How linguistics helps us reconstruct ancient fire mythology

Report on the ‘Martin Burr Fund’ grant offered for a monograph on the Norse God Loki

written by Riccardo Ginevra (Center for Hellenic Studies, Harvard University)

The historical and comparative approach to Indo-European poetic language and myth has developed greatly in the second half of the 20th century, particularly thanks to the efforts of, among others, Rüdiger Schmitt (1967; 1968), Enrico Campanile (1977; 1990) and Calvert Watkins (1995). By means of this methodology, linguists have been able to highlight several parallels between the poetic phraseology, divine names and mythological narratives occurring in texts in the Indo-European languages of oldest attestation, in particular Hittite, Vedic and Epic Sanskrit, Ancient Greek and Latin.

Sources in Old Norse and in the other ancient Germanic languages, however, have been somewhat overlooked by scholars of comparative Indo-European poetics, mainly because, in comparison with, for instance, Hittite and Greek (already attested in the 2nd millennium BCE), these languages were first written down at a relatively late date (the Early Middle Ages). The antiquity of a tradition’s written sources, however, should not be the main criterion in assessing its importance in reconstruction: in contrast, as demonstrated by Calvert Watkins (1995: 414ff.) in his classic discussion of dragon-slaying myths, Germanic traditions often preserve very archaic features of Indo-European poetics and myth. In my opinion, this truth may be even further demonstrated by a linguistic analysis of the names, phraseology, and narratives concerning the mischievous Scandinavian god Loki, a topic to which I have devoted much of my PhD research at the Università per Stranieri di Siena and the University of Cologne between 2015 and 2018.

My interest in Loki actually arose in 2014, when I was still a Master’s student at the Università Cattolica of Milan, writing a thesis on the several parallels between (among others) some Ancient Greek myths involving the god Hermes (the – at least partial – reflex of an ancient fire-god, as already argued by van Berg 2001) and the titan Prometheus (the “Fire Thief” par excellence), and some Norse mythological narratives involving Loki, who had already been analyzed as a fire-deity by Jacob Grimm in his Deutsche Mythologie (1835). Even though Grimm’s interpretation of Loki was later challenged on various grounds (cf. e.g. Liberman 1992:131), the god’s association with fire is supported not only by parallels in Scandinavian folklore (cf. e.g. Heide 2011) and archaeology (cf. Gestsson 1961), but also by the linguistic analysis of his myths both within their Scandinavian context and in comparative Indo-European perspective, as I have proposed in two publications (Ginevra 2018a; 2018b) and as I argue extensively in a corresponding monograph, on which I am currently working and whose preparation has been partially supported by the “Martin Burr Fund” from the Philological Society.

Depiction of Loki from an Icelandic manuscript (16th-century).

As a first example of the inner-Norse and comparative evidence for Loki’s interpretation as a fire-god, I would like to discuss the Norse myth of Loki’s capture and binding (Vǫluspá 35; prose finale of Lokasenna; Gylfaginning 50), whose plot may be briefly summarized as follows: the gods capture Loki and bring him to a cave, where he is chained to three stones and a venomous snake is fixed above him, in such a manner that the reptile’s poison constantly drips into his face; standing by his side, Loki’s wife Sigyn tries to collect the snake’s venom into a basin, but, every time she has to leave her husband’s side in order to pour the liquid away, the snake’s poison drips directly into Loki’s face and the god shakes so violently that the whole earth trembles, causing earthquakes.

Now, according to Norse cosmology, earthquakes were caused by the violent movements of underground fires (cf. Konungs skuggsjá 14 ms. b.,Brenner 1881:30 “You also said that so vast are the fires in the bowels of the land that earthquakes arise out of fire’s violent movements”); furthermore, the poison which torments Loki may be analyzed as a reflex of the Old Norse kenning “poison of the fire” for ‘water’ (cf. Sigvatr Þórðarson, Lausavísa 1.2–4 “snake of the poison of fire” = “snake of water” = “fish”). These considerations allow for the interpretation of Sigyn’s act of collecting and pouring out the poison which drips onto Loki’s face as a mythical representation of the pouring of water or other liquids onto fire, possibly reflecting corresponding pre-Christian Scandinavian fire-rituals which have been reconstructed by Kaliff 2005 on the basis of archaeological evidence.

