Runes in the First Century

Bernard Mees

One of the lacunae in the classical testimonies from antiquity that form the basis
of what we know of the history of the ancient Germanic tribes is the lack of any
clear reference to the use by the Germani of runes. The first reference to the
runic script in a literary source is usually attributed to Venantius Fortunatus, a
poet of the Merovingian period. An Italian, Fortunatus had come to Gaul to seek
employment, and toward the end of the sixth century composed a mocking
description presumably in reference to the use of writing among his Germanic
lords (cf. Page 1999: 100-1):

barbara fraxineis pingatur rhuna tabellis

quodque papyrus agit, virgula plana valet.
Ashen writing-tablets might be embellished with a barbarian rune
And for what papyrus manages, a flat wand will do.
Venantius Fortunatus, Carmina 7, 18 (Ad Flavium), 19–20 = Leo 1881: 173.

Here Fortunatus jests to his friend Flavius that if he cannot manage to write to
him in some civilised manner, he could try runes in the barbarian style. Thus it
is hardly a reference from which we should expect to learn much of runic use in
Merovingian Gaul.
There is no testimony so explicit in its reference to Germanic uses of writing
from classical antiquity. Various Germanic figures we learn may indeed have
been able to write, yet it is usually assumed that these references, going back as
early as the first century A.D., refer to literacy in Latin letters. The references
are also restricted to figures among tribes such as the Marcomanni who were in
direct contact with the Roman world.

201
 The references to Germanic literacy themselves are known exclusively from
two sources: the Annals of Tacitus the great first-century Roman historian1 and
the Histories of Ammianus Marcellinus which describe the years of the late
Empire.
In the first passage from Tacitus, Maroboduus the Marcomannic king writes
a letter to Tiberius asking for help, and in the second it is recalled that Adgan-
destrius, a chief of the Chatti in the time of Arminius, had appealled in writing
to the Roman senate, this time in rather more sinister circumstances. Neither
passage from Tacitus mentions what form of writing is used, but it is difficult
to see how it could be any other than a Latin one, presumably Old Roman Cur-
sive, as both are requests to Rome, plainly a circumstance where Germanic lan-
guage or script would have been inappropriate. Of course it is possible that both
persons knew Latin and the Latin alphabet as it is clear that those who served
in Roman armies had some need of alphabetism and Germanic warriors from the
tribes near the borders of the Empire were already serving in the Roman army
by the first century; but it is equally likely that the letters were in fact penned by
notaries, perhaps slaves who may not even have been of Germanic extraction.

Maroboduo undique deserto non aliud subsidium quam misericordia Caesaris

fuit. transgressus Danuuium, qua Noricam prouinciam praefluit, scripsit Tiberio
non ut profugus aut supplex, sed ex memoria prioris fortunae: nam multis
nationibus clarissimum quondam regem ad se uocantibus Romanam amicitiam
praetulisse. // Maroboduus, completely deserted, was obliged to appeal to the
emperor’s mercy. Crossing the Danube – at the point where it borders on the
province of Noricum – he wrote to Tiberius. His tone was not that of a refugee or
petitioner, but reminiscent of his former greatness. When he had been a powerful
monarch, he said, and many nations had made approaches to him, he had pre-
ferred the friendship of Rome.
Reperio apud scriptores senatoresque eorundem temporum Adgandestrii princi-
pis Chattorum lectas in senatu litteras, quibis mortem Arminii promittesbat, si

1 Tacitus also mentions that “litterarum secreta viri pariter ac feminae ignorant” (“clandestine
[love-]letters are unknown to both men and women alike”) in the 19th book of his Germania
(Winterbottom & Ogilvie 1975: 47), but it is clear he is criticising a Roman practice he
disdains here, rather than commenting on a lack of the knowledge of writing among the
Germani (Wimmer 1887: 65–67 pace Grimm 1821: 30–32).

202 Bernard Mees

 patrandae neci uenenum mitteretur ... // I find from the writings of contemporary
senators that a letter was read in the Senate from a chieftain of the Chatti named
Adgandestrius, offering to kill Arminius if poison were sent for the job ...
Tac., Ann. II, 63,1; II, 88, 1 = Borzsák 1991: 59, 67; translation Grant 1971:
107, 118–19.

Ammianus recounts four similar occurrences in his rather later continuation of

Tacitus’ Histories (Amm. Marc. XXI, 3, 4–5; XXIX, 4, 7; XXXI, 12, 8–9;
XXXI, 15, 5), each of an exchange between a Roman and a Germanic leader. In
one passage Ammianus even specifically mentions a notary. He also mentions
a written message sent from a Germanic chief to his fellow barbarians, however,
although again clearly in a Romanised context:

Bitheridum uero et Hortarium, nationis eiusdem primates, item regere milites

iussit, e quibus Hortarius, proditus relatione Florenti Germaniae ducis contra
rem publicam quaedam ad Macrianum scripsisse barbarosque optimates ueritate
tormentis expressa conflagrauit flamma poenali. // Bitheridus, indeed, and
Hortarius (chiefs of the same nation) he [i.e. the emperor Valentinian] appointed
to commands in the army; but of these Hortarius was betrayed in a report of
Florentius, commander in Germany, of having written certain things to the
detriment of the state to Macrianus and the chiefs of the barbarians, and after the
truth was wrung from him by torture he was burned to death.
Amm. Marc. XXIX, 4, 7 = Seyfarth 1978: II, 114; translation Rolfe
1935–39: III, 244–45.

It seems then that literacy in the broader sense among tribes in direct contact
with the Empire was culturally, orthographically and even linguistically Latin.
It comes as no surprise, then, that even in the early centuries of Roman occu-
pation we have evidence for Germanic patrons ordering inscriptions in the
Roman style placed on altars in Roman Germany (CIL XIII.2: nos 7776–8860
& XIII.4: nos 11981–12086; cf. Ihm 1887: 105–200; Heichelheim 1930).
Mostly dedications to female divinities usually addressed as deae, matres or
matrones, the earliest of these monuments has been dated to the late first century
(Rüger 1987: 10, 12). A number of these inscriptions, most stemming from the
second-third centuries, show evidence for the influence of Germanic language

Runes in the First Century 203

in what are otherwise Latin dedications. Some show the intrusion of a Germanic
dative plural formation -IMS in the place of the usual Roman -IS or (vul-
gar/motions femininum) -(I)ABVS (Much 1887; Ihm 1887: 34–35). Others even
show what appear to be corrections, sometimes taking the form of ligatures or
reduced characters imposed over normal sized letters, as if the dedicator had
noted a mistake in the original and requested that the stonemason correct it
(Hübner 1895: lxvii). One such modification was the addition of a minature H
to a C in the name of a Germanic dedicator, no doubt in order to mark the
spirant enunciation of the relevant phone more clearly:

} A}TRONIS
M
MACHAMARIAF
ETAALLOA

Matronis. M(arcus) Chamari f(ilius) et Allo.

To the mother goddesses. Marcus son of
Chamarus, and Allo

Fig. 1: Germano-Latin dedication from Enzen (Rüger 1981:

no. 8).

Presumably, then, some of the Germanic-speaking dedicators read Latin letters

well enough to criticise inadequate spelling.
Given such evidence for Germani literate in Latin, it comes as no surprise
to witness a recent revival of the thesis most commonly associated with Ludvig
Wimmer, of a Latin origin for the runic script. There is good evidence for
literacy in Latin among the Germani already at the time of the first flowering of
a clearly runic tradition, one which since Haakon Shetelig (1924: 14) first dated
the Øvre Stabu spearhead to c. A.D. 180, has been accorded a late second cen-
tury date by most archaeologists.

204 Bernard Mees

Nevertheless, recent advances in our knowledge of the epigraphic record of
northern Europe shed new light on the question of the earliest Germanic tradi-
tion of literacy. In almost the reverse manner from what we might expect to find
from the testaments of classical authors, evidence of epigraphs of Germanic au-
thorship from the first century have appeared on a fibula of Germanic manu-
facture from Meldorf in the Dithmarschen (Düwel & Gebühr 1981) and on a
pottery sherd, likewise of Germanic manufacture, from Osterrönfeld in the
district of Rendsford (Dietz et al. 1996).

Fig. 2. The Meldorf fibula inscription (c.

