Independent clauses

Hello, Gilbert,

@gilbert posted:

1     I've got a sandwich and an apple.

2     They were poor but hardworking.

Although the sentences above could be said to come from these compound sentences where ellipsis has occurred:

1a. I've got a sandwich and [I've got] an apple.
2a. They were poor but [they were] hardworking.

on a superficial reading they are simple sentences with a compound object (1) and with a compound subject complement (2).

The fact that "I've got a sandwich" and "They were poor" can stand alone does not mean that "an apple" and "hardworking" are independent clauses, that is, unless they are found after a period, in which case ellipsis is evident:

A: I've got a sandwich.
B: And (you've got) an apple.

C: They were poor.
D: But (they were) hardworking.

@gilbert posted:

1     I've got a sandwich and an apple.

2     They were poor but hardworking.

Hi, Gilbert and Gustavo—I'm not sure what makes you inclined to assume ellipsis is at work here. I see a conjunction of two noun phrases as the direct object in (1) and a conjunction of two adjective phrases as the subject complement in (2). Here are two diagrams (old-fashioned and modern) for (1):

In each case, there is only one independent clause. In (1), there would continue to be one independent clause if it were "I've got a sandwich, an apple, a can of soda, a slice of cake, and a napkin." If you want to see ellipsis involving a second independent clause, we can change the example a bit and use gapping:

(3) I've got a sandwich and an apple, and he a slice of pizza and a Coke.

As you know, (3) comes from "I've got a sandwich and an apple, and he has got a slice of pizza and a Coke." Thus, "has got" has been elided in (3).

@David, Moderator posted:

I'm not sure what makes you inclined to assume ellipsis is at work here.

The parsing I wrongly described as superficial should then be regarded as the correct one:

@Gustavo, Co-Moderator posted:

[...] on a superficial reading they are simple sentences with a compound object (1) and with a compound subject complement (2).

Sorry for the confusion.

Thank you Gustavo and David.

Sorry that I tend to overthink things, sometimes. Your explanations have made it very clear.

David, I must admit that I'm kind of hopeless with tree diagrams. Could you kindly recommend a good (and simple) reference that would be helpful to me?

Your help is much appreciated.

@gilbert posted:

David, I must admit that I'm kind of hopeless with tree diagrams. Could you kindly recommend a good (and simple) reference that would be helpful to me?

Hi, Gilbert—Although I didn't learn how to do syntactic tree diagrams from a texbook, there are some good ones out there (not necessarily simple) by two guys with the first name of Andrew: Andrew Radford and Andrew Carnie.

@Gustavo, Co-Moderator posted:

The parsing I wrongly described as superficial should then be regarded as the correct one:

I didn't mean to be too dismissive of the idea that there is ellipsis there. One old-fashioned theory in generative grammar concerns Conjunction Reduction. I believe your and Gilbert's parsings are in keeping with it.

Logically, the sentence "X has Y and Z," implies, if it is true, that the separate propositions "X has Y" and "X has Z" are true, and, if both are true, their conjunction as independent clauses will be true, too: "X has Y, and X has Z."

But logic and grammar are distinct. In order to express what is logically entailed by a given sentence, we often need to expound at some length, using a variety of expressions. Naturally, it won't work to say that all those words are elided.

Yesterday, I tried to think of an argument for why the ellipsis-free parsing is needed, independently of the fact that no one would think there are words missing there ("X has Y and Z" can be said unproblematically in a vacuum).

No argument came to me yesterday, but this morning one has. Some kinds of sentences require, for their proper interpretation, that the coordinate structure itself be parsed as the direct object. Consider these examples:

(4) I don't have a sandwich AND an apple.
(5) I don't have a sandwich or an apple.

Sentence (4) means that it is not the case that I have both a sandwich and an apple. It may be true that I have one or the other. Thus, it is not semantically or logically equivalent to "I don't have a sandwich, and I don't have an apple."

Sentence (5) is semantically equivalent to "I don't have a sandwich, and I don't have an apple"; however, the conjunction used is "or." The sentence does not mean the riddle "(Either) I don't have a sandwich, or I don't have an apple."

@David, Moderator posted:

I didn't mean to be too dismissive of the idea that there is ellipsis there. One old-fashioned theory in generative grammar concerns Conjunction Reduction. I believe your and Gilbert's parsings are in keeping with it.

Funnily enough, I thought that speaking about ellipsis would sound like a groundbreaking analysis. Next time I will stick to my convictions — never in my syntactic analysis exercises since I was in college would I have said that those sentences were anything but simple.

@David, Moderator posted:

Hi, Gilbert—Although I didn't learn how to do syntactic tree diagrams from a texbook, there are some good ones out there (not necessarily simple) by two guys with the first name of Andrew: Andrew Radford and Andrew Carnie.

David, thanks for the reference. I shall check out the two Andrews to know what tree diagrams are all about!

Thanks again.