# Grammar mavens...

**URL:** <https://boards.straightdope.com/t/grammar-mavens/200938>\
**Category:** Factual Questions\
**Created:** [September 11, 2003, 7:24am UTC](https://boards.straightdope.com/t/grammar-mavens/200938 "2003-09-11T07:24:35Z")\
**Posts on this page:** 20\
**Page:** 1

<div class="post-metadata">

**Author:** ![Achilles](https://avatars.discourse-cdn.com/v4/letter/a/dc4da7/32.png) [@Achilles](https://boards.straightdope.com/u/Achilles)\
**Post date:** [September 11, 2003, 7:24am UTC](https://boards.straightdope.com/t/grammar-mavens/200938/1 "2003-09-11T07:24:35Z")

</div>

I need to know which of the following two is correct -

“The report **which** is to be produced…”

or

“The report **that** is to be produced…”

Thanks all.

---

<div class="post-metadata">

**Author:** ![Bryan\_Ekers](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/bryan_ekers/32/183_2.png) [@Bryan\_Ekers](https://boards.straightdope.com/u/Bryan_Ekers)\
**Post date:** [September 11, 2003, 8:10am UTC](https://boards.straightdope.com/t/grammar-mavens/200938/2 "2003-09-11T08:10:50Z")

</div>

Try “The report, once it is produced,…”

---

<div class="post-metadata">

**Author:** ![mhendo](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mhendo/32/3159_2.png) [@mhendo](https://boards.straightdope.com/u/mhendo)\
**Post date:** [September 11, 2003, 8:12am UTC](https://boards.straightdope.com/t/grammar-mavens/200938/3 "2003-09-11T08:12:27Z")

</div>

In order for us to be able to adjudicate on this, you have to give us the rest of the sentence, because we need to know whether the clause in question is restrictive or non-restrictive.

---

<div class="post-metadata">

**Author:** ![Kaitlyn](https://avatars.discourse-cdn.com/v4/letter/k/73ab20/32.png) [@Kaitlyn](https://boards.straightdope.com/u/Kaitlyn)\
**Post date:** [September 11, 2003, 8:12am UTC](https://boards.straightdope.com/t/grammar-mavens/200938/4 "2003-09-11T08:12:42Z")

</div>

It depends upon the context. The entire sentence would need to be included before I could answer accurately, but most likely “that” would be used.

“That” is used to introduce a restrictive modifier, which is not set off with commas. “Which” is used to introduce a non-restrictive modifier, which is set off with commas.

For example:

A. The committee will research the issue and write a report detailing their findings. The report, which is to be produced by the third of November, will be the basis of future decisions.

B. The committe will research the issue. The report that is to be produced will be the basis for future decisions.

If the clause gives extra information about the report, use “which” along with a pair of commas. If it identifies which report, use “that”, but no commas.

I hope that helps. If you post the entire sentence, it would be easier to answer, as context is everything here.

---

<div class="post-metadata">

**Author:** ![mhendo](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mhendo/32/3159_2.png) [@mhendo](https://boards.straightdope.com/u/mhendo)\
**Post date:** [September 11, 2003, 8:14am UTC](https://boards.straightdope.com/t/grammar-mavens/200938/5 "2003-09-11T08:14:57Z")

</div>

I should add that the absence of a comma after “report” suggests that “that” is the correct word, but i can only be sure after seeing the whole sentence.

---

<div class="post-metadata">

**Author:** ![mhendo](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mhendo/32/3159_2.png) [@mhendo](https://boards.straightdope.com/u/mhendo)\
**Post date:** [September 11, 2003, 8:16am UTC](https://boards.straightdope.com/t/grammar-mavens/200938/6 "2003-09-11T08:16:09Z")

</div>

Simulpost, **Number Six**. And with the same advice!

