Edward Thompson: Wartime C.M.E. Discussion

About 290 LNER Pre-Grouping 4-4-0s were withdrawn between 1944 & 1951 - I suspect that those were the largest group to be displaced by B1s.
The new A1 and A2 Pacifics were also being introduced during this period, greatly increasing the overall numbers of Pacifics. On paper, these probably replaced the Atlantics, particularly Ivatt C1 and Raven C7?

 

Had completely forgot the 4-4-0s in my haste to type earlier. Good reminder.

 

Concerning statistics I sort of agree they can be too much of a good thing. There needs to be enough stats & charts to present a convincing expert argument but without overwhelming the average reader who will hopefully be able to enjoy the book without having to cross-check all the adding up.

Nowadays the web allows you to have the best of both worlds. Put the essential information in the book (with Appendices if essential). And include a link to a ‘Simon’s Stats’ web page with as many additional tables, charts and findings as you ultimately desire.

 

I could not support this more. Ideally: both the entire data base as scanned in and the spread sheet from it. Both complete.


.

 

I can't unfortunately publish the original database myself without incurring significant costs from the National Archives.

 

When you say "original database", are you referring to images of the original cards, or the data contained thereon (in digital form)?

I'm not familiar with the intellectual property situation with regard to the LNER archives; I assume ownership would have been transferred to the British Transport Commission when the LNER was nationalized; after that it's a blank to me.

Noel

 

I think that it could make for two publications. The first is the Thompson book, where I think the key statistical material ought to be paired down to the critical material that supports the core argument.

There is a second equally important discussion about availability/costs etc using the statistical evidence. Perhaps for a more hardcore audience (but no more hardcore than say some of the more technical discussions that go on). This is where you would do the diachronic, synchronic, national and even if possible international comparisons (how does UK availability for example compare to what was being achieved in the US, Germany or France for example). A less Thompson and LNER-centric discussion about one of the core questions that all lines were facing in this period. We have for example one of the US lines making fitters dress up in asbestos suits to reduce down time.

It can be very tempting to try to stick everything in, but it can make more sense to split the material. Less is more etc.

 

I can't publish the original archive material: which is held at Kew Gardens, in terms of the physical, original document. The data contained within, I can publish, because my spreadsheet is derived from that data and no such issue exists in reporting on the data itself.

It has been complicated for a number of things (moreso photographs) but we are at a stage where - reasonably - replicating the original archive in print incurs costs from Kew, whereas showing it by way of my derived database, doesn't.

 

So if you wanted to, you could publish the database on a Web site? If there's no way (other than the book) to monetize all the work you put into transcribing it, maybe that would be the thing to do, eventually.

Noel

 

I have had a conversation with my publisher and something may be possible after the book is published. I am currently focusing on completing the edits he requires, so I can start adding captions to the new photographs in the body of the book.

 

This is inevitable. (And incidentally my area of expertise).

Data is raw. The art - not science - of Statistics (or Numerical Analysis or Data Modelling or Data Mining) is to try and create the 'best' interpretation of the raw data allowing for circumstantial variables, not least human error in gathering the raw data in the first place. This may be to illuminate actions in the past or assist in determining the path to follow for the future.

'best' might mean most accurate - which is what I have always tried to do.

But it can also mean - just as with interpretation of historical facts, which has never been regarded as anything but an Art - the approach which best supports the argument or best exonerates the company or which attracts the most remuneration.

Me being a stickler for the 'best accurate estimate' is why I'm typing this in a leaky house in North Wales as opposed to being a highly paid advisor in either public or private service. I'm not bitter - I'm happy and with a clear conscience.

 

I have some considerable sympathy with Simon over all this as I am aware of the considerable financial implications.

One might explore whether Kew might make an exception in this case (unlikely perhaps).

But Simon could share his research (raw documents such as pics of all the relevant documents from Kew) with others such as Pete for a 'peer review' etc. as part of research etc.

Cheers,
Julian

 

The detailed transcription, in machine-readable form (which Simon already confirmed he's free to release) wouldn't be 'good enough'?

Yes, the advantage of scans of original documents would be that that cuts out one potential source of error (transcription mistakes), but there are several disadvantages:

* One would have to deal with Kew's restrictions; a lot of paperwork at the very least, and likely expense (and don't even think of blowing that off, and just rolling the dice; that would likely damage future access for other historians - once bitten, twice shy)
* It's infinitely easier to deal with data, especially voluminous data, in digital form, rather than a bunch of document images
* Using data from the images, there'd still be the issue of transcription errors (just different ones)

Maybe there's something I'm missing, though?

Noel

 

I have shared my research, including references, sources and similar. The books bibliography is extensive and will point everyone in the right places to go.

Regarding my data in the spreadsheet, I have made it available multiple times. I’ve always been happy to share it. The source data is different though, and although I’ve shown my printed scans at lectures, I’ve never been comfortable sharing them over the internet, and I feel like that for a variety of reasons.

I did three solid weeks of work at Kew where everyday I was photographing or scanning documents. It then took three years to convert the data into a spreadsheet. It has then taken a further year to consolidate the data into a reasonable depiction of events, including a graphical analysis. It has been a monumentally large undertaking.

Given the data has been at the heart of my consistent arguments for a reframing of Thompson's reputation overall, I feel strongly that my offer to share the spreadsheet derived from it is enough. If the question is now one of accuracy: I have had the data validated against the original source document by several statisticians, and any errors in transcribing have been rectified. So you can at least be assured of the spreadsheet's accuracy to the source document.

Anyone following this thread over the last four years will have seen my repeated efforts to make the data more readily available. Indeed, on this thread (acknowledged in the book) we created the availability formula to give a percentage figure for all of the L.N.E.R. classes, annually; as the dataset gives us the ability to do this. I have had this work checked, raised this in the thread, and asked for responses many times.

The idea that I haven't shared my research or data enough isn't fair: I have done so, consistently.

 

The valid point in this is that the Data has value - in the case of Kew a financial one that provides income to maintain its services and products. In essence Simon has the same right by virtue that he has compiled a set of Data that has value yet despite it being commercially valuable he has always been willing to supply it FREE OF CHARGE. Whilst that has allowed others to suggest improvements or raise questions that have encouraged Simon to analyse other sources, it is the case that Simon has a product with value that he has been willing to share; the return for him has been the responses from this forum. I have previously suggested that his data could support a second book and it appears that this opinion is shared by his publisher thus I wish him well with this as I await publication of the book in the near future.

 

May I add to this - I put some money in Kew's coffers by way of a membership, donations, and quite a lot of coffees at the café!

I am glad I did the bulk of my archive research pre-covid-19. I am not sure I would have been able to do everything I needed to, now.

I feel strongly that my second book needs to be on Bulleid, purely for branching out further into the tangled web that is the Second World War and locomotive design. However, I have such a huge dataset and source material available to me, that I have begun collating lots of different chapters of graphical analysis on the L.N.E.R.

It may yet become a second book in itself.

 

And here's me hoping youd've been having a crack at the work of Francis Webb next!

 

Never say never...

 

Be fair; Thompson was much influenced by Gresley and so was Bulleid hence it is surely more appropriate to follow up the "Gresley influence" with research of others such as Arthur Peppercorn and J.F. (Freddie) Harrison whose work was similarly influenced. Perhaps confirm my thoughts asto how far 71000 Duke of Gloucester (designed by Harrison) could be described as the last LNER Pacific despite it being built at Crewe ?

 

There you go again with your damned "logic", Fred.