14

I was reading a book called "The Penguin Guide to Punctuation" by R. L. Trask. It seems that the book doesn't explain all the uses of comma. It says "There are four uses of the comma, called the listing comma, the joining comma, the gapping comma and bracketing commas".

Here's a brief summary of the rules as described in book:

The listing comma: Use a listing comma in a list wherever you could conceivably use the word and (or or) instead. Do not use a listing comma anywhere else. Example: The Three Musketeers were Athos, Porthos and Aramis.

The joining comma: Use a joining comma to join two complete sentences with one of the words and, or, but, yet or while. Do not use a joining comma in any other way. Example: Norway has applied to join the EC, and Sweden is expected to do the same.

The gapping comma: Use a gapping comma to show that words have been omitted instead of repeated. Example: Some Norwegians wanted to base their national language on the speech of the capital city; others, on the speech of the rural countryside.

The bracketing comma: Use a pair of bracketing commas to set off a weak interruption. Example: Schliemann, of course, did his digging before modern archaeology was invented.

In the following examples, which type of comma is being used?

After two hours, the train came to the station.
Before I could talk to her, she was gone.

It appears to me that the above usage of commas doesn't meet any of the comma rules described in the book.

Marthaª
  • 32,910
cpx
  • 601
  • @cpx: Yes, thanks. Where I found the question almost frustrating before; I now find it deserving of an upvote. – J.R. Dec 24 '12 at 01:18
  • 1
    Dunno about you, but I would read that totally incorrect advice about the listing comma, and run far, far away from this book. (99% of the time, lists are much clearer if you use a comma between all of the items, even the last one. In the few cases that are exceptions, omitting the last comma isn't as much help as rewriting the sentence.) – Marthaª Dec 24 '12 at 02:56
  • 1
    For anyone who would like to read further, the full text is available here: http://www.informatics.sussex.ac.uk/department/docs/punctuation/node00.html – Barrie England Dec 24 '12 at 08:09

5 Answers5

7

These are bracketing commas, 'used to mark off a weak interruption of the sentence'. As Trask later explains:

Sometimes a weak interruption comes at the beginning or at the end of its sentence. In such a case, one of the two bracketing commas would logically fall at the beginning or the end of the sentence — but we never write a comma at the beginning or at the end of a sentence. As a result, only one of the two bracketing commas is written in this case:

All in all, I think we can say that we've done well.

I think we can say that we've done well, all in all.

As further examples of weak interruptions at the beginning of the sentence, he gives:

Having worked for years in Italy, Susan speaks excellent Italian.

Unlike most nations, Britain has no written constitution.

Although Mercury is closer to the sun, Venus has the higher surface temperature.

After capturing the Aztec capital, Cortés turned his attention to the Pacific.

He also says

In many cases a weak interruption does not absolutely require bracketing commas. Thus either of the following is fine:

Shortly before the war, he was living in Paris.

Shortly before the war he was living in Paris.

Barrie England
  • 140,205
6

According to Trask himself, the commas in your examples are bracketing commas used to separate the bolded interruptions. Bracketing commas can appear as a pair to enclose an interruption in the middle of a sentence. They can also appear alone to separate an interruption at the start or end of a sentence. They serve to add more information to sentences that are complete.

After two hours, the train came to the station.

Before I could talk to her, she was gone.

  • Trask says that bracketing commas are used to set off "weak interruptions". Neither the introductory phrase nor the introductory clause in the OP's example sentences is a weak interruption analogous to "of course", which is totally expendable without losing any meaning. Drop the introductory phrase or clause, and meaning is lost, unless it's kept by adding a separate sentence. –  Dec 24 '12 at 03:41
  • 1
    I certainly don't think these two are "bracketing commas". I don't have his complete text to confirm exactly how he defines his four types, but it's obvious to me of the four types as listed by OP, these must be "joining commas". All they're doing is marking the "artificial join" created by stylistic reordering of component clauses (@Bill's "sentence constituents") which would otherwise have been sequenced as "The train came to the station after two hours" and "She was gone before I could talk to her" (in which form they obviously don't need commas at all). – FumbleFingers Dec 24 '12 at 03:54
  • 2
    Okay - you've convinced me! I'll leave the earlier comment (to my eternal shame! :), but let it be noted that I NO LONGER ENDORSE MY PREVIOUS COMMENT. To quote a specific example from Trask of an "unpaired weak interruption comma", "The pronunciation of English is changing rapidly, we are told." is exactly the same construction as OP's examples here. Top marks to you, and nul points to OP for not reading the very guide he's asked us about. – FumbleFingers Dec 24 '12 at 04:30
4

