what constitutes anacrusis? clause openings, kuhn’s law...

16
What constitutes anacrusis? Clause openings, Kuhn’s Law, and the changing shape of Germanic syntax 1 Patrizia Noel Abstract The shape of Germanic anacrusis is puzzling because the kind of variation involved is too high to be expected in an otherwise constrained versification system. If anacrusis is part of the meter, it is usually non-obligatory in Germanic. Also, the number of syllables in anacrusis can vary considerably. This article demonstrates that the notion of metrical simplicity is directly related to grammatical well-formedness. A grammatical perspective on what constitutes anacrusis is presented. The shape of Germanic anacrusis is suggested to be a relic of the transition from Proto-Indo-European word order to Germanic word order. While the left sentence periphery in Proto-Indo-European is characterised by enclitics in second position (cf. Wackernagel’s Law), Germanic syntax has “proclitic” sentence particles instead (cf. Kuhn’s Law). It will be shown that anacrusis is subject to syntactic change and thus language- specific. A scenario is proposed in which the shape of Germanic anacrusis emerged as a compromise between language change and conservative versification. Furthermore, the controversial issue of initial unstressed syllables in Germanic alliterative verse and their metrical analysis is addressed. 1. The concept of anacrusis The term anacrusis was introduced in the 18th century by Richard Bentley for classical meter and referred to the initial drops in iambic and anapaestic verse. The concept of anacrusis became established later by the works of Bentley’s student Gottfried Hermann, who defines anacrusis as (1799: §35): Eine oder mehrere der Arsis als deren Ursachen vorhergehende Zeitabtheilungen, die ihrer Natur nach Theile einer vorhergegangenen unendlichen Reihe sind, weil ihnen nicht wieder eine Arsis vorhergeht, heissen anakrgsij. [One or more temporal units preceding the arsis as their cause, which are by nature parts of a preceding infinite sequence because there is no arsis preceding them, are called anakrgsij.] Anacrusis is unambiguously defined by Hermann as the position before the first lift. As the following sections on Germanic anacrusis will show, this concept did not remain unchallenged; in German metrical analysis, differences in the concept of anacrusis also pertain to formal aspects. As regards the shape of anacrusis, Hermann (1799: §45) stresses: Die Anacrusis, als ein Teil der Thesis einer unendlichen Reihe, muß ebenfalls ein durchgängig gleichbleibendes Maaß haben. Auch hier würde durch das ungleiche Maaß der Sylben eine Arsis in die Anacrusis kommen. Wenn z.B. die anapästische Anacrusis ∪∪ | -∪∪- Pater Oceanus, ungleiche Sylben enthalten könnte, so würde statt des Anapästen ein ganz anderer Rhythmus entstehen, z.B. 1 I would like to thank Theo Vennemann and Dagmar Hirschberg for their valuable comments as well as Gillian Barnard for correcting my English. Abbreviations: A: Anacrusis; PIE: Proto-Indo-European; OHG: Old High German; MHG Middle High German; NHG: New High German; E: Modern English; Icel.: Icelandic; Fmin: minimal foot; Fext: extended foot; Fpar: parasitic foot.

Upload: others

Post on 03-Aug-2020

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

What constitutes anacrusis? Clause openings, Kuhn’s Law, and the changing shape of Germanic syntax1

Patrizia Noel

Abstract

The shape of Germanic anacrusis is puzzling because the kind of variation involved is too high to be expected in an otherwise constrained versification system. If anacrusis is part of the meter, it is usually non-obligatory in Germanic. Also, the number of syllables in anacrusis can vary considerably. This article demonstrates that the notion of metrical simplicity is directly related to grammatical well-formedness. A grammatical perspective on what constitutes anacrusis is presented. The shape of Germanic anacrusis is suggested to be a relic of the transition from Proto-Indo-European word order to Germanic word order. While the left sentence periphery in Proto-Indo-European is characterised by enclitics in second position (cf. Wackernagel’s Law), Germanic syntax has “proclitic” sentence particles instead (cf. Kuhn’s Law). It will be shown that anacrusis is subject to syntactic change and thus language-specific. A scenario is proposed in which the shape of Germanic anacrusis emerged as a compromise between language change and conservative versification. Furthermore, the controversial issue of initial unstressed syllables in Germanic alliterative verse and their metrical analysis is addressed.

1. The concept of anacrusis

The term anacrusis was introduced in the 18th century by Richard Bentley for classical meter and referred to the initial drops in iambic and anapaestic verse. The concept of anacrusis became established later by the works of Bentley’s student Gottfried Hermann, who defines anacrusis as (1799: §35):

Eine oder mehrere der Arsis als deren Ursachen vorhergehende Zeitabtheilungen, die ihrer Natur nach Theile einer vorhergegangenen unendlichen Reihe sind, weil ihnen nicht wieder eine Arsis vorhergeht, heissen anakrgsij.

[One or more temporal units preceding the arsis as their cause, which are by nature parts of a preceding infinite sequence because there is no arsis preceding them, are called anakrgsij.]

Anacrusis is unambiguously defined by Hermann as the position before the first lift. As the following sections on Germanic anacrusis will show, this concept did not remain unchallenged; in German metrical analysis, differences in the concept of anacrusis also pertain to formal aspects.

As regards the shape of anacrusis, Hermann (1799: §45) stresses:

Die Anacrusis, als ein Teil der Thesis einer unendlichen Reihe, muß ebenfalls ein durchgängig gleichbleibendes Maaß haben. Auch hier würde durch das ungleiche Maaß der Sylben eine Arsis in die Anacrusis kommen. Wenn z.B. die anapästische Anacrusis

∪∪ | −�∪∪−Pater Oceanus,

ungleiche Sylben enthalten könnte, so würde statt des Anapästen ein ganz anderer Rhythmus entstehen, z.B.

1 I would like to thank Theo Vennemann and Dagmar Hirschberg for their valuable comments as well as Gillian Barnard for correcting my English.Abbreviations: A: Anacrusis; PIE: Proto-Indo-European; OHG: Old High German; MHG Middle High German; NHG: New High German; E: Modern English; Icel.: Icelandic; Fmin: minimal foot; Fext: extended foot; Fpar: parasitic foot.

Page 2: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

Patrizia Noel

parens Oceanus∪−� | −�∪∪−

oder

magnus Oceanus−�∪ | −�∪∪−

indem die längere Sylbe nothwendig zur Arsis würde.

[Anacrusis, as part of the thesis of an infinite sequence, has to have a constant measure throughout. If, for instance, the anapaestic anacrusis ∪∪ | −�∪∪− Pater Oceanus, could comprise dissimilar syllables, a different rhythm would emerge instead of an anapaest, e.g. parens Oceanus ∪−� | −�∪∪− or magnus Oceanus −�∪ | −�∪∪− .]

