<<

. 7
( 12)



>>

Er besch¨ digte das Auto.
a
˜He (agent) damaged the car (patient).™
b. passive voice
Das Auto wurde (von ihm) besch¨ digt.
a
˜The car (patient) was damaged (by him [agent]).™
Notice that the passive in German is periphrastic: it is formed using the
past participle of a verb (besch¨ digt ˜damaged™ in the example above)
a
together with a form of werden as an auxiliary (wurde ˜was™ in the example
above).31
In addition to the werden-passive, German also has a sein-passive. These
two passive types differ in their semantics as well as in their auxiliaries.
Semantics 173

(84) a. Das Auto wird gewaschen. (werden-passive)
˜The car is being washed.™
b. Das Auto ist gewaschen. (sein-passive)
˜The car is washed.™

The werden-passive expresses a process, for example, the process of washing
a car, as in (84a). The sein-passive expresses a state that results from a process,
for example, the state resulting from having been washed, as in (84b).
In German (as well as in English) there is also a middle construction, which
has characteristics of both active and passive clauses. The verb in the middle
construction in (85c), for example, has active morphology, like the verb in the
active sentence in (85a). The subject of the middle, however, is a patient, and
is thus similar to the patient subject of the passive in (85b).
(85) a. active
Er liest das Buch.
˜He (agent) reads the book (patient).™
b. passive
Das Buch wird gelesen.
˜The book (patient) is being read.™
c. middle
Das Buch liest sich leicht.
the book reads re¬‚ easily ˜The book (patient) reads easily.™
In the following section (4.6.1) we will look at the semantics of the passive,
and in section 4.6.2 we will look at the semantics of the middle.


4.6.1 The passive
We will focus this discussion on the semantic restrictions on werden-passives.
There are two types of werden passives: personal, which are formed from verbs
with accusative objects in the active, as in (86); and impersonal, which are
formed from verbs without accusative objects, as in (87).32
(86) a. Er zerreißt den Zettel.
he tears-up the note-acc. ˜He tears up the note.™
b. Der Zettel wird zerrissen. ˜The note is torn up.™
¨
(87) a. Sie tanzen uberall. ˜They™re dancing everywhere.™
¨
b. Uberall wird getanzt. ˜There™s dancing everywhere.™
Not all verbs with accusative objects can be used to form personal passives.
Those verbs that most closely ¬t the transitive prototype (action predicates with
agentive subjects and totally affected objects) form the most acceptable personal
174 German

passives (Shannon 1987; Dudenredaktion 2005:553) “ verbs like zerreißen
˜to tear up™, for example.33 Verbs that ¬t this prototype less closely are less
acceptable in the passive. For example, passives formed from verbs like haben
˜to have™, besitzen ˜to own™, wissen ˜to know™, enthalten ˜to contain™, and
umfassen ˜to include™, which are stative and have non-agentive subjects, are
unacceptable.

Ein BMW wird von mir gehabt/besessen.
(88)
˜A BMW is had/owned by me.™
Verbs like erhalten ˜to receive™, bekommen ˜to receive™, and kriegen ˜to get™,
which have goal subjects, do not form good passives.

(89) Ein Brief wurde von ihm erhalten/bekommen.
˜A letter was received by him.™
Verbs with objects that are a part of the subject™s body do not form good passives
(Drosdowski 1984:182).
(90) a. Er sch¨ ttelt den Kopf. ˜He shakes his head.™
u

Der Kopf wird von ihm gesch¨ ttelt. ˜The head is shaken by him.™
u
b.
The (active) subject in (90a) does not ¬t the transitive prototype because it
is not separate from the patient and not unaffected by the action (Shannon
1987); hence the unacceptability of the passive in (90b). A passive with a verb
like sch¨ tteln ˜to shake™ is perfectly acceptable, however, if the entity being
u
shaken is not a part of the subject™s body.
(91) Das Fass wurde gesch¨ ttelt. ˜The barrel was shaken.™
u
Intransitive verbs like tanzen ˜to dance™ and feiern ˜to celebrate™ may be
used to form passives in German. Although these verbs do not ¬t the transitive
prototype, since they lack objects, they are acceptable in passive clauses because
they have highly agentive subjects. Intransitive verbs with subjects that are not
agentive do not form good passives. In the following example (from Zifonun
et al. 1997:1806), the underlying subject is not animate and hence not agentive.
(92) a. Der Baum ¬el. ˜The tree fell.™

Hier wurde (von dem Baum) gefallen. ˜Here it was fallen (by the
b.
tree).™
The passive in (93) is acceptable, but only if it is understood as referring to a
noise made by a human subject; it cannot refer to noise being made by a door.
(93) Es wurde laut gequietscht. (Zifonun et al. 1997:1805)
˜There was loud squeaking.™
The verbs bluten ˜to bleed™ and sterben ˜to die™, typically classi¬ed as non-
agentive predicates, should not be acceptable in impersonal passives. However,
Semantics 175

if these verbs are understood as describing volitional acts, as in the following
(attested) example, impersonal passives are ¬ne:
(94) F¨ r den lieben K¨ nig und Herrn wird alles getan, wird treulich
u o
gek¨ mpft, wird willig geblutet, wird freudig in den Tod gegangen,
a
f¨ r ihn wird mehr als gestorben. (attested; Curme 1960:338)
u
˜Everything is done for the beloved king and lord, battle is loyally
fought, blood is willingly shed, death is entered happily, more is
done for him than simply die.™
In the example here, bluten means something like ˜to shed blood for one™s
country™ and sterben means ˜to die for one™s country™.
A prototype approach to the semantics of the passive, as outlined here brie¬‚y,
is useful for accounting for the interaction of the passive with a range of
predicates. It takes into account not just the basic meaning of a verb, but its
extended meanings as well. It also focuses on the semantic properties of the
arguments that occur with a verb.34


4.6.2 The middle
The term “middle voice” has been used in the linguistic literature to refer to a
wide range of phenomena, from an in¬‚ectional category of the verb in languages
like Greek to re¬‚exive verbs in German like sich verbeugen ˜to bow™ and sich
f¨ rchten ˜to be afraid™.35 We use the term “middle” here to refer to sentences
u
like those in (95b) and (96b). The middle in (95b) is a personal middle, formed
from a transitive verb (a verb with an accusative object). The middle in (96b) is
impersonal; it is formed from a verb that does not have an accusative object.
(95) a. Meine Frau f¨ hrt dieses Rad.
a
˜My wife rides this bike.™
b. Dieses Rad f¨ hrt sich leicht.
a
this bike rides re¬‚ easily ˜This bike rides easily.™
(96) a. Sie singt im Chor.
˜She sings in the choir.™
b. Es singt sich leichter im Chor.
it sings re¬‚ more-easily in-the choir
˜It™s easier to sing in the choir.™
Both types of middles contain a re¬‚exive pronoun. The subject of a personal
middle corresponds to the (accusative) object of its active counterpart; compare
(95a) and (95b). The subject of an impersonal middle is es ˜it™. It is a syntactic
subject, not a place-holder, since it can occur in other than sentence-initial
position.
176 German

(97) Hier sitzt es sich bequem.
˜You can sit here comfortably.™

Both personal and impersonal middles describe a property of some entity.
The middle in (95b) describes a property of a particular bicycle; the middle in
(96b) describes a property of a location (in the choir). Because they describe
properties, middles are stative rather than eventive. Both types of middle also
have an understood agent. A common paraphrase of the middle is an active
clause with man ˜one™ as subject, an overt realization of this understood agent.
The middle in (96b), for example, can be paraphrased as in (98). Notice that
this paraphrase contains the modal verb kann ˜can™, re¬‚ecting the modal notion
of ability or possibility inherent in middles.
(98) Man kann leichter im Chor singen.
˜One can sing more easily in the choir.™
The understood agent in a middle, unlike its passive counterpart, cannot
appear overtly.

Dieses Rad f¨ hrt sich von {einem/den meisten Fahrern} leicht.
a
(99)
˜This bike rides by {one/most riders} easily.™
In general, middles require an adverbial modi¬er. This holds for middles in
English as well as those in German.

