II. Metrics
On this web page you will enter the second part of our compendium on the basic linguistic structures of the Divina Commedia. It deals with metrics, i.e. the verse meter of the work, and discusses ways in which the poet can change the number of syllables in a verse through variable linguistic interventions.
You will find out about such cryptic-seeming terms as sinalefe and dialefe, and why there is a complex connection or disconnection of words behind them. You will learn that sineresis and dieresis assert an analogous syllable arrangement within a word. You will encounter the phenomenon of the hiatus as well as the syllabification of the intervocalic i and master a final metrical subtlety with the falling diphthong.
This will help you know when to join or separate the words of a verse and prepare you to tackle the last missing aspect of your reading skills with accents, the third menu item in our outline.
1. Sinalefe and Dialefe
The Divina Commedia, as is well known, follows the verse meter of the Endecasillabo. It is defined by having a total of eleven metrical syllables in its basic form, the penultimate of which is stressed. In order to achieve this, the author does not have to simply place the words in a banal way so that eleven syllables come together. He has the option of shortening or lengthening the number of syllables to be counted using various linguistic techniques.[16]
We want to go a little deeper into this aspect in order to facilitate the understanding of the following and to make it clear where the original text, which already contains important designations, must be supplemented in order to have complete transparency of the syllable division and the pronunciation that is potentially oriented to it.[17]
To arrive at a correct syllable count or scansion, adjacent syllables are first singly scored. Here, depending on whether they are open syllables (they end with a vowel) or closed syllables (they end with a consonant), vowels or consonants can meet. The critical constellation occurs when two vowels are next to each other. In general, one speaks of sineresis when two consecutive vowels within a word count as one syllable.[18] Sinalefe is given when the final vowel of a word and the initial vowel of the following one in the verse count as one syllable. If, on the other hand, syllabification occurs between two adjacent vowels within a word, causing them to count as two syllables, one speaks of dieresis.[19] In dialefe, syllabification occurs between two words. Here, the final vowel of the first word and the initial vowel of the second word count as two syllables.
It is now significant that within a word the sineresis represents the normal case, i.e., it is not specifically designated.[20] Thus, if the author wants to make a syllable division where, as in the diphthong, there is usually a single syllable, he can, in certain linguistically permitted cases, place a dieresis, i.e., mark a colon above the critical vowel, indicating that there are two syllables here. Within a word, this would provide clarity, i.e., the text would indicate how the syllables are to be valued.[21]
Between two words separated by vowels, however, the text leaves us in the lurch. It is not clear when we have a sinalefe and when a dialefe before us. In most cases one will be allowed to trust some basic rules, such as that there is usually a sinalefe here, especially, for instance, when an unstressed syllable closes the word. The dialefe would be the exception, or appear more often in Dante's work at the end of stressed syllables or in the majority of monosyllabic words.
But our confidence is soon disappointed. For one thing, there is a large number of exceptions that cannot be grasped even in the most differentiated special rules. On the other hand, Dante allows himself a considerably greater freedom of scansion than his classical successors. This has given rise to the impression among linguistic epigones that our poet, notwithstanding a high, even idiosyncratic regularity of his own, made in the Divina Commedia, in certain cases simply followed the necessity of correct scansion, i.e., altered it to produce a correct endcasillabo of eleven metrical syllables.
We do not wish to take sides here, we just want to state that the layman is overwhelmed by mere observation of the verse, i.e., cannot recognize the correct division of the sinalefes and dialefes. And even the expert, who in many cases intuitively makes the right choice, has to count word for word in unorthodox verses in order to find the solution.[22] We know that behind this choice there is a complex but ultimately unsatisfactory set of rules, and we know that the literature has commented exhaustively on the subject.[23] Anyone who has studied the matter in depth knows the articles of the Enciclopedia Dantesca, the works of a Menichetti, Beltrami, and others.[24]
Yet, I would like to single out a book and a database that shed light on the thicket of variants. These are David Robey's Sound and Structure in the Divine Commedy (2000) and his authorized database Italian Narrative Poetry of the Middle Ages and Renaissance (2003).[25] It contains with impressive reliability the sinalefes and dialefes for each verse of the Divina Commedia.
Our edition first made the classification of these metrical signs independently and then checked them against the database. It has been able to confirm their entries without exception and offers the reader a complete, homogeneous metrical breakdown of Dante's poetry.
2. Hiatus and syllable division in the intervocalic i
Let us now turn to two other metrical phenomena that catch our attention only because they have an influence on syllable counting.
