The Sentence Schema (Satsschema)

Swedish word order can feel like a pile of separate rules — V2, where inte goes, where particles go, where objects go. Scandinavian linguistics solves this with one elegant device: the sentence schema (satsschema), a grid of fixed positions that every clause slots into. Once you see the grid, the rules collapse into a single picture, and the most baffling phenomenon for English speakers — why a compound verb splits around inte (har inte läst, "haven't read") — becomes obvious. This page presents the schema for main and subordinate clauses and uses it to predict placement.

The main-clause grid

A Swedish main clause has seven fields, in this fixed order:

FundamentFinite verbSubjectSentence adverbNon-finite verbObjectAdverbial
Igårhadejagfaktiskt intelästbokenhemma.
Jagläseraldrigtidningenpå morgonen.
Nuvillhannogåkahem.

Read the first row as a sentence: Igår hade jag faktiskt inte läst boken hemma — "Yesterday I actually hadn't read the book at home." Every word has a home:

  • Fundament (first position) — exactly one element, freely chosen for emphasis: a subject, an adverbial, an object, even a whole subordinate clause. Here it is igår ("yesterday").
  • Finite verb (second position) — the conjugated verb, locked to slot 2. This is the V2 rule. Here, the auxiliary hade.
  • Subject — moves here when something else took the fundament. Here, jag.
  • Sentence adverbinte, aldrig, alltid, faktiskt, nog, kanske. Here, faktiskt inte.
  • Non-finite verb — supines, infinitives, participles governed by the finite verb. Here, the supine läst.
  • Object — here, boken.
  • Adverbial — manner/place/time complements. Here, hemma.

Igår hade jag faktiskt inte läst boken hemma.

Yesterday I actually hadn't read the book at home. Every word maps to a field: fundament 'Igår', finite 'hade', subject 'jag', adverb 'faktiskt inte', non-finite 'läst', object 'boken', adverbial 'hemma'.

På morgonen dricker jag alltid kaffe i köket.

In the mornings I always drink coffee in the kitchen. Fundament 'På morgonen' → finite 'dricker' → subject 'jag' → adverb 'alltid' → object 'kaffe' → adverbial 'i köket'.

💡
The fundament holds exactly one element, but that element can be as big as a whole subordinate clause. After it, the finite verb is nailed to slot 2 and the subject is pushed to slot 3. This single grid is the V2 rule made visible.

Why compound verbs split

Now the payoff. English keeps a perfect tense glued together — "I have not read" or "I had not read": auxiliary and participle stay adjacent, with "not" tucked between or after. Swedish does something that looks bizarre until you see the grid: the finite part fronts to slot 2, while the non-finite part stays back in slot 5 — and inte sits in the sentence-adverb field between them.

FundamentFinite verbSubjectSentence adverbNon-finite verbObject
Jagharintelästboken.

So Jag har inte läst boken ("I haven't read the book") splits the verb har ... läst around inte. The auxiliary har is the finite verb and must occupy slot 2; the supine läst is non-finite and stays in slot 5; inte lives in the sentence-adverb field, which falls neatly between them. The split is not an exception — it is the grid working exactly as designed.

Jag har inte läst boken än.

I haven't read the book yet. The verb 'har ... läst' splits: finite 'har' at slot 2, supine 'läst' at slot 5, 'inte' between.

Hon hade aldrig sett snö förut.

She had never seen snow before. Finite 'hade' + adverb 'aldrig' + non-finite 'sett'.

Vi kommer nog inte att hinna idag.

We probably won't make it in time today. Two sentence adverbs ('nog inte') sit together between finite 'kommer' and the infinitive marker 'att hinna'.

This also fixes the single most common adverb error. Inte attaches relative to the finite verb: after it in a main clause, before it in a subordinate clause. With a compound verb, "after the finite verb" still means between the auxiliary and the participle — never after the participle. Jag har inte läst is right; Jag har läst inte is wrong.

The subordinate-clause grid

A subordinate clause uses a parallel — but crucially different — grid. The conjunction opens it, the subject comes next, and the sentence adverb moves to before the finite verb (this is the BIFF rule). The non-finite verb still trails behind.

ConjunctionSubjectSentence adverbFinite verbNon-finite verbObjectAdverbial
attjagfaktiskt intehadelästbokenhemma

The same content as before, now embedded: Hon trodde att jag faktiskt inte hade läst boken hemma — "She thought I actually hadn't read the book at home." Compare the two grids side by side and the difference is a single swap: the adverb and the finite verb trade places. In the main clause it is finite verb → adverb; in the subordinate clause it is adverb → finite verb.

Hon trodde att jag faktiskt inte hade läst boken hemma.

She thought I actually hadn't read the book at home. Subordinate grid: conjunction 'att' → subject 'jag' → adverb 'faktiskt inte' → finite 'hade' → non-finite 'läst' → object 'boken' → adverbial 'hemma'.

Han sa att han inte hade hört något.

He said he hadn't heard anything. Inside the clause, 'inte' precedes the finite 'hade'; the supine 'hört' still trails.

Det är konstigt att du aldrig har varit i Stockholm.

It's strange that you've never been to Stockholm. Subordinate order: 'aldrig' before the finite 'har'.

Notice that the non-finite verb (läst, hört, varit) sits after the finite verb in both grids. Only the sentence adverb relocates. That is the entire difference between main and subordinate word order, captured in one move on the schema.

💡
The whole main-vs-subordinate contrast reduces to one swap on the grid: main clause is finite verb → adverb (hade inte läst), subordinate is adverb → finite verb (inte hade läst). The non-finite verb never moves. If you can place inte, you have the whole schema.

Particles and the schema

A verb particle (the stressed second half of a particle verb such as känna igen "recognise", slå på "switch on") behaves like a tiny non-finite element: it sits in the post-verbal field after inte, not before it.

Jag känner inte igen honom.

I don't recognise him. The particle 'igen' follows 'inte', sitting in the non-finite zone — finite 'känner' → adverb 'inte' → particle 'igen'.

Hon slog inte på lampan.

She didn't switch on the lamp. Particle 'på' after 'inte', mirroring the auxiliary–participle split.

Common Mistakes

❌ Jag har läst inte boken.

Incorrect — 'inte' belongs in the sentence-adverb field, between the finite 'har' and the non-finite 'läst', never after the participle.

✅ Jag har inte läst boken.

I haven't read the book.

❌ Han sa att han hade inte hört något.

Incorrect — in a subordinate clause the adverb moves before the FINITE verb: 'inte hade', not 'hade inte'.

✅ Han sa att han inte hade hört något.

He said he hadn't heard anything.

❌ Jag inte läser tidningen på morgonen.

Incorrect — in a main clause 'inte' goes after the finite verb, not before the subject. The verb is locked to slot 2.

✅ Jag läser inte tidningen på morgonen.

I don't read the paper in the mornings.

❌ Jag känner igen inte honom.

Incorrect — the particle 'igen' is part of the non-finite zone and follows 'inte'.

✅ Jag känner inte igen honom.

I don't recognise him.

Key Takeaways

  • A Swedish clause maps onto a grid of fixed fields: fundament | finite verb | subject | sentence adverb | non-finite verb | object | adverbial.
  • The finite verb is locked to slot 2 (V2); the non-finite verb stays back — which is why compound verbs split around inte (har inte läst).
  • Inte sits in the sentence-adverb field: after the finite verb in main clauses, between auxiliary and participle, never after the participle.
  • The subordinate grid is identical except that the sentence adverb moves to before the finite verb (the BIFF rule) — that single swap is the whole main-vs-subordinate difference.
  • Particles behave like non-finite elements and follow inte (känner inte igen).

Now practice Swedish

Reading grammar gets you part of the way. The exercises are where it sticks — free, no signup needed.

Start learning Swedish

Related Topics

  • The V2 Rule (Verb Second)A1The core law of the Swedish main clause: the finite verb occupies the SECOND position, no matter what comes first. Position one — the fundament — can hold the subject, an object, a time or place adverb, or even a whole clause, but only ONE constituent fits there, and the verb follows immediately. Crucially, V2 counts CONSTITUENTS, not words: a five-word time phrase is still 'first', so a long opener still leaves the verb right after it.
  • The BIFF Rule (Subordinate Clause Order)B1Subordinate clauses do NOT have V2. The order is conjunction + subject + sentence-adverb + finite verb, so the sentence adverb (especially 'inte') comes BEFORE the verb — the exact opposite of a main clause, where 'inte' follows it. The mnemonic BIFF stands for 'I Bisats kommer Inte Före Finita verbet' — in a subordinate clause, 'inte' comes before the finite verb. The single diagnostic for clause type is where 'inte' sits: after the verb = main, before the verb = subordinate.
  • Placing inteA2Exactly where inte goes: AFTER the finite verb in a main clause (Han sover inte), after verb+subject when something is fronted (Idag sover han inte), BEFORE the finite verb in a subordinate clause (...att han inte sover), and BETWEEN the two verbs in a compound tense (Han har inte sovit / Han vill inte sova). Plus object shift: a weak pronoun object hops left over inte (Jag känner honom inte).
  • Object and Adverb PlacementB2How Swedish orders the things after the verb: indirect object before direct (gav honom boken), place before time at the end (i Lund nu), and the rule competitors never mention — object shift, where an unstressed pronoun object hops left over inte (Jag såg honom inte) while a full-noun object stays put (Jag såg inte Pelle). This asymmetry is Holmberg's generalisation, and it governs everyday pronoun placement.