The Late Medieval Greek Vernacular Πολιτικὸς Στίχος Poetry: A Modern Linguistic Analysis into Intonation Units

In: Journal of Greek Linguistics

The Late Medieval Greek “vernacular” (12th–15th c.) is one of the least studied stages of the history of the Greek language. The lack of interest by linguists can presumably be ascribed to its major source, i.e. metrical πολιτικὸς στίχος poetry. The language of this type of poetry has been labelled a “Kunstsprache”, because of its oral-formulaic character and because of its mixed idiom incorporating vernacular yet also archaizing elements. In this article, however, I demonstrate that the Late Medieval Greek πολιτικὸς στίχος poetry should not automatically be excluded from linguistic research, given that it clearly possesses a strongly vernacular, i.e. spoken, syntactic base: its underlying syntax runs in a very natural way. This is proven by the fact that we can apply the modern linguistic concept of the Intonation Unit, the basic unit of analysis in contemporary spoken(!) languages, to the texts composed in the πολιτικὸς στίχος: far from having an artificial syntax, the πολιτικὸς στίχος poetry is conceptually made up of short, simple “chunks” of information. More precisely, each verse consists of two (stylized) Intonation Units, demarcated by the fixed caesura, which can thus be equated with an Intonation Unit boundary. This thesis is supported by various arguments, both of a metrical and of a syntactico-semantic nature. Arguments belonging to the former category are the length of each half-line, the possibility of stress on the first syllable of each half-line, the origin of the metre, and especially the avoidance of elision at the caesura. In the second category of (syntactico-semantic) evidence, we can consider the tendency of each half-line for constituting a grammatical sense-unit. I also bring forward some little-studied syntactic features of Late Medieval Greek: first, I pay attention to the distribution of the archaizing “Wackernagel particles”, which do not only appear in second position in the verse, but also occur after the first word/constituent following the caesura and thus further confirm my thesis. The same holds for the position of “corrective afterthoughts”, for the verbs and pronominal objects taking the singular are consistently separated from their plural referents by the caesura. Once the Intonation Unit is thus established as a meaningful methodological tool for the analysis of the πολιτικὸς στίχος poetry, the way is cleared for more linguistic research on the Late Medieval Greek vernacular.

Abstract

The Late Medieval Greek “vernacular” (12th–15th c.) is one of the least studied stages of the history of the Greek language. The lack of interest by linguists can presumably be ascribed to its major source, i.e. metrical πολιτικὸς στίχος poetry. The language of this type of poetry has been labelled a “Kunstsprache”, because of its oral-formulaic character and because of its mixed idiom incorporating vernacular yet also archaizing elements. In this article, however, I demonstrate that the Late Medieval Greek πολιτικὸς στίχος poetry should not automatically be excluded from linguistic research, given that it clearly possesses a strongly vernacular, i.e. spoken, syntactic base: its underlying syntax runs in a very natural way. This is proven by the fact that we can apply the modern linguistic concept of the Intonation Unit, the basic unit of analysis in contemporary spoken(!) languages, to the texts composed in the πολιτικὸς στίχος: far from having an artificial syntax, the πολιτικὸς στίχος poetry is conceptually made up of short, simple “chunks” of information. More precisely, each verse consists of two (stylized) Intonation Units, demarcated by the fixed caesura, which can thus be equated with an Intonation Unit boundary. This thesis is supported by various arguments, both of a metrical and of a syntactico-semantic nature. Arguments belonging to the former category are the length of each half-line, the possibility of stress on the first syllable of each half-line, the origin of the metre, and especially the avoidance of elision at the caesura. In the second category of (syntactico-semantic) evidence, we can consider the tendency of each half-line for constituting a grammatical sense-unit. I also bring forward some little-studied syntactic features of Late Medieval Greek: first, I pay attention to the distribution of the archaizing “Wackernagel particles”, which do not only appear in second position in the verse, but also occur after the first word/constituent following the caesura and thus further confirm my thesis. The same holds for the position of “corrective afterthoughts”, for the verbs and pronominal objects taking the singular are consistently separated from their plural referents by the caesura. Once the Intonation Unit is thus established as a meaningful methodological tool for the analysis of the πολιτικὸς στίχος poetry, the way is cleared for more linguistic research on the Late Medieval Greek vernacular.

1. Introduction

The Greek language has the longest documented history of any Indo-European language, covering nearly 3500 years of written records (Joseph 1987). However, not every period has been equally well been investigated. The language of the Late Medieval Greek (LMG) period (12th to 15th century), the so-called “vernacular” (Horrocks 2010:216), constitutes an underexplored area. An important reason for this neglect is the fact that prose texts are very scarce during this period (Browning 1999:75): the only extent literary texts which have been preserved are poems composed in the accentual fifteen-syllabic metre, known as the πολιτικὸς στίχος (politikos stixos). Consequently, if you want to investigate LMG vernacular, you are almost forced to include πολιτικὸς στίχος poetry: the inclusion of πολιτικὸς στίχος texts is thus “the usual practice for compiling the corpus for this period” (Chila-Markopulou 2004:201).

As such, linguists, who hold on to the principle that prose texts constitute the only (reliable) witnesses to the natural spoken language, have usually disregarded these texts, especially because LMG πολιτικὸς στίχος poetry presents two further idiosyncratic hallmarks. First, although the language of the LMG πολιτικὸς στίχος poetry is generally described as “vernacular”, its idiom is actually mixed: linguistic features belonging to the contemporary spoken language are used side by side with classicizing “dead” elements. Therefore, Toufexis (2008:204) labels it a “non-standardized, so-called mixed or macaronic language incorporating vernacular and learned elements.” Secondly, LMG πολιτικὸς στίχος poetry abounds with (quasi-)identical half-lines, so-called “formulas”. This formulaic character has been the main reason for relating this type of poetry to an oral background: the πολιτικὸς στίχος poets are nowadays assumed to have deliberately adopted (features of) an oral style (e.g. Sifakis 2001; E. Jeffreys 2011). As a consequence, we should not be unduly surprised that this metrical, mixed and oral-formulaic character has withheld linguists from thoroughly investigating the πολιτικὸς στίχος poetry, in the conviction that its discourse is artificial or even distorted.

Nonetheless, in this article, I show that the underlying discourse of the LMG πολιτικὸς στίχος poetry is definitely natural: it possesses a strongly vernacular, i.e. spoken, syntactic base. For this purpose, I rely on a very popular concept within modern linguistics, which has been developed on the basis of modern spoken languages in order to emphasize that we do not speak in complex syntactic sentences, but in short “chunks” of information, i.e. “Intonation Units” (IUs) (Chafe 1993). I argue that we can apply this concept to the πολιτικὸς στίχος poetry, which lends itself naturally to a division into IUs. More precisely, we can equate the fixed caesura after the eighth syllable with an IU boundary and the two standard half-lines of respectively eight and seven syllables thus correspond to two (stylized) IUs.

To prove this syntactic articulation in IUs, I define the modern concept of the Intonation Unit in §4, after briefly discussing in §3 the precise texts which make up my corpus. I then list various arguments from a metrical and from a syntactico-semantic nature in §5, which constitutes the core of this paper. Finally, in §6, I formulate my conclusions. I begin, however, in §2, by going more deeply into the two peculiar yet problematic aspects of LMG vernacular as preserved in the πολιτικὸς στίχος poetry: its mixed idiom (§2.1) and its oral-formulaic character (§2.2), which resulted in a shortage of studies adopting a modern linguistic perspective (§2.3).

2. The Late Medieval Greek Vernacular

2.1. Mixed Idiom

From the 12th century on, the Greeks throw off the classicizing yoke and start to write literature “in an approximation of the spoken language of the period” (Joseph 1990:5; see Hinterberger 2006:1): “Die Verfasser haben jetzt das rhetorische, archaisierende Griechisch aufgegeben und sind dazu übergegangen, in der Volkssprache zu schreiben” (Rosenqvist 2007:170; see Cupane 2003:577). This “break-through of the vernacular tongue into literature” (Browning 1999:72ff.) is associated with the use of the fifteen-syllable πολιτικὸς στίχος. The connection between both is so inextricable, that Beaton (1996:99) even calls the πολιτικὸς στίχος “the natural medium for vernacular narrative”.

