Upper Saxon (Obersächsisch /ɵːpoˁˈsɛksʃ/) refers to a group of dialects spoken by over two million people in the Free State of Saxony in eastern Germany. It is considered one of the eastern branches of Central German (Wiesinger Reference Wiesinger, Besch, Knoop, Putschke and Wiegand1983, Lewis Reference Lewis2009), with major phonological, morphological, and lexical differences from Standard German and other regional dialects.
The transcriptions below reflect the speech of middle-aged speakers from Chemnitz, speaking an urban variety of the local Vorerzgebirgisch /foˁːˈaˁːtskəbʌˁːʃ/ dialect, which is described in Bergmann (Reference Bergmann1990: 292) as transitional between the Meissen (Meißnisch), Vogtland (Vogtländisch), and Ore Mountain (Erzgebirgisch) dialects. Due to both this transitional nature and a lesser degree of influence from Standard German (Hochdeutsch) than what is seen in other urban centers (e.g. Leipzig, Dresden), the Chemnitz dialect is largely intelligible to speakers of other varieties of Upper Saxon while still preserving the most salient phonological and phonetic features recognizable to speakers of other varieties of German as defining characteristics of Upper Saxon.
Of course, as the degree of influence from regional dialects and from Standard German varies greatly across speakers and contexts, this illustration should not be taken to be representative of all speakers in Chemnitz, let alone of all varieties of Upper Saxon. See Bergmann (Reference Bergmann1965) for a detailed historical description of this Vorerzgebirgisch variety, Keller (Reference Keller1960), Bergmann (Reference Bergmann1990: 312), Kügler (Reference Kügler2005: 18; 2007: 11), and Rues et al. (2007: 91–99) for descriptions of and references on other varieties of Upper Saxon, and Kleber (Reference Kleber2011) for the pronunciation of Standard German by Upper Saxon speakers. Examples of Standard German are given for comparison of selected forms, and are based on Rues et al.'s (2007) transcription scheme.
Consonants
Voicing and aspiration
Unlike Standard German, Upper Saxon lacks a voicing contrast. Underlyingly, obstruents are voiceless and sonorants are voiced (Becker Reference Becker1942: 104, 127–128; Bergmann Reference Bergmann1965: 43, 1987: 18, 1990: 309–310; Zimmermann Reference Zimmermann1992: 102–107; Rues et al. Reference Rues, Redecker, Koch, Wallraff and Simpson2007: 94; Kleber Reference Kleber2011). Unaspirated stops /p tk/ can be partially or fully voiced [b d ɡ] between sonorants, as in Figure 1 and Figure 3 below; word-finally, they can vary widely from fully voiced stops to voiceless aspirated stops [pʰ tʰ kʰ], as in Figure 2, but are most commonly realized as voiceless unaspirated stops (Bergmann Reference Bergmann1965: 43; Albrecht Reference Albrecht1983: 11), as in Figure 3. The sole aspiration distinction /kʰ k/ is restricted to initial position (Bergmann Reference Bergmann1965: 43), e.g. /ˈkʰʌsə/ Kasse ‘cash register’ vs. /ˈkʌsə/ Gasse ‘lane’ above. Neighboring varieties of Central and Upper German are also described as having a similar lack of a contrast in either voicing or aspiration (i.e. fortis vs. lenis) outside of velars, e.g. Thuringian (Spangenberg Reference Spangenberg1990: 270), East Franconian (Rowley Reference Rowley1990: 400), and Swabian (Hiller Reference Hiller1995: 34–35).
Spectrogram of the phrase /tiː ˈ
ɵːmɪʃ
ˈtɛmp
n/ die römischen Tempeln ‘the Roman temples’, showing full voicing of /t p/ in Tempeln /ˈtɛmpn/ (Standard German Tempel /ˈtɛmp/).
Spectrogram of the phrase /ən ˈɵːtəs tleːt/ ein rotes Kleid ‘a red dress’, showing full voicing of /t/ in /ˈɵːtəs/ rotes ‘red’ and aspiration of the final /t/ in /tleːt/ Kleid ‘dress’, as well as the coronal articulation of the pre-lateral consonant.
Spectrogram of the phrase /hʌst tʉː ˈʋʌˁːkləʃ ɛʃt/ hast du wirklich Recht ‘(do you think) you're really right?’, showing full voicing of /k/ in /ˈʋʌˁːkləʃ/ wirklich ‘really’ and retraction of the final /t/ in /ɛʃt/ Recht ‘right’ due to the preceding /ʃ/.
