13

We have attempted to find a formal theory for relationships between words, starting from the observation that we can create generalized rules that describe sentences. Recall these sentences from Chapter 1.

Using the sentences in (1), we were able to define categories and then define rules over categories. From this starting point, we continued to abstract further. Generalizing over Phrase Structure Rules led to X’-theory. We further refined this model with theories of binding and movement.

Major themes

Throughout this book, a few topics kept re-occurring.

• Hypothesis testing. Each step along the way required evidence. The theoretical assumptions are built on empirical observation. This is just good science — it shouldn’t be done any other way.
• How much is syntax? That is, where does syntax end and morphology and semantics begin. The answer is not always clear. There are some things, like selection, which appear to be semantic in nature. But they clearly matter for syntax. Likewise, do-support seems to be a morphological “reaction” to syntax.
• There are levels in syntax. That is, sometimes things are pronounced in a different place from where they start. Put another way, we need a theory that can account for silence created by movement.
• Local vs. non-local dependencies. In general, we created a theory that emphasized local interaction: For instance, selection, we discovered had to happen locally. But then there are some things that appear to interact at a distance, e.g., Binding Theory and Movement.

In one sense, the end goal of (generative) syntax is to find the theory that needs the least amount of stuff. What is absolutely necessary, and what can be framed as simply a product of more processes? As I see it, there are two major things that any syntactic theory needs.

• A theory of constituent structure. How do you put pieces together? How do you partition up the words into phrases, and phrases into bigger phrases?
• A theory that encodes non-local dependencies. How do we relate pieces of structure across distance? For instance, how do we encode the relationship between an antecedent and an anaphor? And how do we encode the relationship between a trace and moved element? Ideally, both of binding theory and movement (of all types) can be reduced to a single mechanism.

In Syntax II, we’ll try to design a theory that reduces to our grammar to just the minimal amount of operations. This new approach is called Minimalism.

The Minimalist Program

What is Minimalist syntax? As the name implies, Minimalism (or the Minimalist Program) is an attempt to discover what the roots of grammar are. As our syntactic theory developed in the mid-eighties, we developed a system of modules, or sub-components of the theory, that interacted to derive the empirical data that we were starting to discover. As a theory, Government and Binding (or generally the Principles and Parameters Framework) were extraordinarily in successful in providing explanations for phenomenon.

The problem was that the explanations themselves weren’t very plausible. We had contrived a theory that incorporated an enormous of “stuff,” and we assumed all human children were born with all this stuff. It was a bit like saying that we’re born with a programming language that had a ton of built in functions. Like maybe we’re born with knowledge that Binding Theory exists. We’re all born with the knowledge phrases can move—but also respect islands. But why? Why do we know this? Why is all of this knowledge universal across languages and speakers? Why did things end up this way, and not some other way?

The goal of Minimalism is to ask which things we need and which can be derived. The Minimalist Program is an attempt to ask which things truly make up the core knowledge that we’re born with, and which things are consequences of knowing this information.

It is worth emphasizing that Minimalism is a program. It is a set of assumptions which guide our theoretical research. But as a theory, it is, in many ways, less successful than X’-theory (and the theories that used X’-theory).