---

<div class="post-metadata">

**Author:** ![Achilles](https://avatars.discourse-cdn.com/v4/letter/a/dc4da7/32.png) [@Achilles](https://boards.straightdope.com/u/Achilles)\
**Post date:** [September 11, 2003, 12:11pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/7 "2003-09-11T12:11:50Z")

</div>

the full sentence is “[proprietary]…regarding the annual report which is due before end September 2003.”

From what I have gathered from the above, this should read “that”, not “which”. Gald I was correct for once 🙂

---

<div class="post-metadata">

**Author:** ![trabi](https://avatars.discourse-cdn.com/v4/letter/t/ecae2f/32.png) [@trabi](https://boards.straightdope.com/u/trabi)\
**Post date:** [September 11, 2003, 12:13pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/8 "2003-09-11T12:13:55Z")

</div>

> [@](#):
>
> “That” is used to introduce a restrictive modifier, which is not set off with commas. “Which” is used to introduce a non-restrictive modifier, which is set off with commas.

It’s a myth that it’s compulsory to use ‘that’ to introduce a restrictive clause. However, ‘that’ may not be used to introduce a nonrestrictive relative clause.  
Example: (From Longman’s Guide to English Usage):

I saw the car _that/which_ ran over your cat

I saw the car, which _(not that)_ was very fortunate  
The report which/that is to be published  
(as opposed to the report which/that isn’t going to be published, or the report that’s/which is going to be flushed down the toilet, etc…)  
So it’s a matter of personal taste and/or anality. 😉  
Interestingly, Longman’s Guide to English Usage appears to have a pretty misleading typo in its chapter on this:  
(Underlining mine)

[QUOTE]  
_From Longman’s Guide to English Usage_  
**That** cannot introduce a restrictive relative clause, but **Who, whom, whose** and **which** can introduce either a restrictive or a nonrestrictive one.

[QUOTE]

Should read …cannot introduce a nonrestrictive relative clause…

---

<div class="post-metadata">

**Author:** ![dantheman](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/dantheman/32/9421_2.png) [@dantheman](https://boards.straightdope.com/u/dantheman)\
**Post date:** [September 11, 2003, 12:34pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/9 "2003-09-11T12:34:24Z")

</div>

> [@](#):
>
> \*Originally posted by trabi \*  
> \*\*It’s a myth that it’s compulsory to use ‘that’ to introduce a restrictive clause. However, ‘that’ may not be used to introduce a nonrestrictive relative clause.  
> Example: (From Longman’s Guide to English Usage):
> 
> I saw the car _that/which_ ran over your cat
> 
> I saw the car, which _(not that)_ was very fortunate  
> \*\*

The second of those sentences is an example of poor grammar. The car is not fortunate; _seeing_ the car is. Therefore the sentence is incorrect and should be amended to “Fortunately, I saw the car.”

> [@](#):
>
> The report which/that is to be published  
> (as opposed to the report which/that isn’t going to be published, or the report that’s/which is going to be flushed down the toilet, etc…)  
> So it’s a matter of personal taste and/or anality. 😉

Not at all. “The report that is to be published” indicates that more than one report exists, but only the one to be published is being considered. “The report, which is to be published” indicates that there is only one report. It’s not a matter of personal taste at all, since the use of **that** or of **which** can produce diametrically opposed meanings.

(Of course, it’s worth noting that American English and British English may differ on this point.)

> [@](#):
>
> **Achilles**  
> the full sentence is “[proprietary]…regarding the annual report which is due before end September 2003.”

If you use “that,” then the sentence implies that there are multiple annual reports, one of which is due before end September 2003. I’m assuming there’s only one annual report, so the sentence should instead be:

“[proprietary]…regarding the annual report, which is due before end September 2003.”

---

<div class="post-metadata">

**Author:** ![raygirvan](https://avatars.discourse-cdn.com/v4/letter/r/ac91a4/32.png) [@raygirvan](https://boards.straightdope.com/u/raygirvan)\
**Post date:** [September 11, 2003, 12:46pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/10 "2003-09-11T12:46:19Z")

</div>

I agree with **dantheman**. Asking whether the sentence implies more than one object is a folksy way of identifying whether we’re talking about commenting or defining clauses.

