Talk:Mass killings under communist regimes

Due to the editing restrictions on this article, a sub-page has been created to serve as a collaborative workspace or dumping ground for additional article material.

Estimates Soviet Union

You cite R. J. Rummel's estimate of 110 million deaths due to communism. Rummel surveyed only English language sources published prior to 1990 to arrive at his estimates. In The case of the Soviet Union from 1927-53 he claims 42,672 million victims of communism, 2,000 million famine deaths in 1933 & 333,000 in 1947, and 26,125 million war/famine dead (1941-45), total unnatural deaths 71,330 million. The figures come from Lethal Politics by Rummel, I own a hard copy of the book.
In 1993 the Russian demographers Andreev, Darski and Kharkova (ADK) published a study of the Soviet population for the Russian Academy of Science, Naselenie Sovetskogo Soyuza : 1922-1991. The conclusions of their study are summarized in Demographic Trends and Patterns in the Soviet Union Before 1991. The source of the often cited statistic of 27 million Soviet war dead is this 1993 study by (ADK). In the 1993 study they listed the population balance from 1927-53, 1927 beginning balance 148,656 million; 149,965 million births; gain 1939-45 annexed territories 20,268 million; emigration out of country 822,000; total base population 318,067 million. Total deaths were 191,623 million, including war dead of 26.668 million and infant mortality of about 20 million, leaving population of 191,623 million at the end of 1953. In a 2001 article on the Russian website Demoscope (ADK) estimated 38 million unnatural deaths from 1933 53[19] 7.2 million famine dead in 1933; 2 million purge deaths in 1937-39, 27.2 million war dead 1939-45; 1 million famine dead in 1947 and 1 million repression deaths 1946-53. We have a difference of 33 million deaths between Rummel and ADK, in the case of Rosefielde who claims 45 million unnatural deaths (22 million victims of communism and 23 million war dead) the difference is 7 million. The difference in Rummel and Rosefielde can be attributed to forced labor in remote regions of the USSR by free Soviet citizens building socialism. The workers had Stakhanovite movement work quotas, only they strong survived, the weak perished.--Woogie 10w 19:13, 24 October 2018 (UTC)

Woogie, you are right, but we are mixing two categories: mass killings and unnatural deaths. Only a fraction of authors (e.g.. Rummel, Rosefielde and Valentino) equate them. Others disagree. that means your data are the maximal possible estimate.

In general, I would say, there is a huge problem with sources in the Estimate section.

1. Ref 30 cites Rummel's "Death by government" Rummel is known to use obsolete sources (see Woogie 10w), and his figures have been criticized.

2. a footnote "n" is an introduction to the Black Book (its figures have been widely criticised)

3. a footnote "o" is an opinion of Malia, this opinion has also been criticised, and it is not clear where these figures are taken from (looks like he just repeats what Courtois says). Obviously, he never did his own study, and some sources say he was just obsessed with the idea to get a magic figure of 100 million to advocate a very concrete idea: that Communism was much more murderous than Nazism.

4. a footnote "p" cites Valentino, but it cites it wrongly: Valentino makes a reference to Rummel (without endorsing this figure), and his own summary of the secondary sources available to him give different figures: 21 to 70.

5. a footnote "q" contains (i) a reference to Rummel (again); (ii) a reference to Courtois (again); (iii) a reference to Malia (again); (iv) a reference to Brzezinski (are you really sure he was doing his own estimates? He is not a historian, he was a politician; he definitely cites someone else's figures, probably Rummel's; (v) a reference to White's "Atlas of history" (which is definitely a tertiary source, which is not good; (vi) a reference to Culbertson, which Valentino seems to copypasted from Rummel's book without checking. The reference (both in the Wikipedia article and Valentino's book) is wrong. By copypasting a reference to a non-existing source (just try to google it), Valentino demonstrated his actual professional level (in my field, that mistake is unforgivable).

6. a reference 31 is a reference to Rummel's non peer-reviewed personal blog (interestingly, Rummel seems to take into account new data only when they increase his estimates; the new archival research that reevaluate the data to lower side are igniored by him)

7. a reference 32 is to Rosefielde, arguably, the only good expert in this panopticon. Unfortunately, he is a specialist in Soviet history only, and he is known to produce higher figures than anybody else in this field.

8. a reference "r" is White (this tertiary source is used again)

9. a reference "s" is to some non-peer-reviewed blog (although it, at least, mentions the names of the authors where the dat awere taken from, but most names are odd, whereas Brzesinski's name and Rummel's name have already been mentioned above;

10. a reference "t" and 33 is to a newspaper article that doesn't disclose sources. It is just Kotkin's opinion, but it is unclear where these data were taken from.

To summarise. The whole section is pure cheating: it presents not the best sources, provides obsolete, fake or unreliable data, and even worse, in a series of cross-references, it actually reproduces the same figures twice or trice to create a false impression of abundant studies in these area. If we remove cross-references and get rid of desperately obsolete sources and the sources that jus tangentially mention some figure (so we cannot talk about serious fact-checking and accuracy), the whole list shrinks to a couple of items. That is an indication of a simple fact that good source on this subject are desperately lacking. The explanation is simple: good scholars are not working within this paradigm.

Meanwhile, a lot of fresh and reliable sources (peer-reviewed articles published in top journals) exist that discuss the three major perpetrators of mass killings (Stalin's USST, Mao's China and Pol Pot's Cambodia) separately. According to Valentino, there were almost no mass killings in other countries, so I see absolutely no reason why cannot we present three different modern figures for each of the three regimes instead of presenting obsolete and questionable cumulated figures taken from obscure sources. That is a good example how an intrinsically flawed concept serves as a magnet for various garbage.

In addition, our neutrality policy advises us to avoid segregation of text to create an apparent hierarchy of data. That is exactly what we see in this section. Not only it uses poor and obsolete sources, it segregates criticism from the estimates. That means the policy has been violated twice here: by providing a lot of redundant sources (most of which cite the same data many times), the section gives undue weight to them, and, by creating an apparent hierarchy, the criticism of those figures is understated.

If we remove all sources that just re-group the figures obtained from a couple of authors who assembled them from the literature available to them (not just cite a figure like Kotkin or Brzesinsky), only few sources will remain: Rummel (widely cited in popular media), Courtois, and Valentino. Since the first two authors have been criticized for their figures, they must be supplemented with needed criticism (which should not be put to the bottom.

Moreover, Mann (an author who has been cited in this article as an originator of the term "classicide") says that he does not consider Great Leap famine as "classicide", and considers it as an unintentional consequence of mistakes made by Chinese leadership. It is neither mass killing nor classicide, according to him. Other authors also have reservations on what part of excess deaths under communist regimes should be considered killing. However, a Great Leap famine is responsible for about 50% of all "communism death toll", which means Mann's opinion has a direct relevance to the subject on this section.

In connection to that, I think, we need to change the "Estimates" section as following:

1. Per Valentino, we explain that majority of excess deaths in communist states occurred in China, USSR and Cambodia.

2. We provide recent estimates for each country, and split the numbers by categories: "famine", "civil war", "executions and political repressions".

