The River of Voices

The weekly production report for the Berlin tunnel, dated May 1955, listed the figures in neat columns: 368 conversations recorded, 47 deemed of intelligence value, 12 transcribed in full, 3 forwarded to Washington. A clerk in the Berlin Operations Base added the numbers to the running tally, then filed the carbon copy in a cabinet marked with the operation’s codename. The report was one of dozens generated each week by the machinery of interception. The tunnel had been operational for less than a month, and already the paperwork of surveillance had established its own rhythm.

The routine began each day before dawn. Technicians in the Rudow listening post, a cramped complex of rooms built into the tunnel’s western entrance, checked the recording equipment: racks of tape recorders, amplifiers, and the splicing tools needed to maintain the physical connection to the Soviet cables. The three trunk lines, running from Karlshorst to Schönefeld through the Altglienicke district, carried military communications for the Group of Soviet Forces in Germany. The taps, installed with precision in April 1955, drew from this flow. Technicians changed reels every few hours, labeling each one with date, time, and line number. Magnetic tape accumulated in stacks, then in boxes, then in crates. A requisition form from June 1955 requested an additional two hundred reels of recording wire, listed as “office supplies” for the Berlin Base. The logistics of interception demanded their own bureaucracy.

Translation fell to a team of linguists, most of them native German speakers recruited by the CIA and MI6 for their knowledge of Russian military terminology. They worked in shifts, headphones clamped over their ears, transcribing conversations that analysts had flagged for attention. According to the declassified CIA history, the tunnel eventually recorded roughly 90, 000 Soviet communications over its eleven-month lifespan.

The sheer volume was staggering: sixty-seven thousand hours of Russian and German conversations were sent to London for transcription by a special section staffed by 317 Russian emigres and German linguists. Teleprinter signals, many of them multiplexed, were also collected on magnetic tape and forwarded for processing. Linguists could not process every call. They prioritized based on keywords, speaker identification, and the judgment of supervising officers.

A call discussing troop movements received immediate attention. A call about leave requests or supply shortages might be logged and set aside. The system produced its own logic: the most interesting material rose to the top, while the rest sank into the archive, untranscribed but preserved.

Reports that emerged from this process traveled through secure channels to Washington and London. Each week, a summary arrived on the desks of senior intelligence officers, detailing what the tunnel had revealed about Soviet military dispositions, training schedules, logistics, and personnel matters. The CIA’s Berlin Base and the MI6 station treated the operation as a triumph. William Harvey, the CIA officer who had championed the tunnel from its inception, received the production reports with satisfaction. Peter Lunn, the MI6 station chief in Berlin and Harvey’s counterpart in the joint venture, shared the conviction that they had achieved something unprecedented: a direct line into the Soviet military machine, a penetration that no other operation had matched.

The value, however, was difficult to quantify with precision. Intercepted conversations provided a picture of Soviet military life in East Germany. Officers discussed promotions, transfers, and disciplinary matters. Drivers reported mileage and fuel consumption. Clerks processed paperwork. The tunnel captured the mundane machinery of occupation, the daily business of an army at peace. Occasionally, something more significant appeared: a reference to training exercises, a discussion of readiness levels, a complaint about equipment shortages. Analysts in Berlin and Washington sifted these fragments, looking for patterns that might reveal Soviet intentions. The Cold War demanded such vigilance. The tunnel fed the demand.

Each week of production, each reel of tape, each report sent to headquarters confirmed the wisdom of the investment. The tunnel had cost millions of dollars and thousands of man-hours. The CIA and MI6 had staked their reputations on its viability. That it worked—that the equipment functioned, the taps held, the Soviets did not detect the intrusion—seemed to validate the entire enterprise. Institutional momentum built with each passing month. The operation became, in the minds of its managers, a proven asset. The possibility that it might be something else—a channel for deception, a trap baited with genuine but controlled information—did not fit the narrative of success.

The Soviets, for their part, maintained their routine. The calls continued. Military communications followed the patterns one would expect from a garrison force. Nothing in the intercepts suggested that the Soviets knew of the tunnel or had altered their behavior because of it. This absence of evidence for detection was itself taken as evidence of success. The tunnel remained secret. Confidence of its managers grew.

British experience with Operation Silver, the Vienna tunnel that had preceded the Berlin operation, provided a model for what success looked like. Operation Silver, which ran from 1949 to 1955, had tapped into landline communications of the Soviet Army headquarters in Vienna. Its staff included Peter Lunn, who had brought lessons from Vienna to Berlin. The British monitoring station in Vienna had been disguised as a shop which sold tweed clothing, a cover that held for years. That operation ended only when Austria regained full sovereignty in 1955, and Soviet occupation forces withdrew. Vienna had proven that a tunnel could succeed. Berlin would prove that the model could scale.

The daily routine of the tunnel established its own normalcy. Technicians changed tapes. Linguists transcribed calls. Analysts wrote reports. Couriers carried material to the sending stations. Encrypted cables traveled to Washington and London. The cycle repeated, week after week. The operation became a fact of institutional life, a line item in the budget, a standing entry in production statistics. Sense of triumph that had accompanied its launch settled into a quieter confidence.

Confidence extended to the highest levels of both services. Allen Dulles, the director of Central Intelligence, received regular briefings on the tunnel’s output. CIA leadership, which had approved the operation despite its risks and costs, saw production figures as vindication. The British Secret Intelligence Service, which had partnered with the CIA on the project from its inception, shared the satisfaction. The tunnel represented a rare victory in the intelligence competition with the Soviet Union. Soviets had secured their most sensitive communications by moving from radio to landlines. The West, unable to break Soviet encryption, had found another way in.

Value of the intelligence, however, remained a matter of interpretation. The tunnel produced no smoking guns, no revelations of Soviet war plans, no evidence of imminent aggression. Instead, it produced a steady stream of military routine: logistics, administration, personnel. Analysts who processed the material found it useful for building a picture of Soviet force structure and readiness. The picture corroborated intelligence from defectors and aerial reconnaissance. Soviet forces in East Germany were a garrison army, equipped and trained for defensive operations, not preparing for an imminent offensive. The tunnel confirmed what other intelligence channels indicated. Confirmation was valuable. It reduced uncertainty. It filled gaps.

The gap-filling function became the tunnel’s primary contribution. Other sources—defectors, aerial reconnaissance, human agents—provided pieces of the puzzle. The tunnel provided more pieces. Cumulative effect was a richer, more detailed understanding of Soviet military dispositions. Understanding was not revolutionary. It did not change strategic assessments. But it added depth and texture to the picture. Consumers of intelligence appreciated depth and texture. They appreciated the sense that they could see inside the Soviet military, hear its internal conversations, understand its workings from the inside out.

Sense of penetration was itself a form of intelligence. That the tunnel existed, that it worked, that it remained undetected, told the Anglo-American intelligence services something about Soviet capabilities and procedures. Soviets had not found the tunnel. Their counterintelligence had not detected the intrusion. Their security was not impenetrable. The lesson was clear: with sufficient resources, sufficient ingenuity, and sufficient audacity, Western intelligence could breach Soviet defenses. The tunnel was proof of concept.

Success masked the fundamental question that would later dominate assessments of the operation: whether the intelligence it produced was genuine or controlled. In 1955 and early 1956, the question did not arise with any urgency. Material seemed authentic. It matched expectations. It filled the expected gaps. Possibilities that the Soviets might have allowed the tunnel to operate, might have fed it selected information, might have used it as a channel for deception—these remained speculative, unexamined, unwelcome.

Institutional inertia that the tunnel generated made questioning difficult. The operation had required enormous investment. It had involved multiple agencies, multiple governments, multiple layers of approval. Reputations of officers who had championed it were bound up in its success. Careers of those who worked on it depended on its continuation. Bureaucracy of intelligence production—the clerks, the linguists, the analysts, the couriers—had organized itself around the tunnel’s output. To question the tunnel’s value was to question the judgment of everyone who had approved it, planned it, built it, and operated it.

Barriers were not insurmountable. Intelligence officers were trained to think critically, to question sources, to assess reliability of information. CIA and MI6 had experienced analysts who knew that successful operations could be compromised, that valuable sources could be doubled, that enemies could feed false information through channels that seemed secure. Discipline of intelligence demanded such skepticism. But discipline competed with psychology of success. The tunnel was producing. Production was valuable. Value was visible. Risk of compromise was invisible.

Invisibility of the risk did not mean that no one considered it. Possibility that George Blake, the MI6 officer who had been posted to Berlin in 1955, might have betrayed the operation to the Soviets was not, at that time, a possibility that anyone in the CIA or MI6 entertained. Blake was a trusted officer, a veteran of the Korean War, a man who had been captured by the North Koreans and repatriated as a hero. His posting to Berlin was routine. His access to information about the tunnel was limited but real. Possibility that he was a Soviet agent was not on anyone’s radar. The operation was secure. The operation was succeeding.

Production continued through the summer and fall of 1955. Weekly reports accumulated. Running tally of intercepted conversations grew. Crates of magnetic tape filled storage areas. Analysts identified patterns, noted changes, tracked developments. Soviet military in East Germany went about its business. Calls continued. Tunnel listened.

Flow was not uniform. Some days produced more traffic than others. Some weeks yielded more items of interest. Rhythm of military life dictated rhythm of interception. Training exercises generated spikes in traffic. Holidays produced lulls. Analysts learned to read the rhythm, to anticipate fluctuations, to distinguish significant from routine. Work required patience and attention. It required the ability to listen to hours of mundane conversation in search of the fragment that mattered. Fragment might be a reference to a new piece of equipment, a change in command, a shift in readiness posture. Fragment might be nothing more than a complaint about the quality of rations.

Chaff was part of the value. Mundane traffic, routine calls, administrative trivia—all contributed to the picture. Picture was not built from dramatic revelations. It was built from accumulation of detail. Detail established a baseline. Baseline made deviations visible. Deviations, when they appeared, might indicate something significant. Routine made the non-routine stand out.

Wheat, when it appeared, was duly noted. A conversation about troop movements. A discussion of fuel allocations. A reference to a training exercise. Each item of interest was transcribed, translated, analyzed, and reported. Reports traveled up the chain of command. Consumers of intelligence read them, filed them, integrated them into their assessments. Tunnel had become a reliable source.

Cycle would continue for nearly a year. From May 1955 until April 1956, the tunnel operated, the tapes turned, the reports flowed. Operation would end only when the Soviets “discovered” it, staging a public revelation that made headlines around the world. But in the months between beginning and end, the tunnel was simply there: a fact of intelligence life, a producer of data, a source of information.

Flow created its own reality. For Anglo-American intelligence services, the tunnel was a success. It was producing valuable intelligence. It was penetrating Soviet communications. It was demonstrating Western capability. Reality of success was confirmed by production figures, weekly reports, steady accumulation of material. What managers did not see—what they could not see, given information available to them—was possibility that success itself was part of a larger design.

Design, if there was one, would become apparent only later. In 1956, the tunnel would be exposed. George Blake would be arrested, tried, and imprisoned as a Soviet agent. Full extent of his betrayal would emerge. Question of whether Soviets had allowed the tunnel to operate, and why, would dominate retrospective assessments. Debate would continue for decades. Participants would offer different interpretations. Documents would provide partial answers. Full truth might never be known.

But in 1955, none of this had happened. Tunnel was running. Voices were flowing. Intelligence was being produced. Managers were satisfied. Operation was, by every visible measure, a success. Invisible measures—possibility of compromise, risk of deception, danger of a mole within the system—remained invisible.

Counting went on. Weekly production reports continued. Tallies grew. River of voices widened. Confidence of managers deepened. Operation that had begun as a gamble had become an asset. Asset that had been built at such cost was now paying dividends. Dividends were measured in reels of tape, in transcribed conversations, in reports sent to headquarters.

Reality of that success, however partial or problematic it would later prove to be, had consequences. Tunnel produced intelligence. Intelligence shaped assessments. Assessments informed policy. Policy affected conduct of the Cold War. Chain of causation ran from taps in the clay to decisions in Washington and London. Decisions were made on basis of information that seemed genuine. Information was genuine, in sense that real Soviet officers were having real conversations about real military matters.

Source, however, was under Soviet feet. Cables ran through Soviet-occupied territory. Taps were physically accessible to anyone who knew where to look. Operation depended on Soviet ignorance of its existence. Ignorance was assumed. Assumption was wrong. Soviets knew. Soviets had known from the beginning. Soviets had decided to let the operation run.

Decision would later become the central puzzle of the Berlin tunnel. Why had Moscow allowed it? Obvious answer was to protect George Blake. Blake was a valuable agent, embedded in MI6, with access to British intelligence operations and personnel. Exposing the tunnel would have exposed Blake. KGB, informed by Blake of the tunnel plan from its inception, had faced a choice: shut down the tunnel and preserve the intelligence flow from Blake, or let the tunnel run and risk the possibility that the West would eventually discover the compromise. KGB had chosen to let the tunnel run. Choice made strategic sense. Blake was more valuable than the tunnel. Tunnel could be sacrificed later. Blake could continue to produce.

Choice raised another possibility. If Soviets knew about the tunnel, they could control what it heard. They could feed it selected information, shaping Western perceptions of Soviet military capabilities and intentions. They could use it as a channel for disinformation, planting false indicators of Soviet strength or weakness. They could turn the West’s own operation against it. Possibility of such manipulation would become a central theme of post-operation analysis. Possibility could neither be proven nor dismissed. Soviets had the capability. Soviets had the motive. Soviets had the opportunity. Whether they used it remained an open question.

Open question would be debated for years. Debate would never be fully resolved. Evidence was ambiguous. Tunnel had produced genuine intelligence—conversations that reflected real Soviet military concerns. But intelligence had also been consistent with Soviet interests—confirming that Soviet forces in East Germany were defensive in posture, not preparing for aggression. Had Soviets allowed the tunnel to hear only what they wanted heard? Or had tunnel simply captured the reality of a garrison force that was, in fact, defensive? Two possibilities were indistinguishable from perspective of the intercepts themselves. Material was what it was. Interpretation depended on what one believed about Soviet intentions.

Managers of the tunnel, in 1955 and early 1956, believed that the operation was secure. They believed that intelligence was genuine. They believed that Soviets did not know. Their beliefs were wrong. But wrongness of their beliefs was not yet visible. Beliefs were supported by evidence available to them: operation had not been detected, intelligence was coherent and valuable, production figures were strong. Evidence pointed to success. Success pointed to validity of the beliefs. Circle closed.

Closure would be broken in April 1956, when Soviets staged the discovery of the tunnel. Staging would be a public relations coup, a propaganda victory that highlighted Western espionage and Soviet vigilance. Discovery would also mark the end of the operation and the beginning of the reckoning. Reckoning would take years to unfold. Blake would not be exposed until 1961. Full story of the tunnel’s compromise would emerge only gradually, in fragments, through trials and memoirs and declassified documents.

In months of operation, from May 1955 to early 1956, there was only the production. Production was substantial. Production was valuable. Production was, by standards of intelligence collection, impressive. Tunnel had tapped into Soviet military’s internal communications and had extracted a river of voices. Voices spoke of routine matters, of logistics and personnel, of daily business of an army. Voices were not dramatic. Voices were not revelatory. But voices were real. Or so it seemed.

Seeming was enough. Operation had succeeded by its own metrics. It had produced intelligence. It had avoided detection. It had justified the investment. Managers were satisfied. Consumers were satisfied. Tunnel was a success. Success was real enough to sustain confidence. Confidence was real enough to blind.

Production reports accumulated in files. Weekly tallies grew. River of voices widened into a flood. Each report, each tally, each reel of tape confirmed what managers wanted to believe: that they had achieved something unprecedented, that they had penetrated Soviet defenses, that they were hearing the unguarded conversations of an adversary. Confidence built on this river of data now demands an explanation: why would Moscow allow it?