They have a variety of forms in different languages and in different media.

Usage
Quotations and speech

Single or double quotation marks are used to denote either speech or a quotation. Neither style is an absolute rule though double quotes are preferred, but a publisher's or even an author's style may take precedence. The important rule is that the style of opening and closing quotes must be matched.

For speech within speech, the other is used as inner quotation marks.

Sometimes, quotations are nested in more levels than inner and outer quotation. Nesting levels up to five can be found in the Bible. In these cases, questions arise about the form (and names) of the quotation marks to be used. The most common way is to simply alternate between the two forms.

If such a passage is further quoted in another publication, then all of their forms have to be shifted over by one level.

In most cases, quotations that span multiple paragraphs should be block-quoted, and thus do not require quotation marks. Quotation marks are used for multiple-paragraph quotations in some cases, especially in narratives. The convention in English is to give the first and each subsequent paragraph opening quotes, using closing quotes only for the final paragraph of the quotation. The Spanish convention, though similar, uses closing quotes at the beginning of all subsequent paragraphs beyond the first.

When quoted text is interrupted, such as with the phrase he said, a closing quotation mark is used before the interruption, and an opening quotation mark after. Commas are also often used before and after the interruption, moreso for quotations of speech than for quotations of text.

It is generally considered incorrect to use quotation marks for paraphrased speech: Emphasis and irony

Another important use of quotation marks is to indicate or call attention to ironic or apologetic words. Ironic quotes can also be called scare, sneer, shock, or distance quotes. Ironic quotes are sometimes gestured in verbal speech using air quotes.

Ironic quotes should be used with care. Without the intonational cues of speech, they could obscure the writer's intended meaning. They could also be confused easily with quotations.

In a similar sense, quotes are also used to indicate that the writer realizes that the word is not being used in its (currently) accepted sense.

Quotes are also sometimes used for emphasis in lieu of underlining or italics, most commonly found on signs or placards. This is generally discouraged not only because it is historically an improper usage, but also because it is easily confused with ironic or altered-usage quotation.

Notice how easily the statement above could be construed to imply that the word fresh is not being used with its everyday meaning.

Either quotes or italic type can emphasize that an instance of a word refers to the word itself rather than its associated concept.

In HTML/XHTML, a semi-semantic way to tell regular quotes from distant quote is to use the q tag for the former, while using actual quotations marks for distant quotes. It is only semi-semantic because the behaviour for non-eye-centered media is uncertain. Still, it is safer than not introducing any distinction at all.

Titles of artistic works

Quotation marks, rather than italics, are generally used for the titles of shorter works. Whether these are single or double is again a matter of style:

short fiction, poetry, etc.: Arthur C. Clarke's “The Sentinel� book chapters: The first chapter of 3001: The Final Odyssey is “Comet Cowboy� articles in books, magazines, journals, etc.: “Extra-Terrestrial Relays,� Wireless World, October 1945 album tracks, singles, etc.: David Bowie's “Space Oddity�

Typographical considerations
Punctuation

The American convention is for sentence punctuation to be included inside the quotation marks, even if the punctuation is not part of the quoted sentence, while the British style is to have the punctuation outside the quotation marks for small quoted phrases:

Despite what is sometimes written on discussions of punctuation, British positioning is the same as American in complete quoted speech:

In contrast to most areas of publication, in subjects such as chemistry and software documentation it is conventional to include only the quoted string within the quotes. This avoids ambiguity with regard to whether a punctuation mark belongs to the quote:

In American English, commas and periods always go inside the quotation marks, no matter the circumstance:

This is changing, however, due to the influence of computer science as explained above. (See BNF rules for describing formal languages). Therefore the following is becoming acceptable:

Question marks and exclamation marks must rely on logic to determine whether they go inside or outside:

(Note that in the above sentences, only one punctuation mark is used at the end of each sentence. Regardless of its placement, only one end mark (?, !, or .) can end a sentence in American English, whereas in British English, the combination ?�. is acceptable.)

