32

In my personal experience, many native speakers of U.S. English are familiar with the term "umlaut" as referring to the double dots above a letter, though they are not generally aware of its actual effect on pronunciation. They are typically unfamiliar with the term "dieresis". Therefore, they would describe the dots over the "e" in "Zoë" as an umlaut, despite it being a dieresis and not an umlaut.

What is the effect of the dieresis and the umlaut, and how do they differ?

Here are some examples of the dieresis and umlaut:

  • Dieresis: naïve, noël, Zoë
  • Umlaut: doppelgänger, übermensch
  • Metal umlaut: Motörhead, Mötley Crüe
M. Justin
  • 973
  • The word dieresis contains (ie) a dieresis. – Nigel J Oct 12 '20 at 18:33
  • When did Haïti become Haiti? Unlike naive it does not seem to be pronounced as a dieresis any more. – Weather Vane Oct 12 '20 at 18:49
  • 6
    I'm extremely disappointed you didn't include "Spın̈al Tap" in your examples. :) – IMSoP Oct 13 '20 at 08:22
  • Related questions on ELL: Why does the i in naïve have two dots and this one on ELU. – Decapitated Soul Oct 13 '20 at 08:32
  • I'm putting this as a comment, because I don't have time to research an answer, but I'd say that both diaeresis and umlaut are phenomena of words, and the name of the actual accent mark is trema. – TRiG Oct 13 '20 at 10:36
  • 1
    @TRiG — From what I can see looking at various online sources, "dieresis" & "umlaut" can refer to the linguistic phenomena, as well as the diacritic mark used in its particular context. "Trema" refers to the diacritic mark used in either context. So either is accurate, so long as you don't refer to the dots in an umlaut as a "dieresis" or vice versa. – M. Justin Oct 13 '20 at 14:36
  • 3
    Also, just want to point out that the diacritic mark in question can also serve as neither umlaut nor diaeresis, most notably in Cyrillic letter "ё" ("yo"). – Alice Oct 13 '20 at 20:36
  • 1
    An umlaut is German, a dieresis is French. – Mark Oct 13 '20 at 23:20
  • 1
    Just to confuse matters further, in Swedish ä and ö are neither diereses nor umlauts, but are considered as separate letters of the alphabet (along with å). – Bob says reinstate Monica Oct 14 '20 at 12:13
  • "übermensch" is the most 1933-ish example for a German word I can think of. – Zsolt Szilagy Oct 14 '20 at 16:14
  • @ZsoltSzilagy There was a very limited selection of German loanwords in English that I could find that still had the umlaut, let alone ones that would be at least somewhat commonly known. – M. Justin Oct 14 '20 at 17:55
  • @M.Justin: Understandable. Maybe Götterdämmerung and Doppelgänger might be the only other known ones. – Zsolt Szilagy Oct 14 '20 at 20:42
  • 1
    The [tag:diaeresis] tag wiki needs love, btw. – Martin Schröder Oct 14 '20 at 21:17
  • 1
    @MartinSchröder I created an initial version of the tag usage guidance & wiki, though it could be made more complete: https://english.stackexchange.com/tags/diaeresis/info – M. Justin Oct 15 '20 at 18:19
  • @Mark or New Yorkish https://www.theatlantic.com/culture/archive/2012/04/we-resist-further-cooperation-cooperation/328832/ – llama Oct 15 '20 at 23:25

5 Answers5

28

Bryan Garner, Garner's Modern American Usage, second edition (2003) offers the following succinct discussion of the two punctuation marks (which are really one punctuation mark with two different names and functions) that the posted question asks about:

umlaut; diaeresis. These words denote the same mark consisting of two raised dots (¨) placed over a vowel, but they serve different phonetic functions. An umlaut (pronounced /oom-lowt/) indicates that the vowel has a modified sound especially in German, as in Männer (pronounced /men-ner/) A diaeresis (pronounced /dI-air-ə-sis/ and sometimes spelled dieresis) indicates that the second of two adjacent vowels is pronounced separately, as in naïve.

But the distinction is largely academic: even with modern word-processing capabilities, these marks are often omitted.

Allan Siegal & William Connolly, The New York Times Manual of Style and Usage, revised edition (1999) notes that the mark occasionally appears in imported proper names in which it serves a third purpose: to signal "pronunciation of a normally silent final consonant":

accent marks are used for French, Italian, Spanish, Portuguese and German words and names. ...

...

The umlaut modifies vowel sounds in German, (Götterdämmerung, Düsseldorf). Some news wires replace the umlaut with an e after the affected vowel. Normally undo that spelling, but check before altering a personal name; some individual Germans use the e form. In the Latin languages, the umlaut is known as a dieresis. It denotes separated pronunciation of two adjacent vowels (naïve, Citroën, Noël), or signals pronunciation of a normally silent final consonant (Saint-Saëns, Perrier-Jouët).

