On the early morning of November 11th 2008, about 150 police officers raided the tiny village of Tarnac, in central France, where they arrested members of what the interior ministry described as a cell of anarchist militants suspected of carrying out sabotage attacks against the French national railway network.
The people detained were part of a group of French and foreign nationals who had been living for three years as a community in converted farm buildings in the tiny Limousin region village of 350 inhabitants where they also bought and managed the local grocery shop and ran a library.
Parallel to the raid in Tarnac, alleged accomplices were also arrested in Paris, in Rouen in northern France, and in eastern France, at the end of a surveillance operation that had begun in earnest seven months earlier. Later that day, then French interior minister Michèle Alliot-Marie triumphantly announced the arrests of "ultra-left, anarchist-autonomists" who had targeted the French railway system “as a symbol of the state”.
The French media just as immediately reported that the suspects were members of an organised far-left group which had been involved in a number of violent demonstrations in France and abroad, and most recently a demonstration on November 3rd in Vichy where European Union interior ministers were meeting to discuss immigration controls. “They all come from the ultra-left, none of the suspects belong to society,” said David Pujadas, anchorman of the France 2 TV channel evening news programme.
It was the start of what became known in France as ‘the Tarnac affair’, and which would soon become a judicial fiasco.
In all, of the 20 people arrested, nine were placed under investigation for suspected terrorist-related activity. In 2009, a tenth person was also placed under investigation on the same count. The alleged ringleaders of the Tarnac group, which villagers described in numerous media interviews as a mild and friendly bunch of middle-class city drop-outs living an alternative lifestyle, were Julien Coupat, 34 and his 25-year-old girlfriend Yildune Lévy.
Coupat, the privately-educated son of a Parisian doctor and a pharmaceutical company director and who went to the elite ESSEC business studies higher education school before founding a radical philosophical review, Tiqqun, and Lévy, an archaeologist from a middle-class left-wing Parisian family, had for some while been the object of surveillance by French police services for their far-left sympathies, beginning with an enquiry into the finances behind the purchase by Coupat and others in December 2005 of the farmhouse buildings in Tarnac (and which found no suspicious funding). Coupat was allegedly among the authors - a self-styled ‘invisible committee’ - of an anti-capitalist 2007 book entitled The Coming Insurrection. In January 2008, Coupat and Lévy had travelled to New York for an international meeting of anarchists and far left militants, and it was soon afterwards that surveillance of the pair, and their entourage, intensified, including phone taps placed on the grocery store in Tarnac.
The police swoop on the village on November 11th 2008 followed a series of sabotage attempts on high-speed railway lines in France over the preceding weeks in which purposely-built U-shaped metal bars with hooked ends were attached to the overhead power lines to cause a breakdown of electric current when the train ran into them.
During the night of November 7th-8th, four such hooks were attached to high-speed track overhead cables, on the north-, east- and south-bound lines, damaging several high-speed TGV trains that began ‘sweeping’ the lines in the early hours before normal services began. Meanwhile, several electricity transformer installations along rail tracks in Germany were also damaged by fires. The events caused enormous traffic disruption, with the delays of some 150 TGV services, including Paris-London Eurostar links, delaying the travel of more than 20,000 passengers.
While the disruption caused, and its cost, was significant, there was no apparent attempt to cause danger to lives in what was a sophisticated and organised operation of vandalism.
Responsibility for the sabotage was claimed the following day, November 9th, by an anonymous German group who said it was in retaliation for the death of a French anti-nuclear activist who was run-over by a train carrying nuclear waste during a protest in 2004.
However, it would emerge that the arrests in Tarnac followed a surveillance operation against Coupat and Lévy who had been followed by anti-terrorist police during the night of the sabotage incidents, when their car was shadowed as it roamed countryside close to the east-bound high speed track in the Seine-et Marne département (county) that lies just south-east of Paris. The surveillance officers reported that the couple’s car had been seen parked on a service lane beside the railway track where one of the hooks was placed. Meanwhile, three other members of the Tarnac group were also reported to have been sighted that night close to railway tracks in eastern France.
'Transcription errors' led to 160 kph car chase
On November 14th, three days after the arrests in Tarnac, Paris public prosecuter Jean-Claude Marin gave a press conference announcing that among those detained was “a hard core composed of five people”, of which one had been identified as “the leader of a structure with terrorist aims”, a crime which is punishable by a 20-year jail sentence. “This hard core,” Marin continued in a solemn tone, “had conceived a structure which called itself the ‘invisible cell’ and which had armed struggle as its aim”. The prosecutor said the cell had links with similar groups in Germany, Italy and Greece, and that the “hard core” he referred to had shown by their uncooperative attitude under questioning that they were “anchored in a position of marginalisation”.
“It is not excluded that these people had envisaged more violent actions, including against people,” he concluded.
The authorities, notably interior minister Michèle Alliot-Marie in a speech before parliament, applauded the efficiency of the investigations led by the newly reformed domestic intelligence services, re-named the DCRI, and the police anti-terrorist department, the SDAT. The arrests were a political coup for Alliot-Marie, who claimed responsibility for instigating the surveillance of the Tarnac group, and a boost for the reputations of the DCRI and the SDAT.
But before long the case began to flounder. Four of the nine initially detained were released after 96 hours, followed by three others three weeks later. Lévy was released in January 2009 after serving two months of preventive detention. Coupat was released six months after his arrest.
All, along with one other person, have since remained under investigation in the case, which was finally brought to a close in April this year when Judge Jeanne Duyé, who took over the investigation in 2012, announced the end of her enquiries.
In the intervening years, the case began to fall apart. A key witness against the Tarnac group, who gave a statement claiming that he had been present at meetings where Coupat had raised the “possibility of killing”, later told TF1 TV channel that the police had “added elements to his statement, taken from their files”. Then it was found that phone taps placed during surveillance of the Tarnac group in March 2008 had been made without proper authorisation. The incoherencies of the investigation were the subject of increasing revelations in the media, including by Mediapart, and in 2012 the magistrate then in charge of what had become a largely discredited case, Thierry Fragnoli, asked to be removed from the job because of what he called “a press campaign” against him.
Following Judge Duyé’s announcement on April 14th that she had reached an end to her enquiries, the case file is now with the public prosecutor’s office, which will later this summer advise on who, if anyone, should be charged and sent for trial - after which Duyé will take the final decision whether there is a case for prosecution.
After six years of investigations, the central, remaining evidence against the Tarnac group is a police surveillance report, a document coded in the investigating file as ‘PV 104’ (for ‘Procés Verbal 104’), which is the account of the shadow operation targeting Coupat and Lévy on the night of November 7th-8th. Crucially, it recounts how Coupat and Lévy, were seen parked in their car between 4.05 a.m. and 4.20 a.m. on a service slip road in countryside 75 kilometres east of Paris alongside a high-speed train line where one of the sabotage hooks were found.
PV 104 was drawn up by the French anti-terrorist police section, the SDAT, (Sous-direction anti-terroriste de la police judiciaire). But this, too, has been shown to contain disturbing incoherencies - as Mediapart reported back in December 2009 (article available in French here) - and the lawyers for those placed under investigation have requested, as is their legal right once the case has been closed, for further investigations into these be made before the final decision on whether a trial is to take place is reached.
The lawyers contest not only numerous factual details in the surveillance report, but also the very presence of some of the officers from the SDAT, the Police Judiciaire and the internal intelligence service, the DCRI, who it is recorded took part in the mission that night.
The incoherencies in the surveillance report centre on the recorded timings of events that night. In one example, it is recorded that Coupat and Lévy’s car, and also that of the police which was following them, travelled 27 kilometres of cross-country small roads in just ten minutes, which represents an average speed of 160 kph. The police subsequently claimed this was due to an error in the transcription of the details of the surveillance operation, in which 20 minutes were lost when a recorded time of “3.50 a.m.” should have read 3.30 a.m.
In their formal request for further investigations into the police surveillance report, sent on June 30th, the lawyers noted that the geographic inconsistencies of the report cannot be explained by typing errors. “On the other hand, a latter reconstruction of a sham of surveillance by using an internet site could perfectly explain them,” they wrote.
'Blunders' and 'approximations'
In June this year, French daily Le Monde published a report by one of its journalists who retraced the itinerary of the 20-hour surveillance operation of Coupat’s car, beginning during the day of November 7th, as described by the police officers. With the details of the surveillance report to hand, the reporter criss-crossed the countryside of the Seine-et-Marne département (county) south-east of Paris, and found that there were roads mentioned in the report which did not in fact exist, directions indicated that were wrong, and tunnels described at certain sites which were in fact bridges.
“During a shadow operation, the PV [report] is not written-up in real time,” cautioned a former anti-terrorist police officer, speaking on condition of anonymity. “The police officer is driving, he takes notes as best he can, with his radio set stuck between his legs, and once he arrives back in his office he reconstructs the route.”
Nevertheless, it remains surprising, at least, that none of some 20 SDAT and DCRI officers who had been following the car from 11 a.m. on November 7th failed to witness Yildune Lévy and Julien Coupat placing a metal hook onto the high-speed train track overhead power lines.
According to PV 104, Coupat’s car spent a lengthy moment at Trilport, five kilometres east of the town of Meaux before heading off at 3.50 a.m. The officers report that they then lost track of the vehicle, at 4 a.m., as it travelled along a small country road close to the east-west high-speed rail track.
The surveillance report noted that captain ‘A.L.’, head of the SDAT surveillance group, “understanding that the vehicle being followed, having not been signalled out by his colleagues, had manifestly stopped on the lane following the rail track” made “a pedestrian approach to the site” and found “the presence of the vehicle, parked with its lights out at the entrance to the said service lane at a few metres before the railway bridge.
But according to a separate police version of events, it was the driver of the SDAT surveillance car, ‘A.L.’s deputy, a police lieutenant and who is identified as ‘B.M.’, who carried out the approach by foot. Whatever the case, there is no mention of anyone being seen on the rail tracks below the officers, and even less so two anarchists with head torches in the process of fixing a metal hook onto the overhead cables with the aid of two, two-metre long tubes of PVC – like those astonishingly found as evidence one-and-a-half years later floating in the nearby Marne river.
Already in November 2011, the veracity of the surveillance report was the subject of an investigation – within the investigation – into possible fraud. However, a lack of cooperation from anti-terrorist and intelligence services, citing reasons of official secrecy, limited its progress. That sub-investigation was finally closed on April 7th this year, just as the principle investigation wound to an end, when the suspicions of fraud were dismissed. The incoherencies of the surveillance report were found to be the results of an “imperfect” document that contained “approximations” and “blunders”. The 17 officers from the anti-terrorist police and the intelligence service who were questioned as part of the investigation had, in the written conclusions of the magistrate in charge of the case, “all confirmed their presence at the scenes of the surveillance and confirmed the contents that concerned them” as appeared in the report.
During their questioning by the internal police investigation department, the IGPN, three of those officers involved in the surveillance operation said they had seen, from the car in which they were sitting, Coupat’s Mercedes parked beside the railway line at 4 a.m.. Yet, according to the account given by their colleague and deputy head of the SDAT team, ‘B.M.’, this would have been physically impossible, and which was why an approach by foot to the scene was necessary to spot Coupat and Lévy’s car.
A member of the DCRI intelligence team testified that he followed the couple’s car on its return journey to Paris, beginning at 4.20 a.m. Yet he had also claimed to have remained at the scene to inspect the railway track and had, at 5.10 a.m., observed a shower of sparks when a high-speed train passed the spot – suggesting it had hit the sabotage hook.
Yet more contradictions in the officers’ testimony appeared with that of a member of the Police Judiciaire who had initially claimed to have been based that night at the headquarters of the SDAT, situated in the Paris suburb of Levallois-Perret, west of the capital. But he also claimed during questioning by the IGPN to have been posted at Dhuisy, close to the high-speed rail track where the hook was later found, until 5.30 a.m.
Also striking is the impression given that many of the officers involved in the operation in fact took little part in the action. Just two of them said they followed Coupat and Lévy’s vehicle after it left for Paris at 4.20 a.m., and who include the officer who confusingly also claimed to have inspected the railway track. Four others said they also inspected the railway track and witnessed the shower of sparks as the high-speed train passed by.
In their June letter to Judge Duyé requesting further enquiries into the incoherencies of the surveillance report, the defence lawyers underline that the other officers “don’t follow the vehicle, they don’t help their colleagues check the track or search rubbish bins […] They go about nothing in the night and the cold”.
What mobile phone records revealed
Police lieutenant ‘B.M.’, who was also the officer who signed off the surveillance report, said he had helped “certain colleagues in the operation” search a rubbish bin at 5.30 a.m. in Trilport, where Coupat was allegedly seen throwing away a number of objects a few hours earlier. In the bin, they reportedly found official documents from the French railways company SNCF about its high-speed train tracks, and the packaging for a head torch. In a police report of evidence found during the police raid in Tarnac on November 11th 2008, a head torch of the same model was recorded as having been found in Coupat’s bedroom. However, none of the other officers in the surveillance operation recalled taking part in the search of the bin in Trilport.
Establishing the presence and locations of the surveillance team that night might have been possible by checking mobile phone movement. In March 2010, lieutenant ‘B.M’ handed the magistrate then in charge of the investigation, Judge Thierry Fragnoli, a revised copy of a chart he had himself drawn up indicating “the phone traffic of cells of mobile phone operators Orange, SFR and Bouygues Télécom” in the area of the sabotage attempts, and covering the 50-minute period between 5.10 a.m. and 6 a.m. On it appeared just two brief calls made by the head of the SDAT team, captain ‘A.L.’, at 5.23 a.m. and 5.25 a.m. When asked to provide the original chart indicating the mobile phone traffic from which the altered copy was taken, the SDAT reported it had been lost.
