Raising vs Control Verbs

Two sentences can look identical and be built completely differently. Han prøver å vinne "He tries to win" and Han ser ut til å vinne "He seems to win" share the frame subject + verb + å + infinitive, yet in the first sentence han is the one doing the trying, while in the second han does nothing to the verb ser ut til at all — the seeming is not something he performs. This is the raising vs control distinction, one of the genuinely deep facts of clause structure. Mastering it explains why Det synes å regne "It seems to be raining" is fine but ✱Det prøver å regne is gibberish, and why Han er lett å like means "he's easy to like" and not "he is easy and likes." We assume the basics of infinitive complements (see infinitive clauses); here we get to the structural heart.

The core difference: who gives the subject its role

Every verb hands out semantic roles (theta-roles) to its arguments — prøve "try" needs a trier, love "promise" needs a promiser. The question that separates raising from control is simple:

• Control verbs (prøve, love, ønske, nekte, bestemme seg for) give their subject a real role. The subject genuinely tries / promises / refuses, and it also controls a silent understood subject (PRO) in the infinitive: Han prøver [PRO å vinne] = "He tries [for him to win]."
• Raising verbs (synes, se ut til, late til, vise seg) give their surface subject no role of their own. The subject is semantically an argument only of the embedded infinitive; it has been "raised" up into the matrix subject position for purely syntactic reasons. Han synes å vinne — han is the winner, not the "seemer"; there is no seemer. (Virke "seem" belongs to the same semantic family, but idiomatic Bokmål uses it with a predicative — Han virker sliten — not with a bare infinitive; virker å være is a Swedish-flavoured calque Språkrådet advises against, so the cleaner raising frames are se ut til å, late til å and the literary synes å.)

In the control case, han wears two hats — trier and would-be-comer. In the raising case, han is only the embedded subject (winner, tired one), lifted into the front of the sentence. This abstract difference produces concrete, testable consequences.

The expletive test: det synes å regne vs ✱det prøver å regne

The cleanest diagnostic is the expletive det. A raising verb assigns its subject no role, so it happily hosts a meaningless, role-less det "it" — the same det you see in weather and existential sentences. A control verb needs a real, role-bearing subject, so it rejects expletive det:

Now try the same with a control verb and it collapses, because there is no one to do the trying:

✱Det prøver å regne "It tries to rain" is ungrammatical (or only works as bizarre personification) because prøve demands an agent — something that can intend to try — and expletive det is no agent. So the test is: can you put a meaningless det in the subject slot? Yes → raising. No → control. This single test sorts the two classes instantly, and it works the same way as the English "It seems to rain" (fine) vs "It tries to rain" (only as personification).

The idiom test: meaning preserved under raising

A second classic diagnostic uses idioms. An idiom chunk like det trekker "there's a draught" or katta er ute av sekken "the cat is out of the bag" keeps its idiomatic meaning only if its pieces stay together structurally. A raising verb lets the idiom's subject raise without losing the idiomatic reading, because raising is a purely structural lift. A control verb breaks the idiom, because it would force the raised piece to be a real agent:

Try to embed the same idiom under a control verb and the figurative meaning evaporates — katta prøver å være ute av sekken can only mean a literal cat attempting something. Idiom meaning preserved → raising; idiom meaning destroyed → control. Both the expletive test and the idiom test point the same way, because both probe the same fact: raising verbs don't assign a subject role.

synes: the verb that does both jobs

Synes deserves a note because it leads a double life. As a raising verb it means "seem" and takes an infinitive: Han synes å være fornøyd "He seems to be content" — han is the content one, no one does the "synes-ing." As an ordinary opinion verb it means "think/find" and takes an at-clause: Jeg synes at filmen var god "I think the film was good" — here jeg really holds the opinion (a real role). The raising use of synes "seem" is somewhat (formal/literary) in modern Bokmål; everyday speech prefers virker or ser ut til for "seem." (More on the -s verb family in s-verbs.)

Control verbs and the subject's reflexive

Because a control verb's subject is a real argument, it can bind a reflexive in the embedded clause — the subject is "high" enough to do binding work. This links directly to the binding page: control structures feed subject-oriented reflexives.

Note that bestemme seg for "decide" and nekte "refuse" are inherently control: there is always someone who decides or refuses. You cannot raise an expletive into them (✱Det bestemte seg for å regne), which re-confirms the classification. (For why these reflexives behave as they do, see binding.)

Tough-movement: Han er lett å like

A close cousin of raising is tough-movement, named for English "tough/easy/hard to please." Here a subject that is logically the object of the infinitive surfaces as the subject of an evaluative adjective (lett "easy," vanskelig "hard," umulig "impossible," deilig "lovely"):

In Han er lett å like, han is not the one finding things easy — han is the thing being liked. The construction parallels an impersonal Det er lett å like ham "It is easy to like him," with ham as the object, but tough-movement promotes that object to subject and leaves a gap after the infinitive (å like _). English does the identical thing ("He is easy to like" ↔ "It is easy to like him"), so the structure transfers — but watch the gap: Han er lett å like has no pronoun after like (✱Han er lett å like ham is wrong, a double-marking of the object).

Common Mistakes

1. Treating a raising verb as control by blocking expletive det. Raising verbs welcome meaningless det.

2. Spelling se ut til å and late til å wrong. They are multi-word units: se ut til å, later til å — not run together, and with å before the infinitive.

3. Inserting a subject into a control verb's infinitive. Control infinitives have a silent PRO controlled by the matrix subject — no overt subject.

4. Filling the tough-movement gap with an object pronoun. The promoted subject already is the object; don't repeat it.

5. Using raising synes in casual speech where virker/ser ut til is natural. Raising synes "seem" is formal/literary; everyday speech prefers virker.

Key Takeaways

• Control verbs (prøve, love, ønske, nekte, bestemme seg for) give the subject a real role and control a silent PRO: Han prøver å vinne = he is the trier.
• Raising verbs (synes, se ut til, late til, vise seg) give the surface subject no role of their own; it belongs to the embedded infinitive and is lifted up: Han ser ut til å vinne = he is only the winner. (Virke "seem" is semantically a raiser too, but takes a predicative — Han virker sliten — rather than virker å være.)
• Expletive test: Det synes/ser ut til/later til å … is fine (raising); ✱Det prøver/lover å … fails (control).
• Idiom test: an idiom keeps its meaning under raising, loses it under control.
• synes is raising ("seem," + infinitive, formal) or an opinion verb ("think," + at-clause).
• Tough-movement (Han er lett å like) promotes the infinitive's object to subject and leaves a gap — never fill it.