Annotated Heritage: Runestones and Old Swedish

The runestones of central Sweden — several thousand of them, concentrated in the province of Uppland and raised mostly in the 11th century — are the oldest substantial body of written Swedish. They are also, for a learner, the most revealing, because they show the language at a stage that feels almost foreign: a language with a full case system, where the role of a noun in the sentence was marked by its ending, not by its position. Modern Swedish has shed all of that. Reading a runestone is therefore a study in subtraction — you see, in the carved endings, exactly the machinery that the modern language gave up, and you understand why modern Swedish has to lean so hard on word order instead. We work through one representative stone.

A note on what you are reading

Real runestone inscriptions are written in the younger futhark, a 16-rune alphabet, run together with few word divisions and many spelling shortcuts. To read them, scholars first transliterate the runes into Latin letters, then normalise the result into standard Old Swedish / Old Norse spelling. The stone below is not a photograph of one specific monument; it is a representative, normalised transliteration of the standard Uppland memorial formula, built so that each word can be glossed cleanly. The formula itself — "X had the stone raised in memory of Y" — is utterly standard and appears on hundreds of genuine stones (for instance Uppland inscriptions such as U 448, Ígull ok Bjǫrn létu reisa stein eptir Þorstein, fǫður sinn). Treat the example as a faithful composite of the type, not as a quotation of a single rock.

The inscription

Here is the representative inscription, first in its transliterated/normalised Old Swedish form, then as a modern Swedish rendering.

Two letters in that line will be new even to an advanced learner of modern Swedish: þ (thorn) and the related ð (eth), plus the runic ʀ. We meet them as we go.

The memorial formula, word by word

The whole inscription is a single fixed sentence type — the memorial formula — and almost every genuine stone is a variation on it. Parsing it word by word against modern Swedish is the fastest way to see the lost grammar.

Sven — the sponsor (NOMINATIVE)

Sven is the man who had the stone raised — the grammatical subject, and so it stands in the nominative case, the case of the subject. On many stones the nominative is visible in an ending (a name like Holmgeirr, Ígull); here the short name Sven happens not to show it overtly. The point to hold onto: the job of being the subject is, in principle, marked by case, not merely by coming first. In modern Swedish, Sven is just Sven and we know he is the subject because he sits before the verb — position has taken over the work that case once did.

lét — the verb "had (done)" (older form)

lét is the preteritum (past tense) of láta ("to let, to cause, to have something done"), here "(he) caused / had." This is the engine of the whole formula: the sponsor does not raise the stone with his own hands; he has it raised, lét reisa. The modern reflex is lät ("let, had"), preteritum of låta — you can see the older é where modern Swedish has ä. When the sponsors are plural, the older language even marks number on this verb: one sponsor lét, two or more létu (as in Ígull ok Bjǫrn létu..., "Ígull and Bjǫrn had..."). That number agreement on the verb — singular lét vs plural létu — is itself something modern Swedish has lost; today it is lät whether one person or ten.

reisa stein — "raise the stone" (and the ACCUSATIVE)

reisa is the infinitive "to raise" (modern resa), governed by lét: lét reisa = "had (it) raised." Its object is stein (modern sten, "stone") — and here is the first clear case ending in action. stein is the direct object of reisa, and so stands in the accusative case. In Old Swedish, masculine nouns like stein distinguished a nominative from an accusative; the accusative is what you use for the thing being acted upon. Add the demonstrative þenna ("this," accusative masculine — this stone) and you have stein þenna, "this stone," with both the noun and its demonstrative agreeing in the accusative. Modern Swedish keeps no such distinction: sten is sten whether it is subject or object, and denna sten never changes shape. We know it is the object purely from where it sits — after the verb.

eftiʀ — "in memory of" (and the runic ʀ)

eftiʀ (modern efter, "after, in memory of") is the preposition that introduces the person commemorated. Its final letter, ʀ, is the transliteration of a special rune (the yr-rune) representing a distinct r-like sound — historically a z that had become a kind of r, the famous East Norse ʀ. By the modern period this sound merged with ordinary r, which is why eftiʀ becomes plain efter. You will also see this ʀ as a grammatical ending — it was one of the nominative case markers for many masculine nouns (a name carved ...kirʀ or ...fastr), which is why transliterated runic names so often end in -ʀ or -r. Encountering ʀ on a stone is thus a double signal: a sound modern Swedish has lost, and very often a case ending modern Swedish has lost.

