Applying a transaction cost perspective to decode viking Scandinavia's earliest recorded value relation: insights from the forsa ring’s runic inscription

ABSTRACT

This article reevaluates the inscription of the Viking-era Forsa Ring, which contains Scandinavia’s oldest extant legal codex. The inscription’s fine reads ‘uksa … auk aura tua’, previously translated as ‘ox … and two öre [silver]’ and interpreted as a payment of both ox and silver, suggesting cumbersome transactions. This study applies a transaction cost perspective and draws on economic, legal and etymological contexts to propose that the fine could be paid with either an ox or two öre silver, not mandatorily both. This reinterpretation positions the Forsa Ring as Scandinavia's earliest documented instance of a value relation. The value of an ox at two öre of silver corresponds to the valuation of an ox at 30 pence in Anglo-Saxon Laws during the same period.

KEYWORDS:

• Viking age
• money
• economic history
• runes
• Scandinavia

Introduction

The Scandinavian Viking Age falls between periods lacking written sources and those abundant with them, thus making it a proto-historical era. The oral traditions of the Viking Age are sometimes captured in runes, which Stefan Brink (2005, p. 117) describes as ‘fossilised (oral) language’.

The Swedish runic Forsa Ring, or ‘Forsaringen’ in native Swedish, is an iron artefact dating back to the late 9th or 10^th century, and originally served as a door handle. It is heralded as Scandinavia's earliest extant legal codex (Ruthström, 2002, p. 15). Traditionally, the Forsa Ring has been understood to mandate fines in both cattle and silver: an ox and two öre of silver for the first violation, with penalties doubling for subsequent offenses. This suggests the artefact does not provide information on the price of oxen in silver öre.

However, to pay both in cattle and silver would be logistically challenging, elevating transaction costs considerably. It would differentiate these penalties from later provincial laws in Scandinavia, which either demanded payment in a single commodity or allowed multiple commodities with set exchange rates.

Upon reevaluating the Forsa Ring, this study offers an alternative interpretation. Applying a transaction cost perspective, which is a common approach in New Institutional Economics (Ögren et al., 2022), this study focuses on how various institutional setups reduced these costs. In Old Norse, the connector ‘auk’ originally represented the conjunctive adverb ‘also’, and only later developed into the additive conjunction ‘and’. The rereading suggests that fines might be settled in oxen, but also in silver as an alternative, with an ox priced at two öre, potentially constituting Scandinavia's earliest extant price record. Assuming that one öre in silver weighed 25–26 grams, equivalent to 6 Islamic mithqals (Sperber, 1996, p. 55), this article shows that the inferred ox-silver price relative aligns with similar data from contemporary European regions, particularly England.

The article is structured as follows: It begins with a review of previous interpretations of the Forsa Ring, followed by an exposition of the economic framework applied. It then explores the linguistic intricacies of the Forsa Ring, assessing the potential interpretations of terms and the functioning of a multi-commodity monetary standard. The subsequent sections compare the Forsa Ring's derived value of an ox with similar European and Nordic sources. The conclusion offers avenues for further investigation. In an appendix, a Bayesian approach is applied as a complement to evaluate the probability that the proposed reinterpretation is correct.

Earlier readings of the forsa ring

The inscription on the Forsa Ring begins as follows, with the first, third, and fifth lines in italic representing the runic text (sourced from the Samnordisk runtextdatabas, Hs 7 $), and the subsequent lines providing an English translation (sourced from Peterson, 2006), word-for-word:

uksa tuiskilan auk aura tuo staf at fursta laki

ox double-worth and öre two staff for first time

uksa tuo auk aura fiura at aþru laki

ox two and öre four for second time

in at þriþia laki uksa fiura auk aura ata staf

but third time ox four and öre eight staff

The exact nature of the violation that resulted in the imposition of this fine remains unclear. Already Bugge (1877, p. 9) argued that the word form ‘tuiskilan’ is odd; it should have been written ‘tuikilan’ without an ‘s’ ending ‘tui’. Ruthström's (1990) attempts to unveil an offence behind the term ‘tuiskilan’ through an alternative reading, not as ‘double-worth’, but instead consisting of ‘(a)t uis (s)kilan’, with one t-rune missing, denoting ‘for the restoration of sanctuary in a valid state’. Following Ruthström, Samnordisk runtextdatabas translates the three lines as:

One ox and two ounces of silver (in fine) to the staff for the restoration of a sanctuary in a valid state for the first time; two oxen and four ounces of silver (in fine) for the second time; but for the third time four oxen and eight ounces of silver (in fine).

A problem with the latter translation is that the ounce was a different unit from the öre. Additionally, although the öre was a silver unit, it did not necessarily indicate fine silver. A peculiarity is also the word order when placing ‘for the restoration of sanctuary in a valid state’ after ‘ox’, and not after ‘and öre two’ or after ‘staff’ in the rune inscription. Even if Ruthström’s criticism of the old translation seems to have been accepted, alternative translations have been suggested (Källström, 2011).

The dominant interpretation of the Forsa Ring suggests that the fine was settled by providing an ox plus two öre. Should this prevalent interpretation be sustained, it would imply that the Forsa Ring fails to shed light on the value relation between silver and oxen.

The dating of the Forsa Ring has sparked considerable debate, influenced by how certain terms are interpreted. Many scholars initially thought that this inscription dated back to the medieval era and was influenced by Christianity. Originally, Bugge (1877) estimated its creation to be in the latter half of the twelth century. Hafström (1981, p. 532) posits a date in the first half of the twelth century. Nevertheless, he presumes the legal worth of an ox to be two öre, drawing from thirteenth-century provincial laws. This assumption is debatable, especially given the significant price increase in that century, as discussed later in this article. Hafström also perceives the ‘double-worth ox’ as twice as valuable as a regular ox. This would require a payment of 6 öre initially, followed by 12 öre for the second instance, and 24 öre for the third.

In the late 1970s, Norwegian runologist Aslak Liestøl (1979) revisited the Forsa Ring runes and proposed that the artefact might hail from the early Viking era. Altering the interpretation of one rune changed the meaning of a word from ‘priesthood’ to ‘people’. This supported the hypothesis that the Forsa Ring was crafted before the advent of Christianity. Its style is reminiscent of the renowned Rök Stone from the 9th-century Sweden, though the Forsa Ring is believed to be slightly more recent.

Carl Löfving (2005; 2010) has maintained that the Forsa Ring is a twelth-century artefact, noting the mention of öre, a currency that did not exist in the Viking period. However, opponents argue that during the Viking era, öre denoted a weight of silver (Brink, 2010), a fact also recorded in 9th-century Anglo-Saxon laws (Thorpe, 1840).

The current consensus leans towards the Forsa Ring being crafted in the 10^th century, likely in the early part of the century (Källström, 2010).

Applying a transaction cost perspective on the forsa ring runes

To thoroughly comprehend the monetary aspects of the Forsa Ring inscription, a transaction cost perspective offers valuable insights. The fundamental concept within institutionalism, encompassing both the ‘old’ and ‘new’ schools, revolves around how institutions handle transaction costs through rules and customary practices. These institutions evolve over time, which generates path dependence (Hodgson, 1998, p. 168). An exemplar of this is Ronald Coase's (1937) theory, which posits that firms exist to streamline transaction costs.

From a transactional cost perspective, the mandatory inclusion of both silver and oxen as forms of payment is perplexing. Consider a scenario where a fine is to be paid in both oxen and silver. If the payer possesses only silver, the payer must first purchase oxen using a portion of the silver, a time-consuming process. Subsequently, the recipient may need to convert the received oxen back into silver, also time consuming. This entire process is avoidably cumbersome. All parties involved can enhance their overall utility if payment is permitted in either oxen or silver.

In legal anthropology, as illustrated in The Cheyenne Way (Llewellyn and Hoebel, 1941), the emphasis is on understanding the internal logic of the legal system under study rather than imposing modern Western legal categories.

A significant contribution of this study is to evaluate the Forsa Ring within the context of the monetary and legal system of the Viking Age and early Middle Ages. This system can be described as a multicommodity or parallel standard, where silver and various commodities were fixed in value relative to each other to allow flexibility in payments of fines (below several examples are given). They both served as units of account, not solely as means of payment. Over time, this system evolved into bimetallism (Redish, 2000), as non-metallic goods ceased to function as units of account. However, as late as in the 17th and 18th centuries, the British territories in North America, frequently allowed the payment of fines and taxes not only in silver but also in what was known as ‘country pay’, which included widely available goods such as tobacco, at predetermined values (Michener, 2011).

In recent decades, economic historians have made significant contributions to the study of bimetallism (Edvinsson, 2012; Redish, 2000). The insights derived from this analysis can also be applied to understanding the monetary system of the Viking Age.

A problem with a multi-commodity standard with a fixed value relation between units of account is its instability, due to relative price fluctuations. This is evident from the experiences of bimetallism (Redish, 2000), where the overvalued currency tend to drive out the undervalued currency. To what extend it is feasible for bad money to replace good money in accordance to Gresham’s Law has been debated, and depends on various contexts (Rolnick and Weber, 1986; Selgin, 1996), for example, the transaction costs associated with utilising the good money at a rate above its face value and laws punishing the circulation of good money at a premium.

In the context of the Vikings, the substantial transaction costs associated with deviating from the established ox-silver relationship ensured the inflexible adherence to the established valuation. Evaluating the worth of either commodity outside this fixed relation would have been unwieldy and inefficient. The breakdown of the fixed multicommodity monetary standard likely only occurred when there was a significant shift in the value relationships, necessitating the establishment of a new value relation or the exclusion of one of the commodities from circulation in a process similar to Gresham’s Law. However, value relationships were likely far more stable in the long term during this period than they are today, due to slower technological change, although short-term fluctuations certainly occurred. A possibility may have existed also to readjust the units of account, for example, by allowing ox of lower quality to be used as a means of payment if the purchasing power of silver decreased, thus ensuring that both silver and oxen could be used in fine payments.

While the term ‘monetarisation’ is typically associated with the introduction of physical coins (Myrberg, 2008, p. 47), it is important to recognise that the concept of a monetary standard should encompass non-state-sanctioned as well as state-sanctioned forms of currency. As highlighted by Kenneth Jonsson (2011), the reevaluation of the Forsa Ring's dating to the Viking Age suggests that the economy of that time might have been more sophisticated than previously assumed. During the period roughly spanning from 880 to 970/90, dirham coins were not subjected to testing, such as pecking, implying that they were likely counted instead of weighed. This practice of counting indicates that the dirham coins served not only as a means of payment but also fulfilled the essential monetary function of units of account. If the öre unit was counted, rather than weighed up, in dirhams, it indicates that it functioned as a monetary unit of account already during the Viking Age, before the rise of domestically state-produced coins.

Similar to early modern bimetallism (Kindleberger 1991, p. 151), parallel use of multiple commodity standards before the rise of domestically state-produced coins, which allowed flexibility in the choice of payment means, may have fulfilled all monetary functions for substantial transactions. Notably, bullion remained a viable payment method for larger transactions in medieval Europe even after the introduction of minted coins (Svensson, 2016, p. 1111). In essence, the pre-coinage multi-commodity standard persisted because, akin to the firm of Ronald Coase, it lowered transaction costs.

In medieval legal documents, the terms often specify that a payment could be fulfilled using one type of currency or a variety of different commodities. In Hälsingelagen (Manhelgdsbalken XXXVIII), which governs the region encompassing Forsa, the wergild is set at ‘7 marks in silver or the equivalent value thereof’ (translated by the author). Similarly, Östgötalagen (Dråpabalken XXI) dictates that a particular fine must be satisfied with ‘six marks in pennies, or three mark homespun … or four cattle’ (translated by the author). Predominantly, Swedish medieval legislation expresses fines in monetary terms, albeit with the implicit understanding that payment can be made in either coins or commodities recognised as valid forms of payment. This approach was not unique to Swedish law; it was a common practice across various Scandinavian legal systems. For example, in the Norwegian Gulatingslov (section 223), the oldest among Scandinavian provincial laws with origins tracing back to the 10^th century, there are intricate guidelines specifying the acceptable forms and equivalents for fine payments:

Now, an account will be provided for the payment. Regarding a cow, it should be valued at two and a half öre. If one is to pay with cows, one should not pay with a cow older than 8 winters, unless the other party accepts it. Payments should be made with cows that are unharmed on horns and tail, on eyes and udder, and on all feet.

