The 1990 Strangeways Riot – The Warnings That Never Reached the Governor

On the morning of 1 April 1990, staff at Strangeways quietly kept vulnerable prisoners away from the chapel because they expected violence. Minutes later, they allowed about 300 prisoners to enter that same space without being searched, including men later named as key ringleaders. Brendan O’Friel would later describe the outbreak as an unanticipated ‘explosion of evil’, but the actions taken on the wings before 11.00 am point to prior warning and a breakdown much lower down the chain.

The Accepted History vs. The Narrow Gap

The standard account of Strangeways comes from Lord Justice Woolf’s inquiry, published on 31 January 1991. Woolf did not struggle to find reasons for revolt. The prison system was dangerously overcrowded and lacked any credible grievance process for inmates. Strangeways itself still relied on ‘slopping out’, the practice of prisoners emptying waste buckets by hand because cells had no toilets. Men were held two or three to a cell built for one. That was not a minor complaint. It shaped daily life.

Woolf’s broader conclusion changed the way the British state talked about prisons. The old emphasis on security and control was no longer enough on its own. Justice had to sit alongside it. That meant basic dignity, some sort of fair grievance process and living conditions that did not grind people down every day. On those points, the historical argument is settled.

The broader structural explanation is entirely valid, yet it obscures a distinct operational collapse. Squalid conditions explain why a prison might erupt. But they do not answer how explicit intelligence regarding the upcoming chapel protest was effectively swallowed by the system. The timeline shows local staff taking defensive action on Sunday morning, even as the executive suite remained completely unaware of the threat.

This operational gap requires a closer look. Following the siege, the political establishment preferred to address the vandalism rather than the ignored warnings. Home Secretary Kenneth Baker proposed a new ten-year sentence for prison mutiny. That legislative push fixed public attention on the riot itself. It completely diverted scrutiny away from the breakdown in command on the wings during the final weekend of March 1990.

Command Culture and the Administrative Bottleneck

Strangeways in 1990 sat inside a rigid hierarchy. Governor grades occupied the executive offices. Uniformed officers ran the landings. They unlocked cells, moved prisoners and dealt with the daily pressure face-to-face. On paper, that structure should have helped intelligence travel upwards. In practice, it could trap information halfway.

A Security Information Report, usually shortened to SIR, was the formal way to pass threat intelligence through the system. Any member of staff could use one to record warnings, suspicious conduct or information from informants. Once written, the report should go to the security department and on to the Governing Governor. That was the mechanism designed to turn gossip, rumour or raw warning into an actionable security picture.

The problem is that the warning appears to have stayed verbal. Retrospective accounts say officers received dozens of messages in March 1990 about a protest planned for 1 April in the chapel and that prisoners would be ‘tooled up’, meaning armed with improvised weapons.

The key failure was a catastrophic lack of paperwork rather than missing intelligence. No surviving SIR has surfaced that carried those warnings to O’Friel before the riot. That fits the prison culture described by later observers. Strangeways was said to be a ‘screws nick’, a place where uniformed officers held the real operational grip. If that culture was as strong as described, officers may have chosen to handle the threat locally instead of feeding it into the formal chain. That is not proof of sabotage. It points more towards data siloing and systems failure, where information exists but never reaches the person with the authority to act on it.

Testing the Pre-Riot Intelligence

The warnings described in later accounts were unusually specific. They were not vague claims that the prison felt tense. They allegedly named the date, the location and some of the men expected to lead the protest. Paul Taylor, a remand prisoner, was one of the names later associated with the takeover of the chapel. If officers had that level of detail before Sunday morning, the failure sits squarely in reporting and response, not in intelligence gathering.

The most striking prelude came on Saturday, 31 March 1990. During a film screening in the chapel, a reported ‘dummy run’ took place. According to later witness accounts, Paul Taylor made an inflammatory speech in front of other prisoners. That should have set off alarms. A rehearsal in the same room used the next day for the chapel service is not background noise.

Yet O’Friel later said his Saturday had been entirely ‘ordinary’. That detail matters because it marks the split between the wings and the executive office. If the dummy run happened as later described, the Governor did not hear about it in time.

No contemporaneous incident report reproduced in the available material closes that gap. The trail ends with retrospective testimony. That leaves a sharp contradiction. Either the rehearsal was seen and not escalated, or it has been overstated after the fact. The available material leans heavily towards the first explanation, but the absence of a surviving report means the point remains partly dependent on later witness recollection rather than a dated document from 31 March itself.

The 1 April Threat Assessment Paradox

On the morning of Sunday, 1 April 1990, staff took one precaution that gives the game away. They kept segregated and vulnerable prisoners away from the chapel service. Officers Duffield and Collins are named in the available material as carrying out that local evacuation. Staff do not remove vulnerable men from a routine religious service unless they think violence may break out.

Here is where the operational logic completely collapses. Just minutes after moving those vulnerable men out of harm’s way, the same staff allowed roughly 300 standard inmates to walk straight into the main service. Nobody was patted down for weapons. That entering crowd included the specific ringleaders named in the earlier intelligence. When questioned later, officers justified the lapse by claiming a full search was logistically impractical, adding that they did not want to infringe on the inmates’ religious freedoms.

