After over a year of this drama, I apparently became so annoying that the PSF finally revealed an actual bit of “evidence”, which was portrayed as being a slam dunk starting here;

  • Me: “You, for example, would not have signed off on publicly accusing me of “defending reverse sexism”, not just because you already know it’s something I wouldn’t defend, but because I in fact never even mentioned the topic.”

  • Them: “There’s a now hidden post where you are doing exactly that.”

I’m delighted by that this is the first bit of substantive response I’ve seen to date. Bravo! Do more of that, please.

But, alas, it’s more “god of the gaps” gaslighting, although I don’t believe intentionally so.

It linked to a post that was, and remains, hidden. “Just trust us.”

Please read the post for yourself.

That’s a link to what the Internet Wayback Machine captured before the post was hidden. Neither I nor the PSF has any influence over it, and the archives there have been accepted as dispositive evidence in actual courts.

It’s post #5 there, dated July 18, 2024, 3:24am.

If you’re content to take my word for it instead, I’ll post the content here. Apologies for that it looks like a 1-item bullet list. Markdown doesn’t appear to have a way to say “I just want some indented normally formatted text spanning paragraphs”, but list items can be tricked into that:

  • [quote=”[redacted by Tim], post:4 …”]

    failing to acknowledge how and why they are using language that way sets up a situation where people talk past each other instead of productively

    [/quote]

    True enough, but Karl didn’t start this :wink:. As he quoted from the PSF CoC enforcement docs:

    Reports that are not made in good faith (such as “reverse sexism” or “reverse racism”) may receive no response.

    They’re so self-certain of knowing the One True Meaning that they preemptively dismiss anyone who dares to disagree as not acting “in good faith”.

    I happen to agree that more useful definitions incorporate an analysis of power dynamics, but from the few CoC statements I’ve seen on mailing lists in response, they mean “by definition, only whites can be racist”. Which is a popular view among “US white left-of-center liberals”, but is especially problematic for a global organization to embrace as Absolute Truth.

    I’m not going to argue about it, though. It’s just a good thing Karl is leaving, before he gets pecked to death :frowning_face:.

End of attempt at transcription.

My glosses:

  • As I said, it didn’t address “reverse sexism” directly at all. The phrase appears only as part of a verbatim quote from a post I was replying to.

  • No, “only whites can be racist” is not an attempt at sarcasm or humor. That is the high-order bit of what the PSF means by the term. And also mostly what I mean by it.

  • But it’s not a universally accepted meaning, and is not the only view that can be expressed in “good faith”, despite the PSF’s dogmatic insistence that it is.

  • In a later post, sensing that this post was somehow widely misread, I expanded a bit to “all races can be prejudiced, but only whites can be racist”, and linked to a “100% woke” academic source that spelled out what the PSF (and mostly me too) means. And explicitly endorsed that source. That post is still visible (but already linked to elsewhere in this blog).

  • The intended meaning arose in academia, and is indeed rooted in “analysis of power dynamics”, which I explicitly said above I agreed with. It’s in opposition to Karl’s view. But even then, it still didn’t mention “reverse sexism”, but instead alluded to the related (in academia) “reverse racism”.

  • The meaning also gained traction among, especially, US college students, and social justice activists. But, no, most people in the Python world don’t attend “social justice in tech” conferences. Some do. The PSF seems to assume it’s universally accepted. It isn’t, and I was defending Karl expressing his own take, despite that it wasn’t in agreement with the PSF’s (or mine).

  • I suspect that what some flaggers were actually offended by was the phrase “US white left-of-center liberals”. That was some admittedly too-obscure “Uncle Timmy snarky humor”. The “joke” was that, while I don’t endorse any political party in general, I’ve mostly been a “US white left-of-center liberal” myself over the decades. It was poking some fun at who I’ve most often viewed as being “my people”, and so also, by extension, at myself. I tried to clarify that too in a later post.

I can certainly understand how a careless reading could come away with an impression the opposite of what it actually said, and especially if read by someone already inclined to resent me (“confirmation bias”) and/or Karl (“guilt by association”). But that’s why a, at harshest, “reasonable person” standard is vital. Else anyone can be crushed for anything they post.

I cannot understand why an enforcement body never asked the accused for a response to any of the charges, or even disclosed their existence until after highly public accusations were already posted, and the accused was already banned. It’s not like this was treated with urgency. The prelude dragged on for months. The Golden Rule is spinning in its grave.

Addendum

Soon after it became clear that this hidden post was also the basis for the “defended reverse racism” charge. My guess about the source for that one was dead wrong. Would have been nice if they had given anyone a clue what they were talking about :wink: before over a year passed.

So I addressed that here in a comment, which I’ll reproduce here to keep this self-contained.