Now, one should pay wergild and fines with grain, oxen, and cows that bear calves. One can pay fines with gold or melted silver, if one has it. One can pay with horses, but not with mares; with dapple grey horses, but not with geldings; with any horse that's neither skittish nor dim-eyed, and that does not have a white foreskin or fails to urinate properly, and that has no other faults. One can pay with sheep, but not with goats.

One can pay with ancestral land, but not with purchased land. One can pay with a boat, provided it does not require repairs, and unless it is such an old boat that the first oarlocks on it are worn out. One shouldn't also pay with a boat where the prows are broken off, or one where the planks are patched, unless this was done when the boat was in the shipyard.

One shouldn't pay with any item valued less than an öre unless someone has to receive a lesser fine; then he should accept the payment, unless the fine increases to an öre and the other party, in return, receives a security pledge.

One can pay with weapons that are functional, whole, sturdy, and unharmed. One should not offer the weapons with which the person was killed. One shouldn't pay fines with a sword unless it is adorned with gold or silver.

One can pay with all new broadcloth or linen and clothes; they should be completely new and uncut, unless the other party will accept cut, new cloth. Fines should be paid with men's clothing, not women's; new, not old. One can pay with new and unused trade furs. One can pay with black sheepskin fur and new and uncut attire.

One can pay with all thralls born at home – any that are not younger than 15 winters, unless the other party accepts it. Female thrall should not be included in the wergild.

The quote illustrates the flexibility of payment under the combined precious metal and non-metallic standard. It demonstrates the significant discrepancy between an interpretation of the Forsa Ring that necessitates payment in exact amounts of two distinct goods and the established practices within later Scandinavian legal systems. Applying a transactional cost perspective to this scenario highlights the concept of path dependence, demonstrating how later legal structures can provide critical insights into and shape our understanding of earlier practices, thereby informing the ongoing development of legal and institutional frameworks.

When a payment required a combination of two different units, it was typically due to one unit acting as a fractional component of the other, facilitating the expression of partial amounts. For instance, rather than specifying a payment as 1.25 marks, it would be articulated as 1 mark and 2 öre, given the equivalence of 1 mark to 8 öre. In Norway in the 1020s, individuals intending to relocate to Iceland were required to remit payment of either six cloaks and six ells of home-spun fabric or half a mark of silver (Diplomatarium Islandicum, 1857-76, pp. 65–66). Nonetheless, it is possible to interpret the ell of home-spun as a fractional component of a cloak. One cloak is stated to be equivalent of 12 ells of homespun in Icelandic Grágás (section 246). In this light, the addition of six ells of home-spun to the six cloaks not only fulfills the requirement but also equates the first payment option to the second one, which is 0.5 marks (4 öre) of silver, necessitating the use of cloak fractions.

Conversely, the silver öre did not directly function as a subunit of an ox in the same way. The two represented distinct categories of means of payment: cattle as opposed to silver. Had the öre been intended to act as a fractional value of an ox, we would anticipate its elimination through the doubling phenomena evident in the Forsa Ring. For instance, if two öre equated to half the value of an ox, then a combination of two oxen and four öre should logically equate to five oxen, while four oxen plus eight öre should tally up to ten oxen. However, this is not the pattern we observe in the Forsa Ring, indicating a different relationship between öre and oxen.

Nevertheless, in contrast to fines, rents and taxes were sometimes settled through a variety of commodities. For example, in 1277, peasants in Södermanland, Sweden, were required to remit 20 tön of grain and a monetary payment in pennies equivalent to the grain’s market value (Franzén, Hedegård, and Svensson, 2020, p. 52). This dual-payment system could have served as a strategic measure to mitigate the risks associated with fluctuating grain prices. Taxes were typically linked to the production or income levels of taxpayers, allowing for advanced planning, whereas fines were imposed as one-time payments. The diverse nature of goods or wealth held by those subject to fines necessitated the flexibility to accept various payment methods. To bolster the perceived fairness and acceptance of the fine payment system, it could have been favourable to ensure its adaptability to diverse household units, a factor important in maintaining the institution's legitimacy and sustainability.

The monetary meaning of ‘auk’

When approaching the original Old Norse text of the Forsa Ring, it is crucial to recognise that translation should consider the broader institutional and economic context in which it was composed. Legal customs and language usage are integral components of a community's collective identity (Ögren et al., 2022, p. 180). Language places constraints on individuals similarly as other institutions.

Understanding the Forsa Ring as requiring both an ox and two öre implies interpreting ‘auk’ as ‘plus’ in modern terms. A modern equivalent for an either-or scenario would instead be the disjunctive conjunction ‘or’. However, linguistic practices in oral societies varied from those in written-centric cultures (Brink, 2005). This difference does not imply irrationality or lack of abstract logical thinking of ancient societies. In oral traditions, communication heavily relied on immediate context and non-verbal cues, embedding meaning in the situation at hand. In contrast, written societies necessitate embedding context within the text itself, as the physical and immediate social context is removed or distant, leading to a backgrounding of additional information and potential for misunderstanding if the text is not sufficiently clear (Denny, 1991). In long-term perspective, language practice has evolved from high contextualisation to decontextualisation.

In the earliest written provincial laws of mainland Sweden (excluding Scania and Gotland) from the thirteenth century (for example, Östgötalagen, Dråpabalken XXI), the term ‘ælla’, meaning ‘or’, was employed to denote payments made with either one good or another. However, ‘ælla’ is notably absent from any runestones (Peterson, 2006), although it is important to consider that runestones, serving specific commemorative and legal purposes, might not capture the full extent of everyday linguistic practices (Jesch, 2001, p. 36). The frequent appearance of ‘ælla’ in thirteenth century provincial laws may suggest a later addition to the Swedish language, potentially influenced by the Christianisation of Scandinavia. Another term for ‘or’ is the Old Norse ‘eða’, which still exists in Icelandic, and was used in the thirteenth century provincial Gutnish Law, but not in mainland Swedish provincial laws. During the Viking Age, ‘eða’ was used in two Danish rune stones (DR 209 and DR 230), but is absent from the Swedish runestones (Peterson, 2006), despite that the number of runestones in mainland Sweden is much larger than in Denmark. Even if ‘eða’ likely have been used in mainland Sweden during the Viking Age, this absence might provide some indication of its relative usage compared to ‘auk’.

The term ‘auk’ or ‘ok’ is frequently encountered in runic inscriptions, typically denoting the additive conjunction ‘and’, a word that later evolved into the Swedish ‘och’. However, in a few instances, it could also signify the conjunctive adverb ‘also’ (Peterson, 2006), which was its original meaning in Old Norse (Willson, 2017, p. 529). The Swedish words ‘ock’ and ‘också’, meaning ‘also’, likewise trace their roots back to the term ‘auk’.

When viewed as an additive conjunction, ‘auk’ operates much like ‘and’ in the English phrase ‘apples and oranges’. It links two items or ideas, integrating them into the same list or sequence. The word ‘and’ suggests a sense of equality between the connected elements. For instance, when we say ‘apples and oranges’, both fruits are presented as equally relevant within the context. Neither fruit is inherently emphasised or more significant than the other; they share an equivalent standing. In the context of our runic example, if we interpret ‘auk’ as ‘and’, both the ‘ox’ and the ‘two öre’ would be regarded as equally essential components of the payment. This implies that the complete payment comprises both elements, with neither overshadowing the other in significance.

In contrast to conjunctions, adverbs often serve to modify or qualify the statements they are part of, giving more semantic content. When ‘also’ is used, it is not just a mere connector like ‘and’. Instead, it often introduces an additional element, which might be seen as supplementary or secondary to the primary statement. It serves as a so-called conjunctive adverb (Braunmüller, 1978, p. 100), modifying clauses rather than verbs, adjectives, or other adverbs. Both conjunctions and conjunctive adverbs are connectors. Saying ‘I want apples, also oranges’, might imply that while apples are the primary desire, oranges would be a nice addition but are not the main focus.

In the context of the Forsa Ring, if ‘auk’ is interpreted as ‘also’, then the ‘two öre’ might be perceived as a supplementary statement to the phrase ‘ox’. The ox remains the main method of payment, but there is an additional, perhaps alternative, option in the form of the two öre. This interpretation shifts the weight of importance between the two payment methods, emphasising the ox over the öre. Interpreting the Forsa Ring as a payment of ‘an ox, also two öre’ would entail that the Forsa Ring is not stating that payment must be made of an ox plus two öre, but rather of an ox that also is equivalent to two öre. In reality, payment could likely be made in other goods as well as long as the value of these goods was equivalent to an ox, which, in turn, was priced two öre.

Ancient languages typically feature an abundance of tangible and concrete terminology while lacking in abstract expressions (Lohisse, 1979, p. 75). This reflects the higher level of contextualisation in oral societies. In the initial phases of language development, the application of conjunctions is frequently observed to be less exact compared to their usage in more mature stages of the language (Nielsen, 2017, p. 237). A plausible interpretation posits that terms like ‘auk’ or ‘ok’ could potentially be remnants or frozen instances of an older form, originating from the proto-Germanic noun ‘*auka’, which directly translates to ‘addition’ (Willson, 2017, p. 529). Examining the linguistic lineage, we see this concept echoed in the Old Norse verb ‘auka’, denoting an increase (Kroonen, 2013, p. 42), a nuance preserved in the contemporary Swedish term ‘öka’. The connection to ‘addition’ and ‘increase’ naturally imbues the connective ‘auk’ with an adverbial shade, subtly implying that something is in the process of being supplemented or enhanced. Illustrating this, the compound term ‘aukanafn’ directly translates to ‘additional name’.

The evolution of the Old Norse ‘auka’ or ‘ok’, that later became the Swedish ‘och’, can be described as the grammaticalisation, involving a process of conventionalising ordinary phrases into conjunctions (Nielsen, 2017, p. 237). Grammaticalisation is a linguistic phenomenon in which words that once had specific lexical meanings evolve into grammatical elements. This process, which is institutional, typically involves a shift from a more concrete meaning to a more abstract, grammatical function. It denotes a loss of meaning, so called ‘semantic bleaching’, a kind of decontextualisation. Compared to ‘auk,’ the etymological origins of ‘ælla’ and ‘eða’ remain more uncertain. However, there is a prevailing likelihood that these two words also evolved from adverbial forms, similar to ‘auk’ (Teleman, 2012, p. 65).

It is possible that the Old Norse ‘auk’ term fulfilled several functions in the language (Braunmüller, 1978, p. 114), before a differentiation took place.

Nevertheless, even in modern language, the word ‘and’ can convey various meanings. As Nielsen (2017, p. 240) aptly notes, ‘coordination in the linguistic sense does not always equate to a straightforward addition in its strict mathematical sense’. The meaning can also encompass a natural sequence. For example, in the phrase ‘morning and evening’, it does not imply that morning and evening happen simultaneously. Similarly, in the case of a law addressing payments in ox and silver, it may allude to a sequence of distinct payments involving the two commodities, rather than requiring both to be used in a single transaction.

The field of monetary terminology is rife with historical examples that highlight the potential for confusion between ‘and’ and ‘or’, emphasising the need for contextualising phrases. According to Svenska Akademins Ordbok, n.d. ‘and’ as conjunction could sometimes mean ‘or’. Three examples can illustrate this:

1. Svenska Akademins Ordbok mentions the expression ‘one barrel of rye is valid as 10 and 12 gyllen’ from 1624. Here is not meant that a barrel of rye is paid 10 plus 12 gyllen, but either 10 or 12 gyllen. While it still may be correct to use the conjunction ‘and’ from the point of view or recording two or more income flows that have taken place, from the point of view of just one transaction, it is either 10 or 12 gyllen that is paid. A similar situation occurs for the payment of fines that the Forsa Ring was intended to describe.

2. An earlier example where ‘and’ means ‘or’ can be found in the thirteenth-century Västgötalagen (Rättlösabalken 12, manuscript B 59). The law stipulates a fine when rented animals perish of ‘half a mark for ox and mare’ (translated from Holmbäck and Wessén, 1946, p. 113). However, the intended meaning is that a half-mark is to be paid for either an ox or a mare that perish, given that it would be unprobable for law statements to be formulated for the more unlikely event that both one ox and one mare perish at one time. It is noteworthy considering that ‘or’, manifested as ‘ælla’ in the lexicon of Västgötalagen, was a readily available option in the language of the time.

