by Professor Shlomo Karni / In her delightful book “Eats Shoots & Leaves”, Lynn Truss drives home the point with the following story: A panda walks into a restaurant, orders a sandwich, eats it, then, on his way out, draws a gun and fires two shots. The alarmed waiter asks “Why?” In reply, the panda pulls out a badly punctuated wildlife manual and shows him the entry for ‘panda’: “Large black-and-white mammal, native to China. Eats, shoots and leaves.” “So,” writes Truss, “punctuation really does matter, even if it is only occasionally a matter of life and death.”

It’s The Punctuation,…!

Print Friendly, PDF & Email

Guest post by Professor Shlomo Karni

Shlomo Karni was Professor of Electrical Engineering and Religious Studies at University of New Mexico until his retirement in 1999. His books include Dictionary of Basic Biblical Hebrew:Hebrew-English (Jerusalem: Carta, 2002).

In her delightful book “Eats Shoots & Leaves”, Lynn Truss drives home the point with the following story: A panda walks into a restaurant, orders a sandwich, eats it, then, on his way out, draws a gun and fires two shots. The alarmed waiter asks “Why?” In reply, the panda pulls out a badly punctuated wildlife manual and shows him the entry for ‘panda’: “Large black-and-white mammal, native to China. Eats, shoots and leaves.”

“So,” writes Truss, “punctuation really does matter, even if it is only occasionally a matter of life and death.”

Punctuation in the Bible really does matter. Between the 8th and 10th centuries CE, our sages in Tiberias, the so-called “Massoretes” (from the Hebrew “Massorah”, transmittal from generation to generation), formalized the system of vowelling for the correct pronunciation of the words, and the system of musical notes (trope) for the words. With its various vocal inflections, pauses and stops, the trope gives the vowelled sentence its structure and meaning, much like commas and periods.

As just one example of the usefulness of the trope, consider the story of Sodom and Gemorrah. Lot, Abraham’s nephew, is urged to leave the place before its destruction, but “he delayed.” (Gen. 19:16). The Hebrew word for “delayed” is a very long verb (7 letters), giving immediately a sense of the delay. The trope here is the longest (some 24 separate notes!), further emphasizing the idea of delay.

However, a language is a dynamic, living entity. In its everyday use, certain errors may — and do — crop up. This is true with Biblical Hebrew in several instances.

One such remarkable error is found in the famous chapter 40 of Isaiah, “Comfort ye, comfort ye, My people.” Verse 3 reads , “The voice of him that crieth in the wilderness, ‘prepare ye the way of the Lord’. “ (King James’ version, et al.)

From this source, we have the idiomatic expression: “A voice in the wilderness”, meaning “someone’s words not being heeded.” For example, “Churchill’s early warning of the danger of Nazism was a voice in the wilderness.” Similarly, the motto of Dartmouth College on its seal proclaims nobly, “Vox Clamatis in Deserto.” Ernst Bloch wrote a musical composition named “Voices in the Wilderness”; likewise, Handel’s “Messiah” uses that verse from Isaiah in the same sense.

The trouble with this expression is that… it is wrong. The trope renders this verse as, “A voice calls out: ‘In the desert clear a road for the Lord! In the wilderness level a highway for our God!’”

The symmetry of the two halves — “In the desert clear a road”, and “In the wilderness level a highway” – is a very common structure of Biblical poetry. The trope reflects this, properly stressing the structure.

Yet, despite the symmetry of the sentence and despite the trope, “A voice in the wilderness” remains an established idiom. And, as if this were not enough, the same idiom and mistake exist in Hebrew (whence King James’ version got it) to this very day.

Let us hope that this note does not remain…a voice in the wilderness.

About Shlomo Karni

29 comments

  1. I’m not sure I’m reading it correctly, but it seems that Rashi on 40:3 implies that he’s reading the verse in line with the idiom that you reject. Please advise?

  2. When I was laining Shir Hashirim after the sedarim I noticed that the trop occasionally appear to incorrectly break some of the parallelism. I checked JPS and they do here and there parse lines differently from the masoretic trop. I suppose that this is more common when you have biblical poetry combined with obscure terms and lost idioms.

  3. Interesting example. Scanning the summary of Christian translations available at http://bible.cc/isaiah/40-3.htm one can see both forms in translation.

    It is worth noting there are parallels in each of the 4 Gospels, which means that an analysis of the Greek Gospels would likely shed light on the pre-Massorete oral mesorah.

    A quick scan of Mikraot Gedolot indicated both readings. As Menachem above points out, Rashi seems to connect Kol Koreh and ba’Midbar; whereas, Ibn Ezra seems to see Kol Koreh standalone.

    I always find the mechon-mamre online version useful because הפיסוק כאן לפי טעמי המקרא.

  4. It’s worth keeping in mind that the Masoretic trope serves in many places as an interpretation of the original text. Just like Rashi’s explanations that differ with the earlier Onkelos aren’t necessarily “wrong”, so too it’s not obvious to me that the interpretation that connects bamidmar to kol kore is “wrong” — just different than the Masoretes’ interpretation.

  5. Ah. The New Jerusalem Bible (a scholarly Catholic translation) renders Isaiah 40:3 as:

    “A voice cries. ‘Prepare in the desert
    a way for Yahweh.
    Make a straight highway for our God
    across the wastelands”

    But, it has a footnote:

    “The identity of the voice, obeying the order in verse 2 has been deliberately left mysterious by the prophet. The evangalists, see Matthew 3:3, John 1:23, quote this text in its LXX [Septuagint] form: ‘A voice of one who cries in the dessert’.”

    And, indeed, they keep that form in rendering Matthew and John in their translation, with no further footnotes beyond the cross-reference to Isaiah 40:3.

  6. For the avoidance of doubt, the Christians had a theological reason to prefer the linkage with ba’Midbar; and by the time of the Masoretes, the Jews and Karaites had a theological reason to prefer no linkage.

    It is here where Bible Criticism techniques — often denigrated on this blog — are helpful: e.g. Prof Karni’s observation regarding the (literary) “structure of Biblical poetry”.

  7. “As just one example of the usefulness of the trope . . . The trope here is the longest (some 24 separate notes!), further emphasizing the idea of delay.”

    such “usefulness” is . . .

    a) only for ashkenazim. for sephardim the shalsheles has “normal” legnth. e.g., for iraqis, egyptians and syrians it’s a mere 6 notes, if that much. so did moshe rabbeinu chant it long like ashkenazim or short like sephardim?

    b) post facto. the reason there is a shalsheles is a function of the fact that the shalsheles is a substitute for segol. a segol must be preceded by a zarka, but in cases where syntactic parsing causes a segol to be assigned to the first word of a pasuk, it is converted to a shalsheles. (note that in babylonian notation the symbol for shalsheles incorporated a corrupted shin, which was the babylobian symbol for שרי, i.e., segol)

  8. As I’ve layned over the years, I’ve noticed that very often (more often than not?) the trope is simply the only one that would fit those words.

  9. Rashi has it right, “The holy spirit calls, ‘in the desert clear G-d’s road…” adding, “the Jerusalem road — to return exiles to it.”

    BTW: So do Radak (R’ David Kimchi), and “metsudat Tzion”in my edition of Mikraot G’dolot.

    Thank you, Abba, for the notes (no pun intended) about the various Sephardic shalshelaot.

  10. Why is it that almost every frum child is taught that the trope is MiSinai?

  11. According to oral legend, there used to be a tennis court behind Good Samaritan Hospital which is next door to Hebrew Union College in Cincinnati. There was a sign there that said “doctors only no students allowed.” But the students played there anyway, reading the sign as “doctors only? No, students allowed.” By the time I got to HUC there was no tennis court and no sign, but it was a good example of how punctuation matters.

  12. >Why is it that almost every frum child is taught that the trope is MiSinai?

    Because it seems to be what the Talmud’s position on Nedarim 37b is, which is why it became the “traditional” position (or, that it was from Ezra in its alternate form).

  13. To Mo: If you attach ‘bamidbar’ to ‘kol koreh’, the symmetry of the two halves of the sentence is ruined: where is the parallel to ‘ba-aravah’?

  14. As an example from our own traditions about explanation and position of the trop, compare (contrast) the 13 middot(attributes)of mercy as read, understood, in festival, rosh hashanna, and yom kippur prayers, and how they are understood, as per trop, in the torah, (shemot, 34:6-7)?

  15. >To Mo: If you attach ‘bamidbar’ to ‘kol koreh’, the symmetry of the two halves of the sentence is ruined: where is the parallel to ‘ba-aravah’?

    Although parallelism was not entirely unknown (notably the Rashbam makes note of it) it was not widely recognized by Jewish or non-Jewish exegetes. The first to really delve into it was R. Azariah de Rossi (late 16th century) followed by Robert Lowth (mid 18th). So it’s not surprising that the KJV (and perhaps other translations) would not translate according to the “obvious” parallelism, much less according to the trope, the authority of which was subject of a raging debate among Christians at that time.

  16. Prof. Karni — I think you’re trying to have it both ways on Rashi. My reading of his terse gloss is that he reads “a voice in the wilderness” where the voice is God:
    רוח הקודש קורא במדבר דרך ירושלים

  17. Another point I wanted to raise: Nachum claims he sees how the trope fits the persukim. I, honestly, have never seen any rhyme or reason for the patterns. Does anyone out there have a clue what “algorithm” was used to decide the trope of pesukim? (Why/when kadma v’azla? Why/when gershayim? etc. etc. etc.)

  18. Re “Comfort ye, comfort ye, My people”: this rendering itself is an example of the “eats, shoots and leaves” misplaced-comma phenomenon. “Nahamu nahamu ammi” does not mean “Be comforted, be comforted, O my people.” Instead, it means “Give comfort, give comfort to my people.” In the King James version, there is no comma after the second “comfort ye.”

  19. Meir Weingarten

    similar situation: Arami Oved Avi.
    Is it: Arami, Oved Avi – lavan wanted to kill jacob
    or: Arami Oved, Avi – maybe avraham, a aramean nomad

  20. >Another point I wanted to raise: Nachum claims he sees how the trope fits the persukim. I, honestly, have never seen any rhyme or reason for the patterns. Does anyone out there have a clue what “algorithm” was used to decide the trope of pesukim? (Why/when kadma v’azla? Why/when gershayim? etc. etc. etc.)

    Trope has two functions. One, punctuation. It attaches and detaches words and phrases to greater and lesser degrees. Two, to chant it with a pleasant tune. The specific combinations are informed by both functions. The way people figured it out is mostly through analysis and comparison, since there really aren’t any early Masoretic treatises explaining the rules.

  21. BARUCH:

    “Does anyone out there have a clue what “algorithm” was used to decide the trope of pesukim?”

    the “algorithms” are well established (with the usual amiguous cases, exceptions, etc.) and there is much literature on the topic. for the basics see william wickes in english (free on google books) and r. mordechai breuer in hebrew

    (my second point in the comment i left above is that there is a shalsheles because that it was the “algorithm” calls for)

  22. to be pedantic (what calls for that if not a discussion about grammar!), arami oved avi is not a “comma” issue per se. The question is what part of speech is “oved” – is it an adjective modifying “arami,” with the whole phrase “arami oved” being the subject of the sentence, or is it a verb, with “arami” alone as the subject and “avi” the object. just look at your translation – it’s not “lavan, wanted to kill jacob,” but rather both have no commas.

    That said, I can imagine that the trop could choose one over the other, but I am not sure how the existing trop actually does so – would be happy to hear the explanation if it does.

  23. Actually, Arami Oved Avi does have a syntactical parsing that might be considered relevant. There is a pashta on Arami, which is a low level disjunctive ta’am (roughly a comma), whereas Oved clearly has a munach, which is the conjuctive ta’am to the Zakef Katon on Avi. That being said, the two meanings might both still be possible. For example, it could read “My wandering father was an Aramean” (well, reversed in the Hebrew syntax, but the point is that ‘wandering’ might be connected to father but still mean wandering), or it might alternatively mean “An Aramean – was the destruction of my father” or somesuch.

    Of course, looking to the trope to answer this kind of question is largely irrelevant. Clearly it has been used to mean both things in Jewish exegesis on the phrase. It isn’t even remotely uncommon for exegetical readings (including many in the Talmud) to horribly twist the plain meaning of the sentence, often going so far as to change word roots, change things like subject/object, etc. Changing syntax is peanuts in comparison.

    Furthermore, the trope itself sometimes goes against the clear syntax of the sentence. A classic example is in V’Zot HaBrachah: It say “Ur’tzon Shoch’ni, S’neh”, with the comma being a disjunctive ta’am. This is clearly incorrect – the disjunctive should be after Ur’tzon to yield the meaning “and the Will of the Dweller in the Bush” or somesuch (presumably referring to God’s first revelation to Moshe), rather than the odd “And the Will of the Dweller – Bush”. The explanations I have seen suggest that the trope was intentionally changed from the plain meaning of the sentence to combat the misconception that God actually physically dwells in a burning bush. *shrugs*

    The point is that trope is a useful tool and it often helps understand the underlying syntax of a verse, but it is not always the only authority we should use – there are multiple exegetical readings of many passages, and the trope itself sometimes contradicts the clearly correct syntax.

    Also, I’d just like to add my voice to the chorus of (presumably) ba’alei k’riyah who object to the constant use of the shalshelet in some drash – whether it’s talking about ‘delay’ or any other case (though I wonder how people would do it for ‘Vayishchat’ in Vayikra 8:23). It’s very clear why the shalshelet exists (as mentioned previously, it’s when you need a high level disjunctive ta’am that would normally be done by a segol, but you don’t have a word beforehand to take the zarqa), and the ‘specialness’ of its sound is an artifact of Ashkenazi custom. In fact, it is doubtful that any of the Ashkenazi trope modalities are very old, and they almost certainly have no basis in Mesorah.

  24. EMMA:

    the trop of arami oved avi is discussed at the end of r. mordchai’s breuer’s books on trop. presumably also in simcha kohut and michael perlman’s books, but don’t recall. and in general no one should even think about the grammar of the phrase without consulting prof. richard steiner’s article on it in moshe greenberg festschrift

    “arami oved avi is not a “comma” issue per se”

    i think sometimes we make too much out of “comma issues” with trop. according to the principle of continuous dichotomy continue to parse a pasuk with disjunctive trop (the commas, if you prefer) until we get down to 2-word units. so every three word unit is still going to get a “comma” regardless of whether or not it really belongs syntactically. iirc in some cases similarly structured (or even worded) psukim have different trop patterns because of issues relating to word and syllable counts, not syntax (e.g., in lists)

  25. MATLABFREAK:

    ditto.
    although this type of drash understanding is not entirely limited to vorts and divre torah. there is an article exploring this approach for the rare trops even in an academic journal like JQR (david weissberg in the 1960s), although i thought there were holes here too.

    “it is doubtful that any of the Ashkenazi trope modalities are very old”

    iirc eliyahu levita mentions shalsheles sung like 2 pazers (our sounds like 3 pazers?)

    also could be that some modalities are not that old even for sephardim (or all who accept tiberian mesorah), as babylonian mesorah doesn’t include some of the rare trops. (but this doesn’t mean they didn’t exist orally, as babylonian masoretes in general focussed on recording mostly mafsikim, so we don’t really know about their mesharesim)

    one other pet peeve. making a big deal our of merche kefula and then pausing, when in fact it is a meshares.

  26. Anyone want to read and record Sebastian Muenster’s transcription of Aahkenazi trope from 1524?

    http://onthemainline.blogspot.com/2009/10/what-did-torah-reading-sound-like-500.html

  27. To Meir Weingarten: The first interpretation is wrong – “oved” is in the “kal” (pa’al) stem. “To kill”, using this root verb, would be “ma’avid” – in the hif’il stem

  28. Re: a drash associated with a shalshelet: did you notice the shalshelet over “vayitmahamah” – Lot “lingered” (Gen.19:16)? Is this trope not adding to the sense of lingering ? And how about “Va-yema’en”, when Joseph resisted the advances of Potiphar’s wife (Gen. 39:8)? Does not the shalshelet strengthen the sense of his refusal? Or…? 😉

Leave a Reply

%d bloggers like this: