Do-support in the English language

Table of contents

1. Abstract

2. Introduction
2.1 Thematic Introduction
2.3 Overview
2.4 Review of Previous Research
2.5 Research Question and Hypothesis

3. Economy of Do-Support
3.1 Elliptical Form of the Verb Phrase
3.2 Fronting of the Verb Phrase

4. Yes/ No- Questions
4.1 Affix Hopping onto T in Yes/No- Questions
4.2 Affix Hopping onto V in Yes/No- Questions
4.3 The Usage of Do-Support in Yes/No- Questions

5. Negative Phrases and their Syntactic Structure
5.1 Affix Hopping onto V in Negative Phrases
5.2 Do-Support in Negative Sentences

6. Negative Interrogative Sentences
6.1 T-to-C Movement in Negative Interrogative Sentences
6.2 Do-Support in Negative Interrogative Sentences

7. Infinitives of main verbs in Empathic Sentences
7.1 Infinitives after “Do” in Empathic Sentences

8. Conclusion

9. References

10. Statement of Non-Plagiarism

1. Abstract

The following term paper deals with the two different approaches of the usage of do- support by Adger in Core Syntax- A Minimalist Approach and Radford’s Analyzing English Sentences (Adger , 2003; Radford, 2016). The pivotal question “Why and when do we use the do in some cases and in some not?” should be analyzed in the following. Therefore, the works of both linguists will be regarded in terms of do-support.

We will see that Adger and Radford base their breakdown on Chomsky’s analysis of auxiliaries in syntactic structures which says that we insert do-support when all other rules fail to apply (Adger, 2003). Anyway Adger bases his analysis on the so-called Pronouncing Tense Rule (short: PTR) which will be outlined in chapter 2.1 (Adger, 2003). Radford in contrast represents that we always need to attach a verbal affix to an overt verb stem to satisfy the so-called Attachment Condition which will be introduced in chapter 2.1 (Radford, 2016). If we cannot realize this condition through movement operations like Affix Lowering, do-support needs to kick in (Radford, 2016).

2. Introduction

2.1 Thematic Introduction

Adger and Radford both deal with the usage of the do-support in different sentences. First we set up conditions when do-support is necessary and in which cases it is not. Afterwards different sentences will be regarded, always with the question back in mind “Is do-support essential for this sentence to save its grammar?”. Therefore, the theories of both linguists will be contrasted. Finally, we’ll conclude our analysis and sum up our knowledge.

These sentences, where do-support is used, are projected in complementiser phrases (CP) and tense phrases (TP) (Radford, 2016, p. 289).

Complementiser phrases define the discursive status of sentences like expressions, statements or questions (Adger, 2003). The phrases are headed by a complementiser e.g. “that/if/whether/for”. Complementiser prhases can also be headed by an inverted auxil­iary (e.g. in questions) or a null C constituent which means that they’re empty (Radford, 2016, p. 489).

In contrast, tense phrases define the relation to the verb phrase and determine if some­thing already happened, happens or will happen (Adger, 2003, p. 47).

First we have to define some essential key terms in order to understand the following analysis by Radford and Adger.

Initially the first question is, what is the so-called do-support? The do-support refers to the use of the expletive auxiliary do which has no meaning but is essential to be inte­grated into the sentence to save its grammar. This theory is based on Chomsky’s system of Generative Grammar, which says that do-support is “[...] a last resort strategy, which kick in when the rule which attaches tense to main verbs fails to apply for some reason” (Adger, 2003, p. 185). Radford also bases his analysis on Chomsky’s system and refers to the “economy account of do-support, [...] where use of the more economical auxil- iariless structure leads to a crash” (Radford, 2016, p. 289).

So in which sentences is do-support necessary?

We have to use do-support in elliptical and fronted sentences, in yes-no questions and in negative phrases as well as in negative interrogative phrases and empathic sentences (Adger, 2003; Radford, 2016).

At this point we have to introduce some conditions and operations, which justify the presence of do in those sentences.

Initially we have to introduce the Attachment Condition on which Radfords approach of do-support is based. It states that every verbal affix needs to be attached to an overt verb stem at its phonetic level, which means that it has to be spelled out (Radford, 2016). That condition is relevant for our following analysis to justify the relevance of do-support in cases, where otherwise the derivation leads to a crash. If there is no possi­bility to realize the attachment of the verbal affix to a verb stem through movement op­erations like Affix Lowering, do-support needs to kick in (Radford, 2016).

In contrast, Adger bases his analysis of do-support on the Pronouncing Tense Rule (PTR). It entails that tense features are only checked if v is the head of T’s sister. When this is the case, the tense features on little v are pronounced, they are in a so-called tense chain. Each syntactic object in the chain c-commands the next one. As a result the chain breaks, when c-command doesn’t hold and the PTR is violated (Adger, 2003, p. 192). Now we will regard some conditions and operations which are introduced in order to satisfy the former conditions.

