Lat. scrībere in Germanic
Matteo Tarsi
University of Iceland
The present article deals with the reflexes of Lat. scrībere in Germanic. It is
proposed that the word was borrowed into Germanic at quite an early stage
(1st century AD) as a result of contacts between West-Germanic-speaking
populations and the Romans. Special stress is put on the importance of the
Roman military in introducing the practice of writing among those that
served in the army. Special attention is given to the North Germanic reflexes
of Lat. scrībere in order to tentatively explain the morphological difference
found in that branch of Germanic, where the verb is found both in the first
class of strong verbs and in the second class of weak verbs. It is proposed
that the former conjugation is primary, and that the rise of the latter is due
to later developments such as lexical analogical processes and languageexternal causes. Furthermore, the present study confirms from a different
perspective that English influence on writing is primary in the Old-WestNorse-speaking area. Finally, Schulte’s (2015) proposal is re-read in the light
of terminological evidence from England and Scandinavia.
1.
Introduction1
While PGmc *wrīta- was used to denote the act of carving runes, for writing
on parchment many Germanic vernaculars had recourse to a Latin borrowing,
i.e. Lat. scrībere.2 Two remarkable exceptions to this are on the one hand Gothic,
1. I wish to thank several scholars who contributed with different insights to the shaping of
this article: my mentor and PhD supervisor Jón Axel Harðarson (University of Iceland), Rolf
Bremmer (Leiden University), Fabrizio D. Raschellà (University of Siena), and Seán Vrieland
(University of Copenhagen). Moreover, I wish to thank Rory McTurk (University of Leeds)
for revising my English, and the Managing Editor Stephen Laker and two anonymous
peer-reviewers for their additional helpful suggestions. Any remaining errors are responsibility
of the author.
2. Also PGmc *raista- was used to denote the act of carving runes. In Scandinavian runic
inscriptions, *wrīta- and *raista- appear to have a clear geographical distribution. Whereas the
former is used in the west (Norway, England), in the east (Denmark, Sweden and Gotland)
https://doi.org/10.1075/nowele.00019.tar | Published online: 12 April 2019
NOWELE 72:1 (2019), pp. 42–59. issn 0108-8416 | e‑issn 2212-9715 © John Benjamins Publishing Company
Lat. scrībere in Germanic
which uses the verb mēljan,3 and on the other Old English, in which the native
verb wrītan seems never to have been challenged by a foreign lexical competitor.4
OE scrīfan shows in fact only secondary meanings of the Latin verb, namely ‘to
decree, appoint, judge, doom’, and the ecclesiastical technical meanings ‘to shrive,
impose Church penance (after confession)’ (cf. furthermore ASD: s.v. scrífan).
The aim of this article is to account for the borrowing process that led from Lat.
scrībere to its reflexes in the Germanic languages, and to shed light on the morphological differences that exist in North Germanic.
As noted by Green (1998: 263–264), the West Germanic languages seem to
show a clear divide with respect to the acquisition and use of Lat. scrībere: on
the one hand there is the abovementioned OE scrīfan ‘to decree, appoint, judge,
doom, shrive, impose Church penance (after confession)’, while on the other there
are OLF skrīvan and OHG skrīban ‘to write’. Old Frisian constitutes a middle zone
by showing both meanings, with the legal and religious ones being clearly of Old
English provenance.5
The semantics of the verb in North Germanic seems to show calquing on Old
High German, for OWN skrifa,6 ODan. skriuæ, OSw. skriva, and OGu. skrifa all
mean ‘to write’.7
the latter is found (cf. Samnordisk runtext databas). Moreover, whereas Old Icelandic preserves
both verbs (ríta/rita and rísta/rista), in neither Old Danish nor Old Swedish are there reflexes of
inherited PGmc *wrīta, for Sw. rita is in all probability a Low German loan (cf. SAOB: s.v. rita).
3. With regard to the etymology of Got. mēljan, Lehmann (1986: s.v.) compares the verb to its
Germanic cognates, namely OIce. mæla ‘to paint, portray’, OE gemǣlan ‘to mark, stain’, OFris.
mēlia ‘to paint’, OS and OHG malōn ‘to paint’. Writing as painting is the idea conveyed by the
Gothic verb, as also by e.g. the runic inscriptions from Vetteland, Einang, and Rö (cf. Antonsen
1975, inscr. num. 18, 20, and 26 respectively), where PGmc *faihidōn is used.
4. Cf. by contrast the situation in Old West Nordic, where both ríta (I cl. st.) and rita (II cl. wk)
are used alongside skrifa.
5. Green (1998: 264) says that Old Saxon also showed semantic agreement with both Old High
German and Old English. However, I am not able to find any instance of the Old Saxon verb
in the technical meaning of OE scrīfan. A parallel can possibly be traced to OS biskrīƀan ‘to be
reserved, care’.
6. The terminology used throughout this article with reference to North Germanic is that
of Ottosson (2002). Old Nordic (abbr. ON) thus covers all the North Germanic vernaculars,
whereas Old West Nordic (abbr. OWN) is opposed to Old East Nordic in that it covers only the
western vernaculars. In order to avoid confusion between a narrow and a broad use of the term
“Norse”, the term “Nordic” has been adopted.
7. Note that Ásgeir Blöndal Magnússon (2008: s.v. skrifa) lists two obsolete meanings for the
verb, namely ‘to paint’ and ‘to forbid, prohibit’. Whereas there is copious evidence for the former (cf. ONP: s.v.), the latter, evidence for which could possibly point to a connection between
the Old West Nordic verb and OE scrīfan, is not otherwise recorded. I have personally checked
43
44
Matteo Tarsi
Whereas West Germanic shows division with respect to the semantics of the
verb, North Germanic does so with respect to its morphology. In West Nordic and
Elfdalian the verb is in fact inflected according to the second class of the weak
conjugation (OWN skrifa, Ice. skrifa, Nyn. skriva, Far. skriva, Elfd. skrieva). In Old
Danish, on the other hand, the verb is inflected according to the first class of the
strong conjugation. Old Swedish shows a transitional phase in that it shows both
weak and strong conjugation forms for this verb, as does Old Gutnish, where the
strong conjugation is attested from the 15th century (cf. Snædal 2002: 221).
