May 14th 1940 and few little things

[Rob Philips]

Thanks, Franek. I'm not looking for satisfaction, but for knowledge. Your reply tells me that many Polish aviators produced accounts about the French campaign. That's a start. Now the rest, please.

Regards,

Rob

[Juha]

Rob
IMHO you made a fundamental error here.
IMHO your list should be like:

1. A total of X Polish military aviators participated in the French campaign.
Source:
2. We have found accounts of Y of these aviators.
Sources:
3. of these accounts in Z dissatisfaction is expressed with the French military aviators.
Sources:

4. The dissatisfaction that was expressed can be summarized as follows:
A. About the men
B. About the machines
C. About the organisation(s)
5. Z/Y is a significant percentage. We cut short the definition of "significant" in statistics, and declare =>20% to be significant.

Juha

[Rob Philips]

Thanks, Juha. I consider this as consent with my proposed strategy to deal with the matter. You propose increased clarity, which is exactly what I'm after.

You raise three issues, to which I would like to respond as follows:

1. The paragraph 3. you propose is in fact present in my paragraph 2, after the comma. I am not interested in the total number of accounts produced by the Polish aviators who participated in the French campaign, but in the total number of Polish aviators who produced accounts expressing dissatisfaction with the French. The wording was selected so as to avoid counting multiple accounts by the same aviator. However, your suggestion shall work just as fine.

2. You suggest that relevance be established by looking at the dissatisfaction accounts divided by the total number of accounts. This could work, but I suggested the other way, as Franek is looking for ways to generalize. My proposed way looks at all Polish aviators involved. Your proposed way looks at those of the Poles who produced accounts. That would mean a limitation of the group called "The" Poles, hence a limitation of the scope of the generalization. Consider the situation that the group size would be 200, and that there were 10 notorious rebels, who were always venting dissatisfaction. If a generalization would be made on that basis, then this would not do justice to the 190 others, and therefore not to the group.

3. Perhaps therefore you suggest to raise the significance level from 10 to 20%. That's fine with me. It has always surprised me that in medicine doctors produced theses on the basis of less than 10 cases of a certain type of disease, whilst the group described runs into the millions. A level between 5 and 10% is usually accepted as sufficient to declare significance, but surely a higher percentage shall deliver a clearer case. We are free to make these choices, as long as we explain what we are doing, if we wish to make sense.

An improvement of the proposed strategy would be to shift the significance declaration to the chapter "opinions". In fact this is a nice example of Franek's statement, that much of what we call fact is in fact opinion. Again, the fact here is the agreement about the content of the statement, not the content itself.

Regards,

Rob

[Juha]

Hello Rob
my point is that if we are interested in how common the negative attitude to French performance during the BoF was amongst Polish pilots we must study the group of whose attitude we has knowledge. Then there are those whose attitudes are unknown to us now. And the attitudes of those we cannot study now.

An extreme example would be:
3 million soldiers from country A fought in WWI, nowadays we can find written description on how they felt during the war from 50 000 of those soldiers, of which 45 000 describes mostly negative feelings. Now 45 000 out of 50 000 gives impression that negative attitude to WWI style warfare was very common amongst participants, 45 000 out of 3 000 000 on the other hand means that negative feelings were not common. Of course in the first case we must make clear that we have know ledge of attitudes of only 1/60 of participiants and there might have been tendency of those who were critical to be more likely to vent their feelings.

Juha

[Rob Philips]

Thanks, Juha, good point. The difference with "les poilus de la Grande Guerre" is that ALL pilots were debriefed, which gave them a first opportunity to express whatever they wanted, including dissatisfaction with others. Certainly any knowledge we may gain is limited to the data we can find, and needs to be updated if we find more data. The way out here would be to follow both paths, the ones suggested by you and by me. That would make clear the limitations of any conclusions to which we might come.

Furthermore, studying the complaints can shed light on the foundation of these complaints. If an objectifiable and marked difference in aircraft and/or armament performance would be at the basis of this, then this would tell us something, whether reported by few or by many.

If we want to generalize AND make sense, we need to dive deep into the matter, and we need to make clear what we are doing and how we are doing that. Your point is a good illustration of that.

Regards,

Rob

[Ruy Horta]

Thought 5% was significant in a normal distribution.

But we can go too far in the mathematical approach, imho.

[Franek Grabowski]

Well, there is another issue. Those 45 reports cover virtually all assigned airmen - they were filed by commanders of sections dispersed across France. I estimate that about 75 percent of reports representative for about 75 percent of 190 Polish fighter pilots, contain negative remarks against Frenchmen. The remaining 25 percent does not have any remarks, be it positive or negative. I presume authors did not like to write too much, more than necessary. It is worth to note that in many cases French pilot are described as great friends, and very pleasant people, this not referring at all to their combat abilities.
As an aside note, Lionel asked about GC II/1. Polish pilot of the unit, Marian Trzebiński noted, that the French officers avoided combat at all cost, and that it looked like within the month most Poles had more flights that those officers since the start of the war. He was nonetheless full of admiration for French NCO pilots, who led flights and squadrons instead of officers.
Certainly an extreme situation, nonetheless another piece of rather complicated situation in France.

[Rob Philips]

Thanks, Franek. So the matter is definitively significant. Could you summarize the multitude of possible explanations for this complicated matter? Perhaps, with input from others, this may lead to a well developed view on things.

Regards,

Rob

[Birgir Thorisson]

A few points about the officers of CG II/1, drawn from S. Joanne, Le Bloch MB 152.

The Poles joined on mai 19 1940, when the unit had suffered heavy losses.
Looking through the day-to-day accounts, I do not see the commandant (Robillon) leading the unit.
The leader of 3 escadrille, Capitaine Veniel is very active, especially before may 19th. He was wounded on june 3rd. Of the other 4 officers, 3 were wounded. S/L Fontaine is especially active.
Of the Sous-officiers, (8) 3 were killed before the Poles arrived.
4th Escadrille.
The leader (Capitaine Coiral) was killed on mai 14th. On the same day, 2 of the other 4 officers were wounded. Of the 8 "sous-officiers" 3 had been wounded before the poles joined.
One wonders what 50% attrition in one week does to a unit?

Birgir Thorisson

[Rob Philips]

Thanks, Ruy. I hope that I did not give the impression that I believe that these matters can be transformed into and resolved as mathematical equasions. It is just that mathematics, and I include the rules of logical inference, are tools to make clear what we want to say, in ways that can be checked by the other guy. I believe that a lot of the confusion, and even emotion, and even anger, that arose in this thread came about as a result of unclear language. Statements such as "many aviators", certainly when given in forceful ways, do not lead to a lot of clarity. Statements such as "20% of the aviators involved", and in as far as these statements can be demonstrated to be accurate, give the other guy a better chance to agree, or a better handle to disagree in a bizlike way. I could demonstrate that the extreme opinions vented in this thread are less opposing than believed by their proponents, who fell victim to their own lack of clarity, in other words to their own "method" or style. And that's a pity, leading to a lot of energy lost in the wrong directions.

Regards,

Rob