This is further supported by Sigyn’s name (actually to be scanned as Sígyn, as shown by Skaldic meter), which may be the Old Norse outcome of a Proto-Indo-European formation *sei̯k-én-ih2/sei̯k-n̥-iéh2‘she of the pouring’, a reflex of the root *sei̯k‘pour’ which is widely attested in Vedic Sanskrit in the context of fire-rituals, cf. Rigveda 8.9.7cd “Here he will pour (siñcād – another reflex of the root *sei̯k) the most honeyed soma and the hot milk in the presence of the fire-priest”. The same reconstructed formation *sei̯k-én-ih2may also underlie Vedic °sécanī– ‘pouring’ (upa-sécanī- ‘id.; pouring ladle’ in RV 10.21.02c; 10.105.10a), and it would closely parallel the (Latinized) Celtic river-name and theonym Sēquana (modern-day river Seine in France), which is the reflex of a similar formation *sei̯k-en-eh2 ‘she of the pouring’.

Depiction of the myth of Loki and Sigyn on the Gosforth Cross (mid-11th

As a further example of the comparative evidence for Loki’s identification as the reflex of an Indo-European fire-god, let us take into consideration the Norse myth of Loki’s wager with the dwarf Brokkr (as told in Skáldskaparmál 35): as a punishment for shaving all the hair off the head of Thor’s wife, Loki has to forge beautiful gifts for the gods; after having done so, Loki pledges his own head in a wager with the dwarf-smith Brokkr, challenging him to forge better treasures; when Brokkr succeeds and wins the wager, Loki tries to flee, but is soon captured by Thor, and his lips are finally sewn together by Brokkr.

Several parallels for this narrative may be found in the Sanskrit myth of Fire’s cursing by the seer Bhr̥gu (Mahābhārata 1.5–7; 9.46.12–20), in which the god Fire betrays the trust of Bhr̥gu’s wife and is therefore cursed by the seer to become an “eater of everything”, thus losing control over his own mouth; ashamed, Fire flees, but his hiding place is discovered by the gods, who convince the supreme god Brahma to restrain Fire’s mouth (slightly modifying Bhr̥gu’s curse).

Various phraseological and structural parallels between these Norse and Sanskrit narratives allow for the reconstruction of an Indo-European “Myth of the Mouth of the Fire-god”, in which, after committing a crime against a woman, the fire-god risks losing control over his head or mouth because of a wager or curse; the fire-god flees, but is later found by the other gods, and his mouth is restrained – either literally (Loki) or metaphorically (Fire).

Furthermore, these correspondences allow for the assumption of a direct connection between the Norse god Loki’s rival Brokkr and the Sanskrit god Fire’s opponent Bhr̥gu, whose names may in fact be traced back to *bhr̥g-nó– and *bhŕ̥g-oṷ- respectively, two related formations which are derivatives of one and the same root *(s)bh(h2)g– ‘crackle, roar’, among whose reflexes are, for instance, Homeric σφαραγέομαι ‘crackle, sizzle’ and Vedic Sanskrit sphūrjáyant- ‘crackling, sizzling’ and bhūrjáyant- ‘id.’ (reflexes of *[s]bhr̥h2g-éi̯e- ‘crackle, sizzle’). The possibility of a common etymology for the names of their respective enemies thus supports an identification of Loki as the Norse counterpart to the Sanskrit god Fire.

Depiction of character with his mouth stitched (currently interpreted as a reference to the myth of Loki and Brokkr) on the Snaptun Stone (10th-11th century).