The outline of this article is as follows: In Section 2, a critical account is given
of earlier etymological discussions of the reflexes of this verb in Germanic. Subsequently (Section 3), the issues of age and path of borrowing for this verb are
addressed. Following Rosenfeld (1952), it is proposed that the verb entered Germanic at a very early stage, namely in the period when Germanic soldiers started
to be enlisted in the Roman army, and that it subsequently spread northwards
during Late Antiquity and the Middle Ages, not least as a result of Christian missions. The discussion will then (Section 4) focus on the morphological differences
between the forms of the verb as borrowed in North Germanic. In conclusion
(Section 5), the main points of the discussion will be summed up and two observations made on the writing traditions of Iceland, Norway, and England.
2.
Earlier etymologies: A survey and critical account
Etymological research on the verb for ‘to write’ in Germanic is rooted in 19thcentury historical linguistics. Whereas the reflexes of PGmc *wrīta- originally
denoted the act of carving runes, only in English does it convey the basic meaning
‘to write’. In Continental West Germanic and North Germanic, another verb,
which is ultimately related to PIE *skrei̯bh- ‘to scratch, carve’ (cf. LIV2: 562), has
been adopted (Ger. schreiben, Du. schrijven, Dan. skrive, Sw. skriva, Elfd. skrieva,
Ice. skrifa, Far. skriva, Nyn. skriva).
The dictionaries and other lexical studies which treat the latter verb have
always been divided into two schools of thought. One camp saw the verb as necesÁsgeir Blöndal Magnússon’s dictionary slips at the Department of Lexicography of the Árni
Magnússon Institute for Icelandic Studies, together with Ásgeir Magnússon’s main sources, but
I was unable to find any reference to the source of such a meaning. Moreover, an obscure hapax
legomenon, OIce. skrifnask, appears in the skaldic lexicon, namely in a lausavísa by the 11thcentury skald Sigvatr Þórðarson. Two interpretations exist for this verb: (1) Cleasby/Vigfússon
(1874: s.v. skrifnask) ascribe to it the meaning ‘it is imposed (of penance)’; (2) Judith Jesch (2014)
interprets the verb as meaning ‘it is written to’. In his edition of the skaldic corpus, Finnur Jónsson (1912–1915: B I, p. 253) deems the passage too obscure to allow for an interpretation.
Lat. scrībere in Germanic
sarily, or at least with all probability, native to Germanic, mainly because it shows
the strong verb inflection; the other conceded instead that the verb was a borrowing from Lat. scrībere. Of the former opinion (1) have been scholars such as Kluge
(1889), W. de Vries (1921), Trier (1951), and J. de Vries (1971). The latter camp (2)
has instead obtained the favour of the majority of scholars from the late 19th century to this day (e.g. Zimmer (1892), Grimm (1854–1961), Schröder (1924), Franck/
van Wijk (1936), Falk/Torp (1960), Hellquist (1966), Seebold (1970), Pfeifer (1989),
Philippa et al. (2003–2009), and Durkin (2015)). Table 1 provides a concise synopsis of the surveyed studies along with their key features.
Some recurring points emerge from this survey. Firstly, while scholars have
had different opinions about the origin of the word, i.e. whether it is a native lexeme or a loanword, the latter option seems to have received more serious attention,
while the former view has possibly been compromised by linguistic prejudice in
certain socio-cultural milieux. Secondly, there is the question of how Lat. scrībere
entered the Germanic languages. This is closely bound up with another question
namely, how best to account for OE scrīfan ‘to decree, appoint, judge, doom, shrive’
in the light of the other reflexes of Lat. scrībere in Germanic. As to the word’s origin,
those etymologies that have taken it under scrutiny form three groups, namely
those in which ‘to write’ is considered the primary meaning (A), those in which
this meaning is considered to be secondary (B), and those which envisage a double
borrowing process most likely (C). Groups A and B reflect the aforementioned two
schools of thought: under Group A fall those etymologists who consider the verb
to be a Latin loanword (Zimmer, Schröder, Franck/van Wijk, Hellquist, Seebold,
Philippa et al., and Durkin), whereas under Group B are those for whom the verb
is native to the Germanic lexicon (Kluge 1889, W. de Vries, and J. de Vries). While
still maintaining that the verb is a loan from Latin, the etymologists in Group C
(Grimm and Pfeifer) are unique in considering the meaning ‘to write’ as secondary,
i.e. acquired with the introduction of literacy and manuscript writing: the primary
meaning for them was ‘to order, decide’, which arose together with the borrowing
itself at an early stage, i.e. through contact with the Romans.
The peculiar meanings of OE scrīfan as opposed to Continental West Germanic and North Germanic pose a significant problem for etymology which not
all the etymologists reviewed above have addressed. As it is, the answers offered
boil down to two: the meanings shown by OE scrīfan are either primary, or they
are a secondary semantic development. The former option is taken up only by
Kluge (1889), whereas Zimmer, Grimm, Schröder, Franck/van Wijk, Falk/Torp,
and Pfeifer consider the Old English meanings to be peculiar to that language.
Finally, Durkin does not take a clear position but hints at the fact that the Old
English meanings are secondary, thus derived from ‘to write’, even though they
were also partially present already in Latin.
45
46
Matteo Tarsi
Table 1. A synopsis of earlier etymological studies
Group
Source
Key features
1
Kluge (1889:
s.v. schreiben)
Postulates the existence of PGmc *skrīb-/skrĭb- ‘to impose a
punishment’. The stem was then adopted by the Church as a
technical term, hence OE scrīfan and related terms, e.g. OE
scrift ‘what is prescribed as a punishment, a penalty’, OWN
skrift ‘confession, shrift, penance, penalty’ (an Old English loan,
N/A), and OFris. skrīva ‘to write, impose Church penance’. The
meaning ‘to write’ would have been derived secondarily from
Lat. scrībere following the acquisition of literacy. a
W. de Vries (1921: 94)
Argues that the meaning ‘to write’ has arisen in Gmc due to
semantic influence from Lat. scrībere much in the same way as
did Du. lezen (= Ger. lesen, Dan. læse, Sw. läsa, Ice. lesa etc.)
from Lat. legere.
Trier (1951: 73–75)
Maintains that the verb is native to Gmc. He draws a parallel
with the reflexes of PIE *skrei̯bh- in Baltic.
J. de Vries (1971:
s.v. schrijven)
Considers the meaning ‘to write’ to have been acquired via Lat.
scrībere. The peculiar semantic development of Old English
influenced Old Frisian and Old Saxon.
Zimmer (1892)
The primary meaning of the word is ‘to write’. He explains OE
scrīfan ‘to shrive’ as a verb used in that sense by the Old English
Church, but which nevertheless had its roots in writing
practices. Its meaning would then have spread with English
missionary activity to northern Germany, the Netherlands, and
Scandinavia, leaving southern Germany unaffected, as there is
no trace of such a meaning in the OHG linguistic area.
Grimm (1854–1961:
s.v. schreiben)
Critical of Zimmer’s line of reasoning. Grimms’ dictionary
argues that the direction of semantic change could not have
been from the specialised religious meaning to a generalised
legal meaning but rather the opposite. A double borrowing
process is considered most likely: firstly by West Germanic
populations through contact with the Romans, and more
specifically from the expression scribere milites ‘to enlist
soldiers’; hence the specific meanings that developed along the
centuries in OE, OFris., and OS (and also for OHG furiskrīban
‘prescribe’); and secondly, the meaning ‘to write’ would have
been acquired with the introduction of literacy and manuscript
writing in the 8th century.
Schröder
(1924: 57–58)
PGmc *wrīta- has only been retained by the Anglo-Saxons,
who gave it a new meaning, i.e. ‘to write’ (as opposed to ‘to
carve’). The Latin loanword was introduced to German
differently, namely via the Merovingian chancery in the 8th
century, whence it came to compete with the native
synonym. b
2
Lat. scrībere in Germanic
Table 1. (continued)
Group
Source
Key features
Franck/van Wijk
(1936: s.v. schrijven)
Consider the verb to have been borrowed together with
parchment writing. OE scrīfan is thought to have developed a
specific juridical meaning in Anglo-Frisian and, furthermore, a
peculiar religious meaning in England.
Falk/Torp (1960:
s.v. skrive)
Consider the Old English meaning as secondary. S.v. skrift it is
proposed that the semantic development in OE had been from
‘to prescribe, decide’ (Lat. praescribere) to ‘to impose (Church)
penance’.
Hellquist (1966:
s.v. skriva)
Considers the verb to have been borrowed in connection with
the introduction of literacy and suggests that the North
Germanic verb has been at least partially borrowed from
neighbouring languages.
Seebold (1970:
s.v. skreib-a-)
Argues that the Old English meaning of the verb stems from
Latin usage.
Pfeifer (1989:
s.v. schreiben)
Envisages a double borrowing process from Latin: (1) with the
meanings ‘to order, decide’, hence OE scrīfan but also OS
biskrīƀan ‘to be reserved, care’; (2) with the meaning ‘to write’.
Philippa et al.
(2003–2009:
s.v. schrijven)
Early borrowing from Lat. due to the influence exerted by
Roman culture.
Durkin
(2015: 141–142)
Early Latin loan in Continental Germanic. The meaning ‘to
write’ is probably primary, whereas those of OE scrīfan derive
from it but are nevertheless already present in Latin.
a Kluge/Seebold (2002: s.v. schreiben) treat the word primarily as a loan from Lat. scrībere but still
mention the possibility of Ger. schreiben being native to Germanic.
b This coexistence and competition is preserved in the Heliand (Old Saxon, 9th c.), whereas in the
same century in Otfrid (Rhine Franconian) the two verbs appear to be semantically polarised: skrīban
‘to write’: rīzan ‘to inscribe’ (cf. also Sonderegger 2003: 180–182).
3.
Lat. scrībere in Germanic: Age and path of borrowing
Determining the age and path of the borrowing of Lat. scrībere into Germanic
amounts to tracing the history of writing (with Latin letters) among the Germanic
populations. It is thus a potential milestone that could shed light on the interrelations between Latin and Germanic cultures. The issue to be addressed is as
follows: when did Latin writing become sufficiently known among the Germanicspeaking peoples to merit borrowing of Lat. scrībere? An obvious way towards
answering this would be to consider the borrowing of the verb as having gone
hand in hand with the flourishing of vernacular literature, i.e. in the wake of
Christianisation and as a consequence of the socio-cultural implications of Christianity. Such a hypothesis would imply that Lat. scrībere had been borrowed into
47
48
Matteo Tarsi
Germanic from about the 6th century. However, this hypothesis has a major historical drawback, namely that it completely disregards earlier and substantial contacts between Latin speakers and Germanic tribes. From a linguistic point of
view, it is known that the phonological opposition between long and short vowels
was not preserved in Latin after the 3rd century (Vänäänen 2006: 31). Thereafter,
long and short vowels still existed only at the phonetic level, their length being
dependent on syllable structure. Thus, in theory, Lat. scrībere could still have been
borrowed at a later time with a long root vowel. In the present article, another
hypothesis is favoured, however, namely that Lat. scrībere entered Germanic as a
result of earlier contacts between Germanic tribes and the Romans. This hypothesis needs to fulfil the following minimum requirements:
1. That such contacts actually took place over a considerable period of time.
2. That such contacts were of a kind in which writing was involved.
3. That Germanic/Latin bilingualism is documented in the period during which
Lat. scrībere was borrowed.
Point 1:
Point 2:
We know that the earliest documented close contacts between Romans
and Germanic populations first occurred following Caesar’s conquest of
Gaul (50 BC) and the establishment of the Rhine as a frontier (Adams
2003: 274). These contacts were primarily military and commercial and
lasted at least until the fall of the Western Roman Empire in 476 AD.
If we consider only the military and commercial aspects of these contacts, these two types of contact naturally involved a certain degree of
fluency in both spoken and written Latin. With regard to the military
aspect, the relationship between the Roman and Germanic populations
was close, as members of the Germanic tribes were constantly enrolled
in the Roman army (Carroll 2001: 102–103). It is moreover known that
there was a certain degree of intermingling between Germanic peoples
and the Romans, especially along the frontier (cf. Carroll
2001: 104–108). Also, there are numerous Latin anecdotal accounts
about the Latin fluency of certain members of the Germanic tribes,
e.g. Arminius (18 BC–19 AD, cf. furthermore Adams 2003: 20–21 and
275–279). Moreover, to belong to the Roman army certainly involved
fluency in both spoken and written Latin, at least for some of the soldiers, for writing was an integral part of military organisation (see
Rüger 1998: 357–360). Indeed, a number of Germanic soldiers adopted
Roman names, e.g. Gaius Julius Civilis, the leader of the Batavian revolt
in 69 AD (cf. furthermore Kluge 1913: § 7 and also Birley 2000). Fluency
in written Latin is best seen in surviving military correspondence, i.e.
writing tablets such as those found at Vindolanda, a fort along Hadrian’s
Lat. scrībere in Germanic
Point 3:
Wall garrisoned by Batavian soldiers. Fluency in spoken Latin, on the
other hand, can be assumed as a consequence of the Roman-Germanic
interactions along the limes in particular.
From the points listed above it also follows that a certain degree of bilingualism was present at least among the speakers of one or other variety of Germanic, particularly among those in the army. Germanic/Latin
bilingualism is moreover documented by Latin authors such as, among
others, Tacitus (for a complete account see Adams 2003: 275–279).
The minimum requirements listed under 1–3 above are all met by the present
hypothesis. It can thus be suggested that Lat. scrībere was borrowed into Germanic
at a very early stage (1st century AD),8 following close military as well as commercial contacts between the Romans and the populations native to the newly conquered territories extending north of the Alps to the west bank of the Rhine. It
is unfortunately impossible to get any closer to where the loanword first entered
Germanic. It can however be proposed that this is likely to have happened in
one of the important centres along the Rhine, maybe Cologne, Mainz, or Trier,
although this can be no more than speculation.
Given the phonological structure of the root of the Latin verb, i.e. /skrīb-/, the
primary morphological adaptation of the loanword in such an early period would
naturally have been to the first class of strong verbs, whose root vowel in the present tense was also /ī/. Moreover, it may be supposed that, at least in some cases,
the Latin and West Germanic verbs would have sounded alike (cf. Lat. scrībō,
scrībis, scrībit : WGmc *skrīƀū, *skrīƀis/z(i), skrīƀiþ/đ(i)), thus facilitating the borrowing process. It should, however, be borne in mind that the verb denoted a practice proper to Latin culture and administration, and it would not be surprising if
the Germanic peoples, whenever they started to carve runes (probably not much
later than the acquisition of Lat. scrībere, cf. also Tacitus’ [1962: chapter 10] account
of divination in his De origine et situ Germanorum), had used a native verb to
denote such a practice.
The distribution of the historically attested forms in the Germanic language
family tends to suggest that the borrowing is of some age, at least for West Germanic (but see also below § 4). The fact that PGmc *skrīƀa- is a strong verb has
compelled some scholars to consider the word as native to the Germanic lexicon,
as shown above.
8. Lat. scrībere is of course not the only early Latin loanword in Germanic. Other examples of
early Latin loans are Lat. aureus ‘aureus (kind of Roman gold coin)’ > OIce. eyrir, ODan. øre,
OSw. öre; Lat. cellārium ‘food storage’ > OHG kellari, OS kelleri ‘cellar’, OIce. kjallari; Lat. vīnum
‘wine’ > Got. wein, OE wīn, OHG and OS wīn ‘wine’, and many others (cf. furthermore Kluge
1913: § 8–12, Scardigli 1995: 561–562, and Green 1998: 201–218).
49
50
Matteo Tarsi
A last word must be said on OE scrīfan, which seems to elude the otherwise
well-attested semantics of this verb in Germanic. As seen in Section 2 above, the
Old English reflex of Lat. scrībere shows very specific legal and religious meanings, the religious meaning ‘to impose penance’ being also shared by Old Frisian,
possibly due to Old English influence. As pointed out by Durkin (2015: 141–142),
the legal and religious meanings displayed by OE scrīfan are both derivable from
‘to write’, i.e. they imply the act of writing, and are in part already present in
the semantics of Lat. scrībere. Within the framework of the present hypothesis,
Durkin’s remark stimulates a reflection on whether it is possible to ascribe an otherwise unattested meaning ‘to write’ to the alleged precursor of the Old English
verb, or whether conversely, it can be said that pre-Old English did not know the
verb. The fact that Old English has neither semantically polarised nor established
a synonymic relationship between PGmc *wrīta- and the reflex of Lat. scrībere,
as happened elsewhere in Germanic, strongly suggests that the loanword had not
spread early on in the variety of Germanic spoken by the Anglo-Saxons, since
if it had done so a more obvious reorganisation of the lexicon would have been
expected. Moreover, in line with the principle followed by Dekker (2002: 30–31),
the presence of the meaning ‘to write’ in Old Frisian demonstrates that the religious meaning of the verb is secondary and due to influence from Old English.
This can only mean that the verb, with its meaning ‘to write’, was adopted in Frisia
after the departure of the Anglo-Saxons, possibly from Old Low Franconian. The
rise of OE scrīfan with its legal and religious meanings is thus to be ascribed to a
phase in which Old English was receptive to Latin loans in the semantic sphere
in which OE scrīfan is found.9 Finally, the special position that Old English holds
in this respect is further confirmed by the absence of an otherwise common Germanic semantic loan for ‘to read’. OE lesan means in fact ‘to lease, gather, collect’,
whereas for the practice of reading one observes a semantic development of the
native verb rǣdan, which also belonged, like OE wrītan, to runic literacy.
4.
Lat. scrībere in North Germanic: Strong vs. weak conjugation
The reflexes of Lat. scrībere in North Germanic show both strong and weak conjugation. Whereas Danish consistently inflects the verb according to the first
class of strong verbs, West Nordic (Old Icelandic, Old Norwegian, and Faroese)
9. Older Scots also testifies to this state of things, as it in fact shows two different reflexes of
Lat. scrībere. On the one hand there is OSc. schrive, which corresponds to OE scrīfan, while on
the other there is OSc. scrieve ‘to write (copiously)’ which is undoubtedly a Scandinavian loan
(DSL: s.vv. schrive and scrieve).
Lat. scrībere in Germanic
and Elfdalian nativise the verb as a member of the second class of the weak conjugation. Old Swedish and Old Gutnish show both weak and strong paradigms.
In Old Swedish (skriva), runic evidence bears witness to the coexistence of
the two forms, though it must be admitted that the word is not copiously attested.
On a rune-stone from the latter part of the 12th century (Sm 81) the strong form
skref is observable. On the other hand, the first attestation of the weak form in
runic Swedish is on Sm 23B, which has been dated to 1250–1300. Both strong and
weak paradigms continue to coexist well into the 15th century according to Söderwall’s dictionary (see Sö: s.v. skriva). It is, moreover, worth mentioning that Middle
Swedish forms like skreuadher and skrefuadher clearly show that the verb could
also have a short stem vowel, i.e. /e/ < /i/ (cf. Noreen 1904: § 115, 1).
