ining books ,
whichthe editionof Ca s tellus did not
conta in,
he wa s forced to us e Va t. La t.
2969 ('E'),
whichis a
fifteenth-century copy
of
V,
a nd the
ninth-century
Codex
Hers feldens is
('M'),
whichhe ha d obta ined from the
mona s tery
of Hers feld a nd whichis V's immedia te
a nces tor.13 Only
s ix s heets of M rema inin
ex is tence,'4
a nd its vis it to the
printing-s hop
inBa s el
ma y
well ha ve beenfa ta l. Ghelen's vers ionof the
pa s s a ge
inAmmia nus x x viii. 2. 12 does not
a ppea r
inE. Either it wa s derived from
M,
whichdoes not s urvive; or,
it is a
conjecture
of
Ghelen's own.
And,
of thes e two
pos s ibilities ,
the s econd is
by
fa r the more
proba ble:
for
10o
'
Verum a rticulorum dolore
Frigeridopra epedito,
vel certe
(ut
obtrecta tores
finx ere ma livoli),
morbum
ca us a nte,
ntie
ferventibus
proeliis
interes s et'
(x x x i. 7. 5).
"'
WhenAmmia nus s umma rizes this
ca mpa ign
a t x x x .
7.
8 his s a rca s m ha s wea kened: the S a x ons were
defea ted
'ma lefidoquidem
s ed utili commento'.
12
C. U. Cla rk
(ed.),
Ammia ni Ma rcellini rerum
ges ta rum
libri
qui s upers unt (Berolini, 1910-15).
13
S ee C. U. Cla rk,
The tex t tra dition
of
Ammia nus Ma rcellinus (New Ha ven, Conn., 1904),
Io
a nd pa s s im.
Cla rk, however, thought
tha t M a nd V were
copied
from a commona rchetype (cf.
a ls othe s temma printed
on
p.
v of his editionof Ammia nus ).
This wa s refuted in
1936:
s ee W. S eyfa rth,
Kliox lviii (1967), 214.
14
Thes e s heets conta infour
pa s s a ges :
x x iii.
6.37-x x iii. 6.45;
x x viii.
4. 21-x x viii. 5.2;
x x viii. 5.
II-x x viii.
6.5;
a nd x x x . 2.
5-x x x . 4.
2.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
FOURTH-CENTURY S AXONS
173
Ghelen's a ddictionto
irres pons ible conjecture
is
notorious .15
But it is
neces s a ry
to tes t the
a s s umption
tha t we a re
dea ling
witha
s ix teenth-century conjecture
ra ther tha na na uthentic
rea ding
tra ns mitted from a
ninth-century ma nus cript by a s king
whether a nother
rea ding might
not be more
a ccepta ble.
Now a
conjectura l
s ubs titutionof 'timentur' for 'tumentur'
would,
in
a ny ca s e,
s eem
rea s ona ble; but,
whenone reca lls Ammia nus ' ha bit of
us ing a djectives
in
pa irs ,
V's
'repentini'
s eems to s ta nd
a wkwa rdly, a lone,
a t the end of the s entence. Onthe two occa s ions when
Ammia nus us es
'repentinus '
of ba rba ria n
tribes ,
it
a ppea rs
a s the s econd of two
a djectives :
a t
x x x i. 2. 8 we rea d
tha t,
in
ba ttle,
the Huns 'a d
pernicita tem
s unt
levies
et
repentini';
a t x x ix .
5-
7,
Theodos ius , newly
a rrived in
Africa , s pecula tes
how 'hos tem
ca peret
dis curs a torem et
repentinum'.
Inthe
pres ent pa s s a ge,
the a ltera tionof 'ut' to 'et' would be
s imple.
It needs to be
a s ked whether the 's a x a ves ' of V could be
repla ced by
a nother
a djective.
Now in
this ,
a nd in
the
preceding,
s ectionthe cha ra cteris tic of the
Ma ra tocupreni
whichis
being
s tres s ed is their
dependence,
intheir
ca mpa igns , uponguile.
The word us ed
by
Ammia nus to des cribe the
s ucces s ful
a pplica tion
of
thought
to
milita ry
a ffa irs is
's a ga x '.6 'S a ga ces '
is no more
improba ble pa la eogra phica lly
tha nGhelen's
'S a x ones ';
if it is res tored inthe
tex t,
the
pa s s a ge
would rea d
'qua m
ob ca us a m
pra e
ceteris hos tibus
s a ga ces
timentur et
repentini'.
This would
not conflict withAmmia nus '
s tyle,
a nd would a ccord well withthe
genera l
s ens e of this
s ectionof the na rra tive. Ghelen's tex t s eems to be
s pecula tive
a nd
ought
to be s et
a s ide."7
But the
propos ed
res tora tionmea ns tha t the ta ctics of the
Ma ra tocupreni only
a re
being
des cribed inthis
pa s s a ge;
there is no mentionof the
S a x ons ,
a nd no
grounds
a re
provided
for
a s s igning
to them a n
unus ua lly da ngerous degree
of
unpredicta bility
intheir a tta cks
upon
the
Roma n
empire.
THE S AXONS AND BRITAIN
Two
pa s s a ges
of Ammia nus Ma rcellinus
a ga inprovide
the evidence for the involvement of the
S a x ons inthe
his tory
of Brita ininthe mid-fourth
century.
At x x vi.
4. 5
the his toria nlis ts a
number of ba rba ria ninva s ions :
a mong
them we find tha t 'Picti
S a x ones que
et
S cotti,
et Atta cotti
Britta nos a erumnis vex a vere continuis '. Va lentinia nha s
jus t procla imed
his brother Va lens
emperor.
The da te is
364,
a nd the deductionis ma de tha t a ninva s ionof Brita infrom the
s outh-ea s t,
the northa nd the wes t took
pla ce
inthis
yea r.
The s econd
pa s s a ge
is x x vii. 8.
5-
Ammia nus ha s
begun
this Britis h
cha pter by des cribing
the dis a s trous
culmina tion,
in
367,
of a
'ba rba rica
cons pira tio':
the Roma n
a rmy
inBrita inha d beenthrowninto
dis a rra y,
a nd
Va lentinia nwa s forced to
des pa tch
firs t
S everus ,
then
Jovinus ,
a nd
fina lly
Theodos ius to
remedy
the s itua tion. Ammia nus
continues ,
a little
la ter,
withthe fa mous
pa s s a ge:
'eo
tempore
Picti indua s
gentes
divis i,
Dica lydona s
et
Verturiones , itidemque Atta cotti,
bellicos a hominum
na tio,
et
S cotti, per
divers a
via ga ntes ,
multa
popula ba ntur.
Ga llica nos vero tra ctus Fra nci et
S a x ones ,
is dem
confines , quo quis que erumpere potuit,
terra
viel
ma ri, pra edis
a cerbis
incendiis que,
et
ca ptivorum funeribus omnium,
viola ba nt'. It is a s s umed tha t the ba rba ria ntribes enumera ted
inthis
pa s s a ge
were the
pa rties
to the
'cons pira tio';
tha t
they
ha d routed the Britis h
ga rris on;
a nd tha t
they
met defea t a t the ha nds of Theodos ius inthe
ca mpa igns
whicha re des cribed in
the res t of x x vii. 8 a nd inx x viii.
3.
A
s econd,
more
da ma ging,
inva s ionof ba rba ria ns
s hould,
therefore,
be
a s s igned
to the
yea r 367.
15 For Ghelen's modus
opera ndi, s ee
Irela nd, op. cit. (note 3), es pecia lly48.
16
S ee x x i. 5. I; x x ix . 6. 14; x x x i. 3. 6; a nd x x x i. 7. 6.
17
Ghelenma yha ve beeninfluenced a t this point bythe uns ens a tiona l des criptionof the S a x ons a t x x x . 7. 8
a s
's emper quolibet inex plora toruentes ',
or
by
the ma ritime connota tions of the
des cription
of the
Ma ra tocupreni
a s
brea king
out
'quocumque
ventus
dux era t',
whichis intended
purelymeta phorica lly.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
174
PHILIP BARTHOLOMEW
The firs t of thes e two
pa s s a ges
of Ammia nus ha s
recently
beenthe
s ubject
of a ca reful
ex a mina tion
by
Dr R. S .
O.
Tomlin.18 He s howed tha t a t this
point
Ammia nus is
merely
giving
a
preview
of the ba rba ria ninva s ions whichwere to occur inthe
joint reign
of Va lentinia n
a nd Va lens ;
he is not
reporting
the
milita ry
events of
364,
a nd it is mis ta kento a s cribe hos tile
incurs ions inthis
yea r
to
a ny
of the tribes lis ted
here, including
the
S a x ons ,
onthe ba s is of this
pa s s a ge.
But Tomlinbelieves tha t the 'rea l' S a x oninva s ionfores ha dowed inx x vi.
4. 5
is the
a tta ck whichis
a ppa rently
des cribed inx x vii. 8.
5.
This lea ds to a
difficulty:
for inthe former
pa s s a ge
the S a x ons a re s ta ted to ha ve a tta cked
Brita in,
while inthe la tter the
s phere
of their
opera tions
is s ta ted to be the 'Ga llica nos tra ctus '. Like mos t
commenta tors ,
Tomlinbelieves
tha t this inva s ionmus t ha ve
's pilled
over' into Brita in. But intha t
ca s e,
the Britis hinva s ion
s hould be
ex pected
to ha ve found a mentioninx x x .
7 where,
a fter Va lentinia n's
dea th,
Ammia nus
gives
a
s umma ry
of the
emperor's milita ry
a chievements . Yet we
only find,
a t
x x x .
7. 8,
a reference ba ck to the S a x oninva s ionof north-ea s t Ga ul whichha s beendis cus s ed
a bove. Tomlindecla res tha t this
incons is tency
'ca nnot be res olved'. But
progres s
ca nbe ma de
if we
try
to view x x vii. 8.
