German Casualty discussion

[Ralph J. Whitehead]

Come off it, Ralph, you're well aware that I've given you plenty of reasons in the past why I view your posts as nothing more than excuse making, and you also know that I've clearly stated in several threads that I'm not a defender of Edmonds per se, but I am a severe critic of the nonsense that German casualty lists breed - so your apparent indignation does you no credit at all.

I'm now challenging the very logic of your stance by using your own words, and putting you on the spot i.e. I'm asking you directly, given the masses of research you've done, which German list you would recommend we accept as fact - or have you made so many caveats, for each individual list, that you now find it impossible to answer?

Cheers-salesie.

My response is simple. The thread is not an attempt at excuses. I was under the impression that most scholars would like to see all of the information possible on any subject so a clear and concise conclusion could be drawn and supported. I like to know everything possible about my sources before making a judgment, as failing to do so could allow me to provide less than accurate details. If new materials come up then the research needs to be reviewed.

Each list and report that was prepared during the war and afterward that have been discussed here I accept as fact. Each list is independent of the other when created and while some use details obtained from earlier reports there is nothing in my research that indicates any list is a work of fiction or a misdirection by evil government officials and the like. Use each list and report for what it provides. Utilize them in research. Compare them with opposing lists of the same period of time and criteria.

I have yet to see any solid research at all that any of the information is patently false or contrived. Until I do then I have little to add to the discussion.

Ralph

[salesie]

Perhaps we need a fresh thread so we can get into a discussion about why we should not regard Brigadier General Edmonds as a lying fraudster, pursuing his his own decidely iffy agenda when it came to German casualties. Harsh words, dear reader, but sadly this is precisely the tone and approach adopted by those who attack, on the basis of very little, or no, knowledge, the efforts of those who have studied the matter and who wish only to cast light on the subject - especially Ralph who has probably done more than any other individual in modern times to study and understand the casualty reporting methods.

Jack

I apologise for the tone of my posts, Jack, but certainly not for their content. I have no doubt that Ralph is committed to his research, and he is to be commended for that. But I have to say that not everything depends on oodles of research, and hours of number crunching; in the end any argument derived from said research has to stand or fall on its own intrinsic logic. And if Ralph wishes to contest the rationale of certain arguments then it seems only fair that the logic of his own words should be equally open to challenge. Research as a means to an end is welcome, but research as an end in itself is worthless - hence my "bottom-line" question to Ralph.

Cheers-salesie.

[phil andrade]

Two million dead, seven million casualties overall....there is widespread consensus about this order of magnitude of German military manpower losses in the Great War. Not one hundred per cent agreement, of course : Edmonds would have us believe that the true total of German war dead was nearly four million.

Well he would, wouldn't he ?

Just imagine, for a moment, that we accept that figure of four million dead. Make sure we take into account the many millions of wounded as well - oh, yes, and do not forget that missing caregory of lightly wounded - not to mention the virtually one million who were taken prisoner. The implication herein is that virtually every single man of the 13.2 million German males of military age would have been killed, or died, or suffered wounds of varying severity, or been taken prisoner. What do you reckon ? That's just a bit silly, isn't it ?

Phil (PJA)

[Ralph J. Whitehead]

Truthergw

No new list was presented and no new tally was presented in this thread. It has never been said it is infallible but on the very same issues that could indicate not everything was accurate then this would apply to every list from every country. Same issues apply, documents being tabulated by people. Could errors be found? Yes. Do they negate the list? No.

After looking through the methods and numbers used by Edmonds and his theories and self-admissions of his lack of any effort to actually prove his numbers or reasoning from his own correspondence and you still feel that he presented the numbers accurately then so be it. In my view Edmonds may have produced an excellent set of official records on the BOH but he needs to return to basic math before telling the world how many Germans were killed, etc. It has been demonstrated that he was in error numerous times. Please provide the numbers that show he is correct and their source, thanks.

In most instances, just because your friends all agree with you it does not make you right. In accepting the level of effort needed to complete the SanB and if the numbers are accepted they do not make other lists and sources of information wrong. There is no logic to that idea at all.

Of course, the evil Hun wanted to deceive the world. Have you any idea of the level of bureaucracy that existed in Imperial Germany? Not any less in the decades that followed? Unless you can show me the deceptions, the errors made and not just insinuate they existed then there can be no genuine discussion of the issue. Oddly enough, when I search through numerous German official documents and lists when researching individual men a common thread appears – the high level of detail, information and confirmation of data. I have utilized many of the period records, newspaper reports, current archival data from individual towns and villages and have been able to build a picture of the life and death of a soldier. It would seem, based on your innuendo of conspiracy, all of the people from every source looked at was in on the criminal act and they repeated it millions of time because it seems they had the time and resources to coordinate such a vast network of agents.

If government officials are so devious then all government lists, reports, tabulations for every country in the world is wrong. No facts on any subject can be accepted or is it just German officials during the Great War and in general?

Basics: The VL was printed 6 days a week from almost the start of the war until 1919. It contained different details on German losses as reported through the system in place at the time. At first more details were printed, later it changed but the basics were still there. They were snapshots of the reports filed at the time. Corrections were made as information changed, as all countries did in their accounting system. In the end the lists stopped printing as it was deemed no longer necessary by the government after the war had ended some 13 months earlier and the loss details had been printed.

The ZNA also kept records of monthly reports and numbers of men killed, etc. but did not narrow the focus to units, names, dates, etc. They took the basic numbers as a whole. It should be noted that this was not the main reason they existed, it was more focused on war graves, missing men and their resolution, etc.

The RA and ZNB and the commission writing the SanB used war time information and reports to provide their data. There was no coordination to their efforts, their aims in supplying the data and there was no reason to do so. The lists were used for their specific purposes by the authors and calculators. If you do not know of any factual information to show otherwise then you have no basis to refute this other than innuendo.

Now you add a new theory, a post-war issue on the causes of the war. What has this to do with the list or are the reports and lists all suspect because of association? If you know a person who committed a crime should we lock you up because you know him and nothing else? Stick to the thread discussion. Your analogy is convenient, no proof, no connection to the data collected years earlier. It is as if the war time clerks knew the future and if so why did the war continue then until 1918?

Different ideas are great but if the people putting forth an opinion I would at least expect that they had looked at and researched the documents and books and reports in question. Blanket dismissal without research is useless.

Finally: “gives a clue as to how hard it may be to actually lay one's hands on truthful data where it was in the interests of war time and post war authorities to obscure it.” Sounds like Edmonds. Again, no proof, no facts; just guilt by association.

Ralph

[phil andrade]

Ralph,

There is one glaring disparity in the SB that troubles me, and I would appreciate your comments as to the whys and wherefores of this.

I fully endorse everything you say, and, likewise, I think that Jack is absolutely correct. How any one can believe the fatuous assertions of Edmonds baffles me.