In Old Gutnish (skrifa), the strong conjugation starts to oust the weak in the
15th century (Snædal 2002: 221). Both paradigms are found in a Gutnish inscription (G 55) from 1459.10 In Guta lag (Vrieland 2017), the verb is found only in its
second class weak conjugation. Snædal believes that the emergence of the strong
paradigm in Old Gutnish is due to Danish influence and thus considers the weak
conjugation as primary.
Elfdalian consistently inflects the verb (skrieva) according to the second class
of weak verbs. The diphthong /ie/ bears witness to the fact that the word entered
Elfdalian with a long stem vowel, i.e. /ē/. The only source for such a borrowing
is Middle Swedish, where /e/ in open syllable could yield /ē/ (cf. Noreen
1904: § 115,1).
In West Nordic, i.e. Old Norwegian and Old Icelandic (skrifa),11 the verb is
consistently inflected according to the second class of weak verbs. However, Modern Norwegian, in both Bokmål and Nynorsk, mainly inflects the verb strong,
although the weak inflection is attested in a number of dialects, often coexisting
with the strong paradigm (cf. Venås 1967: 42–43). OWN skrifa is first recorded in
the oldest manuscripts in both Norway and Iceland, where it competes with the
endogenous synonyms ríta (I cl. st.) and rita (II cl. wk). According to the oldest written manuscripts (Larsson 1891 for Old Icelandic, Holtsmark 1955 for Old
Norwegian, see Table 2 and 3), however, the loanword was little used, especially
in comparison with its endogenous counterpart. Whereas in the oldest written
sources Icelandic attests only the endogenous verb in the strong conjugation, in
Norway a transitional period in which both strong and weak conjugations coex-
10. ta en : iak uar · skrivaþ […] betar · aukar·sarfa · han skr-if mik ‘When I was written […]
Pétar of Ocksarve, he wrote me’.
11. Faroese, where the verb is inflected according to the second class of weak verbs, can be
safely left out of the present discussion, as it developed late and is sparsely attested.
51
52
Matteo Tarsi
ist is observable from ca. 1200. Such a coexistence also appears in Icelandic but
approximately half a century later than in Norway (cf. ONP: s.vv. ríta and rita).
Table 2. Skrifa vs. ríta in the oldest Icelandic manuscripts (Larsson 1891) *
skrifa
ríta
Total
Rb
4
÷
43
1
H
Ph III
645
El
1
1
1
1
20
÷
3
19
* Rb = Rímbegla in GKS 1812 4to (from 1192); H = Stock. perg. 15 4to (Icelandic Homily Book, ca. 1200);
Ph III = AM 673 a II 4to (Physiologus, ca. 1200); 645 = AM 645 4to (1225–1250); El = AM 674 a 4to
(Elucidarius, 1150–1200).
Table 3. Skrifa vs. ríta and rita in the oldest Norwegian manuscripts (Holtsmark 1955) **
Total
Hom I
Hom II
Hom III
Ra 81 B
OT
skrifa
3
2
÷
1
÷
÷
ríta
4
÷
4
÷
÷
÷
rita
12
3
6
÷
1
2
** Hom I–III = AM 619 4to (Norwegian Homily Book, ca. 1200, the Roman numbers refer to different
hands); Ra 81 B = NRA 1 b, ca. 1200; OT = DG 4, the last two folia of Ólafs saga Tryggvasonar of Oddr
Snorrason.
In seeking to understand the double nature of the loanword in North Germanic,
we may briefly turn to the historical background of the word’s adoption.
In Roman times, Scandinavia was largely on the periphery of the main commercial routes which helped shape continental Europe before the fall of the Empire.
However, historical sources as well as archaeological finds (e.g. Storgaard 2001;
Imer 2004 and 2010) indicate that southern Scandinavia was far from being an
entirely isolated community. After all, the emergence of runic script in Scandinavia must be seen in the context of cultural contacts at least among Germanic
tribes if not to some extent also involving Roman literacy (cf. Imer 2010 and
Spurkland 2010). At any rate, the adoption of Lat. scrībere in Scandinavia must be
viewed as a relatively late acquisition, since the opposite hypothesis fails to fulfil
the points set out above. Thus, it is here that the borrowing took place in the wake
of Christianity (from the 8th c.), which was undoubtedly followed by the introduction of parchment writing. According to palaeographic studies (notably Brøndum-Nielsen 1944 and Seip 1954), this innovation was introduced to Scandinavia
via continental Europe and the British Isles. As far as the present subject of study
is concerned, only the former source comes into play, as in England the reflex
of Lat. scrībere belonged to an altogether different semantic sphere. In continental Europe there are two areas of interest: namely a Romance-speaking area (i.e.
France) and a Germanic-speaking area (i.e. Northern Germany). From a theo-
Lat. scrībere in Germanic
retical perspective, both areas could have been the source of the loan in Scandinavia, and it is indeed possible that they both were. In France as well as Germany
learned men undoubtedly spoke Latin to each other, and Latin was the language in
which their instruction took place. On the other hand, a certain degree of mutual
intelligibility surely existed among the varieties of Germanic spoken in Northern
Germany and Scandinavia, a factor which could have helped in the process of borrowing: the case of Old Danish is especially relevant here.12 Thus, the possibility
that Lat. scrībere entered the Scandinavian languages directly from the Latin used
in learned circles cannot be entirely ruled out.
We may now attempt to explain the difference in the morphological adaptation of the verb in North Germanic. As mentioned above, the weak inflection was
widespread in Scandinavia and occurs in Old Swedish, Old Gutnish, Old Norwegian, and Old Icelandic, together with the late-attested Faroese and Elfdalian. On
the other hand, we note that in Old Swedish, Old Gutnish, and Norwegian the
verb also shows inflectional forms of the first class of strong verbs. If we make a
comparison – taking into consideration that certain areas do not show the strong
inflection while others, mostly dialectal, show remnants of the weak inflection –
we may conclude that Lat. scrībere was borrowed for the most part as a weak verb
and subsequently underwent a change in inflectional class. In Sweden, Norway,
Iceland, and on Gotland, the primary form in which the verb has been borrowed
must thus have been the weak. In Denmark, by contrast, it is possible that the verb
never inflected weak.13
Elfd. skrieva, a loan from Middle Swedish, bears witness to the fact that the
verb had been borrowed with a long stem vowel in that language, i.e. /ē/, which
subsequently underwent diphthongisation. The Old Swedish form from which
Elfd. skrieva ultimately derives thus had /i/, and we can postulate that, after having
been initially assigned to the second class of weak verbs, it shifted to the first class
of strong verbs by language-internal causes, i.e. analogy (cf. OSw. driva, riva etc.).
In the case of Old Gutnish, by contrast, language-external factors such as language
contact have been proposed (Snædal 2002: 221). In all likelihood, the verb had also
been originally borrowed with /i/ in Old West Nordic. Here, however, in contrast
to Old Swedish and Old Gutnish, the verb never shifted inflectional class. Con-
12. Veturliði Óskarsson (2003: 150 and 172–174) is inclined to ascribe the loan in Icelandic to
Old Saxon or possibly Middle Low German.
13. Strik (2015: 40, 218) rightly points out that a late analogical change must have occurred in
Swedish but assumes that, since the verb is already attested with strong inflection at an early
stage in the Germanic languages, it must have undergone an analogical change from weak to
strong inflection. Unlike the present article, Strik does not entertain the possibility that Lat.
scrībere entered one branch of Germanic as a strong verb and another primarily as a weak verb.
53
54
Matteo Tarsi
versely, it is conceivable that its weak inflection influenced its endogenous counterpart ríta, which switched to the second class of weak verbs, first in Norway and
subsequently in Iceland. One tentative way of explaining why the verb never had
strong inflection in Old West Nordic, at least as far as is known, is that it was a late
loan (late 11th–12th c.),14 and was thus assigned a productive inflectional class (as
was also initially the case in Swedish and Gutnish). It would not have been possible to classify it as a class I strong verb because of the phonological opposition still
extant in Old West Nordic between /i/ and /ī/.
5.
Conclusions
This article has attempted to demonstrate on the one hand the age and the borrowing path of Lat. scrībere in Germanic, and on the other the causes for the
borrowed verb being assigned to the first class of the strong conjugation and the
second class of the weak conjugation in North Germanic. In order to account for
the regular adaptation of the loan in West Germanic, it is necessary to consider
that the verb was borrowed directly from Latin at an early stage (1st century AD)
thanks to the close military and commercial contacts of the Romans with the Germanic tribes absorbed into the Roman state. Evidence for this is suggested by the
fact that, in line with the historically attested Old West Germanic languages, the
Germanic loan calqued the phonemic structure of the Latin word, thus giving
PGmc *skrīƀa-. A likely place for this borrowing to have taken place could have
been one of the main centres along the Rhine: Cologne, Mainz, or Trier, but in
view of the impossibility of pinpointing precisely the place and time of the borrowing, this must remain pure speculation.
As regards the expansion of such a term northwards, a terminus a quo has
been established, namely the departure of the Anglo-Saxons from continental
Europe. This appears to be substantiated both by the peculiar semantics of the
Old English verb (which knows nothing of ‘to write’) and by the attestation of the
14. Albeit preserved in much later manuscripts, the First Grammatical Treatise (1130–1140,
AM 242 fol. from ca. 1350) and Íslendingabók (beginning of the 12th c., AM 113 b fol. from
ca. 1650) may provide evidence of the spread and use of OIce. skrifa vs. ríta/rita. In the First
Grammatical Treatise, the former verb is never used, and there is no reason to take account
of the possibility that its occurrences were changed by later scribes. On the other hand, the
use of ríta/rita is such that the former occurs 19 times whereas the latter occurs 12 times. The
overall occurrence of ríta/rita is 41 times, i.e. if occurrences are taken into account where it
is not possible to determine to which inflectional class the verb belongs. In Íslendingabók, on
the other hand, skrifa is used 4 times, whereas ríta is used only once (<ritiþ>, AM 113 b fol., f.
1v13). OIce. rita does not occur.
Lat. scrībere in Germanic
meaning ‘to write’ in Old Frisian, alongside the religious meaning (but not all the
meanings) also found in Old English.
With regard to the double nature of this verb in North Germanic, the possibility that it could have reached Scandinavia thanks to contacts with the Romans
has been ruled out, as these contacts were not continuous and did not necessarily
involve any writing, least of all writing in Latin, although it seems, admittedly, that
Roman writing was not unknown in the 1st century AD in Scandinavia (cf. furthermore Storgaard 2001; Imer 2004 and 2010, and Spurkland 2010). It has thus
been suggested that the loan must be a relatively late borrowing into North Germanic, probably not arriving on the scene before the advent of Christianity (from
the 8th c.). In order to account for the differences in conjugation in the various
North Germanic languages it has been proposed that: (1) the verb entered North
Germanic with /i/ as the root vowel; (2) Old Saxon or Middle Low German might
have influenced the retention in Old Danish of only the strong conjugation; (3)
the verb had otherwise entered other North Germanic languages as a second class
weak verb; and (4) its shifting to the first class of strong verbs is possibly due to a
synergy of lexical analogical processes and language-external causes. In Old West
Nordic, the verb probably never shifted inflectional class partly because it is a late
loan, partly because of the phonological opposition still extant between /i/ and /ī/.
In Old West Nordic, moreover, the verb possibly influenced a shift in inflectional
class of its native counterpart ríta.
As for the writing tradition in Norway and Iceland on the one hand and in
England on the other, the present study allows for two considerations. Firstly,
palaeographic studies (Seip 1954 and Hreinn Benediktsson 1965) demonstrate that
parchment writing in both Norway and Iceland was influenced to varying degrees
by the insular, i.e. English, tradition. Nevertheless, these same studies also concede that the continental tradition left its mark on Icelandic as well as Norwegian
script. Given that Lat. scrībere cannot have reached Norway, and hence Iceland,
from England, it must have done so via a continental source. On the other hand,
the widespread use of the endogenous terms ríta and rita in the earliest Norwegian and Icelandic written documents suggests that the source for the meaning
‘to write’ comes from England, i.e. wrītan (cf. furthermore Hreinn Benediktsson
1965: 40). In this respect, the witness of the First Grammatical Treatise is crucial.
In fact, the lack of OIce. skrifa in the entire treatise might be interpreted as a conscious intention by its author to establish and consolidate a native term on the
model of the English tradition, which he is known to have followed.
