A summary of comparisons

I imagine a table set before me, in three columns. The table will present property by property the issues I have outlined in the past 6 weeks. I think that the jury should be able to come to a verdict on whether one method uses and explains more of the embedded information than another.

The property column below defines the question or decision. Each strategy must deal with the information, what is used and what is ignored. Both methods (Sequencing - Col 2 vs Reciting Note - Col 3) ignore some information. Each must also deal with the objective decisions, the musical phrases, the ease of perceiving the result, the ease of the use of musical language, the issue of beauty, of harmonic implications, and other issues.

Property / Decision	Sequencing strategy (Burns, Jacobson)	Reciting Note strategy (Haïk-Vantoura)
Distinguish the function of accents below the text from those above	No distinction in function.	Accents below the text determine the reciting note, from c (21 books only) to d, e, f, g, A, B, C (all books). Those above are ornaments.
Observation of every variation in accent placement	Pays attention to differences in placement.	Does not pay attention to differences in placement below the text. Both ב֖ and ב֭ change the reciting note (in this case to g). Ornaments may have some differences depending on placement. Zarqa and zinor are identical.
Double merkha	one accent among many.	considered equivalent to merkha.
Sof pasuk and Paseq	are observed as relating to the accents.	are not part of the music.
Dropping of observed accents	Burns drops the first of a pair sometimes.	Never drops a defined accent. Accents below the text are completely transparent. Sight reading is as easy as solfege.
Pitch	Accents are at a fixed pitch.	Accents define and are relative to the reciting note
Tonality	Jacobson is a single tonality, major for Torah, minor for haftarah. Burns has shades of possible tonal changes.	Observes tonic, supertonic, and subdominant as cadence points. Dominant and sub-mediant allow for tonality to  vary. Multiple modes supported but rationale for their use is subjective. Tonality immediately suggests clear musical shape.
Sequences of accents	are complex slurred combinations of notes, melismas.	below the text define the reciting note and thus allow sequences to be explained as intervals and triads. Melismas may occur with ornaments.
Variations in accent sequences	are explained with great complexity.	are explained as musical phrases.
Agreed senses	Silluq and Atnah as pauses. Recognizes common pairings and preparation.	Silluq (tonic) and Atnah (subdominant) as cadences. Recognizes common pairings (e.g. revia mugrash) and preparation (e.g. ole veyored prepared by revia or zarqa). Interpretation is different because the yored is equivalent to merkha and thus comes to a cadence on the second (supertonic).
Silluq and Metheg (an error in Unicode design).	Distinguished by reader, not mentioned by Burns, because he uses sof pasuk as terminus.	Distinguished by reader. Metheg is not a note and should not pull the reciting note prematurely to the tonic, but silluq in the middle of a phrase is just a note, not a cadence.
Beauty	the human voice and its flexibility.	the human voice and the sense of tonality.
Tone of voice	not noted.	The tonic (e) for the sense of home,  the subdominant (A) for the sense of rest,  the supertonic (f) for a briefer rest, (in the 3 books only) and always prior to the subdominant if both are in the verse, the dominant (B) for proclamation, the submediant (C) for emotional appeal, the lower notes as preparation for the phrase, sometimes leaping a seventh or an octave for joy.
Passing the test of the wandering mode, tonus peregrinus	Does not pass.	Passes perfectly. Psalms 114 in plainsong and chanting tradition is anticipated by the variable reciting note decision. This surviving melody suggests five of the reciting notes (B g A f e). (SHV did not make this connection. It might have shortened her work. That she arrived here is impressive.)
Efficiency	does not use the consistent placement above and below.	uses more information implicit in the accents.
Clarity	takes experience to hear.	takes experience to hear, may allow the noting of allusions between texts.
Connections between texts	denied.	affirmed and obvious, e.g. the attacca noted for Psalms 115, the clear knowledge of the speeches by the narrator of Job, the highlighting of strophe and structure in poetry and story.
Consistency	has many arcane ‘rules’.	an accent always sounds and is always the same definition within the current reciting note.
Hierarchy	emphasized.	not relevant.
Measurement	not defined.	allows the simple concept of syllables per reciting note as a measure of a verse.

Addendum: For something different, have a listen to the rendering of Psalm 1 on the home page of TiberianHebrew.com, and for contrast, consider this. More work to be done.

Related posts The ladder, in five parts Introduction, Going up to the subdominant, Going down from the subdominant, Verses without a rest on the subdominant, and further subdivisions for long verses. And Recitation analysis.

The results of my research in the use of software to read the Hebrew Bible

I have written much on this blog of my progress in deciphering the meaning of the signs in the text of the Hebrew. Now for a simple example of uncertainty, science, and testing of the results. There is much more to do - particularly to teach scholars how to read music.

William Wickes writes in the 19th century on the accents in the Hebrew text and he has studied many manuscripts directly. You know what this would have been like in the 19th century. Travel, study, notes, and so on, a long process. The dedication and industry of such work is commendable and born I am sure of love of the subject (at least).

I won't go into detail on his naming of the accents and so on, but just move immediately to a single claim. He makes the claim (page 23) that "Logically, a verse may be closely connected with the one preceding or following it; but musically and accentually no such connection exists."

This is a testable claim. Science looks at the data and performs experiments to see if general statements such as this can stand scrutiny. Perhaps one exception would prove that such a general rule cannot stand. But there is not just one, there are examples on every page of scripture that deny that general statement. These could be proven by naming the accents that begin the verses of Proverbs 2. But you would not understand and neither would I want you to struggle through such an explanation.

As I demonstrated a few days ago (before I had even heard of Wickes' book though I had considered others such as Jacobson), the music tells us so much more simply than an arcane explanation. The music sings the complex connections between verses. No example I have looked at fails to illustrate this far in excess of what any 'punctuation' could do.

Simple conclusion. Tradition has lost the meaning of the signs. Sure, it knows they are somehow exegetical and musical, but it can no longer communicate this to the masses that would love to hear how that works. It has lost its way. If I have to read a complex treatise with 25 unpronounceable words and contorted 'simplification' about conjunctive and disjunctive hierarchical roles of unreadable diacritics, then I simply won't do it. If I have to listen to transparent and beautiful music and don't have to be told what I am hearing because it is obvious to me, then I will hear and love the result.

[Addendum] Tradition (Helmut Richter) occasionally results in a clear statement of purpose of the cantillation marks, but only clear from its own point of view. Then it states what can only be interpreted as a complete escape clause: "the real nesting levels stay nowhere constant and can jump up and down by arbitrary amounts."

Vantoura is transparent. The rules are easily learned. Sight reading is easily learned. The result is beauty. What's to argue about? Tradition is struggling under a falsely engineered explanation.

This results in explanations like the following: A sequence like officer servant duke king duke servant king servant emperor is fully permissible.

The scholars are not musicians and therefore will argue or explain only from that diacritic point of view. They will ask you to read a book instead of showing you the beauty of the word.

I have almost completed my transcription of the whole Bible. [The first complete set of files in the default mode should be done by the end of the first week in September.] My work is scientific and therefore repeatable by any competent programmer. It need not be lost. It follows a clear set of transcription rules such that no information is lost in the resulting music. (I could even keep the qere-ketiv distinctions but for the most part, though the programming supports this, I have chosen not to.)

If you want to see the work, the xml files are all at this link. To read them as music, you need a music program such as Musescore version 2 or above.

My work is dependent on the work of Suzanne Haik Vantoura, as well as the development of the MUSIC-XML language. Lots of dependencies for which I am grateful. When I began, I had no idea that this would be possible to complete. Now it is clear that it is for I have done it.

The entire Bible, chapter by chapter, is now available in Music XML form on my shared drive.

What are these accents in the OT?

There is little sense I suppose in arguing. It was Will Durant who said that "History is mostly guesswork, and the rest prejudice". I prefer to avoid the arguments in the religious world.

But I do ask: What do all these accents in the Old Testament signify? With thanks to some negative comments on Haik-Vantoura's system in the Academia site, I now know there are a few more resources to study on the accents - e.g. this one from archive.org: A treatise on the accentuation of the three so-called poetical books on the Old Testament, Psalms, Proverbs, and Job, with an appendix containing the treatise, assigned to R. Jehuda Ben-Bil'am, on the same subject, in the original Arabic.

Well, its not in the original Arabic or I would not be reading it. And it's not too likely that I will read the whole thing very quickly in any case. But a few paragraphs maybe. It will be a counterfoil to Jacobson.

One thing is sure, if it were not for Vantoura's system, I would never (and I mean never) have studied or appreciated these signs. But the music is very revealing. And the more I reveal, the more people will see and hear that there is much to her system that reveals the mind of the poet.

The negative comments I received on my paper (copy is here) were these. The first led me directly to the above link (though his link is commercial and incomplete.)

I didn't have time to comment before the session closed. I wanted to point out an error regarding the two sets of te'amim in the Decalogue. What you missed in the paper is that the te'amim have another function, that of parsing the verse into lexical units. For more on this you should have a look at the following classic work by William Wickes: here The two sets of te'amim reflect two reading modes, one to be used in synagogue, and one for study. The reading in synagogue splits the Decalogue more obviously into 10 distinct utterances, and so necessitates a different set of te'amim of accommodate the different lexical parsing.

Now this is news to me and quite different from Vantoura's suggestion. Worth some consideration.

My response: Thank you for the interesting observation. This is certainly a possibility. I did not expect much discussion. Vantoura considered the extra accents as music in any case. I suspect she and her father, a rabbi, did not know this other possible usage. Besides modern (i.e. post 12th century) explanations of the accents, do you have some idea of the origin of these two sets of them in the Decalogue?

The second is less revealing:

I can not congratulate you. The theory of Haïk-Vantoura is based on wrong data. Vantoura knew little about function of the te'amim. She interpreted them incorrectly. Vantoura's reconstruction differ fundamentally with the synagogue practice of reading Torah and with the opinion of the leading researchers on cantillation marks.

This last sentence is a true statement. Unfortunately, the first statements are guesswork and prejudice.

My response: I understand that what you say is true. But I would hope you could give more reason than just disagreement. Vantoura's interpretation is the inference of a musician and it reveals function that is quite brilliant and beautiful. It also shows a carefully coded internally consistent system that is typical of human engineering. What is the right data and how can it be inferred and confirmed as correct?

The test case for Vantoura's methods remains for me their application to Psalm 114 where they resolve into the famous tune: tonus peregrinus. See more in The Songs of Ascents by David Mitchell (Precentor, Holy Trinity, Brussels).


I should also say that her method leads to the ability to sight-read the accents and to perform them with ordinary people. We have done so at St Barnabas on several occasions in both Hebrew and English. One would not want to do this every week, but the results strike people deeply and make the Old Testament lessons and psalms stand out in all their drama, foreignness, and beauty.

Preliminary tables comparing Haik-Vantoura with traditional cantillation

Here’s a couple of preliminary tables comparing Haik-Vantoura with traditional cantillation. The percentages are summarized from the 16.5% of the Bible that I currently have in my database. 

I have noted which signs are conjunctive and which are disjunctive. No sign appears in both columns.

Next I will work through several examples to see if the traditional ‘explanation’ is better ‘explained’ by the musical scheme of Suzanne Haik-Vantoura. I have never seen such confusion in the online resources. (If you seek pain, compare Mechon-Mamre, Wikipedia, and other sources for cantillation and coding.) I do not believe that this has to be so complex.

This first set of signs is the set for the 21 books. Music will separate the text note by note from the prior text, allowing an accent at the equivalent of a ‘bar line’. Each of the signs to the left changes the reciting note. The music for a verse always returns to the tonic E. The mid-point of most verses cadence on subdominant A. The percentage below is calculated as number of verses in the sample (about 15% of TNK) over the total verses in the sample 2527 poetry + 1276 prose = 3803 total, at present.

	per Haik-Vantoura		Usage – 0 means never		Traditional explanation
Unicode (decimal)	Sign	Shape/ Note	Poetry  %/vs	Prose %/vs	Disjunctive (level)	Conjunctive serves
1447	Darga	כ֧ c	0	12		tevir
1435	Tevir	ד֛ d	0	29	2/3
1469	Silluq	עֽ e	100	100	1
1445	Merkha	פ֥ f	67	74		silluq, tifcha
1430/53	Tifcha	ג֖ g#	23	87	2
	Dehi	ג֭	56	0
1425	Atnah	א֑ A	99	75	1
1443	Munah	ב֣ B	82	68	4?	everybody
1444/34	Mahpach	ך֤ C	21	36		pashta
	Yetiv	ך֚	0	6

Traditional cantillation and Haik-Vantoura agree on the importance and the cadences effected for both silluq and atnah. From there on there is little agreement in function apart from the musical accent produced by a change of note or a melisma. For traditional cantillation, the tifcha is a ‘lesser cadence’ (Jacobson 48). For SHV it is the leading tone as a reciting note. It is not a cadence, though it certainly aids pronunciation. Traditional cantillation has no concept of a change in reciting note. This concept is signaled for SHV by the placing of the 8 (or 7 for the poetry) signs below the letters. It is fundamental to her theory.

For the 3 books, drop darga, tevir, add galgal. The remainder of the 7 note poetry scale is identical to the above except sharpen the f. Default mode is E minor but several modes are possible with both sets of signs. The combination of ole-veyored is quite frequent and forms a cadence on the supertonic (in the default mode, an f#).

Unicode (decimal)	Sign	Shape/ Note	Poetry %/vs	Prose %/vs	Disjunctive  (level)	Conjunctive serves
1450	Galgal	ד֪ d#	8	0		qarney para pazer gadol
1451+1445	Ole-veyored	ב֫ב֥	67 (often together)	74 (maybe separately)	2
1437+1431	Geresh-revia	ב֝ב֗	75	30	two in a row

Note 1442 ב֢ is exactly = 1450ב֪ . I have no coded examples of 1442 so far. I don’t know why 1442 exists.

Music joins, conjoins and disjoins its libretto in all sorts of ways. These are part of the mystery of language and creation as well as human creativity. The Word of God is meant to be sung so that in the singer it lives.

	per Haik-Vantoura		Usage – 0 means never			Traditional explanation
Unicode (decimal)	Sign	Shape	Poetry	Prose		Disjunctive (level)	Conjunctive serves
1426	Segol	ש֒	0	4		4?
1427	Shalshelet	שׁ֓	1	0.15		4?
1428	Zaqef Qatan	ז֔	0	70		2
1429	Zaqef Gadol	ז֕	0	40		2
1431	Revia	ר֗	76	30		3
1432/54	Zarqa	ז֘	6	0	there is no distinction in these pairs except timing of the melisma and which books they are used in	3
	Zinor	ז֮	8	4		3
1433/48	Pashta	ף֙	0	63		3
	Qadma	ק֨	11	23		3
1436/7	Geresh	ג֜	0.04	14		4
	Geresh Muqdam	ג֝	66	0		4
1438	Gershayim	ג֞	0	6		4
1439	Qarney Para	ק֟	0	0	like horns	4
1440	Telisha Gedola	ת֠	0	3		4
1441	Paser	ף֡	3	2		4
1449	Telisha Qetana	ט֩	0	6			Zarqa
1451	Ole	ל֫	13	0		2
1452	Illuy	י֬	5	0			=Munah Jacobson p 929
1455	Masora Circle	ס֯ not used	0	0