How do children take a string of speech sounds, chop it up into discrete units (words), and then assign to that chopped up string of sounds a particular meaning? In many sentences, perhaps most, words that are semantically related to one another are also near to each other in the sentence. For example, a verb and its arguments – the nouns or other phrases that the verb selects – are usually in close proximity to each other (at least, they are generally clausemates): in a simple main clause sentence like The student read a book, the verb read selects a subject and a direct object noun phrase (NP), and these NPs are positioned right next to the verb that selects them. This is so regardless of the particular language's basic word order or even the degree of rigidity of word order. Many theories of language acquisition exploit this fact to explain (part of) how children begin to tackle the challenge of integrating form and meaning in their language.

But arguments need not be proximal to their selecting predicate, and adjacent or proximal words need not stand in a semantic, selectional relation to one another. This is because human language allows for the semantic relations between words to span long distances – in principle, infinite distances. This book is about how children begin to figure out how to interpret sentences in which the proximity of words belies their semantic relations – how children determine the underlying syntactic structure of sentences in which semantic relations are long-distance, and how knowing the syntactic structure helps children interpret those semantic relations.