5 a ga ins t
the
ba ckground
of Ammia nus ' methods of
compos ition.
It is the Loeb editionof Ammia nus whichis
norma lly
cons ulted
by
s chola rs interes ted inthis
period
of Roma no-Britis h
his tory.
Rolfe
prints
x x vii. 8.
5
a s a
s epa ra te pa ra gra ph
from the
s ectionwhich
precedes
it.19 This s ta tes tha t Ammia nus ha s des cribed the 'motus a dules centis et
s enes centis ocea ni, s itumque
Britta nia e' a t a nea rlier
point
inhis na rra tive. He continues
by
s a ying
tha t no further a dditiona l informa tionof this kind is
neces s a ry:
'a d ea
qua e diges ta
s unt
s emel,
revolvi
s uperfluum
dux i'. But the nex t
s ection,
the s ectionwhichconta ins the
des cription
of the ba rba ria n
inva s ions , begins
withthe words 'illud ta men
s ufficiet
dici'. Ammia nus ha s
a ba ndoned the res olutionwhichhe ha s
jus t
ma de: he ha s decided to
s upplement
his ea rlier
des cription
of Brita in
by
a brief reference to the a ctivities of certa inba rba ria ntribes . This
reference
belongs clos ely
withthe s ectionwhich
precedes
it.
Every
modern
editor, a pa rt
from
Rolfe,
ha s rea lized this :
they
a ll
print
the two s ections
together
a s a
s ingle pa ra gra ph.20
If the two s ections a re ta ken
together,
it becomes clea r
immedia tely
tha t
they
form a n
a bbrevia ted
digres s ion.
It wa s Ammia nus '
pra ctice,
whenever the s cene of his na rra tive
cha nged,
to
provide
his rea ders witha s much
ba ckground
informa tiona s he could. At
times ,
it is ha rd
to res is t the
feeling
tha t he wa s concerned to s how tha t he ha d done the
ba ckground rea ding
whichwa s
ex pected
of a ma nof
letters ;
his s ources ca n
us ua lly
be
tra ced,
a nd the
digres s ions
conta in
very
little tha t is
origina l.21 But,
however little he
knew,
he
a lwa ys
found it es s entia l
to
s a y s omething.
Releva nce did not ma tter: the
des cription
of Thra ce
(x x vii. 4. 1-14)
is ins erted
a s a nintroductionto the
ca mpa igns
of Va lens
a ga ins t
the Goths in
367-369, when,
in
fa ct,
thes e
ca mpa igns
were
fought entirely
to the northof the Da nube. Nor did the fa ct tha t he ha d
a lrea dy given
a n
a dequa te
trea tment of the
s ubject
concerned ina nea rlier
pa s s a ge
res tra inhim
from
returning
to it: the
ethnogra phica l
rema rks a bout the S a ra cens a t x iv.
4. 3-6
a re
preceded
by
the
words , 's uper quorum
moribus licet ina ctibus
principis Ma rci,
et
pos tea a liquotiens
memini
rettulis s e,
ta mennunc
quoque pa uca
de is dem
ex pedia m ca rptim' (x iv. 4. 2).
This is a
pa ra llel
to
Ammia nus ' references to Brita ina t x x vii. 8.
4-5.
He ha s
given
his
des cription
of Brita ininone
18 Cla s s ica l
Qua rterly,
n.s . x x ix
(1979), 470-8.
19 S ee vol.
3,
p.
52
of his edition
(op. cit.,
note
7).
20
It is
regretta ble tha t,
like
Rolfe,
A. R.
Birley
ha s fa iled tos ee the
s ignifica nce
of x x vii. 8.
4.
Inhis
entry
for Theodos ius inThe Fa s ti
of
Roma nBrita in
(Ox ford, 1981), 333-4,
he
quotes ex tens ively
from this
cha pter
of
Ammia nus ;
but s ection
4
is
repres ented by
three dots . The omis s ionof this s ectionma kes a
proper unders ta nding
of this
pa s s a ge impos s ible.
21
Onthis , s ee T. Momms en,
Hermes x vi
(i88i), 602-36.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
FOURTH-CENTURY S AXONS
175
of the los t books
;22 but,
forced to na rra te s ome more Britis h
events ,
he ha s found the
tempta tion
to
provide
s ome more
ba ckground
informa tionirres is tible. But inthis ca s e he ha d no more
ba ckground
informa tionto
provide.
The
ex ploits
of the Picts a nd their
collea gues
a re in
a ny
ca s e of dubious releva nce to the
ca mpa ign
of Theodos ius which
follows ;
a nd the Fra nks a nd
S a x ons a re a dmitted to ha ve
fought
els ewhere. Like the 'Hoc
tempore'
whichs ta nds a t the
beginning
of x x vi.
4. 5,
the 'eo
tempore'
of x x vii. 8.
5
is of
va gue a pplica tion.23
The ba rba ria n
inva s ions of this
pa s s a ge
a re not tied downto the
yea r 367;
a nd inthe ca s e of the Fra nks a nd
S a x ons a t lea s t
they
ha ve no connectionwithBrita in.
Ina dditionto x x vi.
4. 5,
the tribes of the northa nd the wes t ma ke one other
a ppea ra nce
in
the na rra tive of Ammia nus . In
360,
Julia nwa s forced to s end
Lupicinus
to
Brita in;
this wa s
ma de
neces s a ry,
in
pa rt
a t
lea s t, by
'S cottorum
Pictorumque
gentiumfera rumn
ex curs us '
(x x .
i.
I).
This s eems to ha ve beena
genuine inva s ion,
but
Ammia nus , perha ps
influenced
by
his
hos tility
to
Lupicinus ,
who wa s no friend to
Julia n,
tells us a lmos t
nothing
a bout the
ens uing ca mpa ign.
Inthe a bs ence of
a ny
further inva s ions from thes e
qua rters ,
it mus t be a s s umed tha t this
incurs ion
la y
a t the ba ck of Ammia nus ' mind whenhe wrote x x vii. 8.
5.
But
a ga in
there is a
dis crepa ncy:
the tribes involved in
360
were the S cots a nd the
Picts ,
but the la ter
pa s s a ge s pea ks
of the a ctivities of the
Picts ,
the Atta cotti a nd the S cots .
However,
a t x x . i. i. the
rea ding
'S cottorumin'
is
unlikely
to be
correct,
for the firs t three letters of this word ha ve beenins erted
inV ina nother
ha nd;24 pres uma bly,
V wa s
corrupt
a t this
point,
a nd a nother
s cribe,
whos e
a ltera tions do not
ca rry
the s a me
a uthority, a ttempted
to correct it. A more
proba ble
emenda tionis 'Atta cottorum'. If this is
a ccepted,
the S cots will be the odd tribe out inx x vii. 8.
5.
They
s ta nd
a lone,
a wkwa rd a nd
unqua lified,
a t the end of the
des cription
of the other two
tribes . We ha ve
a lrea dy
s eentha t Ammia nus is fond of
us ing a djectives
in
pa irs .
'Bellicos us ' is
a lmos t
a lwa ys
us ed in
conjunction
witha nother
a djective;
beca us e of its
rhythm,
it is
inva ria bly
the firs t of the two. Onthree
occa s ions ,
it is the firs t of two
a djectives a pplied
to a ba rba ria n
tribe.25 It needs to be a s ked whether the
s tylis tica lly
a wkwa rd 'S cotti' could be
repla ced by
a nother
a djective.
The
pa la eogra phica lly
s imila r 's a eva '
s ugges ts
its elf
immedia tely;
in
Ammia nus it a lmos t
a lwa ys a ppea rs
a s the s econd of two
a djectives .26
If this res tora tionis
ma de,
the tex t would rea d
'itideminque Atta cotti,
bellicos a
hominumi
na tio et
s a eva , per
divers a
va ga ntes ,
multa
popula ba ntur'. Rhythmica lly,
this is
prefera ble
to the
a ccepted tex t;
inthe
a bs ence of further references to the S cots in
Ammnia nus
(with
the
ex ception
of x x vi.
4. 5,
which
will be dis cus s ed
below),
it does no violence to the s ens e. If we ca n
a ccept it,
the
corres pondence
betweenx x . I. I a nd x x vii. 8.
5
will be
complete,
a nd there ca nbe no motive to
ques tion
the
deriva tionof the la ter
pa s s a ge
from the ea rlier.
It is a
rela tively s tra ightforwa rd
ma tter to
s ugges t
a
point
of reference for the ra ids of the
Fra nks a nd S a x ons onthe 'Ga llica nos
tra ctuts '.
Ammia nus will be
a nticipa ting
the S a x onra id
onnorth-ea s t Ga ul des cribed inx x viii.
5. 1-7.
It is
pos s ible
tha t this is
only
the mos t
s triking
of
22
It is wrong toa s s ume tha t this digres s ionwould ha ve proved tobe a mine of informa tiona bout the s ta te of
Brita ininthe fourth
century.
His
des cription
of the 's itus Britta nia e' would
proba bly
ha ve beenderived from
Ptolemy;
the beha viour of the ocea nmentioned ins ection
4,
a s well a s the
peculia rities
of the S tra its of Dover
a lluded toins ection
6,
would ha ve come from
Tima genes
of
Alex a ndria ,
whos e
Periplus ,
writtentowa rds the
end of the firs t
centuryB.C.,
would
proba bly
ha ve ex tended a s fa r a s the mouthof the Rhine. S ee
Momms en,
op.
cit.
(note 21), 612-27.
23 The former
ex pres s ion
is dis cus s ed
byTomlin, op.
cit.
(note 18).
But he thinks
(p. 470,
n.
4)
tha t the 'eo
tempore'
of x x vii. 8.
5
refers toa
precis e
moment intime.
24
Cla rk's tex tua l a ppa ra tus tothis pa s s a ge ma kes this clea r.