(100) This book reads.
(101) a. — Das Buch liest sich.
˜The book reads.™
b. — Es singt sich im Chor.
˜It sings in the choir.™
Goldberg and Ackerman (2001) account for obligatory adverbials in middles
by appealing to pragmatics.36 Without an adverbial modi¬er, middles like those
in (100) and (101) are anomalous because they are uninformative. A middle
that asserts that a book can be read, for example, conveys no new information,
because it is a given that books can be read. A middle that claims that a book
can be read easily, on the other hand, is acceptable, because it conveys new
information. As Goldberg and Ackerman point out, their pragmatic explanation
for obligatory adverbials allows them to account for middles that are acceptable
without such expressions.
(102) It snaps/It zips/ It buttons. (Goldberg and Ackerman 2001:806)
If uttered in response to a question like How do you close this purse?, the
middles in (102) are acceptable. They are informative in this context because
Semantics 177

the predicates themselves (snap, zip, button) are informative, and no adverbial
is necessary.37



Exercises
1. Are the following word pairs examples of synonymy, complementary oppo-
sites, gradable antonyms, converses, hyponymy, or meronymy?
(a) Eigelb, H¨ hnerei; (b) Verb, Zeitwort; (c) vollj¨ hrig, minderj¨ hrig;
u a a
¨
(d) Hund, Schnauzer; (e) Verk¨ ufer, Kunde; (f) uber, unter; (g) gr¨ ßer,
a o
kleiner; (h) Tomate, Paradeiser; (i) laut, leise; (j) Brille, Brillenglas
2. What do the present tense forms express in the following examples? Do they
describe an event that occurs at the moment of speaking, before the moment
of speaking (the historical present), or after the moment of speaking “ or do
they express habitual events or timeless truths?
(a) Sie liest jeden Abend zwei Stunden vor dem Einschlafen.
(b) Wir essen gleich.
(c) Waschb¨ ren sind S¨ ugetiere.
a a
(d) Die Maschine landet in einer Stunde.
(e) Ich sitze gestern in der Mensa, als sich drei Personen an meinen Tisch
setzen.
(f) Wir sind gerade beim Abendessen.
(g) Der Rhein ist der gr¨ ßte Nordseezu¬‚uss.
o
(h) Ich rufe dich heute Abend an.
3. Present perfect or past? Translate the following sentences into English.
¨
(a) Ich bin schon zweimal nach Osterreich ge¬‚ogen.
(b) Letztes Jahr bin ich zweimal unter 2:09 gelaufen.
(c) Noch nie habe ich ihn ohne seine Einkaufst¨ te gesehen.
u
(d) Sie hat nie geraucht.
(e) Das habe ich dir schon hundertmal gesagt.
(f) Sein sch¨ nes B¨ ro hat er nur einmal benutzt.
o u
(g) Ich habe schon oft vor Gericht gelogen.
(h) Wir sind auch einmal jung gewesen.
4. Do the modals in the following sentences express epistemic modality, root
modality, or do they express evidential meaning?
(a) Ich kann nicht verstehen, warum sie so reagiert hat.
(b) Ich soll dieses Motorrad angeblich vor 6 Tagen gekauft haben.
¨
(c) Er durfte dar¨ ber etwas verstimmt sein.
u
(d) Der Baum k¨ nnte krank gewesen sein und daher morsch.
o
(e) Es wird ihr langsam klar, dass es Sabotage gewesen sein muss.
(f) Die Pressefreiheit darf nicht missbraucht werden.
178 German

(g) Wir sollten nicht vergessen, dass ein langer und schwieriger Weg vor
uns liegt.
(h) Er hat f¨ r die Tatzeit ein Alibi: W¨ hrend des Vorfalls will er mit einem
u a
Freund in einer Diskothek gewesen sein.
5. Identify the thematic roles of the NPs in bold in the following sentences
(agent, patient, theme, experiencer, instrument, location,
goal, source, percept).
(a) Ich sp¨ re den Rauch in meiner Lunge.
u
(b) Die Katze sprang von meinem Arm direkt auf Annas.
(c) Das Kind hat ihn mit einem Stock geschlagen.
(d) Ein Frosch sitzt auf einem Stein.
(e) Wir sehen einen Schatten in der Dunkelheit.
(f) In Italien isst man Nudeln mit einer Gabel.
(g) Das Gep¨ ck wird vom Hotel direkt zum Bahnhof gebracht.
a
(h) Die Schokolade zerschmilzt auf meiner Zunge.

notes
1 Philipp von Zesen (1619“1689) was a member of die Fruchtbringende Gesellschaft,
the ¬rst and most famous of the seventeenth-century German language societies that
were dedicated to the preservation and puri¬cation of the German language. See
chapter 5 (section 5.5) for further discussion of these language societies.
2 If heavier meant ˜heavy to a greater degree™, then (2b) would mean ˜This box is
light, but it™s heavy to a greater degree than that one™, in which case it would be as
marginally acceptable as (2a).
3 The examples in (2) through (6) are from Cruse (1986:207).
4 Durrell notes (1988:236) that his tests with native informants con¬rm Cruse™s claim
that gut and schlecht are polar antonyms, but that Bierwisch™s analyses (1967,
1987:230“237) suggest that gut and schlecht are of the overlapping type, like English
good and bad.
5 Gaps in lexical taxonomies are, in general, not uncommon.
6 The term R¨ de refers to male dogs, wolves, foxes, and members of the marten
u
family.
7 There is likely a correlation between these hyponymic facts and grammatical gen-
der. For example, the superordinate that has only a uniquely female hyponym is
grammatically masculine (der Hund); the superordinate that has only a uniquely
male hyponym is grammatically feminine (die Katze).
8 For additional examples of differences between German and English in the area of
lexical semantics, see, for example, Leisi 1967, Hawkins 1986, and Durrell 1988.
9 For further discussion of tense and aspect, see, for example, Comrie 1976, 1985.
10 Schipporeit and Strothmann (1970) classify phrases with seit as up-to-now (UTN)
phrases, phrases that refer to periods of time that begin somewhere in the past and
reach up to the moment of speaking.
11 As mentioned in chapter 2 (section 2.2.4.2), the present perfect is formed using
the past participle and haben ˜to have™ or sein ˜to be™ (in the present tense) as an
Semantics 179

auxiliary. The choice between haben and sein is governed by a verb™s transitivity
and semantics. If a verb is intransitive (if it does not have an accusative object) and
it expresses a change of location or state, it will select the auxiliary sein; otherwise,
it will select haben. For example, intransitive fahren ˜to drive™ expresses a change
of location and thus selects the auxiliary sein: Er ist nach Berlin gefahren ˜He has
driven to Berlin.™ Transitive fahren ˜to drive (someone or something)™, on the other
hand, selects haben: Er hat den Wagen nach Berlin gefahren ˜He has driven the car
to Berlin.™
12 Reichenbach (1947:290) uses diagrams like these to characterize the difference
between the past and the present perfect in English. They also apply to Vater™s
characterization of the past and present perfect in German.
13 This example is based on one in Trier 1965:18.
14 The future is formed with a present tense form of the auxiliary werden ˜to become™
and the in¬nitive of the main verb. The future perfect is formed with a present tense
form of werden and the perfect in¬nitive of the main verb. The perfect in¬nitive is
formed using the past participle of the main verb followed by the in¬nitive of the
auxiliary it requires in the perfect tenses, haben or sein.
15 These and the remaining German examples in this section, but not the glosses or
translations, are from Vater 1994:74“75, unless indicated otherwise.
16 The past perfect is formed with a past form of haben ˜to have™ or sein ˜to be™ as an
auxiliary and the past participle of a verb: hatte gekauft ˜had bought™; war gegangen
˜had gone™.
17 Reference time does not have to be stated explicitly; it can be implicit, as in the
following example (Vater 1994:71): Ich hatte getr¨ umt ˜I had dreamed.™
a
18 See Weinrich 1964 and Weinrich 1993:198“222 for further discussion of die erz¨ hlte a
Welt and die besprochene Welt.
19 For further discussion of the double present perfect and double past perfect,
see Breuer and Dorow 1996, Litvinov and Radˇ enko 1998, and Dudenredaktion
c
2005:520“521.
20 The English translations of these sentences are also unacceptable and provide further
evidence for the incompatibility of states with the progressive.
21 When modality distinctions are marked on the verb, these distinctions are tradition-
ally called moods (Lyons 1995:327). See chapter 2, section 2.2.4.3 for a discussion
of the three moods in German: the indicative, the subjunctive, and the imperative.
22 According to Durrell (2002:302), a sentence with epistemic d¨ rfen means much the
u
same as one with the “future” (werden + in¬nitive) and wohl ˜probably™.
23 See Fagan 2001 for an explanation for why the default form of epistemic modals is
essentially the present indicative.
24 See Fagan 1996 for an explanation for the correlation between epistemic modality
and stative in¬nitives.
25 Verbs that are accomplishments (make a cake) or achievements (win the race) in
Vendler™s (1967) classi¬cation are telic.
26 See Fagan 2001:225“226 for further discussion.
27 There are many terms that are used to refer to these semantic roles: for example,
thematic relations (Gruber 1965, Jackendoff 1972), participant roles (Allan 1986),
thematic roles (Dowty 1986), and semantic roles (G´von 1990). We use the term
±
“thematic role,” since it is widely used in the literature. There is also a certain
180 German

amount of variation in the terms for, de¬nitions of, and number of thematic roles
that have been identi¬ed. The thematic roles presented here largely those proposed
by Jackendoff (1972, 1987).
28 Foley and Van Valin as well as Rappaport and Levin have somewhat different labels
for the thematic roles that are identi¬ed here as goal and source.
29 The examples in (78) and (79) are from Hawkins 1986:58“59. Many examples in
Hawkins™s discussion of the semantic range of subjects in German and English are
from Rohdenburg 1974.
30 See K¨ nig and Nickel 1970, Rohdenburg 1974, and Hawkins 1986:57“61 for further
o
discussion and examples.
31 The passive in German has essentially the same range of tenses and moods as the
active voice; these are indicated by the different forms of the passive auxiliary. See
Durrell 2002:308 for details.
32 If a verb has a dative object (e.g., helfen ˜to help™), this object stays in the dative
case in the passive, and the verb is in its “default” form, the third person singular:
Den Kindern (dative plural) wird (third person singular) geholfen, traumatische
¨
Erfahrungen zu uberwinden ˜The children are being helped to overcome traumatic
experiences.™
33 See Hopper and Thompson 1980 for a discussion of the properties that contribute
to the transitivity of a clause.
34 For further discussion of restrictions on the werden passive in German, see, for exam-
ple, Shannon 1987, Zifonun et al. 1997:1796“1808, and Dudenredaktion 2002:553“
554.
35 Verbs like sich verbeugen ˜to bow™ and sich f¨ rchten ˜to be afraid™ are inherently
u
re¬‚exive: they must and can only be used with a re¬‚exive pronoun. Compare Er
verbeugte sich ˜He bowed™ with — Er verbeugte ihn ˜He bowed him.™ This use of the
re¬‚exive, while widespread in German, is not common in English.
36 Fagan (1992) argues that adverbial phrases in German and English middles are
required by the rule of Middle Formation. Goldberg and Ackerman™s (2001) account
provides an explanation for the presence of these expressions.
37 For further discussion of middle constructions in German, as well as additional
references, see Fagan 1992.
History of the language
5




All languages change over time, phonologically, morphologically, lexically,
syntactically, and semantically. The example in (1a), a form of German from
the ninth century (Old High German), differs signi¬cantly from its modern
German glosses and its modern German translation in (1b).
(1) a. Fater uns¯ r, thu in himilom bist, giuu¯hit s¯ namo th¯n. (Braune et al.
e ± ± ±
1994:34)
Vater unser du in Himmel bist geweiht sei Name dein
b. Vater unser, der du bist im Himmel. Geheiligt werde dein Name.
(Waterman 1991:81)
father our who you are in-the heaven sacred become your name
˜Our Father who art in heaven, hallowed be thy Name.™
We see differences in pronunciation (thu/du),1 differences in morphology
(himil-om/Himmel), lexical/semantic differences (giuu¯hit/geheiligt), and dif-
±
2
ferences in word order (namo th¯n/dein Name). The various changes that
±
German has undergone throughout its history can be used to identify the major
periods of the language that precede the modern period: Old High German
(750“1050); Middle High German (1050“1350); Early New High German
(1350“1650); and New High German (1650“1900). In this chapter we will
look at the important phonological, morphological, and syntactic characteris-
tics of three of these periods, Old High German, Middle High German, and
Early New High German, and brie¬‚y describe developments in the New High
German period. We will begin with a discussion of the prehistoric period and
provide a brief description of German™s ancestors, Proto-Indo-European and
Germanic.

5.1 The prehistoric period

5.1.1 Proto-Indo-European
The languages of the world can be classi¬ed genetically, that is, according to
their development from common ancestors. The ancestor can be an attested
181
182 German

Table 5.1 PIE stops (from Fortson 2004:51)

labial dental palatal velar labiovelar

ˆ
k
voiceless p t k k
ˆ
voiced b d
ˆh
voiced aspirated bh dh h h




language (a language with extant texts); or it can be a reconstructed proto-
language for which no texts exist. For example, the ancestor of Spanish and
French is Latin, a language that is attested. The ancestor of English and German,
Germanic, is not attested.3 Latin and Germanic are themselves genetically
related; the ancestor they have in common is Proto-Indo-European (PIE).
PIE is the parent language of the family of languages known as Indo-
European (IE), which are spoken in the area extending from (and including)
Europe in the west up to the southern portion of the Indian subcontinent in
the east. The branches of PIE that still have living descendents are Celtic,
Germanic, Italic, Albanian, Greek, Balto-Slavic, Armenian, and Indo-Iranian.
Two branches that are now extinct are Anatolian (which includes Hittite) and
Tocharian.
A salient characteristic of PIE phonology is its wealth of stops. Table 5.1
lists the ¬fteen stops reconstructed for PIE.4 In contrast, PIE has a single
fricative, — s.5 PIE also had the liquids — l and — r, and the nasals — n and — m.6
PIE vowels include short — i, — e, — u, — o, and — a, and long — e, — o, and — a.
¯¯ ¯
Although the inventory of PIE sounds is more extensive than presented here,
this simpli¬ed picture is suf¬cient for tracing the development of PIE sounds in
Germanic.
Stress (accent) is an important feature of PIE phonology that ultimately plays
an important/de¬ning role in the history of German. As noted in chapter 1, a
stressed syllable is one that is perceived as more prominent than neighboring
syllables. A language that uses greater volume as the primary signal of promi-
nence is a stress-accent language. One that primarily uses a difference in pitch
to signal prominence is a pitch-accent language. It is generally agreed that
PIE was a pitch-accent language. Vedic Sanskrit and Ancient Greek, for exam-
ple, are among the IE languages with pitch-accent systems (Fortson 2004:62).
Another feature of the accent in PIE is its mobility; the position of the accent
was not predictable, and it could be used to signal differences in meaning.
Compare Russian pis´ l ˜he was writing™ with p´sal ˜he was peeing™ (Fortson
a ±
2004:62).
PIE was a highly in¬‚ected language; words typically consisted of a root
(which carried the basic meaning of the word) plus one or more (derivational)
History of the language 183

suf¬xes plus an (in¬‚ectional) ending. For example, the word — mn9-te!i-s ˜of
thought™ was made up of the root — men- ˜think™, followed by the suf¬x — -t(e)i-,
used to form abstract nouns, followed by the genitive case ending — -s ˜of™
(Fortson 2004:77). Nouns were in¬‚ected for number (singular, dual, plural),
case (eight), and gender (masculine, feminine, neuter).7 Verbs were in¬‚ected
for person (¬rst, second, third), number (singular, dual, plural), tense (present,
imperfect, aorist),8 voice (active, middle), and mood (indicative, imperative,
subjunctive, optative).9 The various grammatical categories of the verb yielded
many different in¬‚ected forms for each verb “ a much more complicated
system of verbal in¬‚ection than we ¬nd in modern English and even in modern
German.
PIE is generally held to be an SOV (subject“object“verb) language, with
verb-¬nal the “default” in most of the earlier IE languages (Fortson 2004:142;
Watkins 1998:68). In Hittite, for example, the verb is clause-¬nal; the only
exception to this basic order is when the verb is moved to clause-initial position
by the process of topicalization (Fortson 2004:142, 144). Various movement
processes have the effect of altering the basic order of elements in a clause.
Topicalization is used to front a constituent “ not just a verb “ for purposes
of emphasis or contrast.10 (Fronting is the movement of an element to the
beginning of a sentence.) Another type of fronting, wh-movement, moves ele-
ments like interrogatives (˜who™, ˜what™), relative pronouns, and subordinating
conjunctions to the complementizer position. In the older IE languages, the
landing site for topicalization was to the left of the complementizer position
(Hale 1987). We see this in the following example (from Fortson 2004:145),
where the topicalized element is underlined and the complementizer is
boldfaced:

(2) Latin
f¯ st¯ di¯ s¯ quid prod¯ geris
eo e± e
˜if you splurge a bit on a holiday™

An important feature of the syntax of older IE languages, discovered by
Jacob Wackernagel (1892), is known as Wackernagel™s Law, the tendency of
clitics to appear in the “second position” in a clause.11 In the following Gothic
example (from Fortson 2004:146), the clitic is underlined.

