Your question is rather odd: punctuation following an ellipsis does not constitute a ligature. For more information about ellipses in general, try this...

http://www.chicagomanualofstyle.org/search.epl?q=ellipsis&site=cmslive

I second what Nick said. Your examples should just have been written as “What did she…?” and “I said, get your filthy…!”.

Given that a ligature is simply defined as being a combination of two or more characters contracted (or of course, legated) into a single character, it's not that odd a question. However, more importantly, note that the CMoS is concerned with proper North American practices, which has very little to do with the historical use of a character throughout the ages.

Several British people I've asked about this noted that they would use "..?" rather than a full ellipsis and then a question mark, so a natural conclusion is that that practice has to have come from somewhere, and this makes it an interesting topic.

Given that the ellipsis has been around for many centuries, printers and authors throughout the ages must have at some point or another decided -even if merely as experiment to save ink- to legate the ellipsis with the dot in the question and exclamation mark.

I would like to know how (not) widespread that became and have no idea where to start looking, so hopefully people on this forum have some ideas on where to start.

In Russian, it should be something like “?..”, “!..”. So I would also love to know how to typeset it properly.

As it doesn't need to be a ligature it would be pointless making it into one.

The interrobang is a ligature replacement for ?!, and probably equally unused, but people don't seem to have much trouble calling that a ligature or accepting its existence in unicode at least. Just because it's not in use now, doesn't mean it hasn't been in use in the past, and I am interested in finding out to what degree it was used. Making a judgement call on whether there is 'a point' to some ligature is a bit silly. There's no point to the ffi ligature other than the aesthetic "it looks good" and the practical "it saves lead and ink". The same would go for an ellipsis paired with interpunction that has a dot at its base. It might look better, or it might save ink, and the moment either one of those is considered valid, there's your point.

Just to demonstrate that a point is trivially attributed: let's say someone is doing a poster design for a play with the title "Did you hear what she... ?" and decides that for stylistic purposes it would be quite nice to make use of the vertical space above the ellipsis by contracting the ellipsis and questionmark and reshaping the questionmark curve so that it overlaps most of the space above the ellipsis. That is now a ligature. This is how they come about.

None of the historical ligatures come from committees deciding to combine letters, they were invented by typesetters and writers (and in some cases, those writers influenced other writers, who in turn ended up influencing the entire written language). If one can think of a ligature that looks good, voila, one may be born. Just because you could approximate it doesn't mean there will never be a point to a "real" one; that's up to the minds of the people who actually use them.

But, we all digress, as that wasn't the point of this thread - it was primarily about getting pointers on where to look in order to find some kind of history of ellipsis ligatures, or at least a discussion of the use of "..?" and "..!" (or as texnic points out for Russian, "?.." and "!..")

So please, no more replies in the form of "it's a silly ligature" or "why would you use it", that is irrelevant. Does anyone know of a documented history of this stylistic choice?

Given that a ligature is simply defined as being a combination of two or more characters contracted (or of course, legated) into a single character...

No, a ligature is a combination of two or more characters represented by a single glyph.

An ellipsis mark consists of three dots, either by themselves... or followed by a fourth dot or other punctuation if they occur at the end of a sentence.... What you refer to as a 'stylistic choice' looks to me like a grammatical error.

Strunk and White weigh in...

The ellipsis itself is three periods (always); it can appear next to other punctuation, including an end-of-sentence period (resulting in four periods). Use four only when the words on either side of the ellipsis make full sentences. You should never use fewer than three nor more than four periods, with only a single exception: when entire lines of poetry are omitted in a block quotation, it's a common practice to replace them with a full line of spaced periods [emphasis added].

There is a Unicode "double dot" character (uni2025, twodotenleader), but I suspect this is supposed to be either the kind of dot leader that was used in price tables, or a "ditto" mark.

If it is an issue of spacing, a kerning pair between ellipsis and question mark would do the job. Why would you ever need it joined? what purpose is served?

The Unicode U+2024 and U+2025 are characters for filling spaces and were included due to compatibility issues.

I never saw any usage of two dots + exclamation/interrogation mark instead of ellipsis + mark. It may be included as a font stylistic option, but I personally dislike this –it will seem an error for almost all readers.

'There is a Unicode "double dot" character (uni2025, twodotenleader)'

It's described as a twodotleader, but the final edition of Hart's Rules says that two full points should be used when someone's initial is not known (for instance, in a bibliography).

The interesting question is how to design the ellipsis glyph. Should the dots be the same size as the period? No less an authority than Matthew Carter sometimes makes them the same, sometimes different. What's the ideal? How should they be spaced, and conceivably kerned, in relation to the period? If you are doing a four-dot ellipse (i.e. an end-of-sentence ellipse) then is it desirable that the dots of the ellipse be different from the dots of the period?

Go ahead, Bill, connect the dots for us ;-)

Most of the time, I make the ellipsis dots the same size as the full stop (period). In display types, I might make them slightly smaller, and also more tightly spaced. In text type, I prefer quite loose spacing of ellipsis dots, since if they are too tight they look like a grey line. I set the left sidebearing to be the same as that of the period, and I set the right sidebearing so that a period after the ellipsis is the same distance from the third dot as the ellipsis dots are from each other.

> How should they be spaced, and conceivably kerned, in relation to the period?

See also: http://typophile.com/node/42278

Thanks Kent, both threads are interesting. I only wish I had documented my own decisions over the years, because I've completely forgotten what I did and why. I sure hope the new Fontlab will have a note field for each glyph (if it doesn't already), as well as multiple background layers . . . .

Bill:I sure hope the new Fontlab will have a note field for each glyph (if it doesn't already)...

It does. Glyph notes are particularly useful when collaborating.