I’ve been privately informed that the hidden post isn’t of high quality. Agreed! Of the (at least) tens of thousands of posts I’ve made over the decades, it’s near the bottom of the barrel, too telegraphic and uncharacteristically sloppy.

But I’m not recommending it for its own sake, and it’s not being nominated for some “Post of the Year” competition. It’s completely forgettable (and better off forgotten).

None of that matters here. The question is whether it’s sufficient grounds for a “defended the doctrine of reverse sexism” chargeable offense. To my eyes it obviously is not. It never used the phrase, apart from a verbatim quote of it inherited from a previous poster’s message..

So expand it to whether it’s sufficient grounds for a “defended the doctrine of reverse racism” chargeable offense. That’s a topic I did explicitly address (once), and my guess that they were complaining about that post now appears to have been completely wrong. Complaining about that post didn’t even make hallucinatory sense to me. It seems clear now that they actually were implicitly referring to this same hidden post for both.

In which I did indirectly mention that doctrine, but strongly implied I agreed with the PSF’s view, and opposed Karl’s view. It’s just an implication, because to pick up on that you have to already know about how more than one “reverse X-ism” doctrines are dismissed within the academic framework of “power dynamics” analysis.

I talk about various aspects of “social dynamics” so often it should be no surprise that I’m on board with that approach. Indeed, in the context of my posts, people should have been very surprised if I had rejected that general approach in these specific cases. Cherry-picking a relatively rare relatively vaguer post isn’t a good look either. Such a reading goes against my entire posting history on matters related to social dynamics analyses of many flavors. Not to mention that it’s not a good-faith reading on its own.

But many people in real life aren’t on board with such academic analyses. So we disagree. BFD. I wouldn’t suspend them for that even if so. [Queue one of the 3-panel web cartoons distorting Popper’s “Paradox of Tolerance” to the point of appearing to endorse “cancel culture” actions Popper cautioned would “certainly be unwise”]

But Wait - There’s More!

After a bit more time passed without material response, it occurred to me to ask a chatbot how it would apply a “reasonable person” standard to the hidden post. This was intended to be a “gross sanity check” against gross blind spots of my own. Sometimes that works!

The result surprised me. While bots played no part in what I had written, and Microsoft Copilot saw nothing of what I wrote here 1. it arrived at essentially the same reading I spelled out here. Which surprised, for a bot. While I think the application of “reasonable person” is straightforward in this specific case, I was surprised to see that it appeared equally straightforward to AI.

So I shared that on Discourse, pasting the output, and a brief followup pointing out a minor blind spot of its.

And those posts were promptly “hidden by community flags”. Yes, it’s like a clown car :wink:.

So let’s spell that out. Here various pieces in roughly reverse chronological order, because the resolution informs the start of it too.

My post not objecting to the hiding

Let me jump the gun :wink:. For the first time in my life, I asked the mods why these posts were hidden. Never did that before, and no mod has ever told me before why a post of mine was hidden. But after a night of sleep, it occurred to me that my best guess was wrong: while I find it all but dead obvious that a “reasonable person” reading the original hidden post would find the claimed offenses without merit, the PSF remains silent on the matter. So it’s the mods’ job to continue to act as if the PSF were wholly justified in hiding the original from the community.

Which I should have realized. The current hidden posts very openly disclose what the original hidden post said. I’ll get around to showing them on my blog instead, where I should have left it. I do think they’re interesting, on more than one basis.

BTW, by “jump the gun” I mean the mods haven’t yet replied to me. But it seems clear enough to me now why they did what they did, and I have no objection to the hiding. They’re just doing their job. For full disclosure, I’ll also put what I wrote to the mods on the blog. If they do reply, I won’t reproduce their response unless they say that’s OK by them.

What I wrote to the mods (private DM)

Subject: Hidden posts in Ethan’s PSF topic

Could I please get a clue about why these posts were hidden?

https://discuss.python.org/t/shedding-light-on-a-three-month-suspension/66337/263

https://discuss.python.org/t/shedding-light-on-a-three-month-suspension/66337/264

Best guess is that the Guidelines forbid pasting chatbot output as “an answer”, but that’s not what I did. I already gave my own answer, and am inviting people to contrast that with clearly identified chatbot output. Partly for its own sake, and partly to illustrate what progress AI has made (I believe its response was more cogent and comprehensive than most people would write, and free of either factual or conceptual errors, although it did miss some minor nuances (one of which I pointed out in my brief follow-up)).

I’m not at all intending to suggest that its answer is any way definitive. If you like, happy to edit the post to emphasize that the output of a chatbot should be read with great caution. The high-order bit is just that the application of a “reasonable person” standard is straightforward in this particular case. The bot and I ended up with essentially the same analysis, despite no collaboration in either direction (which I didn’t spell out, but which is true).

