Use the following persistent identifier: http://nrs.harvard.edu/urn-3:hul.ebook:CHS_BlankenborgR.Rhythm_without_Beat.2014.
Introduction: Sense, Syntax and Prosody
What exactly did Homer leave behind? Was it a long series of hexameters? Yes and no. In writing, both the Iliad and the Odyssey appear as continuous repetitions of stichic verses. The epic narrative is cast in the hexametric mould: every single line fits one hexameter. The metrical shape of the hexameters varies, but all the verses of the Iliad and the Odyssey meet the structural requirements of hexametric poetry: alternating rhythm, the number of feet, of syllables, of theses, and of arses. All the verses scan without serious difficulty; the division of the Homeric text into hexametric lines has never been a problem like the division into lines of, for example, Pindar. The metrical shape of the Iliad and the Odyssey, as well as the metrical units into which the text divides, are clear enough. Not so clear are the intelligible units of sense into which the epic divides. At times whole verses or parts thereof, the hemistichs or the metrical cola, seem to function as the units of sense; then again, often the units of sense do not stop at the boundaries of the verse, the hemistich, or the metrical colon. Homeric syntax appears to use metrical cola and verses as seemingly independent units of sense, but at the same time syntax develops well beyond the metrical cola and the verse. Metrical units, as the prosodic organisation that can be most easily identified, comprise of units of syntax, without automatically being syntactical units themselves. This study aims to further clarify the relation between these two structural aspects of the Homeric epic: prosody and syntax. Both are aspects of “building”. Both aspects are minor scale: they have little or nothing to do with the grander scheme of the Iliadic or Odyssean story. The two aspects do not present themselves in a specific order of importance; they never have.  In the Iliad and the Odyssey, as in most Greek poetry, prosody and syntax are two sides of the same coin. In fact, Greek poetry is about that coin having two inseparable sides: prosody and syntax must come into being together. There is no “first” or “second”: prosody without understandable wording resembles a musical or rhythmical pattern, syntax without sufficient, recognisable prosodic characterisation, patterning and regulation is prose. 
Often the prosodic characterisation, patterning and regulation of the Homeric text is taken for granted, especially in the case of meter: the metrical structure of the repetitive hexameter is seen as a datum. Any other type of minor-scale patterning in the Iliad and the Odyssey is analysed in relation to the metrical surface structure. The syntactical patterns of the text, for example, can be judged considering their accordance, or mismatch, with the repetitive hexametric structure. Do the syntactical units fit recurring metrical shapes like the colon, the hemistich, and the verse? Or does syntax “straddle” the structural boundaries of metrical phrasing? There is an underlying presupposition here: the Homeric wording had to be “grammatised” within the already existing metrical shape. Seductive as this underlying presupposition may be, it presupposes much more than the simple need for “grammatisation” within rigid prosodic formats. It equally presupposes that the single hexametric verse is not just the metrical unity in the Homeric narrative, but also the unity whose prosodic characterisation suggests “audible coherence” on the level of syntax. The single hexameter would be the “frame” that signals both the prosodic and the syntactic unity. This presupposition shows that the prosodic pattern, the metrical shape, is regularly seen as pre-existing compared to the development of wording and syntax. There is no evidence, however, for such pre-existence of any prosodic patterning. I will study the development of syntactical unities without any assumptions concerning the pre-existence of prosodic patterns. It must be studied per se: the syntactical unity must be identified relying solely on grammatical criteria.
It is not so easy, however, to conclude to any level of syntactical “coherence” in Homer. The general point of view still found in many modern studies on Homeric wording and syntax presupposes features on the level of syntax that can be compared to those in written classical Greek. The comparison with written language, however, has its limits: not surprisingly, analysis of the Iliad and the Odyssey as spoken language proves to be much more productive than any comparison to written Greek. The effect of the change in approach towards analysis as spoken language is enormous; syntactical analysis of the Iliad and Odyssey, both familiar and recent, shows the fruits to be harvested from refocusing on Homer’s newly redefined “orality”. In this field, there already exists much, often groundbreaking, research to build upon.  Regardless, however, of the re-establishment of the oral nature of Homeric epic as spoken, partly unplanned, language, the “mismatch” of Homer’s verse and his utterance was not always explained as a result of the pre-existence of the prosodic pattern. There is much concern for the effect on syntax of the clustering of the prosodic pattern’s units: what happens when syntax is not confined to the units of the pattern? What happens if the unit of the prosodic pattern ends in enjambment? Existing explanations dealing with the “mismatch” focus on the exact realisation of verse-end enjambment.  The separation, by the verse end, of constituents that belong to one syntactical unit is described as milder or harsher depending on the level of grammatical completeness at the verse end. I do not agree with this approach to “out-of-line” composition, but the enjambment studies have proven to be a valuable basis for my own views.