Sven Yargs
  • 163,267
  • 2
    This is implied by the second quote, but one thing you could add is if one wants to make an educated guess at whether one is looking at a dieresis or an umlaut, one should look if the diacritic is placed over a vowel that is preceded by another vowel, in which case it is more likely a dieresis. If the vowel in question is preceded by a consonant, then almost certainly an umlaut, since it couldn't serve the purpose of a dieresis in that context. – Todd Wilcox Oct 13 '20 at 07:30
  • @ToddWilcox Usage depends on the original language. For example, in Spanish (and I believe the Portuguese usage is similar) the dieresis is used in the sequences "güe" and "güi" (that is, a u preceded by the consonant g) to indicate that the u isn't silent as it normally would be. My educated guess would rather depend on whether there is a sequence of at least two vowels one of which has the two dots, regardless of whether it's the first one or the second one. In that case, it's likely a dieresis. If there's a single vowel with the two dots, it's an umlaut. – dgstranz Oct 13 '20 at 10:23
  • 16
    "But the distinction is largely academic: even with modern word-processing capabilities, these marks are often omitted." Ouch, this sentence hurts a German native speaker. You cannot write Gotter instead of Götter, this would be simply wrong. And replacing umlauts by appending an e, as in Goetter, is correct but really not in common use. If a modern word processor cannot deal with umlauts, then IMHO it is unusable to write German texts. – rexkogitans Oct 13 '20 at 11:24
  • 3
    For anyone else who was curious about the Saint-Saëns example, Wikipedia says this in a footnote: “The diaeresis on the e dates from a time when the e was not silent, but the diaeresis no longer affects the pronunciation of the name(s) because the e is silent…” – bdesham Oct 13 '20 at 15:23
  • NIt pick: "... pronunciation of two adjacent vowels..." – MPW Oct 13 '20 at 17:42
  • @MPW: Thanks for pointing out the error. I've corrected it. – Sven Yargs Oct 13 '20 at 17:49
  • 1
    to add a little to the 2nd quote, the explanation often given for the history of the umlaut is that it started as a scribe's shorthand for the "e". See here wiki article . "In blackletter handwriting, as used in German manuscripts of the later Middle Ages and also in many printed texts of the early modern period, the superscript e still had a form that would now be recognisable as an e, but in manuscript writing, umlauted vowels could be indicated by two dots since the late medieval period." – Angst Oct 13 '20 at 20:10
  • 7
    @rexkogitans, I suspect the text is referring to loanwords adopted into English, not German text within an English work. For example, doppelganger is typically written in English with no umlaut. – The Photon Oct 13 '20 at 21:12
  • 1
    Might be worth noting that there are also cases of the two raised dots above a letter that are technically neither an umlaut or a dieresis. Albanian for example use ë as a distinct letter with different phonetic behavior from an umlaut or dieresis. It gets interesting too because some languages use what is essentially an umlaut (even in terms of orthographic evolution) but still treat the modified characters as wholly distinct letters (see ä and ö in Swedish for example). – Austin Hemmelgarn Oct 14 '20 at 01:20
  • 1
    @rexkogitans “even with modern word-processing capabilities,” i.e. even though modern word processors make it easy to do it right. No one’s saying the capability is missing, just that people don’t bother even though they could. And that statement only makes sense in English, where most writers don’t really know what that mark means or when it should be used, or how to type one, and most of their audience wouldn’t know either. – KRyan Oct 14 '20 at 13:51
  • "indicates that the vowel has a modified sound especially in German," If there were a comma between "sound" and "especially", this would mean that this statement is especially true in German, i.e. "(this indicates that the vowel has a modified sound) especially in German. Without the comma, it means that the modification of the sound is especially pronounced in German. i.e. "this means that the vowel (has a modified sound especially in German)". Odd for a book discussing linguistics to not give care to use punctuation precisely. – Acccumulation Oct 14 '20 at 20:24
  • 1
    @rexkogitans also as a Dutch speaker, but because of the dieresis. It's necessary because Dutch has multiple two-letter vowel combinations that form a single sound. In cases in which these should be separate sounds instead, the dieresis is needed. To be fair though, I assume the text is specifically talking about usage in English. – tjalling Oct 15 '20 at 10:05
  • Today I learned that I've been pronouncing "dieresis" wrong my whole life. – Hellion Oct 15 '20 at 19:17
11

According to the Wikipedia article for dieresis, the dieresis is used to indicate that two vowels are to be pronounced separately, rather than combined as a single combined sound. On the other hand, an umlaut is used to indicate a sound shift of the vowel, pronouncing it differently than it would without the umlaut.