3. Another example where ‘and’ means ‘or’ is the quoted passage from the Norwegian Gulatinglov in the previous section stating that ‘Now, one should pay wergild and fines with grain, oxen, and cows that bear calves. One can pay fines with gold or melted silver, if one has it.’ The first sentence entails not that payments of fines must include grains, oxen as well as cows for the same transaction, but rather grains, oxen or cows. The second sentence instead use ‘or’ that the fines could be paid in either gold or silver, which serves the same function as ‘and’ in the previous sentence.

The existence of varied interpretations of an abstract term, such as ‘auk’, may pose less of a challenge in an oral society, where communication predominantly transpires in a face-to-face manner, facilitating immediate clarification (Denny, 1991, p. 72; Nygaard, 2021, p. 157). In this context, the subtleties and nuances of language can be better understood through direct interaction and non-verbal cues.

The semantics of the parallel silver and non-metallic monetary standard

Money serves different functions. At some time during antiquity it became common to value various goods and slaves in oxen. There were no oxen used in these transactions, and there could be no oxen in sight. Everybody knew that this was rather an abstract unit of account (Henstra, 2000, p. 4). Although grains could potentially also be used as units of account, the relative prices of these commodities usually fluctuated significantly from year to year. Cattle and precious metals were more stable in value, therefore better fulfilling the monetary function of unit of account in contracts. Nevertheless, long-term shifts, even sudden shocks, occurred that disrupted the system. The function of money as a unit of account must, in turn, be distinguished from its function as means of payment (Edvinsson, 2012). While the current interpretation of the Forsa Ring treats ox and silver solely as means of payment, it is highly likely that the two likely functioned as units of account.

In a multi-commodity standard, there are several units of account with a fixed relation between them. They are used interchangeably (Redish, 2000), although each unit can sometimes be divided into separate subunits (such as one öre into three örtug). For example, it would be appropriate to say that payment could be made in either oxen or öre when referring to the actual means of payment utilised. However, when considering oxen and öre as units of account with a fixed relationship to each other, a payment is represented in both oxen and öre, even if the transaction physically involves only one of them. The importance of cattle and silver may even point towards a parallel silver and non-metallic monetary standard as a variant of the multi-commodity standard. This is also termed ‘commodity-money system’ (Skre, 2011, p. 69).

Although not a multi-commodity standard, we would not state that payment could be made in either two dollars or 200 cents, since two dollars are 200 cents. The payment of two dollars is also a payment of 200 cents. It is payment of two dollars, and also a payment of 200 cents, but not two dollars plus 200 cents, i.e. of 4 dollars. When speaking of dollar and cents, we are not referring to the concrete coins that are used, but to the unit of account (dollar) and its subdivision (cent). Considering this context, when translating the first sentence of the Forsa Ring, the term ‘also’ is more correct logically than ‘or’ or ‘and’ considering the framework of the combined silver and non-metallic monetary standard, where one ox equals two öre silver.

An interesting aspect of the Forsa Ring is that in the first line quoted above, the ox according to an older translation is termed ‘uksa tuiskilan’, normalised as ‘oxa tvisgildr’, i.e. double-worth ox. Hafström (1981) argues that there was an ox that was worth double a common ox, or even a cow, but such high-value oxen were probably rare. It is likely that the law used a standard measure of an ox. In Anglo-Saxon law, the ox was valued at one third more than a cow (Thorpe, 1840, pp. 98 and 151).

In king Edward’s law from early 10^th century, a distinction is made between ‘twelf-hynde man’ and ‘twy-hynde man’. The first man was worth 1200 shillings in ‘wer-gild’, i.e. blood money, and the second 200 shillings, which differentiated between the highest class of Anglo-Saxon aristocracy and the lowest class of freemen (Thorpe, 1840, p. 75):

A twelve-hynde man’s wer is twelve hundred shillings.

A twy-hynde man’s wer is two hundred shillings.

A possibility to consider is that ‘tuiskilan’, if the older translation would be valid, refers to the value of two öre. If this interpretation is correct, which may not be the most likely one given the new translation by Ruthström (1990), the valuation of an ox as two öre appears twice in different forms in the same formulation of the Forsa Ring. Such repetitions may seem unnecessary, but could be interpreted in its institutional context. The authors behind Forsa Ring may have wanted to ensure that a standard ox valued two öre were to be paid, not a worse one. Similar apparent tautologies, ‘something valid as x monetary units is equivalent to x monetary units’, can be observed in the quote from King Edward’s law and later Swedish provincial laws. Pre-modern agents were well aware that cows and oxen were of different qualities. Often a standard was used for an ox or a cow, and then the actual cattle was compared to it (an example is Grágás 246).

A direct link between ox and monetary units also existed in other languages at the time. In 10^th century northern Iberia, a standardised cow and ox, ‘vacca soldare’ and ‘bos solidare’, the one solidus cow and the one solidus ox, were used as units of account alongside other monetary units, despite that oxen were valued at 4–10 solidus in silver (Davies, 2002, p. 169). It is possible that in northern Iberia, the cow or ox were originally valued at one solidus gold coins, and that these units continued to exist despite the debasement of the solidus as a unit of account.

The European context

The hypothesis that an ox was the equivalent of two öre silver in Viking-Age Scandinavia calls for a comparison with similar prices in related market conditions. For example, if the price of an ox would be six öre in Iceland, but only two öre in Norway, and the cost of transporting the ox to Iceland would be two öre, with no other costs and obstacles, oxen would be traded on a massive scale from Norway to Iceland until the price there drops towards four öre. Given the Vikings’ extensive travels and their proficiency in cattle transportation, some level of such price integration appears likely. For instance, The Saga of Erik the Red chronicles the Vikings’ establishment of a colony in ‘Vinland’ (now known as present-day America), where they transported a variety of cattle on their ships. A bull from their stock instilled terror in the local ‘skrælingar’ population, culminating in lethal clashes.

Table 1 lists various prices for an ox, valued in öre of 25–26 gram silver weight, sourced from non-Nordic regions, and includes some derived from other animals and animal products. Subsequently, Table 2 in the following section provides data from Nordic countries. Collectively, these tables convincingly affirm that the reinterpretation proposed in this study is empirically realistic.

Table 1. Estimated prices of an ox in öre of 25.5 gram (6 mithqals) from various sources in non-Nordic countries.

Year	Place	Actual price relation	Assumptions in this table	Ox in öre	Source
301	Rome	Pair of best oxen = 10000 d., cow = 2000 d., pound gold = 72000 d.	1 ox of standard quality = 1.33 cows. 1 g Au = 12 g Ar. Pound = 329 g.	6	Edictum Diocletiani
452	Rome	solidi = 200 pounds pork	1 ox = 230 pounds pork. 1 g Au = 12 g Ar. Solidi = 4.5 g	2	Spufford, 1988, p. 8
499, famine	Rome	1 kg meat = 2.5 folles	1 ox = 230 pound meat. 1 solidi = 360 folles. Solidi = 4.5 g Au. 1 g. Au = 12 g.	3	Spufford, 1988, p. 9
7^th century	Co­logne	ox = 2 solidi, cow = 3 solidi	Solidi = 1.35 Au. 1 g Au = 12 g Ar. 1 ox = 1.33 cows.	1-2	Lex Ribvaria Tit. XXXVLXI
7^th and 8^th cen­turies	Ireland	Milch cow = 24 scruples, 4-calf cow = 30 scruples, 5-calf cow = 40 scruples, 6-calf cow = 50 scruples	1 ox = 4- to 6-calf cow. 1 scruple = 1/24 ounce = 1.15 g Ar.	1-2	Davies, 2020, p. 177, Naismith, 2017, p. 371
796	North Iberia	ox = solidi + tremisses	Solidi [Byzantian] = 4.5 g Au. 1 g Au = 12 g Ar	2	Manzano and Canto, 2020, p. 177
c. 800	Saxons	calv 1 year old = 2 tremiss	1 tremiss = 0.45 g Ag. 1 g Au = 12 g Ar. 1 ox = 4-5 calves 1 year old	2	Lex Saxonum Tit. XIX
10th century	Bulgur on Kama	Murten fur = 2.5 dirham	1 dirham = 2.9 g Ar. 1 ox = 6 murten furs	2	Gieysztor, 1987, p. 482.
10th century	Eng­land	ox = 30 pence	1 öre = 16 pence	2	Thorpe, 1840, pp. 98 and 151
10th century	Byzan­tium	Cattle priced at 3 nomismata	1 nomismata = 4.5 g Au. ox = cattle. 1 g Au = 12 g Ar.	7	Morrisson and Cheynet, 2002, p. 839
939	North Iberia	ox = 5 solidi	Solidi = 8 arienzos = 8 dirhams of 2.7 g Ar [uncertain]	4	Manzano and Canto, 2020, p. 184
991	North Iberia	3 oxen = 18 solidi 7 arienzos	Solidi = 8 arienzos = 8 dirhams of 2.7 g Ar [uncertain]	5	Manzano and Canto, 2020, p. 180
1124-1153	Scotland	cow = 48 English pence	144 sterlings = 8 öre silver. ox = 1.33 cows	4	Leges inter Brettos et Scottos
1155/56	Byzan­tium	cow = 1 nomisma	1 nomisma = 3.4 g Au. 1 g Au = 12 g Ar. 1 ox = 1.33 cows.	2	Morrisson and Cheynet, 2002, p. 839
1162-98	England	ox = 41 pence (average)	144 sterlings = 8 öre silver	2	Farmer, 1956, p. 41
1200-10	England	ox = 78 pence (average)	144 sterlings = 8 öre silver	4	Farmer, 1956, p. 41

Note: The assumptions are mainly based on the same or similar source materials.

Table 2. Estimated prices of an ox in öre of 25.5 gram from various sources in Nordic countries.

Year	Place	Actual price relation	Assumptions in this table	Ox in öre	Source
c. 900–950	Hälsing­land	1 oxen auk 2 öre	‘auk’ means ‘also’. Öre = 25.5 g Ar.	2	Reinterpretation of the Forsa Ring
1020s	Norway	6 cloaks + 6 ells home­spun = 0.5 mark silver [4 öre silver]	1 ox = 90 ells of homespun. 1 cloak = 12 ells of homespun.	5	Diplomatarium Islandicum, 1857-76, pp. 65-66.
eleventh century	Iceland	1 öre silver = 4 öre homespun.	1 öre vadmal = 6 ells of homespun. 1 ox = 90 ells of homespun.	4	Grágás 245
1096	Iceland	1 mark silver = 360 ells of homespun	1 ox = 90 ells of homespun	2	Diplomatarium Islandicum, 1857-76, p. 97.
twelth century	Ice­land	1 ox = 90 ells home­spun. 1 öre silver = 45 ells home­spun		2	Diplomatarium Islandicum, 1857-76, p. 316.
c. 1290	Öster­götland	4 cattle = 6 mark penningar = 288 ells homespun	1 mark silver = 3 mark penningar. Ox = 90/72 cattle = 90 ells of homespun.	5	Östgötalagen, Dråpabalken XXI
1295	Uppland	1 cow = 12 öre	1 öre silver = 3 öre. 1 ox = 1.33 cows.	5	Franzén, Hedegård and Svensson, 2020, p. 130. Svenskt Diplomatariums Huvudkartotek (n.d.) 1616.
1340	Scania	1 ox = 28 öre	1 öre silver = 4.5 Scania öre	6	Svenskt Diplomatariums Huvudkartotek 4582
1503	Stock­holm	1 ox = 32 öre	1 öre silver = 12.77 öre	3	Almquist and Hildebrand, 1915, p. 77
1730s	Sweden	1 ox = 16.5 daler silvermynt (average)	1 riksdaler = 3 daler silvermynt = 25 g Ar.	6	Jörberg, 1972, p. 647

Note: The assumptions are mainly based on the same or similar source materials.

In drawing conclusions about price relations from ancient legislations, caution is paramount. Yet, Jarrick and Wallenberg Bondesson (2016, p. 177) have highlighted that laws from even starkly different cultures can be more comparable than many scholars might initially assume. Numerous value relationships are embedded within laws and decrees. The dates, units of measurements, and relationship to market prices are frequently not clear. As highlighted by Charpentier (2022, p. 167) and Horsnæs, Moesgaard, and Märcher (2018, p. 246), there is a nuanced distinction between established norms and actual practices.

