Editing Germanic languages (section)

====Table of outcomes====
The following table shows the main outcomes of Proto-Germanic vowels and consonants in the various older languages. For vowels, only the outcomes in stressed syllables are shown. Outcomes in unstressed syllables are quite different, vary from language to language and depend on a number of other factors (such as whether the syllable was medial or final, whether the syllable was [[open syllable|open]] or [[closed syllable|closed]] and (in some cases) whether the preceding syllable was [[light syllable|light]] or [[heavy syllable|heavy]]).

Notes:
* ''C-'' means before a vowel (word-initially, or sometimes after a consonant).
* ''-C-'' means between vowels.
* ''-C'' means after a vowel (word-finally or before a consonant). Word-final outcomes generally occurred ''after'' deletion of final short vowels, which occurred shortly after Proto-Germanic and is reflected in the history of all written languages except for [[Proto-Norse]].
* The above three are given in the order ''C-'', ''-C-'', ''-C''. If one is omitted, the previous one applies. For example, ''f, -[v]-'' means that ''[v]'' occurs after a vowel regardless of what follows.
* Something like ''a(…u)'' means "''a'' if /u/ occurs in the next syllable".
* Something like ''a(n)'' means "''a'' if /n/ immediately follows".
* Something like ''(n)a'' means "''a'' if /n/ immediately precedes".

{| class="wikitable" style="text-align: center;"
|+ Development of Germanic sounds
! Proto-Germanic{{sfnp|Ringe|2006}}{{sfnp|König|van der Auwera|1994}} !! (Pre-)Gothic<ref group=lower-alpha name="pre-gothic">The [[Gothic language#Alphabet and transliteration|Gothic writing system]] uses the spelling {{angbr|ai}} to represent vowels that derive primarily from four different sources:
# Proto-Germanic /ai/
# Proto-Germanic {{IPA|/eː/}} and {{IPA|/æː/}} before vowels
# Proto-Germanic /e/ and /i/ before /h/, /hʷ/ and /r/
# Greek {{IPA|/ɛ/}}.
The spelling {{angbr|au}} is similarly used to represent vowels primarily deriving from the following four sources:
# Proto-Germanic /au/
# Proto-Germanic {{IPA|/oː/}} and {{IPA|/uː/}} before vowels
# Proto-Germanic /u/ before /h/, /hʷ/ and /r/
# Greek {{IPA|/ɔ/}}.
It is generally agreed that the outcome of case 2 was pronounced {{IPA|[ɛː/ɔː]}} in Gothic, distinct from the vowels written {{angbr|e}} and {{angbr|o}}, which were pronounced {{IPA|[eː/oː]}}. Likewise, it is generally agreed that the outcomes of cases 3 and 4 were pronounced {{IPA|[ɛ]}} and {{IPA|[ɔ]}} in Gothic.  However, there is some argument over whether the outcomes of case 1 were still pronounced as diphthongs {{IPA|[ai/au]}}, as in Proto-Germanic, or had merged with case 2 as monophthongs {{IPA|[ɛː/ɔː]}}. There is some historical evidence (particularly from Latin spelling variations of ''Gaut-'' vs. ''Gōt-'', used to represent the name of the Goths) that the Proto-Germanic diphthongs had changed into monophthongs shortly before (i.e., within a century of) the time of [[Wulfila]], who designed the [[Gothic alphabet]] and wrote the [[Gothic Bible]] c. 360 AD. This accords with the fact that Wulfila used the same symbols {{angbr|ai/au}} to represent all the outcomes, despite the fact that the spellings {{angbr|aj/aw}} were available to unambiguously represent diphthongs (and, in fact, alternate with {{angbr|ai/au}} in a number of nominal and verbal paradigms). The use of the spelling {{angbr|ai}} to represent a monophthong {{IPA|[ɛ(ː)]}} was evidently in imitation of 4th century Greek, where {{angbr|ai}} likewise stood for {{IPA|[ɛː]}}, and {{angbr|au}} was apparently created by analogy. Consistent with many sources, such as {{Harvcoltxt|Bennett|1980}}, the phonology described here is that of "Pre-Gothic" (i.e., the phonology of Gothic just before the monophthongization of /ai/ and /au/).</ref>{{sfnp|Bennett|1980}}{{sfnp|Wright|1919}} !! Old Norse{{sfnp|Gordon|1927}} !! Old English{{sfnp|Campbell|1959}}{{sfnp|Diamond|1970}}{{sfnp|Lass|Anderson|1975}}{{sfnp|Lass|1994}}{{sfnp|Mitchell|Robinson|1992}}{{sfnp|Robinson|1992}}{{sfnp|Wright|Wright|1925}} !! Old High German{{sfnp|Wright|1906}}{{sfnp|Waterman|1976}}<!-- !! Old Saxon{{sfnp|Helfenstein|1870}} !! Old Low Franconian !! Old Frisian -->
|- style="border-top: 3px solid;"
! a
| rowspan=2 | a || a, ɔ(...u)<ref group=lower-alpha name="norse-u-umlaut">In Old Norse, non-rounded vowels become rounded when a /u/ or /w/ follows in the next syllable, in a process known as [[Old Norse#U-umlaut|u-umlaut]].  Some vowels were affected similarly, but only by a following /w/; this process is sometimes termed ''w-umlaut''. These processes operated after [[i-umlaut]].  ''U-umlaut'' (by a following /u/ or /w/) caused /a/, /ja/ (broken /e/), /aː/, and /e/ to round to /ɔ/ (written ''ǫ''), /jɔ/ (written ''jǫ''), /ɔː/ (written ''ǫ́'' and later unrounded again to /aː/), and /ø/, respectively.  The vowels /i/ and /ai/ rounded to /y/ and /ey/, respectively, only before /w/.  Short /a/ become /ø/ by a combination of i-umlaut and w-umlaut.</ref> || æ, a(...a),<ref group=lower-alpha name="a-umlaut">A process known as [[a-mutation]] or ''a-umlaut'' caused short /u/ to lower to /o/ before a non-high vowel (usually /a/) in the following syllable.  All languages except Gothic were affected, although there are various exceptions in all the languages.  Two similar process later operated:
* In Old High German, /iu/ (from Proto-Germanic /eu/,/iu/) became /io/ before a non-high vowel in the next syllable.
* In Old English, /æ/ (from Proto-Germanic /a/) became /a/ before /a/ in the next syllable.
All of these processes were blocked in an i-umlaut context (i.e. by a following /j/).</ref> a/o(n), æ̆ă(h,rC,lC)<ref group=lower-alpha name="oe-breaking">The diphthongal results are due to [[Old English breaking]].  In general, front vowels break into diphthongs before some subset of ''h'', ''w'', ''rC'', and ''lC'', where ''C'' is a consonant.  The diphthong /æa/ is written ''ea''; /eo/ is written ''eo''; /iu/ is written ''io''; and /iy/ is written ''ie''. All diphthongs umlaut to /iy/ ''ie''. All diphthongs occur both long and short. Note that there is significant dispute about the actual pronunciation of ''io'' and (especially) ''ie''. Their interpretation as /iu/ and /iy/, respectively, follows Lass (1994), ''Old English: A historical linguistic companion''.</ref> || a
|-
! a(...i)<ref group=lower-alpha name="i-umlaut">All languages except Gothic were affected by [[i-umlaut]].  This was the most significant of the various [[Diaeresis (diacritic)#Umlaut|umlaut]] processes operating in the Germanic languages, and caused back vowels to become fronted, and front vowels to be raised, when /i/, /iː/ or /j/ followed in the next syllable.  The term ''i-umlaut'' actually refers to two separate processes that both were triggered in the same environment.  The earlier process raised /e/ and /eu/ to /i/ and /iu/, respectively, and may have operated still in Proto-Germanic (with its effects in Gothic obscured due to later changes).  The later process affected all back vowels and some front vowels; it operated independently in the various languages, occurring at differing times with differing results.  Old English was the earliest and most-affected language, with nearly all vowels affected.  Old High German was the last language to be affected; the only written evidence of the process is with short /a/, which is umlauted to /e/.  However, later evidence suggests that other back vowels were also affected, perhaps still sub-phonemically in Old High German times.  These are indicated with a [[diaeresis (diacritic)|diaeresis]] or "umlaut" symbol (two dots) placed over the affected vowels.</ref>
| e, ø(...u)<ref group=lower-alpha name="norse-u-umlaut" /> || e, æ, ĭy̆(h,rC,lC)<ref group=lower-alpha name="oe-breaking" /> || e, a(hs,ht,Cw)
|-
! ãː
| rowspan=2 | aː  || aː || oː || aː