In addition, end marks always go inside single quotes, too:

When a quoted text not appropriate for block quoting extends beyond the length of its originated paragraph without pause, the paragraph forgoes a closing quotation mark while the following paragraph begins with an opening quotation mark. This continues through all additional paragraphs until a closing quotation mark concludes the quote:

Spacing
In English, when a quotation follows other writing on a line of text, a space precedes the opening quotation mark unless the preceding symbol, such as a dash, requires that there be no space. When a quotation is followed by other writing on a line of text, a space follows the closing quotation mark unless it is immediately followed by other punctuation within the sentence, such as a colon or closing punctuation. These exceptions are ignored by some Asian computer systems that systematically display quotation marks with the included spacing.

In Chinese, the spacing is irrelevant since all characters, including punctuation, are the same width.

There is generally no space between an opening quotation mark and the following word, or a closing quotation mark and the preceding word. When a double quotation mark or a single quotation mark immediately follows the other, proper spacing for legibility requires that a (non-breaking) space be inserted.

In English, opening quotation marks never appear at the end of a line of text, and closing quotation marks never appear at the beginning. As with most punctuation, these marks are wrapped around with the associated word.

Non-language related usage
Straight quotes (or italic straight quotes) are often used to approximate the prime and double prime (e.g., when when signifying inches and feet, or arcminutes and arcseconds). For instance, 5 feet and 6 inches is often written 5' 6", & Forty degrees, Xx proceedings, & Fifty seconds is written 40° Xx' L". When available, however, the prime should be used instead (e.g., 5′ 6″, and 40° 20′ 50″).

Straight single and double quotes are used in most programming languages to delimit strings or literal characters. In some langauges (e.g. pascal) only one type is allowed, in some (e.g. C and its derivatives) both are used with different meanings and in others (e.g. php and Python) both are used interchangeably. In many languages if it is desired to include inside a string the same quotes used to delimit the string the quotes are doubled. For example to represent the string eat 'hot' dogs in pascal you use 'eat hot dogs'.

Glyphs
A list of glyphs used as quotation marks and their Unicode (and HTML) values and names follows. The Unicode standard defines two general character categories, “Ps� (punctuation quote start) and “Pe� (punctuation quote end), for all quotation mark characters. (Warning: Some of these glyphs may not display properly in older browsers, which may substitute other sorts or a square.)

Typewriter quotation marks
"Two-handed" quotation marks were introduced on typewriters to reduce the number of keys on the keyboard, and were inherited by computer keyboards and character sets. However, modern word processors have started to convert text to use curved quotes (see below). Some computer systems designed in the past had proper opening and closing quotes, with a few machines even making a distinction between regular apostrophes (e.g. couldn’t) and apostrophes that show possession (e.g. Dave’s car). However, the standard ASCII character set, which has been used on a wide variety of computers since the 1960s, only made three quotation marks available: ", ', & a dubious backquote ` (also known as the backtick & the letterless grave accent). A Unicode standard includes typographic & the kind of international quotation marks.

Many systems, like the personal computers of the 1980s and early '90s, actually drew straight quotes like curved closing quotes on-screen and in printouts, so text would appear like this (approximately):

The grave accent (`) could then be used to supply single quote marks. This use resulted in fonts with an open quote glyph at the grave accent position. This gives a proper appearance at the cost of semantic correctness. Nothing similar was available for the double-quote, so many people resorted to using sets of two single quotes for punctuation, which would look like the following:

However, the appearance of these characters has varied greatly from font to font. On systems which provide straight quotes and grave accents the appearance is poor. Unicode specifies that ASCII single and double quotes should be vertical rather than angled, which means if such such tricks are used with a font that follows the rules the result will look rather messy (see next sample). On the other hand Unicode also provides the ability to do angled quotes properly.

Quotation marks in English
English curved quotes, also called “book quotes� or “curly quotes�, look like small figures six and nine with the counters filled. They are preferred in formal writing and printed typography. In e-mail and on Usenet they can only be used by using a MIME type with a character set outside of the ISO-8859 series such as a Unicode encoding or one of the Windows-125x series. Whilst not a problem for most modern mail clients this does increases the size of the message and makes the raw message text harder to follow and so some believe it is bad practice (in much the same way that some think that HTML e-mail is a bad thing). A few mail clients send curved quotes using the windows-1252 codes but mark the text as ISO-8859-1 causing problems for decoders that don't make the dubious assumption that C1 control codes in ISO-8859-1 text were meant to be windows-1252 printable characters.