Nonetheless, the appropriateness of the term “vernacular” to describe the language of the LMG πολιτικὸς στίχος poetry has been discussed in Cupane (2003) and Hinterberger (2006), for it cannot simply be equated with the truly spoken tongue. As a matter of fact, the poets still include ancient, learned forms which have long died out in the spoken language (such as the particles; see infra: §5.3.2.).1 Hence, the term “Mischsprache” (Hinterberger 1993:158) has been used to portray this language extremely rich in variants: “The availability of ancient and modern alternatives in vocabulary, morphology, and syntax accounts for what is meant by the ‘mixed’ character of medieval Greek” (Mackridge 2000:134). A prototypical example is the verbal ending of the third person plural: the two-syllable “ancient” ending -ουσι(ν) (-ousi(n)) alternates with the modern monosyllabic -ουν (-oun) (Browning 1999:81; Hinterberger 2001:218; Horrocks 2010:319).2

2.2. Oral-Formulaic Character

To complicate matters, this “hotchpotch” is peppered with (quasi-)identical half-lines, so-called “formulas”. An example of a famous intra- and intertextual formula in my corpus is “μικροί τε καὶ μεγάλοι” (“small ones and big ones”):3

article image

Given that formulas are a phenomenon typical of oral poetry, the LMG πολιτικὸς στίχος poems have been related with an oral tradition, just like the Homeric epics, which are acknowledged to be the result of an oral tradition: “The most convincing evidence that this group of fourteenth-century texts has been produced against a background in oral poetry takes the form of a high ‘formula’ count in certain of them” (E. Jeffreys 2011:470).4 Two extratextual characteristics have reinforced this association, namely the anonymous character of the texts and their fluid textual tradition (see 3.2). However, unlike the Homeric situation, oral composition is nowadays definitely excluded: “The style markers of orality that can be identified in the works of Medieval Geek, so-called vernacular, poetry are not to be taken as indications of an earlier phase of oral composition of such works—for which there is not a shred of evidence—but as signs of the style of traditional folk poetry adopted, to a greater or lesser degree, by literate poets” (Sifakis 2001:67). Indeed, there nowadays exists “a tacit acceptance that the stylistic features and peculiarities of this group of late Byzantine verse texts are best explained against a background of orally composed and orally disseminated poetry” (E. Jeffreys 2011:470).5 In sum, whereas the Homeric epics (and formulas) are a product of “primary orality”, i.e. of a pre-literature society, the adopted oral style of LMG poetry clearly belongs to “secondary orality”. The oral style markers, of which the formulas are the representatives par excellence, are deliberately and self-consciously adopted in texts composed in a literate culture.

As a matter of fact, we cannot even say with certainty that the LMG poetry has been orally performed. Oral recitation of the texts under investigation is considered rather unlikely by most researchers, although they admit that it might depend on the specific text (Agapitos 2004:43) (see 3.2: footnote 13, on the Oxford manuscript of the Achilleis Byzantina).6 This suggests that we should use the term “oral” with some care, for it incorporates different senses.

2.2.1. Conception vs. Medium

Foremost, rather than maintaining a simple mono-dimensional dichotomy (oral vs. written), we should discern the medial sense of “oral” from the conceptional one—or in Shawcross’s words: “a distinction should be made between the physical means by which a piece of literature is composed and the type of discourse employed in that composition” (Shawcross 2005:312). Especially Oesterreicher’s work is revealing in this respect. In a series of articles, Oesterreicher convincingly shows that when dealing with oral/written issues, we should distinguish the medium, to which we can still apply the dichotomy oral (phonic) versus written (graphic), from the linguistic conception of a text (Koch & Oesterreicher 1985; Oesterreicher 1997). With regard to the latter, a continuum constitutes a better way of representation and “orality” in this sense should thus be considered a gradient property: from more spoken to less spoken, with intermediate grades, or from “intimate” to “distant” in Oesterreicher’s terminology, with a scale from “Sprache der Nähe” to “Sprache der Distanz”. As such, a conceptionally oral text “is often medially oral as well, but it is also possible for such a discourse to be written” (Bakker 1997:8), which is the case with regard to the πολιτικὸς στίχος poems, which may then be “oral to a greater or lesser degree, depending on the nature of the conception underlying them” (Bakker 1997:9; see Bakker 1999:30).

2.2.2. Oral ≠ Spoken

However, “oral”, in the sense in which it has been used by Byzantinists, cannot be straightforwardly equated with (more or less) “spoken”: in the above statements by Sifakis and E. Jeffreys, “oral” mainly refers to a stylistic issue (i.e. “formulaic”), rather than to the truly spoken language. Therefore, although an oral style obviously entails many features of the spoken language, “spoken” and “oral” can here not be considered mere synonyms. The above-mentioned formula testifies to this: μικροί τε καὶ μεγάλοι contains the archaizing particle τε, which beyond any doubt no longer belongs to the spoken language at this point. This example demonstrates that not everything oral should automatically be considered vernacular.

2.3. The Linguistic Neglect of Such a “Kunstsprache”

The side-by-side occurrence of vernacular characteristics and archaizing, learned features leads us back to the so-called “Mischstil” (Eideneier 1991:26) of the texts under scrutiny: the many variants, often differing in number of syllables, have been accounted for from this “oral” perspective. More concretely, it is believed that the availability of ancient and modern alternatives must have simplified the composition of fifteen-syllabic verses: “This freedom of choice provided the writer with considerable linguistic flexibility and greatly facilitated the process of verse composition; for instance, where two alternatives were available, one of them might fit the metre and the other might not” (Mackridge 2000:134). A number of scholars have even been tempted to conclude that the choice between certain forms mainly depends on metrical considerations and accordingly emphasizes the (formative) influence of the metre on the language found in the poems. As a result, they speak of a “Koine der Dichtersanger” (Eideneier 1983:132), an “art-language” or a “Kunstsprache” (see Beaton 1996:184ff.). This art-language then preserves obsolete forms under the pressure of having a language adapted to the needs of versification. Eideneier (1983:236) even claims that in this Kunstsprache “die poetische Syntax geht vor der grammatischen Syntax”.

On the other hand, some scholars have counterbalanced this idea by stressing that although they do not deny the functionality of variants for metrical purposes, many of them are genuine—at least in different parts of the Greek empire—and thus inherent to the actual spoken language of the time (even if many of these forms are nowadays no longer found in Standard Modern Greek). Especially Hinterberger (1993:165) is a fervent supporter of this view: he adheres to the “These, dass die Sprache der Volksliteratur weder eine künstliche Sprache noch eine Mischsprache ist.” In a more recent article, he repeats: “η πολυτυπία καθαυτήν δεν είναι στοιχείο μιας Kunstsprache (έντεχνης γλώσσας) ούτε τεχνητής γλώσσας” (Hinterberger 2001:227); the variation is thus pre-existent: “η πολυτυπία (…) προυπάρχει και στον απλό γραπτό (και προφορικό) λόγο της εποχής” (Hinterberger 2001:215). With regard to the above-mentioned pair (-ουσι(ν)/-ουν; see §2.1.), Hinterberger (2006:4) appeals to dialectal evidence to prove that -ουσι(ν) must have belonged to the spoken language of the time: “Also the apparently archaizing verb endings -ουσιν and -ασιν were also used in medieval spoken language as they are today e.g. in Cypriot Greek, besides the “normal” endings -ουν and -αν.”

However, it seems that the damage had already been done: the field of LMG linguistics is still in its infancy. Whether or not explicitly, many scholars have judged the πολιτικὸς στίχος poetry unqualified for thorough linguistic research, considering prose as our only reliable witness to the spoken language. The following statements all testify to this widespread idea:

Unfortunately, he [Pappas 2004] only uses mainly poetic works due to difficulties in finding prose texts of the period he examines (12th–16th), admittedly a problem if one wants to say something about the language of the period.

Thoma 2007:142

Another point (…) concerns the extent to which the corpus is as wide and as representative as possible, since it is restricted to vernacular texts of LMG written in verse. As a consequence, the reliability of the statistical results is compromised.

Chila-Markopoulou 2004:201

I have chosen to study prose texts in order to avoid the question whether considerations of metre may have influenced the choices made by authors and copyists.

Vejleskov 2005:198

One strategy has been to accord greater weight to the evidence of prose texts over poetic ones for showing “real” features of the spoken language, the assumption being that part of the poetic process involves stretching grammatical and lexical boundaries.

Joseph 2000:312

Nevertheless, I am convinced that the underlying discourse of the πολιτικὸς στίχος poetry reflects the actual spoken language: far from radically excluding “metri causa” as a valuable argument to explain certain morphological doublets, I demonstrate here that the syntax of the πολιτικὸς στίχος poetry is not artificial at all. On the contrary, it possesses a strongly vernacular, i.e. spoken, syntactic base. More concretely, the πολιτικὸς στίχος texts are actually conceived as a chain of “Intonation Units” (“IUs”), which are the basic units to analyze modern spoken (!) languages. I turn first, though, to the nature of my corpus and the editorial problems posed by the πολιτικὸς στίχος poetry.

3. Corpus

3.1. The Romances and a Chronicle

My corpus contains eight lengthy πολιτικὸς στίχος texts, all anonymous and all dating from the 13th–15th centuries (the Palaeologan period), i.e. the heyday of LMG vernacular poetry.7 The majority belongs to the genre of the romance, which is the genre par excellence in the Middle Ages (Beaton 1996): Achilleïs Byzantina (AB); Belthandrus & Chrysandza (BC); Ilias Byzantina (IB); Imperius & Margarona (IM); Libistrus & Rodamne (LR) and Phlorius & Platzia Phlora (PP).8 It should be noted that drawing up an exhaustive inventory of the LMG romances is a more arbitrary task than one might think, as Agapitos (2004:12) testifies: “Let us move on to the texts I have been referring to as Byzantine vernacular romances. How many are they and which exactly? The seemingly simple question proves more difficult than one would have expected (…) They float around the general heading ‘romance of chivalry’.”

Although the War of Troy (BT9) has long been neglected, I have also taken this lengthy text into consideration: “There is also another text which (…) despite clearly belonging with the group, has been almost entirely ignored in discussions of these romances. This is the War of Troy (…) it shares the romances’ characteristic features: it is anonymous, in the fifteen-syllable line, with repeated phrases and a fluid textual tradition, and uses a form of the vernacular” (E. Jeffreys 2013:224). Finally, I have also included a text which is traditionally considered among the genre of the chronicle, but which actually meets the same (formal) characteristics as the romances: the Chronicle of Morea (Morea).

3.2. Parallel Versions Leading to Editorial Problems

Admittedly, the selection of these texts might seem a quite arbitrary task. Moreover, as mentioned above (2.2), one of the (extratextual) characteristics of the LMG πολιτικὸς στίχος poetry is a fluid textual tradition. Indeed, many LMG πολιτικὸς στίχος texts survive in a number of manuscripts which differ to such an extent that it is “impossible to collate the readings into one primary version by conventional editorial principles” (E. Jeffreys 2011:469). Agapitos (2006:105) calls this “ἡ ἀνεπάρκεια τῆς στεμματικῆς μεθόδου”. Hence, we best speak of different “versions”, which all have their own validity and require “einer unterschiedlichen editorischen Behandlung” (Moennig 2004:217). The existence of such different versions of one and the same story poses a serious challenge for editors. Consequently, editorial techniques have been problematized (see the discussions in Eideneier, Moennig & Toufexis 2001). A definitive solution is still wanting, as E. Jeffreys (2013:220) has very recently observed: “a consensus has yet to emerge over appropriate editorial solutions (…) if there is a consensus of any sort it is that each text has to be treated as a special case and there are no universally valid solutions.”

With regard to my corpus, BC and IB are preserved in one sole manuscript and accordingly pose few problems. In these cases, I have decided to rely on the most recent edition: Egea (1998) in the case of the former text (Par. gr. 2909; 1350 verses) and Nørgaard & Smith (1975) in the case of the latter (Par. suppl. gr. 926; 1166 verses). However, AB, BT, IM, LR, PP and Morea survive in parallel versions. Here, the question arises regarding which manuscript(s) and which edition we should use.

In the case of Morea, the answer is straightforward: it is generally established that the Havniensis 57 (H) manuscript is “unquestionably the earliest and most authentic” (Lurier 1964:33) of the five manuscripts in which the Greek version of the Chronicle has been handed down to us (see Lurier 1964; Shawcross 2009).10 I have relied on the edition of Schmitt (1904), which is “old but nevertheless reliable” (Aerts 2005:142).11

BT has been preserved in no less than seven manuscripts (Papathomopoulos & E. Jeffreys 1996:xciii–civ). However, thus far, only one edition has been published, presumably because of the text’s length (14,401 verses!): the edition of Papathomopoulos & E. Jeffreys (1996). Papathomopoulos & E. Jeffreys (1996) thought it best to combine all manuscripts using a Lachmannian stemma.

Two manuscripts survive containing the story of PP (Lond. Add. 8241 & Vindob. Theol. gr. 244). The most recent edition is the one by Salas (1998a), who bases his edition on the Viennese version (1867 verses) after a thorough manuscript study.12

With regard to LR, which has been attested in four manuscripts, Agapitos (2006) has combined manuscripts according to his understanding of the author’s language, which has resulted in text “α”. However, I have given preference to the edition of Lendari (2007), who presents a single manuscript, i.e. the Vatican one (Vat. gr. 2391; 4013 verses), which is “our fullest witness” (Beaton 1996:106).

I have treated AB in a similar way: I have chosen to concentrate on the Naples version (Neapol. gr. III.B.27; 1926 verses), edited by Agapitos, Hult & Smith (1999), because it represents “the most coherent and most complex realization of the Byzantine romance of Achilles” (Agapitos 2006:158).13

Finally, with regard to IM, I have applied the same principle. Of the (no less than) five manuscripts in which IM has been handed down, I have relied on the one containing the fullest version of the story, i.e. the Naples manuscript (Neapol. graec. III Β 27), which seems Kriaras’ base manuscript (1955; 893 verses):14 “It is preserved in whole or in part in five manuscripts, of which the earliest is the Naples manuscript that also contains the fullest version of the Tale of Achilles” (Beaton 1996:140).

The aforementioned editions can all be integrally found on the Thesaurus Linguae Graecae (TLG) website, the comprehensive database of Ancient and Medieval Greek texts, which greatly facilitates queries. However, since the TLG does not include the critical apparatus of the editions, we cannot solely rely on it. Rather, we should check the printed editions or preferably even the manuscripts, to exclude with certainty the possibility of editorial intervention. Nonetheless, the reliance on editions has become a common practice with regard to LMG poetry, as Manolessou (2008:68) criticizes: “There is virtually no linguistic research involving Medieval Greek that does not take as its base critical editions of literary texts.” Manolessou (2008:70) goes on by raising the need of new editions: “if the language of literary texts is to be studied thoroughly and reliably, new editions/transcriptions of the texts are necessary, not for the general reader, but for the linguist.” Until the publication of new editions, however, linguists should in my opinion not stand by doing nothing: observations of little-studied linguistic peculiarities (such as the LMG distribution of P2 particles and the phenomenon of “corrective afterthoughts”; see §§5.3.2, 5.3.3) might actually help future editors.

In sum, despite the textual problems, these eight texts constitute a representative corpus of LMG vernacular poetry, totaling almost 35,000 πολιτικοὶ στίχοι.

4. Concept of the Intonation Unit

4.1. Definition

The concept of the IU has principally been elaborated by the American linguist Wallace Chafe in a series of articles (1987, 1988, 1993, 1994, 1998, 2001). Chafe has noticed that natural spoken language does normally not consist of complex grammatical sentences, but of short “chunks” of information: so-called IUs. As its name announces, an IU is primarily defined from a prosodic perspective: it is “a stretch of speech occurring under a single unified intonation contour” (Chafe 1987:22). This sequence consists of an average of four (full) words.15 The criterion which is most often mentioned as demarcating two successive IUs is the breathing pause, but other prosodic clues, such as pitch resetting and final lengthening, also exist (Cruttenden 1997:30).16 Semantically, an IU usually expresses one idea and thus also constitutes a sense-unit. This is the reason why Chafe speaks of “idea units” in his early work.17 Chafe has related the concept to the cognitive side of communication: an IU presents the (limited) amount of information our mind can focus on (both as a speaker and as a listener). From a syntactic point of view, IUs tend to coincide with grammatical units (but do not need to): “These bursts often consist of a single grammatical phrase or clause” (Johnstone 2008:68–71; see Scheppers 2011:28).18 However, one of the most important consequences of Chafe’s findings is that the traditional notion of “sentence” loses much of its importance (see Bakker 1990:3). Despite the non-categorical character of these criteria, the IU is considered a “well established unit of spoken discourse” (Tao 1996:11).

As for the linkage of IUs, linguists have observed that a sequence of IUs generally develops in a simple, non-elaborate manner: in English “fully 50% of the cases with explicit connectives consist of or include the maximally general connective and” (Chafe 1988:10).19 In other words, (complex) subordination is rare in natural spoken language. The information flow can thus be said to progress linearly, i.e. from one chunk to the other, or, applied to written texts from left to right.

4.2. Studied Languages (Including Homeric Greek)

Most research on the IU has concentrated on English data (Chafe 1993, 1994; Croft 1995; Cruttenden 1997). In the last decade, however, a variety of other modern languages are taken into account as well; including Japanese (Clancy, Suzuki, Tao & Thompson 1996; Iwasaki 1993; Matsumoto 2000, 2003), Chinese Mandarin (Tao 1996); Taiwanese Mandarin (Tseng 2006, 2008), Finnish (Helasvuo 2001); Korean (Kim 1999) and Sasak (Wouk 2008).

Moreover, the concept has already been applied to corpus languages, including Ancient Greek. Especially with regard to the poetry of Homer, an analysis in terms of IUs has been successful. As mentioned above (§2.2), it has been acknowledged that the Iliad and Odyssey are the written result of an oral tradition, as is especially reflected in the texts’ highly formulaic character. The dactylic hexameter, the metre of the Homeric epics, is divided into two or more parts by so-called “caesurae”. Traditionally, these parts, labeled “cola”, were considered purely metrical phenomena, having little to do with natural spoken discourse. However, the studies of Fraenkel (1933) and Janse (1998, 2003) have shown that caesurae actually imply breathing boundaries and that cola thus coincide with the chunks of spoken discourse identified by modern linguists, i.e. IUs. In addition, Bakker (1997), who explicitly relates Homeric poetry to spoken language in his book entitled “Poetry in Speech”, has also contributed to a more natural conception of the dactylic hexameter poetry. He fervently argues for the abandonment of the concept of “sentence” in favour of the IU: “The application to the Homeric style of the concept of the fragmentation of oral narrative into idea units has, I think, an immediate appeal” (Bakker 1990:5). Of course, Bakker (1997) realizes that we cannot simply equate the Homeric discourse with natural speech and therefore speaks of “stylized” IUs (Bakker 1997), in the belief that “the principle basically applies” (Bakker 1990:19). Bakker (1997:304) even associates the notion of the IU with the formula: “Formulas derive from the very nature of spoken language, as a regularization of its basic segment, the cognitively determined intonation unit.”20

These Homeric studies can to a certain extent be considered in parallel with my research on the LMG πολιτικὸς στίχος poetry: the Homeric language has also been considered an artificial, mixed language, employing a mixture of the Ionian and Aeolian dialects. Meister’s (1921) monograph is for instance entitled “Die Homerische Kunstsprache”. However, by successfully applying the IU to the dactylic hexameter poetry, scholars have proven that the “flow” of information (in other words, the syntax) runs in a natural, spoken way. This insight has been an important methodological achievement and has led to numerous illuminating discourse-related studies on the Homeric epics (Bakker 1990, 1993, 1997; Slings 1992, 1994; Janse 1998; Allan 2009).