The diaeresis and the umlaut are diacritics marking two distinct phonological phenomena. The diaeresis represents the phenomenon also known as diaeresis or hiatus in which a vowel letter is pronounced separately from an adjacent vowel and not as part of a digraph or diphthong. The umlaut, in contrast, indicates a sound shift. These two diacritics originated separately; the diaeresis is considerably older.

The dieresis is native to English, though it is generally considered to be archaic. On the other hand, the umlaut is a diacritic found in other languages (in particular, Germanic ones).

The grave accent and the diaeresis are the only diacritics native to Modern English (apart from diacritics used in loanwords, such as the acute accent, the cedilla, or the tilde). The use of both, however, is considered to be largely archaic.

Germanic umlaut is a specific historical phenomenon of vowel-fronting in German and other Germanic languages. In German it causes back vowels /a/, /o/ and /u/ to shift forward in the mouth to /ɛ/, /ø/ and /y/, respectively. In modern German orthography, the affected graphemes ⟨a⟩, ⟨o⟩ and ⟨u⟩ are written as ⟨ä⟩, ⟨ö⟩ and ⟨ü⟩, i.e. they are written with the diacritical mark "umlaut", which looks identical to the diaeresis mark.

Some languages have borrowed some of the forms of the German letters Ä, Ö, or Ü, including Azerbaijani, Estonian, Finnish, Hungarian, Karelian, some of the Sami languages, Slovak, Swedish, and Turkish. This indicates sounds similar to the corresponding umlauted letters in German.

[Diaeresis - Wikipedia]

Decapitated Soul
  • 16,598
M. Justin
  • 973
  • 2
    That's basically correct. What's the problem? – John Lawler Oct 12 '20 at 18:27
  • 9
    @JohnLawler Are you referring to the fact that I asked & answered my own question? Since that's an encouraged Stack Exchange activity: https://meta.stackexchange.com/a/17467/308327. In short, I didn't see anything directly addressing this question on this site, it's something I've done some research into, something I would like to see here, and I felt that adding it as a Q&A-style self-answer was a positive contribution to the site. – M. Justin Oct 12 '20 at 18:30
  • 11
    I don't think any discussion of the dieresis in English would be complete without reference to The New Yorker, which is a unique force keeping its use alive in English. "She said that once, in the elevator, he told her he was on the verge of changing that style and would be sending out a memo soon. And then he died. This was in 1978. No one has had the nerve to raise the subject since." https://www.newyorker.com/culture/culture-desk/the-curse-of-the-diaeresis – nohat Oct 12 '20 at 18:42
  • @nohat One of the paragraphs in that article answers this question nicely, as well. "The difference is that an umlaut is a German thing that alters the pronunciation of a vowel, and often changes the meaning of a word. [...] A diaeresis goes over the second vowel and indicates that it forms a separate syllable." – M. Justin Oct 12 '20 at 18:55
  • 1
    Discussions about foreign languages belong not on ELU but often on Linguistics.SE. Just like maths-specific terminology belongs on Mathematics, questions about interpretation of passages in books belong on Literature, questions about politeness conventions belong on Interpersonal Skills (though of course they all use language). And answering one's own question presupposes that reasonable decent research has been done and accompanies the question. – Edwin Ashworth Oct 12 '20 at 19:26
  • @EdwinAshworth — I'll give your point about needing research in the question some further thought in the morning to see how I can best address it in the question. I'm running into the quandary discussed here: https://meta.stackexchange.com/questions/256644/is-a-short-description-of-a-question-ok-if-self-answering. – M. Justin Oct 13 '20 at 05:56
  • 1
    @EdwinAshworth — I've attempted to address your foreign language point by changing the umlaut examples to English words. Granted, they're loanwords from German, but they do show up in English dictionaries with the umlaut spelling. – M. Justin Oct 13 '20 at 05:56
5

The simplest summary is: Umlaut is a German phenomenon, while dieresis is used mostly in French (see examples above).

At a closer look they look similar but convey totally different meanings: Dieresis indicates that two vowels are to be pronounced separately, whereas umlaute originally symbolized ae (ä), oe (ö), or ue (ü). And these vowel combinations actually are vowels of their own in the German language (see above for pronunciation guidelines).

Cheers, Günther

P.S. Actually the dieresis is called a "Trema" in German and most Germans don't know about it before starting to learn French.