AD 1–50). Author’s drawing.
Transliteration: i2þi"h
Fig. 3. The Osterrönfeld sherd inscription
(c. AD 51–100). Author’s drawing.
Transliteration: r-w

Both of these finds are from German Schleswig and though their contents are so
opaque as to be of little significance linguistically, they are both outstanding in
terms of the history of Germanic literacy. These finds are clear evidence that
some level of literacy had penetrated the north of Germania by the first century.
It also seems quite likely that even if neither of the finds are to be accepted as
runic, they are at least likely to represent the forebear of what was to become
runic in the next century. Indeed, although inscriptions on pottery are quite rare
from later times, fibulae are one of the most typical of media for runic inscrip-
tions during the period of the older runic tradition.
Yet these inscriptions are not the earliest examples of Germanic epigraphy
that have come down to us so far. The oldest Germanic inscription, clearly in
North Etruscan characters, was unearthed early in the nineteenth century, al-

Runes in the First Century 205

though it was not recognised as Germanic in language until 1925. This inscrip-
tion on an Etruscan type of helmet found in what seems to have been the re-
mains of a Celtic sanctuary on the border of Noricum and Pannonia probably
dates to the second or early first century B.C., although some have countenanced
a later date, as recent as the mid first century A.D. The inscription from Ženjak,
Negova (Negau) is quite unlike the northern inscriptions because it is of interest
linguistically, and can be translated. Although like a number of other epigraphs
from the area, it appears to have been influenced by Rhaetic grammar, it con-
tains the only clear example of a retained diphthongal pronunciation of IE *ei
in Germanic. It is also key evidence for a North Etruscan origin of the runes
(Marstrander 1925; Nedoma 1995; Markey 2001).
In the third chapter of his Germanic ethnography of A.D. 98, Tacitus
mentions reports of epigraphic remains in Greek letters about the border of
Rhaetia and Germania (i.e. on the German Danube) that until the 1970s could
not be matched by physical evidence. Even now we have no evidence of epi-
graphs in stone such as Tacitus describes, but the evidence on pottery sherds
from the pre-Roman settlement at Manching (Krämer 1982) prove some level
of literacy among the ethnically mixed population of pre-Roman Vindelicia.
One of the inscriptions, on a pottery sherd of local manufacture, is even part of
the alphabet-row, which by the shape of the theta would appear to be the Greek
(perhaps Gallo-Greek) one. Some of the other inscribed sherds are obviously
North Etruscan, but these may represent imports from the East Alpine Fritzens-
Sanzeno culture (whose epigraphic remains are Rhaetic in language). The only
lexically transparent inscription appears on a locally made ceramic and reads
BOIOS, a Celtic or Venetic anthroponym known from both Latin and Venetic
epigraphs (CIL III.2: no. 5417; III, supp. 1: no. 11563; Pellegrini & Proscodimi
1967: no. Es 28). The alphabet used on this occasion could equally be Latin or
North Etruscan, however, the latter option seeming increasingly credible since
the recent publication of a Celtic, though obviously orthographically North
Etruscan inscription from Ptuj which clearly uses B (Eichner et al. 1994). Thus
there is somewhat mixed evidence from the epigraphic record of central Europe
for what we might expect to represent the pre-Roman prototype upon which the
runes were based. Nevertheless, there are structural and orthographical features

206 Bernard Mees

of the Germanic script which clearly point to a North Etruscan source, and the
alphabetism at Manching appears likely to represent its immediate provenance.
In fact Wolfgang Krause may have been correct to stress the evidence of the
inscription on the Castaneda flagon given the many similarities between runic
and the Sondrio or Camunic tradition which it is usually thought to represent
(Krause 1940: 184–85; 1970: 37–39; Mees 2000).2

Fig. 4. An inscriptions from the oppidum of

Manching (c. 150–50 BC). Author’s drawing.
Transliteration: "BOIOS

Fig. 5. Abecedaria from Salitz della Zurla and Piancogno, Val Camonia (perhaps 4rd–3rd
centuries BC). Author’s rendition.

2 My thanks to Tom Markey for the references to the Ptuj inscription and the Camunic finds.

Runes in the First Century 207

Although there are no clear classical references to the runes when literacy is
described among the Germanic peoples by classical authors, for some time it has
been conjectured that Tacitus has recorded in his ethnography of the Germanic
tribes a description that may, nevertheless, apply to another sort of use of the
Germanic script.

auspicia sortesque ut qui maxime observant. sortium consuetudo simplex. virgam

frugiferae arbori decisam in surculos amputant eosque notis quibusdam discretos
super candidam vestam temere ac fortuito spargunt. mox, si publice consultatur,
sacerdos civitatis, sin privatim, ipse pater familiae, precatus deos caelumque
suspiciens, ter singulos tollit, sublatos secundum impressam ante notam interpre-
tatur. // For omens and the casting of lots they have the highest regard. Their
procedure in casting lots is always the same. They cut off a branch of a nut-
bearing tree and slice it into strips; these they mark with different signs and
throw them completely at random onto a white cloth. Then the priest of state, if
the consultation is a public one, or the father of the family if it is private, offers a
prayer to the gods, and looking up at the sky picks up three strips, one at a time,
and reads their meaning from the signs previously scored on them.
Tac., Germ. 10 = Winterbottom & Ogilvie 1975: 42; translation Mattingly
1970: 109.

Modern commentaries on the Germania differ in their interpretations of this

seminal passage. In one recent commentary Allan Lund (1991: 1893–94) relies
on the opinions of the linguist Elmar Seebold (1986: 554ff.) who clearly con-
siders that the reference is to runes. Lund is somewhat more conservative in his
approach, however, in his, the most recent properly historiographical-critical
edition of the Germania (Lund 1988: 140). Nineteenth-century classicists on the
whole tended somewhat to circumspection (e.g. Orellius 1845–48: II, 345, n.).
There were some such as Arthur Murphy (1807: VI, 216–15; and cf. now Rives
1999: 165–66) who attempted to link the passage with the employment of lots
recounted in later, mostly medieval sources, but not ones in which notis quibus-
dam ... impressam play a part. Wilhelm Grimm (1821: 296–320), on the other
hand, thought it likely that Tacitus’ notae were runes, an opinion that was to
come to prove influential among Germanists in the following years. Rochus von
Liliencron & Karl Müllenhoff (1852: 26ff.; cf. Müllenhoff 1900: 223–24,

208 Bernard Mees

226–27) more surely linked the Tacitean notae not merely with the runes, but
also to the question of their origin. Their seminal work launched the German
Urschrift theory, i.e. the notion that the runes grew out of indigenous Germanic
symbols such as the swastika. Despite some acceptance among German acade-
mics, this theory had mostly been relegated to the field of popularisers and
mystics by the turn of the century (Losch 1889; Wilser 1895; 1905; 1912;
Meyer 1896; von List 1907). Thus, in the last decade of the nineteenth century,
reflecting the caution of the Latinists, Eduard Sievers (1891: 239) had not felt
any more confident than to state that the possibility that the notae were runes
could not be ruled out. The following decades saw Germanists promote a more
ambitious assessment of the passage, however. The emerging Sinnbildforschung
of the time, both amateur and academic, and the Urschrift thesis which was so
central to it clearly depended on a link between the runes, symbols such as the
swastika, Tacitus’ notae and the notion of the indigeneity of the Germanic char-
acters. As this notion gained more and more converts amid the growing nation-
alist tide in the Germany of the 1910s and 20s, a new theoretical basis developed
in which the connection between notae and runes might be explained (cf. Mogk
1894; 1900: 400ff.; Petsch 1917; Feist 1919: 251–54; Marstrander 1928: 173;
Arntz 1935: 245–50). The common understanding among non-specialists that
‘rune’ was a valid designation for any mysterious marking (cf. Furneaux 1894:
57, n. 3; Gerber & Greef 1903: 971) led to the development of the theory that
the notae represented runes in the general, not specific sense. Krause (1936a;
1936b; 1938; 1939; 1943: 2–3; and cf. Arntz 1943–44), himself clearly influ-
enced by the Zeitgeist of growing Germanomania, spelt this separation out
clearly, attempting to pare this most extreme aspect of the by then burgeoning
Sinnbildforschung, i.e. the notion that the runes were uralt, stressing the distinc-
tion Lautrune : Begriffsrune (cf. Petsch 1917 who had similarly spoken of
Schriftrunen : Zeichenrunen) and restricting indigeneity and primeval antiquity
only to the latter. Similar sentiments are expressed in the scholarly Germania
commentaries of the time (Fehrle 1929: 80; Reeb 1930: 29; Much 1937: 131;
Anderson 1938: 79). Krause was certain the notae were at the very least
Begriffsrunen, however, in marked contrast to Helmut Arntz who treated all
Sinnbildforschung like a scourge. Nevertheless, the consensus developed among

Runes in the First Century 209

the other continental specialists by the end of the war was that although the
notae were clearly more than indiscriminate markings, Tacitus’ description was
too early for the notae to be runes. The thesis of a North Etruscan origin for the
runes accepted by Krause and Arntz made this quite illogical. Yet many of their
Germanist contemporaries, unaware of the latest developments in runology, still
stressed that Tacitus must have been describing the employment of the signs
from which the runes developed, the forebears of the Begriffsrunen – swastikas
and the like – not the runes (Lautrunen) proper. The underlying assumption
here, of course, was that if runes were as old as the Roman’s description then
these notae would have been runes; but they were not.
The description of the divination procedure in which these notae are men-
tioned may be drawn from the lost Bella Germaniae of the elder Pliny, an author
who, unlike Tacitus himself, is known to have visited Germania. It may even be
that Tacitus here has borrowed from Poseidonios’ lost Germanic ethnography
of the last century B.C. which was rather more clearly plagiarised by Caesar.
Nevertheless, though Eduard Norden (1922: 124, n. 2) notes that the opening
phrase is similar in style to that of Poseidonios, he admits there is no clear
evidence to link this passage to the work of the Greek historian. It is more than
likely, then, that the inscriptions from Meldorf and Osterrönfeld are either con-
temporary to or slightly earlier than the source used by Tacitus in the composi-
tion of his ethnography of the Germani. The inscriptions from Meldorf and
Osterrönfeld intensify the doubt that an origin for the runes in the years before
Roman influence becomes dominant in the material record of the North imme-
diately brings to the argument that the notae are too early to have been runes.
Indeed the notion that Tacitus refers here to pre-runic ideographs should proba-
bly be considered a relic only of Germanomanias past.