3. We explain that different opinia exist among scholars if deaths resulted from mistakes and callous policy should be considered a killings (Mann vs Rummel vs Valentino vs Wheatcroft, etc). It is also should be explained that some authors discuss documented deaths, whereas others include population losses (from all causes, including birth deficit and migration).

4. We provide estimates of total figures (Rummel, Courtois, Valentino), as well as Rosefielde, and we explain all controvercies around these figures (for first two sources).

That would be a good and professional presentation of what various sources say on that account.--Paul Siebert (talk) 19:34, 24 October 2018 (UTC)

Paul you wrote In general, I would say, there is a huge problem with sources in the Estimate section. IMO the problem in a nutshell is that the editors are going to Google Books or the internet to dig up a single solitary statistic from a reliable academic source and then they spin a story around it. I suspect that they don't have a clue about how these statistics are derived or what they represent. I beg you on bended knees please read Population dynamics: consequences of regular and irregular changesby Andreev, Darski and Kharkova (ADK) in Demographic trends and patterns in the Soviet Union before 1991

At this point we are on the level of Fox News when they attempt to discredit Bernie Sanders and Socialism. [20] I would not be surprised if the hack on Fox News got his information from this page.--Woogie 10w 21:27, 24 October 2018 (UTC)

I am reading it currently. Meanwhile, can you please comment on my proposal (at the bottom of my post)?--Paul Siebert (talk) 21:53, 24 October 2018 (UTC)

@Woogie 10w: Can you please make a Wikipedia account? It is important for several reasons. RhinoMind (talk) 14:53, 27 October 2018 (UTC)

Wow, this sounds real serious [21]--Woogie 10w 21:38, 27 October 2018 (UTC)

It seems it was a technical problem with a signature. Actually, a user who made this post was Woogie10w (no space).--Paul Siebert (talk) 16:08, 27 October 2018 (UTC)

@Woogie10w: Thanks. Better try to ping the guy then. RhinoMind (talk) 16:42, 27 October 2018 (UTC)

Alexandra Laignel-Lavastine

I see that this author has been quoted twice in the "Terminology" section. I have three objections to that. First, the quote is taken out of context. These words are in the opening paragraph of the chapter, the author disagrees with these views, and the whole chapter is devoted to debunking these views. As another author summarises Alexandra Laignel-Lavastine's views, "in her opinion, postcommunist Romanian historiography had been captured by (both inter-war and national-communist) ideology." (the quote was taken from Shafir, I added the ref to the text). Second, Romania is not a leader in world historiorgaphy, so I am not sure why should we devote a space to what Romanian (Serbian, Panamian, etc) scholars think. Especially if this thinking "had been captured by (both inter-war and national-communist) ideology". Third, the authors speaks about Romania only, she mentions the rest of Eastern Europe in a previous sentence. I remove both quotes as misinterpreted, misleading and only marginally relevant.--Paul Siebert (talk) 20:35, 25 October 2018 (UTC)

According to Laignel-Lavastine, Romanian historiography is captured by ideology (p. 180-82). Therefore, the information about usage of the Red Holocaust by Romanian intellectuals is hardly helpful. If someone wants to restore the text removed by me, please, don't forget to add the above notion of Laignel-Lavastine (about ideologically captured Roman historiography).--Paul Siebert (talk) 21:23, 25 October 2018 (UTC)

Terminology section

The way this section is written is misleading and redundant. Usually, the goal of a "terminology" section is added at the beginning of monographs and articles to explain the terms that are extensively extensively in this particular book. An example is the Mann's monograph. In contrast, majority of the terms listed in the "Terminology" section are not used in the article (the term "democide" is used when Rummel's views are described). As a result, this section looks totally redundant.

The second problem is that the section makes a lot of undue generalisations. Thus, it creates an absolutely wrong impression that the term "genocide" or "classicide" are applied to all excess mortality events. (A short reservation ("in whole or part") is added at the beginning, but it is totally unclear). In reality, with exception of Rummel's "democide", which, by definition, is identical to "excess deaths", all other terms are applied (or not applied) selectively to each separate event. Thus, mainstream scholars do not apply the term "genocide" to famines as whole (neither in the USSR nor in China). There are two exceptions: a significant fraction of them sees the 1933 famine in some part of Ukraine as genocidal, and many scholars also consider the famine in Cambodia as a part of genocide, but almost no authors describe a famine in China as genocide. Furthermore, many scholars note that in East Asia a social division leads to separation on social groups that have some traits of ethnic groups, so any repression against a separate social group has some traits of genocide. That is specific for Asia, but not for communist states only. And so on, and so forth.

A third problem is that by listing the authors who calls MKucR as, e.g. "genocide", the section does not list those who thinks otherwise. Thus, Mann, Wheatcroft and many other scholars openly object to the attempts to characterise MKucR as genocide, and their opinion (and their rationale) should be presented, otherwise the section looks extremely biased.

A fourth problem is that the section does not explain why all those new terms are needed. Actually, the only goal this section currently serves to is to support the obscure statement in the lead. It is clear why was the term "genocide" proposed: it is a legal category (a crime), and its introduction allows prosecution of a certain type of mass killings. What other terms are needed for? The section does not describe that. Meanwhile, these explanations can be provided for "democide" (Rummel introduced it to create a mathematical model of global state violence), for "politicide" (Harff used it for her own model and the database), for "classicide" (Mann proposed it to contrast the events in communist states with genocide). However, the section does not explain that.

A fifth problem is that the section mixes scholarly terms ("genocide", "politicide", "mass killing") with polemical epithets like "Red Holocaust": this metaphor does not play any positive role in understanding of the events in communist states, it is not used by scholars (except one, who used it only in his popular book, not in peer-reviewed papers), and its usage is condemned as the Holocaust trivialisation. In addition, some terms are used almost exclusively by those authors who coined them, and some authors (e.g. Shaw) clearly say that 'ethnic cleansing', 'democide', urbicide', 'classicide' and the bizarre 'auto-genocide' and other labels do little to further our understanding. Therefore, to avoid discrediting Wikipedia, I propose to remove "Red Holocaust" (and probably "Crimes against humanity", with a reference to a single blog article with zero citations), and to explain that many authors are sceptical about invetion of various labels.

I propose to rewrite the section using the scheme that was used by Mann in his monograph. This book has a "Terminology" section, but, in contrast to this article, Mann discusses the meaning and applicability of a certain term to different instances of violence, and it groups them according to various degrees. Actually, this scheme works here too, because mass mortality (more precisely, "population losses", because most figures presented in this article are based on demographic evidences) in communist states had several components, from the excess deaths resulted from bad policy (for example, partially man made Chinese famine) to a Holocaust style mass murder (killing fields in Kampuchea). Overwhelming majority of authors (except Rummel, who calls everything "democide") apply different terms to each of those events, and we need to explain that.--Paul Siebert (talk) 04:51, 27 October 2018 (UTC)

I agree it does not belong. These terms should be introduced in the article where they arise. For example, when discussing the views of various suthors. For example, the article uses the term "Red Holocaust": 'In 2010, Steven Rosefielde stated that the "Red Holocaust" "still persists in North Korea...."' It would be better to explain what he means and how accepted his terminology is. Alternatively, we could substitute the term mass killings. TFD (talk) 01:23, 30 October 2018 (UTC)

Paul Siebert, this topic's terminology subtopic is unusual in this regard.

