User:Jitse Niesen/Notes

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Some notes for the ArbCom case involving Carl Hewitt (evidence, workshop). No attempt to NPOV is made.

Ruud's RfC mentions`Actor model, Actor model and process calculi, Actor model early history, Actor model implementation, Actor model, mathematical logic, and physics, Actor model theory, Arbiter (electronics), Carl Hewitt, Denotational semantics, Model (abstract), Category:Mathematical model, Planner programming language, Premature, Prematurity, Scientific Community Metaphor, Unbounded nondeterminism, Scheme programming language and Lambda calculus.

Timeline[edit]

User talk:CarlHewitt[edit]

  • 17 Jun: User:Chalst asks User:CarlHewitt whether he is Carl Hewitt.
  • 17 Jun: User:Macrakis asks about the extensive links created in "Feyerabend, Popper, philosophy of science etc."
  • 27 Jun: CSTAR writes: "I'm a litle puzzled by your adding a link to the actor model in the uncertainty principle page."
  • 4 Jul: User:Stan Shebs welcomes Carl and writes: "I hope you've not been too put off by some of your initial experiences."
  • 5 Jul: User:Koffieyahoo writes: "To keep wikipedia articles uncluttered please refrain from adding the actor model to every page that is relvant for the actor model." Some discussion following, spilling over to User talk:Koffieyahoo/Archive 1, where Koffieyahoo writes: "All this is motivated in the summary. Your changes somehow have a feel of vanity and non-NPOV. If you refer to the actor model, you should also refer to any other model of computation which might be slightly relevant, which you don't do, and besides this is why the category structure and the what links here exist."
  • 6 Jul: CSTAR refers to a discussion on his talk page.
  • 8 Jul: Discussion with Koffieyahoo continues: "I [Koffieyahoo] tried adding some constructive comments and questions on the talk pages of the following articles that might interest you: Actor model, Concurrent programming language, Parallel programming, and Lambda calculus." He also mentions microprocessors, X86 and quantum indeterminacy. Carl's replies on his talk page are evasive.
  • 10 Jul: User:82.69.39.138 (signing with Blissex) writes, refering to the assumption that User:CarlHewitt is Carl Hewitt: "I guess that's ill-founded, as several of things written in the Actor model pages are utterly ridiculous … The other possibility is that the real Carl Hewitt is retroactively trying to redefine much of his research in order to claim after-the-fact victory (because unfortunately much of his research as originally defined has had zero impact on the so called real world, which has gone down a rather worse path), and he should not get away with it." Carl replies: "Unfortunately Blissex you have it all wrong. I suggest that you start your own article with your own personal "Alternative view"." CSTAR says that the identity does not really matter, but suggests to "For instance remove references to yourself and colleagues in the intro to the Actor model … it seems too much like self-promotion, unfortunately"
  • 11 Jul: Koffieyahoo complains that Carl hasn't answered his questions.
  • 12 Jul: Conversation with Blissex continues; fairly nasty stuff, Blissex mentions references and quotes from them so he's not ignorant.
  • 13 Jul: Koffieyahoo complains again that Carl hasn't really answered his questions, even though Carl said he did.
  • 15 Jul: Koffieyahoo complains again that Carl hasn't answered his questions.
  • 18 Jul: User:Pavel Vozenilek asks about concurrency languages. Carl disagrees with Pavel's point.
  • 20 Jul: Koffieyahoo writes: "From my perspective you have either not answered or only gave evasive answers wrt my questions and comments. Can you please explain this behaviour?" Carl replies, but not to Koffieyahoo's satisfaction.
  • 8 Aug: Carl advertises a talk he is going to give on Wikipedia on User talk:CarlHewitt.
  • 12 Aug: CSTAR writes: "Before you make any sweeping generalizations about how academia can contribute to WP, please study the following reference to the history of a deleted page: [1]." Not sure what this leads to.
  • 20 Aug: User:Lotsofissues writes: "Overall, it is unwise to mention Wikipedia in any article. Decisions to include should be carefully considered because we would likely violate our own standards against vanity. In this case the Wikipedia vanity is obvious. You are adding details from a forum post discussed by less than 10 Wikipedians looking into solving a general problem raised by contributors, bloggers, etc." Carl moves the message to Talk:Robert McHenry.
  • 2 Sep: User:Alteripse explains his revert of Prematurity.
  • 3 Sep: User:Jmc says: "I have replied at length at Talk:Packet switching"
  • 7 Sep: Koffieyahoo complains on User talk:CarlHewitt: "I’m getting really, really tired of you reverting stuff which I removed under, what I believe are valid arguments."
  • 15 Sep: User:Linas says hi and writes on User talk:CarlHewitt: "We were concerned about some of your edits at Wikipedia talk:WikiProject Physics; the general relativity/quantum mechanics people were not amused". User:Laurascudder explains further.
  • 16 Sep: CfD discussion for Category:Relativistic Information Science. Carl removed the CfD header, likely by accident. EMS complains violently on User talk:CarlHewitt about Carl's behaviour, refering to NOR. Carl makes a strange response: "often attempts are made to suppress interdisciplinary work."
  • 16 Sep: CSTAR says on User talk:CarlHewitt: "EMS I think you are being unjustifiably aggressive here. I am no fan of some of Carl Hewitt's views, but there seems to be a vicious campaign against him. … I also want to make it it clear to EMS where I diverge with Carl Hewitt. My criticisms were very specific about the role and influence of the actor model in some of the historical accountsparticularly in relation to other ideas in computer science. But this is really a matter of opinion and if you look at some of the talk pages, you will see a divergence of views." User:Hillman.
  • 17 Sep: User:Hillman disagrees with CSTAR and Linas also takes exception to CSTAR's comment ("I am rather upset about all of this"). Hillman also says to Carl: "I think it might be time to suggest, Carl, that you remove the specific speculations … you need to be sensitive to our increasing frustration with your tendency to dismiss our objections as some kind of "supression", which they are not."
  • 17 Sep: CSTAR says: "I agree that the historical accounts are bizarre (please see the Actor model talk page, where you will notice I make this point repeatedly)."
  • 17 Sep: Hillman says "The problem is that Carl is resisting all our requests to remove specific seriously misleading claims from two of this articles." Much discussion between CSTAR, EMS, Hillman and Linas.
  • 18 Sep: CarlHewitt writes "You should not read very much into the fact that my colleagues have not yet shown up here. Many have indeed read the material that I have written and have commented on it in email communications among computer scientists. Just about everyone mentioned in the material that I have written knows about it. \\ In fact I recently gave a seminar at MIT on "The Present and Future of the Wikipedia in Academia" that was well received. I can tell you that my colleagues are extremely skeptical of the Wikipedia. They regard my efforts as a pioneering experiment. Typically they do not believe that the Wikipedia has sufficient social mechanisms to make it worthwhile for them to participate other than as casual readers of material of dubious accuracy.
  • 22 Sep: CSTAR says "I retrieved the deleted talk page and put it here. User talk:CSTAR/Relativistic information science discussion"
  • 17 Oct: EMS says "== RfC on quantum indeterminacy == \\ Carl - You can do it yourself. (Or so I think) \\ BTW - Do remember 3RR."
  • 17 Oct: Charles Matthews says "I have been asked to look over the page. It seems to me that you will do much better by finding a compromise with CSTAR, over your choice of quotations, than with further reverts."
  • 30 Oct: User:R.Koot complains about Carl announcing his talk on Talk:Carl Hewitt.
  • 31 Oct: R.Koot complains "You have been adding a lot of links to the Scientific Community Metaphor." and asks " if there is any realtion between the actor model and intelligent agents and, if so, what that relation is?" Furthermoreh Ruud says, refering to an AfD discussion: "Anyone with a B.S. in CS would know this statement is total nonsense. I doubt an professor emiritus would make such a claim. Are you really Carl Hewitt, is this a prank or have you gone insane? Would you care to give an explanation?" It's not clear to me whether the last issue is resolved.
  • 2 Nov: R.Koot warns that he "removed some categories from articles related to the actor model." Carl disagrees.
  • 2 Nov: R.Koot gets blunt: "it is very simple that all you are trying to do is here is write some POV articles on your own research, create as many links to it to promote them and revert all changes anybody makes to them."
  • 3 Nov: EMS concurs that there is too much "crosscategorization"
  • 21 Nov: Carl expands on his remarks on the reaction from colleagues following his talk on Wikipedia. Discussion ensues.
  • 21 Nov: CSTAR seeks further clarification on the connection between the actor model and physics. Carl replies.
  • 22 Nov: R.Koot complains about capitalization of "Actor".

Scientific Community Metaphor and related articles[edit]

Edits by Carl[edit]

This is a list of articles edited by Carl, by number of edits of him, current as of 1 Dec. Deleted articles are not shown. There are some mistakes near the end, but I can't be bothered to fix them. Asterisks denote histories that I looked through, o things I need to check

Draft[edit]

This case revolves in my opinion around the following key principles:

  • Neutral point of view: CarlHewitt is pushing the Actor model and other research of his and his coworkers'. Unfortunately, it is not easy to prove that something is not NPOV.
  • Editors should try to achieve consensus; this implies that they should be willing to compromise. I could not find an eminently quotable policy on this one, but I do think that such a principle exist. For instance, Wikipedia:Consensus starts by saying "Wikipedia works by building consensus".
  • Collegiality. In particular, questions, and the people asking them, should be taken seriously and not be answered in a dismissive or a evasive manner, or by asking a counterquestion. This is partially covered by Wikipedia:Civility, but that policy focuses on more flagrant examples of uncivility.
  • Ownership of articles: CarlHewitt is overly protective of his articles.

I found it extremely hard to summarize the case as it spreads out over several months in which CarlHewitt has made thousands of edits (though his edit count is inflated as he often follows a major edit with dozens of minor edits fixing spelling or formatting). There is no smoking gun, no obviously bad behaviour of CarlHewitt's part; rather, it is a long series of annoyances that add up to make the experience of working with CarlHewitt extremely frustrating (this is also discussed in the evidence presented by EMS). Unfortunately, this means that a lot of evidence has to be presented, and I apologize to the arbitrators for having to read through all of it.

I chose to present the evidence in two parts. In the first section, I take one article (denotational semantics) and give an account of most of the events relating to this article; this has the advantage that the reader can see the context. The disadvantage is that it is a lot of work to read it (not to mention writing it), so in the next section, I present some diffs to illustrate the principles that I listed at the top.

Denotational semantics[edit]

This section discusses the events surrounding denotational semantics (which I'll abbreviate as "den.sem.") and the associated talk page.

  • 22 Aug 16:02, Den.sem.: CarlHewitt adds the PhD thesis by Will Clinger (a student of his) to the references section.

September[edit]

  • 5 Sep 06:53 – 6 Sep 07:27, Den.sem. (158 edits[1]): CarlHewitt adds a large section Denotational semantics of concurrency which consists largely of big quotes from Clinger's thesis. As a consequence, Christopher Strachey and Dana Scott (according to our articles the founders of the field) are relegated to the bottom of the article.
  • 6 Sep 10:05, Talk:Den.sem.: Koffieyahoo says on the talk page that "some of the remarks added to the article regarding denotation semantics and concurrency are in correct. The article suggest atm that denotational semantics is inherently non-concurrent. However, this is not the case." Koffieyahoo refers to a chapter by Ong to back him up.
  • 6 Sep 10:09, Talk:Den.sem.: Koffieyahoo adds that "The article also seems to have been retargeted with a focus on the Actor model, which seems completely unjustified given the wide spread use of denotational semantic through out computer science."
  • 6 Sep 16:08, Talk:Den.sem.: CarlHewitt answers "Can you cite any errors in the article?" I don't understand this answer as did is exactly what KoffieYahoo did.
  • 7 Sep 07:47, Talk:Den.sem.: Koffieyahoo clarifies what he sees as the problem, referring to a specific sentence in Denotational semantics.
  • 7 Sep 21:29, Talk:Den.sem.: CarlHewitt replies that Ong is not an original source. This seems clearly irrelevant to me and only serves to avoid the question.
  • 9 Sep 10:34, Talk:Den.sem.: Koffieyahoo replies "Ong gives [a]n overview. Overviews are an integral part of science […] Didn't respond wrt the errors I pointed out. Please do so (Learn to respond to all arguments. Do proceed to practice this selective responding. It's very annoying. Moreover, learn to respond to the point.)."
  • 9 Sep 14:25, Talk:Den.sem.: CarlHewitt reacts "Not to detract from Ong's work, it is not an original source."
  • 9 Sep 15:47, Den.sem.: CarlHewitt moves two subsections to the new article Unbounded nondeterminism.
  • 9 Sep 19:20 – 12 Sep 05:34 Den.sem. (12 edits): CarlHewitt indents the parts that are quotes from Clinger's thesis and makes some other small changes.
  • 12 Sep 07:27, Den.sem.: User:64.80.151.253 (probably Carl himself) adds another quote from Clinger's thesis.
  • 12 Sep 14:45, Talk:Den.sem.: Koffieyahoo says "Sigh, I don't see why I'm not allowed to cite an overview articles. Particular one that provides you with all the relevant references. Ong cites at least 50 relevant references which I'm not going to repeat here. Moreover, overview or not, this is completely irrevant considering the remarks. So, can you now please answer my questions and remarks." Six minutes later, Koffieyahoo makes one further contribution to Wikipedia, and he has not contributed since (Koffieyahoo's contributions).
  • 12 Sep 16:23, Talk:Den.sem.: TuukkaH says "Overview articles are important references for Wikipedia", referring to WP:NOR.
  • 12 Sep 16:45, Talk:Den.sem.: CarlHewitt agrees "You can cite overview articles. I was simply pointing out that they are not original sources." It seems that this is not how Koffieyahoo saw it.
  • 13 Sep 07:38, Den.sem.: CarlHewitt adds another quote from Clinger's thesis.
  • 15 Sep 09:21, Den.sem.: CarlHewitt adds yet another quote from Clinger's thesis.
  • 19 Sep 11:58, Den.sem.: Jpbowen adds a sentence to the top mentioning Scott and Strachey.

There is now a break for over a month, in which not much happens.

November[edit]

  • 10 Nov 06:44, Talk:Den.sem.: CSTAR writes "Besides evading many of Koffieyahoo's points, I would also point out that the article completely ignores the contributions of J. de Bakker and his collaborators on the denotational semantics of concurrency."
  • 10 Nov 07:33, Talk:Den.sem.: CarlHewitt asks "Is any of de Bakker's work on denotational semantics?"
  • 10 Nov 17:06, Talk:Den.sem.: CSTAR answers affirmatively, giving some references.
  • 11 Nov 13:20, Talk:Den.sem.: CarlHewitt remarks "Samson Abramsky has also done some work in this area." Is he trying to evade CSTAR's point?
  • 11 Nov 17:33, Talk:Den.sem.: CSTAR says that "the article is deficient in references, in coverage of the material and in clarity. In particular, I don't think you're use of extensive direct quotes is helpful."
  • 11 Nov 18:57, Talk:Den.sem.: Charles Matthews chimes in, stating that he "studied with Abramsky for a couple of years at Imperial in the early 1990s. The article has a slanted POV."
  • 11 Nov 21:24, Talk:Den.sem.: CSTAR elaborates "Aside from the intro gives a completely distorted view of denotational semantics. It should at least begin with denotational models for the lambda calculus and relate this to work of Dana Scott and Chris Strachey."
  • 13 Nov 16:35, Talk:Den.sem.: CSTAR complains "One major defect of the article is this: It never mentions the basic point of denotational semantics, which is to associate meaning to expressions satisfying in a compositional way."
  • 14 Nov 18:43. Den.sem.: In response, CarlHewitt writes a section titled Compositionality, again stressing the Actor model.
  • 17 Nov 21:05, Talk:Den.sem.: Chalst enters the discussion. "Although CH[2] started with an article that was not in very good shape, the article as it now stands is so imbalanced that I think WP is worse off than it was. [… Clinger] is not remotely a current authority on the subject […] In defence of CH, he is not a semanticist, and is documenting what he knows about. But the defence is rather weak when seen in the context of CH's pattern of editing elsewhere: the article is being used inappropriately to push the actor model, at the expense of balanced coverage of the topic at hand."
  • 21 Nov 10:13, Talk:Den.sem.: CarlHewitt replies to CSTAR's complaint (13 Nov 16:35): "I have written a section on compositonality that treats it from a modern point of view. See what you think. You should be aware that some of the classical treatments of compositonality in the text books have severe limitations, e.g., they do not handle delayed evaluation."
  • 21 Nov 10:33, Talk:Den.sem.: CarlHewitt responds to Chalst's comment that Clinger is not a current authority that "his dissertation is still a classic work on the subject. It is widely know by researchers in the field. Recent work builds on it and will be published soon." Refering to soon-to-be-published papers is tactic that Carl employs regularly; further down, he says "There will be additional publications on denotational semantics this summer which will require the article to be revised in any case."
  • 21 Nov 16:50, Talk:Den.sem.: CSTAR answers to CarlHewitt's 10:13 reply: "Your new section is not an improvement. […] Saying this treats this from a "modern point of view" is your opinion, which I doubt is widely shared."
  • 21 Nov 18:47, Talk:Den.sem.: CarlHewitt replies: "That the Actor model approach to compositionality is a modern point of view is not just my opinion. See the references in Actor model." The list of references contains about 100 items long, so just pointing to them is not very useful, but CarlHewitt does not seem to realize this. CarlHewitt's replies continues "The reason that the Actor approach has become the modern point of view is that other approaches have run into trouble, e.g. with delayed evaluation."
  • 21 Nov 19:03, Talk:Den.sem.: CSTAR reacts "Baloney. See for example work of Mitch Wand and coworkers,"
  • A technical discussion on whether other approaches can define semantics for delayed evaluation follows.
  • 22 Nov 01:22, Talk:Den.sem.: Trying to close the discussion, CarlHewitt says "I suppose that you [that is, CSTAR] have realized by now that there is no good way using the old text books and papers to write down the denotational semantics of true delayed evaluation (i.e. the version without the strange hack in Scheme). This is one reason that modern treatments of compostional denotational semantics use the Actor model."
  • 22 Nov 02:42, Talk:Den.sem.: CSTAR reacts "Re: I suppose that you have realized by now... I'm sorry I don't believe anything of what you've said. You dismiss both Steele and Sussman as hackers, suggesting their approach was strange. Presumptuous to say the least." The names refer to Guy L. Steele, Jr. and Gerald Jay Sussman.
  • 24 Nov 06:56, Talk:Den.sem.: CarlHewitt apologies for his 22 Nov 01:22 remark, and also for an earlier remark in which he said that CSTAR does not "quite understand the problem here. Denotational semantics is supposed to be a mathematical model theory."

As a result of CarlHewitt's edits, the article denotational semantics currently gives a highly skewed treatment of the field, doing little justice any researchers not associated with Carl Hewitt. Three quarters of the article is about the actor model.

Other evidence[edit]

This section lists examples of what I view as CarlHewitt's uncollegial attitude and the reactions of other editors to CarlHewitt's actions.

CarlHewitt evades questions
  • 13 July 12:56, User talk:Carl Hewitt: Koffieyahoo complains "You aren't answering the questions you're just re-iterating your view on things and that's not an answer".
  • 15 July 13:02, User talk:Carl Hewitt: Koffieyahoo says "Once again, unfortunately, I'd like to ask you to react to my questions and comments on the talk pages of the following topics"
  • 20 July 10:45, User talk:Carl Hewitt: Koffieyahoo writes "From my perspective you have either not answered or only gave evasive answers wrt my questions and comments."
  • 19 Oct 13:20, Wikipedia:Articles for deletion/Actor model, mathematical logic, and quantum physics: User:Linas (who has also presented evidence) states that his "frustration with Hewitt is that he makes assertions and then persistently dodges questions and criticisms."
CarlHewitt answers a question by giving an unhelpful reference
  • 15 Jul 16:51, Talk:Actor model: When a user signing with "Blissex" asks "what kind of read/write thing is present in the lambda calculus...", CarlHewitt answers in condescending manner: "Read Church's paper" (that is his full answer).
  • 19 Oct 04:39, Wikipedia:Articles for deletion/Actor model, mathematical logic, and quantum physics: In reaction to a comment by Ems57fcva that he does not see any connection between the actor model and quantum physics, CarlHewitt writes "You might try reading Computers without Clocks by Ivan Sutherland and Jo Ebergen to gain some intution. Then ask the following question: In terms of quantum physics, how does an arbiter work?". This answer is useless (see the remainder of the discussion), but I find it again very condescending.
  • 21 Nov 18:47, Talk:Denotational semantics[4]: CarlHewitt writes: "That the Actor model approach to compositionality is a modern point of view is not just my opinion. See the references in Actor model." Anybody wishing to verify this statement will be stymied by the fact that the list of references is about 100 items long.
CarlHewitt answers a question with another question instead of offering an answer
  • 6 July 16:24, Talk:Actor model: In reaction Koffieyahoo's remark that while reading the article, he got "the distinct impression that comparisions are trying to show that the actor model is best," Carl Hewitt asks "For what purpose do you think that the article argues that the Actor model is best?"
  • 9 Oct 02:34, Talk:Quantum indeterminacy[3]: In response to a comment by CSTAR that Fuchs' views are too idiosyncratic to rely on, CarlHewitt writes "I agree that Fuchs is not the only view and that the treatment in the introduction should bring out the underlying controversies. Can you see how to do this?"
  • 19 Oct 05:24, Wikipedia:Articles for deletion/Actor model, mathematical logic, and quantum physics: CSTAR implores CarlHewitt "to rely on a more "progressive" dialogue strategy. Please try not to respond to questions with questions". CSTAR also says that CarlHewitt's answer at 19 Oct 04:34, mentioned in the previous paragraph, is not helpful.
CarlHewitt answers a question by referring to publications that will soon appear
  • 31 Oct 03:56, Talk:Scientific Community Metaphor: CarlHewitt says that "we can expect many more publications in the coming years. Of course there is a delay between when work is done and it is published in the literature."
  • 21 Nov 10:33, Talk:Denotational semantics[4]: CarlHewitt writes "Recent work builds on it and will be published soon. […] There will be additional publications on denotational semantics this summer which will require the article to be revised in any case."
CarlHewitt is pushing his own work
CarlHewitt's assumption of ownership
CarlHewitt claims that something is controversial while he's the only one arguing for a particular PoV
  • 17 Oct 22:10, Talk:Quantum indeterminacy[3]: CarlHewitt says "This article has proven to be extremely controversial in practice (although there has been a claim that the subject matter of the article may not be controversial)."
  • 27 Nov 05:34, Talk:Model (abstract): CarlHewitt undoes a revert and explains on the talk page "The section now discusses the controversy in contrast to the previous version which violated NPOV because it didn't discuss the controversy" (the alleged controversy is whether there are abstract models that are not mathematical). However, as the ensuing discussion shows, he has no evidence that there is such a controversy, even though CSTAR asks him at 06:03, 06:12, 07:37, 15:20, and 16:12.

Concluding remarks[edit]

CarlHewitt is clearly an expert on the actor model, and as such, he could prove very useful for Wikipedia. Unfortunately, most (if not all) editors that have interacted with him over some time, have felt considerable frustrations because of it, leading some to leave Wikipedia alltogether. An ideal conclusion of this case would allow Wikipedia to profit from CarlHewitt's expertise without his poisoning the community (the latter has of course a negative effect on Wikipedia).

Notes[edit]

  1. CarlHewitt often makes many edits in succession. For the convenience of the reader, I decided to group all edits together in one diff when they are made consecutively without any other editor intervening.
  2. Here, "CH" refers to User:CarlHewitt, and not to User:Hillman (Chris Hillman) who also presented evidence and signs his posts with "CH".
  3. The conflict surrounding Quantum indeterminacy is treated in more detail in the evidence presented by EMS.
  4. This diff also appears in the section #Denotational semantics.