Fricatives /f s ʃ χ/ can be partially or fully voiced [v z ʒ ʁ] between sonorants, but remain largely voiceless elsewhere. Fricatives /χ h/ occur in complementary distribution: /h/ occurs as the onset of stressed syllables and word-initially while /χ/ occurs elsewhere.
Due to the lack of a voicing contrast, Upper Saxon has many homophone pairs whose Standard German cognates form minimal pairs, as illustrated in the list below. The only vestige of a voicing distinction in Upper Saxon is the aspiration contrast /kʰ k/, which often corresponds to /k ɡ/ in Standard German; however, there are instances of Upper Saxon ‘lenis’ /k/ corresponding to Standard German ‘fortis’ /k/ instead of to ‘lenis’ /ɡ/, e.g. Upper Saxon /koˁːp/ Korb ‘basket’ and Standard German /kɔ
p/ ([kʰɔ
p]). Like most varieties of German, Upper Saxon exhibits only voiceless obstruents word-finally; in Standard German, this is due to final devoicing (Brockhaus Reference Brockhaus1995), while in Upper Saxon, this is due to the lack of a voicing contrast to begin with.
Clusters
Neither voicing nor aspiration is contrastive in clusters (Bergmann Reference Bergmann1965: 110), even in the velar place of articulation, producing homophones where Standard German has (near-) minimal pairs of voicing as in the examples below. In initial clusters with /l/, stops /t k/ do not contrast and can vary freely, e.g. /klʌːs/~/tlʌːs/ Glas ‘glass’ (Goepfert Reference Goepfert1878: 25; Sievers Reference Sievers1885: 160; Bergmann Reference Bergmann1965: 110; Blevins & Grawunder Reference Blevins and Grawunder2009); the /tl/ variant is a salient marker of Chemnitz speakers (Wallner-Zimmer Reference Wallner-Zimmer1999; Blevins & Grawunder Reference Blevins and Grawunder2009: 271–272), although it is widespread throughout eastern variants of Central German and Upper German.Footnote 2 In Figure 2, the spectrogram of the phrase /ən ˈ
ɵːtəs tleːt/ ein rotes Kleid ‘a red dress’ reveals a high-frequency (~5 kHz) concentration of burst energy in the onset of /tleːt/ Kleid ‘dress’, indicating a coronal articulation. Final /ʃ/, including the suffixes spelled -ig/-ich/-isch (e.g. /hɔˁː-ʃ/ haarig ‘hairy’), can create clusters not found in Standard German, e.g. /ɛsʃ/ Essig ‘vinegar’, /sɛks-ʃ/ Sächsisch ‘Saxon’, and /sɞnst-ʃ/ sonstig ‘other’ (Bergmann Reference Bergmann1965: 115; Albrecht Reference Albrecht1983: 19; Gilles Reference Gilles2005: 70).
Rhotic
As in Standard German (Kohler Reference Kohler1999: 87–88), the rhotic // occurs only in onsets. In Upper Saxon, it freely varies between a voiced approximant [], a voiced fricative [ʁ], a devoiced fricative [
] or trill [
], and a voiceless unaspirated uvular stop [q], e.g. /ɞk/ [ɔkʰ] Rock ‘skirt’, /ɵm/ [
m]~[m] Rum ‘rum’, and /a
/ [a]~[qa] rau ‘rough’. What was historically its coda counterpart only surfaces in Upper Saxon as pharyngealization on the preceding vowel, e.g. /ˈmaˁːəə ˈlɞ
tə/ mehrere Leute ‘more people’ vs. /ɪʃ keː nɪʃ maˁːtsʉː ˈʃʉːlə/ ich gehe nicht mehr zu Schule ‘I don't go to school anymore’; in other German varieties known to exhibit pharyngealization, such as Swabian, the coda rhotic is treated as either a pharyngeal approximant [
] or pharyngealization on the preceding vowel [ˁ] (Frey Reference Frey1975: 15–16; Hiller Reference Hiller1995: 35; Rues et al. Reference Rues, Redecker, Koch, Wallraff and Simpson2007: 95–98), but consistently as a surface realization of an underlying consonant //. See section ‘Vowels and diphthongs’ below for details of pharyngealization.
