O with Umlaut (ö): Oe in German, O in Finnish, Ö in Swedish

Is ö a letter of its own or an o with two dots?

Both, and the language decides which. In German, the o with umlaut is an o with a changed sound, and it is filed with the o words. Swedish, Finnish, Estonian, Icelandic, Turkish and Hungarian list ö as a separate letter of the alphabet. Their sorting rules give it a place of its own.

The Unicode CLDR collation rules show where each language files the o umlaut. Each row names the rule that places ö in a sorted list, and the last column gives words and places from AccentPad's facts.

Language	What ö is there	Where a sorted list puts ö	Seen in
German	an o with an umlaut	among the o words: Kohl, Köhler, Kolb	schön, können; Köln, Neukölln
Swedish	the last letter of the alphabet	after z, behind å and ä	Malmö, Linköping, Örebro
Finnish	a letter, named öö	after z, behind å and ä
Estonian	a letter	after w, õ and ä, before ü
Icelandic	the last letter of the alphabet	after z, þ and æ
Turkish	a letter	straight after o, before p	Arnavutköy, Gölbaşı
Hungarian	a short vowel letter	after o, sharing its place with ő

The names of the letters come from Wiktionary's entry for ö (opens in a new tab). The positions come from the CLDR charts for Swedish (opens in a new tab), Finnish (opens in a new tab), Estonian (opens in a new tab), Icelandic (opens in a new tab) and Turkish (opens in a new tab). Where a language makes ö a letter, the dots are part of the spelling. That is what decides the stand-in question below.

When can you write oe instead of ö?

Only in German. There, oe is the accepted stand-in for the o with umlaut when ö cannot be typed. German sorting rules even file ö as oe. Finnish uses a plain o. In Swedish, Turkish and Hungarian no stand-in is safe: drop the dots and the reader sees another word.

Writing in	If ö cannot be typed	What goes wrong otherwise
German	write oe: schoen, Koeln	plain o makes another word: schön (beautiful) becomes schon (already), lösen (to solve) becomes losen (to draw lots)
Finnish	write o, never oe	koe is a real word, test or experiment, so oe can be read as two vowels
Swedish	keep ö	för (too; because) becomes for (went), kör (choir) becomes kor (cows)
Turkish	keep ö	öl (die!) becomes ol (be!), kör (blind) becomes kor (ember)
Hungarian	keep ö	o is another letter, and ő another vowel: öt (five) against őt (him, her)

The German row rests on the CLDR German collation rules (opens in a new tab). Their phonebook order holds the line &OE<<ö<<<Ö: ö sorts as oe and differs from it only to break ties. The word meanings come from Wiktionary's entries for schön (opens in a new tab), schon (opens in a new tab), lösen (opens in a new tab) and losen (opens in a new tab).

Finnish is the trap. Wiktionary's usage note for Finnish ö says to write o under technical limits, not oe as in German. Traficom, the Finnish domain regulator, applies the same rule to .fi names (opens in a new tab). It treats both åäö.fi and aao.fi as identical to a protected name ÅÄÖ. And koe (opens in a new tab) shows why oe misleads, because Finnish already spells the test word with a real o followed by a real e.

The Swedish pairs come from Wiktionary's entries for för (opens in a new tab), for (opens in a new tab), kör (opens in a new tab) and kor (opens in a new tab). The Turkish pair öl and ol is the imperative of to die (opens in a new tab) against the imperative of to be (opens in a new tab).

Will a search for oe or o find ö?

Software follows the same split, so the answer depends on the rules it runs. With accents ignored, German rules match König with konig. Only the German phonebook rules match it with koenig. Swedish, Finnish, Turkish and Hungarian rules match neither, because there ö is a different letter.

Rules in use	konig finds König	koenig finds König	Öberg sorts
German, standard	yes	no	as Oberg, before Oberhof
German, phonebook	no	yes	as Oeberg, between Oder and Oeding
Swedish, Finnish, Icelandic	no	no	after Zander, at the very end
Estonian	no	no	after Wolf, before Xaver
Turkish, Hungarian	no	no	after Ost, before Pohl

Every cell was computed with the Intl.Collator in Node.js, which reads the CLDR data, with the comparison set to ignore accents and case. We re-ran the whole table on 2026-10-13 in Node.js v26.0.0, which runs ICU 78.3 on CLDR 48 data, and every cell held. As a check, the same German comparison matches cafe with Café.