Thus the range of words permitted in anacrusis depends on syllable structure, more precisely on quantity. Choosing magnus or parens instead of Pater is tantamount to a change of metrical pattern. In other words, language directly affects anacrusis since it constrains choices. The range of vocabulary in initial position is evaluated on the basis of prosody.

However, the situation is different in the Germanic tradition, although the Germanic prosodic system was a quantity system, too. Germanic alliterative verse has free anacrusis (Heusler 1894: 30, 31) as a default. This means that in contrast to bound anacrusis2 the existence or absence of anacrusis is optional in the default of alliterative verse. This technique was still most common in early Middle High German poetry. Both the occurrence of anacrusis and its syllabic filling are independent of the rest of the line. Therefore, the shape of Germanic anacrusis has no effect on the inner composition of the verse. It is of no importance whether a Germanic hemistich is preceded by one or more unstressed syllables in anacrusis (Sievers 1885: 214).

One of the results of the variation within Germanic anacrusis is that there is no consensus with respect to what actually constitutes anacrusis. As a consequence, there is also no general opinion on where the line minus anacrusis starts. Since there are no common criteria for anacrusis in alliterative verse, the question was raised of whether anacrusis should be accepted as a category at all (cf. Kendall 1991 and Hoover 1985). Without a concept of what anacrusis is and whether it does or does not exist, it is impossible to identify versification patterns on objective grounds.

Evidently, there are inherent differences between Classical and Germanic poetry in terms of the form of anacrusis. Old Germanic anacrusis shows the following peculiarities: the number of syllables varies within the same poem, syllabic structure is not a constituting factor, and anacrusis is not part of the regular meter since it can be present or absent.

It is hypothesised in this article that the basis for the Germanic pattern is language change. The argument is threefold:

1) Old Germanic syntax is characterised by unstressed function words preceding the first content word in the left periphery of the sentence. This position correlates with anacrusis.

2) The varying number of syllables and words characteristic for Germanic anacrusis is derived from a syntax which contains sequences of initial function words.

3) If the syntax of clause openings changes, anacrusis changes as a result.

The following section deals with the grammatical background that laid the language-specific foundation for Germanic anacrusis. It is this language-specific basis which is directly tied to the issue of metrical simplicity.

2 Typical representatives of bound anacrusis in German literary history are the late Minnesang, the Meistergesang, and also the modern canonical meters with trochaic vs. iambic and dactylic vs. anapestic metrical feet (Heusler 1894: 31).

2

Page 3: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

What constitutes anacrusis?

2. The language-specific shape of Germanic anacrusis

The fact that anacrusis as a concept of metrical description was introduced by scholars with a Germanic linguistic and cultural background was probably not coincidental if one considers the central role it plays in Germanic versification. By contrast, anacrusis has been largely irrelevant in syllable-counting metrical traditions to this day. It is claimed here that Germanic syntax in combination with a conservative metrical tradition is the motivation for the characteristic shape of Germanic anacrusis.

2.1. The varying number of syllables in Germanic anacrusis

In an oral metrical tradition like the Old Germanic one which rests on sung poetry, high variation concerning the filling of anacrusis (i.e. musical upbeats in song) does not appear to be economic – it implies complications in timing. The grammatical reason for the emergence of unstressed syllables before the first lift is a Germanic syntactic innovation. It was first discussed in its diachronic linguistic context by Hans Kuhn.

2.1.1. The grammatical basis: From Proto-Indo-European to Germanic clause openings

In his article “Wortstellung und -betonung im Altgermanischen” (‘Old Germanic word order and stress’) written in 1933, Kuhn focuses on a systematic difference between Proto-Indo-European and Germanic word order. Proto-Indo-European clause openings had been subsumed under Wackernagel’s Law (1892) before. In the Proto-Indo-European sentence, certain words like coordinating conjunctions, certain adverbs, and object pronouns occurred in second position, i.e. after the first word. For instance, the Latin sentence conjunction -que ‘and’ always comes as the second word or second constituent of a sentence: remoramque faciunt rei privatae et publicae (Plautus, Trinummus, 38). The English translation, however, has the sentence conjunction not in second, but in first position: And they caused obstruction to public and private affairs. Kuhn’s achievement is the first adaptation of the Proto-Indo-European state of second position words to Germanic (1933: 6). Sentence particles were enclitic (i.e. “leaning” on the preceding word) to the first word in Proto-Indo-European, but proclitic3 (i.e. “leaning” on the following word) in the Old Germanic languages. He calls the Germanic development Germanisches Satzpartikelgesetz, “Germanic law of sentence particles”, also known as Kuhn’s First Law:

Die satzpartikeln stehen in der ersten senkung des satzes, in der proklise entweder zu seinem ersten oder zweiten betonten worte. (Kuhn 1933: 8)

[‘Sentence particles are found in the first drop of the sentence, i.e. in proclitic position either to the first or to the second stressed word.’]

As Satzpartikeln ‘sentence particles’, Kuhn defines all unstressed or weakly stressed words which refer to the whole sentence and which are therefore syntactically autonomous. The class of Germanic sentence particles consists of “substantive pronouns and sometimes adjective pronouns, many adverbs, finite and sometimes infinite verbs, conjunctions, predicative nouns, and “maybe” vocatives” [transl. PN] (Kuhn 1933: 4-5).

Sentence particles in first drop position occur in anacrusis. However, there are also sentence particles which occur later. For instance, while coordinating sentence conjunctions precede the first content word, anaphoric object pronouns usually follow it. As to the principle behind this order of sentence particles, Kuhn (1933: 100) proposes a solution in terms of phonological prominence:

An die erste stelle hat man im auftakt gern die satzpartikel gesetzt, die den stärksten ton hatte, nach

3 In Germanic, the notion of “proclisis” is rather an analogy to the term “enclisis” than a morphonological fact.

3

Page 4: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

Patrizia Noel

einem starktonigen wort an der satzspitze dagegen die tonschwächste, so daß die satzpartikeln der letzten art meist die zweite stelle des satzes bekamen.

[‘In first position in anacrusis one preferably placed the sentence particle which was most prominent; after a stressed sentence-initial word, however, one placed the weakest one. As a consequence, the sentence particles of the latter kind took the second position in the sentence.’]

