Fake Talmud Quotes: Forgeries, Censorship, and Propaganda
How do false quotations of the Talmud spread? An examination of forgeries, mistranslations, censorship and the use of texts in propaganda.
AI-generated editorial illustration, not a documentary photograph or historical evidence.
How can we distinguish authentic rabbinic texts from mistranslations, disputed variants, and invented quotations?
Online discussions are full of quotations introduced by the words “The Talmud says.” Some refer to actual rules in classical rabbinic literature, including troubling ones. Others have been removed from the surrounding discussion, mistranslated, or attributed to books in which they do not appear at all. Still others arose from conflating different sources or recycling older polemical lists. These cases are often presented with the same degree of certainty, despite having very different evidentiary value.
The task is further complicated by the historical reality that Jewish texts have been banned, burned, censored, and altered in print over the centuries. We must not ignore that historical fact. Yet censorship cannot serve as a catch-all explanation for every quotation whose original source cannot be found.
In the previous article, “Goyim in the Talmud: What Does It Actually Say?”, we examined authentic passages about non-Jews, including provisions that establish legal distinctions. This time, we will not repeat the entire discussion of those rules. Our question is different: how can we establish what was actually written, to whom the words were attributed, and how they reached readers today?
Why Are There So Many Disputed Talmudic Quotes?
The Talmud is not a single, uninterrupted body of teaching. The Mishnah preserves early rabbinic traditions, while the Gemara contains later discussions, objections, explanations, and narratives. There is a Babylonian Talmud and a Jerusalem Talmud. Alongside them are numerous midrashic works, minor tractates, and later legal codes. The Mekhilta, Tractate Soferim, and the Shulchan Arukh are therefore not simply three different names for the Talmud.
This distinction can be lost in the very first retelling. A statement from a midrash becomes “a Talmud quote”; the opinion of an individual teacher is transformed into a supposedly binding rule for all Jews; a passage addressing a particular legal question is presented as a universal statement about the worth of human beings. Sometimes only a single word changes; in other cases, the entire attribution changes.
Historical European polemical literature collected and translated such passages for various religious and political purposes. During the twentieth century, they also appeared in racist propaganda. Digital media have given many old errors renewed circulation: an image containing an alleged quotation may circulate on hundreds of websites without any of them providing the Hebrew original. The National Library of Israel documented just such a chain of transmission in the case of one of the most widely circulated false references, known as Libbre David 37. [1]
Yet the history of propaganda does not mean that every unpleasant rabbinic statement is a forgery. That would be an error in the opposite direction. Careful reading can confirm that a troubling ancient provision is authentic while also rejecting a later fabrication that purports to cite it.
How Do We Verify a Quote?
A credible verification begins with an identifiable bibliographic reference: the title of the work, the tractate, chapter, page, or paragraph and, where relevant, the particular edition. “Talmud, page 37,” without a tractate name, is not enough. Each tractate has its own pagination, and different collections may use different numbering systems.
The next step is to read the original language and the immediate context. A Hebrew or Aramaic sentence may be the opinion of one teacher, an objection, a quotation from an earlier tradition, or the resolution of a debate. Even when the individual words are translated accurately, omitting an objection can create a very different impression. Another crucial distinction concerns terms such as goy, nokhri, oved avodah zarah, and other legal categories. Their scope is not always the same.
The third step is to compare textual witnesses. A manuscript, an early printed edition, and a modern electronic text may differ. Such differences may result from copying errors, editorial choices, censorship, or different lines of transmission. The existence of a variant does not, by itself, establish its cause.
Finally, we must distinguish at least four questions: Does the passage exist? Is it cited at the correct location? Is the translation accurate? And does it actually support the claim attached to it online? Only then does it make sense to consider its subsequent legal significance or its moral implications. Digital libraries such as Sefaria are useful tools, but a single online presentation is no substitute for critical comparison of manuscripts. [2]
The Real History of Talmud Censorship
The censorship of Jewish books in Christian Europe is a documented historical reality. On September 9, 1553, copies of the Talmud were publicly burned in Rome. Jewish printing later faced various restrictions elsewhere in Italy as well. The Basel edition of 1578–1581 is an example of printing under conditions of censorship, with portions of the text omitted or altered. [3]
One tangible source is the Basel edition of the Babylonian Talmud from 1578–1581, described in the digital catalogues of the University Library of Basel. [4] The historian of Hebrew printing Marvin J. Heller specifically discusses alterations to Bava Kamma 38a. According to his published comparison, material that had appeared in Daniel Bomberg's earlier Venetian edition was substantially abridged in the censored Basel edition. Some later editions marked omissions with visible blank spaces; others rearranged the text so that the omission was not obvious. This account draws on Heller's published comparison; we have not independently collated facsimiles of the two editions for this article. [5]
There is also manuscript evidence of organized oversight. The British Library holds Sefer ha-Zikkuk, manuscript Or 6639, a collection of instructions and lists for censoring Hebrew books associated with Domenico Irosolimitano. The catalogue describes 136 folios, a list of works under review, and passages to which censors were to pay attention. Digital images are not available in the current online catalogue, but the catalogue description itself can be checked publicly. [6]
This material establishes the existence of specific institutions, procedures, and interventions. There is no need to invoke some unspecified “secret censorship,” since the historical evidence is publicly documented. It does not, however, establish that every anonymous internet quote once existed and was subsequently removed. To support such a claim, we would need to produce a manuscript or an earlier printed edition containing that quotation.
What Happened to the Terms *Goy* and *Akum*?
In the Hebrew Bible, goy can mean a nation; in many rabbinic texts, it refers to a non-Jew. The term akum (עכו״ם) is an abbreviation generally understood to refer to worshippers of heavenly bodies, or idolaters. The two expressions are not automatically interchangeable: “non-Jew” denotes a broader category of belonging, while “idolater” describes a religious category.
In certain printed editions, broader terms were replaced—through censorship or self-censorship—with expressions that appear to refer to a narrower religious group. Such a change can affect how readers understand the people addressed by a rule. Someone familiar only with an altered edition may conclude that the rule was never directed at non-Jews generally, but exclusively at idolaters. In another case, however, the term for an idolater may genuinely belong to the older text rather than being a later substitution.
It would therefore be wrong to impose a single explanation on every passage. A comparison must specify the manuscript or printed edition being used, the word it contains, and the century in which that reading is attested. Where that comparison has not yet been made, the conclusion must remain limited. Documented cases from the history of printing show that censorial alterations occurred, but we must not guess at the reasons behind every variant. [5]
The reverse caution matters just as much: a textual alteration made under pressure from authorities does not prove that all later translators or rabbis consciously participated in a cover-up. Sometimes they simply inherited a printed text that had already been changed.
An Authentic but Misrepresented Saying: “Kill Even the Best of the Non-Jews”
One of the most difficult cases concerns a saying preserved in different forms within midrashic tradition. The saying is attested in several midrashic and related traditions, including Tractate Soferim 15:8 under Sefaria's numbering; some editions number the passage 15:10. The English translation published there attributes it to Rabbi Shimon bar Yochai and includes the qualification “in wartime.” A commentary note accompanying that translation also points to scholarly arguments that this qualification may have been added later. This is not, by itself, a final determination of the original wording. [7]
It would therefore be incorrect to claim that the saying was fabricated from nothing. The saying exists, and it is explicitly violent. At the same time, it should not simply be attributed to a page of the Babylonian Talmud when the source actually being cited is another collection. A misleading bibliographic label can become the starting point for a wider misunderstanding.
Was the wartime qualification part of the earliest form of the saying? That cannot be established from one edition and an accompanying commentary note. Different formulations and interpretations appear in its history. A 2021 scholarly study by Eric Lawee examines the transmission and medieval interpretation of the saying and shows how it became a subject of controversy. [8]
Nor do the efforts of some later writers to restrict the violent statement to wartime establish that this interpretation was undisputed everywhere and at all times. The more careful conclusion is also the more demanding one: we are dealing with an authentic and troubling tradition, different textual forms, multiple interpretations, and a history of polemical use. We must neither deny the saying's existence nor declare it a general command supposedly binding on all Jews today.
“Whoever Saves One Life Saves an Entire World”: Why Do the Manuscripts Differ?
Even an apparently universal maxim can be disputed at the level of its wording. Mishnah Sanhedrin 4:5 links the value of a single life to that of an entire world. Yet the surviving readings are not identical. Some contain the additional words “of Israel” (mi-Yisrael), while others refer to one human life without that limitation. Among the sources that can be compared are the transmission of the Mishnah, the passage as cited in the Babylonian Talmud, and a version in the Jerusalem Talmud. [9]
This does not prove that a translation omitting “of Israel” is false. Nor does it establish that the version with those words must be the oldest. Assessing that question requires dating the individual textual witnesses and determining their independence from one another. The differences among editions and commentaries warn against drawing definitive conclusions from any single familiar version of the sentence. The Mishnat Eretz Yisrael commentary published alongside the Mishnah on Sefaria specifically notes manuscript witnesses without the words “of Israel,” as well as a difference in the printed text. Determining which reading is original, however, requires a broader text-critical comparison. [9]
The surrounding discussion also matters. The Mishnah addresses the warning given to witnesses in cases that may carry a death sentence and the consequences of taking a life. It goes on to emphasize humanity's common origin in Adam and cautions against exalting one's own lineage. This discussion is broader than the single sentence we encounter on a poster or in a film.
The finding is therefore twofold: a qualified version exists, and a broader version is also attested. Without appropriate critical analysis, we will not declare either one definitively original. In historical research, clearly acknowledged uncertainty is worth more than a confident but unsupported conclusion.
“Libbre David 37”: What Do We Do with a Source That Cannot Be Found?
Online collections sometimes circulate an alleged instruction that a Jew must deliberately lie to a non-Jew about the contents of rabbinic books in order to conceal their teachings. The quote is often accompanied by the bibliographic reference “Libbre David 37.”
The difficulty begins with the title. “Libbre David” has not been identified as a tractate of the Talmud or as a verifiable source containing the alleged sentence. The National Library of Israel has described in detail how this reference appeared in anti-Jewish and Nazi publications and has examined possible connections with similarly named works, such as Divrei David. It found no corresponding passage in the works it examined. [1]
That is strong evidence of an incorrect or unverified attribution; it does not justify the absolute claim that no book with a partly similar title exists anywhere. Likewise, the fact that a real work called Divrei David was published in the seventeenth century does not prove that the online fabrication under discussion originated at that time.
The decisive question is whether the claim can be checked independently. Someone who says a quotation comes from a book must be able to identify an edition and point to the passage in it. If that cannot be done, the quote must not be treated as an authentic source merely because it has been repeated for generations. Indeed, a repeatedly reproduced but imprecise reference may itself be a clue that a statement was copied from an older polemical list.
Quotes Attributed to the Wrong Place: Orach Chayim 539 and Sanhedrin 59a
When the issue is a false attribution, verification does not always require elaborate manuscript scholarship. It may be enough to open the cited passage.
First case: Shulchan Arukh, Orach Chayim 539. Online posts attribute to this chapter a general command to defraud non-Jews. In fact, the chapter concerns trade, loans, and commercial activity during the intermediate days of a festival (chol ha-moed). Some provisions do mention dealings with non-Jews, but the text examined does not contain the alleged general instruction to defraud them. [10]
The type of book matters, too. The Shulchan Arukh is a later legal code, not a Talmudic tractate. Calling it “the Talmud” makes the bibliographic attribution inaccurate from the outset. Rejecting this particular quotation does not mean that no controversial rules about accounting errors exist elsewhere in Jewish literature. We discussed those rules separately in GOY-02 and will return to them in GOY-04.
Second case: Babylonian Talmud, Sanhedrin 59a. This page has been cited for a sentence supposedly comparing the murder of a non-Jew with killing an animal. That formulation does not appear on the cited page. What it does contain is an authentic and strict discussion of non-Jews studying the Torah, together with an objection citing a different position attributed to Rabbi Meir. [11]
By rejecting an incorrect attribution, we have not shown that all related historical questions are simple or unimportant. We have shown something narrower and verifiable: the cited location does not say what someone claims it says. That is a meaningful finding in its own right.
From Historical Polemical Books to Modern Lists
Collections of alleged rabbinic quotations appeared long before social media. Johann Andreas Eisenmenger assembled numerous excerpts from Jewish literature in Entdecktes Judenthum (1700). His work influenced subsequent polemics. In Der Talmudjude (1871), August Rohling continued the tradition of harsh anti-Jewish interpretations; Justinas Bonaventure Pranaitis, with Christianus in Talmude Iudaeorum (1892), contributed another influential polemical collection. [12]
These authors wrote in different circumstances, and their works did not all emerge from one simple chain of direct copying. Independent catalogues and scholarly works provide bibliographic information about their early editions. [12] It would therefore be inappropriate to claim, without comparing the relevant editions, that every later quotation originated with a particular predecessor. It is, however, historically documented that many attributions and interpretations were repeated, sometimes even with identical errors.