Here comes the "BUT" :

The SB is clearly meticulous and comprehensive in its chronicle of wounds and sickness, and accidental injury. To allege that it sets out to distort or surpress the extent of German casualties is, in my opinion, preposterous. Why, then, does it tabulate a total of killed in action and died of wounds that is so much lower than the officially sanctioned figures that were produced by the German government during the war itself ?

Here are some figures that demonstrate my point. By January 31 1917, the German official communique of casualties (VL or ZN ?) revealed that 929,116 soldiers had been killed in action, or had died from wounds. In addition, there were 247,991 reported as prisoners, and 276,278 missing, whose fate was yet to be clarified, and of whom many were surely killed. By October 31 1917, nine months later, these figures were 1,138, 768, 387,979 and 263,043 respectively. The big increase in the number of battle dead and prisoners reveals the intensity of the fighting during that time, although, significantly the number of unnaccounted for missing had dropped as the fate of these missing men had been clarified. It should be noted that additional deaths from disease and other non battle causes were reported too : a total of 59,213 by the end of January 1917. And, of course, these figures do not allow for the fact that an additional half million casualties were subsequently tabulated in the years after the war, and that the inclusion of some of these would necessarilly increase the actual number of casualties at that time.

The SB, as of July 31 1918, exactly a year and a half later, and after some of the most intense and prodigal battles of the war, reported a total of only 1,061,740 combat deaths ( 772,687 KIA, 289,053 DOW), and 771,659 missing/PoWs. The figure of battle deaths here is obviously far too low - it had already been surpassed nine months before that date, by admission of figures that were bound to be incomplete at the time.

George has suggested that this might be because the SB was primarilly concerned with hospital records, and that the tabulation of those killed outright was of secondary importance. It does seem a very anomalous approach for such a colossal work.

With that in mind, I'm wondering whether the problem lies not in the exclusion of the lightly wounded - Edmonds would find all those he needs, and more besides - in the SB : it might be that the record of men killed in action is deficient in that source.

The only solution that I can think of lies in those three quarters of a million plus who are recorded as missing in the SB : but then an awful lot of those would have to be allocated to the killed in order to reconcile the figure of battle dead with the VL/ZN, and we would be left with too few PoWs.

Now I have cited the SB returns for the Battle of the Somme, and used them to refute the Edmonds/Oman argument. The figures are 57,987 killed, 273, 132 wounded and 85,683 missing. Is it feasible that this return is flawed by an understatement, not of the wounded, but of the killed ?

Phil (PJA)

[salesie]

Ralph,

There is one glaring disparity in the SB that troubles me, and I would appreciate your comments as to the whys and wherefores of this.

I fully endorse everything you say, and, likewise, I think that Jack is absolutely correct. How any one can believe the fatuous assertions of Edmonds baffles me.

Here comes the "BUT" :

The SB is clearly meticulous and comprehensive in its chronicle of wounds and sickness, and accidental injury. To allege that it sets out to distort or surpress the extent of German casualties is, in my opinion, preposterous. Why, then, does it tabulate a total of killed in action and died of wounds that is so much lower than the officially sanctioned figures that were produced by the German government during the war itself ?

Phil (PJA)

Perhaps the discrepancy is greater than you suppose, Phil? McRandle & Quirk, in Blood Test Revisited, give a breakdown per calendar year to show the differences between the ZN & Sanitats, Table 11 (They argue that the imbalance between categories doesn't matter because the totals are close, and, therefore, this is validation of the veracity of both lists. I say, of course, that's absolute cobblers [see my review in another thread]).

Table 11 (page 689)

Comparison between the Zentral Nachweiseamt Cumulative Casualties and Sanitats Cumulative Casualties (in thousands)

The respective categories being: KIA/Died, MIA/POW, WIA, Total

12/31/1917

Zentral Nachweiseamt 1272, 666, 3179, 5117

Sanitats Total 635, 664, 3901, 5202

Western Front 462, 521, 2783,

Eastern Front 173, 144, 1118,

12/31/1918

Zentral Nachweiseamt 1621, 861, 4104, 6585

Sanitats Total (estimated) 841, 1163, 5060, 7064

Western Front (estimated) 667, 1019, 3917

Eastern Front (estimated) 174, 144, 1143

Cheers-salesie.

[Latze]

Tom,

There is near unanimity in the Entente camp, including the Americans, that the Germans either did not report casualties or when they did, consistently under reported them.

if that is true, could you point me to a French estimate of German casualties? I am unaware if there is/was a similar controversy in France regarding the German reporting. As my French is very sketchy I do not have a good overview over the discussion there. I think it would be interesting to see what 'their' take on this is. Thanks in advance!

to all:

Regarding the Americans: does anybody know what the "US War Department report from 1924" (cited by Randle & Quirk for example) is exactly? The relevant footnotes in the literature do not lead me to the original. I think it would be highly relevant to the point Tom is trying to make here what it says about the methods used to arrive at their numbers and their relationship to official German counts in 1924.

regards

Matt

Edited 3 February , 2011 by Latze

[ph0ebus]

Daniel, It is a reasonable question to ask, and yes, I have done a great deal of research on regimental records such as the post war histories and the VL printed during the war. I have come across some differences but after a further search it seems that most is a matter of spelling or dates. Some dates in regimentals were not as accurate as others but I will add these are few and far between.

In regard to the comparison of known injuries to the VL I have found no errors as such. It might be that one battalion was prionted one day, the next in a list or two after. The names, and some are mispelled only to be corrected down the road, are found where they should be. The biggest problem I face is that almost every regimental that has a roll of honour does not list wounded men, only killed and fatally wounded, fatally injured, died from illness, etc.

One Bavarian artillery regiment goes as far as listing all killed, wounded, injured, etc. Books such as the listings for the Jewish war dead are helpful as they list names, rank, units and the VL they appeared in. The names are also referenced by towns in each German state so it is easy to trace them back to the VL.

If I can help out in a search for the appropriate VL for your relative please let me know and I will see what I can find.

Ralph

Hi Ralph,

Thanks for your reply...it clarifies thing for me greatly. I will also take you up on your kind offer via PM.

As an addendum, I am somewhat mystified by the 'spirited' replies to your posts in this thread. As someone who professionally reads an enormous amount of research, albeit in the realm of psychology and social work rather than history, I see your approach to this as being quite on point. A good researcher presents the available data, offers an interpretation of that data, points out the strengths and potential weaknesses in the conclusion(s) being drawn, and opens the floor to debate and potential avenues for further research. Given the magnitude of the task you have undertaken, my hat is off to you, and those Pals who engage is this sort of refined re-evaluation of those "facts" the average fellow has over the years perhaps taken at face value without giving what they are reading any substantial critical thought.

-Daniel

[truthergw]

Truthergw

No new list was presented and no new tally was presented in this thread. It has never been said it is infallible but on the very same issues that could indicate not everything was accurate then this would apply to every list from every country. Same issues apply, documents being tabulated by people. Could errors be found? Yes. Do they negate the list? No.