(3) Gothic
fram-uh þamma sokida Peilatus fraletan ina
at-and this sought Pilate release him
˜And at this Pilate sought to release him.™
Hale (1987) shows that several independent processes lead to clitics appearing
in the “Wackernagel position.”12
184 German

Table 5.2 Grimm™s Law

Stage I Stage II Stage III
—p —b — bh
>f >p >b
—t —d — dh
>þ >t >d
— k > x (> h) —g > k — gh > g




5.1.2 Germanic
5.1.2.1 Introduction The Germanic languages are traditionally
divided into three branches: East Germanic (now extinct), made up of Gothic
(Goth.) and the languages of the Vandals and Burgundians; North Germanic,
made up of Old Norse, the ancestor of the modern Scandinavian languages,
Icelandic, Faroese, Norwegian, Danish, and Swedish; and West Germanic,
made up of Old High German, Old Saxon (ancestor of Low German), Old Low
Franconian (earliest attestation of what is now the Dutch linguistic area), Old
Frisian, and Old English (OE).
The earliest speakers of Germanic lived in northern Europe during the ¬rst
half of the ¬rst millennium BC, mainly in southern Scandinavia and along the
shores of the North and Baltic seas. By 500 BC they had migrated beyond the
Rhine in the west, to the Vistula River in the east, and had occupied the Low
German plain in the south (Ramat 1998).
The Germanic tribes came into contact with speakers of various languages.
Germanic (Gmc.) borrowings in Russian point to contact with Balto-Slavic
in the east: Russian xleb ˜bread™ < Gmc. — hlaibaz (> OE hl¯ f > Eng. loaf).13
a
Germanic borrowings in Finnish, a non-IE language, are particularly intriguing,
since the modern forms of these words are very similar to forms that have been
reconstructed for Germanic: Finnish kuningas ˜king™ < Gmc. — kuningaz (> Old
Saxon cuning, OHG kuning ˜king™). Borrowings like these provide evidence for
syllables that have been lost in the daughter languages because of an important
innovation in the phonology of Germanic, the accent shift. In the following
section we will look at this as well as other important features of the phonology
of Germanic.

5.1.2.2 Phonology Grimm™s Law is the name given to one of the
major innovations in the phonology of Germanic that sets it apart from the
other branches of IE.14 This sound change, also known as the First Sound
Shift (Die erste Lautverschiebung), involves a shift (or a series of shifts) in the
articulation of the PIE stops. In Germanic, the voiceless stops are changed to
voiceless fricatives, the voiced stops become voiceless stops, and then voiced
aspirated stops become plain voiced stops,15 as shown in Table 5.2.16 The forms
in (4) through (6) illustrate these changes.
History of the language 185

(4) Stage I

p>f Lat. pater Eng. father

t>þ Lat. tr¯ s
e Eng. three

k > x (> h) Lat. corn¯ u Eng. horn

(5) Stage II

b>p Lat. labium Eng. lip

d>t Lat. duo Eng. two

g>k Lat. ego OE ic (= Eng. I)

Stage III17
(6)

bh > b Sk. bhar- Eng. bear

dh > d Sk. dh¯ - ˜put™
a Eng. do

gh > g Sk. magh´ m ˜wealth™
a German mag ˜am able™

The Germanic re¬‚exes of the PIE voiced aspirated stops are treated as voiced
stop phonemes with two types of allophones. In initial position, these phonemes
are realized as simple voiced stops; intervocalically they appear as voiced frica-
tives. We see this in the English word give (< OE giefan), which derives from
the PIE root — ghabh- (Pokorny 1959“1969:407“408). The voiced fricatives
became voiced stops in many of the Germanic languages (German geben ˜to
give™, for example, corresponds to Eng. give).
There were various exceptions to Grimm™s Law. For example, PIE voiceless
stops did not shift if they followed an s. Compare Lat. stella ˜star™ with Eng.
star, which retains an unshifted PIE — t following — s.18 Another set of exceptions,
also involving the PIE voiceless stops, was explained by the Danish scholar
Karl Verner. Verner discovered that the PIE accent was responsible for the
voiced fricatives that sometimes appeared instead of the expected voiceless
fricatives f, þ, h (< PIE — p, — t, — k). According to Verner™s Law, if PIE — p, — t, and

k occurred word-internally and were not immediately preceded by a stressed
syllable, they became voiced fricatives, not the voiceless fricatives one would
expect given Grimm™s Law. We see this in (7a), where the syllable preceding
PIE — t is not stressed. In (7b), on the other hand, the syllable before PIE — t
is stressed, Verner™s Law does not apply, and we ¬nd the voiceless fricative
predicted by Grimm™s Law.19
PIE — ph2 ter > Gmc. — fad´ r ˜father™
´
¯
(7) a. ¯e
PIE — bhrater- —¯
> Gmc. broþer- ˜brother™
´ ´
¯
b.

Verner™s Law also affected PIE — s, which became — z when in word-internal
ˆo
position and not preceded by a stressed syllable: PIE kas´ n > Gmc. haz´ n o
˜hare™ > German Hase (Fortson 2004:303).
Because stress in PIE and early Germanic was mobile and could appear on
different syllables in different forms of a given word, Verner™s Law resulted in
186 German

alternations between voiced and voiceless fricatives within a single paradigm.
Jacob Grimm, who noticed these alternations but could not explain them,
termed them “Grammatischer Wechsel.” Although some of these alternations
have been leveled out over time, we still ¬nd examples, particularly in the forms
of strong verbs and words derived from these forms. Because of subsequent
sound changes, these alternations appear in Modern German as alternations
between f and b; d and t; h and g; and s and r:20
(8) a. Hefe ˜yeast™ heben ˜to lift™
b. schneiden ˜to cut™ geschnitten ˜cut™
c. ziehen ˜to pull™ zog ˜pulled™
d. Verlust ˜loss™ verloren ˜lost™
Various changes in the PIE vowels took place before or in Germanic. Two
signi¬cant changes were the merger of short — a and — o to a and the merger of
long — a and — o to o:21
¯ ¯¯
ˆo
PIE — okt¯ (u) > German acht ˜eight™
(9) a.
PIE — bhr¯ t¯ r > Goth. broþar ˜brother™22
b. ae
After the changes brought about by Verner™s Law had taken place, the accent,
which was mobile in PIE and early Germanic, became ¬xed on the initial
syllable of the root. Certain pre¬xes remained unstressed (e.g., be"kommen ˜to
receive™, ge"b¨ ren ˜to bear™, ver"geben ˜to forgive™). In addition, the pitch accent
a
found in PIE had evolved into stress accent: prominent syllables in Germanic
were signaled not by a difference in pitch, but by greater muscular energy used
in their production.