5 x x viii. 5. 9, the Burgundia ns ; x x ix . 5. 26, the Ma zices ; x x x i. 16. 3, the Huns a nd the Ala ns , together.
" x iv. 10.
4;
x vi. 12.
23;
x vii. 8.
I;
x x ii.
13. I;
x x iii. 2.
3; x x vi.
5. 7. The ma inex ceptionoccurs a t x x viii. I.
53.
The word is us ed a lone a t x x ix . 1.
38
a nd x x iii.
I.
6;
but inthe la tter ins ta nce the correct
rea ding
s hould
proba bly
be
's ca evumni'.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
176
PHILIP BARTHOLOMEW
a number of incidents whichha d been
continuing,
inthis
a rea ,
for s ome
yea rs ;
for Zos imus
records tha t onJulia n's a rriva l in
Ga ul,
he found tha t the ba rba ria ns ha d
pra ctica lly
rea ched
the cities nea r the
coa s t.27
The reference to thes e coa s ta l cities is reminis cent of Ammia nus
x x viii.
5.
I,
where the S a x ons a tta cked the Roma ns
'Ocea nidifficulta tibus permea tis ',
a nd of the
'Ga llica nos tra ctus ' of x x vii. 8.
5.
There is no
difficulty
a bout the Fra nks :
they
were
neighbours
to the
S a x ons ,28
a nd a re mentioned a s
a cting
in
conjunction
withthem els ewhere.29 It is
proba ble
tha t
they pa rticipa ted
inea chother's
ca mpa igns : pres uma bly,
the Roma ns la cked the
a bility,
a nd the
interes t,
to
dis tinguis h
betweenthe two. Like the reference to the northern
tribes ,
the a llus ionto the a ctivities of the Fra nks a nd S a x ons ha s no connectionwithevents in
Brita inin
367.
A
problem
rema ins . The a ctivities of the ba rba ria ntribes whicha re mentioned inx x vi.
4. 5
a re des cribed a t
grea ter length,
under the
yea rs
when
they
took
pla ce,
a t la ter
points
inthe
na rra tive of Ammia nus . But the s entence 'Picti
S a x ones que
et
S cotti,
et Atta cotti Britta nos
a erumnis
vex a viere
continuis ' ca nnot be a reference forwa rd to x x vii. 8.
5,
s ince this too is a
'ba ckground'
reference to events whichoccurred a t other times a nd other
pla ces .
A
des cription
is
needed,
la ter inAmmia nus ' a ccount of the
reign
of
Va lentinia n,
of a ba rba ria ninva s ionof
Brita infrom overs ea s . No s uchevent is
mentioned;
a nd there is no rea s onfor
s uppos ing
tha t
a ny
s uchinva s iontook
pla ce.
It is
neces s a ry
to a s k wha t
ma nus cript a uthority
there is for the
a tta cks of x x vi.
4. 5.
In
fa ct,
there is none: for the
rea ding
of V is
gibberis h.
The s cribe wrote
'recti
S a x ones que
et s ecuti et a ta citti urita nos a erumnis
viex a viere
continis '.30 Almos t
predicta bly,
the
printed
tex t is Ghelen's .
But,
a s the
ma ny points
of
corres pondence
ma ke
clea r,
he is here
dependent upon
the editionof Ca s tellus . This is a
printed
vers ion
(a t s econd-ha nd)
of the wors t
ma nus cript,
vitia ted s till further
by
the editor's own
conjectures ;
a nd Ghelens eems to ha ve
ta kenthe vers ionof
Ca s tellus ,
a nd a dded s ome
'improvements '
of his own.31 His tex t mus t be
rega rded,
not a s
tes timony
to the events of the fourth
century,
but a s evidence of the
fertility
of his
ima gina tion.
The S a x ons a lone a re recorded in
V;32
Ammia nus is not
likely
to ha ve
included them ina lis t of ba rba ria ntribes who lived ina
completely
different
a rea ;
a nd it
needs to be a s ked whether the tex t
ma y
not be res tored ins ucha
wa y
a s to
pres erve
the S a x ons
but to elimina te the intruders from the north. 'Fra nci
S a x ones que
etia mtum tra ctus Ga llica nos
a erumnis
vex a vere
continuis ' is a
pos s ibility.
Intha t
ca s e,
like x x vii. 8.
5,
the
pres ent pa s s a ge
will
s imply
be a reference forwa rd to the inva s ionof north-ea s t Ga ul of x x viii.
5.
I-7.
The
pres ent
conjecture
s hould not be
pres s ed
too
fa r;
but it needs to be
empha s ized
tha t there a re no certa in
grounds
for
believing
tha t Ammia nus mentioned either Brita inor the tribes of the northin
this
pa s s a ge.
If he did
not,
there is
nothing
to
prevent
us
a s s uming
tha t he intended a n
uns ens a tiona l reference to Continenta l events whichhe did des cribe els ewhere. It follows tha t
the
incons is tency
whichTomlinfound betweenAmmia nus '
s umma ry
of Va lentinia n's
27
Zos imus iii. 3. I1: Tol5 8s Pa pp pous
...
pXpt OX)E86v
TCv
-rrp6s
Ga ?
TTr,
vr6?ECOv 81EM6VTa 5.
28
Zos imus iii. 6. 2: cOpa yKv 6p6pcova cbTois OVTCwrov. Cf. a ls oJulia n, Ora tioni, 34D,
where the Fra nks a nd S a x ons
a re s ta ted to
co-opera te
KTa -r
T
r6UyyEViS .
29
Zos imus iii. I.
I:
KcovTrVTto . ..
e
oECbPVOS ... .()pXyKOU5 pV Kcx i' A a i
a pa vvoOS
K K a iC6OVCX5 igs t8 ToEa ca p&KOVTa TrOEl5
...
Ka TrEi\116Tca S .
30
Tomlin, op. cit. (note 18), dra ws a ttentiontothe
uns a tis fa ctory
s ta te of the
ma nus cript
tra dition
(p. 474,
n.
28).
But he does not
develop
this line of
thought.
31
The deta ils ma ybe found inCla rk. The poor qua lity
of Ca s tellus ' editionis
empha s ized
inhis 'Tex t tra di-
tionof Ammia nus Ma rcellinus '
(cf.
note
13), pp. io a nd
67.
Inhis
prefa ce
tohis editionhe dubs it 'editio
mendos is s ima '.
32
In
fa ct,
'S cotti' wa s ins erted a bove 's ecuti' inV ina nother ha nd. But the s ta tus of this correctionca nnot be
a uthentica ted.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
FOURTH-CENTURY S AXONS
I77
a chievements a nd x x vi.
4. 5
does not ex is t. The
corres pondence
betweenthe ba rba ria na ctivities
mentioned inthe la tter
pa s s a ge
a nd the
genuine
inva s ions des cribed la ter will be
complete.33
COUNT THEODOS IUS AND BRITAIN
It ha s been
a rgued
a bove tha t ba rba ria ninva s ions
pla yed
no
pa rt
inthe Britis hevents which
occurred inthe
yea rs 367
a nd
368.
This is confirmed
by
Ammia nus x x x .
7. 9,
where the
ba rba ria ntribes of x x vii. 8.
5
a re
cons picuous by
their a bs ence from the his toria n's
s umma ry
of Va lentinia n's Britis hs ucces s es .34 But s ucces s es there were: the elus ive
brevity
of x x x .
7. 9
is a reference ba ck to the two
cha pters
x x vii. 8 a nd x x viii.
3,
where the a chievements of Count
Thecdos ius inBrita ina re des cribed a t
length,
but with
ta nta lizing obs curity.
It is
neces s a ry
to
look
a ga in
a t thes e two
cha pters
inorder to formula te a na ns wer to the
ques tion:
Wha t did
Theodos ius do?
Now
throughout
thes e two
cha pters ,
Ammia nus does not los e a
s ingle opportunity
to
a pply
a
la uda tory epithet
to Theodos ius . It is
tempting
to deduce from this tha t the a chievements of
Theodos ius were
pa rticula rly outs ta nding.
But s ucha deductionta kes no a ccount of Ammia nus '
litera ry procedures .
Whenever he mentioned a
genera l,
he felt it
neces s a ry
to
ex pa nd upon
the reference
by
the a dditionof
des criptive epithets
of one kind or a nother. Ins o
doing,
he
drew
upon
a commons tock of
prejudices
whichwere held towa rds the
lea ding figures
of a
s lightly
ea rlier
era ;35
he could
only
reflect the rela tive
popula rity
of his cha ra cters a t the time
he wrote. As
informa tive,
a nd
objective,
a s s es s ments of the a chievements a nd a bilities of the
genera ls
who
a ppea r
inhis
na rra tive,
his
des criptions
a re va lueles s . Mos t of his
genera ls
received
a bus e, or,
a t the
mos t, s everely qua lified pra is e;
Theodos ius
ha ppened
to be
lucky.
Tha t is
rea lly
a ll tha t ca nbe s a id. Before a
proper
a s s es s ment ca nbe ma de of his a chievements
in
Brita in,
a ll the
la uda tory ex pres s ions
whichAmmia nus a tta ches to his na me mus t be
dis counted.
They
a re not
grounds
for
s uppos ing
tha t wha t he did wa s
pa rticula rly s ignifica nt.36
An
inquiry
is bes t s ta rted
by
a nex a mina tionof the s econd of the two
pa s s a ges
which
Ammia nus devotes to Theodos ius '
ca mpa igns
inBrita in. Muchof this
cha pter (s ections 3-6)
is
occupied by
a na ccount of the des tructionof the
puta tive
rebel Va lentinus . This a lone of a ll
thes e events finds its
wa y
into Zos imus
(iv.
12.
2);
a nd it ha s been
pla us ibly s ugges ted by
A. R.
Birley tha t, by compa ris on,
the
milita ry
a ctions underta ken
by
Theodos ius were
"
This
a rgument,
whenta kenin
conjunction
withthe
preceding one, completes
the elimina tionof the S cots
from Ammia nus ' na rra tive. Pa n. La t. x ii.