...................................................

In most instances, just because your friends all agree with you it does not make you right. In accepting the level of effort needed to complete the SanB and if the numbers are accepted they do not make other lists and sources of information wrong. There is no logic to that idea at all.

..............................................................

Ralph

We are in agreement on one aspect of the discussion. No new evidence is being produced. I say I have misgivings as to the absolute veracity of the latest list, you tell me Edmonds deliberately misrepresented German casualties. Until I see some new evidence, I'll sit back and watch this one.

As regards my friends. Some agree with my ideas, some disagree. The vast majority of my friends have no interest whatsoever in the war or casualty counts. Of one thing you can be assured, I post my own ideas and do not seek my friends' approval nor support for these views. I am surprised you seem to think otherwise or indeed that you would think fit to bring my friends into the discussion at all.

[phil andrade]

Perhaps the discrepancy is greater than you suppose, Phil? McRandle & Quirk, in Blood Test Revisited, give a breakdown per calendar year to show the differences between the ZN & Sanitats, Table 11 (They argue that the imbalance between categories doesn't matter because the totals are close, and, therefore, this is validation of the veracity of both lists. I say, of course, that's absolute cobblers [see my review in another thread]).

Cheers-salesie.

The McRandle& Quirk article was a dsappointment to me, salesie. The important fact that the SB return for killed is for confirmed killed in action only, and does not include died of wounds or sickness, is either overlooked, ignored, or not understood by the authors : at least, that's how it appears to me. The ZN figure includes confirmed deaths from all causes, and is therefore literally more than twice as high by the end of 1917. And both sets of figures for deaths, of course, do not include the great numbers of missing who would susequently have to be added to the count of the dead.

The SB did actually compile a total of confirmed deaths for all causes up until July 31 1918. The figure presented was 1,202,042, of which 140,302 were from disease, accidents or other non battle causes. This total was still below the ZN figure for the end of 1917, when the massive loss of life in the 1918 German offensive was still to come. And so, clearly, the SB count of fatalities is very significantly understated. I am convinced that the authors of the SB were aware of this, and were at pains to explain it. Although I cannot read the German, I can see the figures, and the total of just under 2,037,000 deaths is clearly endorsed. I wonder if you, Ralph, might be willing to tell me what the work says about this. Please don't think me presumptuous...I know you've got a lot on your plate. Does the SB explain about the several hundred thousand German battle dead who had fallen, but not been included in the total of killed and died of wounds ? Is the missing figure the key to this ? If, say, half the 771,659 missing as of July 31 1918 were considered to be dead, then the notional total of battle deaths would reach a figure approaching one and a half million, which would equate more with the statistics of the ZN. But, if this is the case, there would be a very significant dicrepancy between the number of German PoWs claimed by the Allies by the end of July 1918 ( about 400,000 in the West and over 150,000 in the East) and the number implicit in the remaining missing in the SB tabulation.

When it comes to counting the wounded , injured and sick, the SB is the Holy Grail. I'm a little wobbly about its figures for killed and missing, and seek some clarification.

Phil (PJA)

[Ralph J. Whitehead]

We are in agreement on one aspect of the discussion. No new evidence is being produced. I say I have misgivings as to the absolute veracity of the latest list, you tell me Edmonds deliberately misrepresented German casualties. Until I see some new evidence, I'll sit back and watch this one.

As regards my friends. Some agree with my ideas, some disagree. The vast majority of my friends have no interest whatsoever in the war or casualty counts. Of one thing you can be assured, I post my own ideas and do not seek my friends' approval nor support for these views. I am surprised you seem to think otherwise or indeed that you would think fit to bring my friends into the discussion at all.

My post did not reference your actual friends, it was related to the earlier comments you made that the allies were in agreement regarding the German losses. This group was my reference to 'friends'.

Phil,

I will copy your recent posts and I will see what you are looking for and get back to you as soon as I can. In the meantime I note you mentioned not reading German. There is a fairly easy fix to that problem. It is not always the best and will require some fine tuning but there are a number of German/English translation programs on-line. Just type in the sections you are looking at and hit translate. The results may be a bit odd as some words do not readily translate into modern German but it can help to understand what the authors were getting at.

I know in the SanB there is a great deal of information explaining each chart or graph, what it represents, etc. Without knowing what they are it is difficult to use many of the graphs as is. Good luck if you try. I will be in touch soon.

Ralph

[Jack Sheldon]

Tom

I have not had a chance to respond to your request for 'new evidence' until now. I am not quite sure what is wrong with the old evidence which, to my mind, is pretty conclusive. Ralph has recently summarised the Williams 1966 article which appeared in the RUSI journal. You may find it helpful to read the whole thing and, if you contact me off-Forum, I can let you have an electronic copy.

For today, I am going to concentrate on the fact that at least some of the British thought on this matter is based on delusion and wishful thinking, taking first as my text a phrase from this sentence by Sir James Edmonds, which appears on p xv of the preface to BOH 1916 Vol II and which I have emphasised in bold.

[Of the German casualties on the Somme] "This gives a gross total for comparison not of 500,000, but 650,000, plus, say, 10,000 in the preliminary bombardment, total 660,000. Edmonds states on the same page that published German casualty figures "exclud[ed]the seven day bombardment" (His emphasis). However, if we turn to page 51 of Vol III of the San B (the existence of which volume, you will recall, I stated that I believed Sir James had deliberately suppressed), we find a table which provides a complete breakdown of casualties suffered by Second German Army from 24 June to 26 November 1916 split into the usual ten day reporting periods.

The relevant entries read as follows: 'Ten Day Period: 21 - 30.6.16; Number of divisions: 14 (24.6 Drum fire and discharges of gas); Establishment: 311,007; [Evacuated]Sick: 4,380; [Evacuated] Wounded: 4,482: Sum of the previous 2: 8,862; KIA and Missing: 2,478; Of whom missing: 1,289; Sum of previous 4: 11,340; Slightly wounded and sick, not struck off unit strength: 3,363.

Square brackets - mine; normal brackets - original German.

This demonstrates that Edmonds was quite wrong to state that the casualty figures for Second Army during the bombardment were not published. It also shows that the German army, just as Ralph has said, recorded every man who reported to his medical officer and was given 'medicine and duties' as the the RAMC would have (and still does) record such minor appearances on sick parade. If we leave out medical evacuations, which is the normal procedure, we are left with KIA 1,189, Missing 1289, Wounded 4,482 = 6,960. Let us also remember that this covers ten days and the bombardment only lasted for seven. I am not going to attempt to say how many of these casualties occured during the non-bombardment three days, because I do not want to be guilty of the same offence as Edmonds - 'Say, 10,000'? Given that you pulled that figure out of your back pocket, Sir James, why not say 15,000, or anything else you fancy? Not only were you guessing, you were wrong anyway; first, because the figures were published, and second, because you probably doubled them.