As these two examples suggest, an analysis of Loki’s mythology from a comparative Germanic and Indo-European perspective allows for the identification of several inherited features of religious onomastics, poetic language, and narrative structures in Scandinavian pre-Christian religion and myth, improving our understanding of the meaning and origin of these ancient texts. Thanks to the partial support of the Philological Society’s “Martin Burr Fund”, my book-length discussion of these topics hopefully will, on the one hand, match the late Martin John Burr’s academic interest in Comparative Philology and Old Germanic languages and cultures, and, on the other hand, contribute to the Philological Society’s mission “to investigate and promote the study and knowledge of the structure, the affinities, and the history of languages”.


van Berg, Paul-Louis. 2001. Hermes and Agni: a fire-god in Greece?. In: M. E. Huld, K. Jones-Bley, A. Della Volpe and M. Robbins Dexter, Proceedings of the Twelfth Annual UCLA Indo-European Conference, 189-204. Washington DC.

Brenner, Oscar. 1881. Speculum regale. Ein altnorwegischer Dialog nach Cod. Arnamagn. 243 Fol. B und den ältesten Fragmenten. München.

Campanile, Enrico. 1977. Ricerche di cultura poetica indoeuropea. Pisa

Campanile, Enrico. 1990. La ricostruzione della cultura indoeuropea. Pisa.

Gestsson, Gisli. 1961. Mynd af Loka Laufeyjarsyni. Árbók Hins íslenzka fornleifafélags 58:47–51.

Ginevra, Riccardo. 2018a. Old Norse Brokkr, Sanskrit Bhr̥gu- and PIE *(s)bh(h2)g– ‘crackle, roar’. In: David M. Goldstein, Stephanie W. Jamison, and Brent Vine (eds.), Proceedings of the 28th Annual UCLA Indo-European Conference, 71–93. Bremen.

Ginevra, Riccardo. 2018b. Old Norse Sígyn (*sei̯k-n̥-i̯éh2– ‘she of the pouring’), Vedic °sécanī– ‘pouring’, Celtic Sēquana and PIE *sei̯k ‘pour’. In: David M. Goldstein, Stephanie W. Jamison, and Brent Vine (eds.), Proceedings of the 29th Annual UCLA Indo-European Conference, 65–76. Bremen.

Heide, Eldar. 2011. Loki, the Vätte, and the Ash Lad: A Study Combining Old Scandinavian and Late Material. Viking and Medieval Scandinavia 7.63–106.

Kaliff, Anders. 2005. The Vedic Agni and Scandinavian Fire Rituals. Current Swedish Archaeology 13.77–97.

Liberman, Anatoly. 1992. Snorri and Saxo on Útgarðaloki, with notes on Loki Laufeyjarson’s character, career, and name. In: Carlo Santini (ed.), Saxo Grammaticus. Tra storiografia e letteratura. Bevagna, 27-29 Settembre 1990, 91–158. Roma.

Schmitt, Rüdiger. 1967. Dichtung und Dichtersprache in indogermanischer Zeit.Wiesbaden.

Schmitt, Rüdiger (ed.). 1968. Indogermanische Dichtersprache. Darmstadt.

Watkins, Calvert. 1995. How to Kill a Dragon: Aspects of Indo-European Poetics. New York.