By applying a criterion of scaled weakness, Kuhn relies on a traditional consensus concerning the phonological motivation of word order patterns (e.g. Delbrück 1888, Wackernagel 1892, Thurneysen 1892). However, since Kuhn remarked that most Germanic sentence particles could be stressed (1933: 5),4 he was aware of problems with this phonological explanation.5

What he nevertheless aimed at was a description of the occurrence and distribution of grammatical elements in the left periphery of the sentence. Therefore, Kuhn had to separate sentence particles from unstressed sentence part particles. Kuhn’s First Law excludes the following construction:

Das satzpartikelgesetz schließt es aus, daß an der satzspitze vor den satzpartikeln ein wort steht, das satzteilpartikeln (artikel, adject. pron., adverbien, präpos. oder eine unbetonte vorsilbe) vor sich hat, weil die satzpartikeln sonst in der zweiten senkung des satzes stehen würden. (Kuhn 1933: 43)

[‘The law of sentence particles forbids words with sentence part particles (article, adjectival pronoun, preposition, or an unstressed prefix) to precede sentence particles at the absolute beginning of a sentence, since sentence particles would otherwise occur in the second drop of the sentence.’]

By his law, Kuhn states that a sentence does not start with a sentence part particle (i.e. an article, adjectival pronoun, preposition, or an unstressed prefix) alone before the content word if the sentence contains a sentence particle. Syntax, therefore, automatically constrains anacrusis in all metrical patterns in which the beginning of the line is identical with the beginning of the clause.

As a textual basis for his analysis, Kuhn chose poetry, because unlike prose it documents stress (“Betonungsverhältnisse”) (1933: 2). The Old English Beowulf, as the oldest Germanic epic, is particularly insightful in terms of early syntactic patterns. It contains about 2000 sentence particles excluding connectors (1933: 10). The construction with Kuhnian sentence particles in drop position after a preceding sentence particle in the first lift position occurs about 50 times. Usually, it is the finite verb that occurs in lift position in these constellations. As a default, the finite verb is preceded by a prefix or by the negation particle ne (or both) in anacrusis (1933: 11).6 These negation particles and the prefixes occur before the first stave (cf. (18) and (18) with sentence particles in inner drop position in italics and anacrusis underlined).7

(18) 109 ne geféah heÜ þæÜre fæÜÛhe8

4 This sounds like a contradiction to the definition of sentence particles as weak particles. However, linguistic laws are neither rules nor universals. They are not without exceptions (cf. Kuhn 1933: 43-48 for a list of exceptions). Prosodic weakness had been the defining criterion for Wackernagel (1892), too, who also notes exceptions.

5 There is no prosodic basis for assuming weak and less weak function words if they display the same syllable structure. For an alternative in terms of subsystem interaction and distribution according to scope cf. Noel Aziz Hanna (forthc. a).

6 Kuhn’s law of sentence particles works only if finite verbs are considered sentence particles. Wackernagel (1892) had claimed for Proto-Indo-European finite verbs that they were enclitic and thus offered a hypothesis on the emergence of verb second in Germanic. Kuhn takes over Wackernagel’s concept. Just as Wackernagel’s hypothesis, Kuhn’s classification is not unproblematic. It is, for instance, unclear why they are, as sentence particles, placed in second position in Germanic declaratives, but clause-finally in other Indo-European languages.

7 The same pattern is known in the Old Saxon Heliand in which it occurs 170 times. The finite verb occurs in lift position and is preceded by the prefix or the negation particle (Kuhn 1933: 16).

8 [‘this feud he did not enjoy’].

4

Page 5: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

What constitutes anacrusis?

(18) 1322 ne frīÛn þuÜ æfter sæÜÛlum9

Kuhn combines his First Law with a second law concerning a particularity of Germanic syntax: The Germanic Law of clause openings (Germanisches Satzspitzengesetz). Kuhn’s Second Law says:

Im satzauftakt müssen satzpartikeln stehen. (Kuhn 1933: 43)

[‘There have to be sentence particles in the opening of a clause.’]

The fact that Germanic sentences, as a default, start with sentence particles paved the way for the Germanic version of anacrusis as a structural position. It also meant that there are clear lexical limitations for words in anacrusis. Suzuki (1995: 146-147) summarizes the findings on the lexical filling of anacrusis in Beowulf (Cable 1974, Ducan 1985):

[I]t is not the case that any arbitrary unstressed syllable qualifies for forming anacrusis. As Cable and Duncan have revealingly shown, most typically a verbal prefix and a negative particle ne (or occasionally a combination of the two) may constitute anacrusis. Out of a total of 89 instances, there are only 15 occurrences (ca. 17% in ratio) that deviate from this regularity.

In Beowulf with its “compact” language, the optional anacrusis contains one to four syllables. The number of syllables in anacrusis is comparatively low as compared to later alliterative verse. But what about the status of the limited inventory of lexical material in anacrusis? There are two possibilities. The first is a convention on the filling of anacrusis which was passed on from generation to generation as a metrical rule or an internalised abstraction (cf. Orton 1999: 287). In this perspective, Kuhn’s laws would refer to constraints on versification that poets had to be consciously aware of. The other possibility is a syntactic pattern that is naturally employed in versification.10 It is this second solution that is investigated in the following sections, the basis being a scenario for the emergence of anacrusis and language change effects in versification.

2.1.2. A hypothetical linguistic scenario for the emergence of Germanic anacrusis

The grammatical basis of polysyllabic anacrusis resulted from the above outlined word order difference between Proto-Indo-European and Germanic. Gothic, as the oldest attested Germanic language, still showed features in the left sentence periphery that were more similar to Proto-Indo-European syntax than to Old English or Old High German syntax. It had sentence conjunctions and other sentence particles that were enclitic to the first word or particle of the sentence.11 With such enclitic sentence particles, the likelihood of a chain of function words in anacrusis is low. If a grammatical system has enclitic particles instead of Kuhnian sentence particles, the typical elements of Germanic anacrusis occur not at the beginning of the line, but in the middle of the line. Thus if alliterative verse had its origins in a system similar to Gothic, the question of long anacrusis with a varying number of syllables would simply not have posed itself to the same degree.

In a scenario in which Germanic alliterative verse emerged on a syntactic basis like Gothic

9 [‘Do not you ask after pleasures’].10 The options do not represent conflicting solutions. The pattern of natural versification can

precede a stage of metrical rule or internalised abstraction in a formulaic mode of composition. Orton (1999: 302), too, argues for unearthing the linguistic foundations of Kuhn’s Laws.

11 In Proto-Indo-European syntax, an initial word is followed by enclitic particles that are followed by the rest of the sentence (Wackernagel 1892). The classification is still considered valid, c.f. Krisch’s formulation (1997: 283):

(1) #X(E)...........# X: one word (default); E: Wackernagel enclitic(2) #X(E1)C(E2).....# E1 and E2: Wackernagel enclitics; E1: enclitic coordinating conjunctions; E2: sentence particles and enclitic pronouns; C: subordinating conjunctions, relative pronouns, question pronouns, verbs.