The function of these commas is to separate sentence constituents (which means that it's what Trask calls "a joining comma").

The first constituent in S1 is an introductory adverbial (prepositional) phrase; the first constituent in S2 is an introductory subordinate adverbial clause. Sometimes they need to be separated to disambiguate the sentence. In S1, the comma isn't needed to disambiguate: the sentence is clear without a comma, so it's optional. In S2, the comma separates two clauses, the subordinate first clause, and the independent (main) second clause. This is a typical example of the joining comma, and an obvious terminological problem because it both joins and separates.

Sometimes the comma provides the reader a cue to pause slightly when reading.

The four categories listed are incomplete. Ignore Trask and look at other explanations of commas. There are plenty that more clearly and completely tell you how to "properly" [intentional scare quotes] use punctuation.

  • 4
    And please, when you ignore Trask, do not run instead to Strunk & White. – MetaEd Dec 24 '12 at 01:11
  • There are much better explanations of how to use commas than Strunk & White's. Here's one. There are others. Contemporary rules & suggestions are more appropriate than ancient rules & regulations. –  Dec 24 '12 at 01:23
  • 2
    @J: That makes Trask's piece of work not so excellent, then. It is, as I said, incomplete. Other grammarians who use the "joining comma" terminology illustrate its use by joining a subordinate & a main clause. In any case, terminology is a pain in the butt because it has to be precisely & clearly defined. When it isn't, it's useless. Terminology alone doesn't necessarily explain anything. Parenthetical remarks are interruptions because they're unnecessary. Subordinate clauses are not "interruptions": they're complete sentences linked to the main clause by a subordinating conjunction. –  Dec 24 '12 at 03:35
  • 1
    @Bill: Check out Jasper's link (to Trask's own text), and my comments thereunder. You're free to disagree with Trask's categorisation scheme, of course (he is pretty highly-rated, though). But if you want to classify OP's examples according to that scheme, it really is beyond dispute that Trask says they're (unpaired) *weak interruption* commas. He gives examples thereof which are syntactically identical to OP's sentences. – FumbleFingers Dec 24 '12 at 04:37
  • @FumbleFingers: Different theories, different terminology. Terminology doesn't explain anything, especially when different theories disagree on what the terminology means. Knowing whether it's a bracketing or a joining comma is, it seems to me, irrelevant. Knowing what it does in the sentence, why it's there, is important. I can't agree that an example of "bracketing commas" is a single comma that separates a subordinate introductory phrase or clause. It says commas, which means 2 commas to me. "Cat" in "cat food" is a noun that functions as an adjective = "noun" means zilch. –  Dec 24 '12 at 07:00
  • 1
    See my answer. These are bracketing commas, not joining commas. Trask describes a joining comma as being ‘used to join two complete sentences into a single sentence, and it must be followed by a suitable connecting word. The connecting words which can be used in this way are and, or, but, while and yet.’ This is clearly not applicable in the OP’s examples. FWIW, I find Trask’s guidance invaluable. Until his death in 2004 he was Professor of Linguistics at the UK’s University of Sussex. He was born in New York State. – Barrie England Dec 24 '12 at 08:29
  • @Barrie: I'm sure you know what you're talking about, but I see all cognitive maps of the linguistic world as subjective. I see all linguistics taxonomies as artificial divisions proposed by linguists who espouse one theory or another. This is not like the theory of evolution, which deals with natural science & explains where species come from & how they evolved (except where there are gaps in the fossil record). Taxonomy does not explain anything; it merely labels things in a logical way. Most of the time it makes sense, but sometimes it doesn't. I haven't read Trask's book. –  Dec 24 '12 at 10:07
  • @JasperLoy: I read what the OP posted & it seemed to me (& to Marthaª) that Trask's categories were screwy. I can "dismiss the works of a native speaker, linguistics professor, & author of several books": I'm not a slave to "authority"--I can think for myself & agree or disagree with Prof Trask. +, I'm opinionated, I edit technical language every day & have for 15+ years, teach writing, & am an amateur linguist (with a dated theoretical linguistics MA). I don't pretend to be an expert. I see some expert amateurs here who know a lot more than I do about linguistics, but I have no fears –  Dec 24 '12 at 13:39
  • about expressing my opinions. I willing to be persuaded that I'm wrong; I'll change my erroneous opinions, but I'll argue my point until I see that I've lost the argument. I won't argue what Trask thinks. You & Barrie & others have convinced me of what he thought, but I think he's wrong, which is why I suggested that the OP ignore him. Barrie finds him helpful. Great. I can't argue with that. I prefer someone who agrees with how I think "commas should be classified". Some here use the CGEL as a bible; I find it overly politicized & prefer other authorities. None are perfect, though. –  Dec 24 '12 at 13:47
  • We aren't so far apart. –  Dec 24 '12 at 13:59
  • @Bill Franke: I don't disagree with a word of your last comment, but I still think if we allow that OP is asking how does Trask* classify these commas?* we have to accept that Jasper's answer is correct. Trask would not /does not call OP's examples "joining commas". – FumbleFingers Dec 24 '12 at 14:50
  • @Fumble: Yes, you're right about that. I just think that Trask is inconsistent and that it doesn't matter what he calls his commas. The OP should ignore him. Jasper's answer is correct: Trask calls them bracketing commas. But that's not obvious from what little of Trask's explanation the OP posted. It also seems that the OP isn't reading this book because he has to for a grammar class but for his own edification. That's why I said the OP should ignore him. If he has to answer a teacher's question about what kind of comma Trask thinks it is, he should use Jasper's answer. Jasper's not wrong. –  Dec 24 '12 at 15:03
1