1&4) Rather than discussing terminology in order to define it for future reference in the same publication, sources discuss the terminology in a similar way to how this article's terminology section is used: to describe the variety of competing or alternative terms used by various academics. For some examples, see excerpts a, c, d, e, f, g, h, j, l, and m, reproduced below (it's likely you have already read these, Paul, but others may not have):

Excerpts

a) Krain 1997, pp. 331-332: "1. The literatures on state-sponsored mass murder and state terrorism have been plagued by definitional problems. Terms such as state-sponsored mass murder and state terrorism can be (and often are) easily confused and therefore need elaboration. The main difference between state-sponsored mass murder and state terrorism, for instance, is one of intentionality. The purpose behind policies of state-sponsored mass murder such as genocide or politicide is to eliminate an entire group (Gurr 1986, 67). The purpose behind policies of state terrorism is to "induce sharp fear and through that agency to effect a desired outcome in a conflict situation" (Gurr 1986, 46). The former requires mass killings to accomplish its goal. The latter's success is dependent on the persuasiveness of the fear tactics used. Mass killings may not be necessary to accomplish the particular goal." ... "2. Genocides are mass killings in which the victim group is defined by association with a particular communal group. Politicides are mass killings in which "victim groups are defined primarily in terms of their hierarchical position or political opposition to the regime and dominant groups" (Harff and Gurr 1988, 360). Interestingly, many of the instances coded by Harff and Gurr as "politicide" are considered by much of the literature to be instances of state terrorism (e.g., Argentina, Chile, El Salvador) (Lopez 1984, 63). Evidently there is some overlap between state terrorism and some kinds of state-sponsored mass murder." c) Karlsson & Schoenhals 2008: "‘Crimes against humanity’ is a linguistically and logically cumbersome term when the aim is to analyse physical violence perpetrated by individual groups, institutions and states against specific victim groups in their own country, which is essentially the case in the context of communist regimes’ crimes against humanity. In addition, it is not in keeping with the terms that have long been used by the academic community. Naturally, the work of creating an inventory includes examining the terms used in practice by researchers in their analyses, and it is reasonable to assume that every time, every society and every paradigm has its own terms to refer to the crimes of communist regimes. Nonetheless, it is possible to establish at this early stage that researchers have long used the word terror to describe the crimes of the Soviet communist regime, regardless of the framework of interpretation to which they adhere. Although the extent to which the mass operations and forced deportations of specific ethnic groups ordered by Stalin before and during the Second World War can be defined as genocide is debated, there is agreement among researchers that the term ‘terror’ is the best reflection of the development of violence in Bolshevik Russia and in the communist Soviet Union under Lenin and Stalin. As a result, terror will be the term most frequently used here in analysing the Soviet communist criminal history. On the other hand, the term terror is seldom used to describe the mass killings in Cambodia between 1975 and 1979, which may be because it is less clear that the actual intention and stated motive of the Khmer Rouge was to terrorise people into submission. The term genocide, however, is relatively widely accepted and established in describing the systematic and selective crimes of the communist regime in Cambodia, although the use of this term is not entirely uncontroversial. Therefore, in analysing the criminal history of Cambodia, this term will be used in precise contexts dealing with the killing of a category of people, whereas more neutral terms such as mass killing and massacre are used to refer to the general use of violence. The terminology used in the Chinese criminal history is dealt with in detail as part of the section on China." ... "In the Soviet case, as Klas-Göran Karlsson so rightly notes, there is an ‘established term’ for the crimes of the regime, namely ‘terror’ – and this is used almost regardless of the general frameworks of interpretation employed by individual researchers. In the same way, he notes that ‘the term genocide is established and accepted as a description of the crimes of the Khmer Rouge’. In the case of the People’s Republic of China, however, there are no equivalent terms that are accepted or generally established in the academic community and that can be made use of in a research inventory. Bibliographies and search engines all speak their own clear language: those who carried out research on Maoism in its day made very limited use of words such as terror and genocide, and neither do these terms appear among the key terms that carry implicit clear explanations and are therefore regularly used by current foreign and Chinese historians." d) Semelin 2009, p. 318: "'Classicide', in counterpoint to genocide, has a certain appeal, but it doesn't convey the fact that communist regimes, beyond their intention of destroying 'classes' - a difficult notion to grasp in itself (what exactly is a 'kulak'?) - end up making political suspicion a rule of government: even within the Party (and perhaps even mainly within the Party). The notion of 'fratricide' is probably more appropriate in this regard. That of 'politicide', which Ted Gurr and Barbara Harff suggest, remains the most intelligent, although it implies by contrast that 'genocide' is not 'political', which is debatable. These authors in effect explain that the aim of politicide is to impose total political domination over a group or a government. Its victims are defined by their position in the social hierarchy or their political opposition to the regime or this dominant group. Such an approach applies well to the political violence of communist powers and more particularly to Pol Pot's Democratic Kampuchea. The French historian Henri Locard in fact emphasises this, identifying with Gurr and Harff's approach in his work on Cambodia. However, the term 'politicide' has little currency among some researchers because it has no legal validity in international law. That is one reason why Jean-Louis Margolin tends to recognise what happened in Cambodia as 'genocide' because, as he points out, to speak of 'politicide' amounts to considering Pol Pot's crimes as less grave than those of Hitler. Again, the weight of justice interferes in the debate about concepts that, once again, argue strongly in favour of using the word genocide. But those so concerned about the issue of legal sanctions should also take into account another legal concept that is just as powerful, and better established: that of crime against humanity. In fact, legal scholars such as Antoine Garapon and David Boyle believe that the violence perpetrated by the Khmer Rouge is much more appropriately categorised under the heading of crime against humanity, even if genocidal tendencies can be identified, particularly against the Muslim minority. This accusation is just as serious as that of genocide (the latter moreover being sometimes considered as a subcategory of the former) and should thus be subject to equally severe sentences. I quite agree with these legal scholars, believing that the notion of 'crime against humanity' is generally better suited to the violence perpetrated by communist regimes, a viewpoint shared by Michael Mann." e) Weiss-Wendt 2008, p. 42: "The field of comparative genocide studies has grown beyond recognition over the past two decades, though more quantitatively than qualitatively. On the surface, everything looks good: the number of books on genocide has triple within less than a decade; the field of comparative genocide studies has its own professional association and journals; more and more colleges and universities offer courses on genocide; several research institutions dedicated to the study of genocide have been established. If we are talking numbers, comparative genocide studies are indeed a success. Upon closer examination, however, genocide scholarship is ridden with contradictions. There is barely any other field of study that enjoys so little consensus on defining principles such as definition of genocide, typology, application of a comparative method, and timeframe. Considering that scholars have always put stress on prevention of genocide, comparative genocide studies have been a failure. Paradoxically, nobody has attempted so far to assess the field of comparative genocide studies as a whole. This is one of the reasons why those who define themselves as genocide scholars have not