5.
2 will be dis cus s ed
below;
but it s hould be
pointed
out here tha t
the references inCla udia n
(De
iii cons .
Honorii, 55;
De iv cons .
Honorii, 33)
which
purport
tos how a 'S cottis h'
involvement in
fourth-century
Britis h
his tory
a re too
deeply
embedded in
pa s s a ges
of
poetica l ex a ggera tion
to
be a llowed tocount a s his torica l evidence. It follows tha t there is no
litera ryjus tifica tion
for a
s ea rch, by
a rcha eologis ts ,
for evidence of Iris hs ettlement inWa les a nd the s outh-wes t a s
ea rly
a s the mid-fourth
century.
34
A ba rba ria ninva s ion
is , however, mentioned inthe
cha pter
s umma ries tox x vii. 8 a nd x x viii.
3.
But thes e
were not written
byAmmia nus ; they
firs t
a ppea r
inthe
s eventeenth-century printed editions ,
a nd s how
only
tha t,
a s a res ult of a
ha s tyrea ding
of the
tex t, mis unders ta ndings
ha d a ris en. Rolfe's
pra ctice,
inthe Loeb
edition,
of
printing
thes e s umma ries a t the
beginning
of ea ch
cha pter (ins tea d
of a t the
beginning
of ea ch
book,
a s ea rlier editors ha d
done)
crea tes a fa ls e
impres s ion. They
a re not evidence for
fourth-century
Brita in.
35
This view is ela bora ted
a bove, p. 171.
36
Evendes criptions of wha t
a ppea r tobe
genuine a ctions ca nnot be relied
upon.
At x x viii.
3. I-2,
Theodos ius
is des cribed a s 'nihil
grega riis impera ns ,
cuius
non
ips e primitia s ,
a la cri
ca pes s eret
mente.
Hocque genere
cum
s trenuii militis munia et
pra ecla ri
ducis cura s
ex pleret
. . .'. This is
s us picious ly
reminis cent of the
des cription
of
Julia na t x vii. I. 2 a s
'omnnis opera e conturma lem,
a uctorita te
ma gnificum ducem, plus
la boris indicere s ibi
qua m
militi,
s icut
pers picue contigit,
a ds uetum'; a nd,
a s Ammia nus a dmits
(x vi.
I.
3),
he found it
difficult,
in
s pea king
of
Julia n,
toa void
s tra ying
into
'la uda tiva
ma teria '.
S imila rly,
the reference a t x x viii.
3. 9
tothe es cort whichhe
received 'a d
us que fretum'
mus t be
rega rded
a s
pure
fiction.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
178
PHILIP BARTHOLOMEW
ins ignifica nt.37
Va lentinus ,
a
Pa nnonia n,
wa s brother-in-la w of the
da ngerous ,
but
influentia l,
Ma x iminus ;
a nd it is
pos s ible
tha t his demis e ha d wide
politica l repercus s ions . Thes e, however,
fa ll outs ide the
s cope
of the
pres ent pa per.
By compa ris on,
the es ta blis hment of Va lentia ca nha ve beenof little
genuine importa nce.
Ammia nus records tha t Theodos ius
'recupera ta m provincia m, qua e
indicionem conces s era t
hos tium,
ita reddidera t s ta tui
pris tino,
ut eodem
referente
et rectorem ha beret
legitimum,
et
Va lentia deinde
voca retur...'
(x x viii. 3. 7);
a nd this ca n
only
mea ntha t Theodos ius wres ted a
province
from
enemy control,
res tored the
pa ra pherna lia
of Roma n
a dminis tra tion,
a nd na med
the
province
'Va lentia '. Inother
words ,
he
ga ve
a new na me to a nold
province.
Almos t a ll
s chola rs , however,
ha ve
dis rega rded
the
pla in
words of
Ammia nus ,
a nd
pla ced grea ter
credence
onthe evidence of the Notitia
Dignita tum.
This
a ppa rently
lis ts Va lentia a s a fifthBritis h
province,
to
complement
the four which
go
ba ck to the a dminis tra tive
cha nges
of the turnof
the
century. Theodos ius ,
it would
s eem,
ma de a s ubs ta ntia l a dminis tra tive innova tion. But the
a uthority
of the Notitia
Dignita tum
a s evidence needs to be
ca refully
ex a mined.
Va lentia is entered inthe wes terns ectionof the Notitia
Dignita tum
four times . It
a ppea rs
bothinthe tex t of the
cha pter
of the Vica r of the
Brita ins ,
a nd a s a
ca ption
to one of the
pictures
of
s trongholds
whichform the illus tra tionto this
cha pter (x x iii.
I
I;
x x iii.
4).
It is lis ted
a ga in
a s one of the
provinces
whichfell under the
juris diction
of the
pra etoria nprefect
of the Ga uls
(iii. 34),
a nd is
duly
recorded inthe 'Index ' to the whole wes terns ection
(i. 77). However,
its
occurrence a s a
ca ption
to a nillus tra tionca nbe
dis rega rded.
Inthe ca s e of the De rebus bellicis
it ha s beens howntha t the
ca ptions
to the
pictures
were
proba bly
a dded a t a
fa irly
la te
s ta ge
inthe tra ns mis s ionof the
tex t;38
their s ource wa s the tex t its elf. Giventhe clos e connection
betweenthe tex tua l
his tory
of the De rebus
bellicis
a nd tha t of the Notitia
Dignita tum,
it is
rea s ona ble to
s uppos e
tha t the s a me
procedure
wa s followed here too.39 This mea ns tha t the
Va lentia of x x iii.
4
will ha ve beena dded to our vers ionof the Notitia
Dignita tum through
the
interventionof a n
ea rly
media eva l s cribe. The Va lentia of i.
77
is
s us pect
for a s imila r rea s on.
Ma nnha s
s ugges ted tha t,
a t s ome
time, perha ps la te,
inthe fifth
century,
the Notitia wa s re-ca s t
to
perform
a new
function,
a na ltera tionwhichis reflected inthe trea tment of
cha pters
v-vii a nd
x lii inthe wes tern
s ection;40
the
pres ent
writer believes tha t a nother
cha nge
of us e
occurred,
a t
s ome time betweenthe
origina l compila tion
of the Notitia a nd the
a da pta tionpos tula ted by
Ma nn, when, a mong
other
cha nges ,
the Index wa s ins erted.41 The
compiler
of the Index will
ha ve been
opera ting long enough
a fter the end of the fourth
century
to ha ve been
ca pa ble
of
mis unders ta nding
the
origina l
Notitia
Dignita tum
from whichhe wa s
working.
The Va lentia
of the
Index , therefore,
ca nnot be
rega rded
with
certa inty
a s
forming pa rt
of the ba s ic la te
fourth-century
document. But no s uchdoubts ca nbe enterta ined
concerning
the
pedigree
of
the two 'Va lentia s ' a t iii.
34
a nd x x iii. 11.42 Yet to ma ke thes e two occurrences of the na me
conform to the informa tion
provided by Ammia nus ,
it is
only neces s a ry
to a s s ume tha t 'nunc'
intervened betweenthe Ma x ima Ca es a riens is a nd the Va lentia of iii.
33-34
a nd the Ma x ima
Ca es a riens is a nd the Va lentia of x x iii. io-I
I;
a nd tha t a t s ome time before the
compila tion
of
the Index it wa s los t. Intha t ca s e the number of Britis h
provinces
lis ted inthe Notitia
Dignita tum
37 op.
cit.
(note 20), 338,
n. 12. A
ga rbled
reference
by
Zos imus a t iv.
35. 3
s hows tha t his
s ource, Euna pius ,
conta ined a fuller a ccount of Theodos ius ' Britis h
ex pedition.
Zos imus chos e
only
the Va lentinus incident for
inclus ioninhis ownwork.
38
Irela nd, op.
cit.
(note 3), 103-4, io6,
o107-8.
9
This
s uppos ition
is
s upported by
the numerous occa s ions onwhichthe
ea rly
media eva l
copyis t
fa iled to
unders ta nd the tex t he ha d infront of
him,
a nd
produced
a
ga rbled
vers ioninthe
ca ptions .
40
S ee his
contribution,
'Wha t wa s the Notitia
Dignita tumn
for
?',
inR. Goodburna nd P. Ba rtholomew (eds .),
As pects of
the Notitia
Dignita tum (Ox ford, 1976), 1-9.
41
This view
obvious ly
needs tobe
a rgued
indeta il els ewhere.
42
It mus t be pointed
out tha t norelia nce a t a ll ca nbe
pla ced
onthe numera l 'V' a t iii.
32.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
FOURTH-CENTURY S AXONS
179
will be reduced to
four,
a nd the a ccount whichAmmia nus
gives
of the es ta blis hment of
Va lentia ca nbe ta kena t its fa ce va lue.
Moreover,
the
province
to whichTheodos ius
ga ve
a
new na me ca nbe s eento be Ma x ima Ca es a riens is .
Now inx x viii.
3
Ammia nus refers to a number of new
milita ry dis pos itions
whichTheodos ius
ma de. Thes e a re
us ua lly
a s s ocia ted withthe a rea of Ha dria n's Wa ll. In
pa rt,
this
a s s umption
is derived from the a ccount whichAmmia nus
gives
of the duties of the 'Arca ni'.43
They
were
s uppos ed
to obta in
milita ry intelligence 'per longa s pa tia
dis currentes '
(x x viii. 3. 8);
but
they
were bribed into
divulging
informa tiona bout the Roma n
a rmy,
a nd Theodos ius removed
them 'a s ta tionibus s uis '. The
'longa s pa tia '
a re
us ua lly thought
to ha ve la into the northof
Ha dria n's
Wa ll;
but if the Picts a nd the Atta cotti a re removed from the
his tory
of the
yea rs 367
a nd
368,
the ma inmotive for the loca tionof the
'longa s pa tia '
inthe fa r north
dis a ppea rs .
