Re: [hrpc] Censorship

[Bill Woodcock <[e-mail]>]

> On Mar 11, 2022, at 10:59 AM, Vittorio Bertola <[e-mail]> wrote:
> 
> 
> 
>> Il 11/03/2022 10:38 Bill Woodcock <[e-mail]> ha scritto:
>> 
>> 
>>> On Mar 11, 2022, at 9:57 AM, Vittorio Bertola <vittorio.bertola=[e-mail]> wrote:
>>> I think that this call for a global, multistakeholder process to block access to Russian military and news websites
>> 
>> I encourage you to all read it for yourselves, because, as one of the authors and one of the signatories, that’s not what we said.
> 
> Indeed people should read it, that's why it was linked.

Thank you, I appreciate that, and I appreciate that you’re arguing in good faith, as I’ve always observed you to do before.  I respect your opinion, and I want to make sure that both you understand what the goals of the letter were, and that I understand your objections or considerations.

> The document itself calls for a public discussion, though it does not specify where.  I am happy to join the public discussion wherever the signatories want it to happen.

I don’t believe any of us had any thought that we had any authority to dub anywhere the “correct” place.  But I think that, more specifically, we all share the view that Internet multistakeholder governance is bottom-up and non-exclusionary.  So there is no “wrong place” except perhaps forums that, as you point out, may deem it off-topic:

> The points you raise could be off topic for this group, but if the chairs think differently, I am happy to continue the discussion in detail here.

I as well, and I believe there are other signatories and authors on this list.  Also happy to try to constrain discussion here to aspects which lie at the intersection of the thoughts we expressed and the topic of this list.

> If all Internet protocols and services were designed in the way we advise in draft-guidelines, then the sanctions proposed by the document would be technically impossible

Can you explain that in more detail?  As someone who works in this field, and who has read https://datatracker.ietf.org/doc/draft-irtf-hrpc-guidelines/ it is not clear to me what portion of the proposed mechanism you believe would be made impossible.  Or for that matter, touched at all.  If you can direct me to something specific, I can give a more considered response.

> Perhaps we should recognize in draft-guidelines that, after all, censorship is sometimes desirable and there should be technical and policy hooks to implement it; so protocol designers should rather think at how to address this need while preventing such practices from becoming overbroad. I would be much in favour of this, as I do think that censorship is sometimes desirable, including in this case; but this did not seem to be the consensus of this group (and of PEARG) in the past.

I feel that others in this group have far deeper intellectual backgrounds upon which to draw on this topic.  Myself, I feel that it’s self-evident that there’s a moral imperative to choose the lesser of two possible evils, when one is compelled to make a choice on behalf of another.  That is, when one is making a choice with consequences only for one’s self, one is free to suffer a greater evil in order to support a rhetorical point.  But when the sacrifice is someone else’s, doing so is anathema.

If viewed in this light, _sufficiently constrained_ sanctions are the proper choice.  They may involve some minor sacrifice on the parts of some actors (if, for example, they are selling services to the Russian military, as Cloudflare recently reiterated their commitment to continuing to do), but relative to kinetic warfare and atrocities, it is an essentially deescalatory response, which is good.  It shows moral conviction and provides some allyship with victims, while simultaneously removing subsidies (in the form of societally-borne common costs) from the aggressor.  This is shunning, or outlawing, as Icelandic law established more than a thousand years ago. If someone flagrantly transgresses societal norms, they cannot expect to benefit from societally-provided goods, at society’s common expense.

Removing Russian military routes from our routing tables reduces a collective cost to society, and transfers those costs back to the Russian military and those who are selling them services, to carry individually.

You raised, initially, the question of whether this constituted “censorship.”  I would distinguish between individual efforts which have the effect of being in the direction of censorship, versus censorship as a measured effect.  The Internet famously “routes around damage” and “censorship is indistinguishable from damage.”  Thus if a network chooses to no longer carry Russian military routes, the Russian military has not been disconnected from the Internet… they simply have to carry a greater share of the cost of their connection individually, and leave less of the cost to be shared communally.  Every single one of some thirty thousand networks would have to withdraw Russian military routes before they became truly unreachable.  So in that sense, I would actually disagree that this is censorship, in that it’s not truly effective as censorship.  It is exactly effective as a sanction, however, since the sanction is intended to transfer the costs of aggressors’ actions back into their own wallets.

                                -Bill