Chapter 8 of Syntactic Structures proposes a test of a grammar’s adequacy from a different angle than anything so far: ambiguity. If a sentence can be understood in more than one way, it ought to admit more than one analysis somewhere in the grammar — and conversely, a string that admits more than one analysis ought, in fact, to be ambiguous. How well this correspondence holds, Chomsky argues, is a measure of how good a grammar is.
Ambiguity Inside a Single Word
The simplest case sits at the level of sound. Spoken English “a name” and “an aim” can come out as exactly the same string of sounds, /əneɪm/. If a grammar had only a phonemic level, this coincidence would go unexplained. But once a level of morphemes — units like “a,” “an,” “aim,” “name” — is established, the fact that this sound sequence analyzes two ways falls out automatically from building the morphology in the simplest way. The discovery of the ambiguity itself becomes the justification for having a morphemic level at all.
The same thing happens at the level of phrase structure. “They are flying planes” can mean either “those specks are aircraft in flight” or “those people are piloting planes.” The two readings are already distinct at the phrase-structure stage — whether “flying” is an adjective modifying “planes,” or part of a progressive verb “are flying” with “planes” as its object.
“The Shooting of the Hunters”: Ambiguity Visible Only at the Transformational Level
But even phrase structure leaves some ambiguity unexplained.
“The shooting of the hunters” can mean either that the hunters did the shooting (“the hunters shoot”) or that they were shot (someone “shoots the hunters”). Yet at the level of phrase structure, both readings come out as exactly the same shape — “the - V+ing - of + NP.” Phrase structure alone has nowhere to locate this ambiguity.
Go down to the transformational level and the two readings pull apart cleanly. The first derives from the kernel sentence “the hunters shoot” by a nominalizing transformation (NP - C - V → the - V+ing - of + NP). The second derives from the transitive kernel “they shoot the hunters” by a different nominalizing transformation. The ambiguity is exactly the difference in which kernel sentence, and which transformation, produced the surface form.
A similar pair: “the picture was painted by a new technique” and “the picture was painted by a real artist.” Identical in phrase structure — “NP - was+Verb+en - by + NP” — yet understood quite differently. The first keeps the instrumental “by” of an active sentence; the second is the agentive “by” the passive transformation inserts. Their transformational histories are entirely different.
Is Grammar Based on Meaning?
Chapter 9 reinforces the argument from the other direction, testing claims — common at the time — that grammar should be built on meaning.
Take the claim that the subject-verb relation corresponds to “actor-action.” Sentences like “I will disregard his incompetence” or “I missed the train” break the parallel claim for verb-object too — the object of “disregard” or “miss” isn’t a “target” acted upon so much as something the action failed to reach or failed to notice. The semantic role shifts sentence by sentence.
The same holds for sound and meaning. “Bachelor” and “unmarried man” are synonyms with entirely different sounds; “bank” (of a river) and “bank” (for savings) share a sound but not a meaning. “Same sound iff same meaning” fails in both directions.
None of this means Chomsky thinks meaning should be ignored.
Grammar is best formulated as a self-contained study independent of semantics.
“Best formulated independently” is not the same as “has nothing to do with.” Indeed, the whole discussion of structural ambiguity in chapter 8 shows a clear correspondence between the form of a sentence (syntax) and how people understand it (a semantic fact). That correspondence just isn’t precise enough to found grammar on. Build syntax first, autonomously, in purely formal terms — then study its correspondence with meaning as its own question. That two-step method is the methodological core of the book.
Closing
The test of constructional homonymity carries a strong claim: any difference that shows up nowhere in the grammar shouldn’t exist as a difference in understanding either. Measured against that test, Chomsky demonstrates — from the concrete fact of ambiguity — that a level beyond phrase structure, the transformation, is necessary. And the book holds two things at once: grammar is not founded on meaning, and the relationship between grammar and meaning is still a serious object of study. That combination is the most easily misread, and most carefully constructed, conclusion of Syntactic Structures.