There have been objections raised about the SanB along the way, but one of the reasons I tend to trust it is that we can go a long way to verifying the above entries, using three methods: One the Verluslisten, which are comprehensive for that period as Ralph has told us; two we can check the regimental histories which provide a lot of information on the subject and three - best of all - we can access giant quantities of primary source information about this time and place. Why? Let us take the example of Serre to Maricourt. The defenders were drawn from contingents of the German army as follows: Serre and north to Gommecourt - Baden; Redan Ridge to the Albert - Bapaume Road - Wuerttemberg(plus Bavarian elements); Albert - Bapaume road to Montauban - Baden; Montauban to Maricourt - less clear cut, because of presence of Prussian troops there during the main part of the bombardment, but they were replaced by Bavarians before 1 July. This means for those interested that solid facts can be extracted from the archives in Karlsruhe (Baden), Stuttgart (Wuerttemberg) and Munich (Bavaria). Note , in addition, that because the Prussian IR 66 was subordinated to 52 Res Div and RIR 99 to 26 Res Div, their records for this period are in Baden and Stuttgart respectively - and not burnt at Potsdam.

Those who have read my Somme book and my Beaumont Hamel and Thiepval guides will already have been exposed to information about this period based on my archival research. Suffice it as an example to repeat that the return for the British for the immense quantities of shells and gas they directed at 51 Res Bde between Serre and St Pierre Divion during the bombardment was: RIR 121 24 KIA, 1 Missing, 122 wounded and RIR 119 20 KIA, 83 wounded. With those figures typical for much of the front, I have no trouble accepting the overall SanB statistics. Furthermore, it is open to any reader of this to go to the archives I have mentioned and do the checks themselves. I am even prepared to provide file numbers to save time if anyone is keen.

This getting rather long, so I am going to break and will return to a contemporary example of British delusion later.

Jack

[Jack Sheldon]

Sir James Edmonds states on p xiv of the Preface I referred to earlier, that in the previous volume of the BOH " ... the compilers had erred in underestimating the German casualties. These were, as British Intelligence at the time averred, actually greater than those of the Allies". This set me wondering about what sort of thing caused the 'averring'; what led them down this path of reasoning? Then I came across the following assessment, contained in a GHQ Summary of Information, Effect of Gas Attacks on the Enemy, 26th - 27th June 1916, 21 July 1916, NA WO157/111, quoted by Duffy Through German Eyes p 125:

"3. Of the troops in the first position, probably about 5 per cent were gassed. From our knowledge of the effects of the gas, it is probable that a high proportion of these cases proved to be serious, including many regarded as light at the time".

Apparently this assessment was based on prisoner interrogation but, if this is an example of the standard of intelligence at the time, then delusional fantasy must have played a large part in their work and it can only have led to accumulated error and systemic exaggeration of German casualties. One sector of the front which was singled out for a major effort with gas was Thiepval from St Pierre Divion to the Leipzig Redoubt, that is to say Sectors C [for Courcelette] 1-9. Let us say, for the sake of the discussion, that each in the normal fashion was the responsibility of one company and that each was at a strength of 225 at the start of the bombardment. I am theorizing here, but it is a reasonable assumption. Doing the maths, the number potentially exposed, therefore, was 225 x 9 = 2,025, 5% of which (rounded down) is 101. We happen to know a lot about the bombardment in this area, because Bundle M43/19 RIR 99 at the Hauptstaatsarchiv in Stuttgart includes inter alia, the minute by minute watch keeper's log of the bombardment, which was included as part of the RIR 99 after action report sent to 26th Res Div on 17 August 1916.

As an example of how detailed our information can be, I am going to upload my translation of the whole thing in bite sized chunks next but, to avoid keeping you in suspense, I can reveal that the immense effort which went into the gas attack on Thiepval was almost a complete failure. It killed one man and slightly gassed a handful. As often as not it blew back into the British trenches. Little wonder, therefore, that the evening SITREP of Second Army for 28 June 1916 (See my German Army on the Somme p 128) stated, 'The enemy's gas tactics, which are being aided by the prevailing west winds, of releasing constantly repeated small clouds of gas is aimed also at gradual attrition. Because of technical mistakes the enemy has so far achieved little through the use of gas'.

If Allied intelligence officers were trying to factor in 5% of the forward strength of the defending regiments as casualties on the basis of this sort of analysis, but could not find them in the German record, no wonder they thought that there was a cover up.

Jack

[Jack Sheldon]

RIR 99 Bombardment 1.doc

Thiepval bombardment Part 1.

RIR 99 Bombardment 2.doc

Part 2

RIR 99 Bombardment 3.doc

Part 3. Some of the references may be a little obscure, but I am willing to provide clarification if needed.

Jack

[Steven Broomfield]

Although I confess that I feel these arguments sometimes resemble mediaeval theologians arguing about the number of angels one could place on the head of a pin, I must say that this discussion has done nothing to convince me that the research by Ralph and Jack is anything other than believable. In this discussion, and all those preceding it, I cannot honestly say the 'opposition' have supplied anything to convince me otherwise.

Yes, there may have been many reasons why the figures might have been fiddled, but there are probably more reasons why they weren't, and until anyone comes up with a case that persuades me otherwise, I must say that Jack and Ralph have my vote.

I also confess myself amazed at the research done by both men, and extremely grateful at their willingness to share it with us.

Thank you both.

[phil andrade]

Please let me second your vote of thanks, Steven.

The work of Ralph Whitehead is heroic, and Jack Sheldon's aslo.

My vote has been very trenchantly anti- Edmonds, and remains so.

It is a fact that the VL included the slightly wounded in the casualty bulletins. The SanB is unimpeachable, in my opinion, when it comes to details of wounds, injuries and sickness.

There is one caveat as far as I'm concerned : the number posted as killed in the SB is, apparently, deficient, and markedly so at that..

Forgive me for reiterating some figures : the total killed and died from wounds, according to the ZN/VL compilations, had already reached 1,138, 768 by the end of October 1917. The SB total in these categories nine months later was only 1,061,740. Of this number, 772,687 were confirmed killed in action, and a further 289,053 were reported as having died from wounds. Note that the number returned as died from wounds equates to about 37.5% of the number posted as killed. Using McRandle and Quirk's monthly SB battle casualty returns, I see that by October 31 1917 some 629,000 German soldiers had been confirmed as killed in action on all fronts. Applying the same ratio for died of wounds as apertains for the figures as of July31 1918, provides a figure of about 235,000 died of wounds : an aggregate of 864,000. This falls short of the official Goverment figures by 274,000, which suggests that the total of killed and died of wounds in the SB might need to be increased by more than thirty per cent in order to make them comply with the tally in the ZN and the details of the VL.