Questioned in April 2013, and under condition that their identities did not appear in the case file, the 12 DCRI intelligence service officers who officially took part in the surveillance operation declined to give the numbers of the mobile phones they were carrying at the time of the events, citing their right to do so under national defence secrecy laws. Meanwhile, the three officers from the SDAT anti-terrorist team involved in the surveillance, also asked for their mobile phone numbers, said they no longer remembered them and also raised concerns over the safety of their families if they were traced.
The defence lawyers managed to obtain a copy of the evidence in the case file relating to information of the traffic recorded by mobile phone relay masts in the region of the surveillance operation. However, this turned out to be spreadsheets drawn up by an officer of the SDAT, and did not contain the original documents provided by the mobile phone operators – and which appear nowhere in the evidence recorded in the case file. Also missing are the official requests for this information that would have been sent to the operators. Meanwhile, a record of traffic on mobile operator SFR’s masts that night was made immediately after the events, on November 9th 2008, by gendarmerie investigators – but which is also contained on ‘Xcel’ spreadsheets.
According to the information available from the different spreadsheets studied by Mediapart, it appears that some 15 mobile phones used by the surveillance officers activated relay masts between 6.43 p.m. on November 7th and 5.25 a.m. on November 8th. To establish this, Mediapart took the only known mobile phone number that was used by the officers that night – that of SDAT group leader captain ‘A.L.’ and traced its communication with other mobile numbers, and the subsequent intercommunication between them all. However, while the mobile traffic based on the spreadsheet information indicates the presence in the area of the majority of the surveillance team, it allows for only a partial corroboration of the timings given in the surveillance report.
Two mobile phones belonging to the police are the first to be recorded, at 6.43 p.m. and the other four minutes later, picked up by the mobile masts at Montreuil-aux-Lions and Coulombs-en-Valois, about 25 kilometres north east of Trilport. These mobile masts again show a presence of police mobile phones at 8.50 p.m., which corresponds with what was written in the surveillance report that describes, during that early evening, how the officers shadowed Coupat and Lévy as they drove around the area following the high-speed rail track.
But strangely, the mobile phones of several of the surveillance team are recorded by the mast at Montreuil-aux-Lions up until 11.12 p.m. whereas, according to the surveillance report, they were supposed to be at Trilport as of 9.50 p.m. when Coupat and Lévy were recorded as having stopped in Trilport and then observed entering a pizzeria.
Later in their report, as already detailed above, the surveillance team record that they lost track of Coupat’s Mercedes at 4 a.m. (November 8th) as it travelled along a section of a minor cross-country road. That was when, captain ‘A.L.’, head of the SDAT surveillance group, or his deputy lieutenant ‘B.M.’, according to another police statement, deducted that Coupat’s car had left the road to park on a service lane adjoining the railway track, and when either officer ‘A.L.’ or ‘B.M.’ made “a pedestrian approach to the site” and found the vehicle, parked with its lights out alongside the east-west high-speed train line.
In a statement given in March 2010, and which appears to confirm the version of events that it was lieutenant ‘B.M.’ who identified the car parked on the service lane, captain ‘A.L.’ said that “the officer in turn informed his colleagues of the localisation of the vehicle and went quickly back to his vehicle to be aboard in case the vehicle being followed began to leave”. According to the surveillance report, Coupat’s car left the site at 4.20 a.m.Yet Mediapart found no trace of any activity of the police mobile phones on the closest mast to the spot until 4.23 a.m.
Faced with the major incoherence of the account first given by the surveillance team that they had left Trilport, shadowing Coupat’s car, at 3.50 a.m. – which would have meant that the vehicles travelled at an average speed of 160 kph to reach the site of the service lane at 4 a.m. – the officers subsequently explained this to be a typing error, and that the real time they left Trilport was 3.30 a.m. But the mobile phone mast records suggest that they may in fact have arrived at the scene 20 minutes later than indicated in their report.
But it is uncertain that the questions surrounding the veracity of the surveillance report will impress the magistrate in charge of the investigation, Judge Duyé. In her concluding notes she commented that the aim of a report on surveillance operations “was not to describe the precise actions of the surveillance operation but to describe the behaviour of the individual or individuals who are the object of the surveillance”.
-------------------------
- The French version of this article can be found here.
English version, and additional reporting, by Graham Tearse