My aim is to achieve the same results with regard to the LMG πολιτικὸς στίχος poetry, our major source for LMG vernacular: in the next section (§5), I apply the concept of the IU to the πολιτικὸς στίχος poetry in order to show that its syntax runs in a very natural way. This leads to a more natural conception of the πολιτικὸς στίχος poetry, which in turn might encourage (discourse-oriented) linguistic research.21

5. An Analysis of the πολιτικὸς στίχος into Intonation Units

5.1. General Information on the πολιτικὸς στίχος

5.1.1. Some Basic Facts

The πολιτικὸς στίχος is based on the pronounced word accent instead of on the no longer existing difference between long and short vowels as in classical metres (which are, however, still artificially used in the Early Medieval Greek period). It consists of fifteen syllables (hence also “dekapentasyllable”), of which in principle only the even ones can be stressed (Lauxtermann 1999). Given that this iambic pattern suits the Greek language very well, the πολιτικὸς στίχος is considered a very natural metre. Rosenqvist (2007:113), for instance, writes the following, when talking about the 12th-century epic Digenis Akritis: “Es ist in “politischen Versen” abgefasst, einem Versmaß, welches der natürlichen Aussprache und dem natürlichen Rhythmus der Volkssprache gut angepasst ist.”22 Furthermore, each verse has a fixed break or so-called “caesura” after the eighth syllable (from now on marked by the sign #), which divides the verse into two half-lines (also called “hemistichs”) of respectively eight and seven syllables.

5.1.2. Starting Point: The Pointlessness of the Concept “Sentence”

When reading LMG πολιτικὸς στίχος poetry, it attracts attention that the concept “sentence” does not make much sense, as the following passage shows:

article image

The insertion of a full stop is an arbitrary affair: the concept “sentence” becomes useless, as Bakker has noted with regard to the Homeric epics: “Instead of sentential arrangements, we see a relation of addition between units” (Bakker 1999:41). Rather, we should conceive these texts as a concatenation of IUs. More concretely, in what follows, I argue that not only the end of each verse constitutes a breathing pause (which is fairly evident) but also that the fixed “caesura” implies a breathing boundary, which is one of the most often-cited criteria to demarcate IUs in modern spoken languages (see 4.1). Accordingly, we can equate the two standard half-lines with two (stylized) IUs.

Interestingly, the need of a unit of speech for an analysis of the πολιτικὸς στίχος texts has already been pointed at by Eideneier (1989:180): “Es ist ja nicht zu verkennen, daß auch Texte im einfachen Erzählstil rhythmischen Gegebenheiten der Spracher unterliegen.” Moreover, the correspondence between a standard metrical unit (i.e. half-line) and an “information unit” (i.e. IU) seems not at all exceptional with regard to medieval “oral” poetry: “Among oral narrative forms of the Middle Ages, the most rigorously paratactic are the verse genres composed formulaically in performance. Here parataxis is the result of prosodic constraints and the desirability, for both the poet and audience, of having information blocks correspond to metrically regular units” (Fleischman 1985:876: footnote 32; my italics). However, to be perfectly clear, I do not want to pronounce upon the medium of the πολιτικὸς στίχος poetry: the discussion below should not be read as an argument in favour of an oral performance (see §2.2). All I assert is that the underlying conception of the texts is a syntactic articulation in IUs.

In order to justify the hypothesis that the fixed caesura constitutes an IU boundary and that the two standard half-lines can thus be compared with two (stylized) IUs, I appeal to metrical (§5.2) and syntactico-semantic (§5.3) arguments. The latter category is subdivided according to three aspects: the tendency of each half-line to correspond to a grammatical sense-unit (§5.3.1), the distribution of the so-called P2 particles (§5.3.2), and the position of corrective afterthoughts (§5.3.3). This last phenomenon is given special attention, given that it has not yet been described.

5.2. Metrical Arguments

By far the most important “metrical” argument why the fixed caesura implies a breathing boundary is the fact that elision (the omission of a vowel) is avoided between the eighth and the ninth syllable, whereas it is allowed elsewhere (Apostolopoulos 1984:211; Lendari 2007:132), for instance:

article image

The opposite of elision, i.e. hiatus, occurs between the vowels of the eighth (αι in the above example) and the ninth (α) syllable. While the presence of elision seems to exclude the possibility of a breathing pause, hiatus is a signal of discontinuous speech and thus of a possible breathing pause. As such, it is reasonable to suggest that the fixed caesura, just like the verse-end, constitutes a breathing boundary.

As just mentioned, each πολιτικὸς στίχος contains fifteen syllables, of which in principle only the even ones can be stressed. Occasionally, however, the (uneven) first and (uneven) ninth syllable may be stressed as well (Apostolopoulos 1984:214). Inasmuch as the ninth is the first syllable following the caesura, this suggests a possible identical beginning of the two half-lines:

article image

Eideneier (1999:104) even relates the length of the half-verses to our average breathing capacity: “Wenn wir von einem menschlichen Atemvolumen für den Vortrag von Versen zwischen 12 und 17 Silben ausgehen (…) ist eine solche Mittelzäsur eine zusätzliche Möglichkeit zur Sinn-gliederung und Pausenmarkierung.” This assumption is supported by the presumable origin of the πολιτικὸς στίχος: a combination of two metres, namely an octosyllable and a heptasyllable (Lauxtermann 1999).

5.3. Syntactico-Semantic Arguments

5.3.1. Grammatical Sense-Unit

Another argument in favour of an interpretation of the πολιτικὸς στίχος as two IUs is its straightforward syntactic structure. A strong enjambment is not only rare between two successive verses, it also hardly occurs between the two standard half-lines (Lendari 2007:132).23 As a consequence, most half-lines represent an independent, autonomous entity, i.e. a syntactic unit by itself. Predominantly, this short unit is a clause or a phrase, for instance:

article image

This observation corresponds to the numerous inquiries which seek the “grammatical equivalent” of the IU in various modern languages (see §4.1).

Moreover, these units do not only constitute a grammatical entity, but also a sense-unit: “each half-line comprises a self-contained unit, in terms of syntax and sense (…) As a general rule, a line of political verse consists of two units” (Beaton 1980:44). Accordingly, the formulaic phrases of the πολιτικὸς στίχος (see §2.2), which are traditionally defined as “prefrabricated sense-blocks”, typically cover one half-line: “the formula must fill either the first or the second half of the political line, the popular meter of early Demotic Greek poetry. It must be either eight or seven syllables long respectively” (M. Jeffreys 1973:175).24

5.3.2. P2 Particles

Perhaps surprisingly, small words such as the particles γάρ (gar), δέ (de), μέν (men) and οὖν (oun) can also provide further evidence for our hypothesis. The occurrence of these particles in πολιτικὸς στίχος poetry constitutes an ideal example of its so-called “mixed” language (see 2.1.), for γάρ, δέ, μέν and οὖν have disappeared from the contemporary spoken language and are thus to be considered “archaic relics” (see Jannaris 1897:400). Given that the particles’ original (multifunctional) discourse role has been reduced, some scholars claim that their presence in πολιτικὸς στίχος poetry is solely driven by metrical reasons: “They [the particles in BT] seem rarely to have a distinctive meaning and are often blatant line fillers” (Papathomopoulos & Jeffreys 1996:lxxvi).25

Nonetheless, the particles (underlined from now on) are not inserted randomly in my corpus. On the contrary, they still follow the Indo-European Wackernagel “Law” (1892), which posits that they should appear after the first word/constituent in the IU (hence “P2 particles”; Janse 1998). Consequently, the P2 particles can serve as diagnostic means to identify IUs: we can derive the beginning of a new IU from their position. As expected, the majority of the particles (84.4%) in my corpus is found after the first word/constituent in the verse, for instance:

article image

However, they also appear after the first word/constituent following the caesura, though less frequently (13.1%), for example:

article image

Both placements can also co-occur:

article image

This results in the following statistics:

Table 1Distribution of P2 particles γάρ, δέ, μέν and οὖν in the eight texts of my corpusTable 1

In this extent, the work by Loudová (2007, 2009), one of the few scholars who has examined the particles in Medieval Greek, should be mentioned: Loudová has already identified the “rhythmical” function of the P2 particles with regard to the Early Medieval Greek chronicles (e.g. 6th century Malalas). She explicitly relates the position of the particles to “rhythmical boundaries”: “I would like to point out one more function of the particles, which is linked closely with their ability to distinguish the discourse boundaries. It seems that thanks to it the particles ensured rhythmical segmentation of the texts intended for recitation (…) This way we can at least partly explain apparent abundance of the particles in the narrative texts where the particles separating the phrasal units occasionally do not even take into consideration sentence syntactic articulation. They served presumably not only as text dividing means while reading but also during the recitation they indicated pauses” (Loudová 2007:7; my italics). Loudová thus acknowledges two different functions of P2 particles, i.e. signalling discourse boundaries and indicating recitation pauses, but she does not make the logical conclusion that both functions coincide in the concept of the IU (boundary). In the same vein, we should also interpret the following statement by Eideneier (1989:190): “Vorlesepraxis (…) da die Partikel δέ die Pause eindeutig hörbar macht.”