5

Page 6: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

Patrizia Noel

with enclitic particles in Wackernagel position, the change from Wackernagel syntax to Kuhnian syntax means an increase of elements in the left sentence periphery before the first stressed word. The first stressed word corresponds to the first lift position in the line for all meters which start lines with sentences. The number of syllables in anacrusis thus increased for reasons of language change.

Meanwhile, the metrical pattern remained conservative: it did not integrate these new “extrametrical” syllables. There was no integration of initial function words to the regular meter, because a principle of preferably assigning stress to content words holds. Just like a typical Proto-Indo-European sentence, regular meter in Germanic alliterative verse usually starts with a stressed content word. Alliterative verse does not just map prosodic rhythm, but also incorporates a distinction between informative and less informative units, which is encoded by staves. The prototypical sentence particle, unlike the prototypical alliterating word, does not encode new information, but serves the function of establishing coherence and supporting discourse organisation. It is therefore unstressed. As the diachronic perspective on Kuhnian particles will show, this principle of versification was gradually given up. The left periphery of the sentence became integrated in the canonical metrical patterns, thus changing the shape of Germanic versification. Also, the lexical material in anacrusis developed according to syntactic change.

2.2. A diachronic perspective on the change of lexical inventory and syntactic position in Germanic anacrusis: the Germanic negation particle

In a synchronic analysis of versification principles, a distinction between naturally employed metrical patterns on the basis of everyday language and the mere continuation of metrical conventions is not unproblematic. However, the influence of linguistic structure on a metrical system becomes conspicuous if the metrical system changes as a consequence of language change. As will be illustrated in the next section, language change went hand in hand with a change in the lexical inventory of anacrusis. Since the role of the negation particle in anacrusis was highlighted by Kuhn and others for Old Germanic poetry (section 2.1.1.), this particle is traced from Old High German to Middle High German, after which it gradually disappeared.

2.2.1. The weakness of the Germanic negation particle

The Germanic negation particle is involved in systematic word order regularities in the left periphery of the sentence from the oldest times on. Its phonological weakness and its preverbal position in a verb second language are factors that made its placement in anacrusis convenient. The correlation of anacrusis and grammatical characteristics is not a trivial one because both weakness and position of the particle are Germanic innovations. Versification exploits these new grammatical structures. The weakness of the Germanic negation particle was not inherited from Proto-Indo-European (Behaghel 1932: §1428):

Im Germanischen steht die Negation + Verbum am Eingang des Satzes, so daß also dieses in Mittelstellung erscheint; das stammt offenbar aus dem Igm., vgl. lat. nequeo, nescio; es wird wohl ursprünglich die Negationspartikel betont gewesen sein, mindestens stärker als das Verbum. (Behaghel 1932: §1428)

[‘In Germanic, the negation + verb are placed at the beginning of the sentence, which means that the verb appears in the middle of the sentence; this is evidently inherited from Indo-European, cf. Lat. nequeo, nescio. Originally the negation particle was probably stressed, and was at least more prominent than the verb.’]

The Germanic negation particle was reduced from the beginnings of its documentations. The Proto-Indo-European negation particle is *ne with full vowel e; in contrast to this unreduced form, the Old High German form is ni with the vowel i indicating that the particle behaves

6

Page 7: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

What constitutes anacrusis?

like an unstressed prefix (Braune/Eggers 1987: § 70); the Middle High German particle was ne with reduced vowel schwa which resulted from systematic absence of stress. The gradual reduction of the German negation particle is summarised in (18):

(18) The gradual reduction of the German negation particle:PIE *ne > OHG ni > MHG ne/en > NHG Ø

As a consequence of its systematic weakness, the Germanic negation particle was later on re-enforced and then replaced by other particles, e.g. E not, Icel. ekki, NHG nicht ‘not’. The systematics of reinterpretation are known as Jespersen’s Cycle (1917). Table 2 gives the development in German:

Phase I ni + VfinPhase II ni + Vfin + (niwiht)

Phase III en/ne + Vfin + niht

Phase IV (ne) + Vfin + niht

Phase V + Vfin + niht

Table 2. Jespersen’s Cycle (cf. Donhauser 1996: 202)

In phase I, Early Old High German has mono-negation with ni. In phase II, Late Old High German has optional niwiht ‘not a thing’. Phase III illustrates the Middle High German situation with obligatory double negation. Phase IV represents the Early New High German situation with optional negation particle ne and obligatory negation particle nicht. In phase V, i.e. after the 16th century, mono-negation with nicht becomes effective.

2.2.2. Changes in the grammar of negation and changes in anacrusis

Apart from the weakening of the old Germanic negation particle, also the preverbal placement of the particle in the left sentence periphery was a Germanic innovation. There neither was a limitation to preverbal placement nor systematic verb second syntax in Proto-Indo-European. This, in turn, means that the peculiar interrelations of Kuhn’s Law with the negation particle are language specific.

Unlike preverbal OHG ni, the new negation particle NHG nicht ‘not’ is neither preverbal nor placed in the left periphery of the sentence. Trivially, the modern German particle is therefore not a good candidate for anacrusis. But also with respect to the situation before mono-negative NHG nicht, there is a change in the capacity of the negation particle for forming anacrusis, which is due to grammatical change. This change is here illustrated by Otfrid von Weissenburg’s Evangelienbuch  (9th century, South-Rhenish Franconian) and the Nibelungenlied (manuscripts A, B, and C12).

In the Old High German alliterative Evangelienbuch,   the  negation particle   is  hardly ever stressed.  This   can  be  deduced   from  Otfrid’s   use  of   diacritics.  The acute marks stressed syllables, while a subscript dot marks elided vowels. Otfrid gives two lines with stressed ni and 23 cases of stressed nist ‘is not’. However, as the statistics in Table 2 illustrate, stressed occurrences of ni are significantly infrequent as compared to unstressed ni.

12 A, Hohenems-Münchener Handschrift, Cod. germ. 34, Bayerische Staatsbibliothek München, second half of the 13th century; B, St. Galler Handschrift, Ms. 857, Stiftsbibliothek St. Gallen, second half of the 13th century; C, Hohenems-Laßbergische or Donaueschinger Handschrift, Ms. 63, Hofbibliothek zu Donaueschingen, probably first half of the 13th century (cf. Becker 1977).

7

Page 8: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

Patrizia Noel

word frequencyni 1111ní 2

níst 23ni 6

Table 2. Statistics: ni with and without acute in Otfrid’s Evangelienbuch

1111 instances of unstressed ni as opposed to two instances of stressed ni demonstrate that this particle was usually unstressed in Old High German. It is therefore unsurprising that ni was placed in anacrusis. Ni was unstressed and occurred in preverbal position in a language in with verb second syntax was prevalent. (18) and (18) give examples with OHG ni in anacrusis:

OHG ni in anacrusis(18) ni brístit thoh in thés thiu mín ni sie sih házzon untar ín, (Otfrid, Evb V.23)13

(18) Ni scríbu ih nu in alawár so sih ther órdo dregit thár (Otfrid, Evb III.1)14

The Middle High German situation concerning the syntactic position of the negation particle is different from Old High German. Thus also the position of the particle in anacrusis changes. In the manuscripts A, B, and C of the strophic Nibelungenlied, a comparison of sentence rhythm in parallel lines is possible because of their strict metrical organisation. Meters are not just a condition a poet has to meet. They evolve from everyday speech and stylise speech rhythm (cf. Miller 1902, Vennemann 1995).

A comparison of the manuscripts A, B, and C of the Nibelungenlied reveals that there are considerable signs of caution in placing the negation particle in a lift position, no matter whether it functions as a sentence negation or as a word negation. There is just one case of unambiguously stressed ne (18); in this line, ne is emphatic.15 When ne is placed in front of a verb that starts with a stressed syllable, ne is integrated in the preceding foot, cf. Hiba: Ahivatkozás forrása nem található). It can also become deleted Hiba: A hivatkozás forrása nemtalálható) or denuclearised (18), which again highlights its weakness.

(18) Emphatic ne (C 1782.2)deich hort der Nibelunge | nie|ne ge|pflach. | Ø16

Fmin Fmin Fmin Ø

(18) ne as part of the preceding foot (A 1691.4)ich lâze ivch daz beschowen, daz | ich ge|logen | niene | hân.17

A Fext Fmin Fext Fmin

(18) Deletion of ne (A: 809.2)|ia [18] ge|diende | Sî|frit | nie alsolhen haz19

FminFext Fmin Fmin

(18) Denuclearisation: loss of the reduced vowel (C 133.3)vñ ovch in ein div frowe, die | er noch | nien[ e ] ge |sach," | Ø20

A Fext Fext Fmin Ø

The data from the Nibelungenlied, with its various strategies of getting ne out of strong metrical positions, demonstrates that there was a certain unease for Middle High German

13 [‘yet still they do not lack hatred towards each other’].14 [‘I will certainly not write in chronological order’].15 Manuscript B 1739 has emphatic nie nie.16 [‘that I did not take care of the hoard of the Nibelungs anymore’].17 [‘I will show you that I did not lie’].18 The particle ne is realised in the parallel manuscripts B and C.19 [‘Yes, Siegfried has never deserved such hatred’].20 [‘And also the very lady, whom he had never seen, [esteemed] him’].

8

Page 9: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

What constitutes anacrusis?

poets of stressing the negation particle. This qualifies it for anacrusis, cf. (18), (18):

MHG ne in anacrusis:(18) ich sol si miden beide; sone | chan mir| nimmer | misse|gan. |21 (B 15.4)

AFext Fext Fext Fmin

(18) si ne | chvndes |niht be|scei|den | baz der gvoten22 (B 12.2)A Fext Fext Fmin Fpar

However, unlike the negation particle in Beowulf, MHG ne never occurs in the first position of a sentence in the Nibelungenlied. While   there are 156 cases of ni in Otfrid’s Evangelienbuch at the absolute beginning of a sentence, the Middle High German particle ne is not accepted in this position in verb second sentences in the Nibelungenlied. If ne occurs in anacrusis, it has to occupy a position later than the first, for instance second position in the case of so and si in (18) and (18) above.

The grammatical reason for the positional change of the negation particle from Old High German to Middle High German is the stabilisation of verb second in the declarative sentence.23 In contrast to the Old High German negation particle, MHG ne could not fill the prefield24 of a sentence on its own any more. If it occurs in anacrusis, it is accompanied by a sentence-initial word to its left.

As a consequence, the Middle High German negation particle ne could not be placed any more in anacrusis in the instances described by Kuhn in terms of Old Germanic. The old construction of anacrusis was affected by the establishment of verb second syntax. Thus anacrusis is subject to language change.

3. Concepts of anacrusis and their consequences for metrical descriptions: basic types vs. minimal requirements

Connected to the question of what constitutes anacrusis is the question of how to describe in metrical terms where the line minus anacrusis begins. Descriptions of Germanic anacrusis expose fuzzy edges of this category. At the centre of discussion is the issue of whether there is a difference between anacrusis and line-initial unstressed syllables in Germanic alliterative verse.

3.1. Anacrusis in alliterative verse I: Sievers’ classification

The most frequently cited model of Germanic alliterative poetry is Sievers’ system of five types. Sievers’ classification rests on his concept of anacrusis. He writes (1885: 214-215):

Der ‘auftakt’ ist eben nur ein vorschlag zu der eigentlichen rhythmischen reihe, der beliebig stehen oder fehlen kann. Just die willkür in seiner anwendung ist es, die diesen vorschlag als ‘auftakt’ charaktierisiert. Ganz anders liegen die dinge, wenn mit dem stehen oder fehlen einer unbetonten silbe im verseingang gewisse modificationen der ganzen verszeile hand in hand gehen [...] Es kann dann meines bedünkens keine rede davon sein, dass in solchem falle die unbetonte eingangssilbe nur die rolle eines auftakts spiele; sie ist dann ein wesentlicher bestandteil des verses und muss als solcher auch von der theorie in die eigentliche rhythmische reihe hineinbezogen werden.

[‘Anacrusis is just an upbeat to the actual rhythmic sequence. It can be present or it can be missing, as desired. It is the arbitrariness of its application that characterises this upbeat as ‘anacrusis’. This stands in contrast to certain modifications of a complete line, which go hand in hand with the presence or absence of an unstressed syllable at the beginning of the line. [...]. From my point of view, it is out of the question that the unstressed syllable plays just the role of anacrusis in such cases; then it is an integral part of the verse and, as such, has to be integrated by the theory in the

21 [‘I shall avoid both of them; so I can never be harmed’].22 [‘she could not give a better vision’]23 Verb first remained as the pattern of imperative sentences and interrogative sentences.24 The pre-field precedes the finite verb in a verb second sentence.

9

Page 10: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

Patrizia Noel

actual rhythmic sequence.’]

Thus Sievers suggests a metrical concept in which syllables in anacrusis and unstressed syllables before the first lift are distinguished. Consequently, the notion of anacrusis in contrast to unstressed initial syllables plays a major role in the classification of Sievers’ five basic types. (18) gives variants of type A, which is constituted by two left-headed measures. The rhythmic pattern of type A fits Germanic prosody, since Germanic prosodic feet are left-headed. In the left column, Sievers lists the variants without anacrusis (“ohne Auftakt”), and in the right column the variants with anacrusis (“mit Auftakt”):

(18) Anacrusis in Sievers’ type A: –Ûx | –Ûx25 (Sievers 1885: 275-276)

Sievers considers tetrasyllabic forms to be basic (for a critique of syllable-based basic patterns in Germanic cf. Heusler 1925: §171, Noel Aziz Hanna forthc. b). As Sievers’ list of variants with up to four syllables in anacrusis in (18) shows, all syllables before the first stress count as anacrusis in type A. However, this is not the case in Sievers’ basic types B (x–Û | x –Û) and C (x–Û | –Ûx), in which there is no anacrusis. In types B and C, the first unstressed syllable constitutes part of the regular meter and therefore belongs to the first measure. (18) gives variants of type B:

25 Countings for first hemistichs of the simple form ; –Û: long and stressed syllable; x: unstressed (long or short) syllable; (x): anacrusis. „Ohne Auftakt“: without anacrusis; “mit Auftakt”: with anacrusis. Numbers: number of hemistichs. Asterisk: problematic cases.

10

Page 11: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

What constitutes anacrusis?

(18) Unstressed initial syllables in Sievers’ type B: –Ûx | –Ûx26 (Sievers 1885: 294)

Since Sievers takes tetrasyllabic lines to be basic and since anacrusis is not part of the regular metrical pattern, it is only consistent that he assigns a special status to unstressed syllables. If the first unstressed syllable was part of anacrusis in tetrasyllabic lines of type B, the regular meter no longer consisted of four syllables. However, there is no grammatical, lexical, or prosodic reason for the separation of anacrusis and unstressed line-initial syllables.27

Just as in Sievers’ type theory, tetrasyllabicity plays a role in metrical descriptions of Middle High German poetry. However, the status of tetrasyllabicity differs from Sievers’ approach: Tetrasyllabicity is not considered the basic pattern, but the minimal requirement. Beyschlag (1959: 19) therefore claims in terms of Middle High German anacrusis:

Die Setzung eines Auftaktes ist meistens frei; jedoch kann die Halbzeile nie unter vier Silben sinken: umfassen die beiden Füße nur drei Silben, dann ist Auftakt erforderlich.“

[Employing anacrusis is usually free; however, the half-line can never drop below four syllables: if the two feet comprise only three syllables, anacrusis is obligatory.]

In research on Middle High German metrics, anacrusis comprises all syllables before the first lift (e.g. Beyschlag 1959: 12, Hoffmann 1981: 3). This principle stands in contrast to Sievers’ concept of alliterative poetry with its unstressed initial syllables and implies the question of whether the Old Germanic tradition was interrupted or Sievers’ description is too complex.

3.2. Anacrusis in alliterative verse II: Heusler’s classification

26 “Zweite Senkung einsilbig”: monosyllabic second drop; “Zweite Senkung zweisilbig”: disyllabic second drop. Numbers in brackets refer to second hemistichs, numbers without brackets to first hemistichs.

27 The concept of anacrusis vs. unstressed line-initial syllables has therefore been criticised (cf. e.g. Hoover 1985: 139, 144).

11

Page 12: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

Patrizia Noel

The interpretation of the first lift as an indicator for the end of anacrusis for Middle High German poetry agrees with Heusler’s theory on alliterative verse (1925-1929). In Heusler’s system, all syllables that precede the first lift occur in anacrusis.

He gives an overview on the default number of syllables in anacrusis in alliterative verse (1925: §216). For instance, in the Eddic long line, anacrusis is usually monosyllabic, while three syllables are rare, and the maximum number of syllables is four. In Beowulf, two syllables are most frequent, four syllables are rare, and in rare cases five syllables occur in second hemistichs. Other Old English poetry has up to seven syllables in anacrusis. In the Old High German Hildebrandslied, anacrusis with up to six syllables is employed. Again, longer anacrusis occurs in second hemistichs. The Old Saxon Genesis has up to 11 syllables in anacrusis, the Old Saxon Heliand, according to Heusler, up to 14 syllables:

(18) Anacrusis in the Old Saxon Heliand (Heliand, VII, 605.2; Heusler 1925: §216)

Saga ûs, undar huilicumu he sî thesaro cunneo afôdit28

The reason why Heusler assumes anacruses of such enormous length is that he consistently follows the criterion of anacrusis containing all syllables before the first lift. He suggests (1925: § 218) that very long anacrusis occurs only in poetry which is not sung.

3.3. Anacrusis from Middle High German to New High German

In Middle High German versification, the change from alliteration to the default of end rhyme had taken place. Anacrusis is exemplified by a strophic form, the first strophe of the Nibelungenlied in (18). Anacrusis is present in 1.1.1 (on-verse), but absent in 1.1.2 (off-verse). It consists of two syllables in 1.4.2 and otherwise of one syllable. The initial hemistich, which corresponds to the beginning of the sentence, conforms to Kuhn’s law in that it starts with the Kuhnian sentence particle uns ‘us’ which, as a function word without sentence stress, occurs in anacrusis29.

(18) 1st strophe of the Nibelungenlied with zero to two syllables in (C 1)30

Uns | ist in | alten | mæ|ren | Ø | wnders | vil ge|seit | Ø |A Fext Fext Fmin Fpar A Fext Fext Fmin Ø

von | heleden |lobe|bæ|ren |, von | grozer | are|beit, | Ø |A Fext Fmin Fmin Fpar A FextFmin Fmin Ø

von | frevde (v)ñ | hochge|ci|ten |, von | weinen | vñ | klagen, | Ø |A Fext Fext Fmin Fpar A FextFmin Fmin Ø

von | kvner | recken | stri|ten | mvget | ir nv | wnder | horen | sagen. |A Fext Fext Fmin Fpar A Fext Fext Fext Fmin

The first strophe of the epic was made with exceptional diligence, since it shows internal rhyme in addition to end rhyme. This extra effort in the first strophe clearly demonstrates that the absence or presence of anacrusis as well as the varying number of syllables in anacrusis did not influence the quality of the meter negatively. As can also be seen, the pattern of starting the line minus anacrusis with a content word continued (wunders ‘wonder’, heleden ‘heroes’, grozer ‘great’, freude ‘joy’, weinen ‘weeping’, kuner ‘keen’) after alliterative verse

28 [‘Tell us of which of these families he sprouted from’].29 The finite verb ist is also a Kuhnian sentence particle. Kuhn acknowledges instances of

stressed (alliterating) sentence particles (see section 2.1.1.above).30 [‘To us in olden story / are wonders many told

Of heroes rich in glory, / of trials manifold:Of joy and festive greeting, / of weeping and of woe,Of keenest warriors meeting, / shall ye now many a wonder know.’]