Syllabic consonants
As in other colloquial varieties of German, final nasals and laterals can be produced as syllabic /
/ following a stop or fricative (Albrecht Reference Albrecht1983: 19), e.g. /ˈɔˁːpa
t/ arbeiten ‘to work’, /ˈmʌnt/ Mantel ‘cloak’. Syllabic nasals share the same place of articulation as the preceding consonant, e.g. /ˈhɛlf/ helfen ‘to help’, /ˈtɵk/ ducken ‘to duck’, and /ˈmʌχ/ [ˈm
χ
] machen ‘to make’. Due to this obligatory assimilation, syllabic nasals can be considered underlyingly underspecified for place. When the syllabic nasal is preceded by another nasal, the two are produced as a single consonant, sometimes with lengthening (Bergmann Reference Bergmann1965: 102), e.g. /kəˈkɞm/ gekommen ‘come’, /kəˈʋɞn/ gewonnen ‘won’, and /kəˈkʌŋ/ gegangen ‘gone’. When the vowel before the two final consonants is long, it is common for the oral consonant to be deleted, e.g. /ˈpɛːp/ beben ‘to shake’ pronounced [pɛːm]. Similar phenomena are also described in colloquial Standard German (Lemke Reference Lemke and Jonach1998; Rues et al. Reference Rues, Redecker, Koch, Wallraff and Simpson2007: 72).
Glottal stop
As in Standard German (Kohler Reference Kohler1999: 86; Rues et al. Reference Rues, Redecker, Koch, Wallraff and Simpson2007: 37), a glottal stop [ʔ] is typically inserted before stressed onset-less vowels both within words, e.g. /teːˈʌːtoˁ/ [teːˈʔɔˁːtoˁ] Theater ‘theater’, /foˁːˈaˁːtskəpʌˁːʃ/ [foˁːˈʔɛː
ˁtskəpɪːˁʃ] Vorerzgebirgisch, and word-initially regardless of stress, e.g. /ʌm ɵːmt ʋʌˁːt ˈɵːmpɵːt kəˈkɛs/ [ʔʌm ʔɵːmt ʋʌˁːt ˈʔɵːmbɵːt kəˈkɛz] Am Abend wird Abendbrot gegessen ‘Supper is eaten in the evening’.
Sibilants
Due to various mergers, Upper Saxon /ʃ/ can correspond to Standard German /k ɡ ʃ ç/, e.g. /ʋaˁːʃ/ Werk ‘work’, /tsʋaˁːʃ/ Zwerg ‘dwarf’, /tɪʃ/ Tisch ‘table’, and /tɪʃ/ dich ‘you’, cf. Standard German /vɛk/, /tsvɛk/ (underlyingly final /ɡ/, cf. plural /tsvɛɡ-ə/ Zwerge), /tɪʃ/, and /dɪç/ (Spangenberg Reference Spangenberg1990: 274). Its retracted articulation carries over into a following /t/, audible in words such as /ʋʊˁːʃt/ [ʋʊˁʃt] Wurst ‘sausage’. In Figure 3, the spectrogram of the word /ɛʃt/ Recht ‘right’ reveals the lower-frequency burst energy (< 3 kHz) of the /t/ following /ʃ/.
Vowels and diphthongs
Upper Saxon has six long non-pharyngealized vowels /iː eː ɛː ʉː ɵː ʌː/. The vowel chart reflects mean formant values collected via spectrographic analysis.
There are also five short non-pharyngealized vowels /ɪ ɛ ɵ ɞ ʌ/. When unstressed, these vowels optionally reduce to [ə]. Because of the varied formant values reduced vowels can take, no attempt is made in assessing their phonemic quality; all reduced vowels are simply transcribed /ə/. Utterance-final /ə/ is often fronted (e.g. /ˈtɪkə f
a
/ dicke Frau ‘fat woman’ vs. /ˈtɪkə/ dicke ‘fat’). The vowel chart reflects mean formant values collected via spectrographic analysis.
Pharyngealization is described in various regional dialects of German (e.g. Swabian, see Frey Reference Frey1975: 15–16; Hiller Reference Hiller1995) as well as in variants of Standard German (Lodge Reference Lodge2003). Upper Saxon has six pharyngealized vowels, five of which are long /aˁː ʌˁː ʊˁː oˁː ɔˁː/ and often pronounced as pharyngealized diphthongs [ɛː
ˁ ɪː
ˁ ʊː
ˁ oː
ˁ ɔː
ˁ] in careful speech; compare /aˁː/ [ɛː
ˁ] er ‘he’ in isolation versus in the phrase /jə maˁː aˁː ˈpʉːstətə/ [jə mɛː
ˁ ʔaˁː ˈpʉːstətə] je mehr er pustete ‘the more he blew’.Footnote 3 The monophthongal pronunciation is common in vowels optionally shortened before coda clusters (see Swabian, Hiller Reference Hiller1995); compare /faˁːt/ [fɛː
ˁt] fährt ‘goes’ with /faˁːtʃ/ [faˁtʃ] fertig ‘ready’. The remaining pharyngealized vowel is short unstressed /oˁ/, which can be variously analyzed as pharyngealized high-mid rounded vowel /ɵˁ/, pharyngealized schwa /əˁ/, a sequence of a schwa-like vowel and the rhotic /ə
/ (/ɔr/ in Bergmann Reference Bergmann1965), or a syllabic rhotic /
/ (vocalic /r/ in Bremer Reference Bremer1968: 171).Footnote 4 This is the counterpart of the Standard German lower-mid central vowel /ɐ/. Acoustically, pharyngealization in Upper Saxon is characterized by a significantly lowered F2 frequency and a slightly raised F3 frequency, similar to pharyngealization in other languages, such as Arabic (Obrecht Reference Obrecht1968, Laufer & Baer Reference Laufer and Baer1988, McCarthy Reference McCarthy and Keating1994, Yeou Reference Yeou2001, Al-Masri & Jongman Reference Al-Masri, Jongman, Agwuele, Warren and Park2003, among others), Hebrew (Laufer & Baer Reference Laufer and Baer1988), and Hongyan Qiang (Evans Reference Evans2006).Footnote 5 The vowel chart reflects mean formant values collected via spectrographic analysis.