5.1.2.3 Morphology and syntax The morphology of Germanic can
be characterized as much simpler than that of PIE. In the declension of
nouns, the eight cases found in PIE are reduced to six: nominative, vocative,23
accusative, dative, genitive, and instrumental (Fortson 2004:304).24 The three
numbers (singular, dual, plural) are reduced to two in nouns: singular, plu-
ral (the dual is retained only in pronouns and verbs). The three grammatical
genders of PIE (masculine, feminine, neuter), however, are preserved.
The Germanic verbal system is also much simpler than the verbal system in
PIE. Only the present and perfect stems remain; the perfect becomes a simple
past tense called the preterite (= the past of Modern German). The optative
becomes the Germanic subjunctive. The middle survives as the passive only in
Gothic.25
An important feature of the verbal system in Germanic is the distinction
between the “strong” and “weak” conjugations.26 The strong conjugation con-
tinues and extends the IE system of ablaut (alternation of the root vowel) to make
past-tense forms (the preterite and past participle): trinken ˜to drink™, trank
˜drank™, getrunken ˜drunk™. The weak conjugation, an innovation in Germanic,
History of the language 187

uses a dental suf¬x for past tense forms: danken ˜to thank™, dankte ˜thanked™,
gedankt ˜thanked™.27
Germanic, like its ancestor PIE, is generally considered to be an SOV lan-
guage. Because it is still a relatively heavily in¬‚ected language, sentence con-
stituents are to a great extent free syntactically and can be used to convey prag-
matic information. There is thus a good deal of variation in surface (actual)
word order, as demonstrated by the oldest runic inscriptions (Ramat 1998:386,
410“411):28
(10) a. SOV
ek hlewagastiz holtijaz horna tawido
I Hlewagastiz of-Holt horn made
b. SVO
haþuwolafa sate staba þria
Haþ. set staves three
c. OVS
haidr runoronu falah-ak
bright runes-sequence hid-I
Other features, however, suggest an underlying SOV order for Germanic. For
example, we ¬nd the Determiner + Determined order that is characteristic of
OV languages (Ramat 1998:411):
(11) a. genitive + noun
hariwulfs stainaz (Runic inscription)
Hariwulf™s stones
b. adjective + noun
afar ni managans dagans (Gothic)
after not many days
Although Germanic is still a highly synthetic (in¬‚ecting) language, we see
the beginnings of features of analytic languages.29 For example, meanings
that were expressed with case in¬‚ections in PIE began to be expressed by
prepositions when these cases were lost. Some of the prepositions that can be
attributed to Germanic are the following (Ramat 1998:408): — frama ˜from™ (>
Goth. fram), — med(i) ˜with™ (> Goth. miþ), — under ˜under™ (Goth. undar), — to,
¯

ta ˜towards™ (> OE t¯ ), — ut ˜out of™ (> Goth. ut).
o¯ ¯


5.2 Old High German

5.2.1 Introduction
Old High German (OHG), the earliest stage of the German language for which
there are extensive written documents, spans the period from roughly 750 to
1050 AD. The term “High” identi¬es the area in which much of the language
188 German

Table 5.3 The High German Consonant Shift

West Germanic English Old High German
— p- path pf pfad
— t- tide ts z¯t (= Zeit ˜time™)
±
— k- corn kx chorn
— pp apple pf aphul
— tt set (OE settan) ts setzan
— kk wake (OE weccean) kx wechen
— -p ship f scif
— -t foot s fuoz
— -k book x buoch



was spoken, the mountainous region in the southern portion of the territory
occupied by the West-Germanic languages, and distinguishes it from the “Low
German” spoken in the lowlands in the north. There was no standard or suprare-
gional variety of OHG; OHG consisted rather of a number of different dialects.
There was also no literary standard that united the dialects. Most of the OHG lit-
erature was produced in monasteries and consisted primarily of religious texts,
many of which were translations from Latin. The oldest surviving “book” in
German is Der Abrogans (c. 765), a Latin“OHG synonym dictionary, named
after the ¬rst entry, Lat. abrogans (OHG dheomodi ˜humble™). Other transla-
tions of Latin texts include the Lord™s Prayer, the Rule of St. Benedict, the
Psalms, and the Gospels. Notker Labeo (c. 950“1022), a Benedictine monk,
translated works by Boethius and Aristotle, among others. Some OHG works
are original compositions, for example, Otfrid von Weissenburg™s Evangelien-
buch (863“871), a gospel harmony in rhyming couplets. There is very little
extant non-Christian original OHG material. One important work is the Hilde-
brandslied, a fragment (sixty-eight lines) of an epic poem in alliterative verse in
an unusual mixture of OHG and Old Saxon, found on the ¬rst and last leaves of
a manuscript dating from the early part of the ninth century. The Merseburger
Zauberspr¨ che (magic spells), generally held to be of pagan origin, were dis-
u
covered in a religious manuscript in Fulda from the ninth or tenth century.


5.2.2 Phonology
The most important sound change that distinguishes OHG from all other West
Germanic languages is known as the Second Sound Shift (Die zweite Lautver-
schiebung) or the High German Consonant Shift. In OHG, the West Germanic
voiceless stops — p — t — k became the affricates pf ts kx when in initial position,
medially following a consonant, or geminated (— pp — tt — kk); elsewhere they
became the fricatives f s x. Table 5.3 provides examples from English and OHG
History of the language 189

that illustrate this sound change. The English words contain unshifted p t k; their
OHG cognates contain the affricates and fricatives that resulted from the sound
shift.30 Because there was no standard orthography, the sounds that resulted
from the Old High German Consonant Shift were not represented consistently.
Examples (12) and (13) illustrate the typical ways in which the affricates and
fricatives resulting from — p — t — k were represented orthographically in OHG
texts.
(12) a. pf <pf> or <ph>
b. ts <z> or <tz>
c. kx <kh> or <ch>
(13) a. f <ff> or <f>
b. s <zz> or <z>
c. x <hh> or <ch> (in later texts)

Notice that the orthography was often ambiguous: <z> could represent the
affricate ts as well as the fricative s; <ch> could represent the affricate kx or
the fricative x.
One consistent exception to the High German Consonant Shift involved the
sound — s. West Germanic — p — t — k were not shifted when they followed — s.31
(14) a. OHG spinnan ˜to spin™
b. OHG gast ˜guest™
c. OHG scuoh ˜shoe™
The High German Consonant Shift began in the southern portion of the
German-speaking territory and eventually petered out by the time it reached
the Low German area. We in fact ¬nd the affricate kx only in the south, in
the Alemannic (Swiss German) area; elsewhere West Germanic — k (initial,
post-consonantal, and geminate) remains unshifted. In OHG dialects that have
unshifted — k we ¬nd, for example, korn ˜grain™ and wecken ˜to wake™ instead of
chorn and wechen. In OHG dialects closer to the Low German area we also ¬nd
unshifted— p- and — t-. In Rhenish Franconian (Frankfurt, Mainz), for example,
West Germanic — p- frequently remains unshifted (Waterman 1991:58). As we
will see in the discussion of Modern German dialects in chapter 6, the extent
to which a given dialect participated in the High German Consonant Shift
determines whether the dialect is classi¬ed as being Upper, Central, or Low
German.
Two additional sound changes that are characteristic of OHG are the change
from — d to t and the change from — þ to d.

(15) a. d>t Eng. deep OHG tiof (= tief)

þ>d
b. Eng. thing OHG ding
190 German

These changes are usually viewed as a chain reaction. After — t had become
ts and s (as a result of the High German Consonant Shift), — d became t,
allowing — þ to become d. The change from — þ to d affected Dutch (Dutch
ding ˜thing™) and eventually spread to Low German, although Old Saxon,
the ancestor of Low German, did not yet show the change (Old Saxon thing
˜thing™).
Two important features of the OHG vowel system are umlaut and the reten-
tion of full vowels in unaccented syllables. Prior to the OHG period, West
Germanic — e had been raised to i preceding an i or j (a non-syllabic i) in
a following syllable. This change, known as West Germanic i-Umlaut, is an
example of assimilation, a sound change where one sound acquires one or more
features of a neighboring sound. In this example, the mid vowel e becomes the
high vowel i (it acquires the feature [+high]) when it precedes the high vowel
i. We see this change in the third person singular form of the present tense of
strong verbs like OHG geban ˜to give™. We also see it in etymologically related
words.

(16) West Germanic i-Umlaut
OHG geban ˜to give™, er gibit ˜he gives™ (< — gebit)
a.
OHG erda ˜earth™, irdisc ˜earthly™ (< — erdisk)
b.

Primary Umlaut, the change from (short) a to e before an i or j in a following
syllable, took place at the beginning of the OHG period.32 We see this change
in the present tense of strong verbs, in comparative forms of adjectives, and in
the plurals of nouns.

(17) Primary Umlaut
OHG faran ˜to drive™, er ferit ˜he drives™ (< — farit)
a.
OHG lang ˜long™, lengiro ˜longer™ (<— langiro)
b.
OHG gast ˜guest™, gesti ˜guests™ (< — gasti)
c.

