The phonology of Standard German is the standard pronunciation or accent of the German language. It deals with current phonology and phonetics as well as with historical developments thereof as well as the geographical variants and the influence of German dialects. While the spelling of German is officially standardised by an international organisation (the Council for German Orthography) the pronunciation has no official standard and relies on a de facto standard documented in reference works such as Deutsches Aussprachewörterbuch (German Pronunciation Dictionary) by Eva-Maria Krech et al., Duden 6 Das Aussprachewörterbuch (Duden volume 6, The Pronunciation Dictionary) by Max Mangold and the training materials of radio and television stations such as Westdeutscher Rundfunk, Deutschlandfunk, or Schweizer Radio und Fernsehen. This standardised pronunciation was invented, rather than coming from any particular German-speaking city. But the pronunciation that Germans usually consider to be closest to the standard is that of Hanover. Standard German is sometimes referred to as Bühnendeutsch (stage German), but the latter has its own definition and is slightly different.

Vowels

Monophthongs

Some scholars treat as an unstressed allophone of. Likewise, some scholars treat as an allophone of the sequence or as a vocalic realization of syllabic /r̩/. The phonemic status of is also debated – see below.

Phonemic status of

The existence of a phoneme in German is disputed. The distinction between the long lax and the long tense does not exist in some varieties of Standard German, and many authors treat the phoneme as peripheral and regard a distinction between it and as a spelling pronunciation. Most commonly, they are merged before an intervocalic, so that potential minimal pairs such as Ähre 'ear of grain' and Ehre 'honor' or Bären 'bears' and Beeren 'berries' are rendered homophonous, as and. Some authors claim that no distinction between and is possible in this position unless in hypercorrect pronunciation, in which Ehre and Beeren may be pronounced and, with a tense. Other authors claim that there is regional variation, a distinction occurring especially in southern variaties of Standard German. In contexts other than before intervocalic, the contrast between and is more stable, so that bäten 'bid, conjunctive', Dänen 'Danes' and Sägen ('saws, n.') may be differentiated from beten 'to pray', dehnen 'to stretch' and Segen 'blessing'. Even here the vowels can merge, but to a tense :. Scholars who question the existence of a phoneme do so for the following reasons:

Diphthongs

Phonemic

Phonetic

The following usually are not counted among the German diphthongs as German speakers often feel they are distinct marks of "foreign words" (Fremdwörter). These appear only in loanwords: In the varieties where speakers vocalize to in the syllable coda, a diphthong ending in may be formed with every stressable vowel:

Consonants

With around 22 to 26 phonemes, the German consonant system has an average number of consonants in comparison with other languages. One of the more noteworthy ones is the unusual affricate.

Ich-Laut and ach-Laut

Ich-Laut is the voiceless palatal fricative (which is found in the word ich 'I'), and ach-Laut is the voiceless velar fricative (which is found in the word ach the interjection 'oh', 'alas'). Laut is the German word for 'sound, phone'. In German, these two sounds are allophones occurring in complementary distribution. The allophone occurs after back vowels and (for instance in Buch 'book'), the allophone after front vowels (for instance in mich 'me/myself') and consonants (for instance in Furcht 'fear', manchmal 'sometimes'). The allophone also appears after vocalized ⟨r⟩ in superregional variants, e.g. in Furcht 'fear'. In southeastern regiolects, the ach-Laut is commonly used here, yielding. In loanwords, the pronunciation of potential fricatives in onsets of stressed syllables varies: in the Northern varieties of standard German, it is, while in Southern varieties, it is , and in Western varieties, it is (for instance in China: vs. vs. ). The diminutive suffix -chen is always pronounced with an ich-Laut. Usually, this ending triggers umlaut (compare for instance Hund 'dog' to Hündchen 'little dog'), so theoretically, it could only occur after front vowels. However, in some comparatively recent coinings, there is no longer an umlaut, for instance in the word Frauchen (a diminutive of Frau 'woman'), so that a back vowel is followed by a, even though normally it would be followed by a , as in rauchen ('to smoke'). This exception to the allophonic distribution may be an effect of the morphemic boundary or an example of phonemicization, where erstwhile allophones undergo a split into separate phonemes. The allophonic distribution of after front vowels and after other vowels is also found in other languages, such as Scots, e.g. licht 'light', dochter 'daughter', and the same distribution is reconstructed for Middle English. However, it is by no means inevitable: Dutch, Yiddish, and many Southern German dialects retain (which can be realized as instead) in all positions. It is thus reasonable to assume that Old High German ih, the ancestor of modern ich, was pronounced with rather than. While it is impossible to know for certain whether Old English words such as niht (modern night) were pronounced with or, is likely (see Old English phonology). Despite the phonetic history, the complementary distribution of and in modern Standard German is better described as backing of after a back vowel, rather than fronting of after a front vowel, because is used in onsets (Chemie 'chemistry') and after consonants (Molch 'newt'), and is thus the underlying form of the phoneme. According to Kohler, the German ach-Laut is further differentiated into two allophones, and : occurs after (for instance in Buch 'book') and after (for instance in Bach 'brook'), while either or may occur after, with predominating. In Western varieties, there is a strong tendency to realize as unrounded or, and the phoneme may be confused or merged with altogether, secondarily leading to hypercorrection effects where is replaced with , for instance in Fisch , which may be realized as. Within German dialects, a large variation exists as to the environments which trigger or prevent one realization or the other.

Fortis–lenis pairs

Various German consonants occur in pairs at the same place of articulation and in the same manner of articulation, namely the pairs, , , ,. These pairs are often called fortis–lenis pairs, since describing them as voiced–voiceless pairs is inadequate. With certain qualifications,, and are also considered fortis–lenis pairs. Fortis-lenis distinction for is unimportant. The fortis stops are aspirated in many varieties. The aspiration is strongest in the onset of a stressed syllable (such as Taler 'thaler'), weaker in the onset of an unstressed syllable (such as Vater 'father'), and weakest in the syllable coda (such as in Saat 'seed'). All fortis consonants, i.e. are fully voiceless. The lenis consonants range from being weakly voiced to almost voiceless after voiceless consonants: Kasbah ('kasbah'), abdanken ('to resign'), rotgelb ('red-yellow'), Abwurf ('dropping'), Absicht ('intention'), Holzjalousie ('wooden jalousie'), wegjagen ('to chase away'), tropfen ('to drop'), Obstjuice ('fruit juice'). states that they are "to a large extent voiced" in all other environments, but some studies have found the stops to be voiceless word/utterance-initially in most dialects (while still contrasting with due to the aspiration of the latter). are voiceless in most southern varieties of German. For clarity, they are often transcribed as. The nature of the phonetic difference between the voiceless lenis consonants and the similarly voiceless fortis consonants is controversial. It is generally described as a difference in articulatory force, and occasionally as a difference in articulatory length; for the most part, it is assumed that one of these characteristics implies the other. In various central and southern varieties, the opposition between fortis and lenis is neutralized in the syllable onset; sometimes just in the onset of stressed syllables, sometimes in all cases. The pair is not considered a fortis–lenis pair, but a simple voiceless–voiced pair, as remains voiced in all varieties, including the Southern varieties that devoice the lenes (with however some exceptions). Generally, the southern is realized as the voiced approximant. However, there are southern varieties which differentiate between a fortis (such as in sträflich 'culpable' from Middle High German stræflich) and a lenis (, such as in höflich 'polite' from Middle High German hovelîch); this is analogous to the opposition of fortis and lenis.

Coda devoicing

In varieties from Northern Germany, lenis stops in the syllable coda are realized as fortis stops. This does not happen in varieties from Southern Germany, Austria or Switzerland. Since the lenis stops are unvoiced or at most variably voiced (as stated above), this cannot be called devoicing in the strict sense of the word because it does not involve the loss of phonetic voice. More accurately, it can be called coda fortition or a neutralization of fortis and lenis sounds in the coda. Fricatives are truly and contrastively voiced in Northern Germany. Therefore, the fricatives undergo coda devoicing in the strict sense of the word. It is disputed whether coda devoicing is due to a constraint which specifically operates on syllable codas or whether it arises from constraints which "protect voicing in privileged positions".

Stress

In German words there is always one syllable carrying main stress, with all other syllables either being unstressed or carrying a secondary stress. The position of the main stress syllable has been a matter of debate. Traditionally, word stress is seen as falling onto the first stem syllable. In recent analyses, there is agreement that main stress is placed onto one of the last three (stressable) syllables. Within this three-syllable window, word stress is put regularly onto the second-to-last syllable, the penultimate syllable. However, syllable quantity may modify this pattern: a heavy final or prefinal syllable, i.e., one with a long vowel or with one or more consonants in the syllable coda, will usually attract main stress. A set of illustrative examples also stems from Japanese loan words, as these cannot be borrowed with their stress patterns (Japanese has a system of pitch accents, completely different from word stress in Germanic languages): A list of Japanese words in German reveals that none of the words with four syllables has initial stress, confirming the three-syllable-window analysis. Secondary stresses precede the main stress if at least two syllables are present, as in ̩Bib-li- ̩o-the-'ka-rin. Suffixes, if containing a stressable vowel, are either stressed (-ei, ion, -al, etc.) or unstressed (-ung, -heit, -isch, etc.) In addition, German uses different stresses for separable prefixes and inseparable prefixes in verbs and words derived from such verbs:

Acquisition

General

Like all infants, German infants go through a babbling stage in the early phases of phonological acquisition, during which they produce the sounds they will later use in their first words. Phoneme inventories begin with stops, nasals, and vowels; (contrasting) short vowels and liquids appear next, followed by fricatives and affricates, and finally all other consonants and consonant clusters. Children begin to produce protowords near the end of their first year. These words do not approximate adult forms, yet have a specific and consistent meaning. Early word productions are phonetically simple and usually follow the syllable structure CV or CVC, although this generalization has been challenged. The first vowels produced are, , and , followed by , , and , with rounded vowels emerging last. German children often use phonological processes to simplify their early word production. For example, they may delete an unstressed syllable (Schokolade 'chocolate' pronounced ), or replace a fricative with a corresponding stop (Dach 'roof' pronounced ). One case study found that a 17-month-old child acquiring German replaced the voiceless velar fricative with the nearest available continuant, or deleted it altogether (Buch 'book' pronounced or ). Prosodically, children prefer bisyllabic words with the pattern strong – weak over monosyllabic words.

Vowel space development

In 2009, Lintfert examined the development of vowel space of German speakers in their first three years of life. During the babbling stage, vowel distribution has no clear pattern. However, stressed and unstressed vowels already show different distributions in the vowel space. Once word production begins, stressed vowels expand in the vowel space, while the F1 – F2 vowel space of unstressed vowels becomes more centralized. The majority of infants are then capable of stable production of F1. The variability of formant frequencies among individuals decreases with age. After 24 months, infants expand their vowel space individually at different rates. However, if the parents' utterances possess a well-defined vowel space, their children produce clearly distinguished vowel classes earlier. By about three years old, children command the production of all vowels, and they attempt to produce the four cardinal vowels,, , and , at the extreme limits of the F1–F2 vowel space (i.e., the height and backness of the vowels are made extreme by the infants).

Nasals

The acquisition of nasals in German differs from that of Dutch, a phonologically closely related language. German children produce proportionately more nasals in onset position (sounds before a vowel in a syllable) than Dutch children do. German children, once they reached 16 months, also produced significantly more nasals in syllables containing schwas, when compared with Dutch-speaking children. This may reflect differences in the languages the children are being exposed to, although the researchers claim that the development of nasals likely cannot be seen apart from the more general phonological system the child is developing.

Phonotactic constraints and reading

A 2006 study examined the acquisition of German in phonologically delayed children (specifically, issues with fronting of velars and stopping of fricatives) and whether they applied phonotactic constraints to word-initial consonant clusters containing these modified consonants. In many cases, the subjects (mean age = 5.1) avoided making phonotactic violations, opting instead for other consonants or clusters in their speech. This suggests that phonotactic constraints do apply to the speech of German children with phonological delay, at least in the case of word-initial consonant clusters. Additional research has also shown that spelling consistencies seen in German raise children's phonemic awareness as they acquire reading skills.

Sound changes

Sound changes and mergers

A common merger is that of at the end of a syllable with or, for instance Krieg ('war'), but Kriege ('wars'); er lag ('he lay'), but wir lagen ('we lay'). This pronunciation is frequent all over central and northern Germany. It is characteristic of regional languages and dialects, particularly Low German in the North, where ⟨g⟩ represents a fricative, becoming voiceless in the syllable coda, as is common in German (final-obstruent devoicing). However common it is, this pronunciation is considered sub-standard. Only in one case, in the grammatical ending -ig (which corresponds to English -y), the fricative pronunciation of final ⟨g⟩ is prescribed by the Siebs standard, for instance wichtig ('important'), Wichtigkeit ('importance'). The merger occurs neither in Austro-Bavarian and Alemannic German nor in the corresponding varieties of Standard German, and therefore in these regions -ig is pronounced. Many speakers do not distinguish the affricate from the simple fricative in the beginning of a word, in which case the verb (er) fährt ('[he] travels') and the noun Pferd ('horse') are both pronounced. This most commonly occurs in northern and western Germany, where the local dialects did not originally have the sound. Some speakers also have peculiar pronunciation for in the middle or end of a word, replacing the in with a voiceless bilabial fricative, i.e. a consonant produced by pressing air flow through the tensed lips. Thereby Tropfen ('drop') becomes, rather than. Many speakers who have a vocalization of after merge this combination with long (i.e. > or > or ). Hereby, Schaf ('sheep') and scharf ('sharp') can both be pronounced or. This merger does not occur where is a front vowel while is realised as a back vowel. Here the words are kept distinct as ('sheep') and ('sharp'). In umlaut forms, the difference usually reoccurs: Schäfer or vs. schärfer. Speakers with this merger also often use (instead of formally normal ) where it stems from original. The word Archen ('arks') is thus pronounced, which makes a minimal pair with Aachen , arguably making the difference between and phonemic, rather than just allophonic, for these speakers. In the standard pronunciation, the vowel qualities, , , , as well as , , , , are all still distinguished even in unstressed syllables. In this latter case, however, many simplify the system in various degrees. For some speakers, this may go so far as to merge all four into one, hence misspellings by schoolchildren such as Bräutegam (instead of Bräutigam) or Portogal (instead of Portugal). In everyday speech, more mergers occur, some of which are universal and some of which are typical for certain regions or dialect backgrounds. Overall, there is a strong tendency of reduction and contraction. For example, long vowels may be shortened, consonant clusters may be simplified, word-final may be dropped in some cases, and the suffix -en may be contracted with preceding consonants, e.g. for haben ('to have'). If the clusters, , , or are followed by another consonant, the stops , and usually lose their phonemic status. Thus while the standard pronunciation distinguishes ganz ('whole') from Gans ('goose'), as well as er sinkt from er singt, the two pairs are homophones for most speakers. The commonest practice is to drop the stop (thus, for both words), but some speakers insert the stop where it is not etymological (, for both words), or they alternate between the two ways. Only a few speakers retain a phonemic distinction.