Decapitated Soul
  • 16,598
Günther
  • 51
3

Other answers have addressed the question pretty well in a general context, so I thought I'd give an answer focussing more on the linguistic history in particular

In this context, the diacritic is not referred to as an umlaut, but rather a diaresis (if indicating that the letter should be pronounced separately, rather than as part of a digraph) or a trema, whilst umlaut refers to a set of partially-grammaticalised partially-lexical vowel changes in the Germanic languages, some of which are shown with a trema in German orthography

There are three types of umlaut, i-umlaut, a-umlaut, and u-umlaut. The first two occurred in all surviving Germanic languages, whilst the latter only occurred in the North Germanic (Scandinavian) languages

I-umlaut affected vowels followed by an i (or a j, which in most Germanic languages has the sound of an English y). Back vowels (a, o, & u) affected by it move forward in the mouth (creating new vowels ä, ö, ü, in English these later became e, e, & i), whilst front vowels (e & i) move up in the mouth if possible (merging into i)

A-umlaut affected vowels followed by an a, and only affects the high vowels (i & u) which were lowered to merge into the mid vowels (e & o)

U-umlaut affected vowels followed by a u (or a w, which later became a v in most Germanic languages). Unrounded vowels (i, e, a) affected by u-umlaut became rounded (ü, ö, å)

In all cases, long vowels behaved pretty much the same as their short equivalents

If this was just easily predictable though, it wouldn't have ever needed to be written explicitly, you'd just know to pronounce an a followed by an i further forwards in your mouth. The reason it came to be written explicitly is because these vowel changes persisted after the vowels that triggered them were lost

As an example, the English word man has the plural men, reflecting an earlier stage where it was mann and manniz with the i triggering i-umlaut. The -iz ending was lost, but the umlaut from a > ä (which later developed into e) was retained

German reduced unstressed i to e in most cases, so most instances of i-umlaut that still had an obvious trigger seemed to have it triggered by an e, and so German scribes started writing an e above a vowel to show that it had been affected by i-umlaut and this eventually evolved into the modern German trema we see today

The diaresis as used in French (to show that the second vowel letter in a sequence should be read separately, rather than as a digraph) is a direct continuation (via Latin) of the Greek trema used in the same way developed during the Hellenistic era (along with most of the rest of the Greek diacritics)

Tristan
  • 860
  • There is also Luxembourgish, which has ë as a stressed schwa, most notably on coins saying Lëtzebuerg, the name of the country – Henry Oct 14 '20 at 14:00
  • @Henry yeah, since its origins (which largely follow the process I described), it then often came to mean "a vowel with all the features of this vowel, except opposite frontness", as a stressed schwa is back of an /e/ but otherwise has the same roundedness and height, and there's no obvious way to otherwise distinguish it it's a reasonable extension. This also gives rise to the ï sometimes seen in transcriptions of Central Asian languages to describe a high back unrounded vowel (i.e. an unrounded /u/) – Tristan Oct 14 '20 at 14:45
  • "diaresis or trema is the term used to refer to the diacritic in almost all instances" — Isn't it only called a "diaeresis" in the context where the letter is sounded separately, and not in other instances (such as the umlaut case)? – M. Justin Oct 15 '20 at 21:10
  • true, I mostly meant that the diacritic isn't referred to as an umlaut in linguistics, but this was not clear. I will edit to clarify – Tristan Oct 16 '20 at 08:49
2

I suggest looking at Babbel.com A Tale Of Two Dots — The History Of The Umlaut And The Diaeresis

The meaning of the word umlaut is revealing: it means “around sound” in German. It was named by the linguist Jacob Grimm, one of the Grimm Brothers. His “around sound” describes a process of sound-change where a vowel’s sound is influenced by another vowel that follows it in the word. The plural of Hant (“hand”) in High German was Hanti, but the i ending influenced the pronunciation of the previous vowel a. So, Hanti became Henti. Eventually, the final i lost its timbre. So, in contemporary German we have Hand and the plural Hände.

The representation of the umlaut in Middle High German was sometimes denoted by adding an e to the affected vowel, either after in regular size or above in a smaller size. The former is still applicable in a name like Goethe — which is never spelled Göthe.

In French, the diaeresis (from the Greek diaíresis [διαίρεσις], meaning “division” or “distinction”) looks exactly the same as the umlaut but has a radically different purpose. Whereas the umlaut represents a sound shift, the diaeresis indicates a specific vowel letter that is not pronounced as part of a digraph or diphthong. In French words such as Noël (Christmas), the two dots are there to remind you not to fuse the two vowels into one sound, but to pronounce the O and the E separately.

In Spanish, … vergüenza (“shame”) or ambigüedad (“ambiguity”), which is usually silent in words such as guitarra (“guitar”) or guionista (“scriptwriter”).

English also has its occasional diaeresis, mostly used in surnames or given names, such as Brontë or Zoë. In some words, such as naïve, it is optional.

The whole article is worth reading.

Greybeard
  • 41,737