Fair enough, except the earlier decision had already interfered with the normal running of the service because staff feared violence. The prison was not behaving as though it expected a peaceful morning.

The resulting contradiction is hard to ignore. The intelligence was credible enough to justify protecting specific targets, yet somehow not serious enough to warrant searching the main crowd. This looks like a classic local containment mindset. The uniformed staff on duty seemingly believed they could manage the danger internally without triggering a full command response. They grossly miscalculated the scale of the threat. Once the chapel filled up, their narrow defensive tactic collapsed almost immediately.

The Missing Hour

The riot began during the 11.00 am service on 1 April. Chaplain Noel Proctor was conducting the sermon when Paul Taylor seized the microphone. According to the material gathered later, he preached resistance to the system, and another prisoner shouted, ‘Fuck your system, fuck your rules.’ Weapons were then produced and officers lost control of the chapel.

O’Friel did not arrive until 11.55 am. By then, nearly an hour had passed since the takeover began. On arrival, he ordered a Governor 5, a governor-grade duty officer, to secure the remand prison gates. He later stated that the remand prison could have been held with an earlier order. That is not a minor remark. It is effectively an admission that the critical window for containment had already been wasted before he got there.

The missing piece is the control room between 11.00 am and 11.55 am. During this massive window, the control room operated without the executive leader. We completely lack the unredacted radio logs from the control room for that period, leaving us in the dark about what specific decisions were made by the unnamed Duty Officer during those critical 55 minutes. Who was giving orders? Who decided when to withdraw? Who chose not to secure the remand area sooner?

Without them, the most important command decisions of the day sit in a blank space. That absence matters because it prevents a full assignment of responsibility. The prison was not lost at 11.00 am by magic. It was lost through a sequence of choices made in that first hour, and the surviving public account does not give a minute-by-minute chain for those choices. No record exists here after O’Friel’s late arrival that explains the earlier commands in full.

The Media Vacuum and Fabricated Atrocities

As the siege dragged on, the press gathered outside the prison walls and filled the air with rumour. Some reports claimed prisoners had been castrated. Others said bodies had been cut up and flushed through the sewers. Those sensational stories were entirely false.

Woolf dealt with them directly. He found that the atrocity stories had no basis in fact. Violence did occur in the early phase of the riot, especially against sex offenders, but the large-scale torture stories presented to the public did not happen. Independent assessments, including the 1991 CPT Report to the United Kingdom Government, documented the actual physical aftermath and separated the real violence from the media fabrications.

That distinction matters. A prison riot does not need invented horrors to be serious.

The press failure also distorted the historical picture. Once sensational claims harden in public memory, later corrections rarely catch up. In this case, the media vacuum outside Strangeways became a machine for speculation. It turned uncertainty into an apparent fact and pushed the real operational questions further into the background.

That secondary fog still hangs over the event. Many people remember Strangeways as a scene of medieval brutality because that was the version shouted loudest at the time. Woolf’s findings stripped out those fabrications, but the cleaner account never travelled with the same force.

The Silo Effect Theory

The strongest explanation in the available material points to the silo effect rather than deliberate inaction by staff. Information gathered on the landings stayed on the landings. Officers appear to have trusted their own local handling of the threat more than the paperwork route designed to alert security staff and the Governor. In a culture where uniformed staff saw themselves as the people who actually ran the jail, that makes grim sense.

There is no proof here that anyone in management wanted the prison to fall. No internal memo has surfaced saying the riot should be allowed to happen. No whistle-blower account in the available material describes a plan to force reform by letting matters explode. That theory has surface appeal because staff had every reason to be angry about overcrowding and conditions, but it breaks down on motive. The officers on duty were the people most likely to be assaulted once control failed. One prison officer would later die in the wider wave of prison unrest that followed.

The available facts point to a chain of reporting failures and misjudgements inside the prison. Officers heard warnings. The middle ranks did not turn them into formal SIRs. A reported rehearsal in the chapel was not carried upward through the reporting chain. Vulnerable prisoners were protected, but the main gathering went ahead. The Governor arrived late and said the remand wing might have been saved with an earlier order.

Each part points towards a chain of misjudgements inside a divided institution. That does not let the system off the hook. Quite the opposite. A prison built on overcrowding, sloping out, and poor legitimacy was already unstable. Add a command culture where warnings can stop at the senior officer level, and you have a structure ready to fail at the worst possible moment. Woolf addressed the wider rot. The narrower collapse on 31 March and 1 April still looks like a case of systems failure, data siloing and evidence mishandling.

Sources

Sources include: the official 1991 report of the public inquiry into prison disturbances (The Woolf Report, Cm 1456); the Prison Reform Trust’s summary of findings regarding the 1990 riots; historical news archives and investigative reporting from The Guardian (1990–2010); published witness testimony and retrospective accounts from former Governing Governor Brendan O’Friel; academic and operational analyses including ‘Strangeways 1990: A serious disturbance’ published by the Centre for Crime and Justice Studies; UK Parliamentary records from Hansard (April 1990) regarding the initial Statement on Strangeways Prison; the final report of the Billy Wright Inquiry regarding historical prison command structures; operational logs and internal intelligence grading data from HM Prison Service (1990); and investigative retrospectives detailing the ‘dummy run’ and the timeline of the chapel takeover.