Talk:2021 Salisbury rail crash

Tunnel location

The tunnel is not east of the city is north north east - or north east for simplicity. Look at this map. You can see the eastern portal just to the right of 3rd Salisbury Sea Scouts and the western portal just to the left of Wordsworth Road. Emergency services are accessing the scene from London road. 10mmsocket (talk) 23:03, 31 October 2021 (UTC)

Headcode

There's a ref pointing to a 1L57 headcode, but most reports I've seen, and the infobox, specify 1L53. (Oh. And 1F32 vs 1F30, too). Not sure what's up with that. --Tagishsimon (talk) 01:52, 1 November 2021 (UTC)

Fixed, thanks. MIDI (talk) 08:40, 1 November 2021 (UTC)

Images

Trains

One of the Realtime Trains refs used suggests that 159102 was at least scheduled to be one of the trains involved. Can someone confirm that's the case, so we can use an image such as File:Westbury_railway_station_MMB_38_159102.jpg? MIDI (talk) 09:16, 1 November 2021 (UTC)

Unit identities are known, but not from useable sources yet. Once they can be identified via RSs, we can add images of both to the infobox, as per the Milavče train crash article. Mjroots (talk) 09:28, 1 November 2021 (UTC)

Units involved are 158 762 and 158 763 (1F30) and 159 102 (1L53). Mjroots (talk) 09:32, 1 November 2021 (UTC)

Quality images - File:Westbury DMU Sidings - GWR 158762 (70807).JPG, File:Bristol Temple Meads - fGWR 158763 arriving from Taunton.JPG and File:Westbury_railway_station_MMB_38_159102.jpg as suggested by MIDI above.Mjroots (talk) 09:41, 1 November 2021 (UTC)

159 102 confirmed by BBC, image added to infobox. Mjroots (talk) 09:52, 1 November 2021 (UTC)

From Newborn baby 'rescued from Salisbury tunnel train crash by hero firefighter' it looks like 158763 was the other. MIDI (talk) 10:36, 1 November 2021 (UTC)

I've swapped the images as the 158763 one has current livery and the photo is from last year. MIDI (talk) 10:38, 1 November 2021 (UTC)

Can we please find a better source than the Daily Mirror? All redtops fail WP:RS, even for what the day is today. Mjroots (talk) 11:04, 1 November 2021 (UTC)

Location

I think Geograph:: Salisbury Tunnel Junction © Roy Hughes cc-by-sa/2.0 might be the right location, albeit looking in the opposite direction of the tunnel... MIDI (talk) 09:20, 1 November 2021 (UTC)

Geograph:: Grateley or West Dean? © Neil Owen cc-by-sa/2.0 gives an alternative POV but still looking away from the tunnel. MIDI (talk) 09:21, 1 November 2021 (UTC)

The GB News story (currently ref [8]) has an excellent photo showing the crash. Junction is immediately before entering tunnel. Mjroots (talk) 09:28, 1 November 2021 (UTC)

A look at the pictures (from the Telegraph article) immediately confirms the location (the "Up"/North-eastern end of the tunnel), and looking at google and at OTT diagrams gives a pretty good idea of the location (159102 is standing upright on the Down line [towards Andover], straddling the diamond crossing across Salisbury Tunnel Jn., with the derailed carriage behind it hiding the branch towards Dean). Of course, I have absolutely no clue what lead to this (beyond finding that "knocked out the signalling" seems like an imprecise layman's terms explanation: surely the signalling is fail-safe and drivers have route knowledge so they should have noticed the lack of a signal when entering/approaching the tunnel - now, of course, all WP:OR from me: the logical hypothetical, based on But in an update on Monday afternoon, British Transport Police (BTP) confirmed this was not the case, and that both trains were travelling in the same direction when one train struck the side of the other, causing it to derail whilst in the tunnel., is that one of the trains was not supposed to be there). In any case, it might be more precise to say the incident happened at the junction, or at least at the north-eastern end of the tunnel. Cheers, RandomCanadian (talk / contribs) 16:16, 1 November 2021 (UTC)

Forum chat is leaning towards one of the trains being unable to stop at a signal due to poor adhesion conditions, leading to both trains being at the junction at the same time. Of course this is all WP:CRYSTAL at the moment and not for publication unless and until a RS reports it. Expect things will be a lot clearer tomorrow. Mjroots (talk) 18:25, 1 November 2021 (UTC)

It seems I didn't make my point clear at the top of this section, and that's my fault – I was looking for an image to illustrate the general location, rather than discuss the precise location of the incident (looks like no more than 50 metres (160 ft) between junction and tunnel portal). This is, however, largely irrelevant now given we've got photos of the trains involved. As you were... MIDI (talk) 18:49, 1 November 2021 (UTC)