Curved and straight quotes are also sometimes referred to as “smart quotes� and "dumb quotes" respectively; these names are in reference to the name of a function found in word processors like Microsoft Word that automatically converts straight quotes typed by the user into curved quotes. This function is necessary because keyboards lack separate quotation marks (due to the fact the ASCII character set didn't have distinct opening and closing quotation marks) A quote followed by a letter generally becomes an opening quote, whereas a quote with a letter or period preceding it and a space after it becomes a closing quote.

Variants of ‘ and “ are: Supporting curved quotes has been a problem in information technology, primarily because the widely-used ASCII character set did not include a representation for them (as discussed above).

Word processors have traditionally offered curved quotes to users, because in printed documents curved quotes are preferred to straight ones. Before Unicode was widely accepted and supported, this meant representing the curved quotes in whatever 8-bit encoding the software and underlying operating system were using â€” but the character sets for Windows and Macintosh used two different pairs of values for curved quotes, and ISO 8859-1 (typically the default character set for the Unices and Linux) had no curved quotes, making cross-platform compatibility a nightmare.

Compounding the problem is the “smart quotes� feature mentioned above, which some word processors (including Microsoft Word and OpenOffice.org) use by default. With this feature turned on, users may not have realised that the ASCII-compatible straight quotes they were typing on their keyboards ended up as something entirely different.

Unicode support has since become the norm for operating systems. Thus, in at least some cases, transferring content containing curved quotes (or any other non-ASCII characters) from a word processor to another application or platform has sometimes been less troublesome, provided all steps in the process (including the clipboard if applicable) are Unicode-aware. But there are many applications which still use the older character sets, or output data using them, and thus problems still occur.

There are other considerations for including curved quotes in the widely-used markup languages HTML, XML, and SGML. If the encoding of the document supports direct representation of the characters, they can be used, but doing so can result in difficulties if the document needs to be edited by someone who is using an editor that cannot support the encoding. For example, many simple text editors only handle a few encodings or assume that the encoding of any file opened is a platform default, so the quote characters may appear as "garbage". HTML inclused a set of entities for curved quotes ‘ (left single), ’ (right single), ‚ (low 9 single), “ (left double) ” (right double) &dbquo; (low 9 double). XML does not define these by default, but specifications based on it can do so, and XHTML does. In addition, while the HTML 4, XHTML and XML specifications allow specifying numeric character references in either hexadecimal or decimal, SGML and older versions of HTML (and many old implementations) only support decimal references. Thus, to represent curly quotes in XML and SGML, it is safest to use the decimal numeric character references. That is, to represent the double curly quotes use “ and ”, and to represent single curly quotes use ‘ and ’. In HTML, it is safest to use the named entity references (“, etc.), although decimal numeric character references are processable by most web browsers (Netscape 4 being a notable exception).

There has been some argument in recent years about the appropriateness of book quotes, since they are perceived by some as distracting. Editors who are against book quotes generally argue for ASCII-style straight quotes.

Quotation marks in Finnish and Swedish

In Finnish and Swedish, right quotes are used to mark both the beginning and the end of a quote. They are referred to as typographical quotes to distinguish them from straight quotes. This usage also translates to angular quotes (sometimes used in magazines), although recently, the notation »…« has won acceptance as well.

Quotation marks in German
Confusingly, what is the “left quote� in English is used as the right quote in Germany and Austria, and a different “low 9 quote� is used for the left instead:

This style of quoting is also used in Georgian, Estonian, Icelandic, Bulgarian (the single quotes are not used), Romanian and in Russian.

Quotation marks in Polish
In the Polish language, double angle quotes are used inside of standard Polish double quotes.

According to current PN-83/P-55366 standard from 1983, Setting rules from composing of Polish texts (Zasady składania tekstów w języku polskim) one can use either „ordinary Polish quotes� or «French quotes» (without space) for first level, and ‚single Polish quotes’ or «French quotes» for second level, which makes three styles of nested quotes: