What is Universal Grammar?

I will use the term “universal grammar” to refer to these properties of human biological endowment. Thus the properties of universal grammar are “biologically necessary,” but in the interesting cases not logically necessary elements of what someone might choose to call a language. (28)

We might discover that there is no language faculty, but only some general modes of learning applied to language or anything else. If so, then universal grammar in my sense is vacuous, in that its questions will find no answers apart from general cognitive principles. But still, universal grammar conceived as a study of the biologically necessary properties of human language (if such exist) is strictly a part of science. (29)

“universal grammar” is taken to be the set of properties, conditions, or whatever that constitute the “initial state” of the language learner, hence the basis on which knowledge of language develops. It by no means follows from such an account that there must be specific elements or rules common to all languages, or what he calls “features” common to all languages, unless we take these “features” in a suitably abstract manner, in which case, the controversy dissolves. (69)

Explanation:

strongly supports a qualitative, explanatory approach, especially in contrast to surface-level, data-driven descriptions.

🔍 Chomsky’s Statement (paraphrased and unpacked)

"Universal grammar" is taken to be the set of properties, conditions, or whatever that constitute the "initial state" of the language learner...

• This confirms: UG is not a list of rules or language features, but rather the initial cognitive structure from which language can emerge.
• It's about mental potential and structural constraints, not just what’s observable on the surface.

"...It by no means follows from such an account that there must be specific elements or rules common to all languages..."

• Chomsky warns against a naïve or superficial universality — e.g., assuming that all languages must have, say, nouns, subjects, or passive constructions.
• This rejects a purely descriptive or typological approach, which looks for surface patterns across languages.

"...unless we take these “features” in a suitably abstract manner, in which case, the controversy dissolves."

• The key is abstraction: universal features may exist, but at a deep, cognitive, or formal level, not necessarily visible in surface grammar.
• For example:
  • All languages may have hierarchical structure — but not all mark it the same way.
  • All languages may have subject-predicate relationships — but not all show it morphologically.

What Universal Grammar is Not