|-
! ãː(...i)<ref group=lower-alpha name="i-umlaut" />
| æː || eː || äː
|-
! æː
| rowspan=2 | eː, ɛː(V) || aː || æː, æa(h)<ref group=lower-alpha name="oe-breaking" /> || aː
|-
! æː(...i)<ref group=lower-alpha name="i-umlaut" />
| æː || æː || äː
|-
! e
| rowspan=2 | i, ɛ(h,hʷ,r) || ja,<ref group=lower-alpha name="norse-breaking">Proto-Germanic /e/ usually became Old Norse /ja/ by a process known as [[vowel breaking]].</ref> jø(...u),<ref group=lower-alpha name="norse-u-umlaut" /> (w,r,l)e, (w,r,l)ø(...u)<ref group=lower-alpha name="norse-u-umlaut" /> || e, ĕŏ(h,w,rC)<ref group=lower-alpha name="oe-breaking" /> || e, i(...u)
|-
! e(...i)<ref group=lower-alpha name="i-umlaut" />
| i, y(...w)<ref group=lower-alpha name="norse-u-umlaut" /> || i || i
|-
! eː
| eː, ɛː(V) || eː || eː || {{not a typo|ie}}
|-
! i
| i, ɛ(h,hʷ,r) || i, y(...w)<ref group=lower-alpha name="norse-u-umlaut" /> || i, ĭŭ(h,w,rC)<ref group=lower-alpha name="oe-breaking" /> || i
|-
! iː
| iː || iː || iː, iu(h) || iː
|-
! oː
| rowspan=2 | oː, ɔː(V) || oː || oː || uo
|-
! oː(...i)<ref group=lower-alpha name="i-umlaut" />
| øː || eː || üö
|-
! u
| rowspan=2 | u, ɔ(h,hʷ,r) || u, o(...a)<ref group=lower-alpha name="a-umlaut" /> || u, o(...a)<ref group=lower-alpha name="a-umlaut" /> || u, o(...a)<ref group=lower-alpha name="a-umlaut" />
|-
! u(...i)<ref group=lower-alpha name="i-umlaut" />
| y || y || ü
|-
! uː
| rowspan=2 | uː, ɔː(V) || uː || uː || uː
|-
! uː(...i)<ref group=lower-alpha name="i-umlaut" />
| yː || yː || üː
|-
! ai
| rowspan=2 | ai<ref group=lower-alpha name="pre-gothic" /> || ei, ey(...w),<ref group=lower-alpha name="norse-u-umlaut" /> aː(h,r)<ref group=lower-alpha name="norse-ai">Before Proto-Germanic /x/, /xʷ/ or /r/, but not before Proto-Germanic /z/ (which only merged with /r/ much later in North Germanic). Cf. Old Norse ''árr'' (masc.) "messenger" < PG *airuz, ''ár'' (fem.) "oar" < PG *airō, vs. ''eir'' (fem.) "honor" < PG *aizō, ''eir'' (neut.) "bronze" < PG *aizan. (All four become ''ār'' in Old English; in Gothic, they become, respectively, ''airus'', (unattested), ''*aiza'', ''*aiz''.) Cf. {{cite web |last1=Köbler |first1=Gerhard |title=Altenglisches Wörterbuch |url=http://www.koeblergerhard.de/germanistischewoerterbuecher/altenglischeswoerterbuch/AENG-A.pdf |archive-url=https://web.archive.org/web/20030418155218/http://www.koeblergerhard.de/germanistischewoerterbuecher/altenglischeswoerterbuch/AENG-A.pdf |archive-date=18 April 2003 |url-status=live}}</ref> || aː || rowspan=2 | ei, eː(r,h,w,#)<ref group=lower-alpha name="ohg-ai">Before /r/, /h/ (including when derived from Proto-Germanic /xʷ/) or /w/, or word-finally.</ref>
|-
! ai(...i)<ref group=lower-alpha name="i-umlaut" />
| ei, æː(h,r) || æː
|-
! au
| rowspan=2 | au<ref group=lower-alpha name="pre-gothic" /> || au, oː(h) || æa || ou, oː(h,T)<ref group=lower-alpha name="ohg-au">Before /h/ (including when derived from Proto-Germanic /xʷ/) or before any [[dental consonant]], i.e. /s/,/z/,/þ/,/t/,/d/,/r/,/l/,/n/.</ref>
|-
! au(...i)<ref group=lower-alpha name="i-umlaut" />
| ey, øː(h) || iy || öü, öː(h,T)<ref group=lower-alpha name="ohg-au" />
|-
! eu
| rowspan=2 | iu || juː, joː(T)<ref group=lower-alpha name="dental">Before any [[dental consonant]], i.e. /s/,/z/,/þ/,/t/,/d/,/r/,/l/,/n/.</ref> || eo || rowspan=2 | io, iu(...i/u)<ref group=lower-alpha name="a-umlaut" />
|-
! eu(...i)<ref group=lower-alpha name="i-umlaut" />
| yː || iy
|-
! p
| p || p || p || pf-, -ff-, -f
|-
! t
| t || t || t || ts-, -ss-, -s<ref group=lower-alpha name="german-s">The result of the [[High German consonant shift]] produced a different sort of ''s'' than the original Proto-Germanic ''s''.  The former was written {{angbr|z}} and the latter {{angbr|s}}.  It is thought that the former was a [[dental consonant|dental]] /s/, somewhat like in English, while the latter was an "[[voiceless apicoalveolar fricative|apicoalveolar]]" sound as in modern European Spanish, sounding somewhere between English /s/ and /ʃ/.{{Harvcoltxt|Joos|1952}}) Modern standard German has /ʃ/ for this sound in some contexts, e.g. initially before a consonant (''schlimm'' cf. English ''slim''; ''Stand'' /ʃtant/, cf. English ''stand''), and after /r/ (''Arsch'', cf. English ''arse'' or ''ass''). A number of modern southern German dialects have /ʃ/ for this sound before all consonants, whether or not word-initially.</ref>
