mannes reord

navigating philological sources in heathen reconstruction

ingwine húnfriþ
harvest mmxxii

Sutton Hoo belt-buckle, 7th c.

lét ðá of bréostum   ðá hé gebolgen wæs
wedergéata léod   word út faran
stearcheort styrmde·   stefn in becóm
heaðotorht hlynnan   under hárne stán·
hete wæs onhréred·   hordweard oncníow
mannes reorde·

«then he loosed from his breast, when he swelled with anger, that man of the weder-géats, words came forth, the stark-hearted one thundered; the voice came in, resounded battle-bright under the gray stone; hated was stirred — the hoard-warden recognized the man’s voice.»

— béowulfbéowulf 2550a-2555a ·mmdl—mmdlv·

🚪 beginning

reconstructing a heathen practice often means a lot of reading & research, & the sources upon which we must rely usually either require a lot of interpretation or are highly technical in nature. the frigg’n heathens podcast have a great episode covering some of the greater challenges with finding sources for heathen reconstruction, as well as some general principles for vetting sources.

since many our primary sources are written in old germanic languages (& those that are not are usually written in latin), we must often rely on the work of philologists in particular to relay information about heathenry to us; for this reason, philological evidence is especially salient for heathens. however, not all philological information is the same, much of it is difficult to interpret (even for experts in the field), & there are often hidden assumptions made when a piece of philological information is applied in a different context from which it originates. moreover, when one encounters philological evidence, it is often not obvious what it means, or could mean, for the practitioner of heathenry.

i have an entire section of my site devoted to sharing my philological articles, some of which contain evidence that might be relevant to heathens. however, the articles therein are approached from an entirely academic point of view, & are written in a more formal style. the goal of this more informal piece is to provide people who are interested in using philological information in their heathen practice with an overview of the different kinds of philological evidence, what difficulties they can present, & how they might be put into practice.

⚓ types of philological information

the corpus of old germanic literature which is relevant to heathens is vast & made up of many disparate pieces of evidence. however, i have attempted to create a rough system for categorizing different types of philological evidence, each of which i will delve into below (along with some examples from an old english heathen perspective). the categories i will explore are:

📜 direct attestations

among the direct attestations we have of heathen sources, the prose writings of figures such as snorri sturluson, the venerable bede, ælfríc of eynsham, & saxo grammaticus are some of the most invaluable. as prose writings, they are also less difficult to translate in such a way that a 21st century audience can still gain as much of an appreciation for the texts as possible without reading them in the original language. however, these texts still pose difficulties because they were not actually penned by heathens themselves, but rather christian authors whose cultural contexts & agendas must be parsed carefully in order to intepret & apply the information in the texts. it is important when being presented with information from these writers to consider these issues, rather than take what they have written about the beliefs of others on blind faith.

aside from the prose writings of these authors, our other main kind of direct attestation is that which we find in poetry. it is important to note that, while many of the earliest dated poems may have been originally composed & recited orally by heathens, all of the extant poetry we have (with the possible exception of some fragments of poetry carved into runestones) was recorded & transmitted to us via christians, & many of these poems show both obvious & dubious evidences of christian insertions or revisions. however, even when christian elements can be confidently determined, there still remain difficulties with interpretation. in addition to being less straightforward than prose in terms of presentation, poetry also uses specialized (& often obscure) language, including words which are difficult to translate.

translators of poetry must constantly weigh the pull of two primary forces on their traductory choices: on the one hand, there is adherence to the literal meaning of the words used in the original poem which the translator aims to transmit to the audience; on the other hand, there is what the translator determines is the intended poetic (usually figurative) meaning which they are, in some part, expected to convey to the audience as well.* available translations of poetry often having a leaning towards preferring one type of meaning over the other, but regardless of which approach is preferred, the real value in a work of poetic translation for the reader who wishes to use these translations as a source for their practice is in the translator’s notes (which are hopefully provided).

*a third force which is also at play when translating poetry is the form of the poetry itself; inasmuch as a poem is an experiential piece of linguistic art, this quality of the original is almost necessarily lost in translation — it is incredibly difficult to completely preserve all the intricate qualities of germanic alliterative verse meter while also adhering to the meaning, whether literal or figurative; some translators may make a partial attempt at preserving alliterative elements, but this always poses the risk of obscuring the meaning.

notes in poetic translations are key to gaining a full appreciation for the poem in translation. if a translator has erred on the side of conveying their estimation of the meaning intended by the original poem, rather than the literal words used, a note can tell the reader what words the original poem used — this can not only let them see what kinds of figurative language were employed (which they may find useful in e.g. prayers or rituals), but also give them the original content which they may end up interpreting differently than the translator did. if a translator has erred on the side of providing a more literal translation of the poem, preserving the figurative uses of vocabulary for the reader to see for themselves, a note can tell the reader about the various ways that other scholars have interpreted that language so as to give the reader some ideas about how they themselves might interpret the text.

it is very important to keep these issues of translation in mind especially when interpreting poetic sources of informations about heathen gods, legendary figures, & beliefs.

ultimately, when presented with any primary source attestation of an entity or concept, whether it be in prose or poetry, the following questions can be very useful:

🔍 allusions

while we do have some direct attestations for heathen beliefs & practices in both prose & poetic sources, much more of our information comes to us in a fragmented state. especially for old english & continental heathens, the greater part of our sources do not come from direct attestations of myths or practices such as we may find for the norse in snorri or saxo, but rather in vague references or allusions to a previously venerated entity, ancient heroic legend, or aspect of belief.

allusions to heathen beliefs range in clarity; on one end of the spectrum, we have something more straightforward such as one of béowulf’s speeches ending with gǽð á wyrd swá hío scel ‘wyrd goes ever as she must’ (454b) which, at the very least, tells us that the poet of béowulf was familiar with the heathen concept of wyrd ‘fate, fortune.’ on the other end of the spectrum, we have the juxtaposition of the poetic description of the narrator’s soul faring forth in the elegiac poem the seafarer & repeated bird imagery & description, which points to a potential belief (or at least poetic concept) of the soul’s ability to travel, & a possible connection with this concept & birds, which might be related distantly to concepts of the soul taking animal shapes (of which we have slightly clearer evidence in the norse tradition, e.g. the fylgja).

both of the above interpretations of poetic allusions are defensible, especially when more evidence is brought to the fore, but it is important to recognize that the latter is much more tenuous than the former, & should therefore require more evidence to support it. unqualified definitive statements such as “there is an old english attestation for an analogue to the norse fylgja” are not accurate representations of the evidence, & a reader should be wary of any philological source which presents information in this way without providing the attestation (at least in their notes).

sometimes allusions can be difficult to discern, especially when they are simply namedrops of a figure we recognize from another source. this is made doubly difficult by the fact that proper names are not capitalized in medieval documents. a good example of this difficulty is the possible mentions of the figure of hell. in a text such as the old english gospel of nicodemus, we at least have context around the mention of séo hell ‘the hell’ which makes it clear that she is a personified figure, but we are more often met with ambiguous mentions of hell, such as béowulf 852b, in which grendel dies in his cave þǽr him hel onféng ‘where hell received him’ — whether hell is a figure or just a place here is up to interpretation. however, the difficulty in determining what is meant here does not preclude it from being evidence of hell as a figure in old english; it just cannot stand on its own as “definitive” evidence, & it is the responsibility of anyone who is sharing it as a source to make that clear.

some key questions to ask when presented with allusions to heathen concepts in primary sources are:

📚 glosses

because the cultures which practice heathenry were for the most part nonliterary cultures, most of the records we have to reconstruct from are written by authors who lived in a context of cultural contact with the latin literary tradition of medieval christianity. as a result, there are some texts from these contexts which were written in both a germanic language & latin; not all of these texts survive, but many of them did serve as sources for the latin-germanic glossaries which early modern philologists compiled before they were lost. these are invaluable resources for contemporary philologists who translate germanic texts, but they can also serve as an indirect source for heathens when utilized responsibly.

the main utility of glosses between a germanic source language & latin for heathens is what they tell us about how certain germanic terms or concepts were viewed in relation to better understood latin terms. for example, the latin (ad)ōrāre ‘to pray [to god]’ is often translated into old english as the reflexive verb gebiddan (made up of biddan ‘to request, ask for’ & the prefix ge- which can indicate a process or result); a reflexive verb is one in which the subject does the action to themself, or where plural subjects engage in an action reciprocrally, e.g. a gospel translation of latin cum ōrātismatthew 6:5 ‘when ye pray’ as ðonne gé éow gebiddon ≈ ‘when ye pray (yourselves).’ the use of a reflexive verb to render the latin verb (which is not reflexive) might indirectly suggest something about how the early english viewed prayer.