Without much justification the inscription on the Meldorf fibula has been
dubbed “proto-runic” by some runologists. Whether it is seen as Latin or runic
seems to depend more on whether the runologist in question prefers a Latin
origin for the runic script than on a strict epigraphic analysis. The inscription is
clearly at the very least preliminary or proto-runic because it appears on an item
that is characteristic of early runic finds, i.e. a Germanic fibula, and is inscribed

210 Bernard Mees

in a decorative fashion typical of later runic testaments. Latin inscriptions are
not found on such items or with such decoration (nor are Greek or North
Etruscan inscriptions for that matter). All accounts so far have also been in
agreement as to the Germanic language of the inscription (Mees 1997). The only
Latin interpretation suggested by the inscription is |D|N, i.e. D(ominus) N(oster),
a style assumed by some Roman emperors (cf. IK no. 282). Yet this description
was only first assumed by Hadrian, i.e. in the second century (Kneissl 1969:
95–96), and is not widely attested in epigraphs until the time of Septimius Seve-
rus (CIL XIII.5, p. 59). The Meldorf fibula is proof of literacy in free Germany
at about the time from which the description of Tacitus stems, and presumably
a linguistically Germanic literacy at that. Such letters may well have been
termed notae by the Roman. Consequently, the main reservation that scholars
have had in the past in accepting that the notae were runes (or more specifically
“Lautrunen”) is no longer warranted.
Yet if the notae were runes we might ask why unlike Fortunatus did Tacitus
not use a Latinised word of Germanic extraction. The use of a foreign term,
however, would be quite unusual in his Germania. Contrary to the habit of some
authors, Tacitus reproduced few barbarian words in his works. Instead, as an
ethnographic sketch for a general audience, in the Germania Tacitus endeavours
to make his prose as natural and unaffected as possible. His use of foreign terms
is limited to the Germanic expression framea ‘a Germanic type of spear’, and
possibly the palaeographically difficult bar(d)itus, if this term is not merely
Latin barritus ‘roar (like that of an elephant)’. As both are descriptions from the
martial sphere, Pliny may well be the source of these terms, as he is known to
have performed armed service on the German frontier.
Tacitus clearly abstains from the use of, or was unable to provide, native
terms for most of what he describes in his Germanic ethnography. It is quite
commonly assumed, for example, that the sacramentum mentioned in the fealty
ceremony of Germ. 13–14 is in fact an oath (a reciprocal concept perhaps not
able to be translated satisfactorily into pre-Christian Latin), and the descriptions
comitatus and comes seems to refer to the trustis and antrustiones of the later
Frankish Salic law (Much 1937: 163; Schlesinger 1953: 235; Wenskus 1961:
357–58; Green 1965: 78, 87–89, 126ff.). Thus it is probably expecting too much

Runes in the First Century 211

of the Roman to expect to find the word rune in his works. In fact the super-
fluous, pseudo-Greek <h> in Fortunatus’ rhuna probably indicates that this term
is a loan of the early medieval period. Nevertheless, notae almost seems an ideal
Latin translation given the idea of signification and even the mystery that the
term can also convey.
Arthur Mentz (1937) offered an updated version of the interpretation of von
Liliencron & Müllenhoff explaining that Tacitus (or his source) did not observe
the use of runes as phonological devices but instead as ideographs. This Be-
griffsrune argument of course fits almost too perfectly with the meaning of nota.
It clearly means ‘sign, indication, mark’ (see Glare 1982: s.v. nota for the full
range of classical uses). Tacitus is well known for his concise style and his
choice of term here is probably quite wilful. Thus in Tacitus’ day, rather than
some nondescript marking (pace the blithe dismissal of Askeberg (1944: 42–43)
who obviously saw himself as something of an iconoclast here), notae in this
passage must represent some sort of indicating character as the term was clearly
used to describe standard abbreviations (such as the notae vulgares which
abbreviated common Roman terms and first names), ciphers and ideographs
such as the consular sign or the mark of banishment. Roman school children
who could recite the notae vulgares of the ABC (such as M for Marcus etc.)
were even deemed notarii (notarians) (Bonner 1977: 168). Nonetheless, a more
sophisticated philological investigation is required before such an identification
might be regarded as proven.
One possible model for Tacitus’ description here is surely the poet Virgil in
whose Aeneid a sybil’s divination is described using this very term. Virgil says
of the sybil: “fata canit foliisque notas et nomina mandat” (“She foretells desti-
ny and entrusts signs and names to leaves”). Although it is not immediately
clear what Virgil means by notae here, a later author, Servius in his fourth-
century commentary on Virgil, explains what this passage meant to him. Not
only does Servius gloss notas by litteras ‘letter’, but he makes explicit how the
notae are being used in the sybil’s divination:

tribus modis futura praedicat: aut voce, aut scriptura aut signis // there are three
ways to predict the future: either summoning, writing or by portents

212 Bernard Mees

 id est quibusdam notis, ut in obelisco Romano videmus, velut alii dicunt notis
litterarum, ut per unam litteram significet aliquid. // that is those signs such as
we see in Roman epigraphs, which some call writing signs, as something can be
signified by one letter.
Serv. ad Aen. III, 444 = Stocker & Travis 1965: 170.

Servius evidently reached a similar conclusion to that of Müllenhoff and Mentz.

A more contemporary author has left us with another similar description, an
author whom Tacitus is known on occasion to have actually aped stylistically.
This is how Cicero describes the method of the legend of the Praenestine oracle
in his De Divinatione (On Divination):

itaque perfracto saxo sortis erupisse in robore insculptas priscarum litterarum

notis. // And so when he had broken open the stone, the lots sprang forth carved
on oak, in ancient characters.
Cic., De Div. II, 85 = Giomini 1975: 118; translation Falconer 1971: 466.

Now Tacitus should have been aware of this passage, and it does not seem likely
that he has plagiarised Cicero here. In the Latin of Cicero’s day litterarae notae
clearly meant ‘written characters’ (Glare 1982: s.v. nota 6b); hence the use of
notae seems to be established as the typical way to describe alphabetical char-
acters of any sort employed in such a divination. But of course the question must
arise as to whether the similarity of the Ciceronian passage to that in the Ger-
mania is more than merely similitude, or whether it does not in fact represent an
ethnographical τόπος (commonplace). Many ancient ethnographical descriptions
of barbarian peoples feature τόποι taken from a common stock of barbarian
practices. Often these descriptions can even be traced back to the influential
Scythian ethnography of Herodotus. Nevertheless, Tacitus has usually proved
a reliable source whenever his claims can be checked (Jankuhn 1966). And
though two descriptions of divinations with cuttings from trees do appear in
Herodotus’ Scythian ethnography, there is no mention of symbols or scrat-
chings, so it does not appear likely that we are dealing with some sort of plagia-
ristic ethnographic Wandermotiv. (The Herodotean description may have in-
spired Tacitus to mention Germanic forms of divination, however.) Instead, in

Runes in the First Century 213

the Ciceronian passage we again discover the use of notae carved in pieces of
wood. Clearly then, nota is the expected Roman term in such a description of
this period. Here, then, as Servius indicates in his commentary on Virgil, nota
unambiguously describes alphabetical characters. Hence the notae of Tacitus
appear likely to have been runes. Indeed, the Germanic letters probably have
their origin in an archaic alphabet such as Cicero describes.
Tacitus and Cicero are clearly both describing sortes, a divination procedure
well known in classical experience (hence sortilege), and one which some
authors trace ultimately to a Greek origin. It is more than merely a procedure of
lot-drawing, and is comparable as a category to auspices, haruspices and other
forms of fortune-telling known in ancient times (Becker 1856: 103ff.; Bouché-
Leclercq 1879–82: I, 195 and IV 145–59; Ehrenberg 1927). It is clearly a tradi-
tion to be associated with alphabetism, and presumably arose first among the
lettered. It probably represents part of the well-attested tendency of ancient and
medieval cultures to associate all manner of religious and magical beliefs to
alphabetism.
Tom Markey (1999: 142–43) has recently shown that nota appears loaned
into Old English in the expression wæl(l)not in Solomon and Saturn (158),
which seems to be a calque on wælr)un (Poem of Elene 28 = ON valrúnar). If the
valrúnar were in fact runes, then the association of rune and nota seems to have
maintained a remarkable longevity, especially given the range of meanings over
which nota could be applied by the Middle Ages. Yet despite the usage of nota
in other comparable passages, it is not impossible that Tacitus’ notae were
another form of marking, much as applies to the valrúnar. Of course Tacitus’
use of notae has been compared to other markings, for example those found on
a collection of rectangular wooden tokens at Kitzbühel (Pittioni 1942). These
markings date to the Hallstatt period, however, and although they have also been
linked to the Mediterranean sortes, not only are they clearly not Germanic, both
their meaning and employment remains obscure. Nevertheless, the use of the
term rúnar in Old Norse for what might not have been Lautrunen at all may
point to the earlier employment of PG *r)unoz to represent other ‘things that
speak without talking’ (cf. Sievers 1891: 239–40). In fact Tacitus may even be
describing some sort of alphabetism or semi-alphabetism belonging to a tradi-

214 Bernard Mees

tion other than that of the Lautrunen. More evidence is needed to move beyond
the likelihood that letters or some sort of substitute are being used in the divi-
nation described by Tacitus.