When payments are made in goods or money, the goods’ value might be underestimated relative to market prices. This discrepancy arises due to the additional costs associated with converting goods into money or other commodities, which the receiver bears. Payers are inclined to settle debts with goods of the lowest feasible quality. Tax values, which tend to be systematically lower than market prices, reflect this tendency even though they might closely track market trends. The valuations assigned for compensations in cases of animal or slave fatalities could either surpass or fall short of their market price. For instance, the Lex Visigothorum (Book VIII, Title IV, XVI) mandates compensating a slave’s death with two slaves.

It is crucial to note that transcription errors are an ever-present risk, as exemplified by the scribe of the Anglo-Saxon Chronicle's omission of a crucial numeral in a sequence (Wareham, 2012, p. 918).

It is known that during times and in countries where use of metallic money was more prevalent amplified access to the most productive mines through conquest and trade, prices in silver or gold were high (Edvinsson and Söderberg, 2011). During the era of the Roman Empire, a high level of monetisation was upheld, which diminished after the decline and fall of the Western Roman Empire. A resurgence possibly occurred in the 9th century but was particularly expedited at the end of the 10th century with new discoveries of silver (Spufford, 1988; Pirenne, 1956), peaking in the eleventh century. This was followed by a diminished level of monetisation in the twelth century. In the thirteenth century, a price revolution transpired, inflating prices expressed in silver 2–3 times in most countries. English silver prices swiftly doubled in a few years around 1200 (see Table 1). After the Black Death, silver prices decreased, and a bullion famine ensued in the fifteenth century, reversing in the sixteenth century with the subsequent silver inflation period.

In Homeric Greece, cattle served as a unit of account, alongside silver and gold (Graeber, 2011, p. 417). One of the earliest valuations of an ox appears in the Iliad, where an ox is the 2nd prize in a foot race, and a half-talent is the 3rd prize. This suggests that an ox was valued at a talent of gold, equivalent to 8.5 grams, which with a gold-silver value ratio of 12 to 1 implies that an ox was valued at 4 öre, i.e. double that of the Forsa Ring.

According to the Edictum Diocletiani from 301 A.D., the price of a cow was 2000 denarii, equivalent to 9 grams of gold, given that 72,000 denarii equalled a Roman pound (328.9 grams) of gold. However, there are indications that by AD 340 the purchasing power of gold may have doubled (Howgego, 1995, pp. 140–141).

An indication of low prices in the 7^th and 8^th centuries comes from Irish laws, where a cow was valued at 1–2 ounces of silver. The ounce was slightly heavier than the öre.

A less clear-cut example dates from the 7^th century, courtesy of Lex Ribvaria, which assessed an ox at the value of two solidi. The complexity here is rooted in the notable devaluation of the solidi during that era; the precise value of the solidi cited in this legal document is ambiguous. The most credible conjecture posits that the combined weight of the two solidi at the time amounted to a mere 2.7 grams of gold. Nonetheless, it is conceivable that the ox was originally exchanged for two solidi, the total gold content of which was 9 grams.

The estimated valuation of an ox at two öre in silver demonstrates a remarkable alignment with records from England, shedding light on this aspect especially considering the substantial Viking presence in the British Isles at that time. During the 10th century in England, an ox was listed at 30 pence, and a cow at 24 pence. With the öre being equivalent to 16 pence, this converts to a price of approximately 1.875 öre for an ox. However, the equivalence of the öre’s value in England to that in Sweden remains uncertain. Intriguingly, more reliable records from the period of 1162–1199 in England reveal an average ox price of 41 pence (equivalent to around 60 gram of silver). This indicates a slight inflation in ox prices compared to the 10th century, although the pence experienced a mild devaluation in silver content over that duration.

Prices remained elevated in the Byzantine Empire well into the year 1000, serving as a continuation of the Roman Empire and representing the most advanced region in Europe during the early Middle Ages. Specifically, in the 10th century, cattle held a substantial value of 3 numismati, equivalently 3 solidi, which translates to 13.5 grams of gold. This valuation notably surpassed the levels observed during antiquity and was considerably higher than the prices inferred from the Forsa Ring. The Vikings, engaged in extensive long-distance trade, an endeavour that would only yield profitability under conditions of significant price disparities. Nevertheless, by the time the twelth century rolled around, the Byzantine price of a cow had diminished to a mere one nomismata, aligning with the price levels seen in England and the Forsa Ring.

Price records from northern Iberia suggest a possible price revolution in the 10th century. However, these records are challenging to interpret due to the significant uncertainty surrounding the units used, as no domestic coins were minted at the time. It is likely that the unit of solidi shifted from representing a gold coin to its devalued equivalence in silver (Manzano and Canto, 2020). In 796, an ox was valued at one solidus and a tremisses (the latter, one third of a solidus). By the 10th century, this value had risen to 5–6 solidi. While the solidus of 796 may have contained 4.5 grams of gold, the 10th-century solidus likely represented an accounting unit for 8 arienzos. The arienzos could be the Umayyad Andalusi dirham, weighing 2.7 grams in silver and imported from southern Iberia. This translates to a price of 4–5 silver öre for an ox. Higher prices are expected later in the 10th century, but the high prices as early as 939 might reflect better access to silver through the import of dirham coins.

The nordic context

Table 2 provides a detailed examination of the estimated value of an ox, expressed in öre for Nordic countries. The Forsa Ring record is the oldest.

As discussed above, the fee instituted in the 1020s for those wishing to emigrate from Norway to Iceland entailed that 78 ells of homespun was valued half a mark of silver. Assuming an ox was valued 90 ells of homespun, as indicated by other sources, one can deduce that the going rate for an ox at the time was roughly 5 öre. This was higher than that recorded in the Forsa Ring, but also higher than in the twelth century sources. Interestingly, there is evidence from another Icelandic source that suggests a high price for an ox, although climatic conditions may have been different in Iceland for cattle breeding.

One cannot discuss the economics of the era without mentioning the significant influx of silver into Scandinavia during the late 10th and early 11^th centuries. This was largely a result of the hefty Danegeld payments, although not all of it went to Scandinavia (Jones, 1991, p. 597). This newfound wealth and economic strength led to a monumental event in Swedish history: the minting of the first domestic coins in 995 AD. After a few decades of coin minting, the practice halted in the 1030s. The subsequent century saw Sweden's economy adapt to this change, likely by resorting to units of account detached from silver and gold. It is conceivable that the Swedish economy, during this period, mirrored the Icelandic model where non-metallic goods like homespun cloth and cattle became standard units of account and trade, as indicated by the later Östgötalagen.

