RŪNA RŪNŌS – Semantic shift and polysemy of the word ‘rune’ from Proto-Germanic to modern examples.
Submitted for publication to the debut edition of Norse Zine.
Introduction
Glossary:
PIE – Proto Indo European PG – Proto Germanic PWG – Proto West Germanic PC – Proto Celtic
The word ’rune’ across various languages carries a notably broad set of definitions, despite being most commonly considered as the name for the various Germanic alphabets employed sometime around or before the 1st century AD to the medieval period, before largely going extinct. The reason for this endonym being employed for runic writing in Germanic languages and thus its ’meaning’ beyond the arbitrary have challenged scholars for centuries, with the latest proposed definition posited by Mees (2014), which reconstructs it as ’counsel’, as a noun from which the many more abstract reflexes derive. This report seeks to investigate the plausibility of this proposition, what it may tell us about the muddled entry of the word into Germanic, and how the legacy of the word continues into modern usage today crosslinguistically.
Research Questions
The present report seeks to examine 3 facets of polysemy regarding the lexeme in question.
1: How well supported is Mees’ (2014) semantic reconstruction for the Proto-Germanic *rūnō?
2: Can a certain mode of directionality be favoured between Celtic and Germanic from the PIE root of the word?
3: Are there any modern uses still employed, or new usages arising for the word or derivations of it in modern languages?
Definition
Fig. 1 Polysemy Chart
The above chart adheres to the PIE origin of *rūnō as determined by Orel (2003).
Although most famously appearing in Old Norse sources to denote the individual letters of the Younger Futhark alphabet, the word rune across variants is attested in writing centuries before Old Norse sound shifts even arose. The two earliest attestations on the record appear with the 4th century Einang stone (Proto-Norse) and the fourth century Gothic edition of the bible, commonly known as Wulfila's bible, which attests the word around 369 CE (Miller, 2019). It is impossible to date which example is first, but they are rough contemporaries. Related terms in the Gothic Bible with affixes are used as translations of 'beguiling' and 'counsel' with a prefix denoting completion.
Corinthians 13:2
𐌾𐌰𐌷 𐌾𐌰𐌱𐌰𐌹 𐌷𐌰𐌱𐌰𐌿 𐍀𐍂𐌰𐌿𐍆𐌴𐍄𐌾𐌰𐌽𐍃 𐌾𐌰𐌷 𐍅𐌹𐍄𐌾𐌰𐌿 𐌰𐌻𐌻𐌰𐌹𐌶𐌴 𐍂𐌿𐌽𐍉𐍃 𐌾𐌰𐌷 𐌰𐌻𐌻 𐌺𐌿𐌽𐌸𐌹 𐌾𐌰𐌷 𐌷𐌰𐌱𐌰𐌿 𐌰𐌻𐌻𐌰 𐌲𐌰𐌻𐌰𐌿𐌱𐌴𐌹𐌽, 𐍃𐍅𐌰𐍃𐍅𐌴 𐍆𐌰𐌹𐍂𐌲𐌿𐌽𐌾𐌰 𐌼𐌹𐌸𐍃𐌰𐍄𐌾𐌰𐌿, 𐌹𐌸 𐍆𐍂𐌹𐌰𐌸𐍅𐌰 𐌽𐌹 𐌷𐌰𐌱𐌰𐌿, 𐌽𐌹 𐍅𐌰𐌹𐌷𐍄𐍃 𐌹𐌼.
jah jabai habau praufētjans jah witjau allaizē runōs jah all kunþi jah habau alla galaubein, swaswē fairgunja miþsatjau, iþ friaþwa ni habau, ni waihts im.
And though I have the gift of prophecy, and understand all mysteries, and all knowledge; and though I have all faith, so that I could remove mountains, and have not charity, I am nothing. (KJV)
Einang Runestone
(Ek go)ðagastiz rūnō faihido.
'I, Goðagastiz carved the rune'.
Here the inscription refers to itself in the singular form, indicating that however the semantic nuances may be understood, they constitute the whole entity and not individual letters.