been able to detect the situation of crisis." f) Williams 2008: "A vital element of the evolution of genocide studies is the increased attention devoted to the mass killing of groups not primarily defined by ethnic or religious identities. Most vulnerable minorities around the world had been so defined when Lemkin was crafting his genocide framework, and when UN member states were drafting the Genocide Convention. Such groups continued to be targeted in the post-Second World War period, as in East Pakistan/ Bangladesh in 1971, or Guatemala between 1978 and 1984. But it became increasingly apparent that political groups were on the receiving end of some of the worst campaigns of mass killing, such as the devastating assault on the Indonesian Communist Party in 1965—1966 (with half a million to one million killed), and the brutal campaigns by Latin American and Asian military regimes against perceived dissidents in the 1970s and 1980s. One result of this re-evaluation was that the mass killing by the Khmer Rouge regime in Cambodia between 1975 and 1978, previously ruled out as genocide or designated an 'auto-genocide' because most victims belonged to the same ethnic-Khmer group as their killers, came to be accepted as a classic instance of twentieth-century genocide. Detailed investigations were also launched into the hecatombs of casualties inflicted under Leninism and Stalinism in the post-revolutionary Soviet Union, and by Mao Zedong's communists in China. In both of these cases—and to some degree in Cambodia as well—the majority of deaths resulted not from direct execution, but from the infliction of 'conditions of life calculated to bring about [the] physical destruction' of a group, in the language of Article II(c) of the Genocide Convention. In particular, the devastating famines that struck the Ukraine and other minority regions of the USSR in the early 1930s, and the even greater death-toll—numbering tens of millions—caused by famine during Mao's 'Great Leap Forward' (1958—1962), were increasingly, though not uncontroversially, depicted as instances of mass killing underpinned by genocidal intent." g) Wayman & Tago 2009, pp. 1-17: "The two important scholars who have created datasets related to this are Rummel (1995) and Harff (2003). Harff (sometimes with Gurr) has studied what she terms 'genocide and politicide', defined to be genocide by killing as understood by the Genocide Convention plus the killing of a political or economic group (Harff & Gurr, 1988); the combined list of genocides is sometimes labeled 'geno-politicide' for short. Rummel (1994, 1995) has a very similar concept, 'democide', which includes such genocide and geno-politicide done by the government forces, plus other killing by government forces, such as random killing not targeted at a particular group. As Rummel (1995: 3-4) says, 'Cold-blooded government killing ... extends beyond genocide'; For example, 'shooting political opponents; or murdering by quota'. Hence, 'to cover all such murder as well as genocide and politicide, I use the concept democide. This is the intentional killing of people by government' (Rummel, 1995: 4). So Rummel has a broader concept than geno-politicide, but one that seems to include geno-politicide as a proper subset." h) Midlarsky 2005, pp. 22, 309, 310: "I distinguish between genocide as the systematic mass murder of people based on ethnoreligious identity, and politicide as the large-scale killing of designated enemies of the state based on socioeconomic or political criteria. Although genocide can be understood to be a species of politicide (but not the converse), in practice, genocidal (i.e., ethnoreligious) killings tap into much deeper historical roots of the human condition. In this distinction, I follow Harff and Gurr 1988, 360." ... "Turning to Cambodia, the mass killings in that country during Pol Pot's murderous regime are often characterized with other seemingly identical circumstances. Cambodia and Rwanda, for example, are typically treated as genocides that differ little from each other in essential characteristics. However, the victimization rates for the two countries are similar only when treated as proportions of the total country population systematically murdered. Although the mass murders in Cambodia are frequently characterized as genocide, I argue that in fact genocidal activity was only a small proportion of the killing and that the vast majority of Cambodians died in a politicide, substantially different in origin from the genocides we have been examining. The matter of etiology lies at the root of my distinction here, not definitional semantics. If we lump the Cambodian case other instances of systematized mass murder, then the sources of all of them become hopelessly muddled." ... "Essentially, I argue that genocides stem from a primitive identification of the "collective enemy" in Carl Schmitt's sense, whereas politicides, at least of the Cambodian variety, are attributable to more detailed ideological considerations. Further, the Cambodian case falls under the rubric of state killings, having a particular affinity with earlier practices in the Soviet Union and China. Indeed, an arc of Communist politicide can be traced from the western portions of the Soviet Union to China and on to Cambodia. Not all Communist states participated in extensive politicide, but the particular circumstances of Cambodia in 1975 lent themselves to the commission of systematic mass murder. Because an element of Cambodian state insecurity existed in this period, especially vis-à-vis Vietnam, a genocidal element is found in the killing of non-Khmer peoples such as the Vietnamese, who comprised a small proportion of the total." j) Staub 2011, p. 100: "In contrast to genocide, I see mass killing as 'killing (or in other ways destroying) members of a group without the intention to eliminate the whole group, or killing large numbers of people' without a focus on group membership." l) Wayman & Tago 2009, pp. 1-17: "Our term, 'mass killing', is used by Valentino (2004: 10), who aptly defines it as 'the intentional killing of a massive number of noncombatants'. The word 'noncombatants' distinguishes mass killing from battle-deaths in war, which occur as combatants fight against each other. The 'massive number' he selects as the threshold to mass killing is 'at least fifty thousand intentional deaths over the course of five or fewer years' (Valentino, 2004: 11-12), which of course averages to at least 10,000 killed per year." ... " One reason for selecting these thresholds of 10,000 and 1,000 deaths per year is that we find that in the Harff data on geno-politicide, which are one of our key datasets, there are many cases of over 10,000 killed per year, but also some in which between 1,000 and 10,000 are killed per year. Therefore, analyzing at a 1,000-death threshold (as well as the 10,000 threshold) insures the inclusion of all the Harff cases. Valentino chooses 50,000 over five years as 'to some extent arbitrary', but a 'relatively high threshold' to create high confidence that mass killing did occur and was deliberate, 'given the generally poor quality of the data available on civilian fatalities' (Valentino, 2004: 12). We believe that our similar results, when we lower the threshold to 1,000 killed per year, are an indication that the data in Harff and in Rummel remain reliable down even one power of ten below Valentino's 'relatively high' selected threshold, and we hope that, in that sense, our results can be seen as a friendly amendment to his work, and that they basically lend confidence, based on empirical statistical backing, for the conceptual direction which he elected to take." ... "Within that constant research design, we then showed that the differences were not due to threshold either (over 10,000 killed per year; over 1,000; or over 1). The only remaining difference is the measure of mass killing itself - democide vs. geno-politicide." m) Semelin 2009, p. 37: "Mann thus establishes a sort of parallel between racial enemies and class enemies, thereby contributing to the debates on comparisons between Nazism and communism. This theory has also been developed by some French historians such as Stéphane Courtois and Jean-Louis Margolin in The Black Book of Communism: they view class genocide as the equivalent to racial genocide. Mann however refuses to use the term 'genocide' to describe the crimes committed under communism. He prefers the terms 'fratricide' and 'classicide', a word he coined to refer to intentional mass killings of entire social classes."

The role of the terminology section in this article is two-fold: to reflect the reliable sources, per WP:BALANCE and WP:IMPARTIAL, and to inform readers of the other terms used. This second reason is particularly important because we are using a descriptive title readers may not be looking for (which was necessary due to the lack of a consensus among the sources on the best term to use).

