James Gillies

How do you get a billion people excited about a giant superconducting proton collider?

Thinking about how to communicate the start-up of the LHC was the focus of everything we were doing since 2004. We saw a huge opportunity to put CERN and particle physics on the map. So we started saying to journalists: “Wanna see this? Well you’d better come soon because it’s going to be closed up — and it’s impressive.” Until 2004 we’d have about 200 visits per year, but in 2007 we had 600.

It’s now clear that 10 September was the only day the LHC was functioning long enough to get protons around in both directions. Does it unnerve you that so much luck was involved when the world was watching?

We took the decision to show people the reality of doing science at this level, and that carries risk. It could have gone wrong, but on the other hand I know the people who run these machines and they’re just an amazing bunch.

How did you feel when you woke up on 10 September?

Slightly terrified. Plan B was for me to stand up in the Globe [where the media were corralled] and tell 340 journalists who had come to CERN that it ain’t going to happen today. As it was, a minor cryogenic outage added a bit of drama, but that drove home to people that the event was live. We hadn’t rehearsed it, and that openness was appreciated.

And by the time you went to bed?

Well, fantastic. It was just the most amazing day. It was amazing to see how much emotion there was in witnessing this big machine coming to life. The event was huge. Eurovision [the broadcast service which beamed footage to television networks and cost CERN about 50,000 CHF] estimated we had exposure to 1bn people. The result now is that the LHC is mentioned without explanation in contexts that have nothing to do with science.

Rumour was that the switch-on date was arranged around BBC presenter Andrew Marr’s holiday plans?

It’s hilarious. The BBC did ask if we could put the date back if Andrew couldn’t make it, and we said “no”. But on the other hand, BBC Radio 4 pulled out all the stops and decided to do something unprecedented in science by devoting a day to the event, so in return we gave them a room just off the CERN Control Centre to use as a studio. The fact that Radio 4 went so big on CERN drove it out to the rest of the BBC, culminating in “Big Bang Day”, and then out to the rest of the UK media and the world.

Did the black-hole Armageddon frenzy aid or hinder your communication efforts?

Ultimately it helped us by generating interest, but it also worried an awful lot of people and that makes me somewhat angry. People were phoning us up genuinely worried about the end of the world and demanding to know who CERN is accountable to. Of course we’re accountable — 20 countries have to say “yes” before we do anything! We ran a strict press accreditation procedure and there was heightened security on the day.

People were phoning us up genuinely worried about the end of the world and demanding to know who CERN is accountable to.

Did CERN handle the issue well?

With hindsight I would have treated the black hole stuff in exactly the same way we dealt with Angels and Demons [the Dan Brown novel in which antimatter is stolen from CERN to destroy the Vatican]. We were very proactive with that — we put up a webpage and had fun with it — but we didn’t envisage the black hole story going as far as it did. On one hand we didn’t want to engage with the scaremongers, but the particle physics community worldwide was slow to pick up and say “this is nonsense”. It’s one thing for CERN to say everything is safe, but we needed other voices, which have since surfaced.

How did you feel when you realized the full extent of the damage caused by the electrical fault on 19 September?

I was genuinely sad, and I think a lot of people at CERN felt the same. When you’ve been so intimately involved with something for such a long time, and when the start up went so well, the incident was a huge shock. But there’s a story to tell here: mishaps like this are part of life when working at the cutting edge of technology and research.

Wouldn’t it have made more sense to test all the LHC circuits before the media event on the 10th?

I don’t think it would have been any easier to live with what happened. In fact, it may have been less easy because at least now we know that the LHC works extremely well.

So the incident wasn’t the result of pressure to switch on before the machine was ready?

No. The timetable was driven by [LHC project director] Lyn [Evans] and the machine operators. The plan was to get some collision data at low energies, then finish testing the hardware to run at higher energies. Had an electrical transformer not broken down three days after the 10th, we would have had that collision data and the incident would have happened later.

We knew the warm up and cool down would take two months minimum so we quickly put that out in a statement, although with hindsight we should have been more cautious.

CERN’s new director general [Rolf-Dieter Heuer] told staff on 12 January, that from now on people would hear about events first from him, not the press. Was there a lack of communication internally following the incident?

It wasn’t organized in a way that it needs to be now that there is such a huge demand for information. We knew the warm up and cool down would take two months minimum so we quickly put that out in a statement, although with hindsight we should have been more cautious because we soon realized there was no way the LHC was going to be back up that year. Internally, people from the machine and management side were giving talks to the experiments, but we could have used our intranet, website and the CERN Bulletin better.

Did CERN try to withhold information from the media?

There’s a great quote in a Salman Rushdie book: whenever information is tightly controlled, rumour becomes a valued source of news. That was happening at CERN. All the way through, the then director-general [Robert Aymar] genuinely wanted to put out factually accurate information as soon as it was available, but CERN probably tried too hard to keep tight control. Although we were quick off the mark with releasing official statements there were long gaps in between. Even though there wasn’t very much to say, there was stuff that could have been said which would have capped those rumours.

Why was the LHC logbook modified retrospectively on the day of the incident?

There’s nothing sinister about it. The person on shift that morning just wrote down what had happened, then someone came along and said: “Everybody can see that, let’s take it away!” It was a wrong decision made in the heat of the moment, but it was naivety rather than anything systematic. It’s an issue we’re probably going to look at. A logbook should be somewhere people can write down whatever they feel, but that’s not necessarily something that should be visible to the whole world.

[Modifying the log book] was a wrong decision made in the heat of the moment, but it was naivety rather than anything systematic.

Were the long awaited official photos of the damage chosen because they were taken after the tunnel had been cleaned up?

No. They were specifically chosen because they showed where the damage was worse, at the end of the helium-induced pressure wave. I think the rumours had led people to expect something more dramatic.

Who ordered links to photos and some presentations to be password protected after they appeared on blogs?

[Aymar] wanted the CERN community to receive the news from him before it was made more widely available, so access to slides was temporarily restricted. People just hadn’t realized how much in the spotlight we are now.

Is it true that people were being threatened with disciplinary action if they circulated pictures of the tunnel before they had been officially released?

It’s true that the former DG wanted to be the one issuing the information. Look, we have to be more effective in the way that we communicate at CERN both internally and externally. The world is watching this. We’ve created what we’ve created, there is a demand for information and we need to provide it. That’s something that the new management is very aware of.

How are the repairs going?

They’re going well, but there’s a lot to do. We need to increase the LHC’s capacity to vent helium in the event of another leak, so already on all the sectors that are warmed up (half the machine) we are changing the valves on all the quadrupole magnets and putting new ones on the dipoles.

Where’s the logic in making only half of the machine safer?

What we’re doing is about as conservative as you can get. The LHC will now be able to vent ten times as much helium as before, and on top of that we’ve got lots of extra monitoring which will allow us to see a similar electrical fault coming.

CERN has a history of overly optimistic LHC timetables. Isn’t the current schedule of first-beam in July/August rather aggressive?

I’m pretty confident there will be collisions this year. There’s a great determination, but caution is the guiding principle. There will be a meeting in Chamonix in early February after which a realistic schedule will be announced.

Does the world have the appetite for “Big Bang Day II”?

It won’t be as big as 10 September. We’re not going to be inviting anyone back for first-beam this year. Journalists are keen to see first collisions even though it may involve camping out with us for a week. The whole process will be webcast. CERN was unwilling to invest in bandwidth before the 10th so the webcast fell over very early in the morning, but we’ve since had companies offering us bandwidth in exchange for having their logo displayed.

This year CERN won’t be chasing the media or the blogs. We’ll be the primary source of news about [the lab].

What do you think the LHC will find and when?

Well, you should probably ask the people who are working on it.

Come on, you’ve got a PhD in physics!

We’re going to find the Higgs particle if it exists, and I can’t think of any reason why it doesn’t. But that will take a year’s worth of good data. What I would really like to see, although it would be a nightmare from a communications point of view, is a flood of supersymmetric particles as soon as we switch on. Some models say that could happen, and if so we’ll have a group of people saying “wait!” and another group saying “look, this is a signal!”

How are you going to manage information when data arrive and rumours spread?

We’ve got protocols in place for the experiments so that if they really feel they are ready to make an announcement then we move very fast and organize a seminar here very quickly. One thing you’ll see this year is that CERN’s official communications won’t be chasing the media or the blogs. We’ll be the primary source of news about CERN.

Is that realistic, given that bloggers can brain-dump a post in a matter of minutes?

I think we’ve got to try. If there’s someone who’s blogging about a “three sigma” effect [meaning there is less than 1% chance it is a statistical fluke] that’s been verified, then there’s no reason why we wouldn’t talk about it as well. But if someone is blogging about a three sigma effect in their own particular analysis which hasn’t gone through the official verification process in their experiment then we will deny it, which may come to releasing a statement.

Are you planning to implement rules on blogging, as the CDF collaboration at Fermilab has done in response to rumours about Higgs sightings?

Yes. Some of the experiments have them already as a result of what happened there.

Isn’t that an attempt to censor information?

It’s an attempt to stop blogs fuelling rumour. Nobody wants to clamp down on people releasing information about results that have passed through official quality control in an experiment.

What’s the best question you’ve ever been asked about the LHC?

Visitors often ask first whether they can ask a really stupid question, and then come out with something profound and unanswerable that goes straight to the core of what we’re doing here.