Principles and parameters theory proposes that human languages share universal grammatical principles (structure dependence, subjacency, binding) with parametric variation at discrete switch points. During acquisition, children set parameters for their language—null-subject or not, pro-drop or obligatory subjects. This framework explains the surprising uniformity of deep structure across languages despite surface diversity.
From your study of Universal Grammar, you know the central claim: humans are born with an innate linguistic endowment that constrains the shape of possible natural languages. Principles and Parameters theory, developed by Noam Chomsky and colleagues in the early 1980s, gives that hypothesis its most precise formal expression. The framework proposes that the innate endowment has exactly two components: principles, which are invariant across all languages, and parameters, which vary but only within a pre-specified set of options. Together, these two components produce the striking combination of deep uniformity and surface diversity that characterizes human languages.
Principles are the absolute constraints that hold without exception across every natural language. Structure dependence is perhaps the clearest example: all grammatical rules operate on hierarchical phrase structure, never on linear sequences of words. When English forms a yes/no question from "The man who is tall is happy," the auxiliary that moves is the one in the main clause — not the first auxiliary encountered reading left to right. Even children, who have never been taught the concept of clausal hierarchy, respect this constraint without error. They never produce "Is the man who tall is happy?" This cannot be explained by the input alone — the constraint must be part of the innate system. Subjacency (restricting the distance over which movement can extract constituents) and the binding principles (governing the reference of pronouns and anaphors) function the same way: they hold everywhere, children respect them from the start, and they cannot be learned from positive evidence alone.
Parameters are where languages diverge. Unlike principles, parameters have more than one value, and children must determine from their input which value their language has set. The null-subject parameter (also called the pro-drop parameter) is the canonical example. Italian and Spanish allow sentences without overt subjects — "Parla bene" is grammatical without "lui" — while English requires an overt subject in nearly every clause. The parameter is binary: a language either allows null subjects or requires overt ones. What makes parametric variation powerful is that it is not a collection of isolated quirks; each parameter setting implicates a cluster of correlated properties. Languages with [+null subject] also tend to allow stylistic subject-verb inversion in declaratives, omit expletive subjects like "it" and "there," and display richer agreement morphology. A child who sets [+null subject] acquires the whole cluster at once, not each property through separate evidence.
This clustering is the framework's most theoretically productive feature and its deepest empirical commitment. It predicts that cross-linguistic variation should not be random — languages should clump into types defined by parameter settings, and those types should co-vary in predictable ways. It also predicts that second-language acquisition should involve parameter resetting, which is harder than first-language acquisition precisely because parameters must be unset from their L1 value — a prediction born out by persistent difficulty English speakers have acquiring null-subject languages and vice versa.
Principles and Parameters theory has been substantially revised since its 1980s formulation — the Minimalist Program, ongoing since the 1990s, aims to derive as much as possible from general principles of computational economy rather than stipulated parameters. But the core intuition endures: the deep structure of human language is constrained by innate universal principles, surface diversity arises from a finite set of parametric choices, and the task of language acquisition is not rule-learning from scratch but parameter-setting within a pre-specified design space. That intuition reshaped linguistics and cognitive science, and it remains the starting point for any serious account of how children acquire grammar.