They might belong a nywhere
inthe Britis hdioces e.
Moreover,
if the 'Arca ni' s hould
rea lly
be
a s s igned
to the fa r
north,
the res ults of their
trea chery ought
to be detecta ble inthe
a rcha eology
of the Wa ll. But
they
a re not: it would s eem tha t not a
s ingle
des truction
la yer s ecurely
da ta ble
to
367
ha s beenidentified. S ome
repa ir
work onthe Wa ll forts a nd onthe curta inca nbe
a s s igned
to the fourth
century;
but more
precis e da ting
is
impos s ible.44
Inthe
circums ta nces ,
it
ma y
be
s ugges ted
tha t the
a s s umption
of a 'Theodos ia n'
pha s e
inthe
his tory
of the Wa ll is
more of a nemba rra s s ment tha na na id to
unders ta nding.
No
s upport
for it ca nbe derived from
Theodos ius ' efforts to
protect
the
'limites '
(x x viii. 3. 7);
the ex is tence of
only
one
genuine
'limes ' inBrita inma kes the
plura l s us picious ,
a nd it s hould be reca lled
tha t,
for a
s cribe,
a
confus ionbetween'limes ' a nd 'miles ' is
fa ta lly
ea s y.45
Inthe
continuing
a bs ence of a rcha eo-
logica l
confirma tionfrom the
Wa ll,
it is wis e to
rega rd
the ca s e for
loca ting a ny
of Theodos ius '
a ctivities inthe northa s not
proven.
But the
province
of Ma x ima Ca es a riens is ca nbe
given
a
proba ble
loca tion: it ha s been
convincingly a rgued by
Ma nntha t it wa s the
principa l province
of the Britis h
dioces e,
s itua ted
inthe s outh-ea s t withLondona s its
ca pita l.46 Now, a pa rt
from Theodos ius '
la nding-pla ce
a t
Richborough,
the
only
Britis htownwhichreceives a mentioninAmmia nus ' a ccount of his
ca mpa igns
is London
(x x vii.
8.
7-8
a nd x x viii.
3. 1); indeed,
the libera tionof Londons eems
to ha ve beenhis firs t
objective,
a nd his
s ubs equent milita ry
a ctions were centred
upon
it. In
the a bs ence of
a ny
further clea r reference
by
Ammia nus to the
loca lity
of the
fighting,
it mus t
be a s s umed to be confined to the
s outh-ea s t,
tha t is to
s a y
to the a rea of Ma x ima
Ca es a riens is ,
the
province
whichTheodos ius did res tore. Now this
province
conta ined the
la rges t
number of
s ignifica nt
urba ns ettlements inBrita in. And this fa ct
permits
us to
ex pa nd
a little
upon
the
'gentes ' who
a ppea r
to ha ve beenTheodos ius ' ma in
opponents
inthe la ter
s ta ges
of his
ca mpa ign. They
were not ba rba ria ninva ders from overs ea s .
They
were not
dis pos s es s ed
rura l
pea s a nts (or Ba ca uda e)
in
revolt;
for otherwis e tra ces of their a ctivities could be
ex pected
to
ha ve beendis covered from the
la rge
number of
fourth-century
villa s whichha ve been
3
At this
point,
V is
hopeles s lycorrupt.
'Arca nos ' is a nemenda tionof
'a rea nos ';
but s ince noother a ncient
writer mentions the
'Arca ni',
their res tora tionhere is
quite
ins ecure. It is unwis e to
a ttempt
toma ke deductions
from the
a ppa rent
form of their na me totheir function. It is a ls o
uns a tis fa ctory
toba s e
s pecula tions a bout the
duties of thes e
troops upon
the 'a rea nos ' of the
ma nus cripts ,
s ince this form occurs nowhere els e.
"
D. J. Breeze a nd B.
Dobs on, Ha dria n's Wa ll
(London, 1976),
221-6.
*
The tex t of Ammia nus x x viii.
5. I provides
a nillus tra tionof the dubious
a uthenticity
of 'limes '.
According
toCla rk a nd the other modern
editors ,
the
object
of the S a x ona tta ck onnorth-ea s t Ga ul in
370
wa s 'Roma num
limitem'. But the
rea ding
of M
(whichs urvives a t this
point)
is 'Roma num
militem';
a nd M is followed
by
V.
'Limniten' a ppea rs only
a s a correctionina ninferior
fifteenth-centuryma nus cript,
a nd inGhelen. This indica tes
the rea dines s withwhichRena is s a nce s chola rs
thought
of 'limites ' inthe contex t of la te Roma n
milita ry
opera tions ;
it a ls os hows how
uns crupulous
Ghelencould be in
s etting
a s ide the
tes timony
of M
(when
he felt
s o
inclined)
infa vour of a
conjecture
of his own. The decis ionof the moderneditors to
a ccept
'limitem' ins tea d
of the better a ttes ted 'militem' mus t
a ppea r dis tinctlyques tiona ble.
46
Antiquity
x x x v
(1961), 318-9.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
I8o
PHILIP BARTHOLOMEW
ex ca va ted,
a nd thes e tra ces a re
la cking.
Their
origin
mus t be looked for inthe towns . It s eems
rea s ona ble to
s ugges t
tha t Theodos ius wa s fa ced withthe revolt of a dis contented urba n
proleta ria t.47
But rebels of this kind did not form the
only oppos ition. Quite ea rly
inhis a ccount of
Theodos ius ' Britis h
ca mpa igns ,
Ammia nus refers to des erters a nd to a
genera l
s ta te of
milita ry
indis cipline (x x vii.
8.
io);
a nd the mentionof towns inthe la ter fourth
century
s hould in
a ny
ca s e remind us
immedia tely
of the
a rmy.
The
qua rtering
of
troops
intowns wa s a
wides prea d
a nd fa milia r
pra ctice,
a ttes ted a t a number of
points
inthe Theodos ia nCode
;48
Brita inca nnot
be
ex pected
to ha ve been
ex empt
from it. S ome
troops ,
of
cours e,
continued to live in
milita ry
ca mps ;
a nd in
des cribing
the work of the
'mens ores ',
who ha d the
job
of
a s s igning qua rters
to
troops , Vegetius
refers
equa lly
to both
ca mps
a nd towns : 'Mens ores
qui
inca s tris
a dpodis mum
demetiuntur
loca ,
in
quibus
tentoria
milites figa nt,
vel
hos pitia
incivita tibus
pra es ta nt' (ii. 7).
The
Theodos ia n'res tora tion' is s ta ted
by
Ammia nus to ha ve beencentred
upon
'civita tes et ca s tra '
(x x viii. 3. 2)
or 'urbes et
pra es idia ria
ca s tra '
(x x viii. 3. 7),
the la tter
being only
a
re-ca s ting
of
the former
phra s e.
Thes e
echo, precis ely,
the
s phere
of
opera tions
which
Vegetius a s s igns
to
the 'mens ores '. If we remember tha t the other
protective
mea s ure des cribed inx x viii.
3. 7 ma y
well ha ve dea lt with
'milites ',
not
'limites ',
the conclus ion
emerges
tha t the firs t efforts of
Theodos ius inBrita inwere directed
a ga ins t
units of the Roma n
a rmy.
At this
point,
it is
perha ps pos s ible
to ex tra ct s ome confirma tionfor the
pres ence
of Roma n
troops
inthe towns of the s outh-ea s t from the Notitia
Dignita tum.
It ha s beens ta ted ea rlier
tha t the
ca ptions
to the
pictures
inthis document
a re,
ina ll
proba bility,
a dditions ma de
by
a n
ea rly
media eva l s cribe onthe ba s is of a ra ther ca reles s
rea ding
of the tex t. Now inthe
cha pter
of the Count of the S a x onS hore we find a n
entry
for the
'Pra efectus legionis
s ecunda e
Augus ta e'
(Oc.
x x viii.
19);
but none of our four
ma nus cripts
of the Notitia records a loca tionfor the
legion.
S eeck
prints 'Rutupis '.
He does s o onthe
s trength
of the occurrence of
'Rutupis '
a s a
ca ption
to one of the ca s tella which
a ppea r
inthe illus tra tiona t the hea d of the
cha pter
(Oc.
x x viii.
9).49
But one
pos s ibility
is tha t the
ea rly
media eva l
copyis t, ha ving pa inted
nine
ca s tella to
corres pond
to the nine units lis ted inthe
cha pter, found,
whenhe ca me to enter the
loca tions of the units inthe
illus tra tion,
tha t the
Legio
II
Augus ta
did not ha ve a loca tion
a s s igned
to it inthe tex t. He therefore invented one. Intha t
ca s e,
there will be no a ncient
litera ry
evidence whichloca tes the
Legio
II
Augus ta
a t
Richborough. Richborough, a nywa y,
is not
big enough
to conta ina
legion;
but there ca nbe no doubt
tha t, by
the
ea rly
fourth
century,
the
Legio
HII
Augus ta
ha d left
Ca erleon.5o
It mus t ha ve beens ta tioned s omewhere in
the
s outh-ea s t;
but the forts of the S a x onS hore
(the
'ca s tra ' of Ammia nus a nd
Vegetius )
a lrea dy
ha d different units
a s s igned
to them. It s eems
proba ble
tha t the
legion
wa s
s plit up
a nd
its men
qua rtered
inthe va rious towns of the a rea .
4
The ex is tence of s ucha
proleta ria t
is one outcome of the
development
whichReece ha s s ketched for
Roma no-Britis h towns inthe fourth
century.
He
s ugges ted (World Archa eology
x ii
(1980), 77-92, es pecia lly
86-7) tha t,
inthe la ter Roma n
period,
the s ucces s ful a nd cultured members of the
upper
cla s s es moved out of
the towns intothe
countrys ide; convers ely,
the les s well-off moved
a wa y
from the
countrys ide
intothe towns .
The la tter would ha ve beenthe
impoveris hed
res idue of the Britis hIron
Age tribes ,
or Ammia nus '
'gentes '.
This move ca nbe detected
a rcha eologica lly;
for inthe la ter fourth
century
the
qua lity
of
occupa tion
inthe
towns
cha nges dra ma tica lly
for the wors e.
They
were not
plea s a nt pla ces
inwhichto
live,
a nd the occurrence
of urba ndis orders would s eem tobe
only
too
proba ble.
48 W.
Goffa rt,
Ba rba ria ns a nd
Roma ns ,
A.D.
418-584 (Princeton, 1980), 42-3.
49 O. S eeck, (ed.),
Notitia
Dignita tum (Berolini, 1876),
i8o-i. The fa ct tha t
'Rutupis ' (a t
line
19)
is
printed
in
ita lics indica tes tha t it ha s been
s upplied by
the editor. The tex t of this
cha pter
of the Notitia
Dignita tum,
a s
it
a ppea rs
inthe Ox ford
ma nus cript,
is illus tra ted inJ.
Mothers ole,
The S a x onS hore
(London, 1924), pla te
fa cing p. 18; 'Rutupis '
is
pla inly
a bs ent. The
pla te fa cing p. 19
s hows the
picture
from the hea d of the
cha pter,
with
'Rutupis '
a dded
a mong
the
ca ptions .
5o
G. C. Boon, Is ca , 3rd ed.
(Ca rdiff, 1972), 62-4.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
FOURTH-CENTURY S AXONS 181
We ca nnow turnba ck to the ea rlies t
pa rt
of Ammia nus ' a ccount of the
yea rs 367-8
in
Brita in. It ha s been
a rgued
a bove tha t s ections
4
a nd
5
of x x vii. 8 s hould be ta ken
clos ely
together,
a nd tha t
they
form a n
unneces s a ry
a nd uninforma tive a s ide. If
they
a re
dis rega rded,
s ection
6,
where the
des cription
of Theodos ius '
ex ploits
is
s ta rted,
ca nbe s eento follow
clos ely
upon
s ection
3,
where Ammia nus records his
a ppointment.
The firs t words of s ection6 a re
'Ad ha ec
prohibenda '; they
mus t refer ba ck to the 'multa et metuenda ' of s ection
3,
to whichthe
a ppointment
of Theodos ius wa s a
res pons e.
Thes e mus t be identified withthe 'ins ta ntes
neces s ita tes ' of the
preceding
s ection. The clue to their na ture cons is ts of the
phra s e
which
Ammia nus
us es ,
ins ection
I,
to des cribe the conditionof Brita in: it ha d been
reduced,
he tells
us ,
'a d ultima m
inopia m'. 'Inopia '
mea ns 'la ck of'. It occurs a t
ma ny
other
points
inthe
na rra tive of
Ammia nus ;
it is s ometimes
qua lified by
a n
a ccompa nying
nouninthe
genitive,
s ometimes not. Thes e
nouns ,
of
cours e, va ry;
but a
highproportion
of them cons is ts of words
mea ning
'food'. 'La ck of
food', then,
is the
predomina nt
s ens e of
inopia .
And on
every
occa s ion
when
inopia
occurs without a
qua lifying noun,
its
mea ning, predicta bly,
is 'la ck of food'.51 And
it is
unqua lified
a t x x vii. 8.
I.
Wha t occurred in
367
wa s a fa mine. We ha ve s eentha t Theodos ius '
firs t a ctions were ta ken
a ga ins t
the
a rmy;
a nd it mus t be
pres umed tha t,
a t the root of the
trouble, la y
a
milita ry food-s horta ge.
This ca us ed a
riot,
which
triggered
off a nins urrection
a mong
the urba n
poor.
A food riot
brought
Theodos ius to
Brita in,
a nd it wa s his
principa l
a chievement there to
quell
it.
The full
ins ignifica nce
of the a chievements of Theodos ius inBrita innow becomes clea r.
The Roma n
a rmy's
dema nd for food wa s
perpetua lly
in
da nger
of
ex ceeding
the
s upply;
the
ris k of
s ta rva tion,
a nd the
s teps
ta kento fores ta ll
it,
form a
recurring
theme inAmmia nus '
His tory. Dea ling
withthe
problems
of the
food-s upply
wa s one of the mos t
ordina ry
ta s ks
withwhicha Roma n
genera l
of the fourth
century
wa s confronted. And we know
why
thos e
problems
a ros e. Cons ta ntine I ha d ins tituted
ma gis tri,
to be the
s upreme
comma nders of the
ca va lry
a nd the
infa ntry;
but the
provis ion
of the a nnona rema ined inthe ha nds of the
pra etoria n
prefects .
At a lower a dminis tra tive
level,
the
provis ion
of food for the
troops depended upon
the outcome of
negotia tions
betweenthe
prefect's deputy,
the
vica r,
a nd the loca l dux .52 A more
certa informula for dis a s ter ca nnot be
ima gined.
S ometimes the
food-s upply
fell into the
uns crupulous
ha nds of the s enior
milita ry officers ,
who mis us ed it for their own
purpos es .53
Onother
occa s ions ,
for one rea s onor
a nother,
the
pra etoria nprefect
fa iled to
provide
the
food tha t wa s needed.54 And this would s eem to be the s itua tionwhichconfronted Theodos ius
in
367.
The officia ls who received the bla me were the dux a nd the comes ma ritimi
tra ctus ;
the
la tter a t lea s t
pa id
for the food
s horta ge
withhis life
(x x vii.
8.
I).
But Theodos ius a ls o found it
neces s a ry
to
repla ce
the vica r
(x x vii.
8.
Io).
His
identity
is unknown. But
pres uma bly,
in
conjunction
withthe
pra etoria nprefect,
he ha d
engineered
the food
s horta ge
which
la y
a t the
51
Thes e occa s ions a re: x iv.
7. 2;
x iv.
Io. 4;
x v.
5. 29;
x vi.
4. 4;
x x ii.
14. I;
x x iv.
3. 14;
x x iv. 8.
2;
a nd x x x i. 8.
4.
'Inopia '
is
qua lified by
a noun
mea ning
food a t x iv. 6.
19;
x viii.
7. 8;
x ix . 10.
I;
x x vi.
3. 6;
x x vi. 8.
3;
x x vii.
5. 7;
x x viii. I.
17;
x x x . 6.
2;
a nd x x x i.
4.
II;
a nd
by
a noun
mea ning
drink a t x iv. 6.
I;
x v.
7.3;
a nd x x ii.
13.4.
At x x v.
7. 14,
food a nd drink occur
together ('potus inopia
et
cibi'). Only
five
pa s s a ges
s how
'inopia ' qua lified by
a noun
witha different
mea ning:
x vii.
7. 5 (a ux iliorum);
x viii. I.
4 (documentorum);
x x vii. 6. 2
(dignorum);
x x x .
I.
8
(na vium);
a nd x x x .
4.
16
(a llega tionum).
52
D. va n
Berchem,
Memoires de la
S ocidtd
na tiona le des
a ntiqua ires
de Fra nce lx x x
(1937), 193-4.
s
cf. Zos imus ii.
33. 5:
"
r61rTMov
TCr
A
rt-CECv pEpOS
E
Kipo "T00
c-rpa -rTlyo0 Ka I
Tc"V
OTTIPETJIopvcoV -Troi-rc XcoPEIV.
Codex Theodos ia nus vii.
4. 3,
is s ued in
357,
records tha t the Count of Africa ha d
wrongfullya ppropria ted
the food
s upply;
inthe s a me
yea r, Ba rba tio, ma gis terpeditum,
burnt
pa rt
of the food
s upply
of the
a rmyfighting
a ga ins t
the
Ala ma nni, pos s ibly
toca us e emba rra s s ment toJulia n
(Ammia nus
x vi. I I.
12-13).
54 In
360, the
pra etoria nprefect Florentius tried toca us e difficulties for
Julia n,
whowa s a t
Pa ris , byfa iling
tos end the a nnona from Vienne
(Ammia nus x x .
4.
6 a nd x x .
4. 9). Ea rlier,
in
354,
food
s upplies
were held
ba ck from the
troops
a t
Ca byllona ,
inorder tha t the life of the
pra etoria nprefect
Rufinus
might
be
put
a t ris k
(Ammia nus
x iv. 10.
3-5).
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
182 PHILIP BARTHOLOMEW
root of the troubles of
367.
If the elus ive a nd
imprecis e ex pres s ion
'ba rba rica
cons pira tio'
ha s
a
mea ning,
it mus t be a reference to their ma levolent
co-opera tion.55
COUNT THEODOS IUS AND THE S AXONS
It ha s beens howntha t Theodos ius did not
fight
the S a x ons inBrita in. Yet he did
fight
them
s omewhere: this is
proved by
a n
ins cription
found nea r
S tobi,
whichrefers to his
milita ry
ex ploits
in
Ea covEIa .56
It is
neces s a ry
to a s k where a nd whenthis
ca mpa ign
took
pla ce.
For
evidence,
we mus t turnto the la s t of the
Pa negyrics
inthe collected XII
Pa negyrici
La tini. 7 This wa s delivered before the
emperor
Theodos ius
by Pa ca tus ,
a t Rome inthe la te
s ummer of
389; a nd,
a t a n
ea rly s ta ge
inhis
s peech,
Pa ca tus a lludes to the
milita ry
dis tinctions
of Count
Theodos ius ,
who wa s the
emperor's
fa ther. He s ta tes
tha t,
ha d the
republica ns ys tem
of
a wa rding
titles to
genera ls
rema ined in
being,
the elder Theodos ius would ha ve beenca lled
S a x onicus ,
S a rma ticus a nd Ala ma nnicus
(Pa n.
La t. x ii.
5. 4).