A customer database is the practical case. If it stores König and a user types Koenig, only a German phonebook collation connects the two. For a Swedish surname such as Östlund, no rule set but the German phonebook one matches Oestlund, and none outside German matches Ostlund.

One more trap sits below the rules. An ö can be stored as one code point, U+00F6, or as a plain o followed by the combining mark U+0308, and the two look the same on screen.

In the same Node.js run, a plain substring search for König missed it in text that held the two-part form, and found it once the text was converted to NFC, the composed form that Unicode's normalization report (opens in a new tab) defines. Normalize names to NFC before you store or compare them.

Which names spell the ö sound without the dots?

Some family names keep an older spelling of the sound, and turning it into ö would misspell the name. The best known is Goethe. Wiktionary's entry for Goethe (opens in a new tab) gives the pronunciation /ˈɡøː.tə/ and lists Göthe only as an alternative form. The oe in Goethe is the name's own spelling, not a typing workaround.

Hungarian has the same pattern. The Hungarian spelling rules (opens in a new tab) list old letters that survive in family names. Two of them stand for ö: eö in Eötvös and ew in Thewrewk, which is said like török. So a search and replace that turns every oe into ö, or every ö into oe, damages real names in both directions. Check names one by one.

Is Hungarian ő the same letter as ö?

No. Hungarian ő is a separate letter from the o with umlaut. It has a different mark, the double acute, and it stands for the long version of the ö sound. The Hungarian spelling rules (opens in a new tab) list o, ó, ö and ő as separate letters of the alphabet. They also warn that vowel length can change a word.

	ö	ő
Unicode name	LATIN SMALL LETTER O WITH DIAERESIS	LATIN SMALL LETTER O WITH DOUBLE ACUTE
Mark	two dots	two slanted strokes, the double acute
Hungarian vowel	short	long
Word pairs	öt (five), tör (to break), elöl (answers where?)	őt (him, her), tőr (dagger), elől (answers from where?)
In a sorted list	shares a place with ő	sorts with ö; the short word comes first only in a tie
Windows Alt code	Alt+0246	none
Places with it	Gödöllő (Hungary) has both	Győr (Hungary)

The meanings and the sounds come from Wiktionary's entries for öt (opens in a new tab), őt (opens in a new tab), tör (opens in a new tab) and tőr (opens in a new tab). Wiktionary writes öt as [ˈøt] and őt as [ˈøːt], with the length mark. The elöl and elől pair, and the rule that the two letters count as equal when sorting, come from the spelling rules. The CLDR Hungarian rules (opens in a new tab) encode it as &O<ö<<<Ö<<ő<<<Ő: ö is a new letter after o, and ő is a variant of ö.

The Alt code gap has a plain cause. Windows Alt codes come from the Windows-1252 code page (opens in a new tab), which holds ö at byte F6, 246 in decimal, and so gives Alt+0246. It has no ő. Unicode's mapping file for Windows-1250 (opens in a new tab) does hold ő, U+0151, at byte F5, but AccentPad computes its Alt codes from Windows-1252 only. Typing ö where Hungarian needs ő is not a harmless shortcut, because the pairs above turn into other words.

Are the two dots on ö an umlaut or a diaeresis?

On a German word, the o with dots is an umlaut; in some other languages the same dots are a diaeresis. Unicode does not tell them apart. It has one character, U+00F6, and calls it o with diaeresis whatever job the dots do.

Duden defines the Umlaut (opens in a new tab) as the change of a vowel such as o to ö, and as the letter that stands for the changed vowel. Its entry for the Trema (opens in a new tab), the diaeresis, describes another job. There the dots sit over one of two vowels side by side, to show that they are said separately. The Unicode Standard (opens in a new tab) says the combining mark U+0308 covers diaeresis, trema and umlaut alike, and does not encode the uses separately.

Dutch shows the second job on an o. In coördinatie (opens in a new tab), Wiktionary splits the word as co‧or‧di‧na‧tie and writes the sound as /ˌkoː.ɔr/. The dots only say that the second o starts a new syllable. It is still an o sound, not the German ö of schön, so oe would add a letter that is not there.