We conceptualize or model a line of Greek hexameter as a grid:
One long syllable is equal to two short syllables, and all long syllables are equal.
It might be compared to musical notation (measure ≃ foot):
– ⏑ ⏑ – ⏑ ⏑ – ⏑ ⏑ – ⏑ ⏑ – ⏑ ⏑ – –
Imagine ties to indicate the boundaries of words, still composed of a sum of quarter and eight notes:
Performance of music does not always match the literal notation. Performers lengthen, shorten, delay notes for various reasons.
Does actual performance of Greek hexameter accord with the grid-based conceptualization?
If not, in what ways does it tend to differ?
Internet Archive
ἦμος δ’ ἠριγένεια φάνη ῥοδοδάκτυλος Ἠώς,
καὶ τότ’ ἔπειτα νομόνδ’ ἐξέσσυτο ἄρσενα μῆλα,
θήλειαι δὲ μέμηκον ἀνήμελκτοι περὶ σηκούς·
ἄνδρα μοι ἔννεπε, μοῦσα, πολύτροπον, ὃς μάλα πολλὰ
πλάγχθη, ἐπεὶ Τροίης ἱερὸν πτολίεθρον ἔπερσεν·
πολλῶν δ’ ἀνθρώπων ἴδεν ἄστεα καὶ νόον ἔγνω,
πολλὰ δ’ ὅ γ’ ἐν πόντῳ πάθεν ἄλγεα ὃν κατὰ θυμόν,
ἀρνύμενος ἥν τε ψυχὴν καὶ νόστον ἑταίρων.
Praat – https://www.fon.hum.uva.nl/praat/
Video demo: https://archive.org/details/sedes-praat-odyssey-9_458-459
Data so far at https://github.com/sasansom/phonetics
work,book_n,line_n,word_n,word,syllable_n,syllable,t_begin,t_end Od.,9,437,1,ἦμος,1,ἦ,3.539302,3.866121 Od.,9,437,1,ἦμος,2,μος,3.866121,4.123347 Od.,9,437,2,δʼ,1,δʼ,4.358528,4.510836 Od.,9,437,3,ἠριγένεια,1,ἠ,4.614314,4.921169 Od.,9,437,3,ἠριγένεια,2,ρι,4.921169,5.124992 Od.,9,437,3,ἠριγένεια,3,γέν,5.124992,5.324269 Od.,9,437,3,ἠριγένεια,4,ει,5.324269,5.645382 Od.,9,437,3,ἠριγένεια,5,α,5.645382,5.785575 Od.,9,437,4,φάνη,1,φά,5.943990,6.042322 Od.,9,437,4,φάνη,2,νη,6.042322,6.544669 Od.,9,437,5,ῥοδοδάκτυλος,1,ῥο,6.786806,7.124554 Od.,9,437,5,ῥοδοδάκτυλος,2,δο,7.124554,7.355420 Od.,9,437,5,ῥοδοδάκτυλος,3,δάκ,7.436495,7.532689 Od.,9,437,5,ῥοδοδάκτυλος,4,τυ,7.746454,7.872108 Od.,9,437,5,ῥοδοδάκτυλος,5,λος,7.872108,8.244817 Od.,9,437,6,Ἠώς,1,Ἠ,8.411553,8.652478 Od.,9,437,6,Ἠώς,2,ώς,8.652478,9.013262
| work | book_n | line_n | word_n | word | syllable_n | syllable | t_begin | t_end |
|---|---|---|---|---|---|---|---|---|
| Od. | 9 | 437 | 1 | ἦμος | 1 | ἦ | 3.539302 | 3.866121 |
| Od. | 9 | 437 | 1 | ἦμος | 2 | μος | 3.866121 | 4.123347 |
| Od. | 9 | 437 | 2 | δʼ | 1 | δʼ | 4.358528 | 4.510836 |
| Od. | 9 | 437 | 3 | ἠριγένεια | 1 | ἠ | 4.614314 | 4.921169 |
| Od. | 9 | 437 | 3 | ἠριγένεια | 2 | ρι | 4.921169 | 5.124992 |
| Od. | 9 | 437 | 3 | ἠριγένεια | 3 | γέν | 5.124992 | 5.324269 |
| Od. | 9 | 437 | 3 | ἠριγένεια | 4 | ει | 5.324269 | 5.645382 |
| Od. | 9 | 437 | 3 | ἠριγένεια | 5 | α | 5.645382 | 5.785575 |
| Od. | 9 | 437 | 4 | φάνη | 1 | φά | 5.943990 | 6.042322 |
| Od. | 9 | 437 | 4 | φάνη | 2 | νη | 6.042322 | 6.544669 |
| Od. | 9 | 437 | 5 | ῥοδοδάκτυλος | 1 | ῥο | 6.786806 | 7.124554 |
| Od. | 9 | 437 | 5 | ῥοδοδάκτυλος | 2 | δο | 7.124554 | 7.355420 |
| Od. | 9 | 437 | 5 | ῥοδοδάκτυλος | 3 | δάκ | 7.436495 | 7.532689 |
| Od. | 9 | 437 | 5 | ῥοδοδάκτυλος | 4 | τυ | 7.746454 | 7.872108 |
| Od. | 9 | 437 | 5 | ῥοδοδάκτυλος | 5 | λος | 7.872108 | 8.244817 |
| Od. | 9 | 437 | 6 | Ἠώς | 1 | Ἠ | 8.411553 | 8.652478 |
| Od. | 9 | 437 | 6 | Ἠώς | 2 | ώς | 8.652478 | 9.013262 |
SEDES does a lot—in particular it gives you the scansion of hexameter texts. (Using an algorithm by Hope Ranker.)
sedes$ src/tei2csv Od. corpus/odyssey.xml | head -n 20 work,book_n,line_n,word_n,word,lemma,sedes,metrical_shape,scanned,num_scansions Od.,1,1,1,ἄνδρα,ἀνήρ,1,–⏑,auto,1 Od.,1,1,2,μοι,ἐγώ,2.5,⏑,auto,1 Od.,1,1,3,ἔννεπε,ἐνέπω,3,–⏑⏑,auto,1 Od.,1,1,4,μοῦσα,Μοῦσα,5,–⏑,auto,1 Od.,1,1,5,πολύτροπον,πολύτροπος,6.5,⏑–⏑⏑,auto,1 Od.,1,1,6,ὃς,ὅς,9,–,auto,1 Od.,1,1,7,μάλα,μάλα,10,⏑⏑,auto,1 Od.,1,1,8,πολλὰ,πολύς,11,––,auto,1 Od.,1,2,1,πλάγχθη,πλάζω,1,–⏑,auto,1 Od.,1,2,2,ἐπεὶ,ἐπεί,2.5,⏑–,auto,1 Od.,1,2,3,τροίης,τρέω,4,––,auto,1 Od.,1,2,4,ἱερὸν,ἱερός,6,⏑⏑–,auto,1 Od.,1,2,5,πτολίεθρον,πτολίεθρον,8,⏑⏑–⏑,auto,1 Od.,1,2,6,ἔπερσεν,πέρθω,10.5,⏑––,auto,1 Od.,1,3,1,πολλῶν,πολύς,1,––,auto,1 Od.,1,3,2,δ’,δέ,3,,auto,1 Od.,1,3,3,ἀνθρώπων,ἄνθρωπος,3,–––,auto,1 Od.,1,3,4,ἴδεν,εἶδον,6,⏑⏑,auto,1 Od.,1,3,5,ἄστεα,ἄστυ,7,–⏑⏑,auto,1
| work | book_n | line_n | word_n | word | lemma | sedes | metrical_shape | scanned | num_scansions |
|---|---|---|---|---|---|---|---|---|---|
| Od. | 1 | 1 | 1 | ἄνδρα | ἀνήρ | 1 | –⏑ | auto | 1 |
| Od. | 1 | 1 | 2 | μοι | ἐγώ | 2.5 | ⏑ | auto | 1 |
| Od. | 1 | 1 | 3 | ἔννεπε | ἐνέπω | 3 | –⏑⏑ | auto | 1 |
| Od. | 1 | 1 | 4 | μοῦσα | Μοῦσα | 5 | –⏑ | auto | 1 |
| Od. | 1 | 1 | 5 | πολύτροπον | πολύτροπος | 6.5 | ⏑–⏑⏑ | auto | 1 |
| Od. | 1 | 1 | 6 | ὃς | ὅς | 9 | – | auto | 1 |
| Od. | 1 | 1 | 7 | μάλα | μάλα | 10 | ⏑⏑ | auto | 1 |
| Od. | 1 | 1 | 8 | πολλὰ | πολύς | 11 | –– | auto | 1 |
| Od. | 1 | 2 | 1 | πλάγχθη | πλάζω | 1 | –⏑ | auto | 1 |
| Od. | 1 | 2 | 2 | ἐπεὶ | ἐπεί | 2.5 | ⏑– | auto | 1 |
| Od. | 1 | 2 | 3 | τροίης | τρέω | 4 | –– | auto | 1 |
| Od. | 1 | 2 | 4 | ἱερὸν | ἱερός | 6 | ⏑⏑– | auto | 1 |
| Od. | 1 | 2 | 5 | πτολίεθρον | πτολίεθρον | 8 | ⏑⏑–⏑ | auto | 1 |
| Od. | 1 | 2 | 6 | ἔπερσεν | πέρθω | 10.5 | ⏑–– | auto | 1 |
| Od. | 1 | 3 | 1 | πολλῶν | πολύς | 1 | –– | auto | 1 |
| Od. | 1 | 3 | 2 | δ’ | δέ | 3 | auto | 1 | |
| Od. | 1 | 3 | 3 | ἀνθρώπων | ἄνθρωπος | 3 | ––– | auto | 1 |
| Od. | 1 | 3 | 4 | ἴδεν | εἶδον | 6 | ⏑⏑ | auto | 1 |
| Od. | 1 | 3 | 5 | ἄστεα | ἄστυ | 7 | –⏑⏑ | auto | 1 |
https://sasansom.github.io/sedes/odyssey_all.html
Align to sedes grid
| Performer | Nominal syllable length | min | mean | max | stdev |
|---|---|---|---|---|---|
| Hélène Kémiktsi | Short | 0.105 | 0.317 | 0.690 | 0.125 |
| Hélène Kémiktsi | Long | 0.088 | 0.378 | 0.779 | 0.145 |
| John F. C. Richards | Short | 0.078 | 0.243 | 0.501 | 0.105 |
| John F. C. Richards | Long | 0.096 | 0.390 | 0.690 | 0.118 |
…the simple tone patterns of these psychological experiments and in the far more complex prosody of speech, the definitional 1:2 ratio is subject to modification reflecting the higher level hierarchical structure of the pattern, specifically the grouping of the pattern elements into rising or falling rhythms and the demarcation of higher level structures. These modifications need to be factored out in evaluating the accuracy not only of 1:2 ratios but also off 1:1 ratios in actual production tokens. Similarly, in music performance, the time values of notes are not precisely respected for a variety of reasons, including accent marking, demarcative decelerando, and expressive factors. It follows that music is abstractly organized in terms of simple integral ratios and concretely implemented by complex nonintegral ratios. Just like quantitative verse.