I've asked at the Wilts WP if anyone can get a photo of the tunnel portal or the crashed trains, given there is a road nearby to take said photo from. Mjroots (talk) 21:34, 1 November 2021 (UTC)

Casualties

Salisbury District Hospital has tweeted that thirteen people were taken to hospital. Four were admitted and one has been discharged. Mjroots (talk) 13:36, 1 November 2021 (UTC)

First train did not hit an object

Accordin to the Daily Telegraph (subscription required, which I don't have, but have seen quoted on a forum), there was no collision with an object in the tunnel. This was a rear-end collision in the tunnel. Mjroots (talk) 14:27, 1 November 2021 (UTC)

Confirmed by the BTP. Mjroots (talk) 14:46, 1 November 2021 (UTC)

Yup, DT says BTP said: "... both trains were travelling in the same direction when one train struck the side of the other, causing it to derail whilst in the tunnel". -- DeFacto (talk). 14:51, 1 November 2021 (UTC)

First para of the "Incident" is now borked. Sentence 1 says train B ran into train A and there was a derailment. Sentence 2 says, then train A was hit by train B. --Tagishsimon (talk) 14:53, 1 November 2021 (UTC)

Sorted. --Tagishsimon (talk) 15:03, 1 November 2021 (UTC)

I've removed the {{fv}} tag from ref [4]. Archive seems to be playing up, working sometimes and sometimes not. However, the live link contains this "The crash happened at 18:46 GMT when one train hit an object in a tunnel, and the second train then collided with it due to signalling problems", which covers what is being stated. Mjroots (talk) 20:13, 1 November 2021 (UTC)

BBC also very clearly has: The trains were travelling in the same direction on different tracks but collided at a Y-shaped junction approaching the tunnel, with one hitting the side of the other causing it to derail, BTP said.
A carriage was initially thought to have derailed after hitting something, and the second train then crashed into it, but Supt Lisa Garrett, of BTP, told a press conference "there was nothing to suggest" the train had struck an object. ([1]) RandomCanadian (talk / contribs) 20:19, 1 November 2021 (UTC)

It's good though, as it is stated that initially it was claimed that the train had hit an object and derailed, followed by BTP's rebuttal of said claim. Mjroots (talk) 20:40, 1 November 2021 (UTC)

Braithwaite

I added this: Graham Braithwaite, a professor of safety and accident investigation at Cranfield University, said that usually the "fail-safe" system in place across the network means that if signalling goes out - trains are told not to pass. Braithwaite told Sky News, "It's an unusual accident, looking at it at face value."" But it was reverted with the edit summary "This is not an article on failsafe systems. Whilst the Prof is doubtless right, it is an inappropriate addition here, not least whilst there is no information on whether failsafe is an issue in this crash." I'm well aware that "This is not an article on failsafe systems", thanks. But I would have thought that an acknowledged expert such as Braithwaite might know what he was talking about. That SkyNews article goes into quite a lot of detail. If there has to be some kind of moratorium on comment here until we have an official interim accident report, so be it. But many RS sources are reporting there was a signalling failure. Martinevans123 (talk) 19:28, 1 November 2021 (UTC)

Wait for now, this is not a straightforward accident as has already been said. We'll know a lot more in the next 24-48 hours. Mjroots (talk) 19:38, 1 November 2021 (UTC)

I agree with the removal by user:Tagishsimon. There is no need to include what is, as of now, speculation, regardless of who said it. Meters (talk) 19:42, 1 November 2021 (UTC)

I think most sources agree that the signalling failed. But they can't yet agree why. Martinevans123 (talk) 19:44, 1 November 2021 (UTC)

@Martinevans123: I don't even think that is currently mentioned in the article, and for good reasons: it is far too early to know (WP:NOTNEWS also applies - doubly so when early news reports were contradicted by later ones). Also, and caveat lector (I am no expert), I already mentioned the fail-safe aspect in my previous comment (look for it in to #Location subsection higher up) [an additional thing I didn't mention is that at night, unlit signals are of course much less conspicuous than by daylight, but again the driver would have expected there to be one...). When we have further confirmation from the RAIB or other trusted sources, this kind of information can probably be included, but until then it remains pure speculation, whether it comes from newspapers or elsewhere. RandomCanadian (talk / contribs) 19:56, 1 November 2021 (UTC)

I doubt I'm up to speed, but we seem to have had scenarios painted of object, derailment, later collision; no object, no prior derailment, just a collision; and, according to Martinevans123, general agreement that the signalling failed (rather than, for instance, SPAD perhaps arising from lack of adhesion). Focussing on one theory in the absence of any informed RS specifying that it was implicated, is speculation and unwise. --Tagishsimon (talk) 20:04, 1 November 2021 (UTC)