“The cat which is on the mat is a tabby”. There’s only one cat; “which” introduces a clause commenting on it.

“The cat that is on the mat is a tabby”. As opposed to the black cat on the chair, the green cat on the ceiling, etc.; “that” introduces a clause defining it in relation to other possible items.

So, unless you have other annual reports due out at different times, “which” is right for the OP.

---

<div class="post-metadata">

**Author:** ![dantheman](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/dantheman/32/9421_2.png) [@dantheman](https://boards.straightdope.com/u/dantheman)\
**Post date:** [September 11, 2003, 12:51pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/11 "2003-09-11T12:51:02Z")

</div>

Don’t forget the comma preceding “which.” The word is introducing a parenthetical statement; without it, the sentence still makes sense:

“The cat is a tabby.”

“The cat, which is on the mat, is a tabby.”

“The cat that is on the mat is a tabby. The cat that is on the floor is a Siamese.”

---

<div class="post-metadata">

**Author:** ![refusal](https://avatars.discourse-cdn.com/v4/letter/r/ac91a4/32.png) [@refusal](https://boards.straightdope.com/u/refusal)\
**Post date:** [September 11, 2003, 1:33pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/12 "2003-09-11T13:33:57Z")

</div>

Fowler says you should use “who/which” in non-defining clauses, but should use normally “that” in defining clauses. Except you can use “who/which” to avoid awkwardness (e.g. don’t say “The dog that that man bought is fat.”, use “which that man bought”), and it is usual to use “who” rather than “that” with people.

However “The report which/that is to be produced regarding the annual report which is due before end September 2003.” isn’t a sentence at all, so you might want to reconsider that. Not to mention the ugliness of the 2 subordinate clauses.

---

<div class="post-metadata">

**Author:** ![trabi](https://avatars.discourse-cdn.com/v4/letter/t/ecae2f/32.png) [@trabi](https://boards.straightdope.com/u/trabi)\
**Post date:** [September 11, 2003, 6:30pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/13 "2003-09-11T18:30:23Z")

</div>

> [@](#):
>
> Not at all. “The report that is to be published” indicates that more than one report exists, but only the one to be published is being considered. “The report, which is to be published” indicates that there is only one report. It’s not a matter of personal taste at all, since the use of that or of which can produce diametrically opposed meanings.

No, it’s not the use of ‘which’ or ‘that’ which differentiates between a defining or non-defining relative clause, but whether or not there is a comma before it.

> [@](#):
>
> The car is not fortunate; seeing the car is. Therefore the sentence is incorrect and should be amended to “Fortunately, I saw the car.”

Er… I don’t think anyone reading that sentence would misunderstand it as meaning that the car is fortunate.

Also, I wouldn’t call it incorrect grammar either (bad grammer looks like this: I will to go to the theatre in today).

It’s even a perfectly common type of phrase, especially in spoken English or narrative: e.g. “I missed the performance, which was a shame because I’d paid 200 quid for the ticket.”

What the hell’s wrong with that? Does anyone really think that it means that the performance was a shame?

Yes, I could say “It was unfortunate that I missed the performance…” but it’s up to me which sentence construction I use.:rolleyes:

---

<div class="post-metadata">

**Author:** ![dantheman](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/dantheman/32/9421_2.png) [@dantheman](https://boards.straightdope.com/u/dantheman)\
**Post date:** [September 11, 2003, 6:40pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/14 "2003-09-11T18:40:41Z")

</div>

> [@](#):
>
> \*Originally posted by trabi \*  
> \*\*No, it’s not the use of ‘which’ or ‘that’ which differentiates between a defining or non-defining relative clause, but whether or not there is a comma before it. \*\*

Not at all true. The function of a clause is not determined by the punctuation surrounding it; that is, adding a comma does not make a clause restrictive or nonrestrictive. If that were true, then there would be no wrong punctuation - only wrong meanings.