(Translation by Needler 1904).

12

Page 13: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

What constitutes anacrusis?

became old-fashioned. This procedure is not an artificial feature of alliterative verse, but is derived from everyday language, since content words bear more information weight than functions words and are thus more likely to be stressed. However, this praxis does not have the quality of a rule, but only of a preferred strategy, as line 1.1.1 and 1.4.2 with the auxiliary ist ‘is’ and the pronoun ir ‘you’ in first lift position demonstrate. In Middle High German (18a) and in Modern German poetry (18b), stressed initial function words are not uncommon.31 Evidently, Kuhn’s Law is no longer effective, since Kuhnian sentence particles do not have to be placed in first drop position any more.

(18) Initial Kuhnian sentence particle (conjunction) in the first lift position(a) | unde | wîse | mich | dar | (Hartmann, Iwein, 541)32

(b) | Und be|kümmert | sieht der | Schiffer | (Goethe, Meeresstille) 33

Although there was a metrical change resulting in the integration of Kuhnian sentence particles to the regular meter, free anacrusis was not lost. It is characteristic of Old High German and Early Middle High German couplets. In Early Middle High German, trisyllabic anacrusis is still frequent, and up to six syllables occur (Paul/Glier 1961: §46, §53; Heusler 1927: §559-560). Even in Middle High German strophic forms, the shape and presence or absence of anacrusis is not necessarily a defining criterion. Despite the then prestigious model of syllable-counting Romance versification, the number of syllables occurring in anacrusis is often not limited. However, in the later Minnesang, presence or absence of anacrusis was increasingly regulated with anacrusis becoming part of regular meter. The result of this development has characterised German canonical forms since the Baroque literary theorist Opitz who established the notion of iambic lines (Heusler 1927: §559; §643-644) with Romance poetry as a structural image:

(18) Ist Liebe lauter nichts / wie daß sie mich entzündet?34 xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

Ist sie dann gleichwol was / wem ist ihr Thun bewust? xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

Ist sie auch gut und recht / wie bringt sie böse Lust? xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

Ist sie nicht gut / wie daß man Frewd' aus ihr empfindet? xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

Lieb' ich ohn allen Zwang / wie ka n ich Schmertzen tragen? xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

Muß ich es thun / was hilfft's daß ich solch Trawren führ'? xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

Heb' ich es ungern an / wer dann befihlt es mir? xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

Thue ich es aber gern' / umb was hab' ich zu klagen? xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

31 Furthermore and in contrast to the Old Germanic tradition, Middle High German verse has instances of content words in anacrusis (Heusler 1927: §561), e.g. MHG kurz ‘short’:

| kurz, | scharf als | igels hût | ez waz | (Wolfram, Parzival 517, 27)[‘it was short and sharp as a hedgehog’s skin’].

32 [‘and direct me there’].33 [‘and distressed the mariner sees’].34 [‘Is love nothing more but that it sets me on fire? Yet if it is something, who is conscious of

its deeds? And be it good and virtuous, how does it bring bad passion? If it is not good, how come that it brings a feeling of joy? If I love without restraint, how can I bear the pain? If I have to do it, of what avail is it to grief? If I start it reluctantly, who orders me to do so? Yet if I like to do so, what to I lament? I waver in the cool wind like the grass beds to and fro around vespers time: I flow like a ship through the wild sea which haunted by waves cannot come to terms. I do not know what I want, I do not want what I know. I am freezing in summer, I am hot in winter.’].

13

Page 14: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

Patrizia Noel

Ich wancke wie das Graß so von den kühlen Winden xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

Umb Vesperzeit bald hin geneiget wird / bald her: xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

Ich walle wie ein Schiff das durch das wilde Meer xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

Von Wellen umbgejagt nicht kan zu Rande finden. xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

Ich weis nicht was ich wil / ich wil nicht was ich weis: xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

Im Sommer ist mir kalt / im Winter ist mir heiß. xxÛ|xxÛÛ|xxÛÛÛ||xxÛÛÛ|xxÛÛÛ|xxÛ

(Martin Opitz, Francisci Petrarchae)

In iambic verse, initial unstressed syllables are usually not conceived of as an instance of fixed anacrusis, but as an integral part of the regular meter. The meter as such no longer starts with the first content word, but is predefined as starting with one unstressed syllable. Although German prosodic rhythm has always been left-headed (trochees and dactyls), the “iambic” versification pattern does not violate rhythmic well-formedness on the level of the sentence. On closer examination with the Germanic tradition as a background, “iambic” patterns work like earlier metrical patterns with bound anacrusis. Thus in New High German canonical meters, anacrusis is disguised behind the established metrical categorisation. Yet although the metrical terminology changed, the grammatical basis of anacrusis did not vanish in the Germanic languages and is therefore reflected in metrical patterns.

In this respect, anacrusis has constantly been employed to this day. Even the presence or absence of anacrusis was not abolished as a feature of versification. Free anacrusis is upheld by “iambic reversal” resulting in dactylic beginnings (e.g. Standop 1972, Tarlinskaya 1987). Although Opitz accepts only iambs and dactyls in his treaty on poetry, his poem in (18) displays several lines with iambic reversal, i.e. initial dactyls, according to everyday language: ll. 2, 3, 4, 5, 6, 7, 8, because the first word in the line is heavier in terms of information structure. Line-initial dactylic feet in iambic verse correspond to lines without anacrusis, thus again permitting unregulated presence and absence of unstressed line-initial syllables in an otherwise strictly-confined metrical pattern. However, while this peculiar feature of Germanic poetry prevailed, another characteristic was lost. The central metrical peculiarity of Old Germanic anacrusis, which was finally given up in the canonical New High German meters, is the traditional variety of the number of syllables. The conservative preservation of not integrating long stretches of syllables before the first content word changed when sentence rhythm was ranked higher that information weight. This again altered the shape of Germanic anacrusis, finally resolving the metrical traces of the transition from Wackernagel’s Law to Kuhn’s Laws.

Conclusion

This article discussed the emergence of optional sequences of syllables in Germanic anacrusis and the decline of this feature on a grammatical basis. Anacrusis is not a static phenomenon in German and Germanic versification. It was suggested that grammatical change from Proto-Indo-European to Germanic syntax, in combination with a conservative tendency to assign the first lift to the first content word, caused the characteristic shape of old Germanic anacrusis. It was, furthermore, claimed that the grammatical relation between Germanic syntax and anacrusis is threefold:

1) Old Germanic syntax is characterised by unstressed function words preceding the first content word in the left periphery of the sentence. This position correlates with anacrusis.