First the so-called Recoverable Condition (Radford, 2016) or Identity condition of VP Ellipsis (ICVE) (Adger, 2003, p. 198) needs to be covered. It says, that “Material can only be deleted if its content is recoverable.”(Radford, 2016, p. 291). This means that syntactic operations can only take place, if there is an identical VP in the same sen­tence, which takes the role of an antecedent. Antecedents are expressions, which refer to each other because they are in a relation. The condition is relevant for the cases, where we deal with elliptical forms of phrases (Radford, 2016, p. 520).

The Inversion Licensing Condition is really important in terms of yes-no questions. If we have an empty C and it has a negative or degree specifier or an interrogative spec­ifier in a root clause, then it is licensed to carry a T-feature which triggers Auxiliary Inversion. Auxiliary Inversion is the process, where the auxiliary is moved to the front of the sentence. As a result a T auxiliary becomes the adjunct of C (Radford, 2016, p. 506).

Another operation is the Duke of York operation. No constituent is allowed to move back into a position where it already has moved out of. In addition, the Excorporation Constraint defines that no head can move out of another head, which it is adjoined to. These conditions will be relevant when we try to move constituents around to satisfy the Attachment Condition (Radford, 2016, p. 293 ).

We will also come across the Head Movement Constraint. Heads can only move be­tween the head of a given structure and its complement, which is the sister of the head (Radford, 2016, p. 273).

Besides, we have to define the Strict Cyclicity Principle which is really important for our movement operations. Cycling can only apply from the head of a structure and the constituents which are c-commanded by the head (Radford, 2016, p. 287).

2.3 Overview

The given term paper deals examines 8 units. First we already introduced essential con­ditions to satisfy the usage of do-support with regards to Adger and Radford (2003, 2016). Afterwards we consider previous research which will be helpful for the follow­ing analysis and on whose basis we formulate our research hypothesis. Then the differ­ent kinds of sentences in which do-support is necessary will be examined. We start with elliptical and fronted ones, followed by yes-no questions, negative phrases, negative interrogative phrases and finally empathic sentences. Therefore, Adger’s and Radford’s analysis will be contrasted on basis of their assumptions. Finally, we’ll conclude our outcome.

2.4 Review of Previous Research

A theory of sentence structures should be as economical as possible. Therefore, do- support in only used as a very last option to save the sentence to be grammatical. In the following we will now compare Adger’s Core Syntax- A Minimalist Approach with Radford’s Analyzing English Sentences. Even though both linguists rely on Chomsky’s analysis of syntactic structures, Adger explains the usage of do-support by the PTR whereas Radford adapts to the Attachment Condition (Adger, 2003; Radford, 2016). We will examine the problems which come across with these adaptations and might not be regarded in their analysis.

2.5 Research Question and Hypothesis

The leading question is “How and when do we use do-support in a sentence?”. We have already defined some principles which have to be satisfied and concluded that do- support is only used when it’s essential for the grammar of the sentence. But how is this realized and how do we get the do on the right position in the sentence? This will be discussed in the following.

3. Economy of Do-Support

Let’s take the example:

(1) You cooked a meal.

Abbildung in dieser Leseprobe nicht enthalten

[CP [C - ] [TP You [T Af [VP [V cook+Af] [ANP a meal]]]

In this example we first lower the tense affix (Af) via Affix Hopping to satisfy the At­tachment Condition. Affix Hopping is the process, where the unattached affix moves to the closest verb stem - in this case, the verb cooked (Radford, 2016, p. 186).

The verb stem cook has got an inflectional affix attached to it. Because of the past tense affix, cook becomes cooked.

In this case do-support isn’t necessary to save the grammar of the sentence. The under­lying conditions are satisfied and so the sentence can stay as it is.

3.1 Elliptical Form of the Verb Phrase

Now let’s take a sentence, where parts of the verb phrase are left out.

(2) You said you cook a meal, and you did.

This sentence is the elliptical form of

(3) You said you cook a meal and you did cook a meal.

First we consider Radford’s approach. He outlines that the VP Ellipsis is the product of the deletion of the PF component, which means that the phonetic expression of cook a meal isn’t spelled out (Radford, 2016, p. 290).

Now two operations occur in the PF component: The VP Ellipsis and the Affix Low­ering. But how can we know which comes first? At this point we have to distinguish between syntactic operations and morphological ones. VP Ellipsis is a syntactic opera­tion. Affix Lowering in contrast is a morphological one. Syntactic operations occur be­fore morphological ones and therefore ellipsis comes first (Radford, 2016).

But if we now follow this order we would violate our Attachment Condition. The in­flectional affix (Af) can’t attach to the verb stem, because the cook had undergone ellip­sis before (Radford, 2016).

Even though Radford suggests turning the operations around and realize Affix Low­ering before VP Ellipsis, we won’t follow this approach. As we stated above, syntactic operations occur before morphological ones. Therefore, this order doesn’t make sense under these determined conditions (Radford, 2016).

As we now learnt, the usage of a tense affix alone in combination with ellipsis leads the derivation to crash (Radford, 2016). So what would happen if we use the do-support in our example?

[...]