Home    Introduction    Talmudic Issues    Essays  

 Pamphlets    Parsha    Original Essays    Contact Us 

Pamphlet 9 - The Letters of the Torah

The Number of Letters

There is a famous saying, from the Zohar Chadash on Song of Songs (74d), that there are 600,000 letters in the Torah. The Megaleh Amukot (186) explained that these letters correspond to the 600,000 Jewish souls that exist (evidently, a person can have part of a soul because there are more than 600,000 Jews). He also suggested that the word Yisrael is an acronym for "Yesh Shishim Ribo Otiyot LaTorah" - "There are 600,000 letters in the Torah". The difficulty with this is that the Gemara in Kiddushin 30a says that there are 5,888 verses in the Torah. Even if each verse had 100 letters, and a quick check will reveal that the average is well below that, the Torah would still have less than 600,000 letters (see Chavot Yair 235).

Furthermore, our Torahs have 304,805 letters. This can be verified by counting and is recorded by the famous 10th century Masorete Aharon Ben Asher in his Dikdukei Taamim. For the Torah mentioned by the Zohar Chadash to have 600,000 letters it must be almost twice as long as our Torah. However, we have ancient Bibles such as the Septuagint (third century BCE) and the Samaritan Torah (from before Ezra) that are almost identical to our texts (we discuss the differences elsewhere). It is inconceivable that there was ever a Torah that was twice as long as the Torah we currently have. Rather than this saying being a statement about ancient Torah scrolls and therefore an indictment of ours, it is a puzzling statement that does not seem to describe any known or possible variant of the Torah.

The puzzle is solved, however, when we remember that this saying is recorded in kabbalistic books. Characteristic of this genre, the statement is referring to mystical issues and not the simple letters of the Torah. Some claim that it refers to the strokes a scribe requires to write the letters while others suggest that it refers to both the written and unwritten portions of a scroll. See R' Reuven Margoliyot's Hamikra Vehamesora ch. 12. Whatever the saying means, we can be certain that it does not mean that there are literally 600,000 letters in the Torah.

Letters and Words in the Torah

  Words Letters
Genesis 20,512 78,064
Exodus 16,723 63,529
Leviticus 11,950 44,790
Numbers 16,368 63,530
Deuteronomy 14,294 54,892




While we are on the subject of letters, let us mention that Rav Saadia Gaon wrote a poem about the letters of the Torah whose total of 792,077 does not match ours of 304,805. However, as R' Chaim Yair Bachrach pointed out in his Chavot Yair (235), Rav Saadia Gaon's list is impossible. The list has the alephs, gimmels, and zayins with almost the same frequency in the Torah while everything we know about the Hebrew language tells us that this is not naturally possible. Aleph is an extremely common letter while gimmel and zayin are not. The Tanach Yehoash has a list of how many times each letter appears in the Torah. Aleph appears 27,057 times while gimmel appears 2,109 times and zayin 2,198 times. Aleph is more than ten times more frequent than either gimmel or zayin. Similarly, we counted in Genesis ch. 1 and found aleph 158 times, gimmel 5 times, and zayin 11 times. From where Rav Saadia Gaon got his list we do not know. But he definitely did not get it from counting letters in his Torah. How he could have used that list and exactly what this sage meant remains a mystery.

Letters in the Torah

  Letters     Letters
א 27,057 ל 21,570
ב 16,344 מ 25,078
ג 2,109 נ 14,107
ד 7,032 ס 1,833
ה 28,052 ע 11,244
ו 30,509 פ 4,805
ז 2,198 צ 4,052
ח 7,187 ק 4,694
ט 1,802 ר 18,109
י 31,522 ש 15,592
כ 11,960 ת 17,949



Is this number of 304,805 letters in the Torah exact? Did G-d give Moshe a Torah with precisely that number of letters? We do not know for sure but we know that it was very close to that number. The reason we cannot be certain is twofold. First, the Gemara in Kiddushin 30a says that we are not experts in chaser and yeter. There are certain vowel sounds in Hebrew that can be spelled with (yeter) or without (chaser) an assisting letter. It is important to note that the presence or absence of this letter make no difference in terms of meaning and pronunciation. The words and verses mean exactly the same whether they are spelled chaser or yeter, which may be how these uncertainties crept in. Because of this, there are certain discrepancies between even good versions of the Torah in this respect. Beginning in the 8th century, the Masoretes tried to standardize the spelling of chaser and yeter words by recording them in their masoretic notes. Surprisingly, even some excellent manuscripts do not follow this Masora precisely (see R' Mordechai Breuer's introduction to The Aleppo Codex and the Accepted Text of the Bible, par. 20). However, this standardization of chaser and yeter came after the talmudic statement that we are not experts in them so the standardization is not final (see Rama, Orach Chaim 143:3). Therefore, there remain differences between texts in terms of chaser and yeter. Again, it is important to emphasize that these minor differences do not change the meaning or pronunciation of the words (see ).

The second reason that there might be slight discrepancies between Torahs is that there are some words whose spelling is a matter of dispute. In the Torah itself, there are two major questions. Genesis 9:29 has a word that may be spelled ויהי or ויהיו. Ashkenazi Torahs have the former and Yemenite Torahs have the latter. The difference is between singular and plural and is insignificant enough to be lost in translation from Hebrew to English. Small as it is, it is still a difference. Similarly, there is a question in Deuteronomy 23:2 whether a word should be spelled דכא or דכה. Here, there is no difference in meaning at all. Some would suggest based on midrashim that there are a handful of other single-letter differences in the Torah but others argue that this is merely a misunderstanding of midrashic techniques (we discuss this at length in our essay on The Text of the Torah).

In the end, out of over 300,000 letters in the Torah, there are at most a dozen or two instances where a letter is under question. This means that the Torah text we have is over 99.99% correct. That is important to remember when discussing this issue.

The Accepted Text of the Torah

Some may wonder whether the less than one hundredth of a percent that is under question presents an halachic problem. How can we make a blessing over the reading of the Torah in synagogue if we are not entirely certain that the Torah has been written correctly? The simple answer is that the Rambam wrote in a responsum (Pe'er Hador, 9) that, for the purposes of synagogue use, even an invalid Torah scroll may be used. While many disagree with this ruling, we rely on it in times of great need (see Rama, Orach Chaim 143:4 and Mishnah Berurah, 29).

However, we do not need to rely on this ruling of the Rambam because of two important halachic concepts. The first is that of majority. In Sofrim 6:4 we are told that this is a valid method of determining an authoritative text of the Torah (we discuss this passage at length elsewhere). By taking well-known, reliable texts we can resolve the few differences based on majority. This is certainly sufficient halachically (Chullin 11a-b) but is also an excellent tool for arriving at the original version of the Torah. All scribes err occasionally but excellent scribes do so only rarely. By taking the majority of readings, we can be fairly certain that the resulting version is based on error-free transmission. The second tool we have is that of tradition - masora. We can rely on good ancient texts because they were accepted as authoritative in their time. Similarly, we can rely on the Masoretic notes because they were written based on intensive study of manuscripts that were ancient even in the days of the Masoretes.

These two principles have been used before. In 1525, Daniel Bomberg's publishing house printed a rabbinic Bible - Mikraot Gedolot - that was arranged by Yaakov ben Chaim. In addition to arranging this edition, Yaakov ben Chaim gathered together the masoretic notes from many different manuscripts into one text that he called Masora Rabbata. Many like to exaggerate his role in the transmission of the Torah because, later in his life, he became an apostate by converting to Christianity, thus embarrassing traditionalists who rely on his work. However, his accomplishments were not original but technical. He helped publish things that had already been written and attempted to publish them as accurately as possible. Yet, his rabbinic bible is still riddled with errors that had to be corrected later. This was done by R' Menachem di Lonzano in his Or Torah and R' Shlomo Yedidiah Nortzi in his Minchat Shai. These two scholars used the tools of majority and tradition to clarify the accepted text of the Bible and their work remains the guide for scribes as codified by R' Shlomo Ganzfried (the author of Kitzur Shulchan Aruch) in his Kesset Sofer. The claim that Yaakov ben Chaim determined the basis of the accepted text is entirely wrong. He contributed to the confusion by printing a mistaken text and to the solution by printing masoretic notes. The true determinators of the accepted text were the authors of Or Torah and Minchat Shai (see Breuer, par. 23).

Recently, R' Mordechai Breuer applied this same methodology to the best and most ancient texts of the Bible available. He used the following versions: The Aleppo Codex, the Leningrad Codex, the British Museum Manuscript, the Cairo Codex, and the two Sasoon Manuscripts of the Bible. Based on the principles of majority and tradition, he arrived at a text of the entire Bible that is consistent with the Masora and is, surprisingly, almost identical to the Aleppo Codex. See his The Aleppo Codex and the Accepted Text of the Bible. His edition of the Bible is already becoming standard in many libraries and synagogues.

Going back to our original question, when we use the halachic principle of majority then there is no problem of making a blessing. Halachically, this Torah is considered acceptable. Similarly, there is no problem in fulfilling the mitzva of writing a Torah scroll. While the Chatam Sofer (Responsa, Orach Chaim, 52) suggested that we do not recite a blessing on the mitzva of writing a Torah scroll because of the doubts regarding chaser and yeter, this has been refuted by later halachists. See his student the Maharam Schick's work on the 613 Mitzvot (613:2-3), Responsa Ginat Vradim (Orach Chaim 2:6), Yabia Omer (vol. 8, Yoreh Deah, 36:3), and Ateret Paz (1:2, Yoreh Deah, he'arot 12:2). With the accepted text based on the majority of manuscripts and Masoretic notes, we can assert that we have confidence that even the less than 0.01% of letters that were in question have been resolved correctly.

However, we cannot be absolutely certain. Therefore, a Torah based on an ancient tradition that was in the minority cannot be summarily rejected. For example, a Torah that has Genesis 9:29 written as ויהיו, based on the minority Yemenite tradition, cannot be considered unacceptable. While it should not be written that way, a Yemenite scribe who followed his tradition and wrote it that way did so based on an ancient masora. We must therefore accept it as a possible version. See the sources quoted by R' Ovadia Yosef in his Yechave Daat 6:56. He cites rulings by R' Avraham ben Harambam, Meiri, Radbaz, and others as precedent. However, as the Meiri wrote in his commentary to Kiddushin 30a, only variations that have traditionally been in question may be considered acceptable ex post facto. For the over 99.99% of the spellings in the Torah in which we are expert, including the thousands of chaser and yeter that have never been questioned, variations are not acceptable.

Verses in the Torah

We have already cited above the Gemara in Kiddushin that there are 5,888 verses in the Torah. Some versions of the Gemara have 8,888. This version is clearly incorrect because it implies a Torah that is over 50% larger than the Torah we have. We can again turn to the ancient Samaritan Torah and Septuagint that do not imply a book that is 150% the size of our Torah. However, this version of 8,888 caused great anguish to many commentators, including the Minchat Shai, who were puzzled that our Torah is thousands of verses shorter than that mentioned in the Gemara. We can say with confidence that this was simply due to a copyist's error.

The Gemara also says that Psalms has an additional eight verses and Chronicles has eight less. With this, we find two puzzles in this Gemara. The first is that our Torahs have 5,845 verses rather than the 5,888 stated in the Gemara. The second is that Psalms and Chronicles do not have anywhere near that number of verses. Psalms has 2,527 verses and Chronicles has 1,764 verses. That is far from being within eight verses of 5,888.

R' Menachem Kasher (Torah Shelemah, vol. 28 addenda ch. 12) quotes an explanation of this Gemara from R' Yehuda Epstein, a student of R' Chaim of Volozhin. R' Epstein pointed out that there are 43 verses from the Torah that are quoted in Psalms and Chronicles - 8 in Psalms and 35 in Chronicles. If these Torah verses that are cited in Psalms and Chronicles are added to the 5,845 verses in the Torah we arrive at the number of 5,888 that the Gemara mentions. While the exact wording of the Gemara is still difficult, the meaning seems to have been elucidated. It is not that Psalms and Chronicles have a few more or less verses than the Torah. Rather, if we add certain Torah verses from these books to the count in the Torah then we arrive at the number cited by the Gemara.

The Middle of the Torah

That same Gemara in Kiddushin states the following: The vav of gachon (Leviticus 11:42) is the middle of the letters of the Torah, darosh darash (Leviticus 10:16) is the middle of the words of the Torah, and the ayin of miyaar (Psalms 80:14) is the middle letter of Psalms. Simply counting the letters and words of these two books shows that everything on the list is incorrect. Does this shed doubt on the authenticity of our books? Not only are they incorrect, but for the vav of gachon to be the middle of the Torah, the Torah would need another 9,667 letters. That is a large number of letters to be missing.

R. Menachem Kasher (ibid.) quoted R. Yitzchak Yosef Zilber (in Shmaatin issue 43) who offered the following explanation. Almost all of the letters of the Torah are written in the standard Hebrew script in the standard size. However, there are some letters that are written in an unusual fashion and some that are written large or small. If one were to count all of the small and large letters in a standard Torah, one would find that there are exactly 16 of these letters. Of these, the ninth, the middle one, is vav of gachon. In other words, the Gemara was not referring to vav of gachon as the middle of all the letters of the Torah. Rather, it was referring to it as the middle of all the unusually large and small letters in the Torah. However, there is another tradition of large and small letters, that of R' Yosef Tov Elem. But, even according to that tradition there are 32 such letters and the sixteenth is vav of gachon. While this explanation seems far-fetched, it is confirmed by noting that there are exactly seven unusually large and small letters in Psalms and the fourth - the middle letter - is ayin of miyaar.

Similarly, there are 77 instances of double words in the Torah (like Avraham Avraham and Lech Lecha). Of those 77 cases, the 39th instance - the middle one - is darosh darash. It is not the middle of all the words in the Torah but it is the middle of all the unusual double-words.

Large and Small Letters in the Torah

1. Genesis 1:1 5. Exodus 34:7 9. Leviticus 11:42 13. Deuteronomy 6:4
2. Genesis 2:4 6. Exodus 34:14 10. Leviticus 13:33 14. Deuteronomy 29:27
3. Genesis 23:2 7. Leviticus 1:1 11. Numbers 14:7 15. Deuteronomy 32:6
4. Genesis 27:46 8. Leviticus 6:2 12. Deuteronomy 6:4 16. Deuteronomy 32:18

The Script of the Torah

The Gemara in Sanhedrin 21b-22a tells us what at first seems very surprising. However, after a careful reading and placing the events in an historical context they do not seem surprising at all.

Mar Zutra and some say Mar Ukva said: Originally the Torah was given to Israel in Ktav Ivri (paleo-Hebrew characters) and in the holy lanugage. It was given again to them in Ezra's time in Ktav Ashurit (Assyrian characters) and in Aramaic. Israel selected for themselves Ktav Ashurit and the Hebrew language... It was taught: Rebbe said: Torah was originally given to Israel in Ktav Ashurit. When they sinned it was changed to Roetz (Ktav Ivri). When they repented, Ktav Ashurit was reintroduced... R' Shimon ben Elazar said in the name of R' Eliezer ben Parta, who said in the name of R' Elazar Hamodai: This writing was never changed...

We see three opinions regarding the script of the Torah. According to Mar Zutra, the Torah was given to Israel in Ktav Ivri and in Hebrew but Ezra changed it to Ktav Ashurit and Aramaic. The people, however, only accepted Ktav Ashurit and Hebrew. According to Rebbe, the Torah was given in Ktav Ashurit but was changed to Ktav Ivri due to the people's sins. According to R' Elazar Hamodai, the script of the Torah never changed.

This passage raises a number of questions. How could Ezra change the script of the Torah? How could he change the Torah's language from Hebrew to Aramaic? Furthermore, if he found the authority to do so, how could the people determine an outcome against his decision? According to Rebbe, why would the script of the Torah change based on whether Israel sinned or repented?

R' Reuven Margoliyot (Margoliyot Hayam, Sanhedrin ad loc,; Hamikra Vehamesora, ch. 9) answers all of these questions with the following historical reflection. It is known that some ancient cultures had one script for sacred purposes and one for everyday use. For example, the Indians only used Sanskrit for religious purposes and not for the mundane. The talmudic sages mentioned in the above passage were debating the extent of this practice of having a script for only holy purposes in Israel. However, according to everyone this was the practice, similar to the talmudic dictum, "Something that is used for the sacred may not be used for the profane" (Avodah Zara 52a).

According to Mar Zutra, the first tablets of the ten commandments were written in Ktav Ashurit (see Responsa Radbaz 3:442) but once Israel sinned with the Golden Calf they were deemed unworthy. They could not be trusted to use Ktav Ashurit for purely sacred matters. Therefore, the second tablets and the Torah scrolls written for general use were in Ktav Ivri. This can, perhaps, be seen from the fact that in Megillah 2b Rav Chisda says that the mem and samech in the tablets were miraculously hanging in the air. This can only happen in Ktav Ashurit and not in Ktav Ivri. However, in the Gemara in Sanhedrin quoted above, Rav Chisda seems to agree with Mar Zutra that the Torah was originally given to Israel in Ktav Ivri. Therefore, it seem that Rav Chisda would have to say that the tablets were in Ktav Ashurit and the Torah in Ktav Ivri. Or, as the Radbaz suggested, everything was originally in Ktav Ashurit but after the sin of the Golden Calf the second tablets and the Torah were in Ktav Ivri. But not all of the Torahs were in Ktav Ivri.

That the original tablets were given in Ktav Ashurit but not the second tablets can be seen hinted in a number of sources. For example, the Gemara in Pesachim 87b says "the tablets broke and the letters floated in the air". Exactly what it means that the letters floated in the air is unclear. However, on that same page the Gemara says, "Three things returned to their origin... the script of the tablets". That sounds like Ktav Ashurit being replaced with Ktav Ivri. Similarly, the Mechilta on Exodus 17:8 says that after the tablets were broken "the heavenly writing returned to its place". We perhaps also see evidence of the disappearance of Ktav Ashurit much later in history. The Tanchuma on Vayeshev 2 says, "What did they do [in response to the Samaritans]? Ezra, Zerubavel, and Yehoshua gathered the community to the sanctuary... and excommunicated the Samaritans with the sacred name of G-d, with the script that was written on the tablets, with the decree of the heavenly court,..." The use of the "script that was written on the tablets" is important for two reasons. First, it seems that this script was unique. Furthermore, we know from the Gemara in Sanhedrin and from other historical sources that the Samaritans used Ktav Ivri. The contrast between the Samaritans and the "script that was written on the tablets" implies that this script was not Ktav Ivri. We thus see that there is ample material supporting the Radbaz's claim that the first tablets were in Ktav Ashurit.

Recall that Mar Zutra said that the Torah was given to Israel in Ktav Ivri. The Ritva deduced from this that the special Torah of Moshe that was kept in the ark and later in the Temple was in Ktav Ashurit. Only Torahs for the people were in Ktav Ivri. The ability to read Ktav Ashurit was maintained by priests and scribes, which is why King Yoshiyahu needed a priest to read to him from Moshe's Torah when it was found in the Temple (2 Kings 22:8-11; Abarbanel). The king had never before seen Ktav Ashurit and his reaction to seeing it fo the first time, and in the Torah scroll that Moshe himself had written, demonstrates the deep religious emotion it evoked. We perhaps find hints of this in Isaiah 8:1 where the prophet is commanded, "Take a large tablet and write on it in common characters". This is must have been referring to Ktav Ivri that was used by the common people (see Rashi). Ktav Ivri had gained such prominence that the existence of ending letters (ךףץןם) was forgotten by the masses and had to be restored (Megillah 2b-3a).

However, Ktav Ashurit was still studied by the priests and scribes, of which Ezra was both. When he saw that Ktav Ashurit was so forgotten that, when it was written on the wall of King Belshatzar of Babylonia, only Daniel could read it (Daniel 5) he realized that it must be reintroduced to the people. Yet, he still had the dilemma that people would then be writing Hebrew in the holy Ktav Ashurit for improper purposes. His solution was to translate the Torah into Aramaic and introduce the Aramaic Torah in Ktav Ashurit into common usage. That way people would become familiar with Ktav Ashurit without using it in their daily Hebrew writing. This is what is meant in Nehemiah 8:8, "So they read from the book, from the law of God, with interpretation." It was interpreted by translation into Aramaic (Megillah 3a). (This translation was later recreated by Onkelos). However, the people had lived their whole lives with a Hebrew Torah and were not ready to change the language of their holiest of books. Therefore, they decided to retain a Hebrew Torah in Ktav Ashurit but conduct their daily business in Aramaic. This would produce the results that Ezra desired because Ktav Ashurit in Hebrew would not be a part of the daily routine.

Rebbe agreed with this historical reconstruction but attributed the original transition from Ktav Ashurit to Ktav Ivri to the idolatrous era of the First Temple rather than the episode of the Golden Calf. According to Rebbe, it is even more plausible that the scholars always retained knowledge of Ktav Ashurit. It was only the masses who were busy with their daily lives and/or idolatrous ways who forgot Ktav Ashurit when the Torahs were changed to Ktav Ivri.

R' Elazar Hamodai does not necessarily disagree that people forgot Ktav Ashurit. He only argued that the Torahs were never changed from one script into another. However, he agreed that people had forgotten Ktav Ashurit, the script used only for sacred purposes, and that Ezra had to re-educate the masses in the holy script (see Teshuvot HaRambam, ed. Blau no. 268).

As a final note, the Gemara in Sanhedrin 22a offers two opinions why the script is called Ktav Ashurit. One is that the Jews brought it back to Israel with them from Babylonia/Assyria (Ashur). The other is that it is a beautiful script (me'usheret). Since the literal translation of Ktav Ashurit is "Assyrian script", we must ask why the Gemara even asks such a basic question. It is called Ktav Ashurit because the Assyrians used it. Furthermore, the view that it is called Ktav Ashurit because the script is beautiful strains credibility. We already know that it is called Ktav Ashurit because it is an Assyrian script, as the words simply mean.

We have seen that many questions can be raised about the validity of our Torahs. However, Judaism, like any other serious thought system, is complex. While by necessity we were taught simplicites in our childhood, we need to sieze all available opportunities to broaden our perspectives and deepen our faiths. Rather than using questions as reasons to reject traditional Judaism, we must use them as opportunities for intellectual and religious growth.

Back Home
Last revised: 2/6/02
© Aishdas 2002