The death toll from last week’s stampede at the annual Hajj pilgrimage to Mecca has passed 700; on Saturday, Saudi Arabia’s health ministry reported 769 deaths and 934 people injured. Basma Attasi reports for Al Jazeera that the stampede occurred when two waves of pilgrims collided — but that there are conflicting reports about why that happened:

One crowd had just finished a ritual in which pilgrims throw pebbles at three stone columns representing the devil – a rite central to Hajj – when it ran into another wave of people heading to perform the rite.

Sources close to the government said that some Hajj tours – which have an assigned time slot for allowing their groups to go to the Jamarat, where the rite is performed – have not abided by the schedule.

… One witness told Al Jazeera that a vehicle belonging to security forces had left the street it was supposed to stationed in to block two way access. Another said the pilgrims forcefully removed a barrier placed on one side of the road.

The interior ministry blamed pilgrims for not abiding by the rules. An official with the information ministry said it happened due to pilgrim’s “illiteracy” and their inability to understand Arabic or English.

A report published in the UK- based Independent, citing tour operators, said that Saudi security forces had blocked roads ahead of the arrival of dignitaries.

Another Al Jazeera article summarizes some of the previous Hajj stampedes. The worst was in 1990, when 1,426 pilgrims died in an overcrowded pedestrian tunnel; others include a 2004 crush that killed 244 and a 2006 stampede that claimed 360 lives.

In a 2011 New Yorker article on crowd disasters, John Seabrook noted that press accounts of these kinds of events often make incorrect assumptions about participants’ states of mind:

In the literature on crowd disasters, there is a striking incongruity between the way these events are depicted in the press and how they actually occur. In popular accounts, they are almost invariably described as “panics.” The crowd is portrayed as a single, unified entity, which acts according to “mob psychology”—a set of primitive instincts (fear, followed by flight) that favor self-preservation over the welfare of others, and cause “stampedes” and “tramplings.” But most crowd disasters are caused by “crazes”—people are usually moving toward something they want, rather than away from something they fear, and, if you’re caught up in a crush, you’re just as likely to die on your feet as under the feet of others, squashed by the pressure of bodies smashing into you. (Investigators collecting evidence in the aftermath of crowd disasters have found steel guardrails capable of withstanding a thousand pounds of pressure bent by crowd force.) In disasters not involving fire, panic is rarely the cause of fatalities, and even when fire is involved, such as in the 1977 Beverly Hills Supper Club fire, in Southgate, Kentucky, research has shown that people continue to help one another, even at the cost of their own lives.

Seabrook’s article focused on the 2008 stampede at a Walmart store in Long Island that killed 34-year-old worker Jdimytai Damour. OSHA cited Walmart under the General Duty Clause, which requires that employers furnish “employment and a place of employment which are free from recognized hazards that are causing or are likely to cause death or serious physical harm to his employees.” The agency fined the company $7,000 for inadequate crowd management (crowd crushes being a recognized hazard), and the company contested the citation. In a ruling issued on the 100th anniversary of the Triangle Shirtwaist Factory Fire, Chief Administrative Law Judge Covette Rooney upheld the citation. Walmart appealed, but then earlier this year decided to withdraw its appeal and pay the fine.

While officials in Saudi Arabia may be attempting to blame particular groups of pilgrims for the latest deadly stampede, they have also recognized the hazards inherent in such large-scale gatherings and invested heavily in changes to reduce the risk of deadly crowd events. In his WIRED piece “The Hajj Stampede is a Fluid Dynamics Problem,” Adam Rogers reports:

After one in 2006, Saudi authorities instituted single-direction pathways, visitor counts, and theme park-like scheduling of visits. The Jamarat Bridge, location of three pillars that represent the devil, at which pilgrims are supposed to throw stones, was the site of a stampede that killed over 1,000 people; today it’s a multi-level, multi-exit complex designed to keep people moving. In the past decade or so, the Saudi government has worked with a wide variety of architects and designers, including the famed international firm Gensler, to improve flow and safety at all of the hajj’s major sites, from the central mosque to the tent city.

At Vox, Zack Beauchamp notes that Hajj sites weren’t build with todays crowds in mind; this year’s total attendance is estimated at around 2 million. While new developments that change the look and feel of ancient holy sites may be controversial, Beauchamp concludes that “if Saudi Arabia wants to prevent future incidents like this one, it will have to find a solution of some kind, whether that means more infrastructure even at the cost of the sanctity of the ancient sites, or limiting participation even further, or something else.”