Þórð — the commemorated (ACCUSATIVE, and thorn)

Þórð is the man being remembered — Thórd. It begins with þ (thorn), the runic and Old Swedish letter for the "th" sound. Old Swedish used þ for both the voiced "th" of English that (the sound otherwise written ð, eth) and the voiceless "th" of English thing; the two letters þ and ð were not consistently distinguished, and in the later 14th century both were replaced in writing by the digraphs th and dh, before the sounds themselves shifted to the modern t and d. That is why Old Swedish Þórð surfaces in modern Swedish as Tord — the þ hardened to t, the ð to d. Grammatically, Þórð is the object of the preposition eftiʀ, and prepositions in Old Swedish governed case: eftir took the accusative (or dative, depending on the construction), so Þórð carries the accusative ending. Once again, the form of the name tells you its role; modern Tord would not change at all.

fǫður sinn — "his father" (and the lost DATIVE/agreement)

The inscription closes by saying who Thórd was to the sponsor: fǫður sinn, "his father." fǫður is faðir ("father," modern far/fader) in an oblique (accusative) case, agreeing with Þórð — it is in apposition to the commemorated man, so it takes the same case he does. sinn is the reflexive possessive "his (own)," masculine accusative, agreeing in case, gender and number with fǫður. So fǫður sinn is a tiny three-way agreement chain: noun and possessive both bend to match the accusative role of Þórð. Modern Swedish flattens all of this to sin far — sin agrees only in gender (sin far, sitt hus), never in case, and far never changes for case at all.

This is also where the dative would surface in other inscriptions. Old Swedish had four cases — nominative, accusative, genitive, dative — and the dative marked, among other things, the indirect object and many prepositional complements. A stone reading gaf ... brú fyrir sál ("gave ... a bridge for the soul") puts sál ("soul") in the dative after the preposition. The dative is the most thoroughly lost of all the cases: nothing of it survives in standard modern Swedish except a few frozen phrases (i ladugården, till salu, man ur huse).

The case system, gathered

Pulling the four cases together makes the scale of the loss concrete. Old Swedish marked every noun, adjective, demonstrative and pronoun for case; modern Swedish marks essentially nothing for case except the possessive -s.

Notice the single survivor: the genitive -s. When you write modern Svens sten ("Sven's stone") or stadens torg ("the town's square"), that -s is the last living fragment of the four-case system the runestones display in full. It is the same -s you carve on a runestone genitive — see The Genitive -s for how this lone case-ending behaves in the modern language, and why it is best understood as a clitic rather than a true case.

Why this matters for modern Swedish

The runestones are not a curiosity off to the side of the language; they explain a core fact about how modern Swedish works. Because the case endings are gone, the modern language has no morphological way to tell a subject from an object — hunden looks identical whether the dog bites or is bitten. The only thing that disambiguates is position: subject before the verb, object after it. That is precisely why Swedish word order is so strict and why the V2 rule is so load-bearing. A language that keeps its cases (Old Swedish, Old Norse, modern Icelandic, German to a degree) can afford freer order, because the endings still say who did what. Swedish spent its cases and bought rigid order with the proceeds. Read against the ballads — which keep verbal agreement but have already lost most of the noun cases — and the history page, the runestone shows you the very beginning of that long exchange.

Common Mistakes

What to notice

• The runestones show Old Swedish with its full case system — nominative, accusative, genitive, dative — where a noun's ending marked its job in the sentence.
• The memorial formula "X lét reisa stein eftiʀ Y" ("X had the stone raised in memory of Y") is utterly standard; almost every stone is a variation on it.
• Modern Swedish has lost all four cases but one: only the genitive -s survives (Svens sten). Everything the lost cases did is now done by word order and prepositions — which is why Swedish word order is so rigid.
• The letters þ (thorn) and ð (eth) are "th" sounds that became modern t and d (Þórð → Tord); the runic ʀ marks both a lost sound and, often, a lost nominative ending.
• Read against the ballad and the history of the language, the runestone is the deepest layer — Swedish before it traded case for position.