My own views focus on the audible demarcation of syntactically coherent units. I am looking for what makes phrases audible in the continuous flow of dactyls. It is possible, I think, to identify the audible markers of phrase termination in our written version of the Homeric epic. What is more, a proper understanding of the audible patterning of phrases will show that there is some guidance on syntactical development from the side of prosody. In this study, I will argue that syntactical development beyond the autonomy of single words or word groups was facilitated by an aspect of Homeric prosody that differs from the metrical surface structure, though it is realised together with meter. My focus will be on the strength of metrical boundaries as phonetically realised pauses. I will deal, in other words, with the combination of metrics and phonetics. My hypothesis will be that phonetically realised pauses demarcate phonological phrases, and that they do so because of the non-demarcative value of the remaining metrical boundaries.
My starting point is the description of Homeric discourse by Chantraine and Bakker, and the acknowledgement of enjambment by Clark.  In Clark’s view, enjambment is clause formation over the verse end. My approach takes one more step: enjambment is one aspect of clause formation over metrical boundaries, including the verse end. I will show that the basis of Homeric discourse is not the metrical unity but the grammatical clause. That grammatical clauses may ignore metrical boundaries is acknowledged by ancient testimonies.  Observations from antiquity concerning “breaking the measure” further acknowledge that the internal syntactical organisation of grammatical clauses is audibly perceptible as such. The boundaries of recurring metrical shapes can be “muted”. Metrical colometry becomes subordinate as the syntactic phrases become audible in the flow of dactyls. I will identify the problems concerning the description of Homeric syntax, and provide evidence from the text of the Homeric epic for a solution to the problems. Subsequently, I will start looking for a reflection in prosody, other than merely the metrical surface structure, for the variable phrasing. This approach is preferable to one that takes the other way around: let the syntactic structure be identified by its own criteria, not by the recurring patterns of prosody. Questions automatically present themselves: how can the emergence of rather free clause formation be reconciled with the recurrence of metrical patterns that is so strong that it determined the general appearance of Homeric discourse? How did these clauses become audible as phrases in the continuous flow of dactyls? In other words, what aspect of Homeric prosody caused the way the grammatical unity starts to deviate from the way the metrical unity starts?
To trace this “aspect of prosody” in the written form of our Iliad and Odyssey, I take into account the analysis of terminal lengthening and final-syllable-retardation by Ruijgh, and the contribution of additional lengthening to prepausal demarcation by Devine and Stephens.  Their findings allow me to identify the audible pause in Homer. I argue that the combination of word end in a syllable (as opposed to word end in a segment), pl us location on a specific metrical position, may allow for either a stronger or a milder audible pause. Location on the metrical thesis results in a stronger pause, location on the arsis in a milder pause. The underlying structural norms of dactylic meter remain intact on the surface, but the different combinations of segmental weight, metrical position, and word-internal position bring out the variance in phrasing in the continuous flow of dactyls. In my analysis, the verse end occupies a special position, as it may be realised in many different ways: the verse final metrical element is not always a metrical arsis. I conclude with a number of examples that visualise the prosodic determination of syntactical development in Homer. In both smaller and larger scale composition, the effect of prosodic reorganisation can be seen as the emergence of syntactical readjustment.