2) This sounds like something you could clarify by adding additional sourced statements. Nothing is perfect, and the article will always be a work in progress. But please edit the section in a way that does not prioritize one reliable source over another. Where there are disputes among the sources, we should describe them.

3) Those who think otherwise would be included for the term they use, we should not list all the sources that do not use a term as part of that term's text.

5) Wikipedia's standard for reliable sources does not restrict itself to academic sources and that is the standard for inclusion that we must follow.

6) We should avoid trying to synthesize this variety into something more coherent than it is among the reliable sources. Using Mann's book as the basis for the section would give undue weight to his opinions on which terms are appropriate and which ones are not. Since there is no consensus among the sources on the best term, it is most neutral to include them all. We should, of course, include any criticism that has been published in reliable sources as well. I don't think we can say what they "overwhelming majority" say without something solid to back that up. My reading has left me with the impression that there is no consensus for the best term, and Valentino says that explicitly (excerpt b: Valentino 2005, p. 9: "Mass killing and Genocide. No generally accepted terminology exists to describe the intentional killing of large numbers of noncombatants."). AmateurEditor (talk) 03:14, 30 October 2018 (UTC)

How does it help readers to explain the term "red holocaust" and say it is rarely used? It seems self-explanatory anyway, hence needs no explanation. In fact it isn't even defined, which is unusual for a terminology section. TFD (talk) 03:53, 30 October 2018 (UTC)

@AmateurEditor. Re 5. Yes, Wikipedia standards do not prohibit usage of non-academic sources. However, this section tells about scholarly terms. We must clearly separate terminology from allegory (the words that are used to explain from the words that are used to condemn). Thus, introduction of the word "genocide" or "politicide" helps scholar to understand something about the events these words describe. In contrast, the words like "Red Holocaust" carry just an emotional load. I suggest to remove all non-scholaraly sources: if some journalist calls MKucR "Red Cannibalism" it can hardly be considered a scholarly term (although meets WP:V). In connection to that, "Red Holocaust" should be removed (the only scholarly sources that discuss it condemn its usage). "Crimes against humanity under communist regimes" was introduced in the review published in some blog (with no citations in google scholar), so it also should be removed. Other non-scholarly sources should be removed too. All mentions of state organisations or NGOs should be removed too: let's limit with terminology sensu stricto. the terms "Crimes against humanity" or "Red Holocaust" belongs to the section about the attempt to condemn the crimes of communist regimes.

I agree we should avoid synthesis. To this end, it should be clearly explained that overwhelming majority of authors cited it this section do not focus on MKucR. Actually, they even do not attempt to propose any comprehensive term for MKucR, because do not see MKucR as a single category.

Thus, Wheatcroft speaks only about Soviet Union. However, the term "repressions" he (as well as all historians of Stalinism) does not even refer to mass killings only, and different scholars applied it to different parts of mass killings. Per Ellman, the term "repressions" is very vague, it includes both lethal and non-lethal actions. Wheatcroft does put collectivisation and dekulakisation deaths (especially, famine) into the "mass killings" category, however, in his view, repressions victims include survived Gulag prisoners.

Another example is Harff. Her global database of genocides and politicides does not include majority of the events Valentino sees as "Communist mass killing" (it starts from 1955, so it does not cover USSR, however, Chinese famine is not included, and that reflects majority views). Exclusion of Chinese famine only immediately decreases the scale of describing events by more than a half.

Obviously, if we decided to explain what terms was used for MKucR, we also have to list the events that are beyond the scope of the proposed terminology, and to tell the reason of that. We also should explain the views of those who think otherwise.

In connection to that, the section should clearly explain who exactly applied which term to which event. Only few authors tried to apply "genocide" to all MKucR, and they should be listed. Others applied it only to Cambodia (the least controversial case), many author applied it to the 1933 famine in some parts of Ukraine (it is known there as Holodomor). All other events are called "genocide" very rarely, and we must explain that.

Semelin's, Mann's etc theorisings should be explained, however, the terminology used by country experts must be described too. Thus, I think it is necessary to make Cambodia, PRC and USSR sections, and explain what terms are used by numerous historians working in these fields. If different terms are used for different events, that must be explained.

That is exactly what we need to get rid of synthesis (which is currently present in this section).--Paul Siebert (talk) 16:02, 30 October 2018 (UTC)

The section is not restricted to scholarly sources any more than the article as a whole is. About the hypothetical example "Red Cannibalism", lets cross that bridge when we come to it. Polemical terms can be notable as well. In fact, they sometimes become the commonly used term for a topic (such as Boston Massacre). I suspect that some oddball journalist using the term would not be considered a reliable source, but if he were, then it actually ought to be included, along with any criticism from other reliable sources that it has received. The use of "Holocaust" for this topic may be an example of this already, but it has definitely been used in a significant way by multiple sources, including academics, and is tied to an academic dispute about the "double genocide theory", apparently. It is appropriate that the criticism is also included and appropriately referenced. I don't know what blog you are referring to about "Crimes against humanity under communist regimes", but there are academics who prefer that term among the current references. If you believe that this is a fringe topic that the majority of scholars reject, then that can be added to the article only if you can support it with direct reference to reliable sources. We cannot assume to know why such-and-such scholar chooses not to study or address any particular topic. Please note that Michael Mann, who you wanted to use to structure things, says (excerpt ab) "All accounts of 20th-century mass murder include the Communist regimes. Some call their deeds genocide, though I shall not." This strongly suggests that the topic is mainstream, rather than fringe, but you are free to present alternate sources. Those alternate sources, however, should positively state your conclusion so that we do not have to interpret why they choose not to say something, which would be original research on our part. I agree that the individuals who used the different terms and how should be explicit in the section. About the terminology used by country experts being included in country-specific sections, I have no objection in principle. AmateurEditor (talk) 02:55, 31 October 2018 (UTC)

AmateurEditor, I disagree. The very name of the section implies some academism. That means each term introduced in this section is supposed to increase our understanding of the subject. Any terms that are used for polemical, political or similar purposes must be clearly separated from scholarly terms. If some journalist or political organisation introduces some word, we must clearly explain the context. For example, if the term is proposed by an anti-communist political organisation, that should be explained clearly. In addition, we should separate universally accepted terms and the terms that are not used widely. Thus, Fein writes about "comminist genocides and 'democide'" (in quotation marks) implying that "democide" has not become an established scholarly term, and is used mostly by Rummel himself.

Second, the whole section is a big logical fallacy. It names the terms, but it does not explain to which events has this term been applied, thereby creating an absolutely false impression that these scholars are trying to propose a better term for Valentino's 70+ million victims. In actuality, this is a worst example of lye. For example, the text says: "Professor Barbara Harff studies "genocide and politicide", sometimes shortened as geno-politicide, in order to include the killing of political, economic, ethnic, and cultural groups," which is correct. However, if one takes a look at the list of Harff's "genocides and politicides", it is easy to see that Harff studies mostly non-communist politicides. In other words, the section creates an absolutely false impression that Harff was trying to propose some terminology of MKucR, although it is absolutely clear she is not separating the events in communist states from other geno/politicides, and the events in communist countries are not the focus of her study. This is just one example of cherry-picking that is intended to mislead a reader by creating an absolutely false impression that the subject (MKucR) is seen by genocide scholars as a separate and notable topic. (I believe, we agree that country experts do not see it as a separate topic).--Paul Siebert (talk) 04:20, 31 October 2018 (UTC)

The quote from Mann is somewhat misleading. He actually states that some communist regimes committed some number of mass murders that are considered as genocide. This statement is absolutely indisputable, because we know at least about one communist mass killing that is recognized as genocide almost universally ("there is at least one sheep in New Zealand that is black at least from one side"). Moreover, I can agree that there was some connection between some mass killings in different communist states, and it might be correct to analyze them in connection to each other. However, that does not mean that each author cited in the Terminology section speaks about the same events. Thus, Rummel claims all excess deaths in the USSR was "democide", and it includes some 6 million deaths in post-Stalin USSR (which is totally ridiculous: his calculations are based on some obsolete data on prisoner population multiplied by annual mortality, which was a product of his own speculations; the fact that several million people were killed in the USSR after 1953 is a totally fringe view, but this figure is implicitly present in this article as a part of Rummel's estimates). Anyway, the article cites Rummel's view that there was 60 million "democide" in the USSR. Then it cites Mann's view that that was not genocide but "classicide". Formally, both statements are correct and attributable to reliable source. However, these two authors are not talking about the same event. Rummel speaks about some unstopping killing machine that was continuously destroying the population of its own country during a 70 years period. Mann speaks about some separate events that happened during some separate periods of time and had some limited scale. Although Mann cites the figure of 6-7 million victims of 1933 famine, he attributes it to mistakes of the Stalinist policy, and does not apply the term "classicide" or "genocide" (or other "-side") to it. He definitely considers famines, as well as some other deaths, as a separate topic (I mean, separate from "classicide"), but the article doesn't make it clear. The article directly misleads people. By the way, the very title of the Mann's book implies that Mann sees mass killings (he is mostly focused on ethnic violence) as a result of democratic transformations that occurred globally in XX century. Interestingly, this article prefers comfortably ignore this thesis. Instead, selectively cited Mann is used to support views of Rummel. If it is not a lie than what is a lie?--Paul Siebert (talk) 15:11, 31 October 2018 (UTC)

The only solution would be to clearly explain which term was applied to what event and by whom. --Paul Siebert (talk) 15:43, 31 October 2018 (UTC)

Actually the solution is status quo ante and not worry about the exceedingly long posts on the talk page. Once the post gets too long, it is clear it will not get consensus support, historically. Collect (talk) 17:45, 31 October 2018 (UTC)

Collect, this is just a part of a long discussion between AmateurEditor and me that started on his talk page. We are discussing possible ways to improve the article because many sources are represented incorrectly and some are ignored almost completely. I encourage you to read in full it before making comments of that kind.

In addition to that, please, keep in mind that we are discussing several questions in parallel. The post does not look exceedingly long if one focuses at one question and (temporarily) ignores others. Do you have any specific comment on some separate questions raised in our discussion?.--Paul Siebert (talk) 17:52, 31 October 2018 (UTC)

Paul Siebert, as I understand you, you see troubling implications that are hinted at and go unrefuted in the article but you don't feel you can fix them on your own (presumably this can't be resolved by changing the name of the Terminology section). If so, then you are right to bring it here for discussion, but your language to describe the issues is full of vague value judgments and unfalsifiable interpretations ("...implies...", "...polemical, political or similar purposes...", "...should be explained clearly...", "...universally accepted...", "... not used widely...", "...implying...", "...established scholarly term...", "...creating an absolutely false impression...", "...worst example of lie...", "...intended to mislead a reader...", "...somewhat misleading...", "...the very title of the Mann's book implies..."). Part of this I think could be solved by you adding what balance or additional clarifying statements you believe to be missing. As long as you do so with solid references to reliable sources (that rely on a literal reading of those references), I don't think anyone would have a problem with that. I certainly would not. Beyond that general point, there are a couple more specific responses I want to make.

1) You said "I believe, we agree that country experts do not see it as a separate topic". No, I repeatedly stated in our discussion on my talk page that we cannot assume to know what people think outside of what they explicitly say, and I stand by that. I do not agree that, because a scholar chooses to study an event in a single country without reference to the wider phenomenon, he rejects the idea of the wider phenomenon. As I said before, another explanation is that the wider phenomenon is simply outside the scope of his interest.

2) About different authors including different events, this is primarily related to the famine debate and we have an entire section of the article about that already, as we should. Different sources/terms do not all have to include all famine deaths or all exclude those deaths in order for them to all be discussing/be applied to the same topic.

3) About this being a continuation of the discussion on my talk page, I thought that was primarily about the Estimates section and your proposal to include a table of estimates from more specialized single country sources. AmateurEditor (talk) 03:40, 1 November 2018 (UTC)

Re 1. Yes, sometimes, the conclusions that we make based on what the sources say are pretty obvious, but we cannot write that, because the sources do not say that directly. Thus, Wheatcroft, Ellman, Davis and Getty are focused on Stalinist repressions, they never mention Communist ideology and see no parallelism with other communist states, however, this fact is insufficient to write in the article that they do not support ideas of Courtois (although they definitely don't). Moreover, whereas it would be absolutely correct to conclude that they ignore Courtois, we cannot write even that: they really ignore Courtois, which means they do not write about him at all. Note, I am perfectly aware of this problem.

However, it seems you don't understand that it would be equally incorrect organise the article in the way that may lead a reader to wrong conclusion. Thus, no evidences have been provided that the views expressed by Courtois, Rummel or Valentino are shared by numerous country experts, so we must do our best to avoid creating an absolutely false impression that country experts use theories of Rummel, Valentino, Malia or Courtois in their work, and that their explanations of these events are based on these theories. In reality, they propose their own explanations, and they are totally different from what Courtois or Rummel say.

This problem can be solved by choosing correct article's structure. These two examples demonstrate bad and good way to arganise text. The bad version is as follows:

Rummel proposed a concept of "democide" that was specific to all totalitarian states and had roots in ideology and concentration of power. "Democide" killed people through direct executions, murder, famine, disease, overwork.

Wheatcroft described mass killings in the USSR as "repressions", whereas Maksudov speaks about "demographic catastrophes".

Kiernan notes that Cambodian genocide had roots in Cambodia's colonial past

Vincent and Dulic note that Rummel's statistical approach has serious flaws."

From that text, a reader may conclude that Rummel proposed some general theory that has been accepted by Wheatcroft, Maksudov and Kiernan, who are working within that paradigm and identified some additional specific features in their countries of interest. Some authors contest Rummel's conclusions because they found some flaws in Rummel's methodology, but they represent minority views. That is a conclusion a reader draws from this text. However, we have absolutely no ground to claim Wheatcroft, Maksudov and Kiernan are working within the paradigm that was set by Rummel. They never wrote about that. In other words, this text is deeply misleading, although each separate statement is correct. However, if we reorganise this text in this way, the problem is resolved. An adequate text contains the same statements organised in different order:

Wheatcroft described mass killings in the USSR as "repressions", whereas Maksudov speaks about "demographic catastrophes".

Kiernan notes that Cambodian genocide had roots in Cambodia's colonial past

Rummel proposed a concept of "democide" that was specific to all totalitarian states and had roots in ideology and concentration of power. "Democide" killed people through direct executions, murder, famine, disease, overwork.

Vincent and Dulic note that Rummel's statistical approach has serious flaws."

See, everything becomes absolutely correct: yes, country experts are studying their countries of interest, and they propose their own explanations. Rummel tries to do some generalisations (and the text does not create an impression that country experts are working under auspices of Rummel's theory), and the criticism of his theory has been duly represented. That is exactly what happens in reality. (Of course, this text is very primitive, it describes a very oversimplified situation, I deliberately did that to demonstrate my point)

I believe I have been able to explain you my point.--Paul Siebert (talk) 04:22, 1 November 2018 (UTC)

Re 2. No, it has relation not only to famines (and the very title of the "Famine debates" section is misleading: there is only one famine (Holodomor) that is a subject of debates; all other famines (Volga famine, post WWII Soviet famine, Chinese famine) are not considered mass killing or genocide by overwhelming majority of scholars, and Kampuchean famine is considered a part of genocide). Thus, civil wars are not considered mass killings/-cides by most authors. Prison camp mortality and forced labor also are not considered as mass killings.

Again, the descriptive approach (see above) will allow us to avoid unjustified generalisations. Like you, I am also want to avoid original research, but I also keep in mind the problem of giving undue weight to few theories proposed by a small fraction of authors, which are not shared by majority of scholars.--Paul Siebert (talk) 04:36, 1 November 2018 (UTC)

The same can be said more briefly: the works of the authors writing about specific subjects can be presented in a context of more general theories only when these theories are universally accepted by country experts, or these authors make explicit reference to these general theories; otherwise, the works of country experts and general theorists should be presented separately: the text must clearly separate the facts and conclusions presented by country experts from attempted generalisations made by genocide scholars.--Paul Siebert (talk) 16:41, 1 November 2018 (UTC)

1) "See, everything becomes absolutely correct: yes, country experts are studying their countries of interest, and they propose their own explanations. Rummel tries to do some generalisations..." But in your example of the correct way to do things (granting that it is just a rough example), you are not stating Wheatcroft's and Maksudov's explanations, you are stating terms they use. Then you switch subjects to an explanation from Kiernan, then you switch back to a term from Rummel that you believe to be an explanation (but isn't). Then you switch to technical criticism of Rummel's statistics. Being more critical of this example, I would tell you that Wheatcroft uses "mass killing" in his publications, not just "repressions" (which includes mass killing), Maksudov's term "demographic catastrophes" appears to be a one-off turn of phrase, rather than any kind of term he uses repeatedly to define something, your definition of "democide" is wrong (it is just murder by a government of any kind, see here, and not a theory of totalitarianism), and Vincent and Dulic's note about statistics methods is only relevant to estimates, not to terms or theories.

I understand you wanting to put the most accurate sources first (although if we knew which were most accurate, why would we bother with any others to begin with), but Wikipedia has no basis for determining accuracy (and neither do we), it only cares about verifiability of content from what it defines as reliable sources and the popularity of that content among the reliable sources. It makes some sense that the country-specific specialists would have the most precise or perhaps the most up-to-date numbers (ignoring that they can and do disagree with each other and that any publication can become out of date over time), but it does not make sense to use country-specific specialists for explanations when they don't actually have them (Wheatcroft seems focused on what happened, i.e. the numbers, not why) or putting them ahead of reliable sources that directly speak to the article's topic in whole, especially when the broader reliable sources are themselves based upon country-specific sources already. You appear to be prioritizing these academic sources because they resonate with what you believe to be True, rather than them being more mainstream or popular in academic circles. Without an evident consensus among the sources, which is the situation we are in, we need an objective way to organize them that does not depend on our own personal preferences. Chronological order is the obvious answer, which is what the article tries to do currently. This is not an objection to including single-country sources, but it is an objection to putting them first because they are single-country sources.

2) "...civil wars are not considered mass killings/-cides by most authors. Prison camp mortality and forced labor also are not considered as mass killings." I believe that all the estimates included here are of the murder/manslaughter type (that is, intentional killing of noncombatants directly or intentional killing of noncombatants indirectly through conditions that would reasonably result in death), but correct me if I am wrong. About the famine section being misleading because it is or should be only about the Holodomor, that is not true. The section right now contains sourced statements of wider debates that include other famines as well.

3) Weren't you going to try adding a table of single-country source estimates? AmateurEditor (talk) 05:03, 2 November 2018 (UTC)

Re 3. Yes, but it may take a significant amount of work, and I would like to make sure I will not be reverted. I already did that once (it was not a table, but something of that kind), but the text was mauled and moved elsewhere.

Re 2. If you are talking about Courtois, he includes not only famines, he also attributes all civil war victims (from both sides) to Communism. Rummel includes de facto all excess death resulted from government's policy (e.g. resulted from good faith strategic blunders) to the "democide" death toll. Look at the Harff's "genocides and politicides" table and tell me what part of Courtois' "Communist death toll" is included there? Actually, all authors who calculated a total "Communist death toll" (Courtois, Rummel, Rosefielde) do that assuming all famines, camp mortality, etc were the result of a calculated policy of the regimes. Meanwhile, the authors who consider, e.g. Chinese famine partially man made but non-intentional do not bother to calculate their own "communist death toll" figures, because they consider the very subject non-existing. The exception is Valentino, who provides a range. However, even he does not explain the origin of the difference between 20 and 70 million: the actual difference is not because the figures are not available but because different authors interpret figures differently. Thus, a current consensus is 15-30 million died during the famine. However, there is no consensus that the famine victims were the victims of "Communist mass killings". Just read the specialised articles, and you will see that country experts put no more responsibility on Mao for the Great famine than on Churchill for the Bengal famine.

Re 1. AmateurEditor, my example is totally artificial, that is not a draft of the text, so your main criticism is both correct and irrelevant. What is important is your misunderstanding of the main point. You write "I understand you wanting to put the most accurate sources first ", and this your understanding of my words is wrong. The problem (and possible solution) is as follows:

We have two groups of authors. The group A studies individual countries, they produce data, provides their own explanations of the events in each country (separately), and they totally ignore the authors from the group B. The group B performs no study of each separate country, they use the works of the group A authors, and they make their own general statements about the subject as whole. If we describe the general views of the group B authors first, and then provide country-specific details from the works authored by the group A scholars, we create an absolutely wrong impression that the group A is working within a paradigm defined by the group B. However, if we present the views of the group A first and add later that some authors (group B) proposed some generalisation, this description would be correct, and a reader would not be mislead.