Instead of focusing on the arcana of punctuation, I suggest you be guided by one primary criterion: Does the comma enable the reader to read the sentence smoothly, without having to back up and re-read it because it did not scan well the first time?

For example, which is easier to scan, A or B?

A: I counseled John a former neighbor of mine not to purchase the house.
B: I counseled John, a former neighbor of mine, not to purchase the house.

I guess you'd say the example includes two "bracketing"commas. As for that last "listing" comma in a series of three or more, be guided by how easily the sentence scans--either with or without that last comma. Which of the following scans best, in your opinion?

A: Joan brought a casserole, some rolls, a tasty dessert, and iced tea to the dinner.
B: Joan brought to the dinner a casserole, some rolls, a tasty dessert, and iced tea.
C. A casserole, some rolls, a tasty dessert, and iced tea were the items Joan brought to the party.
D. Joan brought a casserole, some rolls, a tasty dessert and iced tea to the party.

How should the following be punctuated?:

Ahh coffee the nectar of the gods given to humanity is my favorite beverage.

What about this one?

His middle name Reginald he seldom uses. 

(The sentence states a fact about Reggie; it does
not address someone named Reginald.)

              OR

He seldom uses his middle name Reginald.

I hope the above proves helpful.

rhetorician
  • 19,383
  • Yes, I think this is helpful. It asks for a functional analysis, not an analysis of the taxonomy. Functionalism may be retro (19th century), but it still has some value. Seems to me that "good English" is "English that works", "English that has the effect that its author intended it to have". This is not the "intentional fallacy", because I'm not talking about meaning. It's a functional argument: Does it do what it was intended to do? Maybe other English would have been more effective, but that's not the point either. –  Dec 24 '12 at 13:57
-1

It's a joining comma that adds to sentences with after and before.

RegDwigнt
  • 97,231
Bhushan Firake
  • 223