I cannot believe that I would ever endorse such an "Edmondsian" statement regarding adjustments to German casualty figures ! But there it is : there is, for some reason, a significant understatement in the number returned as killed in battle in these SB tabulations. As far as non fatal battle casualties are concerned, this source is not only authoritative, it is a prodigious achievement and an invaluable source. But, because of this deficiency in the tabulations of the number killed, I have to say that I'm uncomfortable about it.

If we survey the regimental casualties figures that Ralph and Jack have provided in their books, we might notice that, whatever the overall number - be it a few dozen or more than two thousand, there is a high proportion of killed amongst them. In this respect, I find them hard to reconcile with the evidence of the SB, where, in the 1916 fighting on the Western Front, an average of only twelve per cent of the total battle casualties were returned as killed. I don't know how this figure compares with the evidence of the regimental losses, but I'm prepared to bet that if we took an overall sample of a couple of dozen German Infantry regimental casualty returns for, say, the Somme, the proportion of the casualties that were posted as killed would be twice as high, roughly, as that SB figure.

I hope I'm right !

Phil (PJA)

[Ralph J. Whitehead]

In an attempt to address numerous questions and details posted earlier I have broken the answer down to answering in bold. This is simply to differentiate between the two entries and for no other reason.

Ralph

Perhaps the discrepancy is greater than you suppose, Phil? McRandle & Quirk, in Blood Test Revisited, give a breakdown per calendar year to show the differences between the ZN & Sanitats, Table 11 (They argue that the imbalance between categories doesn't matter because the totals are close, and, therefore, this is validation of the veracity of both lists. I say, of course, that's absolute cobblers [see my review in another thread]).

Table 11 (page 689)

Comparison between the Zentral Nachweiseamt Cumulative Casualties and Sanitats Cumulative Casualties (in thousands)

The respective categories being: KIA/Died, MIA/POW, WIA, Total

12/31/1917

Zentral Nachweiseamt 1272, 666, 3179, 5117

Sanitats Total 635, 664, 3901, 5202

Western Front 462, 521, 2783,

Eastern Front 173, 144, 1118,

12/31/1918

Zentral Nachweiseamt 1621, 861, 4104, 6585

Sanitats Total (estimated) 841, 1163, 5060, 7064

Western Front (estimated) 667, 1019, 3917

Eastern Front (estimated) 174, 144, 1143

Are these the actual numbers? Were they taken from a third party source or from the actual reports and really did match the categories and time frames used? Perhaps the biggest question, why do you expect the numbers to match? Two sources, two time frames, two different methods of looking at data.

They probably are correct but in any case why not go to the source and not rely on others. Even Churchill made errors in transcribing numbers. They were small but errors all the same. Is it not possible others also made errors in numbers or possibly the parameters of each list or number used? Trust but verify.

There is one glaring disparity in the SB that troubles me, and I would appreciate your comments as to the whys and wherefores of this. Here comes the "BUT" :

The SB is clearly meticulous and comprehensive in its chronicle of wounds and sickness, and accidental injury. To allege that it sets out to distort or surpress the extent of German casualties is, in my opinion, preposterous. Why, then, does it tabulate a total of killed in action and died of wounds that is so much lower than the officially sanctioned figures that were produced by the German government during the war itself ?

What is the basis for their reports? What time periods are these reports from? Are they representative of the information supplied from the 10 day Truppenkrankenrapport and as such still unresolved as to prisoners/missing, wounded men dying at a later stage, etc. The point is that the way each group was reporting data depends upon the timing the data was being submitted and if it was at a point where all of the end results for each entry was known.

If you are counting report information from an initial 10 day reporting document for the Germans on the Somme for 1 July 1916 your numbers will be completely different than a group counting loss information from the end results of these reports. Many of the MIA for 1 July were eventually found to be killed, captured and in a few cases wounded and in hospital. These were reports from front line units with limited details on many of the men.

I continued to check MIA status for 7 months following 1 July and resolved a great many of the MIA men but there were still a good size group left that were not resolved even by the end of January 1917. I suspect they were eventually resolved but I have not been able to look into further lists due to the time involved.

Here are some figures that demonstrate my point. By January 31 1917, the German official communique of casualties (VL or ZN ?) revealed that 929,116 soldiers had been killed in action, or had died from wounds. In addition, there were 247,991 reported as prisoners, and 276,278 missing, whose fate was yet to be clarified, and of whom many were surely killed.

Phil, I really need to know the source of these numbers. This will provide key details on what information was being collected and the time frame of the losses and their possible resolution (MIA, severely wounded, etc.) In regard to assuming many of the MIA were killed I would be careful to avoid making assumptions until the status is as fully known as possible. It could be months or years before details were obtained or filtered through the system to show if a man was captured, died in enemy hands, killed, etc.

By October 31 1917, nine months later, these figures were 1,138, 768, 387,979 and 263,043 respectively. The big increase in the number of battle dead and prisoners reveals the intensity of the fighting during that time, although, significantly the number of unnaccounted for missing had dropped as the fate of these missing men had been clarified. It should be noted that additional deaths from disease and other non battle causes were reported too : a total of 59,213 by the end of January 1917. And, of course, these figures do not allow for the fact that an additional half million casualties were subsequently tabulated in the years after the war, and that the inclusion of some of these would necessarilly increase the actual number of casualties at that time.

What is the source of these figures? If I cannot look at the source and details on the statistics it is impossible to provide any answer. Be careful about the additional losses added at the end of the war. I have heard that they belong scattered throughout the 4 years and I have heard they represent losses from the final period of fighting. Given the state of the army at the end and the level of accuracy from reports in the earlier periods I tend to believe the losses you mentioned were from the final period of the war when the reports were incomplete from many sources. However, as I mentioned in a previous post there are statistical numbers from losses in the final months. They do not cover all units but do show sizeable numbers for the units represented. I will try to dig this out and put these numbers into the mix.

The SB, as of July 31 1918, exactly a year and a half later, and after some of the most intense and prodigal battles of the war, reported a total of only 1,061,740 combat deaths ( 772,687 KIA, 289,053 DOW), and 771,659 missing/PoWs. The figure of battle deaths here is obviously far too low - it had already been surpassed nine months before that date, by admission of figures that were bound to be incomplete at the time.

Also, the reports on file in these charts were not all resolved as the very high MIA/POW number shows. It is a number that the records from that time showed. As they resolved the numbers would change. What chart did you refer to for these statistics? I would like to look at in the SanB and see what the explanation indicates in the text of the book.

George has suggested that this might be because the SB was primarilly concerned with hospital records, and that the tabulation of those killed outright was of secondary importance. It does seem a very anomalous approach for such a colossal work.

The medical reports did include the men killed and this was as important as reporting all other casualties. The Truppenkrankenrapport has a section specifically addressing the men killed, broken down by rank and included categories such as died from wounds, died by accident, suicide and sickness. I understand that there are many comments regarding the numbers placed into postings on the forum but I have doubts as to how many have actually read the SanB and the explanation notes for the dozens of different charts.

With that in mind, I'm wondering whether the problem lies not in the exclusion of the lightly wounded - Edmonds would find all those he needs, and more besides - in the SB : it might be that the record of men killed in action is deficient in that source.

I would only caution making any assumptions until the charts are reviewed along with the explanation text. If you do not know what is being presented and for what reasons and if there are any parameters provided, restrictions on sources and such then the chart cannot be used fully. Each chart has an explanation as to why it appears, the data used and so on. The information that is contained in the separate lists and reports all depend upon the time frame they cover and if the losses are all known or still in process of being determined.

The only solution that I can think of lies in those three quarters of a million plus who are recorded as missing in the SB : but then an awful lot of those would have to be allocated to the killed in order to reconcile the figure of battle dead with the VL/ZN, and we would be left with too few PoWs.

Do not assume or try to reverse engineer numbers to fit accepted assumptions on numbers. See what they were faced with in their text that accompanies the reports, see what can be done to resolve the status of these men before making assignments of percentages for one group over another.

Now I have cited the SB returns for the Battle of the Somme, and used them to refute the Edmonds/Oman argument. The figures are 57,987 killed, 273, 132 wounded and 85,683 missing. Is it feasible that this return is flawed by an understatement, not of the wounded, but of the killed ?

Yes, the missing men would have to be resolved. 99%+ would finally be shown as either killed or as a prisoner. Few would be shown as in hospital and miscounted, etc. The 85,683 names would have to be traced through the VL system in the months following the battle to see what they show. As each piece of information comes in the names are added as corrections in future lists. I also see a problem with the idea that the old accepted numbers by Edmonds are being compared to the loss reports filed at the time of the fighting. Regardless of the numbers of men killed, wounded or captured the facts of the battle remain the same. It just appears to be an issue that more men should be dead than were reported therefore the German loss numbers must be wrong. Not a very scientific method at all.

The important fact that the SB return for killed is for confirmed killed in action only, and does not include died of wounds or sickness, is either overlooked, ignored, or not understood by the authors : at least, that's how it appears to me. The ZN figure includes confirmed deaths from all causes, and is therefore literally more than twice as high by the end of 1917. And both sets of figures for deaths, of course, do not include the great numbers of missing who would susequently have to be added to the count of the dead.

Phil, where are you getting the details for the SanB restrictions in their reports and which reports are you referring to? This is critical if any response is to be made when looking at your statement, thanks. Along the same vein, do not attempt to have both lists match. They were from different periods, prepared for different reasons and in the case of the ZNB (war time designation) the counting of casualties was not their primary purpose in the war and they issued reports yearly I believe and at times at 6 and 9 moth intervals. It was not the official casualty reporting bureau.

At one point in the war I have read that the ZNB staff was so overwhelmed by numbers as information came in, the numbers of casualties they were looking at for missing and war graves simply became too much for the current staff and more people had to be added to handle the volume. In knowing items such as this as well as all other factors possible you can use the information, with the known restrictions, just as you would any other set of data. No reports or numbers can be taken with blind faith, German, British, U.S., French, etc. Each has a number of issues facing them, each needs to be known or you would simply take all at face value and never know just how accurate they could be. This is by no means any excuse for the lists, their flaws or their strengths, they are simply facts that needs to be factored in to any discussion.

It is the same way that some members look at different figures from all the German sources and declare that since they do not all match then they are all lies. I have seen different numbers from every country, some millions of numbers apart. Using this logic no list from any source is valid, they are all lies and therefore our research into this issue is a waste of time.

U.S. loss numbers for enemy dead in Vietnam are extremely suspicious. Many are overcounted when you learn just how they provided numbers. Many are guesses as we had no true knowledge of the actual losses. Helicopter losses, something you would think was simple. It took off and did not return, it was lost most likely. However, if they could recover the wreckage, no matter how bad the condition was, it was not added to the number of helicopters lost. This last portion comes from memory from news reports at the time or shortly afterward when many of these revelations came out.

The SB did actually compile a total of confirmed deaths for all causes up until July 31 1918. The figure presented was 1,202,042, of which 140,302 were from disease, accidents or other non battle causes. This total was still below the ZN figure for the end of 1917, when the massive loss of life in the 1918 German offensive was still to come. And so, clearly, the SB count of fatalities is very significantly understated. I am convinced that the authors of the SB were aware of this, and were at pains to explain it. Although I cannot read the German, I can see the figures, and the total of just under 2,037,000 deaths is clearly endorsed. I wonder if you, Ralph, might be willing to tell me what the work says about this. Please don't think me presumptuous...I know you've got a lot on your plate.

Phil, just identify the particular section or sections you refer to, thanks. Just keep in mind that the two systems did not look at the loss reports and documents in the same manner at the same time. There is no reason to assume or expect that the numbers published in 1934 after years of review of loss reports, hospital records, etc. would match a set of statistics being produced during the war that did not break them down by where they occurred or against which enemy. The ZNB was simply reporting on the number in general as details were provided to it as the war progressed.

Does the SB explain about the several hundred thousand German battle dead who had fallen, but not been included in the total of killed and died of wounds ? Is the missing figure the key to this ? If, say, half the 771,659 missing as of July 31 1918 were considered to be dead, then the notional total of battle deaths would reach a figure approaching one and a half million, which would equate more with the statistics of the ZN. But, if this is the case, there would be a very significant dicrepancy between the number of German PoWs claimed by the Allies by the end of July 1918 ( about 400,000 in the West and over 150,000 in the East) and the number implicit in the remaining missing in the SB tabulation.

Phil, this portion would take some time to go through, translate and review so I am afraid there will be quite a delay on any response. I agree that the SanB is probably the best direct source for details on German losses as based upon the reports filed during the war. Still, you need to know what information is being presented and in what form. Some fo the comparisons used in the charts can be very confusing if the explanation section is not read. Each list or report has their own method and if you know them or as many as you can then you can utilize the information properly.

I would appreciate any discussion on any of these items but in saying so I would like to see direct research and specific details on issues or problems from the lists and methods used. In short, a better understanding on just what is being discussed and not simply opposition because numbers A and B do not match. Tell me why they do not based on reviewing the actual materials, the reports and do not simply rely upon earlier reviews by people who appear to have many agendas on both sides of the discussion. Churchill often used the larger of any loss number he found for the British losses. Why? It probably made his overall argument that the British lost more men than the Germans during the fighting. Of course he omitted the 1918 battles where the numbers were more equal.

Again, it did not fit his theory and was left out. Without knowing these things can you really accept all of the conclusions at face value? Edmonds and Oman played games with numbers until they reached the point where the Germans lost more than the Allies, especially the British but after looking at their methods and assumptions I have no faith in anything that they came up with. This applies to this particular issue and not anything else. From what I have read and my limited knowledge of the British effort as found in the BOH I am content to accept what is written as a true and accurate record of events. If someone with more expertise refutes any part of it I would be very glad to read their findings and conclusions. After all, we are all looking for the truth about the war, aren’t we?

When it comes to counting the wounded , injured and sick, the SB is the Holy Grail. I'm a little wobbly about its figures for killed and missing, and seek some clarification.

Holy Grail? A real good source once you read and understand it? No to the former, yes to the latter. The VL is the same. Is it perfect? No. Is it a great source of information and detail on the actual men who were casualties? Yes. The lack of perfection in the VL or any other source does not make it useless, it only provides clues to the strength and weakness of any research materials we use.

Phil, get back to me on the lists, charts and numbers you were using so I can take a good look at them and see if there is an explanation for the different numbers, Ok?

Ralph

[geraint]

Ralph. Phil, Jack - excellent stuff. I'm not even able to hold a candle to your contributions. I'm honoured to be able to follow this thread. Well Done.

[salesie]

Jack, there is no doubt that GHQ Intelligence (under Charteris) overestimated the number of German casualties the BEF was inflicting in all engagements, right up until August 1917 when it was "forced" to adopt the War Office's (Macdonogh's) methods. And, it has to be said, I’m ambivalent about Edmonds i.e. I can see that his arithmetic is open to question. I'm also convinced that German unit returns were highly detailed, and, in turn, this detail (individual names etc.) was reflected in the published VL.

But I'm far from convinced that such things count as solid "evidence" for accepting the totals in the several German lists that have been bandied around on this forum. As Tom says, a new (to the forum) list seems to appear every so often, and those that propose said "new" list always seem to take the stance that this is "proof", at last, of the veracity of German casualty reporting; I sometimes wonder just how many lists we will eventually end up with? The latest "final proof" being offered is the Sanitats because it is based on comprehensive physical counts by the German medical services - but, in my opinion, it creates more questions than answers, and, thus, casts further doubt on the veracity of said lists (for reasons given mostly in other threads).

A quick summary, if I may:

GHQ Intelligence did overestimate German casualties for a period of time, and there was a contemporary view in London, amongst some, that certain battles were being fought/continued because of Charteris' "flawed" intelligence - but Macdonogh's widely accepted calculations were still higher than what the Germans were "admitting" to (and let's not forget the premature German class call-ups).

Though I can see that Edmonds' arithmetic is open to question, I can also see that his assertions about German lists (total counts) being consistently understated/misleading may well be valid. As I've said before, given that Edmonds receives due praise for his overall work on the BOH, this "slip-up" in arithmetic seems something of a conundrum. No one is perfect, of course, but why would he include "bad" arithmetic on the basis that "everyone knows the Germans consistently understated their casualties" when the rest of his work is pretty scrupulously done? It seems to me that he may have had access to British Military Intelligence files (Macdonogh's work, which showed higher rates of German attrition, but not as high as GHQ's) but was told he could not divulge his source (for obvious reasons of secrecy)? Pure supposition, of course, but no more so than assuming he was defending Haig's reputation, and, after all, the gross imbalances in the German lists hardly lend them to trustworthiness.

The detail in German returns and the VL, regarding individual names etc. is something of a red-herring in my opinion, in that when looking too closely at the trees then the wood is easily missed i.e. a family could see that their loved one was named in them, but how on earth could they know that he was included in the total count? They couldn't possibly know this, so no complaints from relatives etc. would be made to the authorities (as some claim validates these lists). Names are one thing but total counts are quite another; that's when the names become mere numbers.

I admire your work tremendously, Jack, and you're to be thanked for posting extracts from your research but you have given many caveats in the past as to how "unreliable" German stats may well be, and I have to say that you can't "escape" totally from your own caveats, and that small extracts from certain regiments is looking too closely at the trees when the wood is the problem.

Now for the gross category/total imbalances in and between the several German lists we have been given (looking at the wood not the trees). Some pass this off as inconsequential - Ralph tells us that he accepts each and every one of these lists as being valid because some counted "apples" and some counted "oranges", and, not only that, they counted them at different times. I say that is nonsense because all these lists purport to count/measure the same German casualties, in the same categories, in the same war (an even more important point when we consider that we're also told that German administrators were superbly efficient). Even PJA has started to wobble on this point (but I'm not sure whether that's a good or a bad thing )

Cheers-salesie.

[truthergw]

When I suggested that the latest list might be no more accurate than the previous ones, I was told that Edmonds was a fraud who cooked the books. Now, both of these statements may be true, both may be false, or only one may be true. The fact is, the truth of one statement has no effect on the truth of the other. I have given my reasons for believing that the figures may be wrong and whether or not Edmonds was a liar does not affect these reasons. As I said earlier, I'll reserve judgement until some real corroborative evidence comes to light. Until then, this is simply another list of casualties with yet another set of figures.

[Robert Dunlop]

...there is no doubt that GHQ Intelligence (under Charteris) overestimated the number of German casualties the BEF was inflicting in all engagements, right up until August 1917 when it was "forced" to adopt the War Office's (Macdonogh's) methods.

Lest anyone think that Macdonogh's methods were accurate, it is worth considering Occleshaw's review of this. The information is especially interesting in light of Ralph and Jack's excellent posts (thank you both!):

"[british] manpower studies comprised two inter-related sections, one dealing with the strength and losses of the hostile armies in the field, and the other with reserves in the homeland who could serve with the armies in the field. The War Office and GHQ employed different means of assessing these factors. The former allotted a fixed percentage of casualties to every battalion the enemy had engaged in a specific action. To give an example, a paper by Macdonogh on German losses in the Somme offensive up to 30 September 1916 stated that, since 1 July, 380 individual battalions had been engaged by the enemy on the British section of the Somme, of which 35 had been used twice and 25 others had remained in the line for the whole three months. Macdonogh considered it safe to assume that these 25 battalions had received 100 per cent drafts over this period: it was German practice to replace losses in active and reserve units without battalion movements. Consequently the number of battalions engaged was 380 + 35 + 25; giving a total of 440. The German official casualty lists (at this stage belated and incomplete) [Ralph - presumably this is a reference to VL?] showed an average loss of 43 per cent in infantry. Macdonogh considered that 'We shall therefore probably not be taking too high a figure if we assume an average of 50 p.c. casualties in the 440 infantry battalions engaged.' These calculations put German losses at 220,000 men, to which an estimated 40,000 was added for losses amongst supporting units, such as artillery and pioneers. As the French were heavily engaged in the offensive as well, a similar calculation was made for their section of the front. This figure, added to that on the British section, gave a total of 884 German battalions engaged and a total estimated German loss of 522,000 men. This was only an estimate and Macdonogh admitted 'it is of course not possible to give as yet, any exact estimate of the German losses on the Somme'.

