Korva is a scroll of the internet’s best ideas.
Sentences have hidden structure
A question that fails
Take the sentence Corey talked to Nick this morning. A question whose answer is Nick comes out fine: who did Corey talk to this morning? A question whose answer is morning does not: what did Corey talk to Nick this? Linguists have long been engaged by that difference, and it has led to a consensus that more is going on beneath the surface than words in a row.
The same hidden layer sits behind the old gray house sounding normal and the gray old house sounding somewhat deranged.
Ambiguous headlines
Ambiguity shows the structure too. An old newspaper headline read: Sisters reunited after 17 years at checkout counter. The first reading most people get is that the sisters spent 17 years at the checkout counter, when the meaning is that they were reunited after 17 years. The phrase after 17 years can attach at two different places in the sentence, and each attachment gives a different meaning.
Why does the gray old house sound odd?
English adjectives follow a loose ordering that speakers apply without being taught: opinion, size, age, shape, color, origin, material, and purpose. Age comes before color in that order, which is why old gray house sounds natural. Reversing the pair breaks a pattern that native speakers hold unconsciously. The same ordering is found in many other languages, though not in all of them.
How can one sentence carry two meanings?
A phrase can attach to more than one point in a sentence, and each attachment builds a different structure. The classic textbook case is I saw the man with the telescope, where the telescope belongs either to the viewer or to the man. Linguists draw these sentences as tree diagrams in which the same words hang from different branches. The two trees give two different meanings from an identical string of words.
Why do some questions break and others work?
Phrases like this morning behave as a single unit, and pulling just one piece out of a unit leaves the sentence ungrammatical. The linguist John Robert Ross described restrictions of this kind in his 1967 MIT doctoral dissertation. He called the structures that block such movement islands. Researchers have studied island constraints in many languages since then.
More from Wu Tsai Neurosciences Institute, Stanford
