quotations and punctuation

The question of proper placement of quotation marks relative to other punctuation, though it has received much attention, remains in a cloud of confusion, at least from the American perspective. Logic would dictate that a mark of punctuation should not appear to the left of a closing quotation mark unless it properly belongs to the quotation itself rather than to the enclosing sentence. This is essentially the position held by most British authorities, e.g., the style guide of The Times:

Note that punctuation marks go inside the inverted commas if they relate to the words quoted, outside if they relate to the main sentence, eg, She is going to classes in 'health and beauty'. If the whole sentence is a quotation, the final point goes inside, eg, 'Beauty is truth, truth beauty.'*
Corresponding American sources, such as the Associated Press Stylebook, offer conflicting advice:
Follow these long-established printers' rules:
   — The period and the comma always go within the quotation marks.
   — The dash, the semicolon, the question mark and the exclamation point go within the quotation marks when they apply to the quoted matter only. They go outside when they apply to the whole sentence.
Eric Partridge comments on the controversy:
It is difficult to see why American practice should agree with the British practice, with logic, with good sense, in putting outside of the quotation marks indicating quoted matter all colons, semicolons, dashes, question and exclamation marks that don't belong to the quoted matter, yet should put inside the quotation marks even those commas and periods which don't belong to the quoted matter at all but which concern only the sentences as a whole.
While it is generally true that contemporary British and American commentators tend to split along the same lines as the two style guides referenced above, the special treatment of commas and periods is not an American invention, as Partridge would suggest. In fact, it is not difficult to find instances of it on either side of the Atlantic. Just five years before the above observation appeared in print, Bertrand Russell, an Englishman noted for his grasp of logic, managed to slip the following past his British publisher:
A relation is said to be "one-one" when, if x has the relation in question to y, no other term x' has the same relation to y, and x does not have the relation to any term y' other than y. When only the first of these two conditions is fulfilled, the relation is called "one-many"; when only the second is fulfilled, it is called "many-one."*
Henry Alford, identified by Partridge as "the spiritual father of the Fowler brothers", and whose sentiment toward the American influence was most clearly expressed by his reference to "the process of deterioration which our Queen's English has undergone at the hands of the Americans"*, consistently quoted not only commas and periods, but colons and semicolons as well:
The long "u" has a power; we may say "a unit," "a university," because the first syllable sounds as if it began with "you," and "y" has here the power of a consonant. But the short "u," as in "humble," is not one of those vowels which require a consonant to enunciate them: one could not say "a unlearned man:" and I must therefore still maintain that the occurrence of "thy humble," and "thine unworthy," shows that the "h" was meant to be aspirated in the former case, as we know it was not in the latter.*
Partridge himself includes the following among his examples of "British practice at its most modern and at its best":
'Modern British practice,' he said, 'appears to be more consistent and, in a sense, more ruthless than American.'
Does he really mean to insist that the first comma in this sentence properly belongs to the quoted matter rather than to the sentence as a whole? This example, though in direct contradiction to its author's characterization of correct practice, illustrates perhaps the only aspect of the issue that has been treated fairly consistently across the Atlantic, namely, the placement of punctuation relative to a quotation that is followed by a phrase of attribution. It is remarkable, not only that Partridge has so studiously ignored this aspect of the problem, but that the underlying rules, though implicitly followed by most contemporary writers of English, have never been explicitly laid out, as far as I can tell, until now:
(1) If the quotation would logically end with a question mark, an exclamation point, or a dash (in the case of quoted speech, indicating that the speaker was abruptly interrupted), then that mark appears within the quotation marks.
"Heavens!" he exclaimed.

"But—" he protested.
(2) If the quotation would logically end with a period, then a comma is inserted within the quotation marks and the attributive phrase is followed by a period.
"No man is an island," he mused.
(3) When an interpolated attributive phrase appears within a quotation, thus requiring two sets of quotation marks, the sentence is punctuated according to the mark of punctuation that would otherwise occur at the end of the initial segment of the quotation. If that mark is a question mark or an exclamation point, then it is handled as in (1) above.
"Landlord!" said I. "What sort of a chap is he—does he always keep such late hours?"
If it is a period, colon, semicolon, comma, or dash, then it is placed after the attribution and a comma is inserted to the left of the closing quotation mark.
"Landlord," I whispered, "that aint the harpooner is it?"

"Broke," said I—"broke, do you mean?"
If there would have been no mark at the end of the initial segment, then two commas are used, one placed to the left of the quotation mark and one following the attribution.
'Modern British practice,' he said, 'appears to be more consistent and, in a sense, more ruthless than American.'
It is also remarkable that Partridge, whose writing has been observed to conform to these rules and in particular, includes the last example above, should be so puzzled by the special allowances that American practice makes for commas and periods. Clearly, the basis for this distinction lies in the difference in levels of semantic content between various punctuation marks. A question mark or exclamation point dramatically alters the meaning of the sentence or phrase that it terminates, and it is therefore critical whether it is interpreted as applying to a quotation or to the sentence that contains it. To a lesser extent, a colon or semicolon also carries meaning: it may indicate a semantic relation between the clauses that it separates, as in the present sentence. In contrast, a comma or period is commonly used merely to indicate a pause, and may serve this purpose equally well, it might be argued, from either side of a quotation mark.

It is not our objective, however, to justify this questionable notation, which we grudgingly admit as being the current prevailing American practice. It happens often enough that idiom or convention leaves us no choice but to deny logic and reason. In the present case, however, there is sufficient ambivalence, even among American commentators, to leave the door open for dissent.

Lyle Spencer, however dated (1914), is one noted American authority who sides with logic on this issue:

Quotation marks are regularly put outside other marks of punctuation when those marks refer to the quotation alone; otherwise they are put inside. But most publishers place the quotation marks outside commas and periods even when those marks should not be included, as when a single word or a short phrase is quoted at the close of a sentence.
Strunk and White also recognize the disparity between logic and usage, observing that "typographical usage dictates that the comma be inside the marks, though logically it often seems not to belong there."

Webster's 3rd, while generally supporting the American convention, allows that it may be superseded "when a logical or exact distinction is desired in specialized work in which clarity is nore important than usual (as in this dictionary)".

The Chicago Manual of Style, 13th edition, insists that in general, "periods should be placed within the quotation marks", but concedes that "in works of philosophy and theology, terms having special philosophical or theological meaning are often enclosed in single quotation marks. [Hey, what about semiotics and paleontology?] Following punctuation is placed outside the quotation marks...." The 15th edition, curiously, backs off from this particular exception on the grounds that it "requires extreme authorial precision", but allows that the rule defies logic and need not be followed when "scholarly integrity" is important or when "inaccuracy or ambiguity is intolerable."

Certainly, all of this equivocation leaves room for those of us who believe that clarity, integrity, and precision are generally good things, and that inaccuracy and ambiguity are not, to follow logic in this instance rather than dubious convention. That is, apart from the exceptions noted above in connection with phrases of attribution, which appear to be universally observed, there is no compelling reason not to follow the simple rule proposed by The Times, as displayed at the beginning of this article. If all that is lacking is a firm stand taken by an enlightened American authority, then consider that done.

Speaking of logic, have I already asked the question, "Have you read Raymond Smullyan's 'What is the Name of This Book?'?"?