In Modern German, Primary Umlaut is typically represented orthographically
by <¨ >.
a
West Germanic i-Umlaut and Primary Umlaut, which were both a type of
i-umlaut (i in¬‚uenced a vowel in a preceding syllable), yielded sounds that
already existed in the language, i and e. Secondary Umlaut, also a type of
i-umlaut, yielded sounds that were new to the language. It is assumed that
Secondary Umlaut took place during the OHG period, but was not indicated
orthographically with regularity until later, during the Middle High German
(MHG) period, because of the dif¬culty of representing these sounds with the
letters available in the Latin alphabet. Secondary Umlaut is illustrated here
using MHG words with the “normalized” orthography that is typically used for
MHG texts.
History of the language 191

(18) Secondary Umlaut
¯
a. a>¦ OHG m¯ ri
a MHG m¦re ˜news™
¨
b. o>o OHG mohti MHG m¨ hte ˜would like™
o
¯
c. o>“ OHG sc¯ ni
o MHG sch“ne ˜beautiful™
¨
¨
d. u>u OHG ubir MHG uber ˜over™
¯
e. u > iu [y¦] OHG h¯ sir
u MHG hiuser ˜houses™

The umlaut symbol used today to indicate umlauted vowels originated in the
MHG practice of writing <e> or <i> over the vowels <a>, <o>, and <u>.
The vocalism of unaccented syllables in OHG is another important feature
of the language “ one in particular that distinguishes it from the language of the
MHG period. In OHG, unaccented syllables still retain full vowels. Although
some syllables have been lost because of the Germanic accent shift (Gmc.

kuningaz > OHG kuning ˜king™), the vowels in the remaining unaccented
syllables have not yet been reduced to schwa.33 Compare the endings in the
following OHG words with those of their MHG counterparts (where <e> in
word-¬nal syllables represents [™]).
(19) a. OHG hirti MHG hirte ˜shepherd™
b. OHG hab¯ n
e MHG haben ˜to have™
c. OHG zunga MHG zunge ˜tongue™
d. OHG boto MHG bote ˜messenger™
e. OHG (ih) nimu MHG (ich) nime ˜(I) take™
As we will see in the next section, the presence of full vowels in unaccented
syllables in the OHG period is of signi¬cance for the morphology and syntax
of the language.


5.2.3 Morphology and syntax
Because the vowels of in¬‚ectional endings are still rather strongly differenti-
ated in the OHG period, grammatical distinctions can still be made to a large
extent synthetically (through in¬‚ection rather than through the use of indepen-
dent words). There are some changes in the in¬‚ection of nouns and pronouns,
however. Of the six cases found in Germanic, essentially only four remain:
nominative, accusative, dative, and genitive. The function of the vocative has
been taken over by the nominative. Vestiges of the instrumental remain, but
only in early OHG in the strong nouns and pronouns: OHG tagu ˜day™ (mascu-
line singular instrumental); OHG diu ˜the™ (masculine singular instrumental).
Unlike Gothic, OHG no longer distinguishes between the dual and plural in the
pronouns.34
In addition to these changes in nominal in¬‚ection, we also see some changes
in verbal in¬‚ection. In OHG verbs, the two tenses of Germanic remain: the
192 German

present, which can also be used to express the future, and the preterite, which is
the general tense for the past.35 The three moods of Germanic remain: active,
subjunctive, and imperative. Passive voice, however, which is only partially
retained in Gothic, can no longer be found in OHG in synthetic form. Passive
meaning is expressed analytically in OHG, with an in¬‚ected form of wesan ˜to
be™ or werdan ˜to become™ and a past participle, for example, ist ginoman or
wirdit ginoman ˜is taken™ (Braune 2004:256).
Although OHG is still largely a synthetic language, we see further evidence
of movement towards the use of analytic structures. In addition to the expression
of mood (the passive) by analytic means, we also ¬nd the analytic expression of
tense. For example, although the present is generally used to express the future,
there are a few examples of an analytic future, most using the auxiliary sculan
(sollen) ˜shall™, some using wellen (wollen) ˜to intend™ (Braune 2004:256). A
periphrastic perfect (with the auxiliaries hab¯ n ˜to have™, eigan ˜to have™, or
e
wesan ˜to be™) is used in translations of the Latin perfect, although the preterite
is also used for this purpose.
Because many of the extant OHG texts are translations of Latin, it is often
dif¬cult to gain a clear picture of OHG syntax. For example, if the use of
a feature in OHG does not deviate from the use of that feature in the Latin
original, it is impossible to determine whether the feature can be attributed
to OHG or whether it simply re¬‚ects a feature of Latin. Other factors that
contribute to the dif¬culty in determining the characteristics of OHG syntax
include the diverseness of the various German-speaking regions, the large time
span covered by the OHG period, and the possible in¬‚uence of meter and
rhyme on the language of the OHG texts (Schrodt 2004:vii). In spite of these
dif¬culties, there are some generalizations that can be made about the syntactic
characteristics of the language.
With increasing frequency, the subject of a sentence is realized pronominally
even when it can be determined through context or by linguistic means (Meineke
and Schwerdt 2001:312). The eventual loss of distinctive in¬‚ectional endings
on the verb may have contributed to the increase in the use of subject pronouns.
However, we ¬nd examples where subject pronouns are used (possibly for
emphasis) when the Latin original lacks them and when the verbal endings in
OHG are unambiguous (Meineke and Schwerdt 2001:313).

(20) OHG Dhu minn¯ des reht ˜you loved justice™
o
Lat. dilexisti iustitiam

This suggests that the loss of distinct in¬‚ections cannot have been the sole
factor that contributed to the rise in the use of subject pronouns.
OHG continues the development of a de¬nite article from a demonstrative
pronoun. The de¬nite article is particularly common in Otfrid™s writings, and
History of the language 193

most nouns in Notker™s translations from Latin are preceded by a de¬nite article
(Meineke and Schwerdt 2001:314“315).

(21) zu dero uuiti dero euuigheite (Notker, cited in Schrodt 2004:25)
to the vastness of-the eternity

The position of the ¬nite verb in OHG is a particularly dif¬cult topic, given
the potential in¬‚uence of Latin, the effects of meter, and the role of stylis-
tic factors (e.g., extraposition), all of which have to be taken into account.36
Although there is a good deal of variation and many sentences are dif¬cult
to analyze unequivocally, some generalizations about the position of the ¬nite
verb can be made (Schrodt 2004:197“208). Verb-¬nal position, inherited from
IE, has become a feature typical of subordinate clauses (adverbial elements can
be extraposed to the right of the ¬nite verb, however).

(22) a. daz uuir des dikk¯ m (Braune et al. 1994:34)
e
that we for-it ask
b. extraposition: daz zist¯ rit ist durh unsre suntan (Braune et al.
o
1994:81)
that destroyed is through our sins

Verb-second or middle position is an innovation in OHG, as in the other
Germanic languages, and is the norm in declarative sentences.

(23) Anna hiaz ein wib thar (Otfrid, cited in Schrodt 2004:200)
Anna was-called a woman there

Verb-¬rst position, which has the emphatic and discourse-connective function
found in IE, also has the Modern German function of marking imperatives,
interrogatives, and wishes.

(24) a. emphatic: yrforahtun tho thie liuti thio wuntarlichun dati (Otfrid,
cited in Schrodt 2004:198)
feared then the people the marvelous deed
b. interrogative: pechennest tu mih? (Notker, cited in Schrodt
2004:200)
recognize you me

Although the ¬nite verb appears in various positions on the surface (¬rst,
second, last, etc.), we can assume that OHG still has underlying SOV word
order. In spite of Latin in¬‚uence on OHG translations, we ¬nd, for example,
the adjective + noun word order that is characteristic of OV languages, not
the noun + adjective order that is characteristic of Latin. As a rule, attributive
adjectives precede the noun they modify (Schrodt 2004:37).
194 German

(25) fohem uuortum; friuntlaos man (Hildebrandslied, Braune et al.
1994:84)
few words (dative plural); friendless man

5.3 Middle High German

5.3.1 Introduction
Middle High German (MHG) was spoken from roughly 1050 to 1350.37 Dur-
ing the MHG period we see the development of the ¬rst “standard language”
(Gemeinsprache) in the German-speaking area, the literary language of the
court poets during the golden age of chivalric poetry (1170“1250). This lan-
guage, also known as “Classic Middle High German” (das klassische Mittel-
hochdeutsch), is relatively uniform, but less so than would appear from the
modern editions of MHG texts, which employ a normalized orthography.38
The uniformity of this literary language has its origins in the desire of MHG
poets to achieve as broad a circulation of their work as possible. They avoided
words and pronunciations that were perceived as strongly dialectal. They also
made an effort to use words that would rhyme in any dialect. In general, the
manuscripts of the courtly literature exhibit a mixture of Alemannic and East
Franconian phonology (Paul 1998:14).
Whereas the literature of the OHG period was used primarily for reli-
gious/didactic purposes, the literature of the MHG period, especially that of
“Classic MHG,” was intended as entertainment for the knightly class (Ritter-
stand). The main genres of this period are the courtly epic and lyric poetry.
A famous epic, the Nibelungenlied, is the work of an anonymous poet from
roughly 1200. Important MHG authors of epic poetry include Hartmann von
Aue (c. 1168“1210), who wrote Erec and Iwein; Wolfram von Eschenbach (c.
1170“1220), author of Parzival; and Gottfried von Straßburg (c. 1200), author
of Tristan und Isolde. Lyric poetry included the Minnesang ˜minnesong™, the
song of courtly love. Perhaps the greatest lyric poet of the MHG period was
Walther von der Vogelweide (c. 1170“1230), who composed political verse as
well as Minnelieder.
The prestige of French culture was strong during the MHG period, and
many borrowings found their way into the German language during this period
through contact with French speakers. In addition to loanwords like MHG
ˆ
aventiure (Abenteuer) ˜adventure™ (< Old French aventure) and MHG rˆm ±
˜rhyme™ (< Old French rime), we ¬nd loan translations such as MHG hovesch
˜courtly™ (French courtois) and MHG ritter ˜knight™ (French chevalier). French
suf¬xes also found their way into the language. They ¬rst appeared in loanwords
and then became productive and were used to create new lexical items with
History of the language 195

German stems. The suf¬x -ei in words like B¨ ckerei ˜bakery™ and the verbal
a
suf¬x -ieren in words like buchstabieren ˜to spell™ were introduced into the
language in this way (-ei through words like MHG partˆe ˜part™ and -ieren
±
through words like MHG parlieren ˜to speak™).
With the demise of the knightly class came the end of the in¬‚uence of the
literary language on the development of a German standard. During the next
several hundred years, the language of the chanceries, dialectally less neutral
than the language of court poets, played an important role in shaping the
standard language that emerged during the Early New High German period.


5.3.2 Phonology
One of the most salient features of MHG phonology is the weakening of
unstressed syllables. Recall from the discussion in section 5.2.2 that the full
vowels in the unaccented (typically in¬‚ectional) syllables of OHG are replaced
ˆ
in MHG by schwa (<e>). The famous opening lines of the ¬rst aventiure
˜adventure™ of the Nibelungenlied show this clearly.

(26) Das Nibelungenlied (First Adventure)
Uns ist in alten m¦ren wunders vil geseit
von helden lobeb¦ren, von grˆ zer arebeit,
o
von fr¨ uden, hˆ chgezˆten, von weinen und von klagen,
o o ±
von k¨ ener recken strˆten muget ir nu wunder h“ren sagen.
u ±
(de Boor 1988:3)
Wondrous things are told in ancient tales
Of famous men and bold, of great travails,
Of joy and festive life, of woe and tears,
Of warriors met in strife “ the wonder shall ¬ll your ears!
(Ryder 1962:43)

All the in¬‚ectional syllables in this text contain schwa (e.g., the adjective ending
of alten ˜old™ < OHG alt¯ n, the in¬nitive ending on the verb weinen ˜to cry™
e
< OHG wein¯ n, etc.). The weakening of unstressed vowels is presumed to be
o
a result of the accent shift in Germanic. The ¬xing of the accent on the initial
syllable of the root and, in particular, the change from a pitch accent to a stress
accent led to the weakening (and loss) of unstressed syllables, as these were
articulated with less force than accented syllables.
Another important feature of MHG, Secondary Umlaut, can also be seen in
the ¬rst lines of the Nibelungenlied. The words m¦ren, k¨ ener, and h“ren all
u
contain vowels that were fronted by an i in a following syllable. The word m¦re,
¯
for example, was m¯ ri in OHG. After the a assimilated to i and became ¦, the
a
196 German

i was weakened to schwa. As mentioned in section 5.2.2, it is assumed that
Secondary Umlaut took place during the OHG period, but was not represented
orthographically until MHG times. When the conditioning factor for umlaut
¨ ¨
was lost (when i was reduced to schwa), the new sounds, ¦ o “ u iu [y¦],
acquired the status of phonemes. The sound ¦, for example, was no longer an
¯ ¯
allophone of the phoneme a (the realization of a when it occurred before i in a
following syllable); it was a phoneme in its own right.
In OHG, when the consonant clusters ht, hs, and rw occurred between a
and i, or when i appeared two syllables away from a, Primary Umlaut did
not take place. These barriers to umlaut were overcome in MHG, and this later
¨
fronting of short a, represented orthographically with the symbol a, is subsumed
under Secondary Umlaut. In OHG we ¬nd, for example, mahtig ˜powerful™ and
magadi ˜virgins™; in MHG these appear as m¨ htec and m¨ gede.
a a
Two important changes can be seen in the consonants of MHG, Palatalization
and Final Fortition (Auslautverh¨ rtung). Several lines from Walter von der
a
Vogelweide™s “Under der linden” contain examples of both changes.

(27) a. schˆ ne sanc diu nahtegal
o
beautifully sang the nightingale
b. seht wie rˆ t mir ist der munt
o
see how red me is the mouth

Palatalization refers to the development of a new phoneme, /ʃ/, from the
consonant cluster sk. For example, OHG scif ˜ship™ > MHG schif. In the
example in (27a), we ¬nd MHG sch¨ ne ˜beautifully™, which developed from
o
OHG sc¯ ni. We ¬nd this same change in English, where ship corresponds to
o
German Schiff and ¬sh to German Fisch (< OHG ¬sc). In late MHG, s is
palatalized (> /ʃ/) when it occurs before other consonants (l, m, n, p, t, w).

(28) a. OHG slahan ˜to hit™ > schlagen
b. OHG smal ˜small™ > schmal
c. OHG sn¯ ˜snow™ > Schnee
e

This change began in the Upper German dialects in the south and spread north-
ward, but did not reach all of the northern dialects. It is not represented con-
sistently in the orthography of Modern German (spitz ˜pointed™, Stein ˜stone™),
and in the modern colloquial language of northern Germany, speakers retain
the old (unpalatalized) pronunciation before p and t: [s]pitz, [s]tein.
Final Fortition affected b, d, and g, which became p, t, and c when in
syllable-¬nal position. Although Final Fortition is predictable, it is represented
orthographically in MHG texts, as we see in the words sanc ˜sang™ and munt
˜mouth™ in (27). The effects of Final Fortition are particularly clear if we
compare words with and without in¬‚ectional endings.
History of the language 197

Table 5.4 The declension of OHG tag
˜day™ and MHG tac ˜day™

OHG MHG

Singular Nominative tag tac
Accusative tag tac
Dative tage tage
Genitive tages tages
Instrumental tagu
Plural Nominative taga tage
Accusative taga tage
Dative tagum tagen
Genitive tago tage



(29) a. stoubes, stoup ˜dust™
b. kindes, kint ˜child™
c. tages, tac ˜day™

See Mihm 2004 for a discussion of the history of Final Fortition and the
relationship between Final Fortion in MHG and the modern Fortition found in
the standard language.
As we will see in the following section, the changes in phonology that are
evident in the MHG period, in particular the weakening of unstressed vowels,
were to have far-reaching effects on the make-up of the language.