There is, however, some evidence of indigenous origin that seems to justify an
equation of notae with Lautrunen. There are two lemmas, both of which have
only recently received renewed attention, that might support an identification of
the notae with runes.
First, according to Tacitus, the notae were inscribed on slips or slivers of
wood cut from fruit- (or nut-) bearing trees. Many have suggested the beech
owing to the usual etymology accepted for the Germanic term book (thus, pre-
sumably, the common translation of frugiferae as ‘nut-bearing’ rather than
‘fruit-bearing’). It has long been recognised that this etymology is problematic
phonologically, however, as book is clearly a (decaying) root-consonant stem
in Old English, Old Saxon and thus presumably Proto-Germanic. The root-con-
sonant paradigm is unproductive in Germanic and clearly more archaic than
vocalic-stem formations such as beech, though ‘book’ can hardly have been the
original meaning for PG *b)oks. Instead Sievers (1891: 241–42; cf. Murray et al.
1888–1933: s.v. book) and more recently Seebold (1981: 289–92; in Kluge
1995: s.v. Buch; and cf. Ebbinghaus 1982; 1991; Peeters 1982) recognising book
was originally an athematic stem have proffered a connection with Skt. bh)agá-
‘lot, fate, share’, Av. b)aga- ‘share’ (Mayrhofer 1956–80: s.v. bhága"h, bhájati;
1986ff.: s.v. bhága-, BHAJ) which (unlike beech) is also attested occasionally
as a root-consonant stem, specifically citing the Tacitean passage: i.e. the surculi
when inscribed with notis were called books; ‘lot’ > ‘lot with rune carved on it’
> ‘book’. Sievers later abandoned this etymology in favour of the linkage of
book with beech, swayed by Friedrich Kluge’s observation that a similar
semantic development had occurred with a Sanskrit term for ‘birch’ (Kluge
1883: s.v. Buch; 1890; Sievers 1901: 252), and similar observations from Greek
and Latin practice were adduced later by Johannes Hoops (1911–13: 338–39).
Ernst Leumann (1930: 190) pointed out that Sanskrit seemed to show that a
reconstruction *bh)ag- ‘lot etc.’ is probably to be associated with a lengthened-
grade form of Skt. bhága"h ‘prosperity, happiness, possession, fortune’ (see-

Runes in the First Century 215

mingly a development of the verb bhájati ‘divides, distributes, receives, enjoys’)
and he connected *bh)ag- ‘lot etc.’ to beech as reflecting a semantic evolution
of ‘*Losbaum’ > ‘beech’. By this time, however, Hermann Osthoff (1905) had
already put forward evidence to suggest there was a long diphthong (a length-
ened-grade form with a laryngeal) in the root of the attested IE cognates of
beech (i.e. PG *b)ok)o-, Lat. f)agus ‘beech’, Alb. bung ‘oak’, Gk φηγόl ‘Valonia
oak’ < *bh)a(u)ǵ-; Icel. beyki ‘beech forest’, Russ. dial. buziná ‘elder’ < *bhEuǵ-
; early NHG bûchen, biuchen, ME bouken ‘buck, wash in (hot) lye’, Ukr. byźe
‘lilac’, Kurdish b)uz ‘elm’, Mysian/Lydian μυσόl ‘beech’ < *bh)uǵ-; Russ. boz,
Bulg. bŭz, Cz., Pol. bez ‘elder’ < *bhuǵ-). The Slavic and Kurdish cognates
show clear evidence for u-vocalism in IE beech, and despite the negative assess-
ment of George Lane (1967), the variations first examined by Osthoff might be
explained by laryngeal metathesis: i.e. *bh)a(u)ǵ- < *bheH2uǵ-, *bhEuǵ- <
*bhH2uǵ- and *bh)uǵ- < *bhuH2ǵ-, coupled with a shortening of *bh)uǵ- >
*bhuǵ- in most Slavic dialects (Krogmann 1955–56; Eilers & Mayrhofer 1962;
Friedrich 1970: 106–15). Noting the ablaut first delineated by Osthoff, Wilhelm
Wissmann (1952: 15ff.; and in Marzell 1948–79: s.v. Fagus silvatica; cf. Neu-
mann & Beck 1981) has thus proposed that IE beech was originally a root-
consonant stem and the form that gave us book is merely morphologically con-
servative. The attested forms of beech in Germanic, Latin and Greek would then
represent an original genitive form that developed into a feminine o-stem
*bh)aǵós (and Albanian bung a singulative *bh)aǵ-n-os). But this supposition is
based on an a priori acceptance that Germanic book derives from IE beech. The
full-grade form *bhéH2uǵs (. PG *b)oks) would clearly be the nominative in the
required hysterokinetic paradigm, yet the genitive would be *bhH2uǵós (i.e.
*bhEuǵós or *bh)uǵós) which of course generates the wrong dialectal forms.
None of the numerous cognates for beech shows the root-consonant forms that
would support this nominal ablaut-based thesis, and the feminine o-stems that
are attested are not reconcilable with the root-consonant form assumed to have
produced book unless we assume some irregular (i.e. ad hoc) assimilation has
taken place somewhere in the development of the beech word. In fact the
proposed Germanic assignation of a novel semantic value to what is, pace Kluge
(1890: 212), clearly also a morphologically archaic term seems even more im-

216 Bernard Mees

probable if, rather than to beechen notae, the origin of the term book is to be
connected with the use of beechen writing tablets in the Roman style as was
argued by Hoops (1911–13: 339; cf. Rosenfeld 1952; 1969; Green 1998: 258ff.).
On the other hand, Stefan Zimmer has rejected Seebold’s etymology as the
attested meaning ‘lot, fate’ may not be any more than an Indic development of
an earlier Indo-European root *bhag- ‘share’. If Zimmer is correct when he
dismisses the evidence of Tocharian A p)ak, B p)ake ‘part’ as loans from Middle
Iranian, the lengthened-grade forms of the Indo-Iranian terms may not be of
Indo-European antiquity either. Nevertheless, his explanation for the develop-
ment of the Slavic cognates represented by OCS bogŭ ‘god’ (contrasting with
OCS bogatŭ ‘rich’, ubogŭ ‘poor’, nebogŭ ‘pitiful’) as influenced by an early
Iranian form such as OPer. baga- ‘god’ < ‘share’ is rather ad hoc; the Slavic
evidence might be better explained by accepting an Indo-European development
of a secondary religious meaning of ‘lot, fate’ (preserved both in the Indic full-
grade and the typically more conservative lengthened-grade forms too) from the
root *bhag- ‘share’ which further developed to a connotation of deity in Iranian
and Slavic independently (Zimmer 1984: 196; 207–8; 210, n. 13; 215, n. 58).
This interpretation would also support a Germanic development of *bhag-
‘share’ to book, one which Zimmer dismisses perhaps too readily. Although
both derivations remain somewhat problematic, there are two possible IE models
for Germanic book; and if the relationship of book to the Indic terms for ‘lot,
fate, share’ can be maintained, this would probably underscore the identification
suggested here between Tacitus’ notae and the runes.
The linkage of the beech etymology for book with the Tacitean passage has
been used by some commentators as evidence for a broader, cultic connection
of writing with the beech tree (Gamkrelidze & Ivanov 1984: 623–24 = 1995: I,
533–35). Nevertheless, Tacitus’ description would tend to favour other trees
such as the ash, the rowan, or as in the passage from Cicero, the oak; and we
might have expected the historian to have used the term fagus if there were a
specific connection between the beech and the notae. Indeed, this description is
not at all at odds with the ashen tablets or wands described by Fortunatus some
five centuries later. A special Germanic connection between writing and the
beech, then, seems to be a rather flimsy philological construction. Tacitus’ men-

Runes in the First Century 217

tion of a white cloth and a state priest clearly indicates a developed religious
service: a divination known to the Mediterraneans as sortes, divination with
letters. Nevertheless, there does not seem to be any specific connection in this
tradition with regard to the type of surface used to record the alphabetical char-
acters. The drawing of lots is also ascribed to the Cimbri and the Suebi (Plu-
tarch, Marius 15, 4; Caesar, B.G. 1, 50), but no mention is made of how these
were selected, and so it is unclear where these ceremonies stand in relation to
that recounted by Tacitus.3 It is clear that Tacitus is not merely describing the
drawing of lots with straws or small sticks as occurs in some later Scandinavian
and Anglo-Saxon sources. Instead he seems to recount sortes, a divination
whose fundamental descriptive feature was the employment of written charac-
ters.

The second lemma is the pairing system that underlies the order of the rune-row.
First delineated by Erik Brate (1920; cf. von der Leyen 1930; Seebold 1986:
541–48; 1993: 415–21; Mees 1999: 146–47; 2000: 70–79), it is clearly based
in sequences that are known to stem from ancient letter learning. Pupils in the
ancient world were not only drilled in memorising the alphabet in its usual order
but also in variant orders based around pairs (Bonner 1977: 165–68). There are
a number of Greek and many more Roman examples of one of these systems,
the so-called boustrophedon-paired abecedaria. Other pairings based on other
principles are also known, albeit in somewhat lesser numbers; one such poorly
attested arrangement known only from two ancient testaments using the Greek
alphabet is based around first separating the alphabet into six parts instead of
two.

