The origins of the theophoric week

Early Medieval Europe





(



)

386

–

401

2007

The Author. Journal Compilation ©

2007

Blackwell Publishing Ltd,



Garsington Road, Oxford OX





DQ, UK and



Main Street, Malden, MA



, USA

Blackwell Publishing Ltd

Oxford, UK

EMED

Early Medieval Europe

0963-9462

XXX

Original Articles

The theophoric week in the Germanic languages

Philip Shaw

The origins of the theophoric week in

the Germanic languages





The Germanic terms for the days of the week that contain the names of
Germanic deities are often taken to result from Romano-Germanic inter-
actions in the fourth century AD. Yet this need not be the case: the
linguistic arguments underpinning this view are not decisive. A reassessment
of these names suggests that an early medieval process of transfer in scholarly
Christian contexts may be equally, if not more, plausible.

This article addresses a simple question: when and how were the
planetary names of the days of the week, used in Latin, borrowed into
the Germanic languages? Although the question is a simple one, how we
answer it could have considerable implications for our view of the pre-
Christian religious life of Germanic-speaking groups. It is usually
supposed that the processes of transfer took place around the fourth
century AD, with the adoption of the names by pagan Germanic-
speaking communities. This model can be used as the backbone of an
argument for a group of more or less pan-Germanic deities, worshipped
from at least the ﬁrst century AD onwards. For convenience, we will
refer to these deities as the ‘great gods’. This article will, however, cast
some doubt on claims that these deities were employed in Germanic
names of the days of the week as early as the late Roman period. This
may have implications for the case for the centrality of these deities
among most Germanic groups. The discussion below will suggest that

This article began life as a few paragraphs in my doctoral thesis, and I have therefore accrued
many debts in its production. My supervisors, Joyce Hill, Mary Swan and Ian Wood, all
contributed to the embryonic version of this piece, and I have also beneﬁted from a great
deal of good advice from colleagues in several disciplines at the University of Shefﬁeld. I am
very grateful to Morn Capper and Siân Prosser for reading and commenting on versions of
this paper. I am also indebted to Alaric Hall for stimulating discussions of the problems
involved, and to James Palmer for his very considerable efforts (and great patience) in bringing
this set of essays to fruition. The errors and idiosyncracies of the ﬁnished product are, of
course, all my own work.

The theophoric week in the Germanic languages

387

Early Medieval Europe





(



)

2007

The Author. Journal Compilation ©

2007

Blackwell Publishing Ltd

the fourth-century dating of the transfer cannot be proven, and that an
alternative date and cultural milieu in the seventh and eighth centuries
AD could provide a plausible alternative model.

The names of the days of the week in the Germanic languages

In the late Roman Empire, one of the ways of measuring time was the
seven-day week with which we are familiar today.

The days of this

week were sometimes named after planets, in a pattern now termed
the planetary week:

solis dies

(‘day of the sun’),

lunae dies

(‘day of the

moon’),

martis dies

(‘day of Mars’),

mercurii dies

(‘day of Mercury’),

jovis dies

(‘day of Jove[=Jupiter]’),

veneris dies

(‘day of Venus’),

saturni

dies

(‘day of Saturn’). The planetary week forms the basis for the names

of the days of the week in many Germanic languages to this day. The
great gods are the gods whose names appear in the Germanic version
of the planetary week. In English, for instance, Tuesday derives from
Old English

tiwes dæg

(‘Tiw’s day’), Wednesday from

wodnes dæg

(‘Woden’s day’), Thursday from

unres dæg

(‘Thunor’s day’), and Friday

from

frige dæg

(‘Frig’s day’). The deities Tiw, Woden, Thunor and Frig

have been treated as equivalents for Mars, Mercury, Jove and Venus in
translating the names of the days. This process of borrowing foreign
terms (the names of the days), but translating their individual parts
(the names of the planets/deities and the word

dies

), is known as loan-

translation. Loan-translation also produced Monday and Sunday, which
correspond to

lunae dies

(‘day of the moon’) and

solis dies

(‘day of the

sun’) respectively. Saturday, however, involves a direct borrowing of the
name Saturn from

saturni dies

(‘the day of Saturn’). This pattern of

loan-translations and loanwords from Latin appears to have been common
to many of the medieval Germanic languages, although there are some
signiﬁcant variations.

In order to understand the processes involved in the transfer of the

Roman planetary names into the Germanic languages, we must also
consider the alternatives to the planetary names employed in some
Germanic dialects. The basic pattern derived from the planetary week
is common to all Germanic languages. In some dialects, however, a partial

A good, recent summary of the main points outlined in this paragraph is D.H. Green, ‘Zu
den germanischen Wochentagsnamen in ihren europäischen Beziehungen’, in W. Haubrichs,
W. Kleiber and R. Voß (eds),

Vox Sermo Res: Beiträge zur Sprachreﬂexion, Literatur- und

Sprachgeschichte vom Mittelalter bis zur Neuzeit

(Stuttgart and Leipzig, 2001), pp. 223–35, at

pp. 223–7.

D.H. Green,

Language and History in the Early Germanic World

(Cambridge, 1998), pp. 243–

53. Green, ‘Wochentagsnamen’, pp. 224–8. K. Helm,

Altgermanische Religionsgeschichte

, 2 vols

(Heidelberg, 1953), II.2, pp. 231–3.

388

Philip Shaw

Early Medieval Europe





(



)

2007

The Author. Journal Compilation ©

2007

Blackwell Publishing Ltd

planetary week came to be employed, with non-planetary names for
some days. The commonest non-planetary names in medieval German
dialects were the ancestors of Modern German

Dienstag

(‘Tuesday’),

Mittwoch

(‘Wednesday’) and

Samstag

(‘Saturday’). It has been suggested,

on the basis of the unique Romano-Germanic votive inscription to

Mars

Thingsus

, found on Hadrian’s Wall, that the medieval form

Dingstag

(

Dienstag

) is also a loan-translation of Latin

martis dies

This relies on

identifying

Mars Thingsus

as an alternative name for the god Tiw, but it

is far from clear that this identiﬁcation is correct. This name would repay
further study, but there is not the space here to discuss it further.

Mittwoch

(Old High German

mittawecha

) and

Samstag

(Old High German

sambaztag

) are of more obvious signiﬁcance:

Mittwoch

simply means ‘mid-

week’, and is a loan-translation of the ecclesiastical Latin term

media

hebdomas

, while

Samstag

derives from Latin

sambatum dies

(‘sabbath day’).

Less common alternatives to planetary names include the Greek-

derived Bavarian forms

Ertag

(‘the day of Ares’, i.e. Tuesday),

phinztag

(‘the ﬁfth day’, i.e. Thursday) and

pherintag

(from Greek

paraskeue

, i.e.

Friday), perhaps borrowed from Gothic, which was heavily inﬂuenced
by Greek.

There are also two names which make use of the positioning

of days in relation to other days:

aftermontag

(‘after Monday’ i.e.

Tuesday), which is known mainly from the diocese of Augsburg in the
south of modern Germany, and

sunâbent

(‘sun evening’, i.e. Saturday,

the day before Sunday), evidenced in some Middle High German texts.

One author alone, Notker of Saint Gall, appears to have attempted to
provide an explicitly Christian Germanic term for Sunday, coining
frônetag (‘the lord’s day’) as a loan-translation of Latin dies dominicus.

This probably represents a learned coinage, and may reﬂect personal
preference rather than common usage.

Some of the other uncommon

forms, however, apparently had wider currency: the Bavarian forms, in
particular, seem to suggest processes of adoption of names ultimately
from the Greek-speaking world, although these names subsequently
gave way, for the most part, to the set of names which is most commonly
used in modern German.

The names Dienstag, Mittwoch and Samstag seem to have enjoyed a

wider currency still, and clearly these represent an important deviation
from the wholly planetary week which obtains across many Germanic
languages. The fact that the planetary names were not universal among

Green, ‘Wochentagsnamen’, pp. 224–5.

Green, ‘Wochentagsnamen’, pp. 225–7.

Green, ‘Wochentagsnamen’, pp. 226 and 232.

Green, ‘Wochentagsnamen’, pp. 225 and 227.

Green, ‘Wochentagsnamen’, p. 227.

Green, Language and History, p. 249.

The theophoric week in the Germanic languages

389

Early Medieval Europe

  ()

the medieval Germanic languages suggests that processes of adoption of
names for the days of the week were not uniform across all Germanic-
speaking groups. The overall pattern has been characterized as one in
which planetary names were retained in the north-west, while Christian
alternatives such as Mittwoch and Samstag intruded in more southerly
areas.

There is not space here to list – let alone attempt to account

for – every variation, and this article’s focus is on the planetary names
as part of a broader pattern of evidence for the great gods. It will,
nevertheless, be useful to return to the broad pattern of retention and
replacement of planetary names once their origins and development
have been considered in more detail.

Should we date the Germanic day-names to the

late Roman period?

The fact that this set of names is common to many Germanic languages
can be seen as an indication that these names were transmitted from
Latin into the Germanic languages at an early stage, before they constituted
several distinct languages. The presence of the effects of i-mutation in
the Old English form sæterdæg can also be seen as evidence that the
borrowing of this name took place at an early stage in the development
of Old English.

It has also been suggested that the use of <t> in this

word – where other loanwords sometimes have <d> or <2>, representing
a voiced form of the consonant which developed in Vulgar Latin –
provides evidence for considering it one of a group of ‘very early
loan-words’.

In addition to these internal dating criteria, there are

also external criteria to consider. Good recent discussions of these are
provided by D.H. Green in a chapter on the day-names in his very
important book Language and History in the Early Germanic World, and
in a subsequent article on the topic.

He identiﬁes two major external

dating criteria. One is that the interactions between Latin and Germanic
speakers along the Rhine frontier in the late Roman empire provide a
plausible social milieu for the adoption of these names by Germanic
speakers.

Another, and perhaps more important, criterion is that the

Romance languages generally have Christianized versions of the weekend
days: French, for instance, has samedi and dimanche, deriving from Latin
sabbatum and dominica. This would suggest, then, that the Germanic

Green, ‘Wochentagsnamen’, p. 228.

A. Campbell, Old English Grammar (Oxford, 1959; repr. 1997), §495 and §504.

Campbell, Old English Grammar, §530.

Green, Language and History, pp. 236–53. Green, ‘Wochentagsnamen’.

Green, Language and History, p. 246.

390

Philip Shaw

Early Medieval Europe

  ()

names must have been borrowed before these Christianized names
became the norm in proto-Romance.

All of these dating criteria are open to question. The spread of these

names across many of the Germanic languages could, of course, be the
result of early transmission: but it is hardly impossible – or even
unlikely – that such a spread could occur in the early Middle Ages,
rather than in the late Roman period. The case is much the same for
i-mutation: while i-mutation was clearly operative at an early stage in
the development of Old English, it is by no means clear that this stage
occurred prior to the Anglo-Saxon settlements in England.

We would

be unwise to place great reliance on i-mutation as evidence for a pre-
migration borrowing of the name. Similar difﬁculties arise in relation
to the use of <t> spellings as a dating criterion: this relies on the
assumption that Sæterdæg involved a borrowing from Vulgar Latin, but
this need not have been the case.

The development of a Christian weekend in the Romance languages

is a more signiﬁcant issue. It seems quite clear that the Germanic names
must have been borrowed before this Christian weekend became the
dominant Romance form. When the latter development took place,
however, is hard to determine. The earliest Romance vernacular to
provide large quantities of material is Old French, and, while there are
a few earlier texts in this language, substantial production of Old French
texts does not appear to start until around the beginning of the twelfth
century.