Once again, however, I want to emphasize that I do not draw conclusions about the reality of an oral performance (see §2.2). The distribution of the particles, more precisely that they do not only occur in P2 in the first half-line, but also in the second half-line, merely shows that the two half-lines are conceived as IUs.

5.3.3. Corrective Afterthoughts

A final argument for my hypothesis is furnished by the position of so-called “corrective afterthoughts”. Given that this phenomenon has largely remained unnoticed, I elaborate upon it here. I have selected a sample of verses involving an afterthought, which has led to a lack of grammatical agreement, either between the verb and its subject (§5.3.3.1) or between the pronominal object and its referent (§5.3.3.2).26 We see that the precise position of the corrective afterthoughts supports the interpretation of the caesura in terms of an IU boundary (§5.3.3.3).

Before continuing, I should remark that I have verified the critical apparatus with regard to all the below mentioned examples as to ensure that the lack of grammatical agreement is not from the pen of the editor(s). However, it is rather unlikely that the editions are unreliable in this respect. As a matter of fact, the opposite is much more plausible, namely that editors have “corrected” the “mistakes” concerning agreement and that more examples of corrective afterthoughts are thus to be found in the manuscripts …

5.3.3.1. Lack of Agreement between Verb and Subject

In principle, verbs must agree with their subjects in person (first, second or third) and number (singular or plural). In the following examples, however, the verbal ending does not show morphological agreement with its subject: the verb, which stands in the first half-line, remains singular, while either an additional subject (§5.3.3.1.1) or the multiple subject as a whole (§5.3.3.1.2) is added after the fixed caesura.

Before turning to the examples, it should be noted that the lack of verbal agreement is not always evident in my English translation, for English possesses fewer morphological distinctions than Greek, for instance: the third person singular (e.g. “(I) speak/spoke”) is identical to the third person plural (e.g. “(they) speak/spoke”). Certainly with regard to the subjects added as whole (5.3.3.1.2), it is difficult to render the Greek faithfully, given that English requires that the subject be explicitly expressed. Indeed, as opposed to English, Greek is a so-called “pro-drop” language: it is not necessary to explicitly express the subject in Greek. I have attempted to solve this inconvenience by adding a (provisional) pronoun in my translation between square brackets.

5.3.3.1.1. Additional Subject Added

Let me begin by citing some verses in which an extra subject is added to a preverbal subject:

article image

According to our written norms,27 we would have expected a plural verb ἐθαύμασαν (which corresponds to the singular ἐθαύμασεν as regards number of syllables and accentuation and thus has the same metrical value)—or even a different word order: “the dux Aeneas and all the nobles came.” This also holds for the next examples in which an extra subject is appended to a postverbal subject:

article image

Note that the additional subject is always attached by means of καί, which immediately follows the fixed caesura. Kαί is considered the coordinator par excellence in the Greek Middle Ages (Eideneier 1999; see Toufexis & Thoma 2004). Indeed, in my corpus as well, it occurs extremely frequently, i.e. 14,195 times. Kαί often appears in initial position in the verse (5263 instances) or immediately after the fixed caesura (3927 instances) (see §5.1.2: LR 145–147). In other words: it tends to open the two half-lines (9190 times = almost 65% of the total). In this connection, it is very interesting to note that Eideneier (1999:116) has called καί the “Auftaktpartikel für eine rhythmische Phrase”. Toufexis & Thoma (2004) even explicitly use the term “Intonation Unit marker” to describe (a part of) the functioning of καί in LMG vernacular.28

5.3.3.1.2. Subject Added as a Whole

A composite subject can also as a whole be adjoined; note that the first part typically involves demonstrative pronouns such as αὐτός and ἐκεῖνος:

article image

5.3.3.2. Lack of Agreement between Pronominal Object and Referent

Another structure falling under the construction of the corrective afterthought is the lack of agreement between pronominal objects (singular) and their actual referents (plural):29

article image
article image

5.3.3.3. Breathing Boundary between Thought and Afterthought

In modern linguistics, it has been acknowledged that corrective afterthoughts are usually accompanied by a breathing pause, namely between the actual thought (the “reparandum” in the words of Geluykens 1994:182) and the afterthought, which allows the speaker some time to think and to correct himself—in the words of Lambrecht (1981:76): “afterthought constructions are often preceded by a pause” (see Janse 2008:170).30

With regard to the above examples, it attracts notice that all the corrective afterthoughts are consistently attached immediately after the fixed caesura. Schematically, this results in the following structures: “S+ V sg # καί S” or “V sg + S # καί S” (§5.3.3.1.1), “V sg # S καί S” (§5.3.3.1.2) and “Pronominal Object sg # O καί O” (§5.3.3.2). A second possibility, however, is the addition of the afterthought immediately after the verse-end, which has been acknowledged to constitute an IU boundary (§5.1.2):

article image

In sum, it attracts the attention that the afterthoughts are never inserted inside a half-line. Again, these observations lead one to suspect that the fixed caesura constitutes an IU boundary. Therefore, the corrective afterthoughts can be cited as further evidence for the thesis that we can compare the two half-lines of the πολιτικὸς στίχος with modern IUs. Furthermore, the corrective afterthoughts obviously point to (the adoption of) a spoken discourse (see §2.2): they are best explained as the result of the linear progression of speech, which runs from one IU to the other, or it is applied to written texts from left to right (see §4.1). We see that in all the above examples the corrective afterthought is added to the right side of the verb/pronominal object. It seems that the poet has already pronounced the verb/pronominal object (in singular) before realizing its referent is actually multiple. However, as rewriting was not possible (or at least such is the impression he gives), the poet correctively adds (a part of) the subject/object, often making use of the popular coordinator καί. In this extent, it is significant to note that I have not encountered examples in which the opposite happens, i.e. a verb with a plural ending having a singular subject or a plural pronominal object having a singular referent; nor have I found examples in which a multiple subject precedes and stands thus at the left side of the verb put in singular. As such, it might have become clear that with regard to my corpus in particular, yet also with regard to the πολιτικὸς στίχος poetry in general, the notion of “sentence” is best abandoned in favour of the IU, which points to the naturalness of its underlying conception.

6. Conclusion

Presumably due to its label of “Kunstsprache”, resulting from its mixed idiom and its oral-formulaic character, the LMG πολιτικὸς στίχος poetry, which constitutes our main source for LMG vernacular, has not received much attention from linguists. Nonetheless, the syntax of this type of poetry does not run in an artificial way. On the contrary, it parallels the natural, i.e. spoken, language. This is proven by the fact that we can apply the modern linguistic concept of the IU, the basic unit of spoken (!) languages, to the LMG πολιτικὸς στίχος texts. More precisely, I have argued that the fixed caesura after the eighth syllable implies a breathing boundary, which is one of the most important criteria to demarcate subsequent IUs, and that the two standard half-lines of respectively eight and seven syllables can thus be equated with (stylized) IUs.

From a metrical point of view, especially the avoidance of elision at the caesura can be invoked as an argument for this thesis. Further confirmation is given by different “syntactico-semantic” phenomena: beside the fact that each half-line usually consists of a grammatical sense-unit, the position of both P2 particles and of corrective afterthoughts are strong arguments in favour of our hypothesis. Although the P2 particles have long since disappeared from spoken Greek, they still occur in the πολιτικὸς στίχος poetry and still obey Wackernagel’s Law. As such, they are diagnostic of IU boundaries. Significantly, the P2 particles do not only occur in second position in the verse but also sporadically appear after the first word/constituent following the caesura. Finally, I focused on the little-studied phenomenon of corrective afterthoughts (both in the form of subjects and of objects), which are in modern spoken languages usually preceded by a breathing pause. Their distribution as well corroborates our analysis of the πολιτικὸς στίχος into IUs.

I trust that by applying the concept of the IU I have contributed to a more natural conception of the LMG πολιτικὸς στίχος poetry and that this will in turn trigger a number of illuminating discourse-related studies, as has happened with regard to the Homeric “Kunstsprache”.

References

Aerts, Willem J. 2005. The lexikon to the Chronicle of Morea as a tool for linguistic studies. In Elizabeth Jeffreys and Michael Jeffreys (eds.), Approaches to texts in Early Modern Greek (Neograeca Medii Aevi V), 141–151. Oxford: University of Oxford.

Agapitos, Panagiotis, Hult, Karin, and Ole Smith. 1999. The Byzantine Achilleid. The Naples version. Vienna: Österreichische Akademie der Wissenschaften.

Agapitos, Panagiotis. 2004. SO debate: Comment on P.A. Agapitos, genre, structure and poetics in the Byzantine vernacular romances of love. Symbolae Osloenses 79: 7–54.

Agapitos, Panagiotis. 2006. Ἀφήγησις Λιβίστρου καὶ Ῥοδάμνης Κριτικὴ ἔκδοση τῆς διασκευῆς α. Athens: Μορφωτικό Ἵδρυμα Ἐθνικῆς Τραπέζης.