3 Hans Kuhn (1938: 62–63; cf. Arntz 1944: 293; Green 1998: 264–65) also proposed a
connection to Tacitus’ ceremony might be found in OHG lesan ‘to gather, to read’: this
development is peculiar to OHG, however, and might represent an influence of the seman-
tically similar Latin legere ‘to gather, to read’.

218 Bernard Mees

 α β γ δ ε ζ η h ι i λ μ ν ξ ο π
ω ψ φ χ υ τ σ k π ο + ν μ λ i ι h η ζ ε δ γ
α ω β ψ γ χ δ n ε υ ζ τ η σ h k ι π i ο λ ξ μ ν

Fig. 6. Greek boustrophedon sequence from Egypt (Boak 1921: 189–94).

AXBVCTDSERFQGPHOIN

Fig. 7. Pompeian boustrophedon inscription (CIL IV,

5472 et al.)

[α] ε Ι Ν Ρ Φ –
P αg ιν kn βζ
Β Ζ Κ Ξ C Χ iξ σχ γη λο
Γ Η Λ Ο Τ Ψ τψ δh μπ υω
Δ Θ Μ Π Υ ω
–
P αω βψ γχ δn
Fig. 8. Inscription from Sparta gυ ζτ ησ hk
featuring sextile pairs –
P
(IG V, 365).
ιn iο λξ μν

Fig. 9. Both sextile and boustrophedon

pairs in a Greek schoolbook (P. Vindob.
G 29274, pp. 13–14 = Van Haelst 1979:
no. 136.).

Runes in the First Century 219

Such pairings are also known to have readily taken on religious or mystical
meanings. They are mentioned in Jewish cryptography and alphabet mysticism,
in Gnostic belief and ancient astrology (Irenaeus I, 14, 3; Talmud, Shab. 104a;
Dornseiff 1925: 17–20, 126–33, 137–38). From the North Etruscan sphere there
is even a bilingual bronze inscribed in Latin and Venetic (i.e. orthographically
North Etruscan) letters: two such paired sequences are found where the accom-
panying dedications clearly show that the sequences are meant as a dedication
to the Venetic goddess Reitia by a certain Voltionmnos.
The clearly legible section of the scrambled Latin abecedarium,

---]RF[q]$GPHOINKM[---

consists of pairs grouped from the centre of the Latin abecedarium: F, G, H, I

and K are paired with Q, P, O, N and M. The method of this ordering is part of
a boustrophedon-pair arrangement of the Latin abecedarium. The accompanying
Venetic sequence is also a set of paired letters:

---]NRNPRΦRŚLŚNTR[--- | ---]KNMNMLSRSLΦLXRX[---

They form a section of an ordering commonly found on such tablets at Este

(though here in a somewhat irregular arrangement) that has been shown by
Michel Lejeune to derive from a patterned arrangement of the cultic Venetic
consonantal abecedarium, the VZA. For these two sequences the accompanying
dedications make quite clear the context of such inscriptions as are found at
Este. The Latin dedication is

---]O[---] DEDIT LIBENS MERITO

‘... given willingly and deservedly’, a common Latin votive formula. The Vene-
tic inscription, however, is even more specific:

[vza.]N[.] VO.L.T[.iio.n.]MNO.S. [zo]NA.S.TO KE

LA.X[.sto ša.i.]NATE.I. RE.I.TIIA.I[.] O.P. [vo.]L.TIIO [l]E[no]

220 Bernard Mees

‘This abecedarium Voltionmnos offers and dedicates to Šainate Reitia, willingly
and deservedly’ (Lejeune 1953: 75–78, 1970: 190–97; Pellegrini & Prosdocimi
1967: Es 27; Fogolari & Prosdocimi 1988: 271–74, Marinetti 1990: 112–18).
Many of the other North Etruscan epigraphs from Este appear to be offerings
made with the hope of receiving succor from the (local) goddess Reitia. Thus it
seems that lucky or divine letter sequences were known to the peoples who
employed scripts related to that which found its way north in the last centuries
B.C. to become the Germanic runes. The order of the rune-row clearly stems
from such an arrangement, and this pairing was first noticed in an investigation
of the names of the futhark. Yet as the pairing may have been determined by an
orthographic superstition, and the use of notae in divination is attested by Taci-
tus, it seems that this confluence of alphabetic superstition is more than acciden-
tal. Similar pairs found on Greek pottery sherds have been interpreted as the
product of divinations and the biblical ’ūrūm and tunmīm used for divination in
the Old Testament may well have been alphabetical pairs of this type (Hieron.,
In Jerem. 25, 26; idem, Epist. ad Laetam 107, 4; Heinevetter 1911: 33ff.; Ro-
bertson 1964). Given what we know of the Mediterranean tradition of sortes, it
is hard to see how Tacitus can be speaking of anything other than runes, espe-
cially if Germanic book is related to Indo-Iranian terms for ‘fate’. We may pre-
fer to disdain magical interpretations on the a priori grounds that too much in
past runology has been ascribed to magic, and indeed the more macabre dal-
liances of the racist runic mysticists of the early twentieth century might again
give us pause here. But if the pairing of the runes derives from a mundane alpha-
betic exercise gone somewhat awry, the rune names with their pairs of oppo-
sitions and complements surely suggest a cultic or cosmological context, one if
not ultimately derived from divinatory practice, then one that could easily be put
to that purpose.
The continuation of the tradition of the grouping of staves into three ættir
(first attested in the Vadstena rune-row) into the medieval manuscripts and as
the fundamental principle in the formation of cryptic runes is probably the last
remembrance of the mechanical process employed in this binary system. The
selection of staves to be omitted in the reduction of the older to the younger
futhark also appears to have been governed by the principle of pairing. Yet the

Runes in the First Century 221

transference of the elder R to the third ætt as if to address the subsequent im-
balance caused in the relative sizes of the ættir seems to indicate that the
principle that first governed this pairing must have only been a dim memory by
the eighth century. Some commentators have supposed that the wide range of
meanings of the rune names well represent the world experienced by the ancient
Germani (von der Leyen 1938: 98; Polomé 1991), perhaps here notably in
contrast to letter names in traditions such as Ogham. The pairs also often
represent oppositions and complements that could easily be employed in a
divinatory system. The system of merism and metaphor exhibited by many of
the pairs may have been the only stylistic expression available in the construc-
tion of the earliest of the runic poems given that alliteration would have been a
hindrance to a poem based around acrophonic names. Yet the resulting system
of paired names would have readily been employed in any magical or mystical
practice that became associated with the Germanic letters. The alliterative
naudir nije and þoss ðrettan (Flowers 1986: 266) of later rune magic clearly
developed in such a manner. Evidently, the paedagogical principle was primary
and the less mundane associations merely accretions to the developing Germanic
alphabetical tradition.

Tacitus clearly mentions that three signs were drawn to be interpreted and many
have suggested that this number is reflected in the three ættir which has led to
an overly ambitious attempt to read semantic spheres into the ættir as well. Their
origin remains obscure, but is perhaps not implausibly to be linked with the
stanzaic features of the original runic poems: two names to a line, four lines to
a stanza. Nonetheless, it is not surprising given the many points of comparison
between runic and Irish Ogham practice to discover four selections employed
in an Irish divination using Oghams, as the Irish orthography was separated into
four aicmi (a term which also signifies ‘family’).
In his exposition of the North Etruscan thesis for the origin of the runes, Carl
Marstrander indicated a number of similarities between runic and the Irish
Ogham letters. These parallels he explained as evidence for a Celtic North Etrus-
can source for the runes. Indeed his contemporary, the Ogham epigraphist
Stewart Macalister, also had sought the origin of the Irish orthography in a

222 Bernard Mees

North Italic alphabet, independently from Marstrander (Marstrander 1928:
130ff.; Macalister 1937: 18–29, 1945: iv-ix). Thus another strikingly similar
passage from an early Irish source, the recension of Tochmarc Étaíne (The
Wooing of Étaín) in the Lebor na hUidri (Book of the Dun Cow), may not
merely derive from coincidence:

Ba tromm immorro laisin druid dicheilt Étainiu fair fri se bliadna, co ndernai
iarsin .iiii. flescca ibuir, ocus scripuidh oghumm inntib 7 foillsigthir dó triana
eochraib écsi 7 triana oghumm Etain do bith i Sith Breg Leith iarna breth do
Midir inn. // Now it seemed grievous to the wizard that Étaín should be hidden
from him for six years, so then he made four rods of yew, and he writes an
ogham thereon; and by his keys of knowledge, and by his ogham, it is revealed to
him that Étaín is in the Fairy Mound of Breg Leith, having been carried into it by
Midir.
Tochmarc Étaíne 18 = Windisch 1880: 129; translation Stokes 1891:
440–41, n. 12.