The analysis of Homeric syntax by, among others, Chantraine and Bakker, is discussed in chapter 1. Bakker 1997 describes Homeric discourse as special speech, partly due to the metrical format: grammar emerges in response to metrical circumstances. Often enough the recurring metrical boundaries demarcate the scope of grammatical development and organisation. Clark 1995 points at the frequency and ease with which Homeric composition is extended beyond the hexameter. His approach (as does Bakker 1990) does not focus on grammatical completeness at verse end (like Parry 1929, Kirk 1966, Higbie 1990), or at another metrical boundary simply because it is a metrical boundary. Clark’s main concern is the set of circumstances facilitating enjambment. He identifies runovers and anticipations as semantic “hooks” to link formulas over the verse end. Clark’s analysis, I think, correctly reduces the importance of grammatical completeness at verse end. Homeric composition over the verse end, however, is not restricted to semantic hooks. Nor does it only involve formulas. In chapter 2, I show that composition, including composition over the verse end, answers to the needs of the grammatical clause: often within recurring metrical boundaries, but regularly despite them, even despite the verse end. My modification of the description of Homeric discourse results in modified definitions of clausal grammar and clausal enjambment. The constituents of Homer’s clause and their alignment into larger-scale syntactic units resembles the phonological patterning into minor and major phrases. As I show in chapter 3, the metrical and the phonological shape of Homer’s grammatical clause must be clearly distinguished. Phonology provides some clues for phonetic demarcation. Ancient terminology also modestly sheds light on the issue of phonetic demarcation. An attempt to compare the grammatical phrasing in clauses and the phonological patterning in major phrases clarifies the phonology of “grammatisation” and of metrical boundaries like the verse end. Composition of grammatical clauses that mutes metrical boundaries like verse end is reflected in the perception of syntactical unities as of varying size and location. Specific rhythmical properties of metrical texts appear to take a likeness to their equivalent in prose. Free clause composition “breaks the measure”.
As clause formation may break the measure, metrical boundaries are possibly overestimated as breaks in sense. They may still play a part in the demarcation of syntactical unities but only if combined with another type of realisation. This other type of realisation, the subject of chapter 4, has to do with audible termination: in Greek poetry the additional lengthening at specific metrical positions and the speech retardation towards the end of the utterance. Starting from the analysis of terminal lengthening by Ruijgh 1987, and the contribution of additional lengthening to prepausal demarcation by Devine and Stephens 1994 in addition to more general views on pause like Daitz 1991, I argue that the combination of syllable and metrical position may allow for either a stronger or a milder audible pause. Whereas I deviate from existing studies in looking for metrical-phonetic indicators for pause instead of purely metrical, purely phonological, or semantic indicators, the existing notions concerning terminal lengthening and the foot-internal proportion remain my starting point. I accept the analysis of the foot-internal proportion as defended by Ruijgh, but I believe that it has consequences for the possibilities meter offers to pause in order to audibly demarcate phrases. The difference in realisation of an arsis or thesis as either word-final segment or syllable is crucial. Both rhythmical weights, of the segment as well as the syllable, are perceived but they are perceived differently: the syllable can not be perceived as word-final without disrupting the metrical surface structure; the segment can. Metrical location, on the thesis or the arsis, reflects the phonetic realisation of word or phrase termination: the more phonetic disruption metrical location allows, the stronger the audible pause at word end in a syllable. Stronger and milder audible pauses lead to a type of phrasing, even a kind of rhythmical repetition, that audibly differs from metrical colometry. The verse end occupies a special position: it can be realised as a stronger pause, a milder pause, or as no pause at all.
In the final chapter, chapter 5, a number of examples will show where we see the effect of extended freedom in composition. Prosody motivates the extension of compositional freedom. The alignment of independent minor phrases into major phonological phrases results in syntactical “subordination” of the minor phrases within the major phrase. This phrasal reorganisation of minor phrases is due to the difference between the audible pauses. Prepausal realisation of word-final syllables as milder pauses leads to the muting of metrical boundaries and the combining of smaller scale unities into larger scale syntactical wholes between stronger pauses. Within the grammatical clause, syntactical development through prosodic reorganisation becomes visible in the location of, for example, appositives, attributive elements, and hyperbaton. The “progressive tendency” of Homeric syntax, as suggested by Bakker 1997b, is thus reflected in prosody. The analysis of larger scale composition shows that the tendency is actually elicited by prosodic reorganisation. In the combining and linking of grammatical clauses, prosodic reorganisation results in syntactical readjustment and complexity.
[ back ] 1. Cf. the mixed nature of the comments concerning Iliadic passages by scholars from antiquity. T on Iliad 1.2, as presented in Erbse 1969:6 may serve as an example.
[ back ] 2. Cf. the remarks in Gorgias Helen 9; Demetrius On Style 180–182.
[ back ] 3. Especially the work of Bakker 1997b, 2005, and Oesterreicher 1997.
[ back ] 4. Parry 1929; Kirk 1966; Higbie 1990; Clark 1997.
[ back ] 5. Chantraine 1953; Bakker 1997b; Clark 1997.
[ back ] 6. E.g. in Dionysius of Halicarnassus On Literary Composition 26.
[ back ] 7. Devine & Stephens 1994.