Younger Runes in Manuscripts and Early Printed Works

Munich, Bayerisches Staatsbibliothek, MS Clm. 276 and MS Clm. 641

As of June 2020, the best images of these manuscipts available to me are reproductions of a black and white microfilm, and thus while fairly good are neither in colour nor as sharp as would be optimal. Presumably, the original manuscripts will eventually be covered by the ongoing effort of the Munich DigitiZation Center, MDZ.

MS Clm. 276, Folio 96 verso

MS Clm. 641, Folio 38 recto

Folio 96 verso of MS Clm. 276 from the 14th century contains various Hebrew, Greek and fantasy alphabets. At the top of this collection is a fuþark of the intermediate type with values and names immediately followed by a runic alphabet with values only. They are closely related, but bear clear marks of having been copied separately; the alphabet does not derive directly from the fuþark in the form it has here. The rune names and both sets of values are written in the same hand as the rest of the page and does not seem to be a later addition. However, they are written after the runes, and the layout of the names is badly planned. The length of the names meant that they gradually got out of alignment with their corresponding runes, so the scribe had to resort to writing the names on two rows with lines showing which rune each belongs to.

Folio 38 recto of MS Clm. 641 from the 15th century contains three fantasy alphabets and a runic alphabet of the same origin as that in MS Clm. 276, here too with values in the same hand as the rest of the page. The same fantasy alphabet (not known to me from other such compilations) follows the runic alphabet in both manuscripts. The two manuscripts share many unoriginal features, but although the shape of the runes are generally less well preserved here, there are also enough instances of more original features preserved than could be due to chance. The common ancestor of both manuscripts must therefore have been significantly closer to the true epigraphical form of the runes, but still already have acquired the shared peculiarities of the two surviving manuscripts.

The fuþark

In the following transcription, dotted ‘i’ is used twice, contrasting with the usual ‘ı’. This also occurs sporadically elsewhere on the page, and must be intentional despite not signifying any distinction between the dotted and undotted forms.

ſo

ꞇ

ꝺ

fhe

ꞇhoꝛs

reꞇher

chon

hagalꝺ

noꝺhⸯ

ıhs

ſol

ꞇhur

bıoꝛchi

maꝺhⸯ

laghⸯ

vir

ſolhengeꞇhe

The rune forms are here and there somewhat mangled, but are all clearly of the normal intermediate type, with a long-branch form of s. There are four additional runes: ‘c’ written with what is presumably intended as a short-twig s with a ring at the bottom; ‘e’ which is slightly ambiguous and could either be intended as æ or e with a short horizontal crossbar; ‘p’ which is written as b, where one must assume that the ultimate source had a dot in both loops or at least in the lower one; and finally ‘d’ of the type consisting of a short-twig t with a short horizontal bar across the stem. It is rather unexpected that g, æ and ø are all missing from a fuþark where p is included. The latter two could be explained as not being understood by the author, but g is alongside e the earliest and most frequenty occurring supplemental rune, making its absence quite conspicuous. The distortion of the shapes will be discussed jointly with the shapes of the runic alphabet below.

The values have some oddities. For þ, the value ‘z’ is rather uniquely given; k is given the unusual value ‘q’; and for s the value is written ‘so’. The simplest explanation covering all three of these is that the author combined a source consisting of 1) a fuþark with names, but not values, for the first 17 runes, and values for the remaining supplemental values, and 2) a runic alphabet without explicit sound values but where these were implied from the order of the runes. If the alphabet did not have a sign for ‘z’, but continued with the signs without an equivalent in the original Latin alphabet, the þ-rune would be found in the position where ‘z’ was expected. The k-rune would be found both in the position of ‘k’ and ‘q’, and the latter was chosen either because it was noticed first, or because minor accidental differences in how they were drawn made this a better match than ‘k’. Finally, in such a scenario it would be quite understandable that the writer started copying the rune name instead of just writing the sound value for s.