@Tagishsimon: The BTP ([2]) seems to say that "We are keeping an open mind but at this early stage there has been nothing to suggest the train struck an object or that there was any significant delay between the trains colliding and then one derailing.", which appears to rule out the first two scenarios; in fact even implies that the derailment was caused by the collision (which is not that silly if you think of it: trains don't usually derail out of their own free will...). They do not mention a signal failure. RandomCanadian (talk / contribs) 20:10, 1 November 2021 (UTC)

Sure, but that tells us nothing on the subject of whether or not it has been established that there was a lack of failsafe, which is what this thread is about. --Tagishsimon (talk) 20:23, 1 November 2021 (UTC)

@Tagishsimon: Fail-safe signalling practice is standard across the British railway network (and pretty much worldwide). In the absence of evidence that there was a definitive failure of this (i.e. a "wrong-side failure", extremely rare but not impossible), it is safe to assume it didn't occur, and it is unwise to speculate about it in any case. Cheers, RandomCanadian (talk / contribs) 20:28, 1 November 2021 (UTC)

Oh ffs, RandomCanadian. Thank you for your admonition to the person that removed the speculation & has argued throughout this thread against speculation, that it is unwise to speculate about it. I stand corrected. --Tagishsimon (talk) 20:51, 1 November 2021 (UTC)

Yeah ffs, RandomCanadian. Anyone would think this was an encyclopaedia. (I'll get my coat, and leave it to the experts...or not) (and here's some more speculation to not mention: "A source at Network Rail said the working assumption was a “multiple” failure in signalling allowing both trains into the tunnel at the same time.") Martinevans123 (talk) 22:11, 1 November 2021 (UTC)

Rail editor Nigel Harris has posted an interesting tweet. Said elsewhere he is hopeful that the RAIB will issue a statement this afternoon. Mjroots (talk) 13:01, 2 November 2021 (UTC)

@Mjroots: There's also this which might be interesting at some point. RandomCanadian (talk / contribs) 13:22, 2 November 2021 (UTC)

References

ran into its rear

I see we say of train A, that train B "ran into its rear". That does not seem to be the case; the rear of train A seems mostly unaffected. Train B seems to have had a side-impact with one of the leading coaches of train A, after which the two trains started to zip themselves together. --Tagishsimon (talk) 23:05, 1 November 2021 (UTC)

Sidelong collision now confirmed by BBC. Which raises the question "which of the two trains SPADded? Mjroots (talk) 08:26, 2 November 2021 (UTC)

Too early to know, one could think that the train that was not irreparably late (1L53, only 6 mins behind schedule) was the one given priority, and hence that the other was the one which wasn't supposed to be there. Or it is possible that the other one was given priority, to minimise delay to a succeeding service using the same stock (though that likely wouldn't be an issue on a Sunday evening, would it?). Again, all speculation. When we have actual information, a more detailed diagram of the accident location could be useful: this picture seems to be a decent basis, as there does not appear to have been a change in this area when Salisbury signal box closed in 2016. RandomCanadian (talk / contribs) 13:48, 2 November 2021 (UTC)

It was the SWR train that ran the red. Mjroots (talk) 16:32, 2 November 2021 (UTC)

The Metro is not a reliable source for that, as per discussions linked at WP:METRO. Joseph2302 (talk) 16:38, 2 November 2021 (UTC)

@Joseph2302: changed to BBC now that they've caught up. Mjroots (talk) 16:47, 2 November 2021 (UTC)

Talking of unreliable sources, can we lose that Mirror source please? Mjroots (talk) 16:48, 2 November 2021 (UTC)

I don't know if somebody has a different angle picture. The Telegraph (and every other news outlet) has one (the same) which clearly shows 159102, but the other is obstructed. This confirms it is a class 158, but does not give exact unit number. RandomCanadian (talk / contribs) 16:55, 2 November 2021 (UTC)

The Mirror at https://www.mirror.co.uk/news/uk-news/newborn-baby-rescued-salisbury-tunnel-25343829 has a photo which clearly identifies 158763. Railcam data shows 158762 & 158763, but that is user-supplied data so not always 100% reliable. I would have expected it to have been corrected by now if wrong, but I think that 763 is the one that can be reliably sourced. --David Biddulph (talk) 17:06, 2 November 2021 (UTC)

Even if tabloid Daily Mirror not generally reliable, they can't really get a picture wrong can they: [3]? It looks like the first digit "1" is not visible? But it can't be anything else for this DMU, can it? Martinevans123 (talk) 17:26, 2 November 2021 (UTC)

