Shadow lists the contents of an interesting special issue of Written Language & Literacy.
One of the pieces, The apostrophe: A neglected and misunderstood reading aid, has this to say:
Almost all apostrophes commonly explained as indicating omission can also be explained as marking morpheme boundaries. No apostrophes that do not mark boundaries do occur at all in the earliest texts and in modern formal texts.
Consequently, the apostrophe ought to be defined as having as its one dominant function the indication of morpheme boundaries where for certain reasons this seems necessary….
Furthermore, the apostrophe, which was borrowed into the Latin alphabet from Greek, seems to have indicated a boundary rather than an omission from the start.
This is also how apostrophes are used in Pinyin.
further reading:
- notes for the apostrophe article
- Web site of the author, Daniel Buncic
A glance shows this thesis is written about German and apparently French. It should be pointed out, however, that the opposite is true in English. From what I know of German, I can cite an example:
“Das gibt’s doch nit!” “That can’t be!” In this sentence, “gibt’s” is a contraction of “gibt es”, an example of two morphemes divided by an apostrophe. Note that the “nit”, however, even though it is an informal abbreviation of “nicht”, receives no apostrophe, as would for example “’til” in English, an abbreviation of “until” (a word coincidentally coequal with “till”, e.g. “Wake not ’til the morn.”).
Contrast this with English “shouldn’t”, in which the apostrophe is internal to the “not”, which is itself simply scrunched together with the “should”, not a division of morphemes at all.
In Polish and Hungarian (and I think some other languages in the region) the main purpose of apostrophes is to indicate a mute e in a foreign borrowing, as in:
software’u (Polish, the genitive case, without the apostrophe the word would have to be respelled (softwaru? soft?eru?) My own name gets this to, Mike’a (also genitive case) though some people respell it anyway as Maika)
Iv’e also seen it with mens names ending in y to indicate that it’s not a Polish y (a full vowel, sort of like i in sit) but has a y glide (normally written with j) after it.
Kennedy’ego (pronounced as if written Kennedyjego or Kennedjego)
A Hungarian example is Alice’szel (with Alice).