The rune names are well enough preserved that it is possible to place them within the dialectal variation of Old Norse, despite the orthography being thoroughly confused, especially in the use of the letter ‘h’. This is used in five ways; word-initially as the sound [h] itself, postvocalic to denote vowel length, following ‘c’ to denote its original non-sibilant value, following other normally plosive consonants to denote spirant sounds, and finally spuriously after any consonant without a clear significance. The last two are in conflict, as it neutralises the opposition between Old Norse word-initial /t/ and /θ/, the former here written ‘th’ with a spurious ‘h’, the latter ‘th’ with the ‘h’ denoting spirant pronunciation. The author clearly has no understanding of this contrast, making the same error in the Hebrew and Greek alphabets below. Although originally /tʰ/, Greek theta had long since shifted to /θ/; in Latin it was represented by the digraph ‘th’ based on the ancient pronunciation, and it was precisely this shift in pronunciation that gave rise to the use of ‘h’ as a marker of spirant pronunciation. In the two Greek alphabets, theta it written ‘teta’ without an ‘h’, while tau is written ‘thau’ in the first and ‘tau’ in the second. In Hebrew, the pronunciation of the letter corresponding to Greek theta was not not so straightforward and also changing somewhat over time, but usually rendered as ‘th’ in Latin based on the parallel with Greek and the need to distinguish it from the normal ‘t’. Here the two alphabets in the original hand spells the two names ‘tet’/‘thet’ and ‘thau’. The scribe’s use of ‘h’ is altogether arbitrary, inconsistent and more often wrong than right. In the first rune name it might denote vowel length despite preceding the vowel rather than following it.

The monophtongs in the fifth, sixth and eighth names shows that the names derive from an Old East Norse (OEN) form, ancestral to Old Danish and Old Swedish, rather than Old West Norse (OWN) which is the form usually referred to by the term ‘Old Norse’ when nothing else is specified; often it refers even more narrowly to the normalised dictionary form based on Old Norwegian and Old Icelandic, and usually preferring the older form where these differ. In order to highlight both what are fair representations of the source and what remains more problematic, I have in addition to the normalised name forms and the forms as written in the manuscript added two columns in the table below. Following the classical form of the names is a reconstructed OEN form. As is conventional, vowel length is here marked by macrons rather than with acute accents, but they have the identical meaning. The letters i and þ are used for j and ð which are strictly speaking redundant to mark as separate in OWN as well. Following this is my best guess as to what the author intended to express, following the same convention.

In the discussion of the individual names, comparisons will be made not only to the OWN and OEN forms, but also to the evidence of forms predating this split evidenced in Universiteitsbibliotheek Leiden, Vossianus lat. Q 83 (Leiden), Codex Sangallensis 878 (Abecedarium), The Book of Oghams (Ogham) and the first of the two younger fuþarks in St. John’s College, Oxford, MS 17 (Oxford Y1).

Classical OWN form	Possible OEN form	Presumed intention	Actual spelling
fé	*fē	*fē	fhe	If the oddly placed ‘h’ denotes vowel length, this is a very precise rendering.
úrr	*ūrr	*ur	vr	Length is not expressed neither in the vowel nor the consonant, but their values are correct.
þurs	*þurs	*þors	thors	The vowel is unexpected, but not too dissimilar. A scribe unfamiliar with the language might possibly have heard the short vowel as lower than the long vowel in the preceding name, and interpreted that as a significant difference.
áss/ǫ́ss	*ǭss	*os	os	The peculiar Norse vowel is consistently rendered as ‘o’ in non-Norse sources, and consonant length is rarely marked.
reið	*reþer	*reþer	rether	The first ‘e’ is the monophtongisation of earlier ‘ei’ (phonemically /æi/), the latter is an epenthetic vowel. The final ‘r’ is not original, but introduced by analogy. This development is also seen in Oxford Y1.
kaun	*kon	*kon	chon	The spelling renders the expected monophtongised form, except that the ‘h’ is redundant as a marker of non-sibilant pronunciation of the ‘c’ in front of a back vowel. In Abecedarium the corresponding name is hard to read, but similarly seems to have ‘ch’ in front of an ‘a’, while the alliteration requires a /k/ pronunciation.
hagall	*hagall	*hagald	hagald	The final ‘d’ might be spurious or possibly represent a “hardened” pronunciation of the long final /l/.
nauð(r)	*noþer	*noþer	nodher	The abbreviation sign at the end must be read as ‘er’ based of the form ‘rether’. Unlike in that name, where the final r is added by anology, the same letter is original here, although often lost in Classical OWN. It is preserved in Leiden while Abecedarium is too worn to read at this point and Ogham has a corrupted reading. Oxford Y1 lacks the r, but as it also lacks the ð the name could be corrupted or truncated there. Normalised ð is rendered ‘dh’ here and in the following non-initial cases, contrasting with ‘th’ in the fifth rune. Alongside the erratic use of ‘h’ in general, this strengthens the impression that the scribe had great problems with fricative sounds.
íss	*īss	*īs	ihs	The length of the vowel is denoted by the ‘h’, the length of the consonant is as usual not denoted.
ár	*ār	*ar	ar	The length of the vowel is not denoted. The reason why it is not when it is in the previous name might be due to the consonant also being long there, which might have made the vowel length more salient. On the other hand, vowel length is not denoted before long consonants in the second and fourth runes either, so it might also be rather arbitrary.
sól	*sōl	*sol	sol	The length of the vowel is again not denoted.
týr	*tȳr	*tur	thur	Here the spurious use of ‘h’ nullifies the distinction in initial consonant between this and the third rune of the fuþark. However, the representation of the vowel is the more interesting part. The author clearly has no problem equating a rune or a letter from a fantasy alphabet to ‘y’, but does not associate this letter with a sound value, unlike somebody knowing Old Norse or Old English, and therefore has to use ‘u’ as an approximation for its sound.
bjarkan	*biarkan	*biorki	biorchi	This is the most problematic of all the name forms in this list, with both ‘o’ in the first syllable and ‘i’ in the last being quite baffling. This rune name is the only one which in Classical Old Norse is neither a common noun nor a proper name, but is only recorded as a rune name. It is therefore a possibility that it could at times and places be replaced by a similar word, such as birkinn (“birchen”, “made of birch”). However, the form recorded here could hardly be such a word. The root of the word was originally berk- (with a short e) in Proto-Norse; but a short e would change if followed by a syllable containing i, a or u. Before i, it would change to i by older i-mutation (which had a raising effect rather than the fronting effect of the later “normal” i-mutation). The i in the second syllable would then have to be an original e which changed to i after older i-mutation had taken place, or be a later derivation such as the enclitic definite article. The latter would here have had the form -in, and is in any case a rather late deveopment which we would not expext to see in a rune name at this stage. Before a syllable with a, the e would become ja by breaking, just as in the standard form of the rune name. The third option is breaking caused by a u in the following syllable, which initially yielded jo, but which (at least in OWN) later became jǫ in nouns by analogy with the frequent a:ǫ alternation caused by u-mutation. The root vowel ‘o’ suggests a form of this type, but that would leave no explanation for the final ‘i’.
maðr	*maþer	*maþer	madher	This name shows no anomalities beyond what is described in previous entries.
lǫgr	*lǫǥer	*laǥer	lagher	Classical OWN has the root vowel ǫ here, which is presumably the sound that is rendered as ‘o’ in the fourth rune. However, u-umlaut is the youngest of the Norse umlauts, and was less widespread in OEN than in OWN. Corresponding to the similar OWN word lǫg “law” is Old Danish logh and Old Swedish lagh – the former exhibiting u-umlaut, the latter not. As these examples show, the voiced fricative velar which is spelled g in normalised OWN is typically written with the digraph gh in OEN; I write this with a barred g in the reconstructed forms here.
ýr	*ȳr	*uir	vir	The name is here spelled differently from the coda of the twelfth name, which is phonemically identical. This could of course just be yet another instance of the vagaries of the scribe, but it is tempting to suggest a possible explanation for the exact choice made. At this time, y was a redundant letter without a sound value of its own in the Latin alphabet, but was pronounced as i just like k was a duplicate of c. While the scribe had no reservations in ascribing ‘y’ as the Latin-letter equivalent of runes or letters from fantasy alphabets, he represents the underlying sound as ‘u’ in the name of the t-rune. Doing so here, however, would both render the name identical to that of the u-rune, and would fail to explain the connection to the given value. Therefore the name of the letter y was used: ‘ui’. That this was the name of this letter according to the scribe can be seen clearly in the second Greek alphabet where it is spelled ‘vı’. In the first it is spelled ‘nẏ’, where ‘n’ clearly must be an error for ‘u’ and ‘y’ is equivalent to ‘i’ in the representation of sound.

The alphabet

ꝺ

ꞇ

⫶

ꝺ

Tor Gjerde
i@old.no