A. Hiatus
A hiatus occurs when two consecutive vowels within a word are independent of each other, i.e. form two syllables of their own and are also pronounced separately (Serianni 2011: 20). This distinguishes the hiatus from the diphthong, which binds two vowels together to form one syllable. The grammar knows a number of constellations in which a hiatus is to be assumed:
- if neither of the two vowels is an i or a u: ma-estro, po-eta
- if one of the two vowels is an accented i or u: fa-ìna, pa-ùra
- if the i or u ends a prefix such as ri-, bi-, di-, or su- followed by a vowel: ri-avere, su-accennato.
The hiatus becomes significant because it is not specifically noted in the original text. The reader cannot decide whether this is a monosyllabic diphthong without vowel separation or a two-syllable hiatus with vowel separation. The matter is complicated by the fact that there are words whose double vowel is once valued as monosyllabic, but another time as two-syllabic. An example would be Beatrice, once as a diphthong (Bea-trice), but then also as a hiatus (Be-a-trice).[26] And even a word like creature, whose Latin origin should point to a hiatus and which otherwise appears in the Divina Commedia as such, is once (Paradiso 3, 127) at the beginning of the word treated as a monosyllabic, i.e. as a sineresis. Rules, which want to justify etymologically, when a hiatus and when a diphthong has to take place, obviously reach their limits here. They, as well, cannot be expected of the reader. Our edition designates each hiatus with a vertical bar in blue letters, making it thus recognizable.
B. Intervocalic i
We are left with a related phenomenon, the intervocalic i. Here, an i sits between two vowels, whereby it is assigned to the second vowel and usually begins a new syllable with it. Nevertheless, there are exceptions. In the Divina Commedia, for example, mi-glia-ia is once, as expected, executed as a two-syllable word at the end of the word (3 23 28), i.e., the last i opens a new syllable. Another time, however, we encounter it in analogous position as monosyllabic mi-gliaia (2 22 36). Here, then, the final i combines with the adjacent vowels to form a syllable. The original text does not indicate which of the two forms is present. We have therefore inserted a bar in blue color in the case of a syllable division, analogous to the hiatus. The database of Robey (2003) provided us with excellent, error-free services here as well as there.
C. The Problem of the Falling Diphthong
We want to conclude our consideration of metrics with an interesting phenomenon that has led metricians to different conclusions. It concerns constellations in which a falling diphthong is followed by a vowel. David Robey (2000) has addressed this linguistic aspect with great care, so we may follow his account.
Two solutions are possible here to obtain the correct number of syllables and thus the metric of the hendecasyllable. One is to combine a dieresis with a sinalefe, and the other is to combine a sineresis with a dialefe. Let's take a closer look at an example. The word pair avea and il can serve both the first solution (avëaˆil) and the second solution (aveaˇil). In both cases, two metrical syllables are obtained. Also, there is no accent shift. The first version preserves the sinalefe after the unstressed a, i.e., it satisfies one of the basic rules of vowel linkage; the second avoids the artificial dieresis on the e and accounts for the possibility of vowel separation when the final syllable is stressed.
Which solution is now the correct one? Petrocchi (2003) prefers the first option, Antonio Lanza (1996) opts in a more recent edition of the Divina Commedia for the second option. Now, while Bausi & Martelli (1993: 15) side with Petrocchi, Fasani (1992: 20) allows this to hold only for cases where a falling diphthong ends in an open vowel (a, o, or u) and is replaced by an unstressed vowel. Thus, ïoˆavea is true, but so is ioˇèra, (since the initial a of avea is unstressed, the initial e of era however stressed). The great Menichetti also expresses sympathy for this way in the case of a stressed following syllable.[27]
A closer look at the total of 60 verses affected by this phenomenon reveals a remarkable inconsistency even within the same authors. This is true - as Robey (2000: 116) differentially demonstrates - not only for Petrocchi, but also for Fasani. We have taken a closer look and not only written out all of these verses, but also consulted the current editions of a Sanguineti (2001) and Inglese (2016) in this regard. The result will hardly surprise you.
While Sanguineti does not set any dieresis at all in his edition, i.e. bypasses the phenomenon, Inglese essentially follows the second variant propagated by Lanza in the Inferno and Paradiso, but Petrocchi's first one in the Purgatorio. (Only the author could clarify whether this is an oversight). Here, too, therefore, no clear direction emerges.
We have therefore, in the spirit of Ockham's Razor, opted for Robey's solution and, with two justified exceptions[28], put a sineresis with dialefe in all cases of a falling diphthong followed by a vowel. The author rightly argues that it makes sense to treat the phenomenon uniformly, especially since this second variant is also prominent elsewhere in the Divina Commedia.[29] To us it also appears more visually appealing, i.e. less artificial.
For this adaption we do not need any further special except for a resolution of the dieresis on I in two cases.