Secondly, Schulte (2015) has suggested that “the extension of the fuþark in the
Anglo-Frisian setting is due to close contact with the Christian Church, including manuscript culture and Classical grammatical schooling, whereas these factors were almost entirely absent in pre-Viking-Age Scandinavia”. It could thus be
55
56
Matteo Tarsi
suggested that, if Schulte’s theory is valid (at least with regard to the Anglo-Saxons),15 then it would be substantiated by the fact that Old English uses exclusively
native verbs for activities such as those of writing and reading, whereas in Scandinavia the very same terminology has strong foreign traits. One has to think here
not just of the verb for ‘to write’, but also of the verb for ‘to read’, the history of
which might well have been comparable to that of Lat. scrībere.16
References
Adams, J. N. 2003. Bilingualism and the Latin language. Cambridge: Cambridge University
Press. https://doi.org/10.1017/CBO9780511482960
Antonsen, E. H. 1975. A concise grammar of the older runic inscriptions. Tübingen: Niemeyer.
https://doi.org/10.1515/9783111411583
ASD = An Anglo-Saxon Dictionary: Based on the manuscript collections of the late Joseph
Bosworth. 1972. Edited by T. Northcote Toller. Oxford: Oxford University Press.
Ásgeir Blöndal Magnússon. 2008. Íslensk orðsifjabók. 4th edn. Reykjavik: Orðabók Háskólans.
Birley, A. R. 2000. The Names of the Batavians and Tungrians in the Tabulae Vindolandenses.
In T. Grünewald (ed.), Germania inferior, RGA-E 28, 241–260. Berlin: De Gruyter.
Brøndum-Nielsen, J. 1944. Palæografi A: Danmark og Sverige. Nordisk Kultur XXVIII:A.
Stockholm: Albert Bonniers Förlag.
Carroll, M. 2001. Romans, Celts & Germans. Stroud: Tempus.
Cleasby, R. & G. Vigfusson. 1874. An Icelandic-English dictionary. Oxford: Clarendon.
Dekker, K. 2002. Between Rome and Rustringen: Latin loan words in Old Frisian. In
P. Boersma, P. H. Breuker, L. G. Jansma & G. van der Vaart (eds.), Philologia Frisica anno
1999, 27–56. Ljouwert: Fryske Akademy.
DSL = Dictionary of the Scots language/Dictionar o the Scots leid. 2004. Scottish Language
Dictionaries Ltd. http://www.dsl.ac.uk
Durkin, P. 2015. Borrowed words. Oxford: Oxford University Press.
Falk, H. & A. Torp. 1960. Norwegisch-dänisches etymologisches Wörterbuch. 2nd edn. Oslo:
Universitetsforlaget.
Finnur Jónsson (ed.). 1912–1915. Den norsk-islandske skjaldedigtning. A I–II. Tekst efter
håndskrifterne. B I–II. Rettet tekst. Copenhagen: Møller.
15. Schulte’s theory is developed in his 2012 article, where he focuses on the relationship
between epigraphic literacy and Christianity, in particular in the light of the Byggen runic find.
See especially pp. 157–159.
16. There are several examples in Old Icelandic of the verb ráða ‘to read’ (see Hreinn Benediktsson 1965: 17, fn. 2, and ONP: s.v. ráða rit) alongside those of lesa in the same meaning.
According to Hreinn Benediktsson (1972: 53, fn. 2), the verb undoubtedly derives from runic
terminology. At any rate, the use of ráða is to be linked to the English tradition in much the
same way as we have seen for ríta/rita. On the other hand, the use of lesa is to be linked to the
German tradition, as is that of skrifa.
Lat. scrībere in Germanic
Franck, J. 1936. Franck’s Etymologisch woordenboek van der nederlandsche Taal, 2nd edn by
N. de Wijk. ’s-Gravenhage: Nijhoff.
Green, D. H. 1998. Language and history in the early Germanic world. Cambridge: Cambridge
University Press.
Grimm, J. & W. Grimm. 1854–1961. Deutsches Wörterbuch. Leipzig.
Hellquist, O. 1966. Svensk etymologisk ordbok. 3rd edn. Lund: Gleerup.
Holtsmark, A. 1955. Ordforrådet i de eldste norske håndskrifter til ca. 1250. Oslo: Dybwad.
Hreinn Benediktsson. 1965. Early Icelandic script. Icelandic Manuscripts Series in Folio II.
Reykjavik: The Manuscript Institute of Iceland.
Hreinn Benediktsson. 1972. The First Grammatical Treatise. University of Iceland Publications
in Linguistics 1. Reykjavik: Institute of Nordic Linguistics.
Imer, L. M. 2004. Latin og græsk i romersk jernalder: Fremmed indflydelse på Nordens
tidligste runeskrift. Aarbøger for Nordisk Oldkyndighed og Historie (2004). 63–105.
Imer, L. M. 2010. Runes and Romans in the north. Futhark 1. 41–64.
Jesch, J. (ed.). 2014. Sigvatr Þórðarson, Lausavísur 28. In D. Whaley (ed.), Poetry from the
Kings’ Sagas 1: From mythical times to c. 1035, 734. Skaldic Poetry of the Scandinavian
Middle Ages 1. Turnhout: Brepols.
Kluge, F. 1889. Etymologisches Wörterbuch der deutschen Sprache. 4th edn. Strasbourg: Trübner.
Kluge, F. 1913. Urgermanisch: Vorgeschichte der altgermanischen Dialekte. Grundriss der
Germanischen Philologie 2. 3rd ed. Strasbourg: Trübner.
Kluge, F. 2002. Etymologisches Wörterbuch der deutschen Sprache, 24th edn by E. Seebold.
Berlin: De Gruyter.
Larsson, L. 1891. Ordförrådet i de älsta islänska handskrifterna. Lund: Lindstedts.
Lehmann, W. P. 1986. A Gothic etymological dictionary. Leiden: Brill.
LIV2 = Rix, H. (ed.). 2001. Lexikon der indogermanischen Verben. Die Wurzeln und Ihre
Primärstammbildungen, 2nd edn by M. Kümmel and H. Rix. Wiesbaden: Reichert.
Noreen, A. 1904. Altschwedische Grammatik mit Einschluss des Altgutnischen. Sammlung
kurzer Grammatiken germanischer Dialekte VIII. Halle a. S.: Niemeyer.
ONP = Ordbog over det norrøne prosasprog. Copenhagen: Københavns Universitet. http://onp
.ku.dk
Ottosson, K. 2002. Old Nordic: A definition and delimitation of the period. In O. Bandle et al.
(eds.), The Nordic Languages, vol. I, Handbücher zur Sprach- und
Kommunikationswissenschaft 22.1, 787–793. Berlin: De Gruyter.