Old French uses the Christian weekend forms with great

regularity. This suggests that the transition from a pagan to a Christian
weekend had been completed by the twelfth century, but it is clear that
in the late Roman period the situation was much more mixed. In Latin
inscriptions of the late Roman empire, we ﬁnd both types of weekend.

Green, Language and History, pp. 241–6.

For instance, Joseph B. Voyles, Early Germanic Grammar: Pre-, Proto-, and Post-Germanic
Languages (San Diego, 1992) suggests that i-umlaut of the sort responsible for the form
Sæterdæg should be dated to around 700 AD (pp. 142–3), although no explicit reasoning is
given for specifying this date. Voyles does, however, make a persuasive case for rejecting the
view that i-mutation reﬂects a development in Primitive Germanic (pp. 79–85).

See M. Zink, Littérature française du Moyen Age (Paris, 1992), pp. 27–41 on the earliest texts
that can be termed French, and pp. 71–130 on the literary developments of the late eleventh
and early twelfth centuries.

See, for instance, Inscriptiones Latinae Christianae Veteres, ed. E. Diehl, 3 vols (Berlin, 1961),
nos. 261, 707, 1689, 2104, 3532a, 4391 (dominica); 1312, 1385, 1646, 1706, 3129, 3904 D, 4216,
4408 (sabbato); 543, 577, 2146a, 3650, 4377, 4405, 4406, 4406 A, 4407 (saturni); 1148, 1334,
4372, 4379, 4387, 4388, 4389, 4390, 4553, 4797 (solis). The majority of these inscriptions date
to the fourth and ﬁfth centuries, but no. 1148 provides an example of dies Solis probably as
late as the early seventh century. See also the excellent survey of epigraphic evidence for the
planetary week in M.A. Handley, Death, Society and Culture: Inscriptions and Epitaphs in Gaul
and Spain, AD 300–750, BAR International Series 1135 (Oxford, 2003), pp. 110–16, which
conﬁrms the mixed character of early medieval epigraphic practice with regard to the names
for Saturday and Sunday.

The theophoric week in the Germanic languages

391

Early Medieval Europe

  ()

In early medieval Latin texts we also ﬁnd both still in use, although
some authors make a point of recommending the Christian forms.
Isidore of Seville, for instance, having explained the origins of the
planetary day-names and drawn a careful distinction between the Judaeo-
Christian numerical names and the planetary names used by pagani,
saeculares and gentiles, gives a statement – substantially dependent on
Augustine of Hippo’s Enarrationes in psalmos – of the usage Christians
should adopt: ‘but in the matter of naming the days, the ecclesiastical
habit of speaking comes better from the Christian mouth’.

Caesarius of

Arles also complains of the use of the planetary week, and recommends
the use of numerical names: ‘let us never say day of Mars, day of
Mercury, day of Jove, but let us name them ﬁrst and second or third
day, according to what has been written’.

Like Isidore, Caesarius was

no doubt inﬂuenced by Augustine’s views on the matter, but we might
be unwise to suppose that either author was emptily repeating the now-
irrelevant prohibitions of earlier centuries.

Whether or not Isidore or

Caesarius paid much attention to quotidian speech patterns in the
populace at large, the usage in Spain and France continues up to the
present day to include direct descendants of most of the Latin planetary
day-names. It is clear, therefore, that quotidian speech in early medieval
Spain and France made use of such names.

If we turn to much later sources, moreover, we ﬁnd that Latin usage

apparently remains mixed throughout much of the Middle Ages.
Johannes Beleth, writing in twelfth-century France, discusses the
planetary names as follows:

Among the pagans, indeed, both according to the common people
and the learned, they were labelled in the same way, but there is a
difference in the reasons for the names. The ﬁrst day is called the day
of the Sun, the second the day of the Moon, the third the day of Mars,
the fourth the day of Mercury, the ﬁfth the day of Jove, the sixth the
day of Venus, the seventh the day of Saturn, but this order is back

‘Melius autem in vocabulis dierum de ore Christiano ritus loquendi ecclesiasticus procedit’:
Isidori Hispalensis Episcopi Etymologiarum sive Originum Libri XX, ed. W.M. Lindsay, 2 vols
(Oxford, 1911), I, Book V, c. 30, section 11. My translation. For Augustine’s discussion of
the day-names, see Sancti Aurelii Augustini Ennarationes in Psalmos, ed. E. Dekkers and
I. Fraipont, 3 vols, CCSL 38–40 (Turnhout, 1956), II, 1302–3 (psalm 93, section 3).

‘nunquam dicamus diem Martis, diem Mercurii, diem Iovis; sed primam et secundam vel
tertiam feriam, secundum quod scriptum est, nominemus’: Sancti Caesarii Arelatensis Ser-
mones, ed. G. Morin, CCSL 104, 2nd edn (Turnhout, 1953), p. 785 (no. 193). My translation.

As Yitzhak Hen has pointed out, Caesarius was not greatly concerned with paganism and
superstitions, but does appear to have recognized ‘that the practices he condemned as pagan
were often detached from their pagan-religious meaning’ (‘Paganism and Superstitions in the
Time of Gregory of Tours: Une question mal posée! ’, in K. Mitchell and I. Wood (eds), The
World of Gregory of Tours (Leiden, 2002), pp. 229–40, at p. 231).

392

Philip Shaw

Early Medieval Europe

  ()

to front. For the moon is less than a planet, and Saturn is greater
than all the others. The pagan commons used to name them in this
way because they believed them to be gods, the Sun and the Moon and
Mars and the others, from whose names the philosophers labelled the
days, because these are the names of planets, from whose motions
and characteristics all things, according to them, derive their impulse.

Sacred scripture, however, has not received these names. In the

usage of the common people, nevertheless, they are maintained.

The description of the planetary names here can, of course, be seen

as part of a tradition stemming back at least as far as Isidore of Seville,
as the editor of Beleth’s treatise points out in the apparatus for this
passage. The explicit claim that these names were in popular use in Beleth’s
day need not, however, represent mere transmission of the Isidorian
tradition. Again, since most of the planetary names continue in use in
French up to the present day, we have no reason to doubt Beleth’s
testimony. Indeed, Old French texts from the early twelfth century
onwards substantiate his claim, recording Old French reﬂexes of the
Latin planetary names for the days other than Saturday and Sunday.

These reﬂexes include not just the pattern ‘name+dies’ (which underlies
modern lundi, mardi, mercredi, jeudi and vendredi) but also the pattern
‘dies+name’.

The fact that the two elements of these names could still

be combined in either order suggests that at this period, or at least
shortly before, these names were still analysable as consisting of two
meaningful parts. They were not, therefore, completely fossilized, and
the underlying meanings of the names were apparently still problematic
for French Christian authors in the early twelfth century, as Philippe
de Thaon’s reanalysis of the names in his Comput suggests.

It would

‘Apud gentiles uero, tum secundum uulgus, tum secundum philosophos, eisdem uocabulis
denominantur, sed diuersitas est in causis nominum. Prima dies dicitur dies Solis, secunda
dies Lune, tertia dies Martis, quarta dies Mercurii, quinta dies Iouis, sexta dies Veneris,
septima dies Saturni, ordine tamen retrogrado. Luna enim inferior est planeta, Saturnus
omnibus superior. Vulgus gentilium sic eos nominabat, quia credebat eos deos esse, Solem et
Lunam et Martem et alios, a quorum nominibus dies denominabant philosophi, quia hec
sunt nomina planetarum, ex quorum motibus et naturis omnia suam uegetationem secundum
illos trahunt. Hec autem nomina sacra scriptura non recipit. In usu tamen uulgari habentur.’
Iohannis Beleth Summa De Ecclesiasticis Ofﬁciis, ed. H. Douteil, CCCM 41 and 41A, 2 vols
(Turnhout, 1976), II, 8–9 (c. 3). My translation.

See A. Tobler and E. Lommatzsch, Altfranzösisches Wörterbuch, 11 vols (Wiesbaden, 1925–
2002), s.v. lundi, marsdi, mercredi, juesdi, vendredi.

Tobler and Lommatzsch, Altfranzösisches Wörterbuch, s.v. deluns, demars, demerques, dijuès, divenres.

Philippe de Thaon: Comput (MS BL Cotton Nero A.V), ed. I. Short, Anglo-Norman Text
Society, Plain Texts Series 2 (London, 1984), pp. 11–12 (lines 493–578). Prior to this passage,
Philippe outlines the etymological derivations from the classical gods, and in this passage
proposes a set of unetymological Christian relationships, relating, for instance, marsdi to
martyrie and jusdi to the adjective joius (pp. 10–11, lines 423–82).

The theophoric week in the Germanic languages

393

Early Medieval Europe

  ()

appear, therefore, that by the twelfth century, replacement of the planetary
weekend with a Christian weekend was largely complete. Nevertheless,
the processes of Christian anxiety regarding the planetary names were
still at work. In Portuguese such processes extended the Christian
terminology of numbered feria throughout the week, whereas in French
the more accepting approach implicit in Philippe de Thaon’s appropriation
and re-interpretation of the planetary names evidently won out.

We have, then, little evidence for early medieval Romance usage,

unless we accept Wright’s strong arguments that early medieval Latin
was (in Romance-speaking areas) employed as the written form of the
early Romance languages.

This would then suggest that early medieval

Romance usage was mixed, as early medieval Latin usage was. Even if
we did not accept Wright’s claims, the lack of evidence for Romance
usage should caution us against employing the Christian weekend in
the Romance languages as evidence for the pre-medieval origins of the
Germanic pagan weekend. Overall, the linguistic criteria outlined above
do not support the claim that the Germanic day-names must have been
borrowed from Latin in the period of the late Roman empire. We
cannot, on linguistic grounds, exclude the possibility that the borrowing
took place later than this. It is therefore worth reassessing the evidence
for the socio-religious life of Romano-Germanic communities on the
Rhine frontier in the late imperial period. Such a reassessment may help
us to determine how likely it is that the day-names were borrowed in this
context. As noted above, the late Roman frontier has been identiﬁed as
providing a plausible cultural milieu for the transfer of the day-names,
but, as we shall see, there is reason to doubt that this is the case.

Reassessing the fourth-century Rhine frontier context for

transference of the day-names

There is a substantial quantity of textual evidence, created by pre-
Christian Germanic individuals and groups on the Rhine frontier, of
their religious practice. In this area, there are large numbers of votive
inscriptions to gods and goddesses with Germanic names. Such
Romano-Germanic votive inscriptions begin to appear around the ﬁrst
century AD, and continue to be produced until the fourth century AD.
Tacitus’s oft-repeated claim that the Germani worship Mercury most
of all, and also Hercules and Mars, can usefully be compared with the

Handley, Death, Society and Culture, p. 111 suggests that the use of numbered feria ‘may have
been adopted in western Iberia’ around the mid- to late sixth century AD.

For a summary of the case, see R. Wright, A Sociophilological Study of Late Latin, Utrecht
Studies in Medieval Literacy 10 (Turnhout, 2002), pp. 3–17.

394

Philip Shaw

Early Medieval Europe

  ()

evidence available from such inscriptions.

Although relatively few such

inscriptions survive from Tacitus’s day – with the bulk of production
taking place in the late second and early third centuries AD – the small
number of extant inscriptions from the ﬁrst century AD clearly indicates
that at least some equations of native and Roman gods were already in
use in this area by this date.

Mercury, Mars and Hercules are the

commonest identiﬁcations for non-Roman gods in the region. Derks
has discussed the gods identiﬁed by a non-Roman as well as a Roman
name, and notes that gods equated with Hercules and Mars appear to
have had public cults, while gods equated with Mercury, despite being
numerous, appear to have enjoyed private cults.

There are also

regional patterns to be observed. Hercules, paired with the native name
Magusanus, is common around the northern end of the Rhine, and
appears to be associated speciﬁcally with the Batavi.

Mars, on the other

hand, is concentrated mainly in the region inhabited by the Treveri,
and is paired with several different native names.

Mercury appears over

a wider area, but is not so common in the area in which the cult of
Hercules Magusanus is attested.

In no case of which the present author

is aware is any of these gods called Wodan, Thonar or Tiw, or anything
similar to these names.

There were, then, a number of apparently well-established inter-

pretationes germanicae of Mercury, Mars and Hercules in the Rhine area,
at least some of which existed in Tacitus’s day.

Deities identiﬁed with

Tacitus, Germania, c. 9, in H. Furneaux and J.G.C. Anderson (eds), Cornelii Taciti Opera
Minora (Oxford, 1900; repr. 1958).

T. Derks, Gods, Temples and Ritual Practices: The Transformation of Religious Ideas and Values
in Roman Gaul, Amsterdam Archaeological Studies 2 (Amsterdam, 1998), pp. 87–90 and
84–5.

Derks, Gods, Temples and Ritual, pp. 95–9.

N. Roymans, ‘The Sword or the Plough: Regional Dynamics in the Romanisation of Belgic
Gaul and the Rhineland Area’, in N. Roymans (ed.), From the Sword to the Plough: Three
Studies on the Earliest Romanisation of Northern Gaul, Amsterdam Archaeological Studies 1
(Amsterdam, 1996), pp. 9–126, at pp. 90–1. Derks, Gods, Temples and Ritual, p. 98.

N. Roymans, Tribal Societies in Northern Gaul: An Anthropological Perspective, Cingula 12
(Amsterdam, 1990), pp. 55–7. Derks, Gods, Temples and Ritual, pp. 96–8.

Derks, Gods, Temples and Ritual, p. 99.

Roymans, ‘The Sword or the Plough’, argues that, while Tacitus might have had in mind the
cult of Hercules Magusanus in referring to Germanic peoples worshipping Hercules, it is
more likely ‘that Tacitus was alluding to a native deity or deities, comparable in his opinion
to (the Roman) Hercules’ (p. 94). Roymans notes in support of this view the fact that Tacitus
refers in his Annales (Book II, c. 12; see The Annals of Tacitus: Books 1–6, ed. F.R.D. Goodyear,
Cambridge Classical Texts and Commentaries 23 (Cambridge, 1981), p. 23) to a wood sacred
to Hercules somewhere in northern Germany (p. 94). Roymans is no doubt correct in
thinking that this account demonstrates interpretatio romana of a Germanic god on the part
of Tacitus or one of his informants, but it need not follow that Tacitus was not inﬂuenced
by the Magusanus cult in his treatment of Hercules in the Germania. Derks, Gods, Temples
and Ritual, speciﬁcally rejects a link between the paired names of deities found in the Rhine
area and Tacitus’s interpretatio romana, arguing that positing such a link ‘hold[s] the Roman
authorities responsible for the associations’ (p. 100, n. 103). He does not, however, address

The theophoric week in the Germanic languages

395

Early Medieval Europe

  ()

Mars and Hercules were more geographically restricted than deities
identiﬁed with Mercury. All of this accords remarkably well with Tacitus’s
claims. If, as seems likely, his fullest information on Germanic tribes
came from the Rhine region, then the picture he received would have
been just what has been outlined here: Mercury received the widest cult,
while Hercules and Mars were also both signiﬁcant deities. The existence
of these interpretationes germanicae calls into question the claim that the
names of the days of the week reﬂect Germanic loan-translations of the
Latin names in the late Roman period. It is apparent from these votive
inscriptions that different Germanic groups, even within a relatively
restricted geographical area, equated quite a variety of deities with
Mercury and Mars. It would be surprising indeed if all or most of these
tribes chose to equate two different deities (to whom no extant votive
inscriptions attest) with Mercury and Mars in the speciﬁc context of
naming the days of the week. The simplest way of accounting for this
data is to suppose that the Germanic day-names were borrowed from
Latin in a different cultural context. It is worth considering, therefore,
an alternative model for the transference of the names of the days of
the week into the Germanic languages.

An alternative model for the transference of the day-names

Christine Fell has suggested that the origins of the Old English names
of the days of the week lie in the process of conversion: ‘In teaching the
Anglo-Saxons how to measure time it is evident that the missionaries
simply offered them loan translation, the sun’s and moon’s days
remaining as luminary names, the others translating into whatever the
missionaries and their converts thought the nearest equivalent.’

This

model has much to recommend it, but in conﬁning itself to the Anglo-
Saxon day-names it fails to answer the questions raised by the presence
of these names in other Germanic languages. In attempting to account
for the spread of the planetary week across the Germanic languages, we
need to consider not only processes of conversion and Christianization,
but also the intellectual culture of Christian Germanic societies.

the possibility that the inscriptions reﬂect a pre-existing interpretatio germanica which could
have inﬂuenced Tacitus’s interpretatio romana, although he makes a convincing case for the
equations between Germanic and Roman deities as ‘a speciﬁc selection [of equations]
connected with native perception of the Roman pantheon’ (p. 100).

C.E. Fell, ‘Paganism in Beowulf : A Semantic Fairy-Tale’, in T. Hofstra, L.A.J.R. Houwen
and A.A. MacDonald (eds), Pagans and Christians: The Interplay between Christian Latin and
Traditional Germanic Cultures in Early Medieval Europe, Mediaevalia Groningana 16 (Groningen,
1995), pp. 9–34, at p. 18.

396

Philip Shaw

Early Medieval Europe

  ()

Some of our earliest records of Old English appear in glossaries that

are largely Latin–Latin, but with occasional English glosses. The Épinal-
Erfurt and Corpus glossaries derive in part from the earliest English
glossarial activity, which may have been associated with the school of
Theodore and Hadrian at Canterbury.

Amongst other references to

classical pagan mythology, these texts gloss the names of some of the
planetary deities who feature in the day-names. Michael Herren has seen
in such glossaries evidence of a ‘mini-renaissance’ of classical learning in
southern areas of England in the late seventh and early eighth centuries.

He speciﬁcally notes that ‘the Anglo-Saxons of the late seventh century
were conscious of the relation between the gods and prodigies of Graeco-
Roman mythology and the divinities and wondrous beings of their
native religion and worked out a table of correspondences between the two
mythological systems’.

This is a plausible reading of the Anglo-Saxon

evidence, but it ignores continental evidence for such equations of
Graeco-Roman and Germanic deities.

The earliest references to these Germanic deities on the Continent

belong to the seventh century, and most if not all of such references are
the work of Christians – usually Christian authors with a good classical
education, such as Jonas of Bobbio. Jonas’s Vita sancti Columbani
contains what could be the ﬁrst equation of one of the great gods with
a Graeco-Roman god, apparently equating Wodan with Mercury.

The

earliest references to these deities are, then, predominantly the work of
men who were brought up reading classical Latin texts, men who were
professional Christians whose lives turned around the seven-day week
that regulated monastic life. They did not pluck these deities out of the
air – no doubt they were worshipped in the early Middle Ages in at least
some Germanic areas – but they were clearly in a position to have a
massive impact on the diffusion of these deities in Christian Germanic
culture. The Old English glossaries provide evidence of deliberate,
scholarly efforts to equate Mercury with Wodan and Mars with Tiw
and Jove with Thonar. In such an intellectual climate, we should not
be surprised if Christian scholars also made use of these equations to
loan-translate the Latin days of the week. This might also explain the

M. Lapidge, ‘The School of Theodore and Hadrian’, in M. Lapidge, Anglo-Latin Literature,
600 –899 (London, 1996), pp. 141–68, at pp. 153–5 (originally published in Anglo-Saxon England
15 (1986), pp. 45–72).

M.W. Herren, ‘The Transmission and Reception of Graeco-Roman Mythology in Anglo-
Saxon England, 670–800’, Anglo-Saxon England 27 (1998), pp. 87–103, at p. 102.

Herren, ‘Transmission and Reception’, p. 103.

Jonas of Bobbio, Vita Columbani, I.27, ed. B. Krusch, MGH SRM 4 (Hanover, 1902),
pp. 64–152, at p. 102. Jonas de Bobbio, Vie de Saint Columban et de Ses Disciples, trans. A. de
Vogüé and P. Sangiani, Vie Monastique 19 (Bégrolles-en-Mauges, 1988), p. 159 (n. 9) suggests
that this equation is a later interpolation from Paul the Deacon’s Historia Langobardorum.

The theophoric week in the Germanic languages

397

Early Medieval Europe

  ()

presence of Saturn in untranslated form in the Germanic day-names:
this is the one deity for whom no obvious parallel exists among the
deities known to early medieval Christian authors.

The arguments put forward by Green for the transfer of the planetary

week from Latin to the Germanic languages in the late Roman period
rest, essentially, on an assumption that the planetary week is a pagan
phenomenon, which Christians always sought to replace with Christian
equivalents.

As Handley has clearly demonstrated, however, this was

not the case on the Continent in the late Roman period.

It is also

clear, moreover, that the attitude of the Anglo-Saxons must have been
basically accepting. Had Christian opposition to the planetary week
existed and continued in England, it would eventually have fallen out of
use. The fact that it did not, together with the lack of later complaints
against the usage, suggests that the names survived because they ceased
to be considered objectionable; or, quite possibly, because Anglo-Saxon
society had never, from the outset, considered them objectionable. It is
possible that not all Germanic-speaking societies regarded the planetary
week as unobjectionable. The pattern of a complete planetary week in
the north-west, with only a partially planetary week in more southerly
areas, discussed above, could be used as evidence for Christian opposition
further south. One could argue that the complete planetary week was
initially in use in southerly areas as well, and that early medieval
ecclesiastics attempted to replace it (or at least parts of it) with more
acceptable alternatives. We should consider the possibility, however,
that some southerly Germanic dialects never used a full planetary week.
The Bavarian names discussed above can probably be understood as
reﬂecting a fairly early adoption of Greek names, perhaps via Gothic:
whether or not Ertag indicates that a Greek version of the planetary
week operated side by side with a competing numbered system,
reﬂected by the other Bavarian names, must remain an open question.
This process of adoption of names is clearly distinct from the adoption

Carl Edlund Anderson suggests that the somewhat mysterious Old Norse form laugardagr
(‘bath day’) may also be the result of difﬁculties in ﬁnding an obvious equivalent for saturni
dies (‘Scandinavian Days: Old or New’, pp. 10–11, published at <http://www.carlaz.com/phd/
Scandinavian_Day_Names.pdf>, accessed November 2006). G. Rausing, ‘The Days of the
Week and Dark Age Politics’, Fornvännen 90 (1995), pp. 229–39, at pp. 233–4, suggests, quite
implausibly, that laugardagr derives from an alternative name of the god Loki, used as an
equivalent for Saturn. L. Hermodsson, ‘On Some Names of Days’, Fornvännen 92 (1997),
pp. 202–3, quite rightly rejects this claim on linguistic grounds. J. Ekermann, ‘Second
Thoughts on Some Names of Days’, Fornvännen 93 (1998), pp. 201–2, responds to Hermods-
son with some extraordinary claims, including the suggestion that ‘Saturday’ itself derives
from an otherwise unattested Saxon god called ‘Sater(n)e or Sæter(n)e’ (p. 201).

Green, Language and History, pp. 243–5.

Handley, Death, Society and Culture, pp. 110–16.