Theodos ius is knownto ha ve
pa rticipa ted
intwo
ca mpa igns a ga ins t
the
Ala ma nni,
in
370 (Ammia nus
x x viii.
5. 15),
a nd in
371I,
whenhe
fought
withVa lentinia n
a ga ins t
the Ala ma nnic
king
Ma cria nus
(Ammia nus
x x ix .
4. 5).58 Together,
thes e
ca mpa igns
would ha ve
qua lified
him for the title Ala ma nnicus .
A contex t for S a rma ticus is ha rder to
provide,
s ince no S a rma tia n
ca mpa ign
is des cribed
by
Ammia nus . He
does , however,
s how the
younger
Theodos ius ina ction
a ga ins t
a bra nchof
the S a rma tia ns in
374,
whenhe is des cribed a s dux Moes ia e
(x x ix .
6.
15).
Pa ca tus indica tes
tha t,
a t a n
ea rly point
inhis
ca reer,
the
younger
Theodos ius
a ccompa nied
his fa ther on
ca mpa ign
(x ii.
8.
3), a nd,
s ince
nothing
s hows tha t the future
emperor ca mpa igned a nywhere
els e a t this
period,
it
ma y
be a s s umed tha t his fa ther
joined
him onthe middle Da nube a
yea r
or two
prior
to
374.
Confirma tionof this is
provided by Zos imus ,
who s ta tes tha t the
troops
who were
dis pa tched
to Africa to crus hFirmus ' revolt in
373
ca me from Pa nnonia a nd Moes ia
(iv.
16.
3);
pres uma bly
Theodos ius ,
who comma nded
them,
ha d the s a me
point
of
depa rture.
It
ma y
be
inferred
tha t,
before his
depa rture
for
Africa ,
Theodos ius
enga ged
ina
ca mpa igna ga ins t
s ome
of the ba rba ria ntribes of the Ba lka ns a nd tha t
this ,
to a
pa negyris t,
wa s s ufficient to ea rnhim
the title S a rma ticus .59
But,
to unders ta nd the
a pplica tion
of S a x onicus we mus t look ba ck a t a
s lightly
ea rlier
pa s s a ge
inthe
Pa negyric
where the
a uthor,
a lmos t overwhelmed
by
their
ma gnitude, ex pa nds upon
the a chievements of the elder Theodos ius a t
grea ter length.
In
full,
this
pa s s a ge (x ii. 5. 2)
rea ds a s follows :
'Quid, inqua m, fa cia m? qua e
Rhenus a ut Va ha lis vidit
a ggredia r?
Ia m
s e mihi S a rma tica ca ede
s a nguineus
His ter obiciet. Attrita m
pedes tribus proelis
Brita nnia m
refera m
? S a x o
cons umptus
bellis
na va libus
offeretur.
Reda ctum a d
pa ludes
s ua s S cotum
loqua r? Compuls us
ins olitudines a via s
omnis Ala ma nnus et
uterque
Ma urus occurrent'.
55
One
interes ting pos s ibilityma y
be mentioned here. Codex Theodos ia nus x i.
I. 15,
da ted to
19thMa y, 366,
is a ddres s ed to
Probus , pra etoria nprefect
of the Ga uls . If he ca nbe identified withthe Petronius Probus whos e
a va rice is
contemptuous ly
des cribed
by
Ammia nus a t x x vii. I I.
3
a nd x x x .
5. 5-6,
the
dis a ppea ra nce
of the
corn-s upply
for the Britis h
troops , a rra nged jointly
withhis
deputy,
the Vica r of the
Brita ins ,
will offer noca us e
for
s urpris e.
The mis deeds of Petronius Probus were
very
fa milia r toAmmia nus ' rea ders . His references tothem
a re ironica l a nd
verya mbiguous .
If his rea ders knew tha t Probus wa s
indirectly
involved inthe Britis hevents
of
367,
the la ck of
cla rity
withwhichAmmia nus
begins
his
des cription
of them
ma y
be
rea dily
unders tood.
56 L'Annde
dpigra phique (1931), 53.
Cf. a ls oR.
Egger, Byza ntion
v
(1929-30), 9-32 (=his
Romis che
Antike und
friihes Chris tentum,
Ba nd
I
(Kla genfurt, 1962), 126-43).
57
Reference ha s beenma de tothe editionof E.
Ga lletier,
Pa ndgyriques
la tins ,
tome
3 (Pa ris , 1955).
58
A.
Dema ndt,
Hermes c
(1972), 91-4
a nd
IIo-I.
Theodos ius wa s now
ma gis ter equitum
inthe
wes t,
a nd
held this
pos t
until his dea th.
59
It is
pres uma bly
this
ca mpa ign
towhicha llus ionis ma de inthe
ins cription
from nea r S tobi withthe
phra s e
Xdppa
pya
Aa pb6vcov.
There is no
jus tifica tion
for the a ltera tionof
Aa p66vcov
to
Bprr-rvcov,
put
forwa rd
by
W. Ens s lin
(RE 5A,
col.
1939),
a nd
printed byBirley, op.
cit.
(note 20), 336.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
FOURTH-CENTURY S AXONS
183
The
a ppa rent
effect of this is to
pla ce
the S a x ons inBrita in; but, ta kena s a
whole,
the
pa s s a ge
ra is es a
difficulty.
As we ha ve
s een,
three titles
only, S a x onicus ,
S a rma ticus a nd
Ala ma nnicus ,
a re
a s s igned
to Theodos ius a t x ii.
5. 4.
But a t this
point,
the
pa negyris t
is
hea ping
a s much
pra is e upon
Theodos ius a s the
s ubject
will
a llow;
ha d there been
a ny
ca s e a t a ll for
a s s igning
ex tra titles to
him, they
would ha ve beenrecorded here. But the vers ionof his
ca mpa igns given
inthe ea rlier
pa s s a ge implies tha t, by
the s ta nda rds of a
pa negyris t,
he ha d ea rned the titles
Brita nnicus ,
S coticus a nd Ma uricus . The a bs ence of Ma uricus need not
unduly
trouble
us , s ince,
s trictly s pea king,
the
object
of Theodos ius ' Africa n
ca mpa ign
wa s the
s uppres s ion
of a
us urper,
not the defea t of a ba rba ria ninva s ion.
Moreover,
Theodos ius ' ca reer inAfrica ha d ended
s pecta cula rly
onthe s ca ffold in
Ca rtha ge:
it would be ta ctles s for a
pa negyris t
to s tres s the
a s s ocia ted events ina nora tiondelivered before his s on. But the
dis crepa ncy
over the other two
titles needs to be
ex pla ined.
Now,
the
des cription
of Theodos ius '
ca mpa igns given
inx ii.
5.
2 cons is ts of three s ections .
Ea chs ection
compris es
two
s entences ,
whicha re writtento contra s t withone a nother. In
ea ch
ca s e,
the
point
of the contra s t is
geogra phica l.
The third s ectioncontra s ts the tra ckles s
wildernes s es inha bited
by
the Ala ma na nd the Moor withthe ma rs hes whichform the home-
la nd of the
S cot;
bothha d beenthe s cene of Theodos ius '
milita ry triumphs .60
But the
occurrence of 'the S cot' a t this
point
is
highly s us pect:
it ha s been
a rgued
a bove tha t a 'S cottis h'
cla im to a n
a ppea ra nce
inthe tex t of Ammia nus Ma rcellinus is
ex tremely dubious ,
a nd it is
mos t
unlikely
tha t
a nyone
a t Rome knew whether ma rs hes were a cha ra cteris tic fea ture of the
Iris h
la nds ca pe
or not. But s ome notionof the cha ra cter of the Da nube ba s inwa s
genera l.
It
ma y
be
s ugges ted
tha t 'S cotum' s hould be emended to 'Gothum'.61 There will thenbe a necho
here of Theodos ius ' Ba lka n
ca mpa ign.
If the reference to the Ma urus
ma y
be
dis counted,
both
the
opponents
mentioned inthe third
pa ir
of
contra s ting
s entences now fa ll withinthe
s cope
of
the honorific titles
a s s igned
to Theodos ius a little la ter: his dis mis s a l of the Ala ma ns to their
remote wildernes s es would ha ve ea rned him the title
Ala ma nnicus ;
a nd a s ucces s ful
ca mpa ign
a ga ins t
the
Goth,
whichdrove him ba ck to his na tive
ma rs hes ,
would ha ve merited the title
S a rma ticus . Thes e
ca mpa igns
a re a ttes ted els ewhere.
The firs t s ectionof
contra s ting
s entences ha s rivers a s its theme. The
ex periences
of the
Rhine a nd the Wa a l in
pers ona lized
form a re contra s ted withthos e of the Da nube. The la tter
ha d become 'S a rma tica ca ede
s a nguineus ',
a clea r
reference, a ga in,
to the
ca mpa ign
which
Theodos ius
fought
inthe Ba lka ns a nd ea rned him the title
S a rma ticus ;
it ha d beenthe s cene
of victories a s
grea t
a s thos e the other two rivers ha d witnes s ed. The Rhine will ha ve s eenthe
two
ca mpa igns a ga ins t
the Ala ma nni whicha re a ttes ted in
Ammia nus ; both,
obvious ly,
a re
s umma rized inthe
epithet
Ala ma nnicus . But the
ex perience
whichthe Wa a l ha d
undergone
ca n
only
be ma de clea r
by
a cons idera tionof the centra l s ectioninthis
pa s s a ge.
Now the
point
of the contra s t betweenthe two s entences whichform the centra l s ectionis the
va riety
of the kinds of
milita ry
s ucces s es whichTheodos ius won
a ga ins t
the S a x on. Victories
onla nd ha d been
recorded; they
could be
equa lled,
if not
ca pped, by
victories a t s ea . The a rea
where thes e
triumphs
took
pla ce
is
a ppa rently
s ta ted to be Brita in. But this is
only a ppa rent.