Joseph Loth (1895; cf. Grimm 1821: 307–10; Seebold 1981: 291) has shown
that a number of Insular Celtic terms for ‘fate’ derive from a meaning of ‘casting
wood’, and similar references in Irish literature including terms for sorcery such
as fidlanna (literally ‘wood-planes’) have been assembled by Joseph Vendryes
(1948: 107–11). In fact the Germanic vocabulary similarly gives us Goth. tains,
OHG zein, OE tan, ON teinn ‘lot, slip of wood’ and, as we have seen, arguably
book as well. It is perhaps not surprising then to find here in a medieval Irish
divination that Ogham letters appear in the place where in Cicero and Tacitus
we have notae. In all three passages, characters seem to be inscribed on wooden
staves; and in all three passages the characters on the staves are read for their
divinatory meaning. None describe words, merely characters. There is, it seems,
a common Western Indo-European connection between sticks or slivers or slips
and writing. The use of mysterious or archaic letters in sortes recorded in
Cicero, Tacitus and the Irish Wooing may not be directly linked, but it does
seem likely that they are speaking of alphabetic characters employed in a similar
manner.

Runes in the First Century 223

 We know other beliefs were grafted on to the Germanic alphabetic tradition:
from the healing lore of the rúnatál stanzas of the Eddic Sigrdrífumál in which
Markey finds an ultimately Indo-European model, to the alliterating nine needs
of late sources such as the early modern Galdrabók; witness even the modern
mystical tradition established by Guido (von) List. We should not, then, be sur-
prised to find the Mediterranean custom of sortes reflected in early Germanic
practice. Obviously, the names of the runes were remembered as something
much more than mere mnemonic devices. Although they are treated as curious
toys in the medieval monastic tradition, the grouping in the rune-row inscrip-
tions and the ideographic use of runes point to a former more serious recognition
of the superstitious uses of letters. In fact the Roman account that describes the
first employment of ideographic signs by the Germani appears to make quite
clear how important they were judged in pre-Christian times: in the tenth chapter
of his ethnography of the Germanic tribes, Tacitus seems to describe the inscrip-
tion of runes on specially created slips or slivers of wood which were randomly
interpreted for their ideographic meanings in order to predict future events.

Bibliography

Anderson, John George Clark, ed. 1938. Cornelius Tacitus, De origine et situ Germanorum.
Oxford: Clarendon.
Arntz, Helmut. 1935. Handbuch der Runenkunde. Sammlung kurzer Grammatiker germanischer
Dialekte, B: Erganzungsreihe 3. Halle a. S.: Niemeyer.
— 1943–44. Review of Krause 1943. Beiblatt zur Anglia 54/55: 244–47.
— 1944. Handbuch der Runenkunde. Sammlung kurzer Grammatiker germanischer Dialekte,
B: Erganzungsreihe 3. 2nd ed. Halle a. S.: Niemeyer.
Askeberg, Fritz. 1944. Norden och kontinenten i gammal tid: Studier i forngermansk kultur-
historia. Uppsala: Almqvist & Wiksell.
Becker, Wilhelm Adolph. 1856. Handbuch der römischen Alterthümer. IV. Leipzig: Weidmann.
Boak, Arthur E. R. 1921. “Greek and Coptic School Tablets at the University of Michigan.”
Classical Philology 16: 189–94.
Bonner, Stanley F. 1977. Education in Ancient Rome: From the Elder Cato to the Younger Pliny.
London: Methuen.
Borzsák, Stephanus (István), ed. 1991. Cornelius Tacitus, Libri qui supersunt I, 1: Ab excessu
divi Augusti libri I–VI. Bibliotheca scriptorum Graecorum et Romanorum Teubneriana.
Stutgardia: Teubnerus.

224 Bernard Mees

Bouché-Leclercq, Auguste. 1879–82. Histoire de la divination dans l’antiquité. 4 vols. Paris:
Leroux.
Brate, Erik. 1920. “Runradens ordningsföljd.” Arkiv för nordisk filologi 36: 193–207.
CIL = Corpus inscriptionum Latinarum. Eds. Theodor(us) Mommsen et al. / Academia
Litterarum Regiae Borussica (et seqq.). 17 vols. Berolinum: Reimerus (et seq.), 1863ff.
Dietz, Martina et al. 1996. “Eine frühkaiserzeitliche Scherbe mit Schriftzeichen aus Osterrönfeld,
Kr. Rendsburg-Eckenförde.” Archäologisches Korrespondenzblatt 26: 179–88.
Dornseiff, Franz. 1925. Das Alphabet in Mystik und Magie. Gτοιχgια 7. 2nd ed. Leipzig:
Teubner.
Düwel, Klaus & Gebühr, Michael. 1981. “Die Fibel von Meldorf und die Anfänge der Runen-
schrift.” Zeitschrift für deutsches Altertum und deutsche Literatur 110: 159–75.
Ebbinghaus, Ernst A. 1982. “The Book and the Beech Tree.” General Linguistics 22: 99–103.
— 1991. “Further Thoughts on Gothic boka.” American Journal of Germanic Linguistics and
Literatures 3: 51–56.
Ehrenberg, Viktor. 1927. “Losung.” In Paulys Real-Encyclopädie der classischen Altertumswis-
senschaft. XIII.II. New ed. begun by Georg Wissowa. Stuttgart: Metzler, pp. 1451–1504.
Eichner, Heiner et al. 1994. “Ein römerzeitliches Keramikgefäß aus Ptuj (Pettau, Poetovio) in
Slowenien mit Inschrift in unbekanntem Alphabet und epichorischer (vermutlich keltischer)
Sprache.” Arheološki vestnik 45: 131–42.
Eilers, Wilhelm & Mayrhofer, Manfred. 1962. “Kurdisch b)uz und die indogermanische
„Buchen“-Sippe.” Mitteilungen der Anthropologischen Gesellschaft in Wien 92: 61–92.
Falconer, William A., trans. 1971. Cicero in Twenty-Eight Volumes, XX: De Senetute, De Amici-
tia, De Divinatione. Loeb Classical Library. London: Heinemann.
Feist, Sigmund. 1919. “Runen und Zauberwesen im germanischen Altertum.” Arkiv för nordisk
filologi 35: 243–87.
Fehrle, Eugen, ed. 1929. Publius Cornelius Tacitus, Germania. München: Lehmann.
Flowers, Stephen E. 1986. Runes and Magic: Magical Formulaic Elements in the the Older
Runic Tradition. American University Studies, Series I: Germanic Languages and Literature
53. New York: Frank.
Fogolari, Giulia & Prosdocimi, Aldo L. 1988. I Veneti antichi: Lingua e cultura. Il mito e la
storia: Serie maggiore 2. Padova: Programma.
Friedrich, Paul. 1970. Proto-Indo-European Trees: The Arboreal System of a Prehistoric People.
Chicago: University of Chicago Press.
Furneaux, Henry, ed. 1894. Cornelius Tacitus, De Germania. The Clarendon Press Series. Ox-
ford: Clarendon.
Gamkrelidze, Thomas V. (T’amaz Gamqrelize) & Ivanov, Viačeslav V. 1984. Indoevropeı̆skiı̆
iazyk i indoevropeı̆tsy: Rekonstruktsiia i istoriko-tipologičeskiı̆ analiz praiazyka i proto-
kul’tury. 2 vols. Tbilisi: Izdatel’stvo Tbilisskogo Universiteta.
— 1995. Indo-European and the Indo-Europeans: A Reconstruction and Historical Analysis
of a Proto-Language and a Proto-Culture. Trans. Johanna Nichols. Trends in Linguistics:
Studies and Monographs 80. 2 vols. Berlin: De Gruyter.

Runes in the First Century 225

Gerber, Arnold & Greef, Adolf, eds. 1903. Lexicon Taciteum. Lipsia: Teubnerus.
Giomini, Remus, ed. 1975. M. Tullus Cicero, Scripta quae manserunt omnia, Fasc. 46: De divi-
natione, de fato, Timaeus. Bibliotheca scriptorum Graecorum et Romanorum Teubneriana.
Lipsia: Teubnerus.
Glare, P. G. W., ed. 1982. Oxford Latin Dictionary. Oxford: Clarendon.
Grant, Michael, trans. 1971. Tacitus, The Annals of Imperial Rome. The Penguin Classics. Har-
mondsworth: Penguin.
Green, Dennis Howard. 1965. The Carolingian Lord; Semantic Studies on Four High German
Words: Balder, Frô, Truhtin, Hêrro. Cambridge: Cambridge University Press.
— 1998. Language and History in the Early Germanic World. Cambridge: Cambridge Univer-
sity Press.
Grimm, Wilhelm C. 1821. Ueber deutsche Runen. Göttingen: Dieterich.
Haelst, Joseph van, ed. 1976. Catalogue des papyrus littéraires juifs et chrétiens. Série papyro-
logie 1. Paris: Publications de la Sorbonne.
Heichelheim, Friedrich. 1930. “Matres.” In Paulys Real-Encyclopädie der classischen Altertums-
wissenschaft. XIV.II. New ed. begun by Georg Wissowa. Stuttgart: Metzler, pp. 2213–50.
Heinevetter, Franz. 1911. Würfel- und Buchstabenorakel in Griechenland und Kleinasien. Fest-
gruß des Archäologischen Seminars zum hundertjährigen Jubiläum der Universität Breslau.
Breslau: Grass, Barth & co.
Hoops, Johannes. 1911–13. “Buch.” In Reallexikon der germanischen Altertumskunde. I. Ed.
Johannes Hoops. Straßburg: Trübner, pp. 338–40.
Hübner, Aemilius (Emil). 1895. Exempla scripturae epigraphicae Latinae: A Caesaris dictatoris
morte ad aetatem Iustiniani. Berolinum: Reimerus.
IG = Corpus inscriptionum Graecarum (Inscriptiones Graecae). Eds. Adolph(us) Kirchhoff et
al. / Academia Litterarum Regiae Borussica (et seqq.). 14 vols. Berolinum: Reimerus (et
seq.), 1873ff.
Ihm, Max. 1887. “Der Mütter- oder Matronenkultus und seine Denkmäler.” Jahrbücher des Ver-
eins von Alterthumsfreunden im Rheinlande (Bonner Jahrbücher) 83: 1–200.
IK = Die Goldbrakteaten der Völkerwanderungszeit, 1, 2 – 3, 2: Ikonographischer Katalog 1–3.
Ed. Karl Hauck. Münster Mittelalter-Schriften 24/1, 2 – 3, 2. 3 vols. Münster: Fink,
1985–89.
Jankuhn, Herbert. 1966. “Archäologische Bemerkungen zur Glaubwürdigkeit des Tacitus in der
Germania.” Nachrichten der Akademie der Wissenschaften in Göttingen, Phil.-Hist. Klasse
1966, no. 10: 409–26.
Kluge, Friedrich. 1883. Etymologisches Wörterbuch der deutschen Sprache. Straßburg: Trübner.
— 1890. “AE gaerdas, bócstafa, bóc.” Zeitschrift für deutsches Altertum und deutsche Litte-
ratur 34: 210–13.
— 1995. Etymologisches Wörterbuch der deutschen Sprache. Ed. Elmar Seebold. 23rd ed.
Berlin: De Gruyter.