398

Philip Shaw

Early Medieval Europe

  ()

of the planetary week from Latin, and could have happened at about
the same time, although this need not have been the case.

The adoption of sambatum dies as sambaztag in Old High German

suggests a quite different line of transmission. Green relates it to modern
French samedi, noting the distinction between the Vulgar Latin form
sambatum, which produces the <m> in both samedi and Samstag, and
the literary sabbatum.

As has been shown above, the French form

samedi must have replaced saturni dies sometime before the twelfth
century, but when exactly cannot be determined with any certainty. We
could suppose that the change in the ancestor of French precipitated a
parallel change in some neighbouring Germanic dialects, producing
Old High German sambaztag. On the other hand, as we have seen, it is
likely that both sambatum and saturni forms were in use at the same time
in the ﬁrst few centuries of the early Middle Ages: perhaps Anglo-Saxon
authors chose to loan-translate the planetary variant, while in areas of
what is now southern Germany the explicitly Christian variant was
borrowed instead. This may be partly the result of the particular interest
in engaging with pagan classical literature in England in the seventh and
eighth centuries, and especially a consequence of the emphasis in England
on ﬁnding vernacular equivalents for Roman mythological ﬁgures.

A similar interpretation is possible for the distinction between north-

western Wednesday (‘Woden’s day’) and southern Mittwoch (‘mid-week’),
although in this case the southern loan-translation of the Latin media
hebdomas seems more likely to reﬂect ecclesiastical attempts to provide
a vernacular Christian term. This loan-translation is similar to the
more geographically and chronologically restricted forms sunâbent and
aftermontag, discussed above, which may be the result of less successful
efforts in this direction. Loan-translating media hebdomas serves to
make very clear to the populace at large that the term is neutral, and
does not refer to a pagan god. The fact that Wednesday is most heavily
targeted could reﬂect interactions between pagans and Christians in
the southern German area, since the Lombards and Alamanni provide
the contexts for the earliest explicit accounts of worship of Wodan. The
Lombard ethnogenesis, as it appears in the Origo Gentis Langobardorum,
Fredegar’s Chronicon, and Paul the Deacon’s Historia Langobardorum,
seems to present Wodan as a deity with a special tribal signiﬁcance
for the Lombards.

Jonas of Bobbio’s Vita Sancti Columbani describes

Green, ‘Wochentagsnamen’, p. 226.

Origo Gentis Langobardorum, ed. G. Waitz, MGH Scriptores rerum Langobardicarum et
Italicarum, Saec. VI–IX (Hanover, 1878), pp. 2–6, at pp. 2–3. Fredegarii Chronica, III.65, ed.
B. Krusch, MGH SRM 2 (Hanover, 1888), pp. 18–193, at p. 110. Paul the Deacon, Historia
Langobardorum, I.8–9, ed. L. Bethmann and G. Waitz, MGH Scriptores rerum Langobardi-
carum, pp. 45–187, at pp. 52–3.

The theophoric week in the Germanic languages

399

Early Medieval Europe

  ()

Suevi (that is, Alamanni) worshipping Wodan, and the inscription on
a brooch found in an Alamannic row cemetery at Nordendorf also
suggests a cult of Wodan among pre-Christian Alamanni.

Perhaps,

therefore, early Christians in southern parts of the Germanic-speaking
world may have had a particular interest in providing a Christian name
for Wednesday. In England, by contrast, Woden was enthusiastically
adopted by the Christian English as a royal ancestor ﬁgure.

Given this

reuse of Woden, his appearance in the English names of the days of the
week need not have been problematic for the Christian English.

The existence of alternatives to the pagan names in some Germanic

dialects, suggests that levels of hostility to the planetary week, or to
particular days, differed in various parts of the Germanic-speaking
world. Yet other explanations are also possible. Gad Rausing has argued
that the broad geographical pattern of the names of the days of the
week in fact reﬂects three political centres of gravity of the Migration
Age.

He claims that the completely planetary week reﬂects contacts

between the North Sea area (and especially Scandinavia) and the late
Roman empire, while the Franks adopted the form of the week used by
Latin-speaking inhabitants of Gaul, and the Goths, Bavarians and
Alamanni adopted Greek-derived terms. The idea of a Gothic centre of
gravity that provided lexical items to the Bavarians and Alamanni is
plausible, but we should bear in mind that the Frankish empire also
exercised an inﬂuence in this direction. The Scandinavian and North
Sea centre of gravity is still more problematic: there is quite plausible
evidence that the Scandinavian forms of the days of the week were
secondary borrowings from one of the West Germanic languages, perhaps
Old English.

In many ways, the early medieval Christian Germanic

For the Vita Sancti Columbani, see n. 35, above. For the Nordendorf inscription, see Die
Runeninschriften im älteren Futhark, ed. W. Krause, Abhandlungen der Akademie der Wis-
senschaften in Göttingen, Philologisch-Historische Klasse, ser. 3, 65, 2 vols (Göttingen, 1966),
I, p. 294 (no. 151). See also K. Düwel, ‘Runen und Interpretatio Christiana: zur religionsges-
chichtlichen Stellung der Bügelﬁbel von Nordendorf I’, in N. Kamp and J. Wollasch (eds),
Tradition als historische Kraft (Berlin, 1982), pp. 78–86. Düwel’s claims that this is a Christian
attack on the gods named in the inscription need not demonstrate that there was no cult of
these gods in this area: quite the reverse. If Düwel is correct, this would suggest that there
were cults in the area, acting as foci for low-level Christian resistance of the sort Düwel
appears to envisage.

This may have something to do with Bede’s claim that Hengest and Horsa were the sons of
Woden. See Venerabilis Baedae: Historiam Ecclesiasticam Gentis Anglorum: Historiam Abba-
tum: Epistolam ad Ecgberctum una cum Historia Abbatum Auctore Anonymo, ed. C. Plummer,
2 vols (Oxford, 1896), I, pp. 31–2 (Book I, Ch. 15). The centrality of Woden in later royal
genealogies could be largely a product of the centrality of Bede’s Historia Ecclesiastica in
Anglo-Saxon views of their own origins and early history.

Rausing, ‘The Days of the Week’.

The clearest indicator is the Old Norse form frjádagr, which is most plausibly explained as a
borrowing of the Old English frigedæg, the Old High German frijatac, or a related form from
a more northerly West Germanic dialect. See Helm, Altgermanische Religionsgeschichte, II.2,
pp. 268–9. See also Anderson, ‘Scandinavian Days’, pp. 9–10, arguing against this view.

400

Philip Shaw

Early Medieval Europe

  ()

societies offer more plausible milieux for these complex processes of
cultural transfer – milieux in which the pagan classical past, and the
heathen Germanic past, remained potent cultural models, whose ideas
and terminology were both embraced and rejected in differing social,
political and religious contexts.

The ultimate origins of the Germanic planetary week must, as Fell

suggests, lie in interactions between Christians and pagans. At the same
time, however, we should not underestimate the impact of Christian
culture on the development of this system. In particular, the spread of
this system across Germanic-speaking Christian societies can plausibly
be attributed to the activities of Germanic Christians in conversion and
Christianization. The earliest evidence for equations of the great gods
with Graeco-Roman deities survives from the Anglo-Saxon, Colum-
banian and Frankish intellectual powerhouses that were instrumental in
the Christianization of much of north-western Europe. It would not be
particularly surprising to ﬁnd that a theophoric week, which originally
developed as part of Christian scholarly engagement with the classical
past, became so much part of Christian English and Frankish language
and culture that it spread to other groups through interactions with the
Franks and Anglo-Saxons. With some other groups this process may have
been aided by the existence of pre-Christian cults of some or all of the great
gods among those groups, but this would not be a prerequisite for such
transmission to occur. In the case of southerly areas, worship of Wodan
could, as we have seen, even have functioned as a check on the borrowing
of the term Wednesday, and we should also allow for the possibility of
variant forms existing side by side in some contexts. Christianization
could be responsible, therefore, for the spread of the supposedly pagan
day-names throughout most of the Germanic languages.

Anderson’s arguments for pre-Christian adoption of the day-names in Scandinavia are open
to some criticisms. He claims, for instance, that ‘if the same kind of interpretatio Scandinavica
that could turn wodnesdæg into ó9insdagr were applied to Old English sunnandæg or OSax
sunnondag, then the expected result might have been an Old Norse **sólsdagr’, rather than
the attested sunnudagr (‘Scandinavian Days’, p. 8). While this is true, it is also clear that
contacts between speakers of Old English and Old Norse produced different kinds of trans-
lation, including the sort of phoneme-by-phoneme translation which could have produced
sunnudagr even if the Old Norse word sunna (which Anderson discusses) had not existed (see
M. Townend, Language and History in Viking Age England: Linguistic Relations between Speak-
ers of Old Norse and Old English (Turnhout, 2002), pp. 43–68, for a discussion of evidence
for this phenomenon – for which Townend uses the term ‘switching-code’ – in the Danelaw).
A similar objection can be raised to Anderson’s similar argument regarding Old Norse
mánadagr (p. 9). Nevertheless, Anderson’s emphasis on the problems and uncertainties
surrounding these names is wise. We should emphasize that the claim being made here is not
that adoption of the day-names and conversion or Christianization went hand in hand, but
that the impact of Christian culture on contacts of all sorts between pagan and Christian areas
and communities provides a plausible context for the transmission of these names.

The theophoric week in the Germanic languages

401

Early Medieval Europe

  ()

This reassessment of the transfer of the theophoric day-names from

Latin into the early Germanic languages raises important questions.
Whether or not one accepts the proposed alternative model for this
transfer, it is clear that the traditional model is not without its difﬁculties.
Those who wish to retain the traditional late Roman dating must now
consider how to account for the epigraphic evidence discussed above.
This evidence suggests that Romano-Germanic religious interactions in
the late Roman period involved equations of Roman and Germanic
deities which are apparently incompatible with those employed in the
transfer of the day-names. The model for a later transfer that has been
proposed here may not be the only plausible alternative, and further
work in this area could offer new possibilities. At the same time, the
Germanic theophoric week offers a way of linking together the rather
disparate and patchy evidence for the great gods outside Scandinavia. A
reassessment of the date and circumstances of the origin of this week
therefore suggests the possibility that these deities may not be safely
evidenced across most of the continental Germanic area from the late
Roman period onwards. We might even question whether the sparse but
widely spread medieval evidence for these deities represents essentially
the fading echoes of pre-Christian religious life, or whether some or
much of it is better understood as part of the reuse and re-imagination
of the great gods by early medieval authors.

University of Shefﬁeld

Wyszukiwarka

Podobne podstrony:
Ogden T A new reading on the origins of object relations (2002)
Blacksmith The Origins Of Metallurgy Distinguishing Stone From Metal(1)
Evolution in Brownian space a model for the origin of the bacterial flagellum N J Mtzke
1 The origins of language
1 The Origins of Fiber Optic Communications
The Origins of Parliament
Theories of The Origin of the Moon
Ogden T A new reading on the origins of object relations (2002)
2004 Haramein The Origin of Spin
Ronit Meroz The Middle Eastern Origins of the Kaballah
The Origins of Old English Morphology
The Intellectual Origins Of Modern Catholic Social Teaching On Economics
1953 Note on the Origin of the Asteroids Kuiper
Britain and the Origin of the Vietnam War UK Policy in Indo China, 1943 50
Caananite Origins of the Kabbalah
Mark Jenkins Vampire Forensics Uncovering the Origins of an Enduring Legend
[Mises org]Menger,Carl On The Origins of Money
The Origins of Turkish

więcej podobnych podstron