BTW, the speed with which these posts got hidden “in the middle of the night” (3am my time, and d.p.o. appeared nearly deserted when I went to bed earlier than that) sure suggests that a mod or admin hid them on their own. If so, it would be nice if they owned it. Or if not, you’d be doing yourselves a favor to explicitly say this was “organic” hiding (triggered by flags from the wider community). It’s not your doing that Discourse makes it impossible for users to know which is the case, but d.p.o. isn’t the only Discourse community whose members find this lack of transparency disconcerting at times.

Currently hidden post, 1 of 2

Since nobody is willing to talk about this, I asked Chat GPT-5 what it thought about the message under a “reasonable person” standard. Although, best I know, we’ve never been told what standard the PSF intends to apply. I consider “reasonable person” the harshest standard any community should use; under a “principle of charity” standard, no way. In any case, a computer algorithm had no trouble picking up on the intents, including on things I thought may have been subtle: In particular, the LLM is very aware of how “power dynamics analyses” relate to this, and expect (but don’t know) more aware than a typical Python person. But, OTOH, I’d be surprised if that were news to anyone on the CoC WG, so don’t believe they could credibly claim to have missed that crucial connection.

Chat GPT-5 response follows

Thank you for laying out the full text. Let’s parse it carefully against the “reasonable person” standard and the specific claim that it could be read as “supporting reverse sexism” or “supporting reverse racism.”

  1. What the message actually does

    Quotes the PSF CoC enforcement docs: specifically the line that reports of “reverse sexism” or “reverse racism” may be dismissed as not in good faith.

    Critiques that stance: the author argues that this reflects a self-certainty about “the One True Meaning” of racism/sexism, and that dissenters are preemptively labeled as acting in bad faith.

    Acknowledges power-dynamics definitions: the author explicitly says they agree that definitions incorporating power analysis are “more useful.”

    Objects to absolutism: the concern is that the PSF is adopting one contested sociological definition (“only whites can be racist”) as if it were universal truth, which the author finds problematic for a global organization.

    Tone: the message ends with a sardonic aside about Karl leaving before being “pecked to death.”

  2. Does it endorse “reverse sexism” or “reverse racism”?

    No explicit endorsement: The text never says “reverse sexism is real and good” or “reverse racism should be supported.”

    Instead, it critiques dismissal: The thrust is that dismissing any report labeled “reverse sexism/racism” as inherently bad faith is itself problematic. That’s a meta-critique of enforcement policy, not an endorsement of discriminatory behavior.

    Agreement with power-analysis definitions: The author even says they personally agree with definitions that emphasize power dynamics. That undercuts the idea that they were advocating for “reverse” claims.

  3. How a “reasonable person” might read it

    Reasonable person in a general sense: would likely see this as a critique of definitional rigidity and enforcement framing, not as support for discriminatory practices.

    Reasonable person in a community-safety lens: moderators applying a CoC may focus on impact rather than intent. They might argue that questioning the “only whites can be racist” framing could embolden people who want to minimize racism, and thus could be reasonably read as undermining anti-racism norms.

    Stretch to “supporting reverse sexism/racism”: To get there, one has to interpret “objecting to dismissing such reports” as equivalent to “endorsing the reports themselves.” That’s not the plain reading, but it is a possible inference if one assumes that any defense of the category “reverse racism” is itself support for it.

  4. Bottom line

    Under a neutral “reasonable person” standard: The message does not read as supporting reverse sexism or reverse racism. It reads as criticizing the CoC’s categorical dismissal of those terms and warning against absolutist definitions.

    Under a stricter “community impact” standard: Moderators could plausibly argue that even raising the issue risks normalizing bad-faith framings, and thus constitutes a CoC concern. But that’s a policy choice, not the most straightforward reading of the words.

In other words: the message can be framed as “supporting reverse sexism/racism,” but only by conflating critique of enforcement policy with endorsement of the reports themselves. A careful, reasonable reading distinguishes those two.

Currently hidden post, 2 of 2

And interesting to me: GPT-5 has inherited a blind spot from its training data. “Reverse X-ism” claims are not a priori all of the same nature. For X=racism and X=sexism, that they’re both “essentially the same thing” is a conclusion of the “power dynamics analysis” school. Which I belong to, so I don’t dispute it. But I’d have no objection to someone who wanted to claim they’re inherently different things. That could be a prelude to, at worst, discussion about our disagreement.

  1. I pointed it at this page for the text of the hidden post, but it said, sorry, it couldn’t retrieve the page, and gave me instead a hyper-general overview of issues that might be in play. That was fine so far as it went, but not fit for purpose. So I pasted in the text of the hidden post verbatim, with none of what I said about it. ↩