Mees (2014) seeks to reconstruct the definition of the word *rūnō in proto Germanic, the ancestor of English rune, most often applied to the runic alphabets utilised mostly by Germanic language speakers between at least the first and fourteenth centuries CE. Although beholden to the issue of context making semantic reconstructions somewhat imperfect by default, Mees does provide arguments as to why other, radically different proposed originating subject matters are not supported.
Mees contends that a magic-centered original definition is disputed by the banal nature of early inscriptions, although this is arguably itself contended by Tacitus' allusion to wood strip runic magic carvings used by Germanic tribes for divination, which would not be preserved in the record (Tacitus et al., 1847). On the contrary, Tacitus' account does not satisfactorily specify what the symbols are, although later Eddic poetry may support such a position, such as in the Old Norse poem Voluspá wherein the god Hœnir is said to ‘choose wooden slips for prophecy’ (Larrington, 2014).
Across the scholarly record, several proposed semantic definitions for the PG word have contrasted with Mees’. An opposing proposed etymology relating to 'scratch/dig' proposed by Morris (1985) can be disputed on the basis of reflexes and loans denoting speech, such as Latvian runa (‘speech’), though this example is not included in figure 1 as its date and source of acquisition is particularly poorly understood. Mees' definition is supported over Morris’ by this, as well as by the existence of the related PG ablaut form *raunō, whose Old Frisian and Icelandic descendants point to a definition of 'examination' (Orel, 2003). Whilst not highly adjacent to counsel, it is wholly unrelated to scribal practice. No reflexes of any 'scratch/dig' definition exist anywhere on the corpus of descendent words (see fig. 1).
Rather, a definition broadly pertaining to some form of discrete disclosure appears plausible. Assuming that the word does derive from PIE *h₃rewH, it stands in concurrence with the ‘concreteness effect’ proposed by Lievers et al. (2021) that an abstract noun more likely derives from a concrete one than vice versa, in this case a concept of counsel deriving from a word relating to speech. Later reflexes stray further into more abstract territory of mysteries, whilst also often maintaining an additional definition relating to spoken word. In this respect Mees’ conclusion is a plausible middle ground, certainly compared to any proposed meaning of ‘dig’ or ‘scratch’, but is not without potential criticism. For example, Mees does not include a secondary definition of ‘whisper’, although one seems highly plausible and even likely given the broken continuity between PIE and PWG words relating to speech (see PWG ‘whisper’). It would require fewer assumptions to believe that some definition of spoken word or whisper (compare Latin rumor) was maintained through PG than the alternative, that spoken language disappeared and reappeared as a semantic meaning before and after the PG period. So, while Mees’ ascribed meaning may itself not be problematic, it may also be overly myopic to only assign one.
It may also be questioned whether it is realistic to narrow a semantic reconstruction down to the level of a definition when semantics are far less exact in the regularity of their changes than phonological shifts. Some elements of nuance to the consideration given to the word may have died off before the East Germanic split, rendering Mees’ reconstruction overly reductive. It is, regardless, strengthened by the existence of the derived term *garūnô in PG, translated as ‘counsellor’. With rūnō as the root form, ‘counsel’ becomes an especially appropriate analogue. ‘Counsel’ is also attributed to the Old Norse word on the basis of the related attested term rúni (counsellor, friend) (Orel, 2003, p310), and to establish continuity with examples such as Old Swedish rūna (counsel).
This possible loss can be understood as a semantic analogue to word forms which had neither been attested or reconstructed being discovered in archaeological finds. The recently discovered 5th century Vindelev bracteate in Denmark gives the oldest attestation of the name of the god Odin, albeit not in a form ever reconstructed to that point. The attested wodnaz contradicts the typically reconstructed PN *wodinaz (Imer & Vasshus, 2023). No previously attested evidence was able to predict this instance until it was directly observed. Whether this is a dialectal variant, scribal idiosyncrasy or indication of an otherwise unlikely North/West Germanic loan will likely be the subject of debate in future scholarship, but regardless it stands to highlight the limited power of the comparative method. With semantics being less regular, traceable and predictable than morphology, the possibility of lost or otherwise unreconstructible ancient semantic meanings must be considered at least as much, if not more so as with morphological forms. As a result, any 2,000 year old reconstruction with no attested usages for hundreds of years thereafter must be considered with a high probability of carrying semantic nuance which is simply no longer available to extrapolate.
The abundance of the association with the definition of ‘mystery’ may also strike some viewers of figure 1. The conspicuous lack of such an association in Mees’ PG reconstruction may appear jarring, especially if one is to believe that the Germanic word was loaned into Celtic (as will be discussed later). Such an event would require the ‘mystery’ definition to arise in East Germanic, be loaned into Celtic and travel west back into PWG. This is a highly dubious assumption to make, far more so than simply assuming that ‘mystery’ was an associated definition in PG.
Mees does not explicitly tackle this question pre-emptively, but the focus on connecting it to the name of the written inscriptions may stand as the reason. ‘Counsel’ carries far less opaque connotations than ‘mystery’ and the latter would indicate that the runic inscriptions were named after a primary function of *not* being read, rather than one of disclosing information. Whilst literacy rates in runes (Pollington, 2016) were certainly low in the early Germanic period, it may be deemed unlikely that the inscriptions, whose earliest examples consist largely of names and innocuous indicators, were primarily seen as impenetrable secrets. However, this does not disqualify the possibility of multiple associated meanings of the same word, as rune went on to possess widely in later reflexes.
In summary, Mees’ proposition of ‘counsel’ is itself an adequate semantic association with PG *rūnō, but not an all-encompassing one that satisfactorily covers all likely meanings the word appears to have possessed, based on its apparent immediate loans and later reflexes, as well as the word from which it originated. While ‘counsel’ may have been a semantic definition, even the most common one and that most prominent when applied to the name of the new writing system, it appears that other concurrent meanings such as ‘mystery’ and ‘whisper’ were also likely in use alongside this.
Fig. 2 Semantic Map
The above diagram displays the most commonly occurring definitions across the corpus employed for the creation of figure 1. Note that this is a rough overview designed to briefly demonstrate the most pervasive recurring usages, rather than a comprehensive representation of the entire lexical item corpus, as many boundaries between languages and time periods are arbitrary for the sake of polysemic mapping.
The map follows the parameters laid out by Georgakopoulos & Polis (2021). Connective grey lines vary in width to indicate co-occurrence within a single language, with the width increasing for multiple instances within the myopic corpus gathered for figure 1. For example, whisper and counsel co-occur once, in PWG *garūnō. The size of each circle corresponds to the number of instances of a single definition, topped here by runic letter with five. The results indicate that runic letter sees the most usages, co-occurring frequently with utterance, as mystery does with secret.
Directionality
Past research has not concluded on whether the word entered Proto-Germanic through Proto-Celtic, vice versa, or whether both simply inherited the word independently from PIE. This is due to only the morphology being considered, which does not definitively rule out any of these three possibilities. However, as mentioned previously with regard to Lievers et al.’s (2021) ’concreteness effect’, some patterns of directionality appear regarding abstract nouns more often deriving from concrete nouns than the inverse. One should also consider the path of development which involves the least tumultuous series of changes. As an analogue, consider the polysemy of the English word bug. With a meaning denoting insects dating to the early 17th century, the word nowadays may also denote a craze or trend (as in, ’by the end of the decade, the US had caught the Brit-pop bug’) or an illness (as in, ’she came down with a stomach bug yesterday’). As with the Germanic and Celtic reflexes of *h₃rewH, one may wish to determine if one of these modern definitions came to be through the other, or if they both developed entirely separately from the original word meaning insect. Upon consideration, a most likely path of directionality can be assumed by way of ‘insect’ > ‘illness’ > ‘social craze’. Vermin are associated with and often seen to cause disease, and social trends spread virally in a manner akin to a malady. The definition also gradually becomes more abstract. The alternative of ‘insect’ > ‘social craze’ > ‘illness’ is comparatively problematic and requires more assumptions. The definition changes from concrete noun to abstract noun back to a more concrete definition, violating Lievers et al.’s (2021) concreteness effect. There is also no intuitive transition between meanings, as there is little to no metaphor regarding insects to be garnered from the idea of a social trend. Thus, the former hypothesis must be accepted as the most likely.
With this in consideration, figure 1. indicates that the most likely course of events is a loan from Proto-Germanic into Proto-Celtic, wherein PIE > PG > PC involved a single shift from a broad semantic meaning relating to spoken word, to abstract concepts such as mysteries. This path assumes the least amount of change, as opposed to PC > PG wherein an inherited PIE word pertaining to speech would have to lose all such connotations and then regain them upon entering Germanic. This seems highly unlikely. Since none of the later Celtic reflexes involve speech, it also stands as unlikely that a dual meaning of speech and mystery was preserved through inheritance from Germanic. Dual meaning in PIE is also unlikely, as none of the PIE reflexes in other languages such as Greek or Sanskrit involve abstract concepts comparable to the Celtic examples. Thus, a lineage to Celtic by way of a Germanic loan can be assumed with a fairly high degree of certainty.
Modern and New Usage
In English, the word entered twice through various roots of origin, such as with the word ’robot’. The West Slavic word robota (cognate with German arbeit with metathesis) was first loaned into English through German as robot sometime in the early modern period, then denoting an uncountable system of serfdom, before becoming obsolete by the time the word was again loaned from Czech in the 1920s to describe automatons. In both the case of robot and rune, the later loan greatly overshadowed the use of the older usage. The Old Norse loan rún to modern English rune, denoting runic letters, is by far the most commonly used today. Use of the native West Germanic roun is largely extinct in English, except for in modern Scottish English where it is still employed to mean ’whisper’ by some speakers. The Oxford-English Dictionary records a usage meaning ’secret’ last in 1796. As for whether any new definitions have arisen, the word appears to be garnering a semantic broadening to any sigils, symbols or unreadable/unfamiliar letters, particularly those associated with magic. This appears to be spurred on by the inclusion of such characters across fantasy media. Recent examples include the video games Doom (2016) and Elden Ring (2022), both of which denote their esoteric sigils as runes.
In Swedish, the word is commonly used in reference to runic letters, but also has a more recently recorded definition meaning ’obituary’. Although formerly posited by the author as a potential loan from the Finnish runo (’poem’) rather than a direct Germanic inheritance due to semantic overlap therein, it is more likely another Germanic inheretence stemming from poetic usage. The Finnish word was acquired from Proto-Norse rūnō (’runic inscription’, as a singular whole rather than the individual letters denoted by later Norse usage).
In insular Celtic languages, the word developed notably unique meanings such as ’virtue/essence’ (Welsh rhin) and ’darling’ or ’secret intention’ (Irish rún), possibly stemming from a shared development denoting something personal or dear to someone. The latter is commonly known in Irish-speaking circles from the folk song siúl a rún, although that usage of the word in spoken Irish appears to be fully extinct. It does, however, still survive in daily usage through the word rúnaí (’secretary’), coined as a derivation from the connotation of ’secret’, possibly as a semantic borrowing from English.
Welsh also employs an apparent semantic borrowing from English in modern spoken usage, as evidenced from a forum post (Reddit) utilising rhin fanila for vanilla essence, transplating the human ’essence’ previously indicated by the word to an inanimate ’essence’ used in cooking. Welsh historically also employed a defintion of rhin relating to sorcery or enchantment, first attested in the 14th century, although the origin of this is unclear. GPC (a frequently utilised Welsh Dictionary online resource in philology) records usage of this definition until 1932, although most examples on the site, even for words in modern usage, rarely surpass this period. EDIL – the equivalent resource for Irish, also records a usage pertaining to consciousness or esoteric knowledge of a personal nature. It does seem to corroborate an underlying connective framework of ’that which is privately dear to someone’, but no single English word satisfactorily serves as an analogue.
Conclusion
Whilst morphological and semantic reconstruction of the word ’rune’ back to Proto-Germanic are conclusive, semantic associations are notoriously slippery to pin down, and are far more at the mercy of the indescribable black box of human intention that words exist to serve. Mees (2014) makes solid progress in eliminating unlikely past proposals for the definition of *rūnō, though it does not satisfactorily offer concurrent alternate meanings to its reconstruction, resulting in a definition that likely fit many usages of the word in PG, but not all. The likelihood of a PG usage meaning ’whisper’, ’mystery’ or both is not covered by this reconstruction, even if it does befit the root of the derived form *garūnô (counsellor). Nevertheless, it is a far better supported result by the body of evidence than previous standing publications such as Morris (1985) which sought to answer the same question. Perhaps a focus on a singular ‘original meaning’ is overly prioritised in pursuit of tackling this question, and it ought to be considered that multiple definitions may always be entertained as a possibility, as language from the past was in no way simpler or less varied in its nuance than language today.
Whilst not proven and requiring a significantly more elaborate research project to test conclusions of such, the present research indicates that the question of the directionality between PIE *h₃rewH, PG *rūnō and PC *rūnā may be answerable through observing semantic values rather than morphology. The fact that the PIE and PG examples share a demonstrable association with spoken word, compared to the Celtic examples which have none of such, indicate a loan into PC from PG, with a shift from a concrete noun into an abstract one in accordance with the patterns of directionality explored by Lievers et al. (2021).
Use of the word in surviving reflexes across languages outside of the definition of runic letters is somewhat sparse. Irish no longer uses the word in standard speech except for derivations. Welsh still employs it, with the most semantic deviation from early usage, using rhin to denote both animate and inanimate essence, as well as older definitions shared with the likes of Breton rin. English regularly uses the word to refer to various kinds of esoteric symbols and alphabets, and Scottish English still occasionally employs the native Germanic roun to mean whisper. There is some innovation of the word to be found, although many long standing usages, such as mystery, appear extinct in modern spoken language. Lastly, Latvian still employs it as a very common lexeme for speech, where it has also been grammaticalised into a verb (Konstantīns, 1992).
References
eDIL - https://dil.ie/search?q=r%C3%BAn&search_in=headword
Georgakopoulos, T., & Polis, S. (2021). Lexical diachronic semantic maps: Mapping the evolution of time-related lexemes. Journal of Historical Linguistics, 11(3), 367-420.
Imer, L., & Vasshus, K. S. K. (2023). Lost in transition: The runic bracteates from the Vindelev hoard. NOWELE, 76(1), 60-99. https://doi.org/10.1075/nowele.00074.ime
Konstantīns, K. (1992) “runāt”, in Latviešu Etimoloģijas Vārdnīca[1] (in Latvian), Rīga: AVOTS
Larrington, C. (2014). The Poetic Edda: Translated with an introduction and notes. Oxford University Press.
Mees, B. (2014). The etymology of rune. Beiträge zur Geschichte der deutschen Sprache und Literatur, 136(4), 527-537. https://doi.org/10.1515/bgsl-2014-0046
Morris, R., L. (19850: Northwest Germanic rūn- <rune>. A case of homonymy with Gothic rūna <mystery>, Beiträge zur Geschichte der deutschen Sprache und Literatur 107, pp. 344-358.
Miller, D. Gary (2019). The Oxford Gothic Grammar. Oxford University Press. doi:10.1093/oso/9780198813590.001.0001. ISBN 9780198813590.
Orel, Vladimir (2003) A Handbook of Germanic Etymology, Leiden: page 299, 310
Pollington, S. (2016). Runes: Literacy in the Germanic Iron Age. Anglia Publishing.
Reddit - https://www.reddit.com/r/learnwelsh/comments/11bk8g8/geirfa_ddefnyddiol_feunyddiol_daily_useful/?rdt=47610
Strik Lievers, F., Bolognesi, M. & Winter, B. (2021). The linguistic dimensions of concrete and abstract concepts: lexical category, morphological structure, countability, and etymology. Cognitive Linguistics, 32(4), 641-670. https://doi.org/10.1515/cog-2021-0007
Tacitus, C. & Boetticher, W., Anthon, C., (1847) The Germania and Agricola of Tacitus, with English notes, critical and explanatory, from the best and latest authorities.