The twelth century bore witness to a curious economic trend. The sparse records from this century suggest that prices, in general, were substantially lower than in the surrounding 11^th and 13^th centuries. In Iceland, for instance, an ox's calculated value dropped to two öre. This valuation was consistent with the prices in England during the same period.

According to the Saga of Magnus Erlingsson (King of Norway in 1161–1184), the Archbishop demanded that the farmers give him silver-valued öre in the fines that were due to him. Previously, he had received öre in goods, which were counted in the king's fines. The difference between these öre was that the öre he wanted was twice as valuable as the other. This account signals, if imagining that one öre in goods was originally equal to one silver-valued öre, that the purchasing power of silver doubled compared to goods.

The valuation of an ox in two aura/öre may also give hints about the evolution of the worth of precious metals in the centuries up to the Viking Age. The term ‘aura’ probably originally referred to the late Roman solidi, which was the successor to the gold coin aureus.

Ruthström (1993, p. 98) argues that the öre cannot have its etymological root in the Roman solidus, given that the latter weighed 4.5 grams, which would imply the value equivalent to 36 gram silver with a gold-silver value ratio of 8:1. However, in monetary history, it is often emphasised that some monetary units become fossilised, as ‘ghost money’, losing their original meaning due to the process when bad money usurps the name of better money (Redish, 2000; Sargent and Velde, 2002; Spufford, 1988). For example, the daler was originally minted as a better coin, and was worth 4 mark in common coins in Sweden towards the late sixteenth century. 4 marks therefore was also termed a daler, but during the seventeenth century mark lost in value relative the daler, and the daler came to mean 4 marks and not the daler coin (Edvinsson, 2012). The original daler coin was retermed riksdaler. A similar evolution took place with the term solidi in Western Europe, from being a gold coin to a unit of account for silver coins or bullion of lower value. The öre likely went through such a depreciation after the decline and fall of the Roman Empire. Angela Redish (2000, p. 33) points out ‘that undervaluation in a bimetallic standard could be removed either by depreciating the undervalued metal or by appreciating the overvalued metal. Yet in virtually all instances it was the former that occurred’.

Although, given the dating to the 10th century, it is known that the ‘aura’ in the Forsa Ring referred to a weight of 25–27 gram in silver, the inscription could also be interpreted as valuing an ox in two solidi. Interestingly, in Lex Ribvaria an ox is valued two solidi. A hypothesis could be that, in Scandinavia, sometime in the 5^th or 6^th centuries an ox was fixed in value to as much as two solidi in gold coins, but that gold later became more expensive so that no more fines were paid in gold. Due to the workings of Gresham’s Law, gold was driven out from circulation. During the 5^th and 6^th century solidi coins were quite common in Sweden, while during the 7^th and 8^th centuries precious metal almost disappeared (Skre, 2017, p. 280). One hypothesis could be that the unit of öre lived on through oxen as ghost money, until it got its equivalence in silver, although at a lower value than the original two solidi cold coins, reflecting the increased purchasing power of precious metal after the fall of Western Roman Empire. That this could occur is evident from north Iberia where the terms ‘vacca soldare’ and ‘bos solidare’ were used.

Conclusions

This article gives considerable attention to the semantic nuances of the connective ‘auk’ in the Forsa Ring. If ‘auk’ is construed as an additive conjunction, it implies that fines had to be paid in both oxen and silver. Yet, drawing from its etymological lineage, if ‘auk’ is interpreted as the conjunctive adverb ‘also’, it opens the door to the potential of fines being settled through either oxen or silver exclusively, indicating that an ox was valued at two öre for the purpose of paying a fine. However, even the additive conjunction could sometimes denote ‘or’ in specific contexts. While value relations differed from relative prices, the system that permitted various payment methods to achieve its objective required the fixed value between silver and oxen to align with their relative prices, at least during certain periods.

Should the conclusions drawn in the present study withstand rigorous scrutiny, the ramifications for Scandinavian and European monetary history could be multifaceted:

1. The valuation of an ox at two öre in 10th century Sweden finds parallels with contemporaneous valuations in England, and similarly with those observed in both Scandinavia and England in the twelth century. Nonetheless, the 11^th and 13^th centuries might have been characterised by elevated prices. Such movements suggest prospective price revolutions circa 950–1050 and again in the thirteenth century, bearing resemblances to the seismic shifts of the sixteenth-century price revolution. Concomitantly, each of these pivotal junctures was marked by a significant augmentation in silver influx and coin circulation, arguably both a consequence and a catalyst of profound institutional metamorphoses.

2. The congruence in oxen valuation observed between Northern Sweden and 10th-century England is notably significant. It suggests the existence of a profoundly interwoven European institutional architecture of monetary relations that bridged considerable geographical expanses.

3. The premise that one could reconcile a fine with an ox, also alternatively, with two öre, drawing a subtle equivalence between their values, unveils a dual silver and non-metallic monetary standard. This system recognises oxen and silver not just as mediums of exchange but as units of account, even before coin minting commenced. The actual inherent flexibility of diverse payment modalities media, as evidenced in later provincial Scandinavian legal frameworks, served to diminish transactional complexities and lighten economic duress on individuals in an economy that had not yet converged on a single medium of exchange.

4. An exploration into etymology suggests that the two öre valuation for an ox might have roots in the historical values of two solidi designated for cattle during the late Roman era and its aftermath. The öre might have transformed from a gold- to an oxen-related unit, eventually transitioning into a silver metric. This transition mirrors the path of solidi in various other countries. The significant rise in the purchasing power of precious metals reflect the socio-economic transformations in Western and Northern Europe following the collapse of the Roman Empire. This shift led to a decline in the utilisation of precious metals as currency, possibly influenced in part by the dynamics of Gresham’s Law. Such an observation hints at an extensive pre-history in Scandinavia regarding the function of money as a unit of account.

It is hoped that subsequent research will delve deeper into these propositions. This study seeks to highlight the potential of transaction-cost reasoning in understanding the economic intricacies of proto-historical eras. Archeology can be enriched by economic theory and formalised methodology, opening avenues for more interdisciplinary research into the annals of early economic history.

Disclosure statement

No potential conflict of interest was reported by the author(s).