While not pretending to be exact, this method of estimating enemy losses continued to be used. When the Germans were relieving their divisions more frequently, as in Flanders in 1917, it became necessary to take a somewhat lower figure as a basis of loss. These estimated percentage was reduced to 2000 men per division, or approximately 17 per cent, with losses of supporting arms being estimated at a fifth of this. On this basis the total German casualties on the British front were estimated at 976,000 men."

Interesting to see mention of, what I presume is, VL and how it was interpreted at the time. This fits with Ralph's detailed explanations of how the lists were compiled.

Robert

[Jack Sheldon]

Salesie

You raise an interesting point with regard to Edmonds and his 'arithmetic', so I hope you will allow me time to think about it. I prefer generally to deal with matters which I can back (even caveated!) from my research but, to answer you properly, I need to give an opinion, so I also need time to construct what I hope may be a plausible narrative about what may have laid behind what seems to be a blind spot of his.

Jack

[salesie]

Lest anyone think that Macdonogh's methods were accurate, it is worth considering Occleshaw's review of this. The information is especially interesting in light of Ralph and Jack's excellent posts (thank you both!):

"[british] manpower studies comprised two inter-related sections, one dealing with the strength and losses of the hostile armies in the field, and the other with reserves in the homeland who could serve with the armies in the field. The War Office and GHQ employed different means of assessing these factors. The former allotted a fixed percentage of casualties to every battalion the enemy had engaged in a specific action. To give an example, a paper by Macdonogh on German losses in the Somme offensive up to 30 September 1916 stated that, since 1 July, 380 individual battalions had been engaged by the enemy on the British section of the Somme, of which 35 had been used twice and 25 others had remained in the line for the whole three months. Macdonogh considered it safe to assume that these 25 battalions had received 100 per cent drafts over this period: it was German practice to replace losses in active and reserve units without battalion movements. Consequently the number of battalions engaged was 380 + 35 + 25; giving a total of 440. The German official casualty lists (at this stage belated and incomplete) [Ralph - presumably this is a reference to VL?] showed an average loss of 43 per cent in infantry. Macdonogh considered that 'We shall therefore probably not be taking too high a figure if we assume an average of 50 p.c. casualties in the 440 infantry battalions engaged.' These calculations put German losses at 220,000 men, to which an estimated 40,000 was added for losses amongst supporting units, such as artillery and pioneers. As the French were heavily engaged in the offensive as well, a similar calculation was made for their section of the front. This figure, added to that on the British section, gave a total of 884 German battalions engaged and a total estimated German loss of 522,000 men. This was only an estimate and Macdonogh admitted 'it is of course not possible to give as yet, any exact estimate of the German losses on the Somme'.

While not pretending to be exact, this method of estimating enemy losses continued to be used. When the Germans were relieving their divisions more frequently, as in Flanders in 1917, it became necessary to take a somewhat lower figure as a basis of loss. These estimated percentage was reduced to 2000 men per division, or approximately 17 per cent, with losses of supporting arms being estimated at a fifth of this. On this basis the total German casualties on the British front were estimated at 976,000 men."

Interesting to see mention of, what I presume is, VL and how it was interpreted at the time. This fits with Ralph's detailed explanations of how the lists were compiled.

Robert

I should think it self-evident, Robert, that Macdonogh, on 30th September 1916, couldn't give an exact estimate of German losses on the Somme; the battle was grinding on, and any such estimate would be premature to say the least. And you take this snippet a little out of context i.e. corroboration, to a high degree, sometime after the battle finished would be forthcoming from later captured German paybooks. Not in the way GHQ used them, subtracting roll numbers to assess individual unit casualties (which formed the basis of Charteris' overestimates) but the conscription-class statistics each paybook carried. These useful pieces of intelligence gave a pretty accurate assessment of historical German manpower wastage, giving substantial corroboration to the accuracy, or not, of previous War Office (or GHQ) estimates (something Occleshaw also explains).

It would be wrong to think that Macdonogh was a "one trick intelligence pony" as Charteris was - Macdonogh was in many ways an intelligence genius with his spread of methods, not least in his highly successful setting up of spy/political-subversion rings amongst dissatisfied Germans in Holland which infiltrated Germany itself. He made the Germans look like rank amateurs in the intelligence game - and history proved much of his work, as much as has been revealed that is, to be highly efficient and eerily accurate.

I would say that British Military Intelligence (under Macdonogh) had a pretty good picture of the true level of German casualties - as Occleshaw also says, 'these figures (regarding the apparent wholesale slaughter of German conscription-classes evident in captured German paybooks) should provide a stock answer to those who glory in the persistent argument that German losses were nearly always far lighter than the British.'

Couple all of this with the glaring gross imbalances in and between German lists, lists that purport to count the same German casualties, in the same categories, in the same war, then I would say that, bad arithmetic aside, Edmonds was right to be highly suspicious of German casualty reporting.

Cheers-salesie.

[phil andrade]

Ralph,

You are an expert in the study of statistics by dint of your professional career, and your analysis of German casualty figures is so profound that you have forgotten more about this than I will ever know !

I have to admit that I am out of my depth here, and I can well understand that you might be rather exasperated by my contentions when I fail to back them up with proper source references..

Despite my fixation on some aspects of the casualty figures, I hope that I've got a reasonably good "feel" for the overall order of magnitude of the German ( and other belligerents' ) losses in the Great War. In general terms, there seems to be consensus that German military casualties numbered roughly 7.1 million : of this total, just over 2 million were fatal ; and of these two million dead, pretty well ninety per cent of them were direct battle casualties i.e. they were killed outright on the field or died from their wounds.

I fail to provide the sources for some of the figures that I cite : this is because I see them mentioned in contemporary magazines written during the war itself , when they are referred to as "Official German figures", or when they are displayed in the Zentralnachtweissant compilations in the appendix at the back of Churchill's history....which seem to conform to the figures that those commentators of contemprorary magazines cite. I do hope that I've managed to discern the "wood" from the "trees".

I must try and marshall my sources better, and, when I can eventually open my Sanitatsbericht document (there's something wrong with my computer at the moment) I will use the google translator. The script is in an extraordinary kind of Gothic style, and it might confound my efforts, but I can see enough of the statistics to convince me that the SB openly admit that the total German dead exceeded two million. Clearly there has to be an accounting for the best part of one million dead who have not been tabulated ( many of them obviously due to the exclusion of the last few months of fighting).

In the meantime, thank you - and Jack - very much for offering to help me with the text.

In respect of German regimental losses, Ralph, I've just opened your volume 1 of The Other Side of the Wire, and seen a reference to a German infantry regiment (RIR99) that, in the autumn of 1915, had fought the Fench near Arras and suffered a loss of 98 killed and 203 wounded. This ratio seems more typical than those recorded in the SB, which return a six to one wounded to killed ratio for the Western Front in 1915.

Phil (PJA)

[George Armstrong Custer]

[....] you take this snippet a little out of context [....]

I would say that British Military Intelligence (under Macdonogh) had a pretty good picture of the true level of German casualties [....]

I agree.

George