As in analyses of the Dresden variety of Upper Saxon (Rues et al. Reference Rues, Redecker, Koch, Wallraff and Simpson2007: 92–99), Swabian (Frey Reference Frey1975: 15–16; Hiller Reference Hiller1995), and Standard German (Lodge Reference Lodge2003), these vowels can alternatively be analyzed as sequences of non-pharyngealized vowels followed by a rhotic consonant (Bergmann Reference Bergmann1965) of either uvular /eːiː ʉːʁ ɵː ʌː/ or pharyngeal /eːiː ʉː ɵː ʌː/ articulation. The latter analysis follows that of the pharyngealized short vowels of Swabian [ɛ ɵ ʊ ɔ ɑ], which are interpreted as surface realizations of underlying sequences /e~ɛiuo ɑ/ (Hiller Reference Hiller1995: 45–46). While the current description of Upper Saxon treats pharyngealization as a vocalic feature, it can be easily reinterpreted as the realization of an underlying consonant following German linguistic tradition.
In addition to the pharyngealized diphthongs, Upper Saxon has three non-pharyngealized diphthongs /ɞ
a
a
/. The vowel chart reflects mean formant values collected via spectrographic analysis.
Anticipatory pharyngealization and retraction
Vowels are allophonically pharyngealized when the vowel in the following syllable is underlyingly pharyngealized /aˁː ʌˁː ʊˁː oˁ oˁː ɔˁː/; Rues et al. (2007: 97–98) transcribes the Dresden variety with pharyngealization spreading in both directions onto consonants and vowels alike (e.g. [fʊˁlˁ
ˁnˁ] verloren ‘lost’, [
ˁd
ˁ] worden ‘been’, [
ˁ
ˁ
ˁd┐] Sport ‘sports’). In a separate but superficially similar process, vowels are somewhat retracted when followed by dorsals /k kʰ ŋ χ
/, with strongest retraction preceding uvulars. Contrastive pharyngealization, allophonic pharyngealization, and allophonic retraction are illustrated in the list below. Some retraction is also seen in vowels following dorsal consonants (e.g. /nʉː/ [nʉː] nun ‘now’ vs. /kʰʉː/ [kʰ
ː] Kuh ‘cow’), with the strongest such retraction seen following /
/ (e.g. /
ʉːm/ [
uːm] Ruhm ‘glory’).
Front rounded vowels
As is also reported in other Central German varieties such as those of Berlin (Peters Reference Peters2004: 209) and Thuringia (Spangenberg Reference Spangenberg1990: 270), as well as in Upper German varieties (Chambers & Trudgill Reference Chambers and Trudgill1998: 168) such as Swabian (Frey Reference Frey1975: 47; Russ Reference Russ1990a: 346–347; Hiller Reference Hiller1995: 36–40), the Upper Saxon counterparts of Standard German front rounded vowels /yː øː ʏ œ/ historically merged with the unrounded vowels /iː eː ɪ ɛ/, e.g. /ˈfeːʃ/ Vögel ‘birds’, /ˈkiːə/ Kühe ‘cows’, cf. Standard German /ˈføːɡ/, /ˈkyːə/ (Bergmann Reference Bergmann1965: 57–64, 1990: 309; Albrecht Reference Albrecht1983: 7–8; Zimmermann Reference Zimmermann1992: 103–104; Rues et al. Reference Rues, Redecker, Koch, Wallraff and Simpson2007: 93). However, Upper Saxon speakers fluent in Standard German occasionally produce front rounded /yː ʏ øː œ/ in cognates of particular Standard German words, e.g. /ˈpyːtoˁ/~/ˈpiːtoˁ/ Brüder ‘brothers’, contrasting them with central rounded vowels /ʉː ɵ ɵː ɞ/, e.g. /ˈløːf/ Löwen ‘lions’ vs. /ˈlɵːf/ laufen ‘to walk’, /tʏn/ dünn ‘thin’ vs. /ˈtɵns/ Dunsel ‘idiot’, cf. Standard German /ˈbʁyːdɐ/, /ˈløːv/, /ˈla
f/, /dʏn/, /ˈdʊnz/.
Stress
As in Standard German (Kohler Reference Kohler1999: 87), stress in native roots can be considered primarily initial or penultimate (see Wiese Reference Wiese1996: Section 8), while borrowed and polymorphemic words can exhibit other stress patterns. Because of this potential for ambiguity, stress is transcribed in all polysyllabic examples in this entry. For a detailed account of how stress interacts with intonation in other varieties of Upper Saxon, see Kügler (Reference Kügler2005, 2007) for the Leipzig dialect and Selting (Reference Selting2002a, b), Peters (Reference Peters2004), and Gilles (Reference Gilles2005) for the Dresden dialect.
Transcription of recorded passage
Broad transcription
ˈeːnəs ˈtʌːχəs hʌmʃ toˁ ˈnoˁːtʋɪnt ɵnt tə ˈsɞnə kəˈtsʌŋt, ʋaˁː fɞn ˈpeːttɛn nʉː toˁ ˈʃtaˁːkə ɪs, ɛls ə ˈʋʌntoˁ mɪt nəm ˈʋɔˁːmən ˈmʌnt ʌn foˁˈpeːkʰʌːm. toˁ ˈnoˁːtʋɪnt ɵnt tə ˈsɞnə ʋɔˁːnʃ a
nʃ tʌs toˁ ˈʃtaˁːkə fɞn ˈpeːttɛn ˈmʌntfɞm ˈʋʌntoˁ ˈkiːʃsɞl. toˁ ˈnoˁːtʋɪnt ˈpʉːstətə ʋʌs tʌs tsɞʃ hiːlt ˈʌːpoˁ jə maˁː aˁː ˈpʉːstətə ɵm sɵː maˁː foˁˈkiːʃtə sɪʃ toˁ ˈʋʌntoˁ ɪn ˈsan ˈmʌnt. toˁ ˈnoˁːtʋɪnt kʌːp ɞf. tʌn hʌts tə ˈsɞnə ɵːχ foˁˈsʉːχt mɪt ʌˁːn ˈʋɔˁːm ˈsɞnʃtɔˁːln. ɵnt ɪm nʉː ʃmɪs toˁ ˈʋʌntoˁ ˈsan ˈmʌnt ʋɛʃ. tɔˁː ˈmɵstə toˁ ˈnoˁːtʋɪnt ˈtsʉːkɛːptʌs tə ˈsɞnə toˁ ˈʃtaˁːkə fɞn ˈpeːt ɪs.
Orthographic version (Standard German)
Eines Tages haben sich der Nordwind und die Sonne gezankt, wer von den beiden denn nun der Stärkere ist, als ein Wanderer mit einem warmen Mantel an, vorbeikam. Der Nordwind und die Sonne waren sich einig, dass der Stärkere von den beiden den Mantel vom Wanderer kriegen soll. Der Nordwind pustete was das Zeug hielt, aber je mehr er pustete, um so mehr verkriechte sich der Wanderer in seinen Mantel. Der Nordwind gab auf. Dann hat es die Sonne auch versucht mit ihren warmen Sonnenstrahlen. Und im Nu schmiss der Wanderer seinen Mantel weg. Da musste der Nordwind zugeben, dass die Sonne die Stärkere von den beiden ist.
English translation
One day the North Wind and the Sun were disputing which of the two is the stronger, when a traveler came along in a warm cloak. The North Wind and the Sun agreed that the stronger of the two should take away the cloak from the traveler. The North Wind blew as hard as he could, but the more he blew, the more the traveler held onto his cloak. The North Wind gave up. Then the Sun tried it with her warm rays. And in an instant the traveler took off his cloak. Thus the North Wind had to concede that the Sun is the stronger of the two.
Acknowledgments
We would like to sincerely thank our primary consultant, Rico Weise, and our other speakers for their time and patience, as well as Bruce P. Hayes, Russell G. Schuh, Adrian P. Simpson, Ewa Jaworska, and three anonymous reviewers for their valuable input.