226 Bernard Mees

Kneissl, Peter. 1969. Die Siegestitulatur der römischen Kaiser: Untersuchungen zu den Sieger-
beinamen des ersten und zweiten Jahrhunderts. Hypomnemata 23. Göttingen: Vandenhoeck
& Ruprecht.
Krause, Wolfgang. 1936a. “Sinnbilder und Runen.” Altpreußen 2: 15–24.
— 1936b. “Neue Wege zur Runenforschung.” Forschungen und Fortschritte 12: 443–44.
— 1938. “Die Runen als Begriffszeichen.” In Beiträge zur Runenkunde und nordischen Sprach-
wissenschaft: Gustav Neckel zum 60. Geburtstag. Ed. Kurt H. Schlottig. Leipzig: Harrasso-
witz, pp. 35–53.
— 1939. “Runenmacht.” Geistige Arbeit 6/19: 1–4.
— 1940. “Zum Stand der Runenforschung.” Göttingische gelehrte Anzeigen 202: 181–205.
— 1943. Was man im Runen ritzte. 2nd ed. Halle a. S.: Niemeyer.
— 1970. Runen. Sammlung Goschen 2810. Berlin: De Gruyter.
Krogmann, Willy. 1955–56. “Das Buchenargument.” Zeitschrift für vergleichende Sprachfor-
schung 72: 1–29 & 73: 1–25.
Krämer, Werner. 1982. “Graffiti auf Spätlatènekeramik aus Manching.” Germania: Korrespon-
denzblatt der Römisch-Germanischen Kommission des Deutschen Archäologischen Instituts
60: 489–99.
Kuhn, Hans. 1938. “Das Zeugniss der Sprache über Alter und Ursprung der Runenschrift.” In
Beiträge zur Runenkunde und nordischen Sprachwissenschaft: Gustav Neckel zum 60. Ge-
burtstag. Ed. Kurt H. Schlottig. Leipzig: Harrassowitz, pp. 54–73.
Lane, George S. 1967. “The Beech Argument: A Re-evaluation of the Linguistic Evidence.”
Zeitschrift für vergleichende Sprachforschung 81: 197–212.
Lejeune, Michel. 1953. “Les plaques de bronze votives du sanctuaire vénète d’Este.” Revue des
études anciennes 55: 58–112.
— 1970. Manuel de la langue vénète. Indogermanische Bibliothek, Reihe 1: Lehr- und Hand-
bücher. Heidelberg: Winter.
Leo, Fridericus (Friedrich), ed. 1881. Venantius Honorius Clementianus Fortunatus, presbyterus
Italicus, Opera poetica. Monumenta Germaniae historica: Auctores antiquissima. IV.1. Bero-
linum: Weidmannus.
Leumann, Ernst. 1930. “„Suppletivwesen“ im Nordarischen.” Zeitschrift für vergleichende
Sprachforschung 57: 184–200.
Leyen, Friedrich von der. 1930. “Die germanische Runenreihe und ihre Namen.” Zeitschrift für
Völkerkunde NF 2: 170–82.
— 1938. “Die germanischen Runen.” Deutsche Rundschau 255: 94–104.
Liliencron, Rochus von & Müllenhoff, Karl. 1852. Zur Runenlehre: Zwei Abhandlungen. Halle
a. S.: Schwetschke & Sohn.
List, Guido von. 1907. Das Geheimnis der Runen. Guido von List-Bücherei 1, 1. Gross-Lichter-
felde: Zillman.
Losch, Friedrich. 1889. “Zur Runenlehre.” Germania: Vierteljahrschrift für deutsche Alterthums-
Kunde 34: 397–405.
Loth, Joseph. 1895. “Le sort chez les Germains et les Celtes.” Revue celtique 16: 313–14.

Runes in the First Century 227

Lund, Allan A., ed. 1988. P. Cornelius Tacitus, Germania. Wissenschaftliche Kommentare zu
griechischen und lateinischen Schriftstellern. Heidelberg: Winter.
— 1991. “Zur Gesamtinterpretation der ‘Germania’ des Tacitus.” In Aufstieg und Niedergang
der römischen Welt: Geschichte und Kultur Roms im Spiegel der neueren Forschung. II.
Principat. 33. Sprache und Literatur, 3. Eds. Hildegard Temporini & Wolfgang Haase.
Berlin: De Gruyter, pp. 1858–1988.
Macalister, Robert A. Stewart. 1937. The Secret Languages of Ireland. Ogham, Hisperic, Bear-
lagair na Saer, Bog-Latin, and Cryptography, with Special Reference to the Origin and
Nature of Shelta Language. Cambridge: Cambridge University Press.
— 1945. Corpus inscriptionum Insularum Celticarum I. Dublin: The Stationary Office.
Marinetti, Anna. 1990. “La tavolette alfabetiche di Este.” In Alfabetari e insegnamento della
scrittura in Etruria e nell’Italia antica. Eds. Maristella Pandolfini & Aldo L. Prosdocimi.
Biblioteca di “Studi etruschi” (Istituto di Studi Etruschi et Italici) 20. Firenze: Olschki.
Markey, Thomas L. 1999. “Studies in Runic Origins 2: From Gods to Men.” American Journal
of Germanic Linguistics and Literatures 11: 131–203.
— 2001. “A Tale of Two Helmets: The Negau A and B Inscriptions.” Journal of Indo-European
Studies 29: 69–172.
Marstrander, Carl J. S. 1925. “Les inscriptions des casques de Negau, Styrie.” Symbolae Oslo-
enses 3: 37–64.
— 1928. “Om runene og runenavnenes oprindelse.” Norsk tidskrift for sprogvidenskap 1:
85–188.
Marzell, Heinrich, ed. 1948–79. Wörterbuch der deutschen Pflanzennamen. 5 vols. Leipzig:
Hirzel.
Mattingly, Harold, trans. 1970. Tacitus. Agricola and Germania. Ed. Stanley A. Handford. The
Penguin Classics. Revised ed. London: Penguin.
Mayrhofer, Manfred. 1956–80. Kurzgefaßtes etymologisches Wörterbuch des Altindischen – A
Concise Etymological Sanskrit Dictionary. Indogermanische Bibliothek, Reihe 2: Wörter-
bücher. 4 vols. Heidelberg: Winter.
— 1986ff. Etymologisches Wörterbuch des Altindoarischen. Indogermanische Bibliothek,
Reihe 2: Wörterbücher. 3 vols. Heidelberg: Winter.
Mees, Bernard. 1997. “A New Interpretation of the Meldorf Fibula Inscription.” Zeitschrift für
deutsches Altertum und deutsche Literatur 126: 131–39.
— 1999. “The Celts and the Origin of the Runic Script.” Studia Neophilologica 71: 143–55.
— 2000. “The North Etruscan Thesis of the Origin of the Runes.” Arkiv för nordisk filologi
115: 33–82.
Mentz, Arthur. 1937. “Die notae der Germanen bei Tacitus.” Rheinisches Museum für Philologie
86: 193–205.
Meyer, Richard M. 1896. “Runenstudien. I. Die urgermanischen Runen.” Beiträge zur Ge-
schichte der deutschen Sprache und Literatur 21: 162–84.
Mogk, Eugen. 1894. “Über Los, Zauber und Weissagen bei den Germanen (Eine Bemerkung zum
10. Kapitel der Germania des Tacitus).” In Kleinere Beiträge zur Geschichte von Dozenten