Allan, Rudger. 2009. Orale elementen in de Homerische grammatica: intonatie-eenheid en enjambement. Lampas 42: 136–151.

Apostolopoulos, Photis. 1984. La langue du roman Byzantin Callimaque et Chrysorrhoé. Athens: Aκαδημία Aθηνών.

Bakker, Egbert. 1990. Homeric discourse and enjambment: A cognitive approach. Transactions of the American Philological Association 120: 1–21.

Bakker, Egbert. 1993. Boundaries, topics and the structure of discourse. An investigation of the Ancient Greek particle δέ. Studies in Language 17: 275–311.

Bakker, Egbert. 1997. Poetry in speech: Orality and Homeric discourse. Ithaca (N.Y.): Cornell University Press.

Bakker, Egbert. 1999. How oral is oral composition? In E. Anne Mackay (ed.), Signs of orality. The Oral tradition and its influence in the Greek and Roman world, 29–47. Leiden: Brill.

Beaton, Roderick. 1980. Folk poetry of modern Greece. Cambridge: Cambridge University Press.

Beaton, Roderick. 19962. The medieval Greek romance. Cambridge: Cambridge University Press.

Brazil, David. 19972. The communicative value of intonation in English. Cambridge: Cambridge University Press.

Brown, Gillian, and George Yule. 1983. Discourse analysis. Cambridge: Cambridge University Press.

Browning, Robert. 19992. Medieval and Modern Greek. Cambridge: Cambridge University Press.

Chafe, Wallace. 1987. Cognitive constraints on information flow. In Russell Tomlin (ed.), Coherence and grounding in discourse: Outcome of a symposium, Eugene, Oregon, 21–51. Amsterdam: John Benjamins.

Chafe, Wallace. 1988. Linking intonation units in spoken English. In John Haiman and Sandra A. Thompson (eds.), Clause combining in grammar and discourse, 1–27. Amsterdam: John Benjamins.

Chafe, Wallace. 1993. Prosodic and functional units of language. In Jane A. Edwards and Martin D. Lampert (eds.), Talking data. Transcription and coding in discourse research, 33–43. Hillsdale (N.J.): Erlbaum.

Chafe, Wallace. 1994. Discourse, consciousness and time: The flow and displacement of conscious experience in speaking and writing. Chicago (Ill.): University of Chicago Press.

Chafe, Wallace. 1998. Language and the flow of thought. In Michael Tomasello (ed.), The new psychology of language: Cognitive and functional approaches to language structure, 93–111. Hillsdale (N.J.): Lawrence Erlbaum.

Chafe, Wallace. 2001. The analysis of discourse flow. In Deborah Schiffrin, Deborah Tannen, and Heidi E. Hamilton (eds.), The handbook of discourse analysis, 673–687. Oxford: Blackwell.

Chila-Markopoulou, Despina. 2004. Review of Pappas’ “Variation and morphosyntactic change in Greek: From clitics to affixes”. Journal of Greek Linguistics 5: 199–212.

Clancy, Patricia M., Ryoko Suzuki, Hongyin Tao and Sandra A. Thompson. 1996. The conversational use of reactive tokens in English, Japanese, and Mandarin. Journal of Pragmatics 26 (1): 355–387.

Croft, William. 1995. Intonation units and grammatical structure. Linguistics 33: 839–882.

Cruttenden, Allan. 19972. Intonation. Cambridge: Cambridge University Press.

Crystal, David. 1975. The English tone of voice. London: Edward Arnold.

Cupane, Caroline. 2003. Wie volkstümlich ist die Byzantinische Volksliteratur? Byzantinische Zeitschrift 96: 577–599.

Dawkins, Richard. 1939. Graeco-Barbara. Transactions of the Philological Society 38: 1–30.

de Vries, Mark. 2007. Dislocation and backgrounding. Linguistics in the Netherlands 24 (1): 35–47.

Egea, José M. 1988. Gramática de la Crónica de Morea: un estudio sobre el griego medieval. Vitoria: Instituto de Ciencias de la Antigüedad, Universidad del Pais Vasco.

Egea, José M. 1998. Historia extraordinaria de Beltandro y Crisanza. Granada: Athos-Pérgamos.

Eideneier, Hans. 1983. Leser- oder Hörerkreis? Zur byzantinischen Dichtung in der Volkssprache. Ελληνικά 34: 119–150.

Eideneier, Hans. 1989. KAI als Auftakt zur (rhythmischen) Phrase. Zur verbalisierten Pausenmarkierung im Mittel- und Neugriechischen. Jahrbuch der Österreichischen Byzantinistik 39: 179–200.

Eideneier, Hans. 1991. Byzantinische volkssprachliche Schriftkoine und mündliche Überliefung, Μαντατοφόρος 33: 7–10.

Eideneier, Hans. (1999), Von Rhapsodie zu Rap: Aspekte der griechischen Sprachgeschichte von Homer bis heute. Tübingen: Narr.

Eideneier, Hans, Ulrich Moennig, and Notis Toufexis (2001), Θεωρία και πράξη των εκδόσεων της υστεροβυζαντινής, αναγεννησιακής και μεταβυζαντινής δημώδους γραμματείας. Πρακτικά του Διεθνούς Συνεδρίου Neograeca Medii Aevi IV. Hamburg 28.-31.1. 1999. Herakleion: Πανεπιστημιακές Εκδόσεις Κρήτης.

Eideneier, Hans. 2001. Tι σημαίνει “προφορική παράδοση” στα κείμενα της δημώδους γραμματείας. In Hans Eideneier, Ulrich Moennig, and Notis Toufexis (eds.), Θεωρία και πράξη των εκδόσεων της υστεροβυζαντινής, αναγεννησιακής και μεταβυζαντινής δημώδους γραμματείας. Πρακτικά του Διεθνούς Συνεδρίου Neograeca Medii Aevi IV. Hamburg 28.-31.1. 1999, 47–58. Herakleion: Πανεπιστημιακές Εκδόσεις Κρήτης

Fleischman, Suzanne. 1985. Discourse functions of tense-aspect oppositions in narrative: Toward a theory of grounding. Linguistics 23: 851–882.

Fraenkel, Eduard. 1933. Kolon und Satz: Beobachtugen zur Gliederung des antiken Satzes 1. Nachrichten der Gottinger Gesellschaft der Wissenschaften, philologisch-historische Klasse, 319–354.

Geluykens, Ronald. 1994. The pragmatics of discourse anaphora in English. Evidence from conversational repair. Berlin: Mouton de Gruyter.

Halliday, Michael A.K. 1985. An introduction to Functional Grammar. London: Arnold.

Helasvuo, Marja-Liisa. 2001. Syntax in the making: The emergence of syntactic units in Finnish conversation. Amsterdam: John Benjamins.

Hinterberger, Martin. 1993. Sprachliche Variationsformen in volkssprachlichen metrischen Werken der spätbyzantinischen und frühneugriechischen Zeit. In Nikolaos Panayotakis (ed.), Origini della literatura Neogreca, vol. 1, 158–168. Venice: Instituto Ellenico di Studi Bizantini e Postbizantini di Venezia.

Hinterberger, Martin. 2001. Το φαινόμενο της πολυτυπίας σε δημώδη κείμενα. In Hans Eideneier, Ulrich Moennig, and Notis Toufexis (eds.), Θεωρία και πράξη των εκδόσεων της υστεροβυζαντινής, αναγεννησιακής και μεταβυζαντινής δημώδους γραμματείας. Πρακτικά του Διεθνούς Συνεδρίου Neograeca Medii Aevi IV. Hamburg 28.-31.1. 1999, 215–244. Herakleion: Πανεπιστημιακές Εκδόσεις Κρήτης

Hinterberger, Martin. 2006. How should we define vernacular literature?. Paper given at the conference “Unlocking the Potential of Texts: Interdisciplinary Perspectives on Medieval Greek” at the Centre for Research in the Arts, Social Sciences, and Humanities, University of Cambridge, 18–19 July 2006 (available at: http://www.mml.cam.ac.uk/greek/grammarofmedievalgreek/unlocking/html/Hinterberger.html; 04/02/2014).

Horrocks, Geoffrey. 2010. Greek: A history of the language and its speakers. London: Longman.

Iwasaki, Shoichi. 1993. The structure of the intonation unit in Japanese. In Soonja Choi (ed.) Japanese/Korean Linguistics, vol. 3, 39–53. Stanford: Center for the Study of Language and Information.

Jannaris, Antonius N. (1897), An historical Greek grammar chiefly of the Attic dialect. Hildesheim: Olms.

Janse, Mark. 1991. La phrase segmentée en grec ancien: le témoignage des enclitiques, Bulletin de la Société de Linguistique de Paris 86 (1): 14–16.

Janse, Mark. 1998. Homerische metriek. Orale poëzie in de praktijk. Didactica Classica Gandensia 38: 125–151.

Janse, Mark. 2003. The metrical schemes of the hexameter. Mnemosyne 56: 343–348.

Janse, Mark. 2008. Clitic doubling from Ancient to Asia Minor Greek. In Dalina Kallulli and Liliane Tasmowski (eds.), Clitic doubling in the Balkan languages, 165–202. Philadelphia: John Benjamins.