Third International Workshop on Syntactic Cartography (IWSC2019)

Written by Keith Tse (Ronin Institute, New York)

On what has been a regular fixture in the past few years, the biennial Third International Workshop on Syntactic Cartography (IWSC-3) took place on the 26th-27th October 2019 at Beijing Language and Culture University (BLCU) which happened at the end of Open Access Week in China and was preceded by a separate yet related workshop called ‘Changing Boundaries’ on the 25th. This year’s joint event had a special significance as it was the first IWSC to take place at BLCU after the establishment and inauguration of their Linguistics Department which was celebrated in a similar event at the end of October 2018 (to which the author was also invited to present a poster), and these three days were marked by an impressive number of keynote and invited speakers as well as many local and external presenters who assembled from all around the world to take part in what was to be a rich and dynamic academic forum on cutting-edge issues in biolinguistics and formal cartography using data from a wide range of languages. After a competitive round of abstract reviewing, I was fortunate enough to have my abstract on Chinese Voice alternation (my native language) selected for oral presentation, and my presentation dealt with the use of two famous morphemes in Chinese ba (把) and bei (被) which are widely known to involve object preposing. Since Wang (1959), ba– and bei-constructions have been identified as parallel constructions, since in both constructions the object of the main lexical verb seems to be raised from its base-generated position in the lower VP to a higher position (object i … PRO i), the copy of which can be resumed by a coreferential pronoun:

1) subject BA object i verb (PRO i) (Feng (2002:148))

e.g. 李四 把 壞蛋        殺-了         (他)

        Lisi ba huaidan    sha-le        ta

        Lisi BA scoundrel kill-PERF him

        ‘Lisi killed the scoundrel.’ (Huang, Li, Li (2009:153))

2) object i BEI subject verb (PRO i) (Feng (2002:148))

e.g. 張三         被   李四  打-了       (他)

       Zhangsan bei  Lisi  da-le        ta

       Zhangsan BEI Lisi  hit-PERF him

       ‘Zhangsan was hit by Lisi.’ (Huang, Li, Li (2009:112))

Despite the voluminous work that has been done on Chinese ba and bei-constructions (see Li (2006) and Li (1993) respectively), mainstream movement analyses (Tsao (1987), Feng (1995), Huang (1999)) do not adequately account for their empirical complexities, and my new proposal is that ba and bei are light verb projections denoting Voice (Active and Passive respectively), and the fact that these are merged higher than an optional unaccusative marker gei (給) denoting affectedness (3a-b) (Tang (2001), Cao (2012)) suggests that the preposed object may in fact be merged in an A-position, namely the specifier of gei (Kuo (2010)). In light of the fact that bei can be merged higher than ba which is in turn higher than gei (3c), the cartographic arrangement indicates three distinct A-heads above Asp(ect) to which the lexical verb moves (BEI (Passive) – BA (Active) – GEI (Affect)):

3a) 一-把      火    就      把  阿房-宮           給    廢-了

       yi-ba     huo jiu     ba afang-gong    gei  fei-le

       one-CL fire  then BA afang-palace GEI ruin-PERF

      ‘One torchwas enough to ruin A-fang Palace.’ (Chappell and Shi 2016:471))

3b) 杯子  被    他  給   打-破-了

       beizi bei  ta  gei  da-po-le

       cup   BEI he GEI hit-break-PERF

       ‘The cup got broken by him.’ (adapted from Tang (2001:259))

3c) 他  被    朋友         把  一-個      太太    給  騙-走-了

       ta  bei  pengyou ba yi-ge      taitai  gei  pian-zou-le

       he BEI friend      BA one-CL wife   GEI cheat-go-PERF

       ‘He was cheated of a wife by his friend.’ (Chen (2003:1173))