Middle High German

The Middle High German vowels and developed into the modern Standard German diphthong, whereas and developed into. For example, Middle High German heiz and wîz ('hot' and 'white') became Standard German heiß and weiß. In some dialects, the Middle High German vowels have not changed, e.g. Swiss German heiss and wiiss, while in other dialects or languages, the vowels have changed but the distinction is kept, e.g. Bavarian hoaß and weiß , Ripuarian heeß and wieß (however the Colognian dialect has kept the original [ei] diphthong in heiß), Yiddish הײס heys and װײַס vays. The Middle High German diphthongs, and became the modern Standard German long vowels , and after the Middle High German long vowels changed to diphthongs. Most Upper German dialects retain the diphthongs. A remnant of their former diphthong character is shown when continues to be written ie in German (as in Liebe 'love').

Loanwords

German incorporates a significant number of loanwords from other languages. Loanwords are often adapted to German phonology but to varying degrees, depending on the speaker and the commonness of the word. and do not occur in native German words but are common in a number of French and English loan words. Many speakers replace them with and respectively (especially in Southern Germany, Austria and Switzerland), so that Dschungel (from English jungle) can be pronounced or. Some speakers in Northern and Western Germany merge with, so that Journalist (phonemically ) can be pronounced , or. The realization of as, however, is uncommon.

Loanwords from English

Many English words are used in German, especially in technology and pop culture. Some speakers pronounce them similarly to their native pronunciation, but many speakers change non-native phonemes to similar German phonemes (even if they pronounce them in a rather English manner in an English-language setting):

Loanwords from French

French loanwords, once very numerous, have in part been replaced by native German coinages or more recently English loanwords. Besides, they can also contain the characteristic nasal vowels , , and (always long). However, their status as phonemes is questionable and they are often resolved into sequences either of (short) oral vowel and (in the north), or of (long or short) oral vowel and or sometimes (in the south). For example, Ballon ('balloon') may be realized as or, Parfüm ('perfume') as or and Orange ('orange') as or.

Sample

The sample text is a reading of "The North Wind and the Sun". The phonemic transcription treats every instance of and as and, respectively. The phonetic transcription is a fairly narrow transcription of the educated northern accent. The speaker transcribed in the narrow transcription is 62 years old, and he is reading in a colloquial style. Aspiration, glottal stops and devoicing of the lenes after fortes are not transcribed. The audio file contains the whole fable and was recorded by a much younger speaker.

Phonemic transcription

Phonetic transcription

Orthographic version

Einst stritten sich Nordwind und Sonne, wer von ihnen beiden wohl der Stärkere wäre, als ein Wanderer, der in einen warmen Mantel gehüllt war, des Weges daherkam. Sie wurden einig, daß derjenige für den Stärkeren gelten sollte, der den Wanderer zwingen würde, seinen Mantel abzunehmen. Der Nordwind blies mit aller Macht, aber je mehr er blies, desto fester hüllte sich der Wanderer in seinen Mantel ein. Endlich gab der Nordwind den Kampf auf. Nun erwärmte die Sonne die Luft mit ihren freundlichen Strahlen, und schon nach wenigen Augenblicken zog der Wanderer seinen Mantel aus. Da mußte der Nordwind zugeben, daß die Sonne von ihnen beiden der Stärkere war.