an important thing to be aware of when consulting a dictionary for a germanic language is where certain words are attested. dictionaries such as cleasby & vigfússon’s old icelandic dictionary & bosworth & toller’s old english dictionary are excellent because they provide citations for where a given word is attested; this is actually vital information for reconstructionists because it provides the context for a word being used with a particular meaning. when consulting dictionaries such as these, it is key to be aware of when a word is only attested as a gloss (i.e. there is no attestation of the word in a non-translation environment), as it can indicate the degree to which a dictionary can (or cannot) be used as a source for reconstruction.

for example, we find an entry in bosworth-toller for an old english term sǽ-ælfen which is glossed therein as ‘a sea-elf‚ sea nymph’; however, it would be incredibly hasty to posit the sǽ-ælfen as an old english heathen concept on the basis of this dictionary entry alone. when we look closely at the entry, we can see that it is only attested in the plural form sǽælfenne translating latin naiades in thomas wright’s 1857 publication of latin ~ old english vocabularies. thus, rather than serving as evidence of some heathen concept of sea-nymphs*, it actually suggests something different — that, while there was no unique term that was directly equivalent to the latin naiad (instead we find a compound), the germanic term ælfen was considered comparable enough to the latin term for a kind of nymph to serve as a translation for this type of entity, which can tell us something about the meaning of the term ælf that we see in other contexts; this may even be used to inform one’s praxis regarding these entities.

*there is a term wæterælfenne, which is similarly glossed for latin nymphae, that resembles a medical term found in leech book ii for treating an undetermined condition called wæterælfádl (where ádl refers to ‘illness’). however, as hall (2004)the meanings of elf and elves in medieval england, p 116 notes, it is much more like that this is a compound of the otherwise attested ælfádl prefixed with wæter- as we sometimes see in other medical vocabulary to denote ‘fluid, liquid.’

to conclude this section, here are some questions to ask when presented with a dictionary entry as a source for heathenry:

👩‍👧‍👧 cognates

it is at this point relevant to explain the relationship between the old germanic languages that serve as the primary source languages for heathens. germanic languages descend from a common ancestor called proto-germanic, which was one of the languages that descended from proto-indo-european. the germanic languages are divided into three branches: west germanic, north germanic, & east germanic.* these branches are respectively represented by proto-west-germanic, which is unattested but is reconstructed based on its daughter languages; proto-norse, which is reconstructed internally from old norse, but is also attested in some runic inscriptions; & gothic, which is attested chiefly in wulfila’s translation of (much of) the new testament.

*some linguists group west & north germanic into a larger branch called northwest germanic than in turn separated into west & north germanic.

west germanic divided into multiple attested languages on a dialect continuum from low west germanic languages such as old frankish & old saxon on the one hand to old high german & lombardic on the other; old english developed as a descendant of low west germanic dialects which underwent unique changes as they were separated from the continent. the list below is not comprehensive, but organizes the main attested early germanic languages into their respective branches; an asterisk indicates a purely reconstructed language, as opposed to an attested one.

when a word is attested in multiple members of a branch, linguists can reconstruct the common ancestor to the daughter words, & when multiple branches attest to a word, it can be reconstructed for the proto-language of all of the branches. the comparatively large amount of attested germanic texts have allowed germanic philologists to meticulously map out how sounds might have changed over the centuries & uniquely developed in each branch, so much so that it is often possible to predict precisely what form a cognate (i.e. word from the same origin) should take if it exists in another germanic language (excluding analogical changes to a word’s form). for example, knowing only the old english word hláf & gothic hlaifs ‘bread,’ we can predict based on the known patterns of sound change that an old norse cognate would take the shape hleifr (as it indeed does), as it is the expected result of the proto-germanic *hlaibaz which we can reconstruct on the bases of the cognate forms.

cognates can be useful for looking at shared concepts between different germanic languages, as well as for tracking how concepts may have changed with the development of the unique germanic-speaking cultures that formed over the centuries. one danger to be acutely aware of, however, is that just because a word might be cognate between germanic languages (i.e. come from the same proto-germanic root), that does not mean that it necessarily had the same meaning across languages. one example of this is in reflexes of the proto-germanic word *raginą. in old norse, we see regin appearing independently with the meaning ‘gods, ruling powers’ & in compounds such as reginþing ‘great-council’ & ragnasjǫt ‘the heavens (lit. seat of the powers).’ outside of old norse, however, the meaning ‘gods’ is not directly attested*: gothic ragin can mean ‘law,’ ‘task,’ or ‘counsel,’ but not ‘god,’ while cognate old english regn- only appears as an intensifying prefix in words like regnweard ‘great guard’ & regnmeld ‘a solemn proclamation.’

*however, the widely attested personal name reflexes with the element *ragin- (in descendants such as reynald, rayner, reynard) may point to the possibility that it had sacred or divine connotations.

another example can be found in the comparison between the norse term sif & its cognates in other germanic languages. in old norse, sif has a highly specialized meaning: in poetry, it almost exclusively functions as a proper noun referring to the goddess sif, & in prose, it is a plurale tantum (i.e. a word which only appears in the plural) meaning ‘affinity’ or ‘relations through marriage.’ outside of old norse, however, we see its cognates having a much wider range of meaning, from ‘relationship, kinship,’ to ‘peace’ e.g. old english sibb, gothic sibja. thus, while it is not impossible that there is a goddess in other heathen practices analogous to norse sif, there is no evidence of a cognate name having the same specialized meaning or status as a proper noun in any of the other germanic languages.

it is also important to note that, while these germanic languages come from the same source & share many common words, not all words necessarily survived in every daughter language. for example, the old english noun bedu ‘prayer’ & other west germanic cognates can be compared with the gothic bida ‘prayer, request’ to reconstruct a proto-germanic *bedō, but there does not appear to be any north germanic cognate for the word, even though the related verb biðja ‘to request’ does exist in old norse; old norse instead uses bœn ‘prayer’ (which does have a cognate in old english bén, but these are both from an unrelated root).

the above examples should also serve as a demonstration of the difficulty to make definitive claims on the basis of etymology alone. etymology is one of the key tools for heathens to find connections between our disparate sources. looking at the potential roots of an uncertain word is one way to learn about its potential historical meaning, but it does not define its meaning in the absence of context; an argument that a word’s etymology should determine its meaning (over & above extant usage in context) is called an etymological fallacy. etymologies can provide suggestions for how concepts may have been related or changed over time, & they can even provide a reconstructionist with some ideas for what elements of related attestations they may want to include in their own practice, but etymologies are not equivalent to actual attestations of a word. i will return to this issue in the reconstructed words section below.

when presented with germanic cognates as a source for heathen ideas, many of the questions for evaluating glosses still apply, in addition to the following questions:

🎶 formulae

a formula is a poetic device used by the oral composers of germanic alliterative verse to serve as anchors for their own composition, as familiar motifs for listeners, & as metrical fillers or clichés that made the lines more complete. each individual germanic poetic tradition had a plethora of unique formulae that spanned multiple works, but there were also a number of formulae that spanned languages & even branches of the germanic language family. my master’s thesis was focused on investigating these formulae, the raw data for which i have made available here. there is an extent to which some of these formulae can shed light on shared germanic cultural concepts which may have played a role in heathenry. for example, the formula which coordinates god & elves suggests a close connection between these beings, while the formula wherein a guardian of a particular place or group is referred to as a herdsman suggests that this was a common image or device used to express leadership or guardianship in germanic literature.

however, just as with cognates, it is important to be wary of assuming equivalency across germanic languages. for example, there is a west germanic formula present in old saxon (langa hwîla) & old english (lange hwíle) to express long periods of time. while cognates to both of these words exist in old norse, a cognate formula would be virtually unintelligible — old norse hvíla means ‘bed’ (via a semantic shift from ‘period of time’ → ‘time of rest’ → ‘place of rest’).

a shared verse formula as heathen evidence should undergo much of the same scrutiny as a cognate, as well as the following questions:

✒️ reconstructed compounds

as discussed above, it is often possible to very reliably predict what a term in one germanic language would be based on others. many religious reconstructionists use this to their advantage when looking for terminology to use for concepts & entities in their practice. especially given the vast number of attested cognate words & known patterns of compound-word formation in germanic languages, reconstructionists are often emboldened to form neologisms (newly coined terms) based on existing cognate building blocks on the basis on other germanic languages.

sometimes, this has a firm philological basis. there are many cognate compounds between germanic languages; e.g. the old english compound líchama ‘body-covering’ used for the body has cognates in other west germanic languages with similar meaning, as well as in old norse, which might point to a shared view in germanic languages of the body as some form of vessel for other parts of the self. another example of a shared concept via compound is in the old english term hrímceald ‘frost-cold’ which is only found in one old english source (the wanderer), but is also found many times in old norse poems in relation to ettins & to themes of deprivation or separation.

when building a new germanic compound on the model of another germanic language, it is important to ensure that the elements in your compound bear the same desired meanings & connotations. for example, the cognate elements of the attested norse alfa blót ‘sacrifice for the elves’ exist with roughly corresponding meaning in old english, so *ielfa blót would constitute a reasonable compound reconstruction for an old english heathen who would want to use such a term in their practice.

on the other hand, a compound such as *ésatréow built on the direct cognates to icelandic ásatrú (which itself is not an attested old norse term, but a modern icelandic neologism) seems to ignore the fact that old english tréow refers more typically to trust in a particular person or based on a previously made agreement or oath, as well as grace or favor, but is not attested in reference to a particular religious observance; there do exist attested terms for religious practice & reciprocity (e.g. gield) that would probably be better suited for the intended purpose of “*ésatréow” — a cognate is therefore not always the best source for reconstructing new religious vocabulary.

when looking at a new compound that is built based on cognates to an attested compound in a related germanic language, some questions to be thinking about are:

🪄 reconstructed words

in my categorization, a reconstructed word is a root word that is reconstructed (on the basis of germanic cognates) in a language where that word has no evidence of attestation. a word can reasonably be reconstructed in an old germanic language in which it is not otherwise attested when:

a good example of such a reconstruction is the old english word *díepþu ‘depth’ — this word is not attested directly in old english, but it is a viable reconstruction on the basis of:

it is theoretically possible, given the tools of comparative historical linguistics, to reconstruct words that do not meet these criteria, but without them such a reconstruction lacks a good deal of support; it would seem to a philologist like a cognate is being fabricated simply because one is desired. take for example the aforementioned “missing” old norse cognate to old english bedu & gothic bida ‘prayer’: if one, for whatever reason, desired for there to be a cognate to these words in old norse, they could reconstruct the expected outcome of proto-germanic *bedō in old norse as *bjǫð, but such a word is doubly unsupported — it is not attested in norse* with anything like the meaning found in other germanic languages, & there are also no descendants in any surviving north germanic language to suggest that it would have had that meaning, had it existed.

*coincidentally, an old norse word bjǫð actually does exist, but it appears to be from a different (unrelated) root & means ‘flat land, ground,’ attested here.

reconstructionists, especially in non-norse heathen traditions, are prone to making inferences on the basis of norse evidence, which is understandable considering the size & relative accessibility of the norse corpus. however, it is irresponsible, at least from a philological perspective, to assume with no supporting evidence that a reconstructed cognate for a norse word would (a) exist, (b) mean the same thing, & (c) connote the same heathen concept or entity. in other words, the linguistic ability to reconstruct a hypothetical cognate is not philological evidence, & one should be highly skeptical of any source which presents a reconstructed cognate as such without further supporting evidence.

in summary, the questions to consider when being presented with a reconstructed word in an old germanic language are:

🔑 continuing

what i have presented above are issues & questions to take into consideration when encountering philological evidence for heathen beliefs & practices. how these sources can be adapted & incorporated into one’s own reconstructed heathen practice is highly personal, but trying to answer the questions above for yourself when reading a philological source should take you a long way in developing your own understanding of it & how it may or may not resonate with your practices & beliefs.

of course, it is also important to remember that philological sources are not our only window into heathen religion. archeology, folklore, the environment, & personal experience are other equally important sources for how we can reconstruct our heathen practice, & in some cases may override what philological records indicate for some people. for example, while i stated above the philological problems with importing old norse sif into old english as an equivalent goddess with a cognate name sibb, the personal gnosis of some old english heathens has led them to reconstruct a practice of reciprocity with such a goddess on that basis.

in my own practice, i do heavily weigh philological, archeological, & folkloric evidence, but my beliefs are not wholly determined by this information; heathenry tends to be a highly personal religion, so in my practice i try to keep my academic opinion on the evidence clearly delineated from my own intuitions & personal experiences (which may have less historical/archeological/philological backing, but remain true for me nonetheless). pulling from both pools of understanding is an essential part of my blended reconstructionist & revivalist approach.

my hope for this ramble is not to debunk people’s personal praxis or experience, but to provide some form of criteria for heathens (whether they be philologically enclined or not) to consider whenever they are evaluating the philological evidence that's being presented to them on a regular basis.

thank you for reading. férþú hál·

wæflung

gield