IN THE SUMMER OF 1950, first Julius and then Ethel Rosenberg were arrested on charges of conspiring to commit espionage on behalf of the Soviet Union. Morton Sobell, a former classmate of Julius’, was also arrested and charged with being part of the Rosenberg spy network. Played out against the hysteria generated by the onset of the Korean War, and the Smith Act, and the prosecution of the leadership of the Communist Party of the United States (CPUSA), the Rosenberg trial in March 1951 took a brief two weeks to complete and ended with the jury delivering a guilty verdict. ¹ On 5 April 1951, the presiding judge, Irving Kaufman, sentenced Morton Sobell to thirty years, and Ethel and Julius to death. Their executions were delayed until 19 June 1953 as various appeals were pursued.

      These barebones facts do not adequately convey the controversy surrounding the trial, sentencing, and execution of the Rosenbergs. From the time of their trial to the present, the Rosenbergs have been viewed by some as victims of the Cold War and by others as traitors to their country. The prevailing political climate of the US determines which of these interpretations is in the ascendant. During the repressive 1950s, popular and official views of the case coalesced: it was commonly believed that the Rosenbergs were Communist spies who deserved to die. In the more liberal 1960s and 1970s the Rosenbergs were seen as victims of Cold War hysteria, their trial and execution a miscarriage of justice. By the 1980s, in response to a right-wing shift in American politics, the Rosenberg case was once again subject to revisionist impulses. In the new conservative moment, it was argued that Julius Rosenberg was most assuredly guilty of some kind of espionage, even if Ethel was not. Buttressing this argument were the recently released Venona decrypts, messages between KGB operatives in America and Moscow that, assessed from within this conservative paradigm, confirmed their guilt. For many historians, the Rosenberg case is now closed. This paper argues that the Venona intercepts require far greater scrutiny than they have so far been afforded, that the Rosenbergs’ guilt has not been established, and therefore, that the case is not closed.

On 11 July 1995, the National Security Agency (NSA) announced that it had nearly 3,000 coded and encrypted documents from KGB agents relating to Soviet espionage in the US during the 1940s. These had been decoded, decrypted, translated, and rendered as English plain text ² over the years by several security services of the United States government as part of an enterprise that was given the codename Venona. ³ The NSA indicated that the Venona documents were now being declassified and would be released in batches in ensuing months. The time lag between the public announcement that these documents existed and their declassification and release was necessitated, according to the Agency, by concerns regarding privacy. ⁴ In the interim, to provide a sense of what the project had achieved, the NSA released 49 documents, including all the material related to the Rosenbergs, a cache of 19 decrypted and decoded messages.

      In the next 16 months (between July 1995 and October 1996), the NSA released approximately 2,850 similar documents. In October 1996, to publicize the existence of these documents as well as to mark the official closing of the Venona project, the NSA together with the Central Intelligence Agency (CIA) and the Center for Democracy (associated with the counter-revisionist historian Allen Weinstein) held a conference and media event at the National War College in Washington DC. An assortment of historians, government employees, members of the fourth estate, and other interested parties attended, including Morton Sobell, who had been tried and convicted with the Rosenbergs. ⁵

      Simultaneous with the conference, the NSA and the CIA jointly issued the Robert Louis Benson and Michael Warner edited volume entitled Venona: Soviet Espionage and The American Response 1939-1957, a work intended as a handbook for scholars interested in the Venona project. ⁶ Accompanying Venona was a series of five very short pamphlets summarizing the history of the Venona project (in the first pamphlet of eleven pages) ⁷ and then (in the next four) ⁸ outlining the nature of the documents available through the Venona project. A sixth pamphlet, similar in length and format to the other five, was released somewhat later. ⁹

      Finally, in keeping with their putative desire for transparency with regard to this work, the NSA established a Venona web site through which the entire collection of documents can be viewed. ¹⁰ I use the term "putative" advisedly because, despite (or perhaps because of) the generosity of such an abundant release, there is as yet neither an index for the collection of data in Venona nor one for the entire corpus of approximately 3,000 documents from the Venona project. The absence of a master index allows availability without ease of access. An alphabetical list of all the codenames with the NSA’s correlated given names and the pages on which those names occur would provide evidence of frequency of mention in the Venona documents, and frequency might provide one indicator of relative activity and/or importance of persons identified as being engaged in espionage. ¹¹

      Venona is divided into three parts of unequal length. The shortest, although by no means the least important, is the prefatory gloss to the entire volume, composed of a one page "Foreword" written by William P. Crowell, Deputy Director of the NSA, a 33 page "Preface," a two page list of abbreviations and acronyms, and an eight page chronology. This front material serves two functions. First it attempts to establish the intellectual authority and scholarly authenticity of the work through the formal apparatus of academic writing. Second, it frames the documentary material in the remaining two sections within a specific ideological context. That context is most easily discernible from the chronology, a list of dates that relate to the Venona material. It begins with the first debriefing of Walter Krivitsky (identified as a "Soviet intelligence defector") on 10 January 1939 and ends in 1957, with the following three items:

        17 June

        21 June

        15 November

In terms of the ideological perspective that informs Venona, this conjunction of events is revealing. The Smith Act had nothing to do with acts of espionage. It criminalized forms of speech, that is, it made it illegal to teach and advocate (and to conspire to teach and advocate) the violent overthrow of the American government. ¹³ The Supreme Court, in Yates v. US, advanced a strict interpretation of the First Amendment, a position that Benson and Warner describe as a serious  impediment in the fight against domestic subversion in that it requires evidence of overt acts as distinct from speech about acts (i.e., advocacy). ¹⁴ To include the Yates v. US decision, along with the accompanying explanation of the difficulties that decision apparently created in dealing with American Communists, in the final portion of a list of alleged espionage and subversion creates the impression, implicit throughout Venona, that the CPUSA was an organization devoted to espionage. It suggests, further, that the singular failure of various government security agencies in the period covered by Venona to arrest and convict many American Communists for espionage activities was not because of the insufficiency or absence of evidence against them, but because the Supreme Court was soft on Communists. ¹⁵

      Besides embodying the ideological predisposition of the volume and the project, this finger pointing is also a strategy for avoiding accountability. After 50 years of decoding, decryption, translation, and investigation, the tangible results of the Venona project are remarkably thin. One way to understand the NSA’s insistence that the work of the Venona project did not benefit from computer technology, but was achieved by a labour intensive, time-consuming iterative process of layered decoding that took many years, may be to mitigate this embarrassing fact. Reinforcing this view of the Venona project as requiring herculean human effort, all the Venona documents, both in the volume and the web site, that is, all 3,000 messages, are reproduced from typescript most of which appears to have been composed on manual typewriters.

      The second section of Venona, entitled "The American Response to Soviet Espionage," is a miscellaneous group of 35 US government documents from 1939 to 1960, chronologically arranged, which, according to Benson and Warner, "represent an attempt to gather some of the more interesting, important and revealing original documents available to American policymakers and intelligence officers during the period covered by this volume." ¹⁶ Whether these 35 documents have the standing the editors attribute to them is not clear, since the editors did not indicate the contents of the larger pool of documents from which they gathered this material. The juxtaposition of these documents with the Venona intercepts nevertheless creates an interpretative field that enhances the credibility of both sets of material. By virtue of their proximity to the Venona material, the US government documents urging greater attention to espionage take on an inferential prescience, credibility, and validity. That the US government had serious concerns about domestic spying gives added importance to the Venona intercepts.

      The third and longest part of the volume is composed of selected Venona decrypts (99 in all), which are purportedly among "the most significant and revealing Soviet messages translated by Western analysts." ¹⁷ The 99 documents are prefaced by a note on translation, which is a list of 10 words and phrases the editors characterize as "specialized Soviet intelligence terminology," a jargonized code for the Russian espionage cognoscenti (and an intended verbal barrier for the uninitiated) incorporated into the plain text. ¹⁸ There is, finally, a list of the 99 translated messages and, in italics, the editors’ notations of the names security agency cryptologists associated with each message. Nowhere in Venona is there an explanation of how and why the cryptologists linked each codename with a real name, and why and how, in some cases, the cryptologists concluded that real names were being used rather than codenames There is no discussion of why some codenames changed and others did not. Since so much of the interest in these documents hinges on these identifications, the absence of explanation is a serious lacuna.

      The importance of this omission is in part occluded by the instant impact of viewing such honest looking documents. The releases appear as if they had just been removed from secret government files. Albeit crossed out, many still bear the legible notation "Top Secret." Parts apparently too dangerous for the eyes of ordinary readers are completely blocked out. ¹⁹ Many of the messages are incomplete, the absent portions marked by brackets often containing a note on the number of units missing, although what is meant by a unit in terms of size of omission is unexplained. ²⁰ One or another of the ten code words cited in the introductory material, "fellow countryman," for instance, are retained and reproduced in upper case letters in the main body of the text of many of the releases, with the NSA translation added in brackets, thereby retaining aspects of the foreign codeness of the original documents. Producing the same effect, some Russian code words are left untranslated and reproduced using the Roman equivalents of Cyrillic letters. These typographical features help create an impression of authenticity.

      Footnote letters and numbers have been interpolated into the text, and anonymous footnotes, sometimes of a length far in excess of the message, are added to the bottom of the message as if they flowed automatically from the text instead of being material added by translators and/or editors. The footnotes contain the crucial information of the names; sometimes there is an indication that the identification is only probable; sometimes there is an apparently gratuitous amplification as when it is noted that William Perl is also known as Mutterperl, a fact not in the message. ²¹

      The most effective authenticating detail is the reproduction of the material as typescript, the overwhelming majority of which was generated on manual typewriters, revealing all the unevenness of that crude technology. ²² A smoother, neater, right and left hand justified word processed Venona message would not convey the same immediacy and visual authority. Here then, we are left to infer, is the NSA’s actual working copy. The preliminary quality of the documents, with their irregular typing and their occasional crossings out are, however, matters of surface. Between the acquisition of this material in the 1940s and its appearance in Benson and Warner’s Venona in 1996, the communiqués were decoded and reworked using an iterative process that involved re-writing as new material was decoded.

      The complex history of that revisionary process – as perceived errors were corrected, slightly different words added or subtracted, that is, the whole messy business of translation and editing of texts – is almost entirely suppressed. A concern about how words and phrases were selected is important because even small changes of words can alter the meaning of these documents enormously. If, for example, instead of "recruited," the messages said "met," it would undermine the notion that a spy ring was being formed. What if "bonuses" carried with it the sense of "charitable donation" or "contribution"? What if "bonus" had no fixed meaning? That concern is occasionally reflected in the notes, as when, in a footnote in "Washington [Naval-GRU] 2505-12 to Moscow, 31 December 1942," a translator points out that "MATERIAL is often used in the sense of ‘documents’or ‘documentary material,’" but, in the context of this intercept, "appears to mean ‘information’." In the same set of notes, there is the comment that "KhoZYaJSTVO is very difficult to translate out of context. It can mean ‘economy’, ‘farm’, ‘establishment’, ‘household’." ²³ Such formal acknowledgements of the uncertainties of translation are rare.

      A related question involves the order in which the releases were translated and read. Although the documents in their present condition are arranged chronologically, they are, in fact, the end result of a lengthy process that did not proceed chronologically. The documents, for the most part, carry only two dates, the date the message was sent, and another date that is unexplained, but which may be one of the dates (presumably the last date) the message was worked on. There is no record or notation on the Venona messages of all the dates on which the partial decryptions and translations were made, footnotes added or amended, and names confirmed. There is no indication of who worked on which documents. The absence of this information from the messages reproduced in Venona tends to encourage a perception of stability and certainty about the plain text that a more heavily annotated version, with its accumulated evidence of choices made, might not so readily convey.

      The concern about the chronology of the decryption and translation is not only about the way the appearance of authenticity and authority is constructed; it is also about how the NSA developed its version of the Rosenberg story. Given the need to find a spy ring to justify the Venona project, it is crucially important to ensure that the desire for certain readings to exist did not help to create those readings. It would therefore be helpful to know that the documents that are now being presented as precursors to the arrests of Fuchs, ²⁴ Gold, ²⁵ Greenglass, ²⁶ Sobell, ²⁷ and the Rosenbergs were all translated in the form they now appear before the arrests. Otherwise, it can be argued that the arrests influenced the translations of the Venona releases. ²⁸

Amongst the 3,000 decrypts are 19 messages related directly to the Rosenbergs, identifiable as such because the name of Julius Rosenberg is provided in the Venona translators’ footnotes as the person designated by the codename ANTENNA or LIBERAL in the messages. ²⁹ Of these, twelve appear in Venona. Let us take these documents at face value, assuming they are exactly what the NSA and the CIA say they are, authentic and unaltered KGB traffic. Let us accept that they have been accurately decoded, decrypted, and translated. Let us put aside questions of chronology. What do they tell us about the activities of Ethel and Julius Rosenberg?

      Most of the Rosenberg messages concern Julius. He is described as having a wife, Ethel, a woman of strong politics and sickly disposition:

Information on LIBERAL’S wife. Surname that of her husband, first name ETHEL, 29 years old. Married five years. Finished secondary school. A FELLOWCOUNTRYMAN since 1938. Sufficiently well developed politically. Knows about her husband’s work and the role of METR and NIL. In view of delicate health does not work. Is characterized positively and as a devoted person. ³⁰

This communication, which is the only one that mentions Ethel by name, ³¹ indicates that she was known to and approved by the KGB, and it associates her with marriage, physical weakness, graduation from "secondary school," communism (FELLOWCOUNTRYMAN is defined as meaning a communist), knowledge of her husband’s work and "the role of METR [identified as Joel Barr or Al Sarant] and NIL [unidentified]" but not with any acts of espionage.

      Most of the Rosenberg traffic is not, however, concerned with details of domesticity. Of the nineteen Rosenberg related messages, several that were not included in Venona concern the acquisition and use of cameras, which were unavailable in New York and had to be purchased in Mexico and posted back to the United States. Of the twelve KGB messages in Venona, Julius Rosenberg most often appears in relation to the recruitment of friends (Albert Sarant) and relatives (Ruth Greenglass). ³² What they are recruited for is not explained. Despite the seemingly pointless message about Ethel, it is difficult to imagine that the KGB would be busy transmitting in encrypted code reports on totally innocuous activity. Those who appear in the KGB traffic are presumed to be guilty of something.

      As corroboration of guilt, there are a few messages in the Rosenberg collection connected to the payment of bonuses. ³³ These not only point a finger of wrongdoing at those receiving such payment, they also advance the main theme of Venona, that Americans were willing (if not totally cost free) tools of the KGB. Document 55, New York 1314 to Moscow, 14 September 1944 "William Perl, again," is typical of this motif:

Until recently GNOM was paid only the expenses connected with his coming to Tyre. Judging by an appraisal of the material received and the rest [1 group garbled] sent by us GNOM deserves remuneration for material no less valuable than that given by the rest of the members of LIBERAL’s group who were given a bonus by you. Please agree to paying him 500 dollars. ³⁴

GNOM, we are told, is William Perl and LIBERAL is Julius Rosenberg. "The material received" is never specified. What "the rest [1 group garbled]" refers to remains undefined. The names in this message (and elsewhere in Venona) are clear; the actions, typically, are a blur. Although the contexts in which Rosenberg, Greenglass, Sarant, and Perl are discussed suggest they are not innocent, what they are guilty of is never stated. ³⁵

      We are invited by the NSA and the CIA to accept the Venona traffic as definitive evidence that a Rosenberg spy ring existed. Whatever Julius Rosenberg was engaged in, nowhere in these documents do we find the corroboration that he committed "the crime of the century," the theft of the secret of the atomic bomb. These messages, taken at face, suggest that Julius was engaged in some form of espionage. Yet in the absence of precise knowledge about what information was transferred to the Soviets, the messages convey the appearance of guilt without the certainty. There is always the possibility that some, much, most, or all the information that Julius supplied to the Soviets was not secret; in that case, he might have been engaged in unauthorized technology transfer but not necessarily in espionage.

      Aside from being married to Julius, apparently knowledgeable about his work and a recommender of Ethel’s sister-in-law as clever, Ethel stays at home and does no work at all. Moreover, there are other messages, unrelated to the Rosenbergs, that point to a scientist working at Los Alamos, codenamed MLAD, who provided the USSR with information about the atomic bomb. At what point the government began to suspect MLAD’s activity is unclear. MLAD has been identified as Theodore Hall, and, unlike the Rosenbergs, he has acknowledged that he passed information regarding the atomic bomb to the Russians. ³⁶ MLAD was never charged or arrested.

      If the Venona documents are accepted at face value, as the uncensored communication between KGB agents working in the United States to their counterparts in Moscow, then what they tell us is that, if there was atomic espionage, it was not Ethel and Julius Rosenberg who were engaged in it. If, moreover, the Venona intercepts were the basis for the arrest, trial, and execution of the Rosenbergs, as is now alleged by the FBI, then it is not farfetched to suspect that Ethel and Julius Rosenberg were framed, a position that has been advanced by Morton Sobell, ³⁷ the Rosenberg children, ³⁸ and Walter and Miriam Schneir, ³⁹ among others.

Naturally enough, the Venona project and its interest in the Rosenbergs did not come from political terra incognita, nor did it fall on virgin political soil. The release of the Venona decrypts is embedded in an ongoing debate about the nature of the Cold War, and needs to be understood as justifying and advancing the official version of that period as its history continues to be scrutinized and contested from a variety of revisionary perspectives. In other words, whatever the status of its truth claims, the documents and the publication are part of a political debate, framed by a particular reading of the recent past, and brought forth in the mid-to-late 1990s in a way that reinforces that reading.

      Although the stated goal of the Venona project was to provide a panoramic view of Soviet espionage in the United States during the 1940s, the early release of the Rosenberg messages provides strong evidence that the NSA and the CIA were especially concerned with influencing the way the Rosenberg case is now being interpreted. By the time of the Venona releases, in drama, poetry, art, ⁴⁰ and, most dazzlingly, in such fiction as E.L. Doctorow’s The Book of Daniel ⁴¹ and Robert Coover’s The Public Burning, ⁴² the Rosenbergs were understood to be victims of the Cold War whose guilt has never been established and whose punishment far outweighed any crime they may have committed.

      In the domain of scholarly prose and commercial non-fiction, the findings are more divided. Over time, two positions on the Rosenbergs have evolved, each with its own variations and modulations. On the one hand, those who accept official history judge the Rosenbergs guilty of having passed the secret of the atom bomb to the Russians; even though they were charged with conspiracy to commit espionage, a distinction of major legal import in terms of the rules of evidence. ⁴³ On the other hand, revisionist historians, civil libertarians, and others argue that the Rosenbergs were convicted by the hysteria of the time, that there were many procedural irregularities in their trial, and that their sentence was unnecessarily harsh because they had not been proven guilty, or because they were innocent, or because what they (and especially Ethel) were convicted of did not warrant the death penalty.

      By the early 1970s, as revisionist American history was gaining ascendancy in the academy, this second version of the Rosenberg case threatened to topple the official view. The Rosenberg children, Michael and Robert Meeropol, began a long and convoluted process, not yet complete, of extracting all the files related to their parents from the FBI and other agencies of government under the then newly enacted Freedom of Information Act. To date, this effort has not produced the smoking gun to prove conclusively that the Rosenbergs were framed, and it may have been naive to expect that such definitive evidence now (or ever) exists in a format that does not require interpretation. Nevertheless, research using the material released under Freedom of Information requests confirms and amplifies the contention that there were significant procedural irregularities that prevented the Rosenbergs from receiving a fair trial.

      Central to a revisionist reading of the Rosenberg case is the work of Walter and Miriam Schneir, whose book, Invitation to an Inquest, appears both in hardback and paperback, thus making it, unlike previous studies of the Rosenberg case, available to a mass market. Moreover, it bears the hallmark of respectability, mainstream commercial publishers, starting with Doubleday who put out the first 1967 edition and ending, in 1983, with a fourth edition published by Pantheon. The thesis the Schneirs persuasively advance in all editions is that the Rosenbergs were framed and convicted of a crime that did not take place. In their 1983 edition, the Schneirs incorporate into their argument material obtained from government files under the Freedom of Information Act. They also address the persistent rumors that began circulating from about the time of the Rosenberg execution, of important evidence, suppressed for reasons of state, that, if released, would prove the Rosenbergs guilty. They note that, despite repeated FBI claims of a "Rosenberg spy ring," the Justice Department made no arrests, and that a Justice Department report concedes that "investigation of all logical leads has, so far, failed to produce any appreciable results." ⁴⁴ The Schneirs further note that, "in early 1957, the Department of Justice abandoned the entire project." ⁴⁵ Not unreasonably, the Schneirs interpret the failure to make arrests as evidence that there was no spy ring. The Schneirs end their 1983 edition by linking the Rosenberg case to the Dreyfus case. Implicit in the analogy is a belief that, like Dreyfus, the Rosenbergs were innocent, and deserve exoneration.

      In 1983, the same year as the Schneirs’ fourth edition of Invitation to an Inquest appeared, Ronald Radosh and Joyce Milton published their study of the Rosenberg case, The Rosenberg File. A Search for Truth. ⁴⁶ Using previously unavailable material, primarily from FBI files released under the Freedom of Information Act, Radosh and Milton spruced up and modernized the official version of the Rosenberg case. ⁴⁷ At the time of its publication, this work was heralded as definitive. The reasons for such acclaim are easy to find. The book is crisply written, and finds fault both with the government’s handling of the case (in particular, the use of Ethel Rosenberg as a lever to extract a confession from her husband) and with the Rosenbergs’ diehard communism. Thus, its conclusion that Julius was guilty, that Ethel knew what he was up to, and that American Communists were involved in extensive spying for the Soviet Union, seemed balanced, moderate, and reasonable. This attempt to split the difference in the Rosenberg case evokes the atmospherics of fairness without actually sacrificing the effects of bias. There is surely a moral as well as a legal disproportion between the actions of individuals, even if criminal, and the deformation of the law by the apparatus of the state in efforts to prosecute such individuals. This disproportion is never adequately addressed in The Rosenberg File. The emotional weight of the Radosh and Milton line of argument is towards a view of the Rosenbergs as guilty, if not exactly as charged, at least of something.

      The reviews in the establishment press – The New York Times Book Review, The New York Review of Books, The Times Literary Supplement, The New Yorker – were uniformly favourable. ⁴⁸ And then the battle of the books began. Responding to these reviews, and to The Rosenberg File, in the same magazines and literary supplements, ⁴⁹ and in the independent and socialist press, ⁵⁰ scholars and partisans raised serious questions about Radosh and Milton’s documentation, accuracy, selectivity, omissions, and faulty reasoning. These exchanges continued for well over a year. The emotional high point of this debate came relatively early, however, at a 1983 happening at New York City’s Town Hall, entitled "Were the Rosenbergs Framed?" ⁵¹ Radosh and Milton and Walter and Miriam Schneir, playing to a packed audience, confronted each other’s  versions of the Rosenberg case in often angry exchanges. Writing with less emotion in the scholarly journal, New York History, in the longest and most thoughtful commentary on The Rosenberg File, Edward Pessen concluded that the work "falls far short of being a reliable, let alone definitive, book on the subject." ⁵² As the United States moved to the right in the 1980s and 1990s, the concerns of Pessen and many of the other participants in this exchange did not gain the hearing they warranted.

      By 1983, then, the Schneirs and Radosh and Milton had provided each side of the Rosenberg controversy with analysis and information sufficient to encourage further debate without, however, delivering the long-awaited knockout punch. Between 1983 and the release of the Venona messages, the Soviet Union collapsed. There was, at that time, a strong expectation that the files of the KGB would be opened and unsettled issues like the Rosenberg case would, in all likelihood, be resolved. The wished for research cornucopia did not, however, materialize. And, if it had, it would undoubtedly have provoked the same kinds of questions about authenticity and provenance that the Venona messages elicit.

      What followed within Russia was a vast dislocation that left employees and former employees of many state agencies unemployed, underemployed, and/or poverty-stricken. Some KGB agents (whether real or alleged is hard to know) cottoned on (or were nudged along by academic entrepreneurs) to the value of the confessional mode in the west, and rushed into print with "revelations." Such information needs to be understood at least in part as a supplementary retirement package for incomes made meager by the demise of the Cold War and the USSR. This is not to say that what these Soviet agents have to say is valueless; it is just that knowing how to evaluate such interventions requires care. Perhaps the best example of the difficulties with the Russian "tell all" genre is the book by Pavel and Anotoly Sudoplatov entitled Special Tasks: The Memoirs of an Unwanted Witness — A Soviet Spymaster, published in 1994, a little more than a year before the release of the first Venona documents. In a chapter on "Atomic Spies," the Sudoplatovs accuse four leading atomic physicists associated with the Manhattan Project, (Neils Bohr, Enrico Fermi, Leo Szilard, and J. Robert Oppenheimer), of having supplied vital information about the atomic bomb to the Soviet Union. As to the Rosenbergs, according to the Sudoplatovs, they were very minor players. Here were entirely too many spies, and the wrong ones to boot. In terms of resolving questions about the Rosenbergs, then, the Sudoplatov comments were useless. Indeed, the controversy created by the Sudoplatov "revelations" made it clear that information emanating from Russia would not automatically be seen as reliable, much less persuasive. ⁵³

      This, then, was the state of play regarding the Rosenberg case at the time of the first Venona releases.

The initial impact of the Venona releases can be gauged by the shift in the position of the Schneirs. Writing in The Nation magazine in August 1995, less than a month after the first Venona releases, the Schneirs replace their belief in the innocence of the Rosenbergs with a distressed acceptance that "during World War II Julius ran a spy ring composed of young fellow Communists, including friends and college classmates whom he had recruited." They then comment on the failure of the Venona releases to corroborate the evidence used against the Rosenbergs during their trial: there were "no drawing of lens molds, no sketch of ‘the atom bomb itself,’ no Jell-O box recognition device or password using Julius’s name — in short, none of the testimony so essential in convicting Julius is verified." Since the Venona releases do not confirm the evidence offered at the trial, and since the evidence of the trial was shaky to begin with, the Schneirs’ original argument that the Rosenbergs were framed still retains its persuasive power. But they take no comfort in the solidity of their basic position. The force of the Venona releases is nowhere more apparent than in the Schneirs’ highly emotive concluding remarks. In their penultimate paragraph, they say that they now believe the leadership of the American Communist Party knew about, and by implication condoned espionage, a position that is the cornerstone of counter-revisionist histories of the Cold War. They base their conclusion on a Venona document dated 5 April 1945:

If [6 groups unrecovered] LIBERAL’s membership of the FELLOWCOUNTRYMEN’s ASSOCIATION [ZEMLYaChESTVO] [5 groups unrecovered] and precise information about him through the leadership of the FELLOWCOUNTRYMEN [ZAEMLYaKI] does not exist. The supposition is to the presence in [{number unreadable} groups unrecovered] D.B. was reported by LIBERAL himself to the leadership of the FELLOWCOUNTRYMEN.

It is difficult to see how this incomplete and incomprehensible communication can be used as confirmation of anything. Nevertheless, the Schneirs bestow on this message more coherence and intelligibility than even the NSA was prepared to give it; since it was not included in the Benson and Warner Venona, supposedly containg the most important of the Venona intercepts.

      They end, lamely, with a wringing of hands: "This is not a pretty story," they say. ⁵⁴ "We know that our account will be painful news for many people, as it is for us." ⁵⁵ But even if the Schneirs now believe that Julius Rosenberg committed some low-level espionage, they do not believe that Ethel Rosenberg did. The release of the Venona documents reinforces the argument the Schneirs had been making since the late 1960s, that the American government conducted a show trial and then a murder. Such reconfirmation of their position ought to be the occasion for angry demands that the Rosenberg case be reopened, but it is not. The initial response of the Schneirs to the Venona releases seems a failure of nerve, inexplicable except in terms of the right-wing drift of American politics, which encourages even sophisticated critics of American domestic policy to read indeterminate texts of unverified provenance as proof positive of extensive Cold War Communist subversion.

      If the Schneirs’ initial response to the release of the Venona documents was a premature capitulation, the response of Radosh and Milton was undisguised triumphalism. In 1997, under the prestigious imprimatur of Yale University Press, the second edition of The Rosenberg File appeared, virtually unchanged from its first edition. A new introduction positions the work in relation to the material that has appeared since the first edition in 1983, and in particular to the Venona releases. Radosh and Milton make no attempt to address the serious concerns about documentation, accuracy, and selectivity raised by the reviewers of their first edition.

      For Radosh and Milton, the Venona releases represent the final word on the Rosenberg case. In their opinion, the documents demonstrate conclusively the guilt of Julius, who, "far from being a political dissenter prosecuted for his espousal of peace and socialism ... was an agent of the Soviet Union, dedicated to obtaining military secrets." ⁵⁶ What exactly it means to be "dedicated to obtaining military secrets" they do not say. Does someone so "dedicated" conspire to commit espionage, or does that person actually commit espionage, or does that person, perhaps, do nothing at all except believe that obtaining secrets for the USSR might be a good idea?

      Julius Rosenberg was charged with a specific crime, conspiracy to commit espionage, and in particular, amongst the overt acts, of having conspired with David and Ruth Greenglass to steal atomic secrets and transmit them to the Soviet Union. With their choice of the word "dedication," Radosh and Milton remove the need for any evidence of an overt act. "Dedication" criminalizes a state of mind.

      And what of Ethel? Even if Julius was guilty of conspiring to commit espionage, Ethel was not. But the logic of "dedication" makes Ethel guilty too. The subtext of the Radosh and Milton position proceeds: (1) since American Communists were dedicated to espionage, and (2) since Julius and Ethel were dedicated Communists, it follows that (3) both Julius and Ethel were dedicated to espionage. Radosh and Milton conclude:

The decision to prosecute Ethel Rosenberg on a capital charge, in an effort to put pressure on her husband, is hardly surprising. Although we continue to feel that the use of the death penalty in this context was improper and unfair, the Venona releases show that, overall, our justice system functioned with integrity under trying circumstances. ⁵⁷

The brutality and shallowness of this judgement compromises the Radosh and Milton work. Their caveat about unfairness is a throwaway line. They know that Julius was executed before Ethel. Julius died without providing the government with names of his alleged accomplices; once Julius was dead, how could the government possibly justify the execution of Ethel? If she was no longer a lever and was executed anyway, then the machinery of state was guilty not only of fabricating evidence to convict her, but of murder. How these circumstances illustrate the integrity of the American justice system as Radosh and Milton contend is unclear. Notwithstanding such concerns, Radosh and Milton’s work has become as foundational a text for such subsequent counter-revisionist studies of the Cold War as Haynes and Klehr’s influential and highly regarded Venona. ⁵⁸

      Although Radosh and Milton regard their work "as the most careful and balanced assessment of this important episode in the early Cold War era," it is, like all other studies of the recent past, provisional in nature, subject to critique, deconstruction, and revision. ⁵⁹ That process has already begun. Ellen Schrecker’s Many Are The Crimes: McCarthyism in America uses the Venona releases in a more critical and judicious manner than do either the Schneirs or Radosh and Milton. ⁶⁰ Like them, she accepts their authenticity. But unlike them, she queries several of their underlying assumptions. For example, she wonders about the importance of the espionage. "Was the espionage, which unquestionably occurred, such a serious threat to the nation’s security that it required the development of a politically repressive internal security system?" she asks. Her answer is that it did not. She notes that not all espionage activities were equally serious, and not every piece of information that found its way to the Soviet Union was a military secret. Finally, she points out that "the KGB officers stationed in the United States may have been trying to make themselves look good to their Moscow superiors by portraying some of their casual contacts as having been more deeply involved with the Soviet cause than they actually were." ⁶¹ Nevertheless, Schrecker’s assessment of the Rosenberg case is heavily indebted to her reading of the Venona releases:

The Venona releases also show that the KGB was ... pleased with Julius Rosenberg and his work. According to these documents Rosenberg, a mechanical engineer, was an active agent who recruited about ten of this friends, CCNY classmates ... into an espionage ring ....The documents do not identify all of Rosenberg’s people, but the ones they do, like Joel Barr, Alfred Sarant, Max Elitcher, Michael Sidorovich and William Perl, have long been connected to the case. During the war these scientists and engineers gave Rosenberg information about the weapons they were working on that he then photographed and handed to the KGB. ⁶²

Here she derives from the Venona releases a clarity and specificity that they simply do not have. The releases do not say that all of these classmates of Julius Rosenberg (those identified by name and those, after fifty years of investigation, still unknown) passed information to him about the weapons they were working on.

      Precisely because the Venona documents are so vague they invite readers to play "connect the dots" and superimpose on these disconnected and incomplete communications a narrative continuity that derives not from their intrinsic meaning, but from prior knowledge of the Rosenberg story. In other words, when Schrecker says that "the Venona documents ... show," what she means is that if the Venona documents are read in relation to already existing versions of the Rosenberg case, then they illustrate the case. Take, for example, Schrecker’s acceptance of the spy ring, a group supposedly made up of Joel Barr, Alfred Sarant, Max Elitcher, and others. Schrecker says she finds it credible that the Venona documents associate these men with Julius Rosenberg’s spy ring because they have long been connected to the Rosenberg case. It is not necessarily that their names in the Venona documents confirm their role in the Rosenberg story, but the other way round. It is just as possible that because Barr, Sarant, Elitcher, and the others were friends and classmates of Ethel and Julius Rosenberg, they were sucked into the investigation and, once implicated, they were assumed guilty by their past associations. They were then available to have their real names correlated with code names, particularly since the code names have few identifying particulars. As I have argued earlier in this paper, without further clarification about when the Venona releases were translated, the correlation between real and code names may well have been established after rather than before the arrest of Julius and Ethel Rosenberg, in which case the names in the Venona releases cannot be used as corroboration of a spy ring.

      So reading the Rosenbergs after Venona is not very different from reading the Rosenbergs before Venona, except that the revisionist approach to the case has been temporarily muted by an increasingly noisy right-wing counter-revisionism. Even so, the official version of the Rosenberg case continues to unravel. While accepting, like Schrecker, that the Venona messages demonstrate the guilt of Julius, the most recent contribution to the literature of the Rosenberg case, Sam Roberts’ The Untold Story of Atomic Spy David Greenglass and How He Sent His Sister, Ethel Rosenberg, to the Electric Chair, denies the guilt of Ethel. Roberts’ interviews with David Greenglass confirm what the Rosenbergs and their supporters have long contended, that Greenglass perjured himself when he testified that Ethel typed the secrets of the atom bomb. ⁶³ Since Greenglass’ testimony provided the only evidence that Ethel had participated in an overt act, Greenglass’ admission to Roberts that he lied undermines the credibility of all his other statements at the trial. With Greenglass’ testimony in tatters, the official "case" against the Rosenbergs collapses.

The guilt of Julius now hinges on nineteen Venona messages. This seems a flimsy basis on which to declare the Rosenberg case closed. Further examination of the accuracy of these messages and analysis of their contexts may very well further qualify their meaning. Some, even many, of the Venona releases may be exactly what they appear to be. But it does not follow that all 3,000 are exactly what the NSA, the CIA, Allen Weinstein, Radosh and Milton, and Haynes and Klehr say they are, if for no other reason than that neither the US translators and decrypters nor the KGB and their informants are infallible.

      There is general agreement that the process of decoding was complex and difficult. Indeed, the code has not yet been completely broken since components of varying length within the supposedly decoded messages are still not decoded. As I understand the process from a conversation with an NSA spokesperson in 1999, the messages were in Roman letters because American telegraph services would not transmit material in any other form. These letters correlated to numbers, which in turn correlated to Cyrillic letters. The Cyrillic letters presumably were combined into Russian words, which were then encrypted by the interpolation of random units. These messages, decoded and decrypted, then had to be translated into English. It taxes credibility to believe that the production of English plain text versions of the Venona intercepts are entirely accurate.

      As well as inaccuracies of translation, there is always the potential for errors of transmission. Did those supplying information to the KGB always communicate complete and unvarnished truth? Did the KGB agents always understand the information they were receiving? And, finally, did they always transmit that information accurately, given that they too had to code and encrypt data? Take, for example, one of the first messages translated by American cryptographers. The intercept "New York 1699 to Moscow, 2 December 1944" provides a list of seventeen scientists engaged in "the problem," that is, American atomic research:

Enumerates [the following] scientists who are working on the problem — Hans BETHE, Niels BOHR, Enrico FERMI, John NEWMAN, Bruno ROSSI, George KISTIAKOWSKI, Emilio SEGRE, G. I. TAYLOR, William PENNEY, Arthur COMPTON, Ernest LAWRENCE, Harold UREY, Hans STANARN, Edward TELLER, Percy BRIDGEMAN, Werner EISENBERG, STRASSENMAN. ⁶⁴

Fifteen of those mentioned were involved in the American atom bomb project. Two of them, Werner Eisenberg and Strassenman, had no connection with the project. ⁶⁵ Eisenberg was, according to West, actually Werner Heisenberg, who not only was not invovled in the American project, but was the 1932 Nobel Prize winner in physics who remained in Germany during World War II. ⁶⁶ Eisenberg and Strassenman are mistakenly linked to the other fifteen either by the informant or by the KGB agent. What such an error demonstrates is that the Venona documents need to be read cautiously and critically. This concern about textual accuracy would obtain even if there were no ideological predisposition by the employees of the NSA to read this material in a particular way.

      Scrutiny of text is one way in which the Venona messages may be reassessed; study of context is another. The Venona messages need to be read in relation to FBI and other US government agency files; they also need to be read in relation to KGB and other Russian government files. One of the great mysteries of Venona is that, through William Weisband, who worked on Venona and was thought to be a Soviet agent, and Kim Philby, who was a Soviet agent and, according to Benson and Warner, "received actual translations and analyses [of the Venona material] on a regular basis," the Soviets knew, or ought reasonably to have known, that their codes were broken. ⁶⁷ So why did they continue to use them? Finding the appropriate contexts to answer this and the other questions provoked by the Venona intercepts will undoubtedly influence not only how the Venona intercepts are read, but also how the Rosenberg case is understood. Without those contexts, the Venona material and what it is supposed to tell us about the Rosenbergs must be approached with great caution.

Notes

¹ The Foley Square courthouse in lower Manhattan in which the Rosenbergs were tried was also the site, just weeks before, of the highly publicized Smith Act trials of the leaders of the CPUSA, creating a strong visual link between the two trials, which reinforced their ideological connections. For a political overview of this period, see David Caute, The Great Fear. The Anti-Communist Purge Under Truman and Eisenhower (New York 1978).

²The problems with achieving an authoritative plain text did not end with decoding, decryption, and translation. According to Haynes and Klehr, "National Security practices on transliterating Russian words and names from the Cyrillic to the Latin alphabet changed several times. Further, a portion of work was done by British linguists, who rendered the translations in British English rather than American English."  John Earl Haynes and Harvey Klehr, Venona. Decoding Soviet Espionage in America (New Haven and London 1999), ix. Haynes and Klehr raise this matter to explain their editing of the plain texts to produce a "single standard of anglicization" so that readers will not be left to wonder if the "‘Anatolii’ of one document is the same person as the ‘Anotoly’ of another." Haynes and Flehr, Decoding Soviet Espionage, ix. Their exercise of linguistic standardization rests on the premise that all references to "Anatolii" and to "Anotoly" are to the same person, although why this assumption should be made is not explained. Indeed, according to Benson and Warner, "the KGB occasionally re-used covernames; consequently, a single covername can designate two different persons." Robert Louis Benson and Michael Warner, eds., Venona Soviet Espionage and The American Response 1939-1957 (Washington D.C. 1996), 191. Difficulties created by differences between British and American English are not resolved by standardizing only the spelling of names. British and American English differ in relation to usage as well, and without knowing the nationality of the translator of an intercept (or portion of an intercept), it is impossible to know whether the translation has been nuanced by the nationality of the translator. Other problems with generating an accurate plain text are taken up elsewhere in this paper, particularly in the concluding section.

³Robert Louis Benson, Introductory History of Venona and Guide to the Translations (Fort George G. Meade, MD 1995), 1. See also "Venona Chronology," ttp://www.nsa.gov/docs/ venona/venona_chron.html (13 December 2001).

⁴Benson and Warner, Venona, 191. This caveat about privacy would suggest that the plain text versions of the intercepts were subject to alteration and editing as late as 1995-6.

⁵Sobell’s impressions of the proceedings and their significance in relation to his conviction are recorded on the H-DIPLO web site. There he notes that he is not definitely identified with any cover name, although he is tentatively associated with RELE in three. In a fourth, message 943 of 4 July 1944, RELE is described as having an artificial leg and is unidentified. Sobell, who does not have an artificial leg, then wonders "why, if I was supposed to be a major player in this spy ring (J. Edgar Hoover urged that I be given the death penalty) can they now not identify me in any of the 2200 messages?" Morton Sobell, "Sobell on ‘Venona and the Rosenbergs,’" 27 May 1997, 3, ttp://www2.h-net.msu.edu/~diplo/Sobell.htm (13 December 2001). For the reaction of Michael Meeropol, see "Subject: Michael Meeropol Statement on Ethel and Julius Rosenberg," ttp://www.english.upenn.edu/~afilreis/50s/ meeropol-on-rosenbergs.html (13 December 2001).

⁶Benson and Warner, Venona, back cover.

⁷Robert Louis Benson, Introductory History of Venona and Guide to the Translations (Fort George G. Meade, MD 1995).

⁸Robert Louis Benson, Venona Historical Monograph #2: The 1942-43 New York-Moscow KGB Messages (Fort George G. Meade, MD 1995); Robert Louis Benson, Venona Historical Monograph #3: The 1944-45 New York and Washington-Moscow KGB Messages (Fort George G. Meade, MD 1995); Robert Louis Benson, Venona Historical Monograph #4: The KGB in San Francisco and Mexico City. The GRU in New York and Washington (Fort George Meade, MD 1995); and Robert Louis Benson, Venona Historical Monograph #5: The KGB and GRU in Europe, South America and Australia (Fort George G. Meade, MD 1995).

⁹Robert Louis Benson, Venona Historical Monograph #6: New Releases, Special Reports, and Project Shutdown (Fort George G. Meade, MD 1997).

¹⁰ttp://www/nsa.gov/docs/venona.

¹¹Haynes and Klehr go some way to remedying this deficiency by providing in one of their appendices an alphabetical list of 349 names of persons (US citizens and others) "who had a covert relationship to Soviet intelligence that is confirmed in the Venona traffic." Haynes and Kehr, Decoding Soviet Espionage, 339. The list includes both codenames and real names. Footnotes direct the reader to endnotes which provide references to the relevant intercepts. Despite its usefulness, this method of correlation of names and documents does not provide easy means for assessing relative frequency of mention.

¹²Benson and Warner, Venona, xliv.

¹³Telford Taylor, Grand Inquest. The Story of Congressional Investigations (New York 1955), 138.

¹⁴David Caute, The Great Fear, 208.

¹⁵With the exception of the Rosenbergs and Sobell there were no other Americans convicted of espionage (or conspiracy to commit espionage) in the 1950s; Rudolf Abel, whose conviction ends this chronology, was a Russian operative, not an American communist.

¹⁶Benson and Warner, Venona, 1.

¹⁷Benson and Warner, Venona, back cover.

¹⁸This list of definitions creates precise and stable meaning for such crucial words as "fellow countrymen." Benson and Warner define that term as meaning "members of the local Communist Party," Benson and Warner, Venona, 192,  but it could just as plausibly be read as a catch-all phrase for those with Communist sympathies however shaded. The counter-revisionist argument that American communists were engaged in extensive espionage activity on behalf of the Soviet Union is strengthened by the NSA’s narrow and unambiguous definition of "fellow countrymen." The intercepts themselves, however, do not appear to require such fixed meaning.

¹⁹Benson and Warner explain that the release of these intercepts "involved careful consideration of the privacy interests of individuals mentioned," but this claim is difficult to assess without knowing what names are concealed,  Benson and Warner, Venona, 191. What is clear is that the privacy interests of some appear to be dealt with differently from those of others. For example, consider the intercept "New York 1657 to Moscow, 27 November 1944." The codename METR is associated with both Joel Barr and Alfred Sarant, and so could be either. Benson and Warner, Venona, 381. Despite the uncertainty, both names are provided. Likewise, in "Washington [Naval-GRU] 2505-12 to Moscow, 31