OK, we'll live with it for now, but as soon as I find a better source, it'll be replaced. Mjroots (talk) 17:30, 2 November 2021 (UTC)

The RSP entry reads "The Daily Mirror is a tabloid newspaper that publishes tabloid journalism. There is no consensus on whether its reliability is comparable to that of British tabloids such as the Daily Mail and The Sun." Sure, it's not an ideal one, but it is reliable for the simple information which it is used to support (it's certainly far from being that bad that it is blacklissted, so that leaves room for case-by-case allowances like here). RandomCanadian (talk / contribs) 17:49, 2 November 2021 (UTC)

I agree. The only way to make that photo "more reliable" would be to show the "1". It's not an unsourced opinion, It's not a factual mistake. Martinevans123 (talk) 17:53, 2 November 2021 (UTC)

The RAIB report talks about the "rear two carriages" of the GWR train being derailed, this confirms that the train was composed of not just the 2-car unit 158763; it therefore seems very likely that Railcam data was correct in referring also to 158762. --David Biddulph (talk) 12:56, 3 November 2021 (UTC)

Lines closed

The article says: "Railway lines through the tunnel are expected to remain closed until 8 November at the earliest." The supporting BBC source says: "The crash has caused major disruption, with lines through the city expected to remain closed until at least the end of Monday." So we are all agreed these two sentences are synonymous? Thanks. Martinevans123 (talk) 14:38, 2 November 2021 (UTC)

I think that all boils down to whether the "tunnel" and the "city" can be considered to be one and the same thing. -- DeFacto (talk). 14:46, 2 November 2021 (UTC)

That sounded a slight stretch to me. Especially if a reader has no detailed knowledge of all the rail lines in and out of Salisbury. Or does the infobox graphic explain it all perfectly? Martinevans123 (talk) 14:50, 2 November 2021 (UTC)

The BBC wording is somewhat woolly. Services from Salisbury to the West are already running, but it isn't unreasonable to expect the tunnel to be out of action at least until the date quoted. The recovery and restoration tasks will not be easy. --David Biddulph (talk) 14:57, 2 November 2021 (UTC)

It's clear that the correct information (and one which can be easily verified) is that the lines through the tunnel (and the adjoining junction) are closed (pending removal of the involved units, and probably some repairs to the tracks (which look displaced/warped in some pictures), the trackbed and the pointwork), but not the station itself. RandomCanadian (talk / contribs) 15:08, 2 November 2021 (UTC)

Sure. Just too bad the source doesn't say that. Which was why I tagged it. Martinevans123 (talk) 15:09, 2 November 2021 (UTC)

Martinevans123 Network Rail Wessex has posted an update on Twitter a short while ago: [4] The railway remains closed between Salisbury and Basingstoke/Romsey (implying it is open on the other side) and It is likely that some of the more-damaged train carriages will need to be pulled from the tunnel and lifted away by cranes. We will then have to replace any damaged track, points and signalling [...]... This is also consistent with SWR's post indicating there is single-line working [N.B. this is basically "services as normal", as it is indeed a single-track line; but here of course they are using this to explain the impact of this on pre-existing delays caused by other conditions] between Exeter and Salisbury (implying of course that trains are indeed getting there). RandomCanadian (talk / contribs) 15:29, 2 November 2021 (UTC)

That looks a much better source. Next Monday looks a bit ambitious to me. At least it's not leaves on the line this time. Martinevans123 (talk) 15:36, 2 November 2021 (UTC) p.s. now confused again, alas... "as it is indeed a single-track line". What, just through the tunnel?

Martinevans123 Exeter to Salisbury is on the western side of the station, completely different direction from the tunnel (see this helpful diagram from a simulation software: the accident happened at the eastern [right] end of the tunnel; Exeter is on the other side, further down the line on the same side of the station as Wilton Jn, after Gillingham [on the left/south-western side of the station]). I'm understanding the tweet to mean that services which use this line (which does not require going anywhere near the tunnel) are running as normal, albeit with delays due to other conditions. RandomCanadian (talk / contribs) 15:43, 2 November 2021 (UTC)

I see, thanks. That must be Gillingham, not Gillingham? Martinevans123 (talk) 15:49, 2 November 2021 (UTC)

Yeah, the obvious typo is not there on the original panel. Anyway, so long everybody understands. Maybe the diagram provided in the article (Template:2021 Salisbury rail crash RDT) could be clarified - although it already labels in which direction Exeter is. RandomCanadian (talk / contribs) 15:53, 2 November 2021 (UTC)

Or maybe we can find a source that says "Railway lines through the tunnel are expected to remain closed until 8 November at the earliest." Martinevans123 (talk) 16:28, 2 November 2021 (UTC)

We've already got one in the article. Unless I've misread the calendar. "until Monday" = "Until 8 November" (which is more encyclopedic). The way the date keeps getting pushed back every few days was why I added the "at the earliest" bit to the sentence. Mjroots (talk) 16:39, 2 November 2021 (UTC)

The source says "lines through the city". It's not "wrong", just a bit vague. Obviously it's the trains in the tunnel which have caused some of the lines through the city to be closed. Well, it's obvious to us, anyway. Martinevans123 (talk) 16:55, 2 November 2021 (UTC)

Categories

I added the Category:Railway accidents involving a disregarded signal to the article, which DeFacto removed. The fact is, that is how the category is named. The signal was passed at red, i.e. it's instruction was disregarded. Whether or not this was intentional is immaterial. The category should be restored to the article. Mjroots (talk) 17:29, 2 November 2021 (UTC)

The text of that category says "This category is for railway accidents, or significant incidents, where a train driver/engineer disobeyed a signal at danger despite reading it correctly", so it is clearly not applicable in this case. --David Biddulph (talk) 17:32, 2 November 2021 (UTC)

(e/c) We all assume it wasn't disregarded. Poor adhesion was to blame. Oddly worded Cat? Martinevans123 (talk) 17:33, 2 November 2021 (UTC)

Assuming the driver saw that it was red, and understood what red meant, then yes, he disregarded the signal, so the category is correct. This one seems to be more Stonegate than Wootton Bassett though. Mjroots (talk) 17:42, 2 November 2021 (UTC)

"Disregarded" still implies that this was intentional, as the category description does, which does not seem to be accurate here. RandomCanadian (talk / contribs) 17:44, 2 November 2021 (UTC)

@Mjroots: You presumably haven't read the RAIB statement: "Initial evidence indicates that the South Western train driver applied the brakes as it approached the junction and the red signal, but the train was unable to stop before passing the signal. This evidence suggests that the most likely cause of this was wheelslide, almost certainly a result of low adhesion between the wheels and the track." He didn't disregard the signal; he applied the brakes. --David Biddulph (talk) 17:47, 2 November 2021 (UTC)

Of course I've read the statement. Which is why I've been careful to separate intent and event. Mjroots (talk) 17:53, 2 November 2021 (UTC)

Agree that the category should probably be renamed, although what that name should be remains unclear: "Railway accidents involving a signal passed a danger" could work (disregarding issues about English national variants in terminology). "passed without authority" could work but is jargon, ... RandomCanadian (talk / contribs) 17:44, 2 November 2021 (UTC)

I've nominated the category for renaming. Discussion at Wikipedia:Categories for discussion/Log/2021 November 2#Category:Railway accidents involving a disregarded signal. Mjroots (talk) 18:30, 2 November 2021 (UTC)

I removed it because "disregarded" means "ignored", and the only participant in the process capable of ignoring the signal was the driver, but the article does not support that notion. And indeed the cited source says the brakes were applied, so the signal was clearly not ignored. -- DeFacto (talk). 19:33, 2 November 2021 (UTC)

As the result of the discussion was to rename the category, I've added the renamed category to the article. I trust this is satisfactory. Mjroots (talk) 06:12, 11 December 2021 (UTC)

Diagram

Is the diagram in the RAIB investigation report here protected by copyright? The page says: "All content is available under the Open Government Licence v3.0, except where otherwise stated." Thanks. Martinevans123 (talk) 11:12, 3 November 2021 (UTC)

Does Commons:Template:OGL3 help? I'm aware of OGL but it's not something I've come across in the context of Commons... MIDI (talk) 12:10, 3 November 2021 (UTC)

Thanks. There's also Template:OGL-3.0, which suggests it's ok to upload. If so, there must be some other examples around. Martinevans123 (talk) 12:13, 3 November 2021 (UTC)

Here is some more information, looks fine to me: meta:Open Government Licence --PhiH (talk) 12:25, 3 November 2021 (UTC)

Do we need it though? I've expanded the RDT to add new info from the RAIB news item. Mjroots (talk) 13:08, 3 November 2021 (UTC)

Maybe not. I just thought it was very clear and offered a level detail that's not possible in the RDT. No strong view. Martinevans123 (talk) 13:19, 3 November 2021 (UTC)

For clarity it might be helpful. The RDT diagram is already a bit crowded to my opinion. Alternatively, something like the diagram at 2015 Wootton Bassett rail incident (with the exact layout and the path of each train) might be equally useful. RandomCanadian (talk / contribs) 18:08, 3 November 2021 (UTC)

The diagram will get very complicated if we go down that route. All lines are double track, except the Laverstock Loop. Salisbury station has several roads, bay platforms, and a postal depot. Then there's the disused station and the disused branch to add in. It's fine as it is IMvHO. Mjroots (talk) 20:09, 3 November 2021 (UTC)

@Mjroots: If we go that route, we don't need to go as far as either the disused station on one side or Salisbury station on the other: up as far as slightly beyond Laverstock loop would be sufficient. I'll work on a draft one in my userspace and see if it gets too complex too quickly. Cheers, RandomCanadian (talk / contribs) 20:29, 3 November 2021 (UTC)

Well here's the RAIB one anyway. I thought it was a useful, ready-made, freely available resource. Martinevans123 (talk) 20:14, 3 November 2021 (UTC)

Staff

Do we have numbers for staff on the two trains involved? Obviously two drivers, but what about guards? Forum chat is that Mark Hopwood of Network Rail has briefed that there were three staff. Not clear whether this is referring to both trains, or one of the trains with one of the three being "on the cushions" (i.e. on duty but travelling as a passenger to Salisbury, where he was to start his next driving turn). Mjroots (talk) 12:40, 5 November 2021 (UTC)

Reporting

Please note that the Mail Online is not being used as a reference. The reference for that section is the report in Rail. Mjroots (talk) 08:25, 17 November 2021 (UTC)

@RandomCanadian: Your edits turning the "notes" into "references" mean that we now have an unreliable source used as a reference. Which is exactly what I was trying to avoid. Please return that section to its original format. The Rail editorial is the reference, notes are there to illustrate what is being reported. Mjroots (talk) 18:47, 17 November 2021 (UTC)

@Mjroots: These are being used as references, are they not? Giving the primary source (to "illustrate what is being referenced") is still giving a reference. FWIW, there are other outlets beyond the Daily Fail which reported this exact story. Ex. Telegraph (citing anonymous "sources", which at least don't appear to be making the same kind of ridiculous claims as the DM "automatic obstruction warning", really? Without even bothering to explain what this is referring to or how not every possible obstruction will be detected by the track circuit... - especially if there's already a train on it? If anything, citing the Fail here is just a good way to show how unreliable they are). Apparently, much of this erroneous reporting was probably based on the Daily Fail story - I assume that would be why your source is reporting it, and it would also be an acceptable link to a primary source being discussed by a secondary source - if you really want to stick the point home, just stick in a few more "erroneously" or the like in there.

If you don't want to give a link to the Daily Fail, then it's also possible to keep the content [which is from the secondary source, anyway] but remove the link to the primary source, if that's what you prefer. RandomCanadian (talk / contribs) 02:23, 18 November 2021 (UTC)

Article is about a rail crash, not how the press incorrectly reported on it. Like many evolving news stories, there was incorrect reporting in the initial aftermath, is fairly common for these type of events. That the Daily Mail got it wrong is not notable, given that it is considered unreliable to be banned from use as a Wikpedia source per WP:DAILYMAIL. In this case some garbage information was given to the press, and surprise surprise articles with garbage followed. For that reason I have deleted the section. Ghemptoude (talk) 02:35, 18 November 2021 (UTC)

Ghemptoude How the press incorrectly reported it is an interesting aspect (as judged not by my or your opinion, but by the fact there was coverage of exactly this in an independent secondary source - which speaks directly against your assertion that this is "not notable"). Whether the Daily Fail usually gets it wrong doesn't mean that other more reliable sources reporting on the Daily Fail and others being wrong are not worthy of interest. RandomCanadian (talk / contribs) 02:56, 18 November 2021 (UTC)

@RandomCanadian: - This is how the article was immediately before you editited the section in question. The two newspapers and the NR and NR/TOC statements are not being used as references. The reference is Harris's editorial. That section would also work if I were to remove all four notes. I included them as illustratory, supplemental material. Per WP:DAILYMAIL, we cannot use the Daily Mail as a source. There is no ban on it being used as a note. I'd really like that section to be returned to its original form. Rail 944 has further criticism of the reporting on pages 44-45 – a NR director appeared on Radio 4's Today programme and failed to correct information that he knew to be incorrect. Mjroots (talk) 06:18, 18 November 2021 (UTC)

@Ghemptoude: - the point is not that initial reporting was vastly wide of the mark. It is the reason why such reporting was vastly wide of the mark. As with any major accident, be it a train crash, plane crash or shipwreck, initial reporting often contains inaccuracies, which are corrected as newer information comes to light. It is rare that official information is leaked in the way that it was here. Harris's criticism is not aimed at the media, who ran with what they were given, but at NR and how it handled the leak. Mjroots (talk) 06:24, 18 November 2021 (UTC)

Mjroots The difference between "Daily Fail in {{efn}}" and "Daily Fail between ref tags" seems purely academic to me. As for the rest, quoting the relevant entry at RSP, "As a result, the Daily Mail should not be used for determining notability, nor should it be used as a source in articles. The Daily Mail may be used in rare cases in an about-self fashion. [...] The restriction is often incorrectly interpreted as a "ban" on the Daily Mail." Using the Daily Fail as a primary source link to support "Daily Fail said X" (not as a source to support "X"), when this information is also reported in a secondary source, seems like one of these cases where not interpreting every rule as absolute is more in line with writing a better article - although, of course, if the reason for the inaccuracy is not the Daily Fail's fabrications being picked up, but an inaccurate early official source, then it would make more sense to discuss that instead of discussing the Daily Fail and other [reliable or not] sources. RandomCanadian (talk / contribs) 13:47, 18 November 2021 (UTC)

Is the dispute at this time about whether to mention the initial wrong reporting of the story? That's reliably sourced, albeit entirely to Rail. Has this aspect of the story been picked up beyond the enthusiast/trade press? Mackensen (talk) 11:19, 18 November 2021 (UTC)

@Mackensen: - no, the dispute is the removal of notes and converting them to references. One of the notes was a Mail Online article (since removed), which highlighted really stupid remarks made by a Nework Rail manager. In the form I originally created the section, the entire paragraph was referenced to Harris's editorial, with the four notes expanding on the criticism for those readers interested. Had the four notes been removed, the section would still have been valid. Mjroots (talk) 13:22, 18 November 2021 (UTC)

My suggestion would be to trim the "reporting of the story" section heavily and just go with something like "initial reports suggested that the GWR train derailed after hitting an object but were found to be inaccurate". Inaccurate reporting in the immediate of an incident is par for the course, as is tabloids publishing fiction when there aren't enough facts to make a story. My way of thinking about lasting significance is to imagine the event you're writing about happened 20 years ago, and to avoid the minute-by-minute reporting. HJ Mitchell | Penny for your thoughts? 11:54, 18 November 2021 (UTC)

@HJ Mitchell: - As I said above, inaccurate initial reporting is to be expected, and is accepted. In this case, it is the severity of the inaccurate initial reporting, and the reason for it, which is notable. This is far beyond the norm. Mjroots (talk) 13:22, 18 November 2021 (UTC)

Agree with HJ Mitchell's suggestion, a brief mention at most is all that is required. This was no different to any other evolving stories, there was a fair bit of misinformation, but once the dust has settled the actual facts were established, and the article should be written to reflect this rather than a few hacks expressing their opinions. Ghemptoude (talk) 00:14, 6 December 2021 (UTC)

What is actually being said

Initial reports of the accident, based on a leaked Network Rail log, led to highly inaccurate reporting of the accident in its immediate aftermath. The Evening Standard reported that a derailed locomotive had been left sitting duck for seven minutes after the rear carriage of a train derailed and was then run into by another train after signalling failed. Rail editor Nigel Harris pointed out that the only facts in the story were the location and services involved. The leak led to the MailOnline reporting similarly, with a "senior Network Rail manager" quoted as saying "According to my system, the signalling system was aware seven minutes before impact. It should've automatically stopped the train. It should've automatically set all signals to red. If the driver didn't see the signal, the system should've made the train stop." An initial statement put out by Network Rail also contained incorrect facts. Although it was quickly corrected, Harris said that "the genie was already out of the bottle". The Network Rail log initially reported that the GWR driver had reported hitting an obstruction and derailing. The log was updated seven minutes later to record the collision, which is where the press got their "seven minutes sitting duck" scenario. Harris called for Network Rail (and its to-be successor Great British Railways) to be given responsibility for public statements of national importance. Under the current system, Train Operating Companies, the British Transport Police, and the Office of Rail and Road can all issue statements.

How it looks with the expanatory footnotes

Initial reports of the accident, based on a leaked Network Rail log, led to highly inaccurate reporting of the accident in its immediate aftermath. The Evening Standard reported that a derailed locomotive had been left sitting duck for seven minutes after the rear carriage of a train derailed and was then run into by another train after signalling failed. Rail editor Nigel Harris pointed out that the only facts in the story were the location and services involved. The leak led to the MailOnline reporting similarly, with a "senior Network Rail manager" quoted as saying "According to my system, the signalling system was aware seven minutes before impact. It should've automatically stopped the train. It should've automatically set all signals to red. If the driver didn't see the signal, the system should've made the train stop." An initial statement put out by Network Rail also contained incorrect facts. Although it was quickly corrected, Harris said that "the genie was already out of the bottle". The Network Rail log initially reported that the GWR driver had reported hitting an obstruction and derailing. The log was updated seven minutes later to record the collision, which is where the press got their "seven minutes sitting duck" scenario. Harris called for Network Rail (and its to-be successor Great British Railways) to be given responsibility for public statements of national importance. Under the current system, Train Operating Companies, the British Transport Police, and the Office of Rail and Road can all issue statements.

Do you see the difference that using notes to expand on what is being said makes. Mjroots (talk) 14:17, 18 November 2021 (UTC)

What I see is that these are truly references which are just oddly formatted. Having "The Evening Standard reported ..." would logically require a reference and not an explanatory footnote (which is what efn is usually used for). Idem with the Fail. The other two are just quotes from primary sources (themselves quoted in a secondary one), which is again usually just as well given using |quote=, without any issue. If the issue is that of avoiding having the DM as a source, then the only solution is just to cut it out instead of pretending that changing ref tags to efn is actually meaningful here. RandomCanadian (talk / contribs) 15:48, 18 November 2021 (UTC)

@RandomCanadian: is the paragraph referenced solely to Rail acceptable? The remarks by the Network Rail manager are given in full in the editorial, so we can still link them to a reliable source. Mjroots (talk) 16:16, 18 November 2021 (UTC)

"Acceptable"? Probably. "Ideal" or even "good"? Not quite so; as it makes little sense to be referring to a primary [such as an early newspapers report] source (which is discussed by a secondary source) without giving it's citation if said citation is available - this is independent of whether the source is reliable or not. A source is usually at least reliable for its own content (i.e. WP:ABOUTSELF and WP:PRIMARY - A primary source may be used on Wikipedia only to make straightforward, descriptive statements of facts; so sourcing "Bad source wrote X" to [Bad Source] - with additional commentary from a more acceptable source, as here, is ok). RandomCanadian (talk / contribs) 04:52, 19 November 2021 (UTC)

I've reinstated the full text, referenced to Rail, and added in Wolmar's criticism and praise. Hopefully we can put this to bed now. Mjroots (talk) 05:21, 19 November 2021 (UTC)

Notes

References

Interim report

The interim report has been published. There's new photos and diagrams that might be worth extracting. Mjroots (talk) 11:16, 21 February 2022 (UTC)

Reporting II

@Grandtubetrains: it's supposed to be BRD, not BR∞. I'm now invoking the D part.

Re your second removal of text from the "Reporting of the story" (same as first removal) - This was widely reported on in the Rail press at the time, with much criticism. I reject your claim in that edit summary that "it's not entirely neutral". Where the mainstream press was praised by the rail press, then that is given due coverage. Your gutting of the first paragraph means that much context is lost. Mjroots (talk) 12:36, 13 May 2022 (UTC)

I agree with what was there, i.e. pre-deletion, and that the reasons for removal were wrong. My only observation is that the quote should not be in italics. 10mmsocket (talk) 17:04, 13 May 2022 (UTC)

I agree that "Much of the subarticle is in the wrong chronological order, and it reads like a story rather than a wiki article." Can I suggest that we resequence the subsection Reporting of the story without deleting or changing any content, and then discuss whether any text needs to be changed? Hallucegenia (talk) 20:09, 13 May 2022 (UTC)

I remain to believe with my point regarding the sub article - it is written oddly, with a single paragraph making it look clunky (until after some alterations were made to the MailOnline senior manager "quote"). In retrospect, deleting some of the text was wrong but to me it looks as though it is written like a biography or as part of a news article. The majority of the text is unsourced, the only two being the RAIL editors' criticisms - can we get this sorted as well? Grandtubetrains (talk) 21:51, 13 May 2022 (UTC)

@Hallucegenia: Am amenable to your suggestion that the reporting section could be moved. Would suggest between Aftermath and Investigations would be a good move.

@Grandtubetrains: Where is there any uncited text in the version of the article before your last edit? Mjroots (talk) 05:36, 14 May 2022 (UTC)

If we move the reporting section, the image of the RHTT would sit better in the investigations section. Mjroots (talk) 06:00, 14 May 2022 (UTC)

To put a cat amongst the pigeons, MailOnLine is not a reliable source. Murgatroyd49 (talk) 10:53, 14 May 2022 (UTC)

Agreed, but in this case it's amplifying the fact that the Mail is a P.O.S. by giving an example of its bad reporting. Perhaps it's permissible here? 10mmsocket (talk) 11:25, 14 May 2022 (UTC)

@Murgatroyd49: the MailOnLine is not being used as a source, per the extensive discussion in sections above. Rail is most definitely a reliable source, and it is their criticism of MailOnLine that is being remarked upon. Mjroots (talk) 13:13, 14 May 2022 (UTC)

I've moved the Reporting section up, and moved the image. Mjroots (talk) 16:05, 14 May 2022 (UTC)