XX is the author of the following post on the syntax of imperative constructions in English. I love her enthusiasm for this subject, so I command you to enjoy her essay. Do it! — MLR
Syntactic theory attempts to find a consistent way to document the structures of human language: a relatively brave, exciting, and yet intimidating intellectual goal. As any student of linguistics will tell you, however, there still seem to be a large number of constructions that remain unexplained (which is understandable, as finding a syntactic model that accounts for any possible construction in any possible human language is obviously a little tricky). A few months ago, it occurred to me that imperative constructions in English worked in a way that seemed to run counter to previously established rules of syntactic theory, especially rules that are English-specific. As this initial curiosity eventually turned into a larger research project, I had the opportunity to find a way we can explain how and why imperatives in English are possible.
First, imperatives can be defined as constructions such as, “Wash the dishes,” “Please take a seat,” “Be on time tomorrow,” and “Don’t be a fool”, utterances that represent commands, requests, instructions, wishes, etc. Unlike other languages, English does not provide specific morphological markers for imperatives. Remember high school Spanish or French class when you had to conjugate the verb in a special way if you were speaking in the subjunctive or in the imperative? This change of mood is not shown through the verb, it seems, in English since the base form of the verb is used. In the example, “Be on time tomorrow”, the base form be is used; this is our proof that there has got to be something else going on in the construction that indicates to a native speaker when an utterance is an imperative and when it is not. If the indication is not provided in a morphological change, the only other option is that it exists in the syntactic structure.
So, what seems to be the big problem? Well, syntacticians have established an English-specific rule called the Extended Projection Principle (EPP), which basically says that every sentence needs a subject to be acceptable. With imperative constructions in English, it seems at first glance that either the subject might be optional or implied. This doesn’t seem to be a problem for someone studying, say, pragmatics (why do we care if context dictates conversational implicature, right?) but according to the rules of syntax we need to be able to explain these constructions within the established model that explains (hopefully) the structures of human language.
Using the most contemporary model Syntacticians have come up with (although it’s still not perfect, but the best we have so far), X-bar theory shows a precedent of using specific features to generate movement in the structure tree. For example, to explain how questions starting with wh-words (what, why, where, when, etc.), the [+wh] feature generates NP movement that complies with both the X-bar model and grammatical/acceptable wh-questions in English. For example, “What are you saying?” starts as “You are saying what?” and the [+wh] feature moves the what from the end of the structure to the beginning. Seeing this precedent in syntactic theory, I couldn’t help but wonder if imperatives might work in a similar way.
And, this boys and girls is where we find the magical, wonderful feature [+imp] generates movement within the structure! Here is an illustration that helps explain its mystical properties:
In the illustration on the right, we can see that the imperative “call me tomorrow,” works if we start with the verb in the V position. It moves to the T position, where the [+imp] feature moves it to the C position, which explains why the base form of the verb occurs first in the sentence. There is movement from the specifier position of the VP to the specifier position of the TP, which fills the subject position of the sentence (thus satisfying the EPP).
So, in order to comply with syntactic theory, imperatives must have an additional [+imp] feature to provide for the seemingly missing subject. The V –> T movement, T –> C movement, and VP –> TP movement satisfies the EPP by providing something in the subject position of the diagram. Since this amazing [+imp] feature explains how imperatives can work in English, we can now be one step closer to finding a consistent way to document the structures of human language. Amazing, right? Now we just have to hope that X-bar theory isn’t completely thrown out in the future… because then we’ll need to explain all of this all over again. :]
Of course, there are issues of theta grid requirements and how negated imperatives work in both semantic and syntactic theory… but we’ll just have to get into those later.