2) The varying number of syllables and words characteristic for Germanic anacrusis is derived from a syntax which contains sequences of initial function words.

14

Page 15: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

What constitutes anacrusis?

3) If the syntax of clause openings changes, anacrusis changes as a result.

The question of what constitutes Old Germanic anacrusis is a language-specific issue – while unstressed function words follow the first content word in Proto-Indo-European, they precede it in Germanic. Thus Germanic anacrusis is ‘simple’ in terms of its applicability in terms of Germanic syntax. The special shape of Germanic anacrusis represents a good match between the grammatical and the metrical system, not despite but because of its variation.

ReferencesBecker, P. J. (1977), Handschriften und Frühdrucke mittelhochdeutscher Epen. Wiesbaden: Reichert.Behaghel, O. (1932), Deutsche Syntax. Vol IV. Heidelberg: Carl Winters Universitätsbuchhandlung.Beyschlag, S. (1959), Die Metrik der mittelhochdeutschen Blütezeit. 3rd ed. Nürnberg: Carl.Braune, W. (1987), Althochdeutsche Grammatik. 14. Auflage bearbeitet von H. Eggers. Tübingen:

Niemeyer.Cable, Th. (1974), The meter and melody of Beowulf. Urbana: University of Illinois Press.Delbrück, B. (1888), Syntaktische Forschungen V: Altindische Syntax. Tübingen: Niemeyer.Donhauser, K. (1996), Negationssyntax in der deutschen Sprachgeschichte: Grammatikalisierung oder

Degrammatikalisierung? In Deutsch – typologisch, E. Lang and G. Zifonun (eds.), 201-217. Berlin: Walter de Gruyter.

Duncan, E.W. (1985), Stress, meter an alliteration in Old English poetry. PhD dissertation, Univerity of Texas at Austin.

Hermann, G. (1799), Handbuch der Metrik. Leipzig: Gerhard Fleischer d.J.Heusler, A. (1925-1929), Deutsche Versgeschichte: Mit Einschluss des altenglischen und

altnordischen Stabreimverses. Vol. 1 1925, vol. 2 1927, vol. 3 1929. Berlin: Walter de Gruyter.Heusler, A. (1894), Über germanischen Versbau. Berlin: Weidmann.Hoffmann, W. (1981), Altdeutsche Metrik. 2nd ed. Stuttgart: Metzler.Hoover, D.L. (1985), A new theory of English meter. New York: Lang.Jespersen, O. (1917), Negation in English and other languages [Historisk-filologiske Meddelelser I,5].

Kopenhagen: Høst.Kelle, J. (1963), Otfrid von Weißenburg: Evangelienbuch. Vol. 1. Regensburg: Manz. [First published

1856.]Kendall, C.B. (1991), The metrical grammar of "Beowulf". Cambridge: Cambridge University Press.Krisch,  Th.   (1997),  Zum Hyperbaton   in  altindogermanischen  Sprachen.   In:  Wolfgang Meid  (Hg.), 

Sprache   und   Kultur   der   Indogermanen.   Akten   der   X.   Fachtagung   der   Indogermanischen  Gesellschaft.   Innsbruck,   22.–28.   September   1996,   351–384.   Innsbruck:   Institut   für Sprachwissenschaft.

Kuhn, H. (1933), Zur Wortstellung und ­betonung im Altgermanischen.  Beiträge zur Geschichte der  deutschen Sprache und Literatur 51, 1­109.

Miller, C.W.E. (1902), The Relation of the Rhythm of Poetry to that of the Spoken Language with Special Reference to Ancient Greek. In: Studies in Honor of Basil L. Gildersleeve. Baltimore, MD. 

Das Nibelungenlied. Paralleldruck der Handschriften A, B und C nebst Lesarten der übrigen Handschriften. Edited by M. S. Batts 1971. Tübingen: Niemeyer.

The Nibelungenlied. Translated into rhymed English verse in the metre of the original by G.H. Needler 1904. New York: Holt.

Noel Aziz Hanna, P. (forthc. a), Wackernagels Gesetz im Deutschen: Zur Interaktion von Syntax, Phonologie und Informationsstruktur.

Noel Aziz Hanna, P. (forthc. b), Layers of versification in Beowulf. To appear in Anglia.Orton, P. (1999), Anglo-Saxon attitudes to Kuhn’s Law. The Review of Englisch Studies. New Series

50, 287-303.Otfrid von Weißenburg, Evangelienbuch. Edited by O. Erdmann [Altdeutsche Textbibliothek 49]. 6th

ed. by D. Wolff 1973. Tübingen: Niemeyer.Paul, H. Wiehl, P., and Grosse, S. (1989), Mittelhochdeutsche Grammatik. Tübingen: Niemeyer.Paul, O. and I. Glier (1961), Deutsche Metrik. München: Huber.Sievers, E. (1885), Zur rhythmik des germanischen alliterationsverses I und II. Beiträge zur Geschichte  

der deutschen Sprache und Literatur 10, 209­314, 451­545.Standop, E. (1972), Die Metrik auf Abwegen: Eine Kritik der Halle­Keyser­Theorie. In: Linguistische 

Berichte 19, 1–19.Suzuki, S. (1995), Anacrusis inthe meter of Beowulf. Studies in Philology 92, 141­163.Tarlinskaja, M. (1987), Meter and Language: Binary and ternary meters in English and Russian.  In: 

Style 21, 626–649.Thurneysen,  R.  (1892),  Zur Stellung des  Verbums im Altfranzösischen.  Zeitschrift   für romanische 

Philologie 16, 289–307.Vennemann, Th. (1995), Der Zusammenbruch der Quantität im Spätmittelalter und sein Einfluß auf die 

Metrik.   In:   Hans   Fix   (Hg.),  Quantitätsproblematik   und   Metrik:   Greifswalder   Symposion   zur  germanischen   Grammatik,   185­223.  (Amsterdamer   Beiträge   zur   Älteren   Germanistik,   42.) 

15

Page 16: What constitutes anacrusis? Clause openings, Kuhn’s Law ...arsmetrica.elte.hu/articles/noel/0910_noel.pdf · could be stressed (1933: 5),4 he was aware of problems with this phonological

Patrizia Noel

Amsterdam: Rodopi.Wackernagel,   J.   (1892),   Über   ein   Gesetz   der   indogermanischen   Wortstellung.  Indogermanische  

Forschungen 1, 333–436.

16