> [@](#):
>
> Er… I don’t think anyone reading that sentence would misunderstand it as meaning that the car is fortunate.

You’re confusing the meaning of a sentence with the correction construction of a sentence. I’m not implying that people would misunderstand the meaning; I am saying that the sentence is constructed using poor grammar.

> [@](#):
>
> Also, I wouldn’t call it incorrect grammar either (bad grammer looks like this: I will to go to the theatre in today).

That’s right. That’s incorrect grammar. Incorrect grammar, however, does not take only one form.

> [@](#):
>
> It’s even a perfectly common type of phrase, especially in spoken English or narrative: e.g. “I missed the performance, which was a shame because I’d paid 200 quid for the ticket.”

You’re confusing proper grammar with conventional speaking. I understand your precise meaning. It’s not grammatically correct, but it’s not important that it be grammatically correct when speaking in an informal setting (as opposed to speaking formally, or writing).

---

<div class="post-metadata">

**Author:** ![trabi](https://avatars.discourse-cdn.com/v4/letter/t/ecae2f/32.png) [@trabi](https://boards.straightdope.com/u/trabi)\
**Post date:** [September 11, 2003, 7:24pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/15 "2003-09-11T19:24:39Z")

</div>

> [@](#):
>
> The function of a clause is not determined by the punctuation surrounding it; that is, adding a comma does not make a clause restrictive or nonrestrictive. If that were true, then there would be no wrong punctuation - only wrong meanings.

So [this site](http://esl.about.com/library/grammar/blgr_relative_decide.htm) and[all](http://tolearnenglish.com/free/news/0grelatives2.htm) [these](http://pedagogie.ac-aix-marseille.fr/disciplines/anglais/englishonline/GrammarAndHelp/RelativePronoun/RelPronDoc.htm) [cites](http://pioneer.netserv.chula.ac.th/~pkanchan/html/pronoun.htm) (to name just a few) are wrong?

> [@](#):
>
> Correct punctuation is essential in non-defining relative clauses. If the non-defining relative clause occurs in the middle of a sentence, a comma is put before the relative pronoun and at the end of the clause. If the non-defining relative clause occurs at the end of a sentence, a comma is put before the relative pronoun.

Adding a comma does make a clause nonrestrictive. It’s a simple fact of life. If There’s no way I can persuade you of this then we’ll have to agree to differ. 🙂

> [@](#):
>
> Incorrect grammar, however, does not take only one form.

…And nor does correct grammar, which is exactly my point.

> [@](#):
>
> You’re confusing proper grammar with conventional speaking. I understand your precise meaning. It’s not grammatically correct, but it’s not important that it be grammatically correct when speaking in an informal setting (as opposed to speaking formally, or writing).

Are you saying that grammar doesn’t apply when speaking, even if you’re speaking conventionally? In that case, we might as well say that contractions (I’ll, he’s, etc.) are bad grammar, because they express a spoken form of English. What about writing dialogue? It is YOU who are confusing proper grammar with what you consider to be elegant sentence construction.

---

<div class="post-metadata">

**Author:** ![mhendo](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mhendo/32/3159_2.png) [@mhendo](https://boards.straightdope.com/u/mhendo)\
**Post date:** [September 11, 2003, 8:24pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/16 "2003-09-11T20:24:47Z")

</div>

> [@](#):
>
> \*Originally posted by raygirvan \*  
> \*\*I agree with **dantheman**. Asking whether the sentence implies more than one object is a folksy way of identifying whether we’re talking about commenting or defining clauses.
> 
> “The cat which is on the mat is a tabby”. There’s only one cat; “which” introduces a clause commenting on it.
> 
> “The cat that is on the mat is a tabby”. As opposed to the black cat on the chair, the green cat on the ceiling, etc.; “that” introduces a clause defining it in relation to other possible items.
> 
> So, unless you have other annual reports due out at different times, “which” is right for the OP. \*\*

This strategy might be “folksy,” but it is also incorrect. The use of a restrictive or a non-restrictive clause does not depend on whether you can see another cat at the time of making the statement. Nor does it depend on whether there is another report being prepared at the same time. The purpose of a restrictive clause is simply to distinguish the subject of the sentence from similar items with which it might be confused.