'Brita nnia m' is a
conjecture by
a
huma nis t,
ins erted into a ninferior
ma nus cript.
All the other
6o
In
quoting
this
pa s s a ge, Birley(op.
cit.
(note 20), 333)
omits the s econd
s entence, dea ling
withthe Ala ma n
a nd the
Moor,
from the third
pa ir
of contra s ts . The effect of this is
highlymis lea ding,
s ince it is
implied
tha t
the S cottis hvictories
belong
withthe
S a x on,
a nd tha t
both, a ppa rently,
s hould be
a s s igned
toBrita in.
61 The
rea ding
of the
ma nus cripts
is 'S cottum'
(not 'S cotum'),
whichwould ma ke a
corruption
from Gothum
ra ther more
proba ble.
At x ii. II.
4,
the
ma nus cripts
rea d
Cothus ;
but the occurrence of the words 'Hunus ' a nd
'Ha la nus ' inclos e connectionwith'Cothus ' ma ke the emenda tiontoGothus certa in. It would s eem tha t inboth
ca s es the initia l 'G' beca me
corrupted
to'C' a t a n
ea rlys ta ge
inthe tra ns mis s ionof the tex t. The 'S ' a t x ii.
5.
2
is
perha ps
a nerroneous
repetition
of the fina l 's ' inthe
preceding word,
's ua s '.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
184
PHILIP BARTHOLOMEW
ma nus cripts , including
the two whicha re direct
copies
of the los t
a rchetype,
rea d 'Ba ta via m'.62
It is
impos s ible
to
jus tify
its deletion
by
the editors of the modern
printed
editions . And we ha ve
s eentha t whenJulia n
ca mpa igned a ga ins t
the S a x ons in
358,
it wa s
precis ely upon
Ba ta via tha t
his
ca mpa ign
wa s centred. The S a x ons ca nnot ha ve moved fa r inthe
intervening period.
Like
Julia nbefore
him,
Theodos ius
fought
a s ucces s ful
ca mpa igna ga ins t
the S a x ons inthe a rea of
Ba ta via . Ba ta via mus t be the a covEia of the S tobi
ins cription;
a nd it wa s his victories here
tha t ea rned Theodos ius the title S a x onicus .63
The loca tionof Ba ta via ena bles us to unders ta nd wha t the
Pa negyris t
mea nt whenhe credited
Theodos ius withs ucces s es in'bellis na va libus '
a ga ins t
the S a x on. The whole a rea of Ba ta via
is cris s -cros s ed withwa ter-cours es of
every s ize;
Theodos ius mus t ha ve beenforced to under-
ta ke
a ga ins t
the S a x onthe kind of
a mphibious opera tions which,
on other
occa s ions ,
Ammia nus des cribes the Roma ns a s
ex ecuting
onthe Rhine a nd the
Tigris .
And Ba ta via is
bordered
by
the Lek to the northa nd the Wa a l to the s outh. The
point
of the
Pa negyris t's
ea rlier reference to the Wa a l now becomes
pla in:
victories
a ga ins t
S a x ons onthe Wa a l a re
linked withvictories
a ga ins t
Ala ma ns onthe Rhine ina
compa ris on
withthe s ucces s es
Theodos ius recorded
a ga ins t
the S a rma tia ns onthe Da nube. The two la tter
ca mpa igns
a re
clea rly
a tta ched to
epithets ;
the former
joins
the
ca mpa igns
of the s econd s ectionof this
pa s s a ge
in
giving
s ubs ta nce to the
epithet
S a x onicus . There a re no interna l
dis crepa ncies
inthis
pa rt
of Pa ca tus ' ora tion.
A
s ingle piece
of evidence rema ins to be dis cus s ed. This is the
s urviving
member of a n
identica l
pa ir
of
ins criptions
whichdedica ted the
bridge
of S a int Ba rtholomew inRome64
(C.LL. vi, 1175
=
LL.S .
771).
It lis ts the titles won
by
the
emperors Va lentinia n,
Va lens a nd
Gra tia nintheir
ca mpa igns a ga ins t
the ba rba ria ns .
Among
them is Fra ncicus ma x imus . We
ha ve s eentha t the Fra nks were clos e
neighbours
of the
S a x ons ;65
a nd it ha s been
pointed
out
ea rlier tha t the Roma ns did not
a lwa ys
s ucceed in
ma king
a clea r dis tinctionbetweenthe two.
It
ma y
be
s ugges ted
tha t the
a doption
of the title Fra ncicus ma x imus wa s
jus tified by
Theodos ius '
s ucces s ful
ca mpa igna ga ins t
the S a x ons inBa ta via . The da te of this
ca mpa ign
now becomes
a ppa rent;
for it ha s beendemons tra ted tha t Theodos ius ' Britis h
ca mpa igns
s hould be confined
to the
yea rs 367
a nd
368,66
s o tha t
369 a ppea rs
a s a
ga p
inhis ca reer. And the
ins cription
a t
Rome ca nbe
da ted,
withcons idera ble
precis ion,
to December
369. 369, therefore,
is the
yea r
inwhichTheodos ius '
ca mpa igna ga ins t
the S a x ons took
pla ce.67
THE S AXON S HORE
It is
neces s a ry
to end this
pa per
ona na nti-clima x . At the
beginning
of his a ccount of the
troubles of
367,
Ammia nus
s a ys
tha t their outbrea k wa s ma rked
by
the dea thof
Necta ridus ,
62
This ca nbe s eenclea rlyinGa lletier's tex tua l a ppa ra tus .
S ee a ls othe introductiontotome
I
of his edition
of the
Pa ndgyriques
la tins
(Pa ris , 1949),
where the s temma is
printed
on
p.
lv.
63
Anglo-S a x on a rcha eologis ts , intheir s ea rchfor prototypes for the ea rlies t
pottery
of the inva s ion
period,
ha ve concentra ted their efforts onthe a rea betweenthe Ems a nd the Wes er.
They
ha ve beeninduced todos o
by
the
tes timony
of Bede. But if the
interpreta tion
of the a ncient evidence offered here is
correct,
their
inves tiga tions
s hould be centred ona na rea a t lea s t
150
miles tothe s outh-wes t.
64 This
bridge,
whichlinked the
right
ba nk of the Tiber tothe Is ola Tiberina ,
ha s now been
repla ced by
the
Ponte Ces tio. The
ins cription
wa s ins erted intothe
pa ra pet
of the new
bridge. (F. Coa relli,
Guida
a rcheologica
di Roma
(1974), 313).
65
Their
geogra phica l overla p
is illus tra ted
by
the
entry
inJerome's Chronicle s .a .
373:
'S a x ones ca es i Deus one
in
regione
Fra ncorum'.
Oros ius , dra wing
onthis
entry, pla ces
their defea t 'in
ips is
Fra ncorum
finibus ' (vii.
32. io).
This
ca mpa ign
is
s lightly
la ter.
66
R.
Tomlin,
Brita nnia v
(1974), 303-9.
The
a ttempted
refuta tionof Tomlin's
a rguments by
R. C.
Blockley
(Brita nnia
x i
(1980), 223-5)
is
unconvincing.
67
It ca n
ha rdly
ha ve beena
la s ting
s ucces s . The inva s ionof the S a x ons des cribed
by
Ammia nus a t x x viii.
5.
1-7
occurred inthe
very
nex t
yea r.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions
FOURTH-CENTURY S AXONS
185
comes ma ritimi tra ctus .68 This officia l bea rs a
s triking
res embla nce to the comes litoris S a x onici
of the Notitia
Dignita tum.
Anidentifica tionis a ttra ctive: but inthe a bs ence of
a ny
further
evidence,
it is not
pos s ible
to a dd
a nything
to the blea k
s umma ry
of Profes s or
Wilkes ,
who
wrote tha t thes e two offices
'ma y,
or
ma y not,
be the s a me
thing'.69
Yet, s uppos e
it is a s s umed tha t the two counts a re identica l. It mus t thenbe a s ked wha t their
job wa s ; or, put
ina nother
wa y,
we need to know the
mea ning
of 'S a x ons hore'. The two
pos s ibilities
a re well-known: it
might
mea nthe a rea of a defens ive
s ys tem des igned
to
repel
S a x on
a tta cks ; or,
it
ma y
refer to the a rea of ea rlies t S a x ons ettlement inBrita in. Ina bbrevia ted
form,
thes e a re the 'S hore a tta cked
by'
a nd the 'S hore s ettled
by' hypothes es ;
a nd it is
neces s a ry
to choos e betweenthe two. The former ha s received a lmos t univers a l
a ccepta nce.
Yet it
depends
to a cons idera ble
degree
onthe
a s s umption
tha t one of the chief
ingredients
of the
events of
367
wa s a ninva s ion
by
S a x ons . It ha s been
a rgued
tha t this
a s s umption
is fa ls e. And
there is no other his torica l evidence to s how S a x oninva s ions
ta king pla ce during
the fourth
century.70
We would
a ppea r
to be left witha crude choice betweenthe
hypothes es ,
withno
evidence to
s upport
a decis ioninfa vour of one or the other. For the time
being
a t
lea s t,
the
mea ning
of the term 'S a x ons hore' mus t rema in
wha t,
in
rea lity,
it
a lwa ys
ha s been: a n
open
ques tion.
As hmolea n
Libra ry, Ox ford
68
Infa ct, this is bynomea ns certa in: s ee Cla rk's tex tua l a ppa ra tus tox x vii. 8. I.
"9
His tory
lx iv
(1979), 259.
70 It
would,
of
cours e,
be mis ta kento
a rgue, from the demis e of the comes ma ritimi
tra ctus , tha t the S a x ons
'mus t' ha ve
pla yed
a
pa rt
inthe Britis hevents of
367.
This content downloaded from 147.143.2.5 on Sun, 11 Aug 2013 06:49:57 AM
All use subject to JSTOR Terms and Conditions

Anda mungkin juga menyukai