- [16]We do not want to discuss here those that are anchored in the syntax of the Italian language and are thus recognizable in the text, such as phonosyntactic forms that influence the verse structure of an extensional (prostesi, epentesi, epitesi) or restrictive (aferesi and elisione, apocope, sincope) nature. We are concerned only with those possibilities that do not appear in the text, such as the sinalefe, the dialefe, the hiatus and the intervocalic i.
- [17]The handbook by Richard Lansing (2010): 477 contains a very concise account, which we will follow for now.
- [18]The stricter special case involves two vowels that follow each other, which should actually be counted separately, i.e. as two syllables, but count as one.
- [19]Here, too, there is the stricter interpretation. It occurs when both vowels, such as in the diphthong , usually count as one syllable, but are counted as two syllables.
- [20]Two special cases, the so-called hiatus and the intervocalic i, will be discussed later.
- [21]We have deliberately spoken here of numerical evaluation and not pronunciation, since the pronunciation is independent of the evaluation, i.e. lies in the freedom of the reader. This means that a Dieresis can be spoken like a vowel as well, yes mostly is spoken so, and with the Dialefe the vowels can be stretched differently, from the quasi monosyllabic pronunciation of a Diphthong up to the clear syllable division.
- [22]Even then, as Aldo Menichetti (2013) points out in chapter 5 on Sinalefe and Dialefe, using the example of verse Inf. 18, 117 (che non parëa s'era laico o cherco versus che non parea s'era laïco o cherco), there remain verses in which not only one scansion is justifiable. This is tantamount to saying that there can be incompatible and yet fully adequate metrical solutions for a verse. In such cases, it sometimes helps to consult the complete work, i.e., to examine how the author otherwise usually treats our two metrical figures. Sometimes, however, the decision remains subjective. "Del resto," writes Menichetti (2013), "in determinati casi, nemmeno il metricista più meticoloso ed esperto può illudersi di dare risposte assolutamente inoppugnabili." Here often only the repeated reading aloud of the critical variants can lead to the one that sounds more harmonious, appears more authentic and therefore finally pleases best (see also the next footnote).
- [23]A meticulous methodical review must be able to describe the hierarchy of the decision criteria that apply here. There is a complex interaction of basic rules, context-dependent diversions and preferences of the poet, whereby also the verse type or the position of the accents, which is changed by our metrical signs, must be taken into account. The whole thing is a dialectical process that begins with an initial determination, then, by scansion of the work, becomes acquainted with the author's typical patterns and preferences, and eventually, in a second round, compares what has been done once again with what has been learned. Finally, the result is a text that contains a specific sequence of word combinations, albeit one that can be modified in individual cases, and that is intended to be readable, sonorous, or as I would like to say - musical. The reader will find the result of this work in our color-coded edition. In a later, partially annotated supplementary version, those rare, likewise practicable alternatives can additionally be made available in the form of footnotes.
- [24]The standard work by Aldo Menichetti (1993), which is still valid today, the textbook by Pietro G. Beltrami (2011) and the historically informative volume by Francesco Bausi and Mario Martelli (1993) should be mentioned here. For the dedicated beginner, I would recommend without hesitation the very clear treatise by Giuseppe Sangirardi and Francesco De Rosa (2002), which is unfortunately out-of-print.
- [25] https://ota.bodleian.ox.ac.uk/repository/xmlui/handle/20.500.12024/2455.
- [26]Cf. Robey (2000: 108): "While we have ... 21 [instances] of Bëatrice, we have 43 instances of Beatrice with synaeresis." While here the vowel separation is at least visible in the original text as Dieresis (Bë-a-trice), this is omitted in the following example, which contains the same critical vowel combination "ea". The reader perceives that a vowel separation can occur once via Dieresis explicitly, another time via Hiatus implicitly, i.e. not shown in the original text, which confirms the observation made at the beginning that the outsider cannot know which linguistic fact is present for one and the same vowel combination.
- [27]Menichetti (1993: 249): „Quando poi l’attacco vocalico è rilevato, come in «che ’l cibo ne soleaˇèssere addotto» Inf. 33 44 … si ha decisamente dialefe, non dieresi.”
- [28]Paradiso 3 87 and Paradiso 27 44. In the first example crïa there is a hiatus, but here Petrocchi resolves it as dieresis. The second case pïo reads more elegantly with this metrical sign. Also, we find it consistently in a comparable constellation (adjective or proper noun without final rhyme position).
- [29]Robey (2000: 116/117): „The important issue is that final accented falling diphthongs followed by an initial vowel regularly count as two syllables in the Divine Comedy, and it seems reasonable to conclude that they should all be treated the same way, either as diaereses followed by sinalefe or as synaereses followed by dialefe; my own preference … is for the latter, as the simpler and more consensual solution.”