Due to the tightness of time as there were so many presentations that each presentation was only allocated twenty minutes including Q&A, only one question was allowed for my presentation, and it was made by Professor Marcel den Dikken who asked whether it was possible to use ba and bei with set idioms, and if so, whether this would suggest that the object in the idiom could be raised via movement rather than be generated as new arguments by bei/ba/gei, which might pose as a counter-example to my analysis. This reminded me of Li’s (2006) analysis where she does explicitly use phrasal idioms in Mandarin Chinese (e.g. 佔便宜 ‘to take advantage of’, 開刀 ‘to have an operation’, 幽默 ‘to be humorous’, 小便 ‘to have a pee’) to support her movement analysis as all such idioms are permissible in ba– and, by extension, bei-constructions, though she also recognises that there are constraints on ba-constructions as the raised object must have a certain thematic relationship with the lexical verb which is implicit in gei-insertion, namely affectedness, and this also applies to set idioms (他把便宜(給)佔去了 ‘he took advantage of it’, 他把刀(給)開完了 ‘he finished the operation’, 別把默(給)幽壞了 ‘don’t humour badly’, 你趕快把便(給)小了吧 ‘hurry up peeing’). I pursued this discussion with Professor den Dikken afterwards and discussed some of the technical details with him and Professor Ian Roberts whose first book I cited, and these discussions clarified certain technical details in my analysis. As the invited speakers were invited to the dinner banquet, non-invited presenters such as myself returned to our accommodation, and since most of us stayed at the same hotel in the vicinity of BLCU, I was able to say goodbye to most participants and all speakers upon their sober return from the banquet. I held further discussions with Dr. Joseph Perry and Professor Roberts about the nature of IWSC-3 and how impressed we were by this year’s edition, which is a tribute to the local organisers and all the participants, and as I made my way to the airport, I left our capital feeling not only a sense of mission accomplished but also a job well done. 

I would hence like to place special thanks to members of the Philological Society, especially Professor Klaus Fischer and Professor Peter Austin, for accepting my application to the Martin Burr Fund and to the patrons of the Martin Burr Bequest for their generous sponsorship of my participation in such a prestigious international conference at which I was able to share my research with so many distinguished members of our field. I am now in the process of writing this up for the forthcoming publication of the conference proceedings in which I shall express all my ideas with more clarity and purpose.


Cao, D-G. (2012): ‘ “被” 的雙重語法地位和被字句的生成’. Dangdai Yuyanxue 13(1):73-81. 

Chappell, H. and Shi, D-X. (2016): ‘Major Non-Canonical Clause Types: ba, bei and ditransitives’, in Shi, D-X. and Huang, C-H. (eds), A Reference Grammar of Chinese, Cambridge: Cambridge University Press, pp. 451-483. 

Chen, P. (2003): ‘Indefinite determiner introducing definite referent: a special use of yi “one” + classifier in Chinese’. Lingua 113(12):1169-1184. 

Feng, S-L. (1995): ‘管約理論與漢語的被動句 (GB theory and passive sentences in Chinese)’. Zhongguo Yuyanxue Luncong 1:1-28. 

Feng, S-L. (2002): ‘韻律結構與把字句的來源 (Prosodic structure and the origin of ba construction)’. In Triskova, H. (ed), Tone, stress and rhythm in spoken Chinese (Journal of Chinese Linguistic Monographs 17), Hong Kong: Chinese University Press, pp. 119-168.

Huang, C-T. (1999): ‘Chinese Passives in Comparative Perspective’. Tsinghua Journal of Chinese Studies 29(4):423-509. 

Huang, C-T., Li, A., Li, Y-F. (2009): The Syntax of Chinese. Cambridge: Cambridge University Press. 

Kuo, P-J. (2010): ‘Transitivity and the BA construction’. Taiwan Journal of Linguistics 8(1):95-128. 

Li, A. (2006): ‘Chinese Ba’, in Everaert, M. and van Riemsdijk, H. (eds), Blackwell Companion to Syntax: Volume I, pp. 374-468. 

Li, S. (1993): 現代漢語被字句研究 (Xiandai Hanyu Beiziju yanjiu). Beijing: Beijing University Press. 

Tang, S-W. (2001): ‘A complementation approach to Chinese passives and its consequences’. Linguistics 39(2):257-295. 

Tsao, F-F. (1987): ‘A Topic-Comment Approach to the Ba Construction’. Journal of Chinese Linguistics 15:1-54. 

Wang, H. (1959): 把字句和被字句 (Baziju he Beiziju). Shanghai: Shanghai Education Publications.