While there may be no other report being produced right now at **Achilles** ’s company, this does not mean that “which” is automatically the correct word to use in this case. Take these two sentences:

**"The report that will be presented at the meeting is the product of three months of hard work."**

This is an example of a restrictive clause, because it restricts the scope of the sentence to one particular report–even if there may or may not be other reports underway in the company.

**The report, which will be presented at the meeting, is the product of three months of hard work."**

The words between the commas constitute an example of a non-restrictive clause, giving additional, non-defining information.

And it is here that i must agree with **dantheman** on the issue of punctuation. In correct grammar, “that” should introduce a restrictive clause, and “which” should introduce a non-restrictive clause. As **dantheman** says, the punctuation does not dictate the type of clause being used; however, the type of clause being used does definitely dictate the punctuation. As Bryan Garner says in _A Dictionary of Modern American Usage_, “Your choice, then, is between comma-_which_ and _that_.” He also adds a general rule of thumb: “Use **that** whenever you can.”

He gives an excellent example, using two almost-identical sentences:

1. “All the cars that were purchased before 1995 need to have their airbags replaced.”

2. “All the cars, which were purchased before 1995, need to have their airbags replaced.”

Two quite different meanings, brought about by a change in one word, and a _necessary_ change in punctuation. But the change in the clause _causes_ the change in punctuation, not the other way around. **Trabi** argues that “Adding a comma does make a clause nonrestrictive. It’s a simple fact of life.” Well, no. Because adding a comma to the first sentence would make it read:

“All the cars, that were purchased before 1995, need to have their airbags replaced.”

This is just BAD grammar, not an example of a non-restrictive clause.

Garner concludes his entry: “The distinction between _that_ and _which_ makes good sense. It enhances clarity. And the best American editors follow it.”

---

<div class="post-metadata">

**Author:** ![jormundgondir](https://avatars.discourse-cdn.com/v4/letter/j/dfb087/32.png) [@jormundgondir](https://boards.straightdope.com/u/jormundgondir)\
**Post date:** [September 11, 2003, 9:03pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/17 "2003-09-11T21:03:05Z")

</div>

I assume just “the report to be produced” is right out?

---

<div class="post-metadata">

**Author:** ![Quartz](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/quartz/32/267_2.png) [@Quartz](https://boards.straightdope.com/u/Quartz)\
**Post date:** [September 11, 2003, 9:06pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/18 "2003-09-11T21:06:20Z")

</div>

> [@](#):
>
> … regarding the annual report which is due before end September 2003

The phrase ‘which is’ is superfluous. And you should clear up the sentence. I’d suggest:

_… the annual report due before 30th September 2003_

---

<div class="post-metadata">

**Author:** ![dantheman](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/dantheman/32/9421_2.png) [@dantheman](https://boards.straightdope.com/u/dantheman)\
**Post date:** [September 11, 2003, 9:09pm UTC](https://boards.straightdope.com/t/grammar-mavens/200938/19 "2003-09-11T21:09:58Z")

</div>

> [@](#):
>
> \*Originally posted by trabi \*  
> So [this site](http://esl.about.com/library/grammar/blgr_relative_decide.htm) and[all](http://tolearnenglish.com/free/news/0grelatives2.htm) [these](http://pedagogie.ac-aix-marseille.fr/disciplines/anglais/englishonline/GrammarAndHelp/RelativePronoun/RelPronDoc.htm) [cites](http://pioneer.netserv.chula.ac.th/~pkanchan/html/pronoun.htm) (to name just a few) are wrong?