During the twentieth century, Nazi propagandists also made use of such material. The National Library of Israel describes how antisemitic publications presented alleged passages as evidence of supposed characteristics shared by Jews as a group. That is an unsupported inference: even a wholly authentic historical text cannot, on its own, establish what every member of a community thinks or does. [1]
Analysis of polemical literature must avoid the opposite oversimplification. The fact that an author wrote anti-Jewish propaganda is not, by itself, proof that every quotation the author used was fabricated. The appropriate method is to verify each quotation, translation, and accompanying claim individually. Criticism of a discriminatory historical rule may be justified; attributing a nonexistent rule to a source is not.
How Can Readers Check a Disputed Quote for Themselves?
Before sharing a quotation or treating it as evidence, we should ask a few practical questions. Does it name a specific work? Can we open the correct page or chapter? Is the source the Mishnah, the Gemara, a midrash, or a later legal code? Is the original-language text available? Are the words really spoken by the person to whom they are attributed? And have we read at least a few paragraphs before and after the passage?
Where several versions exist, we should compare them rather than choose whichever best supports our initial assumptions. If we invoke censorship, we should identify the specific earlier and later sources. If a quotation circulates only as an image and its bibliographic reference cannot be verified, we should label it unidentified, not a documented rule. If it is authentic but harsh or discriminatory, we should say so plainly—without concealing it and without assigning responsibility to millions of people living today.
Four separate judgments are especially useful:
- Textual authenticity: Does the passage exist in a verifiable source?
- Accuracy of attribution: Is it really in the cited book at the cited location?
- Accuracy of translation and context: Is its meaning represented fairly, including objections and qualifications?
- Scope of the conclusion: What does the text establish about its own time—and what does it not establish about the present?
The examples we have considered yield different answers. The saying about “the best of the non-Jews” belongs to an authentic but complex midrashic tradition. The maxim about saving a life has several attested textual forms. “Libbre David 37” is an unidentified attribution with a documented history in propaganda. The alleged instructions attributed to Orach Chayim 539 and Sanhedrin 59a, meanwhile, do not appear at the cited locations.
The conclusion is not that the Talmud must be defended unconditionally or condemned unconditionally. It is that authenticity, context, and historical responsibility are separate questions. One can firmly criticize a genuinely discriminatory provision while rejecting a fabricated quotation. That consistency is what distinguishes research from propaganda.
Sources and further reading
- Daniel Lipson, “Into the Depths of Evil: How the Nazis 'Recruited' the Talmud for Anti-Semitic Propaganda,” National Library of Israel (2018): <source>. Source
- Sefaria, digital library of Jewish texts: <source>. Source
- Jewish Encyclopedia, “Talmud,” section on censorship, the burnings of 1553, and the Basel edition of 1578–1581: <source>. This is a historical encyclopedic synthesis, not an independent examination of Inquisition records. Source
- Universitätsbibliothek Basel, [Talmud Bavli], Basel, 1578–1581, e-rara catalogue: <source>. Source
- Marvin J. Heller, “Who can discern his errors? Misdates, Errors, Deceptions, and other Variations in and about Hebrew Books, Intentional and Otherwise: Revisited,” The Seforim Blog (2016), especially the section on Bava Kamma 38a: <source>. Source
- British Library, Sefer ha-Zikkuk, Or 6639, archival catalogue: <source>. Source
- Tractate Soferim 15:8, edition and translation on Sefaria (differently numbered in some editions): <source>. Source
- Eric Lawee, “The Best of Snakes…: A Polemical Midrash in the Rashi Supercommentary Tradition,” in Polemical and Exegetical Polarities in Medieval Jewish Cultures (2021), pp. 317–346: <source>. Source
- Mishnah Sanhedrin 4:5, text: <source>; Mishnat Eretz Yisrael commentary on the same passage (discussing manuscript witnesses and printed variants): <source>. Determining the original reading conclusively would require a separate text-critical study. Source 1 Source 2
- Shulchan Arukh, Orach Chayim 539: <source>. Source
- Babylonian Talmud, Sanhedrin 59a: <source>. Source
- Primary and bibliographic information: Johann Andreas Eisenmenger, Entdecktes Judenthum (1700), biographical entry in Neue Deutsche Biographie: <source>; August Rohling, Der Talmudjude (Münster, 1871), Bavarian State Library catalogue: <source>; Justinas Bonaventure Pranaitis, Christianus in Talmude Iudaeorum (Petropoli, 1892), National Library of Finland catalogue: <source>. For connections among polemical collections, see also [1]. Source 1 Source 2 Source 3