|-
! k
| k || k || k, tʃ(i,e,æ)-, -k-, -(i)tʃ-, -tʃ(i)-<ref group=lower-alpha name="oe-palatal">Old English [[Palatalization (sound change)|palatalizes]] /k,g,ɣ/ to /tʃ,dʒ,j/ near a front vowel.  The sounds /k/ and /ɣ/ palatalized initially before any front vowel.  Elsewhere /ɣ/ palatalized before /j/ or ''before or after'' any front vowel, where /k/ and /g/ (which occurred only in the combinations /gg/, /ng/) palatalized before /j/, or either before or after /i,iː/.</ref> || k-, -xx-, -x
|-
! kʷ
| kʷ || kv, -k || kw-, -k-, -(i)tʃ-, -tʃ(i)-<ref group=lower-alpha name="oe-palatal" /> || kw-, -xx-, -x
|-
! b-, -[β]-<ref group=lower-alpha name="voiced">Voiced fricatives were originally allophones of voiced stops, when occurring after a vowel or after certain consonants (and for /g/, also initially — hard [g] occurred only in the combinations /gg/, /ng/).  In Old Norse and Old English, voiceless fricatives became voiced between vowels (and finally after a vowel in Old Norse); as a result, voiced fricatives were reanalyzed as allophones of voiceless fricatives.  In Old High German, all voiced fricatives hardened into stops.</ref>
| b-, -[β]-, -f || b-, -[v]- || b-, -[v]-, -f || b
|-
! d-, -[ð]-<ref group=lower-alpha name="voiced" />
| d-, -[ð]-, -þ || d-, -[ð]- || d || t
|-
! [ɣ]-, -[ɣ]-<ref group=lower-alpha name="voiced" />
| g-, -[ɣ]-, -[x] || g-, -[ɣ]- || g-, j(æ,e,i)-, -[ɣ]-, -j(æ,e,i)-, -(æ,e,i)j-<ref group=lower-alpha name="oe-palatal" /> || g
|-
! f
| f || f, -[v]- || f, -[v]-, -f || f, p
|-
! þ
| þ || þ, -[ð]- || þ, -[ð]-, -þ || d
|-
! x
| h || h, -∅- || h, -∅-, -h || h
|-
! xʷ
| hʷ || xv, -∅- || hw, -∅-, -h || hw, -h-
|-
! s
| s || s-, -[z]- || s-, -[z]-, -s || ṣ-, -[ẓ]-, -ṣ<ref group=lower-alpha name="german-s" />
|-
! z
| -z-, -s || r || -r-, -∅ || -r-, -∅
|-
! r<ref group=lower-alpha name="r">In the early periods of the various languages, the sound written /r/ may have been strongly [[velarized]], as in modern [[American English]] (Lass 1994); this is one possible explanation for the various processes were triggered by ''h'' (probably {{IPA|[x]}}) and ''r''.</ref>
| r || r || r || r
|-
! l
| l || l || l || l
|-
! n
| n || n-, -∅(s,p,t,k),<ref group=lower-alpha name="lost-n">Old English and Old Norse lose /n/ before certain consonants, with the previous vowel lengthened (in Old Norse, the following consonant is also lengthened).</ref> -∅<ref group=lower-alpha>/n/ lost finally and before /s,p,t,k/, but not before other consonants.</ref> || n, -∅(f,s,þ)<ref group=lower-alpha name="lost-n" /> || n
|-
! m
| m || m || m || m
|-
! j<ref group=lower-alpha name="jw">Proto-Germanic /j/ and /w/ were often lost between vowels in all languages, often with /j/ or /w/ later reappearing to break the hiatus, and not always corresponding to the sound previously present.  After a consonant, Gothic consistently preserved /j/ and /w/, but most languages deleted /j/ (after triggering [[i-umlaut]]), and /w/ sometimes disappeared.  The loss of /j/ after a consonant occurred in the various languages at different times and to differing degrees.  For example, /j/ was still present in most circumstances in written Old Saxon, and was still present in Old Norse when a short vowel preceded and a back vowel followed; but in Old English and Old High German, /j/ only remained after an /r/ preceded by a short vowel.</ref>
| j || ∅-, -j-, -∅ || j || j
|-
! w<ref group=lower-alpha name="jw" />
| w || ∅-, v-(a,e,i), -v-, -∅ || w || w
|}

{{Reflist|group=lower-alpha}}