🙂 Well, strange as it may seem, there are many incorrect sites on the Internet, particularly those that discuss grammar.

> [@](#):
>
> Adding a comma does make a clause nonrestrictive. It’s a simple fact of life. If There’s no way I can persuade you of this then we’ll have to agree to differ. 🙂

**mhendo** explained it better than I could when he (he?) said “the punctuation does not dictate the type of clause being used; however, the type of clause being used does definitely dictate the punctuation.”

You see? It works the other way around. Punctuation can alter the meaning of a sentence, but it doesn’t change a clause from restrictive to nonrestrictive.

> [@](#):
>
> …And nor does correct grammar, which is exactly my point.

If that was your point, it’s a good one. Grammar is not like mathematics.

> [@](#):
>
> Are you saying that grammar doesn’t apply when speaking, even if you’re speaking conventionally?

No. I’m saying it’s often not as important when the words are spoken as when the words are written. For example, when I call up a friend and do not use correct grammar, he’ll still get the meaning of my speech (unless I’m rambling incoherently). However, because writing is two-dimensional (i.e., it’s tougher to gague tone and unstated meaning than for spoken word), in formal speeches and in most forms of writing - except when vernacular or slang is purposely employed) it’s important that the correct meaning be conveyed concisely. To do this, we use correct grammar.

---

<div class="post-metadata">

**Author:** ![trabi](https://avatars.discourse-cdn.com/v4/letter/t/ecae2f/32.png) [@trabi](https://boards.straightdope.com/u/trabi)\
**Post date:** [September 12, 2003, 10:31am UTC](https://boards.straightdope.com/t/grammar-mavens/200938/20 "2003-09-12T10:31:00Z")

</div>

Okay, I haven’t expressed myself clearly enough. First of all, here is a summary of what I have been trying to say, to avoid further confusion:

1. All nonrestrictive relative clauses are introduced by a comma.

2. Nonrestrictive clauses may not use ‘that.’

3. All restrictive clauses omit the comma

4. Although some style guides and usage guides (especially ones for US editors, it would seem) prefer the use of ‘that’ for restrictive clauses, this is a matter of preference, and not written in stone, therefore

5. restrictive clauses may use a comma followed by either ‘that’ or ‘which,’ depending on preference\*.

6. So logically, _in the case of a sentence using ‘which’_, the existance of a comma is what defines whether or not it is a restrictive clause, and

7. The use of ‘that’ does not entail a restrictive clause regardless of the punctuation. All it does for certain is rule out a nonrestrictive clause.

\*[SUB]although, for example, ‘which’ may not be used to substitute ‘who’ in the case of a person, whereas ‘that’ may[/SUB].  
I had been intending to list these points at the end, but this post got so long that I figured nobody would read that far.

Here a a few of my observations re the last couple of posts: (apologies for the length… ☹ )

> [@](#):
>
> Trabi argues that “Adding a comma does make a clause nonrestrictive. It’s a simple fact of life.” Well, no. Because adding a comma to the first sentence would make it read:  
> “All the cars, that were purchased before 1995, need to have their airbags replaced.”

Actually, in an earlier post I pointed out that the word ‘that’ may not be used to introduce a non-defining relative clause. But yes, adding a comma makes that particular example sentence incorrect.

> [@](#):
>
> While there may be no other report being produced right now at Achilles’s company, this does not mean that “which” is automatically the correct word to use in this case. Take these two sentences:
> 
> “The report that will be presented at the meeting is the product of three months of hard work.”
> 
> This is an example of a restrictive clause, because it restricts the scope of the sentence to one particular report–even if there may or may not be other reports underway in the company.
> 
> The report, which will be presented at the meeting, is the product of three months of hard work."

No-one said that ‘which’ is automatically the right word to use. I merely said that if, as the lack of a comma suggested, it is to be a restrictive relative clause, either ‘that’ or ‘which’ can be used.