--Paul Siebert (talk) 14:48, 2 November 2018 (UTC)

Just one more thing

I have taken the time to read the article. Good job, the editors have created a reliable and credible article based on sources in the English language. As Columbo would say just one more thing,the Russian government today does not like to see this topic discussed. Best keep an eye on the article, ballbusters in Russia might try to gangbang it. --Woogie 10w 22:17, 29 October 2018 (UTC)

Just one more thing, the mass killing figures are hypothetical, the official vital statistics in the Soviet Union (1927-53) for live births and natural deaths are fragmentary. Historians push a POV by manipulating the estimated vital statistics. I disagree when you say "peer reviewed" because only demographers or others who specialize in numerical analysis should be considered peers capable of reviewing numerical data. Grover Furr and Rudolph Rummel are academic sources that push a political POV. They are both numerical illiterates who disregard and dismiss demographic analysis. The mass killings article should include a section on demographic analysis. I just recieved Demographic trends and patterns in the Soviet Union before 1991, edited by Wolfgang Lutz, Sergei Scherbov, and Andrei Volkov. This is a reliable source that I am reading. --Woogie 10w 12:53, 2 November 2018 (UTC)

Woogie, you will be surprised, but the data published by Rummel have not passed a peer-review procedure. That means you just confirm what I say: peer-review filter is working fine.

Actually, I disagree with your words about Rummel. He was a good researcher, but his contribution is greatly misunderstood and misinterpreted. He never tried to produce accurate data, and he had never be interested in going into details. His approach can be summarised as follows:

"We need to know how many people were killed by each regime to perform statistical analysis and find correlations between regime's type and violence. To do that, we need a reliable statistics. Let's take all available data, define all excess domestic deaths as "democide" and calculate a median value of deaths caused by each regime based on the all sources available. We know some sources are unreliable, but, assuming that they are equally unreliable for each regime, our approach can provide us with a satisfactory data set for our calculations."

This approach was absolutely correct, and the introduction of factor analysis in the study of violence was a significant step in our understanding of that subject. However, when we are using Rummel's data we must keep in mind their limitations. Their accuracy strongly depends on the initial data set. Thus, his estimates of Cambodia deaths were astonishingly accurate, but his estimates of deaths in the USSR is a total bullshit. The reason is that accurate data for Cambodia were available to Rummel already in late 70s, whereas only rough estimates for the USSR were available by that time.

The only problems with Rummel are: (i) his refusal to take into account fresh archival data for the USSR; (ii) his ahistorism (he attributes deaths only to the current regime and ignores a historical context; (iii) his obsession with statistics (correlation does not mean causation).

To summarise, Rummel can and should be used in the article, but his views should be represented duly, without giving too much weight.--Paul Siebert (talk) 14:25, 2 November 2018 (UTC)

Paul, peer-review procedure by whom? other numerical illiterates. Synder and Wheatcroft cherrypick statistics from various sources that they do not tie into a population analysis,it is amazing that their estimates are regurgitated without question. ADK in Russia however does provide a detailed credible population analysis. At least Rosefielde provides a detailed explanation of how he derived his figures. Rummel took numerous English language sources and came up with a range of losses from high to low, he estimated a mid point which he hangs his hat on. We can see in black and white Rummels sources, his analysis tells us that the widely cited "reliable sources" of the cold war era are as soft as shit. Wheatcroft for example claims 5.7 million famine dead based on "the Lormier corrections" He does not tell us that Lormier maintained that his 1946 population model was hypothetical. Just one more thing if your broker sent you a statement with a beginning and ending balance with no listing of the detailed activity, what would you do? I would confront the guy and say Ehi,where's the rest? [22] --Woogie 10w 17:20, 2 November 2018 (UTC)

Woogie, I am not sure I fully understand you. I agree with your opinion about Rummel, although the problem with his estimates not only in the Cold era sources, but because in a situation when a median is taken from the data that are limited by zero at the low side and have no limit at the high side will inevitable lead to the final figure that is skewed to a high side. That means not only the data are lousy, the very procedure id flawed.

However, I do not understand your criticism of Wheatcroft or Snyder. They separate killing from population losses. For example, both Rosefielde and Maksudov independently came to a conclusion that there was a serious demographic catastrophe in 1990s Russia. Does it fall into a "democratic mass killing" category? They don't say that. Hovewer, Maksudov applies the term "demographic catastrophe" to 1990s population decline, to 1932-33, to 1941-47. In other words, we must clearly explain that population losses and mass killings are seen as two different categories by majority of scholars.

In other words, I think Snyder or Wheatcroft do not cherry-pick statistics, they select those data that seem relevant to the subject of their study.

By the way, you criticise them all, but who, in your opinion, is a good source for population dynamics? ADK?--Paul Siebert (talk) 18:26, 2 November 2018 (UTC)

Paul please do not misunderstand me, I am not trolling here. My remarks off the cuff can be a bit flippant at times, for that I apologize. My objective is to made editors think in terms of a total population base and how the human losses are allocated. Snyder and Wheatcroft pontificate and expect that readers accept their statistics without question. We need first to understand the population dynamics of the Soviet Union before we judge the data in Snyder and Wheatcroft. Maksudov and ADK however provide a demographic overview and then discuss the unnatural deaths due to famines, repression and the war. As editors we use only reliable sources, Snyder and Wheatcroft are reliable but we can contrast them with Maksudov and ADK who analyze the unnatural deaths in the context of the entire population. --Woogie 10w 19:57, 2 November 2018 (UTC)

Maksudov [23]----ADK [24] Google translate does a decent job for us. --Woogie 10w 20:03, 2 November 2018 (UTC)

I think I understand your point well: you say that demographic data set a high limit for mass killing estimates, so any Rummel's claim about 60+ million "democide" in the USSR is not confirmed by modern demographical data. With regard to ADK&Maksudov vs Wheatcroft&Snyder, I don't see much contradiction between them when we talk about, e.g. the 1933 famine. If I understand correct the population losses were about 5 million, according to ADK, right?--Paul Siebert (talk) 20:36, 2 November 2018 (UTC)

Interesting, ADK cite UNO's prognosis for life prospective life expectancy in USSR for 2005-7 (it was made in 1988, noone expected USSR would dissolve in 3 years). They predicted life expectancy would be 70.4/78.2 (World Population Prospcets. 1988. N.Y. I9H9, V. 555). I am just wondering if modern Marxists tried to attribute a decline of life expectancy in post-Soviet states (which is far lower that these UNO figures) to democratic transformations? Can we consider them "mass killing under capitalist regimes"? Of course, I am not serious, I just want to demonstrate how ridiculous some claims of Rummel&Co are.--Paul Siebert (talk) 20:36, 2 November 2018 (UTC)

I think this piece of text should be removed from the article:

"R.J. Rummel wrote in 1993 that "Even were we to have total access to all communist archives we still would not be able to calculate precisely how many the communists murdered. Consider that even in spite of the archival statistics and detailed reports of survivors, the best experts still disagree by over 40 percent on the total number of Jews killed by the Nazis. We cannot expect near this accuracy for the victims of communism. We can, however, get a probable order of magnitude and a relative approximation of these deaths within a most likely range." "

This statement belongs to the author who never did any archival research and worked with secondary sources only. He was not a demographer either. That means this is his personal opinion, and it is relevant only to the discussion of his own estimates. Any objections?--Paul Siebert (talk) 20:50, 2 November 2018 (UTC)

WorldNetDaily Source

I removed a source published in WorldNetDaily, as I understand WND to be a notoriously unreliable source. This was reverted with the summary "he meets the standard for reliable sources." Please discuss. –dlthewave ☎ 11:26, 2 November 2018 (UTC)

I posted the question on WP:RSN.--Paul Siebert (talk) 16:03, 2 November 2018 (UTC)