5.3.3 Morphology and syntax
If we compare the OHG paradigm for the noun tag ˜day™ with the paradigm
for its counterpart in MHG, tac, we see how the weakening of unstressed
syllables has led to a simpli¬cation in the in¬‚ectional endings of MHG (see
Table 5.4). In the OHG paradigm there are seven different endings (includ-
ing the zero-ending in the nominative and accusative singular). In the MHG
paradigm there are four different endings. A comparison of the OHG and
MHG paradigms for the preterite indicative of the strong verb nehmen ˜to
take™ also reveals a simpli¬cation in in¬‚ectional endings (see Table 5.5). The
¬rst and third person plural endings, which are distinct in OHG, are iden-
tical in MHG. The various functions of these in¬‚ectional endings (the sig-
naling of case, person and number distinctions, etc.) is increasingly carried
out by analytic means in the MHG period “ through the use of articles with
nouns, personal pronouns with verbs, prepositions, and auxiliary verbs (Paul
1998:21).
198 German

Table 5.5 The preterite indicative of nehmen
˜to take™ in OHG and MHG

OHG MHG

Singular 1st person nam nam
2nd person n¯ mi
a n¦me
3rd person nam nam
Plural 1st person n¯ mum
a nˆ men
a
2nd person n¯ mut
a nˆ met
a
3rd person n¯ mun
a nˆ men
a



In MHG we see an increase in the number of different tense forms through
the use of analytic tenses. In addition to the present and the preterite, we also
see a present perfect and a past perfect. The present perfect is made up of a
present form of haben ˜to have™ or sˆn ˜to be™ and a past participle.
±
(30) er sprach: ˜dˆ hˆ n ich freude vil verlorn™ (Parzival, cited in Paul
aa
1998:294)
he spoke there have I joy much lost
The past perfect is formed using a preterite form of haben or sˆn and a past
±
participle.
er (got) hie in (Abraham) da er tate, also in sin wib gebeten hete
(31)
(Paul 1998:294)
he (God) commanded him (Abraham) that he do as him his wife had
asked
In OHG, the preterite was used to express past perfect meaning and could also
be used to express present perfect meaning (analytic means were used as well;
see the discussion in section 5.2.3).
Although the present tense is used in MHG to express future meaning,
an analytic future is also used: sol ˜shall™, wil ˜intend™, or muo ˜must™ +
in¬nitive.
(32) und dˆ nˆ ch sol ich schouwen die sch“nen juncvrouwen (Paul
aa
1998:296)
and there after shall I see the beautiful virgins

This analytic future can convey modal as well as temporal information. A
periphrastic future with werden ˜to become™ + in¬nitive is not yet common. In
MHG werden is used with a present participle to express inchoative meaning.
(33) Pinte schire vliende wart (Paul 1998:311)
˜Pinte suddenly began to ¬‚ee.™
History of the language 199

Because of the increased use of articles, subject pronouns, prepositions,
and auxiliary verbs during the MHG period, the language in general can be
characterized as increasingly analytic, thus carrying on the trend begun in
Germanic. The syntax of the language is becoming more stable (Paul 1998:22),
with an increasing tendency for verb-second word order in main declarative
clauses and verb-¬nal order in embedded clauses (Roelcke 1997:148).
(34) a. Sˆn name was erkennelich: er hiez der herre Heinrich (Hartmann
±
von Aue, cited in Schmidt 2000:292)
his name was well-known he was-called the Herr Heinrich
¨
b. daz wir ir niemer vergessen mugen (Schmidt 2000:293)
that we them never forget can
In OHG, the clitic ni ˜not™ was used preverbally to negate a sentence, either
alone or with a negative inde¬nite pronoun.
(35) a. ni mag ther man iouuiht intphahen (Schrodt 2004:136)
not can the man something receive
b. nioman nist in thinemo cunne (Schrodt 2004:136)
no-one not-is in your family
In MHG, the use of ne (< OHG ni) by itself becomes less common; ne (en when
cliticized onto the front of a word) together with niht (< ni wiht ˜not some-
thing™), originally used for emphasis, becomes the regular means of expressing
negation (Paul 1998:338“339).
(36) des enmac diu sch“ne niht getuon (Schmidt 2000:295)
it not-can the beauty not do
In MHG, sentences can in fact contain multiple expressions of negation. These
do not cancel each other out, however; the sentence remains negated.
(37) daz iu nieman niht entuot (Paul 1998:400)
that you no-one not not-do ˜that no-one do it to you™
Eventually, niht takes over the function of negation from en (Paul 1998:398).

5.4 Early New High German

5.4.1 Introduction
Early New High German (ENHG) is generally held to span the period from
1350 to 1650. During this period we see the beginning of the development
of a German standard language. German was used with greater frequency as
the language of the legal documents produced in the chanceries (Kanzleien)
during this period. Although each chancery had its local writing practices,
scribes in the larger chanceries attempted to avoid dialectal characteristics
200 German

that were strongly local. Two widely used written languages with roots in
chancery languages emerged during the ENHG period: Das Gemeine Deutsch,
the Upper German literary dialect from the imperial chancery in Vienna, which
served as the standard written language of southern Germany and Austria; and
Ostmitteldeutsch, the language of the Saxon chancery, which had its roots in the
lingua franca that had developed from the various dialects of the speakers who
had migrated to the east between 1150 and 1350. Ostmitteldeutsch served as
the basis for Martin Luther™s German and ultimately for the German standard
language.
Two important developments hastened the emergence of a standard lan-
guage: the substitution of paper for the more expensive parchment at the end
of the fourteenth century; and the invention of printing with moveable type
(by Johannes Gutenberg) in the middle of the ¬fteenth century. These devel-
opments made the printed word affordable and widely available, whereas the
written word had once been accessible to only a privileged few. The ¬rst books
printed in Germany were in Latin, but towards the end of the seventeenth cen-
tury, German overtook Latin as the dominant language. Printers of German
books developed local printing languages (Druckersprachen) and over time
sought to avoid sounds and expressions that were considered strongly dialectal
in an effort to increase their readership in other dialect areas. Printers turned to
the language of the larger chanceries, which already had wide appeal. The lan-
guage of the imperial chancery, das Gemeine Deutsch, was popular early on: all
fourteen of the pre-Luther German Bibles followed the writing practices of das
Gemeine Deutsch (Waterman 1991:128). Luther™s successful translation of the
Bible, however, resulted in the increase in prestige of the language of the Saxon
chancery, on which it was based. A third literary dialect, Swiss German, enjoyed
wide appeal and retained its independence through the ENHG period. Luther™s
German was rejected in Switzerland for religious reasons (many Swiss were
followers of Calvin and Zwingli), and the language of the imperial chancery
was rejected for political reasons, although Swiss printers did adopt the “new”
Bavarian diphthongs (see the discussion in section 5.4.2) in an effort to increase
their readership.
Although we cannot view Martin Luther (1483“1546) as the creator of New
High German, his translation of the Bible had a major impact on the language.
He sought to create a translation that was idiomatic, natural, and accessible to
ordinary people. As he explains in his “Sendbrief vom Dolmetschen” (˜Circular
letter on translation™), his goal in translating was to reproduce the sense of the
text clearly and vividly (Schmidt 2000:114“115). It is estimated that Hans
Lufft™s press in Wittenberg sold over 100,000 copies of his Bible between 1534
and 1583 (Waterman 1991:130). In part because Luther™s Bible was so widely
read and accepted, the language on which it was based, Ostmitteldeutsch, came
to serve as the basis for the modern standard.
History of the language 201


5.4.2 Phonology
The phonological differences between MHG and ENHG lie primarily in
the vowels. The two major phonological changes are Diphthongization and
Monophthongization.
Diphthongization changed the MHG long vowels ˆ, u, and iu [y¦] to the
±ˆ
diphthongs ei (ai), au, and eu (¨ u).
a
(38) MHG NHG
a. ˆ > ei
± zˆt
± Zeit
ˆ
b. u > au mˆ s
u Maus
c. iu [y¦] > eu diutsch Deutsch
This process began in the twelfth century in the Bavarian dialect area and spread
northward to include the Central German dialects; the Alemannic dialect area
did not undergo this change (in Swiss German one pronounces auf Schwei-
zerdeutsch as [u¦]f Schw[i¦]zerd[y¦]tsch). Diphthongization is also absent in
Low German. The three new High German diphthongs eventually merge in
¨
pronunciation with the old MHG diphthongs, ei, ou, and ou, which were inher-
ited from Germanic. Thus only the MHG forms of the words in which these
diphthongs appear reveal their differing histories.
(39) a. MHG wˆ (weiß) ˜white™ MHG wei (weiß) ˜(I) know™

<<

. 7
( 12)



>>