228 Bernard Mees

 der Leipziger Hochschule: Festschrift zum deutschen Historikertage in Leipzig, Ostern
1894. Leipzig, pp. 81–90.
— 1900. “Mythologie” In Grundriß der germanischen Philologie. III. Ed. Hermann Paul. 2nd
ed.. Straßburg: Trübner, pp. 230–406.
Morandi, Allesandro. 1998. “Epigrafia camuna. Osservazioni su alcuni aspetti della documenta-
zione.“ Revue belge de philologie et d’histoire 76: 99–124.
Much, Rudolf. 1887. “Germanische Dative aus der Römerzeit.” Zeitschrift für deutsches Altertum
und deutsche Litteratur 31: 354–58.
— 1937. Die Germania des Tacitus. Germanische Bibliothek, Abt. 1: Sammlung germanischer
Lehr- und Handbucher, Reihe 5: Altertumskunde 3. Heidelberg: Winter.
Müllenhoff, Karl. 1900. Deutsche Altertumskunde IV: Die Germania des Tacitus. Berlin: Weid-
mann.
Murphy, Arthur. 1807. The Works of Cornelius Tacitus with an Essay on his Life and Genius,
Notes, Supplements, & c. New ed. 8 vols. London: Stockdale.
Murry, James A. H. et al., eds. 1888–1933. A New English Dictionary on Historical Principles;
Founded Mainly on the Materials Collected by The Philological Society (Oxford English
Dictionary). 10 vols plus supplement. Oxford: Clarendon.
Nedoma, Robert. 1995. Die Inschrift auf dem Helm B von Negau: Möglichkeiten und Grenzen
der Deutung norditalischer epigraphischer Denkmäler. Philologica Germanica 17. Wien:
Fassbaender.
Neumann, Günter & Beck, Heinrich. 1981. “Buche.” In Johannes Hoops, Reallexikon der ger-
manischen Altertumkunde. IV. Eds. Heinrich Beck et al. 2nd ed. Berlin: De Gruyter, pp.
55–58.
Norden, Eduard. 1922. Die germanische Urgeschichte in Tacitus Germania. 2nd ed. Leipzig:
Teubner.
Orellius, I. Gaspar (J. Kaspar von Orelli), ed. 1845–48. C. Cornelius Tacitus, Opera quae super-
sunt ad fidum codicum Mediceorum ab Io. Georgio Baitero denuo excussorum cetero-
rumque optimorum librorum rencensuit atque interpretatus est. 2 vols. Turicum: Fuesslinus
et soc.
Osthoff, Hermann. 1905. “Zwei Artikel zum Ablaut der a) u-Basen. 1. Zur Geschichte des
Buchennamens.” Beiträge zur Kunde der indogermanischen Sprachen 29: 149–58.
Page, Raymond I. 1999. An Introduction to English Runes. 2nd ed. London: Boydell.
Peeters, Christian. 1982. “Proto-Germanic ‘Book’.” General Linguistics 22: 266.
Pellegrini, Giovanni B. & Prosdocimi, Aldo L. 1967. La lingua venetica. 2 vols. Padova: Istituto
di Glottologia dell’Università di Padova.
Petsch, Robert. 1917. “Über Zeichenrunen und Verwandtes.” Zeitschrift für den deutschen Unter-
richt 31: 433–49.
Pittioni, Richard. 1942. “Zur Frage nach der Herkunft der Runen und ihrer Verdankerung in der
Kultur der europäischen Bronzezeit.” Beiträge zur Geschichte der deutschen Sprache und
Literatur 65: 373–84.

Runes in the First Century 229

Polomé, Edgar C. 1991. “The Names of the Runes.” In Old English Runes and Their Continental
Background. Ed. Alfred Bammesberger. Anglistiche Forschungen 217. Heidelberg: Winter,
pp. 421–38.
Reeb, Wilhelm, ed. 1930. Tacitus Germania. Leipzig: Teubner.
Rives, James B., trans. 1999. Tacitus, Germania. Clarendon Ancient History Series. Oxford:
Clarendon.
Robertson, Edward. 1964. “The ’Ur)um and Tunmı)m; What Were They?” Vetus Testamentum 14:
67–74.
Rolfe, John C., trans. 1935–39. Ammianus Marcellinus in Three Volumes. Loeb Classical Libra-
ry. London: Heinemann.
Rosenfeld, Hellmut. 1952. “Buch, Schrift und lateinische Schriftkenntnis bei den Germanen vor
der christlichen Mission.” Rheinisches Museum für Philologie 95: 193–209.
— 1969. “Buch und Buchstabe: Zur Sprach- und Kulturgeschichte des Schrift- und Buchwesens
von den Germanen bis auf Gutenberg.” Gutenberg Jahrbuch 1969: 338–44.
Rüger, Christoph B. 1981. “Inschriftenfunde der Jahre 1975–79 aus dem Rheinland.” Epi-
graphische Studien 12: 287–307.
— 1987. “Beobachtungen zu den epigraphischen Belege der Muttergottheiten in den latein-
ischen Provinzen des Imperium Romanum.” In Matronen und verwandte Gottheiten: Ergeb-
nisse eines Kolloquims veranstaltet von der Göttinger Akademiekommission für die Alter-
tumskunde Mittel- und Nordeuropas. Eds. Gerhard Bauchhenß & Günter Neumann. Beihefte
der Bonner Jahrbücher 44. Köln: Rheinland-Verlag, pp. 1–30.
Schlesinger, Walter. 1953. “Herrschaft und Gefolgschaft in der germanisch-deutschen Ver-
fassungsgeschichte.” Historische Zeitschrift 176: 225–75.
Seebold, Elmar. 1981. Etymologie: Eine Einführung am Beispiel der deutschen Sprache.
Beck’sche Elementarbücher. München: Beck.
— 1986. “Was haben die Germanen mit den Runen gemacht? Und wieviel haben sie davon von
ihren antiken Vorbilden gelernt?” In Germanic Dialects: Linguistic and Philological Investi-
gations. Eds. Bela Brogyanyi & Thomas Krömmelbein. Amsterdam Studies in the Theory
and History of Linguistic Science, Series IV: Current Issues in Linguistic Theory 38.
Amsterdam: Benjamins, pp. 525–80.
— 1993. “Fuþark, Beith-Luis-Nion, He Lamedh, Abğad und Alphabet: Über die Systematik der
Zeichenaufzählung bei Buchstaben-Schriften.” In Sprachen und Schriften des antiken Mittel-
meerraums: Festschrift für Jürgen Untermann zum 65. Geburtstag. Eds. Frank Heidermanns
et al. Innsbrucker Beiträge zur Sprachwissenschaft 78. Innsbruck: Institut für Sprachwissen-
schaft der Universität Innsbruck, pp. 411–44.
Seyfarth, Wolfgang, ed. 1978. Ammianus Marcellinus, Rerum gestarum libri qui supersunt. Bib-
liotheca scriptorum Graecorum et Romanorum Teubneriana. 2 vols. Lipsia: Teubnerus.
Shetelig, Haakon, 1924. “Arkeologiske Tidsbestemmelser av ældre norske Runeindskrifter.” In
Norges Indskrifter med de ældre Runer. III. Eds. Sophus Bugge & Magnus Olsen. Norges
Indskrifter indtil Reformationen I. Christiania: Brøgger, pp. 1–76.

230 Bernard Mees

Sievers, Eduard. 1891. “Runen und Runeninschriften.” In Grundriß der germanischen Philo-
logie. I. Ed. Hermann Paul. Straßburg: Trübner, pp. 238–50.
— 1901. “Runen und Runeninschriften.” In Grundriß der germanischen Philologie. I. Ed.
Hermann Paul. 2nd ed. Straßburg: Trübner, pp. 248–62.
Stocker, Arthur F. & Travis, Albert F. (eds.) 1965. Servianorum in Vergilii carmina commentari-
orum. Editio Harvardiana. III. The Commentaries on Aeneid III – V. Oxonia: Universitatus.
Stokes, Whitley. 1891. “Adamnán’s Second Vision.” Revue celtique 12: 420–43.
Vendryes, Joseph. 1948. “L’écriture ogamique et ses origines.” Etudes celtiques 4: 83–116.
Wenskus, Reinhard. 1961. Stammesbildung und Verfassung: Das Werden des frühmittelälter-
lichen gentes. Köln: Böhlau.
Wilser, Ludwig. 1895. “Alter und Ursprung der Runenschrift.” Korrespondenzblatt der Gesamt-
vereins der deutschen Geschichts- und Alterthumsvereine 43: 137–43.
— 1905. Zur Runenkunde: Zwei Abhandlungen. Leipzig: Akademische Verlag für Kunst und
Wissenschaft.
— 1912. “Ursprung und Entwicklung der Buchstabenschrift.” Mannus 4: 123–29.
Wimmer, Ludwig F. A. 1887. Die Runenschrift. Trans. Ferdinand Holthausen. Rev. ed. Berlin:
Weidmann.
Windisch, Ernst, ed. 1880. Irische Texte mit Wörterbuch I. Leipzig: Hirzel.
Winterbottom, Michael & Ogilvie, Robert M., eds. 1975. Cornelius Tacitus, Opera Minora.
Scriptorum classicorum bibliotheca Oxoniensis. Oxonia: Clarendonus.
Wissmann, Wilhelm. 1952. Der Name der Buche. Deutsche Akademie der Wissenschaften zu
Berlin: Vorträge und Schriften 50. Berlin: Akademie-Verlag.
Zimmer, Stefan. 1984. “Iran. baga-: Ein Gottesname?” Münchener Studien zur Sprachwissens-
chaft 43: 187–215.

Runes in the First Century 231