Identifying the Most Quoted Sages in the Talmud's Aggada: A Programmatic and Quantitative Study
Also: word frequency
This piece examines the frequency of citation of particular Amoraic sages within aggadic sugyot in the Talmud through a quantitative analysis of Aggadic passages.1 Using Ein Yaakov, the classic anthology of the aggadic material in the Talmud Bavli, I programmatically extracted statements attributed to sages. In particular, I extracted all statements beginning with "[and] said Rabbi/Rav”. I was able to extract 4,015 statements this way. (The link to my spreadsheet, with the full info, is at the end of this piece.)
My aim is to quantify the contributions of named tradents in the Talmud.
“A person who is responsible for preserving and handing down traditions, especially oral traditions. The term is generally used in the context of the Bible and rabbinic Jewish traditions.”
I extracted from Ein Yaakov (the collection of aggadic passages in the Talmud, which is around a third of the Talmud) all statements starting with “[and] said Rabbi/Rav” ([ve-]amar Rav/Rabbi - אמר רב / רבי).2
I hyperlinked to the Wikipedia entry, and marked the amoraic generation. B = Babylonia, E = Eretz Yisrael.
Here’s the list of the ten most-cited sages, based on my sample, in order of number of citations. I rounded the number of statements to the nearest ten, and the percentage to the nearest whole number.
Rabbi Yohanan (E2). 7% (280 statements).
Rav (B1). 6% (240 statements). In 3/5 of these statements, he’s cited by Rav Yehuda (B2).
Rabbi Elazar (ben Pedat) (E2/3). 5% (270 statements).
Rava (B4). 4% (180 statements).
Abaye (B4). 4% (160 statements).
Rabbi Yitzchak (E3). 3% (120 statements). Surprisingly, for someone cited so often in the Talmud, almost no biographical information is provided in the Talmud. The Talmud isn’t even sure what his father’s name was. We also don’t know who his teacher was. See the discussion in the Hebrew Wikipedia entry: רבי יצחק – ויקיפדיה. Essentially all we know is who he cites, and who cites him. See Toldot Tannaim Ve’Amoraim in his entry.
Shmuel (B1). 2% (110 statements).
Rabbi Yehoshua ben Levi (E1). 2% (100 statements).
Rav Yehuda (not citing Rav). 2% (90 statements).
Based on this, these ten sages are responsible for 34% of all aggadic statements in the Talmud.
Appendix # 1 - word frequency - ten most common words in the aggadic portion of Talmud
אמר (said) - 10,888 occurrences
רבי (Rabbi) - 6,086 occurrences
ליה (to him) - 4,456 occurrences
לא (not) - 4,337 occurrences
רב (Rav) - 3,556 occurrences
על (on) - 3,068 occurrences
את (the) - 2,826 occurrences
לו (to him) - 2,736 occurrences
אלא (but) - 2,439 occurrences
שנאמר (as it is said) - 2,435 occurrences
Compare with the top ten in Tanakh, according to here (there, too, amar is the most frequent word; interestingly, that’s the only word that appears in the top ten of both lists):
Appendix # 2 - Full data
Link to the full list of statements, in this Google Sheet : Programmatic and Quantitative Analysis of talmudic aggadic statements (memrot).
Also as an Excel file here (may not be up-to-date):
This database of statements contains ~82,000 words (including Biblical references, in parentheses). The entire text of Ein Yaakov contains ~411,000 words. So it’s ~20% of the aggadic portion of the Talmud.4
Appendix #3 - the next ~100 most frequently cited sages in aggadic Talmud, in order of frequency, based on the sample. With number of citations in the sample
Interestingly, there are few Tanaim in this list, essentially only: Rabbi Yosei; Rabbi Shimon ben Yochai; Rabbi Meir; Rabbi Akiva; Rabbi Yehoshua; Rabbi Eliezer; Rabban Shimon ben Gamliel; Rabbi Yehuda. At least part of this is simply an artefact of the pattern analyzed to extract statements.
רב פפא 77
רב יוסף 67
שמואל 66
רבי חנינא 53
רבי אבהו 52
רב הונא 49
רב נחמן בר יצחק 47
רב אשי 45
רבי שמואל בר נחמני, אמר רבי יונתן 41
רבי חיא בר אבא, אמר רבי יוחנן 40
רבי יוסי 38
רב יהודה, אמר שמואל 38
רבי יהודה 28
רב נחמן 27
עולא 26
רבי זירא 25
רבה בר בר חנה, אמר רבי יוחנן 24
רבי שמעון בן לקיש 22
רב כהנא 21
רבי יוחנן משום רבי שמעון בן יוחאי 20
רבי לוי 20
רבה בר בר חנה 20
רבי יוסי בר חנינא 18
רב אחא בר יעקב 18
רבי אמי 18
רבה 17
רבי עקיבא 17
רבינא 16
רבי חמא בר חנינא 14
רבן שמעון בן גמליאל 13
רבה בר רב הונא 13
רבי אושעיא 12
רבי מאיר 12
רב ששת 12
רבי חיא בר אבא 11
רב אסי 11
רבי חמא ברבי חנינא 10
רבי ינאי 10
רבי אבא 10
רב גידל, אמר רב 10
רבי יונתן 9
רבי אסי 9
רבי חנינא בר פפא 9
רב קטינא 9
רבי שמעון 9
רבי ירמיה בן אלעזר 9
רבי חלבו 8
רבי יוסי ברבי חנינא 8
רב המנונא 8
רבא בר מחסיא, אמר רב חמא בר גוריא, אמר רב 8
רבי שמואל בר נחמני 8
רב חנן בר רבא, אמר רב 8
רבי יוחנן משום רבי מאיר 8
רבי אבא בר כהנא 7
רב משרשיא 7
רב יהודה אמר רב 7
אמימר 7
רבי חלבו, אמר רב הונא 6
רבי יוחנן משום רבי יוסי 6
רב חסדא, אמר מר עוקבא 6
רב יצחק 6
רבי ירמיה 6
רבי חנין 6
רב אמי 6
רב ששת משום רבי אלעזר בן עזריה 6
רבי אלעזר ברבי יוסי 6
רב חנא בר ביזנא, אמר רבי שמעון חסידא 5
רב אדא בר אהבה 5
רבי יהושע 5
רב זוטרא בר טוביה, אמר רב 5
רבי ברכיה 5
רבי נתן 5
רבי אבין 5
רבי חיא 5
רב הונא בריה דרב יהושע 5
רב נחמן, אמר רבה בר אבוה 5
רב יהודה בריה דרב שמואל בר שילת משמיה דרב 5
רבן יוחנן בן זכאי 5
רבי אחא בר חנינא 5
רב מתנה 5
רבי תנחום 5
רבי תנחום בר חנילאי 5
רבי נחמן 5
רבי זירא, אמר רב חסדא 5
רבי ישמעאל בן אלישע 4
רבי אבהו, אמר רבי יוחנן 4
רב חיא בר אשי, אמר רב 4
For the the names that appear the most in whole Talmud, including Mishnah, based on the list of ~90k appearances collected by Michael Sperling and Michael Satlow, “AllNameReferences” (Excel file, 2020), these are the top ten (count of “Rabbi Name after Link” - I marked the name, and the count, using a pivot table; the same calculation is make by them in a different publication: “Reference Counts By Afterlinkid” [2020]):
רבא 5231
רב 4675
רבי יוחנן 4341
רבי יהודה (ברבי אילעאי) 4016
שמואל 3111
אביי 3049
רבי שמעון (בן יוחאי) 2520
רבי מאיר 2477
רבי יהודה הנשיא 2315
רבי אליעזר (בן הורקנוס) 2006
From these, the names that overlap with my list are Rabbi Yohanan, Rava, Rav, and Abaye. Five of the top ten based on Satlow’s count are tanaim, while the top ten in my list doesn’t have any tanaim.
For other recent studies that take a quantitative approach to Talmudic names, see these studies (cited also in my “An Annotated Bibliography of Works Relating to Biographies of Talmudic Sages“, p. 4 - “Other literature”): Michael Satlow, “Naming Rabbis: A Digital List” (2017); ibid., “The Rabbinic Citation Network”, AJS Review (2020); and their many additional tables and graphs based on that same research project [“Tractate Rabbi Reference Count“; “Rabbi Intervals“; “rabbi quote with ids“; “RabbiQuotes“]; Maayan Zhitomirsky-Geffet and Gila Prebor, “SageBook: toward a cross-generational social network for the Jewish sages’ prosopography”, Digital Scholarship in the Humanities, 34(3) (2019), pp. 676–695; Josh Waxman, “A graph database of scholastic relationships in the Babylonian Talmud”, Digital Scholarship in the Humanities, Volume 36, Issue Supplement_2, (October 2021), Pages ii277–ii289; and my many recent pieces, which can be found at my Academia.edu page, especially my “Comprehensive Table of Amoraim: Index of Amoraim in Albeck's "Mevo LeTalmudim"; In Comparison With Toldot Tannaim VeAmoraim and Hebrew Wikipedia“ and “Sages and Scholars: A Digitized Index and Automated Analysis of Figures in the Mishnah, Talmud, and Late Antique Midrash“ (based on Toldot Tannaim VeAmoraim).
Regex pattern I started with:
'(.אמר ר[^.]*\.)'This regex extracts statement starting with ‘[and] said Rabbi/Rav’, until the first period punctuation (.). It relies on the punctuation provided in the Sefaria edition of Ein Yaakov (which is quite reliable; I’d estimate it has over 95% accuracy, for our purposes).
That simplified regex does not work for names with no prefixed honorific, such as Abaye, Shmuel, Ulla etc. (For a list of such names, see my overview of names in the Talmud, section “No Honorific".) For that reason, I had to add a number of additional patterns to match. (There are additional patterns that would extract additional relevant statements, but I don't believe it would add a significant amount, and there would also be false positives.) Here’s the final regex:
'(.אמר ר[^.]*\.|.אמר אבי[^.]*\.|.אמר עולא[^.]*\.|.אמר שמואל[^.]*\.)'In general, the Talmud is especially amenable to pattern matching, since it's highly formulaic. See on this Elman, “Orality”. And see also my piece on formulaic terms as a form of punctuation.
See my pieces on my Academia page for word counts of entire Talmud Bavli and indvidual tractates, and word counts of individual chapters of Talmud Bavli.