Jeffreys, Michael. 1973. Formulas in the Chronicle of the Morea. Dumbarton Oak Papers 27: 163–195.

Jeffreys, Elizabeth, and Michael Jeffreys. 1979. The traditional style of early demotic verse. Byzantine and Modern Greek Studies 5: 115–139.

Jeffreys, Michael. 1993. Proposals for the debate on the question of oral influence in Early Modern Greek poetry. In Nikolaos Panayotakis (ed.), Origini della literatura Neogreca., vol. 1, 251–266. Venice: Instituto Ellenico di Studi Bizantini e Postbizantini di Venezia.

Jeffreys, Elizabeth. 2011. Medieval Greek epic poetry. In Karl Reichl (ed.), Medieval oral literature, 459–484. Berlin and Boston: Walter De Gruyter.

Jeffreys, Elizabeth. 2013. Byzantine romances: Eastern or western? In Marina Brownlee and Dimitri Gondicas (eds.), Renaissance encounters: Greek East and Latin West, 217–234. Leiden: Brill.

Johnstone, Barbara. 20082. Discourse Analysis. Malden (MA): Blackwell.

Joseph, Brian D. 1987. Greek. In Bernard Comrie (ed.), The world’s major languages, 410–439. London: Croom Helm.

Joseph, Brian D. 1990. Morphology and universals in syntactic change: Evidence from Medieval and Modern Greek. New York and London: Garland Publishing Inc.

Joseph, Brian D. 2000. Textual Authenticity: Evidence From Medieval Greek. In Susan Herring, Piet van Reenen, and Lene Schoesler (eds.), Textual parameters in ancient languages, 309–329. Amsterdam: John Benjamins.

Kim, Kyu-Hyun. 1999. Phrasal unit boundaries and organization of turns and sequences in Korean conversation, Human Studies 22 (2–4): 425–446.

Koch, Peter, and Wulf Oesterreicher. 1985. Sprache der Nähe—Sprache der Distanz: Mündlichkeit und Schriftlichkeit im Spannungsfeld von Sprachtheorie und Sprachgeschichte, Romanistisches Jahrbuch 36: 15–43.

Kriaras, Emmanouel. 1955. Βυζαντινὰ Ἱπποτικὰ Μυθιστορήματα. Athens: Aetos.

Lambrecht, Knud. 1981. Topic, antitopic and verb agreement in non-standard French. Amsterdam: John Benjamins.

Lauxtermann, Marc. 1999. The spring of rhythm: An essay on the political verse and other Byzantine metres. Vienna: Verlag der Österreichischen Akademie der Wissenschaften.

Lendari, Tina. 2007. Livistros and Rodamne. The Vatican version. Critical edition with introduction, commentary and index-glossary. Athens: Μορφωτικό Ἵδρυμα Ἐθνικῆς Τραπέζης.

Loudová, Katerina. 2007. On functional analysis of Greek sentence structure. Paper given at the international workshop “The Aspects of Linguistics in Classical Philology”, Bratislava, 13–15 April 2007.

Loudová, Katerina. 2009. Cohesive textual means in Early Byzantine chronicles as a mirror of the language register. In Katerina Loudová and Marie Žáková (eds.), Early European languages in the eyes of modern linguistics, 189–202. Brno: Masarykova univerzita.

Lurier, Harold E. 1964. Crusaders as conquerors: The Chronicle of Morea. New York: Columbia University Press.

Mackridge, Peter. 1990. Orality in Medieval and Modern Greek poetry: Introduction, Byzantine and Modern Greek Studies 14: 123–128.

Mackridge, Peter. 1993. An editorial problem in Medieval Greek texts. The position of the object clitic pronoun in the Escorial Digenes Akrites. In Nikolaos Panayotakis (ed.), Origini della literatura Neogreca, vol. 1, 325–342. Venice: Instituto Ellenico di Studi Bizantini e Postbizantini di Venezia.

Mackridge, Peter. 2000. The position of the weak object pronoun in Medieval and Modern Greek. Jazyk i rečevaja dejatel’nost’ 3 (1): 133–151.

Manolessou, Io. 2008. On historical linguistics, linguistic variation and Medieval Greek. Byzantine and Modern Greek Studies 32 (1): 63–79.

Matsumoto, Kazuko. 2000. Intonation units, clauses and preferred argument structure in conversational Japanese. Language Sciences 22 (1): 63–86.

Matsumoto, Kazuko. 2003. Unit linking in conversational Japanese. Language Sciences 25: 433–455.

Meister, Karl. 1921. Die Homerische Kunstsprache. Leipzig: Teubner.

Moennig, Ulrich. 2004. Die Erzählung von Alexander und Semiramis: kritische Ausgabe mit einer Einleitung, Übersetzung und einem Wörterverzeichnis. Berlin: De Gruyter.

Nespor, Marina and Irene Vogel. 1986. Prosodic phonology. Dordrecht: Foris.

Nørgaard, Lars and Ole Smith. 1975. A Byzantine Iliad. The text of Par. Suppl. Gr. 926. Copenhagen: Museum Tusculanum.

Oesterreicher, Wolf. 1997. Types of orality in Text. In Egbert Bakker and Ahuvia Kahane (eds.), Written voices, spoken signs, 190–214. Cambridge (Mass.): Harvard University Press.

Papathomopoulos, Manolis and Elizabeth Jeffreys. 1996. Ὁ Πόλεμος τῆς Τρωάδος: Κριτικὴ ἔκδοση μὲ εἰσαγωγὴ καὶ πίνακες. Athens: Μορφωτικό Ἵδρυμα Ἐθνικῆς Τραπέζης.

Rosenqvist, Jan Olof. 2007. Die byzantinische Literatur. Vom 6. Jahrhundert bis zum Fall Konstantinopels 1453. Berlin and New York: Walter de Gruyter.

Salas, Francisco Javier Ortolá. 1998a. Florio y Platzia Flora: una novela bizantina de época paleólogica. Madrid: Universidad de Cádiz.

Salas, Francisco Javier Ortolá. 1998b. Las ediciones de la novela bizantina de época paleóloga Florio y Platzia Flora. Habis 29: 323–333.

Scheppers, Frank. 2011. The colon hypothesis. Word order, discourse segmentation and discourse coherence in Ancient Greek. Brussels: VUB Press.

Schmitt, John. 1904. The Chronicle of Morea: A history in political verse, relating the establishment of feudalism in Greece by the Franks in the thirteenth century. London: Methuen.

Selkirk, Elisabeth. 1984. Phonology and syntax. The relation between sound and structure. Cambridge (Mass.): MIT Press.

Shawcross, Teresa. 2005. Oral residue and narrative structure in the Chronicle of Morea. Byzantion 75: 310–333.

Shawcross, Teresa. 2009. The Chronicle of Morea: Historiography in crusader Greece. Oxford: Oxford University Press.

Sifakis, Grigoris. 2001. Looking for the tracks of oral tradition in Medieval and Early Modern Greek poetic works. Journal of the Hellenic Diaspora 27 (1–2): 61–86.

Slings, Simon R. 1992. Written and spoken language: An exercise in the pragmatics of the Greek sentence. Classical Philology 87 (2): 95–109.

Slings, Simon R. 1994. Een tandje lager: Aanzetten voor een orale grammatica van Homerus, Lampas 27 (5): 411–427.

Tao, Hongyin. 1996. Units in Mandarin conversation. Amsterdam: John Benjamins.

Thoma, Chrystalla. 2007. Distribution and function of clitic object pronouns in popular 16th–18th century Greek narratives. A synchronic and diachronic perspective. In Jochen Rehbein, Christiane Hohenstein, and Lukas Pietsch (eds.), Connectivity in grammar and discourse, 139–163. Amsterdam: John Benjamins.

Toufexis, Notis, and Chrystalla Thoma. 2004. The distribution of the discourse marker “kai” in Greek narratives. Paper given at the “Logos Conference”, controlling language, the Greek experience-interdisciplinary international conference organized by the Centre for Hellenic Studies at King’s College London.

Toufexis, Notis. 2008. Diglossia and register variation in Medieval Greek, Byzantine and Modern Greek Studies 32 (2): 203–217.

Tseng, Shu-Chuan. 2006. Linguistic markings of units in spontaneous Mandarin. ISCSLP Lecture Notes in Artificial Intelligence, 43–54.

Tseng, Shu-Chuan. 2008. Spoken corpora and analysis of natural speech. Taiwan Journal of Linguistics 6 (2): 1–25.

Vejleskov, Peter. 2005. The position of the weak object pronoun in the Greek Portulans and in the Chronicle of the Turkish Sultans. In Elizabeth Jeffreys and Michael Jeffreys (eds.), Approaches to texts in Early Modern Greek (Neograeca Medii Aevi V), 197–209. Oxford: University of Oxford.

Wackernagel, Jacob. 1892. Über ein Gesetz der indogermanischen Wortstellung. Indogermanische Forschungen 1: 333–446.

Wills, Lawrence M. 2002. The quest of the historical gospel: Mark, John and the origins of the gospel genre. London and New York: Routledge.

Wouk, Fay. 2008. The syntax of intonation units in Sasak. Studies in Language 32 (1): 137–162.

*

My work was funded by the Research Foundation of Flanders (FWO) (grant no. B/13006/01).