Philippa, M. et al. (eds.). 2004–2009. Etymologisch woordenboek van het Nederlands.
Amsterdam: Amsterdam University Press. https://doi.org/10.5117/9789053566534
Pfeifer, W. 1989. Etymologisches Wörterbuch des Deutschen. Berlin: Akademie-Verlag.
Rosenfeld, H. 1952. Buch, Schrift und lateinische Sprachkenntnis bei den Germanen vor der
christlichen Mission. Rheinisches Museum für Philologie (Neue Folge) 95(3). 193–209.
Rüger, C. B. 1998. Lateinische Schriftlichkeit im römischen Grenzgebiet gegen die Germanen.
In K. Düwel (ed.), Runeninschriften als Quellen interdisziplinärer Forschung, RGA-E 15,
357–375. Berlin: De Gruyter. https://doi.org/10.1515/9783110821901.357
Samnordisk runtextdatabas. Uppsala: Uppsala Universitet. http://www.nordiska.uu.se/forskn
/samnord.htm
SAOB = Svenska Akademiens ordbok. Stockholm: Svenska Akademien. http://www.svenska.se
/saob
Scardigli, P. 1995. Fremde Einflüsse im Germanischen. In H. Beck et al. (eds.), Reallexikon der
Germanischen Altertumskunde, vol. 9, 2nd edn, 552–563. Berlin: De Gruyter.
57
58
Matteo Tarsi
Schröder, E. 1924. Bunte Lese II. Zeitschrift für deutsches Altertum und deutsche Literatur 61/1.
57–60.
Schulte, M. 2012. Pragmatic runic literacy in Scandinavia c. 800–1300: With a particular focus
on the Bryggen material. In K. Zilmer & J. Jesch (eds.), Epigraphic literacy and Christian
identity, 155–182. Turnhout: Brepols. https://doi.org/10.1484/M.USML‑EB.5.100476
Schulte, M. 2015. Runology and historical sociolinguistics: On runic writing and its social
history in the first millennium. Journal of Historical Sociolinguistics 1(1). 87–110.
https://doi.org/10.1515/jhsl‑2015‑0004
Seip, D. A. 1954. Palæografi B: Norge og Island. Nordisk Kultur XXVIII:B. Stockholm: Albert
Bonniers Förlag.
Seebold, E. 1970. Vergleichendes und etymologisches Wörterbuch der germanischen starken
Verben. The Hague: Mouton. https://doi.org/10.1515/9783110821956
Snædal, Th. 2002. Medan världen vakar: studier i det gotländska runinskrifternas språk och
kronologi. Runrön 16. Uppsala: Institutionen för nordiska språk.
Sonderegger, S. 2003. Althochdeutsche Sprache und Literatur. 3rd edn. Berlin: De Gruyter.
https://doi.org/10.1515/9783110907957
Sö = Söderwall, K. F. 1842–1924. Ordbok öfver svenska medeltids-språket. Stockholm: Svenska
fornskriftsällskapet.
Spurkland, T. 2010. The Older Fuþark and Roman script literacy. Futhark 1. 65–84.
Storgaard, B. 2001. Himlingøje: Barbarian empire or Roman implantation. In B. Storgaard
(ed.), Military aspects of the aristocracy in Barbaricum in the Roman and early migration
periods, 95–111. Copenhagen: Danish National Museum.
Strik, O. 2015. Modelling analogical change: A history of Swedish and Frisian verb inflection.
Groningen: Rijksuniversiteit Groningen. (Doctoral dissertation)
Tacitus, P. C. 1962. Germania: Text. Aschendorffs Sammlung lateinischer und griechischer
Klassiker. 7th edn by W. Franzmeier. Münster: Aschendorff.
Trier, J. 1951. Lehm: Etymologien zum Fachwerk. Marburg: Simons Verlag.
Väänänen, V. 2006. Introduction au latin vulgaire. Reprint of the 3rd edn. Paris: Klincksieck.
Venås, K. 1967. Sterke verb i norske målføre. Oslo: Universitetsforlaget.
Veturliði Óskarsson. 2003. Middelnedertyske låneord i islandsk diplomsprog frem til år 1500.
Bibliotheca Arnamagnæana XLIII. Copenhagen: Reitzels Forlag.
Vrieland, S. D. 2017. Old Gutnish in a Danish hand: Studies in the B manuscript of Guta lag.
Copenhagen: Københavns Universitet, Det Humanistiske Fakultet. (Doctoral
dissertation)
Vries, J. de. 1971. Nederlands etymologisch woordenboek. Leiden: Brill.
Vries, W. de. 1921. Etymologische aanteekeningen. Tijdschrift voor Nederlandse Taal- en
Letterkunde 40. 89–111.
Zimmer, H. 1892. Aus der Bedeutungsgeschichte von schreiben und schrift. Zeitschrift für
deutsches Altertum und deutsche Literatur 36. 145–150.
Lat. scrībere in Germanic
Manuscripts
Den Arnamagnæanske Samling, Copenhagen
AM 242 fol.
AM 619 4to
AM 645 4to
AM 674 a 4to
Kungliga Biblioteket, Stockholm
Stock. perg. 15 fol.
Riksarkivet, Oslo
NRA 1 b
Stofnun Árna Magnússonar í íslenskum fræðum, Reykjavik
AM 113 b fol.
AM 673 a II 4to
GKS 1812 4to
Uppsala Universitetsbibliotek, Uppsala
DG 4
Address for correspondence
Matteo Tarsi
University of Iceland
c/o Stofnun Árna Magnússonar í íslenskum fræðum
Laugavegi 13, IS-101 Reykjavik
Iceland
[email protected]
Publication history
Published online: 12 April 2019
Erratum: In footnote 2 of the original article, published 12 April 2019, two occurrences of
‘*raista’ were incorrectly recorded as ‘*reista’. These have been corrected in the current version.
We apologize for this error.
59
Lat. scrībere in Germanic
Errata corrige
Abstract:
It is proposed that the former conjugation is
primary, and that the rise of the latter ... causes. >
It is proposed that the latter conjugation (i.e. II cl.
wk) is primary, and that the rise of the former (i.e.
I cl. st.) ... causes.
p. 42, footnote 2:
*raista- > *rīsta- (twice)
p. 43 footnote 2:
*wrīta > *wrīta-.
p. 51:
… short stem vowel … long stem vowel … >
… short root vowel … long root vowel …
p. 53:
… long stem vowel … > … long root vowel …