And yes, one of the examples is restrictive, one is nonrestrictive. However, if you replace ‘that’ with ‘which’ in the first example, it remains restrictive. Conversely, if you remove the comma (actually, both of them) in the second example, it becomes restrictive.

> [@](#):
>
> mhendo explained it better than I could when he (he?) said “the punctuation does not dictate the type of clause being used; however, the type of clause being used does definitely dictate the punctuation.”

Yes, that’s fair enough - chicken and egg - but it doesn’t automatically follow that, as you put it:

> [@](#):
>
> …It works the other way around. Punctuation can alter the meaning of a sentence, but it doesn’t change a clause from restrictive to nonrestrictive.

Yes it can (as long as it doesn’t result in a nonrestrictive clause using ‘that’). Changing the punctuation can alter the meaning of the sentence by changing a restrictive clause to a nonrestrictive one.

e.g.:

a. The report which I wiped my bottom with…

b. The report, which I wiped my bottom with…

> [@](#):
>
> Garner concludes his entry: “The distinction between that and which makes good sense. It enhances clarity. And the best American editors follow it.”

For a start, I am neither American, nor an editor. Also this line: “The distinction between that and which makes good sense” is the giveaway here. The guy is obviously writing in support of a grammatical convention that he happens to prefer. “It _enhances_ clarity,” he says. If someone says that ABS enhances braking performance, it doesn’t make it compulsory to fit it to your car.

Another quote from the Longman’s Guide:

> [@](#):
>
> “…where there may be a danger of misunderstanding a restrictive as a nonrestrictive clause, it may be wise to use that.”

Doesn’t this suggest to anyone that we have a choice?

BTW:

> [@](#):
>
> 🙂 Well, strange as it may seem, there are many incorrect sites on the Internet, particularly those that discuss grammar.

Yes, but the source of reference I originally cited was Longman’s Guide to English Usage. I also consulted Longman’s Students Guide to English Grammar. I only posted the internet sites (not to say that they aren’t correct) to save me leafing through a lot of grammar guides and typing them all in manually…

> [@](#):
>
> (i.e., it’s tougher to gague tone and unstated meaning than for spoken word), in formal speeches and in most forms of writing - except when vernacular or slang is purposely employed) it’s important that the correct meaning be conveyed concisely. To do this, we use correct grammar.

Sure. Mhendo’s example sentence: “All the cars, that were purchased before 1995, need to have their airbags replaced.”

IS bad grammer. It doesn’t follow the rules, it has a nonrestrictive clause introduced by ‘that.’ The only way we know it’s incorrect, by the way, is because it has a comma, which pretty much proves my point that it’s the comma which defines it as a nonrestrictive clause. If it didn’t have one we’d just assume it was a defining relative clause using ‘that.’

But,

“I lost my key, which was an irritation” is not “bad grammar,” and I wouldn’t call it slang or vernacular either. It in fact uses the rules of grammar to achieve a sentence construction which, although Dantheman may find clumsy, could quite possibly be desirable - even in written form - for reasons of emphasis, for example, or in a narrative - i.e. precisely to make it easier to “gauge tone.”

Longman’s again (since you are so wary of internet sources):

"In some nonrestrictive clauses the relative pronoun _which_ refers back not to a noun but to a larger unit: \*He failed his driving test three times,\ ***which must be discouraging.** (his having failed his driving test three times must be discouraging). Where it may be unclear which unit the relative pronoun refers to, it would be better to rephrase the sentence: \*He failed his driving test,\ ***which annoyed him** (Did the failure or the test annoy him?). Rephrase: _He failed his driving test, and the failure annoyed him._  
[/QUOTE]

Okay, apart from the fact that the rephrased sentence is really clumsy (although not grammatically incorrect), nothing here suggests that the aforementioned sentence structure is “ungrammatical.”

[Next page](https://boards.straightdope.com/t/grammar-mavens/200938.md?page=2)