1

Note that this mixed language has often been negatively assessed in the past; Dawkins (1939:2) for instance speaks of a “Graeco-barbaric” mix.

2

For more on the “mixed” character of Medieval Greek, see articles by Hinterberger (1993, 2001) and Eideneier (2001).

3

All English translations are my own.

4

Especially M. and E. Jeffreys have done pioneering work concerning formulas in the πολιτικὸς στίχος poetry, in particular with regard to the Chronicle of Morea (e.g. M. Jeffreys 1973) and the War of Troy (e.g. E. & M. Jeffreys 1979).

5

Note that length limitations force me to reproduce this debate in a very sketchy way. A summary of it can be found in M. Jeffreys (1993) and in Beaton (1996:164–188).

6

With regard to Belthandrus & Chrysandza for instance, Agapitos (2004:43) concludes that “it is difficult to decide whether this fictitious audience setting was intended to function within a performative situation of actual recitation, as in L&R, or if the audience setting here already reflects a literary convention”.

7

Note that the dating of the original texts may differ substantially from that of the manuscripts and that researchers do not agree on the latter, see Agapitos (2004:12–13).

8

I have excluded Kallimachos & Chrysorrhoi, for this text is generally acknowledged to be less vernacular than the others: “The language in which this romance is written contains a much higher proportion of learned forms, superimposed on a vernacular base, than we find in any of the other romances of this period” (Beaton 1996:118; see Horrocks 2010:345). Moreover, it is the only text whose author is known.

Note that IM and PP are not considered original Greek romances, but adaptations from western models (see Beaton 1996).

9

This abbreviation is derived from its Latin title “Bellum Troianum”.

10

By choosing H, I follow Egea (1988), who has written a grammar of the Chronicle and mainly uses examples from H (if not: ‘P’ is indicated), as well as Lurier (1964), who has based his English translation on H.

11

Morea (Havniensis): 9219 verses. The edition of Schmitt also contains the Parisinus manuscript, which is presented in a parallel way.

12

See the separate article by Salas (1998b) on the mistakes made in previous editions: the edition of PP by Kriaras (1955) especially is severely criticized. Salas’ principle reads as follows: “Con nuestra edición pretendemos aunar, por tanto, los aciertos de las anteriores pero bajo criterios filológicamente conservadores, manteniendo en la medida de lo posible el texto trasmitido por los manuscritos y otorgándole mayor protagonismo al vienés, el más susceptible de mejora después de haber sido maltratado por el pruritus emendandi de los editores que nos precedieron” (Salas 1998a: 68; my italics).

13

See Beaton (1996:104): “we can be fairly confident that the N version gives us the substance if not the wording of the original poem”. The other manuscripts are the Brit. Mus. addit. 8241 and the Oxon. Bodl. Auct. T.5.24. The latter, which is twice as short as the Naples version (M. Jeffreys 1975:189), possibly reflects a real oral performance (Agapitos 2006:162; see. 2.2.).

14

This is suggested by the following scheme:

1–35: NOV

36–414: NOVH

415–745: NOVHG

746–824: NOVG

825–860: NOV

861–893: NV

with O standing for the Bodl. misc. 287; V for the Vindob. theol. gr. 297 (= Nessel 244); and G and H for the Palat. gr. 426 (G: 65r–72v; H: 73r–93v) (Kriaras 1955:232–233).

15

This number is based on English. However, as Chafe himself observes, the actual number depends on the structure of each individual language.

16

To be perfectly clear: the occurrence of a breathing pause is neither a necessary nor a sufficient criterion to assume an IU boundary.

17

Other common terms are: “sentence segment” (Janse 1991); “intonation groups” (Cruttenden 1997); “intonation(al) phrases” (Selkirk 1984; Nespor & Vogel 1986); “tone units” (Crystal 1975; Brazil 1997); “tone groups” (Halliday 1985). By using the term “Intonation Unit” and its well-known abbreviation “IU”, I follow Chafe (in his later work) and most discourse researchers.

18

Croft (1995) calculated that 97% of the English IUs in his corpus are grammatical units. Much research has been done on the IU–grammatical unit relationship.

19

See Brown & Yule (1983:16): “In spoken language the largely paratactically organised chunks are related by and, but, then and more rarely if”.

20

Very recently, Scheppers (2011) has applied the concept of the IU to other Ancient Greek texts which contain many spoken echoes, namely the Platonic dialogues and the Lysian orations.

21

I must admit that a number of researchers seem to have intuitively felt that we can derive trustworthy syntactic information from the LMG poetry; see Beaton (1996:215): “from a linguistic viewpoint, Peter Mackridge has further been able to demonstrate that the language of the E version [Escorial manuscript of the epic Digenis Akritis, which is also composed in the πολιτικὸς στίχος] is not the hotchpotch that has usually been assumed but actually obeys a consistent set of rules.” Indeed, Mackridge (1993:338) has observed a clear regularity in the placement of object clitic/weak pronouns: “the redactor and/or scribe of the Escorial Digenes Akrites displays a remarkable consistency in his application of the rules governing the position of the clitic pronoun”.

22

Aristotle finds the iambic rhythm to be very natural for the Greek language (i.e., Classical Attic; Poetics 1449a, 24–28; Rhetoric 1408b, 33–35). Moreover, the iambic trimeter is in ancient times used for the spoken parts in tragedy and comedy.

23

However, we need to have a more explicit account of “the types of syntactical structure that may be interrupted at the midline caesura and at the end of a line” (Mackridge 1990:202).

24

See Papathomopoulos & E. Jeffreys (1996:lxxxi): “‘Formulas’ will be used here to indicate a phrase filling a complete half-line”, Beaton (1980:44): “A formula in the Greek tradition is never more nor less than a half-line”, and Beaton (1980:52): there exist “two formula systems, each the length of a half-line”.

25

See Apostolopoulos (1984:210): “la présence de certaines particules conjonctives (γάρ, δέ, γε, τοίνυν, οὖν, etc.), privées de tout rôle fonctionnel, ne s’explique que par des besoins métriques”.

26

From now on, the verbs/pronominal objects (in singular) are put in bold, while their (plural) referents are underlined.

27

To be perfectly clear, I do not claim that these verses are considered “wrong” according to the written standards of the time. As a matter of fact, a codified vernacular standard was still absent: “there was no formalization or standardization and no grammatical tradition in the vernacular” (Manolessou 2008:73).

28

An anonymous referee suggests an alternative explanation for the high frequency of καί (which is not necessarily incompatible with the above one, though): καί constitutes a prototypical marker of narrativity, going back to the New Testament and popular prose Greek romances like the Tale of Alexander and the Life of Aesop (see Wills 2002 for basic bibliography).

29

In modern linguistics, this construction would be considered a case of “clitic doubling”, as the (clitic/weak) pronoun actually “doubles” the noun. According to Janse’s (2008) typology of Modern Greek clitic doubling, all the above examples belong to the type of “right-dislocations”.

30

See de Vries (2007:11): “an afterthought (…) has an independent intonation-contour.”

  • 10

    By choosing H, I follow Egea (1988), who has written a grammar of the Chronicle and mainly uses examples from H (if not: ‘P’ is indicated), as well as Lurier (1964), who has based his English translation on H.

  • 12

    See the separate article by Salas (1998b) on the mistakes made in previous editions: the edition of PP by Kriaras (1955) especially is severely criticized. Salas’ principle reads as follows: “Con nuestra edición pretendemos aunar, por tanto, los aciertos de las anteriores pero bajo criterios filológicamente conservadores, manteniendo en la medida de lo posible el texto trasmitido por los manuscritos y otorgándole mayor protagonismo al vienés, el más susceptible de mejora después de haber sido maltratado por el pruritus emendandi de los editores que nos precedieron” (Salas 1998a: 68; my italics).

  • 18

    Croft (1995) calculated that 97% of the English IUs in his corpus are grammatical units. Much research has been done on the IU–grammatical unit relationship.

  • 20

    Very recently, Scheppers (2011) has applied the concept of the IU to other Ancient Greek texts which contain many spoken echoes, namely the Platonic dialogues and the Lysian orations.

If the inline PDF is not rendering correctly, you can download the PDF file here.

  • 10

    By choosing H, I follow Egea (1988), who has written a grammar of the Chronicle and mainly uses examples from H (if not: ‘P’ is indicated), as well as Lurier (1964), who has based his English translation on H.

  • 12

    See the separate article by Salas (1998b) on the mistakes made in previous editions: the edition of PP by Kriaras (1955) especially is severely criticized. Salas’ principle reads as follows: “Con nuestra edición pretendemos aunar, por tanto, los aciertos de las anteriores pero bajo criterios filológicamente conservadores, manteniendo en la medida de lo posible el texto trasmitido por los manuscritos y otorgándole mayor protagonismo al vienés, el más susceptible de mejora después de haber sido maltratado por el pruritus emendandi de los editores que nos precedieron” (Salas 1998a: 68; my italics).

  • 18

    Croft (1995) calculated that 97% of the English IUs in his corpus are grammatical units. Much research has been done on the IU–grammatical unit relationship.

  • 20

    Very recently, Scheppers (2011) has applied the concept of the IU to other Ancient Greek texts which contain many spoken echoes, namely the Platonic dialogues and the Lysian orations.

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 42 0 0
Full Text Views 371 158 9
PDF Downloads 30 15 1